diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,56034 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 5.233889434085705,
+  "eval_steps": 500,
+  "global_step": 80000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0006542361792607131,
+      "grad_norm": 2.3467204570770264,
+      "learning_rate": 1.0000000000000001e-07,
+      "loss": 0.208,
+      "step": 10
+    },
+    {
+      "epoch": 0.0013084723585214263,
+      "grad_norm": 2.2877755165100098,
+      "learning_rate": 2.111111111111111e-07,
+      "loss": 0.2277,
+      "step": 20
+    },
+    {
+      "epoch": 0.001962708537782139,
+      "grad_norm": 2.4381093978881836,
+      "learning_rate": 3.222222222222222e-07,
+      "loss": 0.2252,
+      "step": 30
+    },
+    {
+      "epoch": 0.0026169447170428526,
+      "grad_norm": 2.922689914703369,
+      "learning_rate": 4.3333333333333335e-07,
+      "loss": 0.2161,
+      "step": 40
+    },
+    {
+      "epoch": 0.0032711808963035655,
+      "grad_norm": 2.4064548015594482,
+      "learning_rate": 5.444444444444444e-07,
+      "loss": 0.2299,
+      "step": 50
+    },
+    {
+      "epoch": 0.003925417075564278,
+      "grad_norm": 2.436875581741333,
+      "learning_rate": 6.555555555555556e-07,
+      "loss": 0.2218,
+      "step": 60
+    },
+    {
+      "epoch": 0.004579653254824992,
+      "grad_norm": 2.463653802871704,
+      "learning_rate": 7.666666666666667e-07,
+      "loss": 0.2234,
+      "step": 70
+    },
+    {
+      "epoch": 0.005233889434085705,
+      "grad_norm": 1.9082249402999878,
+      "learning_rate": 8.777777777777779e-07,
+      "loss": 0.2155,
+      "step": 80
+    },
+    {
+      "epoch": 0.005888125613346418,
+      "grad_norm": 1.9711594581604004,
+      "learning_rate": 9.888888888888888e-07,
+      "loss": 0.2167,
+      "step": 90
+    },
+    {
+      "epoch": 0.006542361792607131,
+      "grad_norm": 2.022458076477051,
+      "learning_rate": 1.1e-06,
+      "loss": 0.2089,
+      "step": 100
+    },
+    {
+      "epoch": 0.007196597971867844,
+      "grad_norm": 2.3458659648895264,
+      "learning_rate": 1.2111111111111111e-06,
+      "loss": 0.2131,
+      "step": 110
+    },
+    {
+      "epoch": 0.007850834151128557,
+      "grad_norm": 2.2194032669067383,
+      "learning_rate": 1.3222222222222222e-06,
+      "loss": 0.2217,
+      "step": 120
+    },
+    {
+      "epoch": 0.00850507033038927,
+      "grad_norm": 2.2126388549804688,
+      "learning_rate": 1.4333333333333333e-06,
+      "loss": 0.2191,
+      "step": 130
+    },
+    {
+      "epoch": 0.009159306509649984,
+      "grad_norm": 2.212867021560669,
+      "learning_rate": 1.5444444444444446e-06,
+      "loss": 0.2201,
+      "step": 140
+    },
+    {
+      "epoch": 0.009813542688910697,
+      "grad_norm": 2.3964221477508545,
+      "learning_rate": 1.6555555555555557e-06,
+      "loss": 0.2154,
+      "step": 150
+    },
+    {
+      "epoch": 0.01046777886817141,
+      "grad_norm": 2.0090246200561523,
+      "learning_rate": 1.7666666666666668e-06,
+      "loss": 0.2154,
+      "step": 160
+    },
+    {
+      "epoch": 0.011122015047432123,
+      "grad_norm": 2.2598977088928223,
+      "learning_rate": 1.877777777777778e-06,
+      "loss": 0.2057,
+      "step": 170
+    },
+    {
+      "epoch": 0.011776251226692836,
+      "grad_norm": 2.0792648792266846,
+      "learning_rate": 1.988888888888889e-06,
+      "loss": 0.1998,
+      "step": 180
+    },
+    {
+      "epoch": 0.012430487405953549,
+      "grad_norm": 2.0464093685150146,
+      "learning_rate": 2.1000000000000002e-06,
+      "loss": 0.1958,
+      "step": 190
+    },
+    {
+      "epoch": 0.013084723585214262,
+      "grad_norm": 1.8751633167266846,
+      "learning_rate": 2.2111111111111113e-06,
+      "loss": 0.1874,
+      "step": 200
+    },
+    {
+      "epoch": 0.013738959764474975,
+      "grad_norm": 1.9019556045532227,
+      "learning_rate": 2.322222222222222e-06,
+      "loss": 0.191,
+      "step": 210
+    },
+    {
+      "epoch": 0.014393195943735688,
+      "grad_norm": 1.7033251523971558,
+      "learning_rate": 2.4333333333333335e-06,
+      "loss": 0.1798,
+      "step": 220
+    },
+    {
+      "epoch": 0.015047432122996402,
+      "grad_norm": 1.5006437301635742,
+      "learning_rate": 2.5444444444444446e-06,
+      "loss": 0.1863,
+      "step": 230
+    },
+    {
+      "epoch": 0.015701668302257114,
+      "grad_norm": 1.7847758531570435,
+      "learning_rate": 2.6555555555555556e-06,
+      "loss": 0.1897,
+      "step": 240
+    },
+    {
+      "epoch": 0.016355904481517827,
+      "grad_norm": 2.0124351978302,
+      "learning_rate": 2.7666666666666667e-06,
+      "loss": 0.1816,
+      "step": 250
+    },
+    {
+      "epoch": 0.01701014066077854,
+      "grad_norm": 1.3791583776474,
+      "learning_rate": 2.877777777777778e-06,
+      "loss": 0.1895,
+      "step": 260
+    },
+    {
+      "epoch": 0.017664376840039256,
+      "grad_norm": 1.6091564893722534,
+      "learning_rate": 2.988888888888889e-06,
+      "loss": 0.174,
+      "step": 270
+    },
+    {
+      "epoch": 0.01831861301929997,
+      "grad_norm": 1.2440049648284912,
+      "learning_rate": 3.1e-06,
+      "loss": 0.1737,
+      "step": 280
+    },
+    {
+      "epoch": 0.018972849198560682,
+      "grad_norm": 1.4476827383041382,
+      "learning_rate": 3.2111111111111115e-06,
+      "loss": 0.1656,
+      "step": 290
+    },
+    {
+      "epoch": 0.019627085377821395,
+      "grad_norm": 1.3262630701065063,
+      "learning_rate": 3.3222222222222226e-06,
+      "loss": 0.1588,
+      "step": 300
+    },
+    {
+      "epoch": 0.020281321557082108,
+      "grad_norm": 1.3778283596038818,
+      "learning_rate": 3.4333333333333336e-06,
+      "loss": 0.1677,
+      "step": 310
+    },
+    {
+      "epoch": 0.02093555773634282,
+      "grad_norm": 1.322625756263733,
+      "learning_rate": 3.5444444444444447e-06,
+      "loss": 0.1776,
+      "step": 320
+    },
+    {
+      "epoch": 0.021589793915603533,
+      "grad_norm": 1.2029178142547607,
+      "learning_rate": 3.655555555555556e-06,
+      "loss": 0.1759,
+      "step": 330
+    },
+    {
+      "epoch": 0.022244030094864246,
+      "grad_norm": 1.393506646156311,
+      "learning_rate": 3.766666666666667e-06,
+      "loss": 0.1781,
+      "step": 340
+    },
+    {
+      "epoch": 0.02289826627412496,
+      "grad_norm": 1.2977313995361328,
+      "learning_rate": 3.877777777777778e-06,
+      "loss": 0.1674,
+      "step": 350
+    },
+    {
+      "epoch": 0.023552502453385672,
+      "grad_norm": 1.1824816465377808,
+      "learning_rate": 3.9888888888888895e-06,
+      "loss": 0.165,
+      "step": 360
+    },
+    {
+      "epoch": 0.024206738632646385,
+      "grad_norm": 1.364259123802185,
+      "learning_rate": 4.1000000000000006e-06,
+      "loss": 0.1652,
+      "step": 370
+    },
+    {
+      "epoch": 0.024860974811907098,
+      "grad_norm": 1.3480299711227417,
+      "learning_rate": 4.211111111111112e-06,
+      "loss": 0.156,
+      "step": 380
+    },
+    {
+      "epoch": 0.02551521099116781,
+      "grad_norm": 1.1253418922424316,
+      "learning_rate": 4.322222222222223e-06,
+      "loss": 0.1666,
+      "step": 390
+    },
+    {
+      "epoch": 0.026169447170428524,
+      "grad_norm": 0.9541553258895874,
+      "learning_rate": 4.433333333333334e-06,
+      "loss": 0.1679,
+      "step": 400
+    },
+    {
+      "epoch": 0.026823683349689237,
+      "grad_norm": 1.3034117221832275,
+      "learning_rate": 4.544444444444445e-06,
+      "loss": 0.167,
+      "step": 410
+    },
+    {
+      "epoch": 0.02747791952894995,
+      "grad_norm": 1.1535122394561768,
+      "learning_rate": 4.655555555555556e-06,
+      "loss": 0.1562,
+      "step": 420
+    },
+    {
+      "epoch": 0.028132155708210663,
+      "grad_norm": 1.2631272077560425,
+      "learning_rate": 4.766666666666667e-06,
+      "loss": 0.1656,
+      "step": 430
+    },
+    {
+      "epoch": 0.028786391887471376,
+      "grad_norm": 0.9639325737953186,
+      "learning_rate": 4.877777777777778e-06,
+      "loss": 0.1696,
+      "step": 440
+    },
+    {
+      "epoch": 0.029440628066732092,
+      "grad_norm": 1.205918550491333,
+      "learning_rate": 4.988888888888889e-06,
+      "loss": 0.161,
+      "step": 450
+    },
+    {
+      "epoch": 0.030094864245992805,
+      "grad_norm": 1.0042939186096191,
+      "learning_rate": 5.1e-06,
+      "loss": 0.1651,
+      "step": 460
+    },
+    {
+      "epoch": 0.030749100425253518,
+      "grad_norm": 1.0380401611328125,
+      "learning_rate": 5.211111111111111e-06,
+      "loss": 0.162,
+      "step": 470
+    },
+    {
+      "epoch": 0.03140333660451423,
+      "grad_norm": 0.9904937148094177,
+      "learning_rate": 5.3222222222222225e-06,
+      "loss": 0.1663,
+      "step": 480
+    },
+    {
+      "epoch": 0.03205757278377494,
+      "grad_norm": 1.1420738697052002,
+      "learning_rate": 5.4333333333333335e-06,
+      "loss": 0.1566,
+      "step": 490
+    },
+    {
+      "epoch": 0.03271180896303565,
+      "grad_norm": 1.0128347873687744,
+      "learning_rate": 5.544444444444445e-06,
+      "loss": 0.1576,
+      "step": 500
+    },
+    {
+      "epoch": 0.033366045142296366,
+      "grad_norm": 0.9660897254943848,
+      "learning_rate": 5.655555555555556e-06,
+      "loss": 0.1549,
+      "step": 510
+    },
+    {
+      "epoch": 0.03402028132155708,
+      "grad_norm": 1.0083342790603638,
+      "learning_rate": 5.766666666666667e-06,
+      "loss": 0.1575,
+      "step": 520
+    },
+    {
+      "epoch": 0.03467451750081779,
+      "grad_norm": 1.092995047569275,
+      "learning_rate": 5.877777777777778e-06,
+      "loss": 0.162,
+      "step": 530
+    },
+    {
+      "epoch": 0.03532875368007851,
+      "grad_norm": 0.9964866638183594,
+      "learning_rate": 5.988888888888889e-06,
+      "loss": 0.1518,
+      "step": 540
+    },
+    {
+      "epoch": 0.035982989859339225,
+      "grad_norm": 1.2313237190246582,
+      "learning_rate": 6.1e-06,
+      "loss": 0.151,
+      "step": 550
+    },
+    {
+      "epoch": 0.03663722603859994,
+      "grad_norm": 1.1173750162124634,
+      "learning_rate": 6.211111111111111e-06,
+      "loss": 0.1651,
+      "step": 560
+    },
+    {
+      "epoch": 0.03729146221786065,
+      "grad_norm": 1.0656248331069946,
+      "learning_rate": 6.322222222222222e-06,
+      "loss": 0.151,
+      "step": 570
+    },
+    {
+      "epoch": 0.037945698397121363,
+      "grad_norm": 1.1378906965255737,
+      "learning_rate": 6.433333333333334e-06,
+      "loss": 0.1666,
+      "step": 580
+    },
+    {
+      "epoch": 0.038599934576382076,
+      "grad_norm": 1.271768569946289,
+      "learning_rate": 6.544444444444444e-06,
+      "loss": 0.1557,
+      "step": 590
+    },
+    {
+      "epoch": 0.03925417075564279,
+      "grad_norm": 1.0535553693771362,
+      "learning_rate": 6.655555555555556e-06,
+      "loss": 0.1423,
+      "step": 600
+    },
+    {
+      "epoch": 0.0399084069349035,
+      "grad_norm": 1.0603430271148682,
+      "learning_rate": 6.766666666666667e-06,
+      "loss": 0.1556,
+      "step": 610
+    },
+    {
+      "epoch": 0.040562643114164215,
+      "grad_norm": 3.6667890548706055,
+      "learning_rate": 6.877777777777778e-06,
+      "loss": 0.1488,
+      "step": 620
+    },
+    {
+      "epoch": 0.04121687929342493,
+      "grad_norm": 0.9498468041419983,
+      "learning_rate": 6.9888888888888895e-06,
+      "loss": 0.155,
+      "step": 630
+    },
+    {
+      "epoch": 0.04187111547268564,
+      "grad_norm": 1.061012625694275,
+      "learning_rate": 7.1e-06,
+      "loss": 0.1559,
+      "step": 640
+    },
+    {
+      "epoch": 0.042525351651946354,
+      "grad_norm": 1.0795965194702148,
+      "learning_rate": 7.211111111111112e-06,
+      "loss": 0.1494,
+      "step": 650
+    },
+    {
+      "epoch": 0.04317958783120707,
+      "grad_norm": 0.8958123922348022,
+      "learning_rate": 7.322222222222222e-06,
+      "loss": 0.1435,
+      "step": 660
+    },
+    {
+      "epoch": 0.04383382401046778,
+      "grad_norm": 0.9391831755638123,
+      "learning_rate": 7.433333333333334e-06,
+      "loss": 0.1466,
+      "step": 670
+    },
+    {
+      "epoch": 0.04448806018972849,
+      "grad_norm": 1.4169477224349976,
+      "learning_rate": 7.544444444444444e-06,
+      "loss": 0.1394,
+      "step": 680
+    },
+    {
+      "epoch": 0.045142296368989206,
+      "grad_norm": 1.0921765565872192,
+      "learning_rate": 7.655555555555556e-06,
+      "loss": 0.1522,
+      "step": 690
+    },
+    {
+      "epoch": 0.04579653254824992,
+      "grad_norm": 0.8971364498138428,
+      "learning_rate": 7.766666666666666e-06,
+      "loss": 0.1567,
+      "step": 700
+    },
+    {
+      "epoch": 0.04645076872751063,
+      "grad_norm": 0.9771297574043274,
+      "learning_rate": 7.877777777777778e-06,
+      "loss": 0.1562,
+      "step": 710
+    },
+    {
+      "epoch": 0.047105004906771344,
+      "grad_norm": 1.0776768922805786,
+      "learning_rate": 7.988888888888888e-06,
+      "loss": 0.1485,
+      "step": 720
+    },
+    {
+      "epoch": 0.04775924108603206,
+      "grad_norm": 1.0369359254837036,
+      "learning_rate": 8.1e-06,
+      "loss": 0.1524,
+      "step": 730
+    },
+    {
+      "epoch": 0.04841347726529277,
+      "grad_norm": 1.057263731956482,
+      "learning_rate": 8.21111111111111e-06,
+      "loss": 0.1542,
+      "step": 740
+    },
+    {
+      "epoch": 0.04906771344455348,
+      "grad_norm": 1.065109372138977,
+      "learning_rate": 8.322222222222223e-06,
+      "loss": 0.1539,
+      "step": 750
+    },
+    {
+      "epoch": 0.049721949623814196,
+      "grad_norm": 1.0507162809371948,
+      "learning_rate": 8.433333333333333e-06,
+      "loss": 0.1383,
+      "step": 760
+    },
+    {
+      "epoch": 0.05037618580307491,
+      "grad_norm": 0.9928821325302124,
+      "learning_rate": 8.544444444444445e-06,
+      "loss": 0.1459,
+      "step": 770
+    },
+    {
+      "epoch": 0.05103042198233562,
+      "grad_norm": 1.1054047346115112,
+      "learning_rate": 8.655555555555555e-06,
+      "loss": 0.1455,
+      "step": 780
+    },
+    {
+      "epoch": 0.051684658161596335,
+      "grad_norm": 0.8906415104866028,
+      "learning_rate": 8.766666666666667e-06,
+      "loss": 0.1351,
+      "step": 790
+    },
+    {
+      "epoch": 0.05233889434085705,
+      "grad_norm": 0.891897439956665,
+      "learning_rate": 8.877777777777777e-06,
+      "loss": 0.1536,
+      "step": 800
+    },
+    {
+      "epoch": 0.05299313052011776,
+      "grad_norm": 0.8980157971382141,
+      "learning_rate": 8.988888888888889e-06,
+      "loss": 0.1366,
+      "step": 810
+    },
+    {
+      "epoch": 0.053647366699378474,
+      "grad_norm": 0.9560266733169556,
+      "learning_rate": 9.100000000000001e-06,
+      "loss": 0.1432,
+      "step": 820
+    },
+    {
+      "epoch": 0.05430160287863919,
+      "grad_norm": 0.9071044921875,
+      "learning_rate": 9.211111111111111e-06,
+      "loss": 0.1437,
+      "step": 830
+    },
+    {
+      "epoch": 0.0549558390578999,
+      "grad_norm": 1.1102005243301392,
+      "learning_rate": 9.322222222222223e-06,
+      "loss": 0.1446,
+      "step": 840
+    },
+    {
+      "epoch": 0.05561007523716061,
+      "grad_norm": 0.9674533605575562,
+      "learning_rate": 9.433333333333335e-06,
+      "loss": 0.1534,
+      "step": 850
+    },
+    {
+      "epoch": 0.056264311416421325,
+      "grad_norm": 0.8930101990699768,
+      "learning_rate": 9.544444444444445e-06,
+      "loss": 0.154,
+      "step": 860
+    },
+    {
+      "epoch": 0.05691854759568204,
+      "grad_norm": 1.0106406211853027,
+      "learning_rate": 9.655555555555557e-06,
+      "loss": 0.1438,
+      "step": 870
+    },
+    {
+      "epoch": 0.05757278377494275,
+      "grad_norm": 1.0581188201904297,
+      "learning_rate": 9.766666666666667e-06,
+      "loss": 0.1465,
+      "step": 880
+    },
+    {
+      "epoch": 0.058227019954203464,
+      "grad_norm": 1.138856291770935,
+      "learning_rate": 9.87777777777778e-06,
+      "loss": 0.1439,
+      "step": 890
+    },
+    {
+      "epoch": 0.058881256133464184,
+      "grad_norm": 1.0272846221923828,
+      "learning_rate": 9.98888888888889e-06,
+      "loss": 0.1627,
+      "step": 900
+    },
+    {
+      "epoch": 0.0595354923127249,
+      "grad_norm": 1.071536898612976,
+      "learning_rate": 1.0100000000000002e-05,
+      "loss": 0.1544,
+      "step": 910
+    },
+    {
+      "epoch": 0.06018972849198561,
+      "grad_norm": 0.9497654438018799,
+      "learning_rate": 1.0211111111111112e-05,
+      "loss": 0.1433,
+      "step": 920
+    },
+    {
+      "epoch": 0.06084396467124632,
+      "grad_norm": 0.9319833517074585,
+      "learning_rate": 1.0322222222222224e-05,
+      "loss": 0.151,
+      "step": 930
+    },
+    {
+      "epoch": 0.061498200850507036,
+      "grad_norm": 0.8812747001647949,
+      "learning_rate": 1.0433333333333334e-05,
+      "loss": 0.1307,
+      "step": 940
+    },
+    {
+      "epoch": 0.06215243702976775,
+      "grad_norm": 0.9672985672950745,
+      "learning_rate": 1.0544444444444444e-05,
+      "loss": 0.1453,
+      "step": 950
+    },
+    {
+      "epoch": 0.06280667320902845,
+      "grad_norm": 0.9534893035888672,
+      "learning_rate": 1.0655555555555556e-05,
+      "loss": 0.1365,
+      "step": 960
+    },
+    {
+      "epoch": 0.06346090938828917,
+      "grad_norm": 0.9244397282600403,
+      "learning_rate": 1.0766666666666666e-05,
+      "loss": 0.1472,
+      "step": 970
+    },
+    {
+      "epoch": 0.06411514556754988,
+      "grad_norm": 0.9510551691055298,
+      "learning_rate": 1.0877777777777778e-05,
+      "loss": 0.1315,
+      "step": 980
+    },
+    {
+      "epoch": 0.0647693817468106,
+      "grad_norm": 1.0477503538131714,
+      "learning_rate": 1.0988888888888889e-05,
+      "loss": 0.137,
+      "step": 990
+    },
+    {
+      "epoch": 0.0654236179260713,
+      "grad_norm": 0.7996601462364197,
+      "learning_rate": 1.11e-05,
+      "loss": 0.1334,
+      "step": 1000
+    },
+    {
+      "epoch": 0.06607785410533203,
+      "grad_norm": 0.9080948829650879,
+      "learning_rate": 1.121111111111111e-05,
+      "loss": 0.1443,
+      "step": 1010
+    },
+    {
+      "epoch": 0.06673209028459273,
+      "grad_norm": 1.1308685541152954,
+      "learning_rate": 1.1322222222222223e-05,
+      "loss": 0.1424,
+      "step": 1020
+    },
+    {
+      "epoch": 0.06738632646385345,
+      "grad_norm": 0.9781653881072998,
+      "learning_rate": 1.1433333333333333e-05,
+      "loss": 0.1332,
+      "step": 1030
+    },
+    {
+      "epoch": 0.06804056264311416,
+      "grad_norm": 0.9940928220748901,
+      "learning_rate": 1.1544444444444445e-05,
+      "loss": 0.1449,
+      "step": 1040
+    },
+    {
+      "epoch": 0.06869479882237488,
+      "grad_norm": 1.109440803527832,
+      "learning_rate": 1.1655555555555555e-05,
+      "loss": 0.1373,
+      "step": 1050
+    },
+    {
+      "epoch": 0.06934903500163558,
+      "grad_norm": 0.7790347337722778,
+      "learning_rate": 1.1766666666666667e-05,
+      "loss": 0.1441,
+      "step": 1060
+    },
+    {
+      "epoch": 0.0700032711808963,
+      "grad_norm": 0.9548224210739136,
+      "learning_rate": 1.1877777777777777e-05,
+      "loss": 0.1512,
+      "step": 1070
+    },
+    {
+      "epoch": 0.07065750736015702,
+      "grad_norm": 0.9746776223182678,
+      "learning_rate": 1.1988888888888889e-05,
+      "loss": 0.1412,
+      "step": 1080
+    },
+    {
+      "epoch": 0.07131174353941773,
+      "grad_norm": 1.2580838203430176,
+      "learning_rate": 1.2100000000000001e-05,
+      "loss": 0.153,
+      "step": 1090
+    },
+    {
+      "epoch": 0.07196597971867845,
+      "grad_norm": 1.0998780727386475,
+      "learning_rate": 1.2211111111111111e-05,
+      "loss": 0.1396,
+      "step": 1100
+    },
+    {
+      "epoch": 0.07262021589793916,
+      "grad_norm": 0.9261326789855957,
+      "learning_rate": 1.2322222222222223e-05,
+      "loss": 0.1431,
+      "step": 1110
+    },
+    {
+      "epoch": 0.07327445207719988,
+      "grad_norm": 0.9783452749252319,
+      "learning_rate": 1.2433333333333335e-05,
+      "loss": 0.147,
+      "step": 1120
+    },
+    {
+      "epoch": 0.07392868825646058,
+      "grad_norm": 0.9785774946212769,
+      "learning_rate": 1.2544444444444445e-05,
+      "loss": 0.1418,
+      "step": 1130
+    },
+    {
+      "epoch": 0.0745829244357213,
+      "grad_norm": 1.0473101139068604,
+      "learning_rate": 1.2655555555555557e-05,
+      "loss": 0.1397,
+      "step": 1140
+    },
+    {
+      "epoch": 0.07523716061498201,
+      "grad_norm": 0.972935140132904,
+      "learning_rate": 1.276666666666667e-05,
+      "loss": 0.1394,
+      "step": 1150
+    },
+    {
+      "epoch": 0.07589139679424273,
+      "grad_norm": 0.9036497473716736,
+      "learning_rate": 1.2877777777777778e-05,
+      "loss": 0.1465,
+      "step": 1160
+    },
+    {
+      "epoch": 0.07654563297350343,
+      "grad_norm": 1.1967897415161133,
+      "learning_rate": 1.298888888888889e-05,
+      "loss": 0.1332,
+      "step": 1170
+    },
+    {
+      "epoch": 0.07719986915276415,
+      "grad_norm": 0.875632643699646,
+      "learning_rate": 1.3100000000000002e-05,
+      "loss": 0.14,
+      "step": 1180
+    },
+    {
+      "epoch": 0.07785410533202486,
+      "grad_norm": 0.9649753570556641,
+      "learning_rate": 1.3211111111111114e-05,
+      "loss": 0.1324,
+      "step": 1190
+    },
+    {
+      "epoch": 0.07850834151128558,
+      "grad_norm": 0.8261013627052307,
+      "learning_rate": 1.3322222222222222e-05,
+      "loss": 0.1333,
+      "step": 1200
+    },
+    {
+      "epoch": 0.07916257769054628,
+      "grad_norm": 1.0080054998397827,
+      "learning_rate": 1.3433333333333334e-05,
+      "loss": 0.1367,
+      "step": 1210
+    },
+    {
+      "epoch": 0.079816813869807,
+      "grad_norm": 1.0562273263931274,
+      "learning_rate": 1.3544444444444446e-05,
+      "loss": 0.1329,
+      "step": 1220
+    },
+    {
+      "epoch": 0.08047105004906771,
+      "grad_norm": 0.9354407787322998,
+      "learning_rate": 1.3655555555555558e-05,
+      "loss": 0.134,
+      "step": 1230
+    },
+    {
+      "epoch": 0.08112528622832843,
+      "grad_norm": 0.9830518364906311,
+      "learning_rate": 1.3766666666666666e-05,
+      "loss": 0.135,
+      "step": 1240
+    },
+    {
+      "epoch": 0.08177952240758914,
+      "grad_norm": 0.951457679271698,
+      "learning_rate": 1.3877777777777778e-05,
+      "loss": 0.1339,
+      "step": 1250
+    },
+    {
+      "epoch": 0.08243375858684986,
+      "grad_norm": 0.887077271938324,
+      "learning_rate": 1.398888888888889e-05,
+      "loss": 0.1436,
+      "step": 1260
+    },
+    {
+      "epoch": 0.08308799476611056,
+      "grad_norm": 0.9000831842422485,
+      "learning_rate": 1.4099999999999999e-05,
+      "loss": 0.1367,
+      "step": 1270
+    },
+    {
+      "epoch": 0.08374223094537128,
+      "grad_norm": 1.0145295858383179,
+      "learning_rate": 1.421111111111111e-05,
+      "loss": 0.1491,
+      "step": 1280
+    },
+    {
+      "epoch": 0.08439646712463199,
+      "grad_norm": 0.8529645800590515,
+      "learning_rate": 1.4322222222222223e-05,
+      "loss": 0.1278,
+      "step": 1290
+    },
+    {
+      "epoch": 0.08505070330389271,
+      "grad_norm": 1.141903042793274,
+      "learning_rate": 1.4433333333333335e-05,
+      "loss": 0.1319,
+      "step": 1300
+    },
+    {
+      "epoch": 0.08570493948315341,
+      "grad_norm": 1.0339152812957764,
+      "learning_rate": 1.4544444444444443e-05,
+      "loss": 0.1362,
+      "step": 1310
+    },
+    {
+      "epoch": 0.08635917566241413,
+      "grad_norm": 0.9347769021987915,
+      "learning_rate": 1.4655555555555555e-05,
+      "loss": 0.1382,
+      "step": 1320
+    },
+    {
+      "epoch": 0.08701341184167484,
+      "grad_norm": 1.0565035343170166,
+      "learning_rate": 1.4766666666666667e-05,
+      "loss": 0.1321,
+      "step": 1330
+    },
+    {
+      "epoch": 0.08766764802093556,
+      "grad_norm": 0.9495023488998413,
+      "learning_rate": 1.4877777777777779e-05,
+      "loss": 0.1486,
+      "step": 1340
+    },
+    {
+      "epoch": 0.08832188420019627,
+      "grad_norm": 0.9335034489631653,
+      "learning_rate": 1.498888888888889e-05,
+      "loss": 0.1413,
+      "step": 1350
+    },
+    {
+      "epoch": 0.08897612037945699,
+      "grad_norm": 0.8525314331054688,
+      "learning_rate": 1.51e-05,
+      "loss": 0.1265,
+      "step": 1360
+    },
+    {
+      "epoch": 0.08963035655871769,
+      "grad_norm": 1.1363369226455688,
+      "learning_rate": 1.5211111111111111e-05,
+      "loss": 0.1397,
+      "step": 1370
+    },
+    {
+      "epoch": 0.09028459273797841,
+      "grad_norm": 0.892582893371582,
+      "learning_rate": 1.5322222222222225e-05,
+      "loss": 0.1314,
+      "step": 1380
+    },
+    {
+      "epoch": 0.09093882891723912,
+      "grad_norm": 1.058522343635559,
+      "learning_rate": 1.5433333333333334e-05,
+      "loss": 0.1356,
+      "step": 1390
+    },
+    {
+      "epoch": 0.09159306509649984,
+      "grad_norm": 0.9127554297447205,
+      "learning_rate": 1.5544444444444445e-05,
+      "loss": 0.1368,
+      "step": 1400
+    },
+    {
+      "epoch": 0.09224730127576054,
+      "grad_norm": 1.5529781579971313,
+      "learning_rate": 1.5655555555555557e-05,
+      "loss": 0.1406,
+      "step": 1410
+    },
+    {
+      "epoch": 0.09290153745502126,
+      "grad_norm": 0.9582400321960449,
+      "learning_rate": 1.576666666666667e-05,
+      "loss": 0.1315,
+      "step": 1420
+    },
+    {
+      "epoch": 0.09355577363428197,
+      "grad_norm": 0.9931496977806091,
+      "learning_rate": 1.5877777777777778e-05,
+      "loss": 0.1327,
+      "step": 1430
+    },
+    {
+      "epoch": 0.09421000981354269,
+      "grad_norm": 0.9021315574645996,
+      "learning_rate": 1.598888888888889e-05,
+      "loss": 0.1335,
+      "step": 1440
+    },
+    {
+      "epoch": 0.09486424599280341,
+      "grad_norm": 0.9401276707649231,
+      "learning_rate": 1.6100000000000002e-05,
+      "loss": 0.1272,
+      "step": 1450
+    },
+    {
+      "epoch": 0.09551848217206411,
+      "grad_norm": 0.9633158445358276,
+      "learning_rate": 1.6211111111111114e-05,
+      "loss": 0.1359,
+      "step": 1460
+    },
+    {
+      "epoch": 0.09617271835132483,
+      "grad_norm": 1.0396912097930908,
+      "learning_rate": 1.6322222222222222e-05,
+      "loss": 0.1298,
+      "step": 1470
+    },
+    {
+      "epoch": 0.09682695453058554,
+      "grad_norm": 0.9414869546890259,
+      "learning_rate": 1.6433333333333334e-05,
+      "loss": 0.1288,
+      "step": 1480
+    },
+    {
+      "epoch": 0.09748119070984626,
+      "grad_norm": 0.8435006737709045,
+      "learning_rate": 1.6544444444444446e-05,
+      "loss": 0.1303,
+      "step": 1490
+    },
+    {
+      "epoch": 0.09813542688910697,
+      "grad_norm": 0.9831594824790955,
+      "learning_rate": 1.6655555555555558e-05,
+      "loss": 0.1439,
+      "step": 1500
+    },
+    {
+      "epoch": 0.09878966306836769,
+      "grad_norm": 1.0466912984848022,
+      "learning_rate": 1.6766666666666667e-05,
+      "loss": 0.1277,
+      "step": 1510
+    },
+    {
+      "epoch": 0.09944389924762839,
+      "grad_norm": 0.9995728731155396,
+      "learning_rate": 1.687777777777778e-05,
+      "loss": 0.1297,
+      "step": 1520
+    },
+    {
+      "epoch": 0.10009813542688911,
+      "grad_norm": 1.018913745880127,
+      "learning_rate": 1.698888888888889e-05,
+      "loss": 0.1341,
+      "step": 1530
+    },
+    {
+      "epoch": 0.10075237160614982,
+      "grad_norm": 1.1046700477600098,
+      "learning_rate": 1.7100000000000002e-05,
+      "loss": 0.1332,
+      "step": 1540
+    },
+    {
+      "epoch": 0.10140660778541054,
+      "grad_norm": 1.0834091901779175,
+      "learning_rate": 1.721111111111111e-05,
+      "loss": 0.1349,
+      "step": 1550
+    },
+    {
+      "epoch": 0.10206084396467124,
+      "grad_norm": 1.079143762588501,
+      "learning_rate": 1.7322222222222223e-05,
+      "loss": 0.1398,
+      "step": 1560
+    },
+    {
+      "epoch": 0.10271508014393196,
+      "grad_norm": 1.0492314100265503,
+      "learning_rate": 1.7433333333333335e-05,
+      "loss": 0.1411,
+      "step": 1570
+    },
+    {
+      "epoch": 0.10336931632319267,
+      "grad_norm": 1.171302318572998,
+      "learning_rate": 1.7544444444444443e-05,
+      "loss": 0.1312,
+      "step": 1580
+    },
+    {
+      "epoch": 0.10402355250245339,
+      "grad_norm": 1.0378974676132202,
+      "learning_rate": 1.7655555555555555e-05,
+      "loss": 0.1254,
+      "step": 1590
+    },
+    {
+      "epoch": 0.1046777886817141,
+      "grad_norm": 0.9647935032844543,
+      "learning_rate": 1.7766666666666667e-05,
+      "loss": 0.1444,
+      "step": 1600
+    },
+    {
+      "epoch": 0.10533202486097482,
+      "grad_norm": 0.9934017658233643,
+      "learning_rate": 1.787777777777778e-05,
+      "loss": 0.1344,
+      "step": 1610
+    },
+    {
+      "epoch": 0.10598626104023552,
+      "grad_norm": 1.0105055570602417,
+      "learning_rate": 1.7988888888888888e-05,
+      "loss": 0.1397,
+      "step": 1620
+    },
+    {
+      "epoch": 0.10664049721949624,
+      "grad_norm": 1.05049729347229,
+      "learning_rate": 1.81e-05,
+      "loss": 0.1348,
+      "step": 1630
+    },
+    {
+      "epoch": 0.10729473339875695,
+      "grad_norm": 1.0454671382904053,
+      "learning_rate": 1.821111111111111e-05,
+      "loss": 0.1239,
+      "step": 1640
+    },
+    {
+      "epoch": 0.10794896957801767,
+      "grad_norm": 1.0142239332199097,
+      "learning_rate": 1.8322222222222223e-05,
+      "loss": 0.1148,
+      "step": 1650
+    },
+    {
+      "epoch": 0.10860320575727837,
+      "grad_norm": 1.1496508121490479,
+      "learning_rate": 1.8433333333333332e-05,
+      "loss": 0.1308,
+      "step": 1660
+    },
+    {
+      "epoch": 0.1092574419365391,
+      "grad_norm": 0.9551987051963806,
+      "learning_rate": 1.8544444444444444e-05,
+      "loss": 0.1339,
+      "step": 1670
+    },
+    {
+      "epoch": 0.1099116781157998,
+      "grad_norm": 0.8923566937446594,
+      "learning_rate": 1.8655555555555556e-05,
+      "loss": 0.1286,
+      "step": 1680
+    },
+    {
+      "epoch": 0.11056591429506052,
+      "grad_norm": 1.0510790348052979,
+      "learning_rate": 1.8766666666666668e-05,
+      "loss": 0.1362,
+      "step": 1690
+    },
+    {
+      "epoch": 0.11122015047432122,
+      "grad_norm": 0.9121619462966919,
+      "learning_rate": 1.8877777777777776e-05,
+      "loss": 0.1273,
+      "step": 1700
+    },
+    {
+      "epoch": 0.11187438665358194,
+      "grad_norm": 1.1024173498153687,
+      "learning_rate": 1.8988888888888888e-05,
+      "loss": 0.1253,
+      "step": 1710
+    },
+    {
+      "epoch": 0.11252862283284265,
+      "grad_norm": 1.032799482345581,
+      "learning_rate": 1.91e-05,
+      "loss": 0.1313,
+      "step": 1720
+    },
+    {
+      "epoch": 0.11318285901210337,
+      "grad_norm": 1.0584205389022827,
+      "learning_rate": 1.9211111111111112e-05,
+      "loss": 0.1342,
+      "step": 1730
+    },
+    {
+      "epoch": 0.11383709519136408,
+      "grad_norm": 0.8857603669166565,
+      "learning_rate": 1.932222222222222e-05,
+      "loss": 0.1308,
+      "step": 1740
+    },
+    {
+      "epoch": 0.1144913313706248,
+      "grad_norm": 0.9453005194664001,
+      "learning_rate": 1.9433333333333332e-05,
+      "loss": 0.1181,
+      "step": 1750
+    },
+    {
+      "epoch": 0.1151455675498855,
+      "grad_norm": 1.0007144212722778,
+      "learning_rate": 1.9544444444444444e-05,
+      "loss": 0.1233,
+      "step": 1760
+    },
+    {
+      "epoch": 0.11579980372914622,
+      "grad_norm": 1.0010242462158203,
+      "learning_rate": 1.9655555555555556e-05,
+      "loss": 0.1218,
+      "step": 1770
+    },
+    {
+      "epoch": 0.11645403990840693,
+      "grad_norm": 0.9568318128585815,
+      "learning_rate": 1.9766666666666668e-05,
+      "loss": 0.127,
+      "step": 1780
+    },
+    {
+      "epoch": 0.11710827608766765,
+      "grad_norm": 0.9792290925979614,
+      "learning_rate": 1.9877777777777777e-05,
+      "loss": 0.135,
+      "step": 1790
+    },
+    {
+      "epoch": 0.11776251226692837,
+      "grad_norm": 1.0170482397079468,
+      "learning_rate": 1.998888888888889e-05,
+      "loss": 0.1401,
+      "step": 1800
+    },
+    {
+      "epoch": 0.11841674844618907,
+      "grad_norm": 0.949272632598877,
+      "learning_rate": 2.01e-05,
+      "loss": 0.1319,
+      "step": 1810
+    },
+    {
+      "epoch": 0.1190709846254498,
+      "grad_norm": 0.8818572163581848,
+      "learning_rate": 2.0211111111111113e-05,
+      "loss": 0.1277,
+      "step": 1820
+    },
+    {
+      "epoch": 0.1197252208047105,
+      "grad_norm": 0.9168890118598938,
+      "learning_rate": 2.0322222222222225e-05,
+      "loss": 0.1168,
+      "step": 1830
+    },
+    {
+      "epoch": 0.12037945698397122,
+      "grad_norm": 1.180184006690979,
+      "learning_rate": 2.0433333333333336e-05,
+      "loss": 0.1281,
+      "step": 1840
+    },
+    {
+      "epoch": 0.12103369316323193,
+      "grad_norm": 1.0016313791275024,
+      "learning_rate": 2.054444444444445e-05,
+      "loss": 0.1351,
+      "step": 1850
+    },
+    {
+      "epoch": 0.12168792934249265,
+      "grad_norm": 0.9246931672096252,
+      "learning_rate": 2.0655555555555557e-05,
+      "loss": 0.1311,
+      "step": 1860
+    },
+    {
+      "epoch": 0.12234216552175335,
+      "grad_norm": 1.3594558238983154,
+      "learning_rate": 2.076666666666667e-05,
+      "loss": 0.1414,
+      "step": 1870
+    },
+    {
+      "epoch": 0.12299640170101407,
+      "grad_norm": 0.9774754643440247,
+      "learning_rate": 2.087777777777778e-05,
+      "loss": 0.1155,
+      "step": 1880
+    },
+    {
+      "epoch": 0.12365063788027478,
+      "grad_norm": 1.1299134492874146,
+      "learning_rate": 2.0988888888888893e-05,
+      "loss": 0.1167,
+      "step": 1890
+    },
+    {
+      "epoch": 0.1243048740595355,
+      "grad_norm": 0.9047524333000183,
+      "learning_rate": 2.11e-05,
+      "loss": 0.1306,
+      "step": 1900
+    },
+    {
+      "epoch": 0.1249591102387962,
+      "grad_norm": 1.080834150314331,
+      "learning_rate": 2.1211111111111113e-05,
+      "loss": 0.1234,
+      "step": 1910
+    },
+    {
+      "epoch": 0.1256133464180569,
+      "grad_norm": 0.8716378211975098,
+      "learning_rate": 2.1322222222222225e-05,
+      "loss": 0.1265,
+      "step": 1920
+    },
+    {
+      "epoch": 0.12626758259731763,
+      "grad_norm": 0.9985718727111816,
+      "learning_rate": 2.1433333333333334e-05,
+      "loss": 0.1284,
+      "step": 1930
+    },
+    {
+      "epoch": 0.12692181877657835,
+      "grad_norm": 1.0242950916290283,
+      "learning_rate": 2.1544444444444446e-05,
+      "loss": 0.1267,
+      "step": 1940
+    },
+    {
+      "epoch": 0.12757605495583907,
+      "grad_norm": 1.1003845930099487,
+      "learning_rate": 2.1655555555555558e-05,
+      "loss": 0.1273,
+      "step": 1950
+    },
+    {
+      "epoch": 0.12823029113509976,
+      "grad_norm": 0.8267995119094849,
+      "learning_rate": 2.176666666666667e-05,
+      "loss": 0.1261,
+      "step": 1960
+    },
+    {
+      "epoch": 0.12888452731436048,
+      "grad_norm": 1.2606829404830933,
+      "learning_rate": 2.1877777777777778e-05,
+      "loss": 0.1166,
+      "step": 1970
+    },
+    {
+      "epoch": 0.1295387634936212,
+      "grad_norm": 0.8288832306861877,
+      "learning_rate": 2.198888888888889e-05,
+      "loss": 0.1128,
+      "step": 1980
+    },
+    {
+      "epoch": 0.13019299967288192,
+      "grad_norm": 1.1182180643081665,
+      "learning_rate": 2.2100000000000002e-05,
+      "loss": 0.1281,
+      "step": 1990
+    },
+    {
+      "epoch": 0.1308472358521426,
+      "grad_norm": 0.8783948421478271,
+      "learning_rate": 2.2211111111111114e-05,
+      "loss": 0.1277,
+      "step": 2000
+    },
+    {
+      "epoch": 0.13150147203140333,
+      "grad_norm": 1.0937625169754028,
+      "learning_rate": 2.2322222222222222e-05,
+      "loss": 0.1294,
+      "step": 2010
+    },
+    {
+      "epoch": 0.13215570821066405,
+      "grad_norm": 0.9571072459220886,
+      "learning_rate": 2.2433333333333334e-05,
+      "loss": 0.1109,
+      "step": 2020
+    },
+    {
+      "epoch": 0.13280994438992477,
+      "grad_norm": 1.1977931261062622,
+      "learning_rate": 2.2544444444444446e-05,
+      "loss": 0.1331,
+      "step": 2030
+    },
+    {
+      "epoch": 0.13346418056918546,
+      "grad_norm": 0.9299104809761047,
+      "learning_rate": 2.2655555555555558e-05,
+      "loss": 0.1351,
+      "step": 2040
+    },
+    {
+      "epoch": 0.13411841674844618,
+      "grad_norm": 0.8753429651260376,
+      "learning_rate": 2.2766666666666667e-05,
+      "loss": 0.1365,
+      "step": 2050
+    },
+    {
+      "epoch": 0.1347726529277069,
+      "grad_norm": 1.0041842460632324,
+      "learning_rate": 2.287777777777778e-05,
+      "loss": 0.1303,
+      "step": 2060
+    },
+    {
+      "epoch": 0.13542688910696762,
+      "grad_norm": 1.1017142534255981,
+      "learning_rate": 2.298888888888889e-05,
+      "loss": 0.124,
+      "step": 2070
+    },
+    {
+      "epoch": 0.13608112528622832,
+      "grad_norm": 1.038955569267273,
+      "learning_rate": 2.3100000000000002e-05,
+      "loss": 0.1275,
+      "step": 2080
+    },
+    {
+      "epoch": 0.13673536146548904,
+      "grad_norm": 0.8835306763648987,
+      "learning_rate": 2.321111111111111e-05,
+      "loss": 0.1292,
+      "step": 2090
+    },
+    {
+      "epoch": 0.13738959764474976,
+      "grad_norm": 1.067949652671814,
+      "learning_rate": 2.3322222222222223e-05,
+      "loss": 0.1308,
+      "step": 2100
+    },
+    {
+      "epoch": 0.13804383382401048,
+      "grad_norm": 0.9632567167282104,
+      "learning_rate": 2.3433333333333335e-05,
+      "loss": 0.1389,
+      "step": 2110
+    },
+    {
+      "epoch": 0.13869807000327117,
+      "grad_norm": 0.9398028254508972,
+      "learning_rate": 2.3544444444444447e-05,
+      "loss": 0.1242,
+      "step": 2120
+    },
+    {
+      "epoch": 0.1393523061825319,
+      "grad_norm": 0.9722338914871216,
+      "learning_rate": 2.3655555555555555e-05,
+      "loss": 0.1179,
+      "step": 2130
+    },
+    {
+      "epoch": 0.1400065423617926,
+      "grad_norm": 0.8309763669967651,
+      "learning_rate": 2.3766666666666667e-05,
+      "loss": 0.1232,
+      "step": 2140
+    },
+    {
+      "epoch": 0.14066077854105333,
+      "grad_norm": 0.8105255961418152,
+      "learning_rate": 2.387777777777778e-05,
+      "loss": 0.1308,
+      "step": 2150
+    },
+    {
+      "epoch": 0.14131501472031405,
+      "grad_norm": 0.8585493564605713,
+      "learning_rate": 2.398888888888889e-05,
+      "loss": 0.1196,
+      "step": 2160
+    },
+    {
+      "epoch": 0.14196925089957474,
+      "grad_norm": 0.9327337145805359,
+      "learning_rate": 2.41e-05,
+      "loss": 0.1212,
+      "step": 2170
+    },
+    {
+      "epoch": 0.14262348707883546,
+      "grad_norm": 0.867504358291626,
+      "learning_rate": 2.421111111111111e-05,
+      "loss": 0.1131,
+      "step": 2180
+    },
+    {
+      "epoch": 0.14327772325809618,
+      "grad_norm": 0.8426870703697205,
+      "learning_rate": 2.4322222222222224e-05,
+      "loss": 0.1239,
+      "step": 2190
+    },
+    {
+      "epoch": 0.1439319594373569,
+      "grad_norm": 0.9085756540298462,
+      "learning_rate": 2.4433333333333335e-05,
+      "loss": 0.1288,
+      "step": 2200
+    },
+    {
+      "epoch": 0.1445861956166176,
+      "grad_norm": 0.9379426836967468,
+      "learning_rate": 2.4544444444444444e-05,
+      "loss": 0.1168,
+      "step": 2210
+    },
+    {
+      "epoch": 0.1452404317958783,
+      "grad_norm": 0.9728325009346008,
+      "learning_rate": 2.4655555555555556e-05,
+      "loss": 0.1237,
+      "step": 2220
+    },
+    {
+      "epoch": 0.14589466797513903,
+      "grad_norm": 0.9717603921890259,
+      "learning_rate": 2.4766666666666668e-05,
+      "loss": 0.1172,
+      "step": 2230
+    },
+    {
+      "epoch": 0.14654890415439975,
+      "grad_norm": 0.9177740812301636,
+      "learning_rate": 2.4877777777777776e-05,
+      "loss": 0.1172,
+      "step": 2240
+    },
+    {
+      "epoch": 0.14720314033366044,
+      "grad_norm": 1.4036056995391846,
+      "learning_rate": 2.498888888888889e-05,
+      "loss": 0.125,
+      "step": 2250
+    },
+    {
+      "epoch": 0.14785737651292116,
+      "grad_norm": 1.0242791175842285,
+      "learning_rate": 2.51e-05,
+      "loss": 0.1333,
+      "step": 2260
+    },
+    {
+      "epoch": 0.14851161269218188,
+      "grad_norm": 0.8376342058181763,
+      "learning_rate": 2.5211111111111112e-05,
+      "loss": 0.1227,
+      "step": 2270
+    },
+    {
+      "epoch": 0.1491658488714426,
+      "grad_norm": 0.9228038191795349,
+      "learning_rate": 2.5322222222222224e-05,
+      "loss": 0.1296,
+      "step": 2280
+    },
+    {
+      "epoch": 0.1498200850507033,
+      "grad_norm": 0.8739307522773743,
+      "learning_rate": 2.5433333333333336e-05,
+      "loss": 0.1086,
+      "step": 2290
+    },
+    {
+      "epoch": 0.15047432122996401,
+      "grad_norm": 0.9078662991523743,
+      "learning_rate": 2.5544444444444445e-05,
+      "loss": 0.1307,
+      "step": 2300
+    },
+    {
+      "epoch": 0.15112855740922473,
+      "grad_norm": 1.0658289194107056,
+      "learning_rate": 2.5655555555555557e-05,
+      "loss": 0.128,
+      "step": 2310
+    },
+    {
+      "epoch": 0.15178279358848545,
+      "grad_norm": 0.8755213618278503,
+      "learning_rate": 2.5766666666666665e-05,
+      "loss": 0.1305,
+      "step": 2320
+    },
+    {
+      "epoch": 0.15243702976774615,
+      "grad_norm": 1.0628185272216797,
+      "learning_rate": 2.5877777777777777e-05,
+      "loss": 0.1244,
+      "step": 2330
+    },
+    {
+      "epoch": 0.15309126594700687,
+      "grad_norm": 1.0583535432815552,
+      "learning_rate": 2.598888888888889e-05,
+      "loss": 0.1257,
+      "step": 2340
+    },
+    {
+      "epoch": 0.15374550212626759,
+      "grad_norm": 0.9537503719329834,
+      "learning_rate": 2.61e-05,
+      "loss": 0.1241,
+      "step": 2350
+    },
+    {
+      "epoch": 0.1543997383055283,
+      "grad_norm": 1.0985156297683716,
+      "learning_rate": 2.6211111111111113e-05,
+      "loss": 0.1359,
+      "step": 2360
+    },
+    {
+      "epoch": 0.155053974484789,
+      "grad_norm": 0.9252329468727112,
+      "learning_rate": 2.6322222222222225e-05,
+      "loss": 0.1317,
+      "step": 2370
+    },
+    {
+      "epoch": 0.15570821066404972,
+      "grad_norm": 1.1681864261627197,
+      "learning_rate": 2.6433333333333333e-05,
+      "loss": 0.1242,
+      "step": 2380
+    },
+    {
+      "epoch": 0.15636244684331044,
+      "grad_norm": 1.2066115140914917,
+      "learning_rate": 2.6544444444444445e-05,
+      "loss": 0.1152,
+      "step": 2390
+    },
+    {
+      "epoch": 0.15701668302257116,
+      "grad_norm": 1.1337898969650269,
+      "learning_rate": 2.6655555555555557e-05,
+      "loss": 0.1334,
+      "step": 2400
+    },
+    {
+      "epoch": 0.15767091920183185,
+      "grad_norm": 0.9998272657394409,
+      "learning_rate": 2.676666666666667e-05,
+      "loss": 0.1189,
+      "step": 2410
+    },
+    {
+      "epoch": 0.15832515538109257,
+      "grad_norm": 0.8952450752258301,
+      "learning_rate": 2.687777777777778e-05,
+      "loss": 0.1227,
+      "step": 2420
+    },
+    {
+      "epoch": 0.1589793915603533,
+      "grad_norm": 1.0793391466140747,
+      "learning_rate": 2.6988888888888893e-05,
+      "loss": 0.1198,
+      "step": 2430
+    },
+    {
+      "epoch": 0.159633627739614,
+      "grad_norm": 1.016587734222412,
+      "learning_rate": 2.7100000000000005e-05,
+      "loss": 0.1208,
+      "step": 2440
+    },
+    {
+      "epoch": 0.1602878639188747,
+      "grad_norm": 1.0260379314422607,
+      "learning_rate": 2.7211111111111113e-05,
+      "loss": 0.1258,
+      "step": 2450
+    },
+    {
+      "epoch": 0.16094210009813542,
+      "grad_norm": 0.9341796636581421,
+      "learning_rate": 2.7322222222222222e-05,
+      "loss": 0.1172,
+      "step": 2460
+    },
+    {
+      "epoch": 0.16159633627739614,
+      "grad_norm": 1.0528295040130615,
+      "learning_rate": 2.7433333333333334e-05,
+      "loss": 0.124,
+      "step": 2470
+    },
+    {
+      "epoch": 0.16225057245665686,
+      "grad_norm": 0.831092894077301,
+      "learning_rate": 2.7544444444444446e-05,
+      "loss": 0.1143,
+      "step": 2480
+    },
+    {
+      "epoch": 0.16290480863591755,
+      "grad_norm": 1.0132439136505127,
+      "learning_rate": 2.7655555555555558e-05,
+      "loss": 0.134,
+      "step": 2490
+    },
+    {
+      "epoch": 0.16355904481517827,
+      "grad_norm": 1.041785717010498,
+      "learning_rate": 2.776666666666667e-05,
+      "loss": 0.1328,
+      "step": 2500
+    },
+    {
+      "epoch": 0.164213280994439,
+      "grad_norm": 1.1009939908981323,
+      "learning_rate": 2.787777777777778e-05,
+      "loss": 0.1262,
+      "step": 2510
+    },
+    {
+      "epoch": 0.1648675171736997,
+      "grad_norm": 1.0970548391342163,
+      "learning_rate": 2.7988888888888893e-05,
+      "loss": 0.132,
+      "step": 2520
+    },
+    {
+      "epoch": 0.16552175335296043,
+      "grad_norm": 1.0594844818115234,
+      "learning_rate": 2.8100000000000005e-05,
+      "loss": 0.1267,
+      "step": 2530
+    },
+    {
+      "epoch": 0.16617598953222112,
+      "grad_norm": 1.1462485790252686,
+      "learning_rate": 2.821111111111111e-05,
+      "loss": 0.1214,
+      "step": 2540
+    },
+    {
+      "epoch": 0.16683022571148184,
+      "grad_norm": 1.044067621231079,
+      "learning_rate": 2.8322222222222222e-05,
+      "loss": 0.1287,
+      "step": 2550
+    },
+    {
+      "epoch": 0.16748446189074256,
+      "grad_norm": 1.0372158288955688,
+      "learning_rate": 2.8433333333333334e-05,
+      "loss": 0.1297,
+      "step": 2560
+    },
+    {
+      "epoch": 0.16813869807000328,
+      "grad_norm": 1.0640093088150024,
+      "learning_rate": 2.8544444444444446e-05,
+      "loss": 0.1266,
+      "step": 2570
+    },
+    {
+      "epoch": 0.16879293424926398,
+      "grad_norm": 0.9317982792854309,
+      "learning_rate": 2.8655555555555558e-05,
+      "loss": 0.1283,
+      "step": 2580
+    },
+    {
+      "epoch": 0.1694471704285247,
+      "grad_norm": 1.5074639320373535,
+      "learning_rate": 2.876666666666667e-05,
+      "loss": 0.1105,
+      "step": 2590
+    },
+    {
+      "epoch": 0.17010140660778542,
+      "grad_norm": 1.1034104824066162,
+      "learning_rate": 2.8877777777777782e-05,
+      "loss": 0.1186,
+      "step": 2600
+    },
+    {
+      "epoch": 0.17075564278704614,
+      "grad_norm": 0.9475632905960083,
+      "learning_rate": 2.8988888888888887e-05,
+      "loss": 0.1191,
+      "step": 2610
+    },
+    {
+      "epoch": 0.17140987896630683,
+      "grad_norm": 0.9753337502479553,
+      "learning_rate": 2.91e-05,
+      "loss": 0.115,
+      "step": 2620
+    },
+    {
+      "epoch": 0.17206411514556755,
+      "grad_norm": 1.0091173648834229,
+      "learning_rate": 2.921111111111111e-05,
+      "loss": 0.1132,
+      "step": 2630
+    },
+    {
+      "epoch": 0.17271835132482827,
+      "grad_norm": 0.8938674330711365,
+      "learning_rate": 2.9322222222222223e-05,
+      "loss": 0.1221,
+      "step": 2640
+    },
+    {
+      "epoch": 0.173372587504089,
+      "grad_norm": 0.9849511384963989,
+      "learning_rate": 2.9433333333333335e-05,
+      "loss": 0.1161,
+      "step": 2650
+    },
+    {
+      "epoch": 0.17402682368334968,
+      "grad_norm": 0.9725368618965149,
+      "learning_rate": 2.9544444444444447e-05,
+      "loss": 0.1117,
+      "step": 2660
+    },
+    {
+      "epoch": 0.1746810598626104,
+      "grad_norm": 0.8807811737060547,
+      "learning_rate": 2.965555555555556e-05,
+      "loss": 0.1251,
+      "step": 2670
+    },
+    {
+      "epoch": 0.17533529604187112,
+      "grad_norm": 1.1785404682159424,
+      "learning_rate": 2.976666666666667e-05,
+      "loss": 0.1219,
+      "step": 2680
+    },
+    {
+      "epoch": 0.17598953222113184,
+      "grad_norm": 0.9174340963363647,
+      "learning_rate": 2.9877777777777776e-05,
+      "loss": 0.1099,
+      "step": 2690
+    },
+    {
+      "epoch": 0.17664376840039253,
+      "grad_norm": 1.0297743082046509,
+      "learning_rate": 2.9988888888888888e-05,
+      "loss": 0.1163,
+      "step": 2700
+    },
+    {
+      "epoch": 0.17729800457965325,
+      "grad_norm": 0.8429158926010132,
+      "learning_rate": 3.01e-05,
+      "loss": 0.1147,
+      "step": 2710
+    },
+    {
+      "epoch": 0.17795224075891397,
+      "grad_norm": 1.0993781089782715,
+      "learning_rate": 3.0211111111111112e-05,
+      "loss": 0.1241,
+      "step": 2720
+    },
+    {
+      "epoch": 0.1786064769381747,
+      "grad_norm": 1.0947730541229248,
+      "learning_rate": 3.0322222222222224e-05,
+      "loss": 0.1137,
+      "step": 2730
+    },
+    {
+      "epoch": 0.17926071311743538,
+      "grad_norm": 1.0554438829421997,
+      "learning_rate": 3.0433333333333336e-05,
+      "loss": 0.1211,
+      "step": 2740
+    },
+    {
+      "epoch": 0.1799149492966961,
+      "grad_norm": 0.987632155418396,
+      "learning_rate": 3.054444444444445e-05,
+      "loss": 0.1207,
+      "step": 2750
+    },
+    {
+      "epoch": 0.18056918547595682,
+      "grad_norm": 1.0157499313354492,
+      "learning_rate": 3.065555555555556e-05,
+      "loss": 0.1182,
+      "step": 2760
+    },
+    {
+      "epoch": 0.18122342165521754,
+      "grad_norm": 0.8833321332931519,
+      "learning_rate": 3.0766666666666665e-05,
+      "loss": 0.1202,
+      "step": 2770
+    },
+    {
+      "epoch": 0.18187765783447823,
+      "grad_norm": 1.0868935585021973,
+      "learning_rate": 3.087777777777778e-05,
+      "loss": 0.1207,
+      "step": 2780
+    },
+    {
+      "epoch": 0.18253189401373895,
+      "grad_norm": 1.1105583906173706,
+      "learning_rate": 3.098888888888889e-05,
+      "loss": 0.1242,
+      "step": 2790
+    },
+    {
+      "epoch": 0.18318613019299967,
+      "grad_norm": 1.014053225517273,
+      "learning_rate": 3.1100000000000004e-05,
+      "loss": 0.1176,
+      "step": 2800
+    },
+    {
+      "epoch": 0.1838403663722604,
+      "grad_norm": 1.2380222082138062,
+      "learning_rate": 3.121111111111111e-05,
+      "loss": 0.1231,
+      "step": 2810
+    },
+    {
+      "epoch": 0.1844946025515211,
+      "grad_norm": 0.9214850664138794,
+      "learning_rate": 3.132222222222223e-05,
+      "loss": 0.1258,
+      "step": 2820
+    },
+    {
+      "epoch": 0.1851488387307818,
+      "grad_norm": 0.9159259796142578,
+      "learning_rate": 3.1433333333333336e-05,
+      "loss": 0.1199,
+      "step": 2830
+    },
+    {
+      "epoch": 0.18580307491004253,
+      "grad_norm": 1.0562981367111206,
+      "learning_rate": 3.154444444444445e-05,
+      "loss": 0.12,
+      "step": 2840
+    },
+    {
+      "epoch": 0.18645731108930325,
+      "grad_norm": 0.9648650288581848,
+      "learning_rate": 3.165555555555555e-05,
+      "loss": 0.1168,
+      "step": 2850
+    },
+    {
+      "epoch": 0.18711154726856394,
+      "grad_norm": 1.028402328491211,
+      "learning_rate": 3.176666666666667e-05,
+      "loss": 0.1281,
+      "step": 2860
+    },
+    {
+      "epoch": 0.18776578344782466,
+      "grad_norm": 0.9670931696891785,
+      "learning_rate": 3.187777777777778e-05,
+      "loss": 0.1236,
+      "step": 2870
+    },
+    {
+      "epoch": 0.18842001962708538,
+      "grad_norm": 0.9317355751991272,
+      "learning_rate": 3.198888888888889e-05,
+      "loss": 0.113,
+      "step": 2880
+    },
+    {
+      "epoch": 0.1890742558063461,
+      "grad_norm": 0.978927493095398,
+      "learning_rate": 3.21e-05,
+      "loss": 0.1223,
+      "step": 2890
+    },
+    {
+      "epoch": 0.18972849198560682,
+      "grad_norm": 0.9678372740745544,
+      "learning_rate": 3.2211111111111116e-05,
+      "loss": 0.1246,
+      "step": 2900
+    },
+    {
+      "epoch": 0.1903827281648675,
+      "grad_norm": 0.856540322303772,
+      "learning_rate": 3.2322222222222225e-05,
+      "loss": 0.1192,
+      "step": 2910
+    },
+    {
+      "epoch": 0.19103696434412823,
+      "grad_norm": 1.0986219644546509,
+      "learning_rate": 3.243333333333333e-05,
+      "loss": 0.1189,
+      "step": 2920
+    },
+    {
+      "epoch": 0.19169120052338895,
+      "grad_norm": 0.9947769641876221,
+      "learning_rate": 3.254444444444444e-05,
+      "loss": 0.1222,
+      "step": 2930
+    },
+    {
+      "epoch": 0.19234543670264967,
+      "grad_norm": 0.9733787178993225,
+      "learning_rate": 3.265555555555556e-05,
+      "loss": 0.1214,
+      "step": 2940
+    },
+    {
+      "epoch": 0.19299967288191036,
+      "grad_norm": 1.0199573040008545,
+      "learning_rate": 3.2766666666666666e-05,
+      "loss": 0.1304,
+      "step": 2950
+    },
+    {
+      "epoch": 0.19365390906117108,
+      "grad_norm": 0.8536534905433655,
+      "learning_rate": 3.287777777777778e-05,
+      "loss": 0.1093,
+      "step": 2960
+    },
+    {
+      "epoch": 0.1943081452404318,
+      "grad_norm": 0.8914840817451477,
+      "learning_rate": 3.298888888888889e-05,
+      "loss": 0.1064,
+      "step": 2970
+    },
+    {
+      "epoch": 0.19496238141969252,
+      "grad_norm": 0.9319343566894531,
+      "learning_rate": 3.3100000000000005e-05,
+      "loss": 0.118,
+      "step": 2980
+    },
+    {
+      "epoch": 0.1956166175989532,
+      "grad_norm": 1.0728380680084229,
+      "learning_rate": 3.3211111111111114e-05,
+      "loss": 0.1202,
+      "step": 2990
+    },
+    {
+      "epoch": 0.19627085377821393,
+      "grad_norm": 0.9401952624320984,
+      "learning_rate": 3.332222222222222e-05,
+      "loss": 0.1173,
+      "step": 3000
+    },
+    {
+      "epoch": 0.19692508995747465,
+      "grad_norm": 0.9090434312820435,
+      "learning_rate": 3.343333333333333e-05,
+      "loss": 0.1285,
+      "step": 3010
+    },
+    {
+      "epoch": 0.19757932613673537,
+      "grad_norm": 1.0409756898880005,
+      "learning_rate": 3.3544444444444446e-05,
+      "loss": 0.1172,
+      "step": 3020
+    },
+    {
+      "epoch": 0.19823356231599606,
+      "grad_norm": 1.026477336883545,
+      "learning_rate": 3.3655555555555554e-05,
+      "loss": 0.1277,
+      "step": 3030
+    },
+    {
+      "epoch": 0.19888779849525678,
+      "grad_norm": 1.1095457077026367,
+      "learning_rate": 3.376666666666667e-05,
+      "loss": 0.1131,
+      "step": 3040
+    },
+    {
+      "epoch": 0.1995420346745175,
+      "grad_norm": 1.1334766149520874,
+      "learning_rate": 3.387777777777778e-05,
+      "loss": 0.1327,
+      "step": 3050
+    },
+    {
+      "epoch": 0.20019627085377822,
+      "grad_norm": 0.8890497088432312,
+      "learning_rate": 3.3988888888888894e-05,
+      "loss": 0.1209,
+      "step": 3060
+    },
+    {
+      "epoch": 0.20085050703303892,
+      "grad_norm": 0.8564671874046326,
+      "learning_rate": 3.41e-05,
+      "loss": 0.1148,
+      "step": 3070
+    },
+    {
+      "epoch": 0.20150474321229964,
+      "grad_norm": 0.9905887842178345,
+      "learning_rate": 3.421111111111111e-05,
+      "loss": 0.1045,
+      "step": 3080
+    },
+    {
+      "epoch": 0.20215897939156036,
+      "grad_norm": 1.105401873588562,
+      "learning_rate": 3.432222222222222e-05,
+      "loss": 0.1101,
+      "step": 3090
+    },
+    {
+      "epoch": 0.20281321557082108,
+      "grad_norm": 1.2804734706878662,
+      "learning_rate": 3.4433333333333335e-05,
+      "loss": 0.1264,
+      "step": 3100
+    },
+    {
+      "epoch": 0.20346745175008177,
+      "grad_norm": 0.8999707698822021,
+      "learning_rate": 3.454444444444444e-05,
+      "loss": 0.1089,
+      "step": 3110
+    },
+    {
+      "epoch": 0.2041216879293425,
+      "grad_norm": 1.0667651891708374,
+      "learning_rate": 3.465555555555556e-05,
+      "loss": 0.1171,
+      "step": 3120
+    },
+    {
+      "epoch": 0.2047759241086032,
+      "grad_norm": 0.9714770913124084,
+      "learning_rate": 3.476666666666667e-05,
+      "loss": 0.1118,
+      "step": 3130
+    },
+    {
+      "epoch": 0.20543016028786393,
+      "grad_norm": 1.0278006792068481,
+      "learning_rate": 3.487777777777778e-05,
+      "loss": 0.1239,
+      "step": 3140
+    },
+    {
+      "epoch": 0.20608439646712462,
+      "grad_norm": 1.0025670528411865,
+      "learning_rate": 3.498888888888889e-05,
+      "loss": 0.1171,
+      "step": 3150
+    },
+    {
+      "epoch": 0.20673863264638534,
+      "grad_norm": 1.191396713256836,
+      "learning_rate": 3.51e-05,
+      "loss": 0.1208,
+      "step": 3160
+    },
+    {
+      "epoch": 0.20739286882564606,
+      "grad_norm": 0.9984032511711121,
+      "learning_rate": 3.5211111111111115e-05,
+      "loss": 0.1164,
+      "step": 3170
+    },
+    {
+      "epoch": 0.20804710500490678,
+      "grad_norm": 0.863865852355957,
+      "learning_rate": 3.532222222222222e-05,
+      "loss": 0.1043,
+      "step": 3180
+    },
+    {
+      "epoch": 0.20870134118416747,
+      "grad_norm": 1.0260671377182007,
+      "learning_rate": 3.543333333333333e-05,
+      "loss": 0.1231,
+      "step": 3190
+    },
+    {
+      "epoch": 0.2093555773634282,
+      "grad_norm": 1.0903806686401367,
+      "learning_rate": 3.554444444444445e-05,
+      "loss": 0.1176,
+      "step": 3200
+    },
+    {
+      "epoch": 0.2100098135426889,
+      "grad_norm": 1.0024579763412476,
+      "learning_rate": 3.5655555555555556e-05,
+      "loss": 0.1153,
+      "step": 3210
+    },
+    {
+      "epoch": 0.21066404972194963,
+      "grad_norm": 1.0972620248794556,
+      "learning_rate": 3.576666666666667e-05,
+      "loss": 0.1191,
+      "step": 3220
+    },
+    {
+      "epoch": 0.21131828590121032,
+      "grad_norm": 0.980950653553009,
+      "learning_rate": 3.587777777777778e-05,
+      "loss": 0.1071,
+      "step": 3230
+    },
+    {
+      "epoch": 0.21197252208047104,
+      "grad_norm": 1.0454057455062866,
+      "learning_rate": 3.598888888888889e-05,
+      "loss": 0.1213,
+      "step": 3240
+    },
+    {
+      "epoch": 0.21262675825973176,
+      "grad_norm": 1.0699424743652344,
+      "learning_rate": 3.61e-05,
+      "loss": 0.1195,
+      "step": 3250
+    },
+    {
+      "epoch": 0.21328099443899248,
+      "grad_norm": 0.956729531288147,
+      "learning_rate": 3.621111111111111e-05,
+      "loss": 0.1157,
+      "step": 3260
+    },
+    {
+      "epoch": 0.2139352306182532,
+      "grad_norm": 1.1061006784439087,
+      "learning_rate": 3.632222222222223e-05,
+      "loss": 0.1204,
+      "step": 3270
+    },
+    {
+      "epoch": 0.2145894667975139,
+      "grad_norm": 1.060990571975708,
+      "learning_rate": 3.6433333333333336e-05,
+      "loss": 0.117,
+      "step": 3280
+    },
+    {
+      "epoch": 0.21524370297677461,
+      "grad_norm": 1.000768780708313,
+      "learning_rate": 3.654444444444445e-05,
+      "loss": 0.1206,
+      "step": 3290
+    },
+    {
+      "epoch": 0.21589793915603533,
+      "grad_norm": 0.9216058850288391,
+      "learning_rate": 3.665555555555556e-05,
+      "loss": 0.1127,
+      "step": 3300
+    },
+    {
+      "epoch": 0.21655217533529605,
+      "grad_norm": 0.9647179841995239,
+      "learning_rate": 3.676666666666667e-05,
+      "loss": 0.1204,
+      "step": 3310
+    },
+    {
+      "epoch": 0.21720641151455675,
+      "grad_norm": 1.077235221862793,
+      "learning_rate": 3.687777777777778e-05,
+      "loss": 0.1343,
+      "step": 3320
+    },
+    {
+      "epoch": 0.21786064769381747,
+      "grad_norm": 1.0138746500015259,
+      "learning_rate": 3.698888888888889e-05,
+      "loss": 0.1187,
+      "step": 3330
+    },
+    {
+      "epoch": 0.2185148838730782,
+      "grad_norm": 0.9533920884132385,
+      "learning_rate": 3.71e-05,
+      "loss": 0.1069,
+      "step": 3340
+    },
+    {
+      "epoch": 0.2191691200523389,
+      "grad_norm": 0.8543539643287659,
+      "learning_rate": 3.7211111111111116e-05,
+      "loss": 0.1246,
+      "step": 3350
+    },
+    {
+      "epoch": 0.2198233562315996,
+      "grad_norm": 0.88579922914505,
+      "learning_rate": 3.7322222222222224e-05,
+      "loss": 0.1059,
+      "step": 3360
+    },
+    {
+      "epoch": 0.22047759241086032,
+      "grad_norm": 0.9485725164413452,
+      "learning_rate": 3.743333333333334e-05,
+      "loss": 0.1074,
+      "step": 3370
+    },
+    {
+      "epoch": 0.22113182859012104,
+      "grad_norm": 1.2649180889129639,
+      "learning_rate": 3.754444444444445e-05,
+      "loss": 0.1173,
+      "step": 3380
+    },
+    {
+      "epoch": 0.22178606476938176,
+      "grad_norm": 0.9635019302368164,
+      "learning_rate": 3.765555555555556e-05,
+      "loss": 0.1057,
+      "step": 3390
+    },
+    {
+      "epoch": 0.22244030094864245,
+      "grad_norm": 1.0199334621429443,
+      "learning_rate": 3.7766666666666665e-05,
+      "loss": 0.1113,
+      "step": 3400
+    },
+    {
+      "epoch": 0.22309453712790317,
+      "grad_norm": 1.4301012754440308,
+      "learning_rate": 3.787777777777778e-05,
+      "loss": 0.1258,
+      "step": 3410
+    },
+    {
+      "epoch": 0.2237487733071639,
+      "grad_norm": 1.0970239639282227,
+      "learning_rate": 3.798888888888889e-05,
+      "loss": 0.1142,
+      "step": 3420
+    },
+    {
+      "epoch": 0.2244030094864246,
+      "grad_norm": 1.0526765584945679,
+      "learning_rate": 3.8100000000000005e-05,
+      "loss": 0.1119,
+      "step": 3430
+    },
+    {
+      "epoch": 0.2250572456656853,
+      "grad_norm": 0.9532793164253235,
+      "learning_rate": 3.821111111111111e-05,
+      "loss": 0.1206,
+      "step": 3440
+    },
+    {
+      "epoch": 0.22571148184494602,
+      "grad_norm": 1.0817819833755493,
+      "learning_rate": 3.832222222222223e-05,
+      "loss": 0.1198,
+      "step": 3450
+    },
+    {
+      "epoch": 0.22636571802420674,
+      "grad_norm": 1.0234626531600952,
+      "learning_rate": 3.843333333333334e-05,
+      "loss": 0.117,
+      "step": 3460
+    },
+    {
+      "epoch": 0.22701995420346746,
+      "grad_norm": 0.9513489007949829,
+      "learning_rate": 3.8544444444444445e-05,
+      "loss": 0.1212,
+      "step": 3470
+    },
+    {
+      "epoch": 0.22767419038272815,
+      "grad_norm": 0.869707465171814,
+      "learning_rate": 3.8655555555555554e-05,
+      "loss": 0.1036,
+      "step": 3480
+    },
+    {
+      "epoch": 0.22832842656198887,
+      "grad_norm": 0.8939826488494873,
+      "learning_rate": 3.876666666666667e-05,
+      "loss": 0.1139,
+      "step": 3490
+    },
+    {
+      "epoch": 0.2289826627412496,
+      "grad_norm": 0.8776119351387024,
+      "learning_rate": 3.887777777777778e-05,
+      "loss": 0.1133,
+      "step": 3500
+    },
+    {
+      "epoch": 0.2296368989205103,
+      "grad_norm": 1.042763590812683,
+      "learning_rate": 3.898888888888889e-05,
+      "loss": 0.106,
+      "step": 3510
+    },
+    {
+      "epoch": 0.230291135099771,
+      "grad_norm": 1.07195246219635,
+      "learning_rate": 3.91e-05,
+      "loss": 0.1103,
+      "step": 3520
+    },
+    {
+      "epoch": 0.23094537127903172,
+      "grad_norm": 0.9736886024475098,
+      "learning_rate": 3.921111111111112e-05,
+      "loss": 0.1118,
+      "step": 3530
+    },
+    {
+      "epoch": 0.23159960745829244,
+      "grad_norm": 1.0525206327438354,
+      "learning_rate": 3.932222222222222e-05,
+      "loss": 0.1201,
+      "step": 3540
+    },
+    {
+      "epoch": 0.23225384363755316,
+      "grad_norm": 0.9918597340583801,
+      "learning_rate": 3.9433333333333334e-05,
+      "loss": 0.1199,
+      "step": 3550
+    },
+    {
+      "epoch": 0.23290807981681386,
+      "grad_norm": 1.1127550601959229,
+      "learning_rate": 3.954444444444444e-05,
+      "loss": 0.1211,
+      "step": 3560
+    },
+    {
+      "epoch": 0.23356231599607458,
+      "grad_norm": 1.1743565797805786,
+      "learning_rate": 3.965555555555556e-05,
+      "loss": 0.1142,
+      "step": 3570
+    },
+    {
+      "epoch": 0.2342165521753353,
+      "grad_norm": 1.09585702419281,
+      "learning_rate": 3.9766666666666667e-05,
+      "loss": 0.1281,
+      "step": 3580
+    },
+    {
+      "epoch": 0.23487078835459602,
+      "grad_norm": 0.9824903011322021,
+      "learning_rate": 3.987777777777778e-05,
+      "loss": 0.1139,
+      "step": 3590
+    },
+    {
+      "epoch": 0.23552502453385674,
+      "grad_norm": 1.1848162412643433,
+      "learning_rate": 3.998888888888889e-05,
+      "loss": 0.1074,
+      "step": 3600
+    },
+    {
+      "epoch": 0.23617926071311743,
+      "grad_norm": 1.1790823936462402,
+      "learning_rate": 4.0100000000000006e-05,
+      "loss": 0.1141,
+      "step": 3610
+    },
+    {
+      "epoch": 0.23683349689237815,
+      "grad_norm": 0.9774477481842041,
+      "learning_rate": 4.021111111111111e-05,
+      "loss": 0.1095,
+      "step": 3620
+    },
+    {
+      "epoch": 0.23748773307163887,
+      "grad_norm": 1.0515410900115967,
+      "learning_rate": 4.032222222222222e-05,
+      "loss": 0.1156,
+      "step": 3630
+    },
+    {
+      "epoch": 0.2381419692508996,
+      "grad_norm": 0.9490527510643005,
+      "learning_rate": 4.043333333333333e-05,
+      "loss": 0.1243,
+      "step": 3640
+    },
+    {
+      "epoch": 0.23879620543016028,
+      "grad_norm": 0.9454949498176575,
+      "learning_rate": 4.054444444444445e-05,
+      "loss": 0.1138,
+      "step": 3650
+    },
+    {
+      "epoch": 0.239450441609421,
+      "grad_norm": 0.9614177346229553,
+      "learning_rate": 4.0655555555555555e-05,
+      "loss": 0.1097,
+      "step": 3660
+    },
+    {
+      "epoch": 0.24010467778868172,
+      "grad_norm": 0.9757283329963684,
+      "learning_rate": 4.076666666666667e-05,
+      "loss": 0.1151,
+      "step": 3670
+    },
+    {
+      "epoch": 0.24075891396794244,
+      "grad_norm": 0.9931315779685974,
+      "learning_rate": 4.087777777777778e-05,
+      "loss": 0.1203,
+      "step": 3680
+    },
+    {
+      "epoch": 0.24141315014720313,
+      "grad_norm": 0.9971386194229126,
+      "learning_rate": 4.0988888888888894e-05,
+      "loss": 0.1139,
+      "step": 3690
+    },
+    {
+      "epoch": 0.24206738632646385,
+      "grad_norm": 1.0257045030593872,
+      "learning_rate": 4.11e-05,
+      "loss": 0.1319,
+      "step": 3700
+    },
+    {
+      "epoch": 0.24272162250572457,
+      "grad_norm": 1.130571722984314,
+      "learning_rate": 4.121111111111111e-05,
+      "loss": 0.1159,
+      "step": 3710
+    },
+    {
+      "epoch": 0.2433758586849853,
+      "grad_norm": 0.893582284450531,
+      "learning_rate": 4.132222222222222e-05,
+      "loss": 0.1225,
+      "step": 3720
+    },
+    {
+      "epoch": 0.24403009486424598,
+      "grad_norm": 0.9939897656440735,
+      "learning_rate": 4.1433333333333335e-05,
+      "loss": 0.1016,
+      "step": 3730
+    },
+    {
+      "epoch": 0.2446843310435067,
+      "grad_norm": 0.9984415173530579,
+      "learning_rate": 4.1544444444444444e-05,
+      "loss": 0.1094,
+      "step": 3740
+    },
+    {
+      "epoch": 0.24533856722276742,
+      "grad_norm": 1.0292317867279053,
+      "learning_rate": 4.165555555555556e-05,
+      "loss": 0.1147,
+      "step": 3750
+    },
+    {
+      "epoch": 0.24599280340202814,
+      "grad_norm": 0.9206924438476562,
+      "learning_rate": 4.176666666666667e-05,
+      "loss": 0.1269,
+      "step": 3760
+    },
+    {
+      "epoch": 0.24664703958128883,
+      "grad_norm": 1.1015186309814453,
+      "learning_rate": 4.187777777777778e-05,
+      "loss": 0.1162,
+      "step": 3770
+    },
+    {
+      "epoch": 0.24730127576054955,
+      "grad_norm": 0.950849175453186,
+      "learning_rate": 4.198888888888889e-05,
+      "loss": 0.1153,
+      "step": 3780
+    },
+    {
+      "epoch": 0.24795551193981027,
+      "grad_norm": 1.0211883783340454,
+      "learning_rate": 4.21e-05,
+      "loss": 0.1259,
+      "step": 3790
+    },
+    {
+      "epoch": 0.248609748119071,
+      "grad_norm": 1.113376259803772,
+      "learning_rate": 4.2211111111111115e-05,
+      "loss": 0.1198,
+      "step": 3800
+    },
+    {
+      "epoch": 0.2492639842983317,
+      "grad_norm": 1.0689711570739746,
+      "learning_rate": 4.2322222222222224e-05,
+      "loss": 0.1199,
+      "step": 3810
+    },
+    {
+      "epoch": 0.2499182204775924,
+      "grad_norm": 1.1541417837142944,
+      "learning_rate": 4.243333333333334e-05,
+      "loss": 0.1148,
+      "step": 3820
+    },
+    {
+      "epoch": 0.2505724566568531,
+      "grad_norm": 0.9707340598106384,
+      "learning_rate": 4.254444444444445e-05,
+      "loss": 0.1298,
+      "step": 3830
+    },
+    {
+      "epoch": 0.2512266928361138,
+      "grad_norm": 1.2857780456542969,
+      "learning_rate": 4.2655555555555556e-05,
+      "loss": 0.1298,
+      "step": 3840
+    },
+    {
+      "epoch": 0.25188092901537457,
+      "grad_norm": 0.8688610792160034,
+      "learning_rate": 4.2766666666666665e-05,
+      "loss": 0.1169,
+      "step": 3850
+    },
+    {
+      "epoch": 0.25253516519463526,
+      "grad_norm": 1.0786139965057373,
+      "learning_rate": 4.287777777777778e-05,
+      "loss": 0.1279,
+      "step": 3860
+    },
+    {
+      "epoch": 0.25318940137389595,
+      "grad_norm": 1.1422791481018066,
+      "learning_rate": 4.298888888888889e-05,
+      "loss": 0.1114,
+      "step": 3870
+    },
+    {
+      "epoch": 0.2538436375531567,
+      "grad_norm": 0.9638531804084778,
+      "learning_rate": 4.3100000000000004e-05,
+      "loss": 0.1144,
+      "step": 3880
+    },
+    {
+      "epoch": 0.2544978737324174,
+      "grad_norm": 1.1392393112182617,
+      "learning_rate": 4.321111111111111e-05,
+      "loss": 0.1087,
+      "step": 3890
+    },
+    {
+      "epoch": 0.25515210991167814,
+      "grad_norm": 0.9770271182060242,
+      "learning_rate": 4.332222222222223e-05,
+      "loss": 0.1305,
+      "step": 3900
+    },
+    {
+      "epoch": 0.25580634609093883,
+      "grad_norm": 0.9613223075866699,
+      "learning_rate": 4.3433333333333336e-05,
+      "loss": 0.1093,
+      "step": 3910
+    },
+    {
+      "epoch": 0.2564605822701995,
+      "grad_norm": 0.9954821467399597,
+      "learning_rate": 4.354444444444445e-05,
+      "loss": 0.1201,
+      "step": 3920
+    },
+    {
+      "epoch": 0.25711481844946027,
+      "grad_norm": 1.0677788257598877,
+      "learning_rate": 4.3655555555555554e-05,
+      "loss": 0.1102,
+      "step": 3930
+    },
+    {
+      "epoch": 0.25776905462872096,
+      "grad_norm": 1.1282657384872437,
+      "learning_rate": 4.376666666666667e-05,
+      "loss": 0.118,
+      "step": 3940
+    },
+    {
+      "epoch": 0.25842329080798165,
+      "grad_norm": 1.0023534297943115,
+      "learning_rate": 4.387777777777778e-05,
+      "loss": 0.1144,
+      "step": 3950
+    },
+    {
+      "epoch": 0.2590775269872424,
+      "grad_norm": 0.9646608233451843,
+      "learning_rate": 4.398888888888889e-05,
+      "loss": 0.1106,
+      "step": 3960
+    },
+    {
+      "epoch": 0.2597317631665031,
+      "grad_norm": 1.139906406402588,
+      "learning_rate": 4.41e-05,
+      "loss": 0.1134,
+      "step": 3970
+    },
+    {
+      "epoch": 0.26038599934576384,
+      "grad_norm": 0.960753858089447,
+      "learning_rate": 4.4211111111111117e-05,
+      "loss": 0.1142,
+      "step": 3980
+    },
+    {
+      "epoch": 0.26104023552502453,
+      "grad_norm": 1.0693498849868774,
+      "learning_rate": 4.4322222222222225e-05,
+      "loss": 0.1091,
+      "step": 3990
+    },
+    {
+      "epoch": 0.2616944717042852,
+      "grad_norm": 1.1489149332046509,
+      "learning_rate": 4.443333333333334e-05,
+      "loss": 0.107,
+      "step": 4000
+    },
+    {
+      "epoch": 0.262348707883546,
+      "grad_norm": 0.9810730218887329,
+      "learning_rate": 4.454444444444444e-05,
+      "loss": 0.1197,
+      "step": 4010
+    },
+    {
+      "epoch": 0.26300294406280667,
+      "grad_norm": 0.9745346307754517,
+      "learning_rate": 4.465555555555556e-05,
+      "loss": 0.1158,
+      "step": 4020
+    },
+    {
+      "epoch": 0.2636571802420674,
+      "grad_norm": 1.056600570678711,
+      "learning_rate": 4.4766666666666666e-05,
+      "loss": 0.1102,
+      "step": 4030
+    },
+    {
+      "epoch": 0.2643114164213281,
+      "grad_norm": 1.090558409690857,
+      "learning_rate": 4.487777777777778e-05,
+      "loss": 0.1151,
+      "step": 4040
+    },
+    {
+      "epoch": 0.2649656526005888,
+      "grad_norm": 0.9222038984298706,
+      "learning_rate": 4.498888888888889e-05,
+      "loss": 0.1103,
+      "step": 4050
+    },
+    {
+      "epoch": 0.26561988877984954,
+      "grad_norm": 1.0304709672927856,
+      "learning_rate": 4.5100000000000005e-05,
+      "loss": 0.1169,
+      "step": 4060
+    },
+    {
+      "epoch": 0.26627412495911024,
+      "grad_norm": 1.0132149457931519,
+      "learning_rate": 4.5211111111111114e-05,
+      "loss": 0.1074,
+      "step": 4070
+    },
+    {
+      "epoch": 0.26692836113837093,
+      "grad_norm": 1.0685970783233643,
+      "learning_rate": 4.532222222222223e-05,
+      "loss": 0.1087,
+      "step": 4080
+    },
+    {
+      "epoch": 0.2675825973176317,
+      "grad_norm": 1.0187596082687378,
+      "learning_rate": 4.543333333333333e-05,
+      "loss": 0.1225,
+      "step": 4090
+    },
+    {
+      "epoch": 0.26823683349689237,
+      "grad_norm": 1.0164580345153809,
+      "learning_rate": 4.5544444444444446e-05,
+      "loss": 0.119,
+      "step": 4100
+    },
+    {
+      "epoch": 0.2688910696761531,
+      "grad_norm": 0.9964419007301331,
+      "learning_rate": 4.5655555555555555e-05,
+      "loss": 0.1077,
+      "step": 4110
+    },
+    {
+      "epoch": 0.2695453058554138,
+      "grad_norm": 1.1001616716384888,
+      "learning_rate": 4.576666666666667e-05,
+      "loss": 0.1203,
+      "step": 4120
+    },
+    {
+      "epoch": 0.2701995420346745,
+      "grad_norm": 0.984094500541687,
+      "learning_rate": 4.587777777777778e-05,
+      "loss": 0.1219,
+      "step": 4130
+    },
+    {
+      "epoch": 0.27085377821393525,
+      "grad_norm": 0.8646033406257629,
+      "learning_rate": 4.5988888888888894e-05,
+      "loss": 0.1144,
+      "step": 4140
+    },
+    {
+      "epoch": 0.27150801439319594,
+      "grad_norm": 0.8888686299324036,
+      "learning_rate": 4.61e-05,
+      "loss": 0.1216,
+      "step": 4150
+    },
+    {
+      "epoch": 0.27216225057245663,
+      "grad_norm": 0.9775686264038086,
+      "learning_rate": 4.621111111111111e-05,
+      "loss": 0.1119,
+      "step": 4160
+    },
+    {
+      "epoch": 0.2728164867517174,
+      "grad_norm": 1.1479732990264893,
+      "learning_rate": 4.632222222222222e-05,
+      "loss": 0.1159,
+      "step": 4170
+    },
+    {
+      "epoch": 0.27347072293097807,
+      "grad_norm": 0.9950860738754272,
+      "learning_rate": 4.6433333333333335e-05,
+      "loss": 0.1151,
+      "step": 4180
+    },
+    {
+      "epoch": 0.2741249591102388,
+      "grad_norm": 1.044150948524475,
+      "learning_rate": 4.6544444444444443e-05,
+      "loss": 0.1029,
+      "step": 4190
+    },
+    {
+      "epoch": 0.2747791952894995,
+      "grad_norm": 0.9072064757347107,
+      "learning_rate": 4.665555555555556e-05,
+      "loss": 0.1217,
+      "step": 4200
+    },
+    {
+      "epoch": 0.2754334314687602,
+      "grad_norm": 1.0930536985397339,
+      "learning_rate": 4.676666666666667e-05,
+      "loss": 0.1161,
+      "step": 4210
+    },
+    {
+      "epoch": 0.27608766764802095,
+      "grad_norm": 1.104744791984558,
+      "learning_rate": 4.687777777777778e-05,
+      "loss": 0.1127,
+      "step": 4220
+    },
+    {
+      "epoch": 0.27674190382728164,
+      "grad_norm": 0.8702093958854675,
+      "learning_rate": 4.698888888888889e-05,
+      "loss": 0.1117,
+      "step": 4230
+    },
+    {
+      "epoch": 0.27739614000654234,
+      "grad_norm": 1.0562191009521484,
+      "learning_rate": 4.71e-05,
+      "loss": 0.1172,
+      "step": 4240
+    },
+    {
+      "epoch": 0.2780503761858031,
+      "grad_norm": 1.0193010568618774,
+      "learning_rate": 4.721111111111111e-05,
+      "loss": 0.1117,
+      "step": 4250
+    },
+    {
+      "epoch": 0.2787046123650638,
+      "grad_norm": 1.2876737117767334,
+      "learning_rate": 4.7322222222222224e-05,
+      "loss": 0.1174,
+      "step": 4260
+    },
+    {
+      "epoch": 0.2793588485443245,
+      "grad_norm": 1.0730646848678589,
+      "learning_rate": 4.743333333333333e-05,
+      "loss": 0.1136,
+      "step": 4270
+    },
+    {
+      "epoch": 0.2800130847235852,
+      "grad_norm": 1.0272183418273926,
+      "learning_rate": 4.754444444444445e-05,
+      "loss": 0.1042,
+      "step": 4280
+    },
+    {
+      "epoch": 0.2806673209028459,
+      "grad_norm": 0.9593018293380737,
+      "learning_rate": 4.7655555555555556e-05,
+      "loss": 0.1068,
+      "step": 4290
+    },
+    {
+      "epoch": 0.28132155708210665,
+      "grad_norm": 1.1273571252822876,
+      "learning_rate": 4.776666666666667e-05,
+      "loss": 0.1183,
+      "step": 4300
+    },
+    {
+      "epoch": 0.28197579326136735,
+      "grad_norm": 0.9116014838218689,
+      "learning_rate": 4.787777777777778e-05,
+      "loss": 0.1114,
+      "step": 4310
+    },
+    {
+      "epoch": 0.2826300294406281,
+      "grad_norm": 0.9448710680007935,
+      "learning_rate": 4.798888888888889e-05,
+      "loss": 0.1155,
+      "step": 4320
+    },
+    {
+      "epoch": 0.2832842656198888,
+      "grad_norm": 0.9872847199440002,
+      "learning_rate": 4.8100000000000004e-05,
+      "loss": 0.1072,
+      "step": 4330
+    },
+    {
+      "epoch": 0.2839385017991495,
+      "grad_norm": 0.9878488779067993,
+      "learning_rate": 4.821111111111111e-05,
+      "loss": 0.1187,
+      "step": 4340
+    },
+    {
+      "epoch": 0.2845927379784102,
+      "grad_norm": 1.1943411827087402,
+      "learning_rate": 4.832222222222223e-05,
+      "loss": 0.1299,
+      "step": 4350
+    },
+    {
+      "epoch": 0.2852469741576709,
+      "grad_norm": 1.0091590881347656,
+      "learning_rate": 4.8433333333333336e-05,
+      "loss": 0.1205,
+      "step": 4360
+    },
+    {
+      "epoch": 0.2859012103369316,
+      "grad_norm": 1.0240885019302368,
+      "learning_rate": 4.8544444444444445e-05,
+      "loss": 0.11,
+      "step": 4370
+    },
+    {
+      "epoch": 0.28655544651619236,
+      "grad_norm": 0.8469128012657166,
+      "learning_rate": 4.865555555555556e-05,
+      "loss": 0.1223,
+      "step": 4380
+    },
+    {
+      "epoch": 0.28720968269545305,
+      "grad_norm": 0.932327389717102,
+      "learning_rate": 4.876666666666667e-05,
+      "loss": 0.1061,
+      "step": 4390
+    },
+    {
+      "epoch": 0.2878639188747138,
+      "grad_norm": 0.9431477785110474,
+      "learning_rate": 4.887777777777778e-05,
+      "loss": 0.1064,
+      "step": 4400
+    },
+    {
+      "epoch": 0.2885181550539745,
+      "grad_norm": 0.9958844780921936,
+      "learning_rate": 4.898888888888889e-05,
+      "loss": 0.112,
+      "step": 4410
+    },
+    {
+      "epoch": 0.2891723912332352,
+      "grad_norm": 1.146600365638733,
+      "learning_rate": 4.91e-05,
+      "loss": 0.1189,
+      "step": 4420
+    },
+    {
+      "epoch": 0.28982662741249593,
+      "grad_norm": 1.0450646877288818,
+      "learning_rate": 4.9211111111111116e-05,
+      "loss": 0.1268,
+      "step": 4430
+    },
+    {
+      "epoch": 0.2904808635917566,
+      "grad_norm": 1.2781046628952026,
+      "learning_rate": 4.9322222222222225e-05,
+      "loss": 0.123,
+      "step": 4440
+    },
+    {
+      "epoch": 0.2911350997710173,
+      "grad_norm": 0.9840906858444214,
+      "learning_rate": 4.943333333333334e-05,
+      "loss": 0.1139,
+      "step": 4450
+    },
+    {
+      "epoch": 0.29178933595027806,
+      "grad_norm": 0.9794589281082153,
+      "learning_rate": 4.954444444444445e-05,
+      "loss": 0.1176,
+      "step": 4460
+    },
+    {
+      "epoch": 0.29244357212953875,
+      "grad_norm": 1.2850165367126465,
+      "learning_rate": 4.965555555555556e-05,
+      "loss": 0.1245,
+      "step": 4470
+    },
+    {
+      "epoch": 0.2930978083087995,
+      "grad_norm": 0.9761756062507629,
+      "learning_rate": 4.9766666666666666e-05,
+      "loss": 0.1052,
+      "step": 4480
+    },
+    {
+      "epoch": 0.2937520444880602,
+      "grad_norm": 0.9061872959136963,
+      "learning_rate": 4.987777777777778e-05,
+      "loss": 0.1102,
+      "step": 4490
+    },
+    {
+      "epoch": 0.2944062806673209,
+      "grad_norm": 1.1952383518218994,
+      "learning_rate": 4.998888888888889e-05,
+      "loss": 0.1046,
+      "step": 4500
+    },
+    {
+      "epoch": 0.29506051684658163,
+      "grad_norm": 1.018266201019287,
+      "learning_rate": 5.0100000000000005e-05,
+      "loss": 0.1126,
+      "step": 4510
+    },
+    {
+      "epoch": 0.2957147530258423,
+      "grad_norm": 1.0872063636779785,
+      "learning_rate": 5.0211111111111107e-05,
+      "loss": 0.1123,
+      "step": 4520
+    },
+    {
+      "epoch": 0.296368989205103,
+      "grad_norm": 1.1259245872497559,
+      "learning_rate": 5.032222222222223e-05,
+      "loss": 0.1138,
+      "step": 4530
+    },
+    {
+      "epoch": 0.29702322538436376,
+      "grad_norm": 0.939315676689148,
+      "learning_rate": 5.043333333333333e-05,
+      "loss": 0.112,
+      "step": 4540
+    },
+    {
+      "epoch": 0.29767746156362446,
+      "grad_norm": 0.9471316933631897,
+      "learning_rate": 5.054444444444445e-05,
+      "loss": 0.116,
+      "step": 4550
+    },
+    {
+      "epoch": 0.2983316977428852,
+      "grad_norm": 1.0100241899490356,
+      "learning_rate": 5.0655555555555554e-05,
+      "loss": 0.105,
+      "step": 4560
+    },
+    {
+      "epoch": 0.2989859339221459,
+      "grad_norm": 1.010973572731018,
+      "learning_rate": 5.0766666666666676e-05,
+      "loss": 0.0968,
+      "step": 4570
+    },
+    {
+      "epoch": 0.2996401701014066,
+      "grad_norm": 1.0246537923812866,
+      "learning_rate": 5.087777777777778e-05,
+      "loss": 0.1301,
+      "step": 4580
+    },
+    {
+      "epoch": 0.30029440628066734,
+      "grad_norm": 1.0651960372924805,
+      "learning_rate": 5.098888888888889e-05,
+      "loss": 0.1184,
+      "step": 4590
+    },
+    {
+      "epoch": 0.30094864245992803,
+      "grad_norm": 1.0061384439468384,
+      "learning_rate": 5.11e-05,
+      "loss": 0.1238,
+      "step": 4600
+    },
+    {
+      "epoch": 0.3016028786391887,
+      "grad_norm": 1.0411393642425537,
+      "learning_rate": 5.121111111111111e-05,
+      "loss": 0.1066,
+      "step": 4610
+    },
+    {
+      "epoch": 0.30225711481844947,
+      "grad_norm": 1.1767691373825073,
+      "learning_rate": 5.1322222222222226e-05,
+      "loss": 0.1176,
+      "step": 4620
+    },
+    {
+      "epoch": 0.30291135099771016,
+      "grad_norm": 1.1108002662658691,
+      "learning_rate": 5.1433333333333334e-05,
+      "loss": 0.1231,
+      "step": 4630
+    },
+    {
+      "epoch": 0.3035655871769709,
+      "grad_norm": 0.9178494811058044,
+      "learning_rate": 5.154444444444445e-05,
+      "loss": 0.1219,
+      "step": 4640
+    },
+    {
+      "epoch": 0.3042198233562316,
+      "grad_norm": 1.046704649925232,
+      "learning_rate": 5.165555555555556e-05,
+      "loss": 0.108,
+      "step": 4650
+    },
+    {
+      "epoch": 0.3048740595354923,
+      "grad_norm": 1.1278116703033447,
+      "learning_rate": 5.1766666666666674e-05,
+      "loss": 0.1205,
+      "step": 4660
+    },
+    {
+      "epoch": 0.30552829571475304,
+      "grad_norm": 1.049989104270935,
+      "learning_rate": 5.187777777777778e-05,
+      "loss": 0.1195,
+      "step": 4670
+    },
+    {
+      "epoch": 0.30618253189401373,
+      "grad_norm": 1.1788783073425293,
+      "learning_rate": 5.1988888888888884e-05,
+      "loss": 0.112,
+      "step": 4680
+    },
+    {
+      "epoch": 0.3068367680732745,
+      "grad_norm": 1.094664454460144,
+      "learning_rate": 5.2100000000000006e-05,
+      "loss": 0.1104,
+      "step": 4690
+    },
+    {
+      "epoch": 0.30749100425253517,
+      "grad_norm": 1.0641672611236572,
+      "learning_rate": 5.221111111111111e-05,
+      "loss": 0.1209,
+      "step": 4700
+    },
+    {
+      "epoch": 0.30814524043179586,
+      "grad_norm": 0.9625852108001709,
+      "learning_rate": 5.232222222222223e-05,
+      "loss": 0.1171,
+      "step": 4710
+    },
+    {
+      "epoch": 0.3087994766110566,
+      "grad_norm": 0.9965457916259766,
+      "learning_rate": 5.243333333333333e-05,
+      "loss": 0.1152,
+      "step": 4720
+    },
+    {
+      "epoch": 0.3094537127903173,
+      "grad_norm": 1.1874282360076904,
+      "learning_rate": 5.2544444444444454e-05,
+      "loss": 0.1069,
+      "step": 4730
+    },
+    {
+      "epoch": 0.310107948969578,
+      "grad_norm": 1.071972370147705,
+      "learning_rate": 5.2655555555555555e-05,
+      "loss": 0.1207,
+      "step": 4740
+    },
+    {
+      "epoch": 0.31076218514883874,
+      "grad_norm": 1.1899669170379639,
+      "learning_rate": 5.2766666666666664e-05,
+      "loss": 0.1003,
+      "step": 4750
+    },
+    {
+      "epoch": 0.31141642132809944,
+      "grad_norm": 1.1113231182098389,
+      "learning_rate": 5.287777777777778e-05,
+      "loss": 0.1101,
+      "step": 4760
+    },
+    {
+      "epoch": 0.3120706575073602,
+      "grad_norm": 0.9851582646369934,
+      "learning_rate": 5.298888888888889e-05,
+      "loss": 0.1246,
+      "step": 4770
+    },
+    {
+      "epoch": 0.3127248936866209,
+      "grad_norm": 1.1461477279663086,
+      "learning_rate": 5.31e-05,
+      "loss": 0.1136,
+      "step": 4780
+    },
+    {
+      "epoch": 0.31337912986588157,
+      "grad_norm": 1.0127012729644775,
+      "learning_rate": 5.321111111111111e-05,
+      "loss": 0.1082,
+      "step": 4790
+    },
+    {
+      "epoch": 0.3140333660451423,
+      "grad_norm": 1.0114226341247559,
+      "learning_rate": 5.332222222222223e-05,
+      "loss": 0.123,
+      "step": 4800
+    },
+    {
+      "epoch": 0.314687602224403,
+      "grad_norm": 1.0806419849395752,
+      "learning_rate": 5.3433333333333336e-05,
+      "loss": 0.1239,
+      "step": 4810
+    },
+    {
+      "epoch": 0.3153418384036637,
+      "grad_norm": 0.9436666369438171,
+      "learning_rate": 5.3544444444444444e-05,
+      "loss": 0.1127,
+      "step": 4820
+    },
+    {
+      "epoch": 0.31599607458292445,
+      "grad_norm": 0.9551024436950684,
+      "learning_rate": 5.365555555555556e-05,
+      "loss": 0.1146,
+      "step": 4830
+    },
+    {
+      "epoch": 0.31665031076218514,
+      "grad_norm": 0.9468205571174622,
+      "learning_rate": 5.376666666666667e-05,
+      "loss": 0.1099,
+      "step": 4840
+    },
+    {
+      "epoch": 0.3173045469414459,
+      "grad_norm": 1.0313448905944824,
+      "learning_rate": 5.387777777777778e-05,
+      "loss": 0.1108,
+      "step": 4850
+    },
+    {
+      "epoch": 0.3179587831207066,
+      "grad_norm": 1.025217890739441,
+      "learning_rate": 5.398888888888889e-05,
+      "loss": 0.1245,
+      "step": 4860
+    },
+    {
+      "epoch": 0.31861301929996727,
+      "grad_norm": 0.9557814002037048,
+      "learning_rate": 5.410000000000001e-05,
+      "loss": 0.1099,
+      "step": 4870
+    },
+    {
+      "epoch": 0.319267255479228,
+      "grad_norm": 1.0116934776306152,
+      "learning_rate": 5.4211111111111116e-05,
+      "loss": 0.1068,
+      "step": 4880
+    },
+    {
+      "epoch": 0.3199214916584887,
+      "grad_norm": 1.0298535823822021,
+      "learning_rate": 5.432222222222223e-05,
+      "loss": 0.1261,
+      "step": 4890
+    },
+    {
+      "epoch": 0.3205757278377494,
+      "grad_norm": 1.6107592582702637,
+      "learning_rate": 5.443333333333333e-05,
+      "loss": 0.1166,
+      "step": 4900
+    },
+    {
+      "epoch": 0.32122996401701015,
+      "grad_norm": 0.9493816494941711,
+      "learning_rate": 5.454444444444444e-05,
+      "loss": 0.1025,
+      "step": 4910
+    },
+    {
+      "epoch": 0.32188420019627084,
+      "grad_norm": 1.0891478061676025,
+      "learning_rate": 5.465555555555556e-05,
+      "loss": 0.1259,
+      "step": 4920
+    },
+    {
+      "epoch": 0.3225384363755316,
+      "grad_norm": 1.0789752006530762,
+      "learning_rate": 5.4766666666666665e-05,
+      "loss": 0.1092,
+      "step": 4930
+    },
+    {
+      "epoch": 0.3231926725547923,
+      "grad_norm": 0.8456171154975891,
+      "learning_rate": 5.487777777777778e-05,
+      "loss": 0.1094,
+      "step": 4940
+    },
+    {
+      "epoch": 0.323846908734053,
+      "grad_norm": 1.1265510320663452,
+      "learning_rate": 5.498888888888889e-05,
+      "loss": 0.1105,
+      "step": 4950
+    },
+    {
+      "epoch": 0.3245011449133137,
+      "grad_norm": 1.0033330917358398,
+      "learning_rate": 5.5100000000000004e-05,
+      "loss": 0.1138,
+      "step": 4960
+    },
+    {
+      "epoch": 0.3251553810925744,
+      "grad_norm": 0.9532373547554016,
+      "learning_rate": 5.521111111111111e-05,
+      "loss": 0.1134,
+      "step": 4970
+    },
+    {
+      "epoch": 0.3258096172718351,
+      "grad_norm": 1.070513367652893,
+      "learning_rate": 5.532222222222222e-05,
+      "loss": 0.1054,
+      "step": 4980
+    },
+    {
+      "epoch": 0.32646385345109585,
+      "grad_norm": 1.1604113578796387,
+      "learning_rate": 5.543333333333334e-05,
+      "loss": 0.1089,
+      "step": 4990
+    },
+    {
+      "epoch": 0.32711808963035655,
+      "grad_norm": 0.9671838283538818,
+      "learning_rate": 5.5544444444444445e-05,
+      "loss": 0.1114,
+      "step": 5000
+    },
+    {
+      "epoch": 0.3277723258096173,
+      "grad_norm": 1.0246025323867798,
+      "learning_rate": 5.565555555555556e-05,
+      "loss": 0.1155,
+      "step": 5010
+    },
+    {
+      "epoch": 0.328426561988878,
+      "grad_norm": 1.1723190546035767,
+      "learning_rate": 5.576666666666667e-05,
+      "loss": 0.1086,
+      "step": 5020
+    },
+    {
+      "epoch": 0.3290807981681387,
+      "grad_norm": 1.158227801322937,
+      "learning_rate": 5.5877777777777785e-05,
+      "loss": 0.1142,
+      "step": 5030
+    },
+    {
+      "epoch": 0.3297350343473994,
+      "grad_norm": 1.070424199104309,
+      "learning_rate": 5.598888888888889e-05,
+      "loss": 0.1075,
+      "step": 5040
+    },
+    {
+      "epoch": 0.3303892705266601,
+      "grad_norm": 1.1170357465744019,
+      "learning_rate": 5.610000000000001e-05,
+      "loss": 0.1242,
+      "step": 5050
+    },
+    {
+      "epoch": 0.33104350670592086,
+      "grad_norm": 1.009589672088623,
+      "learning_rate": 5.621111111111112e-05,
+      "loss": 0.1076,
+      "step": 5060
+    },
+    {
+      "epoch": 0.33169774288518156,
+      "grad_norm": 1.002823829650879,
+      "learning_rate": 5.632222222222222e-05,
+      "loss": 0.1146,
+      "step": 5070
+    },
+    {
+      "epoch": 0.33235197906444225,
+      "grad_norm": 0.9945969581604004,
+      "learning_rate": 5.643333333333334e-05,
+      "loss": 0.1155,
+      "step": 5080
+    },
+    {
+      "epoch": 0.333006215243703,
+      "grad_norm": 0.9207086563110352,
+      "learning_rate": 5.654444444444444e-05,
+      "loss": 0.116,
+      "step": 5090
+    },
+    {
+      "epoch": 0.3336604514229637,
+      "grad_norm": 0.8761827945709229,
+      "learning_rate": 5.6655555555555565e-05,
+      "loss": 0.1119,
+      "step": 5100
+    },
+    {
+      "epoch": 0.3343146876022244,
+      "grad_norm": 1.128303050994873,
+      "learning_rate": 5.6766666666666666e-05,
+      "loss": 0.1134,
+      "step": 5110
+    },
+    {
+      "epoch": 0.33496892378148513,
+      "grad_norm": 0.9229829907417297,
+      "learning_rate": 5.687777777777779e-05,
+      "loss": 0.1085,
+      "step": 5120
+    },
+    {
+      "epoch": 0.3356231599607458,
+      "grad_norm": 1.0717655420303345,
+      "learning_rate": 5.698888888888889e-05,
+      "loss": 0.1154,
+      "step": 5130
+    },
+    {
+      "epoch": 0.33627739614000657,
+      "grad_norm": 1.0167633295059204,
+      "learning_rate": 5.71e-05,
+      "loss": 0.1185,
+      "step": 5140
+    },
+    {
+      "epoch": 0.33693163231926726,
+      "grad_norm": 1.1209583282470703,
+      "learning_rate": 5.7211111111111114e-05,
+      "loss": 0.1182,
+      "step": 5150
+    },
+    {
+      "epoch": 0.33758586849852795,
+      "grad_norm": 0.9364834427833557,
+      "learning_rate": 5.732222222222222e-05,
+      "loss": 0.1024,
+      "step": 5160
+    },
+    {
+      "epoch": 0.3382401046777887,
+      "grad_norm": 1.077996850013733,
+      "learning_rate": 5.743333333333334e-05,
+      "loss": 0.1076,
+      "step": 5170
+    },
+    {
+      "epoch": 0.3388943408570494,
+      "grad_norm": 0.946433424949646,
+      "learning_rate": 5.7544444444444447e-05,
+      "loss": 0.1176,
+      "step": 5180
+    },
+    {
+      "epoch": 0.3395485770363101,
+      "grad_norm": 0.9579144716262817,
+      "learning_rate": 5.765555555555556e-05,
+      "loss": 0.1077,
+      "step": 5190
+    },
+    {
+      "epoch": 0.34020281321557083,
+      "grad_norm": 0.9725630283355713,
+      "learning_rate": 5.776666666666667e-05,
+      "loss": 0.1123,
+      "step": 5200
+    },
+    {
+      "epoch": 0.3408570493948315,
+      "grad_norm": 1.1612457036972046,
+      "learning_rate": 5.787777777777777e-05,
+      "loss": 0.1079,
+      "step": 5210
+    },
+    {
+      "epoch": 0.34151128557409227,
+      "grad_norm": 1.1368802785873413,
+      "learning_rate": 5.7988888888888894e-05,
+      "loss": 0.1145,
+      "step": 5220
+    },
+    {
+      "epoch": 0.34216552175335296,
+      "grad_norm": 1.130491852760315,
+      "learning_rate": 5.8099999999999996e-05,
+      "loss": 0.1127,
+      "step": 5230
+    },
+    {
+      "epoch": 0.34281975793261366,
+      "grad_norm": 1.1472586393356323,
+      "learning_rate": 5.821111111111112e-05,
+      "loss": 0.1067,
+      "step": 5240
+    },
+    {
+      "epoch": 0.3434739941118744,
+      "grad_norm": 1.2415672540664673,
+      "learning_rate": 5.832222222222222e-05,
+      "loss": 0.1035,
+      "step": 5250
+    },
+    {
+      "epoch": 0.3441282302911351,
+      "grad_norm": 0.952216625213623,
+      "learning_rate": 5.843333333333334e-05,
+      "loss": 0.1211,
+      "step": 5260
+    },
+    {
+      "epoch": 0.3447824664703958,
+      "grad_norm": 0.9842739701271057,
+      "learning_rate": 5.8544444444444444e-05,
+      "loss": 0.1081,
+      "step": 5270
+    },
+    {
+      "epoch": 0.34543670264965654,
+      "grad_norm": 0.921035885810852,
+      "learning_rate": 5.8655555555555566e-05,
+      "loss": 0.1063,
+      "step": 5280
+    },
+    {
+      "epoch": 0.3460909388289172,
+      "grad_norm": 1.028622031211853,
+      "learning_rate": 5.876666666666667e-05,
+      "loss": 0.1114,
+      "step": 5290
+    },
+    {
+      "epoch": 0.346745175008178,
+      "grad_norm": 1.103540062904358,
+      "learning_rate": 5.8877777777777776e-05,
+      "loss": 0.1144,
+      "step": 5300
+    },
+    {
+      "epoch": 0.34739941118743867,
+      "grad_norm": 0.9521011710166931,
+      "learning_rate": 5.898888888888889e-05,
+      "loss": 0.1104,
+      "step": 5310
+    },
+    {
+      "epoch": 0.34805364736669936,
+      "grad_norm": 1.1249808073043823,
+      "learning_rate": 5.91e-05,
+      "loss": 0.1235,
+      "step": 5320
+    },
+    {
+      "epoch": 0.3487078835459601,
+      "grad_norm": 1.0736548900604248,
+      "learning_rate": 5.9211111111111115e-05,
+      "loss": 0.1114,
+      "step": 5330
+    },
+    {
+      "epoch": 0.3493621197252208,
+      "grad_norm": 0.9119420647621155,
+      "learning_rate": 5.9322222222222224e-05,
+      "loss": 0.1234,
+      "step": 5340
+    },
+    {
+      "epoch": 0.3500163559044815,
+      "grad_norm": 0.9707286953926086,
+      "learning_rate": 5.943333333333334e-05,
+      "loss": 0.1139,
+      "step": 5350
+    },
+    {
+      "epoch": 0.35067059208374224,
+      "grad_norm": 0.8778756260871887,
+      "learning_rate": 5.954444444444445e-05,
+      "loss": 0.1119,
+      "step": 5360
+    },
+    {
+      "epoch": 0.35132482826300293,
+      "grad_norm": 0.9317549467086792,
+      "learning_rate": 5.9655555555555556e-05,
+      "loss": 0.1162,
+      "step": 5370
+    },
+    {
+      "epoch": 0.3519790644422637,
+      "grad_norm": 0.9935291409492493,
+      "learning_rate": 5.976666666666667e-05,
+      "loss": 0.115,
+      "step": 5380
+    },
+    {
+      "epoch": 0.35263330062152437,
+      "grad_norm": 0.8970289826393127,
+      "learning_rate": 5.987777777777778e-05,
+      "loss": 0.1094,
+      "step": 5390
+    },
+    {
+      "epoch": 0.35328753680078506,
+      "grad_norm": 1.034538745880127,
+      "learning_rate": 5.9988888888888895e-05,
+      "loss": 0.1097,
+      "step": 5400
+    },
+    {
+      "epoch": 0.3539417729800458,
+      "grad_norm": 0.9625052809715271,
+      "learning_rate": 6.0100000000000004e-05,
+      "loss": 0.1254,
+      "step": 5410
+    },
+    {
+      "epoch": 0.3545960091593065,
+      "grad_norm": 1.0226140022277832,
+      "learning_rate": 6.021111111111112e-05,
+      "loss": 0.1089,
+      "step": 5420
+    },
+    {
+      "epoch": 0.35525024533856725,
+      "grad_norm": 0.9254044890403748,
+      "learning_rate": 6.032222222222222e-05,
+      "loss": 0.1233,
+      "step": 5430
+    },
+    {
+      "epoch": 0.35590448151782794,
+      "grad_norm": 0.9888849854469299,
+      "learning_rate": 6.043333333333333e-05,
+      "loss": 0.1031,
+      "step": 5440
+    },
+    {
+      "epoch": 0.35655871769708863,
+      "grad_norm": 0.9793877601623535,
+      "learning_rate": 6.0544444444444445e-05,
+      "loss": 0.1127,
+      "step": 5450
+    },
+    {
+      "epoch": 0.3572129538763494,
+      "grad_norm": 0.9079664945602417,
+      "learning_rate": 6.0655555555555553e-05,
+      "loss": 0.1088,
+      "step": 5460
+    },
+    {
+      "epoch": 0.3578671900556101,
+      "grad_norm": 1.0613657236099243,
+      "learning_rate": 6.076666666666667e-05,
+      "loss": 0.1047,
+      "step": 5470
+    },
+    {
+      "epoch": 0.35852142623487077,
+      "grad_norm": 1.1257914304733276,
+      "learning_rate": 6.087777777777778e-05,
+      "loss": 0.1197,
+      "step": 5480
+    },
+    {
+      "epoch": 0.3591756624141315,
+      "grad_norm": 0.9039554595947266,
+      "learning_rate": 6.098888888888889e-05,
+      "loss": 0.1103,
+      "step": 5490
+    },
+    {
+      "epoch": 0.3598298985933922,
+      "grad_norm": 1.026854395866394,
+      "learning_rate": 6.110000000000001e-05,
+      "loss": 0.1116,
+      "step": 5500
+    },
+    {
+      "epoch": 0.36048413477265295,
+      "grad_norm": 1.1896244287490845,
+      "learning_rate": 6.121111111111112e-05,
+      "loss": 0.1136,
+      "step": 5510
+    },
+    {
+      "epoch": 0.36113837095191365,
+      "grad_norm": 0.9056310057640076,
+      "learning_rate": 6.132222222222223e-05,
+      "loss": 0.1134,
+      "step": 5520
+    },
+    {
+      "epoch": 0.36179260713117434,
+      "grad_norm": 1.078660249710083,
+      "learning_rate": 6.143333333333333e-05,
+      "loss": 0.1063,
+      "step": 5530
+    },
+    {
+      "epoch": 0.3624468433104351,
+      "grad_norm": 1.1922471523284912,
+      "learning_rate": 6.154444444444446e-05,
+      "loss": 0.1116,
+      "step": 5540
+    },
+    {
+      "epoch": 0.3631010794896958,
+      "grad_norm": 1.298832893371582,
+      "learning_rate": 6.165555555555556e-05,
+      "loss": 0.1252,
+      "step": 5550
+    },
+    {
+      "epoch": 0.36375531566895647,
+      "grad_norm": 0.9783083200454712,
+      "learning_rate": 6.176666666666667e-05,
+      "loss": 0.1143,
+      "step": 5560
+    },
+    {
+      "epoch": 0.3644095518482172,
+      "grad_norm": 1.0835789442062378,
+      "learning_rate": 6.187777777777777e-05,
+      "loss": 0.1189,
+      "step": 5570
+    },
+    {
+      "epoch": 0.3650637880274779,
+      "grad_norm": 1.1260126829147339,
+      "learning_rate": 6.198888888888889e-05,
+      "loss": 0.1092,
+      "step": 5580
+    },
+    {
+      "epoch": 0.36571802420673866,
+      "grad_norm": 1.1071107387542725,
+      "learning_rate": 6.21e-05,
+      "loss": 0.1108,
+      "step": 5590
+    },
+    {
+      "epoch": 0.36637226038599935,
+      "grad_norm": 1.383540153503418,
+      "learning_rate": 6.221111111111111e-05,
+      "loss": 0.1208,
+      "step": 5600
+    },
+    {
+      "epoch": 0.36702649656526004,
+      "grad_norm": 1.0578371286392212,
+      "learning_rate": 6.232222222222222e-05,
+      "loss": 0.1105,
+      "step": 5610
+    },
+    {
+      "epoch": 0.3676807327445208,
+      "grad_norm": 0.9242531061172485,
+      "learning_rate": 6.243333333333334e-05,
+      "loss": 0.1026,
+      "step": 5620
+    },
+    {
+      "epoch": 0.3683349689237815,
+      "grad_norm": 1.0212069749832153,
+      "learning_rate": 6.254444444444445e-05,
+      "loss": 0.1121,
+      "step": 5630
+    },
+    {
+      "epoch": 0.3689892051030422,
+      "grad_norm": 1.075881004333496,
+      "learning_rate": 6.265555555555555e-05,
+      "loss": 0.107,
+      "step": 5640
+    },
+    {
+      "epoch": 0.3696434412823029,
+      "grad_norm": 0.993894100189209,
+      "learning_rate": 6.276666666666667e-05,
+      "loss": 0.1202,
+      "step": 5650
+    },
+    {
+      "epoch": 0.3702976774615636,
+      "grad_norm": 1.1158571243286133,
+      "learning_rate": 6.287777777777779e-05,
+      "loss": 0.1151,
+      "step": 5660
+    },
+    {
+      "epoch": 0.37095191364082436,
+      "grad_norm": 0.9299028515815735,
+      "learning_rate": 6.29888888888889e-05,
+      "loss": 0.1203,
+      "step": 5670
+    },
+    {
+      "epoch": 0.37160614982008505,
+      "grad_norm": 0.9454100131988525,
+      "learning_rate": 6.31e-05,
+      "loss": 0.1105,
+      "step": 5680
+    },
+    {
+      "epoch": 0.37226038599934574,
+      "grad_norm": 0.9291905164718628,
+      "learning_rate": 6.32111111111111e-05,
+      "loss": 0.1141,
+      "step": 5690
+    },
+    {
+      "epoch": 0.3729146221786065,
+      "grad_norm": 0.9003360867500305,
+      "learning_rate": 6.332222222222223e-05,
+      "loss": 0.1115,
+      "step": 5700
+    },
+    {
+      "epoch": 0.3735688583578672,
+      "grad_norm": 1.0188775062561035,
+      "learning_rate": 6.343333333333333e-05,
+      "loss": 0.1199,
+      "step": 5710
+    },
+    {
+      "epoch": 0.3742230945371279,
+      "grad_norm": 0.9012176990509033,
+      "learning_rate": 6.354444444444445e-05,
+      "loss": 0.1091,
+      "step": 5720
+    },
+    {
+      "epoch": 0.3748773307163886,
+      "grad_norm": 0.977641224861145,
+      "learning_rate": 6.365555555555555e-05,
+      "loss": 0.1138,
+      "step": 5730
+    },
+    {
+      "epoch": 0.3755315668956493,
+      "grad_norm": 0.9474180340766907,
+      "learning_rate": 6.376666666666668e-05,
+      "loss": 0.1085,
+      "step": 5740
+    },
+    {
+      "epoch": 0.37618580307491006,
+      "grad_norm": 0.8891827464103699,
+      "learning_rate": 6.387777777777778e-05,
+      "loss": 0.1183,
+      "step": 5750
+    },
+    {
+      "epoch": 0.37684003925417076,
+      "grad_norm": 1.0365437269210815,
+      "learning_rate": 6.398888888888888e-05,
+      "loss": 0.1166,
+      "step": 5760
+    },
+    {
+      "epoch": 0.37749427543343145,
+      "grad_norm": 0.8813347220420837,
+      "learning_rate": 6.41e-05,
+      "loss": 0.1007,
+      "step": 5770
+    },
+    {
+      "epoch": 0.3781485116126922,
+      "grad_norm": 0.888383150100708,
+      "learning_rate": 6.421111111111111e-05,
+      "loss": 0.1133,
+      "step": 5780
+    },
+    {
+      "epoch": 0.3788027477919529,
+      "grad_norm": 1.1097888946533203,
+      "learning_rate": 6.432222222222223e-05,
+      "loss": 0.1318,
+      "step": 5790
+    },
+    {
+      "epoch": 0.37945698397121363,
+      "grad_norm": 0.928402304649353,
+      "learning_rate": 6.443333333333333e-05,
+      "loss": 0.1141,
+      "step": 5800
+    },
+    {
+      "epoch": 0.3801112201504743,
+      "grad_norm": 1.0633147954940796,
+      "learning_rate": 6.454444444444445e-05,
+      "loss": 0.1198,
+      "step": 5810
+    },
+    {
+      "epoch": 0.380765456329735,
+      "grad_norm": 1.11026132106781,
+      "learning_rate": 6.465555555555556e-05,
+      "loss": 0.1234,
+      "step": 5820
+    },
+    {
+      "epoch": 0.38141969250899577,
+      "grad_norm": 0.9502913951873779,
+      "learning_rate": 6.476666666666666e-05,
+      "loss": 0.1043,
+      "step": 5830
+    },
+    {
+      "epoch": 0.38207392868825646,
+      "grad_norm": 1.1175204515457153,
+      "learning_rate": 6.487777777777778e-05,
+      "loss": 0.1091,
+      "step": 5840
+    },
+    {
+      "epoch": 0.38272816486751715,
+      "grad_norm": 0.9505729079246521,
+      "learning_rate": 6.498888888888888e-05,
+      "loss": 0.1067,
+      "step": 5850
+    },
+    {
+      "epoch": 0.3833824010467779,
+      "grad_norm": 1.11201810836792,
+      "learning_rate": 6.510000000000001e-05,
+      "loss": 0.1123,
+      "step": 5860
+    },
+    {
+      "epoch": 0.3840366372260386,
+      "grad_norm": 1.0022423267364502,
+      "learning_rate": 6.521111111111111e-05,
+      "loss": 0.1076,
+      "step": 5870
+    },
+    {
+      "epoch": 0.38469087340529934,
+      "grad_norm": 1.1039308309555054,
+      "learning_rate": 6.532222222222223e-05,
+      "loss": 0.1086,
+      "step": 5880
+    },
+    {
+      "epoch": 0.38534510958456003,
+      "grad_norm": 0.970061182975769,
+      "learning_rate": 6.543333333333333e-05,
+      "loss": 0.1192,
+      "step": 5890
+    },
+    {
+      "epoch": 0.3859993457638207,
+      "grad_norm": 1.0175315141677856,
+      "learning_rate": 6.554444444444446e-05,
+      "loss": 0.1301,
+      "step": 5900
+    },
+    {
+      "epoch": 0.38665358194308147,
+      "grad_norm": 1.1553181409835815,
+      "learning_rate": 6.565555555555556e-05,
+      "loss": 0.1175,
+      "step": 5910
+    },
+    {
+      "epoch": 0.38730781812234216,
+      "grad_norm": 0.9421599507331848,
+      "learning_rate": 6.576666666666666e-05,
+      "loss": 0.1147,
+      "step": 5920
+    },
+    {
+      "epoch": 0.38796205430160285,
+      "grad_norm": 1.103137731552124,
+      "learning_rate": 6.587777777777778e-05,
+      "loss": 0.1268,
+      "step": 5930
+    },
+    {
+      "epoch": 0.3886162904808636,
+      "grad_norm": 1.0151023864746094,
+      "learning_rate": 6.598888888888889e-05,
+      "loss": 0.1103,
+      "step": 5940
+    },
+    {
+      "epoch": 0.3892705266601243,
+      "grad_norm": 1.017514944076538,
+      "learning_rate": 6.610000000000001e-05,
+      "loss": 0.1168,
+      "step": 5950
+    },
+    {
+      "epoch": 0.38992476283938504,
+      "grad_norm": 0.9031379222869873,
+      "learning_rate": 6.621111111111111e-05,
+      "loss": 0.1191,
+      "step": 5960
+    },
+    {
+      "epoch": 0.39057899901864573,
+      "grad_norm": 1.1167620420455933,
+      "learning_rate": 6.632222222222222e-05,
+      "loss": 0.1032,
+      "step": 5970
+    },
+    {
+      "epoch": 0.3912332351979064,
+      "grad_norm": 0.9715782403945923,
+      "learning_rate": 6.643333333333334e-05,
+      "loss": 0.1231,
+      "step": 5980
+    },
+    {
+      "epoch": 0.3918874713771672,
+      "grad_norm": 1.0662617683410645,
+      "learning_rate": 6.654444444444444e-05,
+      "loss": 0.1122,
+      "step": 5990
+    },
+    {
+      "epoch": 0.39254170755642787,
+      "grad_norm": 0.9010646343231201,
+      "learning_rate": 6.665555555555556e-05,
+      "loss": 0.1109,
+      "step": 6000
+    },
+    {
+      "epoch": 0.39319594373568856,
+      "grad_norm": 0.9664989709854126,
+      "learning_rate": 6.676666666666667e-05,
+      "loss": 0.1195,
+      "step": 6010
+    },
+    {
+      "epoch": 0.3938501799149493,
+      "grad_norm": 1.0604381561279297,
+      "learning_rate": 6.687777777777779e-05,
+      "loss": 0.1013,
+      "step": 6020
+    },
+    {
+      "epoch": 0.39450441609421,
+      "grad_norm": 1.0185221433639526,
+      "learning_rate": 6.698888888888889e-05,
+      "loss": 0.1087,
+      "step": 6030
+    },
+    {
+      "epoch": 0.39515865227347075,
+      "grad_norm": 1.0667742490768433,
+      "learning_rate": 6.71e-05,
+      "loss": 0.1108,
+      "step": 6040
+    },
+    {
+      "epoch": 0.39581288845273144,
+      "grad_norm": 0.9968790411949158,
+      "learning_rate": 6.721111111111112e-05,
+      "loss": 0.1254,
+      "step": 6050
+    },
+    {
+      "epoch": 0.39646712463199213,
+      "grad_norm": 0.9908486604690552,
+      "learning_rate": 6.732222222222224e-05,
+      "loss": 0.1098,
+      "step": 6060
+    },
+    {
+      "epoch": 0.3971213608112529,
+      "grad_norm": 1.073398470878601,
+      "learning_rate": 6.743333333333334e-05,
+      "loss": 0.1034,
+      "step": 6070
+    },
+    {
+      "epoch": 0.39777559699051357,
+      "grad_norm": 0.9481722712516785,
+      "learning_rate": 6.754444444444444e-05,
+      "loss": 0.1141,
+      "step": 6080
+    },
+    {
+      "epoch": 0.39842983316977426,
+      "grad_norm": 1.1808781623840332,
+      "learning_rate": 6.765555555555555e-05,
+      "loss": 0.113,
+      "step": 6090
+    },
+    {
+      "epoch": 0.399084069349035,
+      "grad_norm": 1.1544404029846191,
+      "learning_rate": 6.776666666666667e-05,
+      "loss": 0.1118,
+      "step": 6100
+    },
+    {
+      "epoch": 0.3997383055282957,
+      "grad_norm": 1.1222143173217773,
+      "learning_rate": 6.787777777777778e-05,
+      "loss": 0.1001,
+      "step": 6110
+    },
+    {
+      "epoch": 0.40039254170755645,
+      "grad_norm": 0.9140275716781616,
+      "learning_rate": 6.798888888888889e-05,
+      "loss": 0.1146,
+      "step": 6120
+    },
+    {
+      "epoch": 0.40104677788681714,
+      "grad_norm": 0.9539370536804199,
+      "learning_rate": 6.81e-05,
+      "loss": 0.1135,
+      "step": 6130
+    },
+    {
+      "epoch": 0.40170101406607783,
+      "grad_norm": 1.0647083520889282,
+      "learning_rate": 6.821111111111112e-05,
+      "loss": 0.1102,
+      "step": 6140
+    },
+    {
+      "epoch": 0.4023552502453386,
+      "grad_norm": 0.9687780141830444,
+      "learning_rate": 6.832222222222222e-05,
+      "loss": 0.1164,
+      "step": 6150
+    },
+    {
+      "epoch": 0.4030094864245993,
+      "grad_norm": 1.0673967599868774,
+      "learning_rate": 6.843333333333333e-05,
+      "loss": 0.1166,
+      "step": 6160
+    },
+    {
+      "epoch": 0.40366372260386,
+      "grad_norm": 1.0471779108047485,
+      "learning_rate": 6.854444444444445e-05,
+      "loss": 0.11,
+      "step": 6170
+    },
+    {
+      "epoch": 0.4043179587831207,
+      "grad_norm": 1.0822899341583252,
+      "learning_rate": 6.865555555555556e-05,
+      "loss": 0.1246,
+      "step": 6180
+    },
+    {
+      "epoch": 0.4049721949623814,
+      "grad_norm": 0.7718848586082458,
+      "learning_rate": 6.876666666666667e-05,
+      "loss": 0.1116,
+      "step": 6190
+    },
+    {
+      "epoch": 0.40562643114164215,
+      "grad_norm": 1.1657593250274658,
+      "learning_rate": 6.887777777777778e-05,
+      "loss": 0.1087,
+      "step": 6200
+    },
+    {
+      "epoch": 0.40628066732090284,
+      "grad_norm": 0.8355668783187866,
+      "learning_rate": 6.89888888888889e-05,
+      "loss": 0.1003,
+      "step": 6210
+    },
+    {
+      "epoch": 0.40693490350016354,
+      "grad_norm": 0.9746671319007874,
+      "learning_rate": 6.91e-05,
+      "loss": 0.1282,
+      "step": 6220
+    },
+    {
+      "epoch": 0.4075891396794243,
+      "grad_norm": 1.1945754289627075,
+      "learning_rate": 6.921111111111111e-05,
+      "loss": 0.1219,
+      "step": 6230
+    },
+    {
+      "epoch": 0.408243375858685,
+      "grad_norm": 1.1995118856430054,
+      "learning_rate": 6.932222222222222e-05,
+      "loss": 0.1212,
+      "step": 6240
+    },
+    {
+      "epoch": 0.4088976120379457,
+      "grad_norm": 1.1618099212646484,
+      "learning_rate": 6.943333333333335e-05,
+      "loss": 0.117,
+      "step": 6250
+    },
+    {
+      "epoch": 0.4095518482172064,
+      "grad_norm": 0.8268874883651733,
+      "learning_rate": 6.954444444444445e-05,
+      "loss": 0.1141,
+      "step": 6260
+    },
+    {
+      "epoch": 0.4102060843964671,
+      "grad_norm": 0.8862093091011047,
+      "learning_rate": 6.965555555555556e-05,
+      "loss": 0.122,
+      "step": 6270
+    },
+    {
+      "epoch": 0.41086032057572786,
+      "grad_norm": 1.2067736387252808,
+      "learning_rate": 6.976666666666666e-05,
+      "loss": 0.1089,
+      "step": 6280
+    },
+    {
+      "epoch": 0.41151455675498855,
+      "grad_norm": 1.0685234069824219,
+      "learning_rate": 6.987777777777779e-05,
+      "loss": 0.1147,
+      "step": 6290
+    },
+    {
+      "epoch": 0.41216879293424924,
+      "grad_norm": 0.9467248916625977,
+      "learning_rate": 6.99888888888889e-05,
+      "loss": 0.1282,
+      "step": 6300
+    },
+    {
+      "epoch": 0.41282302911351,
+      "grad_norm": 1.2121899127960205,
+      "learning_rate": 7.01e-05,
+      "loss": 0.1052,
+      "step": 6310
+    },
+    {
+      "epoch": 0.4134772652927707,
+      "grad_norm": 1.1744288206100464,
+      "learning_rate": 7.021111111111111e-05,
+      "loss": 0.1119,
+      "step": 6320
+    },
+    {
+      "epoch": 0.4141315014720314,
+      "grad_norm": 1.219565749168396,
+      "learning_rate": 7.032222222222223e-05,
+      "loss": 0.1203,
+      "step": 6330
+    },
+    {
+      "epoch": 0.4147857376512921,
+      "grad_norm": 0.9562080502510071,
+      "learning_rate": 7.043333333333334e-05,
+      "loss": 0.1157,
+      "step": 6340
+    },
+    {
+      "epoch": 0.4154399738305528,
+      "grad_norm": 1.188444972038269,
+      "learning_rate": 7.054444444444444e-05,
+      "loss": 0.1153,
+      "step": 6350
+    },
+    {
+      "epoch": 0.41609421000981356,
+      "grad_norm": 1.1564666032791138,
+      "learning_rate": 7.065555555555556e-05,
+      "loss": 0.1135,
+      "step": 6360
+    },
+    {
+      "epoch": 0.41674844618907425,
+      "grad_norm": 0.9498192071914673,
+      "learning_rate": 7.076666666666667e-05,
+      "loss": 0.119,
+      "step": 6370
+    },
+    {
+      "epoch": 0.41740268236833494,
+      "grad_norm": 1.0807279348373413,
+      "learning_rate": 7.087777777777778e-05,
+      "loss": 0.1176,
+      "step": 6380
+    },
+    {
+      "epoch": 0.4180569185475957,
+      "grad_norm": 1.169202446937561,
+      "learning_rate": 7.098888888888889e-05,
+      "loss": 0.1206,
+      "step": 6390
+    },
+    {
+      "epoch": 0.4187111547268564,
+      "grad_norm": 1.1617976427078247,
+      "learning_rate": 7.11e-05,
+      "loss": 0.1017,
+      "step": 6400
+    },
+    {
+      "epoch": 0.41936539090611713,
+      "grad_norm": 0.8993281126022339,
+      "learning_rate": 7.121111111111112e-05,
+      "loss": 0.1053,
+      "step": 6410
+    },
+    {
+      "epoch": 0.4200196270853778,
+      "grad_norm": 1.145491600036621,
+      "learning_rate": 7.132222222222222e-05,
+      "loss": 0.1247,
+      "step": 6420
+    },
+    {
+      "epoch": 0.4206738632646385,
+      "grad_norm": 1.1302369832992554,
+      "learning_rate": 7.143333333333334e-05,
+      "loss": 0.1165,
+      "step": 6430
+    },
+    {
+      "epoch": 0.42132809944389926,
+      "grad_norm": 1.070312261581421,
+      "learning_rate": 7.154444444444444e-05,
+      "loss": 0.112,
+      "step": 6440
+    },
+    {
+      "epoch": 0.42198233562315995,
+      "grad_norm": 1.261004090309143,
+      "learning_rate": 7.165555555555556e-05,
+      "loss": 0.1202,
+      "step": 6450
+    },
+    {
+      "epoch": 0.42263657180242065,
+      "grad_norm": 1.149681806564331,
+      "learning_rate": 7.176666666666667e-05,
+      "loss": 0.1116,
+      "step": 6460
+    },
+    {
+      "epoch": 0.4232908079816814,
+      "grad_norm": 1.0995161533355713,
+      "learning_rate": 7.187777777777777e-05,
+      "loss": 0.1068,
+      "step": 6470
+    },
+    {
+      "epoch": 0.4239450441609421,
+      "grad_norm": 1.2060518264770508,
+      "learning_rate": 7.198888888888889e-05,
+      "loss": 0.1178,
+      "step": 6480
+    },
+    {
+      "epoch": 0.42459928034020283,
+      "grad_norm": 0.885783851146698,
+      "learning_rate": 7.21e-05,
+      "loss": 0.1114,
+      "step": 6490
+    },
+    {
+      "epoch": 0.4252535165194635,
+      "grad_norm": 1.0212535858154297,
+      "learning_rate": 7.221111111111112e-05,
+      "loss": 0.102,
+      "step": 6500
+    },
+    {
+      "epoch": 0.4259077526987242,
+      "grad_norm": 0.9369866847991943,
+      "learning_rate": 7.232222222222222e-05,
+      "loss": 0.1018,
+      "step": 6510
+    },
+    {
+      "epoch": 0.42656198887798497,
+      "grad_norm": 1.102473497390747,
+      "learning_rate": 7.243333333333334e-05,
+      "loss": 0.1163,
+      "step": 6520
+    },
+    {
+      "epoch": 0.42721622505724566,
+      "grad_norm": 0.9343268871307373,
+      "learning_rate": 7.254444444444445e-05,
+      "loss": 0.1118,
+      "step": 6530
+    },
+    {
+      "epoch": 0.4278704612365064,
+      "grad_norm": 0.9922140836715698,
+      "learning_rate": 7.265555555555555e-05,
+      "loss": 0.1312,
+      "step": 6540
+    },
+    {
+      "epoch": 0.4285246974157671,
+      "grad_norm": 0.92384934425354,
+      "learning_rate": 7.276666666666667e-05,
+      "loss": 0.1152,
+      "step": 6550
+    },
+    {
+      "epoch": 0.4291789335950278,
+      "grad_norm": 0.8126742243766785,
+      "learning_rate": 7.287777777777778e-05,
+      "loss": 0.1132,
+      "step": 6560
+    },
+    {
+      "epoch": 0.42983316977428854,
+      "grad_norm": 0.8990784287452698,
+      "learning_rate": 7.29888888888889e-05,
+      "loss": 0.1172,
+      "step": 6570
+    },
+    {
+      "epoch": 0.43048740595354923,
+      "grad_norm": 0.7989428043365479,
+      "learning_rate": 7.31e-05,
+      "loss": 0.1146,
+      "step": 6580
+    },
+    {
+      "epoch": 0.4311416421328099,
+      "grad_norm": 1.0440112352371216,
+      "learning_rate": 7.321111111111112e-05,
+      "loss": 0.1257,
+      "step": 6590
+    },
+    {
+      "epoch": 0.43179587831207067,
+      "grad_norm": 0.9922072887420654,
+      "learning_rate": 7.332222222222223e-05,
+      "loss": 0.116,
+      "step": 6600
+    },
+    {
+      "epoch": 0.43245011449133136,
+      "grad_norm": 0.951362133026123,
+      "learning_rate": 7.343333333333333e-05,
+      "loss": 0.1205,
+      "step": 6610
+    },
+    {
+      "epoch": 0.4331043506705921,
+      "grad_norm": 1.096359133720398,
+      "learning_rate": 7.354444444444445e-05,
+      "loss": 0.1092,
+      "step": 6620
+    },
+    {
+      "epoch": 0.4337585868498528,
+      "grad_norm": 0.9648379683494568,
+      "learning_rate": 7.365555555555555e-05,
+      "loss": 0.1121,
+      "step": 6630
+    },
+    {
+      "epoch": 0.4344128230291135,
+      "grad_norm": 0.9251676797866821,
+      "learning_rate": 7.376666666666667e-05,
+      "loss": 0.1139,
+      "step": 6640
+    },
+    {
+      "epoch": 0.43506705920837424,
+      "grad_norm": 1.0393186807632446,
+      "learning_rate": 7.387777777777778e-05,
+      "loss": 0.1136,
+      "step": 6650
+    },
+    {
+      "epoch": 0.43572129538763493,
+      "grad_norm": 0.870494544506073,
+      "learning_rate": 7.39888888888889e-05,
+      "loss": 0.1052,
+      "step": 6660
+    },
+    {
+      "epoch": 0.4363755315668956,
+      "grad_norm": 0.9582804441452026,
+      "learning_rate": 7.41e-05,
+      "loss": 0.1218,
+      "step": 6670
+    },
+    {
+      "epoch": 0.4370297677461564,
+      "grad_norm": 1.166326642036438,
+      "learning_rate": 7.421111111111111e-05,
+      "loss": 0.1126,
+      "step": 6680
+    },
+    {
+      "epoch": 0.43768400392541706,
+      "grad_norm": 0.9711355566978455,
+      "learning_rate": 7.432222222222223e-05,
+      "loss": 0.1186,
+      "step": 6690
+    },
+    {
+      "epoch": 0.4383382401046778,
+      "grad_norm": 1.110071063041687,
+      "learning_rate": 7.443333333333333e-05,
+      "loss": 0.1117,
+      "step": 6700
+    },
+    {
+      "epoch": 0.4389924762839385,
+      "grad_norm": 0.836465060710907,
+      "learning_rate": 7.454444444444445e-05,
+      "loss": 0.1108,
+      "step": 6710
+    },
+    {
+      "epoch": 0.4396467124631992,
+      "grad_norm": 1.120985507965088,
+      "learning_rate": 7.465555555555556e-05,
+      "loss": 0.1172,
+      "step": 6720
+    },
+    {
+      "epoch": 0.44030094864245994,
+      "grad_norm": 1.0334186553955078,
+      "learning_rate": 7.476666666666668e-05,
+      "loss": 0.1142,
+      "step": 6730
+    },
+    {
+      "epoch": 0.44095518482172064,
+      "grad_norm": 1.0058670043945312,
+      "learning_rate": 7.487777777777778e-05,
+      "loss": 0.1191,
+      "step": 6740
+    },
+    {
+      "epoch": 0.44160942100098133,
+      "grad_norm": 1.1377954483032227,
+      "learning_rate": 7.49888888888889e-05,
+      "loss": 0.1083,
+      "step": 6750
+    },
+    {
+      "epoch": 0.4422636571802421,
+      "grad_norm": 1.0749398469924927,
+      "learning_rate": 7.510000000000001e-05,
+      "loss": 0.1275,
+      "step": 6760
+    },
+    {
+      "epoch": 0.44291789335950277,
+      "grad_norm": 1.0065230131149292,
+      "learning_rate": 7.521111111111111e-05,
+      "loss": 0.1192,
+      "step": 6770
+    },
+    {
+      "epoch": 0.4435721295387635,
+      "grad_norm": 1.1602001190185547,
+      "learning_rate": 7.532222222222223e-05,
+      "loss": 0.1137,
+      "step": 6780
+    },
+    {
+      "epoch": 0.4442263657180242,
+      "grad_norm": 1.0304006338119507,
+      "learning_rate": 7.543333333333333e-05,
+      "loss": 0.1206,
+      "step": 6790
+    },
+    {
+      "epoch": 0.4448806018972849,
+      "grad_norm": 1.0605281591415405,
+      "learning_rate": 7.554444444444446e-05,
+      "loss": 0.1171,
+      "step": 6800
+    },
+    {
+      "epoch": 0.44553483807654565,
+      "grad_norm": 0.8925848007202148,
+      "learning_rate": 7.565555555555556e-05,
+      "loss": 0.1091,
+      "step": 6810
+    },
+    {
+      "epoch": 0.44618907425580634,
+      "grad_norm": 0.9806022047996521,
+      "learning_rate": 7.576666666666667e-05,
+      "loss": 0.1168,
+      "step": 6820
+    },
+    {
+      "epoch": 0.44684331043506703,
+      "grad_norm": 1.2420053482055664,
+      "learning_rate": 7.587777777777778e-05,
+      "loss": 0.117,
+      "step": 6830
+    },
+    {
+      "epoch": 0.4474975466143278,
+      "grad_norm": 1.1187241077423096,
+      "learning_rate": 7.598888888888889e-05,
+      "loss": 0.1286,
+      "step": 6840
+    },
+    {
+      "epoch": 0.44815178279358847,
+      "grad_norm": 1.2027949094772339,
+      "learning_rate": 7.61e-05,
+      "loss": 0.1178,
+      "step": 6850
+    },
+    {
+      "epoch": 0.4488060189728492,
+      "grad_norm": 1.1143661737442017,
+      "learning_rate": 7.621111111111111e-05,
+      "loss": 0.115,
+      "step": 6860
+    },
+    {
+      "epoch": 0.4494602551521099,
+      "grad_norm": 0.9064378142356873,
+      "learning_rate": 7.632222222222222e-05,
+      "loss": 0.1141,
+      "step": 6870
+    },
+    {
+      "epoch": 0.4501144913313706,
+      "grad_norm": 1.2292486429214478,
+      "learning_rate": 7.643333333333334e-05,
+      "loss": 0.1234,
+      "step": 6880
+    },
+    {
+      "epoch": 0.45076872751063135,
+      "grad_norm": 1.2885360717773438,
+      "learning_rate": 7.654444444444445e-05,
+      "loss": 0.1205,
+      "step": 6890
+    },
+    {
+      "epoch": 0.45142296368989204,
+      "grad_norm": 0.9081943035125732,
+      "learning_rate": 7.665555555555556e-05,
+      "loss": 0.1226,
+      "step": 6900
+    },
+    {
+      "epoch": 0.4520771998691528,
+      "grad_norm": 1.1508632898330688,
+      "learning_rate": 7.676666666666667e-05,
+      "loss": 0.1219,
+      "step": 6910
+    },
+    {
+      "epoch": 0.4527314360484135,
+      "grad_norm": 0.971246063709259,
+      "learning_rate": 7.687777777777779e-05,
+      "loss": 0.1315,
+      "step": 6920
+    },
+    {
+      "epoch": 0.4533856722276742,
+      "grad_norm": 1.2282114028930664,
+      "learning_rate": 7.698888888888889e-05,
+      "loss": 0.1183,
+      "step": 6930
+    },
+    {
+      "epoch": 0.4540399084069349,
+      "grad_norm": 1.111584186553955,
+      "learning_rate": 7.71e-05,
+      "loss": 0.1193,
+      "step": 6940
+    },
+    {
+      "epoch": 0.4546941445861956,
+      "grad_norm": 1.0776195526123047,
+      "learning_rate": 7.72111111111111e-05,
+      "loss": 0.1157,
+      "step": 6950
+    },
+    {
+      "epoch": 0.4553483807654563,
+      "grad_norm": 1.1044821739196777,
+      "learning_rate": 7.732222222222223e-05,
+      "loss": 0.1191,
+      "step": 6960
+    },
+    {
+      "epoch": 0.45600261694471705,
+      "grad_norm": 0.9981603622436523,
+      "learning_rate": 7.743333333333334e-05,
+      "loss": 0.1231,
+      "step": 6970
+    },
+    {
+      "epoch": 0.45665685312397775,
+      "grad_norm": 1.0244512557983398,
+      "learning_rate": 7.754444444444445e-05,
+      "loss": 0.1175,
+      "step": 6980
+    },
+    {
+      "epoch": 0.4573110893032385,
+      "grad_norm": 0.9602757692337036,
+      "learning_rate": 7.765555555555555e-05,
+      "loss": 0.1095,
+      "step": 6990
+    },
+    {
+      "epoch": 0.4579653254824992,
+      "grad_norm": 1.00235116481781,
+      "learning_rate": 7.776666666666667e-05,
+      "loss": 0.1066,
+      "step": 7000
+    },
+    {
+      "epoch": 0.4586195616617599,
+      "grad_norm": 0.9261903762817383,
+      "learning_rate": 7.787777777777778e-05,
+      "loss": 0.1073,
+      "step": 7010
+    },
+    {
+      "epoch": 0.4592737978410206,
+      "grad_norm": 0.9496639370918274,
+      "learning_rate": 7.798888888888889e-05,
+      "loss": 0.1291,
+      "step": 7020
+    },
+    {
+      "epoch": 0.4599280340202813,
+      "grad_norm": 0.983691394329071,
+      "learning_rate": 7.81e-05,
+      "loss": 0.1113,
+      "step": 7030
+    },
+    {
+      "epoch": 0.460582270199542,
+      "grad_norm": 1.006894826889038,
+      "learning_rate": 7.821111111111112e-05,
+      "loss": 0.1102,
+      "step": 7040
+    },
+    {
+      "epoch": 0.46123650637880276,
+      "grad_norm": 0.907843291759491,
+      "learning_rate": 7.832222222222223e-05,
+      "loss": 0.1161,
+      "step": 7050
+    },
+    {
+      "epoch": 0.46189074255806345,
+      "grad_norm": 1.1818556785583496,
+      "learning_rate": 7.843333333333333e-05,
+      "loss": 0.1193,
+      "step": 7060
+    },
+    {
+      "epoch": 0.4625449787373242,
+      "grad_norm": 1.0127677917480469,
+      "learning_rate": 7.854444444444445e-05,
+      "loss": 0.1149,
+      "step": 7070
+    },
+    {
+      "epoch": 0.4631992149165849,
+      "grad_norm": 0.9791194796562195,
+      "learning_rate": 7.865555555555556e-05,
+      "loss": 0.1251,
+      "step": 7080
+    },
+    {
+      "epoch": 0.4638534510958456,
+      "grad_norm": 1.0158543586730957,
+      "learning_rate": 7.876666666666667e-05,
+      "loss": 0.1137,
+      "step": 7090
+    },
+    {
+      "epoch": 0.46450768727510633,
+      "grad_norm": 1.2994464635849,
+      "learning_rate": 7.887777777777778e-05,
+      "loss": 0.1277,
+      "step": 7100
+    },
+    {
+      "epoch": 0.465161923454367,
+      "grad_norm": 1.0386182069778442,
+      "learning_rate": 7.89888888888889e-05,
+      "loss": 0.1398,
+      "step": 7110
+    },
+    {
+      "epoch": 0.4658161596336277,
+      "grad_norm": 0.8721251487731934,
+      "learning_rate": 7.910000000000001e-05,
+      "loss": 0.1043,
+      "step": 7120
+    },
+    {
+      "epoch": 0.46647039581288846,
+      "grad_norm": 1.1872791051864624,
+      "learning_rate": 7.921111111111111e-05,
+      "loss": 0.1087,
+      "step": 7130
+    },
+    {
+      "epoch": 0.46712463199214915,
+      "grad_norm": 1.0702382326126099,
+      "learning_rate": 7.932222222222223e-05,
+      "loss": 0.1136,
+      "step": 7140
+    },
+    {
+      "epoch": 0.4677788681714099,
+      "grad_norm": 1.5441776514053345,
+      "learning_rate": 7.943333333333333e-05,
+      "loss": 0.1156,
+      "step": 7150
+    },
+    {
+      "epoch": 0.4684331043506706,
+      "grad_norm": 1.2157565355300903,
+      "learning_rate": 7.954444444444445e-05,
+      "loss": 0.1146,
+      "step": 7160
+    },
+    {
+      "epoch": 0.4690873405299313,
+      "grad_norm": 0.9208630323410034,
+      "learning_rate": 7.965555555555556e-05,
+      "loss": 0.1057,
+      "step": 7170
+    },
+    {
+      "epoch": 0.46974157670919203,
+      "grad_norm": 1.2033475637435913,
+      "learning_rate": 7.976666666666666e-05,
+      "loss": 0.1173,
+      "step": 7180
+    },
+    {
+      "epoch": 0.4703958128884527,
+      "grad_norm": 1.084647297859192,
+      "learning_rate": 7.987777777777778e-05,
+      "loss": 0.1103,
+      "step": 7190
+    },
+    {
+      "epoch": 0.47105004906771347,
+      "grad_norm": 1.0333460569381714,
+      "learning_rate": 7.99888888888889e-05,
+      "loss": 0.1151,
+      "step": 7200
+    },
+    {
+      "epoch": 0.47170428524697416,
+      "grad_norm": 0.7924304604530334,
+      "learning_rate": 8.010000000000001e-05,
+      "loss": 0.1198,
+      "step": 7210
+    },
+    {
+      "epoch": 0.47235852142623486,
+      "grad_norm": 0.9112861156463623,
+      "learning_rate": 8.021111111111111e-05,
+      "loss": 0.1116,
+      "step": 7220
+    },
+    {
+      "epoch": 0.4730127576054956,
+      "grad_norm": 0.8361260890960693,
+      "learning_rate": 8.032222222222223e-05,
+      "loss": 0.1169,
+      "step": 7230
+    },
+    {
+      "epoch": 0.4736669937847563,
+      "grad_norm": 1.0955017805099487,
+      "learning_rate": 8.043333333333334e-05,
+      "loss": 0.1171,
+      "step": 7240
+    },
+    {
+      "epoch": 0.474321229964017,
+      "grad_norm": 0.775754988193512,
+      "learning_rate": 8.054444444444444e-05,
+      "loss": 0.1109,
+      "step": 7250
+    },
+    {
+      "epoch": 0.47497546614327774,
+      "grad_norm": 0.9245556592941284,
+      "learning_rate": 8.065555555555556e-05,
+      "loss": 0.1185,
+      "step": 7260
+    },
+    {
+      "epoch": 0.47562970232253843,
+      "grad_norm": 0.9954939484596252,
+      "learning_rate": 8.076666666666667e-05,
+      "loss": 0.1199,
+      "step": 7270
+    },
+    {
+      "epoch": 0.4762839385017992,
+      "grad_norm": 0.9574825167655945,
+      "learning_rate": 8.087777777777779e-05,
+      "loss": 0.1154,
+      "step": 7280
+    },
+    {
+      "epoch": 0.47693817468105987,
+      "grad_norm": 1.0296094417572021,
+      "learning_rate": 8.098888888888889e-05,
+      "loss": 0.1226,
+      "step": 7290
+    },
+    {
+      "epoch": 0.47759241086032056,
+      "grad_norm": 1.061394453048706,
+      "learning_rate": 8.11e-05,
+      "loss": 0.1142,
+      "step": 7300
+    },
+    {
+      "epoch": 0.4782466470395813,
+      "grad_norm": 1.1966540813446045,
+      "learning_rate": 8.121111111111112e-05,
+      "loss": 0.1252,
+      "step": 7310
+    },
+    {
+      "epoch": 0.478900883218842,
+      "grad_norm": 1.1094386577606201,
+      "learning_rate": 8.132222222222222e-05,
+      "loss": 0.12,
+      "step": 7320
+    },
+    {
+      "epoch": 0.4795551193981027,
+      "grad_norm": 1.1346668004989624,
+      "learning_rate": 8.143333333333334e-05,
+      "loss": 0.121,
+      "step": 7330
+    },
+    {
+      "epoch": 0.48020935557736344,
+      "grad_norm": 1.08058500289917,
+      "learning_rate": 8.154444444444444e-05,
+      "loss": 0.1175,
+      "step": 7340
+    },
+    {
+      "epoch": 0.48086359175662413,
+      "grad_norm": 1.0835899114608765,
+      "learning_rate": 8.165555555555557e-05,
+      "loss": 0.1139,
+      "step": 7350
+    },
+    {
+      "epoch": 0.4815178279358849,
+      "grad_norm": 1.1137175559997559,
+      "learning_rate": 8.176666666666667e-05,
+      "loss": 0.1182,
+      "step": 7360
+    },
+    {
+      "epoch": 0.48217206411514557,
+      "grad_norm": 0.8468176126480103,
+      "learning_rate": 8.187777777777779e-05,
+      "loss": 0.1112,
+      "step": 7370
+    },
+    {
+      "epoch": 0.48282630029440626,
+      "grad_norm": 0.984216034412384,
+      "learning_rate": 8.198888888888889e-05,
+      "loss": 0.1158,
+      "step": 7380
+    },
+    {
+      "epoch": 0.483480536473667,
+      "grad_norm": 1.0112940073013306,
+      "learning_rate": 8.21e-05,
+      "loss": 0.1184,
+      "step": 7390
+    },
+    {
+      "epoch": 0.4841347726529277,
+      "grad_norm": 1.1188663244247437,
+      "learning_rate": 8.221111111111112e-05,
+      "loss": 0.1205,
+      "step": 7400
+    },
+    {
+      "epoch": 0.4847890088321884,
+      "grad_norm": 1.0517669916152954,
+      "learning_rate": 8.232222222222222e-05,
+      "loss": 0.1175,
+      "step": 7410
+    },
+    {
+      "epoch": 0.48544324501144914,
+      "grad_norm": 0.9220474362373352,
+      "learning_rate": 8.243333333333334e-05,
+      "loss": 0.1072,
+      "step": 7420
+    },
+    {
+      "epoch": 0.48609748119070983,
+      "grad_norm": 0.9419307708740234,
+      "learning_rate": 8.254444444444445e-05,
+      "loss": 0.1173,
+      "step": 7430
+    },
+    {
+      "epoch": 0.4867517173699706,
+      "grad_norm": 0.7772601842880249,
+      "learning_rate": 8.265555555555557e-05,
+      "loss": 0.113,
+      "step": 7440
+    },
+    {
+      "epoch": 0.4874059535492313,
+      "grad_norm": 0.9354446530342102,
+      "learning_rate": 8.276666666666667e-05,
+      "loss": 0.1158,
+      "step": 7450
+    },
+    {
+      "epoch": 0.48806018972849197,
+      "grad_norm": 1.0952316522598267,
+      "learning_rate": 8.287777777777777e-05,
+      "loss": 0.1092,
+      "step": 7460
+    },
+    {
+      "epoch": 0.4887144259077527,
+      "grad_norm": 0.9121578335762024,
+      "learning_rate": 8.29888888888889e-05,
+      "loss": 0.1232,
+      "step": 7470
+    },
+    {
+      "epoch": 0.4893686620870134,
+      "grad_norm": 1.195021152496338,
+      "learning_rate": 8.31e-05,
+      "loss": 0.117,
+      "step": 7480
+    },
+    {
+      "epoch": 0.4900228982662741,
+      "grad_norm": 1.1168640851974487,
+      "learning_rate": 8.321111111111112e-05,
+      "loss": 0.1202,
+      "step": 7490
+    },
+    {
+      "epoch": 0.49067713444553485,
+      "grad_norm": 1.1179988384246826,
+      "learning_rate": 8.332222222222222e-05,
+      "loss": 0.1173,
+      "step": 7500
+    },
+    {
+      "epoch": 0.49133137062479554,
+      "grad_norm": 1.026973843574524,
+      "learning_rate": 8.343333333333335e-05,
+      "loss": 0.1139,
+      "step": 7510
+    },
+    {
+      "epoch": 0.4919856068040563,
+      "grad_norm": 1.1670565605163574,
+      "learning_rate": 8.354444444444445e-05,
+      "loss": 0.12,
+      "step": 7520
+    },
+    {
+      "epoch": 0.492639842983317,
+      "grad_norm": 0.9871100187301636,
+      "learning_rate": 8.365555555555556e-05,
+      "loss": 0.1218,
+      "step": 7530
+    },
+    {
+      "epoch": 0.49329407916257767,
+      "grad_norm": 1.137722134590149,
+      "learning_rate": 8.376666666666667e-05,
+      "loss": 0.1178,
+      "step": 7540
+    },
+    {
+      "epoch": 0.4939483153418384,
+      "grad_norm": 1.0345990657806396,
+      "learning_rate": 8.387777777777778e-05,
+      "loss": 0.1226,
+      "step": 7550
+    },
+    {
+      "epoch": 0.4946025515210991,
+      "grad_norm": 1.140154480934143,
+      "learning_rate": 8.39888888888889e-05,
+      "loss": 0.1297,
+      "step": 7560
+    },
+    {
+      "epoch": 0.49525678770035986,
+      "grad_norm": 0.97530198097229,
+      "learning_rate": 8.41e-05,
+      "loss": 0.1194,
+      "step": 7570
+    },
+    {
+      "epoch": 0.49591102387962055,
+      "grad_norm": 1.1377121210098267,
+      "learning_rate": 8.421111111111111e-05,
+      "loss": 0.1295,
+      "step": 7580
+    },
+    {
+      "epoch": 0.49656526005888124,
+      "grad_norm": 1.4145655632019043,
+      "learning_rate": 8.432222222222223e-05,
+      "loss": 0.1145,
+      "step": 7590
+    },
+    {
+      "epoch": 0.497219496238142,
+      "grad_norm": 1.2647418975830078,
+      "learning_rate": 8.443333333333334e-05,
+      "loss": 0.1254,
+      "step": 7600
+    },
+    {
+      "epoch": 0.4978737324174027,
+      "grad_norm": 1.0249871015548706,
+      "learning_rate": 8.454444444444445e-05,
+      "loss": 0.1249,
+      "step": 7610
+    },
+    {
+      "epoch": 0.4985279685966634,
+      "grad_norm": 0.8847797513008118,
+      "learning_rate": 8.465555555555556e-05,
+      "loss": 0.1039,
+      "step": 7620
+    },
+    {
+      "epoch": 0.4991822047759241,
+      "grad_norm": 1.2074919939041138,
+      "learning_rate": 8.476666666666668e-05,
+      "loss": 0.1109,
+      "step": 7630
+    },
+    {
+      "epoch": 0.4998364409551848,
+      "grad_norm": 0.9957761764526367,
+      "learning_rate": 8.487777777777778e-05,
+      "loss": 0.1224,
+      "step": 7640
+    },
+    {
+      "epoch": 0.5004906771344455,
+      "grad_norm": 0.8962732553482056,
+      "learning_rate": 8.498888888888889e-05,
+      "loss": 0.1247,
+      "step": 7650
+    },
+    {
+      "epoch": 0.5011449133137063,
+      "grad_norm": 0.8505379557609558,
+      "learning_rate": 8.510000000000001e-05,
+      "loss": 0.1152,
+      "step": 7660
+    },
+    {
+      "epoch": 0.501799149492967,
+      "grad_norm": 1.0067418813705444,
+      "learning_rate": 8.521111111111112e-05,
+      "loss": 0.1152,
+      "step": 7670
+    },
+    {
+      "epoch": 0.5024533856722276,
+      "grad_norm": 0.9401522278785706,
+      "learning_rate": 8.532222222222223e-05,
+      "loss": 0.1306,
+      "step": 7680
+    },
+    {
+      "epoch": 0.5031076218514884,
+      "grad_norm": 0.9951933026313782,
+      "learning_rate": 8.543333333333333e-05,
+      "loss": 0.1126,
+      "step": 7690
+    },
+    {
+      "epoch": 0.5037618580307491,
+      "grad_norm": 1.0896522998809814,
+      "learning_rate": 8.554444444444444e-05,
+      "loss": 0.1158,
+      "step": 7700
+    },
+    {
+      "epoch": 0.5044160942100098,
+      "grad_norm": 0.9648086428642273,
+      "learning_rate": 8.565555555555556e-05,
+      "loss": 0.1115,
+      "step": 7710
+    },
+    {
+      "epoch": 0.5050703303892705,
+      "grad_norm": 1.1799818277359009,
+      "learning_rate": 8.576666666666667e-05,
+      "loss": 0.1303,
+      "step": 7720
+    },
+    {
+      "epoch": 0.5057245665685313,
+      "grad_norm": 1.237389326095581,
+      "learning_rate": 8.587777777777778e-05,
+      "loss": 0.1238,
+      "step": 7730
+    },
+    {
+      "epoch": 0.5063788027477919,
+      "grad_norm": 0.7409398555755615,
+      "learning_rate": 8.598888888888889e-05,
+      "loss": 0.1184,
+      "step": 7740
+    },
+    {
+      "epoch": 0.5070330389270526,
+      "grad_norm": 0.9186582565307617,
+      "learning_rate": 8.61e-05,
+      "loss": 0.1135,
+      "step": 7750
+    },
+    {
+      "epoch": 0.5076872751063134,
+      "grad_norm": 1.130561351776123,
+      "learning_rate": 8.621111111111112e-05,
+      "loss": 0.1213,
+      "step": 7760
+    },
+    {
+      "epoch": 0.5083415112855741,
+      "grad_norm": 1.1196184158325195,
+      "learning_rate": 8.632222222222222e-05,
+      "loss": 0.1171,
+      "step": 7770
+    },
+    {
+      "epoch": 0.5089957474648348,
+      "grad_norm": 0.9427690505981445,
+      "learning_rate": 8.643333333333334e-05,
+      "loss": 0.1145,
+      "step": 7780
+    },
+    {
+      "epoch": 0.5096499836440955,
+      "grad_norm": 0.9436381459236145,
+      "learning_rate": 8.654444444444445e-05,
+      "loss": 0.113,
+      "step": 7790
+    },
+    {
+      "epoch": 0.5103042198233563,
+      "grad_norm": 1.0007132291793823,
+      "learning_rate": 8.665555555555556e-05,
+      "loss": 0.1174,
+      "step": 7800
+    },
+    {
+      "epoch": 0.5109584560026169,
+      "grad_norm": 1.1906782388687134,
+      "learning_rate": 8.676666666666667e-05,
+      "loss": 0.1363,
+      "step": 7810
+    },
+    {
+      "epoch": 0.5116126921818777,
+      "grad_norm": 0.9627518653869629,
+      "learning_rate": 8.687777777777779e-05,
+      "loss": 0.1123,
+      "step": 7820
+    },
+    {
+      "epoch": 0.5122669283611384,
+      "grad_norm": 1.0318536758422852,
+      "learning_rate": 8.69888888888889e-05,
+      "loss": 0.1138,
+      "step": 7830
+    },
+    {
+      "epoch": 0.512921164540399,
+      "grad_norm": 0.9794100522994995,
+      "learning_rate": 8.71e-05,
+      "loss": 0.1184,
+      "step": 7840
+    },
+    {
+      "epoch": 0.5135754007196598,
+      "grad_norm": 1.2829060554504395,
+      "learning_rate": 8.72111111111111e-05,
+      "loss": 0.1206,
+      "step": 7850
+    },
+    {
+      "epoch": 0.5142296368989205,
+      "grad_norm": 0.9050964117050171,
+      "learning_rate": 8.732222222222223e-05,
+      "loss": 0.1117,
+      "step": 7860
+    },
+    {
+      "epoch": 0.5148838730781812,
+      "grad_norm": 1.2302874326705933,
+      "learning_rate": 8.743333333333334e-05,
+      "loss": 0.1118,
+      "step": 7870
+    },
+    {
+      "epoch": 0.5155381092574419,
+      "grad_norm": 1.3270689249038696,
+      "learning_rate": 8.754444444444445e-05,
+      "loss": 0.1188,
+      "step": 7880
+    },
+    {
+      "epoch": 0.5161923454367027,
+      "grad_norm": 1.0173133611679077,
+      "learning_rate": 8.765555555555555e-05,
+      "loss": 0.1187,
+      "step": 7890
+    },
+    {
+      "epoch": 0.5168465816159633,
+      "grad_norm": 1.0799261331558228,
+      "learning_rate": 8.776666666666668e-05,
+      "loss": 0.1172,
+      "step": 7900
+    },
+    {
+      "epoch": 0.517500817795224,
+      "grad_norm": 1.013592004776001,
+      "learning_rate": 8.787777777777778e-05,
+      "loss": 0.1156,
+      "step": 7910
+    },
+    {
+      "epoch": 0.5181550539744848,
+      "grad_norm": 1.0565035343170166,
+      "learning_rate": 8.79888888888889e-05,
+      "loss": 0.1271,
+      "step": 7920
+    },
+    {
+      "epoch": 0.5188092901537456,
+      "grad_norm": 0.8640610575675964,
+      "learning_rate": 8.81e-05,
+      "loss": 0.1136,
+      "step": 7930
+    },
+    {
+      "epoch": 0.5194635263330062,
+      "grad_norm": 0.942816972732544,
+      "learning_rate": 8.821111111111112e-05,
+      "loss": 0.114,
+      "step": 7940
+    },
+    {
+      "epoch": 0.5201177625122669,
+      "grad_norm": 1.1157722473144531,
+      "learning_rate": 8.832222222222223e-05,
+      "loss": 0.1148,
+      "step": 7950
+    },
+    {
+      "epoch": 0.5207719986915277,
+      "grad_norm": 1.1118143796920776,
+      "learning_rate": 8.843333333333333e-05,
+      "loss": 0.1162,
+      "step": 7960
+    },
+    {
+      "epoch": 0.5214262348707883,
+      "grad_norm": 1.1708115339279175,
+      "learning_rate": 8.854444444444445e-05,
+      "loss": 0.1273,
+      "step": 7970
+    },
+    {
+      "epoch": 0.5220804710500491,
+      "grad_norm": 1.163822054862976,
+      "learning_rate": 8.865555555555556e-05,
+      "loss": 0.1164,
+      "step": 7980
+    },
+    {
+      "epoch": 0.5227347072293098,
+      "grad_norm": 0.847159206867218,
+      "learning_rate": 8.876666666666668e-05,
+      "loss": 0.1161,
+      "step": 7990
+    },
+    {
+      "epoch": 0.5233889434085705,
+      "grad_norm": 0.9535149335861206,
+      "learning_rate": 8.887777777777778e-05,
+      "loss": 0.1193,
+      "step": 8000
+    },
+    {
+      "epoch": 0.5240431795878312,
+      "grad_norm": 0.9705567359924316,
+      "learning_rate": 8.898888888888888e-05,
+      "loss": 0.1217,
+      "step": 8010
+    },
+    {
+      "epoch": 0.524697415767092,
+      "grad_norm": 1.0848075151443481,
+      "learning_rate": 8.910000000000001e-05,
+      "loss": 0.1274,
+      "step": 8020
+    },
+    {
+      "epoch": 0.5253516519463526,
+      "grad_norm": 1.0525822639465332,
+      "learning_rate": 8.921111111111111e-05,
+      "loss": 0.1471,
+      "step": 8030
+    },
+    {
+      "epoch": 0.5260058881256133,
+      "grad_norm": 1.1972851753234863,
+      "learning_rate": 8.932222222222223e-05,
+      "loss": 0.1224,
+      "step": 8040
+    },
+    {
+      "epoch": 0.5266601243048741,
+      "grad_norm": 1.103018045425415,
+      "learning_rate": 8.943333333333333e-05,
+      "loss": 0.1312,
+      "step": 8050
+    },
+    {
+      "epoch": 0.5273143604841348,
+      "grad_norm": 1.3622784614562988,
+      "learning_rate": 8.954444444444446e-05,
+      "loss": 0.12,
+      "step": 8060
+    },
+    {
+      "epoch": 0.5279685966633955,
+      "grad_norm": 1.1434608697891235,
+      "learning_rate": 8.965555555555556e-05,
+      "loss": 0.1083,
+      "step": 8070
+    },
+    {
+      "epoch": 0.5286228328426562,
+      "grad_norm": 1.039299726486206,
+      "learning_rate": 8.976666666666666e-05,
+      "loss": 0.1276,
+      "step": 8080
+    },
+    {
+      "epoch": 0.529277069021917,
+      "grad_norm": 1.0581910610198975,
+      "learning_rate": 8.987777777777778e-05,
+      "loss": 0.1217,
+      "step": 8090
+    },
+    {
+      "epoch": 0.5299313052011776,
+      "grad_norm": 1.0329546928405762,
+      "learning_rate": 8.998888888888889e-05,
+      "loss": 0.1211,
+      "step": 8100
+    },
+    {
+      "epoch": 0.5305855413804383,
+      "grad_norm": 1.1167445182800293,
+      "learning_rate": 9.010000000000001e-05,
+      "loss": 0.1129,
+      "step": 8110
+    },
+    {
+      "epoch": 0.5312397775596991,
+      "grad_norm": 0.8185030817985535,
+      "learning_rate": 9.021111111111111e-05,
+      "loss": 0.1084,
+      "step": 8120
+    },
+    {
+      "epoch": 0.5318940137389597,
+      "grad_norm": 0.9566468000411987,
+      "learning_rate": 9.032222222222223e-05,
+      "loss": 0.117,
+      "step": 8130
+    },
+    {
+      "epoch": 0.5325482499182205,
+      "grad_norm": 1.1805304288864136,
+      "learning_rate": 9.043333333333334e-05,
+      "loss": 0.1243,
+      "step": 8140
+    },
+    {
+      "epoch": 0.5332024860974812,
+      "grad_norm": 1.2267868518829346,
+      "learning_rate": 9.054444444444446e-05,
+      "loss": 0.1197,
+      "step": 8150
+    },
+    {
+      "epoch": 0.5338567222767419,
+      "grad_norm": 1.0790307521820068,
+      "learning_rate": 9.065555555555556e-05,
+      "loss": 0.114,
+      "step": 8160
+    },
+    {
+      "epoch": 0.5345109584560026,
+      "grad_norm": 0.9891142249107361,
+      "learning_rate": 9.076666666666667e-05,
+      "loss": 0.1138,
+      "step": 8170
+    },
+    {
+      "epoch": 0.5351651946352634,
+      "grad_norm": 0.8925236463546753,
+      "learning_rate": 9.087777777777779e-05,
+      "loss": 0.1327,
+      "step": 8180
+    },
+    {
+      "epoch": 0.535819430814524,
+      "grad_norm": 1.0021766424179077,
+      "learning_rate": 9.098888888888889e-05,
+      "loss": 0.1213,
+      "step": 8190
+    },
+    {
+      "epoch": 0.5364736669937847,
+      "grad_norm": 0.9427801966667175,
+      "learning_rate": 9.11e-05,
+      "loss": 0.1073,
+      "step": 8200
+    },
+    {
+      "epoch": 0.5371279031730455,
+      "grad_norm": 1.0449068546295166,
+      "learning_rate": 9.121111111111112e-05,
+      "loss": 0.1078,
+      "step": 8210
+    },
+    {
+      "epoch": 0.5377821393523062,
+      "grad_norm": 1.049320936203003,
+      "learning_rate": 9.132222222222224e-05,
+      "loss": 0.1147,
+      "step": 8220
+    },
+    {
+      "epoch": 0.5384363755315669,
+      "grad_norm": 1.2729623317718506,
+      "learning_rate": 9.143333333333334e-05,
+      "loss": 0.1382,
+      "step": 8230
+    },
+    {
+      "epoch": 0.5390906117108276,
+      "grad_norm": 0.9762771725654602,
+      "learning_rate": 9.154444444444444e-05,
+      "loss": 0.1162,
+      "step": 8240
+    },
+    {
+      "epoch": 0.5397448478900884,
+      "grad_norm": 0.8867229223251343,
+      "learning_rate": 9.165555555555555e-05,
+      "loss": 0.1203,
+      "step": 8250
+    },
+    {
+      "epoch": 0.540399084069349,
+      "grad_norm": 0.9900445938110352,
+      "learning_rate": 9.176666666666667e-05,
+      "loss": 0.131,
+      "step": 8260
+    },
+    {
+      "epoch": 0.5410533202486097,
+      "grad_norm": 1.091282844543457,
+      "learning_rate": 9.187777777777779e-05,
+      "loss": 0.1243,
+      "step": 8270
+    },
+    {
+      "epoch": 0.5417075564278705,
+      "grad_norm": 0.931505024433136,
+      "learning_rate": 9.198888888888889e-05,
+      "loss": 0.1051,
+      "step": 8280
+    },
+    {
+      "epoch": 0.5423617926071311,
+      "grad_norm": 0.9861153960227966,
+      "learning_rate": 9.21e-05,
+      "loss": 0.1134,
+      "step": 8290
+    },
+    {
+      "epoch": 0.5430160287863919,
+      "grad_norm": 0.8883926272392273,
+      "learning_rate": 9.221111111111112e-05,
+      "loss": 0.1148,
+      "step": 8300
+    },
+    {
+      "epoch": 0.5436702649656526,
+      "grad_norm": 0.9315553903579712,
+      "learning_rate": 9.232222222222223e-05,
+      "loss": 0.1204,
+      "step": 8310
+    },
+    {
+      "epoch": 0.5443245011449133,
+      "grad_norm": 0.8906738758087158,
+      "learning_rate": 9.243333333333333e-05,
+      "loss": 0.1149,
+      "step": 8320
+    },
+    {
+      "epoch": 0.544978737324174,
+      "grad_norm": 1.268044114112854,
+      "learning_rate": 9.254444444444445e-05,
+      "loss": 0.1198,
+      "step": 8330
+    },
+    {
+      "epoch": 0.5456329735034348,
+      "grad_norm": 0.9080297946929932,
+      "learning_rate": 9.265555555555557e-05,
+      "loss": 0.1209,
+      "step": 8340
+    },
+    {
+      "epoch": 0.5462872096826955,
+      "grad_norm": 0.8843944072723389,
+      "learning_rate": 9.276666666666667e-05,
+      "loss": 0.1144,
+      "step": 8350
+    },
+    {
+      "epoch": 0.5469414458619561,
+      "grad_norm": 1.117506980895996,
+      "learning_rate": 9.287777777777778e-05,
+      "loss": 0.1256,
+      "step": 8360
+    },
+    {
+      "epoch": 0.5475956820412169,
+      "grad_norm": 0.9390103816986084,
+      "learning_rate": 9.29888888888889e-05,
+      "loss": 0.1206,
+      "step": 8370
+    },
+    {
+      "epoch": 0.5482499182204776,
+      "grad_norm": 1.0785554647445679,
+      "learning_rate": 9.310000000000001e-05,
+      "loss": 0.1265,
+      "step": 8380
+    },
+    {
+      "epoch": 0.5489041543997383,
+      "grad_norm": 1.078389048576355,
+      "learning_rate": 9.321111111111112e-05,
+      "loss": 0.1207,
+      "step": 8390
+    },
+    {
+      "epoch": 0.549558390578999,
+      "grad_norm": 0.811019241809845,
+      "learning_rate": 9.332222222222222e-05,
+      "loss": 0.1087,
+      "step": 8400
+    },
+    {
+      "epoch": 0.5502126267582598,
+      "grad_norm": 0.9671462774276733,
+      "learning_rate": 9.343333333333335e-05,
+      "loss": 0.123,
+      "step": 8410
+    },
+    {
+      "epoch": 0.5508668629375204,
+      "grad_norm": 0.9795189499855042,
+      "learning_rate": 9.354444444444445e-05,
+      "loss": 0.1119,
+      "step": 8420
+    },
+    {
+      "epoch": 0.5515210991167812,
+      "grad_norm": 1.1247087717056274,
+      "learning_rate": 9.365555555555556e-05,
+      "loss": 0.129,
+      "step": 8430
+    },
+    {
+      "epoch": 0.5521753352960419,
+      "grad_norm": 1.5826585292816162,
+      "learning_rate": 9.376666666666666e-05,
+      "loss": 0.1226,
+      "step": 8440
+    },
+    {
+      "epoch": 0.5528295714753025,
+      "grad_norm": 1.0107616186141968,
+      "learning_rate": 9.38777777777778e-05,
+      "loss": 0.1237,
+      "step": 8450
+    },
+    {
+      "epoch": 0.5534838076545633,
+      "grad_norm": 0.9465190172195435,
+      "learning_rate": 9.39888888888889e-05,
+      "loss": 0.1169,
+      "step": 8460
+    },
+    {
+      "epoch": 0.554138043833824,
+      "grad_norm": 1.0885943174362183,
+      "learning_rate": 9.41e-05,
+      "loss": 0.1235,
+      "step": 8470
+    },
+    {
+      "epoch": 0.5547922800130847,
+      "grad_norm": 1.2744386196136475,
+      "learning_rate": 9.421111111111111e-05,
+      "loss": 0.1282,
+      "step": 8480
+    },
+    {
+      "epoch": 0.5554465161923454,
+      "grad_norm": 1.011501669883728,
+      "learning_rate": 9.432222222222223e-05,
+      "loss": 0.1157,
+      "step": 8490
+    },
+    {
+      "epoch": 0.5561007523716062,
+      "grad_norm": 0.9409123063087463,
+      "learning_rate": 9.443333333333334e-05,
+      "loss": 0.1171,
+      "step": 8500
+    },
+    {
+      "epoch": 0.5567549885508669,
+      "grad_norm": 0.8566303253173828,
+      "learning_rate": 9.454444444444444e-05,
+      "loss": 0.1134,
+      "step": 8510
+    },
+    {
+      "epoch": 0.5574092247301276,
+      "grad_norm": 1.0621939897537231,
+      "learning_rate": 9.465555555555556e-05,
+      "loss": 0.1213,
+      "step": 8520
+    },
+    {
+      "epoch": 0.5580634609093883,
+      "grad_norm": 0.7857542037963867,
+      "learning_rate": 9.476666666666668e-05,
+      "loss": 0.1287,
+      "step": 8530
+    },
+    {
+      "epoch": 0.558717697088649,
+      "grad_norm": 1.0616250038146973,
+      "learning_rate": 9.487777777777779e-05,
+      "loss": 0.1235,
+      "step": 8540
+    },
+    {
+      "epoch": 0.5593719332679097,
+      "grad_norm": 1.1893281936645508,
+      "learning_rate": 9.498888888888889e-05,
+      "loss": 0.1181,
+      "step": 8550
+    },
+    {
+      "epoch": 0.5600261694471704,
+      "grad_norm": 1.0680195093154907,
+      "learning_rate": 9.51e-05,
+      "loss": 0.1144,
+      "step": 8560
+    },
+    {
+      "epoch": 0.5606804056264312,
+      "grad_norm": 1.200814962387085,
+      "learning_rate": 9.521111111111112e-05,
+      "loss": 0.1286,
+      "step": 8570
+    },
+    {
+      "epoch": 0.5613346418056918,
+      "grad_norm": 1.1205694675445557,
+      "learning_rate": 9.532222222222222e-05,
+      "loss": 0.1192,
+      "step": 8580
+    },
+    {
+      "epoch": 0.5619888779849526,
+      "grad_norm": 0.915143609046936,
+      "learning_rate": 9.543333333333334e-05,
+      "loss": 0.1227,
+      "step": 8590
+    },
+    {
+      "epoch": 0.5626431141642133,
+      "grad_norm": 0.9382200241088867,
+      "learning_rate": 9.554444444444444e-05,
+      "loss": 0.1272,
+      "step": 8600
+    },
+    {
+      "epoch": 0.563297350343474,
+      "grad_norm": 0.8184203505516052,
+      "learning_rate": 9.565555555555557e-05,
+      "loss": 0.1188,
+      "step": 8610
+    },
+    {
+      "epoch": 0.5639515865227347,
+      "grad_norm": 0.9451429843902588,
+      "learning_rate": 9.576666666666667e-05,
+      "loss": 0.1126,
+      "step": 8620
+    },
+    {
+      "epoch": 0.5646058227019954,
+      "grad_norm": 0.9278766512870789,
+      "learning_rate": 9.587777777777777e-05,
+      "loss": 0.113,
+      "step": 8630
+    },
+    {
+      "epoch": 0.5652600588812562,
+      "grad_norm": 0.9145293831825256,
+      "learning_rate": 9.598888888888889e-05,
+      "loss": 0.1229,
+      "step": 8640
+    },
+    {
+      "epoch": 0.5659142950605168,
+      "grad_norm": 0.8851084113121033,
+      "learning_rate": 9.61e-05,
+      "loss": 0.1117,
+      "step": 8650
+    },
+    {
+      "epoch": 0.5665685312397776,
+      "grad_norm": 0.9469485878944397,
+      "learning_rate": 9.621111111111112e-05,
+      "loss": 0.1198,
+      "step": 8660
+    },
+    {
+      "epoch": 0.5672227674190383,
+      "grad_norm": 0.9359365105628967,
+      "learning_rate": 9.632222222222222e-05,
+      "loss": 0.1169,
+      "step": 8670
+    },
+    {
+      "epoch": 0.567877003598299,
+      "grad_norm": 1.2340917587280273,
+      "learning_rate": 9.643333333333334e-05,
+      "loss": 0.1279,
+      "step": 8680
+    },
+    {
+      "epoch": 0.5685312397775597,
+      "grad_norm": 0.9072956442832947,
+      "learning_rate": 9.654444444444445e-05,
+      "loss": 0.1193,
+      "step": 8690
+    },
+    {
+      "epoch": 0.5691854759568205,
+      "grad_norm": 0.9480854868888855,
+      "learning_rate": 9.665555555555555e-05,
+      "loss": 0.1143,
+      "step": 8700
+    },
+    {
+      "epoch": 0.5698397121360811,
+      "grad_norm": 0.9658315181732178,
+      "learning_rate": 9.676666666666667e-05,
+      "loss": 0.125,
+      "step": 8710
+    },
+    {
+      "epoch": 0.5704939483153418,
+      "grad_norm": 1.0352250337600708,
+      "learning_rate": 9.687777777777778e-05,
+      "loss": 0.1239,
+      "step": 8720
+    },
+    {
+      "epoch": 0.5711481844946026,
+      "grad_norm": 1.0940804481506348,
+      "learning_rate": 9.69888888888889e-05,
+      "loss": 0.1283,
+      "step": 8730
+    },
+    {
+      "epoch": 0.5718024206738632,
+      "grad_norm": 0.9814597964286804,
+      "learning_rate": 9.71e-05,
+      "loss": 0.1155,
+      "step": 8740
+    },
+    {
+      "epoch": 0.572456656853124,
+      "grad_norm": 1.007461667060852,
+      "learning_rate": 9.721111111111112e-05,
+      "loss": 0.122,
+      "step": 8750
+    },
+    {
+      "epoch": 0.5731108930323847,
+      "grad_norm": 0.9953462481498718,
+      "learning_rate": 9.732222222222222e-05,
+      "loss": 0.1322,
+      "step": 8760
+    },
+    {
+      "epoch": 0.5737651292116454,
+      "grad_norm": 0.9330369830131531,
+      "learning_rate": 9.743333333333335e-05,
+      "loss": 0.1129,
+      "step": 8770
+    },
+    {
+      "epoch": 0.5744193653909061,
+      "grad_norm": 0.9782475233078003,
+      "learning_rate": 9.754444444444445e-05,
+      "loss": 0.142,
+      "step": 8780
+    },
+    {
+      "epoch": 0.5750736015701668,
+      "grad_norm": 1.0962268114089966,
+      "learning_rate": 9.765555555555555e-05,
+      "loss": 0.1143,
+      "step": 8790
+    },
+    {
+      "epoch": 0.5757278377494276,
+      "grad_norm": 1.130919337272644,
+      "learning_rate": 9.776666666666667e-05,
+      "loss": 0.1282,
+      "step": 8800
+    },
+    {
+      "epoch": 0.5763820739286882,
+      "grad_norm": 0.9477802515029907,
+      "learning_rate": 9.787777777777778e-05,
+      "loss": 0.12,
+      "step": 8810
+    },
+    {
+      "epoch": 0.577036310107949,
+      "grad_norm": 1.0506799221038818,
+      "learning_rate": 9.79888888888889e-05,
+      "loss": 0.1187,
+      "step": 8820
+    },
+    {
+      "epoch": 0.5776905462872097,
+      "grad_norm": 0.8452275991439819,
+      "learning_rate": 9.81e-05,
+      "loss": 0.1176,
+      "step": 8830
+    },
+    {
+      "epoch": 0.5783447824664704,
+      "grad_norm": 1.2657310962677002,
+      "learning_rate": 9.821111111111111e-05,
+      "loss": 0.1319,
+      "step": 8840
+    },
+    {
+      "epoch": 0.5789990186457311,
+      "grad_norm": 0.8468209505081177,
+      "learning_rate": 9.832222222222223e-05,
+      "loss": 0.1193,
+      "step": 8850
+    },
+    {
+      "epoch": 0.5796532548249919,
+      "grad_norm": 1.0563205480575562,
+      "learning_rate": 9.843333333333333e-05,
+      "loss": 0.127,
+      "step": 8860
+    },
+    {
+      "epoch": 0.5803074910042525,
+      "grad_norm": 0.9747239351272583,
+      "learning_rate": 9.854444444444445e-05,
+      "loss": 0.1217,
+      "step": 8870
+    },
+    {
+      "epoch": 0.5809617271835132,
+      "grad_norm": 1.1098183393478394,
+      "learning_rate": 9.865555555555556e-05,
+      "loss": 0.1313,
+      "step": 8880
+    },
+    {
+      "epoch": 0.581615963362774,
+      "grad_norm": 1.0163518190383911,
+      "learning_rate": 9.876666666666668e-05,
+      "loss": 0.1259,
+      "step": 8890
+    },
+    {
+      "epoch": 0.5822701995420346,
+      "grad_norm": 1.0436562299728394,
+      "learning_rate": 9.887777777777778e-05,
+      "loss": 0.1302,
+      "step": 8900
+    },
+    {
+      "epoch": 0.5829244357212954,
+      "grad_norm": 0.9520934820175171,
+      "learning_rate": 9.89888888888889e-05,
+      "loss": 0.1343,
+      "step": 8910
+    },
+    {
+      "epoch": 0.5835786719005561,
+      "grad_norm": 1.1144458055496216,
+      "learning_rate": 9.910000000000001e-05,
+      "loss": 0.1188,
+      "step": 8920
+    },
+    {
+      "epoch": 0.5842329080798168,
+      "grad_norm": 0.8635526299476624,
+      "learning_rate": 9.921111111111113e-05,
+      "loss": 0.1263,
+      "step": 8930
+    },
+    {
+      "epoch": 0.5848871442590775,
+      "grad_norm": 0.9793311357498169,
+      "learning_rate": 9.932222222222223e-05,
+      "loss": 0.1183,
+      "step": 8940
+    },
+    {
+      "epoch": 0.5855413804383383,
+      "grad_norm": 1.0002834796905518,
+      "learning_rate": 9.943333333333333e-05,
+      "loss": 0.1212,
+      "step": 8950
+    },
+    {
+      "epoch": 0.586195616617599,
+      "grad_norm": 1.0948991775512695,
+      "learning_rate": 9.954444444444446e-05,
+      "loss": 0.1257,
+      "step": 8960
+    },
+    {
+      "epoch": 0.5868498527968596,
+      "grad_norm": 1.0232096910476685,
+      "learning_rate": 9.965555555555556e-05,
+      "loss": 0.1238,
+      "step": 8970
+    },
+    {
+      "epoch": 0.5875040889761204,
+      "grad_norm": 0.8366304039955139,
+      "learning_rate": 9.976666666666667e-05,
+      "loss": 0.1203,
+      "step": 8980
+    },
+    {
+      "epoch": 0.5881583251553811,
+      "grad_norm": 1.152336835861206,
+      "learning_rate": 9.987777777777778e-05,
+      "loss": 0.1273,
+      "step": 8990
+    },
+    {
+      "epoch": 0.5888125613346418,
+      "grad_norm": 0.9613927006721497,
+      "learning_rate": 9.998888888888889e-05,
+      "loss": 0.1119,
+      "step": 9000
+    },
+    {
+      "epoch": 0.5894667975139025,
+      "grad_norm": 0.9294191598892212,
+      "learning_rate": 9.99999993165094e-05,
+      "loss": 0.1197,
+      "step": 9010
+    },
+    {
+      "epoch": 0.5901210336931633,
+      "grad_norm": 1.0190068483352661,
+      "learning_rate": 9.999999695382584e-05,
+      "loss": 0.119,
+      "step": 9020
+    },
+    {
+      "epoch": 0.5907752698724239,
+      "grad_norm": 0.9870291948318481,
+      "learning_rate": 9.999999290351126e-05,
+      "loss": 0.1316,
+      "step": 9030
+    },
+    {
+      "epoch": 0.5914295060516847,
+      "grad_norm": 1.160510778427124,
+      "learning_rate": 9.999998716556578e-05,
+      "loss": 0.1278,
+      "step": 9040
+    },
+    {
+      "epoch": 0.5920837422309454,
+      "grad_norm": 0.9017443656921387,
+      "learning_rate": 9.99999797399896e-05,
+      "loss": 0.1154,
+      "step": 9050
+    },
+    {
+      "epoch": 0.592737978410206,
+      "grad_norm": 1.122379183769226,
+      "learning_rate": 9.999997062678298e-05,
+      "loss": 0.1227,
+      "step": 9060
+    },
+    {
+      "epoch": 0.5933922145894668,
+      "grad_norm": 0.9526849985122681,
+      "learning_rate": 9.999995982594624e-05,
+      "loss": 0.1196,
+      "step": 9070
+    },
+    {
+      "epoch": 0.5940464507687275,
+      "grad_norm": 1.065383791923523,
+      "learning_rate": 9.999994733747969e-05,
+      "loss": 0.1252,
+      "step": 9080
+    },
+    {
+      "epoch": 0.5947006869479883,
+      "grad_norm": 1.0271326303482056,
+      "learning_rate": 9.99999331613838e-05,
+      "loss": 0.1256,
+      "step": 9090
+    },
+    {
+      "epoch": 0.5953549231272489,
+      "grad_norm": 0.8906413316726685,
+      "learning_rate": 9.999991729765906e-05,
+      "loss": 0.1383,
+      "step": 9100
+    },
+    {
+      "epoch": 0.5960091593065097,
+      "grad_norm": 1.0225988626480103,
+      "learning_rate": 9.999989974630596e-05,
+      "loss": 0.1247,
+      "step": 9110
+    },
+    {
+      "epoch": 0.5966633954857704,
+      "grad_norm": 0.9085574746131897,
+      "learning_rate": 9.999988050732512e-05,
+      "loss": 0.1344,
+      "step": 9120
+    },
+    {
+      "epoch": 0.597317631665031,
+      "grad_norm": 0.9044840931892395,
+      "learning_rate": 9.999985958071718e-05,
+      "loss": 0.1308,
+      "step": 9130
+    },
+    {
+      "epoch": 0.5979718678442918,
+      "grad_norm": 0.9104439616203308,
+      "learning_rate": 9.999983696648286e-05,
+      "loss": 0.1253,
+      "step": 9140
+    },
+    {
+      "epoch": 0.5986261040235525,
+      "grad_norm": 0.998406708240509,
+      "learning_rate": 9.99998126646229e-05,
+      "loss": 0.1134,
+      "step": 9150
+    },
+    {
+      "epoch": 0.5992803402028132,
+      "grad_norm": 1.0714468955993652,
+      "learning_rate": 9.999978667513815e-05,
+      "loss": 0.1238,
+      "step": 9160
+    },
+    {
+      "epoch": 0.5999345763820739,
+      "grad_norm": 0.9690518975257874,
+      "learning_rate": 9.999975899802944e-05,
+      "loss": 0.1192,
+      "step": 9170
+    },
+    {
+      "epoch": 0.6005888125613347,
+      "grad_norm": 0.9514327645301819,
+      "learning_rate": 9.999972963329775e-05,
+      "loss": 0.1211,
+      "step": 9180
+    },
+    {
+      "epoch": 0.6012430487405953,
+      "grad_norm": 0.9661425948143005,
+      "learning_rate": 9.999969858094407e-05,
+      "loss": 0.1241,
+      "step": 9190
+    },
+    {
+      "epoch": 0.6018972849198561,
+      "grad_norm": 1.0832891464233398,
+      "learning_rate": 9.999966584096941e-05,
+      "loss": 0.1271,
+      "step": 9200
+    },
+    {
+      "epoch": 0.6025515210991168,
+      "grad_norm": 1.053916573524475,
+      "learning_rate": 9.999963141337492e-05,
+      "loss": 0.1342,
+      "step": 9210
+    },
+    {
+      "epoch": 0.6032057572783774,
+      "grad_norm": 0.9167242646217346,
+      "learning_rate": 9.999959529816173e-05,
+      "loss": 0.1221,
+      "step": 9220
+    },
+    {
+      "epoch": 0.6038599934576382,
+      "grad_norm": 0.8978278636932373,
+      "learning_rate": 9.999955749533107e-05,
+      "loss": 0.1217,
+      "step": 9230
+    },
+    {
+      "epoch": 0.6045142296368989,
+      "grad_norm": 0.9800412654876709,
+      "learning_rate": 9.999951800488422e-05,
+      "loss": 0.1178,
+      "step": 9240
+    },
+    {
+      "epoch": 0.6051684658161597,
+      "grad_norm": 0.9557292461395264,
+      "learning_rate": 9.999947682682251e-05,
+      "loss": 0.1342,
+      "step": 9250
+    },
+    {
+      "epoch": 0.6058227019954203,
+      "grad_norm": 0.9261611104011536,
+      "learning_rate": 9.999943396114732e-05,
+      "loss": 0.1208,
+      "step": 9260
+    },
+    {
+      "epoch": 0.6064769381746811,
+      "grad_norm": 0.8848803639411926,
+      "learning_rate": 9.999938940786011e-05,
+      "loss": 0.1287,
+      "step": 9270
+    },
+    {
+      "epoch": 0.6071311743539418,
+      "grad_norm": 0.818735659122467,
+      "learning_rate": 9.999934316696238e-05,
+      "loss": 0.1135,
+      "step": 9280
+    },
+    {
+      "epoch": 0.6077854105332025,
+      "grad_norm": 1.051778793334961,
+      "learning_rate": 9.99992952384557e-05,
+      "loss": 0.1175,
+      "step": 9290
+    },
+    {
+      "epoch": 0.6084396467124632,
+      "grad_norm": 0.9308475255966187,
+      "learning_rate": 9.999924562234167e-05,
+      "loss": 0.1114,
+      "step": 9300
+    },
+    {
+      "epoch": 0.609093882891724,
+      "grad_norm": 1.1484242677688599,
+      "learning_rate": 9.999919431862197e-05,
+      "loss": 0.1254,
+      "step": 9310
+    },
+    {
+      "epoch": 0.6097481190709846,
+      "grad_norm": 0.9219973683357239,
+      "learning_rate": 9.999914132729832e-05,
+      "loss": 0.1187,
+      "step": 9320
+    },
+    {
+      "epoch": 0.6104023552502453,
+      "grad_norm": 0.9428659677505493,
+      "learning_rate": 9.999908664837255e-05,
+      "loss": 0.1182,
+      "step": 9330
+    },
+    {
+      "epoch": 0.6110565914295061,
+      "grad_norm": 0.8808643221855164,
+      "learning_rate": 9.999903028184646e-05,
+      "loss": 0.1195,
+      "step": 9340
+    },
+    {
+      "epoch": 0.6117108276087667,
+      "grad_norm": 1.1182430982589722,
+      "learning_rate": 9.999897222772198e-05,
+      "loss": 0.1329,
+      "step": 9350
+    },
+    {
+      "epoch": 0.6123650637880275,
+      "grad_norm": 1.0213252305984497,
+      "learning_rate": 9.999891248600107e-05,
+      "loss": 0.1229,
+      "step": 9360
+    },
+    {
+      "epoch": 0.6130192999672882,
+      "grad_norm": 1.048762321472168,
+      "learning_rate": 9.999885105668571e-05,
+      "loss": 0.1213,
+      "step": 9370
+    },
+    {
+      "epoch": 0.613673536146549,
+      "grad_norm": 0.9131613969802856,
+      "learning_rate": 9.999878793977801e-05,
+      "loss": 0.1153,
+      "step": 9380
+    },
+    {
+      "epoch": 0.6143277723258096,
+      "grad_norm": 1.397139310836792,
+      "learning_rate": 9.999872313528009e-05,
+      "loss": 0.1203,
+      "step": 9390
+    },
+    {
+      "epoch": 0.6149820085050703,
+      "grad_norm": 1.120139479637146,
+      "learning_rate": 9.999865664319414e-05,
+      "loss": 0.1216,
+      "step": 9400
+    },
+    {
+      "epoch": 0.6156362446843311,
+      "grad_norm": 1.1770591735839844,
+      "learning_rate": 9.999858846352242e-05,
+      "loss": 0.1135,
+      "step": 9410
+    },
+    {
+      "epoch": 0.6162904808635917,
+      "grad_norm": 1.0336472988128662,
+      "learning_rate": 9.99985185962672e-05,
+      "loss": 0.1197,
+      "step": 9420
+    },
+    {
+      "epoch": 0.6169447170428525,
+      "grad_norm": 1.1690295934677124,
+      "learning_rate": 9.999844704143084e-05,
+      "loss": 0.1329,
+      "step": 9430
+    },
+    {
+      "epoch": 0.6175989532221132,
+      "grad_norm": 0.9790360331535339,
+      "learning_rate": 9.999837379901578e-05,
+      "loss": 0.114,
+      "step": 9440
+    },
+    {
+      "epoch": 0.6182531894013739,
+      "grad_norm": 0.918538510799408,
+      "learning_rate": 9.999829886902446e-05,
+      "loss": 0.1204,
+      "step": 9450
+    },
+    {
+      "epoch": 0.6189074255806346,
+      "grad_norm": 0.986890435218811,
+      "learning_rate": 9.999822225145945e-05,
+      "loss": 0.1291,
+      "step": 9460
+    },
+    {
+      "epoch": 0.6195616617598954,
+      "grad_norm": 0.8274875283241272,
+      "learning_rate": 9.99981439463233e-05,
+      "loss": 0.126,
+      "step": 9470
+    },
+    {
+      "epoch": 0.620215897939156,
+      "grad_norm": 0.8591904640197754,
+      "learning_rate": 9.999806395361867e-05,
+      "loss": 0.1184,
+      "step": 9480
+    },
+    {
+      "epoch": 0.6208701341184167,
+      "grad_norm": 0.9025493860244751,
+      "learning_rate": 9.999798227334827e-05,
+      "loss": 0.1169,
+      "step": 9490
+    },
+    {
+      "epoch": 0.6215243702976775,
+      "grad_norm": 0.9297366142272949,
+      "learning_rate": 9.999789890551483e-05,
+      "loss": 0.1313,
+      "step": 9500
+    },
+    {
+      "epoch": 0.6221786064769381,
+      "grad_norm": 0.843330979347229,
+      "learning_rate": 9.999781385012116e-05,
+      "loss": 0.1178,
+      "step": 9510
+    },
+    {
+      "epoch": 0.6228328426561989,
+      "grad_norm": 0.8769571781158447,
+      "learning_rate": 9.999772710717018e-05,
+      "loss": 0.1191,
+      "step": 9520
+    },
+    {
+      "epoch": 0.6234870788354596,
+      "grad_norm": 0.996464192867279,
+      "learning_rate": 9.999763867666479e-05,
+      "loss": 0.1184,
+      "step": 9530
+    },
+    {
+      "epoch": 0.6241413150147204,
+      "grad_norm": 0.9058352112770081,
+      "learning_rate": 9.999754855860795e-05,
+      "loss": 0.1246,
+      "step": 9540
+    },
+    {
+      "epoch": 0.624795551193981,
+      "grad_norm": 0.8626279830932617,
+      "learning_rate": 9.999745675300271e-05,
+      "loss": 0.1175,
+      "step": 9550
+    },
+    {
+      "epoch": 0.6254497873732418,
+      "grad_norm": 0.8973715901374817,
+      "learning_rate": 9.999736325985221e-05,
+      "loss": 0.1085,
+      "step": 9560
+    },
+    {
+      "epoch": 0.6261040235525025,
+      "grad_norm": 0.9920212030410767,
+      "learning_rate": 9.999726807915956e-05,
+      "loss": 0.1203,
+      "step": 9570
+    },
+    {
+      "epoch": 0.6267582597317631,
+      "grad_norm": 0.8262133598327637,
+      "learning_rate": 9.999717121092802e-05,
+      "loss": 0.1214,
+      "step": 9580
+    },
+    {
+      "epoch": 0.6274124959110239,
+      "grad_norm": 0.7645416259765625,
+      "learning_rate": 9.999707265516079e-05,
+      "loss": 0.1211,
+      "step": 9590
+    },
+    {
+      "epoch": 0.6280667320902846,
+      "grad_norm": 1.0003942251205444,
+      "learning_rate": 9.999697241186126e-05,
+      "loss": 0.1167,
+      "step": 9600
+    },
+    {
+      "epoch": 0.6287209682695453,
+      "grad_norm": 1.0682168006896973,
+      "learning_rate": 9.999687048103278e-05,
+      "loss": 0.1182,
+      "step": 9610
+    },
+    {
+      "epoch": 0.629375204448806,
+      "grad_norm": 0.7786453366279602,
+      "learning_rate": 9.999676686267881e-05,
+      "loss": 0.1241,
+      "step": 9620
+    },
+    {
+      "epoch": 0.6300294406280668,
+      "grad_norm": 0.9212088584899902,
+      "learning_rate": 9.999666155680281e-05,
+      "loss": 0.1219,
+      "step": 9630
+    },
+    {
+      "epoch": 0.6306836768073274,
+      "grad_norm": 0.8398849964141846,
+      "learning_rate": 9.999655456340839e-05,
+      "loss": 0.1217,
+      "step": 9640
+    },
+    {
+      "epoch": 0.6313379129865881,
+      "grad_norm": 0.7940874099731445,
+      "learning_rate": 9.999644588249912e-05,
+      "loss": 0.1156,
+      "step": 9650
+    },
+    {
+      "epoch": 0.6319921491658489,
+      "grad_norm": 0.9876879453659058,
+      "learning_rate": 9.999633551407867e-05,
+      "loss": 0.1307,
+      "step": 9660
+    },
+    {
+      "epoch": 0.6326463853451095,
+      "grad_norm": 0.9693999886512756,
+      "learning_rate": 9.999622345815081e-05,
+      "loss": 0.1232,
+      "step": 9670
+    },
+    {
+      "epoch": 0.6333006215243703,
+      "grad_norm": 0.8915924429893494,
+      "learning_rate": 9.999610971471925e-05,
+      "loss": 0.1255,
+      "step": 9680
+    },
+    {
+      "epoch": 0.633954857703631,
+      "grad_norm": 0.9316652417182922,
+      "learning_rate": 9.999599428378789e-05,
+      "loss": 0.1258,
+      "step": 9690
+    },
+    {
+      "epoch": 0.6346090938828918,
+      "grad_norm": 0.8917170763015747,
+      "learning_rate": 9.99958771653606e-05,
+      "loss": 0.1246,
+      "step": 9700
+    },
+    {
+      "epoch": 0.6352633300621524,
+      "grad_norm": 1.0505290031433105,
+      "learning_rate": 9.999575835944133e-05,
+      "loss": 0.1109,
+      "step": 9710
+    },
+    {
+      "epoch": 0.6359175662414132,
+      "grad_norm": 1.0103509426116943,
+      "learning_rate": 9.999563786603412e-05,
+      "loss": 0.125,
+      "step": 9720
+    },
+    {
+      "epoch": 0.6365718024206739,
+      "grad_norm": 0.8918554186820984,
+      "learning_rate": 9.999551568514298e-05,
+      "loss": 0.1281,
+      "step": 9730
+    },
+    {
+      "epoch": 0.6372260385999345,
+      "grad_norm": 0.8537850975990295,
+      "learning_rate": 9.999539181677208e-05,
+      "loss": 0.1301,
+      "step": 9740
+    },
+    {
+      "epoch": 0.6378802747791953,
+      "grad_norm": 1.0393998622894287,
+      "learning_rate": 9.99952662609256e-05,
+      "loss": 0.1183,
+      "step": 9750
+    },
+    {
+      "epoch": 0.638534510958456,
+      "grad_norm": 0.9914324283599854,
+      "learning_rate": 9.999513901760775e-05,
+      "loss": 0.1348,
+      "step": 9760
+    },
+    {
+      "epoch": 0.6391887471377167,
+      "grad_norm": 1.1659291982650757,
+      "learning_rate": 9.999501008682286e-05,
+      "loss": 0.124,
+      "step": 9770
+    },
+    {
+      "epoch": 0.6398429833169774,
+      "grad_norm": 0.9629908800125122,
+      "learning_rate": 9.999487946857525e-05,
+      "loss": 0.1177,
+      "step": 9780
+    },
+    {
+      "epoch": 0.6404972194962382,
+      "grad_norm": 0.8898019194602966,
+      "learning_rate": 9.999474716286934e-05,
+      "loss": 0.1233,
+      "step": 9790
+    },
+    {
+      "epoch": 0.6411514556754988,
+      "grad_norm": 0.9645368456840515,
+      "learning_rate": 9.99946131697096e-05,
+      "loss": 0.1172,
+      "step": 9800
+    },
+    {
+      "epoch": 0.6418056918547596,
+      "grad_norm": 1.1151118278503418,
+      "learning_rate": 9.999447748910056e-05,
+      "loss": 0.1154,
+      "step": 9810
+    },
+    {
+      "epoch": 0.6424599280340203,
+      "grad_norm": 1.1241856813430786,
+      "learning_rate": 9.999434012104678e-05,
+      "loss": 0.1182,
+      "step": 9820
+    },
+    {
+      "epoch": 0.643114164213281,
+      "grad_norm": 0.883420467376709,
+      "learning_rate": 9.999420106555291e-05,
+      "loss": 0.126,
+      "step": 9830
+    },
+    {
+      "epoch": 0.6437684003925417,
+      "grad_norm": 1.023460865020752,
+      "learning_rate": 9.999406032262362e-05,
+      "loss": 0.1242,
+      "step": 9840
+    },
+    {
+      "epoch": 0.6444226365718024,
+      "grad_norm": 1.0635135173797607,
+      "learning_rate": 9.99939178922637e-05,
+      "loss": 0.1383,
+      "step": 9850
+    },
+    {
+      "epoch": 0.6450768727510632,
+      "grad_norm": 0.738158643245697,
+      "learning_rate": 9.999377377447794e-05,
+      "loss": 0.1199,
+      "step": 9860
+    },
+    {
+      "epoch": 0.6457311089303238,
+      "grad_norm": 1.1745080947875977,
+      "learning_rate": 9.999362796927119e-05,
+      "loss": 0.1239,
+      "step": 9870
+    },
+    {
+      "epoch": 0.6463853451095846,
+      "grad_norm": 1.2103548049926758,
+      "learning_rate": 9.999348047664838e-05,
+      "loss": 0.1281,
+      "step": 9880
+    },
+    {
+      "epoch": 0.6470395812888453,
+      "grad_norm": 0.9273943901062012,
+      "learning_rate": 9.999333129661451e-05,
+      "loss": 0.1274,
+      "step": 9890
+    },
+    {
+      "epoch": 0.647693817468106,
+      "grad_norm": 0.9063096642494202,
+      "learning_rate": 9.999318042917459e-05,
+      "loss": 0.1161,
+      "step": 9900
+    },
+    {
+      "epoch": 0.6483480536473667,
+      "grad_norm": 1.0374435186386108,
+      "learning_rate": 9.999302787433372e-05,
+      "loss": 0.1253,
+      "step": 9910
+    },
+    {
+      "epoch": 0.6490022898266274,
+      "grad_norm": 0.8295223116874695,
+      "learning_rate": 9.999287363209703e-05,
+      "loss": 0.1196,
+      "step": 9920
+    },
+    {
+      "epoch": 0.6496565260058881,
+      "grad_norm": 0.9288245439529419,
+      "learning_rate": 9.999271770246975e-05,
+      "loss": 0.1272,
+      "step": 9930
+    },
+    {
+      "epoch": 0.6503107621851488,
+      "grad_norm": 1.091799259185791,
+      "learning_rate": 9.999256008545714e-05,
+      "loss": 0.1181,
+      "step": 9940
+    },
+    {
+      "epoch": 0.6509649983644096,
+      "grad_norm": 0.8960475325584412,
+      "learning_rate": 9.999240078106452e-05,
+      "loss": 0.1307,
+      "step": 9950
+    },
+    {
+      "epoch": 0.6516192345436702,
+      "grad_norm": 0.9426436424255371,
+      "learning_rate": 9.999223978929727e-05,
+      "loss": 0.116,
+      "step": 9960
+    },
+    {
+      "epoch": 0.652273470722931,
+      "grad_norm": 0.9542483687400818,
+      "learning_rate": 9.999207711016081e-05,
+      "loss": 0.1197,
+      "step": 9970
+    },
+    {
+      "epoch": 0.6529277069021917,
+      "grad_norm": 0.8808081746101379,
+      "learning_rate": 9.999191274366064e-05,
+      "loss": 0.1261,
+      "step": 9980
+    },
+    {
+      "epoch": 0.6535819430814525,
+      "grad_norm": 0.8422145247459412,
+      "learning_rate": 9.99917466898023e-05,
+      "loss": 0.112,
+      "step": 9990
+    },
+    {
+      "epoch": 0.6542361792607131,
+      "grad_norm": 0.956287682056427,
+      "learning_rate": 9.999157894859142e-05,
+      "loss": 0.1356,
+      "step": 10000
+    },
+    {
+      "epoch": 0.6548904154399738,
+      "grad_norm": 0.9550445079803467,
+      "learning_rate": 9.99914095200336e-05,
+      "loss": 0.1277,
+      "step": 10010
+    },
+    {
+      "epoch": 0.6555446516192346,
+      "grad_norm": 0.9725729823112488,
+      "learning_rate": 9.999123840413465e-05,
+      "loss": 0.1257,
+      "step": 10020
+    },
+    {
+      "epoch": 0.6561988877984952,
+      "grad_norm": 0.8097350001335144,
+      "learning_rate": 9.999106560090028e-05,
+      "loss": 0.1199,
+      "step": 10030
+    },
+    {
+      "epoch": 0.656853123977756,
+      "grad_norm": 0.9223546981811523,
+      "learning_rate": 9.999089111033633e-05,
+      "loss": 0.1223,
+      "step": 10040
+    },
+    {
+      "epoch": 0.6575073601570167,
+      "grad_norm": 0.908068060874939,
+      "learning_rate": 9.99907149324487e-05,
+      "loss": 0.1124,
+      "step": 10050
+    },
+    {
+      "epoch": 0.6581615963362774,
+      "grad_norm": 0.9873263835906982,
+      "learning_rate": 9.999053706724335e-05,
+      "loss": 0.1229,
+      "step": 10060
+    },
+    {
+      "epoch": 0.6588158325155381,
+      "grad_norm": 0.870583713054657,
+      "learning_rate": 9.999035751472625e-05,
+      "loss": 0.1113,
+      "step": 10070
+    },
+    {
+      "epoch": 0.6594700686947988,
+      "grad_norm": 0.9837010502815247,
+      "learning_rate": 9.99901762749035e-05,
+      "loss": 0.1281,
+      "step": 10080
+    },
+    {
+      "epoch": 0.6601243048740595,
+      "grad_norm": 1.0344374179840088,
+      "learning_rate": 9.998999334778118e-05,
+      "loss": 0.1355,
+      "step": 10090
+    },
+    {
+      "epoch": 0.6607785410533202,
+      "grad_norm": 1.1033861637115479,
+      "learning_rate": 9.99898087333655e-05,
+      "loss": 0.1213,
+      "step": 10100
+    },
+    {
+      "epoch": 0.661432777232581,
+      "grad_norm": 0.9438520073890686,
+      "learning_rate": 9.998962243166266e-05,
+      "loss": 0.1358,
+      "step": 10110
+    },
+    {
+      "epoch": 0.6620870134118417,
+      "grad_norm": 0.9945731163024902,
+      "learning_rate": 9.998943444267896e-05,
+      "loss": 0.1261,
+      "step": 10120
+    },
+    {
+      "epoch": 0.6627412495911024,
+      "grad_norm": 0.9855345487594604,
+      "learning_rate": 9.998924476642074e-05,
+      "loss": 0.1285,
+      "step": 10130
+    },
+    {
+      "epoch": 0.6633954857703631,
+      "grad_norm": 1.0155123472213745,
+      "learning_rate": 9.998905340289442e-05,
+      "loss": 0.1156,
+      "step": 10140
+    },
+    {
+      "epoch": 0.6640497219496239,
+      "grad_norm": 0.953682541847229,
+      "learning_rate": 9.998886035210643e-05,
+      "loss": 0.1144,
+      "step": 10150
+    },
+    {
+      "epoch": 0.6647039581288845,
+      "grad_norm": 1.0266921520233154,
+      "learning_rate": 9.99886656140633e-05,
+      "loss": 0.1202,
+      "step": 10160
+    },
+    {
+      "epoch": 0.6653581943081452,
+      "grad_norm": 0.8357493281364441,
+      "learning_rate": 9.998846918877162e-05,
+      "loss": 0.1136,
+      "step": 10170
+    },
+    {
+      "epoch": 0.666012430487406,
+      "grad_norm": 0.9822357892990112,
+      "learning_rate": 9.9988271076238e-05,
+      "loss": 0.1327,
+      "step": 10180
+    },
+    {
+      "epoch": 0.6666666666666666,
+      "grad_norm": 0.8407633304595947,
+      "learning_rate": 9.998807127646915e-05,
+      "loss": 0.1299,
+      "step": 10190
+    },
+    {
+      "epoch": 0.6673209028459274,
+      "grad_norm": 0.8804751038551331,
+      "learning_rate": 9.998786978947177e-05,
+      "loss": 0.1204,
+      "step": 10200
+    },
+    {
+      "epoch": 0.6679751390251881,
+      "grad_norm": 1.0064154863357544,
+      "learning_rate": 9.99876666152527e-05,
+      "loss": 0.1181,
+      "step": 10210
+    },
+    {
+      "epoch": 0.6686293752044488,
+      "grad_norm": 1.158360481262207,
+      "learning_rate": 9.998746175381879e-05,
+      "loss": 0.1163,
+      "step": 10220
+    },
+    {
+      "epoch": 0.6692836113837095,
+      "grad_norm": 1.060184359550476,
+      "learning_rate": 9.998725520517693e-05,
+      "loss": 0.1213,
+      "step": 10230
+    },
+    {
+      "epoch": 0.6699378475629703,
+      "grad_norm": 1.123124122619629,
+      "learning_rate": 9.998704696933413e-05,
+      "loss": 0.1184,
+      "step": 10240
+    },
+    {
+      "epoch": 0.6705920837422309,
+      "grad_norm": 0.8490195870399475,
+      "learning_rate": 9.998683704629739e-05,
+      "loss": 0.1236,
+      "step": 10250
+    },
+    {
+      "epoch": 0.6712463199214916,
+      "grad_norm": 0.8913207650184631,
+      "learning_rate": 9.99866254360738e-05,
+      "loss": 0.1195,
+      "step": 10260
+    },
+    {
+      "epoch": 0.6719005561007524,
+      "grad_norm": 0.9286946654319763,
+      "learning_rate": 9.998641213867051e-05,
+      "loss": 0.1168,
+      "step": 10270
+    },
+    {
+      "epoch": 0.6725547922800131,
+      "grad_norm": 0.9744524359703064,
+      "learning_rate": 9.998619715409471e-05,
+      "loss": 0.1134,
+      "step": 10280
+    },
+    {
+      "epoch": 0.6732090284592738,
+      "grad_norm": 0.911208987236023,
+      "learning_rate": 9.998598048235369e-05,
+      "loss": 0.1115,
+      "step": 10290
+    },
+    {
+      "epoch": 0.6738632646385345,
+      "grad_norm": 1.0315773487091064,
+      "learning_rate": 9.99857621234547e-05,
+      "loss": 0.1189,
+      "step": 10300
+    },
+    {
+      "epoch": 0.6745175008177953,
+      "grad_norm": 1.1607714891433716,
+      "learning_rate": 9.998554207740517e-05,
+      "loss": 0.1238,
+      "step": 10310
+    },
+    {
+      "epoch": 0.6751717369970559,
+      "grad_norm": 0.7840601801872253,
+      "learning_rate": 9.99853203442125e-05,
+      "loss": 0.1196,
+      "step": 10320
+    },
+    {
+      "epoch": 0.6758259731763167,
+      "grad_norm": 1.062690019607544,
+      "learning_rate": 9.998509692388416e-05,
+      "loss": 0.1227,
+      "step": 10330
+    },
+    {
+      "epoch": 0.6764802093555774,
+      "grad_norm": 1.104194164276123,
+      "learning_rate": 9.998487181642772e-05,
+      "loss": 0.1231,
+      "step": 10340
+    },
+    {
+      "epoch": 0.677134445534838,
+      "grad_norm": 1.0216797590255737,
+      "learning_rate": 9.998464502185076e-05,
+      "loss": 0.118,
+      "step": 10350
+    },
+    {
+      "epoch": 0.6777886817140988,
+      "grad_norm": 0.9924758076667786,
+      "learning_rate": 9.998441654016095e-05,
+      "loss": 0.1245,
+      "step": 10360
+    },
+    {
+      "epoch": 0.6784429178933595,
+      "grad_norm": 0.9383712410926819,
+      "learning_rate": 9.9984186371366e-05,
+      "loss": 0.116,
+      "step": 10370
+    },
+    {
+      "epoch": 0.6790971540726202,
+      "grad_norm": 0.8061214089393616,
+      "learning_rate": 9.998395451547367e-05,
+      "loss": 0.1172,
+      "step": 10380
+    },
+    {
+      "epoch": 0.6797513902518809,
+      "grad_norm": 0.9712622761726379,
+      "learning_rate": 9.998372097249177e-05,
+      "loss": 0.1206,
+      "step": 10390
+    },
+    {
+      "epoch": 0.6804056264311417,
+      "grad_norm": 0.8583802580833435,
+      "learning_rate": 9.998348574242821e-05,
+      "loss": 0.121,
+      "step": 10400
+    },
+    {
+      "epoch": 0.6810598626104023,
+      "grad_norm": 0.8582515120506287,
+      "learning_rate": 9.99832488252909e-05,
+      "loss": 0.1189,
+      "step": 10410
+    },
+    {
+      "epoch": 0.681714098789663,
+      "grad_norm": 0.9043430685997009,
+      "learning_rate": 9.998301022108789e-05,
+      "loss": 0.1108,
+      "step": 10420
+    },
+    {
+      "epoch": 0.6823683349689238,
+      "grad_norm": 0.9330876469612122,
+      "learning_rate": 9.998276992982717e-05,
+      "loss": 0.1178,
+      "step": 10430
+    },
+    {
+      "epoch": 0.6830225711481845,
+      "grad_norm": 0.9934483170509338,
+      "learning_rate": 9.99825279515169e-05,
+      "loss": 0.1206,
+      "step": 10440
+    },
+    {
+      "epoch": 0.6836768073274452,
+      "grad_norm": 0.8056890964508057,
+      "learning_rate": 9.998228428616523e-05,
+      "loss": 0.1283,
+      "step": 10450
+    },
+    {
+      "epoch": 0.6843310435067059,
+      "grad_norm": 0.9143165349960327,
+      "learning_rate": 9.998203893378037e-05,
+      "loss": 0.1322,
+      "step": 10460
+    },
+    {
+      "epoch": 0.6849852796859667,
+      "grad_norm": 0.9717006683349609,
+      "learning_rate": 9.998179189437062e-05,
+      "loss": 0.1274,
+      "step": 10470
+    },
+    {
+      "epoch": 0.6856395158652273,
+      "grad_norm": 1.0283546447753906,
+      "learning_rate": 9.99815431679443e-05,
+      "loss": 0.1193,
+      "step": 10480
+    },
+    {
+      "epoch": 0.6862937520444881,
+      "grad_norm": 0.9242172241210938,
+      "learning_rate": 9.998129275450983e-05,
+      "loss": 0.1284,
+      "step": 10490
+    },
+    {
+      "epoch": 0.6869479882237488,
+      "grad_norm": 0.8820136785507202,
+      "learning_rate": 9.998104065407565e-05,
+      "loss": 0.124,
+      "step": 10500
+    },
+    {
+      "epoch": 0.6876022244030094,
+      "grad_norm": 0.8272663354873657,
+      "learning_rate": 9.998078686665026e-05,
+      "loss": 0.116,
+      "step": 10510
+    },
+    {
+      "epoch": 0.6882564605822702,
+      "grad_norm": 0.9149497151374817,
+      "learning_rate": 9.998053139224224e-05,
+      "loss": 0.1145,
+      "step": 10520
+    },
+    {
+      "epoch": 0.6889106967615309,
+      "grad_norm": 0.8796846866607666,
+      "learning_rate": 9.99802742308602e-05,
+      "loss": 0.1186,
+      "step": 10530
+    },
+    {
+      "epoch": 0.6895649329407916,
+      "grad_norm": 1.0413655042648315,
+      "learning_rate": 9.998001538251282e-05,
+      "loss": 0.1149,
+      "step": 10540
+    },
+    {
+      "epoch": 0.6902191691200523,
+      "grad_norm": 0.9910515546798706,
+      "learning_rate": 9.997975484720887e-05,
+      "loss": 0.1258,
+      "step": 10550
+    },
+    {
+      "epoch": 0.6908734052993131,
+      "grad_norm": 0.9942857623100281,
+      "learning_rate": 9.997949262495709e-05,
+      "loss": 0.1236,
+      "step": 10560
+    },
+    {
+      "epoch": 0.6915276414785738,
+      "grad_norm": 0.8903499841690063,
+      "learning_rate": 9.997922871576638e-05,
+      "loss": 0.113,
+      "step": 10570
+    },
+    {
+      "epoch": 0.6921818776578345,
+      "grad_norm": 0.9105699062347412,
+      "learning_rate": 9.997896311964561e-05,
+      "loss": 0.1276,
+      "step": 10580
+    },
+    {
+      "epoch": 0.6928361138370952,
+      "grad_norm": 1.2962309122085571,
+      "learning_rate": 9.997869583660375e-05,
+      "loss": 0.1285,
+      "step": 10590
+    },
+    {
+      "epoch": 0.693490350016356,
+      "grad_norm": 1.1062681674957275,
+      "learning_rate": 9.997842686664985e-05,
+      "loss": 0.117,
+      "step": 10600
+    },
+    {
+      "epoch": 0.6941445861956166,
+      "grad_norm": 0.9840047955513,
+      "learning_rate": 9.997815620979297e-05,
+      "loss": 0.1329,
+      "step": 10610
+    },
+    {
+      "epoch": 0.6947988223748773,
+      "grad_norm": 0.7675462961196899,
+      "learning_rate": 9.997788386604224e-05,
+      "loss": 0.1115,
+      "step": 10620
+    },
+    {
+      "epoch": 0.6954530585541381,
+      "grad_norm": 0.9856576323509216,
+      "learning_rate": 9.997760983540686e-05,
+      "loss": 0.108,
+      "step": 10630
+    },
+    {
+      "epoch": 0.6961072947333987,
+      "grad_norm": 1.1004915237426758,
+      "learning_rate": 9.997733411789607e-05,
+      "loss": 0.1158,
+      "step": 10640
+    },
+    {
+      "epoch": 0.6967615309126595,
+      "grad_norm": 0.9109641313552856,
+      "learning_rate": 9.99770567135192e-05,
+      "loss": 0.1198,
+      "step": 10650
+    },
+    {
+      "epoch": 0.6974157670919202,
+      "grad_norm": 0.8749265074729919,
+      "learning_rate": 9.997677762228558e-05,
+      "loss": 0.1166,
+      "step": 10660
+    },
+    {
+      "epoch": 0.6980700032711809,
+      "grad_norm": 0.9863060712814331,
+      "learning_rate": 9.997649684420465e-05,
+      "loss": 0.1151,
+      "step": 10670
+    },
+    {
+      "epoch": 0.6987242394504416,
+      "grad_norm": 0.9862884879112244,
+      "learning_rate": 9.997621437928588e-05,
+      "loss": 0.1199,
+      "step": 10680
+    },
+    {
+      "epoch": 0.6993784756297023,
+      "grad_norm": 0.8637502193450928,
+      "learning_rate": 9.997593022753881e-05,
+      "loss": 0.118,
+      "step": 10690
+    },
+    {
+      "epoch": 0.700032711808963,
+      "grad_norm": 0.8936694860458374,
+      "learning_rate": 9.997564438897304e-05,
+      "loss": 0.1226,
+      "step": 10700
+    },
+    {
+      "epoch": 0.7006869479882237,
+      "grad_norm": 0.8638026714324951,
+      "learning_rate": 9.997535686359819e-05,
+      "loss": 0.1084,
+      "step": 10710
+    },
+    {
+      "epoch": 0.7013411841674845,
+      "grad_norm": 0.8587387800216675,
+      "learning_rate": 9.997506765142399e-05,
+      "loss": 0.1229,
+      "step": 10720
+    },
+    {
+      "epoch": 0.7019954203467452,
+      "grad_norm": 0.894414484500885,
+      "learning_rate": 9.997477675246019e-05,
+      "loss": 0.1046,
+      "step": 10730
+    },
+    {
+      "epoch": 0.7026496565260059,
+      "grad_norm": 0.8560695648193359,
+      "learning_rate": 9.997448416671661e-05,
+      "loss": 0.1083,
+      "step": 10740
+    },
+    {
+      "epoch": 0.7033038927052666,
+      "grad_norm": 0.9901031851768494,
+      "learning_rate": 9.997418989420313e-05,
+      "loss": 0.116,
+      "step": 10750
+    },
+    {
+      "epoch": 0.7039581288845274,
+      "grad_norm": 0.96174556016922,
+      "learning_rate": 9.997389393492966e-05,
+      "loss": 0.1185,
+      "step": 10760
+    },
+    {
+      "epoch": 0.704612365063788,
+      "grad_norm": 0.802387535572052,
+      "learning_rate": 9.997359628890623e-05,
+      "loss": 0.1202,
+      "step": 10770
+    },
+    {
+      "epoch": 0.7052666012430487,
+      "grad_norm": 0.7916472554206848,
+      "learning_rate": 9.997329695614286e-05,
+      "loss": 0.1103,
+      "step": 10780
+    },
+    {
+      "epoch": 0.7059208374223095,
+      "grad_norm": 1.1118965148925781,
+      "learning_rate": 9.997299593664966e-05,
+      "loss": 0.1263,
+      "step": 10790
+    },
+    {
+      "epoch": 0.7065750736015701,
+      "grad_norm": 1.2766551971435547,
+      "learning_rate": 9.997269323043678e-05,
+      "loss": 0.1283,
+      "step": 10800
+    },
+    {
+      "epoch": 0.7072293097808309,
+      "grad_norm": 0.9842929244041443,
+      "learning_rate": 9.997238883751446e-05,
+      "loss": 0.1167,
+      "step": 10810
+    },
+    {
+      "epoch": 0.7078835459600916,
+      "grad_norm": 0.9710572957992554,
+      "learning_rate": 9.997208275789294e-05,
+      "loss": 0.1074,
+      "step": 10820
+    },
+    {
+      "epoch": 0.7085377821393523,
+      "grad_norm": 1.014204978942871,
+      "learning_rate": 9.997177499158258e-05,
+      "loss": 0.1155,
+      "step": 10830
+    },
+    {
+      "epoch": 0.709192018318613,
+      "grad_norm": 0.851264238357544,
+      "learning_rate": 9.997146553859375e-05,
+      "loss": 0.1154,
+      "step": 10840
+    },
+    {
+      "epoch": 0.7098462544978738,
+      "grad_norm": 0.9606887102127075,
+      "learning_rate": 9.997115439893692e-05,
+      "loss": 0.1306,
+      "step": 10850
+    },
+    {
+      "epoch": 0.7105004906771345,
+      "grad_norm": 0.9389902949333191,
+      "learning_rate": 9.997084157262256e-05,
+      "loss": 0.1104,
+      "step": 10860
+    },
+    {
+      "epoch": 0.7111547268563951,
+      "grad_norm": 0.78615403175354,
+      "learning_rate": 9.997052705966126e-05,
+      "loss": 0.1112,
+      "step": 10870
+    },
+    {
+      "epoch": 0.7118089630356559,
+      "grad_norm": 0.884412407875061,
+      "learning_rate": 9.99702108600636e-05,
+      "loss": 0.1246,
+      "step": 10880
+    },
+    {
+      "epoch": 0.7124631992149166,
+      "grad_norm": 0.8281691074371338,
+      "learning_rate": 9.996989297384029e-05,
+      "loss": 0.1097,
+      "step": 10890
+    },
+    {
+      "epoch": 0.7131174353941773,
+      "grad_norm": 1.054056167602539,
+      "learning_rate": 9.996957340100203e-05,
+      "loss": 0.1206,
+      "step": 10900
+    },
+    {
+      "epoch": 0.713771671573438,
+      "grad_norm": 0.8140998482704163,
+      "learning_rate": 9.996925214155962e-05,
+      "loss": 0.1207,
+      "step": 10910
+    },
+    {
+      "epoch": 0.7144259077526988,
+      "grad_norm": 0.8529103398323059,
+      "learning_rate": 9.99689291955239e-05,
+      "loss": 0.1118,
+      "step": 10920
+    },
+    {
+      "epoch": 0.7150801439319594,
+      "grad_norm": 1.020617127418518,
+      "learning_rate": 9.996860456290576e-05,
+      "loss": 0.1215,
+      "step": 10930
+    },
+    {
+      "epoch": 0.7157343801112201,
+      "grad_norm": 0.8196330070495605,
+      "learning_rate": 9.996827824371618e-05,
+      "loss": 0.1094,
+      "step": 10940
+    },
+    {
+      "epoch": 0.7163886162904809,
+      "grad_norm": 0.9389585256576538,
+      "learning_rate": 9.996795023796617e-05,
+      "loss": 0.1175,
+      "step": 10950
+    },
+    {
+      "epoch": 0.7170428524697415,
+      "grad_norm": 0.9360471367835999,
+      "learning_rate": 9.996762054566679e-05,
+      "loss": 0.1138,
+      "step": 10960
+    },
+    {
+      "epoch": 0.7176970886490023,
+      "grad_norm": 0.9107570052146912,
+      "learning_rate": 9.996728916682915e-05,
+      "loss": 0.1124,
+      "step": 10970
+    },
+    {
+      "epoch": 0.718351324828263,
+      "grad_norm": 0.8106933236122131,
+      "learning_rate": 9.996695610146449e-05,
+      "loss": 0.124,
+      "step": 10980
+    },
+    {
+      "epoch": 0.7190055610075237,
+      "grad_norm": 1.0682817697525024,
+      "learning_rate": 9.9966621349584e-05,
+      "loss": 0.1277,
+      "step": 10990
+    },
+    {
+      "epoch": 0.7196597971867844,
+      "grad_norm": 1.0235021114349365,
+      "learning_rate": 9.9966284911199e-05,
+      "loss": 0.1229,
+      "step": 11000
+    },
+    {
+      "epoch": 0.7203140333660452,
+      "grad_norm": 1.0049972534179688,
+      "learning_rate": 9.996594678632085e-05,
+      "loss": 0.1182,
+      "step": 11010
+    },
+    {
+      "epoch": 0.7209682695453059,
+      "grad_norm": 0.8805012106895447,
+      "learning_rate": 9.996560697496094e-05,
+      "loss": 0.116,
+      "step": 11020
+    },
+    {
+      "epoch": 0.7216225057245665,
+      "grad_norm": 0.8662134408950806,
+      "learning_rate": 9.996526547713077e-05,
+      "loss": 0.117,
+      "step": 11030
+    },
+    {
+      "epoch": 0.7222767419038273,
+      "grad_norm": 1.0079435110092163,
+      "learning_rate": 9.996492229284185e-05,
+      "loss": 0.1213,
+      "step": 11040
+    },
+    {
+      "epoch": 0.722930978083088,
+      "grad_norm": 1.034039855003357,
+      "learning_rate": 9.996457742210576e-05,
+      "loss": 0.1116,
+      "step": 11050
+    },
+    {
+      "epoch": 0.7235852142623487,
+      "grad_norm": 0.9001430869102478,
+      "learning_rate": 9.996423086493414e-05,
+      "loss": 0.1155,
+      "step": 11060
+    },
+    {
+      "epoch": 0.7242394504416094,
+      "grad_norm": 0.8581667542457581,
+      "learning_rate": 9.996388262133869e-05,
+      "loss": 0.1147,
+      "step": 11070
+    },
+    {
+      "epoch": 0.7248936866208702,
+      "grad_norm": 0.9072811007499695,
+      "learning_rate": 9.996353269133118e-05,
+      "loss": 0.1164,
+      "step": 11080
+    },
+    {
+      "epoch": 0.7255479228001308,
+      "grad_norm": 0.7761887311935425,
+      "learning_rate": 9.99631810749234e-05,
+      "loss": 0.1107,
+      "step": 11090
+    },
+    {
+      "epoch": 0.7262021589793916,
+      "grad_norm": 1.0262295007705688,
+      "learning_rate": 9.996282777212723e-05,
+      "loss": 0.1209,
+      "step": 11100
+    },
+    {
+      "epoch": 0.7268563951586523,
+      "grad_norm": 0.9890956282615662,
+      "learning_rate": 9.996247278295458e-05,
+      "loss": 0.1215,
+      "step": 11110
+    },
+    {
+      "epoch": 0.7275106313379129,
+      "grad_norm": 0.8441636562347412,
+      "learning_rate": 9.996211610741745e-05,
+      "loss": 0.1076,
+      "step": 11120
+    },
+    {
+      "epoch": 0.7281648675171737,
+      "grad_norm": 0.9701690673828125,
+      "learning_rate": 9.996175774552788e-05,
+      "loss": 0.1305,
+      "step": 11130
+    },
+    {
+      "epoch": 0.7288191036964344,
+      "grad_norm": 0.8142655491828918,
+      "learning_rate": 9.996139769729795e-05,
+      "loss": 0.1164,
+      "step": 11140
+    },
+    {
+      "epoch": 0.7294733398756952,
+      "grad_norm": 1.1642175912857056,
+      "learning_rate": 9.99610359627398e-05,
+      "loss": 0.1196,
+      "step": 11150
+    },
+    {
+      "epoch": 0.7301275760549558,
+      "grad_norm": 0.864878237247467,
+      "learning_rate": 9.996067254186568e-05,
+      "loss": 0.1156,
+      "step": 11160
+    },
+    {
+      "epoch": 0.7307818122342166,
+      "grad_norm": 1.0916894674301147,
+      "learning_rate": 9.996030743468783e-05,
+      "loss": 0.1235,
+      "step": 11170
+    },
+    {
+      "epoch": 0.7314360484134773,
+      "grad_norm": 0.9421660304069519,
+      "learning_rate": 9.995994064121859e-05,
+      "loss": 0.116,
+      "step": 11180
+    },
+    {
+      "epoch": 0.732090284592738,
+      "grad_norm": 0.8942557573318481,
+      "learning_rate": 9.995957216147031e-05,
+      "loss": 0.1136,
+      "step": 11190
+    },
+    {
+      "epoch": 0.7327445207719987,
+      "grad_norm": 0.8067488670349121,
+      "learning_rate": 9.995920199545546e-05,
+      "loss": 0.112,
+      "step": 11200
+    },
+    {
+      "epoch": 0.7333987569512594,
+      "grad_norm": 0.9836673140525818,
+      "learning_rate": 9.99588301431865e-05,
+      "loss": 0.1128,
+      "step": 11210
+    },
+    {
+      "epoch": 0.7340529931305201,
+      "grad_norm": 0.7709094285964966,
+      "learning_rate": 9.995845660467602e-05,
+      "loss": 0.1205,
+      "step": 11220
+    },
+    {
+      "epoch": 0.7347072293097808,
+      "grad_norm": 0.7916396856307983,
+      "learning_rate": 9.99580813799366e-05,
+      "loss": 0.1201,
+      "step": 11230
+    },
+    {
+      "epoch": 0.7353614654890416,
+      "grad_norm": 1.0031894445419312,
+      "learning_rate": 9.995770446898092e-05,
+      "loss": 0.1181,
+      "step": 11240
+    },
+    {
+      "epoch": 0.7360157016683022,
+      "grad_norm": 1.0334042310714722,
+      "learning_rate": 9.995732587182168e-05,
+      "loss": 0.1191,
+      "step": 11250
+    },
+    {
+      "epoch": 0.736669937847563,
+      "grad_norm": 0.9869486689567566,
+      "learning_rate": 9.995694558847169e-05,
+      "loss": 0.1129,
+      "step": 11260
+    },
+    {
+      "epoch": 0.7373241740268237,
+      "grad_norm": 0.7451587915420532,
+      "learning_rate": 9.995656361894377e-05,
+      "loss": 0.1216,
+      "step": 11270
+    },
+    {
+      "epoch": 0.7379784102060843,
+      "grad_norm": 0.9288752675056458,
+      "learning_rate": 9.99561799632508e-05,
+      "loss": 0.1257,
+      "step": 11280
+    },
+    {
+      "epoch": 0.7386326463853451,
+      "grad_norm": 0.972141683101654,
+      "learning_rate": 9.995579462140574e-05,
+      "loss": 0.1124,
+      "step": 11290
+    },
+    {
+      "epoch": 0.7392868825646058,
+      "grad_norm": 1.0879883766174316,
+      "learning_rate": 9.995540759342161e-05,
+      "loss": 0.1216,
+      "step": 11300
+    },
+    {
+      "epoch": 0.7399411187438666,
+      "grad_norm": 0.9992396235466003,
+      "learning_rate": 9.995501887931146e-05,
+      "loss": 0.117,
+      "step": 11310
+    },
+    {
+      "epoch": 0.7405953549231272,
+      "grad_norm": 0.951146125793457,
+      "learning_rate": 9.99546284790884e-05,
+      "loss": 0.106,
+      "step": 11320
+    },
+    {
+      "epoch": 0.741249591102388,
+      "grad_norm": 0.9258811473846436,
+      "learning_rate": 9.995423639276562e-05,
+      "loss": 0.1253,
+      "step": 11330
+    },
+    {
+      "epoch": 0.7419038272816487,
+      "grad_norm": 0.8994291424751282,
+      "learning_rate": 9.995384262035637e-05,
+      "loss": 0.1141,
+      "step": 11340
+    },
+    {
+      "epoch": 0.7425580634609094,
+      "grad_norm": 1.0479780435562134,
+      "learning_rate": 9.99534471618739e-05,
+      "loss": 0.1228,
+      "step": 11350
+    },
+    {
+      "epoch": 0.7432122996401701,
+      "grad_norm": 0.8784400820732117,
+      "learning_rate": 9.99530500173316e-05,
+      "loss": 0.1111,
+      "step": 11360
+    },
+    {
+      "epoch": 0.7438665358194309,
+      "grad_norm": 0.9299617409706116,
+      "learning_rate": 9.995265118674284e-05,
+      "loss": 0.1127,
+      "step": 11370
+    },
+    {
+      "epoch": 0.7445207719986915,
+      "grad_norm": 0.9754354357719421,
+      "learning_rate": 9.99522506701211e-05,
+      "loss": 0.1186,
+      "step": 11380
+    },
+    {
+      "epoch": 0.7451750081779522,
+      "grad_norm": 1.0314409732818604,
+      "learning_rate": 9.99518484674799e-05,
+      "loss": 0.1145,
+      "step": 11390
+    },
+    {
+      "epoch": 0.745829244357213,
+      "grad_norm": 0.7628955841064453,
+      "learning_rate": 9.99514445788328e-05,
+      "loss": 0.1124,
+      "step": 11400
+    },
+    {
+      "epoch": 0.7464834805364736,
+      "grad_norm": 0.9075015783309937,
+      "learning_rate": 9.995103900419348e-05,
+      "loss": 0.1159,
+      "step": 11410
+    },
+    {
+      "epoch": 0.7471377167157344,
+      "grad_norm": 1.0489757061004639,
+      "learning_rate": 9.995063174357555e-05,
+      "loss": 0.1226,
+      "step": 11420
+    },
+    {
+      "epoch": 0.7477919528949951,
+      "grad_norm": 0.9378239512443542,
+      "learning_rate": 9.995022279699281e-05,
+      "loss": 0.1209,
+      "step": 11430
+    },
+    {
+      "epoch": 0.7484461890742558,
+      "grad_norm": 0.8588927388191223,
+      "learning_rate": 9.994981216445905e-05,
+      "loss": 0.1204,
+      "step": 11440
+    },
+    {
+      "epoch": 0.7491004252535165,
+      "grad_norm": 0.760901689529419,
+      "learning_rate": 9.994939984598813e-05,
+      "loss": 0.111,
+      "step": 11450
+    },
+    {
+      "epoch": 0.7497546614327772,
+      "grad_norm": 0.78853440284729,
+      "learning_rate": 9.994898584159397e-05,
+      "loss": 0.114,
+      "step": 11460
+    },
+    {
+      "epoch": 0.750408897612038,
+      "grad_norm": 0.9857150316238403,
+      "learning_rate": 9.994857015129056e-05,
+      "loss": 0.1131,
+      "step": 11470
+    },
+    {
+      "epoch": 0.7510631337912986,
+      "grad_norm": 0.8129025101661682,
+      "learning_rate": 9.994815277509188e-05,
+      "loss": 0.1038,
+      "step": 11480
+    },
+    {
+      "epoch": 0.7517173699705594,
+      "grad_norm": 0.8603448867797852,
+      "learning_rate": 9.994773371301207e-05,
+      "loss": 0.119,
+      "step": 11490
+    },
+    {
+      "epoch": 0.7523716061498201,
+      "grad_norm": 0.797516942024231,
+      "learning_rate": 9.994731296506525e-05,
+      "loss": 0.1257,
+      "step": 11500
+    },
+    {
+      "epoch": 0.7530258423290808,
+      "grad_norm": 0.7435304522514343,
+      "learning_rate": 9.994689053126564e-05,
+      "loss": 0.1089,
+      "step": 11510
+    },
+    {
+      "epoch": 0.7536800785083415,
+      "grad_norm": 0.8212321996688843,
+      "learning_rate": 9.994646641162745e-05,
+      "loss": 0.1115,
+      "step": 11520
+    },
+    {
+      "epoch": 0.7543343146876023,
+      "grad_norm": 0.8718414306640625,
+      "learning_rate": 9.994604060616506e-05,
+      "loss": 0.1215,
+      "step": 11530
+    },
+    {
+      "epoch": 0.7549885508668629,
+      "grad_norm": 0.8560036420822144,
+      "learning_rate": 9.99456131148928e-05,
+      "loss": 0.1098,
+      "step": 11540
+    },
+    {
+      "epoch": 0.7556427870461236,
+      "grad_norm": 1.0013700723648071,
+      "learning_rate": 9.994518393782513e-05,
+      "loss": 0.1199,
+      "step": 11550
+    },
+    {
+      "epoch": 0.7562970232253844,
+      "grad_norm": 0.8606430292129517,
+      "learning_rate": 9.994475307497649e-05,
+      "loss": 0.1022,
+      "step": 11560
+    },
+    {
+      "epoch": 0.756951259404645,
+      "grad_norm": 0.9482471942901611,
+      "learning_rate": 9.994432052636145e-05,
+      "loss": 0.1203,
+      "step": 11570
+    },
+    {
+      "epoch": 0.7576054955839058,
+      "grad_norm": 1.0482877492904663,
+      "learning_rate": 9.994388629199463e-05,
+      "loss": 0.1214,
+      "step": 11580
+    },
+    {
+      "epoch": 0.7582597317631665,
+      "grad_norm": 0.8290185928344727,
+      "learning_rate": 9.994345037189063e-05,
+      "loss": 0.1192,
+      "step": 11590
+    },
+    {
+      "epoch": 0.7589139679424273,
+      "grad_norm": 0.8499178886413574,
+      "learning_rate": 9.994301276606424e-05,
+      "loss": 0.1183,
+      "step": 11600
+    },
+    {
+      "epoch": 0.7595682041216879,
+      "grad_norm": 0.9083896279335022,
+      "learning_rate": 9.994257347453015e-05,
+      "loss": 0.1098,
+      "step": 11610
+    },
+    {
+      "epoch": 0.7602224403009487,
+      "grad_norm": 0.9857131242752075,
+      "learning_rate": 9.994213249730325e-05,
+      "loss": 0.1212,
+      "step": 11620
+    },
+    {
+      "epoch": 0.7608766764802094,
+      "grad_norm": 0.930366575717926,
+      "learning_rate": 9.99416898343984e-05,
+      "loss": 0.1173,
+      "step": 11630
+    },
+    {
+      "epoch": 0.76153091265947,
+      "grad_norm": 0.9344758987426758,
+      "learning_rate": 9.994124548583053e-05,
+      "loss": 0.1246,
+      "step": 11640
+    },
+    {
+      "epoch": 0.7621851488387308,
+      "grad_norm": 0.8181567192077637,
+      "learning_rate": 9.994079945161466e-05,
+      "loss": 0.1207,
+      "step": 11650
+    },
+    {
+      "epoch": 0.7628393850179915,
+      "grad_norm": 0.7763569951057434,
+      "learning_rate": 9.994035173176582e-05,
+      "loss": 0.1251,
+      "step": 11660
+    },
+    {
+      "epoch": 0.7634936211972522,
+      "grad_norm": 1.0689425468444824,
+      "learning_rate": 9.993990232629915e-05,
+      "loss": 0.136,
+      "step": 11670
+    },
+    {
+      "epoch": 0.7641478573765129,
+      "grad_norm": 0.9611626267433167,
+      "learning_rate": 9.993945123522978e-05,
+      "loss": 0.1179,
+      "step": 11680
+    },
+    {
+      "epoch": 0.7648020935557737,
+      "grad_norm": 0.9553081393241882,
+      "learning_rate": 9.9938998458573e-05,
+      "loss": 0.1153,
+      "step": 11690
+    },
+    {
+      "epoch": 0.7654563297350343,
+      "grad_norm": 0.9695801138877869,
+      "learning_rate": 9.993854399634402e-05,
+      "loss": 0.1305,
+      "step": 11700
+    },
+    {
+      "epoch": 0.766110565914295,
+      "grad_norm": 1.1359174251556396,
+      "learning_rate": 9.993808784855823e-05,
+      "loss": 0.1135,
+      "step": 11710
+    },
+    {
+      "epoch": 0.7667648020935558,
+      "grad_norm": 0.8524360060691833,
+      "learning_rate": 9.9937630015231e-05,
+      "loss": 0.1204,
+      "step": 11720
+    },
+    {
+      "epoch": 0.7674190382728164,
+      "grad_norm": 1.0540283918380737,
+      "learning_rate": 9.993717049637779e-05,
+      "loss": 0.1255,
+      "step": 11730
+    },
+    {
+      "epoch": 0.7680732744520772,
+      "grad_norm": 0.8269158601760864,
+      "learning_rate": 9.99367092920141e-05,
+      "loss": 0.1145,
+      "step": 11740
+    },
+    {
+      "epoch": 0.7687275106313379,
+      "grad_norm": 0.9928538203239441,
+      "learning_rate": 9.993624640215552e-05,
+      "loss": 0.1245,
+      "step": 11750
+    },
+    {
+      "epoch": 0.7693817468105987,
+      "grad_norm": 0.9222886562347412,
+      "learning_rate": 9.993578182681767e-05,
+      "loss": 0.1098,
+      "step": 11760
+    },
+    {
+      "epoch": 0.7700359829898593,
+      "grad_norm": 1.1276291608810425,
+      "learning_rate": 9.993531556601621e-05,
+      "loss": 0.1114,
+      "step": 11770
+    },
+    {
+      "epoch": 0.7706902191691201,
+      "grad_norm": 0.9698314070701599,
+      "learning_rate": 9.993484761976688e-05,
+      "loss": 0.1103,
+      "step": 11780
+    },
+    {
+      "epoch": 0.7713444553483808,
+      "grad_norm": 1.0448344945907593,
+      "learning_rate": 9.993437798808549e-05,
+      "loss": 0.1034,
+      "step": 11790
+    },
+    {
+      "epoch": 0.7719986915276414,
+      "grad_norm": 0.7689130902290344,
+      "learning_rate": 9.99339066709879e-05,
+      "loss": 0.112,
+      "step": 11800
+    },
+    {
+      "epoch": 0.7726529277069022,
+      "grad_norm": 0.8706762790679932,
+      "learning_rate": 9.993343366849e-05,
+      "loss": 0.1238,
+      "step": 11810
+    },
+    {
+      "epoch": 0.7733071638861629,
+      "grad_norm": 0.9579262733459473,
+      "learning_rate": 9.993295898060775e-05,
+      "loss": 0.1092,
+      "step": 11820
+    },
+    {
+      "epoch": 0.7739614000654236,
+      "grad_norm": 0.9516409039497375,
+      "learning_rate": 9.993248260735717e-05,
+      "loss": 0.1148,
+      "step": 11830
+    },
+    {
+      "epoch": 0.7746156362446843,
+      "grad_norm": 0.7520780563354492,
+      "learning_rate": 9.993200454875436e-05,
+      "loss": 0.1155,
+      "step": 11840
+    },
+    {
+      "epoch": 0.7752698724239451,
+      "grad_norm": 0.8823644518852234,
+      "learning_rate": 9.993152480481545e-05,
+      "loss": 0.1093,
+      "step": 11850
+    },
+    {
+      "epoch": 0.7759241086032057,
+      "grad_norm": 1.0025229454040527,
+      "learning_rate": 9.993104337555663e-05,
+      "loss": 0.1207,
+      "step": 11860
+    },
+    {
+      "epoch": 0.7765783447824665,
+      "grad_norm": 0.9853441715240479,
+      "learning_rate": 9.993056026099415e-05,
+      "loss": 0.1264,
+      "step": 11870
+    },
+    {
+      "epoch": 0.7772325809617272,
+      "grad_norm": 0.8879414200782776,
+      "learning_rate": 9.99300754611443e-05,
+      "loss": 0.1252,
+      "step": 11880
+    },
+    {
+      "epoch": 0.777886817140988,
+      "grad_norm": 0.8338192105293274,
+      "learning_rate": 9.992958897602344e-05,
+      "loss": 0.1187,
+      "step": 11890
+    },
+    {
+      "epoch": 0.7785410533202486,
+      "grad_norm": 0.872117280960083,
+      "learning_rate": 9.992910080564803e-05,
+      "loss": 0.1105,
+      "step": 11900
+    },
+    {
+      "epoch": 0.7791952894995093,
+      "grad_norm": 0.9396425485610962,
+      "learning_rate": 9.992861095003454e-05,
+      "loss": 0.1131,
+      "step": 11910
+    },
+    {
+      "epoch": 0.7798495256787701,
+      "grad_norm": 1.1532872915267944,
+      "learning_rate": 9.992811940919946e-05,
+      "loss": 0.1247,
+      "step": 11920
+    },
+    {
+      "epoch": 0.7805037618580307,
+      "grad_norm": 1.065436601638794,
+      "learning_rate": 9.992762618315942e-05,
+      "loss": 0.127,
+      "step": 11930
+    },
+    {
+      "epoch": 0.7811579980372915,
+      "grad_norm": 0.9578902721405029,
+      "learning_rate": 9.992713127193106e-05,
+      "loss": 0.1106,
+      "step": 11940
+    },
+    {
+      "epoch": 0.7818122342165522,
+      "grad_norm": 0.8533861637115479,
+      "learning_rate": 9.992663467553108e-05,
+      "loss": 0.1244,
+      "step": 11950
+    },
+    {
+      "epoch": 0.7824664703958129,
+      "grad_norm": 0.9745835661888123,
+      "learning_rate": 9.992613639397624e-05,
+      "loss": 0.1104,
+      "step": 11960
+    },
+    {
+      "epoch": 0.7831207065750736,
+      "grad_norm": 0.8613539338111877,
+      "learning_rate": 9.992563642728335e-05,
+      "loss": 0.1044,
+      "step": 11970
+    },
+    {
+      "epoch": 0.7837749427543343,
+      "grad_norm": 1.0119600296020508,
+      "learning_rate": 9.992513477546931e-05,
+      "loss": 0.1181,
+      "step": 11980
+    },
+    {
+      "epoch": 0.784429178933595,
+      "grad_norm": 0.856393039226532,
+      "learning_rate": 9.992463143855102e-05,
+      "loss": 0.1172,
+      "step": 11990
+    },
+    {
+      "epoch": 0.7850834151128557,
+      "grad_norm": 0.8078387379646301,
+      "learning_rate": 9.992412641654551e-05,
+      "loss": 0.1332,
+      "step": 12000
+    },
+    {
+      "epoch": 0.7857376512921165,
+      "grad_norm": 0.818047046661377,
+      "learning_rate": 9.99236197094698e-05,
+      "loss": 0.1156,
+      "step": 12010
+    },
+    {
+      "epoch": 0.7863918874713771,
+      "grad_norm": 0.8972421288490295,
+      "learning_rate": 9.992311131734098e-05,
+      "loss": 0.1162,
+      "step": 12020
+    },
+    {
+      "epoch": 0.7870461236506379,
+      "grad_norm": 0.9026692509651184,
+      "learning_rate": 9.992260124017623e-05,
+      "loss": 0.1239,
+      "step": 12030
+    },
+    {
+      "epoch": 0.7877003598298986,
+      "grad_norm": 0.8012493252754211,
+      "learning_rate": 9.992208947799276e-05,
+      "loss": 0.1224,
+      "step": 12040
+    },
+    {
+      "epoch": 0.7883545960091594,
+      "grad_norm": 0.7508116960525513,
+      "learning_rate": 9.992157603080785e-05,
+      "loss": 0.1145,
+      "step": 12050
+    },
+    {
+      "epoch": 0.78900883218842,
+      "grad_norm": 0.9431008100509644,
+      "learning_rate": 9.992106089863883e-05,
+      "loss": 0.1209,
+      "step": 12060
+    },
+    {
+      "epoch": 0.7896630683676807,
+      "grad_norm": 0.9398839473724365,
+      "learning_rate": 9.992054408150307e-05,
+      "loss": 0.1195,
+      "step": 12070
+    },
+    {
+      "epoch": 0.7903173045469415,
+      "grad_norm": 0.8729955554008484,
+      "learning_rate": 9.992002557941804e-05,
+      "loss": 0.1151,
+      "step": 12080
+    },
+    {
+      "epoch": 0.7909715407262021,
+      "grad_norm": 0.8589168787002563,
+      "learning_rate": 9.991950539240122e-05,
+      "loss": 0.1228,
+      "step": 12090
+    },
+    {
+      "epoch": 0.7916257769054629,
+      "grad_norm": 0.7919473052024841,
+      "learning_rate": 9.991898352047016e-05,
+      "loss": 0.1113,
+      "step": 12100
+    },
+    {
+      "epoch": 0.7922800130847236,
+      "grad_norm": 0.8259612321853638,
+      "learning_rate": 9.99184599636425e-05,
+      "loss": 0.1143,
+      "step": 12110
+    },
+    {
+      "epoch": 0.7929342492639843,
+      "grad_norm": 0.9801688194274902,
+      "learning_rate": 9.99179347219359e-05,
+      "loss": 0.1148,
+      "step": 12120
+    },
+    {
+      "epoch": 0.793588485443245,
+      "grad_norm": 0.9544298648834229,
+      "learning_rate": 9.991740779536808e-05,
+      "loss": 0.1276,
+      "step": 12130
+    },
+    {
+      "epoch": 0.7942427216225058,
+      "grad_norm": 0.946922242641449,
+      "learning_rate": 9.991687918395686e-05,
+      "loss": 0.1174,
+      "step": 12140
+    },
+    {
+      "epoch": 0.7948969578017664,
+      "grad_norm": 1.27000892162323,
+      "learning_rate": 9.991634888772003e-05,
+      "loss": 0.1163,
+      "step": 12150
+    },
+    {
+      "epoch": 0.7955511939810271,
+      "grad_norm": 0.7371228933334351,
+      "learning_rate": 9.991581690667553e-05,
+      "loss": 0.117,
+      "step": 12160
+    },
+    {
+      "epoch": 0.7962054301602879,
+      "grad_norm": 0.8912835717201233,
+      "learning_rate": 9.99152832408413e-05,
+      "loss": 0.1134,
+      "step": 12170
+    },
+    {
+      "epoch": 0.7968596663395485,
+      "grad_norm": 0.8078799843788147,
+      "learning_rate": 9.991474789023534e-05,
+      "loss": 0.1217,
+      "step": 12180
+    },
+    {
+      "epoch": 0.7975139025188093,
+      "grad_norm": 0.8640681505203247,
+      "learning_rate": 9.991421085487573e-05,
+      "loss": 0.1029,
+      "step": 12190
+    },
+    {
+      "epoch": 0.79816813869807,
+      "grad_norm": 1.075920820236206,
+      "learning_rate": 9.991367213478062e-05,
+      "loss": 0.1155,
+      "step": 12200
+    },
+    {
+      "epoch": 0.7988223748773308,
+      "grad_norm": 1.051973581314087,
+      "learning_rate": 9.991313172996815e-05,
+      "loss": 0.1233,
+      "step": 12210
+    },
+    {
+      "epoch": 0.7994766110565914,
+      "grad_norm": 0.9649173617362976,
+      "learning_rate": 9.991258964045659e-05,
+      "loss": 0.1247,
+      "step": 12220
+    },
+    {
+      "epoch": 0.8001308472358521,
+      "grad_norm": 0.9157887697219849,
+      "learning_rate": 9.991204586626424e-05,
+      "loss": 0.1288,
+      "step": 12230
+    },
+    {
+      "epoch": 0.8007850834151129,
+      "grad_norm": 0.8497806787490845,
+      "learning_rate": 9.991150040740944e-05,
+      "loss": 0.1091,
+      "step": 12240
+    },
+    {
+      "epoch": 0.8014393195943735,
+      "grad_norm": 1.007943034172058,
+      "learning_rate": 9.99109532639106e-05,
+      "loss": 0.1104,
+      "step": 12250
+    },
+    {
+      "epoch": 0.8020935557736343,
+      "grad_norm": 0.8589887022972107,
+      "learning_rate": 9.991040443578618e-05,
+      "loss": 0.1219,
+      "step": 12260
+    },
+    {
+      "epoch": 0.802747791952895,
+      "grad_norm": 1.0197535753250122,
+      "learning_rate": 9.990985392305473e-05,
+      "loss": 0.1156,
+      "step": 12270
+    },
+    {
+      "epoch": 0.8034020281321557,
+      "grad_norm": 0.8951653242111206,
+      "learning_rate": 9.99093017257348e-05,
+      "loss": 0.111,
+      "step": 12280
+    },
+    {
+      "epoch": 0.8040562643114164,
+      "grad_norm": 1.0193089246749878,
+      "learning_rate": 9.990874784384506e-05,
+      "loss": 0.1139,
+      "step": 12290
+    },
+    {
+      "epoch": 0.8047105004906772,
+      "grad_norm": 0.8544113636016846,
+      "learning_rate": 9.990819227740418e-05,
+      "loss": 0.1092,
+      "step": 12300
+    },
+    {
+      "epoch": 0.8053647366699378,
+      "grad_norm": 0.9851740598678589,
+      "learning_rate": 9.990763502643094e-05,
+      "loss": 0.1163,
+      "step": 12310
+    },
+    {
+      "epoch": 0.8060189728491985,
+      "grad_norm": 1.2490851879119873,
+      "learning_rate": 9.990707609094412e-05,
+      "loss": 0.1177,
+      "step": 12320
+    },
+    {
+      "epoch": 0.8066732090284593,
+      "grad_norm": 0.8917890787124634,
+      "learning_rate": 9.990651547096259e-05,
+      "loss": 0.1218,
+      "step": 12330
+    },
+    {
+      "epoch": 0.80732744520772,
+      "grad_norm": 1.0979591608047485,
+      "learning_rate": 9.990595316650528e-05,
+      "loss": 0.114,
+      "step": 12340
+    },
+    {
+      "epoch": 0.8079816813869807,
+      "grad_norm": 0.89876788854599,
+      "learning_rate": 9.990538917759117e-05,
+      "loss": 0.1308,
+      "step": 12350
+    },
+    {
+      "epoch": 0.8086359175662414,
+      "grad_norm": 0.9439699053764343,
+      "learning_rate": 9.990482350423929e-05,
+      "loss": 0.1131,
+      "step": 12360
+    },
+    {
+      "epoch": 0.8092901537455022,
+      "grad_norm": 0.9236170053482056,
+      "learning_rate": 9.990425614646874e-05,
+      "loss": 0.114,
+      "step": 12370
+    },
+    {
+      "epoch": 0.8099443899247628,
+      "grad_norm": 1.0061261653900146,
+      "learning_rate": 9.990368710429866e-05,
+      "loss": 0.1216,
+      "step": 12380
+    },
+    {
+      "epoch": 0.8105986261040236,
+      "grad_norm": 0.9143770933151245,
+      "learning_rate": 9.990311637774827e-05,
+      "loss": 0.1159,
+      "step": 12390
+    },
+    {
+      "epoch": 0.8112528622832843,
+      "grad_norm": 0.8878925442695618,
+      "learning_rate": 9.990254396683683e-05,
+      "loss": 0.1078,
+      "step": 12400
+    },
+    {
+      "epoch": 0.8119070984625449,
+      "grad_norm": 1.0045684576034546,
+      "learning_rate": 9.990196987158364e-05,
+      "loss": 0.1112,
+      "step": 12410
+    },
+    {
+      "epoch": 0.8125613346418057,
+      "grad_norm": 0.7909841537475586,
+      "learning_rate": 9.990139409200812e-05,
+      "loss": 0.1122,
+      "step": 12420
+    },
+    {
+      "epoch": 0.8132155708210664,
+      "grad_norm": 0.8620671629905701,
+      "learning_rate": 9.990081662812966e-05,
+      "loss": 0.121,
+      "step": 12430
+    },
+    {
+      "epoch": 0.8138698070003271,
+      "grad_norm": 0.8675190210342407,
+      "learning_rate": 9.990023747996777e-05,
+      "loss": 0.1225,
+      "step": 12440
+    },
+    {
+      "epoch": 0.8145240431795878,
+      "grad_norm": 1.0223960876464844,
+      "learning_rate": 9.9899656647542e-05,
+      "loss": 0.1217,
+      "step": 12450
+    },
+    {
+      "epoch": 0.8151782793588486,
+      "grad_norm": 0.850107729434967,
+      "learning_rate": 9.989907413087196e-05,
+      "loss": 0.1081,
+      "step": 12460
+    },
+    {
+      "epoch": 0.8158325155381092,
+      "grad_norm": 0.8945133090019226,
+      "learning_rate": 9.98984899299773e-05,
+      "loss": 0.1133,
+      "step": 12470
+    },
+    {
+      "epoch": 0.81648675171737,
+      "grad_norm": 0.8991668820381165,
+      "learning_rate": 9.989790404487773e-05,
+      "loss": 0.1104,
+      "step": 12480
+    },
+    {
+      "epoch": 0.8171409878966307,
+      "grad_norm": 0.7645007967948914,
+      "learning_rate": 9.989731647559304e-05,
+      "loss": 0.1179,
+      "step": 12490
+    },
+    {
+      "epoch": 0.8177952240758914,
+      "grad_norm": 0.7892476916313171,
+      "learning_rate": 9.989672722214307e-05,
+      "loss": 0.1097,
+      "step": 12500
+    },
+    {
+      "epoch": 0.8184494602551521,
+      "grad_norm": 0.882279098033905,
+      "learning_rate": 9.989613628454769e-05,
+      "loss": 0.1278,
+      "step": 12510
+    },
+    {
+      "epoch": 0.8191036964344128,
+      "grad_norm": 1.140187382698059,
+      "learning_rate": 9.989554366282684e-05,
+      "loss": 0.1163,
+      "step": 12520
+    },
+    {
+      "epoch": 0.8197579326136736,
+      "grad_norm": 0.8200512528419495,
+      "learning_rate": 9.989494935700054e-05,
+      "loss": 0.1209,
+      "step": 12530
+    },
+    {
+      "epoch": 0.8204121687929342,
+      "grad_norm": 0.9112234711647034,
+      "learning_rate": 9.989435336708886e-05,
+      "loss": 0.1156,
+      "step": 12540
+    },
+    {
+      "epoch": 0.821066404972195,
+      "grad_norm": 0.7766229510307312,
+      "learning_rate": 9.98937556931119e-05,
+      "loss": 0.1101,
+      "step": 12550
+    },
+    {
+      "epoch": 0.8217206411514557,
+      "grad_norm": 1.0036638975143433,
+      "learning_rate": 9.989315633508983e-05,
+      "loss": 0.1233,
+      "step": 12560
+    },
+    {
+      "epoch": 0.8223748773307163,
+      "grad_norm": 0.9545724391937256,
+      "learning_rate": 9.989255529304287e-05,
+      "loss": 0.1198,
+      "step": 12570
+    },
+    {
+      "epoch": 0.8230291135099771,
+      "grad_norm": 0.8219319581985474,
+      "learning_rate": 9.989195256699133e-05,
+      "loss": 0.1183,
+      "step": 12580
+    },
+    {
+      "epoch": 0.8236833496892378,
+      "grad_norm": 0.7603002190589905,
+      "learning_rate": 9.989134815695556e-05,
+      "loss": 0.1088,
+      "step": 12590
+    },
+    {
+      "epoch": 0.8243375858684985,
+      "grad_norm": 0.7731127738952637,
+      "learning_rate": 9.989074206295592e-05,
+      "loss": 0.1181,
+      "step": 12600
+    },
+    {
+      "epoch": 0.8249918220477592,
+      "grad_norm": 1.0608632564544678,
+      "learning_rate": 9.98901342850129e-05,
+      "loss": 0.1248,
+      "step": 12610
+    },
+    {
+      "epoch": 0.82564605822702,
+      "grad_norm": 0.9366574287414551,
+      "learning_rate": 9.988952482314702e-05,
+      "loss": 0.1197,
+      "step": 12620
+    },
+    {
+      "epoch": 0.8263002944062807,
+      "grad_norm": 0.8143541812896729,
+      "learning_rate": 9.988891367737882e-05,
+      "loss": 0.1033,
+      "step": 12630
+    },
+    {
+      "epoch": 0.8269545305855414,
+      "grad_norm": 1.157392144203186,
+      "learning_rate": 9.988830084772896e-05,
+      "loss": 0.1086,
+      "step": 12640
+    },
+    {
+      "epoch": 0.8276087667648021,
+      "grad_norm": 0.9264316558837891,
+      "learning_rate": 9.98876863342181e-05,
+      "loss": 0.11,
+      "step": 12650
+    },
+    {
+      "epoch": 0.8282630029440629,
+      "grad_norm": 0.9766997694969177,
+      "learning_rate": 9.988707013686698e-05,
+      "loss": 0.11,
+      "step": 12660
+    },
+    {
+      "epoch": 0.8289172391233235,
+      "grad_norm": 0.9637208580970764,
+      "learning_rate": 9.988645225569643e-05,
+      "loss": 0.1112,
+      "step": 12670
+    },
+    {
+      "epoch": 0.8295714753025842,
+      "grad_norm": 0.9603047370910645,
+      "learning_rate": 9.988583269072727e-05,
+      "loss": 0.1092,
+      "step": 12680
+    },
+    {
+      "epoch": 0.830225711481845,
+      "grad_norm": 0.8850484490394592,
+      "learning_rate": 9.988521144198043e-05,
+      "loss": 0.1286,
+      "step": 12690
+    },
+    {
+      "epoch": 0.8308799476611056,
+      "grad_norm": 1.1651179790496826,
+      "learning_rate": 9.988458850947689e-05,
+      "loss": 0.1156,
+      "step": 12700
+    },
+    {
+      "epoch": 0.8315341838403664,
+      "grad_norm": 1.13690185546875,
+      "learning_rate": 9.988396389323764e-05,
+      "loss": 0.1165,
+      "step": 12710
+    },
+    {
+      "epoch": 0.8321884200196271,
+      "grad_norm": 0.8526743054389954,
+      "learning_rate": 9.988333759328379e-05,
+      "loss": 0.1144,
+      "step": 12720
+    },
+    {
+      "epoch": 0.8328426561988878,
+      "grad_norm": 0.8829309344291687,
+      "learning_rate": 9.988270960963648e-05,
+      "loss": 0.1118,
+      "step": 12730
+    },
+    {
+      "epoch": 0.8334968923781485,
+      "grad_norm": 0.8683731555938721,
+      "learning_rate": 9.988207994231689e-05,
+      "loss": 0.1123,
+      "step": 12740
+    },
+    {
+      "epoch": 0.8341511285574092,
+      "grad_norm": 0.9223302602767944,
+      "learning_rate": 9.988144859134627e-05,
+      "loss": 0.1257,
+      "step": 12750
+    },
+    {
+      "epoch": 0.8348053647366699,
+      "grad_norm": 1.061397910118103,
+      "learning_rate": 9.988081555674596e-05,
+      "loss": 0.125,
+      "step": 12760
+    },
+    {
+      "epoch": 0.8354596009159306,
+      "grad_norm": 1.0193485021591187,
+      "learning_rate": 9.98801808385373e-05,
+      "loss": 0.1168,
+      "step": 12770
+    },
+    {
+      "epoch": 0.8361138370951914,
+      "grad_norm": 0.9661241769790649,
+      "learning_rate": 9.987954443674173e-05,
+      "loss": 0.1085,
+      "step": 12780
+    },
+    {
+      "epoch": 0.8367680732744521,
+      "grad_norm": 0.9959526658058167,
+      "learning_rate": 9.98789063513807e-05,
+      "loss": 0.1082,
+      "step": 12790
+    },
+    {
+      "epoch": 0.8374223094537128,
+      "grad_norm": 0.7342641353607178,
+      "learning_rate": 9.987826658247579e-05,
+      "loss": 0.1187,
+      "step": 12800
+    },
+    {
+      "epoch": 0.8380765456329735,
+      "grad_norm": 0.9340495467185974,
+      "learning_rate": 9.987762513004856e-05,
+      "loss": 0.1022,
+      "step": 12810
+    },
+    {
+      "epoch": 0.8387307818122343,
+      "grad_norm": 0.9463670253753662,
+      "learning_rate": 9.98769819941207e-05,
+      "loss": 0.1246,
+      "step": 12820
+    },
+    {
+      "epoch": 0.8393850179914949,
+      "grad_norm": 0.8426292538642883,
+      "learning_rate": 9.987633717471385e-05,
+      "loss": 0.1077,
+      "step": 12830
+    },
+    {
+      "epoch": 0.8400392541707556,
+      "grad_norm": 1.0582283735275269,
+      "learning_rate": 9.987569067184983e-05,
+      "loss": 0.1215,
+      "step": 12840
+    },
+    {
+      "epoch": 0.8406934903500164,
+      "grad_norm": 0.9140653610229492,
+      "learning_rate": 9.987504248555047e-05,
+      "loss": 0.1193,
+      "step": 12850
+    },
+    {
+      "epoch": 0.841347726529277,
+      "grad_norm": 0.843436598777771,
+      "learning_rate": 9.98743926158376e-05,
+      "loss": 0.1243,
+      "step": 12860
+    },
+    {
+      "epoch": 0.8420019627085378,
+      "grad_norm": 0.8201360702514648,
+      "learning_rate": 9.987374106273318e-05,
+      "loss": 0.1106,
+      "step": 12870
+    },
+    {
+      "epoch": 0.8426561988877985,
+      "grad_norm": 1.0431125164031982,
+      "learning_rate": 9.987308782625919e-05,
+      "loss": 0.1182,
+      "step": 12880
+    },
+    {
+      "epoch": 0.8433104350670592,
+      "grad_norm": 0.8388211727142334,
+      "learning_rate": 9.98724329064377e-05,
+      "loss": 0.1093,
+      "step": 12890
+    },
+    {
+      "epoch": 0.8439646712463199,
+      "grad_norm": 0.8879048228263855,
+      "learning_rate": 9.987177630329081e-05,
+      "loss": 0.1004,
+      "step": 12900
+    },
+    {
+      "epoch": 0.8446189074255807,
+      "grad_norm": 0.857945442199707,
+      "learning_rate": 9.987111801684068e-05,
+      "loss": 0.1107,
+      "step": 12910
+    },
+    {
+      "epoch": 0.8452731436048413,
+      "grad_norm": 0.7848942875862122,
+      "learning_rate": 9.987045804710951e-05,
+      "loss": 0.1112,
+      "step": 12920
+    },
+    {
+      "epoch": 0.845927379784102,
+      "grad_norm": 0.8898484706878662,
+      "learning_rate": 9.98697963941196e-05,
+      "loss": 0.114,
+      "step": 12930
+    },
+    {
+      "epoch": 0.8465816159633628,
+      "grad_norm": 1.1552084684371948,
+      "learning_rate": 9.986913305789328e-05,
+      "loss": 0.1143,
+      "step": 12940
+    },
+    {
+      "epoch": 0.8472358521426235,
+      "grad_norm": 0.8362104296684265,
+      "learning_rate": 9.986846803845291e-05,
+      "loss": 0.1142,
+      "step": 12950
+    },
+    {
+      "epoch": 0.8478900883218842,
+      "grad_norm": 1.0865123271942139,
+      "learning_rate": 9.9867801335821e-05,
+      "loss": 0.1147,
+      "step": 12960
+    },
+    {
+      "epoch": 0.8485443245011449,
+      "grad_norm": 0.8024462461471558,
+      "learning_rate": 9.986713295001997e-05,
+      "loss": 0.1101,
+      "step": 12970
+    },
+    {
+      "epoch": 0.8491985606804057,
+      "grad_norm": 0.960501492023468,
+      "learning_rate": 9.986646288107243e-05,
+      "loss": 0.1215,
+      "step": 12980
+    },
+    {
+      "epoch": 0.8498527968596663,
+      "grad_norm": 0.7977132797241211,
+      "learning_rate": 9.9865791129001e-05,
+      "loss": 0.1175,
+      "step": 12990
+    },
+    {
+      "epoch": 0.850507033038927,
+      "grad_norm": 0.913974404335022,
+      "learning_rate": 9.986511769382834e-05,
+      "loss": 0.1115,
+      "step": 13000
+    },
+    {
+      "epoch": 0.8511612692181878,
+      "grad_norm": 0.86736661195755,
+      "learning_rate": 9.986444257557717e-05,
+      "loss": 0.1113,
+      "step": 13010
+    },
+    {
+      "epoch": 0.8518155053974484,
+      "grad_norm": 0.8728740811347961,
+      "learning_rate": 9.98637657742703e-05,
+      "loss": 0.1113,
+      "step": 13020
+    },
+    {
+      "epoch": 0.8524697415767092,
+      "grad_norm": 0.8732495903968811,
+      "learning_rate": 9.986308728993056e-05,
+      "loss": 0.1129,
+      "step": 13030
+    },
+    {
+      "epoch": 0.8531239777559699,
+      "grad_norm": 0.9430647492408752,
+      "learning_rate": 9.986240712258085e-05,
+      "loss": 0.1065,
+      "step": 13040
+    },
+    {
+      "epoch": 0.8537782139352306,
+      "grad_norm": 0.849603533744812,
+      "learning_rate": 9.986172527224413e-05,
+      "loss": 0.1198,
+      "step": 13050
+    },
+    {
+      "epoch": 0.8544324501144913,
+      "grad_norm": 0.75860595703125,
+      "learning_rate": 9.986104173894342e-05,
+      "loss": 0.1063,
+      "step": 13060
+    },
+    {
+      "epoch": 0.8550866862937521,
+      "grad_norm": 0.916583240032196,
+      "learning_rate": 9.986035652270178e-05,
+      "loss": 0.1179,
+      "step": 13070
+    },
+    {
+      "epoch": 0.8557409224730128,
+      "grad_norm": 0.7773154377937317,
+      "learning_rate": 9.985966962354235e-05,
+      "loss": 0.1102,
+      "step": 13080
+    },
+    {
+      "epoch": 0.8563951586522734,
+      "grad_norm": 0.7246621251106262,
+      "learning_rate": 9.985898104148831e-05,
+      "loss": 0.1108,
+      "step": 13090
+    },
+    {
+      "epoch": 0.8570493948315342,
+      "grad_norm": 0.7535557746887207,
+      "learning_rate": 9.985829077656291e-05,
+      "loss": 0.1099,
+      "step": 13100
+    },
+    {
+      "epoch": 0.8577036310107949,
+      "grad_norm": 0.8509471416473389,
+      "learning_rate": 9.98575988287894e-05,
+      "loss": 0.0995,
+      "step": 13110
+    },
+    {
+      "epoch": 0.8583578671900556,
+      "grad_norm": 0.7079533934593201,
+      "learning_rate": 9.98569051981912e-05,
+      "loss": 0.1189,
+      "step": 13120
+    },
+    {
+      "epoch": 0.8590121033693163,
+      "grad_norm": 0.9748969674110413,
+      "learning_rate": 9.985620988479169e-05,
+      "loss": 0.1168,
+      "step": 13130
+    },
+    {
+      "epoch": 0.8596663395485771,
+      "grad_norm": 0.9905074834823608,
+      "learning_rate": 9.985551288861435e-05,
+      "loss": 0.1155,
+      "step": 13140
+    },
+    {
+      "epoch": 0.8603205757278377,
+      "grad_norm": 0.7859392762184143,
+      "learning_rate": 9.98548142096827e-05,
+      "loss": 0.1111,
+      "step": 13150
+    },
+    {
+      "epoch": 0.8609748119070985,
+      "grad_norm": 0.7785911560058594,
+      "learning_rate": 9.985411384802031e-05,
+      "loss": 0.1159,
+      "step": 13160
+    },
+    {
+      "epoch": 0.8616290480863592,
+      "grad_norm": 1.0877043008804321,
+      "learning_rate": 9.985341180365084e-05,
+      "loss": 0.1197,
+      "step": 13170
+    },
+    {
+      "epoch": 0.8622832842656198,
+      "grad_norm": 0.8900846838951111,
+      "learning_rate": 9.985270807659798e-05,
+      "loss": 0.1181,
+      "step": 13180
+    },
+    {
+      "epoch": 0.8629375204448806,
+      "grad_norm": 0.7488992810249329,
+      "learning_rate": 9.985200266688546e-05,
+      "loss": 0.1058,
+      "step": 13190
+    },
+    {
+      "epoch": 0.8635917566241413,
+      "grad_norm": 0.7488667964935303,
+      "learning_rate": 9.985129557453713e-05,
+      "loss": 0.1136,
+      "step": 13200
+    },
+    {
+      "epoch": 0.864245992803402,
+      "grad_norm": 1.060916543006897,
+      "learning_rate": 9.985058679957681e-05,
+      "loss": 0.1169,
+      "step": 13210
+    },
+    {
+      "epoch": 0.8649002289826627,
+      "grad_norm": 1.009759783744812,
+      "learning_rate": 9.984987634202847e-05,
+      "loss": 0.1154,
+      "step": 13220
+    },
+    {
+      "epoch": 0.8655544651619235,
+      "grad_norm": 1.0468559265136719,
+      "learning_rate": 9.984916420191607e-05,
+      "loss": 0.1205,
+      "step": 13230
+    },
+    {
+      "epoch": 0.8662087013411842,
+      "grad_norm": 0.9158945083618164,
+      "learning_rate": 9.984845037926362e-05,
+      "loss": 0.1076,
+      "step": 13240
+    },
+    {
+      "epoch": 0.8668629375204449,
+      "grad_norm": 0.893125057220459,
+      "learning_rate": 9.984773487409527e-05,
+      "loss": 0.1189,
+      "step": 13250
+    },
+    {
+      "epoch": 0.8675171736997056,
+      "grad_norm": 0.9089499711990356,
+      "learning_rate": 9.984701768643512e-05,
+      "loss": 0.1033,
+      "step": 13260
+    },
+    {
+      "epoch": 0.8681714098789663,
+      "grad_norm": 0.9404613375663757,
+      "learning_rate": 9.984629881630738e-05,
+      "loss": 0.1225,
+      "step": 13270
+    },
+    {
+      "epoch": 0.868825646058227,
+      "grad_norm": 0.8523469567298889,
+      "learning_rate": 9.984557826373635e-05,
+      "loss": 0.1146,
+      "step": 13280
+    },
+    {
+      "epoch": 0.8694798822374877,
+      "grad_norm": 0.9582688212394714,
+      "learning_rate": 9.984485602874632e-05,
+      "loss": 0.1278,
+      "step": 13290
+    },
+    {
+      "epoch": 0.8701341184167485,
+      "grad_norm": 0.8448380827903748,
+      "learning_rate": 9.984413211136167e-05,
+      "loss": 0.1224,
+      "step": 13300
+    },
+    {
+      "epoch": 0.8707883545960091,
+      "grad_norm": 0.8751958608627319,
+      "learning_rate": 9.984340651160685e-05,
+      "loss": 0.1088,
+      "step": 13310
+    },
+    {
+      "epoch": 0.8714425907752699,
+      "grad_norm": 0.7682243585586548,
+      "learning_rate": 9.984267922950634e-05,
+      "loss": 0.1048,
+      "step": 13320
+    },
+    {
+      "epoch": 0.8720968269545306,
+      "grad_norm": 0.9508857727050781,
+      "learning_rate": 9.984195026508469e-05,
+      "loss": 0.1124,
+      "step": 13330
+    },
+    {
+      "epoch": 0.8727510631337912,
+      "grad_norm": 1.0506585836410522,
+      "learning_rate": 9.98412196183665e-05,
+      "loss": 0.1266,
+      "step": 13340
+    },
+    {
+      "epoch": 0.873405299313052,
+      "grad_norm": 1.0263482332229614,
+      "learning_rate": 9.984048728937643e-05,
+      "loss": 0.1313,
+      "step": 13350
+    },
+    {
+      "epoch": 0.8740595354923127,
+      "grad_norm": 0.8746235370635986,
+      "learning_rate": 9.98397532781392e-05,
+      "loss": 0.1066,
+      "step": 13360
+    },
+    {
+      "epoch": 0.8747137716715735,
+      "grad_norm": 0.797562301158905,
+      "learning_rate": 9.98390175846796e-05,
+      "loss": 0.1052,
+      "step": 13370
+    },
+    {
+      "epoch": 0.8753680078508341,
+      "grad_norm": 1.0147572755813599,
+      "learning_rate": 9.983828020902244e-05,
+      "loss": 0.1073,
+      "step": 13380
+    },
+    {
+      "epoch": 0.8760222440300949,
+      "grad_norm": 1.0348169803619385,
+      "learning_rate": 9.983754115119261e-05,
+      "loss": 0.1168,
+      "step": 13390
+    },
+    {
+      "epoch": 0.8766764802093556,
+      "grad_norm": 0.8402387499809265,
+      "learning_rate": 9.983680041121509e-05,
+      "loss": 0.1213,
+      "step": 13400
+    },
+    {
+      "epoch": 0.8773307163886163,
+      "grad_norm": 0.8802344799041748,
+      "learning_rate": 9.983605798911484e-05,
+      "loss": 0.1121,
+      "step": 13410
+    },
+    {
+      "epoch": 0.877984952567877,
+      "grad_norm": 0.7464808821678162,
+      "learning_rate": 9.983531388491691e-05,
+      "loss": 0.1184,
+      "step": 13420
+    },
+    {
+      "epoch": 0.8786391887471378,
+      "grad_norm": 0.8094585537910461,
+      "learning_rate": 9.983456809864646e-05,
+      "loss": 0.1138,
+      "step": 13430
+    },
+    {
+      "epoch": 0.8792934249263984,
+      "grad_norm": 0.7302922010421753,
+      "learning_rate": 9.983382063032864e-05,
+      "loss": 0.1124,
+      "step": 13440
+    },
+    {
+      "epoch": 0.8799476611056591,
+      "grad_norm": 0.8307011723518372,
+      "learning_rate": 9.983307147998868e-05,
+      "loss": 0.1155,
+      "step": 13450
+    },
+    {
+      "epoch": 0.8806018972849199,
+      "grad_norm": 0.8752557635307312,
+      "learning_rate": 9.983232064765187e-05,
+      "loss": 0.1097,
+      "step": 13460
+    },
+    {
+      "epoch": 0.8812561334641805,
+      "grad_norm": 0.9539555907249451,
+      "learning_rate": 9.983156813334354e-05,
+      "loss": 0.1194,
+      "step": 13470
+    },
+    {
+      "epoch": 0.8819103696434413,
+      "grad_norm": 0.9456961750984192,
+      "learning_rate": 9.983081393708911e-05,
+      "loss": 0.1129,
+      "step": 13480
+    },
+    {
+      "epoch": 0.882564605822702,
+      "grad_norm": 1.089903473854065,
+      "learning_rate": 9.983005805891401e-05,
+      "loss": 0.1114,
+      "step": 13490
+    },
+    {
+      "epoch": 0.8832188420019627,
+      "grad_norm": 0.9316903948783875,
+      "learning_rate": 9.982930049884377e-05,
+      "loss": 0.1132,
+      "step": 13500
+    },
+    {
+      "epoch": 0.8838730781812234,
+      "grad_norm": 0.8528069853782654,
+      "learning_rate": 9.982854125690395e-05,
+      "loss": 0.1188,
+      "step": 13510
+    },
+    {
+      "epoch": 0.8845273143604842,
+      "grad_norm": 0.8563151955604553,
+      "learning_rate": 9.982778033312019e-05,
+      "loss": 0.1122,
+      "step": 13520
+    },
+    {
+      "epoch": 0.8851815505397449,
+      "grad_norm": 0.7810546159744263,
+      "learning_rate": 9.982701772751816e-05,
+      "loss": 0.1076,
+      "step": 13530
+    },
+    {
+      "epoch": 0.8858357867190055,
+      "grad_norm": 0.7881696820259094,
+      "learning_rate": 9.982625344012361e-05,
+      "loss": 0.1293,
+      "step": 13540
+    },
+    {
+      "epoch": 0.8864900228982663,
+      "grad_norm": 0.7264038324356079,
+      "learning_rate": 9.982548747096235e-05,
+      "loss": 0.1156,
+      "step": 13550
+    },
+    {
+      "epoch": 0.887144259077527,
+      "grad_norm": 1.067252516746521,
+      "learning_rate": 9.982471982006019e-05,
+      "loss": 0.1142,
+      "step": 13560
+    },
+    {
+      "epoch": 0.8877984952567877,
+      "grad_norm": 0.9019017815589905,
+      "learning_rate": 9.982395048744307e-05,
+      "loss": 0.1066,
+      "step": 13570
+    },
+    {
+      "epoch": 0.8884527314360484,
+      "grad_norm": 1.020104169845581,
+      "learning_rate": 9.982317947313695e-05,
+      "loss": 0.1193,
+      "step": 13580
+    },
+    {
+      "epoch": 0.8891069676153092,
+      "grad_norm": 0.8307511806488037,
+      "learning_rate": 9.982240677716788e-05,
+      "loss": 0.1249,
+      "step": 13590
+    },
+    {
+      "epoch": 0.8897612037945698,
+      "grad_norm": 0.9394966959953308,
+      "learning_rate": 9.98216323995619e-05,
+      "loss": 0.1158,
+      "step": 13600
+    },
+    {
+      "epoch": 0.8904154399738305,
+      "grad_norm": 0.9355982542037964,
+      "learning_rate": 9.982085634034515e-05,
+      "loss": 0.1243,
+      "step": 13610
+    },
+    {
+      "epoch": 0.8910696761530913,
+      "grad_norm": 0.9257426857948303,
+      "learning_rate": 9.982007859954386e-05,
+      "loss": 0.1079,
+      "step": 13620
+    },
+    {
+      "epoch": 0.8917239123323519,
+      "grad_norm": 1.074717402458191,
+      "learning_rate": 9.981929917718426e-05,
+      "loss": 0.1144,
+      "step": 13630
+    },
+    {
+      "epoch": 0.8923781485116127,
+      "grad_norm": 0.8805826902389526,
+      "learning_rate": 9.981851807329264e-05,
+      "loss": 0.117,
+      "step": 13640
+    },
+    {
+      "epoch": 0.8930323846908734,
+      "grad_norm": 1.0279000997543335,
+      "learning_rate": 9.98177352878954e-05,
+      "loss": 0.1132,
+      "step": 13650
+    },
+    {
+      "epoch": 0.8936866208701341,
+      "grad_norm": 1.029099941253662,
+      "learning_rate": 9.981695082101893e-05,
+      "loss": 0.1125,
+      "step": 13660
+    },
+    {
+      "epoch": 0.8943408570493948,
+      "grad_norm": 1.044093132019043,
+      "learning_rate": 9.981616467268973e-05,
+      "loss": 0.1163,
+      "step": 13670
+    },
+    {
+      "epoch": 0.8949950932286556,
+      "grad_norm": 0.9173268675804138,
+      "learning_rate": 9.98153768429343e-05,
+      "loss": 0.1133,
+      "step": 13680
+    },
+    {
+      "epoch": 0.8956493294079163,
+      "grad_norm": 0.917991578578949,
+      "learning_rate": 9.981458733177928e-05,
+      "loss": 0.1043,
+      "step": 13690
+    },
+    {
+      "epoch": 0.8963035655871769,
+      "grad_norm": 0.9324220418930054,
+      "learning_rate": 9.981379613925129e-05,
+      "loss": 0.1145,
+      "step": 13700
+    },
+    {
+      "epoch": 0.8969578017664377,
+      "grad_norm": 0.9457996487617493,
+      "learning_rate": 9.981300326537704e-05,
+      "loss": 0.1145,
+      "step": 13710
+    },
+    {
+      "epoch": 0.8976120379456984,
+      "grad_norm": 0.9472202062606812,
+      "learning_rate": 9.981220871018329e-05,
+      "loss": 0.1176,
+      "step": 13720
+    },
+    {
+      "epoch": 0.8982662741249591,
+      "grad_norm": 0.8660651445388794,
+      "learning_rate": 9.981141247369685e-05,
+      "loss": 0.1161,
+      "step": 13730
+    },
+    {
+      "epoch": 0.8989205103042198,
+      "grad_norm": 0.9130904674530029,
+      "learning_rate": 9.981061455594461e-05,
+      "loss": 0.1095,
+      "step": 13740
+    },
+    {
+      "epoch": 0.8995747464834806,
+      "grad_norm": 1.0127148628234863,
+      "learning_rate": 9.980981495695349e-05,
+      "loss": 0.1115,
+      "step": 13750
+    },
+    {
+      "epoch": 0.9002289826627412,
+      "grad_norm": 0.8180867433547974,
+      "learning_rate": 9.980901367675048e-05,
+      "loss": 0.1051,
+      "step": 13760
+    },
+    {
+      "epoch": 0.900883218842002,
+      "grad_norm": 0.8807599544525146,
+      "learning_rate": 9.980821071536264e-05,
+      "loss": 0.1317,
+      "step": 13770
+    },
+    {
+      "epoch": 0.9015374550212627,
+      "grad_norm": 0.8612924218177795,
+      "learning_rate": 9.980740607281707e-05,
+      "loss": 0.1147,
+      "step": 13780
+    },
+    {
+      "epoch": 0.9021916912005233,
+      "grad_norm": 0.9025987386703491,
+      "learning_rate": 9.980659974914091e-05,
+      "loss": 0.1103,
+      "step": 13790
+    },
+    {
+      "epoch": 0.9028459273797841,
+      "grad_norm": 0.8930261731147766,
+      "learning_rate": 9.980579174436138e-05,
+      "loss": 0.1101,
+      "step": 13800
+    },
+    {
+      "epoch": 0.9035001635590448,
+      "grad_norm": 0.8868409395217896,
+      "learning_rate": 9.980498205850577e-05,
+      "loss": 0.1085,
+      "step": 13810
+    },
+    {
+      "epoch": 0.9041543997383056,
+      "grad_norm": 0.7958750128746033,
+      "learning_rate": 9.980417069160139e-05,
+      "loss": 0.1128,
+      "step": 13820
+    },
+    {
+      "epoch": 0.9048086359175662,
+      "grad_norm": 1.0320643186569214,
+      "learning_rate": 9.980335764367563e-05,
+      "loss": 0.1179,
+      "step": 13830
+    },
+    {
+      "epoch": 0.905462872096827,
+      "grad_norm": 0.9353842735290527,
+      "learning_rate": 9.980254291475595e-05,
+      "loss": 0.1012,
+      "step": 13840
+    },
+    {
+      "epoch": 0.9061171082760877,
+      "grad_norm": 0.8081122636795044,
+      "learning_rate": 9.980172650486983e-05,
+      "loss": 0.1129,
+      "step": 13850
+    },
+    {
+      "epoch": 0.9067713444553483,
+      "grad_norm": 0.7655879259109497,
+      "learning_rate": 9.980090841404482e-05,
+      "loss": 0.1043,
+      "step": 13860
+    },
+    {
+      "epoch": 0.9074255806346091,
+      "grad_norm": 0.886113703250885,
+      "learning_rate": 9.980008864230854e-05,
+      "loss": 0.1072,
+      "step": 13870
+    },
+    {
+      "epoch": 0.9080798168138698,
+      "grad_norm": 0.9072959423065186,
+      "learning_rate": 9.979926718968868e-05,
+      "loss": 0.1132,
+      "step": 13880
+    },
+    {
+      "epoch": 0.9087340529931305,
+      "grad_norm": 1.0498781204223633,
+      "learning_rate": 9.979844405621295e-05,
+      "loss": 0.1179,
+      "step": 13890
+    },
+    {
+      "epoch": 0.9093882891723912,
+      "grad_norm": 0.9316992163658142,
+      "learning_rate": 9.979761924190911e-05,
+      "loss": 0.1127,
+      "step": 13900
+    },
+    {
+      "epoch": 0.910042525351652,
+      "grad_norm": 0.9296565055847168,
+      "learning_rate": 9.979679274680504e-05,
+      "loss": 0.11,
+      "step": 13910
+    },
+    {
+      "epoch": 0.9106967615309126,
+      "grad_norm": 1.0889675617218018,
+      "learning_rate": 9.979596457092861e-05,
+      "loss": 0.1027,
+      "step": 13920
+    },
+    {
+      "epoch": 0.9113509977101734,
+      "grad_norm": 0.742987871170044,
+      "learning_rate": 9.979513471430779e-05,
+      "loss": 0.1047,
+      "step": 13930
+    },
+    {
+      "epoch": 0.9120052338894341,
+      "grad_norm": 0.8663235902786255,
+      "learning_rate": 9.979430317697056e-05,
+      "loss": 0.1226,
+      "step": 13940
+    },
+    {
+      "epoch": 0.9126594700686947,
+      "grad_norm": 1.0001239776611328,
+      "learning_rate": 9.979346995894504e-05,
+      "loss": 0.1174,
+      "step": 13950
+    },
+    {
+      "epoch": 0.9133137062479555,
+      "grad_norm": 1.0531165599822998,
+      "learning_rate": 9.979263506025929e-05,
+      "loss": 0.1148,
+      "step": 13960
+    },
+    {
+      "epoch": 0.9139679424272162,
+      "grad_norm": 0.8810727000236511,
+      "learning_rate": 9.979179848094153e-05,
+      "loss": 0.1153,
+      "step": 13970
+    },
+    {
+      "epoch": 0.914622178606477,
+      "grad_norm": 0.8835552930831909,
+      "learning_rate": 9.979096022102e-05,
+      "loss": 0.1089,
+      "step": 13980
+    },
+    {
+      "epoch": 0.9152764147857376,
+      "grad_norm": 0.9073536396026611,
+      "learning_rate": 9.979012028052297e-05,
+      "loss": 0.1254,
+      "step": 13990
+    },
+    {
+      "epoch": 0.9159306509649984,
+      "grad_norm": 0.8263078331947327,
+      "learning_rate": 9.97892786594788e-05,
+      "loss": 0.1104,
+      "step": 14000
+    },
+    {
+      "epoch": 0.9165848871442591,
+      "grad_norm": 0.9409777522087097,
+      "learning_rate": 9.978843535791588e-05,
+      "loss": 0.1119,
+      "step": 14010
+    },
+    {
+      "epoch": 0.9172391233235198,
+      "grad_norm": 0.8402562737464905,
+      "learning_rate": 9.978759037586272e-05,
+      "loss": 0.1134,
+      "step": 14020
+    },
+    {
+      "epoch": 0.9178933595027805,
+      "grad_norm": 0.8777570724487305,
+      "learning_rate": 9.978674371334782e-05,
+      "loss": 0.1077,
+      "step": 14030
+    },
+    {
+      "epoch": 0.9185475956820413,
+      "grad_norm": 1.0120798349380493,
+      "learning_rate": 9.978589537039972e-05,
+      "loss": 0.1179,
+      "step": 14040
+    },
+    {
+      "epoch": 0.9192018318613019,
+      "grad_norm": 0.7693895101547241,
+      "learning_rate": 9.97850453470471e-05,
+      "loss": 0.1248,
+      "step": 14050
+    },
+    {
+      "epoch": 0.9198560680405626,
+      "grad_norm": 0.8541615009307861,
+      "learning_rate": 9.978419364331863e-05,
+      "loss": 0.1226,
+      "step": 14060
+    },
+    {
+      "epoch": 0.9205103042198234,
+      "grad_norm": 1.0362035036087036,
+      "learning_rate": 9.978334025924307e-05,
+      "loss": 0.1274,
+      "step": 14070
+    },
+    {
+      "epoch": 0.921164540399084,
+      "grad_norm": 0.9056493639945984,
+      "learning_rate": 9.97824851948492e-05,
+      "loss": 0.1163,
+      "step": 14080
+    },
+    {
+      "epoch": 0.9218187765783448,
+      "grad_norm": 0.9539541602134705,
+      "learning_rate": 9.97816284501659e-05,
+      "loss": 0.1117,
+      "step": 14090
+    },
+    {
+      "epoch": 0.9224730127576055,
+      "grad_norm": 0.8763441443443298,
+      "learning_rate": 9.978077002522208e-05,
+      "loss": 0.1187,
+      "step": 14100
+    },
+    {
+      "epoch": 0.9231272489368663,
+      "grad_norm": 1.0391874313354492,
+      "learning_rate": 9.977990992004672e-05,
+      "loss": 0.1131,
+      "step": 14110
+    },
+    {
+      "epoch": 0.9237814851161269,
+      "grad_norm": 0.8817654252052307,
+      "learning_rate": 9.977904813466885e-05,
+      "loss": 0.1106,
+      "step": 14120
+    },
+    {
+      "epoch": 0.9244357212953876,
+      "grad_norm": 0.7447488903999329,
+      "learning_rate": 9.977818466911754e-05,
+      "loss": 0.1042,
+      "step": 14130
+    },
+    {
+      "epoch": 0.9250899574746484,
+      "grad_norm": 0.9143378734588623,
+      "learning_rate": 9.977731952342198e-05,
+      "loss": 0.1299,
+      "step": 14140
+    },
+    {
+      "epoch": 0.925744193653909,
+      "grad_norm": 0.9226405024528503,
+      "learning_rate": 9.97764526976113e-05,
+      "loss": 0.1058,
+      "step": 14150
+    },
+    {
+      "epoch": 0.9263984298331698,
+      "grad_norm": 0.7661489844322205,
+      "learning_rate": 9.977558419171485e-05,
+      "loss": 0.1109,
+      "step": 14160
+    },
+    {
+      "epoch": 0.9270526660124305,
+      "grad_norm": 0.8209551572799683,
+      "learning_rate": 9.977471400576185e-05,
+      "loss": 0.1158,
+      "step": 14170
+    },
+    {
+      "epoch": 0.9277069021916912,
+      "grad_norm": 0.9053903818130493,
+      "learning_rate": 9.977384213978173e-05,
+      "loss": 0.1041,
+      "step": 14180
+    },
+    {
+      "epoch": 0.9283611383709519,
+      "grad_norm": 0.9400367140769958,
+      "learning_rate": 9.97729685938039e-05,
+      "loss": 0.1077,
+      "step": 14190
+    },
+    {
+      "epoch": 0.9290153745502127,
+      "grad_norm": 0.9870936274528503,
+      "learning_rate": 9.977209336785783e-05,
+      "loss": 0.1053,
+      "step": 14200
+    },
+    {
+      "epoch": 0.9296696107294733,
+      "grad_norm": 0.9150540828704834,
+      "learning_rate": 9.977121646197309e-05,
+      "loss": 0.1115,
+      "step": 14210
+    },
+    {
+      "epoch": 0.930323846908734,
+      "grad_norm": 1.1972026824951172,
+      "learning_rate": 9.977033787617927e-05,
+      "loss": 0.1049,
+      "step": 14220
+    },
+    {
+      "epoch": 0.9309780830879948,
+      "grad_norm": 0.9222517013549805,
+      "learning_rate": 9.9769457610506e-05,
+      "loss": 0.1094,
+      "step": 14230
+    },
+    {
+      "epoch": 0.9316323192672554,
+      "grad_norm": 0.763886034488678,
+      "learning_rate": 9.976857566498303e-05,
+      "loss": 0.1083,
+      "step": 14240
+    },
+    {
+      "epoch": 0.9322865554465162,
+      "grad_norm": 0.8632875084877014,
+      "learning_rate": 9.976769203964011e-05,
+      "loss": 0.1082,
+      "step": 14250
+    },
+    {
+      "epoch": 0.9329407916257769,
+      "grad_norm": 1.0797144174575806,
+      "learning_rate": 9.976680673450704e-05,
+      "loss": 0.1119,
+      "step": 14260
+    },
+    {
+      "epoch": 0.9335950278050377,
+      "grad_norm": 0.7785837054252625,
+      "learning_rate": 9.976591974961376e-05,
+      "loss": 0.1061,
+      "step": 14270
+    },
+    {
+      "epoch": 0.9342492639842983,
+      "grad_norm": 0.9709704518318176,
+      "learning_rate": 9.976503108499014e-05,
+      "loss": 0.1245,
+      "step": 14280
+    },
+    {
+      "epoch": 0.934903500163559,
+      "grad_norm": 0.8645631670951843,
+      "learning_rate": 9.976414074066622e-05,
+      "loss": 0.1041,
+      "step": 14290
+    },
+    {
+      "epoch": 0.9355577363428198,
+      "grad_norm": 0.8737794756889343,
+      "learning_rate": 9.976324871667204e-05,
+      "loss": 0.1068,
+      "step": 14300
+    },
+    {
+      "epoch": 0.9362119725220804,
+      "grad_norm": 0.9471864104270935,
+      "learning_rate": 9.97623550130377e-05,
+      "loss": 0.1132,
+      "step": 14310
+    },
+    {
+      "epoch": 0.9368662087013412,
+      "grad_norm": 0.9491965770721436,
+      "learning_rate": 9.976145962979337e-05,
+      "loss": 0.1086,
+      "step": 14320
+    },
+    {
+      "epoch": 0.9375204448806019,
+      "grad_norm": 1.008070468902588,
+      "learning_rate": 9.976056256696928e-05,
+      "loss": 0.1133,
+      "step": 14330
+    },
+    {
+      "epoch": 0.9381746810598626,
+      "grad_norm": 0.8952319025993347,
+      "learning_rate": 9.975966382459572e-05,
+      "loss": 0.1053,
+      "step": 14340
+    },
+    {
+      "epoch": 0.9388289172391233,
+      "grad_norm": 0.8078186511993408,
+      "learning_rate": 9.975876340270298e-05,
+      "loss": 0.1079,
+      "step": 14350
+    },
+    {
+      "epoch": 0.9394831534183841,
+      "grad_norm": 1.273657202720642,
+      "learning_rate": 9.975786130132148e-05,
+      "loss": 0.1111,
+      "step": 14360
+    },
+    {
+      "epoch": 0.9401373895976447,
+      "grad_norm": 0.9750173091888428,
+      "learning_rate": 9.975695752048168e-05,
+      "loss": 0.1207,
+      "step": 14370
+    },
+    {
+      "epoch": 0.9407916257769054,
+      "grad_norm": 0.7701619267463684,
+      "learning_rate": 9.975605206021406e-05,
+      "loss": 0.1098,
+      "step": 14380
+    },
+    {
+      "epoch": 0.9414458619561662,
+      "grad_norm": 0.8776541352272034,
+      "learning_rate": 9.97551449205492e-05,
+      "loss": 0.1126,
+      "step": 14390
+    },
+    {
+      "epoch": 0.9421000981354269,
+      "grad_norm": 0.8586710691452026,
+      "learning_rate": 9.975423610151771e-05,
+      "loss": 0.1052,
+      "step": 14400
+    },
+    {
+      "epoch": 0.9427543343146876,
+      "grad_norm": 0.9918507933616638,
+      "learning_rate": 9.975332560315026e-05,
+      "loss": 0.1105,
+      "step": 14410
+    },
+    {
+      "epoch": 0.9434085704939483,
+      "grad_norm": 0.8050119280815125,
+      "learning_rate": 9.97524134254776e-05,
+      "loss": 0.1009,
+      "step": 14420
+    },
+    {
+      "epoch": 0.9440628066732091,
+      "grad_norm": 0.8901785612106323,
+      "learning_rate": 9.975149956853049e-05,
+      "loss": 0.1104,
+      "step": 14430
+    },
+    {
+      "epoch": 0.9447170428524697,
+      "grad_norm": 1.129379391670227,
+      "learning_rate": 9.975058403233981e-05,
+      "loss": 0.122,
+      "step": 14440
+    },
+    {
+      "epoch": 0.9453712790317305,
+      "grad_norm": 0.946667492389679,
+      "learning_rate": 9.974966681693642e-05,
+      "loss": 0.1093,
+      "step": 14450
+    },
+    {
+      "epoch": 0.9460255152109912,
+      "grad_norm": 1.037529468536377,
+      "learning_rate": 9.974874792235131e-05,
+      "loss": 0.12,
+      "step": 14460
+    },
+    {
+      "epoch": 0.9466797513902518,
+      "grad_norm": 0.9789792895317078,
+      "learning_rate": 9.97478273486155e-05,
+      "loss": 0.1238,
+      "step": 14470
+    },
+    {
+      "epoch": 0.9473339875695126,
+      "grad_norm": 0.9181520938873291,
+      "learning_rate": 9.974690509576005e-05,
+      "loss": 0.1214,
+      "step": 14480
+    },
+    {
+      "epoch": 0.9479882237487733,
+      "grad_norm": 0.8259325623512268,
+      "learning_rate": 9.974598116381608e-05,
+      "loss": 0.114,
+      "step": 14490
+    },
+    {
+      "epoch": 0.948642459928034,
+      "grad_norm": 0.8979782462120056,
+      "learning_rate": 9.974505555281476e-05,
+      "loss": 0.117,
+      "step": 14500
+    },
+    {
+      "epoch": 0.9492966961072947,
+      "grad_norm": 0.9124277234077454,
+      "learning_rate": 9.974412826278738e-05,
+      "loss": 0.1052,
+      "step": 14510
+    },
+    {
+      "epoch": 0.9499509322865555,
+      "grad_norm": 1.0142040252685547,
+      "learning_rate": 9.974319929376522e-05,
+      "loss": 0.1166,
+      "step": 14520
+    },
+    {
+      "epoch": 0.9506051684658161,
+      "grad_norm": 0.973535418510437,
+      "learning_rate": 9.974226864577961e-05,
+      "loss": 0.1059,
+      "step": 14530
+    },
+    {
+      "epoch": 0.9512594046450769,
+      "grad_norm": 0.8721091151237488,
+      "learning_rate": 9.974133631886198e-05,
+      "loss": 0.1119,
+      "step": 14540
+    },
+    {
+      "epoch": 0.9519136408243376,
+      "grad_norm": 1.0169627666473389,
+      "learning_rate": 9.97404023130438e-05,
+      "loss": 0.1231,
+      "step": 14550
+    },
+    {
+      "epoch": 0.9525678770035984,
+      "grad_norm": 0.861275315284729,
+      "learning_rate": 9.973946662835658e-05,
+      "loss": 0.1046,
+      "step": 14560
+    },
+    {
+      "epoch": 0.953222113182859,
+      "grad_norm": 1.012182593345642,
+      "learning_rate": 9.973852926483194e-05,
+      "loss": 0.112,
+      "step": 14570
+    },
+    {
+      "epoch": 0.9538763493621197,
+      "grad_norm": 0.804041862487793,
+      "learning_rate": 9.973759022250147e-05,
+      "loss": 0.1076,
+      "step": 14580
+    },
+    {
+      "epoch": 0.9545305855413805,
+      "grad_norm": 0.799505889415741,
+      "learning_rate": 9.97366495013969e-05,
+      "loss": 0.1137,
+      "step": 14590
+    },
+    {
+      "epoch": 0.9551848217206411,
+      "grad_norm": 0.8227983713150024,
+      "learning_rate": 9.973570710154998e-05,
+      "loss": 0.0982,
+      "step": 14600
+    },
+    {
+      "epoch": 0.9558390578999019,
+      "grad_norm": 0.9317682385444641,
+      "learning_rate": 9.973476302299249e-05,
+      "loss": 0.1103,
+      "step": 14610
+    },
+    {
+      "epoch": 0.9564932940791626,
+      "grad_norm": 0.8401086926460266,
+      "learning_rate": 9.973381726575632e-05,
+      "loss": 0.1188,
+      "step": 14620
+    },
+    {
+      "epoch": 0.9571475302584233,
+      "grad_norm": 0.9187564849853516,
+      "learning_rate": 9.97328698298734e-05,
+      "loss": 0.1242,
+      "step": 14630
+    },
+    {
+      "epoch": 0.957801766437684,
+      "grad_norm": 1.0764808654785156,
+      "learning_rate": 9.973192071537567e-05,
+      "loss": 0.112,
+      "step": 14640
+    },
+    {
+      "epoch": 0.9584560026169447,
+      "grad_norm": 0.947796642780304,
+      "learning_rate": 9.97309699222952e-05,
+      "loss": 0.1048,
+      "step": 14650
+    },
+    {
+      "epoch": 0.9591102387962054,
+      "grad_norm": 0.8203085660934448,
+      "learning_rate": 9.973001745066408e-05,
+      "loss": 0.1156,
+      "step": 14660
+    },
+    {
+      "epoch": 0.9597644749754661,
+      "grad_norm": 0.7854580879211426,
+      "learning_rate": 9.972906330051444e-05,
+      "loss": 0.1097,
+      "step": 14670
+    },
+    {
+      "epoch": 0.9604187111547269,
+      "grad_norm": 0.904597818851471,
+      "learning_rate": 9.97281074718785e-05,
+      "loss": 0.1062,
+      "step": 14680
+    },
+    {
+      "epoch": 0.9610729473339875,
+      "grad_norm": 0.8809137940406799,
+      "learning_rate": 9.972714996478851e-05,
+      "loss": 0.1027,
+      "step": 14690
+    },
+    {
+      "epoch": 0.9617271835132483,
+      "grad_norm": 0.8244183659553528,
+      "learning_rate": 9.972619077927679e-05,
+      "loss": 0.1032,
+      "step": 14700
+    },
+    {
+      "epoch": 0.962381419692509,
+      "grad_norm": 0.7864908576011658,
+      "learning_rate": 9.972522991537573e-05,
+      "loss": 0.0975,
+      "step": 14710
+    },
+    {
+      "epoch": 0.9630356558717698,
+      "grad_norm": 0.8401934504508972,
+      "learning_rate": 9.972426737311774e-05,
+      "loss": 0.1166,
+      "step": 14720
+    },
+    {
+      "epoch": 0.9636898920510304,
+      "grad_norm": 0.8961406350135803,
+      "learning_rate": 9.972330315253534e-05,
+      "loss": 0.1121,
+      "step": 14730
+    },
+    {
+      "epoch": 0.9643441282302911,
+      "grad_norm": 0.9105174541473389,
+      "learning_rate": 9.972233725366102e-05,
+      "loss": 0.1047,
+      "step": 14740
+    },
+    {
+      "epoch": 0.9649983644095519,
+      "grad_norm": 0.9146085977554321,
+      "learning_rate": 9.972136967652746e-05,
+      "loss": 0.1086,
+      "step": 14750
+    },
+    {
+      "epoch": 0.9656526005888125,
+      "grad_norm": 0.8616661429405212,
+      "learning_rate": 9.972040042116724e-05,
+      "loss": 0.1113,
+      "step": 14760
+    },
+    {
+      "epoch": 0.9663068367680733,
+      "grad_norm": 0.9188449382781982,
+      "learning_rate": 9.971942948761313e-05,
+      "loss": 0.1171,
+      "step": 14770
+    },
+    {
+      "epoch": 0.966961072947334,
+      "grad_norm": 0.9797625541687012,
+      "learning_rate": 9.971845687589786e-05,
+      "loss": 0.1142,
+      "step": 14780
+    },
+    {
+      "epoch": 0.9676153091265947,
+      "grad_norm": 0.9933854341506958,
+      "learning_rate": 9.97174825860543e-05,
+      "loss": 0.1118,
+      "step": 14790
+    },
+    {
+      "epoch": 0.9682695453058554,
+      "grad_norm": 0.9022259712219238,
+      "learning_rate": 9.971650661811529e-05,
+      "loss": 0.1073,
+      "step": 14800
+    },
+    {
+      "epoch": 0.9689237814851162,
+      "grad_norm": 0.824038028717041,
+      "learning_rate": 9.971552897211381e-05,
+      "loss": 0.1161,
+      "step": 14810
+    },
+    {
+      "epoch": 0.9695780176643768,
+      "grad_norm": 0.7811712622642517,
+      "learning_rate": 9.971454964808284e-05,
+      "loss": 0.1043,
+      "step": 14820
+    },
+    {
+      "epoch": 0.9702322538436375,
+      "grad_norm": 1.1899689435958862,
+      "learning_rate": 9.971356864605544e-05,
+      "loss": 0.1191,
+      "step": 14830
+    },
+    {
+      "epoch": 0.9708864900228983,
+      "grad_norm": 0.9993957281112671,
+      "learning_rate": 9.971258596606472e-05,
+      "loss": 0.1146,
+      "step": 14840
+    },
+    {
+      "epoch": 0.971540726202159,
+      "grad_norm": 0.8771776556968689,
+      "learning_rate": 9.971160160814386e-05,
+      "loss": 0.1151,
+      "step": 14850
+    },
+    {
+      "epoch": 0.9721949623814197,
+      "grad_norm": 0.9334481358528137,
+      "learning_rate": 9.971061557232606e-05,
+      "loss": 0.1068,
+      "step": 14860
+    },
+    {
+      "epoch": 0.9728491985606804,
+      "grad_norm": 0.9948738813400269,
+      "learning_rate": 9.970962785864461e-05,
+      "loss": 0.1052,
+      "step": 14870
+    },
+    {
+      "epoch": 0.9735034347399412,
+      "grad_norm": 0.8575040102005005,
+      "learning_rate": 9.970863846713286e-05,
+      "loss": 0.1025,
+      "step": 14880
+    },
+    {
+      "epoch": 0.9741576709192018,
+      "grad_norm": 1.019413948059082,
+      "learning_rate": 9.970764739782419e-05,
+      "loss": 0.1091,
+      "step": 14890
+    },
+    {
+      "epoch": 0.9748119070984625,
+      "grad_norm": 0.8023679852485657,
+      "learning_rate": 9.970665465075205e-05,
+      "loss": 0.1044,
+      "step": 14900
+    },
+    {
+      "epoch": 0.9754661432777233,
+      "grad_norm": 0.8333026766777039,
+      "learning_rate": 9.970566022594996e-05,
+      "loss": 0.1055,
+      "step": 14910
+    },
+    {
+      "epoch": 0.9761203794569839,
+      "grad_norm": 0.9163589477539062,
+      "learning_rate": 9.97046641234515e-05,
+      "loss": 0.0995,
+      "step": 14920
+    },
+    {
+      "epoch": 0.9767746156362447,
+      "grad_norm": 0.90887850522995,
+      "learning_rate": 9.970366634329024e-05,
+      "loss": 0.1066,
+      "step": 14930
+    },
+    {
+      "epoch": 0.9774288518155054,
+      "grad_norm": 1.0364896059036255,
+      "learning_rate": 9.970266688549991e-05,
+      "loss": 0.1136,
+      "step": 14940
+    },
+    {
+      "epoch": 0.9780830879947661,
+      "grad_norm": 1.5073118209838867,
+      "learning_rate": 9.970166575011422e-05,
+      "loss": 0.1048,
+      "step": 14950
+    },
+    {
+      "epoch": 0.9787373241740268,
+      "grad_norm": 0.8328016400337219,
+      "learning_rate": 9.970066293716695e-05,
+      "loss": 0.1149,
+      "step": 14960
+    },
+    {
+      "epoch": 0.9793915603532876,
+      "grad_norm": 0.8772666454315186,
+      "learning_rate": 9.969965844669197e-05,
+      "loss": 0.1238,
+      "step": 14970
+    },
+    {
+      "epoch": 0.9800457965325482,
+      "grad_norm": 1.083786964416504,
+      "learning_rate": 9.969865227872317e-05,
+      "loss": 0.1016,
+      "step": 14980
+    },
+    {
+      "epoch": 0.9807000327118089,
+      "grad_norm": 0.9459251165390015,
+      "learning_rate": 9.969764443329452e-05,
+      "loss": 0.1047,
+      "step": 14990
+    },
+    {
+      "epoch": 0.9813542688910697,
+      "grad_norm": 0.8524743914604187,
+      "learning_rate": 9.969663491044003e-05,
+      "loss": 0.1097,
+      "step": 15000
+    },
+    {
+      "epoch": 0.9820085050703304,
+      "grad_norm": 0.9077591896057129,
+      "learning_rate": 9.969562371019379e-05,
+      "loss": 0.1087,
+      "step": 15010
+    },
+    {
+      "epoch": 0.9826627412495911,
+      "grad_norm": 0.9452090263366699,
+      "learning_rate": 9.969461083258991e-05,
+      "loss": 0.11,
+      "step": 15020
+    },
+    {
+      "epoch": 0.9833169774288518,
+      "grad_norm": 0.9411560297012329,
+      "learning_rate": 9.969359627766258e-05,
+      "loss": 0.1065,
+      "step": 15030
+    },
+    {
+      "epoch": 0.9839712136081126,
+      "grad_norm": 0.8572501540184021,
+      "learning_rate": 9.969258004544606e-05,
+      "loss": 0.1103,
+      "step": 15040
+    },
+    {
+      "epoch": 0.9846254497873732,
+      "grad_norm": 1.1027193069458008,
+      "learning_rate": 9.969156213597464e-05,
+      "loss": 0.1198,
+      "step": 15050
+    },
+    {
+      "epoch": 0.985279685966634,
+      "grad_norm": 0.780197024345398,
+      "learning_rate": 9.969054254928267e-05,
+      "loss": 0.1102,
+      "step": 15060
+    },
+    {
+      "epoch": 0.9859339221458947,
+      "grad_norm": 0.8345792889595032,
+      "learning_rate": 9.968952128540456e-05,
+      "loss": 0.1166,
+      "step": 15070
+    },
+    {
+      "epoch": 0.9865881583251553,
+      "grad_norm": 0.8777825236320496,
+      "learning_rate": 9.968849834437481e-05,
+      "loss": 0.1066,
+      "step": 15080
+    },
+    {
+      "epoch": 0.9872423945044161,
+      "grad_norm": 0.933706521987915,
+      "learning_rate": 9.968747372622793e-05,
+      "loss": 0.1241,
+      "step": 15090
+    },
+    {
+      "epoch": 0.9878966306836768,
+      "grad_norm": 0.7910727858543396,
+      "learning_rate": 9.968644743099848e-05,
+      "loss": 0.1132,
+      "step": 15100
+    },
+    {
+      "epoch": 0.9885508668629375,
+      "grad_norm": 0.7466284036636353,
+      "learning_rate": 9.968541945872114e-05,
+      "loss": 0.1079,
+      "step": 15110
+    },
+    {
+      "epoch": 0.9892051030421982,
+      "grad_norm": 0.8247948884963989,
+      "learning_rate": 9.968438980943057e-05,
+      "loss": 0.1025,
+      "step": 15120
+    },
+    {
+      "epoch": 0.989859339221459,
+      "grad_norm": 0.8794564008712769,
+      "learning_rate": 9.968335848316157e-05,
+      "loss": 0.1155,
+      "step": 15130
+    },
+    {
+      "epoch": 0.9905135754007197,
+      "grad_norm": 0.794762909412384,
+      "learning_rate": 9.968232547994891e-05,
+      "loss": 0.1075,
+      "step": 15140
+    },
+    {
+      "epoch": 0.9911678115799804,
+      "grad_norm": 0.8396781086921692,
+      "learning_rate": 9.968129079982747e-05,
+      "loss": 0.1137,
+      "step": 15150
+    },
+    {
+      "epoch": 0.9918220477592411,
+      "grad_norm": 0.7761490941047668,
+      "learning_rate": 9.968025444283215e-05,
+      "loss": 0.1065,
+      "step": 15160
+    },
+    {
+      "epoch": 0.9924762839385018,
+      "grad_norm": 1.025817632675171,
+      "learning_rate": 9.967921640899797e-05,
+      "loss": 0.1023,
+      "step": 15170
+    },
+    {
+      "epoch": 0.9931305201177625,
+      "grad_norm": 0.8451741337776184,
+      "learning_rate": 9.967817669835995e-05,
+      "loss": 0.1135,
+      "step": 15180
+    },
+    {
+      "epoch": 0.9937847562970232,
+      "grad_norm": 0.7876104712486267,
+      "learning_rate": 9.967713531095317e-05,
+      "loss": 0.1145,
+      "step": 15190
+    },
+    {
+      "epoch": 0.994438992476284,
+      "grad_norm": 1.0193179845809937,
+      "learning_rate": 9.967609224681281e-05,
+      "loss": 0.1023,
+      "step": 15200
+    },
+    {
+      "epoch": 0.9950932286555446,
+      "grad_norm": 0.8599629998207092,
+      "learning_rate": 9.967504750597405e-05,
+      "loss": 0.1134,
+      "step": 15210
+    },
+    {
+      "epoch": 0.9957474648348054,
+      "grad_norm": 0.9450410604476929,
+      "learning_rate": 9.967400108847213e-05,
+      "loss": 0.1128,
+      "step": 15220
+    },
+    {
+      "epoch": 0.9964017010140661,
+      "grad_norm": 0.841467022895813,
+      "learning_rate": 9.967295299434243e-05,
+      "loss": 0.1072,
+      "step": 15230
+    },
+    {
+      "epoch": 0.9970559371933267,
+      "grad_norm": 0.7316820621490479,
+      "learning_rate": 9.967190322362029e-05,
+      "loss": 0.1059,
+      "step": 15240
+    },
+    {
+      "epoch": 0.9977101733725875,
+      "grad_norm": 0.8536782264709473,
+      "learning_rate": 9.967085177634115e-05,
+      "loss": 0.1026,
+      "step": 15250
+    },
+    {
+      "epoch": 0.9983644095518482,
+      "grad_norm": 0.8215930461883545,
+      "learning_rate": 9.966979865254047e-05,
+      "loss": 0.1131,
+      "step": 15260
+    },
+    {
+      "epoch": 0.9990186457311089,
+      "grad_norm": 1.0290292501449585,
+      "learning_rate": 9.966874385225385e-05,
+      "loss": 0.1151,
+      "step": 15270
+    },
+    {
+      "epoch": 0.9996728819103696,
+      "grad_norm": 1.0046546459197998,
+      "learning_rate": 9.966768737551685e-05,
+      "loss": 0.1078,
+      "step": 15280
+    },
+    {
+      "epoch": 1.0003271180896303,
+      "grad_norm": 1.0391733646392822,
+      "learning_rate": 9.966662922236515e-05,
+      "loss": 0.1175,
+      "step": 15290
+    },
+    {
+      "epoch": 1.000981354268891,
+      "grad_norm": 0.8953946828842163,
+      "learning_rate": 9.966556939283445e-05,
+      "loss": 0.1119,
+      "step": 15300
+    },
+    {
+      "epoch": 1.0016355904481518,
+      "grad_norm": 0.8209472298622131,
+      "learning_rate": 9.966450788696053e-05,
+      "loss": 0.1263,
+      "step": 15310
+    },
+    {
+      "epoch": 1.0022898266274125,
+      "grad_norm": 0.9228883981704712,
+      "learning_rate": 9.966344470477922e-05,
+      "loss": 0.1051,
+      "step": 15320
+    },
+    {
+      "epoch": 1.0029440628066733,
+      "grad_norm": 0.8755295276641846,
+      "learning_rate": 9.966237984632641e-05,
+      "loss": 0.105,
+      "step": 15330
+    },
+    {
+      "epoch": 1.003598298985934,
+      "grad_norm": 0.727872371673584,
+      "learning_rate": 9.966131331163803e-05,
+      "loss": 0.1011,
+      "step": 15340
+    },
+    {
+      "epoch": 1.0042525351651945,
+      "grad_norm": 0.8583409190177917,
+      "learning_rate": 9.96602451007501e-05,
+      "loss": 0.1096,
+      "step": 15350
+    },
+    {
+      "epoch": 1.0049067713444553,
+      "grad_norm": 0.971300482749939,
+      "learning_rate": 9.965917521369865e-05,
+      "loss": 0.1031,
+      "step": 15360
+    },
+    {
+      "epoch": 1.005561007523716,
+      "grad_norm": 0.8533384799957275,
+      "learning_rate": 9.96581036505198e-05,
+      "loss": 0.1199,
+      "step": 15370
+    },
+    {
+      "epoch": 1.0062152437029768,
+      "grad_norm": 0.8822680711746216,
+      "learning_rate": 9.96570304112497e-05,
+      "loss": 0.1087,
+      "step": 15380
+    },
+    {
+      "epoch": 1.0068694798822375,
+      "grad_norm": 0.714950442314148,
+      "learning_rate": 9.965595549592462e-05,
+      "loss": 0.112,
+      "step": 15390
+    },
+    {
+      "epoch": 1.0075237160614983,
+      "grad_norm": 0.9944178462028503,
+      "learning_rate": 9.96548789045808e-05,
+      "loss": 0.1072,
+      "step": 15400
+    },
+    {
+      "epoch": 1.008177952240759,
+      "grad_norm": 0.8967353105545044,
+      "learning_rate": 9.96538006372546e-05,
+      "loss": 0.1119,
+      "step": 15410
+    },
+    {
+      "epoch": 1.0088321884200195,
+      "grad_norm": 0.7786975502967834,
+      "learning_rate": 9.96527206939824e-05,
+      "loss": 0.1114,
+      "step": 15420
+    },
+    {
+      "epoch": 1.0094864245992803,
+      "grad_norm": 0.8993393182754517,
+      "learning_rate": 9.965163907480066e-05,
+      "loss": 0.1047,
+      "step": 15430
+    },
+    {
+      "epoch": 1.010140660778541,
+      "grad_norm": 0.8233774304389954,
+      "learning_rate": 9.965055577974588e-05,
+      "loss": 0.1033,
+      "step": 15440
+    },
+    {
+      "epoch": 1.0107948969578018,
+      "grad_norm": 0.7405421733856201,
+      "learning_rate": 9.964947080885464e-05,
+      "loss": 0.1214,
+      "step": 15450
+    },
+    {
+      "epoch": 1.0114491331370625,
+      "grad_norm": 0.7713890075683594,
+      "learning_rate": 9.964838416216354e-05,
+      "loss": 0.1069,
+      "step": 15460
+    },
+    {
+      "epoch": 1.0121033693163233,
+      "grad_norm": 0.7736397981643677,
+      "learning_rate": 9.964729583970927e-05,
+      "loss": 0.1251,
+      "step": 15470
+    },
+    {
+      "epoch": 1.0127576054955838,
+      "grad_norm": 0.7739368081092834,
+      "learning_rate": 9.964620584152857e-05,
+      "loss": 0.105,
+      "step": 15480
+    },
+    {
+      "epoch": 1.0134118416748445,
+      "grad_norm": 0.8225721120834351,
+      "learning_rate": 9.964511416765821e-05,
+      "loss": 0.1196,
+      "step": 15490
+    },
+    {
+      "epoch": 1.0140660778541053,
+      "grad_norm": 0.7423897981643677,
+      "learning_rate": 9.964402081813504e-05,
+      "loss": 0.1159,
+      "step": 15500
+    },
+    {
+      "epoch": 1.014720314033366,
+      "grad_norm": 0.8205366730690002,
+      "learning_rate": 9.9642925792996e-05,
+      "loss": 0.1092,
+      "step": 15510
+    },
+    {
+      "epoch": 1.0153745502126268,
+      "grad_norm": 0.9027040600776672,
+      "learning_rate": 9.964182909227799e-05,
+      "loss": 0.1122,
+      "step": 15520
+    },
+    {
+      "epoch": 1.0160287863918875,
+      "grad_norm": 0.7824887037277222,
+      "learning_rate": 9.964073071601808e-05,
+      "loss": 0.1046,
+      "step": 15530
+    },
+    {
+      "epoch": 1.0166830225711483,
+      "grad_norm": 0.9918166995048523,
+      "learning_rate": 9.963963066425331e-05,
+      "loss": 0.1141,
+      "step": 15540
+    },
+    {
+      "epoch": 1.0173372587504088,
+      "grad_norm": 0.9670895934104919,
+      "learning_rate": 9.963852893702081e-05,
+      "loss": 0.1055,
+      "step": 15550
+    },
+    {
+      "epoch": 1.0179914949296696,
+      "grad_norm": 0.8161757588386536,
+      "learning_rate": 9.96374255343578e-05,
+      "loss": 0.109,
+      "step": 15560
+    },
+    {
+      "epoch": 1.0186457311089303,
+      "grad_norm": 0.9068933129310608,
+      "learning_rate": 9.963632045630147e-05,
+      "loss": 0.1047,
+      "step": 15570
+    },
+    {
+      "epoch": 1.019299967288191,
+      "grad_norm": 0.902734100818634,
+      "learning_rate": 9.963521370288917e-05,
+      "loss": 0.1019,
+      "step": 15580
+    },
+    {
+      "epoch": 1.0199542034674518,
+      "grad_norm": 0.9012474417686462,
+      "learning_rate": 9.963410527415823e-05,
+      "loss": 0.1052,
+      "step": 15590
+    },
+    {
+      "epoch": 1.0206084396467126,
+      "grad_norm": 0.7533280253410339,
+      "learning_rate": 9.963299517014608e-05,
+      "loss": 0.1101,
+      "step": 15600
+    },
+    {
+      "epoch": 1.021262675825973,
+      "grad_norm": 0.8915255069732666,
+      "learning_rate": 9.963188339089015e-05,
+      "loss": 0.0986,
+      "step": 15610
+    },
+    {
+      "epoch": 1.0219169120052338,
+      "grad_norm": 0.8015055060386658,
+      "learning_rate": 9.963076993642802e-05,
+      "loss": 0.1187,
+      "step": 15620
+    },
+    {
+      "epoch": 1.0225711481844946,
+      "grad_norm": 0.9138506054878235,
+      "learning_rate": 9.962965480679721e-05,
+      "loss": 0.106,
+      "step": 15630
+    },
+    {
+      "epoch": 1.0232253843637553,
+      "grad_norm": 0.8131011724472046,
+      "learning_rate": 9.962853800203541e-05,
+      "loss": 0.1278,
+      "step": 15640
+    },
+    {
+      "epoch": 1.023879620543016,
+      "grad_norm": 0.8767716288566589,
+      "learning_rate": 9.96274195221803e-05,
+      "loss": 0.1054,
+      "step": 15650
+    },
+    {
+      "epoch": 1.0245338567222768,
+      "grad_norm": 0.8304327726364136,
+      "learning_rate": 9.962629936726962e-05,
+      "loss": 0.0943,
+      "step": 15660
+    },
+    {
+      "epoch": 1.0251880929015376,
+      "grad_norm": 0.8598764538764954,
+      "learning_rate": 9.96251775373412e-05,
+      "loss": 0.108,
+      "step": 15670
+    },
+    {
+      "epoch": 1.025842329080798,
+      "grad_norm": 0.8592802286148071,
+      "learning_rate": 9.962405403243287e-05,
+      "loss": 0.1186,
+      "step": 15680
+    },
+    {
+      "epoch": 1.0264965652600588,
+      "grad_norm": 0.9450241327285767,
+      "learning_rate": 9.962292885258259e-05,
+      "loss": 0.1196,
+      "step": 15690
+    },
+    {
+      "epoch": 1.0271508014393196,
+      "grad_norm": 0.8318309187889099,
+      "learning_rate": 9.962180199782831e-05,
+      "loss": 0.1173,
+      "step": 15700
+    },
+    {
+      "epoch": 1.0278050376185803,
+      "grad_norm": 0.8451339602470398,
+      "learning_rate": 9.962067346820808e-05,
+      "loss": 0.119,
+      "step": 15710
+    },
+    {
+      "epoch": 1.028459273797841,
+      "grad_norm": 0.8618002533912659,
+      "learning_rate": 9.961954326375998e-05,
+      "loss": 0.1103,
+      "step": 15720
+    },
+    {
+      "epoch": 1.0291135099771018,
+      "grad_norm": 0.7853378653526306,
+      "learning_rate": 9.961841138452217e-05,
+      "loss": 0.1144,
+      "step": 15730
+    },
+    {
+      "epoch": 1.0297677461563624,
+      "grad_norm": 0.8953865766525269,
+      "learning_rate": 9.961727783053285e-05,
+      "loss": 0.1148,
+      "step": 15740
+    },
+    {
+      "epoch": 1.030421982335623,
+      "grad_norm": 0.9555432200431824,
+      "learning_rate": 9.961614260183028e-05,
+      "loss": 0.1188,
+      "step": 15750
+    },
+    {
+      "epoch": 1.0310762185148838,
+      "grad_norm": 0.9691207408905029,
+      "learning_rate": 9.961500569845275e-05,
+      "loss": 0.1037,
+      "step": 15760
+    },
+    {
+      "epoch": 1.0317304546941446,
+      "grad_norm": 0.9222931265830994,
+      "learning_rate": 9.961386712043868e-05,
+      "loss": 0.1007,
+      "step": 15770
+    },
+    {
+      "epoch": 1.0323846908734053,
+      "grad_norm": 0.838117241859436,
+      "learning_rate": 9.961272686782646e-05,
+      "loss": 0.1034,
+      "step": 15780
+    },
+    {
+      "epoch": 1.033038927052666,
+      "grad_norm": 1.437838077545166,
+      "learning_rate": 9.961158494065461e-05,
+      "loss": 0.1285,
+      "step": 15790
+    },
+    {
+      "epoch": 1.0336931632319266,
+      "grad_norm": 0.8757314085960388,
+      "learning_rate": 9.961044133896166e-05,
+      "loss": 0.1098,
+      "step": 15800
+    },
+    {
+      "epoch": 1.0343473994111874,
+      "grad_norm": 0.8826759457588196,
+      "learning_rate": 9.96092960627862e-05,
+      "loss": 0.1102,
+      "step": 15810
+    },
+    {
+      "epoch": 1.035001635590448,
+      "grad_norm": 0.820978581905365,
+      "learning_rate": 9.96081491121669e-05,
+      "loss": 0.1053,
+      "step": 15820
+    },
+    {
+      "epoch": 1.0356558717697089,
+      "grad_norm": 0.8541067242622375,
+      "learning_rate": 9.960700048714244e-05,
+      "loss": 0.1115,
+      "step": 15830
+    },
+    {
+      "epoch": 1.0363101079489696,
+      "grad_norm": 0.7648676633834839,
+      "learning_rate": 9.960585018775164e-05,
+      "loss": 0.1015,
+      "step": 15840
+    },
+    {
+      "epoch": 1.0369643441282304,
+      "grad_norm": 0.8757931590080261,
+      "learning_rate": 9.960469821403329e-05,
+      "loss": 0.1006,
+      "step": 15850
+    },
+    {
+      "epoch": 1.037618580307491,
+      "grad_norm": 0.9553504586219788,
+      "learning_rate": 9.960354456602628e-05,
+      "loss": 0.1076,
+      "step": 15860
+    },
+    {
+      "epoch": 1.0382728164867516,
+      "grad_norm": 1.0022895336151123,
+      "learning_rate": 9.960238924376954e-05,
+      "loss": 0.1076,
+      "step": 15870
+    },
+    {
+      "epoch": 1.0389270526660124,
+      "grad_norm": 0.8987383842468262,
+      "learning_rate": 9.96012322473021e-05,
+      "loss": 0.112,
+      "step": 15880
+    },
+    {
+      "epoch": 1.0395812888452731,
+      "grad_norm": 0.8711419105529785,
+      "learning_rate": 9.960007357666297e-05,
+      "loss": 0.1046,
+      "step": 15890
+    },
+    {
+      "epoch": 1.0402355250245339,
+      "grad_norm": 0.6669361591339111,
+      "learning_rate": 9.95989132318913e-05,
+      "loss": 0.1021,
+      "step": 15900
+    },
+    {
+      "epoch": 1.0408897612037946,
+      "grad_norm": 0.7434557676315308,
+      "learning_rate": 9.959775121302621e-05,
+      "loss": 0.1104,
+      "step": 15910
+    },
+    {
+      "epoch": 1.0415439973830554,
+      "grad_norm": 0.8554294109344482,
+      "learning_rate": 9.959658752010695e-05,
+      "loss": 0.104,
+      "step": 15920
+    },
+    {
+      "epoch": 1.042198233562316,
+      "grad_norm": 0.9452110528945923,
+      "learning_rate": 9.959542215317278e-05,
+      "loss": 0.1057,
+      "step": 15930
+    },
+    {
+      "epoch": 1.0428524697415766,
+      "grad_norm": 0.9260740876197815,
+      "learning_rate": 9.959425511226304e-05,
+      "loss": 0.1065,
+      "step": 15940
+    },
+    {
+      "epoch": 1.0435067059208374,
+      "grad_norm": 0.939108669757843,
+      "learning_rate": 9.959308639741714e-05,
+      "loss": 0.1098,
+      "step": 15950
+    },
+    {
+      "epoch": 1.0441609421000981,
+      "grad_norm": 0.7533574104309082,
+      "learning_rate": 9.95919160086745e-05,
+      "loss": 0.11,
+      "step": 15960
+    },
+    {
+      "epoch": 1.0448151782793589,
+      "grad_norm": 0.7909335494041443,
+      "learning_rate": 9.959074394607464e-05,
+      "loss": 0.1032,
+      "step": 15970
+    },
+    {
+      "epoch": 1.0454694144586196,
+      "grad_norm": 0.6894845366477966,
+      "learning_rate": 9.958957020965712e-05,
+      "loss": 0.0998,
+      "step": 15980
+    },
+    {
+      "epoch": 1.0461236506378804,
+      "grad_norm": 0.8080494403839111,
+      "learning_rate": 9.958839479946154e-05,
+      "loss": 0.1073,
+      "step": 15990
+    },
+    {
+      "epoch": 1.046777886817141,
+      "grad_norm": 0.8611446022987366,
+      "learning_rate": 9.958721771552759e-05,
+      "loss": 0.1055,
+      "step": 16000
+    },
+    {
+      "epoch": 1.0474321229964016,
+      "grad_norm": 0.9711137413978577,
+      "learning_rate": 9.958603895789501e-05,
+      "loss": 0.1015,
+      "step": 16010
+    },
+    {
+      "epoch": 1.0480863591756624,
+      "grad_norm": 0.9332404732704163,
+      "learning_rate": 9.958485852660356e-05,
+      "loss": 0.1092,
+      "step": 16020
+    },
+    {
+      "epoch": 1.0487405953549231,
+      "grad_norm": 0.7202958464622498,
+      "learning_rate": 9.958367642169308e-05,
+      "loss": 0.0976,
+      "step": 16030
+    },
+    {
+      "epoch": 1.049394831534184,
+      "grad_norm": 0.9648202657699585,
+      "learning_rate": 9.958249264320349e-05,
+      "loss": 0.1103,
+      "step": 16040
+    },
+    {
+      "epoch": 1.0500490677134446,
+      "grad_norm": 0.8752630949020386,
+      "learning_rate": 9.958130719117475e-05,
+      "loss": 0.1095,
+      "step": 16050
+    },
+    {
+      "epoch": 1.0507033038927052,
+      "grad_norm": 1.0610864162445068,
+      "learning_rate": 9.958012006564686e-05,
+      "loss": 0.1026,
+      "step": 16060
+    },
+    {
+      "epoch": 1.051357540071966,
+      "grad_norm": 0.9164272546768188,
+      "learning_rate": 9.957893126665987e-05,
+      "loss": 0.112,
+      "step": 16070
+    },
+    {
+      "epoch": 1.0520117762512267,
+      "grad_norm": 0.8324456214904785,
+      "learning_rate": 9.957774079425395e-05,
+      "loss": 0.1134,
+      "step": 16080
+    },
+    {
+      "epoch": 1.0526660124304874,
+      "grad_norm": 1.0304049253463745,
+      "learning_rate": 9.957654864846924e-05,
+      "loss": 0.101,
+      "step": 16090
+    },
+    {
+      "epoch": 1.0533202486097482,
+      "grad_norm": 0.9455142021179199,
+      "learning_rate": 9.9575354829346e-05,
+      "loss": 0.1024,
+      "step": 16100
+    },
+    {
+      "epoch": 1.053974484789009,
+      "grad_norm": 1.337161660194397,
+      "learning_rate": 9.95741593369245e-05,
+      "loss": 0.1067,
+      "step": 16110
+    },
+    {
+      "epoch": 1.0546287209682697,
+      "grad_norm": 0.9535005688667297,
+      "learning_rate": 9.957296217124513e-05,
+      "loss": 0.1078,
+      "step": 16120
+    },
+    {
+      "epoch": 1.0552829571475302,
+      "grad_norm": 0.849690854549408,
+      "learning_rate": 9.957176333234828e-05,
+      "loss": 0.1083,
+      "step": 16130
+    },
+    {
+      "epoch": 1.055937193326791,
+      "grad_norm": 0.8453637361526489,
+      "learning_rate": 9.957056282027439e-05,
+      "loss": 0.105,
+      "step": 16140
+    },
+    {
+      "epoch": 1.0565914295060517,
+      "grad_norm": 0.9851493835449219,
+      "learning_rate": 9.956936063506402e-05,
+      "loss": 0.1029,
+      "step": 16150
+    },
+    {
+      "epoch": 1.0572456656853124,
+      "grad_norm": 0.8924856781959534,
+      "learning_rate": 9.956815677675772e-05,
+      "loss": 0.1031,
+      "step": 16160
+    },
+    {
+      "epoch": 1.0578999018645732,
+      "grad_norm": 1.1166882514953613,
+      "learning_rate": 9.956695124539613e-05,
+      "loss": 0.1058,
+      "step": 16170
+    },
+    {
+      "epoch": 1.058554138043834,
+      "grad_norm": 0.7543265223503113,
+      "learning_rate": 9.956574404101994e-05,
+      "loss": 0.0973,
+      "step": 16180
+    },
+    {
+      "epoch": 1.0592083742230944,
+      "grad_norm": 0.7494108080863953,
+      "learning_rate": 9.95645351636699e-05,
+      "loss": 0.104,
+      "step": 16190
+    },
+    {
+      "epoch": 1.0598626104023552,
+      "grad_norm": 0.9787102937698364,
+      "learning_rate": 9.956332461338683e-05,
+      "loss": 0.1027,
+      "step": 16200
+    },
+    {
+      "epoch": 1.060516846581616,
+      "grad_norm": 0.957390546798706,
+      "learning_rate": 9.956211239021154e-05,
+      "loss": 0.1086,
+      "step": 16210
+    },
+    {
+      "epoch": 1.0611710827608767,
+      "grad_norm": 1.1353367567062378,
+      "learning_rate": 9.9560898494185e-05,
+      "loss": 0.1035,
+      "step": 16220
+    },
+    {
+      "epoch": 1.0618253189401374,
+      "grad_norm": 1.0974490642547607,
+      "learning_rate": 9.955968292534814e-05,
+      "loss": 0.1087,
+      "step": 16230
+    },
+    {
+      "epoch": 1.0624795551193982,
+      "grad_norm": 1.101291537284851,
+      "learning_rate": 9.955846568374201e-05,
+      "loss": 0.1154,
+      "step": 16240
+    },
+    {
+      "epoch": 1.0631337912986587,
+      "grad_norm": 0.8823967576026917,
+      "learning_rate": 9.955724676940769e-05,
+      "loss": 0.1043,
+      "step": 16250
+    },
+    {
+      "epoch": 1.0637880274779195,
+      "grad_norm": 0.7924269437789917,
+      "learning_rate": 9.955602618238633e-05,
+      "loss": 0.1018,
+      "step": 16260
+    },
+    {
+      "epoch": 1.0644422636571802,
+      "grad_norm": 1.0062264204025269,
+      "learning_rate": 9.955480392271911e-05,
+      "loss": 0.1101,
+      "step": 16270
+    },
+    {
+      "epoch": 1.065096499836441,
+      "grad_norm": 0.8188719749450684,
+      "learning_rate": 9.95535799904473e-05,
+      "loss": 0.0963,
+      "step": 16280
+    },
+    {
+      "epoch": 1.0657507360157017,
+      "grad_norm": 1.1326643228530884,
+      "learning_rate": 9.955235438561222e-05,
+      "loss": 0.1052,
+      "step": 16290
+    },
+    {
+      "epoch": 1.0664049721949624,
+      "grad_norm": 0.7720603942871094,
+      "learning_rate": 9.95511271082552e-05,
+      "loss": 0.1034,
+      "step": 16300
+    },
+    {
+      "epoch": 1.0670592083742232,
+      "grad_norm": 0.8074955344200134,
+      "learning_rate": 9.954989815841771e-05,
+      "loss": 0.1159,
+      "step": 16310
+    },
+    {
+      "epoch": 1.0677134445534837,
+      "grad_norm": 0.7520632147789001,
+      "learning_rate": 9.954866753614118e-05,
+      "loss": 0.1035,
+      "step": 16320
+    },
+    {
+      "epoch": 1.0683676807327445,
+      "grad_norm": 0.8165426254272461,
+      "learning_rate": 9.95474352414672e-05,
+      "loss": 0.1079,
+      "step": 16330
+    },
+    {
+      "epoch": 1.0690219169120052,
+      "grad_norm": 0.8366930484771729,
+      "learning_rate": 9.954620127443733e-05,
+      "loss": 0.1116,
+      "step": 16340
+    },
+    {
+      "epoch": 1.069676153091266,
+      "grad_norm": 0.7152531147003174,
+      "learning_rate": 9.954496563509323e-05,
+      "loss": 0.0996,
+      "step": 16350
+    },
+    {
+      "epoch": 1.0703303892705267,
+      "grad_norm": 0.9950698614120483,
+      "learning_rate": 9.954372832347661e-05,
+      "loss": 0.1074,
+      "step": 16360
+    },
+    {
+      "epoch": 1.0709846254497875,
+      "grad_norm": 0.9378706812858582,
+      "learning_rate": 9.954248933962919e-05,
+      "loss": 0.1058,
+      "step": 16370
+    },
+    {
+      "epoch": 1.0716388616290482,
+      "grad_norm": 0.8824446201324463,
+      "learning_rate": 9.954124868359287e-05,
+      "loss": 0.1022,
+      "step": 16380
+    },
+    {
+      "epoch": 1.0722930978083087,
+      "grad_norm": 0.8328707218170166,
+      "learning_rate": 9.954000635540946e-05,
+      "loss": 0.1081,
+      "step": 16390
+    },
+    {
+      "epoch": 1.0729473339875695,
+      "grad_norm": 0.9733028411865234,
+      "learning_rate": 9.953876235512091e-05,
+      "loss": 0.1012,
+      "step": 16400
+    },
+    {
+      "epoch": 1.0736015701668302,
+      "grad_norm": 0.8833745718002319,
+      "learning_rate": 9.953751668276921e-05,
+      "loss": 0.1173,
+      "step": 16410
+    },
+    {
+      "epoch": 1.074255806346091,
+      "grad_norm": 0.9558156728744507,
+      "learning_rate": 9.953626933839641e-05,
+      "loss": 0.1014,
+      "step": 16420
+    },
+    {
+      "epoch": 1.0749100425253517,
+      "grad_norm": 0.9810806512832642,
+      "learning_rate": 9.95350203220446e-05,
+      "loss": 0.1074,
+      "step": 16430
+    },
+    {
+      "epoch": 1.0755642787046125,
+      "grad_norm": 0.810119092464447,
+      "learning_rate": 9.953376963375596e-05,
+      "loss": 0.1046,
+      "step": 16440
+    },
+    {
+      "epoch": 1.076218514883873,
+      "grad_norm": 0.9398297071456909,
+      "learning_rate": 9.953251727357267e-05,
+      "loss": 0.1014,
+      "step": 16450
+    },
+    {
+      "epoch": 1.0768727510631337,
+      "grad_norm": 0.7809659838676453,
+      "learning_rate": 9.953126324153701e-05,
+      "loss": 0.1171,
+      "step": 16460
+    },
+    {
+      "epoch": 1.0775269872423945,
+      "grad_norm": 1.0472731590270996,
+      "learning_rate": 9.953000753769135e-05,
+      "loss": 0.1062,
+      "step": 16470
+    },
+    {
+      "epoch": 1.0781812234216552,
+      "grad_norm": 1.0119749307632446,
+      "learning_rate": 9.9528750162078e-05,
+      "loss": 0.0979,
+      "step": 16480
+    },
+    {
+      "epoch": 1.078835459600916,
+      "grad_norm": 0.8965128064155579,
+      "learning_rate": 9.952749111473946e-05,
+      "loss": 0.1093,
+      "step": 16490
+    },
+    {
+      "epoch": 1.0794896957801767,
+      "grad_norm": 0.9507614970207214,
+      "learning_rate": 9.95262303957182e-05,
+      "loss": 0.0978,
+      "step": 16500
+    },
+    {
+      "epoch": 1.0801439319594373,
+      "grad_norm": 0.728331446647644,
+      "learning_rate": 9.952496800505679e-05,
+      "loss": 0.1016,
+      "step": 16510
+    },
+    {
+      "epoch": 1.080798168138698,
+      "grad_norm": 0.896195113658905,
+      "learning_rate": 9.952370394279781e-05,
+      "loss": 0.1106,
+      "step": 16520
+    },
+    {
+      "epoch": 1.0814524043179587,
+      "grad_norm": 0.9140453338623047,
+      "learning_rate": 9.952243820898395e-05,
+      "loss": 0.1173,
+      "step": 16530
+    },
+    {
+      "epoch": 1.0821066404972195,
+      "grad_norm": 0.798197329044342,
+      "learning_rate": 9.95211708036579e-05,
+      "loss": 0.095,
+      "step": 16540
+    },
+    {
+      "epoch": 1.0827608766764802,
+      "grad_norm": 0.7951470613479614,
+      "learning_rate": 9.951990172686248e-05,
+      "loss": 0.1061,
+      "step": 16550
+    },
+    {
+      "epoch": 1.083415112855741,
+      "grad_norm": 0.809076189994812,
+      "learning_rate": 9.951863097864052e-05,
+      "loss": 0.0965,
+      "step": 16560
+    },
+    {
+      "epoch": 1.0840693490350017,
+      "grad_norm": 0.8439357876777649,
+      "learning_rate": 9.951735855903488e-05,
+      "loss": 0.1074,
+      "step": 16570
+    },
+    {
+      "epoch": 1.0847235852142623,
+      "grad_norm": 1.051507830619812,
+      "learning_rate": 9.951608446808852e-05,
+      "loss": 0.1089,
+      "step": 16580
+    },
+    {
+      "epoch": 1.085377821393523,
+      "grad_norm": 0.9256449937820435,
+      "learning_rate": 9.951480870584445e-05,
+      "loss": 0.1052,
+      "step": 16590
+    },
+    {
+      "epoch": 1.0860320575727838,
+      "grad_norm": 0.7667525410652161,
+      "learning_rate": 9.951353127234574e-05,
+      "loss": 0.1128,
+      "step": 16600
+    },
+    {
+      "epoch": 1.0866862937520445,
+      "grad_norm": 0.821887731552124,
+      "learning_rate": 9.951225216763549e-05,
+      "loss": 0.1031,
+      "step": 16610
+    },
+    {
+      "epoch": 1.0873405299313053,
+      "grad_norm": 0.8057878017425537,
+      "learning_rate": 9.951097139175687e-05,
+      "loss": 0.1135,
+      "step": 16620
+    },
+    {
+      "epoch": 1.087994766110566,
+      "grad_norm": 0.9914917349815369,
+      "learning_rate": 9.950968894475313e-05,
+      "loss": 0.0971,
+      "step": 16630
+    },
+    {
+      "epoch": 1.0886490022898265,
+      "grad_norm": 0.8095641136169434,
+      "learning_rate": 9.950840482666755e-05,
+      "loss": 0.1118,
+      "step": 16640
+    },
+    {
+      "epoch": 1.0893032384690873,
+      "grad_norm": 0.8370475172996521,
+      "learning_rate": 9.950711903754345e-05,
+      "loss": 0.1098,
+      "step": 16650
+    },
+    {
+      "epoch": 1.089957474648348,
+      "grad_norm": 0.8619472980499268,
+      "learning_rate": 9.950583157742426e-05,
+      "loss": 0.1036,
+      "step": 16660
+    },
+    {
+      "epoch": 1.0906117108276088,
+      "grad_norm": 0.9595767855644226,
+      "learning_rate": 9.950454244635341e-05,
+      "loss": 0.1139,
+      "step": 16670
+    },
+    {
+      "epoch": 1.0912659470068695,
+      "grad_norm": 1.2058511972427368,
+      "learning_rate": 9.950325164437442e-05,
+      "loss": 0.1025,
+      "step": 16680
+    },
+    {
+      "epoch": 1.0919201831861303,
+      "grad_norm": 0.9937551617622375,
+      "learning_rate": 9.950195917153086e-05,
+      "loss": 0.1148,
+      "step": 16690
+    },
+    {
+      "epoch": 1.0925744193653908,
+      "grad_norm": 0.8375882506370544,
+      "learning_rate": 9.950066502786637e-05,
+      "loss": 0.0937,
+      "step": 16700
+    },
+    {
+      "epoch": 1.0932286555446515,
+      "grad_norm": 0.9841185212135315,
+      "learning_rate": 9.94993692134246e-05,
+      "loss": 0.1113,
+      "step": 16710
+    },
+    {
+      "epoch": 1.0938828917239123,
+      "grad_norm": 1.1531639099121094,
+      "learning_rate": 9.949807172824929e-05,
+      "loss": 0.1074,
+      "step": 16720
+    },
+    {
+      "epoch": 1.094537127903173,
+      "grad_norm": 0.8746184706687927,
+      "learning_rate": 9.949677257238428e-05,
+      "loss": 0.1053,
+      "step": 16730
+    },
+    {
+      "epoch": 1.0951913640824338,
+      "grad_norm": 1.0337003469467163,
+      "learning_rate": 9.949547174587337e-05,
+      "loss": 0.1127,
+      "step": 16740
+    },
+    {
+      "epoch": 1.0958456002616945,
+      "grad_norm": 0.9409461617469788,
+      "learning_rate": 9.949416924876047e-05,
+      "loss": 0.1085,
+      "step": 16750
+    },
+    {
+      "epoch": 1.0964998364409553,
+      "grad_norm": 0.7191810011863708,
+      "learning_rate": 9.949286508108957e-05,
+      "loss": 0.096,
+      "step": 16760
+    },
+    {
+      "epoch": 1.0971540726202158,
+      "grad_norm": 0.7731035351753235,
+      "learning_rate": 9.949155924290466e-05,
+      "loss": 0.102,
+      "step": 16770
+    },
+    {
+      "epoch": 1.0978083087994766,
+      "grad_norm": 0.9170941710472107,
+      "learning_rate": 9.949025173424984e-05,
+      "loss": 0.1042,
+      "step": 16780
+    },
+    {
+      "epoch": 1.0984625449787373,
+      "grad_norm": 0.8782985210418701,
+      "learning_rate": 9.948894255516923e-05,
+      "loss": 0.099,
+      "step": 16790
+    },
+    {
+      "epoch": 1.099116781157998,
+      "grad_norm": 0.9650111794471741,
+      "learning_rate": 9.948763170570702e-05,
+      "loss": 0.1106,
+      "step": 16800
+    },
+    {
+      "epoch": 1.0997710173372588,
+      "grad_norm": 1.0858937501907349,
+      "learning_rate": 9.948631918590746e-05,
+      "loss": 0.1097,
+      "step": 16810
+    },
+    {
+      "epoch": 1.1004252535165195,
+      "grad_norm": 0.8748927712440491,
+      "learning_rate": 9.948500499581484e-05,
+      "loss": 0.1124,
+      "step": 16820
+    },
+    {
+      "epoch": 1.1010794896957803,
+      "grad_norm": 0.9151651859283447,
+      "learning_rate": 9.94836891354735e-05,
+      "loss": 0.1082,
+      "step": 16830
+    },
+    {
+      "epoch": 1.1017337258750408,
+      "grad_norm": 0.9073200225830078,
+      "learning_rate": 9.948237160492791e-05,
+      "loss": 0.1106,
+      "step": 16840
+    },
+    {
+      "epoch": 1.1023879620543016,
+      "grad_norm": 0.8182057738304138,
+      "learning_rate": 9.948105240422247e-05,
+      "loss": 0.1098,
+      "step": 16850
+    },
+    {
+      "epoch": 1.1030421982335623,
+      "grad_norm": 0.7684637904167175,
+      "learning_rate": 9.947973153340178e-05,
+      "loss": 0.1042,
+      "step": 16860
+    },
+    {
+      "epoch": 1.103696434412823,
+      "grad_norm": 0.8614751100540161,
+      "learning_rate": 9.947840899251036e-05,
+      "loss": 0.0928,
+      "step": 16870
+    },
+    {
+      "epoch": 1.1043506705920838,
+      "grad_norm": 1.0662237405776978,
+      "learning_rate": 9.947708478159288e-05,
+      "loss": 0.1048,
+      "step": 16880
+    },
+    {
+      "epoch": 1.1050049067713446,
+      "grad_norm": 0.8045850396156311,
+      "learning_rate": 9.947575890069404e-05,
+      "loss": 0.1133,
+      "step": 16890
+    },
+    {
+      "epoch": 1.105659142950605,
+      "grad_norm": 0.9639503955841064,
+      "learning_rate": 9.947443134985857e-05,
+      "loss": 0.1188,
+      "step": 16900
+    },
+    {
+      "epoch": 1.1063133791298658,
+      "grad_norm": 0.8630536794662476,
+      "learning_rate": 9.94731021291313e-05,
+      "loss": 0.0985,
+      "step": 16910
+    },
+    {
+      "epoch": 1.1069676153091266,
+      "grad_norm": 0.8716172575950623,
+      "learning_rate": 9.947177123855708e-05,
+      "loss": 0.1138,
+      "step": 16920
+    },
+    {
+      "epoch": 1.1076218514883873,
+      "grad_norm": 0.8983464241027832,
+      "learning_rate": 9.947043867818084e-05,
+      "loss": 0.1056,
+      "step": 16930
+    },
+    {
+      "epoch": 1.108276087667648,
+      "grad_norm": 0.889821469783783,
+      "learning_rate": 9.946910444804755e-05,
+      "loss": 0.1008,
+      "step": 16940
+    },
+    {
+      "epoch": 1.1089303238469088,
+      "grad_norm": 0.8643634915351868,
+      "learning_rate": 9.946776854820224e-05,
+      "loss": 0.1147,
+      "step": 16950
+    },
+    {
+      "epoch": 1.1095845600261693,
+      "grad_norm": 0.8867340683937073,
+      "learning_rate": 9.946643097869002e-05,
+      "loss": 0.096,
+      "step": 16960
+    },
+    {
+      "epoch": 1.11023879620543,
+      "grad_norm": 0.859140932559967,
+      "learning_rate": 9.946509173955603e-05,
+      "loss": 0.1037,
+      "step": 16970
+    },
+    {
+      "epoch": 1.1108930323846908,
+      "grad_norm": 1.0756253004074097,
+      "learning_rate": 9.946375083084545e-05,
+      "loss": 0.11,
+      "step": 16980
+    },
+    {
+      "epoch": 1.1115472685639516,
+      "grad_norm": 0.7562744617462158,
+      "learning_rate": 9.946240825260356e-05,
+      "loss": 0.1174,
+      "step": 16990
+    },
+    {
+      "epoch": 1.1122015047432123,
+      "grad_norm": 0.8000214099884033,
+      "learning_rate": 9.946106400487567e-05,
+      "loss": 0.1084,
+      "step": 17000
+    },
+    {
+      "epoch": 1.112855740922473,
+      "grad_norm": 0.9482758641242981,
+      "learning_rate": 9.945971808770716e-05,
+      "loss": 0.1032,
+      "step": 17010
+    },
+    {
+      "epoch": 1.1135099771017338,
+      "grad_norm": 0.8377864360809326,
+      "learning_rate": 9.945837050114345e-05,
+      "loss": 0.0981,
+      "step": 17020
+    },
+    {
+      "epoch": 1.1141642132809944,
+      "grad_norm": 0.9127662181854248,
+      "learning_rate": 9.945702124523002e-05,
+      "loss": 0.1082,
+      "step": 17030
+    },
+    {
+      "epoch": 1.114818449460255,
+      "grad_norm": 1.0239959955215454,
+      "learning_rate": 9.945567032001243e-05,
+      "loss": 0.1188,
+      "step": 17040
+    },
+    {
+      "epoch": 1.1154726856395158,
+      "grad_norm": 1.2458797693252563,
+      "learning_rate": 9.945431772553626e-05,
+      "loss": 0.122,
+      "step": 17050
+    },
+    {
+      "epoch": 1.1161269218187766,
+      "grad_norm": 1.0552146434783936,
+      "learning_rate": 9.945296346184716e-05,
+      "loss": 0.1069,
+      "step": 17060
+    },
+    {
+      "epoch": 1.1167811579980373,
+      "grad_norm": 0.7637792825698853,
+      "learning_rate": 9.945160752899085e-05,
+      "loss": 0.0978,
+      "step": 17070
+    },
+    {
+      "epoch": 1.117435394177298,
+      "grad_norm": 1.0093464851379395,
+      "learning_rate": 9.94502499270131e-05,
+      "loss": 0.121,
+      "step": 17080
+    },
+    {
+      "epoch": 1.1180896303565586,
+      "grad_norm": 0.7700073719024658,
+      "learning_rate": 9.944889065595972e-05,
+      "loss": 0.1018,
+      "step": 17090
+    },
+    {
+      "epoch": 1.1187438665358194,
+      "grad_norm": 0.8178618550300598,
+      "learning_rate": 9.94475297158766e-05,
+      "loss": 0.096,
+      "step": 17100
+    },
+    {
+      "epoch": 1.1193981027150801,
+      "grad_norm": 0.8623060584068298,
+      "learning_rate": 9.944616710680967e-05,
+      "loss": 0.1043,
+      "step": 17110
+    },
+    {
+      "epoch": 1.1200523388943409,
+      "grad_norm": 0.7678028345108032,
+      "learning_rate": 9.944480282880493e-05,
+      "loss": 0.1027,
+      "step": 17120
+    },
+    {
+      "epoch": 1.1207065750736016,
+      "grad_norm": 0.92671138048172,
+      "learning_rate": 9.944343688190842e-05,
+      "loss": 0.0993,
+      "step": 17130
+    },
+    {
+      "epoch": 1.1213608112528624,
+      "grad_norm": 0.9388012290000916,
+      "learning_rate": 9.944206926616624e-05,
+      "loss": 0.1094,
+      "step": 17140
+    },
+    {
+      "epoch": 1.1220150474321229,
+      "grad_norm": 0.7402985095977783,
+      "learning_rate": 9.944069998162455e-05,
+      "loss": 0.1004,
+      "step": 17150
+    },
+    {
+      "epoch": 1.1226692836113836,
+      "grad_norm": 0.8221508860588074,
+      "learning_rate": 9.943932902832959e-05,
+      "loss": 0.1123,
+      "step": 17160
+    },
+    {
+      "epoch": 1.1233235197906444,
+      "grad_norm": 0.8503665924072266,
+      "learning_rate": 9.94379564063276e-05,
+      "loss": 0.1153,
+      "step": 17170
+    },
+    {
+      "epoch": 1.1239777559699051,
+      "grad_norm": 0.7901976704597473,
+      "learning_rate": 9.943658211566493e-05,
+      "loss": 0.0984,
+      "step": 17180
+    },
+    {
+      "epoch": 1.1246319921491659,
+      "grad_norm": 0.8279184103012085,
+      "learning_rate": 9.943520615638797e-05,
+      "loss": 0.116,
+      "step": 17190
+    },
+    {
+      "epoch": 1.1252862283284266,
+      "grad_norm": 0.9072709679603577,
+      "learning_rate": 9.943382852854313e-05,
+      "loss": 0.1076,
+      "step": 17200
+    },
+    {
+      "epoch": 1.1259404645076874,
+      "grad_norm": 0.9476553797721863,
+      "learning_rate": 9.943244923217695e-05,
+      "loss": 0.102,
+      "step": 17210
+    },
+    {
+      "epoch": 1.126594700686948,
+      "grad_norm": 0.8165683746337891,
+      "learning_rate": 9.943106826733597e-05,
+      "loss": 0.1099,
+      "step": 17220
+    },
+    {
+      "epoch": 1.1272489368662086,
+      "grad_norm": 0.9307758212089539,
+      "learning_rate": 9.942968563406679e-05,
+      "loss": 0.0967,
+      "step": 17230
+    },
+    {
+      "epoch": 1.1279031730454694,
+      "grad_norm": 0.8825275897979736,
+      "learning_rate": 9.942830133241609e-05,
+      "loss": 0.1106,
+      "step": 17240
+    },
+    {
+      "epoch": 1.1285574092247301,
+      "grad_norm": 0.6875970959663391,
+      "learning_rate": 9.942691536243058e-05,
+      "loss": 0.1083,
+      "step": 17250
+    },
+    {
+      "epoch": 1.1292116454039909,
+      "grad_norm": 0.8216149806976318,
+      "learning_rate": 9.942552772415706e-05,
+      "loss": 0.1088,
+      "step": 17260
+    },
+    {
+      "epoch": 1.1298658815832516,
+      "grad_norm": 0.7268308997154236,
+      "learning_rate": 9.942413841764235e-05,
+      "loss": 0.0968,
+      "step": 17270
+    },
+    {
+      "epoch": 1.1305201177625124,
+      "grad_norm": 0.8889786601066589,
+      "learning_rate": 9.942274744293336e-05,
+      "loss": 0.0991,
+      "step": 17280
+    },
+    {
+      "epoch": 1.131174353941773,
+      "grad_norm": 0.8577876091003418,
+      "learning_rate": 9.942135480007701e-05,
+      "loss": 0.1016,
+      "step": 17290
+    },
+    {
+      "epoch": 1.1318285901210337,
+      "grad_norm": 0.9688170552253723,
+      "learning_rate": 9.941996048912035e-05,
+      "loss": 0.106,
+      "step": 17300
+    },
+    {
+      "epoch": 1.1324828263002944,
+      "grad_norm": 0.9323129653930664,
+      "learning_rate": 9.941856451011039e-05,
+      "loss": 0.1081,
+      "step": 17310
+    },
+    {
+      "epoch": 1.1331370624795551,
+      "grad_norm": 0.8865740895271301,
+      "learning_rate": 9.941716686309428e-05,
+      "loss": 0.1051,
+      "step": 17320
+    },
+    {
+      "epoch": 1.133791298658816,
+      "grad_norm": 0.912921667098999,
+      "learning_rate": 9.941576754811919e-05,
+      "loss": 0.1075,
+      "step": 17330
+    },
+    {
+      "epoch": 1.1344455348380766,
+      "grad_norm": 0.7206995487213135,
+      "learning_rate": 9.941436656523236e-05,
+      "loss": 0.107,
+      "step": 17340
+    },
+    {
+      "epoch": 1.1350997710173372,
+      "grad_norm": 0.8259555697441101,
+      "learning_rate": 9.941296391448103e-05,
+      "loss": 0.1092,
+      "step": 17350
+    },
+    {
+      "epoch": 1.135754007196598,
+      "grad_norm": 0.7500777244567871,
+      "learning_rate": 9.94115595959126e-05,
+      "loss": 0.1002,
+      "step": 17360
+    },
+    {
+      "epoch": 1.1364082433758587,
+      "grad_norm": 1.0846604108810425,
+      "learning_rate": 9.941015360957445e-05,
+      "loss": 0.1082,
+      "step": 17370
+    },
+    {
+      "epoch": 1.1370624795551194,
+      "grad_norm": 0.8614764213562012,
+      "learning_rate": 9.940874595551404e-05,
+      "loss": 0.1099,
+      "step": 17380
+    },
+    {
+      "epoch": 1.1377167157343802,
+      "grad_norm": 1.0448957681655884,
+      "learning_rate": 9.940733663377885e-05,
+      "loss": 0.1172,
+      "step": 17390
+    },
+    {
+      "epoch": 1.138370951913641,
+      "grad_norm": 0.8377976417541504,
+      "learning_rate": 9.940592564441649e-05,
+      "loss": 0.1061,
+      "step": 17400
+    },
+    {
+      "epoch": 1.1390251880929014,
+      "grad_norm": 1.1926016807556152,
+      "learning_rate": 9.940451298747456e-05,
+      "loss": 0.1077,
+      "step": 17410
+    },
+    {
+      "epoch": 1.1396794242721622,
+      "grad_norm": 0.9103050231933594,
+      "learning_rate": 9.940309866300075e-05,
+      "loss": 0.1148,
+      "step": 17420
+    },
+    {
+      "epoch": 1.140333660451423,
+      "grad_norm": 0.8122014999389648,
+      "learning_rate": 9.940168267104279e-05,
+      "loss": 0.0984,
+      "step": 17430
+    },
+    {
+      "epoch": 1.1409878966306837,
+      "grad_norm": 0.8694342374801636,
+      "learning_rate": 9.94002650116485e-05,
+      "loss": 0.0997,
+      "step": 17440
+    },
+    {
+      "epoch": 1.1416421328099444,
+      "grad_norm": 0.8882526159286499,
+      "learning_rate": 9.939884568486571e-05,
+      "loss": 0.1021,
+      "step": 17450
+    },
+    {
+      "epoch": 1.1422963689892052,
+      "grad_norm": 0.85523521900177,
+      "learning_rate": 9.939742469074229e-05,
+      "loss": 0.12,
+      "step": 17460
+    },
+    {
+      "epoch": 1.142950605168466,
+      "grad_norm": 0.9386106729507446,
+      "learning_rate": 9.939600202932626e-05,
+      "loss": 0.11,
+      "step": 17470
+    },
+    {
+      "epoch": 1.1436048413477264,
+      "grad_norm": 1.0228506326675415,
+      "learning_rate": 9.939457770066563e-05,
+      "loss": 0.0956,
+      "step": 17480
+    },
+    {
+      "epoch": 1.1442590775269872,
+      "grad_norm": 0.8513170480728149,
+      "learning_rate": 9.939315170480843e-05,
+      "loss": 0.1032,
+      "step": 17490
+    },
+    {
+      "epoch": 1.144913313706248,
+      "grad_norm": 0.8414477109909058,
+      "learning_rate": 9.939172404180284e-05,
+      "loss": 0.1,
+      "step": 17500
+    },
+    {
+      "epoch": 1.1455675498855087,
+      "grad_norm": 0.9836424589157104,
+      "learning_rate": 9.939029471169703e-05,
+      "loss": 0.0972,
+      "step": 17510
+    },
+    {
+      "epoch": 1.1462217860647694,
+      "grad_norm": 0.6796431541442871,
+      "learning_rate": 9.938886371453924e-05,
+      "loss": 0.103,
+      "step": 17520
+    },
+    {
+      "epoch": 1.1468760222440302,
+      "grad_norm": 0.7944294810295105,
+      "learning_rate": 9.938743105037777e-05,
+      "loss": 0.1015,
+      "step": 17530
+    },
+    {
+      "epoch": 1.147530258423291,
+      "grad_norm": 0.9507654309272766,
+      "learning_rate": 9.938599671926097e-05,
+      "loss": 0.1041,
+      "step": 17540
+    },
+    {
+      "epoch": 1.1481844946025515,
+      "grad_norm": 0.8638071417808533,
+      "learning_rate": 9.938456072123727e-05,
+      "loss": 0.1065,
+      "step": 17550
+    },
+    {
+      "epoch": 1.1488387307818122,
+      "grad_norm": 0.8793585896492004,
+      "learning_rate": 9.938312305635514e-05,
+      "loss": 0.0983,
+      "step": 17560
+    },
+    {
+      "epoch": 1.149492966961073,
+      "grad_norm": 0.8482651114463806,
+      "learning_rate": 9.938168372466307e-05,
+      "loss": 0.1042,
+      "step": 17570
+    },
+    {
+      "epoch": 1.1501472031403337,
+      "grad_norm": 0.889570415019989,
+      "learning_rate": 9.93802427262097e-05,
+      "loss": 0.1028,
+      "step": 17580
+    },
+    {
+      "epoch": 1.1508014393195944,
+      "grad_norm": 1.002482295036316,
+      "learning_rate": 9.93788000610436e-05,
+      "loss": 0.1081,
+      "step": 17590
+    },
+    {
+      "epoch": 1.151455675498855,
+      "grad_norm": 0.8683956265449524,
+      "learning_rate": 9.937735572921352e-05,
+      "loss": 0.1026,
+      "step": 17600
+    },
+    {
+      "epoch": 1.1521099116781157,
+      "grad_norm": 0.8563847541809082,
+      "learning_rate": 9.937590973076818e-05,
+      "loss": 0.1024,
+      "step": 17610
+    },
+    {
+      "epoch": 1.1527641478573765,
+      "grad_norm": 1.01363205909729,
+      "learning_rate": 9.937446206575639e-05,
+      "loss": 0.1075,
+      "step": 17620
+    },
+    {
+      "epoch": 1.1534183840366372,
+      "grad_norm": 0.998652458190918,
+      "learning_rate": 9.937301273422703e-05,
+      "loss": 0.1069,
+      "step": 17630
+    },
+    {
+      "epoch": 1.154072620215898,
+      "grad_norm": 0.9310935139656067,
+      "learning_rate": 9.937156173622899e-05,
+      "loss": 0.1146,
+      "step": 17640
+    },
+    {
+      "epoch": 1.1547268563951587,
+      "grad_norm": 0.9854075312614441,
+      "learning_rate": 9.937010907181125e-05,
+      "loss": 0.1147,
+      "step": 17650
+    },
+    {
+      "epoch": 1.1553810925744195,
+      "grad_norm": 1.0847511291503906,
+      "learning_rate": 9.936865474102289e-05,
+      "loss": 0.1066,
+      "step": 17660
+    },
+    {
+      "epoch": 1.15603532875368,
+      "grad_norm": 1.0146312713623047,
+      "learning_rate": 9.936719874391291e-05,
+      "loss": 0.1077,
+      "step": 17670
+    },
+    {
+      "epoch": 1.1566895649329407,
+      "grad_norm": 1.1995619535446167,
+      "learning_rate": 9.936574108053054e-05,
+      "loss": 0.1099,
+      "step": 17680
+    },
+    {
+      "epoch": 1.1573438011122015,
+      "grad_norm": 0.8147319555282593,
+      "learning_rate": 9.936428175092491e-05,
+      "loss": 0.1107,
+      "step": 17690
+    },
+    {
+      "epoch": 1.1579980372914622,
+      "grad_norm": 1.276146650314331,
+      "learning_rate": 9.936282075514534e-05,
+      "loss": 0.1206,
+      "step": 17700
+    },
+    {
+      "epoch": 1.158652273470723,
+      "grad_norm": 0.8899917602539062,
+      "learning_rate": 9.93613580932411e-05,
+      "loss": 0.1001,
+      "step": 17710
+    },
+    {
+      "epoch": 1.1593065096499837,
+      "grad_norm": 1.1056793928146362,
+      "learning_rate": 9.935989376526156e-05,
+      "loss": 0.1106,
+      "step": 17720
+    },
+    {
+      "epoch": 1.1599607458292445,
+      "grad_norm": 0.9014807343482971,
+      "learning_rate": 9.935842777125615e-05,
+      "loss": 0.1002,
+      "step": 17730
+    },
+    {
+      "epoch": 1.160614982008505,
+      "grad_norm": 0.9449447393417358,
+      "learning_rate": 9.935696011127438e-05,
+      "loss": 0.1039,
+      "step": 17740
+    },
+    {
+      "epoch": 1.1612692181877657,
+      "grad_norm": 0.8191094994544983,
+      "learning_rate": 9.935549078536574e-05,
+      "loss": 0.1099,
+      "step": 17750
+    },
+    {
+      "epoch": 1.1619234543670265,
+      "grad_norm": 0.7725281715393066,
+      "learning_rate": 9.935401979357986e-05,
+      "loss": 0.096,
+      "step": 17760
+    },
+    {
+      "epoch": 1.1625776905462872,
+      "grad_norm": 0.8660596609115601,
+      "learning_rate": 9.935254713596637e-05,
+      "loss": 0.1046,
+      "step": 17770
+    },
+    {
+      "epoch": 1.163231926725548,
+      "grad_norm": 1.0237655639648438,
+      "learning_rate": 9.935107281257498e-05,
+      "loss": 0.1121,
+      "step": 17780
+    },
+    {
+      "epoch": 1.1638861629048087,
+      "grad_norm": 1.306152582168579,
+      "learning_rate": 9.934959682345546e-05,
+      "loss": 0.1007,
+      "step": 17790
+    },
+    {
+      "epoch": 1.1645403990840693,
+      "grad_norm": 0.8070068359375,
+      "learning_rate": 9.934811916865763e-05,
+      "loss": 0.1075,
+      "step": 17800
+    },
+    {
+      "epoch": 1.16519463526333,
+      "grad_norm": 0.9843347668647766,
+      "learning_rate": 9.934663984823133e-05,
+      "loss": 0.1066,
+      "step": 17810
+    },
+    {
+      "epoch": 1.1658488714425908,
+      "grad_norm": 0.8091020584106445,
+      "learning_rate": 9.934515886222655e-05,
+      "loss": 0.0987,
+      "step": 17820
+    },
+    {
+      "epoch": 1.1665031076218515,
+      "grad_norm": 0.9831781983375549,
+      "learning_rate": 9.934367621069322e-05,
+      "loss": 0.1053,
+      "step": 17830
+    },
+    {
+      "epoch": 1.1671573438011122,
+      "grad_norm": 0.9934841394424438,
+      "learning_rate": 9.934219189368143e-05,
+      "loss": 0.1143,
+      "step": 17840
+    },
+    {
+      "epoch": 1.167811579980373,
+      "grad_norm": 0.9186519980430603,
+      "learning_rate": 9.934070591124127e-05,
+      "loss": 0.108,
+      "step": 17850
+    },
+    {
+      "epoch": 1.1684658161596335,
+      "grad_norm": 0.9979751706123352,
+      "learning_rate": 9.933921826342286e-05,
+      "loss": 0.1146,
+      "step": 17860
+    },
+    {
+      "epoch": 1.1691200523388943,
+      "grad_norm": 0.8182001709938049,
+      "learning_rate": 9.933772895027644e-05,
+      "loss": 0.0996,
+      "step": 17870
+    },
+    {
+      "epoch": 1.169774288518155,
+      "grad_norm": 0.9773245453834534,
+      "learning_rate": 9.933623797185228e-05,
+      "loss": 0.1056,
+      "step": 17880
+    },
+    {
+      "epoch": 1.1704285246974158,
+      "grad_norm": 0.8775014281272888,
+      "learning_rate": 9.933474532820071e-05,
+      "loss": 0.1121,
+      "step": 17890
+    },
+    {
+      "epoch": 1.1710827608766765,
+      "grad_norm": 0.8223180174827576,
+      "learning_rate": 9.933325101937207e-05,
+      "loss": 0.0976,
+      "step": 17900
+    },
+    {
+      "epoch": 1.1717369970559373,
+      "grad_norm": 0.8393526673316956,
+      "learning_rate": 9.933175504541686e-05,
+      "loss": 0.1034,
+      "step": 17910
+    },
+    {
+      "epoch": 1.172391233235198,
+      "grad_norm": 0.8224334716796875,
+      "learning_rate": 9.933025740638551e-05,
+      "loss": 0.1118,
+      "step": 17920
+    },
+    {
+      "epoch": 1.1730454694144585,
+      "grad_norm": 0.7815266847610474,
+      "learning_rate": 9.932875810232863e-05,
+      "loss": 0.1002,
+      "step": 17930
+    },
+    {
+      "epoch": 1.1736997055937193,
+      "grad_norm": 0.9696399569511414,
+      "learning_rate": 9.932725713329678e-05,
+      "loss": 0.1071,
+      "step": 17940
+    },
+    {
+      "epoch": 1.17435394177298,
+      "grad_norm": 0.9797826409339905,
+      "learning_rate": 9.932575449934062e-05,
+      "loss": 0.1058,
+      "step": 17950
+    },
+    {
+      "epoch": 1.1750081779522408,
+      "grad_norm": 0.90691077709198,
+      "learning_rate": 9.93242502005109e-05,
+      "loss": 0.1041,
+      "step": 17960
+    },
+    {
+      "epoch": 1.1756624141315015,
+      "grad_norm": 0.8655257821083069,
+      "learning_rate": 9.932274423685839e-05,
+      "loss": 0.1105,
+      "step": 17970
+    },
+    {
+      "epoch": 1.1763166503107623,
+      "grad_norm": 0.9878515005111694,
+      "learning_rate": 9.932123660843389e-05,
+      "loss": 0.119,
+      "step": 17980
+    },
+    {
+      "epoch": 1.176970886490023,
+      "grad_norm": 1.0736156702041626,
+      "learning_rate": 9.931972731528831e-05,
+      "loss": 0.1112,
+      "step": 17990
+    },
+    {
+      "epoch": 1.1776251226692835,
+      "grad_norm": 0.9092381596565247,
+      "learning_rate": 9.931821635747259e-05,
+      "loss": 0.1154,
+      "step": 18000
+    },
+    {
+      "epoch": 1.1782793588485443,
+      "grad_norm": 0.9012355208396912,
+      "learning_rate": 9.931670373503771e-05,
+      "loss": 0.1123,
+      "step": 18010
+    },
+    {
+      "epoch": 1.178933595027805,
+      "grad_norm": 0.8360759019851685,
+      "learning_rate": 9.931518944803477e-05,
+      "loss": 0.1086,
+      "step": 18020
+    },
+    {
+      "epoch": 1.1795878312070658,
+      "grad_norm": 1.0164974927902222,
+      "learning_rate": 9.931367349651484e-05,
+      "loss": 0.1137,
+      "step": 18030
+    },
+    {
+      "epoch": 1.1802420673863265,
+      "grad_norm": 0.7645466327667236,
+      "learning_rate": 9.93121558805291e-05,
+      "loss": 0.0971,
+      "step": 18040
+    },
+    {
+      "epoch": 1.180896303565587,
+      "grad_norm": 0.8536872863769531,
+      "learning_rate": 9.931063660012875e-05,
+      "loss": 0.1113,
+      "step": 18050
+    },
+    {
+      "epoch": 1.1815505397448478,
+      "grad_norm": 0.6680562496185303,
+      "learning_rate": 9.930911565536513e-05,
+      "loss": 0.0963,
+      "step": 18060
+    },
+    {
+      "epoch": 1.1822047759241086,
+      "grad_norm": 0.9181721210479736,
+      "learning_rate": 9.930759304628952e-05,
+      "loss": 0.1102,
+      "step": 18070
+    },
+    {
+      "epoch": 1.1828590121033693,
+      "grad_norm": 0.9609587788581848,
+      "learning_rate": 9.930606877295334e-05,
+      "loss": 0.0972,
+      "step": 18080
+    },
+    {
+      "epoch": 1.18351324828263,
+      "grad_norm": 0.8478082418441772,
+      "learning_rate": 9.930454283540802e-05,
+      "loss": 0.0914,
+      "step": 18090
+    },
+    {
+      "epoch": 1.1841674844618908,
+      "grad_norm": 0.7738621830940247,
+      "learning_rate": 9.930301523370507e-05,
+      "loss": 0.11,
+      "step": 18100
+    },
+    {
+      "epoch": 1.1848217206411515,
+      "grad_norm": 0.8491454720497131,
+      "learning_rate": 9.930148596789605e-05,
+      "loss": 0.1067,
+      "step": 18110
+    },
+    {
+      "epoch": 1.185475956820412,
+      "grad_norm": 0.9434249401092529,
+      "learning_rate": 9.92999550380326e-05,
+      "loss": 0.1046,
+      "step": 18120
+    },
+    {
+      "epoch": 1.1861301929996728,
+      "grad_norm": 1.3511039018630981,
+      "learning_rate": 9.929842244416636e-05,
+      "loss": 0.1087,
+      "step": 18130
+    },
+    {
+      "epoch": 1.1867844291789336,
+      "grad_norm": 0.9221027493476868,
+      "learning_rate": 9.929688818634907e-05,
+      "loss": 0.1059,
+      "step": 18140
+    },
+    {
+      "epoch": 1.1874386653581943,
+      "grad_norm": 0.8466002941131592,
+      "learning_rate": 9.929535226463253e-05,
+      "loss": 0.1008,
+      "step": 18150
+    },
+    {
+      "epoch": 1.188092901537455,
+      "grad_norm": 0.9640662670135498,
+      "learning_rate": 9.929381467906858e-05,
+      "loss": 0.1137,
+      "step": 18160
+    },
+    {
+      "epoch": 1.1887471377167158,
+      "grad_norm": 0.9665100574493408,
+      "learning_rate": 9.929227542970908e-05,
+      "loss": 0.0976,
+      "step": 18170
+    },
+    {
+      "epoch": 1.1894013738959766,
+      "grad_norm": 1.027797818183899,
+      "learning_rate": 9.929073451660602e-05,
+      "loss": 0.101,
+      "step": 18180
+    },
+    {
+      "epoch": 1.190055610075237,
+      "grad_norm": 0.7733204364776611,
+      "learning_rate": 9.928919193981141e-05,
+      "loss": 0.105,
+      "step": 18190
+    },
+    {
+      "epoch": 1.1907098462544978,
+      "grad_norm": 0.874467670917511,
+      "learning_rate": 9.928764769937729e-05,
+      "loss": 0.105,
+      "step": 18200
+    },
+    {
+      "epoch": 1.1913640824337586,
+      "grad_norm": 0.8329676985740662,
+      "learning_rate": 9.92861017953558e-05,
+      "loss": 0.094,
+      "step": 18210
+    },
+    {
+      "epoch": 1.1920183186130193,
+      "grad_norm": 0.9815132021903992,
+      "learning_rate": 9.928455422779913e-05,
+      "loss": 0.1056,
+      "step": 18220
+    },
+    {
+      "epoch": 1.19267255479228,
+      "grad_norm": 0.8479218482971191,
+      "learning_rate": 9.92830049967595e-05,
+      "loss": 0.1049,
+      "step": 18230
+    },
+    {
+      "epoch": 1.1933267909715408,
+      "grad_norm": 1.3466743230819702,
+      "learning_rate": 9.928145410228919e-05,
+      "loss": 0.1031,
+      "step": 18240
+    },
+    {
+      "epoch": 1.1939810271508013,
+      "grad_norm": 0.9225918650627136,
+      "learning_rate": 9.927990154444056e-05,
+      "loss": 0.1104,
+      "step": 18250
+    },
+    {
+      "epoch": 1.194635263330062,
+      "grad_norm": 1.143968939781189,
+      "learning_rate": 9.927834732326602e-05,
+      "loss": 0.1122,
+      "step": 18260
+    },
+    {
+      "epoch": 1.1952894995093228,
+      "grad_norm": 0.9355697631835938,
+      "learning_rate": 9.927679143881802e-05,
+      "loss": 0.0969,
+      "step": 18270
+    },
+    {
+      "epoch": 1.1959437356885836,
+      "grad_norm": 0.9907440543174744,
+      "learning_rate": 9.927523389114908e-05,
+      "loss": 0.1212,
+      "step": 18280
+    },
+    {
+      "epoch": 1.1965979718678443,
+      "grad_norm": 0.8513557314872742,
+      "learning_rate": 9.927367468031175e-05,
+      "loss": 0.0925,
+      "step": 18290
+    },
+    {
+      "epoch": 1.197252208047105,
+      "grad_norm": 0.790827751159668,
+      "learning_rate": 9.927211380635869e-05,
+      "loss": 0.0964,
+      "step": 18300
+    },
+    {
+      "epoch": 1.1979064442263656,
+      "grad_norm": 1.2124277353286743,
+      "learning_rate": 9.927055126934256e-05,
+      "loss": 0.0996,
+      "step": 18310
+    },
+    {
+      "epoch": 1.1985606804056264,
+      "grad_norm": 1.0649062395095825,
+      "learning_rate": 9.92689870693161e-05,
+      "loss": 0.0976,
+      "step": 18320
+    },
+    {
+      "epoch": 1.199214916584887,
+      "grad_norm": 1.0294406414031982,
+      "learning_rate": 9.926742120633214e-05,
+      "loss": 0.1023,
+      "step": 18330
+    },
+    {
+      "epoch": 1.1998691527641479,
+      "grad_norm": 0.8996325135231018,
+      "learning_rate": 9.92658536804435e-05,
+      "loss": 0.0976,
+      "step": 18340
+    },
+    {
+      "epoch": 1.2005233889434086,
+      "grad_norm": 0.8989105820655823,
+      "learning_rate": 9.926428449170308e-05,
+      "loss": 0.1096,
+      "step": 18350
+    },
+    {
+      "epoch": 1.2011776251226693,
+      "grad_norm": 0.7884949445724487,
+      "learning_rate": 9.926271364016386e-05,
+      "loss": 0.1116,
+      "step": 18360
+    },
+    {
+      "epoch": 1.20183186130193,
+      "grad_norm": 0.8274450302124023,
+      "learning_rate": 9.926114112587886e-05,
+      "loss": 0.1015,
+      "step": 18370
+    },
+    {
+      "epoch": 1.2024860974811906,
+      "grad_norm": 1.003177285194397,
+      "learning_rate": 9.925956694890115e-05,
+      "loss": 0.1233,
+      "step": 18380
+    },
+    {
+      "epoch": 1.2031403336604514,
+      "grad_norm": 0.9113197326660156,
+      "learning_rate": 9.925799110928388e-05,
+      "loss": 0.1148,
+      "step": 18390
+    },
+    {
+      "epoch": 1.2037945698397121,
+      "grad_norm": 0.8547518849372864,
+      "learning_rate": 9.925641360708021e-05,
+      "loss": 0.1136,
+      "step": 18400
+    },
+    {
+      "epoch": 1.2044488060189729,
+      "grad_norm": 0.7965877056121826,
+      "learning_rate": 9.925483444234341e-05,
+      "loss": 0.0907,
+      "step": 18410
+    },
+    {
+      "epoch": 1.2051030421982336,
+      "grad_norm": 0.9823962450027466,
+      "learning_rate": 9.925325361512679e-05,
+      "loss": 0.0995,
+      "step": 18420
+    },
+    {
+      "epoch": 1.2057572783774944,
+      "grad_norm": 0.8240478038787842,
+      "learning_rate": 9.925167112548365e-05,
+      "loss": 0.1053,
+      "step": 18430
+    },
+    {
+      "epoch": 1.206411514556755,
+      "grad_norm": 0.8721191883087158,
+      "learning_rate": 9.925008697346748e-05,
+      "loss": 0.1069,
+      "step": 18440
+    },
+    {
+      "epoch": 1.2070657507360156,
+      "grad_norm": 0.9270473718643188,
+      "learning_rate": 9.924850115913169e-05,
+      "loss": 0.1148,
+      "step": 18450
+    },
+    {
+      "epoch": 1.2077199869152764,
+      "grad_norm": 0.7241986989974976,
+      "learning_rate": 9.924691368252983e-05,
+      "loss": 0.1004,
+      "step": 18460
+    },
+    {
+      "epoch": 1.2083742230945371,
+      "grad_norm": 0.8268777132034302,
+      "learning_rate": 9.924532454371549e-05,
+      "loss": 0.0935,
+      "step": 18470
+    },
+    {
+      "epoch": 1.2090284592737979,
+      "grad_norm": 0.8865098357200623,
+      "learning_rate": 9.924373374274228e-05,
+      "loss": 0.1058,
+      "step": 18480
+    },
+    {
+      "epoch": 1.2096826954530586,
+      "grad_norm": 0.8308011293411255,
+      "learning_rate": 9.924214127966391e-05,
+      "loss": 0.1022,
+      "step": 18490
+    },
+    {
+      "epoch": 1.2103369316323191,
+      "grad_norm": 0.8910857439041138,
+      "learning_rate": 9.924054715453414e-05,
+      "loss": 0.0998,
+      "step": 18500
+    },
+    {
+      "epoch": 1.21099116781158,
+      "grad_norm": 0.8274952173233032,
+      "learning_rate": 9.923895136740676e-05,
+      "loss": 0.1031,
+      "step": 18510
+    },
+    {
+      "epoch": 1.2116454039908406,
+      "grad_norm": 1.0838192701339722,
+      "learning_rate": 9.923735391833564e-05,
+      "loss": 0.1024,
+      "step": 18520
+    },
+    {
+      "epoch": 1.2122996401701014,
+      "grad_norm": 0.8322728276252747,
+      "learning_rate": 9.92357548073747e-05,
+      "loss": 0.1234,
+      "step": 18530
+    },
+    {
+      "epoch": 1.2129538763493621,
+      "grad_norm": 0.8210560083389282,
+      "learning_rate": 9.923415403457789e-05,
+      "loss": 0.1051,
+      "step": 18540
+    },
+    {
+      "epoch": 1.2136081125286229,
+      "grad_norm": 0.8708292245864868,
+      "learning_rate": 9.92325515999993e-05,
+      "loss": 0.0952,
+      "step": 18550
+    },
+    {
+      "epoch": 1.2142623487078836,
+      "grad_norm": 0.8271515965461731,
+      "learning_rate": 9.923094750369293e-05,
+      "loss": 0.1003,
+      "step": 18560
+    },
+    {
+      "epoch": 1.2149165848871442,
+      "grad_norm": 0.7365519404411316,
+      "learning_rate": 9.9229341745713e-05,
+      "loss": 0.1061,
+      "step": 18570
+    },
+    {
+      "epoch": 1.215570821066405,
+      "grad_norm": 1.1527734994888306,
+      "learning_rate": 9.922773432611366e-05,
+      "loss": 0.1127,
+      "step": 18580
+    },
+    {
+      "epoch": 1.2162250572456657,
+      "grad_norm": 0.7652072310447693,
+      "learning_rate": 9.922612524494919e-05,
+      "loss": 0.1048,
+      "step": 18590
+    },
+    {
+      "epoch": 1.2168792934249264,
+      "grad_norm": 0.8632088899612427,
+      "learning_rate": 9.92245145022739e-05,
+      "loss": 0.101,
+      "step": 18600
+    },
+    {
+      "epoch": 1.2175335296041871,
+      "grad_norm": 0.8901262283325195,
+      "learning_rate": 9.922290209814214e-05,
+      "loss": 0.0977,
+      "step": 18610
+    },
+    {
+      "epoch": 1.218187765783448,
+      "grad_norm": 0.8626736402511597,
+      "learning_rate": 9.922128803260833e-05,
+      "loss": 0.0963,
+      "step": 18620
+    },
+    {
+      "epoch": 1.2188420019627086,
+      "grad_norm": 0.8524816632270813,
+      "learning_rate": 9.921967230572699e-05,
+      "loss": 0.1004,
+      "step": 18630
+    },
+    {
+      "epoch": 1.2194962381419692,
+      "grad_norm": 0.6984155774116516,
+      "learning_rate": 9.92180549175526e-05,
+      "loss": 0.1046,
+      "step": 18640
+    },
+    {
+      "epoch": 1.22015047432123,
+      "grad_norm": 0.9021031856536865,
+      "learning_rate": 9.92164358681398e-05,
+      "loss": 0.096,
+      "step": 18650
+    },
+    {
+      "epoch": 1.2208047105004907,
+      "grad_norm": 0.8669722676277161,
+      "learning_rate": 9.921481515754321e-05,
+      "loss": 0.1071,
+      "step": 18660
+    },
+    {
+      "epoch": 1.2214589466797514,
+      "grad_norm": 0.8371648788452148,
+      "learning_rate": 9.921319278581752e-05,
+      "loss": 0.1051,
+      "step": 18670
+    },
+    {
+      "epoch": 1.2221131828590122,
+      "grad_norm": 0.8591510653495789,
+      "learning_rate": 9.921156875301751e-05,
+      "loss": 0.0984,
+      "step": 18680
+    },
+    {
+      "epoch": 1.222767419038273,
+      "grad_norm": 0.7590886950492859,
+      "learning_rate": 9.920994305919801e-05,
+      "loss": 0.1065,
+      "step": 18690
+    },
+    {
+      "epoch": 1.2234216552175334,
+      "grad_norm": 0.9001472592353821,
+      "learning_rate": 9.920831570441387e-05,
+      "loss": 0.0941,
+      "step": 18700
+    },
+    {
+      "epoch": 1.2240758913967942,
+      "grad_norm": 1.0100491046905518,
+      "learning_rate": 9.920668668872002e-05,
+      "loss": 0.0984,
+      "step": 18710
+    },
+    {
+      "epoch": 1.224730127576055,
+      "grad_norm": 0.9002352952957153,
+      "learning_rate": 9.920505601217143e-05,
+      "loss": 0.1018,
+      "step": 18720
+    },
+    {
+      "epoch": 1.2253843637553157,
+      "grad_norm": 0.8079519867897034,
+      "learning_rate": 9.920342367482318e-05,
+      "loss": 0.1088,
+      "step": 18730
+    },
+    {
+      "epoch": 1.2260385999345764,
+      "grad_norm": 0.9378504157066345,
+      "learning_rate": 9.920178967673031e-05,
+      "loss": 0.1105,
+      "step": 18740
+    },
+    {
+      "epoch": 1.2266928361138372,
+      "grad_norm": 0.9143635034561157,
+      "learning_rate": 9.920015401794803e-05,
+      "loss": 0.0919,
+      "step": 18750
+    },
+    {
+      "epoch": 1.2273470722930977,
+      "grad_norm": 0.8441551923751831,
+      "learning_rate": 9.91985166985315e-05,
+      "loss": 0.097,
+      "step": 18760
+    },
+    {
+      "epoch": 1.2280013084723584,
+      "grad_norm": 0.8805584907531738,
+      "learning_rate": 9.919687771853601e-05,
+      "loss": 0.1003,
+      "step": 18770
+    },
+    {
+      "epoch": 1.2286555446516192,
+      "grad_norm": 1.0052324533462524,
+      "learning_rate": 9.919523707801687e-05,
+      "loss": 0.1009,
+      "step": 18780
+    },
+    {
+      "epoch": 1.22930978083088,
+      "grad_norm": 0.9868443608283997,
+      "learning_rate": 9.919359477702948e-05,
+      "loss": 0.1163,
+      "step": 18790
+    },
+    {
+      "epoch": 1.2299640170101407,
+      "grad_norm": 1.144353985786438,
+      "learning_rate": 9.919195081562924e-05,
+      "loss": 0.1081,
+      "step": 18800
+    },
+    {
+      "epoch": 1.2306182531894014,
+      "grad_norm": 1.0622575283050537,
+      "learning_rate": 9.919030519387164e-05,
+      "loss": 0.0981,
+      "step": 18810
+    },
+    {
+      "epoch": 1.2312724893686622,
+      "grad_norm": 0.8485022783279419,
+      "learning_rate": 9.918865791181224e-05,
+      "loss": 0.1029,
+      "step": 18820
+    },
+    {
+      "epoch": 1.2319267255479227,
+      "grad_norm": 1.022147536277771,
+      "learning_rate": 9.918700896950664e-05,
+      "loss": 0.1053,
+      "step": 18830
+    },
+    {
+      "epoch": 1.2325809617271835,
+      "grad_norm": 0.8381478190422058,
+      "learning_rate": 9.918535836701047e-05,
+      "loss": 0.1024,
+      "step": 18840
+    },
+    {
+      "epoch": 1.2332351979064442,
+      "grad_norm": 0.7995995879173279,
+      "learning_rate": 9.918370610437948e-05,
+      "loss": 0.0961,
+      "step": 18850
+    },
+    {
+      "epoch": 1.233889434085705,
+      "grad_norm": 1.0238806009292603,
+      "learning_rate": 9.91820521816694e-05,
+      "loss": 0.1049,
+      "step": 18860
+    },
+    {
+      "epoch": 1.2345436702649657,
+      "grad_norm": 1.0174421072006226,
+      "learning_rate": 9.91803965989361e-05,
+      "loss": 0.0893,
+      "step": 18870
+    },
+    {
+      "epoch": 1.2351979064442264,
+      "grad_norm": 1.1415332555770874,
+      "learning_rate": 9.917873935623542e-05,
+      "loss": 0.1049,
+      "step": 18880
+    },
+    {
+      "epoch": 1.2358521426234872,
+      "grad_norm": 0.7645314931869507,
+      "learning_rate": 9.91770804536233e-05,
+      "loss": 0.1065,
+      "step": 18890
+    },
+    {
+      "epoch": 1.2365063788027477,
+      "grad_norm": 0.9801765084266663,
+      "learning_rate": 9.917541989115578e-05,
+      "loss": 0.1038,
+      "step": 18900
+    },
+    {
+      "epoch": 1.2371606149820085,
+      "grad_norm": 0.9121055603027344,
+      "learning_rate": 9.917375766888883e-05,
+      "loss": 0.1095,
+      "step": 18910
+    },
+    {
+      "epoch": 1.2378148511612692,
+      "grad_norm": 0.7746009230613708,
+      "learning_rate": 9.917209378687862e-05,
+      "loss": 0.0968,
+      "step": 18920
+    },
+    {
+      "epoch": 1.23846908734053,
+      "grad_norm": 0.8752132058143616,
+      "learning_rate": 9.917042824518127e-05,
+      "loss": 0.1014,
+      "step": 18930
+    },
+    {
+      "epoch": 1.2391233235197907,
+      "grad_norm": 0.8438636660575867,
+      "learning_rate": 9.916876104385303e-05,
+      "loss": 0.1035,
+      "step": 18940
+    },
+    {
+      "epoch": 1.2397775596990512,
+      "grad_norm": 1.1780970096588135,
+      "learning_rate": 9.916709218295015e-05,
+      "loss": 0.1077,
+      "step": 18950
+    },
+    {
+      "epoch": 1.240431795878312,
+      "grad_norm": 0.9799401164054871,
+      "learning_rate": 9.916542166252895e-05,
+      "loss": 0.1137,
+      "step": 18960
+    },
+    {
+      "epoch": 1.2410860320575727,
+      "grad_norm": 0.6996291875839233,
+      "learning_rate": 9.916374948264584e-05,
+      "loss": 0.1007,
+      "step": 18970
+    },
+    {
+      "epoch": 1.2417402682368335,
+      "grad_norm": 0.7507869601249695,
+      "learning_rate": 9.916207564335726e-05,
+      "loss": 0.0997,
+      "step": 18980
+    },
+    {
+      "epoch": 1.2423945044160942,
+      "grad_norm": 0.8153190016746521,
+      "learning_rate": 9.916040014471968e-05,
+      "loss": 0.0974,
+      "step": 18990
+    },
+    {
+      "epoch": 1.243048740595355,
+      "grad_norm": 0.7431339621543884,
+      "learning_rate": 9.915872298678966e-05,
+      "loss": 0.092,
+      "step": 19000
+    },
+    {
+      "epoch": 1.2437029767746157,
+      "grad_norm": 1.1183733940124512,
+      "learning_rate": 9.915704416962382e-05,
+      "loss": 0.1045,
+      "step": 19010
+    },
+    {
+      "epoch": 1.2443572129538762,
+      "grad_norm": 1.058430790901184,
+      "learning_rate": 9.915536369327883e-05,
+      "loss": 0.1204,
+      "step": 19020
+    },
+    {
+      "epoch": 1.245011449133137,
+      "grad_norm": 0.9192705750465393,
+      "learning_rate": 9.91536815578114e-05,
+      "loss": 0.113,
+      "step": 19030
+    },
+    {
+      "epoch": 1.2456656853123977,
+      "grad_norm": 0.899825394153595,
+      "learning_rate": 9.91519977632783e-05,
+      "loss": 0.0965,
+      "step": 19040
+    },
+    {
+      "epoch": 1.2463199214916585,
+      "grad_norm": 0.8516882061958313,
+      "learning_rate": 9.915031230973637e-05,
+      "loss": 0.1042,
+      "step": 19050
+    },
+    {
+      "epoch": 1.2469741576709192,
+      "grad_norm": 0.8238911032676697,
+      "learning_rate": 9.914862519724251e-05,
+      "loss": 0.1051,
+      "step": 19060
+    },
+    {
+      "epoch": 1.24762839385018,
+      "grad_norm": 0.8177777528762817,
+      "learning_rate": 9.914693642585364e-05,
+      "loss": 0.1074,
+      "step": 19070
+    },
+    {
+      "epoch": 1.2482826300294407,
+      "grad_norm": 1.2330362796783447,
+      "learning_rate": 9.914524599562677e-05,
+      "loss": 0.1078,
+      "step": 19080
+    },
+    {
+      "epoch": 1.2489368662087013,
+      "grad_norm": 1.126373529434204,
+      "learning_rate": 9.914355390661896e-05,
+      "loss": 0.1013,
+      "step": 19090
+    },
+    {
+      "epoch": 1.249591102387962,
+      "grad_norm": 0.8344969153404236,
+      "learning_rate": 9.914186015888733e-05,
+      "loss": 0.1009,
+      "step": 19100
+    },
+    {
+      "epoch": 1.2502453385672228,
+      "grad_norm": 0.8182376027107239,
+      "learning_rate": 9.914016475248904e-05,
+      "loss": 0.0968,
+      "step": 19110
+    },
+    {
+      "epoch": 1.2508995747464835,
+      "grad_norm": 0.9014137387275696,
+      "learning_rate": 9.913846768748132e-05,
+      "loss": 0.1089,
+      "step": 19120
+    },
+    {
+      "epoch": 1.2515538109257442,
+      "grad_norm": 0.9868068695068359,
+      "learning_rate": 9.913676896392144e-05,
+      "loss": 0.1141,
+      "step": 19130
+    },
+    {
+      "epoch": 1.2522080471050048,
+      "grad_norm": 1.0246328115463257,
+      "learning_rate": 9.913506858186673e-05,
+      "loss": 0.105,
+      "step": 19140
+    },
+    {
+      "epoch": 1.2528622832842657,
+      "grad_norm": 0.783169686794281,
+      "learning_rate": 9.913336654137459e-05,
+      "loss": 0.0996,
+      "step": 19150
+    },
+    {
+      "epoch": 1.2535165194635263,
+      "grad_norm": 0.8389064073562622,
+      "learning_rate": 9.913166284250251e-05,
+      "loss": 0.089,
+      "step": 19160
+    },
+    {
+      "epoch": 1.254170755642787,
+      "grad_norm": 0.8309929966926575,
+      "learning_rate": 9.912995748530792e-05,
+      "loss": 0.0968,
+      "step": 19170
+    },
+    {
+      "epoch": 1.2548249918220478,
+      "grad_norm": 0.9536809325218201,
+      "learning_rate": 9.912825046984842e-05,
+      "loss": 0.1027,
+      "step": 19180
+    },
+    {
+      "epoch": 1.2554792280013085,
+      "grad_norm": 1.0172020196914673,
+      "learning_rate": 9.912654179618164e-05,
+      "loss": 0.1111,
+      "step": 19190
+    },
+    {
+      "epoch": 1.2561334641805693,
+      "grad_norm": 0.8524767160415649,
+      "learning_rate": 9.912483146436522e-05,
+      "loss": 0.0941,
+      "step": 19200
+    },
+    {
+      "epoch": 1.2567877003598298,
+      "grad_norm": 0.9486467242240906,
+      "learning_rate": 9.912311947445692e-05,
+      "loss": 0.1148,
+      "step": 19210
+    },
+    {
+      "epoch": 1.2574419365390905,
+      "grad_norm": 0.9282219409942627,
+      "learning_rate": 9.91214058265145e-05,
+      "loss": 0.1087,
+      "step": 19220
+    },
+    {
+      "epoch": 1.2580961727183513,
+      "grad_norm": 0.9173193573951721,
+      "learning_rate": 9.911969052059579e-05,
+      "loss": 0.1129,
+      "step": 19230
+    },
+    {
+      "epoch": 1.258750408897612,
+      "grad_norm": 0.8629287481307983,
+      "learning_rate": 9.911797355675874e-05,
+      "loss": 0.109,
+      "step": 19240
+    },
+    {
+      "epoch": 1.2594046450768728,
+      "grad_norm": 0.8385807871818542,
+      "learning_rate": 9.911625493506124e-05,
+      "loss": 0.1052,
+      "step": 19250
+    },
+    {
+      "epoch": 1.2600588812561335,
+      "grad_norm": 0.7391451597213745,
+      "learning_rate": 9.911453465556133e-05,
+      "loss": 0.1045,
+      "step": 19260
+    },
+    {
+      "epoch": 1.2607131174353943,
+      "grad_norm": 0.8053876757621765,
+      "learning_rate": 9.911281271831707e-05,
+      "loss": 0.1016,
+      "step": 19270
+    },
+    {
+      "epoch": 1.2613673536146548,
+      "grad_norm": 0.8911042809486389,
+      "learning_rate": 9.911108912338657e-05,
+      "loss": 0.1077,
+      "step": 19280
+    },
+    {
+      "epoch": 1.2620215897939155,
+      "grad_norm": 0.8589571118354797,
+      "learning_rate": 9.910936387082802e-05,
+      "loss": 0.1093,
+      "step": 19290
+    },
+    {
+      "epoch": 1.2626758259731763,
+      "grad_norm": 0.8094229102134705,
+      "learning_rate": 9.910763696069965e-05,
+      "loss": 0.1096,
+      "step": 19300
+    },
+    {
+      "epoch": 1.263330062152437,
+      "grad_norm": 0.8509666323661804,
+      "learning_rate": 9.910590839305973e-05,
+      "loss": 0.1072,
+      "step": 19310
+    },
+    {
+      "epoch": 1.2639842983316978,
+      "grad_norm": 0.9664101600646973,
+      "learning_rate": 9.910417816796662e-05,
+      "loss": 0.1019,
+      "step": 19320
+    },
+    {
+      "epoch": 1.2646385345109585,
+      "grad_norm": 0.8401138782501221,
+      "learning_rate": 9.910244628547872e-05,
+      "loss": 0.1015,
+      "step": 19330
+    },
+    {
+      "epoch": 1.2652927706902193,
+      "grad_norm": 0.7719675898551941,
+      "learning_rate": 9.910071274565449e-05,
+      "loss": 0.1116,
+      "step": 19340
+    },
+    {
+      "epoch": 1.2659470068694798,
+      "grad_norm": 1.0611908435821533,
+      "learning_rate": 9.909897754855242e-05,
+      "loss": 0.0963,
+      "step": 19350
+    },
+    {
+      "epoch": 1.2666012430487406,
+      "grad_norm": 0.9017995595932007,
+      "learning_rate": 9.90972406942311e-05,
+      "loss": 0.1046,
+      "step": 19360
+    },
+    {
+      "epoch": 1.2672554792280013,
+      "grad_norm": 0.7480419278144836,
+      "learning_rate": 9.909550218274915e-05,
+      "loss": 0.0987,
+      "step": 19370
+    },
+    {
+      "epoch": 1.267909715407262,
+      "grad_norm": 0.949092447757721,
+      "learning_rate": 9.909376201416522e-05,
+      "loss": 0.1129,
+      "step": 19380
+    },
+    {
+      "epoch": 1.2685639515865228,
+      "grad_norm": 0.8834240436553955,
+      "learning_rate": 9.909202018853809e-05,
+      "loss": 0.1007,
+      "step": 19390
+    },
+    {
+      "epoch": 1.2692181877657833,
+      "grad_norm": 0.8566089272499084,
+      "learning_rate": 9.909027670592652e-05,
+      "loss": 0.1074,
+      "step": 19400
+    },
+    {
+      "epoch": 1.2698724239450443,
+      "grad_norm": 0.8526421785354614,
+      "learning_rate": 9.908853156638937e-05,
+      "loss": 0.116,
+      "step": 19410
+    },
+    {
+      "epoch": 1.2705266601243048,
+      "grad_norm": 0.8203496932983398,
+      "learning_rate": 9.908678476998555e-05,
+      "loss": 0.098,
+      "step": 19420
+    },
+    {
+      "epoch": 1.2711808963035656,
+      "grad_norm": 0.8378643989562988,
+      "learning_rate": 9.908503631677399e-05,
+      "loss": 0.1081,
+      "step": 19430
+    },
+    {
+      "epoch": 1.2718351324828263,
+      "grad_norm": 1.0013712644577026,
+      "learning_rate": 9.908328620681373e-05,
+      "loss": 0.1006,
+      "step": 19440
+    },
+    {
+      "epoch": 1.272489368662087,
+      "grad_norm": 0.9886232018470764,
+      "learning_rate": 9.908153444016385e-05,
+      "loss": 0.118,
+      "step": 19450
+    },
+    {
+      "epoch": 1.2731436048413478,
+      "grad_norm": 0.9862875938415527,
+      "learning_rate": 9.907978101688344e-05,
+      "loss": 0.1166,
+      "step": 19460
+    },
+    {
+      "epoch": 1.2737978410206083,
+      "grad_norm": 0.935799241065979,
+      "learning_rate": 9.907802593703173e-05,
+      "loss": 0.0968,
+      "step": 19470
+    },
+    {
+      "epoch": 1.274452077199869,
+      "grad_norm": 0.8719044923782349,
+      "learning_rate": 9.90762692006679e-05,
+      "loss": 0.0947,
+      "step": 19480
+    },
+    {
+      "epoch": 1.2751063133791298,
+      "grad_norm": 0.7218495607376099,
+      "learning_rate": 9.90745108078513e-05,
+      "loss": 0.0911,
+      "step": 19490
+    },
+    {
+      "epoch": 1.2757605495583906,
+      "grad_norm": 0.8420484066009521,
+      "learning_rate": 9.907275075864127e-05,
+      "loss": 0.1092,
+      "step": 19500
+    },
+    {
+      "epoch": 1.2764147857376513,
+      "grad_norm": 0.808154821395874,
+      "learning_rate": 9.907098905309718e-05,
+      "loss": 0.0971,
+      "step": 19510
+    },
+    {
+      "epoch": 1.277069021916912,
+      "grad_norm": 0.7025526165962219,
+      "learning_rate": 9.906922569127853e-05,
+      "loss": 0.108,
+      "step": 19520
+    },
+    {
+      "epoch": 1.2777232580961728,
+      "grad_norm": 0.9783757925033569,
+      "learning_rate": 9.906746067324481e-05,
+      "loss": 0.1065,
+      "step": 19530
+    },
+    {
+      "epoch": 1.2783774942754333,
+      "grad_norm": 0.8030864000320435,
+      "learning_rate": 9.906569399905561e-05,
+      "loss": 0.1054,
+      "step": 19540
+    },
+    {
+      "epoch": 1.279031730454694,
+      "grad_norm": 0.8669215440750122,
+      "learning_rate": 9.906392566877057e-05,
+      "loss": 0.0978,
+      "step": 19550
+    },
+    {
+      "epoch": 1.2796859666339548,
+      "grad_norm": 0.8483967781066895,
+      "learning_rate": 9.906215568244935e-05,
+      "loss": 0.1098,
+      "step": 19560
+    },
+    {
+      "epoch": 1.2803402028132156,
+      "grad_norm": 0.7597243785858154,
+      "learning_rate": 9.906038404015171e-05,
+      "loss": 0.106,
+      "step": 19570
+    },
+    {
+      "epoch": 1.2809944389924763,
+      "grad_norm": 0.9336119294166565,
+      "learning_rate": 9.905861074193745e-05,
+      "loss": 0.0995,
+      "step": 19580
+    },
+    {
+      "epoch": 1.2816486751717369,
+      "grad_norm": 0.8763530254364014,
+      "learning_rate": 9.905683578786641e-05,
+      "loss": 0.0976,
+      "step": 19590
+    },
+    {
+      "epoch": 1.2823029113509978,
+      "grad_norm": 0.9109959602355957,
+      "learning_rate": 9.905505917799851e-05,
+      "loss": 0.1021,
+      "step": 19600
+    },
+    {
+      "epoch": 1.2829571475302584,
+      "grad_norm": 1.0181946754455566,
+      "learning_rate": 9.90532809123937e-05,
+      "loss": 0.1142,
+      "step": 19610
+    },
+    {
+      "epoch": 1.283611383709519,
+      "grad_norm": 1.0679359436035156,
+      "learning_rate": 9.905150099111202e-05,
+      "loss": 0.1152,
+      "step": 19620
+    },
+    {
+      "epoch": 1.2842656198887799,
+      "grad_norm": 0.8650368452072144,
+      "learning_rate": 9.904971941421355e-05,
+      "loss": 0.0966,
+      "step": 19630
+    },
+    {
+      "epoch": 1.2849198560680406,
+      "grad_norm": 1.0523502826690674,
+      "learning_rate": 9.90479361817584e-05,
+      "loss": 0.104,
+      "step": 19640
+    },
+    {
+      "epoch": 1.2855740922473013,
+      "grad_norm": 0.976112961769104,
+      "learning_rate": 9.904615129380676e-05,
+      "loss": 0.0967,
+      "step": 19650
+    },
+    {
+      "epoch": 1.2862283284265619,
+      "grad_norm": 0.8600479364395142,
+      "learning_rate": 9.904436475041891e-05,
+      "loss": 0.105,
+      "step": 19660
+    },
+    {
+      "epoch": 1.2868825646058226,
+      "grad_norm": 1.066514015197754,
+      "learning_rate": 9.904257655165512e-05,
+      "loss": 0.1022,
+      "step": 19670
+    },
+    {
+      "epoch": 1.2875368007850834,
+      "grad_norm": 0.9571858644485474,
+      "learning_rate": 9.904078669757575e-05,
+      "loss": 0.1039,
+      "step": 19680
+    },
+    {
+      "epoch": 1.2881910369643441,
+      "grad_norm": 0.8673845529556274,
+      "learning_rate": 9.90389951882412e-05,
+      "loss": 0.0921,
+      "step": 19690
+    },
+    {
+      "epoch": 1.2888452731436049,
+      "grad_norm": 0.995274007320404,
+      "learning_rate": 9.903720202371198e-05,
+      "loss": 0.1015,
+      "step": 19700
+    },
+    {
+      "epoch": 1.2894995093228656,
+      "grad_norm": 0.9427167773246765,
+      "learning_rate": 9.903540720404856e-05,
+      "loss": 0.0934,
+      "step": 19710
+    },
+    {
+      "epoch": 1.2901537455021264,
+      "grad_norm": 0.8194208145141602,
+      "learning_rate": 9.903361072931156e-05,
+      "loss": 0.0943,
+      "step": 19720
+    },
+    {
+      "epoch": 1.2908079816813869,
+      "grad_norm": 0.8158044815063477,
+      "learning_rate": 9.903181259956161e-05,
+      "loss": 0.1123,
+      "step": 19730
+    },
+    {
+      "epoch": 1.2914622178606476,
+      "grad_norm": 0.7874143123626709,
+      "learning_rate": 9.903001281485937e-05,
+      "loss": 0.0954,
+      "step": 19740
+    },
+    {
+      "epoch": 1.2921164540399084,
+      "grad_norm": 1.0694737434387207,
+      "learning_rate": 9.902821137526564e-05,
+      "loss": 0.1028,
+      "step": 19750
+    },
+    {
+      "epoch": 1.2927706902191691,
+      "grad_norm": 0.9189092516899109,
+      "learning_rate": 9.902640828084118e-05,
+      "loss": 0.1016,
+      "step": 19760
+    },
+    {
+      "epoch": 1.2934249263984299,
+      "grad_norm": 0.7503056526184082,
+      "learning_rate": 9.902460353164687e-05,
+      "loss": 0.0958,
+      "step": 19770
+    },
+    {
+      "epoch": 1.2940791625776906,
+      "grad_norm": 1.0281898975372314,
+      "learning_rate": 9.90227971277436e-05,
+      "loss": 0.0997,
+      "step": 19780
+    },
+    {
+      "epoch": 1.2947333987569514,
+      "grad_norm": 0.817094624042511,
+      "learning_rate": 9.902098906919239e-05,
+      "loss": 0.1076,
+      "step": 19790
+    },
+    {
+      "epoch": 1.295387634936212,
+      "grad_norm": 0.7925615906715393,
+      "learning_rate": 9.901917935605423e-05,
+      "loss": 0.0986,
+      "step": 19800
+    },
+    {
+      "epoch": 1.2960418711154726,
+      "grad_norm": 0.8204017877578735,
+      "learning_rate": 9.901736798839018e-05,
+      "loss": 0.0976,
+      "step": 19810
+    },
+    {
+      "epoch": 1.2966961072947334,
+      "grad_norm": 0.7727102637290955,
+      "learning_rate": 9.901555496626145e-05,
+      "loss": 0.0922,
+      "step": 19820
+    },
+    {
+      "epoch": 1.2973503434739941,
+      "grad_norm": 0.9235280156135559,
+      "learning_rate": 9.901374028972916e-05,
+      "loss": 0.1103,
+      "step": 19830
+    },
+    {
+      "epoch": 1.2980045796532549,
+      "grad_norm": 0.8759993314743042,
+      "learning_rate": 9.901192395885461e-05,
+      "loss": 0.0985,
+      "step": 19840
+    },
+    {
+      "epoch": 1.2986588158325154,
+      "grad_norm": 0.7741413116455078,
+      "learning_rate": 9.901010597369907e-05,
+      "loss": 0.0985,
+      "step": 19850
+    },
+    {
+      "epoch": 1.2993130520117764,
+      "grad_norm": 0.8373982906341553,
+      "learning_rate": 9.900828633432393e-05,
+      "loss": 0.1054,
+      "step": 19860
+    },
+    {
+      "epoch": 1.299967288191037,
+      "grad_norm": 0.708233654499054,
+      "learning_rate": 9.90064650407906e-05,
+      "loss": 0.1,
+      "step": 19870
+    },
+    {
+      "epoch": 1.3006215243702977,
+      "grad_norm": 0.8219735026359558,
+      "learning_rate": 9.900464209316054e-05,
+      "loss": 0.1052,
+      "step": 19880
+    },
+    {
+      "epoch": 1.3012757605495584,
+      "grad_norm": 0.7485021352767944,
+      "learning_rate": 9.900281749149531e-05,
+      "loss": 0.0986,
+      "step": 19890
+    },
+    {
+      "epoch": 1.3019299967288191,
+      "grad_norm": 0.8410215973854065,
+      "learning_rate": 9.900099123585646e-05,
+      "loss": 0.1012,
+      "step": 19900
+    },
+    {
+      "epoch": 1.30258423290808,
+      "grad_norm": 0.9873031377792358,
+      "learning_rate": 9.899916332630565e-05,
+      "loss": 0.1031,
+      "step": 19910
+    },
+    {
+      "epoch": 1.3032384690873404,
+      "grad_norm": 0.7580830454826355,
+      "learning_rate": 9.899733376290458e-05,
+      "loss": 0.1084,
+      "step": 19920
+    },
+    {
+      "epoch": 1.3038927052666012,
+      "grad_norm": 0.8297830820083618,
+      "learning_rate": 9.899550254571499e-05,
+      "loss": 0.1034,
+      "step": 19930
+    },
+    {
+      "epoch": 1.304546941445862,
+      "grad_norm": 0.888883113861084,
+      "learning_rate": 9.899366967479868e-05,
+      "loss": 0.0951,
+      "step": 19940
+    },
+    {
+      "epoch": 1.3052011776251227,
+      "grad_norm": 0.7784373760223389,
+      "learning_rate": 9.899183515021755e-05,
+      "loss": 0.107,
+      "step": 19950
+    },
+    {
+      "epoch": 1.3058554138043834,
+      "grad_norm": 0.7872931361198425,
+      "learning_rate": 9.898999897203347e-05,
+      "loss": 0.0941,
+      "step": 19960
+    },
+    {
+      "epoch": 1.3065096499836442,
+      "grad_norm": 0.8830298781394958,
+      "learning_rate": 9.898816114030846e-05,
+      "loss": 0.1035,
+      "step": 19970
+    },
+    {
+      "epoch": 1.307163886162905,
+      "grad_norm": 0.8231805562973022,
+      "learning_rate": 9.898632165510455e-05,
+      "loss": 0.1061,
+      "step": 19980
+    },
+    {
+      "epoch": 1.3078181223421654,
+      "grad_norm": 0.8644198179244995,
+      "learning_rate": 9.89844805164838e-05,
+      "loss": 0.1068,
+      "step": 19990
+    },
+    {
+      "epoch": 1.3084723585214262,
+      "grad_norm": 0.7423153519630432,
+      "learning_rate": 9.898263772450836e-05,
+      "loss": 0.102,
+      "step": 20000
+    },
+    {
+      "epoch": 1.309126594700687,
+      "grad_norm": 0.9147112965583801,
+      "learning_rate": 9.898079327924044e-05,
+      "loss": 0.1079,
+      "step": 20010
+    },
+    {
+      "epoch": 1.3097808308799477,
+      "grad_norm": 0.8351669907569885,
+      "learning_rate": 9.897894718074229e-05,
+      "loss": 0.1015,
+      "step": 20020
+    },
+    {
+      "epoch": 1.3104350670592084,
+      "grad_norm": 0.9719964861869812,
+      "learning_rate": 9.897709942907623e-05,
+      "loss": 0.096,
+      "step": 20030
+    },
+    {
+      "epoch": 1.311089303238469,
+      "grad_norm": 1.0228232145309448,
+      "learning_rate": 9.897525002430458e-05,
+      "loss": 0.1035,
+      "step": 20040
+    },
+    {
+      "epoch": 1.31174353941773,
+      "grad_norm": 0.8931130766868591,
+      "learning_rate": 9.897339896648985e-05,
+      "loss": 0.1063,
+      "step": 20050
+    },
+    {
+      "epoch": 1.3123977755969904,
+      "grad_norm": 0.992167592048645,
+      "learning_rate": 9.897154625569443e-05,
+      "loss": 0.0942,
+      "step": 20060
+    },
+    {
+      "epoch": 1.3130520117762512,
+      "grad_norm": 0.7140622735023499,
+      "learning_rate": 9.89696918919809e-05,
+      "loss": 0.0967,
+      "step": 20070
+    },
+    {
+      "epoch": 1.313706247955512,
+      "grad_norm": 0.7795881032943726,
+      "learning_rate": 9.896783587541186e-05,
+      "loss": 0.105,
+      "step": 20080
+    },
+    {
+      "epoch": 1.3143604841347727,
+      "grad_norm": 0.8552742004394531,
+      "learning_rate": 9.896597820604992e-05,
+      "loss": 0.0931,
+      "step": 20090
+    },
+    {
+      "epoch": 1.3150147203140334,
+      "grad_norm": 0.7586219310760498,
+      "learning_rate": 9.89641188839578e-05,
+      "loss": 0.1137,
+      "step": 20100
+    },
+    {
+      "epoch": 1.315668956493294,
+      "grad_norm": 0.9405565857887268,
+      "learning_rate": 9.896225790919826e-05,
+      "loss": 0.1114,
+      "step": 20110
+    },
+    {
+      "epoch": 1.3163231926725547,
+      "grad_norm": 0.929673969745636,
+      "learning_rate": 9.89603952818341e-05,
+      "loss": 0.0999,
+      "step": 20120
+    },
+    {
+      "epoch": 1.3169774288518155,
+      "grad_norm": 0.8156664371490479,
+      "learning_rate": 9.89585310019282e-05,
+      "loss": 0.1036,
+      "step": 20130
+    },
+    {
+      "epoch": 1.3176316650310762,
+      "grad_norm": 0.9651734232902527,
+      "learning_rate": 9.895666506954347e-05,
+      "loss": 0.104,
+      "step": 20140
+    },
+    {
+      "epoch": 1.318285901210337,
+      "grad_norm": 1.05132257938385,
+      "learning_rate": 9.895479748474292e-05,
+      "loss": 0.1016,
+      "step": 20150
+    },
+    {
+      "epoch": 1.3189401373895977,
+      "grad_norm": 0.8636537790298462,
+      "learning_rate": 9.895292824758955e-05,
+      "loss": 0.1022,
+      "step": 20160
+    },
+    {
+      "epoch": 1.3195943735688584,
+      "grad_norm": 0.7908622026443481,
+      "learning_rate": 9.895105735814647e-05,
+      "loss": 0.1048,
+      "step": 20170
+    },
+    {
+      "epoch": 1.320248609748119,
+      "grad_norm": 0.8223440051078796,
+      "learning_rate": 9.894918481647684e-05,
+      "loss": 0.0958,
+      "step": 20180
+    },
+    {
+      "epoch": 1.3209028459273797,
+      "grad_norm": 0.9097592234611511,
+      "learning_rate": 9.894731062264383e-05,
+      "loss": 0.1036,
+      "step": 20190
+    },
+    {
+      "epoch": 1.3215570821066405,
+      "grad_norm": 0.8000107407569885,
+      "learning_rate": 9.894543477671072e-05,
+      "loss": 0.0993,
+      "step": 20200
+    },
+    {
+      "epoch": 1.3222113182859012,
+      "grad_norm": 0.8129693865776062,
+      "learning_rate": 9.894355727874083e-05,
+      "loss": 0.113,
+      "step": 20210
+    },
+    {
+      "epoch": 1.322865554465162,
+      "grad_norm": 0.7431675791740417,
+      "learning_rate": 9.894167812879751e-05,
+      "loss": 0.1008,
+      "step": 20220
+    },
+    {
+      "epoch": 1.3235197906444227,
+      "grad_norm": 1.0079487562179565,
+      "learning_rate": 9.893979732694421e-05,
+      "loss": 0.1075,
+      "step": 20230
+    },
+    {
+      "epoch": 1.3241740268236835,
+      "grad_norm": 0.7943064570426941,
+      "learning_rate": 9.89379148732444e-05,
+      "loss": 0.1154,
+      "step": 20240
+    },
+    {
+      "epoch": 1.324828263002944,
+      "grad_norm": 0.7072292566299438,
+      "learning_rate": 9.893603076776162e-05,
+      "loss": 0.113,
+      "step": 20250
+    },
+    {
+      "epoch": 1.3254824991822047,
+      "grad_norm": 0.8471561074256897,
+      "learning_rate": 9.893414501055947e-05,
+      "loss": 0.1068,
+      "step": 20260
+    },
+    {
+      "epoch": 1.3261367353614655,
+      "grad_norm": 0.842134416103363,
+      "learning_rate": 9.893225760170157e-05,
+      "loss": 0.1066,
+      "step": 20270
+    },
+    {
+      "epoch": 1.3267909715407262,
+      "grad_norm": 0.836654543876648,
+      "learning_rate": 9.893036854125166e-05,
+      "loss": 0.099,
+      "step": 20280
+    },
+    {
+      "epoch": 1.327445207719987,
+      "grad_norm": 0.8249868750572205,
+      "learning_rate": 9.892847782927348e-05,
+      "loss": 0.0977,
+      "step": 20290
+    },
+    {
+      "epoch": 1.3280994438992475,
+      "grad_norm": 0.7410039305686951,
+      "learning_rate": 9.892658546583087e-05,
+      "loss": 0.114,
+      "step": 20300
+    },
+    {
+      "epoch": 1.3287536800785085,
+      "grad_norm": 0.7557688355445862,
+      "learning_rate": 9.892469145098766e-05,
+      "loss": 0.0889,
+      "step": 20310
+    },
+    {
+      "epoch": 1.329407916257769,
+      "grad_norm": 1.1154882907867432,
+      "learning_rate": 9.892279578480783e-05,
+      "loss": 0.0977,
+      "step": 20320
+    },
+    {
+      "epoch": 1.3300621524370297,
+      "grad_norm": 1.0034886598587036,
+      "learning_rate": 9.892089846735533e-05,
+      "loss": 0.0942,
+      "step": 20330
+    },
+    {
+      "epoch": 1.3307163886162905,
+      "grad_norm": 0.837894082069397,
+      "learning_rate": 9.891899949869418e-05,
+      "loss": 0.1187,
+      "step": 20340
+    },
+    {
+      "epoch": 1.3313706247955512,
+      "grad_norm": 0.8454033732414246,
+      "learning_rate": 9.891709887888853e-05,
+      "loss": 0.1037,
+      "step": 20350
+    },
+    {
+      "epoch": 1.332024860974812,
+      "grad_norm": 0.8291163444519043,
+      "learning_rate": 9.89151966080025e-05,
+      "loss": 0.1154,
+      "step": 20360
+    },
+    {
+      "epoch": 1.3326790971540725,
+      "grad_norm": 0.9798288941383362,
+      "learning_rate": 9.89132926861003e-05,
+      "loss": 0.1007,
+      "step": 20370
+    },
+    {
+      "epoch": 1.3333333333333333,
+      "grad_norm": 1.067610740661621,
+      "learning_rate": 9.891138711324619e-05,
+      "loss": 0.0952,
+      "step": 20380
+    },
+    {
+      "epoch": 1.333987569512594,
+      "grad_norm": 0.8580688834190369,
+      "learning_rate": 9.89094798895045e-05,
+      "loss": 0.0965,
+      "step": 20390
+    },
+    {
+      "epoch": 1.3346418056918548,
+      "grad_norm": 0.9280464053153992,
+      "learning_rate": 9.890757101493958e-05,
+      "loss": 0.1037,
+      "step": 20400
+    },
+    {
+      "epoch": 1.3352960418711155,
+      "grad_norm": 0.8157536387443542,
+      "learning_rate": 9.890566048961587e-05,
+      "loss": 0.097,
+      "step": 20410
+    },
+    {
+      "epoch": 1.3359502780503762,
+      "grad_norm": 0.8738055229187012,
+      "learning_rate": 9.890374831359787e-05,
+      "loss": 0.0991,
+      "step": 20420
+    },
+    {
+      "epoch": 1.336604514229637,
+      "grad_norm": 0.9500564932823181,
+      "learning_rate": 9.89018344869501e-05,
+      "loss": 0.1093,
+      "step": 20430
+    },
+    {
+      "epoch": 1.3372587504088975,
+      "grad_norm": 0.8507956266403198,
+      "learning_rate": 9.889991900973717e-05,
+      "loss": 0.1148,
+      "step": 20440
+    },
+    {
+      "epoch": 1.3379129865881583,
+      "grad_norm": 0.8578815460205078,
+      "learning_rate": 9.889800188202374e-05,
+      "loss": 0.1032,
+      "step": 20450
+    },
+    {
+      "epoch": 1.338567222767419,
+      "grad_norm": 0.8666653633117676,
+      "learning_rate": 9.889608310387449e-05,
+      "loss": 0.1048,
+      "step": 20460
+    },
+    {
+      "epoch": 1.3392214589466798,
+      "grad_norm": 0.7998878359794617,
+      "learning_rate": 9.88941626753542e-05,
+      "loss": 0.103,
+      "step": 20470
+    },
+    {
+      "epoch": 1.3398756951259405,
+      "grad_norm": 0.8280777931213379,
+      "learning_rate": 9.889224059652771e-05,
+      "loss": 0.107,
+      "step": 20480
+    },
+    {
+      "epoch": 1.340529931305201,
+      "grad_norm": 0.9405080676078796,
+      "learning_rate": 9.889031686745987e-05,
+      "loss": 0.1129,
+      "step": 20490
+    },
+    {
+      "epoch": 1.341184167484462,
+      "grad_norm": 1.0205219984054565,
+      "learning_rate": 9.88883914882156e-05,
+      "loss": 0.101,
+      "step": 20500
+    },
+    {
+      "epoch": 1.3418384036637225,
+      "grad_norm": 0.8715194463729858,
+      "learning_rate": 9.888646445885991e-05,
+      "loss": 0.1148,
+      "step": 20510
+    },
+    {
+      "epoch": 1.3424926398429833,
+      "grad_norm": 0.9229245185852051,
+      "learning_rate": 9.888453577945784e-05,
+      "loss": 0.0971,
+      "step": 20520
+    },
+    {
+      "epoch": 1.343146876022244,
+      "grad_norm": 0.9182971119880676,
+      "learning_rate": 9.888260545007448e-05,
+      "loss": 0.0908,
+      "step": 20530
+    },
+    {
+      "epoch": 1.3438011122015048,
+      "grad_norm": 0.9435850977897644,
+      "learning_rate": 9.888067347077499e-05,
+      "loss": 0.1113,
+      "step": 20540
+    },
+    {
+      "epoch": 1.3444553483807655,
+      "grad_norm": 0.9361865520477295,
+      "learning_rate": 9.887873984162457e-05,
+      "loss": 0.1119,
+      "step": 20550
+    },
+    {
+      "epoch": 1.345109584560026,
+      "grad_norm": 0.9154676198959351,
+      "learning_rate": 9.887680456268848e-05,
+      "loss": 0.1064,
+      "step": 20560
+    },
+    {
+      "epoch": 1.345763820739287,
+      "grad_norm": 1.0098940134048462,
+      "learning_rate": 9.887486763403207e-05,
+      "loss": 0.0951,
+      "step": 20570
+    },
+    {
+      "epoch": 1.3464180569185475,
+      "grad_norm": 0.7353399991989136,
+      "learning_rate": 9.88729290557207e-05,
+      "loss": 0.0975,
+      "step": 20580
+    },
+    {
+      "epoch": 1.3470722930978083,
+      "grad_norm": 0.7690742611885071,
+      "learning_rate": 9.88709888278198e-05,
+      "loss": 0.0931,
+      "step": 20590
+    },
+    {
+      "epoch": 1.347726529277069,
+      "grad_norm": 0.780418872833252,
+      "learning_rate": 9.886904695039484e-05,
+      "loss": 0.0957,
+      "step": 20600
+    },
+    {
+      "epoch": 1.3483807654563298,
+      "grad_norm": 0.817748486995697,
+      "learning_rate": 9.88671034235114e-05,
+      "loss": 0.1011,
+      "step": 20610
+    },
+    {
+      "epoch": 1.3490350016355905,
+      "grad_norm": 0.9198586940765381,
+      "learning_rate": 9.886515824723505e-05,
+      "loss": 0.0976,
+      "step": 20620
+    },
+    {
+      "epoch": 1.349689237814851,
+      "grad_norm": 0.8738722205162048,
+      "learning_rate": 9.886321142163147e-05,
+      "loss": 0.1132,
+      "step": 20630
+    },
+    {
+      "epoch": 1.3503434739941118,
+      "grad_norm": 0.7624223828315735,
+      "learning_rate": 9.886126294676634e-05,
+      "loss": 0.1022,
+      "step": 20640
+    },
+    {
+      "epoch": 1.3509977101733726,
+      "grad_norm": 0.9689080715179443,
+      "learning_rate": 9.885931282270545e-05,
+      "loss": 0.1014,
+      "step": 20650
+    },
+    {
+      "epoch": 1.3516519463526333,
+      "grad_norm": 0.7586194276809692,
+      "learning_rate": 9.885736104951462e-05,
+      "loss": 0.0962,
+      "step": 20660
+    },
+    {
+      "epoch": 1.352306182531894,
+      "grad_norm": 0.8439247012138367,
+      "learning_rate": 9.88554076272597e-05,
+      "loss": 0.0948,
+      "step": 20670
+    },
+    {
+      "epoch": 1.3529604187111548,
+      "grad_norm": 1.0176705121994019,
+      "learning_rate": 9.885345255600666e-05,
+      "loss": 0.1043,
+      "step": 20680
+    },
+    {
+      "epoch": 1.3536146548904155,
+      "grad_norm": 0.753044843673706,
+      "learning_rate": 9.885149583582148e-05,
+      "loss": 0.0954,
+      "step": 20690
+    },
+    {
+      "epoch": 1.354268891069676,
+      "grad_norm": 0.9865116477012634,
+      "learning_rate": 9.884953746677019e-05,
+      "loss": 0.1019,
+      "step": 20700
+    },
+    {
+      "epoch": 1.3549231272489368,
+      "grad_norm": 0.8925600051879883,
+      "learning_rate": 9.88475774489189e-05,
+      "loss": 0.1057,
+      "step": 20710
+    },
+    {
+      "epoch": 1.3555773634281976,
+      "grad_norm": 0.8571544289588928,
+      "learning_rate": 9.884561578233375e-05,
+      "loss": 0.1052,
+      "step": 20720
+    },
+    {
+      "epoch": 1.3562315996074583,
+      "grad_norm": 0.7800583839416504,
+      "learning_rate": 9.884365246708098e-05,
+      "loss": 0.0942,
+      "step": 20730
+    },
+    {
+      "epoch": 1.356885835786719,
+      "grad_norm": 0.9630329608917236,
+      "learning_rate": 9.884168750322684e-05,
+      "loss": 0.1151,
+      "step": 20740
+    },
+    {
+      "epoch": 1.3575400719659796,
+      "grad_norm": 1.2140957117080688,
+      "learning_rate": 9.883972089083766e-05,
+      "loss": 0.1038,
+      "step": 20750
+    },
+    {
+      "epoch": 1.3581943081452406,
+      "grad_norm": 0.7877089977264404,
+      "learning_rate": 9.883775262997981e-05,
+      "loss": 0.0968,
+      "step": 20760
+    },
+    {
+      "epoch": 1.358848544324501,
+      "grad_norm": 0.9409202933311462,
+      "learning_rate": 9.883578272071971e-05,
+      "loss": 0.1003,
+      "step": 20770
+    },
+    {
+      "epoch": 1.3595027805037618,
+      "grad_norm": 1.0525962114334106,
+      "learning_rate": 9.883381116312389e-05,
+      "loss": 0.0987,
+      "step": 20780
+    },
+    {
+      "epoch": 1.3601570166830226,
+      "grad_norm": 0.8905682563781738,
+      "learning_rate": 9.883183795725885e-05,
+      "loss": 0.0976,
+      "step": 20790
+    },
+    {
+      "epoch": 1.3608112528622833,
+      "grad_norm": 0.860859215259552,
+      "learning_rate": 9.882986310319124e-05,
+      "loss": 0.1083,
+      "step": 20800
+    },
+    {
+      "epoch": 1.361465489041544,
+      "grad_norm": 1.0243405103683472,
+      "learning_rate": 9.882788660098768e-05,
+      "loss": 0.0915,
+      "step": 20810
+    },
+    {
+      "epoch": 1.3621197252208046,
+      "grad_norm": 1.0279419422149658,
+      "learning_rate": 9.882590845071487e-05,
+      "loss": 0.1119,
+      "step": 20820
+    },
+    {
+      "epoch": 1.3627739614000653,
+      "grad_norm": 1.0260818004608154,
+      "learning_rate": 9.882392865243961e-05,
+      "loss": 0.105,
+      "step": 20830
+    },
+    {
+      "epoch": 1.363428197579326,
+      "grad_norm": 0.9079784750938416,
+      "learning_rate": 9.882194720622873e-05,
+      "loss": 0.0935,
+      "step": 20840
+    },
+    {
+      "epoch": 1.3640824337585868,
+      "grad_norm": 0.8685601949691772,
+      "learning_rate": 9.881996411214906e-05,
+      "loss": 0.0976,
+      "step": 20850
+    },
+    {
+      "epoch": 1.3647366699378476,
+      "grad_norm": 0.7832821011543274,
+      "learning_rate": 9.88179793702676e-05,
+      "loss": 0.1072,
+      "step": 20860
+    },
+    {
+      "epoch": 1.3653909061171083,
+      "grad_norm": 0.6875680685043335,
+      "learning_rate": 9.88159929806513e-05,
+      "loss": 0.0963,
+      "step": 20870
+    },
+    {
+      "epoch": 1.366045142296369,
+      "grad_norm": 0.8509207367897034,
+      "learning_rate": 9.881400494336719e-05,
+      "loss": 0.1061,
+      "step": 20880
+    },
+    {
+      "epoch": 1.3666993784756296,
+      "grad_norm": 0.7752100229263306,
+      "learning_rate": 9.88120152584824e-05,
+      "loss": 0.1041,
+      "step": 20890
+    },
+    {
+      "epoch": 1.3673536146548904,
+      "grad_norm": 0.7963639497756958,
+      "learning_rate": 9.88100239260641e-05,
+      "loss": 0.1042,
+      "step": 20900
+    },
+    {
+      "epoch": 1.368007850834151,
+      "grad_norm": 0.8341352939605713,
+      "learning_rate": 9.880803094617948e-05,
+      "loss": 0.0957,
+      "step": 20910
+    },
+    {
+      "epoch": 1.3686620870134119,
+      "grad_norm": 0.7604305744171143,
+      "learning_rate": 9.88060363188958e-05,
+      "loss": 0.1033,
+      "step": 20920
+    },
+    {
+      "epoch": 1.3693163231926726,
+      "grad_norm": 0.8400859236717224,
+      "learning_rate": 9.880404004428039e-05,
+      "loss": 0.0955,
+      "step": 20930
+    },
+    {
+      "epoch": 1.3699705593719333,
+      "grad_norm": 0.8343890905380249,
+      "learning_rate": 9.880204212240065e-05,
+      "loss": 0.0925,
+      "step": 20940
+    },
+    {
+      "epoch": 1.370624795551194,
+      "grad_norm": 0.9283037781715393,
+      "learning_rate": 9.880004255332399e-05,
+      "loss": 0.1094,
+      "step": 20950
+    },
+    {
+      "epoch": 1.3712790317304546,
+      "grad_norm": 1.0284000635147095,
+      "learning_rate": 9.879804133711792e-05,
+      "loss": 0.0951,
+      "step": 20960
+    },
+    {
+      "epoch": 1.3719332679097154,
+      "grad_norm": 1.0388516187667847,
+      "learning_rate": 9.879603847384997e-05,
+      "loss": 0.106,
+      "step": 20970
+    },
+    {
+      "epoch": 1.3725875040889761,
+      "grad_norm": 0.9273998141288757,
+      "learning_rate": 9.879403396358775e-05,
+      "loss": 0.0964,
+      "step": 20980
+    },
+    {
+      "epoch": 1.3732417402682369,
+      "grad_norm": 0.8812925815582275,
+      "learning_rate": 9.879202780639892e-05,
+      "loss": 0.1022,
+      "step": 20990
+    },
+    {
+      "epoch": 1.3738959764474976,
+      "grad_norm": 0.8167784214019775,
+      "learning_rate": 9.87900200023512e-05,
+      "loss": 0.113,
+      "step": 21000
+    },
+    {
+      "epoch": 1.3745502126267581,
+      "grad_norm": 0.961060106754303,
+      "learning_rate": 9.878801055151232e-05,
+      "loss": 0.0984,
+      "step": 21010
+    },
+    {
+      "epoch": 1.375204448806019,
+      "grad_norm": 1.0047707557678223,
+      "learning_rate": 9.878599945395015e-05,
+      "loss": 0.0912,
+      "step": 21020
+    },
+    {
+      "epoch": 1.3758586849852796,
+      "grad_norm": 1.0364177227020264,
+      "learning_rate": 9.878398670973256e-05,
+      "loss": 0.1161,
+      "step": 21030
+    },
+    {
+      "epoch": 1.3765129211645404,
+      "grad_norm": 0.7439346313476562,
+      "learning_rate": 9.878197231892747e-05,
+      "loss": 0.1047,
+      "step": 21040
+    },
+    {
+      "epoch": 1.3771671573438011,
+      "grad_norm": 0.9560453295707703,
+      "learning_rate": 9.877995628160288e-05,
+      "loss": 0.0911,
+      "step": 21050
+    },
+    {
+      "epoch": 1.3778213935230619,
+      "grad_norm": 0.8289201855659485,
+      "learning_rate": 9.877793859782683e-05,
+      "loss": 0.092,
+      "step": 21060
+    },
+    {
+      "epoch": 1.3784756297023226,
+      "grad_norm": 0.9151121377944946,
+      "learning_rate": 9.877591926766743e-05,
+      "loss": 0.0996,
+      "step": 21070
+    },
+    {
+      "epoch": 1.3791298658815832,
+      "grad_norm": 0.8370741605758667,
+      "learning_rate": 9.877389829119284e-05,
+      "loss": 0.0936,
+      "step": 21080
+    },
+    {
+      "epoch": 1.379784102060844,
+      "grad_norm": 0.9695648550987244,
+      "learning_rate": 9.877187566847125e-05,
+      "loss": 0.0963,
+      "step": 21090
+    },
+    {
+      "epoch": 1.3804383382401046,
+      "grad_norm": 0.9655565619468689,
+      "learning_rate": 9.876985139957098e-05,
+      "loss": 0.1032,
+      "step": 21100
+    },
+    {
+      "epoch": 1.3810925744193654,
+      "grad_norm": 0.9897701144218445,
+      "learning_rate": 9.876782548456029e-05,
+      "loss": 0.1066,
+      "step": 21110
+    },
+    {
+      "epoch": 1.3817468105986261,
+      "grad_norm": 1.0255415439605713,
+      "learning_rate": 9.876579792350762e-05,
+      "loss": 0.1065,
+      "step": 21120
+    },
+    {
+      "epoch": 1.3824010467778869,
+      "grad_norm": 0.9819886684417725,
+      "learning_rate": 9.876376871648137e-05,
+      "loss": 0.0994,
+      "step": 21130
+    },
+    {
+      "epoch": 1.3830552829571476,
+      "grad_norm": 0.9188646078109741,
+      "learning_rate": 9.876173786355003e-05,
+      "loss": 0.107,
+      "step": 21140
+    },
+    {
+      "epoch": 1.3837095191364082,
+      "grad_norm": 0.8057947158813477,
+      "learning_rate": 9.875970536478213e-05,
+      "loss": 0.0964,
+      "step": 21150
+    },
+    {
+      "epoch": 1.384363755315669,
+      "grad_norm": 0.8052737712860107,
+      "learning_rate": 9.875767122024634e-05,
+      "loss": 0.1005,
+      "step": 21160
+    },
+    {
+      "epoch": 1.3850179914949297,
+      "grad_norm": 0.8417797684669495,
+      "learning_rate": 9.875563543001125e-05,
+      "loss": 0.0942,
+      "step": 21170
+    },
+    {
+      "epoch": 1.3856722276741904,
+      "grad_norm": 1.070704698562622,
+      "learning_rate": 9.875359799414561e-05,
+      "loss": 0.099,
+      "step": 21180
+    },
+    {
+      "epoch": 1.3863264638534512,
+      "grad_norm": 0.871180534362793,
+      "learning_rate": 9.875155891271817e-05,
+      "loss": 0.0993,
+      "step": 21190
+    },
+    {
+      "epoch": 1.3869807000327117,
+      "grad_norm": 0.9643855690956116,
+      "learning_rate": 9.874951818579776e-05,
+      "loss": 0.1023,
+      "step": 21200
+    },
+    {
+      "epoch": 1.3876349362119726,
+      "grad_norm": 0.8685986399650574,
+      "learning_rate": 9.874747581345328e-05,
+      "loss": 0.0985,
+      "step": 21210
+    },
+    {
+      "epoch": 1.3882891723912332,
+      "grad_norm": 0.7610760927200317,
+      "learning_rate": 9.874543179575362e-05,
+      "loss": 0.0978,
+      "step": 21220
+    },
+    {
+      "epoch": 1.388943408570494,
+      "grad_norm": 0.7402287721633911,
+      "learning_rate": 9.874338613276781e-05,
+      "loss": 0.1047,
+      "step": 21230
+    },
+    {
+      "epoch": 1.3895976447497547,
+      "grad_norm": 0.8294671773910522,
+      "learning_rate": 9.874133882456489e-05,
+      "loss": 0.0976,
+      "step": 21240
+    },
+    {
+      "epoch": 1.3902518809290154,
+      "grad_norm": 1.0377382040023804,
+      "learning_rate": 9.873928987121394e-05,
+      "loss": 0.1077,
+      "step": 21250
+    },
+    {
+      "epoch": 1.3909061171082762,
+      "grad_norm": 0.8378622531890869,
+      "learning_rate": 9.873723927278414e-05,
+      "loss": 0.0988,
+      "step": 21260
+    },
+    {
+      "epoch": 1.3915603532875367,
+      "grad_norm": 0.7670599222183228,
+      "learning_rate": 9.87351870293447e-05,
+      "loss": 0.1022,
+      "step": 21270
+    },
+    {
+      "epoch": 1.3922145894667974,
+      "grad_norm": 0.7315048575401306,
+      "learning_rate": 9.873313314096488e-05,
+      "loss": 0.0977,
+      "step": 21280
+    },
+    {
+      "epoch": 1.3928688256460582,
+      "grad_norm": 0.960702121257782,
+      "learning_rate": 9.873107760771401e-05,
+      "loss": 0.0966,
+      "step": 21290
+    },
+    {
+      "epoch": 1.393523061825319,
+      "grad_norm": 0.7483635544776917,
+      "learning_rate": 9.872902042966147e-05,
+      "loss": 0.0883,
+      "step": 21300
+    },
+    {
+      "epoch": 1.3941772980045797,
+      "grad_norm": 0.8449085354804993,
+      "learning_rate": 9.872696160687669e-05,
+      "loss": 0.1002,
+      "step": 21310
+    },
+    {
+      "epoch": 1.3948315341838404,
+      "grad_norm": 0.8629240393638611,
+      "learning_rate": 9.872490113942918e-05,
+      "loss": 0.0977,
+      "step": 21320
+    },
+    {
+      "epoch": 1.3954857703631012,
+      "grad_norm": 0.8286533951759338,
+      "learning_rate": 9.872283902738845e-05,
+      "loss": 0.0968,
+      "step": 21330
+    },
+    {
+      "epoch": 1.3961400065423617,
+      "grad_norm": 0.8153654932975769,
+      "learning_rate": 9.872077527082413e-05,
+      "loss": 0.1113,
+      "step": 21340
+    },
+    {
+      "epoch": 1.3967942427216224,
+      "grad_norm": 0.7683613896369934,
+      "learning_rate": 9.871870986980587e-05,
+      "loss": 0.0887,
+      "step": 21350
+    },
+    {
+      "epoch": 1.3974484789008832,
+      "grad_norm": 1.0355619192123413,
+      "learning_rate": 9.871664282440339e-05,
+      "loss": 0.1074,
+      "step": 21360
+    },
+    {
+      "epoch": 1.398102715080144,
+      "grad_norm": 0.8770645260810852,
+      "learning_rate": 9.871457413468644e-05,
+      "loss": 0.0993,
+      "step": 21370
+    },
+    {
+      "epoch": 1.3987569512594047,
+      "grad_norm": 0.709586501121521,
+      "learning_rate": 9.871250380072487e-05,
+      "loss": 0.094,
+      "step": 21380
+    },
+    {
+      "epoch": 1.3994111874386654,
+      "grad_norm": 0.7333911061286926,
+      "learning_rate": 9.871043182258852e-05,
+      "loss": 0.1037,
+      "step": 21390
+    },
+    {
+      "epoch": 1.4000654236179262,
+      "grad_norm": 0.7974872589111328,
+      "learning_rate": 9.870835820034736e-05,
+      "loss": 0.0953,
+      "step": 21400
+    },
+    {
+      "epoch": 1.4007196597971867,
+      "grad_norm": 0.8069256544113159,
+      "learning_rate": 9.870628293407138e-05,
+      "loss": 0.0922,
+      "step": 21410
+    },
+    {
+      "epoch": 1.4013738959764475,
+      "grad_norm": 0.8490813374519348,
+      "learning_rate": 9.870420602383059e-05,
+      "loss": 0.1018,
+      "step": 21420
+    },
+    {
+      "epoch": 1.4020281321557082,
+      "grad_norm": 0.9025737643241882,
+      "learning_rate": 9.870212746969514e-05,
+      "loss": 0.0992,
+      "step": 21430
+    },
+    {
+      "epoch": 1.402682368334969,
+      "grad_norm": 0.7554858922958374,
+      "learning_rate": 9.870004727173514e-05,
+      "loss": 0.1055,
+      "step": 21440
+    },
+    {
+      "epoch": 1.4033366045142297,
+      "grad_norm": 0.9373718500137329,
+      "learning_rate": 9.869796543002083e-05,
+      "loss": 0.1052,
+      "step": 21450
+    },
+    {
+      "epoch": 1.4039908406934902,
+      "grad_norm": 0.916695237159729,
+      "learning_rate": 9.869588194462249e-05,
+      "loss": 0.103,
+      "step": 21460
+    },
+    {
+      "epoch": 1.4046450768727512,
+      "grad_norm": 1.047790288925171,
+      "learning_rate": 9.869379681561041e-05,
+      "loss": 0.0934,
+      "step": 21470
+    },
+    {
+      "epoch": 1.4052993130520117,
+      "grad_norm": 0.8690340518951416,
+      "learning_rate": 9.869171004305497e-05,
+      "loss": 0.0933,
+      "step": 21480
+    },
+    {
+      "epoch": 1.4059535492312725,
+      "grad_norm": 0.8632329106330872,
+      "learning_rate": 9.868962162702664e-05,
+      "loss": 0.1004,
+      "step": 21490
+    },
+    {
+      "epoch": 1.4066077854105332,
+      "grad_norm": 0.8703600168228149,
+      "learning_rate": 9.868753156759587e-05,
+      "loss": 0.0961,
+      "step": 21500
+    },
+    {
+      "epoch": 1.407262021589794,
+      "grad_norm": 0.9060193300247192,
+      "learning_rate": 9.868543986483325e-05,
+      "loss": 0.1021,
+      "step": 21510
+    },
+    {
+      "epoch": 1.4079162577690547,
+      "grad_norm": 1.0863100290298462,
+      "learning_rate": 9.868334651880932e-05,
+      "loss": 0.1095,
+      "step": 21520
+    },
+    {
+      "epoch": 1.4085704939483152,
+      "grad_norm": 0.837095320224762,
+      "learning_rate": 9.868125152959477e-05,
+      "loss": 0.0981,
+      "step": 21530
+    },
+    {
+      "epoch": 1.409224730127576,
+      "grad_norm": 1.0157208442687988,
+      "learning_rate": 9.867915489726034e-05,
+      "loss": 0.102,
+      "step": 21540
+    },
+    {
+      "epoch": 1.4098789663068367,
+      "grad_norm": 0.9771285057067871,
+      "learning_rate": 9.867705662187673e-05,
+      "loss": 0.1063,
+      "step": 21550
+    },
+    {
+      "epoch": 1.4105332024860975,
+      "grad_norm": 0.9457305669784546,
+      "learning_rate": 9.867495670351483e-05,
+      "loss": 0.094,
+      "step": 21560
+    },
+    {
+      "epoch": 1.4111874386653582,
+      "grad_norm": 0.90764981508255,
+      "learning_rate": 9.867285514224547e-05,
+      "loss": 0.0918,
+      "step": 21570
+    },
+    {
+      "epoch": 1.411841674844619,
+      "grad_norm": 0.8110638856887817,
+      "learning_rate": 9.867075193813959e-05,
+      "loss": 0.1003,
+      "step": 21580
+    },
+    {
+      "epoch": 1.4124959110238797,
+      "grad_norm": 1.1416128873825073,
+      "learning_rate": 9.866864709126821e-05,
+      "loss": 0.1112,
+      "step": 21590
+    },
+    {
+      "epoch": 1.4131501472031402,
+      "grad_norm": 1.088036060333252,
+      "learning_rate": 9.866654060170234e-05,
+      "loss": 0.1099,
+      "step": 21600
+    },
+    {
+      "epoch": 1.413804383382401,
+      "grad_norm": 0.9464057087898254,
+      "learning_rate": 9.866443246951308e-05,
+      "loss": 0.0975,
+      "step": 21610
+    },
+    {
+      "epoch": 1.4144586195616617,
+      "grad_norm": 0.8685123324394226,
+      "learning_rate": 9.866232269477162e-05,
+      "loss": 0.0942,
+      "step": 21620
+    },
+    {
+      "epoch": 1.4151128557409225,
+      "grad_norm": 0.7753238677978516,
+      "learning_rate": 9.866021127754915e-05,
+      "loss": 0.0973,
+      "step": 21630
+    },
+    {
+      "epoch": 1.4157670919201832,
+      "grad_norm": 0.8483602404594421,
+      "learning_rate": 9.865809821791692e-05,
+      "loss": 0.1016,
+      "step": 21640
+    },
+    {
+      "epoch": 1.4164213280994438,
+      "grad_norm": 0.712035596370697,
+      "learning_rate": 9.865598351594627e-05,
+      "loss": 0.0985,
+      "step": 21650
+    },
+    {
+      "epoch": 1.4170755642787047,
+      "grad_norm": 0.87371426820755,
+      "learning_rate": 9.865386717170856e-05,
+      "loss": 0.0952,
+      "step": 21660
+    },
+    {
+      "epoch": 1.4177298004579653,
+      "grad_norm": 1.058411717414856,
+      "learning_rate": 9.865174918527525e-05,
+      "loss": 0.0956,
+      "step": 21670
+    },
+    {
+      "epoch": 1.418384036637226,
+      "grad_norm": 0.7664424777030945,
+      "learning_rate": 9.864962955671779e-05,
+      "loss": 0.0889,
+      "step": 21680
+    },
+    {
+      "epoch": 1.4190382728164868,
+      "grad_norm": 0.9568914771080017,
+      "learning_rate": 9.864750828610776e-05,
+      "loss": 0.102,
+      "step": 21690
+    },
+    {
+      "epoch": 1.4196925089957475,
+      "grad_norm": 0.8429945707321167,
+      "learning_rate": 9.864538537351675e-05,
+      "loss": 0.1018,
+      "step": 21700
+    },
+    {
+      "epoch": 1.4203467451750083,
+      "grad_norm": 1.3158056735992432,
+      "learning_rate": 9.864326081901639e-05,
+      "loss": 0.0934,
+      "step": 21710
+    },
+    {
+      "epoch": 1.4210009813542688,
+      "grad_norm": 0.8673258423805237,
+      "learning_rate": 9.864113462267841e-05,
+      "loss": 0.0917,
+      "step": 21720
+    },
+    {
+      "epoch": 1.4216552175335295,
+      "grad_norm": 1.0829505920410156,
+      "learning_rate": 9.863900678457457e-05,
+      "loss": 0.1176,
+      "step": 21730
+    },
+    {
+      "epoch": 1.4223094537127903,
+      "grad_norm": 0.9344543218612671,
+      "learning_rate": 9.86368773047767e-05,
+      "loss": 0.0962,
+      "step": 21740
+    },
+    {
+      "epoch": 1.422963689892051,
+      "grad_norm": 0.9904654026031494,
+      "learning_rate": 9.863474618335666e-05,
+      "loss": 0.1118,
+      "step": 21750
+    },
+    {
+      "epoch": 1.4236179260713118,
+      "grad_norm": 0.864806592464447,
+      "learning_rate": 9.863261342038639e-05,
+      "loss": 0.0914,
+      "step": 21760
+    },
+    {
+      "epoch": 1.4242721622505725,
+      "grad_norm": 0.8543987274169922,
+      "learning_rate": 9.863047901593786e-05,
+      "loss": 0.1041,
+      "step": 21770
+    },
+    {
+      "epoch": 1.4249263984298333,
+      "grad_norm": 0.8456289172172546,
+      "learning_rate": 9.862834297008314e-05,
+      "loss": 0.1148,
+      "step": 21780
+    },
+    {
+      "epoch": 1.4255806346090938,
+      "grad_norm": 0.9336904287338257,
+      "learning_rate": 9.862620528289431e-05,
+      "loss": 0.1032,
+      "step": 21790
+    },
+    {
+      "epoch": 1.4262348707883545,
+      "grad_norm": 0.9453654885292053,
+      "learning_rate": 9.862406595444351e-05,
+      "loss": 0.1059,
+      "step": 21800
+    },
+    {
+      "epoch": 1.4268891069676153,
+      "grad_norm": 1.0138676166534424,
+      "learning_rate": 9.862192498480299e-05,
+      "loss": 0.0979,
+      "step": 21810
+    },
+    {
+      "epoch": 1.427543343146876,
+      "grad_norm": 1.177456259727478,
+      "learning_rate": 9.861978237404496e-05,
+      "loss": 0.0932,
+      "step": 21820
+    },
+    {
+      "epoch": 1.4281975793261368,
+      "grad_norm": 0.792131781578064,
+      "learning_rate": 9.861763812224177e-05,
+      "loss": 0.0958,
+      "step": 21830
+    },
+    {
+      "epoch": 1.4288518155053975,
+      "grad_norm": 0.7728269696235657,
+      "learning_rate": 9.86154922294658e-05,
+      "loss": 0.0975,
+      "step": 21840
+    },
+    {
+      "epoch": 1.4295060516846583,
+      "grad_norm": 0.7983576059341431,
+      "learning_rate": 9.861334469578946e-05,
+      "loss": 0.0987,
+      "step": 21850
+    },
+    {
+      "epoch": 1.4301602878639188,
+      "grad_norm": 0.8768244385719299,
+      "learning_rate": 9.861119552128523e-05,
+      "loss": 0.0961,
+      "step": 21860
+    },
+    {
+      "epoch": 1.4308145240431795,
+      "grad_norm": 1.0018515586853027,
+      "learning_rate": 9.86090447060257e-05,
+      "loss": 0.0906,
+      "step": 21870
+    },
+    {
+      "epoch": 1.4314687602224403,
+      "grad_norm": 0.9329746961593628,
+      "learning_rate": 9.86068922500834e-05,
+      "loss": 0.0925,
+      "step": 21880
+    },
+    {
+      "epoch": 1.432122996401701,
+      "grad_norm": 0.8282088041305542,
+      "learning_rate": 9.860473815353102e-05,
+      "loss": 0.1014,
+      "step": 21890
+    },
+    {
+      "epoch": 1.4327772325809618,
+      "grad_norm": 0.7623408436775208,
+      "learning_rate": 9.860258241644126e-05,
+      "loss": 0.1107,
+      "step": 21900
+    },
+    {
+      "epoch": 1.4334314687602223,
+      "grad_norm": 0.8091941475868225,
+      "learning_rate": 9.860042503888687e-05,
+      "loss": 0.0936,
+      "step": 21910
+    },
+    {
+      "epoch": 1.4340857049394833,
+      "grad_norm": 0.7656387686729431,
+      "learning_rate": 9.859826602094068e-05,
+      "loss": 0.091,
+      "step": 21920
+    },
+    {
+      "epoch": 1.4347399411187438,
+      "grad_norm": 1.1164404153823853,
+      "learning_rate": 9.859610536267556e-05,
+      "loss": 0.1016,
+      "step": 21930
+    },
+    {
+      "epoch": 1.4353941772980046,
+      "grad_norm": 0.9509071707725525,
+      "learning_rate": 9.859394306416444e-05,
+      "loss": 0.0934,
+      "step": 21940
+    },
+    {
+      "epoch": 1.4360484134772653,
+      "grad_norm": 0.9692185521125793,
+      "learning_rate": 9.859177912548028e-05,
+      "loss": 0.1056,
+      "step": 21950
+    },
+    {
+      "epoch": 1.436702649656526,
+      "grad_norm": 0.9519806504249573,
+      "learning_rate": 9.858961354669616e-05,
+      "loss": 0.1038,
+      "step": 21960
+    },
+    {
+      "epoch": 1.4373568858357868,
+      "grad_norm": 1.0304690599441528,
+      "learning_rate": 9.858744632788514e-05,
+      "loss": 0.1002,
+      "step": 21970
+    },
+    {
+      "epoch": 1.4380111220150473,
+      "grad_norm": 0.7147766947746277,
+      "learning_rate": 9.858527746912039e-05,
+      "loss": 0.1132,
+      "step": 21980
+    },
+    {
+      "epoch": 1.438665358194308,
+      "grad_norm": 0.9351019263267517,
+      "learning_rate": 9.85831069704751e-05,
+      "loss": 0.0982,
+      "step": 21990
+    },
+    {
+      "epoch": 1.4393195943735688,
+      "grad_norm": 1.303455114364624,
+      "learning_rate": 9.858093483202254e-05,
+      "loss": 0.112,
+      "step": 22000
+    },
+    {
+      "epoch": 1.4399738305528296,
+      "grad_norm": 0.778627872467041,
+      "learning_rate": 9.857876105383602e-05,
+      "loss": 0.1028,
+      "step": 22010
+    },
+    {
+      "epoch": 1.4406280667320903,
+      "grad_norm": 1.0590648651123047,
+      "learning_rate": 9.85765856359889e-05,
+      "loss": 0.1024,
+      "step": 22020
+    },
+    {
+      "epoch": 1.441282302911351,
+      "grad_norm": 0.9106878042221069,
+      "learning_rate": 9.857440857855462e-05,
+      "loss": 0.0975,
+      "step": 22030
+    },
+    {
+      "epoch": 1.4419365390906118,
+      "grad_norm": 0.8523755669593811,
+      "learning_rate": 9.857222988160667e-05,
+      "loss": 0.105,
+      "step": 22040
+    },
+    {
+      "epoch": 1.4425907752698723,
+      "grad_norm": 0.8732689619064331,
+      "learning_rate": 9.857004954521858e-05,
+      "loss": 0.0981,
+      "step": 22050
+    },
+    {
+      "epoch": 1.443245011449133,
+      "grad_norm": 0.9251449704170227,
+      "learning_rate": 9.856786756946392e-05,
+      "loss": 0.0962,
+      "step": 22060
+    },
+    {
+      "epoch": 1.4438992476283938,
+      "grad_norm": 0.7228383421897888,
+      "learning_rate": 9.856568395441637e-05,
+      "loss": 0.0949,
+      "step": 22070
+    },
+    {
+      "epoch": 1.4445534838076546,
+      "grad_norm": 0.7796227931976318,
+      "learning_rate": 9.856349870014961e-05,
+      "loss": 0.0964,
+      "step": 22080
+    },
+    {
+      "epoch": 1.4452077199869153,
+      "grad_norm": 0.8933011889457703,
+      "learning_rate": 9.856131180673742e-05,
+      "loss": 0.0917,
+      "step": 22090
+    },
+    {
+      "epoch": 1.4458619561661759,
+      "grad_norm": 0.6613444685935974,
+      "learning_rate": 9.855912327425359e-05,
+      "loss": 0.0944,
+      "step": 22100
+    },
+    {
+      "epoch": 1.4465161923454368,
+      "grad_norm": 0.8646618723869324,
+      "learning_rate": 9.8556933102772e-05,
+      "loss": 0.0989,
+      "step": 22110
+    },
+    {
+      "epoch": 1.4471704285246973,
+      "grad_norm": 1.0366922616958618,
+      "learning_rate": 9.855474129236657e-05,
+      "loss": 0.0964,
+      "step": 22120
+    },
+    {
+      "epoch": 1.447824664703958,
+      "grad_norm": 0.78702712059021,
+      "learning_rate": 9.855254784311129e-05,
+      "loss": 0.1023,
+      "step": 22130
+    },
+    {
+      "epoch": 1.4484789008832188,
+      "grad_norm": 0.685092568397522,
+      "learning_rate": 9.855035275508017e-05,
+      "loss": 0.0978,
+      "step": 22140
+    },
+    {
+      "epoch": 1.4491331370624796,
+      "grad_norm": 0.9662759900093079,
+      "learning_rate": 9.854815602834733e-05,
+      "loss": 0.101,
+      "step": 22150
+    },
+    {
+      "epoch": 1.4497873732417403,
+      "grad_norm": 0.8414245843887329,
+      "learning_rate": 9.854595766298692e-05,
+      "loss": 0.0887,
+      "step": 22160
+    },
+    {
+      "epoch": 1.4504416094210009,
+      "grad_norm": 0.9167496562004089,
+      "learning_rate": 9.854375765907309e-05,
+      "loss": 0.097,
+      "step": 22170
+    },
+    {
+      "epoch": 1.4510958456002616,
+      "grad_norm": 0.8719131946563721,
+      "learning_rate": 9.854155601668013e-05,
+      "loss": 0.0972,
+      "step": 22180
+    },
+    {
+      "epoch": 1.4517500817795224,
+      "grad_norm": 1.0566436052322388,
+      "learning_rate": 9.853935273588236e-05,
+      "loss": 0.1119,
+      "step": 22190
+    },
+    {
+      "epoch": 1.452404317958783,
+      "grad_norm": 0.8643470406532288,
+      "learning_rate": 9.853714781675414e-05,
+      "loss": 0.1097,
+      "step": 22200
+    },
+    {
+      "epoch": 1.4530585541380439,
+      "grad_norm": 0.8405054211616516,
+      "learning_rate": 9.853494125936989e-05,
+      "loss": 0.1023,
+      "step": 22210
+    },
+    {
+      "epoch": 1.4537127903173046,
+      "grad_norm": 0.8226882219314575,
+      "learning_rate": 9.853273306380407e-05,
+      "loss": 0.1037,
+      "step": 22220
+    },
+    {
+      "epoch": 1.4543670264965654,
+      "grad_norm": 0.9032091498374939,
+      "learning_rate": 9.853052323013124e-05,
+      "loss": 0.09,
+      "step": 22230
+    },
+    {
+      "epoch": 1.4550212626758259,
+      "grad_norm": 0.7893581986427307,
+      "learning_rate": 9.852831175842596e-05,
+      "loss": 0.1065,
+      "step": 22240
+    },
+    {
+      "epoch": 1.4556754988550866,
+      "grad_norm": 0.8432521820068359,
+      "learning_rate": 9.85260986487629e-05,
+      "loss": 0.0988,
+      "step": 22250
+    },
+    {
+      "epoch": 1.4563297350343474,
+      "grad_norm": 0.8194817304611206,
+      "learning_rate": 9.852388390121675e-05,
+      "loss": 0.1014,
+      "step": 22260
+    },
+    {
+      "epoch": 1.4569839712136081,
+      "grad_norm": 0.7900116443634033,
+      "learning_rate": 9.852166751586225e-05,
+      "loss": 0.1056,
+      "step": 22270
+    },
+    {
+      "epoch": 1.4576382073928689,
+      "grad_norm": 0.7171338200569153,
+      "learning_rate": 9.851944949277423e-05,
+      "loss": 0.0906,
+      "step": 22280
+    },
+    {
+      "epoch": 1.4582924435721296,
+      "grad_norm": 0.7563906311988831,
+      "learning_rate": 9.851722983202753e-05,
+      "loss": 0.0959,
+      "step": 22290
+    },
+    {
+      "epoch": 1.4589466797513904,
+      "grad_norm": 0.7728096842765808,
+      "learning_rate": 9.851500853369709e-05,
+      "loss": 0.1029,
+      "step": 22300
+    },
+    {
+      "epoch": 1.4596009159306509,
+      "grad_norm": 0.9391055107116699,
+      "learning_rate": 9.851278559785788e-05,
+      "loss": 0.0979,
+      "step": 22310
+    },
+    {
+      "epoch": 1.4602551521099116,
+      "grad_norm": 0.8876420855522156,
+      "learning_rate": 9.851056102458492e-05,
+      "loss": 0.0963,
+      "step": 22320
+    },
+    {
+      "epoch": 1.4609093882891724,
+      "grad_norm": 1.0009467601776123,
+      "learning_rate": 9.85083348139533e-05,
+      "loss": 0.098,
+      "step": 22330
+    },
+    {
+      "epoch": 1.4615636244684331,
+      "grad_norm": 0.9100044369697571,
+      "learning_rate": 9.850610696603817e-05,
+      "loss": 0.0935,
+      "step": 22340
+    },
+    {
+      "epoch": 1.4622178606476939,
+      "grad_norm": 1.0415489673614502,
+      "learning_rate": 9.850387748091471e-05,
+      "loss": 0.104,
+      "step": 22350
+    },
+    {
+      "epoch": 1.4628720968269544,
+      "grad_norm": 0.9339014887809753,
+      "learning_rate": 9.850164635865819e-05,
+      "loss": 0.1056,
+      "step": 22360
+    },
+    {
+      "epoch": 1.4635263330062154,
+      "grad_norm": 0.9061875343322754,
+      "learning_rate": 9.84994135993439e-05,
+      "loss": 0.1065,
+      "step": 22370
+    },
+    {
+      "epoch": 1.464180569185476,
+      "grad_norm": 0.7294244766235352,
+      "learning_rate": 9.849717920304719e-05,
+      "loss": 0.1058,
+      "step": 22380
+    },
+    {
+      "epoch": 1.4648348053647366,
+      "grad_norm": 0.9181326031684875,
+      "learning_rate": 9.849494316984352e-05,
+      "loss": 0.0996,
+      "step": 22390
+    },
+    {
+      "epoch": 1.4654890415439974,
+      "grad_norm": 0.7590892910957336,
+      "learning_rate": 9.849270549980832e-05,
+      "loss": 0.1013,
+      "step": 22400
+    },
+    {
+      "epoch": 1.4661432777232581,
+      "grad_norm": 0.7892376780509949,
+      "learning_rate": 9.849046619301713e-05,
+      "loss": 0.0953,
+      "step": 22410
+    },
+    {
+      "epoch": 1.466797513902519,
+      "grad_norm": 0.7668907642364502,
+      "learning_rate": 9.848822524954553e-05,
+      "loss": 0.1001,
+      "step": 22420
+    },
+    {
+      "epoch": 1.4674517500817794,
+      "grad_norm": 0.6292420625686646,
+      "learning_rate": 9.848598266946918e-05,
+      "loss": 0.0882,
+      "step": 22430
+    },
+    {
+      "epoch": 1.4681059862610402,
+      "grad_norm": 0.8127971887588501,
+      "learning_rate": 9.848373845286376e-05,
+      "loss": 0.1029,
+      "step": 22440
+    },
+    {
+      "epoch": 1.468760222440301,
+      "grad_norm": 0.9663169980049133,
+      "learning_rate": 9.848149259980499e-05,
+      "loss": 0.1066,
+      "step": 22450
+    },
+    {
+      "epoch": 1.4694144586195617,
+      "grad_norm": 0.8952876925468445,
+      "learning_rate": 9.847924511036872e-05,
+      "loss": 0.1101,
+      "step": 22460
+    },
+    {
+      "epoch": 1.4700686947988224,
+      "grad_norm": 0.9786347150802612,
+      "learning_rate": 9.847699598463079e-05,
+      "loss": 0.0865,
+      "step": 22470
+    },
+    {
+      "epoch": 1.4707229309780832,
+      "grad_norm": 1.0925096273422241,
+      "learning_rate": 9.847474522266708e-05,
+      "loss": 0.1013,
+      "step": 22480
+    },
+    {
+      "epoch": 1.471377167157344,
+      "grad_norm": 0.7084100842475891,
+      "learning_rate": 9.84724928245536e-05,
+      "loss": 0.0975,
+      "step": 22490
+    },
+    {
+      "epoch": 1.4720314033366044,
+      "grad_norm": 0.8973484039306641,
+      "learning_rate": 9.847023879036637e-05,
+      "loss": 0.0956,
+      "step": 22500
+    },
+    {
+      "epoch": 1.4726856395158652,
+      "grad_norm": 1.1839032173156738,
+      "learning_rate": 9.846798312018146e-05,
+      "loss": 0.0958,
+      "step": 22510
+    },
+    {
+      "epoch": 1.473339875695126,
+      "grad_norm": 1.004971981048584,
+      "learning_rate": 9.846572581407502e-05,
+      "loss": 0.1028,
+      "step": 22520
+    },
+    {
+      "epoch": 1.4739941118743867,
+      "grad_norm": 0.854030966758728,
+      "learning_rate": 9.846346687212322e-05,
+      "loss": 0.1093,
+      "step": 22530
+    },
+    {
+      "epoch": 1.4746483480536474,
+      "grad_norm": 0.9426572918891907,
+      "learning_rate": 9.846120629440231e-05,
+      "loss": 0.1084,
+      "step": 22540
+    },
+    {
+      "epoch": 1.475302584232908,
+      "grad_norm": 0.7341740131378174,
+      "learning_rate": 9.84589440809886e-05,
+      "loss": 0.0945,
+      "step": 22550
+    },
+    {
+      "epoch": 1.475956820412169,
+      "grad_norm": 0.7816892862319946,
+      "learning_rate": 9.845668023195841e-05,
+      "loss": 0.0945,
+      "step": 22560
+    },
+    {
+      "epoch": 1.4766110565914294,
+      "grad_norm": 1.015047311782837,
+      "learning_rate": 9.845441474738821e-05,
+      "loss": 0.0912,
+      "step": 22570
+    },
+    {
+      "epoch": 1.4772652927706902,
+      "grad_norm": 1.0794068574905396,
+      "learning_rate": 9.845214762735444e-05,
+      "loss": 0.092,
+      "step": 22580
+    },
+    {
+      "epoch": 1.477919528949951,
+      "grad_norm": 0.8350719809532166,
+      "learning_rate": 9.84498788719336e-05,
+      "loss": 0.1005,
+      "step": 22590
+    },
+    {
+      "epoch": 1.4785737651292117,
+      "grad_norm": 0.7004404664039612,
+      "learning_rate": 9.844760848120229e-05,
+      "loss": 0.0958,
+      "step": 22600
+    },
+    {
+      "epoch": 1.4792280013084724,
+      "grad_norm": 0.9362242221832275,
+      "learning_rate": 9.844533645523714e-05,
+      "loss": 0.0953,
+      "step": 22610
+    },
+    {
+      "epoch": 1.479882237487733,
+      "grad_norm": 1.0264415740966797,
+      "learning_rate": 9.844306279411482e-05,
+      "loss": 0.0954,
+      "step": 22620
+    },
+    {
+      "epoch": 1.4805364736669937,
+      "grad_norm": 0.9986675977706909,
+      "learning_rate": 9.84407874979121e-05,
+      "loss": 0.1014,
+      "step": 22630
+    },
+    {
+      "epoch": 1.4811907098462544,
+      "grad_norm": 0.8744516372680664,
+      "learning_rate": 9.843851056670574e-05,
+      "loss": 0.0955,
+      "step": 22640
+    },
+    {
+      "epoch": 1.4818449460255152,
+      "grad_norm": 0.8320896029472351,
+      "learning_rate": 9.843623200057263e-05,
+      "loss": 0.0927,
+      "step": 22650
+    },
+    {
+      "epoch": 1.482499182204776,
+      "grad_norm": 0.9211897850036621,
+      "learning_rate": 9.843395179958965e-05,
+      "loss": 0.1046,
+      "step": 22660
+    },
+    {
+      "epoch": 1.4831534183840367,
+      "grad_norm": 0.9335985779762268,
+      "learning_rate": 9.84316699638338e-05,
+      "loss": 0.0968,
+      "step": 22670
+    },
+    {
+      "epoch": 1.4838076545632974,
+      "grad_norm": 0.9013245105743408,
+      "learning_rate": 9.842938649338205e-05,
+      "loss": 0.0968,
+      "step": 22680
+    },
+    {
+      "epoch": 1.484461890742558,
+      "grad_norm": 1.0684399604797363,
+      "learning_rate": 9.842710138831148e-05,
+      "loss": 0.0904,
+      "step": 22690
+    },
+    {
+      "epoch": 1.4851161269218187,
+      "grad_norm": 1.023200273513794,
+      "learning_rate": 9.842481464869927e-05,
+      "loss": 0.099,
+      "step": 22700
+    },
+    {
+      "epoch": 1.4857703631010795,
+      "grad_norm": 0.9011964797973633,
+      "learning_rate": 9.842252627462254e-05,
+      "loss": 0.0982,
+      "step": 22710
+    },
+    {
+      "epoch": 1.4864245992803402,
+      "grad_norm": 0.9003528356552124,
+      "learning_rate": 9.842023626615857e-05,
+      "loss": 0.0916,
+      "step": 22720
+    },
+    {
+      "epoch": 1.487078835459601,
+      "grad_norm": 0.8671954274177551,
+      "learning_rate": 9.841794462338463e-05,
+      "loss": 0.0844,
+      "step": 22730
+    },
+    {
+      "epoch": 1.4877330716388617,
+      "grad_norm": 1.0183930397033691,
+      "learning_rate": 9.841565134637808e-05,
+      "loss": 0.1009,
+      "step": 22740
+    },
+    {
+      "epoch": 1.4883873078181225,
+      "grad_norm": 0.77434903383255,
+      "learning_rate": 9.841335643521632e-05,
+      "loss": 0.0979,
+      "step": 22750
+    },
+    {
+      "epoch": 1.489041543997383,
+      "grad_norm": 0.9142956733703613,
+      "learning_rate": 9.841105988997682e-05,
+      "loss": 0.0911,
+      "step": 22760
+    },
+    {
+      "epoch": 1.4896957801766437,
+      "grad_norm": 0.9437901973724365,
+      "learning_rate": 9.840876171073707e-05,
+      "loss": 0.101,
+      "step": 22770
+    },
+    {
+      "epoch": 1.4903500163559045,
+      "grad_norm": 0.8872728943824768,
+      "learning_rate": 9.840646189757468e-05,
+      "loss": 0.0982,
+      "step": 22780
+    },
+    {
+      "epoch": 1.4910042525351652,
+      "grad_norm": 0.7070728540420532,
+      "learning_rate": 9.840416045056724e-05,
+      "loss": 0.0918,
+      "step": 22790
+    },
+    {
+      "epoch": 1.491658488714426,
+      "grad_norm": 0.9747138619422913,
+      "learning_rate": 9.840185736979244e-05,
+      "loss": 0.0953,
+      "step": 22800
+    },
+    {
+      "epoch": 1.4923127248936865,
+      "grad_norm": 0.9739376902580261,
+      "learning_rate": 9.839955265532801e-05,
+      "loss": 0.1069,
+      "step": 22810
+    },
+    {
+      "epoch": 1.4929669610729475,
+      "grad_norm": 0.8057883977890015,
+      "learning_rate": 9.839724630725175e-05,
+      "loss": 0.0882,
+      "step": 22820
+    },
+    {
+      "epoch": 1.493621197252208,
+      "grad_norm": 0.9417016506195068,
+      "learning_rate": 9.839493832564149e-05,
+      "loss": 0.1076,
+      "step": 22830
+    },
+    {
+      "epoch": 1.4942754334314687,
+      "grad_norm": 0.7495242953300476,
+      "learning_rate": 9.839262871057515e-05,
+      "loss": 0.091,
+      "step": 22840
+    },
+    {
+      "epoch": 1.4949296696107295,
+      "grad_norm": 0.818233847618103,
+      "learning_rate": 9.839031746213068e-05,
+      "loss": 0.1056,
+      "step": 22850
+    },
+    {
+      "epoch": 1.4955839057899902,
+      "grad_norm": 0.8192006945610046,
+      "learning_rate": 9.838800458038609e-05,
+      "loss": 0.0928,
+      "step": 22860
+    },
+    {
+      "epoch": 1.496238141969251,
+      "grad_norm": 0.9962643980979919,
+      "learning_rate": 9.838569006541944e-05,
+      "loss": 0.097,
+      "step": 22870
+    },
+    {
+      "epoch": 1.4968923781485115,
+      "grad_norm": 0.8884673714637756,
+      "learning_rate": 9.838337391730886e-05,
+      "loss": 0.0924,
+      "step": 22880
+    },
+    {
+      "epoch": 1.4975466143277723,
+      "grad_norm": 0.7692938446998596,
+      "learning_rate": 9.83810561361325e-05,
+      "loss": 0.1011,
+      "step": 22890
+    },
+    {
+      "epoch": 1.498200850507033,
+      "grad_norm": 1.0276055335998535,
+      "learning_rate": 9.837873672196863e-05,
+      "loss": 0.097,
+      "step": 22900
+    },
+    {
+      "epoch": 1.4988550866862937,
+      "grad_norm": 0.9219088554382324,
+      "learning_rate": 9.83764156748955e-05,
+      "loss": 0.0843,
+      "step": 22910
+    },
+    {
+      "epoch": 1.4995093228655545,
+      "grad_norm": 0.7710665464401245,
+      "learning_rate": 9.837409299499149e-05,
+      "loss": 0.0997,
+      "step": 22920
+    },
+    {
+      "epoch": 1.500163559044815,
+      "grad_norm": 0.7890987396240234,
+      "learning_rate": 9.837176868233496e-05,
+      "loss": 0.1039,
+      "step": 22930
+    },
+    {
+      "epoch": 1.500817795224076,
+      "grad_norm": 0.7325422167778015,
+      "learning_rate": 9.836944273700439e-05,
+      "loss": 0.0917,
+      "step": 22940
+    },
+    {
+      "epoch": 1.5014720314033365,
+      "grad_norm": 0.8291808366775513,
+      "learning_rate": 9.836711515907827e-05,
+      "loss": 0.1017,
+      "step": 22950
+    },
+    {
+      "epoch": 1.5021262675825973,
+      "grad_norm": 0.7755422592163086,
+      "learning_rate": 9.836478594863516e-05,
+      "loss": 0.0887,
+      "step": 22960
+    },
+    {
+      "epoch": 1.502780503761858,
+      "grad_norm": 0.8843225836753845,
+      "learning_rate": 9.836245510575368e-05,
+      "loss": 0.1055,
+      "step": 22970
+    },
+    {
+      "epoch": 1.5034347399411188,
+      "grad_norm": 1.0079516172409058,
+      "learning_rate": 9.836012263051252e-05,
+      "loss": 0.1042,
+      "step": 22980
+    },
+    {
+      "epoch": 1.5040889761203795,
+      "grad_norm": 0.6518259048461914,
+      "learning_rate": 9.835778852299039e-05,
+      "loss": 0.1012,
+      "step": 22990
+    },
+    {
+      "epoch": 1.50474321229964,
+      "grad_norm": 0.7768208980560303,
+      "learning_rate": 9.835545278326606e-05,
+      "loss": 0.0852,
+      "step": 23000
+    },
+    {
+      "epoch": 1.505397448478901,
+      "grad_norm": 0.6538995504379272,
+      "learning_rate": 9.835311541141839e-05,
+      "loss": 0.0882,
+      "step": 23010
+    },
+    {
+      "epoch": 1.5060516846581615,
+      "grad_norm": 0.8296500444412231,
+      "learning_rate": 9.835077640752626e-05,
+      "loss": 0.0938,
+      "step": 23020
+    },
+    {
+      "epoch": 1.5067059208374223,
+      "grad_norm": 0.9206341505050659,
+      "learning_rate": 9.834843577166863e-05,
+      "loss": 0.0974,
+      "step": 23030
+    },
+    {
+      "epoch": 1.507360157016683,
+      "grad_norm": 0.779178261756897,
+      "learning_rate": 9.83460935039245e-05,
+      "loss": 0.1044,
+      "step": 23040
+    },
+    {
+      "epoch": 1.5080143931959438,
+      "grad_norm": 0.9146134853363037,
+      "learning_rate": 9.834374960437291e-05,
+      "loss": 0.0909,
+      "step": 23050
+    },
+    {
+      "epoch": 1.5086686293752045,
+      "grad_norm": 0.9325788617134094,
+      "learning_rate": 9.834140407309298e-05,
+      "loss": 0.0969,
+      "step": 23060
+    },
+    {
+      "epoch": 1.509322865554465,
+      "grad_norm": 0.9949310421943665,
+      "learning_rate": 9.833905691016389e-05,
+      "loss": 0.0969,
+      "step": 23070
+    },
+    {
+      "epoch": 1.509977101733726,
+      "grad_norm": 0.8266165256500244,
+      "learning_rate": 9.833670811566485e-05,
+      "loss": 0.0924,
+      "step": 23080
+    },
+    {
+      "epoch": 1.5106313379129865,
+      "grad_norm": 0.8005866408348083,
+      "learning_rate": 9.833435768967514e-05,
+      "loss": 0.0931,
+      "step": 23090
+    },
+    {
+      "epoch": 1.5112855740922473,
+      "grad_norm": 0.9937982559204102,
+      "learning_rate": 9.833200563227411e-05,
+      "loss": 0.0897,
+      "step": 23100
+    },
+    {
+      "epoch": 1.511939810271508,
+      "grad_norm": 0.8967451453208923,
+      "learning_rate": 9.832965194354113e-05,
+      "loss": 0.1015,
+      "step": 23110
+    },
+    {
+      "epoch": 1.5125940464507688,
+      "grad_norm": 0.7568728923797607,
+      "learning_rate": 9.832729662355566e-05,
+      "loss": 0.0945,
+      "step": 23120
+    },
+    {
+      "epoch": 1.5132482826300295,
+      "grad_norm": 0.8389810919761658,
+      "learning_rate": 9.832493967239716e-05,
+      "loss": 0.0958,
+      "step": 23130
+    },
+    {
+      "epoch": 1.51390251880929,
+      "grad_norm": 0.7844974398612976,
+      "learning_rate": 9.832258109014522e-05,
+      "loss": 0.0963,
+      "step": 23140
+    },
+    {
+      "epoch": 1.514556754988551,
+      "grad_norm": 0.7761784791946411,
+      "learning_rate": 9.832022087687944e-05,
+      "loss": 0.0913,
+      "step": 23150
+    },
+    {
+      "epoch": 1.5152109911678115,
+      "grad_norm": 1.0485121011734009,
+      "learning_rate": 9.831785903267949e-05,
+      "loss": 0.1062,
+      "step": 23160
+    },
+    {
+      "epoch": 1.5158652273470723,
+      "grad_norm": 0.8118041157722473,
+      "learning_rate": 9.831549555762507e-05,
+      "loss": 0.0926,
+      "step": 23170
+    },
+    {
+      "epoch": 1.516519463526333,
+      "grad_norm": 1.069348931312561,
+      "learning_rate": 9.831313045179595e-05,
+      "loss": 0.0977,
+      "step": 23180
+    },
+    {
+      "epoch": 1.5171736997055936,
+      "grad_norm": 0.8095777034759521,
+      "learning_rate": 9.8310763715272e-05,
+      "loss": 0.0975,
+      "step": 23190
+    },
+    {
+      "epoch": 1.5178279358848545,
+      "grad_norm": 0.6841381788253784,
+      "learning_rate": 9.830839534813305e-05,
+      "loss": 0.0924,
+      "step": 23200
+    },
+    {
+      "epoch": 1.518482172064115,
+      "grad_norm": 0.7451679706573486,
+      "learning_rate": 9.830602535045908e-05,
+      "loss": 0.0989,
+      "step": 23210
+    },
+    {
+      "epoch": 1.5191364082433758,
+      "grad_norm": 0.7688360214233398,
+      "learning_rate": 9.830365372233006e-05,
+      "loss": 0.1033,
+      "step": 23220
+    },
+    {
+      "epoch": 1.5197906444226366,
+      "grad_norm": 0.7664195895195007,
+      "learning_rate": 9.830128046382605e-05,
+      "loss": 0.0857,
+      "step": 23230
+    },
+    {
+      "epoch": 1.5204448806018973,
+      "grad_norm": 0.7574120163917542,
+      "learning_rate": 9.829890557502714e-05,
+      "loss": 0.0912,
+      "step": 23240
+    },
+    {
+      "epoch": 1.521099116781158,
+      "grad_norm": 0.8044725656509399,
+      "learning_rate": 9.82965290560135e-05,
+      "loss": 0.093,
+      "step": 23250
+    },
+    {
+      "epoch": 1.5217533529604186,
+      "grad_norm": 0.9325090050697327,
+      "learning_rate": 9.829415090686535e-05,
+      "loss": 0.0893,
+      "step": 23260
+    },
+    {
+      "epoch": 1.5224075891396796,
+      "grad_norm": 0.9916201829910278,
+      "learning_rate": 9.829177112766294e-05,
+      "loss": 0.0998,
+      "step": 23270
+    },
+    {
+      "epoch": 1.52306182531894,
+      "grad_norm": 0.8349946141242981,
+      "learning_rate": 9.828938971848663e-05,
+      "loss": 0.0991,
+      "step": 23280
+    },
+    {
+      "epoch": 1.5237160614982008,
+      "grad_norm": 0.8166770339012146,
+      "learning_rate": 9.828700667941675e-05,
+      "loss": 0.0983,
+      "step": 23290
+    },
+    {
+      "epoch": 1.5243702976774616,
+      "grad_norm": 0.8615389466285706,
+      "learning_rate": 9.828462201053376e-05,
+      "loss": 0.0957,
+      "step": 23300
+    },
+    {
+      "epoch": 1.5250245338567223,
+      "grad_norm": 0.8796236515045166,
+      "learning_rate": 9.828223571191814e-05,
+      "loss": 0.0974,
+      "step": 23310
+    },
+    {
+      "epoch": 1.525678770035983,
+      "grad_norm": 0.8417108654975891,
+      "learning_rate": 9.827984778365045e-05,
+      "loss": 0.1001,
+      "step": 23320
+    },
+    {
+      "epoch": 1.5263330062152436,
+      "grad_norm": 0.8357135653495789,
+      "learning_rate": 9.827745822581128e-05,
+      "loss": 0.1045,
+      "step": 23330
+    },
+    {
+      "epoch": 1.5269872423945046,
+      "grad_norm": 0.7510766386985779,
+      "learning_rate": 9.827506703848128e-05,
+      "loss": 0.1021,
+      "step": 23340
+    },
+    {
+      "epoch": 1.527641478573765,
+      "grad_norm": 1.0681124925613403,
+      "learning_rate": 9.827267422174115e-05,
+      "loss": 0.0953,
+      "step": 23350
+    },
+    {
+      "epoch": 1.5282957147530258,
+      "grad_norm": 0.8562158942222595,
+      "learning_rate": 9.82702797756717e-05,
+      "loss": 0.0992,
+      "step": 23360
+    },
+    {
+      "epoch": 1.5289499509322866,
+      "grad_norm": 0.9547275900840759,
+      "learning_rate": 9.826788370035368e-05,
+      "loss": 0.1004,
+      "step": 23370
+    },
+    {
+      "epoch": 1.529604187111547,
+      "grad_norm": 0.8317331075668335,
+      "learning_rate": 9.826548599586802e-05,
+      "loss": 0.1016,
+      "step": 23380
+    },
+    {
+      "epoch": 1.530258423290808,
+      "grad_norm": 0.9036648273468018,
+      "learning_rate": 9.82630866622956e-05,
+      "loss": 0.1068,
+      "step": 23390
+    },
+    {
+      "epoch": 1.5309126594700686,
+      "grad_norm": 0.9889810085296631,
+      "learning_rate": 9.826068569971745e-05,
+      "loss": 0.0998,
+      "step": 23400
+    },
+    {
+      "epoch": 1.5315668956493296,
+      "grad_norm": 0.8592228293418884,
+      "learning_rate": 9.825828310821459e-05,
+      "loss": 0.0989,
+      "step": 23410
+    },
+    {
+      "epoch": 1.53222113182859,
+      "grad_norm": 0.8532178401947021,
+      "learning_rate": 9.82558788878681e-05,
+      "loss": 0.097,
+      "step": 23420
+    },
+    {
+      "epoch": 1.5328753680078508,
+      "grad_norm": 0.9547865390777588,
+      "learning_rate": 9.825347303875916e-05,
+      "loss": 0.0851,
+      "step": 23430
+    },
+    {
+      "epoch": 1.5335296041871116,
+      "grad_norm": 0.833165168762207,
+      "learning_rate": 9.825106556096894e-05,
+      "loss": 0.1017,
+      "step": 23440
+    },
+    {
+      "epoch": 1.5341838403663721,
+      "grad_norm": 1.1163369417190552,
+      "learning_rate": 9.824865645457872e-05,
+      "loss": 0.0899,
+      "step": 23450
+    },
+    {
+      "epoch": 1.534838076545633,
+      "grad_norm": 0.8560509085655212,
+      "learning_rate": 9.824624571966981e-05,
+      "loss": 0.0965,
+      "step": 23460
+    },
+    {
+      "epoch": 1.5354923127248936,
+      "grad_norm": 0.8199770450592041,
+      "learning_rate": 9.824383335632357e-05,
+      "loss": 0.0981,
+      "step": 23470
+    },
+    {
+      "epoch": 1.5361465489041544,
+      "grad_norm": 0.8420577645301819,
+      "learning_rate": 9.824141936462144e-05,
+      "loss": 0.0892,
+      "step": 23480
+    },
+    {
+      "epoch": 1.536800785083415,
+      "grad_norm": 0.8969340324401855,
+      "learning_rate": 9.823900374464487e-05,
+      "loss": 0.1015,
+      "step": 23490
+    },
+    {
+      "epoch": 1.5374550212626759,
+      "grad_norm": 1.0403921604156494,
+      "learning_rate": 9.823658649647544e-05,
+      "loss": 0.0966,
+      "step": 23500
+    },
+    {
+      "epoch": 1.5381092574419366,
+      "grad_norm": 0.866157591342926,
+      "learning_rate": 9.82341676201947e-05,
+      "loss": 0.098,
+      "step": 23510
+    },
+    {
+      "epoch": 1.5387634936211971,
+      "grad_norm": 0.8696637153625488,
+      "learning_rate": 9.82317471158843e-05,
+      "loss": 0.0963,
+      "step": 23520
+    },
+    {
+      "epoch": 1.539417729800458,
+      "grad_norm": 0.7463034987449646,
+      "learning_rate": 9.822932498362593e-05,
+      "loss": 0.1031,
+      "step": 23530
+    },
+    {
+      "epoch": 1.5400719659797186,
+      "grad_norm": 0.8361960053443909,
+      "learning_rate": 9.822690122350138e-05,
+      "loss": 0.0929,
+      "step": 23540
+    },
+    {
+      "epoch": 1.5407262021589794,
+      "grad_norm": 0.8186551928520203,
+      "learning_rate": 9.822447583559242e-05,
+      "loss": 0.1008,
+      "step": 23550
+    },
+    {
+      "epoch": 1.5413804383382401,
+      "grad_norm": 0.8494019508361816,
+      "learning_rate": 9.822204881998093e-05,
+      "loss": 0.0876,
+      "step": 23560
+    },
+    {
+      "epoch": 1.5420346745175009,
+      "grad_norm": 0.9531555771827698,
+      "learning_rate": 9.821962017674881e-05,
+      "loss": 0.094,
+      "step": 23570
+    },
+    {
+      "epoch": 1.5426889106967616,
+      "grad_norm": 0.6968259215354919,
+      "learning_rate": 9.821718990597808e-05,
+      "loss": 0.0877,
+      "step": 23580
+    },
+    {
+      "epoch": 1.5433431468760221,
+      "grad_norm": 0.864067018032074,
+      "learning_rate": 9.821475800775072e-05,
+      "loss": 0.0921,
+      "step": 23590
+    },
+    {
+      "epoch": 1.5439973830552831,
+      "grad_norm": 0.6976091861724854,
+      "learning_rate": 9.821232448214883e-05,
+      "loss": 0.0921,
+      "step": 23600
+    },
+    {
+      "epoch": 1.5446516192345436,
+      "grad_norm": 0.83482426404953,
+      "learning_rate": 9.820988932925455e-05,
+      "loss": 0.1032,
+      "step": 23610
+    },
+    {
+      "epoch": 1.5453058554138044,
+      "grad_norm": 0.8034452199935913,
+      "learning_rate": 9.820745254915005e-05,
+      "loss": 0.0894,
+      "step": 23620
+    },
+    {
+      "epoch": 1.5459600915930651,
+      "grad_norm": 0.7795824408531189,
+      "learning_rate": 9.820501414191763e-05,
+      "loss": 0.0911,
+      "step": 23630
+    },
+    {
+      "epoch": 1.5466143277723257,
+      "grad_norm": 0.7786241769790649,
+      "learning_rate": 9.820257410763953e-05,
+      "loss": 0.0885,
+      "step": 23640
+    },
+    {
+      "epoch": 1.5472685639515866,
+      "grad_norm": 0.8625045418739319,
+      "learning_rate": 9.820013244639816e-05,
+      "loss": 0.0974,
+      "step": 23650
+    },
+    {
+      "epoch": 1.5479228001308472,
+      "grad_norm": 0.7645002603530884,
+      "learning_rate": 9.81976891582759e-05,
+      "loss": 0.097,
+      "step": 23660
+    },
+    {
+      "epoch": 1.548577036310108,
+      "grad_norm": 0.978847861289978,
+      "learning_rate": 9.819524424335524e-05,
+      "loss": 0.0958,
+      "step": 23670
+    },
+    {
+      "epoch": 1.5492312724893686,
+      "grad_norm": 0.9067314267158508,
+      "learning_rate": 9.819279770171867e-05,
+      "loss": 0.0973,
+      "step": 23680
+    },
+    {
+      "epoch": 1.5498855086686294,
+      "grad_norm": 0.8743839263916016,
+      "learning_rate": 9.819034953344881e-05,
+      "loss": 0.1002,
+      "step": 23690
+    },
+    {
+      "epoch": 1.5505397448478901,
+      "grad_norm": 0.7451719641685486,
+      "learning_rate": 9.818789973862825e-05,
+      "loss": 0.0889,
+      "step": 23700
+    },
+    {
+      "epoch": 1.5511939810271507,
+      "grad_norm": 0.833907961845398,
+      "learning_rate": 9.818544831733971e-05,
+      "loss": 0.092,
+      "step": 23710
+    },
+    {
+      "epoch": 1.5518482172064116,
+      "grad_norm": 0.8655661344528198,
+      "learning_rate": 9.81829952696659e-05,
+      "loss": 0.0924,
+      "step": 23720
+    },
+    {
+      "epoch": 1.5525024533856722,
+      "grad_norm": 0.8889705538749695,
+      "learning_rate": 9.818054059568966e-05,
+      "loss": 0.0926,
+      "step": 23730
+    },
+    {
+      "epoch": 1.553156689564933,
+      "grad_norm": 0.9665441513061523,
+      "learning_rate": 9.81780842954938e-05,
+      "loss": 0.105,
+      "step": 23740
+    },
+    {
+      "epoch": 1.5538109257441937,
+      "grad_norm": 0.8035663962364197,
+      "learning_rate": 9.817562636916125e-05,
+      "loss": 0.1007,
+      "step": 23750
+    },
+    {
+      "epoch": 1.5544651619234544,
+      "grad_norm": 0.8797777891159058,
+      "learning_rate": 9.817316681677496e-05,
+      "loss": 0.1106,
+      "step": 23760
+    },
+    {
+      "epoch": 1.5551193981027152,
+      "grad_norm": 0.799263060092926,
+      "learning_rate": 9.817070563841795e-05,
+      "loss": 0.0978,
+      "step": 23770
+    },
+    {
+      "epoch": 1.5557736342819757,
+      "grad_norm": 0.8599784970283508,
+      "learning_rate": 9.81682428341733e-05,
+      "loss": 0.1008,
+      "step": 23780
+    },
+    {
+      "epoch": 1.5564278704612367,
+      "grad_norm": 0.8856527805328369,
+      "learning_rate": 9.816577840412414e-05,
+      "loss": 0.0869,
+      "step": 23790
+    },
+    {
+      "epoch": 1.5570821066404972,
+      "grad_norm": 0.6909691691398621,
+      "learning_rate": 9.81633123483536e-05,
+      "loss": 0.0982,
+      "step": 23800
+    },
+    {
+      "epoch": 1.557736342819758,
+      "grad_norm": 0.8058643937110901,
+      "learning_rate": 9.8160844666945e-05,
+      "loss": 0.0944,
+      "step": 23810
+    },
+    {
+      "epoch": 1.5583905789990187,
+      "grad_norm": 1.1985560655593872,
+      "learning_rate": 9.815837535998156e-05,
+      "loss": 0.1016,
+      "step": 23820
+    },
+    {
+      "epoch": 1.5590448151782792,
+      "grad_norm": 0.8352839946746826,
+      "learning_rate": 9.815590442754666e-05,
+      "loss": 0.1072,
+      "step": 23830
+    },
+    {
+      "epoch": 1.5596990513575402,
+      "grad_norm": 0.9324560761451721,
+      "learning_rate": 9.815343186972369e-05,
+      "loss": 0.1034,
+      "step": 23840
+    },
+    {
+      "epoch": 1.5603532875368007,
+      "grad_norm": 0.8299842476844788,
+      "learning_rate": 9.81509576865961e-05,
+      "loss": 0.0935,
+      "step": 23850
+    },
+    {
+      "epoch": 1.5610075237160617,
+      "grad_norm": 0.9215813875198364,
+      "learning_rate": 9.814848187824742e-05,
+      "loss": 0.0986,
+      "step": 23860
+    },
+    {
+      "epoch": 1.5616617598953222,
+      "grad_norm": 0.8809934854507446,
+      "learning_rate": 9.81460044447612e-05,
+      "loss": 0.0994,
+      "step": 23870
+    },
+    {
+      "epoch": 1.562315996074583,
+      "grad_norm": 0.8337189555168152,
+      "learning_rate": 9.814352538622106e-05,
+      "loss": 0.09,
+      "step": 23880
+    },
+    {
+      "epoch": 1.5629702322538437,
+      "grad_norm": 0.8376109004020691,
+      "learning_rate": 9.814104470271068e-05,
+      "loss": 0.103,
+      "step": 23890
+    },
+    {
+      "epoch": 1.5636244684331042,
+      "grad_norm": 0.7608706951141357,
+      "learning_rate": 9.813856239431378e-05,
+      "loss": 0.1002,
+      "step": 23900
+    },
+    {
+      "epoch": 1.5642787046123652,
+      "grad_norm": 0.7375048995018005,
+      "learning_rate": 9.813607846111416e-05,
+      "loss": 0.0901,
+      "step": 23910
+    },
+    {
+      "epoch": 1.5649329407916257,
+      "grad_norm": 0.7545016407966614,
+      "learning_rate": 9.813359290319563e-05,
+      "loss": 0.1055,
+      "step": 23920
+    },
+    {
+      "epoch": 1.5655871769708865,
+      "grad_norm": 0.7844988703727722,
+      "learning_rate": 9.813110572064212e-05,
+      "loss": 0.0992,
+      "step": 23930
+    },
+    {
+      "epoch": 1.5662414131501472,
+      "grad_norm": 0.7850669622421265,
+      "learning_rate": 9.812861691353757e-05,
+      "loss": 0.0912,
+      "step": 23940
+    },
+    {
+      "epoch": 1.566895649329408,
+      "grad_norm": 1.1129896640777588,
+      "learning_rate": 9.812612648196598e-05,
+      "loss": 0.0932,
+      "step": 23950
+    },
+    {
+      "epoch": 1.5675498855086687,
+      "grad_norm": 0.9184969663619995,
+      "learning_rate": 9.81236344260114e-05,
+      "loss": 0.1044,
+      "step": 23960
+    },
+    {
+      "epoch": 1.5682041216879292,
+      "grad_norm": 0.790431559085846,
+      "learning_rate": 9.812114074575793e-05,
+      "loss": 0.1014,
+      "step": 23970
+    },
+    {
+      "epoch": 1.5688583578671902,
+      "grad_norm": 0.7852354645729065,
+      "learning_rate": 9.811864544128978e-05,
+      "loss": 0.0822,
+      "step": 23980
+    },
+    {
+      "epoch": 1.5695125940464507,
+      "grad_norm": 0.7458804249763489,
+      "learning_rate": 9.811614851269114e-05,
+      "loss": 0.0996,
+      "step": 23990
+    },
+    {
+      "epoch": 1.5701668302257115,
+      "grad_norm": 0.7905547022819519,
+      "learning_rate": 9.81136499600463e-05,
+      "loss": 0.0905,
+      "step": 24000
+    },
+    {
+      "epoch": 1.5708210664049722,
+      "grad_norm": 0.8950849771499634,
+      "learning_rate": 9.811114978343961e-05,
+      "loss": 0.0945,
+      "step": 24010
+    },
+    {
+      "epoch": 1.571475302584233,
+      "grad_norm": 0.9239217638969421,
+      "learning_rate": 9.810864798295541e-05,
+      "loss": 0.0896,
+      "step": 24020
+    },
+    {
+      "epoch": 1.5721295387634937,
+      "grad_norm": 0.8260421752929688,
+      "learning_rate": 9.810614455867818e-05,
+      "loss": 0.0826,
+      "step": 24030
+    },
+    {
+      "epoch": 1.5727837749427542,
+      "grad_norm": 0.8214983940124512,
+      "learning_rate": 9.810363951069241e-05,
+      "loss": 0.0918,
+      "step": 24040
+    },
+    {
+      "epoch": 1.5734380111220152,
+      "grad_norm": 0.7926087975502014,
+      "learning_rate": 9.810113283908266e-05,
+      "loss": 0.0858,
+      "step": 24050
+    },
+    {
+      "epoch": 1.5740922473012757,
+      "grad_norm": 0.8653861284255981,
+      "learning_rate": 9.809862454393352e-05,
+      "loss": 0.1005,
+      "step": 24060
+    },
+    {
+      "epoch": 1.5747464834805365,
+      "grad_norm": 0.846734881401062,
+      "learning_rate": 9.809611462532964e-05,
+      "loss": 0.0908,
+      "step": 24070
+    },
+    {
+      "epoch": 1.5754007196597972,
+      "grad_norm": 1.0754064321517944,
+      "learning_rate": 9.809360308335578e-05,
+      "loss": 0.1067,
+      "step": 24080
+    },
+    {
+      "epoch": 1.5760549558390577,
+      "grad_norm": 0.7325643301010132,
+      "learning_rate": 9.809108991809668e-05,
+      "loss": 0.0927,
+      "step": 24090
+    },
+    {
+      "epoch": 1.5767091920183187,
+      "grad_norm": 0.8486270904541016,
+      "learning_rate": 9.808857512963717e-05,
+      "loss": 0.1034,
+      "step": 24100
+    },
+    {
+      "epoch": 1.5773634281975792,
+      "grad_norm": 0.8908068537712097,
+      "learning_rate": 9.808605871806213e-05,
+      "loss": 0.1011,
+      "step": 24110
+    },
+    {
+      "epoch": 1.57801766437684,
+      "grad_norm": 0.9737763404846191,
+      "learning_rate": 9.808354068345649e-05,
+      "loss": 0.1027,
+      "step": 24120
+    },
+    {
+      "epoch": 1.5786719005561007,
+      "grad_norm": 0.9432488083839417,
+      "learning_rate": 9.808102102590526e-05,
+      "loss": 0.1,
+      "step": 24130
+    },
+    {
+      "epoch": 1.5793261367353615,
+      "grad_norm": 0.8621255159378052,
+      "learning_rate": 9.807849974549347e-05,
+      "loss": 0.0851,
+      "step": 24140
+    },
+    {
+      "epoch": 1.5799803729146222,
+      "grad_norm": 0.8411361575126648,
+      "learning_rate": 9.807597684230623e-05,
+      "loss": 0.106,
+      "step": 24150
+    },
+    {
+      "epoch": 1.5806346090938828,
+      "grad_norm": 0.8948150873184204,
+      "learning_rate": 9.807345231642868e-05,
+      "loss": 0.1006,
+      "step": 24160
+    },
+    {
+      "epoch": 1.5812888452731437,
+      "grad_norm": 1.042441725730896,
+      "learning_rate": 9.807092616794605e-05,
+      "loss": 0.1131,
+      "step": 24170
+    },
+    {
+      "epoch": 1.5819430814524043,
+      "grad_norm": 1.0746121406555176,
+      "learning_rate": 9.806839839694358e-05,
+      "loss": 0.0919,
+      "step": 24180
+    },
+    {
+      "epoch": 1.582597317631665,
+      "grad_norm": 0.8730924725532532,
+      "learning_rate": 9.806586900350658e-05,
+      "loss": 0.0961,
+      "step": 24190
+    },
+    {
+      "epoch": 1.5832515538109257,
+      "grad_norm": 0.7895834445953369,
+      "learning_rate": 9.806333798772047e-05,
+      "loss": 0.0892,
+      "step": 24200
+    },
+    {
+      "epoch": 1.5839057899901865,
+      "grad_norm": 0.786361038684845,
+      "learning_rate": 9.806080534967065e-05,
+      "loss": 0.0953,
+      "step": 24210
+    },
+    {
+      "epoch": 1.5845600261694472,
+      "grad_norm": 0.8377049565315247,
+      "learning_rate": 9.80582710894426e-05,
+      "loss": 0.1049,
+      "step": 24220
+    },
+    {
+      "epoch": 1.5852142623487078,
+      "grad_norm": 0.7862370014190674,
+      "learning_rate": 9.805573520712186e-05,
+      "loss": 0.1034,
+      "step": 24230
+    },
+    {
+      "epoch": 1.5858684985279687,
+      "grad_norm": 0.8017224073410034,
+      "learning_rate": 9.805319770279404e-05,
+      "loss": 0.0996,
+      "step": 24240
+    },
+    {
+      "epoch": 1.5865227347072293,
+      "grad_norm": 0.8779584765434265,
+      "learning_rate": 9.805065857654476e-05,
+      "loss": 0.1031,
+      "step": 24250
+    },
+    {
+      "epoch": 1.58717697088649,
+      "grad_norm": 0.8980815410614014,
+      "learning_rate": 9.804811782845974e-05,
+      "loss": 0.0956,
+      "step": 24260
+    },
+    {
+      "epoch": 1.5878312070657508,
+      "grad_norm": 1.0139700174331665,
+      "learning_rate": 9.804557545862474e-05,
+      "loss": 0.0996,
+      "step": 24270
+    },
+    {
+      "epoch": 1.5884854432450113,
+      "grad_norm": 0.8047105073928833,
+      "learning_rate": 9.804303146712555e-05,
+      "loss": 0.1042,
+      "step": 24280
+    },
+    {
+      "epoch": 1.5891396794242723,
+      "grad_norm": 0.7328765988349915,
+      "learning_rate": 9.804048585404806e-05,
+      "loss": 0.0914,
+      "step": 24290
+    },
+    {
+      "epoch": 1.5897939156035328,
+      "grad_norm": 0.866847813129425,
+      "learning_rate": 9.803793861947816e-05,
+      "loss": 0.0952,
+      "step": 24300
+    },
+    {
+      "epoch": 1.5904481517827938,
+      "grad_norm": 0.8152327537536621,
+      "learning_rate": 9.803538976350189e-05,
+      "loss": 0.0915,
+      "step": 24310
+    },
+    {
+      "epoch": 1.5911023879620543,
+      "grad_norm": 0.7884990572929382,
+      "learning_rate": 9.80328392862052e-05,
+      "loss": 0.0914,
+      "step": 24320
+    },
+    {
+      "epoch": 1.591756624141315,
+      "grad_norm": 1.0479592084884644,
+      "learning_rate": 9.803028718767423e-05,
+      "loss": 0.1004,
+      "step": 24330
+    },
+    {
+      "epoch": 1.5924108603205758,
+      "grad_norm": 0.8851988911628723,
+      "learning_rate": 9.80277334679951e-05,
+      "loss": 0.0979,
+      "step": 24340
+    },
+    {
+      "epoch": 1.5930650964998363,
+      "grad_norm": 0.8954311609268188,
+      "learning_rate": 9.8025178127254e-05,
+      "loss": 0.0914,
+      "step": 24350
+    },
+    {
+      "epoch": 1.5937193326790973,
+      "grad_norm": 0.8873519897460938,
+      "learning_rate": 9.80226211655372e-05,
+      "loss": 0.0978,
+      "step": 24360
+    },
+    {
+      "epoch": 1.5943735688583578,
+      "grad_norm": 0.8193051815032959,
+      "learning_rate": 9.8020062582931e-05,
+      "loss": 0.1028,
+      "step": 24370
+    },
+    {
+      "epoch": 1.5950278050376185,
+      "grad_norm": 0.8333913683891296,
+      "learning_rate": 9.801750237952172e-05,
+      "loss": 0.0951,
+      "step": 24380
+    },
+    {
+      "epoch": 1.5956820412168793,
+      "grad_norm": 0.8090634942054749,
+      "learning_rate": 9.801494055539584e-05,
+      "loss": 0.1018,
+      "step": 24390
+    },
+    {
+      "epoch": 1.59633627739614,
+      "grad_norm": 0.8224121332168579,
+      "learning_rate": 9.801237711063978e-05,
+      "loss": 0.099,
+      "step": 24400
+    },
+    {
+      "epoch": 1.5969905135754008,
+      "grad_norm": 1.034574031829834,
+      "learning_rate": 9.800981204534006e-05,
+      "loss": 0.1089,
+      "step": 24410
+    },
+    {
+      "epoch": 1.5976447497546613,
+      "grad_norm": 0.8645883202552795,
+      "learning_rate": 9.800724535958328e-05,
+      "loss": 0.094,
+      "step": 24420
+    },
+    {
+      "epoch": 1.5982989859339223,
+      "grad_norm": 0.8233097195625305,
+      "learning_rate": 9.800467705345607e-05,
+      "loss": 0.097,
+      "step": 24430
+    },
+    {
+      "epoch": 1.5989532221131828,
+      "grad_norm": 0.9189668893814087,
+      "learning_rate": 9.800210712704512e-05,
+      "loss": 0.1041,
+      "step": 24440
+    },
+    {
+      "epoch": 1.5996074582924436,
+      "grad_norm": 0.8542932271957397,
+      "learning_rate": 9.799953558043715e-05,
+      "loss": 0.0997,
+      "step": 24450
+    },
+    {
+      "epoch": 1.6002616944717043,
+      "grad_norm": 0.7710226774215698,
+      "learning_rate": 9.799696241371898e-05,
+      "loss": 0.1006,
+      "step": 24460
+    },
+    {
+      "epoch": 1.600915930650965,
+      "grad_norm": 0.937626302242279,
+      "learning_rate": 9.799438762697744e-05,
+      "loss": 0.1013,
+      "step": 24470
+    },
+    {
+      "epoch": 1.6015701668302258,
+      "grad_norm": 0.8287016153335571,
+      "learning_rate": 9.799181122029946e-05,
+      "loss": 0.0906,
+      "step": 24480
+    },
+    {
+      "epoch": 1.6022244030094863,
+      "grad_norm": 0.8471201062202454,
+      "learning_rate": 9.798923319377199e-05,
+      "loss": 0.0886,
+      "step": 24490
+    },
+    {
+      "epoch": 1.6028786391887473,
+      "grad_norm": 0.8005667924880981,
+      "learning_rate": 9.798665354748205e-05,
+      "loss": 0.0949,
+      "step": 24500
+    },
+    {
+      "epoch": 1.6035328753680078,
+      "grad_norm": 0.7234973311424255,
+      "learning_rate": 9.798407228151667e-05,
+      "loss": 0.0955,
+      "step": 24510
+    },
+    {
+      "epoch": 1.6041871115472686,
+      "grad_norm": 0.8595584630966187,
+      "learning_rate": 9.798148939596303e-05,
+      "loss": 0.0885,
+      "step": 24520
+    },
+    {
+      "epoch": 1.6048413477265293,
+      "grad_norm": 0.8628395199775696,
+      "learning_rate": 9.797890489090829e-05,
+      "loss": 0.0917,
+      "step": 24530
+    },
+    {
+      "epoch": 1.6054955839057898,
+      "grad_norm": 0.698773980140686,
+      "learning_rate": 9.797631876643967e-05,
+      "loss": 0.0888,
+      "step": 24540
+    },
+    {
+      "epoch": 1.6061498200850508,
+      "grad_norm": 0.7893082499504089,
+      "learning_rate": 9.797373102264448e-05,
+      "loss": 0.108,
+      "step": 24550
+    },
+    {
+      "epoch": 1.6068040562643113,
+      "grad_norm": 0.9154890775680542,
+      "learning_rate": 9.797114165961006e-05,
+      "loss": 0.1043,
+      "step": 24560
+    },
+    {
+      "epoch": 1.607458292443572,
+      "grad_norm": 0.9815983176231384,
+      "learning_rate": 9.796855067742378e-05,
+      "loss": 0.1072,
+      "step": 24570
+    },
+    {
+      "epoch": 1.6081125286228328,
+      "grad_norm": 0.8936287760734558,
+      "learning_rate": 9.796595807617313e-05,
+      "loss": 0.1025,
+      "step": 24580
+    },
+    {
+      "epoch": 1.6087667648020936,
+      "grad_norm": 0.953900933265686,
+      "learning_rate": 9.796336385594557e-05,
+      "loss": 0.0932,
+      "step": 24590
+    },
+    {
+      "epoch": 1.6094210009813543,
+      "grad_norm": 0.8990422487258911,
+      "learning_rate": 9.796076801682871e-05,
+      "loss": 0.0958,
+      "step": 24600
+    },
+    {
+      "epoch": 1.6100752371606148,
+      "grad_norm": 0.7732602953910828,
+      "learning_rate": 9.795817055891016e-05,
+      "loss": 0.0885,
+      "step": 24610
+    },
+    {
+      "epoch": 1.6107294733398758,
+      "grad_norm": 0.913487434387207,
+      "learning_rate": 9.795557148227756e-05,
+      "loss": 0.1008,
+      "step": 24620
+    },
+    {
+      "epoch": 1.6113837095191363,
+      "grad_norm": 0.9833926558494568,
+      "learning_rate": 9.795297078701867e-05,
+      "loss": 0.0969,
+      "step": 24630
+    },
+    {
+      "epoch": 1.612037945698397,
+      "grad_norm": 0.9888587594032288,
+      "learning_rate": 9.795036847322124e-05,
+      "loss": 0.1004,
+      "step": 24640
+    },
+    {
+      "epoch": 1.6126921818776578,
+      "grad_norm": 0.7544565796852112,
+      "learning_rate": 9.794776454097314e-05,
+      "loss": 0.0883,
+      "step": 24650
+    },
+    {
+      "epoch": 1.6133464180569186,
+      "grad_norm": 0.8678017258644104,
+      "learning_rate": 9.794515899036222e-05,
+      "loss": 0.0974,
+      "step": 24660
+    },
+    {
+      "epoch": 1.6140006542361793,
+      "grad_norm": 1.1111491918563843,
+      "learning_rate": 9.794255182147644e-05,
+      "loss": 0.0926,
+      "step": 24670
+    },
+    {
+      "epoch": 1.6146548904154399,
+      "grad_norm": 0.8813068270683289,
+      "learning_rate": 9.793994303440382e-05,
+      "loss": 0.0952,
+      "step": 24680
+    },
+    {
+      "epoch": 1.6153091265947008,
+      "grad_norm": 0.8787200450897217,
+      "learning_rate": 9.793733262923238e-05,
+      "loss": 0.1085,
+      "step": 24690
+    },
+    {
+      "epoch": 1.6159633627739614,
+      "grad_norm": 0.7039420008659363,
+      "learning_rate": 9.793472060605024e-05,
+      "loss": 0.099,
+      "step": 24700
+    },
+    {
+      "epoch": 1.616617598953222,
+      "grad_norm": 0.7518815994262695,
+      "learning_rate": 9.793210696494559e-05,
+      "loss": 0.0981,
+      "step": 24710
+    },
+    {
+      "epoch": 1.6172718351324828,
+      "grad_norm": 0.8269907832145691,
+      "learning_rate": 9.79294917060066e-05,
+      "loss": 0.089,
+      "step": 24720
+    },
+    {
+      "epoch": 1.6179260713117434,
+      "grad_norm": 0.8977356553077698,
+      "learning_rate": 9.792687482932158e-05,
+      "loss": 0.0868,
+      "step": 24730
+    },
+    {
+      "epoch": 1.6185803074910043,
+      "grad_norm": 0.7739750742912292,
+      "learning_rate": 9.792425633497883e-05,
+      "loss": 0.0873,
+      "step": 24740
+    },
+    {
+      "epoch": 1.6192345436702649,
+      "grad_norm": 0.9570137858390808,
+      "learning_rate": 9.792163622306676e-05,
+      "loss": 0.091,
+      "step": 24750
+    },
+    {
+      "epoch": 1.6198887798495258,
+      "grad_norm": 0.817414402961731,
+      "learning_rate": 9.791901449367378e-05,
+      "loss": 0.0953,
+      "step": 24760
+    },
+    {
+      "epoch": 1.6205430160287864,
+      "grad_norm": 0.7275998592376709,
+      "learning_rate": 9.791639114688837e-05,
+      "loss": 0.1121,
+      "step": 24770
+    },
+    {
+      "epoch": 1.6211972522080471,
+      "grad_norm": 0.828654944896698,
+      "learning_rate": 9.791376618279913e-05,
+      "loss": 0.0982,
+      "step": 24780
+    },
+    {
+      "epoch": 1.6218514883873079,
+      "grad_norm": 1.096555471420288,
+      "learning_rate": 9.791113960149458e-05,
+      "loss": 0.0948,
+      "step": 24790
+    },
+    {
+      "epoch": 1.6225057245665684,
+      "grad_norm": 0.7662184834480286,
+      "learning_rate": 9.790851140306345e-05,
+      "loss": 0.0966,
+      "step": 24800
+    },
+    {
+      "epoch": 1.6231599607458294,
+      "grad_norm": 0.7542035579681396,
+      "learning_rate": 9.790588158759441e-05,
+      "loss": 0.1002,
+      "step": 24810
+    },
+    {
+      "epoch": 1.6238141969250899,
+      "grad_norm": 0.8306244015693665,
+      "learning_rate": 9.790325015517622e-05,
+      "loss": 0.0923,
+      "step": 24820
+    },
+    {
+      "epoch": 1.6244684331043506,
+      "grad_norm": 0.8431245684623718,
+      "learning_rate": 9.790061710589771e-05,
+      "loss": 0.092,
+      "step": 24830
+    },
+    {
+      "epoch": 1.6251226692836114,
+      "grad_norm": 1.4137808084487915,
+      "learning_rate": 9.789798243984775e-05,
+      "loss": 0.0905,
+      "step": 24840
+    },
+    {
+      "epoch": 1.6257769054628721,
+      "grad_norm": 0.9978521466255188,
+      "learning_rate": 9.789534615711527e-05,
+      "loss": 0.0955,
+      "step": 24850
+    },
+    {
+      "epoch": 1.6264311416421329,
+      "grad_norm": 0.766831636428833,
+      "learning_rate": 9.789270825778923e-05,
+      "loss": 0.0896,
+      "step": 24860
+    },
+    {
+      "epoch": 1.6270853778213934,
+      "grad_norm": 0.8945890665054321,
+      "learning_rate": 9.78900687419587e-05,
+      "loss": 0.0953,
+      "step": 24870
+    },
+    {
+      "epoch": 1.6277396140006544,
+      "grad_norm": 0.9509278535842896,
+      "learning_rate": 9.788742760971274e-05,
+      "loss": 0.0944,
+      "step": 24880
+    },
+    {
+      "epoch": 1.628393850179915,
+      "grad_norm": 0.752797544002533,
+      "learning_rate": 9.788478486114052e-05,
+      "loss": 0.0847,
+      "step": 24890
+    },
+    {
+      "epoch": 1.6290480863591756,
+      "grad_norm": 0.8730521202087402,
+      "learning_rate": 9.788214049633123e-05,
+      "loss": 0.0901,
+      "step": 24900
+    },
+    {
+      "epoch": 1.6297023225384364,
+      "grad_norm": 0.9180911183357239,
+      "learning_rate": 9.78794945153741e-05,
+      "loss": 0.0958,
+      "step": 24910
+    },
+    {
+      "epoch": 1.6303565587176971,
+      "grad_norm": 0.9941650629043579,
+      "learning_rate": 9.787684691835849e-05,
+      "loss": 0.0961,
+      "step": 24920
+    },
+    {
+      "epoch": 1.6310107948969579,
+      "grad_norm": 0.9088178873062134,
+      "learning_rate": 9.787419770537371e-05,
+      "loss": 0.1026,
+      "step": 24930
+    },
+    {
+      "epoch": 1.6316650310762184,
+      "grad_norm": 0.8117355108261108,
+      "learning_rate": 9.787154687650923e-05,
+      "loss": 0.0913,
+      "step": 24940
+    },
+    {
+      "epoch": 1.6323192672554794,
+      "grad_norm": 0.7052477598190308,
+      "learning_rate": 9.786889443185449e-05,
+      "loss": 0.0911,
+      "step": 24950
+    },
+    {
+      "epoch": 1.63297350343474,
+      "grad_norm": 0.8779917359352112,
+      "learning_rate": 9.7866240371499e-05,
+      "loss": 0.0997,
+      "step": 24960
+    },
+    {
+      "epoch": 1.6336277396140007,
+      "grad_norm": 0.7817448377609253,
+      "learning_rate": 9.786358469553238e-05,
+      "loss": 0.0967,
+      "step": 24970
+    },
+    {
+      "epoch": 1.6342819757932614,
+      "grad_norm": 0.8375545144081116,
+      "learning_rate": 9.786092740404424e-05,
+      "loss": 0.1043,
+      "step": 24980
+    },
+    {
+      "epoch": 1.634936211972522,
+      "grad_norm": 1.144739031791687,
+      "learning_rate": 9.78582684971243e-05,
+      "loss": 0.1022,
+      "step": 24990
+    },
+    {
+      "epoch": 1.635590448151783,
+      "grad_norm": 1.2085695266723633,
+      "learning_rate": 9.785560797486227e-05,
+      "loss": 0.101,
+      "step": 25000
+    },
+    {
+      "epoch": 1.6362446843310434,
+      "grad_norm": 0.6930639147758484,
+      "learning_rate": 9.785294583734796e-05,
+      "loss": 0.0947,
+      "step": 25010
+    },
+    {
+      "epoch": 1.6368989205103042,
+      "grad_norm": 0.921447217464447,
+      "learning_rate": 9.785028208467123e-05,
+      "loss": 0.0978,
+      "step": 25020
+    },
+    {
+      "epoch": 1.637553156689565,
+      "grad_norm": 1.0130419731140137,
+      "learning_rate": 9.784761671692202e-05,
+      "loss": 0.0903,
+      "step": 25030
+    },
+    {
+      "epoch": 1.6382073928688257,
+      "grad_norm": 0.7007921934127808,
+      "learning_rate": 9.784494973419022e-05,
+      "loss": 0.0962,
+      "step": 25040
+    },
+    {
+      "epoch": 1.6388616290480864,
+      "grad_norm": 0.7728266716003418,
+      "learning_rate": 9.784228113656591e-05,
+      "loss": 0.0994,
+      "step": 25050
+    },
+    {
+      "epoch": 1.639515865227347,
+      "grad_norm": 0.8407819867134094,
+      "learning_rate": 9.783961092413914e-05,
+      "loss": 0.089,
+      "step": 25060
+    },
+    {
+      "epoch": 1.640170101406608,
+      "grad_norm": 0.9866244792938232,
+      "learning_rate": 9.783693909700002e-05,
+      "loss": 0.0996,
+      "step": 25070
+    },
+    {
+      "epoch": 1.6408243375858684,
+      "grad_norm": 0.7183135151863098,
+      "learning_rate": 9.783426565523877e-05,
+      "loss": 0.0873,
+      "step": 25080
+    },
+    {
+      "epoch": 1.6414785737651292,
+      "grad_norm": 0.886570394039154,
+      "learning_rate": 9.78315905989456e-05,
+      "loss": 0.1016,
+      "step": 25090
+    },
+    {
+      "epoch": 1.64213280994439,
+      "grad_norm": 0.887607753276825,
+      "learning_rate": 9.782891392821078e-05,
+      "loss": 0.0904,
+      "step": 25100
+    },
+    {
+      "epoch": 1.6427870461236507,
+      "grad_norm": 0.9125217795372009,
+      "learning_rate": 9.78262356431247e-05,
+      "loss": 0.0913,
+      "step": 25110
+    },
+    {
+      "epoch": 1.6434412823029114,
+      "grad_norm": 0.9861580729484558,
+      "learning_rate": 9.782355574377775e-05,
+      "loss": 0.0996,
+      "step": 25120
+    },
+    {
+      "epoch": 1.644095518482172,
+      "grad_norm": 1.0295411348342896,
+      "learning_rate": 9.782087423026036e-05,
+      "loss": 0.0995,
+      "step": 25130
+    },
+    {
+      "epoch": 1.644749754661433,
+      "grad_norm": 0.7599952816963196,
+      "learning_rate": 9.781819110266304e-05,
+      "loss": 0.0966,
+      "step": 25140
+    },
+    {
+      "epoch": 1.6454039908406934,
+      "grad_norm": 0.7582678198814392,
+      "learning_rate": 9.781550636107637e-05,
+      "loss": 0.0931,
+      "step": 25150
+    },
+    {
+      "epoch": 1.6460582270199542,
+      "grad_norm": 0.9556160569190979,
+      "learning_rate": 9.781282000559095e-05,
+      "loss": 0.1033,
+      "step": 25160
+    },
+    {
+      "epoch": 1.646712463199215,
+      "grad_norm": 0.8655904531478882,
+      "learning_rate": 9.781013203629748e-05,
+      "loss": 0.1,
+      "step": 25170
+    },
+    {
+      "epoch": 1.6473666993784757,
+      "grad_norm": 0.7407214045524597,
+      "learning_rate": 9.780744245328666e-05,
+      "loss": 0.0947,
+      "step": 25180
+    },
+    {
+      "epoch": 1.6480209355577364,
+      "grad_norm": 0.6862826943397522,
+      "learning_rate": 9.780475125664927e-05,
+      "loss": 0.0812,
+      "step": 25190
+    },
+    {
+      "epoch": 1.648675171736997,
+      "grad_norm": 0.862034022808075,
+      "learning_rate": 9.780205844647616e-05,
+      "loss": 0.0928,
+      "step": 25200
+    },
+    {
+      "epoch": 1.649329407916258,
+      "grad_norm": 0.8481500744819641,
+      "learning_rate": 9.77993640228582e-05,
+      "loss": 0.0957,
+      "step": 25210
+    },
+    {
+      "epoch": 1.6499836440955185,
+      "grad_norm": 1.0111854076385498,
+      "learning_rate": 9.779666798588637e-05,
+      "loss": 0.0879,
+      "step": 25220
+    },
+    {
+      "epoch": 1.6506378802747792,
+      "grad_norm": 0.6928382515907288,
+      "learning_rate": 9.779397033565164e-05,
+      "loss": 0.0917,
+      "step": 25230
+    },
+    {
+      "epoch": 1.65129211645404,
+      "grad_norm": 0.8151880502700806,
+      "learning_rate": 9.779127107224505e-05,
+      "loss": 0.1028,
+      "step": 25240
+    },
+    {
+      "epoch": 1.6519463526333005,
+      "grad_norm": 0.8269695043563843,
+      "learning_rate": 9.778857019575774e-05,
+      "loss": 0.1064,
+      "step": 25250
+    },
+    {
+      "epoch": 1.6526005888125614,
+      "grad_norm": 0.8382137417793274,
+      "learning_rate": 9.778586770628084e-05,
+      "loss": 0.0939,
+      "step": 25260
+    },
+    {
+      "epoch": 1.653254824991822,
+      "grad_norm": 0.8290400505065918,
+      "learning_rate": 9.778316360390558e-05,
+      "loss": 0.0899,
+      "step": 25270
+    },
+    {
+      "epoch": 1.6539090611710827,
+      "grad_norm": 0.8005163073539734,
+      "learning_rate": 9.778045788872324e-05,
+      "loss": 0.0892,
+      "step": 25280
+    },
+    {
+      "epoch": 1.6545632973503435,
+      "grad_norm": 0.9442570209503174,
+      "learning_rate": 9.777775056082514e-05,
+      "loss": 0.1007,
+      "step": 25290
+    },
+    {
+      "epoch": 1.6552175335296042,
+      "grad_norm": 0.7411683797836304,
+      "learning_rate": 9.777504162030267e-05,
+      "loss": 0.1008,
+      "step": 25300
+    },
+    {
+      "epoch": 1.655871769708865,
+      "grad_norm": 0.8269376158714294,
+      "learning_rate": 9.777233106724722e-05,
+      "loss": 0.0943,
+      "step": 25310
+    },
+    {
+      "epoch": 1.6565260058881255,
+      "grad_norm": 0.6999377608299255,
+      "learning_rate": 9.776961890175034e-05,
+      "loss": 0.106,
+      "step": 25320
+    },
+    {
+      "epoch": 1.6571802420673865,
+      "grad_norm": 0.8918529152870178,
+      "learning_rate": 9.776690512390352e-05,
+      "loss": 0.0957,
+      "step": 25330
+    },
+    {
+      "epoch": 1.657834478246647,
+      "grad_norm": 0.7899738550186157,
+      "learning_rate": 9.776418973379838e-05,
+      "loss": 0.0943,
+      "step": 25340
+    },
+    {
+      "epoch": 1.6584887144259077,
+      "grad_norm": 0.7438578009605408,
+      "learning_rate": 9.776147273152659e-05,
+      "loss": 0.0977,
+      "step": 25350
+    },
+    {
+      "epoch": 1.6591429506051685,
+      "grad_norm": 0.7523559331893921,
+      "learning_rate": 9.77587541171798e-05,
+      "loss": 0.0979,
+      "step": 25360
+    },
+    {
+      "epoch": 1.6597971867844292,
+      "grad_norm": 0.8759346008300781,
+      "learning_rate": 9.775603389084985e-05,
+      "loss": 0.1003,
+      "step": 25370
+    },
+    {
+      "epoch": 1.66045142296369,
+      "grad_norm": 0.7944019436836243,
+      "learning_rate": 9.775331205262847e-05,
+      "loss": 0.0903,
+      "step": 25380
+    },
+    {
+      "epoch": 1.6611056591429505,
+      "grad_norm": 0.809095561504364,
+      "learning_rate": 9.775058860260759e-05,
+      "loss": 0.1002,
+      "step": 25390
+    },
+    {
+      "epoch": 1.6617598953222115,
+      "grad_norm": 0.8237670660018921,
+      "learning_rate": 9.774786354087913e-05,
+      "loss": 0.0944,
+      "step": 25400
+    },
+    {
+      "epoch": 1.662414131501472,
+      "grad_norm": 0.8633749485015869,
+      "learning_rate": 9.774513686753504e-05,
+      "loss": 0.1029,
+      "step": 25410
+    },
+    {
+      "epoch": 1.6630683676807327,
+      "grad_norm": 0.7542330622673035,
+      "learning_rate": 9.774240858266735e-05,
+      "loss": 0.0886,
+      "step": 25420
+    },
+    {
+      "epoch": 1.6637226038599935,
+      "grad_norm": 0.8325693607330322,
+      "learning_rate": 9.773967868636818e-05,
+      "loss": 0.0929,
+      "step": 25430
+    },
+    {
+      "epoch": 1.664376840039254,
+      "grad_norm": 0.9114111065864563,
+      "learning_rate": 9.773694717872963e-05,
+      "loss": 0.0922,
+      "step": 25440
+    },
+    {
+      "epoch": 1.665031076218515,
+      "grad_norm": 0.732631266117096,
+      "learning_rate": 9.773421405984394e-05,
+      "loss": 0.0999,
+      "step": 25450
+    },
+    {
+      "epoch": 1.6656853123977755,
+      "grad_norm": 0.7207270860671997,
+      "learning_rate": 9.773147932980334e-05,
+      "loss": 0.0991,
+      "step": 25460
+    },
+    {
+      "epoch": 1.6663395485770363,
+      "grad_norm": 0.7336596250534058,
+      "learning_rate": 9.772874298870012e-05,
+      "loss": 0.096,
+      "step": 25470
+    },
+    {
+      "epoch": 1.666993784756297,
+      "grad_norm": 0.9117507338523865,
+      "learning_rate": 9.772600503662665e-05,
+      "loss": 0.0961,
+      "step": 25480
+    },
+    {
+      "epoch": 1.6676480209355578,
+      "grad_norm": 0.7073248028755188,
+      "learning_rate": 9.772326547367534e-05,
+      "loss": 0.0996,
+      "step": 25490
+    },
+    {
+      "epoch": 1.6683022571148185,
+      "grad_norm": 1.0052919387817383,
+      "learning_rate": 9.772052429993868e-05,
+      "loss": 0.1011,
+      "step": 25500
+    },
+    {
+      "epoch": 1.668956493294079,
+      "grad_norm": 0.8882834911346436,
+      "learning_rate": 9.771778151550917e-05,
+      "loss": 0.0876,
+      "step": 25510
+    },
+    {
+      "epoch": 1.66961072947334,
+      "grad_norm": 0.8214176893234253,
+      "learning_rate": 9.771503712047937e-05,
+      "loss": 0.0875,
+      "step": 25520
+    },
+    {
+      "epoch": 1.6702649656526005,
+      "grad_norm": 0.9225919842720032,
+      "learning_rate": 9.771229111494194e-05,
+      "loss": 0.0981,
+      "step": 25530
+    },
+    {
+      "epoch": 1.6709192018318613,
+      "grad_norm": 0.8596377968788147,
+      "learning_rate": 9.770954349898956e-05,
+      "loss": 0.0932,
+      "step": 25540
+    },
+    {
+      "epoch": 1.671573438011122,
+      "grad_norm": 0.7050160765647888,
+      "learning_rate": 9.770679427271496e-05,
+      "loss": 0.1051,
+      "step": 25550
+    },
+    {
+      "epoch": 1.6722276741903828,
+      "grad_norm": 1.0160956382751465,
+      "learning_rate": 9.770404343621094e-05,
+      "loss": 0.1,
+      "step": 25560
+    },
+    {
+      "epoch": 1.6728819103696435,
+      "grad_norm": 0.7766900062561035,
+      "learning_rate": 9.770129098957035e-05,
+      "loss": 0.0909,
+      "step": 25570
+    },
+    {
+      "epoch": 1.673536146548904,
+      "grad_norm": 0.9387828707695007,
+      "learning_rate": 9.769853693288608e-05,
+      "loss": 0.1045,
+      "step": 25580
+    },
+    {
+      "epoch": 1.674190382728165,
+      "grad_norm": 1.0153820514678955,
+      "learning_rate": 9.76957812662511e-05,
+      "loss": 0.0992,
+      "step": 25590
+    },
+    {
+      "epoch": 1.6748446189074255,
+      "grad_norm": 0.7925774455070496,
+      "learning_rate": 9.769302398975841e-05,
+      "loss": 0.0978,
+      "step": 25600
+    },
+    {
+      "epoch": 1.6754988550866863,
+      "grad_norm": 1.1735292673110962,
+      "learning_rate": 9.769026510350108e-05,
+      "loss": 0.1079,
+      "step": 25610
+    },
+    {
+      "epoch": 1.676153091265947,
+      "grad_norm": 0.7813097834587097,
+      "learning_rate": 9.768750460757223e-05,
+      "loss": 0.0903,
+      "step": 25620
+    },
+    {
+      "epoch": 1.6768073274452078,
+      "grad_norm": 0.8915982842445374,
+      "learning_rate": 9.768474250206504e-05,
+      "loss": 0.0922,
+      "step": 25630
+    },
+    {
+      "epoch": 1.6774615636244685,
+      "grad_norm": 0.8092440366744995,
+      "learning_rate": 9.768197878707273e-05,
+      "loss": 0.0983,
+      "step": 25640
+    },
+    {
+      "epoch": 1.678115799803729,
+      "grad_norm": 0.9614975452423096,
+      "learning_rate": 9.767921346268858e-05,
+      "loss": 0.103,
+      "step": 25650
+    },
+    {
+      "epoch": 1.67877003598299,
+      "grad_norm": 0.6998571753501892,
+      "learning_rate": 9.767644652900594e-05,
+      "loss": 0.0918,
+      "step": 25660
+    },
+    {
+      "epoch": 1.6794242721622505,
+      "grad_norm": 0.7789784073829651,
+      "learning_rate": 9.76736779861182e-05,
+      "loss": 0.0903,
+      "step": 25670
+    },
+    {
+      "epoch": 1.6800785083415113,
+      "grad_norm": 0.7573550939559937,
+      "learning_rate": 9.767090783411878e-05,
+      "loss": 0.0869,
+      "step": 25680
+    },
+    {
+      "epoch": 1.680732744520772,
+      "grad_norm": 0.8103246688842773,
+      "learning_rate": 9.766813607310122e-05,
+      "loss": 0.0959,
+      "step": 25690
+    },
+    {
+      "epoch": 1.6813869807000326,
+      "grad_norm": 0.7668115496635437,
+      "learning_rate": 9.766536270315903e-05,
+      "loss": 0.1045,
+      "step": 25700
+    },
+    {
+      "epoch": 1.6820412168792935,
+      "grad_norm": 0.7344299554824829,
+      "learning_rate": 9.766258772438586e-05,
+      "loss": 0.098,
+      "step": 25710
+    },
+    {
+      "epoch": 1.682695453058554,
+      "grad_norm": 0.9814144968986511,
+      "learning_rate": 9.765981113687534e-05,
+      "loss": 0.1032,
+      "step": 25720
+    },
+    {
+      "epoch": 1.6833496892378148,
+      "grad_norm": 0.9255467653274536,
+      "learning_rate": 9.765703294072121e-05,
+      "loss": 0.0991,
+      "step": 25730
+    },
+    {
+      "epoch": 1.6840039254170756,
+      "grad_norm": 0.8468577861785889,
+      "learning_rate": 9.765425313601724e-05,
+      "loss": 0.0978,
+      "step": 25740
+    },
+    {
+      "epoch": 1.6846581615963363,
+      "grad_norm": 0.8852717876434326,
+      "learning_rate": 9.765147172285725e-05,
+      "loss": 0.1103,
+      "step": 25750
+    },
+    {
+      "epoch": 1.685312397775597,
+      "grad_norm": 0.6971980333328247,
+      "learning_rate": 9.764868870133511e-05,
+      "loss": 0.0939,
+      "step": 25760
+    },
+    {
+      "epoch": 1.6859666339548576,
+      "grad_norm": 0.9403474926948547,
+      "learning_rate": 9.764590407154476e-05,
+      "loss": 0.1042,
+      "step": 25770
+    },
+    {
+      "epoch": 1.6866208701341185,
+      "grad_norm": 0.8653038740158081,
+      "learning_rate": 9.76431178335802e-05,
+      "loss": 0.097,
+      "step": 25780
+    },
+    {
+      "epoch": 1.687275106313379,
+      "grad_norm": 0.8561285138130188,
+      "learning_rate": 9.764032998753547e-05,
+      "loss": 0.0911,
+      "step": 25790
+    },
+    {
+      "epoch": 1.6879293424926398,
+      "grad_norm": 0.7791575789451599,
+      "learning_rate": 9.763754053350465e-05,
+      "loss": 0.0927,
+      "step": 25800
+    },
+    {
+      "epoch": 1.6885835786719006,
+      "grad_norm": 0.7109609842300415,
+      "learning_rate": 9.76347494715819e-05,
+      "loss": 0.0879,
+      "step": 25810
+    },
+    {
+      "epoch": 1.6892378148511613,
+      "grad_norm": 0.7509888410568237,
+      "learning_rate": 9.763195680186143e-05,
+      "loss": 0.0846,
+      "step": 25820
+    },
+    {
+      "epoch": 1.689892051030422,
+      "grad_norm": 0.727888286113739,
+      "learning_rate": 9.762916252443751e-05,
+      "loss": 0.0913,
+      "step": 25830
+    },
+    {
+      "epoch": 1.6905462872096826,
+      "grad_norm": 0.8703082203865051,
+      "learning_rate": 9.762636663940443e-05,
+      "loss": 0.0902,
+      "step": 25840
+    },
+    {
+      "epoch": 1.6912005233889436,
+      "grad_norm": 0.8051765561103821,
+      "learning_rate": 9.762356914685658e-05,
+      "loss": 0.093,
+      "step": 25850
+    },
+    {
+      "epoch": 1.691854759568204,
+      "grad_norm": 0.8503502607345581,
+      "learning_rate": 9.762077004688836e-05,
+      "loss": 0.0913,
+      "step": 25860
+    },
+    {
+      "epoch": 1.6925089957474648,
+      "grad_norm": 0.7812398672103882,
+      "learning_rate": 9.761796933959428e-05,
+      "loss": 0.0876,
+      "step": 25870
+    },
+    {
+      "epoch": 1.6931632319267256,
+      "grad_norm": 0.8080005049705505,
+      "learning_rate": 9.761516702506886e-05,
+      "loss": 0.0872,
+      "step": 25880
+    },
+    {
+      "epoch": 1.693817468105986,
+      "grad_norm": 0.7798691391944885,
+      "learning_rate": 9.761236310340665e-05,
+      "loss": 0.0895,
+      "step": 25890
+    },
+    {
+      "epoch": 1.694471704285247,
+      "grad_norm": 0.7314333915710449,
+      "learning_rate": 9.760955757470233e-05,
+      "loss": 0.0925,
+      "step": 25900
+    },
+    {
+      "epoch": 1.6951259404645076,
+      "grad_norm": 0.8456811904907227,
+      "learning_rate": 9.760675043905058e-05,
+      "loss": 0.1005,
+      "step": 25910
+    },
+    {
+      "epoch": 1.6957801766437686,
+      "grad_norm": 0.6989087462425232,
+      "learning_rate": 9.760394169654615e-05,
+      "loss": 0.106,
+      "step": 25920
+    },
+    {
+      "epoch": 1.696434412823029,
+      "grad_norm": 1.0372560024261475,
+      "learning_rate": 9.760113134728384e-05,
+      "loss": 0.0865,
+      "step": 25930
+    },
+    {
+      "epoch": 1.6970886490022898,
+      "grad_norm": 1.2391959428787231,
+      "learning_rate": 9.75983193913585e-05,
+      "loss": 0.1008,
+      "step": 25940
+    },
+    {
+      "epoch": 1.6977428851815506,
+      "grad_norm": 0.8071995973587036,
+      "learning_rate": 9.759550582886506e-05,
+      "loss": 0.094,
+      "step": 25950
+    },
+    {
+      "epoch": 1.6983971213608111,
+      "grad_norm": 0.8598625063896179,
+      "learning_rate": 9.759269065989848e-05,
+      "loss": 0.0938,
+      "step": 25960
+    },
+    {
+      "epoch": 1.699051357540072,
+      "grad_norm": 0.853850245475769,
+      "learning_rate": 9.758987388455377e-05,
+      "loss": 0.087,
+      "step": 25970
+    },
+    {
+      "epoch": 1.6997055937193326,
+      "grad_norm": 0.8352051973342896,
+      "learning_rate": 9.7587055502926e-05,
+      "loss": 0.0959,
+      "step": 25980
+    },
+    {
+      "epoch": 1.7003598298985934,
+      "grad_norm": 0.7932831048965454,
+      "learning_rate": 9.758423551511031e-05,
+      "loss": 0.0852,
+      "step": 25990
+    },
+    {
+      "epoch": 1.701014066077854,
+      "grad_norm": 0.8113059997558594,
+      "learning_rate": 9.758141392120188e-05,
+      "loss": 0.099,
+      "step": 26000
+    },
+    {
+      "epoch": 1.7016683022571149,
+      "grad_norm": 0.9581121802330017,
+      "learning_rate": 9.757859072129594e-05,
+      "loss": 0.0961,
+      "step": 26010
+    },
+    {
+      "epoch": 1.7023225384363756,
+      "grad_norm": 0.897538423538208,
+      "learning_rate": 9.757576591548778e-05,
+      "loss": 0.0892,
+      "step": 26020
+    },
+    {
+      "epoch": 1.7029767746156361,
+      "grad_norm": 1.0049359798431396,
+      "learning_rate": 9.757293950387275e-05,
+      "loss": 0.1075,
+      "step": 26030
+    },
+    {
+      "epoch": 1.703631010794897,
+      "grad_norm": 1.0087964534759521,
+      "learning_rate": 9.757011148654625e-05,
+      "loss": 0.1066,
+      "step": 26040
+    },
+    {
+      "epoch": 1.7042852469741576,
+      "grad_norm": 0.9875262379646301,
+      "learning_rate": 9.756728186360373e-05,
+      "loss": 0.0973,
+      "step": 26050
+    },
+    {
+      "epoch": 1.7049394831534184,
+      "grad_norm": 0.837603747844696,
+      "learning_rate": 9.75644506351407e-05,
+      "loss": 0.1014,
+      "step": 26060
+    },
+    {
+      "epoch": 1.7055937193326791,
+      "grad_norm": 0.7701033353805542,
+      "learning_rate": 9.756161780125271e-05,
+      "loss": 0.0923,
+      "step": 26070
+    },
+    {
+      "epoch": 1.7062479555119399,
+      "grad_norm": 0.8201101422309875,
+      "learning_rate": 9.755878336203539e-05,
+      "loss": 0.1013,
+      "step": 26080
+    },
+    {
+      "epoch": 1.7069021916912006,
+      "grad_norm": 0.8958601951599121,
+      "learning_rate": 9.755594731758441e-05,
+      "loss": 0.0874,
+      "step": 26090
+    },
+    {
+      "epoch": 1.7075564278704611,
+      "grad_norm": 0.8118382692337036,
+      "learning_rate": 9.755310966799546e-05,
+      "loss": 0.0847,
+      "step": 26100
+    },
+    {
+      "epoch": 1.708210664049722,
+      "grad_norm": 0.8663161993026733,
+      "learning_rate": 9.755027041336439e-05,
+      "loss": 0.0967,
+      "step": 26110
+    },
+    {
+      "epoch": 1.7088649002289826,
+      "grad_norm": 0.7478710412979126,
+      "learning_rate": 9.754742955378696e-05,
+      "loss": 0.0861,
+      "step": 26120
+    },
+    {
+      "epoch": 1.7095191364082434,
+      "grad_norm": 0.9378357529640198,
+      "learning_rate": 9.75445870893591e-05,
+      "loss": 0.0999,
+      "step": 26130
+    },
+    {
+      "epoch": 1.7101733725875041,
+      "grad_norm": 0.8898491263389587,
+      "learning_rate": 9.754174302017671e-05,
+      "loss": 0.096,
+      "step": 26140
+    },
+    {
+      "epoch": 1.7108276087667647,
+      "grad_norm": 0.9357932806015015,
+      "learning_rate": 9.753889734633583e-05,
+      "loss": 0.0884,
+      "step": 26150
+    },
+    {
+      "epoch": 1.7114818449460256,
+      "grad_norm": 0.7866037487983704,
+      "learning_rate": 9.753605006793249e-05,
+      "loss": 0.0864,
+      "step": 26160
+    },
+    {
+      "epoch": 1.7121360811252861,
+      "grad_norm": 0.938848614692688,
+      "learning_rate": 9.75332011850628e-05,
+      "loss": 0.0993,
+      "step": 26170
+    },
+    {
+      "epoch": 1.712790317304547,
+      "grad_norm": 0.7685279846191406,
+      "learning_rate": 9.753035069782288e-05,
+      "loss": 0.0905,
+      "step": 26180
+    },
+    {
+      "epoch": 1.7134445534838076,
+      "grad_norm": 0.7881633043289185,
+      "learning_rate": 9.7527498606309e-05,
+      "loss": 0.0956,
+      "step": 26190
+    },
+    {
+      "epoch": 1.7140987896630684,
+      "grad_norm": 0.8515254259109497,
+      "learning_rate": 9.752464491061738e-05,
+      "loss": 0.0971,
+      "step": 26200
+    },
+    {
+      "epoch": 1.7147530258423291,
+      "grad_norm": 1.0867912769317627,
+      "learning_rate": 9.752178961084438e-05,
+      "loss": 0.1,
+      "step": 26210
+    },
+    {
+      "epoch": 1.7154072620215897,
+      "grad_norm": 0.8622691631317139,
+      "learning_rate": 9.751893270708631e-05,
+      "loss": 0.0981,
+      "step": 26220
+    },
+    {
+      "epoch": 1.7160614982008506,
+      "grad_norm": 0.8179900050163269,
+      "learning_rate": 9.751607419943966e-05,
+      "loss": 0.0891,
+      "step": 26230
+    },
+    {
+      "epoch": 1.7167157343801112,
+      "grad_norm": 0.9989860653877258,
+      "learning_rate": 9.75132140880009e-05,
+      "loss": 0.0906,
+      "step": 26240
+    },
+    {
+      "epoch": 1.717369970559372,
+      "grad_norm": 0.8796525597572327,
+      "learning_rate": 9.751035237286654e-05,
+      "loss": 0.0915,
+      "step": 26250
+    },
+    {
+      "epoch": 1.7180242067386327,
+      "grad_norm": 0.7103739380836487,
+      "learning_rate": 9.750748905413321e-05,
+      "loss": 0.0898,
+      "step": 26260
+    },
+    {
+      "epoch": 1.7186784429178934,
+      "grad_norm": 0.8540217876434326,
+      "learning_rate": 9.75046241318975e-05,
+      "loss": 0.0971,
+      "step": 26270
+    },
+    {
+      "epoch": 1.7193326790971541,
+      "grad_norm": 0.8470193147659302,
+      "learning_rate": 9.750175760625616e-05,
+      "loss": 0.0885,
+      "step": 26280
+    },
+    {
+      "epoch": 1.7199869152764147,
+      "grad_norm": 0.8234091401100159,
+      "learning_rate": 9.749888947730592e-05,
+      "loss": 0.0924,
+      "step": 26290
+    },
+    {
+      "epoch": 1.7206411514556756,
+      "grad_norm": 0.7980528473854065,
+      "learning_rate": 9.749601974514358e-05,
+      "loss": 0.0907,
+      "step": 26300
+    },
+    {
+      "epoch": 1.7212953876349362,
+      "grad_norm": 0.8894267082214355,
+      "learning_rate": 9.749314840986603e-05,
+      "loss": 0.0888,
+      "step": 26310
+    },
+    {
+      "epoch": 1.721949623814197,
+      "grad_norm": 0.9575929641723633,
+      "learning_rate": 9.749027547157015e-05,
+      "loss": 0.098,
+      "step": 26320
+    },
+    {
+      "epoch": 1.7226038599934577,
+      "grad_norm": 0.9381182193756104,
+      "learning_rate": 9.748740093035293e-05,
+      "loss": 0.0851,
+      "step": 26330
+    },
+    {
+      "epoch": 1.7232580961727182,
+      "grad_norm": 0.8991712927818298,
+      "learning_rate": 9.748452478631139e-05,
+      "loss": 0.1015,
+      "step": 26340
+    },
+    {
+      "epoch": 1.7239123323519792,
+      "grad_norm": 0.9150116443634033,
+      "learning_rate": 9.74816470395426e-05,
+      "loss": 0.095,
+      "step": 26350
+    },
+    {
+      "epoch": 1.7245665685312397,
+      "grad_norm": 0.9800270199775696,
+      "learning_rate": 9.74787676901437e-05,
+      "loss": 0.0929,
+      "step": 26360
+    },
+    {
+      "epoch": 1.7252208047105007,
+      "grad_norm": 0.8880261182785034,
+      "learning_rate": 9.747588673821187e-05,
+      "loss": 0.0904,
+      "step": 26370
+    },
+    {
+      "epoch": 1.7258750408897612,
+      "grad_norm": 0.8214205503463745,
+      "learning_rate": 9.747300418384436e-05,
+      "loss": 0.0933,
+      "step": 26380
+    },
+    {
+      "epoch": 1.726529277069022,
+      "grad_norm": 0.9119507074356079,
+      "learning_rate": 9.747012002713846e-05,
+      "loss": 0.0881,
+      "step": 26390
+    },
+    {
+      "epoch": 1.7271835132482827,
+      "grad_norm": 0.8520674109458923,
+      "learning_rate": 9.746723426819151e-05,
+      "loss": 0.0915,
+      "step": 26400
+    },
+    {
+      "epoch": 1.7278377494275432,
+      "grad_norm": 0.836610734462738,
+      "learning_rate": 9.74643469071009e-05,
+      "loss": 0.0977,
+      "step": 26410
+    },
+    {
+      "epoch": 1.7284919856068042,
+      "grad_norm": 1.0091043710708618,
+      "learning_rate": 9.746145794396412e-05,
+      "loss": 0.0978,
+      "step": 26420
+    },
+    {
+      "epoch": 1.7291462217860647,
+      "grad_norm": 0.77135169506073,
+      "learning_rate": 9.745856737887866e-05,
+      "loss": 0.099,
+      "step": 26430
+    },
+    {
+      "epoch": 1.7298004579653254,
+      "grad_norm": 0.9269153475761414,
+      "learning_rate": 9.745567521194207e-05,
+      "loss": 0.0883,
+      "step": 26440
+    },
+    {
+      "epoch": 1.7304546941445862,
+      "grad_norm": 0.8424230813980103,
+      "learning_rate": 9.7452781443252e-05,
+      "loss": 0.0935,
+      "step": 26450
+    },
+    {
+      "epoch": 1.731108930323847,
+      "grad_norm": 0.7595318555831909,
+      "learning_rate": 9.744988607290611e-05,
+      "loss": 0.0949,
+      "step": 26460
+    },
+    {
+      "epoch": 1.7317631665031077,
+      "grad_norm": 0.7194068431854248,
+      "learning_rate": 9.744698910100211e-05,
+      "loss": 0.099,
+      "step": 26470
+    },
+    {
+      "epoch": 1.7324174026823682,
+      "grad_norm": 0.8426297307014465,
+      "learning_rate": 9.74440905276378e-05,
+      "loss": 0.0878,
+      "step": 26480
+    },
+    {
+      "epoch": 1.7330716388616292,
+      "grad_norm": 0.8402136564254761,
+      "learning_rate": 9.744119035291101e-05,
+      "loss": 0.094,
+      "step": 26490
+    },
+    {
+      "epoch": 1.7337258750408897,
+      "grad_norm": 0.9715792536735535,
+      "learning_rate": 9.743828857691963e-05,
+      "loss": 0.094,
+      "step": 26500
+    },
+    {
+      "epoch": 1.7343801112201505,
+      "grad_norm": 0.8805036544799805,
+      "learning_rate": 9.74353851997616e-05,
+      "loss": 0.1009,
+      "step": 26510
+    },
+    {
+      "epoch": 1.7350343473994112,
+      "grad_norm": 0.9216516017913818,
+      "learning_rate": 9.743248022153491e-05,
+      "loss": 0.0903,
+      "step": 26520
+    },
+    {
+      "epoch": 1.735688583578672,
+      "grad_norm": 0.7972573041915894,
+      "learning_rate": 9.742957364233763e-05,
+      "loss": 0.089,
+      "step": 26530
+    },
+    {
+      "epoch": 1.7363428197579327,
+      "grad_norm": 0.8343783617019653,
+      "learning_rate": 9.742666546226784e-05,
+      "loss": 0.082,
+      "step": 26540
+    },
+    {
+      "epoch": 1.7369970559371932,
+      "grad_norm": 1.03281831741333,
+      "learning_rate": 9.74237556814237e-05,
+      "loss": 0.0882,
+      "step": 26550
+    },
+    {
+      "epoch": 1.7376512921164542,
+      "grad_norm": 0.9120453596115112,
+      "learning_rate": 9.742084429990344e-05,
+      "loss": 0.0942,
+      "step": 26560
+    },
+    {
+      "epoch": 1.7383055282957147,
+      "grad_norm": 0.752260684967041,
+      "learning_rate": 9.741793131780532e-05,
+      "loss": 0.105,
+      "step": 26570
+    },
+    {
+      "epoch": 1.7389597644749755,
+      "grad_norm": 0.8960766792297363,
+      "learning_rate": 9.741501673522767e-05,
+      "loss": 0.1,
+      "step": 26580
+    },
+    {
+      "epoch": 1.7396140006542362,
+      "grad_norm": 0.9580662250518799,
+      "learning_rate": 9.741210055226883e-05,
+      "loss": 0.0972,
+      "step": 26590
+    },
+    {
+      "epoch": 1.7402682368334967,
+      "grad_norm": 0.8709340691566467,
+      "learning_rate": 9.740918276902726e-05,
+      "loss": 0.0925,
+      "step": 26600
+    },
+    {
+      "epoch": 1.7409224730127577,
+      "grad_norm": 0.8603937029838562,
+      "learning_rate": 9.740626338560146e-05,
+      "loss": 0.099,
+      "step": 26610
+    },
+    {
+      "epoch": 1.7415767091920182,
+      "grad_norm": 0.8637518286705017,
+      "learning_rate": 9.740334240208992e-05,
+      "loss": 0.1037,
+      "step": 26620
+    },
+    {
+      "epoch": 1.742230945371279,
+      "grad_norm": 0.954017698764801,
+      "learning_rate": 9.740041981859126e-05,
+      "loss": 0.0986,
+      "step": 26630
+    },
+    {
+      "epoch": 1.7428851815505397,
+      "grad_norm": 0.9584331512451172,
+      "learning_rate": 9.739749563520413e-05,
+      "loss": 0.0905,
+      "step": 26640
+    },
+    {
+      "epoch": 1.7435394177298005,
+      "grad_norm": 0.8306182026863098,
+      "learning_rate": 9.73945698520272e-05,
+      "loss": 0.0921,
+      "step": 26650
+    },
+    {
+      "epoch": 1.7441936539090612,
+      "grad_norm": 0.9190852642059326,
+      "learning_rate": 9.739164246915926e-05,
+      "loss": 0.0971,
+      "step": 26660
+    },
+    {
+      "epoch": 1.7448478900883218,
+      "grad_norm": 0.8570190072059631,
+      "learning_rate": 9.738871348669907e-05,
+      "loss": 0.0914,
+      "step": 26670
+    },
+    {
+      "epoch": 1.7455021262675827,
+      "grad_norm": 0.7864238023757935,
+      "learning_rate": 9.738578290474554e-05,
+      "loss": 0.0974,
+      "step": 26680
+    },
+    {
+      "epoch": 1.7461563624468432,
+      "grad_norm": 0.9038723707199097,
+      "learning_rate": 9.738285072339755e-05,
+      "loss": 0.0914,
+      "step": 26690
+    },
+    {
+      "epoch": 1.746810598626104,
+      "grad_norm": 1.1634936332702637,
+      "learning_rate": 9.73799169427541e-05,
+      "loss": 0.0965,
+      "step": 26700
+    },
+    {
+      "epoch": 1.7474648348053647,
+      "grad_norm": 0.830987811088562,
+      "learning_rate": 9.737698156291418e-05,
+      "loss": 0.1034,
+      "step": 26710
+    },
+    {
+      "epoch": 1.7481190709846255,
+      "grad_norm": 0.7684716582298279,
+      "learning_rate": 9.737404458397688e-05,
+      "loss": 0.0894,
+      "step": 26720
+    },
+    {
+      "epoch": 1.7487733071638862,
+      "grad_norm": 1.0796003341674805,
+      "learning_rate": 9.737110600604135e-05,
+      "loss": 0.1027,
+      "step": 26730
+    },
+    {
+      "epoch": 1.7494275433431468,
+      "grad_norm": 0.8072198629379272,
+      "learning_rate": 9.736816582920674e-05,
+      "loss": 0.0956,
+      "step": 26740
+    },
+    {
+      "epoch": 1.7500817795224077,
+      "grad_norm": 0.9745908379554749,
+      "learning_rate": 9.736522405357231e-05,
+      "loss": 0.0923,
+      "step": 26750
+    },
+    {
+      "epoch": 1.7507360157016683,
+      "grad_norm": 0.9096083641052246,
+      "learning_rate": 9.736228067923735e-05,
+      "loss": 0.0863,
+      "step": 26760
+    },
+    {
+      "epoch": 1.751390251880929,
+      "grad_norm": 0.7566020488739014,
+      "learning_rate": 9.73593357063012e-05,
+      "loss": 0.0921,
+      "step": 26770
+    },
+    {
+      "epoch": 1.7520444880601898,
+      "grad_norm": 0.9019480347633362,
+      "learning_rate": 9.735638913486327e-05,
+      "loss": 0.0888,
+      "step": 26780
+    },
+    {
+      "epoch": 1.7526987242394503,
+      "grad_norm": 0.8336036205291748,
+      "learning_rate": 9.735344096502302e-05,
+      "loss": 0.0952,
+      "step": 26790
+    },
+    {
+      "epoch": 1.7533529604187112,
+      "grad_norm": 0.8614187240600586,
+      "learning_rate": 9.735049119687993e-05,
+      "loss": 0.0946,
+      "step": 26800
+    },
+    {
+      "epoch": 1.7540071965979718,
+      "grad_norm": 0.9070919752120972,
+      "learning_rate": 9.73475398305336e-05,
+      "loss": 0.0853,
+      "step": 26810
+    },
+    {
+      "epoch": 1.7546614327772327,
+      "grad_norm": 0.8516311049461365,
+      "learning_rate": 9.734458686608361e-05,
+      "loss": 0.0972,
+      "step": 26820
+    },
+    {
+      "epoch": 1.7553156689564933,
+      "grad_norm": 1.2225861549377441,
+      "learning_rate": 9.734163230362965e-05,
+      "loss": 0.1026,
+      "step": 26830
+    },
+    {
+      "epoch": 1.755969905135754,
+      "grad_norm": 0.8379483222961426,
+      "learning_rate": 9.733867614327145e-05,
+      "loss": 0.0914,
+      "step": 26840
+    },
+    {
+      "epoch": 1.7566241413150148,
+      "grad_norm": 1.0506983995437622,
+      "learning_rate": 9.733571838510878e-05,
+      "loss": 0.089,
+      "step": 26850
+    },
+    {
+      "epoch": 1.7572783774942753,
+      "grad_norm": 0.8693654537200928,
+      "learning_rate": 9.733275902924146e-05,
+      "loss": 0.1042,
+      "step": 26860
+    },
+    {
+      "epoch": 1.7579326136735363,
+      "grad_norm": 0.8979597091674805,
+      "learning_rate": 9.732979807576941e-05,
+      "loss": 0.0866,
+      "step": 26870
+    },
+    {
+      "epoch": 1.7585868498527968,
+      "grad_norm": 1.0074806213378906,
+      "learning_rate": 9.732683552479252e-05,
+      "loss": 0.1029,
+      "step": 26880
+    },
+    {
+      "epoch": 1.7592410860320575,
+      "grad_norm": 0.7297909259796143,
+      "learning_rate": 9.732387137641084e-05,
+      "loss": 0.1038,
+      "step": 26890
+    },
+    {
+      "epoch": 1.7598953222113183,
+      "grad_norm": 0.7585257887840271,
+      "learning_rate": 9.732090563072437e-05,
+      "loss": 0.1036,
+      "step": 26900
+    },
+    {
+      "epoch": 1.760549558390579,
+      "grad_norm": 0.7857770919799805,
+      "learning_rate": 9.731793828783323e-05,
+      "loss": 0.0912,
+      "step": 26910
+    },
+    {
+      "epoch": 1.7612037945698398,
+      "grad_norm": 0.7956094741821289,
+      "learning_rate": 9.731496934783759e-05,
+      "loss": 0.0953,
+      "step": 26920
+    },
+    {
+      "epoch": 1.7618580307491003,
+      "grad_norm": 0.9437527656555176,
+      "learning_rate": 9.731199881083763e-05,
+      "loss": 0.0901,
+      "step": 26930
+    },
+    {
+      "epoch": 1.7625122669283613,
+      "grad_norm": 0.733113169670105,
+      "learning_rate": 9.730902667693365e-05,
+      "loss": 0.0945,
+      "step": 26940
+    },
+    {
+      "epoch": 1.7631665031076218,
+      "grad_norm": 0.9366680383682251,
+      "learning_rate": 9.730605294622593e-05,
+      "loss": 0.1043,
+      "step": 26950
+    },
+    {
+      "epoch": 1.7638207392868825,
+      "grad_norm": 0.963205873966217,
+      "learning_rate": 9.730307761881487e-05,
+      "loss": 0.1127,
+      "step": 26960
+    },
+    {
+      "epoch": 1.7644749754661433,
+      "grad_norm": 0.8156423568725586,
+      "learning_rate": 9.730010069480088e-05,
+      "loss": 0.088,
+      "step": 26970
+    },
+    {
+      "epoch": 1.765129211645404,
+      "grad_norm": 1.1396329402923584,
+      "learning_rate": 9.729712217428444e-05,
+      "loss": 0.0834,
+      "step": 26980
+    },
+    {
+      "epoch": 1.7657834478246648,
+      "grad_norm": 0.7981417775154114,
+      "learning_rate": 9.72941420573661e-05,
+      "loss": 0.0932,
+      "step": 26990
+    },
+    {
+      "epoch": 1.7664376840039253,
+      "grad_norm": 0.8251798748970032,
+      "learning_rate": 9.729116034414641e-05,
+      "loss": 0.085,
+      "step": 27000
+    },
+    {
+      "epoch": 1.7670919201831863,
+      "grad_norm": 0.8225140571594238,
+      "learning_rate": 9.728817703472604e-05,
+      "loss": 0.0906,
+      "step": 27010
+    },
+    {
+      "epoch": 1.7677461563624468,
+      "grad_norm": 0.8406652808189392,
+      "learning_rate": 9.728519212920568e-05,
+      "loss": 0.0907,
+      "step": 27020
+    },
+    {
+      "epoch": 1.7684003925417076,
+      "grad_norm": 0.7829009294509888,
+      "learning_rate": 9.728220562768607e-05,
+      "loss": 0.0933,
+      "step": 27030
+    },
+    {
+      "epoch": 1.7690546287209683,
+      "grad_norm": 0.6848630309104919,
+      "learning_rate": 9.727921753026802e-05,
+      "loss": 0.0981,
+      "step": 27040
+    },
+    {
+      "epoch": 1.7697088649002288,
+      "grad_norm": 0.7441691756248474,
+      "learning_rate": 9.727622783705239e-05,
+      "loss": 0.0891,
+      "step": 27050
+    },
+    {
+      "epoch": 1.7703631010794898,
+      "grad_norm": 0.8913453221321106,
+      "learning_rate": 9.727323654814009e-05,
+      "loss": 0.0987,
+      "step": 27060
+    },
+    {
+      "epoch": 1.7710173372587503,
+      "grad_norm": 0.7993099093437195,
+      "learning_rate": 9.727024366363206e-05,
+      "loss": 0.0904,
+      "step": 27070
+    },
+    {
+      "epoch": 1.771671573438011,
+      "grad_norm": 0.821160078048706,
+      "learning_rate": 9.726724918362935e-05,
+      "loss": 0.0987,
+      "step": 27080
+    },
+    {
+      "epoch": 1.7723258096172718,
+      "grad_norm": 0.7259571552276611,
+      "learning_rate": 9.7264253108233e-05,
+      "loss": 0.0973,
+      "step": 27090
+    },
+    {
+      "epoch": 1.7729800457965326,
+      "grad_norm": 0.8779374361038208,
+      "learning_rate": 9.726125543754417e-05,
+      "loss": 0.0934,
+      "step": 27100
+    },
+    {
+      "epoch": 1.7736342819757933,
+      "grad_norm": 0.7990883588790894,
+      "learning_rate": 9.725825617166402e-05,
+      "loss": 0.0946,
+      "step": 27110
+    },
+    {
+      "epoch": 1.7742885181550538,
+      "grad_norm": 0.7716463804244995,
+      "learning_rate": 9.725525531069377e-05,
+      "loss": 0.0956,
+      "step": 27120
+    },
+    {
+      "epoch": 1.7749427543343148,
+      "grad_norm": 1.0061755180358887,
+      "learning_rate": 9.725225285473473e-05,
+      "loss": 0.0804,
+      "step": 27130
+    },
+    {
+      "epoch": 1.7755969905135753,
+      "grad_norm": 0.9990642070770264,
+      "learning_rate": 9.724924880388824e-05,
+      "loss": 0.1089,
+      "step": 27140
+    },
+    {
+      "epoch": 1.776251226692836,
+      "grad_norm": 0.8501982688903809,
+      "learning_rate": 9.72462431582557e-05,
+      "loss": 0.0938,
+      "step": 27150
+    },
+    {
+      "epoch": 1.7769054628720968,
+      "grad_norm": 0.8450808525085449,
+      "learning_rate": 9.724323591793851e-05,
+      "loss": 0.0846,
+      "step": 27160
+    },
+    {
+      "epoch": 1.7775596990513576,
+      "grad_norm": 0.6056027412414551,
+      "learning_rate": 9.724022708303824e-05,
+      "loss": 0.0928,
+      "step": 27170
+    },
+    {
+      "epoch": 1.7782139352306183,
+      "grad_norm": 0.7293991446495056,
+      "learning_rate": 9.723721665365639e-05,
+      "loss": 0.0927,
+      "step": 27180
+    },
+    {
+      "epoch": 1.7788681714098789,
+      "grad_norm": 0.8087413311004639,
+      "learning_rate": 9.723420462989461e-05,
+      "loss": 0.0866,
+      "step": 27190
+    },
+    {
+      "epoch": 1.7795224075891398,
+      "grad_norm": 0.9952290654182434,
+      "learning_rate": 9.723119101185455e-05,
+      "loss": 0.0868,
+      "step": 27200
+    },
+    {
+      "epoch": 1.7801766437684003,
+      "grad_norm": 0.6939850449562073,
+      "learning_rate": 9.722817579963789e-05,
+      "loss": 0.0953,
+      "step": 27210
+    },
+    {
+      "epoch": 1.780830879947661,
+      "grad_norm": 0.7775471806526184,
+      "learning_rate": 9.722515899334647e-05,
+      "loss": 0.0985,
+      "step": 27220
+    },
+    {
+      "epoch": 1.7814851161269218,
+      "grad_norm": 0.824198842048645,
+      "learning_rate": 9.722214059308208e-05,
+      "loss": 0.0992,
+      "step": 27230
+    },
+    {
+      "epoch": 1.7821393523061824,
+      "grad_norm": 0.8829124569892883,
+      "learning_rate": 9.72191205989466e-05,
+      "loss": 0.095,
+      "step": 27240
+    },
+    {
+      "epoch": 1.7827935884854433,
+      "grad_norm": 0.8177944421768188,
+      "learning_rate": 9.721609901104194e-05,
+      "loss": 0.094,
+      "step": 27250
+    },
+    {
+      "epoch": 1.7834478246647039,
+      "grad_norm": 1.004304051399231,
+      "learning_rate": 9.721307582947014e-05,
+      "loss": 0.091,
+      "step": 27260
+    },
+    {
+      "epoch": 1.7841020608439648,
+      "grad_norm": 0.8551913499832153,
+      "learning_rate": 9.721005105433319e-05,
+      "loss": 0.0965,
+      "step": 27270
+    },
+    {
+      "epoch": 1.7847562970232254,
+      "grad_norm": 0.7581825256347656,
+      "learning_rate": 9.720702468573321e-05,
+      "loss": 0.0934,
+      "step": 27280
+    },
+    {
+      "epoch": 1.785410533202486,
+      "grad_norm": 0.761549711227417,
+      "learning_rate": 9.720399672377234e-05,
+      "loss": 0.0978,
+      "step": 27290
+    },
+    {
+      "epoch": 1.7860647693817469,
+      "grad_norm": 0.8848084211349487,
+      "learning_rate": 9.72009671685528e-05,
+      "loss": 0.0811,
+      "step": 27300
+    },
+    {
+      "epoch": 1.7867190055610074,
+      "grad_norm": 0.8144980669021606,
+      "learning_rate": 9.719793602017681e-05,
+      "loss": 0.0844,
+      "step": 27310
+    },
+    {
+      "epoch": 1.7873732417402683,
+      "grad_norm": 0.8796505928039551,
+      "learning_rate": 9.71949032787467e-05,
+      "loss": 0.0904,
+      "step": 27320
+    },
+    {
+      "epoch": 1.7880274779195289,
+      "grad_norm": 0.8658892512321472,
+      "learning_rate": 9.719186894436484e-05,
+      "loss": 0.0914,
+      "step": 27330
+    },
+    {
+      "epoch": 1.7886817140987896,
+      "grad_norm": 0.9171237349510193,
+      "learning_rate": 9.718883301713363e-05,
+      "loss": 0.0923,
+      "step": 27340
+    },
+    {
+      "epoch": 1.7893359502780504,
+      "grad_norm": 0.7952425479888916,
+      "learning_rate": 9.718579549715555e-05,
+      "loss": 0.1069,
+      "step": 27350
+    },
+    {
+      "epoch": 1.7899901864573111,
+      "grad_norm": 0.8469095230102539,
+      "learning_rate": 9.718275638453312e-05,
+      "loss": 0.0942,
+      "step": 27360
+    },
+    {
+      "epoch": 1.7906444226365719,
+      "grad_norm": 0.8001088500022888,
+      "learning_rate": 9.717971567936892e-05,
+      "loss": 0.0931,
+      "step": 27370
+    },
+    {
+      "epoch": 1.7912986588158324,
+      "grad_norm": 0.9924861192703247,
+      "learning_rate": 9.71766733817656e-05,
+      "loss": 0.0924,
+      "step": 27380
+    },
+    {
+      "epoch": 1.7919528949950934,
+      "grad_norm": 0.79388028383255,
+      "learning_rate": 9.71736294918258e-05,
+      "loss": 0.0949,
+      "step": 27390
+    },
+    {
+      "epoch": 1.7926071311743539,
+      "grad_norm": 0.8403041362762451,
+      "learning_rate": 9.71705840096523e-05,
+      "loss": 0.084,
+      "step": 27400
+    },
+    {
+      "epoch": 1.7932613673536146,
+      "grad_norm": 0.9250532984733582,
+      "learning_rate": 9.716753693534791e-05,
+      "loss": 0.092,
+      "step": 27410
+    },
+    {
+      "epoch": 1.7939156035328754,
+      "grad_norm": 0.8198657035827637,
+      "learning_rate": 9.716448826901541e-05,
+      "loss": 0.096,
+      "step": 27420
+    },
+    {
+      "epoch": 1.7945698397121361,
+      "grad_norm": 0.7525361180305481,
+      "learning_rate": 9.716143801075775e-05,
+      "loss": 0.0925,
+      "step": 27430
+    },
+    {
+      "epoch": 1.7952240758913969,
+      "grad_norm": 0.7295049428939819,
+      "learning_rate": 9.715838616067786e-05,
+      "loss": 0.0912,
+      "step": 27440
+    },
+    {
+      "epoch": 1.7958783120706574,
+      "grad_norm": 0.9026070833206177,
+      "learning_rate": 9.715533271887876e-05,
+      "loss": 0.0941,
+      "step": 27450
+    },
+    {
+      "epoch": 1.7965325482499184,
+      "grad_norm": 0.9286572933197021,
+      "learning_rate": 9.715227768546354e-05,
+      "loss": 0.0895,
+      "step": 27460
+    },
+    {
+      "epoch": 1.797186784429179,
+      "grad_norm": 0.6571982502937317,
+      "learning_rate": 9.714922106053526e-05,
+      "loss": 0.0912,
+      "step": 27470
+    },
+    {
+      "epoch": 1.7978410206084396,
+      "grad_norm": 0.7618511915206909,
+      "learning_rate": 9.71461628441971e-05,
+      "loss": 0.0883,
+      "step": 27480
+    },
+    {
+      "epoch": 1.7984952567877004,
+      "grad_norm": 0.8936156630516052,
+      "learning_rate": 9.714310303655234e-05,
+      "loss": 0.0926,
+      "step": 27490
+    },
+    {
+      "epoch": 1.799149492966961,
+      "grad_norm": 0.7239774465560913,
+      "learning_rate": 9.71400416377042e-05,
+      "loss": 0.0846,
+      "step": 27500
+    },
+    {
+      "epoch": 1.7998037291462219,
+      "grad_norm": 0.7614884376525879,
+      "learning_rate": 9.713697864775601e-05,
+      "loss": 0.1061,
+      "step": 27510
+    },
+    {
+      "epoch": 1.8004579653254824,
+      "grad_norm": 0.8885960578918457,
+      "learning_rate": 9.713391406681118e-05,
+      "loss": 0.0838,
+      "step": 27520
+    },
+    {
+      "epoch": 1.8011122015047432,
+      "grad_norm": 0.7284700274467468,
+      "learning_rate": 9.713084789497315e-05,
+      "loss": 0.0917,
+      "step": 27530
+    },
+    {
+      "epoch": 1.801766437684004,
+      "grad_norm": 0.7352108955383301,
+      "learning_rate": 9.712778013234538e-05,
+      "loss": 0.0833,
+      "step": 27540
+    },
+    {
+      "epoch": 1.8024206738632647,
+      "grad_norm": 0.7528089880943298,
+      "learning_rate": 9.712471077903144e-05,
+      "loss": 0.0889,
+      "step": 27550
+    },
+    {
+      "epoch": 1.8030749100425254,
+      "grad_norm": 0.895668625831604,
+      "learning_rate": 9.712163983513491e-05,
+      "loss": 0.0957,
+      "step": 27560
+    },
+    {
+      "epoch": 1.803729146221786,
+      "grad_norm": 0.7745876908302307,
+      "learning_rate": 9.711856730075948e-05,
+      "loss": 0.1034,
+      "step": 27570
+    },
+    {
+      "epoch": 1.804383382401047,
+      "grad_norm": 0.9422993659973145,
+      "learning_rate": 9.711549317600881e-05,
+      "loss": 0.0915,
+      "step": 27580
+    },
+    {
+      "epoch": 1.8050376185803074,
+      "grad_norm": 0.8201385140419006,
+      "learning_rate": 9.711241746098669e-05,
+      "loss": 0.092,
+      "step": 27590
+    },
+    {
+      "epoch": 1.8056918547595682,
+      "grad_norm": 0.8114073872566223,
+      "learning_rate": 9.710934015579693e-05,
+      "loss": 0.0939,
+      "step": 27600
+    },
+    {
+      "epoch": 1.806346090938829,
+      "grad_norm": 0.742363452911377,
+      "learning_rate": 9.710626126054338e-05,
+      "loss": 0.0994,
+      "step": 27610
+    },
+    {
+      "epoch": 1.8070003271180897,
+      "grad_norm": 0.9187309145927429,
+      "learning_rate": 9.710318077532998e-05,
+      "loss": 0.0937,
+      "step": 27620
+    },
+    {
+      "epoch": 1.8076545632973504,
+      "grad_norm": 0.9056969881057739,
+      "learning_rate": 9.71000987002607e-05,
+      "loss": 0.0975,
+      "step": 27630
+    },
+    {
+      "epoch": 1.808308799476611,
+      "grad_norm": 0.7385011911392212,
+      "learning_rate": 9.709701503543954e-05,
+      "loss": 0.1049,
+      "step": 27640
+    },
+    {
+      "epoch": 1.808963035655872,
+      "grad_norm": 0.8550733327865601,
+      "learning_rate": 9.709392978097061e-05,
+      "loss": 0.0927,
+      "step": 27650
+    },
+    {
+      "epoch": 1.8096172718351324,
+      "grad_norm": 0.8086677193641663,
+      "learning_rate": 9.709084293695806e-05,
+      "loss": 0.0882,
+      "step": 27660
+    },
+    {
+      "epoch": 1.8102715080143932,
+      "grad_norm": 0.8741077184677124,
+      "learning_rate": 9.708775450350605e-05,
+      "loss": 0.0907,
+      "step": 27670
+    },
+    {
+      "epoch": 1.810925744193654,
+      "grad_norm": 0.7944101095199585,
+      "learning_rate": 9.708466448071884e-05,
+      "loss": 0.084,
+      "step": 27680
+    },
+    {
+      "epoch": 1.8115799803729145,
+      "grad_norm": 0.9104844331741333,
+      "learning_rate": 9.708157286870072e-05,
+      "loss": 0.1128,
+      "step": 27690
+    },
+    {
+      "epoch": 1.8122342165521754,
+      "grad_norm": 0.8403039574623108,
+      "learning_rate": 9.707847966755604e-05,
+      "loss": 0.0835,
+      "step": 27700
+    },
+    {
+      "epoch": 1.812888452731436,
+      "grad_norm": 0.8909661769866943,
+      "learning_rate": 9.707538487738918e-05,
+      "loss": 0.0925,
+      "step": 27710
+    },
+    {
+      "epoch": 1.813542688910697,
+      "grad_norm": 0.8909949660301208,
+      "learning_rate": 9.707228849830465e-05,
+      "loss": 0.0944,
+      "step": 27720
+    },
+    {
+      "epoch": 1.8141969250899574,
+      "grad_norm": 1.029722809791565,
+      "learning_rate": 9.706919053040692e-05,
+      "loss": 0.1,
+      "step": 27730
+    },
+    {
+      "epoch": 1.8148511612692182,
+      "grad_norm": 0.7977635264396667,
+      "learning_rate": 9.706609097380058e-05,
+      "loss": 0.0894,
+      "step": 27740
+    },
+    {
+      "epoch": 1.815505397448479,
+      "grad_norm": 0.9570349454879761,
+      "learning_rate": 9.706298982859021e-05,
+      "loss": 0.0977,
+      "step": 27750
+    },
+    {
+      "epoch": 1.8161596336277395,
+      "grad_norm": 0.8739515542984009,
+      "learning_rate": 9.705988709488052e-05,
+      "loss": 0.1012,
+      "step": 27760
+    },
+    {
+      "epoch": 1.8168138698070004,
+      "grad_norm": 0.8953158259391785,
+      "learning_rate": 9.705678277277622e-05,
+      "loss": 0.1025,
+      "step": 27770
+    },
+    {
+      "epoch": 1.817468105986261,
+      "grad_norm": 0.8194689154624939,
+      "learning_rate": 9.70536768623821e-05,
+      "loss": 0.0959,
+      "step": 27780
+    },
+    {
+      "epoch": 1.8181223421655217,
+      "grad_norm": 1.037415623664856,
+      "learning_rate": 9.705056936380296e-05,
+      "loss": 0.0908,
+      "step": 27790
+    },
+    {
+      "epoch": 1.8187765783447825,
+      "grad_norm": 1.1118489503860474,
+      "learning_rate": 9.704746027714372e-05,
+      "loss": 0.0972,
+      "step": 27800
+    },
+    {
+      "epoch": 1.8194308145240432,
+      "grad_norm": 0.847061276435852,
+      "learning_rate": 9.704434960250931e-05,
+      "loss": 0.0857,
+      "step": 27810
+    },
+    {
+      "epoch": 1.820085050703304,
+      "grad_norm": 1.020371437072754,
+      "learning_rate": 9.704123734000473e-05,
+      "loss": 0.1029,
+      "step": 27820
+    },
+    {
+      "epoch": 1.8207392868825645,
+      "grad_norm": 1.1115399599075317,
+      "learning_rate": 9.703812348973501e-05,
+      "loss": 0.0988,
+      "step": 27830
+    },
+    {
+      "epoch": 1.8213935230618254,
+      "grad_norm": 0.8873423337936401,
+      "learning_rate": 9.703500805180527e-05,
+      "loss": 0.0971,
+      "step": 27840
+    },
+    {
+      "epoch": 1.822047759241086,
+      "grad_norm": 0.7394191026687622,
+      "learning_rate": 9.703189102632064e-05,
+      "loss": 0.0968,
+      "step": 27850
+    },
+    {
+      "epoch": 1.8227019954203467,
+      "grad_norm": 0.965347945690155,
+      "learning_rate": 9.702877241338635e-05,
+      "loss": 0.0818,
+      "step": 27860
+    },
+    {
+      "epoch": 1.8233562315996075,
+      "grad_norm": 0.8741346597671509,
+      "learning_rate": 9.702565221310766e-05,
+      "loss": 0.1045,
+      "step": 27870
+    },
+    {
+      "epoch": 1.8240104677788682,
+      "grad_norm": 0.7647256255149841,
+      "learning_rate": 9.702253042558986e-05,
+      "loss": 0.0899,
+      "step": 27880
+    },
+    {
+      "epoch": 1.824664703958129,
+      "grad_norm": 0.8359330296516418,
+      "learning_rate": 9.701940705093835e-05,
+      "loss": 0.095,
+      "step": 27890
+    },
+    {
+      "epoch": 1.8253189401373895,
+      "grad_norm": 0.9208669066429138,
+      "learning_rate": 9.701628208925855e-05,
+      "loss": 0.101,
+      "step": 27900
+    },
+    {
+      "epoch": 1.8259731763166505,
+      "grad_norm": 0.9185352325439453,
+      "learning_rate": 9.70131555406559e-05,
+      "loss": 0.0914,
+      "step": 27910
+    },
+    {
+      "epoch": 1.826627412495911,
+      "grad_norm": 0.8828587532043457,
+      "learning_rate": 9.701002740523597e-05,
+      "loss": 0.0881,
+      "step": 27920
+    },
+    {
+      "epoch": 1.8272816486751717,
+      "grad_norm": 0.8667250275611877,
+      "learning_rate": 9.700689768310434e-05,
+      "loss": 0.0945,
+      "step": 27930
+    },
+    {
+      "epoch": 1.8279358848544325,
+      "grad_norm": 0.9793092012405396,
+      "learning_rate": 9.700376637436662e-05,
+      "loss": 0.1026,
+      "step": 27940
+    },
+    {
+      "epoch": 1.828590121033693,
+      "grad_norm": 0.8025292158126831,
+      "learning_rate": 9.70006334791285e-05,
+      "loss": 0.0846,
+      "step": 27950
+    },
+    {
+      "epoch": 1.829244357212954,
+      "grad_norm": 0.9942945241928101,
+      "learning_rate": 9.699749899749576e-05,
+      "loss": 0.09,
+      "step": 27960
+    },
+    {
+      "epoch": 1.8298985933922145,
+      "grad_norm": 0.8808379769325256,
+      "learning_rate": 9.699436292957414e-05,
+      "loss": 0.0887,
+      "step": 27970
+    },
+    {
+      "epoch": 1.8305528295714752,
+      "grad_norm": 0.8378937840461731,
+      "learning_rate": 9.699122527546955e-05,
+      "loss": 0.0885,
+      "step": 27980
+    },
+    {
+      "epoch": 1.831207065750736,
+      "grad_norm": 0.7837172150611877,
+      "learning_rate": 9.698808603528786e-05,
+      "loss": 0.0812,
+      "step": 27990
+    },
+    {
+      "epoch": 1.8318613019299967,
+      "grad_norm": 0.7706719636917114,
+      "learning_rate": 9.698494520913503e-05,
+      "loss": 0.0887,
+      "step": 28000
+    },
+    {
+      "epoch": 1.8325155381092575,
+      "grad_norm": 1.1025103330612183,
+      "learning_rate": 9.69818027971171e-05,
+      "loss": 0.0896,
+      "step": 28010
+    },
+    {
+      "epoch": 1.833169774288518,
+      "grad_norm": 0.9141650795936584,
+      "learning_rate": 9.697865879934009e-05,
+      "loss": 0.1002,
+      "step": 28020
+    },
+    {
+      "epoch": 1.833824010467779,
+      "grad_norm": 0.7713013887405396,
+      "learning_rate": 9.697551321591014e-05,
+      "loss": 0.1056,
+      "step": 28030
+    },
+    {
+      "epoch": 1.8344782466470395,
+      "grad_norm": 0.8746879696846008,
+      "learning_rate": 9.697236604693343e-05,
+      "loss": 0.1024,
+      "step": 28040
+    },
+    {
+      "epoch": 1.8351324828263003,
+      "grad_norm": 0.8916308283805847,
+      "learning_rate": 9.696921729251617e-05,
+      "loss": 0.0951,
+      "step": 28050
+    },
+    {
+      "epoch": 1.835786719005561,
+      "grad_norm": 0.7855492830276489,
+      "learning_rate": 9.696606695276464e-05,
+      "loss": 0.0857,
+      "step": 28060
+    },
+    {
+      "epoch": 1.8364409551848218,
+      "grad_norm": 0.8872267603874207,
+      "learning_rate": 9.696291502778519e-05,
+      "loss": 0.0942,
+      "step": 28070
+    },
+    {
+      "epoch": 1.8370951913640825,
+      "grad_norm": 0.8778836727142334,
+      "learning_rate": 9.695976151768419e-05,
+      "loss": 0.0933,
+      "step": 28080
+    },
+    {
+      "epoch": 1.837749427543343,
+      "grad_norm": 0.8608387112617493,
+      "learning_rate": 9.695660642256807e-05,
+      "loss": 0.0811,
+      "step": 28090
+    },
+    {
+      "epoch": 1.838403663722604,
+      "grad_norm": 0.9035623073577881,
+      "learning_rate": 9.695344974254336e-05,
+      "loss": 0.1077,
+      "step": 28100
+    },
+    {
+      "epoch": 1.8390578999018645,
+      "grad_norm": 1.024272084236145,
+      "learning_rate": 9.695029147771655e-05,
+      "loss": 0.0941,
+      "step": 28110
+    },
+    {
+      "epoch": 1.8397121360811253,
+      "grad_norm": 0.9248595833778381,
+      "learning_rate": 9.69471316281943e-05,
+      "loss": 0.1043,
+      "step": 28120
+    },
+    {
+      "epoch": 1.840366372260386,
+      "grad_norm": 0.7749372720718384,
+      "learning_rate": 9.694397019408322e-05,
+      "loss": 0.0868,
+      "step": 28130
+    },
+    {
+      "epoch": 1.8410206084396468,
+      "grad_norm": 0.88482666015625,
+      "learning_rate": 9.694080717549004e-05,
+      "loss": 0.101,
+      "step": 28140
+    },
+    {
+      "epoch": 1.8416748446189075,
+      "grad_norm": 0.8212443590164185,
+      "learning_rate": 9.693764257252149e-05,
+      "loss": 0.0838,
+      "step": 28150
+    },
+    {
+      "epoch": 1.842329080798168,
+      "grad_norm": 0.7402942180633545,
+      "learning_rate": 9.693447638528443e-05,
+      "loss": 0.0853,
+      "step": 28160
+    },
+    {
+      "epoch": 1.842983316977429,
+      "grad_norm": 0.8029096126556396,
+      "learning_rate": 9.693130861388569e-05,
+      "loss": 0.081,
+      "step": 28170
+    },
+    {
+      "epoch": 1.8436375531566895,
+      "grad_norm": 0.9067233204841614,
+      "learning_rate": 9.69281392584322e-05,
+      "loss": 0.0928,
+      "step": 28180
+    },
+    {
+      "epoch": 1.8442917893359503,
+      "grad_norm": 0.959036111831665,
+      "learning_rate": 9.692496831903092e-05,
+      "loss": 0.0982,
+      "step": 28190
+    },
+    {
+      "epoch": 1.844946025515211,
+      "grad_norm": 1.1850526332855225,
+      "learning_rate": 9.692179579578893e-05,
+      "loss": 0.0981,
+      "step": 28200
+    },
+    {
+      "epoch": 1.8456002616944716,
+      "grad_norm": 0.9170256853103638,
+      "learning_rate": 9.691862168881325e-05,
+      "loss": 0.0931,
+      "step": 28210
+    },
+    {
+      "epoch": 1.8462544978737325,
+      "grad_norm": 0.894639790058136,
+      "learning_rate": 9.691544599821105e-05,
+      "loss": 0.0802,
+      "step": 28220
+    },
+    {
+      "epoch": 1.846908734052993,
+      "grad_norm": 0.8013020157814026,
+      "learning_rate": 9.69122687240895e-05,
+      "loss": 0.0937,
+      "step": 28230
+    },
+    {
+      "epoch": 1.8475629702322538,
+      "grad_norm": 0.93445885181427,
+      "learning_rate": 9.690908986655586e-05,
+      "loss": 0.0911,
+      "step": 28240
+    },
+    {
+      "epoch": 1.8482172064115145,
+      "grad_norm": 0.7632222771644592,
+      "learning_rate": 9.69059094257174e-05,
+      "loss": 0.0953,
+      "step": 28250
+    },
+    {
+      "epoch": 1.8488714425907753,
+      "grad_norm": 0.9063422083854675,
+      "learning_rate": 9.690272740168149e-05,
+      "loss": 0.0885,
+      "step": 28260
+    },
+    {
+      "epoch": 1.849525678770036,
+      "grad_norm": 0.652265191078186,
+      "learning_rate": 9.689954379455552e-05,
+      "loss": 0.0847,
+      "step": 28270
+    },
+    {
+      "epoch": 1.8501799149492966,
+      "grad_norm": 0.8725261688232422,
+      "learning_rate": 9.689635860444696e-05,
+      "loss": 0.092,
+      "step": 28280
+    },
+    {
+      "epoch": 1.8508341511285575,
+      "grad_norm": 0.7449600696563721,
+      "learning_rate": 9.689317183146329e-05,
+      "loss": 0.0878,
+      "step": 28290
+    },
+    {
+      "epoch": 1.851488387307818,
+      "grad_norm": 0.8442542552947998,
+      "learning_rate": 9.68899834757121e-05,
+      "loss": 0.0938,
+      "step": 28300
+    },
+    {
+      "epoch": 1.8521426234870788,
+      "grad_norm": 0.9524549841880798,
+      "learning_rate": 9.6886793537301e-05,
+      "loss": 0.1059,
+      "step": 28310
+    },
+    {
+      "epoch": 1.8527968596663396,
+      "grad_norm": 0.9726237654685974,
+      "learning_rate": 9.688360201633763e-05,
+      "loss": 0.0942,
+      "step": 28320
+    },
+    {
+      "epoch": 1.8534510958456003,
+      "grad_norm": 0.991133987903595,
+      "learning_rate": 9.688040891292976e-05,
+      "loss": 0.1008,
+      "step": 28330
+    },
+    {
+      "epoch": 1.854105332024861,
+      "grad_norm": 0.9573308229446411,
+      "learning_rate": 9.687721422718512e-05,
+      "loss": 0.0898,
+      "step": 28340
+    },
+    {
+      "epoch": 1.8547595682041216,
+      "grad_norm": 1.1355527639389038,
+      "learning_rate": 9.687401795921156e-05,
+      "loss": 0.095,
+      "step": 28350
+    },
+    {
+      "epoch": 1.8554138043833825,
+      "grad_norm": 0.7798793315887451,
+      "learning_rate": 9.687082010911698e-05,
+      "loss": 0.0883,
+      "step": 28360
+    },
+    {
+      "epoch": 1.856068040562643,
+      "grad_norm": 0.5980040431022644,
+      "learning_rate": 9.686762067700928e-05,
+      "loss": 0.0922,
+      "step": 28370
+    },
+    {
+      "epoch": 1.8567222767419038,
+      "grad_norm": 0.9269242286682129,
+      "learning_rate": 9.686441966299649e-05,
+      "loss": 0.0893,
+      "step": 28380
+    },
+    {
+      "epoch": 1.8573765129211646,
+      "grad_norm": 0.6925560235977173,
+      "learning_rate": 9.68612170671866e-05,
+      "loss": 0.0957,
+      "step": 28390
+    },
+    {
+      "epoch": 1.858030749100425,
+      "grad_norm": 0.8151103258132935,
+      "learning_rate": 9.685801288968777e-05,
+      "loss": 0.0925,
+      "step": 28400
+    },
+    {
+      "epoch": 1.858684985279686,
+      "grad_norm": 0.9811034798622131,
+      "learning_rate": 9.685480713060808e-05,
+      "loss": 0.1071,
+      "step": 28410
+    },
+    {
+      "epoch": 1.8593392214589466,
+      "grad_norm": 0.9472489953041077,
+      "learning_rate": 9.68515997900558e-05,
+      "loss": 0.089,
+      "step": 28420
+    },
+    {
+      "epoch": 1.8599934576382076,
+      "grad_norm": 0.8960433602333069,
+      "learning_rate": 9.684839086813913e-05,
+      "loss": 0.1004,
+      "step": 28430
+    },
+    {
+      "epoch": 1.860647693817468,
+      "grad_norm": 0.9435994625091553,
+      "learning_rate": 9.684518036496641e-05,
+      "loss": 0.0928,
+      "step": 28440
+    },
+    {
+      "epoch": 1.8613019299967288,
+      "grad_norm": 0.8257802128791809,
+      "learning_rate": 9.6841968280646e-05,
+      "loss": 0.0994,
+      "step": 28450
+    },
+    {
+      "epoch": 1.8619561661759896,
+      "grad_norm": 0.8301753997802734,
+      "learning_rate": 9.683875461528632e-05,
+      "loss": 0.086,
+      "step": 28460
+    },
+    {
+      "epoch": 1.86261040235525,
+      "grad_norm": 0.8830812573432922,
+      "learning_rate": 9.683553936899583e-05,
+      "loss": 0.0972,
+      "step": 28470
+    },
+    {
+      "epoch": 1.863264638534511,
+      "grad_norm": 0.7733142971992493,
+      "learning_rate": 9.683232254188305e-05,
+      "loss": 0.1017,
+      "step": 28480
+    },
+    {
+      "epoch": 1.8639188747137716,
+      "grad_norm": 0.7784489393234253,
+      "learning_rate": 9.682910413405657e-05,
+      "loss": 0.0959,
+      "step": 28490
+    },
+    {
+      "epoch": 1.8645731108930323,
+      "grad_norm": 0.9781941771507263,
+      "learning_rate": 9.6825884145625e-05,
+      "loss": 0.0977,
+      "step": 28500
+    },
+    {
+      "epoch": 1.865227347072293,
+      "grad_norm": 0.9205982685089111,
+      "learning_rate": 9.682266257669703e-05,
+      "loss": 0.1062,
+      "step": 28510
+    },
+    {
+      "epoch": 1.8658815832515538,
+      "grad_norm": 1.1955499649047852,
+      "learning_rate": 9.681943942738141e-05,
+      "loss": 0.1022,
+      "step": 28520
+    },
+    {
+      "epoch": 1.8665358194308146,
+      "grad_norm": 0.8939968943595886,
+      "learning_rate": 9.681621469778692e-05,
+      "loss": 0.0958,
+      "step": 28530
+    },
+    {
+      "epoch": 1.8671900556100751,
+      "grad_norm": 0.7560677528381348,
+      "learning_rate": 9.681298838802242e-05,
+      "loss": 0.0943,
+      "step": 28540
+    },
+    {
+      "epoch": 1.867844291789336,
+      "grad_norm": 0.8495007753372192,
+      "learning_rate": 9.680976049819677e-05,
+      "loss": 0.0946,
+      "step": 28550
+    },
+    {
+      "epoch": 1.8684985279685966,
+      "grad_norm": 0.8184399008750916,
+      "learning_rate": 9.680653102841895e-05,
+      "loss": 0.0948,
+      "step": 28560
+    },
+    {
+      "epoch": 1.8691527641478574,
+      "grad_norm": 1.0168190002441406,
+      "learning_rate": 9.680329997879795e-05,
+      "loss": 0.0905,
+      "step": 28570
+    },
+    {
+      "epoch": 1.869807000327118,
+      "grad_norm": 0.8939940929412842,
+      "learning_rate": 9.680006734944283e-05,
+      "loss": 0.0849,
+      "step": 28580
+    },
+    {
+      "epoch": 1.8704612365063789,
+      "grad_norm": 0.9259584546089172,
+      "learning_rate": 9.679683314046269e-05,
+      "loss": 0.0986,
+      "step": 28590
+    },
+    {
+      "epoch": 1.8711154726856396,
+      "grad_norm": 0.8003655076026917,
+      "learning_rate": 9.679359735196671e-05,
+      "loss": 0.087,
+      "step": 28600
+    },
+    {
+      "epoch": 1.8717697088649001,
+      "grad_norm": 0.8092601895332336,
+      "learning_rate": 9.67903599840641e-05,
+      "loss": 0.0937,
+      "step": 28610
+    },
+    {
+      "epoch": 1.872423945044161,
+      "grad_norm": 0.757056713104248,
+      "learning_rate": 9.678712103686413e-05,
+      "loss": 0.0857,
+      "step": 28620
+    },
+    {
+      "epoch": 1.8730781812234216,
+      "grad_norm": 0.903866171836853,
+      "learning_rate": 9.678388051047611e-05,
+      "loss": 0.0948,
+      "step": 28630
+    },
+    {
+      "epoch": 1.8737324174026824,
+      "grad_norm": 0.8220004439353943,
+      "learning_rate": 9.678063840500944e-05,
+      "loss": 0.0914,
+      "step": 28640
+    },
+    {
+      "epoch": 1.8743866535819431,
+      "grad_norm": 0.8988321423530579,
+      "learning_rate": 9.677739472057354e-05,
+      "loss": 0.0916,
+      "step": 28650
+    },
+    {
+      "epoch": 1.8750408897612036,
+      "grad_norm": 0.8908348679542542,
+      "learning_rate": 9.677414945727787e-05,
+      "loss": 0.0882,
+      "step": 28660
+    },
+    {
+      "epoch": 1.8756951259404646,
+      "grad_norm": 0.7631081938743591,
+      "learning_rate": 9.677090261523201e-05,
+      "loss": 0.0929,
+      "step": 28670
+    },
+    {
+      "epoch": 1.8763493621197251,
+      "grad_norm": 0.7455977201461792,
+      "learning_rate": 9.676765419454552e-05,
+      "loss": 0.0902,
+      "step": 28680
+    },
+    {
+      "epoch": 1.8770035982989859,
+      "grad_norm": 0.9186832904815674,
+      "learning_rate": 9.676440419532804e-05,
+      "loss": 0.0948,
+      "step": 28690
+    },
+    {
+      "epoch": 1.8776578344782466,
+      "grad_norm": 0.8906334042549133,
+      "learning_rate": 9.676115261768928e-05,
+      "loss": 0.0814,
+      "step": 28700
+    },
+    {
+      "epoch": 1.8783120706575074,
+      "grad_norm": 0.9550861716270447,
+      "learning_rate": 9.675789946173897e-05,
+      "loss": 0.0939,
+      "step": 28710
+    },
+    {
+      "epoch": 1.8789663068367681,
+      "grad_norm": 0.7002192735671997,
+      "learning_rate": 9.675464472758695e-05,
+      "loss": 0.0872,
+      "step": 28720
+    },
+    {
+      "epoch": 1.8796205430160287,
+      "grad_norm": 0.9470962882041931,
+      "learning_rate": 9.675138841534303e-05,
+      "loss": 0.0865,
+      "step": 28730
+    },
+    {
+      "epoch": 1.8802747791952896,
+      "grad_norm": 0.8719517588615417,
+      "learning_rate": 9.674813052511715e-05,
+      "loss": 0.0844,
+      "step": 28740
+    },
+    {
+      "epoch": 1.8809290153745502,
+      "grad_norm": 0.8722051978111267,
+      "learning_rate": 9.674487105701926e-05,
+      "loss": 0.0913,
+      "step": 28750
+    },
+    {
+      "epoch": 1.881583251553811,
+      "grad_norm": 0.8734009265899658,
+      "learning_rate": 9.674161001115938e-05,
+      "loss": 0.0878,
+      "step": 28760
+    },
+    {
+      "epoch": 1.8822374877330716,
+      "grad_norm": 0.8401185274124146,
+      "learning_rate": 9.673834738764759e-05,
+      "loss": 0.0889,
+      "step": 28770
+    },
+    {
+      "epoch": 1.8828917239123324,
+      "grad_norm": 0.8197746872901917,
+      "learning_rate": 9.6735083186594e-05,
+      "loss": 0.0991,
+      "step": 28780
+    },
+    {
+      "epoch": 1.8835459600915931,
+      "grad_norm": 0.8320461511611938,
+      "learning_rate": 9.673181740810876e-05,
+      "loss": 0.0823,
+      "step": 28790
+    },
+    {
+      "epoch": 1.8842001962708537,
+      "grad_norm": 0.9948931336402893,
+      "learning_rate": 9.672855005230214e-05,
+      "loss": 0.0862,
+      "step": 28800
+    },
+    {
+      "epoch": 1.8848544324501146,
+      "grad_norm": 0.8380158543586731,
+      "learning_rate": 9.67252811192844e-05,
+      "loss": 0.0835,
+      "step": 28810
+    },
+    {
+      "epoch": 1.8855086686293752,
+      "grad_norm": 0.8905105590820312,
+      "learning_rate": 9.672201060916589e-05,
+      "loss": 0.1009,
+      "step": 28820
+    },
+    {
+      "epoch": 1.886162904808636,
+      "grad_norm": 0.8555837869644165,
+      "learning_rate": 9.671873852205699e-05,
+      "loss": 0.0983,
+      "step": 28830
+    },
+    {
+      "epoch": 1.8868171409878967,
+      "grad_norm": 0.7890472412109375,
+      "learning_rate": 9.671546485806813e-05,
+      "loss": 0.102,
+      "step": 28840
+    },
+    {
+      "epoch": 1.8874713771671572,
+      "grad_norm": 0.8302340507507324,
+      "learning_rate": 9.671218961730981e-05,
+      "loss": 0.1047,
+      "step": 28850
+    },
+    {
+      "epoch": 1.8881256133464182,
+      "grad_norm": 0.912526547908783,
+      "learning_rate": 9.670891279989261e-05,
+      "loss": 0.0886,
+      "step": 28860
+    },
+    {
+      "epoch": 1.8887798495256787,
+      "grad_norm": 0.9973543882369995,
+      "learning_rate": 9.670563440592709e-05,
+      "loss": 0.0974,
+      "step": 28870
+    },
+    {
+      "epoch": 1.8894340857049396,
+      "grad_norm": 0.8430191874504089,
+      "learning_rate": 9.670235443552391e-05,
+      "loss": 0.084,
+      "step": 28880
+    },
+    {
+      "epoch": 1.8900883218842002,
+      "grad_norm": 0.7398623824119568,
+      "learning_rate": 9.669907288879379e-05,
+      "loss": 0.0874,
+      "step": 28890
+    },
+    {
+      "epoch": 1.890742558063461,
+      "grad_norm": 0.8663063049316406,
+      "learning_rate": 9.669578976584748e-05,
+      "loss": 0.0852,
+      "step": 28900
+    },
+    {
+      "epoch": 1.8913967942427217,
+      "grad_norm": 0.8032066226005554,
+      "learning_rate": 9.669250506679582e-05,
+      "loss": 0.1002,
+      "step": 28910
+    },
+    {
+      "epoch": 1.8920510304219822,
+      "grad_norm": 0.7833549976348877,
+      "learning_rate": 9.668921879174965e-05,
+      "loss": 0.0991,
+      "step": 28920
+    },
+    {
+      "epoch": 1.8927052666012432,
+      "grad_norm": 0.8108584880828857,
+      "learning_rate": 9.66859309408199e-05,
+      "loss": 0.0876,
+      "step": 28930
+    },
+    {
+      "epoch": 1.8933595027805037,
+      "grad_norm": 0.6968421936035156,
+      "learning_rate": 9.668264151411755e-05,
+      "loss": 0.098,
+      "step": 28940
+    },
+    {
+      "epoch": 1.8940137389597644,
+      "grad_norm": 1.000785231590271,
+      "learning_rate": 9.66793505117536e-05,
+      "loss": 0.0879,
+      "step": 28950
+    },
+    {
+      "epoch": 1.8946679751390252,
+      "grad_norm": 0.9198935627937317,
+      "learning_rate": 9.667605793383916e-05,
+      "loss": 0.1009,
+      "step": 28960
+    },
+    {
+      "epoch": 1.895322211318286,
+      "grad_norm": 0.9679718613624573,
+      "learning_rate": 9.667276378048535e-05,
+      "loss": 0.0987,
+      "step": 28970
+    },
+    {
+      "epoch": 1.8959764474975467,
+      "grad_norm": 0.9554332494735718,
+      "learning_rate": 9.666946805180336e-05,
+      "loss": 0.101,
+      "step": 28980
+    },
+    {
+      "epoch": 1.8966306836768072,
+      "grad_norm": 0.9619019031524658,
+      "learning_rate": 9.666617074790442e-05,
+      "loss": 0.103,
+      "step": 28990
+    },
+    {
+      "epoch": 1.8972849198560682,
+      "grad_norm": 0.8431410789489746,
+      "learning_rate": 9.666287186889983e-05,
+      "loss": 0.097,
+      "step": 29000
+    },
+    {
+      "epoch": 1.8979391560353287,
+      "grad_norm": 0.9269760847091675,
+      "learning_rate": 9.665957141490096e-05,
+      "loss": 0.0859,
+      "step": 29010
+    },
+    {
+      "epoch": 1.8985933922145894,
+      "grad_norm": 0.9166405200958252,
+      "learning_rate": 9.665626938601917e-05,
+      "loss": 0.0887,
+      "step": 29020
+    },
+    {
+      "epoch": 1.8992476283938502,
+      "grad_norm": 0.973930835723877,
+      "learning_rate": 9.665296578236593e-05,
+      "loss": 0.0897,
+      "step": 29030
+    },
+    {
+      "epoch": 1.899901864573111,
+      "grad_norm": 0.9470407366752625,
+      "learning_rate": 9.664966060405275e-05,
+      "loss": 0.0925,
+      "step": 29040
+    },
+    {
+      "epoch": 1.9005561007523717,
+      "grad_norm": 0.762370228767395,
+      "learning_rate": 9.664635385119117e-05,
+      "loss": 0.0958,
+      "step": 29050
+    },
+    {
+      "epoch": 1.9012103369316322,
+      "grad_norm": 0.8602021932601929,
+      "learning_rate": 9.66430455238928e-05,
+      "loss": 0.091,
+      "step": 29060
+    },
+    {
+      "epoch": 1.9018645731108932,
+      "grad_norm": 0.8808605074882507,
+      "learning_rate": 9.663973562226934e-05,
+      "loss": 0.0837,
+      "step": 29070
+    },
+    {
+      "epoch": 1.9025188092901537,
+      "grad_norm": 0.7062970399856567,
+      "learning_rate": 9.663642414643248e-05,
+      "loss": 0.0916,
+      "step": 29080
+    },
+    {
+      "epoch": 1.9031730454694145,
+      "grad_norm": 0.7167387008666992,
+      "learning_rate": 9.6633111096494e-05,
+      "loss": 0.0866,
+      "step": 29090
+    },
+    {
+      "epoch": 1.9038272816486752,
+      "grad_norm": 0.91849684715271,
+      "learning_rate": 9.662979647256572e-05,
+      "loss": 0.0882,
+      "step": 29100
+    },
+    {
+      "epoch": 1.9044815178279357,
+      "grad_norm": 0.7310143709182739,
+      "learning_rate": 9.662648027475952e-05,
+      "loss": 0.0947,
+      "step": 29110
+    },
+    {
+      "epoch": 1.9051357540071967,
+      "grad_norm": 0.7814351916313171,
+      "learning_rate": 9.66231625031873e-05,
+      "loss": 0.102,
+      "step": 29120
+    },
+    {
+      "epoch": 1.9057899901864572,
+      "grad_norm": 1.0291553735733032,
+      "learning_rate": 9.661984315796111e-05,
+      "loss": 0.0924,
+      "step": 29130
+    },
+    {
+      "epoch": 1.906444226365718,
+      "grad_norm": 0.6880286335945129,
+      "learning_rate": 9.661652223919293e-05,
+      "loss": 0.0787,
+      "step": 29140
+    },
+    {
+      "epoch": 1.9070984625449787,
+      "grad_norm": 0.7895125150680542,
+      "learning_rate": 9.661319974699487e-05,
+      "loss": 0.0845,
+      "step": 29150
+    },
+    {
+      "epoch": 1.9077526987242395,
+      "grad_norm": 0.846792995929718,
+      "learning_rate": 9.660987568147907e-05,
+      "loss": 0.0884,
+      "step": 29160
+    },
+    {
+      "epoch": 1.9084069349035002,
+      "grad_norm": 1.0212069749832153,
+      "learning_rate": 9.660655004275772e-05,
+      "loss": 0.0912,
+      "step": 29170
+    },
+    {
+      "epoch": 1.9090611710827607,
+      "grad_norm": 0.8728976845741272,
+      "learning_rate": 9.660322283094309e-05,
+      "loss": 0.0873,
+      "step": 29180
+    },
+    {
+      "epoch": 1.9097154072620217,
+      "grad_norm": 0.9119880199432373,
+      "learning_rate": 9.659989404614746e-05,
+      "loss": 0.0921,
+      "step": 29190
+    },
+    {
+      "epoch": 1.9103696434412822,
+      "grad_norm": 0.7712875008583069,
+      "learning_rate": 9.65965636884832e-05,
+      "loss": 0.1135,
+      "step": 29200
+    },
+    {
+      "epoch": 1.911023879620543,
+      "grad_norm": 0.8132479786872864,
+      "learning_rate": 9.659323175806271e-05,
+      "loss": 0.0848,
+      "step": 29210
+    },
+    {
+      "epoch": 1.9116781157998037,
+      "grad_norm": 1.06707763671875,
+      "learning_rate": 9.658989825499845e-05,
+      "loss": 0.0922,
+      "step": 29220
+    },
+    {
+      "epoch": 1.9123323519790645,
+      "grad_norm": 0.9323616623878479,
+      "learning_rate": 9.658656317940293e-05,
+      "loss": 0.1036,
+      "step": 29230
+    },
+    {
+      "epoch": 1.9129865881583252,
+      "grad_norm": 0.8813486099243164,
+      "learning_rate": 9.658322653138873e-05,
+      "loss": 0.096,
+      "step": 29240
+    },
+    {
+      "epoch": 1.9136408243375858,
+      "grad_norm": 0.9012136459350586,
+      "learning_rate": 9.657988831106847e-05,
+      "loss": 0.0916,
+      "step": 29250
+    },
+    {
+      "epoch": 1.9142950605168467,
+      "grad_norm": 0.8024969696998596,
+      "learning_rate": 9.657654851855483e-05,
+      "loss": 0.0926,
+      "step": 29260
+    },
+    {
+      "epoch": 1.9149492966961073,
+      "grad_norm": 0.7480851411819458,
+      "learning_rate": 9.657320715396051e-05,
+      "loss": 0.0975,
+      "step": 29270
+    },
+    {
+      "epoch": 1.915603532875368,
+      "grad_norm": 0.8564775586128235,
+      "learning_rate": 9.65698642173983e-05,
+      "loss": 0.0857,
+      "step": 29280
+    },
+    {
+      "epoch": 1.9162577690546287,
+      "grad_norm": 0.79481440782547,
+      "learning_rate": 9.656651970898105e-05,
+      "loss": 0.0852,
+      "step": 29290
+    },
+    {
+      "epoch": 1.9169120052338893,
+      "grad_norm": 0.8546467423439026,
+      "learning_rate": 9.656317362882164e-05,
+      "loss": 0.0975,
+      "step": 29300
+    },
+    {
+      "epoch": 1.9175662414131502,
+      "grad_norm": 0.8550596833229065,
+      "learning_rate": 9.6559825977033e-05,
+      "loss": 0.0848,
+      "step": 29310
+    },
+    {
+      "epoch": 1.9182204775924108,
+      "grad_norm": 0.8139176368713379,
+      "learning_rate": 9.655647675372812e-05,
+      "loss": 0.0843,
+      "step": 29320
+    },
+    {
+      "epoch": 1.9188747137716717,
+      "grad_norm": 0.847602128982544,
+      "learning_rate": 9.655312595902004e-05,
+      "loss": 0.0865,
+      "step": 29330
+    },
+    {
+      "epoch": 1.9195289499509323,
+      "grad_norm": 0.7845019698143005,
+      "learning_rate": 9.654977359302189e-05,
+      "loss": 0.0959,
+      "step": 29340
+    },
+    {
+      "epoch": 1.920183186130193,
+      "grad_norm": 0.8975633382797241,
+      "learning_rate": 9.654641965584678e-05,
+      "loss": 0.0842,
+      "step": 29350
+    },
+    {
+      "epoch": 1.9208374223094538,
+      "grad_norm": 0.872686505317688,
+      "learning_rate": 9.654306414760796e-05,
+      "loss": 0.0976,
+      "step": 29360
+    },
+    {
+      "epoch": 1.9214916584887143,
+      "grad_norm": 1.081772804260254,
+      "learning_rate": 9.653970706841864e-05,
+      "loss": 0.0854,
+      "step": 29370
+    },
+    {
+      "epoch": 1.9221458946679753,
+      "grad_norm": 0.8015065789222717,
+      "learning_rate": 9.653634841839216e-05,
+      "loss": 0.0988,
+      "step": 29380
+    },
+    {
+      "epoch": 1.9228001308472358,
+      "grad_norm": 0.8011831045150757,
+      "learning_rate": 9.653298819764187e-05,
+      "loss": 0.0838,
+      "step": 29390
+    },
+    {
+      "epoch": 1.9234543670264965,
+      "grad_norm": 0.8473259210586548,
+      "learning_rate": 9.65296264062812e-05,
+      "loss": 0.0897,
+      "step": 29400
+    },
+    {
+      "epoch": 1.9241086032057573,
+      "grad_norm": 0.883358895778656,
+      "learning_rate": 9.652626304442361e-05,
+      "loss": 0.0868,
+      "step": 29410
+    },
+    {
+      "epoch": 1.924762839385018,
+      "grad_norm": 1.020606279373169,
+      "learning_rate": 9.652289811218261e-05,
+      "loss": 0.094,
+      "step": 29420
+    },
+    {
+      "epoch": 1.9254170755642788,
+      "grad_norm": 0.7326846718788147,
+      "learning_rate": 9.65195316096718e-05,
+      "loss": 0.0892,
+      "step": 29430
+    },
+    {
+      "epoch": 1.9260713117435393,
+      "grad_norm": 1.0169782638549805,
+      "learning_rate": 9.651616353700479e-05,
+      "loss": 0.088,
+      "step": 29440
+    },
+    {
+      "epoch": 1.9267255479228003,
+      "grad_norm": 0.9561719298362732,
+      "learning_rate": 9.651279389429526e-05,
+      "loss": 0.0906,
+      "step": 29450
+    },
+    {
+      "epoch": 1.9273797841020608,
+      "grad_norm": 0.8069345355033875,
+      "learning_rate": 9.650942268165698e-05,
+      "loss": 0.1029,
+      "step": 29460
+    },
+    {
+      "epoch": 1.9280340202813215,
+      "grad_norm": 1.1819461584091187,
+      "learning_rate": 9.65060498992037e-05,
+      "loss": 0.0911,
+      "step": 29470
+    },
+    {
+      "epoch": 1.9286882564605823,
+      "grad_norm": 0.7691984176635742,
+      "learning_rate": 9.650267554704924e-05,
+      "loss": 0.0923,
+      "step": 29480
+    },
+    {
+      "epoch": 1.929342492639843,
+      "grad_norm": 0.8642717599868774,
+      "learning_rate": 9.649929962530756e-05,
+      "loss": 0.0843,
+      "step": 29490
+    },
+    {
+      "epoch": 1.9299967288191038,
+      "grad_norm": 0.7647484540939331,
+      "learning_rate": 9.649592213409253e-05,
+      "loss": 0.0902,
+      "step": 29500
+    },
+    {
+      "epoch": 1.9306509649983643,
+      "grad_norm": 0.7451311349868774,
+      "learning_rate": 9.64925430735182e-05,
+      "loss": 0.0902,
+      "step": 29510
+    },
+    {
+      "epoch": 1.9313052011776253,
+      "grad_norm": 0.820818305015564,
+      "learning_rate": 9.648916244369863e-05,
+      "loss": 0.0895,
+      "step": 29520
+    },
+    {
+      "epoch": 1.9319594373568858,
+      "grad_norm": 1.151523232460022,
+      "learning_rate": 9.648578024474789e-05,
+      "loss": 0.092,
+      "step": 29530
+    },
+    {
+      "epoch": 1.9326136735361465,
+      "grad_norm": 0.9633611440658569,
+      "learning_rate": 9.648239647678017e-05,
+      "loss": 0.0947,
+      "step": 29540
+    },
+    {
+      "epoch": 1.9332679097154073,
+      "grad_norm": 0.7115769386291504,
+      "learning_rate": 9.647901113990964e-05,
+      "loss": 0.0909,
+      "step": 29550
+    },
+    {
+      "epoch": 1.9339221458946678,
+      "grad_norm": 0.7603625655174255,
+      "learning_rate": 9.647562423425061e-05,
+      "loss": 0.0846,
+      "step": 29560
+    },
+    {
+      "epoch": 1.9345763820739288,
+      "grad_norm": 1.0014585256576538,
+      "learning_rate": 9.647223575991735e-05,
+      "loss": 0.0954,
+      "step": 29570
+    },
+    {
+      "epoch": 1.9352306182531893,
+      "grad_norm": 0.6401631832122803,
+      "learning_rate": 9.646884571702428e-05,
+      "loss": 0.0848,
+      "step": 29580
+    },
+    {
+      "epoch": 1.93588485443245,
+      "grad_norm": 0.8242216110229492,
+      "learning_rate": 9.64654541056858e-05,
+      "loss": 0.0834,
+      "step": 29590
+    },
+    {
+      "epoch": 1.9365390906117108,
+      "grad_norm": 0.9129480719566345,
+      "learning_rate": 9.646206092601636e-05,
+      "loss": 0.0794,
+      "step": 29600
+    },
+    {
+      "epoch": 1.9371933267909716,
+      "grad_norm": 0.8380915522575378,
+      "learning_rate": 9.645866617813053e-05,
+      "loss": 0.1019,
+      "step": 29610
+    },
+    {
+      "epoch": 1.9378475629702323,
+      "grad_norm": 0.7691987156867981,
+      "learning_rate": 9.645526986214286e-05,
+      "loss": 0.0872,
+      "step": 29620
+    },
+    {
+      "epoch": 1.9385017991494928,
+      "grad_norm": 0.9295036792755127,
+      "learning_rate": 9.6451871978168e-05,
+      "loss": 0.0978,
+      "step": 29630
+    },
+    {
+      "epoch": 1.9391560353287538,
+      "grad_norm": 0.8664805889129639,
+      "learning_rate": 9.644847252632065e-05,
+      "loss": 0.0909,
+      "step": 29640
+    },
+    {
+      "epoch": 1.9398102715080143,
+      "grad_norm": 0.9215870499610901,
+      "learning_rate": 9.644507150671554e-05,
+      "loss": 0.0918,
+      "step": 29650
+    },
+    {
+      "epoch": 1.940464507687275,
+      "grad_norm": 0.8502547740936279,
+      "learning_rate": 9.644166891946745e-05,
+      "loss": 0.086,
+      "step": 29660
+    },
+    {
+      "epoch": 1.9411187438665358,
+      "grad_norm": 0.7667042016983032,
+      "learning_rate": 9.643826476469124e-05,
+      "loss": 0.0816,
+      "step": 29670
+    },
+    {
+      "epoch": 1.9417729800457966,
+      "grad_norm": 0.8172445297241211,
+      "learning_rate": 9.64348590425018e-05,
+      "loss": 0.0926,
+      "step": 29680
+    },
+    {
+      "epoch": 1.9424272162250573,
+      "grad_norm": 0.8571412563323975,
+      "learning_rate": 9.643145175301409e-05,
+      "loss": 0.0836,
+      "step": 29690
+    },
+    {
+      "epoch": 1.9430814524043178,
+      "grad_norm": 0.8760371804237366,
+      "learning_rate": 9.642804289634311e-05,
+      "loss": 0.0851,
+      "step": 29700
+    },
+    {
+      "epoch": 1.9437356885835788,
+      "grad_norm": 0.8582215905189514,
+      "learning_rate": 9.642463247260391e-05,
+      "loss": 0.085,
+      "step": 29710
+    },
+    {
+      "epoch": 1.9443899247628393,
+      "grad_norm": 0.915457010269165,
+      "learning_rate": 9.642122048191164e-05,
+      "loss": 0.091,
+      "step": 29720
+    },
+    {
+      "epoch": 1.9450441609421,
+      "grad_norm": 0.8546823859214783,
+      "learning_rate": 9.641780692438142e-05,
+      "loss": 0.0907,
+      "step": 29730
+    },
+    {
+      "epoch": 1.9456983971213608,
+      "grad_norm": 0.9210653305053711,
+      "learning_rate": 9.641439180012848e-05,
+      "loss": 0.0893,
+      "step": 29740
+    },
+    {
+      "epoch": 1.9463526333006214,
+      "grad_norm": 0.7805274128913879,
+      "learning_rate": 9.641097510926809e-05,
+      "loss": 0.0863,
+      "step": 29750
+    },
+    {
+      "epoch": 1.9470068694798823,
+      "grad_norm": 0.8311102390289307,
+      "learning_rate": 9.640755685191556e-05,
+      "loss": 0.083,
+      "step": 29760
+    },
+    {
+      "epoch": 1.9476611056591429,
+      "grad_norm": 0.9258164763450623,
+      "learning_rate": 9.640413702818629e-05,
+      "loss": 0.0849,
+      "step": 29770
+    },
+    {
+      "epoch": 1.9483153418384038,
+      "grad_norm": 0.8838931918144226,
+      "learning_rate": 9.64007156381957e-05,
+      "loss": 0.0955,
+      "step": 29780
+    },
+    {
+      "epoch": 1.9489695780176644,
+      "grad_norm": 0.9353017210960388,
+      "learning_rate": 9.639729268205926e-05,
+      "loss": 0.0864,
+      "step": 29790
+    },
+    {
+      "epoch": 1.949623814196925,
+      "grad_norm": 0.8417748808860779,
+      "learning_rate": 9.639386815989252e-05,
+      "loss": 0.104,
+      "step": 29800
+    },
+    {
+      "epoch": 1.9502780503761858,
+      "grad_norm": 0.8374495506286621,
+      "learning_rate": 9.639044207181105e-05,
+      "loss": 0.1021,
+      "step": 29810
+    },
+    {
+      "epoch": 1.9509322865554464,
+      "grad_norm": 0.8973211646080017,
+      "learning_rate": 9.63870144179305e-05,
+      "loss": 0.1022,
+      "step": 29820
+    },
+    {
+      "epoch": 1.9515865227347073,
+      "grad_norm": 0.9601650834083557,
+      "learning_rate": 9.638358519836656e-05,
+      "loss": 0.1001,
+      "step": 29830
+    },
+    {
+      "epoch": 1.9522407589139679,
+      "grad_norm": 0.9467968344688416,
+      "learning_rate": 9.638015441323496e-05,
+      "loss": 0.0787,
+      "step": 29840
+    },
+    {
+      "epoch": 1.9528949950932286,
+      "grad_norm": 1.110438585281372,
+      "learning_rate": 9.637672206265152e-05,
+      "loss": 0.1028,
+      "step": 29850
+    },
+    {
+      "epoch": 1.9535492312724894,
+      "grad_norm": 1.1199315786361694,
+      "learning_rate": 9.63732881467321e-05,
+      "loss": 0.0869,
+      "step": 29860
+    },
+    {
+      "epoch": 1.95420346745175,
+      "grad_norm": 0.796136736869812,
+      "learning_rate": 9.636985266559258e-05,
+      "loss": 0.0975,
+      "step": 29870
+    },
+    {
+      "epoch": 1.9548577036310109,
+      "grad_norm": 0.8099967241287231,
+      "learning_rate": 9.63664156193489e-05,
+      "loss": 0.0837,
+      "step": 29880
+    },
+    {
+      "epoch": 1.9555119398102714,
+      "grad_norm": 0.7563862204551697,
+      "learning_rate": 9.636297700811712e-05,
+      "loss": 0.0844,
+      "step": 29890
+    },
+    {
+      "epoch": 1.9561661759895324,
+      "grad_norm": 1.2173089981079102,
+      "learning_rate": 9.635953683201325e-05,
+      "loss": 0.0896,
+      "step": 29900
+    },
+    {
+      "epoch": 1.9568204121687929,
+      "grad_norm": 0.7833836078643799,
+      "learning_rate": 9.635609509115344e-05,
+      "loss": 0.0898,
+      "step": 29910
+    },
+    {
+      "epoch": 1.9574746483480536,
+      "grad_norm": 0.6414048075675964,
+      "learning_rate": 9.635265178565385e-05,
+      "loss": 0.0914,
+      "step": 29920
+    },
+    {
+      "epoch": 1.9581288845273144,
+      "grad_norm": 0.7374199032783508,
+      "learning_rate": 9.63492069156307e-05,
+      "loss": 0.0874,
+      "step": 29930
+    },
+    {
+      "epoch": 1.9587831207065751,
+      "grad_norm": 1.0497411489486694,
+      "learning_rate": 9.634576048120027e-05,
+      "loss": 0.086,
+      "step": 29940
+    },
+    {
+      "epoch": 1.9594373568858359,
+      "grad_norm": 0.7760803699493408,
+      "learning_rate": 9.634231248247886e-05,
+      "loss": 0.091,
+      "step": 29950
+    },
+    {
+      "epoch": 1.9600915930650964,
+      "grad_norm": 0.7324857711791992,
+      "learning_rate": 9.633886291958287e-05,
+      "loss": 0.0913,
+      "step": 29960
+    },
+    {
+      "epoch": 1.9607458292443574,
+      "grad_norm": 0.9544631838798523,
+      "learning_rate": 9.633541179262874e-05,
+      "loss": 0.0936,
+      "step": 29970
+    },
+    {
+      "epoch": 1.9614000654236179,
+      "grad_norm": 0.8895130157470703,
+      "learning_rate": 9.633195910173294e-05,
+      "loss": 0.0883,
+      "step": 29980
+    },
+    {
+      "epoch": 1.9620543016028786,
+      "grad_norm": 0.8787074089050293,
+      "learning_rate": 9.632850484701199e-05,
+      "loss": 0.0989,
+      "step": 29990
+    },
+    {
+      "epoch": 1.9627085377821394,
+      "grad_norm": 0.9097030162811279,
+      "learning_rate": 9.632504902858253e-05,
+      "loss": 0.0864,
+      "step": 30000
+    },
+    {
+      "epoch": 1.9633627739614,
+      "grad_norm": 0.9169159531593323,
+      "learning_rate": 9.632159164656114e-05,
+      "loss": 0.101,
+      "step": 30010
+    },
+    {
+      "epoch": 1.9640170101406609,
+      "grad_norm": 0.8284802436828613,
+      "learning_rate": 9.631813270106458e-05,
+      "loss": 0.0929,
+      "step": 30020
+    },
+    {
+      "epoch": 1.9646712463199214,
+      "grad_norm": 0.69024258852005,
+      "learning_rate": 9.631467219220955e-05,
+      "loss": 0.0962,
+      "step": 30030
+    },
+    {
+      "epoch": 1.9653254824991822,
+      "grad_norm": 0.8356816172599792,
+      "learning_rate": 9.631121012011288e-05,
+      "loss": 0.0838,
+      "step": 30040
+    },
+    {
+      "epoch": 1.965979718678443,
+      "grad_norm": 0.8797650933265686,
+      "learning_rate": 9.630774648489141e-05,
+      "loss": 0.0877,
+      "step": 30050
+    },
+    {
+      "epoch": 1.9666339548577036,
+      "grad_norm": 0.9771427512168884,
+      "learning_rate": 9.630428128666204e-05,
+      "loss": 0.1115,
+      "step": 30060
+    },
+    {
+      "epoch": 1.9672881910369644,
+      "grad_norm": 0.791826605796814,
+      "learning_rate": 9.630081452554174e-05,
+      "loss": 0.0912,
+      "step": 30070
+    },
+    {
+      "epoch": 1.967942427216225,
+      "grad_norm": 0.7139357924461365,
+      "learning_rate": 9.629734620164753e-05,
+      "loss": 0.084,
+      "step": 30080
+    },
+    {
+      "epoch": 1.968596663395486,
+      "grad_norm": 0.9226008653640747,
+      "learning_rate": 9.629387631509646e-05,
+      "loss": 0.0954,
+      "step": 30090
+    },
+    {
+      "epoch": 1.9692508995747464,
+      "grad_norm": 0.9457941055297852,
+      "learning_rate": 9.629040486600567e-05,
+      "loss": 0.0924,
+      "step": 30100
+    },
+    {
+      "epoch": 1.9699051357540072,
+      "grad_norm": 0.6717526912689209,
+      "learning_rate": 9.628693185449228e-05,
+      "loss": 0.0829,
+      "step": 30110
+    },
+    {
+      "epoch": 1.970559371933268,
+      "grad_norm": 0.8272339105606079,
+      "learning_rate": 9.628345728067359e-05,
+      "loss": 0.0924,
+      "step": 30120
+    },
+    {
+      "epoch": 1.9712136081125287,
+      "grad_norm": 0.8791738748550415,
+      "learning_rate": 9.62799811446668e-05,
+      "loss": 0.0819,
+      "step": 30130
+    },
+    {
+      "epoch": 1.9718678442917894,
+      "grad_norm": 0.899770200252533,
+      "learning_rate": 9.627650344658929e-05,
+      "loss": 0.0892,
+      "step": 30140
+    },
+    {
+      "epoch": 1.97252208047105,
+      "grad_norm": 0.8326700329780579,
+      "learning_rate": 9.627302418655844e-05,
+      "loss": 0.0937,
+      "step": 30150
+    },
+    {
+      "epoch": 1.973176316650311,
+      "grad_norm": 0.9431321024894714,
+      "learning_rate": 9.626954336469166e-05,
+      "loss": 0.0902,
+      "step": 30160
+    },
+    {
+      "epoch": 1.9738305528295714,
+      "grad_norm": 0.9222663044929504,
+      "learning_rate": 9.626606098110643e-05,
+      "loss": 0.0893,
+      "step": 30170
+    },
+    {
+      "epoch": 1.9744847890088322,
+      "grad_norm": 0.7812402248382568,
+      "learning_rate": 9.62625770359203e-05,
+      "loss": 0.0853,
+      "step": 30180
+    },
+    {
+      "epoch": 1.975139025188093,
+      "grad_norm": 0.8364319801330566,
+      "learning_rate": 9.625909152925088e-05,
+      "loss": 0.0936,
+      "step": 30190
+    },
+    {
+      "epoch": 1.9757932613673534,
+      "grad_norm": 1.1527427434921265,
+      "learning_rate": 9.62556044612158e-05,
+      "loss": 0.0859,
+      "step": 30200
+    },
+    {
+      "epoch": 1.9764474975466144,
+      "grad_norm": 0.8224945068359375,
+      "learning_rate": 9.625211583193275e-05,
+      "loss": 0.0813,
+      "step": 30210
+    },
+    {
+      "epoch": 1.977101733725875,
+      "grad_norm": 0.81529700756073,
+      "learning_rate": 9.62486256415195e-05,
+      "loss": 0.0915,
+      "step": 30220
+    },
+    {
+      "epoch": 1.977755969905136,
+      "grad_norm": 0.839607834815979,
+      "learning_rate": 9.624513389009385e-05,
+      "loss": 0.09,
+      "step": 30230
+    },
+    {
+      "epoch": 1.9784102060843964,
+      "grad_norm": 0.7841370105743408,
+      "learning_rate": 9.624164057777363e-05,
+      "loss": 0.0921,
+      "step": 30240
+    },
+    {
+      "epoch": 1.9790644422636572,
+      "grad_norm": 0.8300483226776123,
+      "learning_rate": 9.623814570467678e-05,
+      "loss": 0.083,
+      "step": 30250
+    },
+    {
+      "epoch": 1.979718678442918,
+      "grad_norm": 0.7854914665222168,
+      "learning_rate": 9.623464927092123e-05,
+      "loss": 0.0886,
+      "step": 30260
+    },
+    {
+      "epoch": 1.9803729146221785,
+      "grad_norm": 0.7365754842758179,
+      "learning_rate": 9.623115127662504e-05,
+      "loss": 0.0808,
+      "step": 30270
+    },
+    {
+      "epoch": 1.9810271508014394,
+      "grad_norm": 0.8063216209411621,
+      "learning_rate": 9.622765172190624e-05,
+      "loss": 0.0885,
+      "step": 30280
+    },
+    {
+      "epoch": 1.9816813869807,
+      "grad_norm": 0.9440072774887085,
+      "learning_rate": 9.622415060688294e-05,
+      "loss": 0.0916,
+      "step": 30290
+    },
+    {
+      "epoch": 1.9823356231599607,
+      "grad_norm": 0.9911891222000122,
+      "learning_rate": 9.622064793167336e-05,
+      "loss": 0.0785,
+      "step": 30300
+    },
+    {
+      "epoch": 1.9829898593392215,
+      "grad_norm": 0.8785775303840637,
+      "learning_rate": 9.621714369639567e-05,
+      "loss": 0.0886,
+      "step": 30310
+    },
+    {
+      "epoch": 1.9836440955184822,
+      "grad_norm": 0.9227697849273682,
+      "learning_rate": 9.621363790116819e-05,
+      "loss": 0.0839,
+      "step": 30320
+    },
+    {
+      "epoch": 1.984298331697743,
+      "grad_norm": 1.0479682683944702,
+      "learning_rate": 9.621013054610922e-05,
+      "loss": 0.0892,
+      "step": 30330
+    },
+    {
+      "epoch": 1.9849525678770035,
+      "grad_norm": 0.957587718963623,
+      "learning_rate": 9.620662163133715e-05,
+      "loss": 0.0773,
+      "step": 30340
+    },
+    {
+      "epoch": 1.9856068040562644,
+      "grad_norm": 1.071298360824585,
+      "learning_rate": 9.620311115697043e-05,
+      "loss": 0.0865,
+      "step": 30350
+    },
+    {
+      "epoch": 1.986261040235525,
+      "grad_norm": 0.8113812208175659,
+      "learning_rate": 9.619959912312752e-05,
+      "loss": 0.0953,
+      "step": 30360
+    },
+    {
+      "epoch": 1.9869152764147857,
+      "grad_norm": 0.7876493334770203,
+      "learning_rate": 9.6196085529927e-05,
+      "loss": 0.0948,
+      "step": 30370
+    },
+    {
+      "epoch": 1.9875695125940465,
+      "grad_norm": 0.9819501638412476,
+      "learning_rate": 9.619257037748742e-05,
+      "loss": 0.0929,
+      "step": 30380
+    },
+    {
+      "epoch": 1.9882237487733072,
+      "grad_norm": 0.8558057546615601,
+      "learning_rate": 9.618905366592745e-05,
+      "loss": 0.0884,
+      "step": 30390
+    },
+    {
+      "epoch": 1.988877984952568,
+      "grad_norm": 0.7599206566810608,
+      "learning_rate": 9.618553539536579e-05,
+      "loss": 0.0896,
+      "step": 30400
+    },
+    {
+      "epoch": 1.9895322211318285,
+      "grad_norm": 0.8661050200462341,
+      "learning_rate": 9.618201556592117e-05,
+      "loss": 0.0754,
+      "step": 30410
+    },
+    {
+      "epoch": 1.9901864573110895,
+      "grad_norm": 0.7166074514389038,
+      "learning_rate": 9.617849417771244e-05,
+      "loss": 0.086,
+      "step": 30420
+    },
+    {
+      "epoch": 1.99084069349035,
+      "grad_norm": 0.7684576511383057,
+      "learning_rate": 9.61749712308584e-05,
+      "loss": 0.0905,
+      "step": 30430
+    },
+    {
+      "epoch": 1.9914949296696107,
+      "grad_norm": 0.8405929803848267,
+      "learning_rate": 9.617144672547798e-05,
+      "loss": 0.09,
+      "step": 30440
+    },
+    {
+      "epoch": 1.9921491658488715,
+      "grad_norm": 0.9315941333770752,
+      "learning_rate": 9.616792066169013e-05,
+      "loss": 0.091,
+      "step": 30450
+    },
+    {
+      "epoch": 1.992803402028132,
+      "grad_norm": 0.7858991622924805,
+      "learning_rate": 9.616439303961391e-05,
+      "loss": 0.095,
+      "step": 30460
+    },
+    {
+      "epoch": 1.993457638207393,
+      "grad_norm": 0.6836791038513184,
+      "learning_rate": 9.616086385936833e-05,
+      "loss": 0.0867,
+      "step": 30470
+    },
+    {
+      "epoch": 1.9941118743866535,
+      "grad_norm": 0.8010273575782776,
+      "learning_rate": 9.615733312107255e-05,
+      "loss": 0.0946,
+      "step": 30480
+    },
+    {
+      "epoch": 1.9947661105659142,
+      "grad_norm": 0.9536259770393372,
+      "learning_rate": 9.615380082484571e-05,
+      "loss": 0.0926,
+      "step": 30490
+    },
+    {
+      "epoch": 1.995420346745175,
+      "grad_norm": 0.880687952041626,
+      "learning_rate": 9.615026697080707e-05,
+      "loss": 0.1006,
+      "step": 30500
+    },
+    {
+      "epoch": 1.9960745829244357,
+      "grad_norm": 0.9787200689315796,
+      "learning_rate": 9.614673155907587e-05,
+      "loss": 0.0929,
+      "step": 30510
+    },
+    {
+      "epoch": 1.9967288191036965,
+      "grad_norm": 0.906243085861206,
+      "learning_rate": 9.614319458977145e-05,
+      "loss": 0.0931,
+      "step": 30520
+    },
+    {
+      "epoch": 1.997383055282957,
+      "grad_norm": 0.7984228730201721,
+      "learning_rate": 9.613965606301321e-05,
+      "loss": 0.0822,
+      "step": 30530
+    },
+    {
+      "epoch": 1.998037291462218,
+      "grad_norm": 0.8581556081771851,
+      "learning_rate": 9.613611597892059e-05,
+      "loss": 0.0877,
+      "step": 30540
+    },
+    {
+      "epoch": 1.9986915276414785,
+      "grad_norm": 0.724999725818634,
+      "learning_rate": 9.613257433761303e-05,
+      "loss": 0.0897,
+      "step": 30550
+    },
+    {
+      "epoch": 1.9993457638207393,
+      "grad_norm": 0.6740709543228149,
+      "learning_rate": 9.612903113921011e-05,
+      "loss": 0.0895,
+      "step": 30560
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.8931834101676941,
+      "learning_rate": 9.612548638383141e-05,
+      "loss": 0.0857,
+      "step": 30570
+    },
+    {
+      "epoch": 2.0006542361792605,
+      "grad_norm": 0.8923904299736023,
+      "learning_rate": 9.612194007159657e-05,
+      "loss": 0.0848,
+      "step": 30580
+    },
+    {
+      "epoch": 2.0013084723585215,
+      "grad_norm": 0.8426147699356079,
+      "learning_rate": 9.61183922026253e-05,
+      "loss": 0.0931,
+      "step": 30590
+    },
+    {
+      "epoch": 2.001962708537782,
+      "grad_norm": 0.7947473526000977,
+      "learning_rate": 9.611484277703733e-05,
+      "loss": 0.0989,
+      "step": 30600
+    },
+    {
+      "epoch": 2.002616944717043,
+      "grad_norm": 0.8306055665016174,
+      "learning_rate": 9.61112917949525e-05,
+      "loss": 0.0867,
+      "step": 30610
+    },
+    {
+      "epoch": 2.0032711808963035,
+      "grad_norm": 0.9573779702186584,
+      "learning_rate": 9.610773925649062e-05,
+      "loss": 0.0943,
+      "step": 30620
+    },
+    {
+      "epoch": 2.0039254170755645,
+      "grad_norm": 0.8894043564796448,
+      "learning_rate": 9.610418516177164e-05,
+      "loss": 0.0891,
+      "step": 30630
+    },
+    {
+      "epoch": 2.004579653254825,
+      "grad_norm": 0.8106326460838318,
+      "learning_rate": 9.610062951091547e-05,
+      "loss": 0.0892,
+      "step": 30640
+    },
+    {
+      "epoch": 2.0052338894340855,
+      "grad_norm": 0.9291239976882935,
+      "learning_rate": 9.609707230404217e-05,
+      "loss": 0.0835,
+      "step": 30650
+    },
+    {
+      "epoch": 2.0058881256133465,
+      "grad_norm": 0.7362892627716064,
+      "learning_rate": 9.609351354127178e-05,
+      "loss": 0.0838,
+      "step": 30660
+    },
+    {
+      "epoch": 2.006542361792607,
+      "grad_norm": 0.7407713532447815,
+      "learning_rate": 9.608995322272442e-05,
+      "loss": 0.0951,
+      "step": 30670
+    },
+    {
+      "epoch": 2.007196597971868,
+      "grad_norm": 0.9148522019386292,
+      "learning_rate": 9.608639134852028e-05,
+      "loss": 0.0895,
+      "step": 30680
+    },
+    {
+      "epoch": 2.0078508341511285,
+      "grad_norm": 0.7129114866256714,
+      "learning_rate": 9.608282791877955e-05,
+      "loss": 0.0866,
+      "step": 30690
+    },
+    {
+      "epoch": 2.008505070330389,
+      "grad_norm": 1.1984944343566895,
+      "learning_rate": 9.607926293362253e-05,
+      "loss": 0.0941,
+      "step": 30700
+    },
+    {
+      "epoch": 2.00915930650965,
+      "grad_norm": 0.7906424403190613,
+      "learning_rate": 9.607569639316953e-05,
+      "loss": 0.0811,
+      "step": 30710
+    },
+    {
+      "epoch": 2.0098135426889105,
+      "grad_norm": 0.9424049258232117,
+      "learning_rate": 9.607212829754094e-05,
+      "loss": 0.0872,
+      "step": 30720
+    },
+    {
+      "epoch": 2.0104677788681715,
+      "grad_norm": 0.9077602028846741,
+      "learning_rate": 9.60685586468572e-05,
+      "loss": 0.0946,
+      "step": 30730
+    },
+    {
+      "epoch": 2.011122015047432,
+      "grad_norm": 0.867131769657135,
+      "learning_rate": 9.606498744123877e-05,
+      "loss": 0.0903,
+      "step": 30740
+    },
+    {
+      "epoch": 2.011776251226693,
+      "grad_norm": 0.7888831496238708,
+      "learning_rate": 9.606141468080623e-05,
+      "loss": 0.0829,
+      "step": 30750
+    },
+    {
+      "epoch": 2.0124304874059535,
+      "grad_norm": 1.2040573358535767,
+      "learning_rate": 9.605784036568011e-05,
+      "loss": 0.1066,
+      "step": 30760
+    },
+    {
+      "epoch": 2.013084723585214,
+      "grad_norm": 0.864673376083374,
+      "learning_rate": 9.605426449598112e-05,
+      "loss": 0.0929,
+      "step": 30770
+    },
+    {
+      "epoch": 2.013738959764475,
+      "grad_norm": 0.7464171648025513,
+      "learning_rate": 9.60506870718299e-05,
+      "loss": 0.0877,
+      "step": 30780
+    },
+    {
+      "epoch": 2.0143931959437356,
+      "grad_norm": 0.8971266150474548,
+      "learning_rate": 9.604710809334723e-05,
+      "loss": 0.096,
+      "step": 30790
+    },
+    {
+      "epoch": 2.0150474321229965,
+      "grad_norm": 0.9102427959442139,
+      "learning_rate": 9.60435275606539e-05,
+      "loss": 0.0975,
+      "step": 30800
+    },
+    {
+      "epoch": 2.015701668302257,
+      "grad_norm": 0.8453085422515869,
+      "learning_rate": 9.603994547387074e-05,
+      "loss": 0.0989,
+      "step": 30810
+    },
+    {
+      "epoch": 2.016355904481518,
+      "grad_norm": 0.9234716892242432,
+      "learning_rate": 9.603636183311871e-05,
+      "loss": 0.089,
+      "step": 30820
+    },
+    {
+      "epoch": 2.0170101406607785,
+      "grad_norm": 0.8139061331748962,
+      "learning_rate": 9.60327766385187e-05,
+      "loss": 0.0854,
+      "step": 30830
+    },
+    {
+      "epoch": 2.017664376840039,
+      "grad_norm": 0.8283985257148743,
+      "learning_rate": 9.602918989019176e-05,
+      "loss": 0.0881,
+      "step": 30840
+    },
+    {
+      "epoch": 2.0183186130193,
+      "grad_norm": 0.780783474445343,
+      "learning_rate": 9.602560158825896e-05,
+      "loss": 0.0888,
+      "step": 30850
+    },
+    {
+      "epoch": 2.0189728491985606,
+      "grad_norm": 0.6656304001808167,
+      "learning_rate": 9.602201173284139e-05,
+      "loss": 0.0792,
+      "step": 30860
+    },
+    {
+      "epoch": 2.0196270853778215,
+      "grad_norm": 0.7495284080505371,
+      "learning_rate": 9.601842032406023e-05,
+      "loss": 0.0935,
+      "step": 30870
+    },
+    {
+      "epoch": 2.020281321557082,
+      "grad_norm": 0.730228066444397,
+      "learning_rate": 9.60148273620367e-05,
+      "loss": 0.083,
+      "step": 30880
+    },
+    {
+      "epoch": 2.020935557736343,
+      "grad_norm": 0.8227654099464417,
+      "learning_rate": 9.601123284689206e-05,
+      "loss": 0.0904,
+      "step": 30890
+    },
+    {
+      "epoch": 2.0215897939156036,
+      "grad_norm": 0.8423411250114441,
+      "learning_rate": 9.600763677874764e-05,
+      "loss": 0.0854,
+      "step": 30900
+    },
+    {
+      "epoch": 2.022244030094864,
+      "grad_norm": 0.9482555985450745,
+      "learning_rate": 9.600403915772484e-05,
+      "loss": 0.0973,
+      "step": 30910
+    },
+    {
+      "epoch": 2.022898266274125,
+      "grad_norm": 0.7677321434020996,
+      "learning_rate": 9.600043998394506e-05,
+      "loss": 0.0847,
+      "step": 30920
+    },
+    {
+      "epoch": 2.0235525024533856,
+      "grad_norm": 0.7692193984985352,
+      "learning_rate": 9.599683925752979e-05,
+      "loss": 0.1019,
+      "step": 30930
+    },
+    {
+      "epoch": 2.0242067386326466,
+      "grad_norm": 0.870507001876831,
+      "learning_rate": 9.599323697860055e-05,
+      "loss": 0.0897,
+      "step": 30940
+    },
+    {
+      "epoch": 2.024860974811907,
+      "grad_norm": 0.933172345161438,
+      "learning_rate": 9.598963314727894e-05,
+      "loss": 0.0922,
+      "step": 30950
+    },
+    {
+      "epoch": 2.0255152109911676,
+      "grad_norm": 0.8484349250793457,
+      "learning_rate": 9.598602776368661e-05,
+      "loss": 0.0927,
+      "step": 30960
+    },
+    {
+      "epoch": 2.0261694471704286,
+      "grad_norm": 0.7965954542160034,
+      "learning_rate": 9.598242082794524e-05,
+      "loss": 0.0983,
+      "step": 30970
+    },
+    {
+      "epoch": 2.026823683349689,
+      "grad_norm": 0.7531421184539795,
+      "learning_rate": 9.597881234017657e-05,
+      "loss": 0.0841,
+      "step": 30980
+    },
+    {
+      "epoch": 2.02747791952895,
+      "grad_norm": 0.9377996325492859,
+      "learning_rate": 9.597520230050242e-05,
+      "loss": 0.1034,
+      "step": 30990
+    },
+    {
+      "epoch": 2.0281321557082106,
+      "grad_norm": 0.8710869550704956,
+      "learning_rate": 9.597159070904458e-05,
+      "loss": 0.0923,
+      "step": 31000
+    },
+    {
+      "epoch": 2.0287863918874716,
+      "grad_norm": 0.9062069058418274,
+      "learning_rate": 9.596797756592502e-05,
+      "loss": 0.0931,
+      "step": 31010
+    },
+    {
+      "epoch": 2.029440628066732,
+      "grad_norm": 1.0289934873580933,
+      "learning_rate": 9.596436287126565e-05,
+      "loss": 0.0977,
+      "step": 31020
+    },
+    {
+      "epoch": 2.0300948642459926,
+      "grad_norm": 0.9156734943389893,
+      "learning_rate": 9.596074662518848e-05,
+      "loss": 0.0874,
+      "step": 31030
+    },
+    {
+      "epoch": 2.0307491004252536,
+      "grad_norm": 0.9219240546226501,
+      "learning_rate": 9.595712882781558e-05,
+      "loss": 0.099,
+      "step": 31040
+    },
+    {
+      "epoch": 2.031403336604514,
+      "grad_norm": 0.9066693186759949,
+      "learning_rate": 9.595350947926907e-05,
+      "loss": 0.09,
+      "step": 31050
+    },
+    {
+      "epoch": 2.032057572783775,
+      "grad_norm": 0.8575799465179443,
+      "learning_rate": 9.594988857967106e-05,
+      "loss": 0.0895,
+      "step": 31060
+    },
+    {
+      "epoch": 2.0327118089630356,
+      "grad_norm": 0.977655291557312,
+      "learning_rate": 9.594626612914383e-05,
+      "loss": 0.0949,
+      "step": 31070
+    },
+    {
+      "epoch": 2.0333660451422966,
+      "grad_norm": 0.8753423094749451,
+      "learning_rate": 9.594264212780962e-05,
+      "loss": 0.0801,
+      "step": 31080
+    },
+    {
+      "epoch": 2.034020281321557,
+      "grad_norm": 0.7495356798171997,
+      "learning_rate": 9.593901657579075e-05,
+      "loss": 0.096,
+      "step": 31090
+    },
+    {
+      "epoch": 2.0346745175008176,
+      "grad_norm": 0.8719765543937683,
+      "learning_rate": 9.593538947320959e-05,
+      "loss": 0.0975,
+      "step": 31100
+    },
+    {
+      "epoch": 2.0353287536800786,
+      "grad_norm": 0.7684383988380432,
+      "learning_rate": 9.593176082018855e-05,
+      "loss": 0.0857,
+      "step": 31110
+    },
+    {
+      "epoch": 2.035982989859339,
+      "grad_norm": 1.1163846254348755,
+      "learning_rate": 9.592813061685015e-05,
+      "loss": 0.1057,
+      "step": 31120
+    },
+    {
+      "epoch": 2.0366372260386,
+      "grad_norm": 0.8532420992851257,
+      "learning_rate": 9.592449886331687e-05,
+      "loss": 0.0969,
+      "step": 31130
+    },
+    {
+      "epoch": 2.0372914622178606,
+      "grad_norm": 0.8979114890098572,
+      "learning_rate": 9.592086555971131e-05,
+      "loss": 0.0875,
+      "step": 31140
+    },
+    {
+      "epoch": 2.037945698397121,
+      "grad_norm": 0.8417396545410156,
+      "learning_rate": 9.591723070615612e-05,
+      "loss": 0.0906,
+      "step": 31150
+    },
+    {
+      "epoch": 2.038599934576382,
+      "grad_norm": 1.090818166732788,
+      "learning_rate": 9.591359430277396e-05,
+      "loss": 0.0861,
+      "step": 31160
+    },
+    {
+      "epoch": 2.0392541707556426,
+      "grad_norm": 0.6975811123847961,
+      "learning_rate": 9.590995634968759e-05,
+      "loss": 0.0902,
+      "step": 31170
+    },
+    {
+      "epoch": 2.0399084069349036,
+      "grad_norm": 0.996051549911499,
+      "learning_rate": 9.590631684701979e-05,
+      "loss": 0.0854,
+      "step": 31180
+    },
+    {
+      "epoch": 2.040562643114164,
+      "grad_norm": 0.9619652628898621,
+      "learning_rate": 9.590267579489338e-05,
+      "loss": 0.0975,
+      "step": 31190
+    },
+    {
+      "epoch": 2.041216879293425,
+      "grad_norm": 0.8321636319160461,
+      "learning_rate": 9.589903319343129e-05,
+      "loss": 0.0883,
+      "step": 31200
+    },
+    {
+      "epoch": 2.0418711154726856,
+      "grad_norm": 0.8303142189979553,
+      "learning_rate": 9.589538904275645e-05,
+      "loss": 0.086,
+      "step": 31210
+    },
+    {
+      "epoch": 2.042525351651946,
+      "grad_norm": 0.7571492195129395,
+      "learning_rate": 9.589174334299189e-05,
+      "loss": 0.0857,
+      "step": 31220
+    },
+    {
+      "epoch": 2.043179587831207,
+      "grad_norm": 0.6824153065681458,
+      "learning_rate": 9.588809609426061e-05,
+      "loss": 0.0876,
+      "step": 31230
+    },
+    {
+      "epoch": 2.0438338240104676,
+      "grad_norm": 0.907193660736084,
+      "learning_rate": 9.588444729668575e-05,
+      "loss": 0.0847,
+      "step": 31240
+    },
+    {
+      "epoch": 2.0444880601897286,
+      "grad_norm": 0.9690275192260742,
+      "learning_rate": 9.588079695039047e-05,
+      "loss": 0.096,
+      "step": 31250
+    },
+    {
+      "epoch": 2.045142296368989,
+      "grad_norm": 0.8058456182479858,
+      "learning_rate": 9.587714505549796e-05,
+      "loss": 0.0905,
+      "step": 31260
+    },
+    {
+      "epoch": 2.04579653254825,
+      "grad_norm": 0.9792216420173645,
+      "learning_rate": 9.587349161213148e-05,
+      "loss": 0.0956,
+      "step": 31270
+    },
+    {
+      "epoch": 2.0464507687275106,
+      "grad_norm": 0.8495068550109863,
+      "learning_rate": 9.586983662041434e-05,
+      "loss": 0.0937,
+      "step": 31280
+    },
+    {
+      "epoch": 2.047105004906771,
+      "grad_norm": 0.9974722862243652,
+      "learning_rate": 9.586618008046992e-05,
+      "loss": 0.086,
+      "step": 31290
+    },
+    {
+      "epoch": 2.047759241086032,
+      "grad_norm": 0.8604632616043091,
+      "learning_rate": 9.586252199242166e-05,
+      "loss": 0.0877,
+      "step": 31300
+    },
+    {
+      "epoch": 2.0484134772652927,
+      "grad_norm": 0.803986668586731,
+      "learning_rate": 9.585886235639299e-05,
+      "loss": 0.0971,
+      "step": 31310
+    },
+    {
+      "epoch": 2.0490677134445536,
+      "grad_norm": 0.6995072960853577,
+      "learning_rate": 9.585520117250744e-05,
+      "loss": 0.0965,
+      "step": 31320
+    },
+    {
+      "epoch": 2.049721949623814,
+      "grad_norm": 0.9633769989013672,
+      "learning_rate": 9.585153844088858e-05,
+      "loss": 0.0906,
+      "step": 31330
+    },
+    {
+      "epoch": 2.050376185803075,
+      "grad_norm": 0.7706478238105774,
+      "learning_rate": 9.584787416166006e-05,
+      "loss": 0.0924,
+      "step": 31340
+    },
+    {
+      "epoch": 2.0510304219823356,
+      "grad_norm": 0.7993813753128052,
+      "learning_rate": 9.584420833494555e-05,
+      "loss": 0.0945,
+      "step": 31350
+    },
+    {
+      "epoch": 2.051684658161596,
+      "grad_norm": 0.7742961645126343,
+      "learning_rate": 9.584054096086877e-05,
+      "loss": 0.0973,
+      "step": 31360
+    },
+    {
+      "epoch": 2.052338894340857,
+      "grad_norm": 0.7687771320343018,
+      "learning_rate": 9.583687203955352e-05,
+      "loss": 0.0907,
+      "step": 31370
+    },
+    {
+      "epoch": 2.0529931305201177,
+      "grad_norm": 0.7839317917823792,
+      "learning_rate": 9.583320157112362e-05,
+      "loss": 0.0888,
+      "step": 31380
+    },
+    {
+      "epoch": 2.0536473666993786,
+      "grad_norm": 0.891391396522522,
+      "learning_rate": 9.582952955570297e-05,
+      "loss": 0.0929,
+      "step": 31390
+    },
+    {
+      "epoch": 2.054301602878639,
+      "grad_norm": 0.699019730091095,
+      "learning_rate": 9.58258559934155e-05,
+      "loss": 0.0879,
+      "step": 31400
+    },
+    {
+      "epoch": 2.0549558390578997,
+      "grad_norm": 0.765687108039856,
+      "learning_rate": 9.582218088438522e-05,
+      "loss": 0.0901,
+      "step": 31410
+    },
+    {
+      "epoch": 2.0556100752371607,
+      "grad_norm": 0.7088882327079773,
+      "learning_rate": 9.581850422873615e-05,
+      "loss": 0.0878,
+      "step": 31420
+    },
+    {
+      "epoch": 2.056264311416421,
+      "grad_norm": 0.9518930315971375,
+      "learning_rate": 9.58148260265924e-05,
+      "loss": 0.0831,
+      "step": 31430
+    },
+    {
+      "epoch": 2.056918547595682,
+      "grad_norm": 0.6729696393013,
+      "learning_rate": 9.581114627807812e-05,
+      "loss": 0.0931,
+      "step": 31440
+    },
+    {
+      "epoch": 2.0575727837749427,
+      "grad_norm": 0.8468174338340759,
+      "learning_rate": 9.58074649833175e-05,
+      "loss": 0.0958,
+      "step": 31450
+    },
+    {
+      "epoch": 2.0582270199542037,
+      "grad_norm": 0.8482375144958496,
+      "learning_rate": 9.580378214243482e-05,
+      "loss": 0.09,
+      "step": 31460
+    },
+    {
+      "epoch": 2.058881256133464,
+      "grad_norm": 0.7801223397254944,
+      "learning_rate": 9.580009775555435e-05,
+      "loss": 0.0862,
+      "step": 31470
+    },
+    {
+      "epoch": 2.0595354923127247,
+      "grad_norm": 0.7654222249984741,
+      "learning_rate": 9.579641182280049e-05,
+      "loss": 0.0897,
+      "step": 31480
+    },
+    {
+      "epoch": 2.0601897284919857,
+      "grad_norm": 0.9105095267295837,
+      "learning_rate": 9.57927243442976e-05,
+      "loss": 0.0818,
+      "step": 31490
+    },
+    {
+      "epoch": 2.060843964671246,
+      "grad_norm": 0.8952615261077881,
+      "learning_rate": 9.578903532017017e-05,
+      "loss": 0.0898,
+      "step": 31500
+    },
+    {
+      "epoch": 2.061498200850507,
+      "grad_norm": 0.9962287545204163,
+      "learning_rate": 9.578534475054272e-05,
+      "loss": 0.0912,
+      "step": 31510
+    },
+    {
+      "epoch": 2.0621524370297677,
+      "grad_norm": 0.9723460674285889,
+      "learning_rate": 9.578165263553982e-05,
+      "loss": 0.0816,
+      "step": 31520
+    },
+    {
+      "epoch": 2.0628066732090287,
+      "grad_norm": 0.8590718507766724,
+      "learning_rate": 9.577795897528605e-05,
+      "loss": 0.0866,
+      "step": 31530
+    },
+    {
+      "epoch": 2.063460909388289,
+      "grad_norm": 0.8612424731254578,
+      "learning_rate": 9.577426376990613e-05,
+      "loss": 0.0951,
+      "step": 31540
+    },
+    {
+      "epoch": 2.0641151455675497,
+      "grad_norm": 0.8139830231666565,
+      "learning_rate": 9.577056701952474e-05,
+      "loss": 0.0836,
+      "step": 31550
+    },
+    {
+      "epoch": 2.0647693817468107,
+      "grad_norm": 0.7661455869674683,
+      "learning_rate": 9.57668687242667e-05,
+      "loss": 0.0822,
+      "step": 31560
+    },
+    {
+      "epoch": 2.065423617926071,
+      "grad_norm": 0.7846705317497253,
+      "learning_rate": 9.576316888425681e-05,
+      "loss": 0.089,
+      "step": 31570
+    },
+    {
+      "epoch": 2.066077854105332,
+      "grad_norm": 0.9100658893585205,
+      "learning_rate": 9.575946749961992e-05,
+      "loss": 0.078,
+      "step": 31580
+    },
+    {
+      "epoch": 2.0667320902845927,
+      "grad_norm": 0.8876975178718567,
+      "learning_rate": 9.575576457048102e-05,
+      "loss": 0.0923,
+      "step": 31590
+    },
+    {
+      "epoch": 2.0673863264638532,
+      "grad_norm": 0.9435741901397705,
+      "learning_rate": 9.575206009696507e-05,
+      "loss": 0.0967,
+      "step": 31600
+    },
+    {
+      "epoch": 2.068040562643114,
+      "grad_norm": 0.7191019058227539,
+      "learning_rate": 9.574835407919709e-05,
+      "loss": 0.0965,
+      "step": 31610
+    },
+    {
+      "epoch": 2.0686947988223747,
+      "grad_norm": 0.7950155735015869,
+      "learning_rate": 9.574464651730219e-05,
+      "loss": 0.0873,
+      "step": 31620
+    },
+    {
+      "epoch": 2.0693490350016357,
+      "grad_norm": 0.7821862697601318,
+      "learning_rate": 9.574093741140549e-05,
+      "loss": 0.0901,
+      "step": 31630
+    },
+    {
+      "epoch": 2.070003271180896,
+      "grad_norm": 0.7659751176834106,
+      "learning_rate": 9.57372267616322e-05,
+      "loss": 0.0866,
+      "step": 31640
+    },
+    {
+      "epoch": 2.070657507360157,
+      "grad_norm": 0.8592097759246826,
+      "learning_rate": 9.573351456810755e-05,
+      "loss": 0.0871,
+      "step": 31650
+    },
+    {
+      "epoch": 2.0713117435394177,
+      "grad_norm": 0.9904183149337769,
+      "learning_rate": 9.572980083095684e-05,
+      "loss": 0.0931,
+      "step": 31660
+    },
+    {
+      "epoch": 2.0719659797186782,
+      "grad_norm": 0.8140203952789307,
+      "learning_rate": 9.572608555030543e-05,
+      "loss": 0.0921,
+      "step": 31670
+    },
+    {
+      "epoch": 2.072620215897939,
+      "grad_norm": 1.019144058227539,
+      "learning_rate": 9.57223687262787e-05,
+      "loss": 0.0873,
+      "step": 31680
+    },
+    {
+      "epoch": 2.0732744520771997,
+      "grad_norm": 0.8662041425704956,
+      "learning_rate": 9.571865035900213e-05,
+      "loss": 0.0883,
+      "step": 31690
+    },
+    {
+      "epoch": 2.0739286882564607,
+      "grad_norm": 0.7000994682312012,
+      "learning_rate": 9.571493044860121e-05,
+      "loss": 0.0819,
+      "step": 31700
+    },
+    {
+      "epoch": 2.0745829244357212,
+      "grad_norm": 0.9109962582588196,
+      "learning_rate": 9.571120899520148e-05,
+      "loss": 0.0957,
+      "step": 31710
+    },
+    {
+      "epoch": 2.075237160614982,
+      "grad_norm": 0.7647160887718201,
+      "learning_rate": 9.570748599892858e-05,
+      "loss": 0.0833,
+      "step": 31720
+    },
+    {
+      "epoch": 2.0758913967942427,
+      "grad_norm": 0.8084906339645386,
+      "learning_rate": 9.570376145990814e-05,
+      "loss": 0.086,
+      "step": 31730
+    },
+    {
+      "epoch": 2.0765456329735033,
+      "grad_norm": 0.9867994785308838,
+      "learning_rate": 9.57000353782659e-05,
+      "loss": 0.0799,
+      "step": 31740
+    },
+    {
+      "epoch": 2.0771998691527642,
+      "grad_norm": 0.9291912317276001,
+      "learning_rate": 9.569630775412762e-05,
+      "loss": 0.0904,
+      "step": 31750
+    },
+    {
+      "epoch": 2.0778541053320247,
+      "grad_norm": 0.9430752992630005,
+      "learning_rate": 9.569257858761909e-05,
+      "loss": 0.0868,
+      "step": 31760
+    },
+    {
+      "epoch": 2.0785083415112857,
+      "grad_norm": 0.7106003165245056,
+      "learning_rate": 9.568884787886621e-05,
+      "loss": 0.0909,
+      "step": 31770
+    },
+    {
+      "epoch": 2.0791625776905462,
+      "grad_norm": 1.0632734298706055,
+      "learning_rate": 9.56851156279949e-05,
+      "loss": 0.0883,
+      "step": 31780
+    },
+    {
+      "epoch": 2.079816813869807,
+      "grad_norm": 0.8719366192817688,
+      "learning_rate": 9.568138183513111e-05,
+      "loss": 0.0842,
+      "step": 31790
+    },
+    {
+      "epoch": 2.0804710500490677,
+      "grad_norm": 0.806563675403595,
+      "learning_rate": 9.567764650040087e-05,
+      "loss": 0.0993,
+      "step": 31800
+    },
+    {
+      "epoch": 2.0811252862283283,
+      "grad_norm": 0.8879052996635437,
+      "learning_rate": 9.567390962393029e-05,
+      "loss": 0.0953,
+      "step": 31810
+    },
+    {
+      "epoch": 2.0817795224075892,
+      "grad_norm": 0.8709869384765625,
+      "learning_rate": 9.567017120584545e-05,
+      "loss": 0.087,
+      "step": 31820
+    },
+    {
+      "epoch": 2.0824337585868498,
+      "grad_norm": 0.7416978478431702,
+      "learning_rate": 9.566643124627258e-05,
+      "loss": 0.0861,
+      "step": 31830
+    },
+    {
+      "epoch": 2.0830879947661107,
+      "grad_norm": 0.8955287337303162,
+      "learning_rate": 9.566268974533789e-05,
+      "loss": 0.0933,
+      "step": 31840
+    },
+    {
+      "epoch": 2.0837422309453713,
+      "grad_norm": 0.8596556186676025,
+      "learning_rate": 9.565894670316767e-05,
+      "loss": 0.0911,
+      "step": 31850
+    },
+    {
+      "epoch": 2.084396467124632,
+      "grad_norm": 0.8234604597091675,
+      "learning_rate": 9.565520211988823e-05,
+      "loss": 0.0864,
+      "step": 31860
+    },
+    {
+      "epoch": 2.0850507033038927,
+      "grad_norm": 0.862963855266571,
+      "learning_rate": 9.5651455995626e-05,
+      "loss": 0.0942,
+      "step": 31870
+    },
+    {
+      "epoch": 2.0857049394831533,
+      "grad_norm": 0.8025641441345215,
+      "learning_rate": 9.56477083305074e-05,
+      "loss": 0.0871,
+      "step": 31880
+    },
+    {
+      "epoch": 2.0863591756624142,
+      "grad_norm": 0.7878461480140686,
+      "learning_rate": 9.564395912465893e-05,
+      "loss": 0.0889,
+      "step": 31890
+    },
+    {
+      "epoch": 2.0870134118416748,
+      "grad_norm": 0.8542391061782837,
+      "learning_rate": 9.564020837820713e-05,
+      "loss": 0.0936,
+      "step": 31900
+    },
+    {
+      "epoch": 2.0876676480209357,
+      "grad_norm": 0.7547128796577454,
+      "learning_rate": 9.56364560912786e-05,
+      "loss": 0.0922,
+      "step": 31910
+    },
+    {
+      "epoch": 2.0883218842001963,
+      "grad_norm": 0.7220647931098938,
+      "learning_rate": 9.563270226400001e-05,
+      "loss": 0.0832,
+      "step": 31920
+    },
+    {
+      "epoch": 2.088976120379457,
+      "grad_norm": 0.9882017374038696,
+      "learning_rate": 9.562894689649802e-05,
+      "loss": 0.0984,
+      "step": 31930
+    },
+    {
+      "epoch": 2.0896303565587178,
+      "grad_norm": 0.949703574180603,
+      "learning_rate": 9.562518998889942e-05,
+      "loss": 0.0903,
+      "step": 31940
+    },
+    {
+      "epoch": 2.0902845927379783,
+      "grad_norm": 1.0962672233581543,
+      "learning_rate": 9.562143154133099e-05,
+      "loss": 0.0848,
+      "step": 31950
+    },
+    {
+      "epoch": 2.0909388289172393,
+      "grad_norm": 0.9452337622642517,
+      "learning_rate": 9.561767155391961e-05,
+      "loss": 0.0901,
+      "step": 31960
+    },
+    {
+      "epoch": 2.0915930650965,
+      "grad_norm": 0.9834058880805969,
+      "learning_rate": 9.561391002679217e-05,
+      "loss": 0.0849,
+      "step": 31970
+    },
+    {
+      "epoch": 2.0922473012757608,
+      "grad_norm": 0.8670421838760376,
+      "learning_rate": 9.561014696007565e-05,
+      "loss": 0.0926,
+      "step": 31980
+    },
+    {
+      "epoch": 2.0929015374550213,
+      "grad_norm": 0.8581560254096985,
+      "learning_rate": 9.560638235389704e-05,
+      "loss": 0.0923,
+      "step": 31990
+    },
+    {
+      "epoch": 2.093555773634282,
+      "grad_norm": 0.9156831502914429,
+      "learning_rate": 9.560261620838342e-05,
+      "loss": 0.0951,
+      "step": 32000
+    },
+    {
+      "epoch": 2.0942100098135428,
+      "grad_norm": 0.6964516639709473,
+      "learning_rate": 9.559884852366191e-05,
+      "loss": 0.0958,
+      "step": 32010
+    },
+    {
+      "epoch": 2.0948642459928033,
+      "grad_norm": 0.859331488609314,
+      "learning_rate": 9.559507929985968e-05,
+      "loss": 0.0784,
+      "step": 32020
+    },
+    {
+      "epoch": 2.0955184821720643,
+      "grad_norm": 0.8123635649681091,
+      "learning_rate": 9.559130853710395e-05,
+      "loss": 0.0899,
+      "step": 32030
+    },
+    {
+      "epoch": 2.096172718351325,
+      "grad_norm": 0.7182052731513977,
+      "learning_rate": 9.558753623552197e-05,
+      "loss": 0.0874,
+      "step": 32040
+    },
+    {
+      "epoch": 2.0968269545305853,
+      "grad_norm": 0.8559288382530212,
+      "learning_rate": 9.558376239524109e-05,
+      "loss": 0.0788,
+      "step": 32050
+    },
+    {
+      "epoch": 2.0974811907098463,
+      "grad_norm": 0.8461546301841736,
+      "learning_rate": 9.557998701638868e-05,
+      "loss": 0.0858,
+      "step": 32060
+    },
+    {
+      "epoch": 2.098135426889107,
+      "grad_norm": 0.8831964135169983,
+      "learning_rate": 9.557621009909218e-05,
+      "loss": 0.1007,
+      "step": 32070
+    },
+    {
+      "epoch": 2.098789663068368,
+      "grad_norm": 0.8673757910728455,
+      "learning_rate": 9.557243164347907e-05,
+      "loss": 0.0876,
+      "step": 32080
+    },
+    {
+      "epoch": 2.0994438992476283,
+      "grad_norm": 0.9492565393447876,
+      "learning_rate": 9.556865164967685e-05,
+      "loss": 0.0919,
+      "step": 32090
+    },
+    {
+      "epoch": 2.1000981354268893,
+      "grad_norm": 0.8999154567718506,
+      "learning_rate": 9.556487011781314e-05,
+      "loss": 0.087,
+      "step": 32100
+    },
+    {
+      "epoch": 2.10075237160615,
+      "grad_norm": 0.7991196513175964,
+      "learning_rate": 9.556108704801558e-05,
+      "loss": 0.0878,
+      "step": 32110
+    },
+    {
+      "epoch": 2.1014066077854103,
+      "grad_norm": 0.8117789030075073,
+      "learning_rate": 9.555730244041182e-05,
+      "loss": 0.0961,
+      "step": 32120
+    },
+    {
+      "epoch": 2.1020608439646713,
+      "grad_norm": 0.8746902346611023,
+      "learning_rate": 9.555351629512963e-05,
+      "loss": 0.0885,
+      "step": 32130
+    },
+    {
+      "epoch": 2.102715080143932,
+      "grad_norm": 1.0785053968429565,
+      "learning_rate": 9.55497286122968e-05,
+      "loss": 0.0928,
+      "step": 32140
+    },
+    {
+      "epoch": 2.103369316323193,
+      "grad_norm": 0.9100284576416016,
+      "learning_rate": 9.554593939204117e-05,
+      "loss": 0.0856,
+      "step": 32150
+    },
+    {
+      "epoch": 2.1040235525024533,
+      "grad_norm": 0.7973883748054504,
+      "learning_rate": 9.554214863449065e-05,
+      "loss": 0.0826,
+      "step": 32160
+    },
+    {
+      "epoch": 2.1046777886817143,
+      "grad_norm": 0.9235144257545471,
+      "learning_rate": 9.553835633977316e-05,
+      "loss": 0.0876,
+      "step": 32170
+    },
+    {
+      "epoch": 2.105332024860975,
+      "grad_norm": 0.7327439785003662,
+      "learning_rate": 9.553456250801671e-05,
+      "loss": 0.0962,
+      "step": 32180
+    },
+    {
+      "epoch": 2.1059862610402353,
+      "grad_norm": 0.7771127223968506,
+      "learning_rate": 9.553076713934936e-05,
+      "loss": 0.0884,
+      "step": 32190
+    },
+    {
+      "epoch": 2.1066404972194963,
+      "grad_norm": 0.8690281510353088,
+      "learning_rate": 9.552697023389922e-05,
+      "loss": 0.0966,
+      "step": 32200
+    },
+    {
+      "epoch": 2.107294733398757,
+      "grad_norm": 0.9444182515144348,
+      "learning_rate": 9.552317179179444e-05,
+      "loss": 0.1007,
+      "step": 32210
+    },
+    {
+      "epoch": 2.107948969578018,
+      "grad_norm": 0.7808245420455933,
+      "learning_rate": 9.551937181316322e-05,
+      "loss": 0.0893,
+      "step": 32220
+    },
+    {
+      "epoch": 2.1086032057572783,
+      "grad_norm": 0.830314576625824,
+      "learning_rate": 9.55155702981338e-05,
+      "loss": 0.104,
+      "step": 32230
+    },
+    {
+      "epoch": 2.1092574419365393,
+      "grad_norm": 0.7720317244529724,
+      "learning_rate": 9.551176724683453e-05,
+      "loss": 0.0856,
+      "step": 32240
+    },
+    {
+      "epoch": 2.1099116781158,
+      "grad_norm": 0.8897857666015625,
+      "learning_rate": 9.550796265939377e-05,
+      "loss": 0.0877,
+      "step": 32250
+    },
+    {
+      "epoch": 2.1105659142950604,
+      "grad_norm": 0.8322727084159851,
+      "learning_rate": 9.550415653593989e-05,
+      "loss": 0.0817,
+      "step": 32260
+    },
+    {
+      "epoch": 2.1112201504743213,
+      "grad_norm": 0.8915346264839172,
+      "learning_rate": 9.550034887660143e-05,
+      "loss": 0.0801,
+      "step": 32270
+    },
+    {
+      "epoch": 2.111874386653582,
+      "grad_norm": 0.859603762626648,
+      "learning_rate": 9.549653968150682e-05,
+      "loss": 0.0939,
+      "step": 32280
+    },
+    {
+      "epoch": 2.112528622832843,
+      "grad_norm": 0.7257876396179199,
+      "learning_rate": 9.54927289507847e-05,
+      "loss": 0.0931,
+      "step": 32290
+    },
+    {
+      "epoch": 2.1131828590121033,
+      "grad_norm": 1.2223953008651733,
+      "learning_rate": 9.548891668456367e-05,
+      "loss": 0.1026,
+      "step": 32300
+    },
+    {
+      "epoch": 2.113837095191364,
+      "grad_norm": 0.8096550107002258,
+      "learning_rate": 9.54851028829724e-05,
+      "loss": 0.0877,
+      "step": 32310
+    },
+    {
+      "epoch": 2.114491331370625,
+      "grad_norm": 0.9068686366081238,
+      "learning_rate": 9.548128754613963e-05,
+      "loss": 0.1137,
+      "step": 32320
+    },
+    {
+      "epoch": 2.1151455675498854,
+      "grad_norm": 0.8379555940628052,
+      "learning_rate": 9.54774706741941e-05,
+      "loss": 0.0789,
+      "step": 32330
+    },
+    {
+      "epoch": 2.1157998037291463,
+      "grad_norm": 1.188866376876831,
+      "learning_rate": 9.547365226726468e-05,
+      "loss": 0.0919,
+      "step": 32340
+    },
+    {
+      "epoch": 2.116454039908407,
+      "grad_norm": 0.776324450969696,
+      "learning_rate": 9.546983232548023e-05,
+      "loss": 0.0906,
+      "step": 32350
+    },
+    {
+      "epoch": 2.117108276087668,
+      "grad_norm": 0.7669685482978821,
+      "learning_rate": 9.546601084896971e-05,
+      "loss": 0.0905,
+      "step": 32360
+    },
+    {
+      "epoch": 2.1177625122669284,
+      "grad_norm": 0.8013961911201477,
+      "learning_rate": 9.546218783786207e-05,
+      "loss": 0.091,
+      "step": 32370
+    },
+    {
+      "epoch": 2.118416748446189,
+      "grad_norm": 0.8074659705162048,
+      "learning_rate": 9.545836329228637e-05,
+      "loss": 0.0855,
+      "step": 32380
+    },
+    {
+      "epoch": 2.11907098462545,
+      "grad_norm": 0.8233710527420044,
+      "learning_rate": 9.545453721237167e-05,
+      "loss": 0.0879,
+      "step": 32390
+    },
+    {
+      "epoch": 2.1197252208047104,
+      "grad_norm": 0.8255857229232788,
+      "learning_rate": 9.545070959824716e-05,
+      "loss": 0.0876,
+      "step": 32400
+    },
+    {
+      "epoch": 2.1203794569839713,
+      "grad_norm": 0.8304885625839233,
+      "learning_rate": 9.544688045004197e-05,
+      "loss": 0.0903,
+      "step": 32410
+    },
+    {
+      "epoch": 2.121033693163232,
+      "grad_norm": 0.9255449175834656,
+      "learning_rate": 9.544304976788541e-05,
+      "loss": 0.0913,
+      "step": 32420
+    },
+    {
+      "epoch": 2.121687929342493,
+      "grad_norm": 0.8845074772834778,
+      "learning_rate": 9.543921755190671e-05,
+      "loss": 0.0849,
+      "step": 32430
+    },
+    {
+      "epoch": 2.1223421655217534,
+      "grad_norm": 0.9310880303382874,
+      "learning_rate": 9.543538380223527e-05,
+      "loss": 0.0988,
+      "step": 32440
+    },
+    {
+      "epoch": 2.122996401701014,
+      "grad_norm": 0.953241765499115,
+      "learning_rate": 9.543154851900045e-05,
+      "loss": 0.0909,
+      "step": 32450
+    },
+    {
+      "epoch": 2.123650637880275,
+      "grad_norm": 0.8089374303817749,
+      "learning_rate": 9.542771170233173e-05,
+      "loss": 0.0877,
+      "step": 32460
+    },
+    {
+      "epoch": 2.1243048740595354,
+      "grad_norm": 0.8584646582603455,
+      "learning_rate": 9.542387335235861e-05,
+      "loss": 0.0873,
+      "step": 32470
+    },
+    {
+      "epoch": 2.1249591102387964,
+      "grad_norm": 0.7349388599395752,
+      "learning_rate": 9.542003346921063e-05,
+      "loss": 0.0909,
+      "step": 32480
+    },
+    {
+      "epoch": 2.125613346418057,
+      "grad_norm": 0.9695256352424622,
+      "learning_rate": 9.541619205301739e-05,
+      "loss": 0.102,
+      "step": 32490
+    },
+    {
+      "epoch": 2.1262675825973174,
+      "grad_norm": 0.6882131695747375,
+      "learning_rate": 9.541234910390857e-05,
+      "loss": 0.0823,
+      "step": 32500
+    },
+    {
+      "epoch": 2.1269218187765784,
+      "grad_norm": 0.8628367185592651,
+      "learning_rate": 9.540850462201387e-05,
+      "loss": 0.0998,
+      "step": 32510
+    },
+    {
+      "epoch": 2.127576054955839,
+      "grad_norm": 0.7853913903236389,
+      "learning_rate": 9.540465860746305e-05,
+      "loss": 0.0905,
+      "step": 32520
+    },
+    {
+      "epoch": 2.1282302911351,
+      "grad_norm": 0.9766311049461365,
+      "learning_rate": 9.540081106038591e-05,
+      "loss": 0.0928,
+      "step": 32530
+    },
+    {
+      "epoch": 2.1288845273143604,
+      "grad_norm": 0.7391847372055054,
+      "learning_rate": 9.539696198091235e-05,
+      "loss": 0.084,
+      "step": 32540
+    },
+    {
+      "epoch": 2.1295387634936214,
+      "grad_norm": 0.9374977946281433,
+      "learning_rate": 9.539311136917227e-05,
+      "loss": 0.0979,
+      "step": 32550
+    },
+    {
+      "epoch": 2.130192999672882,
+      "grad_norm": 0.8064951300621033,
+      "learning_rate": 9.53892592252956e-05,
+      "loss": 0.0788,
+      "step": 32560
+    },
+    {
+      "epoch": 2.1308472358521424,
+      "grad_norm": 0.8344107866287231,
+      "learning_rate": 9.538540554941242e-05,
+      "loss": 0.0878,
+      "step": 32570
+    },
+    {
+      "epoch": 2.1315014720314034,
+      "grad_norm": 0.8620904684066772,
+      "learning_rate": 9.538155034165277e-05,
+      "loss": 0.0801,
+      "step": 32580
+    },
+    {
+      "epoch": 2.132155708210664,
+      "grad_norm": 0.764388918876648,
+      "learning_rate": 9.537769360214678e-05,
+      "loss": 0.0829,
+      "step": 32590
+    },
+    {
+      "epoch": 2.132809944389925,
+      "grad_norm": 0.7205486297607422,
+      "learning_rate": 9.537383533102462e-05,
+      "loss": 0.0815,
+      "step": 32600
+    },
+    {
+      "epoch": 2.1334641805691854,
+      "grad_norm": 0.7179217338562012,
+      "learning_rate": 9.536997552841653e-05,
+      "loss": 0.097,
+      "step": 32610
+    },
+    {
+      "epoch": 2.1341184167484464,
+      "grad_norm": 0.8533260226249695,
+      "learning_rate": 9.536611419445276e-05,
+      "loss": 0.0957,
+      "step": 32620
+    },
+    {
+      "epoch": 2.134772652927707,
+      "grad_norm": 0.9501368403434753,
+      "learning_rate": 9.536225132926366e-05,
+      "loss": 0.0954,
+      "step": 32630
+    },
+    {
+      "epoch": 2.1354268891069674,
+      "grad_norm": 0.8990276455879211,
+      "learning_rate": 9.535838693297963e-05,
+      "loss": 0.0895,
+      "step": 32640
+    },
+    {
+      "epoch": 2.1360811252862284,
+      "grad_norm": 0.9564955830574036,
+      "learning_rate": 9.535452100573108e-05,
+      "loss": 0.0933,
+      "step": 32650
+    },
+    {
+      "epoch": 2.136735361465489,
+      "grad_norm": 0.9729709625244141,
+      "learning_rate": 9.53506535476485e-05,
+      "loss": 0.1018,
+      "step": 32660
+    },
+    {
+      "epoch": 2.13738959764475,
+      "grad_norm": 0.8804893493652344,
+      "learning_rate": 9.534678455886241e-05,
+      "loss": 0.0903,
+      "step": 32670
+    },
+    {
+      "epoch": 2.1380438338240104,
+      "grad_norm": 0.8837690353393555,
+      "learning_rate": 9.534291403950341e-05,
+      "loss": 0.0917,
+      "step": 32680
+    },
+    {
+      "epoch": 2.1386980700032714,
+      "grad_norm": 0.9111039042472839,
+      "learning_rate": 9.533904198970218e-05,
+      "loss": 0.0803,
+      "step": 32690
+    },
+    {
+      "epoch": 2.139352306182532,
+      "grad_norm": 0.7728859186172485,
+      "learning_rate": 9.533516840958934e-05,
+      "loss": 0.081,
+      "step": 32700
+    },
+    {
+      "epoch": 2.1400065423617924,
+      "grad_norm": 0.8555951118469238,
+      "learning_rate": 9.533129329929568e-05,
+      "loss": 0.0857,
+      "step": 32710
+    },
+    {
+      "epoch": 2.1406607785410534,
+      "grad_norm": 0.761581540107727,
+      "learning_rate": 9.5327416658952e-05,
+      "loss": 0.0819,
+      "step": 32720
+    },
+    {
+      "epoch": 2.141315014720314,
+      "grad_norm": 0.7355059385299683,
+      "learning_rate": 9.532353848868914e-05,
+      "loss": 0.0839,
+      "step": 32730
+    },
+    {
+      "epoch": 2.141969250899575,
+      "grad_norm": 0.8941879868507385,
+      "learning_rate": 9.531965878863797e-05,
+      "loss": 0.0905,
+      "step": 32740
+    },
+    {
+      "epoch": 2.1426234870788354,
+      "grad_norm": 0.7693347334861755,
+      "learning_rate": 9.531577755892947e-05,
+      "loss": 0.0845,
+      "step": 32750
+    },
+    {
+      "epoch": 2.1432777232580964,
+      "grad_norm": 0.8309009671211243,
+      "learning_rate": 9.531189479969462e-05,
+      "loss": 0.0811,
+      "step": 32760
+    },
+    {
+      "epoch": 2.143931959437357,
+      "grad_norm": 0.8143077492713928,
+      "learning_rate": 9.530801051106449e-05,
+      "loss": 0.0857,
+      "step": 32770
+    },
+    {
+      "epoch": 2.1445861956166175,
+      "grad_norm": 0.9192956686019897,
+      "learning_rate": 9.53041246931702e-05,
+      "loss": 0.0907,
+      "step": 32780
+    },
+    {
+      "epoch": 2.1452404317958784,
+      "grad_norm": 0.9745957255363464,
+      "learning_rate": 9.530023734614286e-05,
+      "loss": 0.0898,
+      "step": 32790
+    },
+    {
+      "epoch": 2.145894667975139,
+      "grad_norm": 1.038955569267273,
+      "learning_rate": 9.529634847011373e-05,
+      "loss": 0.0901,
+      "step": 32800
+    },
+    {
+      "epoch": 2.1465489041544,
+      "grad_norm": 0.9316650032997131,
+      "learning_rate": 9.529245806521402e-05,
+      "loss": 0.078,
+      "step": 32810
+    },
+    {
+      "epoch": 2.1472031403336604,
+      "grad_norm": 0.9526748061180115,
+      "learning_rate": 9.528856613157509e-05,
+      "loss": 0.0921,
+      "step": 32820
+    },
+    {
+      "epoch": 2.147857376512921,
+      "grad_norm": 0.9869240522384644,
+      "learning_rate": 9.528467266932826e-05,
+      "loss": 0.0872,
+      "step": 32830
+    },
+    {
+      "epoch": 2.148511612692182,
+      "grad_norm": 0.9078952670097351,
+      "learning_rate": 9.528077767860497e-05,
+      "loss": 0.0867,
+      "step": 32840
+    },
+    {
+      "epoch": 2.1491658488714425,
+      "grad_norm": 1.0087946653366089,
+      "learning_rate": 9.527688115953668e-05,
+      "loss": 0.0835,
+      "step": 32850
+    },
+    {
+      "epoch": 2.1498200850507034,
+      "grad_norm": 0.8324182629585266,
+      "learning_rate": 9.527298311225493e-05,
+      "loss": 0.084,
+      "step": 32860
+    },
+    {
+      "epoch": 2.150474321229964,
+      "grad_norm": 0.7914925217628479,
+      "learning_rate": 9.526908353689123e-05,
+      "loss": 0.09,
+      "step": 32870
+    },
+    {
+      "epoch": 2.151128557409225,
+      "grad_norm": 0.8204302787780762,
+      "learning_rate": 9.526518243357725e-05,
+      "loss": 0.0969,
+      "step": 32880
+    },
+    {
+      "epoch": 2.1517827935884855,
+      "grad_norm": 0.8147171139717102,
+      "learning_rate": 9.526127980244466e-05,
+      "loss": 0.0867,
+      "step": 32890
+    },
+    {
+      "epoch": 2.152437029767746,
+      "grad_norm": 0.8153404593467712,
+      "learning_rate": 9.525737564362517e-05,
+      "loss": 0.078,
+      "step": 32900
+    },
+    {
+      "epoch": 2.153091265947007,
+      "grad_norm": 0.9826894402503967,
+      "learning_rate": 9.525346995725057e-05,
+      "loss": 0.0961,
+      "step": 32910
+    },
+    {
+      "epoch": 2.1537455021262675,
+      "grad_norm": 0.911864697933197,
+      "learning_rate": 9.524956274345268e-05,
+      "loss": 0.0884,
+      "step": 32920
+    },
+    {
+      "epoch": 2.1543997383055284,
+      "grad_norm": 0.9221819639205933,
+      "learning_rate": 9.524565400236335e-05,
+      "loss": 0.0984,
+      "step": 32930
+    },
+    {
+      "epoch": 2.155053974484789,
+      "grad_norm": 0.8833640217781067,
+      "learning_rate": 9.524174373411456e-05,
+      "loss": 0.0844,
+      "step": 32940
+    },
+    {
+      "epoch": 2.1557082106640495,
+      "grad_norm": 0.9329414963722229,
+      "learning_rate": 9.523783193883825e-05,
+      "loss": 0.0762,
+      "step": 32950
+    },
+    {
+      "epoch": 2.1563624468433105,
+      "grad_norm": 0.8548219799995422,
+      "learning_rate": 9.523391861666649e-05,
+      "loss": 0.0892,
+      "step": 32960
+    },
+    {
+      "epoch": 2.157016683022571,
+      "grad_norm": 0.9100401401519775,
+      "learning_rate": 9.523000376773132e-05,
+      "loss": 0.0935,
+      "step": 32970
+    },
+    {
+      "epoch": 2.157670919201832,
+      "grad_norm": 0.8718044757843018,
+      "learning_rate": 9.522608739216493e-05,
+      "loss": 0.0816,
+      "step": 32980
+    },
+    {
+      "epoch": 2.1583251553810925,
+      "grad_norm": 0.8366650938987732,
+      "learning_rate": 9.522216949009946e-05,
+      "loss": 0.079,
+      "step": 32990
+    },
+    {
+      "epoch": 2.1589793915603535,
+      "grad_norm": 0.9176976084709167,
+      "learning_rate": 9.52182500616672e-05,
+      "loss": 0.089,
+      "step": 33000
+    },
+    {
+      "epoch": 2.159633627739614,
+      "grad_norm": 1.1271123886108398,
+      "learning_rate": 9.521432910700039e-05,
+      "loss": 0.0919,
+      "step": 33010
+    },
+    {
+      "epoch": 2.1602878639188745,
+      "grad_norm": 0.935555100440979,
+      "learning_rate": 9.521040662623139e-05,
+      "loss": 0.0968,
+      "step": 33020
+    },
+    {
+      "epoch": 2.1609421000981355,
+      "grad_norm": 0.7743455767631531,
+      "learning_rate": 9.52064826194926e-05,
+      "loss": 0.0987,
+      "step": 33030
+    },
+    {
+      "epoch": 2.161596336277396,
+      "grad_norm": 0.7663136720657349,
+      "learning_rate": 9.520255708691646e-05,
+      "loss": 0.0865,
+      "step": 33040
+    },
+    {
+      "epoch": 2.162250572456657,
+      "grad_norm": 0.851262092590332,
+      "learning_rate": 9.519863002863548e-05,
+      "loss": 0.0805,
+      "step": 33050
+    },
+    {
+      "epoch": 2.1629048086359175,
+      "grad_norm": 1.051383137702942,
+      "learning_rate": 9.519470144478219e-05,
+      "loss": 0.099,
+      "step": 33060
+    },
+    {
+      "epoch": 2.1635590448151785,
+      "grad_norm": 0.7119585871696472,
+      "learning_rate": 9.519077133548922e-05,
+      "loss": 0.0792,
+      "step": 33070
+    },
+    {
+      "epoch": 2.164213280994439,
+      "grad_norm": 0.7884335517883301,
+      "learning_rate": 9.518683970088918e-05,
+      "loss": 0.087,
+      "step": 33080
+    },
+    {
+      "epoch": 2.1648675171736995,
+      "grad_norm": 0.80777907371521,
+      "learning_rate": 9.51829065411148e-05,
+      "loss": 0.0824,
+      "step": 33090
+    },
+    {
+      "epoch": 2.1655217533529605,
+      "grad_norm": 1.0060641765594482,
+      "learning_rate": 9.517897185629882e-05,
+      "loss": 0.0832,
+      "step": 33100
+    },
+    {
+      "epoch": 2.166175989532221,
+      "grad_norm": 0.9100120067596436,
+      "learning_rate": 9.517503564657407e-05,
+      "loss": 0.0868,
+      "step": 33110
+    },
+    {
+      "epoch": 2.166830225711482,
+      "grad_norm": 0.8094409704208374,
+      "learning_rate": 9.517109791207337e-05,
+      "loss": 0.0823,
+      "step": 33120
+    },
+    {
+      "epoch": 2.1674844618907425,
+      "grad_norm": 1.0569559335708618,
+      "learning_rate": 9.516715865292967e-05,
+      "loss": 0.0803,
+      "step": 33130
+    },
+    {
+      "epoch": 2.1681386980700035,
+      "grad_norm": 0.7555559277534485,
+      "learning_rate": 9.51632178692759e-05,
+      "loss": 0.0821,
+      "step": 33140
+    },
+    {
+      "epoch": 2.168792934249264,
+      "grad_norm": 0.9259467720985413,
+      "learning_rate": 9.515927556124507e-05,
+      "loss": 0.0833,
+      "step": 33150
+    },
+    {
+      "epoch": 2.1694471704285245,
+      "grad_norm": 0.9541639685630798,
+      "learning_rate": 9.515533172897028e-05,
+      "loss": 0.0959,
+      "step": 33160
+    },
+    {
+      "epoch": 2.1701014066077855,
+      "grad_norm": 0.7850528955459595,
+      "learning_rate": 9.51513863725846e-05,
+      "loss": 0.0848,
+      "step": 33170
+    },
+    {
+      "epoch": 2.170755642787046,
+      "grad_norm": 0.921647310256958,
+      "learning_rate": 9.514743949222122e-05,
+      "loss": 0.0864,
+      "step": 33180
+    },
+    {
+      "epoch": 2.171409878966307,
+      "grad_norm": 0.8776901364326477,
+      "learning_rate": 9.514349108801337e-05,
+      "loss": 0.1069,
+      "step": 33190
+    },
+    {
+      "epoch": 2.1720641151455675,
+      "grad_norm": 0.8369219303131104,
+      "learning_rate": 9.513954116009429e-05,
+      "loss": 0.0923,
+      "step": 33200
+    },
+    {
+      "epoch": 2.1727183513248285,
+      "grad_norm": 0.8588852882385254,
+      "learning_rate": 9.51355897085973e-05,
+      "loss": 0.0969,
+      "step": 33210
+    },
+    {
+      "epoch": 2.173372587504089,
+      "grad_norm": 0.8265673518180847,
+      "learning_rate": 9.513163673365581e-05,
+      "loss": 0.0817,
+      "step": 33220
+    },
+    {
+      "epoch": 2.1740268236833495,
+      "grad_norm": 0.8380441069602966,
+      "learning_rate": 9.512768223540321e-05,
+      "loss": 0.0827,
+      "step": 33230
+    },
+    {
+      "epoch": 2.1746810598626105,
+      "grad_norm": 0.7195032835006714,
+      "learning_rate": 9.512372621397298e-05,
+      "loss": 0.0922,
+      "step": 33240
+    },
+    {
+      "epoch": 2.175335296041871,
+      "grad_norm": 0.8967887759208679,
+      "learning_rate": 9.511976866949864e-05,
+      "loss": 0.0861,
+      "step": 33250
+    },
+    {
+      "epoch": 2.175989532221132,
+      "grad_norm": 0.8154705762863159,
+      "learning_rate": 9.51158096021138e-05,
+      "loss": 0.0936,
+      "step": 33260
+    },
+    {
+      "epoch": 2.1766437684003925,
+      "grad_norm": 0.8090271353721619,
+      "learning_rate": 9.511184901195204e-05,
+      "loss": 0.084,
+      "step": 33270
+    },
+    {
+      "epoch": 2.177298004579653,
+      "grad_norm": 0.8615165948867798,
+      "learning_rate": 9.510788689914707e-05,
+      "loss": 0.0891,
+      "step": 33280
+    },
+    {
+      "epoch": 2.177952240758914,
+      "grad_norm": 0.8889568448066711,
+      "learning_rate": 9.510392326383262e-05,
+      "loss": 0.0916,
+      "step": 33290
+    },
+    {
+      "epoch": 2.1786064769381746,
+      "grad_norm": 0.9349720478057861,
+      "learning_rate": 9.509995810614247e-05,
+      "loss": 0.0903,
+      "step": 33300
+    },
+    {
+      "epoch": 2.1792607131174355,
+      "grad_norm": 0.8932443261146545,
+      "learning_rate": 9.509599142621047e-05,
+      "loss": 0.0844,
+      "step": 33310
+    },
+    {
+      "epoch": 2.179914949296696,
+      "grad_norm": 0.8143129944801331,
+      "learning_rate": 9.509202322417047e-05,
+      "loss": 0.087,
+      "step": 33320
+    },
+    {
+      "epoch": 2.180569185475957,
+      "grad_norm": 0.8007728457450867,
+      "learning_rate": 9.508805350015643e-05,
+      "loss": 0.0925,
+      "step": 33330
+    },
+    {
+      "epoch": 2.1812234216552175,
+      "grad_norm": 0.8921264410018921,
+      "learning_rate": 9.508408225430237e-05,
+      "loss": 0.0925,
+      "step": 33340
+    },
+    {
+      "epoch": 2.181877657834478,
+      "grad_norm": 0.9129869341850281,
+      "learning_rate": 9.508010948674227e-05,
+      "loss": 0.0911,
+      "step": 33350
+    },
+    {
+      "epoch": 2.182531894013739,
+      "grad_norm": 0.9659237265586853,
+      "learning_rate": 9.507613519761022e-05,
+      "loss": 0.0915,
+      "step": 33360
+    },
+    {
+      "epoch": 2.1831861301929996,
+      "grad_norm": 0.8064131736755371,
+      "learning_rate": 9.507215938704043e-05,
+      "loss": 0.0845,
+      "step": 33370
+    },
+    {
+      "epoch": 2.1838403663722605,
+      "grad_norm": 0.9400182366371155,
+      "learning_rate": 9.506818205516705e-05,
+      "loss": 0.0886,
+      "step": 33380
+    },
+    {
+      "epoch": 2.184494602551521,
+      "grad_norm": 0.8010856509208679,
+      "learning_rate": 9.506420320212433e-05,
+      "loss": 0.0912,
+      "step": 33390
+    },
+    {
+      "epoch": 2.1851488387307816,
+      "grad_norm": 0.7543482184410095,
+      "learning_rate": 9.506022282804656e-05,
+      "loss": 0.0765,
+      "step": 33400
+    },
+    {
+      "epoch": 2.1858030749100426,
+      "grad_norm": 0.8534349799156189,
+      "learning_rate": 9.505624093306809e-05,
+      "loss": 0.0913,
+      "step": 33410
+    },
+    {
+      "epoch": 2.186457311089303,
+      "grad_norm": 0.8279025554656982,
+      "learning_rate": 9.505225751732333e-05,
+      "loss": 0.0815,
+      "step": 33420
+    },
+    {
+      "epoch": 2.187111547268564,
+      "grad_norm": 0.8194913268089294,
+      "learning_rate": 9.504827258094673e-05,
+      "loss": 0.0924,
+      "step": 33430
+    },
+    {
+      "epoch": 2.1877657834478246,
+      "grad_norm": 0.9188143014907837,
+      "learning_rate": 9.504428612407277e-05,
+      "loss": 0.0892,
+      "step": 33440
+    },
+    {
+      "epoch": 2.1884200196270855,
+      "grad_norm": 0.6940245032310486,
+      "learning_rate": 9.504029814683603e-05,
+      "loss": 0.0797,
+      "step": 33450
+    },
+    {
+      "epoch": 2.189074255806346,
+      "grad_norm": 0.8881454467773438,
+      "learning_rate": 9.503630864937112e-05,
+      "loss": 0.0794,
+      "step": 33460
+    },
+    {
+      "epoch": 2.1897284919856066,
+      "grad_norm": 0.8755089044570923,
+      "learning_rate": 9.503231763181266e-05,
+      "loss": 0.0877,
+      "step": 33470
+    },
+    {
+      "epoch": 2.1903827281648676,
+      "grad_norm": 0.7432528734207153,
+      "learning_rate": 9.502832509429538e-05,
+      "loss": 0.0857,
+      "step": 33480
+    },
+    {
+      "epoch": 2.191036964344128,
+      "grad_norm": 0.7518520355224609,
+      "learning_rate": 9.502433103695405e-05,
+      "loss": 0.0862,
+      "step": 33490
+    },
+    {
+      "epoch": 2.191691200523389,
+      "grad_norm": 0.7373502254486084,
+      "learning_rate": 9.502033545992347e-05,
+      "loss": 0.0839,
+      "step": 33500
+    },
+    {
+      "epoch": 2.1923454367026496,
+      "grad_norm": 0.8541769981384277,
+      "learning_rate": 9.501633836333847e-05,
+      "loss": 0.0858,
+      "step": 33510
+    },
+    {
+      "epoch": 2.1929996728819106,
+      "grad_norm": 0.8180940747261047,
+      "learning_rate": 9.501233974733402e-05,
+      "loss": 0.0857,
+      "step": 33520
+    },
+    {
+      "epoch": 2.193653909061171,
+      "grad_norm": 0.6392052173614502,
+      "learning_rate": 9.500833961204504e-05,
+      "loss": 0.0873,
+      "step": 33530
+    },
+    {
+      "epoch": 2.1943081452404316,
+      "grad_norm": 0.8578611016273499,
+      "learning_rate": 9.500433795760656e-05,
+      "loss": 0.0951,
+      "step": 33540
+    },
+    {
+      "epoch": 2.1949623814196926,
+      "grad_norm": 0.7681283354759216,
+      "learning_rate": 9.500033478415364e-05,
+      "loss": 0.0896,
+      "step": 33550
+    },
+    {
+      "epoch": 2.195616617598953,
+      "grad_norm": 1.0089340209960938,
+      "learning_rate": 9.499633009182141e-05,
+      "loss": 0.0847,
+      "step": 33560
+    },
+    {
+      "epoch": 2.196270853778214,
+      "grad_norm": 0.9923962950706482,
+      "learning_rate": 9.499232388074503e-05,
+      "loss": 0.0849,
+      "step": 33570
+    },
+    {
+      "epoch": 2.1969250899574746,
+      "grad_norm": 0.9396527409553528,
+      "learning_rate": 9.498831615105974e-05,
+      "loss": 0.0817,
+      "step": 33580
+    },
+    {
+      "epoch": 2.1975793261367356,
+      "grad_norm": 0.7648763060569763,
+      "learning_rate": 9.498430690290078e-05,
+      "loss": 0.0966,
+      "step": 33590
+    },
+    {
+      "epoch": 2.198233562315996,
+      "grad_norm": 0.8955914974212646,
+      "learning_rate": 9.498029613640349e-05,
+      "loss": 0.1008,
+      "step": 33600
+    },
+    {
+      "epoch": 2.1988877984952566,
+      "grad_norm": 1.131964921951294,
+      "learning_rate": 9.497628385170323e-05,
+      "loss": 0.0985,
+      "step": 33610
+    },
+    {
+      "epoch": 2.1995420346745176,
+      "grad_norm": 0.738293468952179,
+      "learning_rate": 9.497227004893544e-05,
+      "loss": 0.0794,
+      "step": 33620
+    },
+    {
+      "epoch": 2.200196270853778,
+      "grad_norm": 0.7371350526809692,
+      "learning_rate": 9.496825472823559e-05,
+      "loss": 0.0758,
+      "step": 33630
+    },
+    {
+      "epoch": 2.200850507033039,
+      "grad_norm": 0.8156116604804993,
+      "learning_rate": 9.496423788973922e-05,
+      "loss": 0.0796,
+      "step": 33640
+    },
+    {
+      "epoch": 2.2015047432122996,
+      "grad_norm": 0.832074761390686,
+      "learning_rate": 9.496021953358189e-05,
+      "loss": 0.0778,
+      "step": 33650
+    },
+    {
+      "epoch": 2.2021589793915606,
+      "grad_norm": 0.9009979963302612,
+      "learning_rate": 9.495619965989924e-05,
+      "loss": 0.0953,
+      "step": 33660
+    },
+    {
+      "epoch": 2.202813215570821,
+      "grad_norm": 0.9261824488639832,
+      "learning_rate": 9.495217826882694e-05,
+      "loss": 0.0941,
+      "step": 33670
+    },
+    {
+      "epoch": 2.2034674517500816,
+      "grad_norm": 0.815540611743927,
+      "learning_rate": 9.494815536050075e-05,
+      "loss": 0.0858,
+      "step": 33680
+    },
+    {
+      "epoch": 2.2041216879293426,
+      "grad_norm": 0.8841716647148132,
+      "learning_rate": 9.494413093505643e-05,
+      "loss": 0.0789,
+      "step": 33690
+    },
+    {
+      "epoch": 2.204775924108603,
+      "grad_norm": 0.9141961932182312,
+      "learning_rate": 9.494010499262982e-05,
+      "loss": 0.0936,
+      "step": 33700
+    },
+    {
+      "epoch": 2.205430160287864,
+      "grad_norm": 0.7224341630935669,
+      "learning_rate": 9.49360775333568e-05,
+      "loss": 0.0938,
+      "step": 33710
+    },
+    {
+      "epoch": 2.2060843964671246,
+      "grad_norm": 0.9036959409713745,
+      "learning_rate": 9.493204855737332e-05,
+      "loss": 0.0769,
+      "step": 33720
+    },
+    {
+      "epoch": 2.206738632646385,
+      "grad_norm": 0.9776110649108887,
+      "learning_rate": 9.492801806481535e-05,
+      "loss": 0.0925,
+      "step": 33730
+    },
+    {
+      "epoch": 2.207392868825646,
+      "grad_norm": 0.6493380665779114,
+      "learning_rate": 9.492398605581896e-05,
+      "loss": 0.0948,
+      "step": 33740
+    },
+    {
+      "epoch": 2.2080471050049066,
+      "grad_norm": 0.8940728902816772,
+      "learning_rate": 9.491995253052022e-05,
+      "loss": 0.0949,
+      "step": 33750
+    },
+    {
+      "epoch": 2.2087013411841676,
+      "grad_norm": 0.9045453071594238,
+      "learning_rate": 9.491591748905527e-05,
+      "loss": 0.0905,
+      "step": 33760
+    },
+    {
+      "epoch": 2.209355577363428,
+      "grad_norm": 0.7662135362625122,
+      "learning_rate": 9.49118809315603e-05,
+      "loss": 0.0852,
+      "step": 33770
+    },
+    {
+      "epoch": 2.210009813542689,
+      "grad_norm": 0.7903282642364502,
+      "learning_rate": 9.490784285817158e-05,
+      "loss": 0.0792,
+      "step": 33780
+    },
+    {
+      "epoch": 2.2106640497219496,
+      "grad_norm": 0.9852091073989868,
+      "learning_rate": 9.490380326902537e-05,
+      "loss": 0.0826,
+      "step": 33790
+    },
+    {
+      "epoch": 2.21131828590121,
+      "grad_norm": 0.9630650281906128,
+      "learning_rate": 9.489976216425804e-05,
+      "loss": 0.0994,
+      "step": 33800
+    },
+    {
+      "epoch": 2.211972522080471,
+      "grad_norm": 0.7870905995368958,
+      "learning_rate": 9.4895719544006e-05,
+      "loss": 0.0926,
+      "step": 33810
+    },
+    {
+      "epoch": 2.2126267582597317,
+      "grad_norm": 0.8298508524894714,
+      "learning_rate": 9.489167540840567e-05,
+      "loss": 0.0815,
+      "step": 33820
+    },
+    {
+      "epoch": 2.2132809944389926,
+      "grad_norm": 0.8366494178771973,
+      "learning_rate": 9.488762975759357e-05,
+      "loss": 0.1,
+      "step": 33830
+    },
+    {
+      "epoch": 2.213935230618253,
+      "grad_norm": 0.9144079685211182,
+      "learning_rate": 9.488358259170622e-05,
+      "loss": 0.0881,
+      "step": 33840
+    },
+    {
+      "epoch": 2.2145894667975137,
+      "grad_norm": 0.8029112815856934,
+      "learning_rate": 9.487953391088027e-05,
+      "loss": 0.0793,
+      "step": 33850
+    },
+    {
+      "epoch": 2.2152437029767746,
+      "grad_norm": 0.9416177272796631,
+      "learning_rate": 9.487548371525234e-05,
+      "loss": 0.0827,
+      "step": 33860
+    },
+    {
+      "epoch": 2.215897939156035,
+      "grad_norm": 0.7527725696563721,
+      "learning_rate": 9.487143200495914e-05,
+      "loss": 0.0921,
+      "step": 33870
+    },
+    {
+      "epoch": 2.216552175335296,
+      "grad_norm": 1.0076922178268433,
+      "learning_rate": 9.486737878013745e-05,
+      "loss": 0.0852,
+      "step": 33880
+    },
+    {
+      "epoch": 2.2172064115145567,
+      "grad_norm": 1.0222320556640625,
+      "learning_rate": 9.486332404092403e-05,
+      "loss": 0.09,
+      "step": 33890
+    },
+    {
+      "epoch": 2.2178606476938176,
+      "grad_norm": 0.9221048951148987,
+      "learning_rate": 9.485926778745579e-05,
+      "loss": 0.091,
+      "step": 33900
+    },
+    {
+      "epoch": 2.218514883873078,
+      "grad_norm": 0.8487015962600708,
+      "learning_rate": 9.485521001986962e-05,
+      "loss": 0.0853,
+      "step": 33910
+    },
+    {
+      "epoch": 2.2191691200523387,
+      "grad_norm": 0.8243840336799622,
+      "learning_rate": 9.485115073830245e-05,
+      "loss": 0.0884,
+      "step": 33920
+    },
+    {
+      "epoch": 2.2198233562315997,
+      "grad_norm": 0.8864408731460571,
+      "learning_rate": 9.48470899428913e-05,
+      "loss": 0.1056,
+      "step": 33930
+    },
+    {
+      "epoch": 2.22047759241086,
+      "grad_norm": 1.3254485130310059,
+      "learning_rate": 9.484302763377328e-05,
+      "loss": 0.0855,
+      "step": 33940
+    },
+    {
+      "epoch": 2.221131828590121,
+      "grad_norm": 1.0001471042633057,
+      "learning_rate": 9.483896381108548e-05,
+      "loss": 0.0901,
+      "step": 33950
+    },
+    {
+      "epoch": 2.2217860647693817,
+      "grad_norm": 0.8576057553291321,
+      "learning_rate": 9.483489847496503e-05,
+      "loss": 0.0802,
+      "step": 33960
+    },
+    {
+      "epoch": 2.2224403009486426,
+      "grad_norm": 0.8410997986793518,
+      "learning_rate": 9.48308316255492e-05,
+      "loss": 0.0777,
+      "step": 33970
+    },
+    {
+      "epoch": 2.223094537127903,
+      "grad_norm": 1.0738435983657837,
+      "learning_rate": 9.482676326297522e-05,
+      "loss": 0.093,
+      "step": 33980
+    },
+    {
+      "epoch": 2.2237487733071637,
+      "grad_norm": 0.7275444865226746,
+      "learning_rate": 9.482269338738038e-05,
+      "loss": 0.0788,
+      "step": 33990
+    },
+    {
+      "epoch": 2.2244030094864247,
+      "grad_norm": 0.7812187075614929,
+      "learning_rate": 9.481862199890213e-05,
+      "loss": 0.0866,
+      "step": 34000
+    },
+    {
+      "epoch": 2.225057245665685,
+      "grad_norm": 0.7161325216293335,
+      "learning_rate": 9.481454909767784e-05,
+      "loss": 0.0865,
+      "step": 34010
+    },
+    {
+      "epoch": 2.225711481844946,
+      "grad_norm": 0.9743297696113586,
+      "learning_rate": 9.481047468384499e-05,
+      "loss": 0.0829,
+      "step": 34020
+    },
+    {
+      "epoch": 2.2263657180242067,
+      "grad_norm": 0.9673652648925781,
+      "learning_rate": 9.480639875754108e-05,
+      "loss": 0.0828,
+      "step": 34030
+    },
+    {
+      "epoch": 2.2270199542034677,
+      "grad_norm": 0.8574245572090149,
+      "learning_rate": 9.480232131890371e-05,
+      "loss": 0.0832,
+      "step": 34040
+    },
+    {
+      "epoch": 2.227674190382728,
+      "grad_norm": 1.0479404926300049,
+      "learning_rate": 9.479824236807051e-05,
+      "loss": 0.0825,
+      "step": 34050
+    },
+    {
+      "epoch": 2.2283284265619887,
+      "grad_norm": 0.9028341174125671,
+      "learning_rate": 9.479416190517914e-05,
+      "loss": 0.0804,
+      "step": 34060
+    },
+    {
+      "epoch": 2.2289826627412497,
+      "grad_norm": 0.7553834915161133,
+      "learning_rate": 9.479007993036733e-05,
+      "loss": 0.0798,
+      "step": 34070
+    },
+    {
+      "epoch": 2.22963689892051,
+      "grad_norm": 0.9540044069290161,
+      "learning_rate": 9.478599644377284e-05,
+      "loss": 0.092,
+      "step": 34080
+    },
+    {
+      "epoch": 2.230291135099771,
+      "grad_norm": 1.0096077919006348,
+      "learning_rate": 9.478191144553352e-05,
+      "loss": 0.0833,
+      "step": 34090
+    },
+    {
+      "epoch": 2.2309453712790317,
+      "grad_norm": 0.8619536757469177,
+      "learning_rate": 9.477782493578725e-05,
+      "loss": 0.0983,
+      "step": 34100
+    },
+    {
+      "epoch": 2.2315996074582927,
+      "grad_norm": 0.8202316761016846,
+      "learning_rate": 9.477373691467195e-05,
+      "loss": 0.0795,
+      "step": 34110
+    },
+    {
+      "epoch": 2.232253843637553,
+      "grad_norm": 0.8354489207267761,
+      "learning_rate": 9.47696473823256e-05,
+      "loss": 0.0805,
+      "step": 34120
+    },
+    {
+      "epoch": 2.2329080798168137,
+      "grad_norm": 1.0788700580596924,
+      "learning_rate": 9.476555633888625e-05,
+      "loss": 0.0929,
+      "step": 34130
+    },
+    {
+      "epoch": 2.2335623159960747,
+      "grad_norm": 0.8058347702026367,
+      "learning_rate": 9.476146378449197e-05,
+      "loss": 0.0913,
+      "step": 34140
+    },
+    {
+      "epoch": 2.234216552175335,
+      "grad_norm": 0.8646268844604492,
+      "learning_rate": 9.475736971928088e-05,
+      "loss": 0.0819,
+      "step": 34150
+    },
+    {
+      "epoch": 2.234870788354596,
+      "grad_norm": 0.8829324841499329,
+      "learning_rate": 9.475327414339121e-05,
+      "loss": 0.0907,
+      "step": 34160
+    },
+    {
+      "epoch": 2.2355250245338567,
+      "grad_norm": 0.8465047478675842,
+      "learning_rate": 9.474917705696114e-05,
+      "loss": 0.0884,
+      "step": 34170
+    },
+    {
+      "epoch": 2.2361792607131172,
+      "grad_norm": 0.8332507014274597,
+      "learning_rate": 9.474507846012901e-05,
+      "loss": 0.0899,
+      "step": 34180
+    },
+    {
+      "epoch": 2.236833496892378,
+      "grad_norm": 0.890265941619873,
+      "learning_rate": 9.474097835303311e-05,
+      "loss": 0.0799,
+      "step": 34190
+    },
+    {
+      "epoch": 2.2374877330716387,
+      "grad_norm": 0.9846871495246887,
+      "learning_rate": 9.473687673581186e-05,
+      "loss": 0.0828,
+      "step": 34200
+    },
+    {
+      "epoch": 2.2381419692508997,
+      "grad_norm": 0.7665266394615173,
+      "learning_rate": 9.47327736086037e-05,
+      "loss": 0.0792,
+      "step": 34210
+    },
+    {
+      "epoch": 2.2387962054301602,
+      "grad_norm": 0.82786625623703,
+      "learning_rate": 9.472866897154712e-05,
+      "loss": 0.0888,
+      "step": 34220
+    },
+    {
+      "epoch": 2.239450441609421,
+      "grad_norm": 0.8131442666053772,
+      "learning_rate": 9.472456282478065e-05,
+      "loss": 0.0819,
+      "step": 34230
+    },
+    {
+      "epoch": 2.2401046777886817,
+      "grad_norm": 0.8473891615867615,
+      "learning_rate": 9.47204551684429e-05,
+      "loss": 0.0905,
+      "step": 34240
+    },
+    {
+      "epoch": 2.2407589139679422,
+      "grad_norm": 0.8712317943572998,
+      "learning_rate": 9.471634600267247e-05,
+      "loss": 0.0955,
+      "step": 34250
+    },
+    {
+      "epoch": 2.241413150147203,
+      "grad_norm": 0.7699024081230164,
+      "learning_rate": 9.471223532760812e-05,
+      "loss": 0.0799,
+      "step": 34260
+    },
+    {
+      "epoch": 2.2420673863264637,
+      "grad_norm": 0.7678673267364502,
+      "learning_rate": 9.470812314338855e-05,
+      "loss": 0.0846,
+      "step": 34270
+    },
+    {
+      "epoch": 2.2427216225057247,
+      "grad_norm": 0.8645968437194824,
+      "learning_rate": 9.470400945015258e-05,
+      "loss": 0.0767,
+      "step": 34280
+    },
+    {
+      "epoch": 2.2433758586849852,
+      "grad_norm": 0.9479940533638,
+      "learning_rate": 9.469989424803907e-05,
+      "loss": 0.0875,
+      "step": 34290
+    },
+    {
+      "epoch": 2.2440300948642458,
+      "grad_norm": 0.7931011319160461,
+      "learning_rate": 9.469577753718689e-05,
+      "loss": 0.0817,
+      "step": 34300
+    },
+    {
+      "epoch": 2.2446843310435067,
+      "grad_norm": 0.9428794980049133,
+      "learning_rate": 9.469165931773498e-05,
+      "loss": 0.0824,
+      "step": 34310
+    },
+    {
+      "epoch": 2.2453385672227673,
+      "grad_norm": 0.9132287502288818,
+      "learning_rate": 9.468753958982238e-05,
+      "loss": 0.0886,
+      "step": 34320
+    },
+    {
+      "epoch": 2.2459928034020282,
+      "grad_norm": 0.9737602472305298,
+      "learning_rate": 9.468341835358809e-05,
+      "loss": 0.0937,
+      "step": 34330
+    },
+    {
+      "epoch": 2.2466470395812888,
+      "grad_norm": 0.918177604675293,
+      "learning_rate": 9.467929560917128e-05,
+      "loss": 0.0813,
+      "step": 34340
+    },
+    {
+      "epoch": 2.2473012757605497,
+      "grad_norm": 0.7503966093063354,
+      "learning_rate": 9.467517135671104e-05,
+      "loss": 0.0886,
+      "step": 34350
+    },
+    {
+      "epoch": 2.2479555119398102,
+      "grad_norm": 0.83729487657547,
+      "learning_rate": 9.467104559634663e-05,
+      "loss": 0.0983,
+      "step": 34360
+    },
+    {
+      "epoch": 2.2486097481190708,
+      "grad_norm": 1.0300636291503906,
+      "learning_rate": 9.466691832821725e-05,
+      "loss": 0.0959,
+      "step": 34370
+    },
+    {
+      "epoch": 2.2492639842983317,
+      "grad_norm": 0.8676597476005554,
+      "learning_rate": 9.466278955246225e-05,
+      "loss": 0.0833,
+      "step": 34380
+    },
+    {
+      "epoch": 2.2499182204775923,
+      "grad_norm": 0.9559263586997986,
+      "learning_rate": 9.465865926922098e-05,
+      "loss": 0.0978,
+      "step": 34390
+    },
+    {
+      "epoch": 2.2505724566568532,
+      "grad_norm": 0.7178763151168823,
+      "learning_rate": 9.465452747863281e-05,
+      "loss": 0.0909,
+      "step": 34400
+    },
+    {
+      "epoch": 2.2512266928361138,
+      "grad_norm": 0.7472977638244629,
+      "learning_rate": 9.465039418083723e-05,
+      "loss": 0.0954,
+      "step": 34410
+    },
+    {
+      "epoch": 2.2518809290153747,
+      "grad_norm": 0.766511082649231,
+      "learning_rate": 9.464625937597377e-05,
+      "loss": 0.0853,
+      "step": 34420
+    },
+    {
+      "epoch": 2.2525351651946353,
+      "grad_norm": 0.9715924263000488,
+      "learning_rate": 9.464212306418194e-05,
+      "loss": 0.0931,
+      "step": 34430
+    },
+    {
+      "epoch": 2.253189401373896,
+      "grad_norm": 0.8225154280662537,
+      "learning_rate": 9.463798524560141e-05,
+      "loss": 0.0927,
+      "step": 34440
+    },
+    {
+      "epoch": 2.2538436375531568,
+      "grad_norm": 0.7842962145805359,
+      "learning_rate": 9.463384592037178e-05,
+      "loss": 0.0843,
+      "step": 34450
+    },
+    {
+      "epoch": 2.2544978737324173,
+      "grad_norm": 0.817205011844635,
+      "learning_rate": 9.46297050886328e-05,
+      "loss": 0.0868,
+      "step": 34460
+    },
+    {
+      "epoch": 2.2551521099116782,
+      "grad_norm": 0.8969591856002808,
+      "learning_rate": 9.462556275052425e-05,
+      "loss": 0.0888,
+      "step": 34470
+    },
+    {
+      "epoch": 2.2558063460909388,
+      "grad_norm": 0.924565851688385,
+      "learning_rate": 9.46214189061859e-05,
+      "loss": 0.0763,
+      "step": 34480
+    },
+    {
+      "epoch": 2.2564605822701997,
+      "grad_norm": 0.7565225958824158,
+      "learning_rate": 9.461727355575764e-05,
+      "loss": 0.0943,
+      "step": 34490
+    },
+    {
+      "epoch": 2.2571148184494603,
+      "grad_norm": 0.9678487181663513,
+      "learning_rate": 9.461312669937938e-05,
+      "loss": 0.1002,
+      "step": 34500
+    },
+    {
+      "epoch": 2.257769054628721,
+      "grad_norm": 1.0543937683105469,
+      "learning_rate": 9.460897833719111e-05,
+      "loss": 0.0908,
+      "step": 34510
+    },
+    {
+      "epoch": 2.2584232908079818,
+      "grad_norm": 0.8663365840911865,
+      "learning_rate": 9.460482846933283e-05,
+      "loss": 0.0793,
+      "step": 34520
+    },
+    {
+      "epoch": 2.2590775269872423,
+      "grad_norm": 0.9217035174369812,
+      "learning_rate": 9.460067709594459e-05,
+      "loss": 0.0862,
+      "step": 34530
+    },
+    {
+      "epoch": 2.2597317631665033,
+      "grad_norm": 0.8142322301864624,
+      "learning_rate": 9.459652421716654e-05,
+      "loss": 0.0895,
+      "step": 34540
+    },
+    {
+      "epoch": 2.260385999345764,
+      "grad_norm": 0.7949792146682739,
+      "learning_rate": 9.459236983313884e-05,
+      "loss": 0.0831,
+      "step": 34550
+    },
+    {
+      "epoch": 2.2610402355250248,
+      "grad_norm": 1.0977342128753662,
+      "learning_rate": 9.45882139440017e-05,
+      "loss": 0.079,
+      "step": 34560
+    },
+    {
+      "epoch": 2.2616944717042853,
+      "grad_norm": 0.8566722869873047,
+      "learning_rate": 9.458405654989542e-05,
+      "loss": 0.0799,
+      "step": 34570
+    },
+    {
+      "epoch": 2.262348707883546,
+      "grad_norm": 0.9690262079238892,
+      "learning_rate": 9.457989765096028e-05,
+      "loss": 0.0845,
+      "step": 34580
+    },
+    {
+      "epoch": 2.2630029440628068,
+      "grad_norm": 0.684556245803833,
+      "learning_rate": 9.45757372473367e-05,
+      "loss": 0.0934,
+      "step": 34590
+    },
+    {
+      "epoch": 2.2636571802420673,
+      "grad_norm": 0.804113507270813,
+      "learning_rate": 9.457157533916508e-05,
+      "loss": 0.0828,
+      "step": 34600
+    },
+    {
+      "epoch": 2.2643114164213283,
+      "grad_norm": 0.8007469177246094,
+      "learning_rate": 9.456741192658589e-05,
+      "loss": 0.0922,
+      "step": 34610
+    },
+    {
+      "epoch": 2.264965652600589,
+      "grad_norm": 1.1300758123397827,
+      "learning_rate": 9.456324700973966e-05,
+      "loss": 0.0841,
+      "step": 34620
+    },
+    {
+      "epoch": 2.2656198887798498,
+      "grad_norm": 0.7691265344619751,
+      "learning_rate": 9.4559080588767e-05,
+      "loss": 0.088,
+      "step": 34630
+    },
+    {
+      "epoch": 2.2662741249591103,
+      "grad_norm": 0.6707723140716553,
+      "learning_rate": 9.455491266380849e-05,
+      "loss": 0.0842,
+      "step": 34640
+    },
+    {
+      "epoch": 2.266928361138371,
+      "grad_norm": 0.848480224609375,
+      "learning_rate": 9.455074323500484e-05,
+      "loss": 0.0938,
+      "step": 34650
+    },
+    {
+      "epoch": 2.267582597317632,
+      "grad_norm": 0.9524247050285339,
+      "learning_rate": 9.454657230249675e-05,
+      "loss": 0.086,
+      "step": 34660
+    },
+    {
+      "epoch": 2.2682368334968923,
+      "grad_norm": 0.8271149396896362,
+      "learning_rate": 9.454239986642502e-05,
+      "loss": 0.086,
+      "step": 34670
+    },
+    {
+      "epoch": 2.2688910696761533,
+      "grad_norm": 0.9054349064826965,
+      "learning_rate": 9.453822592693049e-05,
+      "loss": 0.0883,
+      "step": 34680
+    },
+    {
+      "epoch": 2.269545305855414,
+      "grad_norm": 0.9008976221084595,
+      "learning_rate": 9.453405048415402e-05,
+      "loss": 0.084,
+      "step": 34690
+    },
+    {
+      "epoch": 2.2701995420346743,
+      "grad_norm": 0.7716291546821594,
+      "learning_rate": 9.452987353823654e-05,
+      "loss": 0.0858,
+      "step": 34700
+    },
+    {
+      "epoch": 2.2708537782139353,
+      "grad_norm": 0.7925845384597778,
+      "learning_rate": 9.452569508931908e-05,
+      "loss": 0.0862,
+      "step": 34710
+    },
+    {
+      "epoch": 2.271508014393196,
+      "grad_norm": 0.7443545460700989,
+      "learning_rate": 9.452151513754262e-05,
+      "loss": 0.0817,
+      "step": 34720
+    },
+    {
+      "epoch": 2.272162250572457,
+      "grad_norm": 0.7486448884010315,
+      "learning_rate": 9.451733368304825e-05,
+      "loss": 0.0779,
+      "step": 34730
+    },
+    {
+      "epoch": 2.2728164867517173,
+      "grad_norm": 0.8984254002571106,
+      "learning_rate": 9.451315072597713e-05,
+      "loss": 0.0854,
+      "step": 34740
+    },
+    {
+      "epoch": 2.273470722930978,
+      "grad_norm": 0.880130410194397,
+      "learning_rate": 9.450896626647041e-05,
+      "loss": 0.09,
+      "step": 34750
+    },
+    {
+      "epoch": 2.274124959110239,
+      "grad_norm": 0.8903000354766846,
+      "learning_rate": 9.450478030466938e-05,
+      "loss": 0.0876,
+      "step": 34760
+    },
+    {
+      "epoch": 2.2747791952894993,
+      "grad_norm": 0.8735268115997314,
+      "learning_rate": 9.450059284071529e-05,
+      "loss": 0.1018,
+      "step": 34770
+    },
+    {
+      "epoch": 2.2754334314687603,
+      "grad_norm": 0.7384045720100403,
+      "learning_rate": 9.449640387474948e-05,
+      "loss": 0.0849,
+      "step": 34780
+    },
+    {
+      "epoch": 2.276087667648021,
+      "grad_norm": 0.8934576511383057,
+      "learning_rate": 9.449221340691333e-05,
+      "loss": 0.0834,
+      "step": 34790
+    },
+    {
+      "epoch": 2.276741903827282,
+      "grad_norm": 1.060774564743042,
+      "learning_rate": 9.448802143734831e-05,
+      "loss": 0.0932,
+      "step": 34800
+    },
+    {
+      "epoch": 2.2773961400065423,
+      "grad_norm": 0.825049638748169,
+      "learning_rate": 9.448382796619589e-05,
+      "loss": 0.0854,
+      "step": 34810
+    },
+    {
+      "epoch": 2.278050376185803,
+      "grad_norm": 0.8421124815940857,
+      "learning_rate": 9.44796329935976e-05,
+      "loss": 0.0836,
+      "step": 34820
+    },
+    {
+      "epoch": 2.278704612365064,
+      "grad_norm": 0.7823123335838318,
+      "learning_rate": 9.447543651969506e-05,
+      "loss": 0.0852,
+      "step": 34830
+    },
+    {
+      "epoch": 2.2793588485443244,
+      "grad_norm": 1.0496047735214233,
+      "learning_rate": 9.447123854462989e-05,
+      "loss": 0.0891,
+      "step": 34840
+    },
+    {
+      "epoch": 2.2800130847235853,
+      "grad_norm": 0.8308643698692322,
+      "learning_rate": 9.44670390685438e-05,
+      "loss": 0.0831,
+      "step": 34850
+    },
+    {
+      "epoch": 2.280667320902846,
+      "grad_norm": 0.8488014936447144,
+      "learning_rate": 9.44628380915785e-05,
+      "loss": 0.0911,
+      "step": 34860
+    },
+    {
+      "epoch": 2.281321557082107,
+      "grad_norm": 0.9563519358634949,
+      "learning_rate": 9.445863561387582e-05,
+      "loss": 0.0957,
+      "step": 34870
+    },
+    {
+      "epoch": 2.2819757932613673,
+      "grad_norm": 0.8364535570144653,
+      "learning_rate": 9.44544316355776e-05,
+      "loss": 0.0821,
+      "step": 34880
+    },
+    {
+      "epoch": 2.282630029440628,
+      "grad_norm": 0.918013870716095,
+      "learning_rate": 9.445022615682571e-05,
+      "loss": 0.0789,
+      "step": 34890
+    },
+    {
+      "epoch": 2.283284265619889,
+      "grad_norm": 0.7931444048881531,
+      "learning_rate": 9.44460191777621e-05,
+      "loss": 0.0908,
+      "step": 34900
+    },
+    {
+      "epoch": 2.2839385017991494,
+      "grad_norm": 0.9308255314826965,
+      "learning_rate": 9.44418106985288e-05,
+      "loss": 0.0819,
+      "step": 34910
+    },
+    {
+      "epoch": 2.2845927379784103,
+      "grad_norm": 1.0205659866333008,
+      "learning_rate": 9.443760071926784e-05,
+      "loss": 0.0908,
+      "step": 34920
+    },
+    {
+      "epoch": 2.285246974157671,
+      "grad_norm": 0.8457900881767273,
+      "learning_rate": 9.44333892401213e-05,
+      "loss": 0.0996,
+      "step": 34930
+    },
+    {
+      "epoch": 2.285901210336932,
+      "grad_norm": 0.8027629852294922,
+      "learning_rate": 9.442917626123136e-05,
+      "loss": 0.0913,
+      "step": 34940
+    },
+    {
+      "epoch": 2.2865554465161924,
+      "grad_norm": 0.8995937705039978,
+      "learning_rate": 9.442496178274019e-05,
+      "loss": 0.0892,
+      "step": 34950
+    },
+    {
+      "epoch": 2.287209682695453,
+      "grad_norm": 0.9904996752738953,
+      "learning_rate": 9.442074580479004e-05,
+      "loss": 0.0921,
+      "step": 34960
+    },
+    {
+      "epoch": 2.287863918874714,
+      "grad_norm": 0.9785271286964417,
+      "learning_rate": 9.441652832752324e-05,
+      "loss": 0.0801,
+      "step": 34970
+    },
+    {
+      "epoch": 2.2885181550539744,
+      "grad_norm": 0.8374738097190857,
+      "learning_rate": 9.441230935108212e-05,
+      "loss": 0.0855,
+      "step": 34980
+    },
+    {
+      "epoch": 2.2891723912332353,
+      "grad_norm": 0.8071938157081604,
+      "learning_rate": 9.440808887560907e-05,
+      "loss": 0.0967,
+      "step": 34990
+    },
+    {
+      "epoch": 2.289826627412496,
+      "grad_norm": 1.0699573755264282,
+      "learning_rate": 9.440386690124656e-05,
+      "loss": 0.0865,
+      "step": 35000
+    },
+    {
+      "epoch": 2.290480863591757,
+      "grad_norm": 1.1547331809997559,
+      "learning_rate": 9.43996434281371e-05,
+      "loss": 0.085,
+      "step": 35010
+    },
+    {
+      "epoch": 2.2911350997710174,
+      "grad_norm": 0.8544806838035583,
+      "learning_rate": 9.439541845642322e-05,
+      "loss": 0.0808,
+      "step": 35020
+    },
+    {
+      "epoch": 2.291789335950278,
+      "grad_norm": 0.8750424385070801,
+      "learning_rate": 9.439119198624755e-05,
+      "loss": 0.0826,
+      "step": 35030
+    },
+    {
+      "epoch": 2.292443572129539,
+      "grad_norm": 1.0498063564300537,
+      "learning_rate": 9.438696401775271e-05,
+      "loss": 0.0912,
+      "step": 35040
+    },
+    {
+      "epoch": 2.2930978083087994,
+      "grad_norm": 0.7838863730430603,
+      "learning_rate": 9.438273455108144e-05,
+      "loss": 0.0985,
+      "step": 35050
+    },
+    {
+      "epoch": 2.2937520444880604,
+      "grad_norm": 0.7744349241256714,
+      "learning_rate": 9.437850358637648e-05,
+      "loss": 0.0799,
+      "step": 35060
+    },
+    {
+      "epoch": 2.294406280667321,
+      "grad_norm": 0.7421576380729675,
+      "learning_rate": 9.437427112378063e-05,
+      "loss": 0.0907,
+      "step": 35070
+    },
+    {
+      "epoch": 2.295060516846582,
+      "grad_norm": 0.7675076723098755,
+      "learning_rate": 9.437003716343676e-05,
+      "loss": 0.0768,
+      "step": 35080
+    },
+    {
+      "epoch": 2.2957147530258424,
+      "grad_norm": 0.6893305778503418,
+      "learning_rate": 9.436580170548777e-05,
+      "loss": 0.076,
+      "step": 35090
+    },
+    {
+      "epoch": 2.296368989205103,
+      "grad_norm": 0.7479074001312256,
+      "learning_rate": 9.436156475007662e-05,
+      "loss": 0.0824,
+      "step": 35100
+    },
+    {
+      "epoch": 2.297023225384364,
+      "grad_norm": 0.7704381346702576,
+      "learning_rate": 9.435732629734633e-05,
+      "loss": 0.0882,
+      "step": 35110
+    },
+    {
+      "epoch": 2.2976774615636244,
+      "grad_norm": 0.9246371984481812,
+      "learning_rate": 9.435308634743992e-05,
+      "loss": 0.0812,
+      "step": 35120
+    },
+    {
+      "epoch": 2.2983316977428854,
+      "grad_norm": 0.8029863834381104,
+      "learning_rate": 9.434884490050053e-05,
+      "loss": 0.0976,
+      "step": 35130
+    },
+    {
+      "epoch": 2.298985933922146,
+      "grad_norm": 0.8177743554115295,
+      "learning_rate": 9.434460195667133e-05,
+      "loss": 0.0784,
+      "step": 35140
+    },
+    {
+      "epoch": 2.2996401701014064,
+      "grad_norm": 0.8609619140625,
+      "learning_rate": 9.434035751609551e-05,
+      "loss": 0.1004,
+      "step": 35150
+    },
+    {
+      "epoch": 2.3002944062806674,
+      "grad_norm": 0.7648687958717346,
+      "learning_rate": 9.433611157891633e-05,
+      "loss": 0.085,
+      "step": 35160
+    },
+    {
+      "epoch": 2.300948642459928,
+      "grad_norm": 0.9224421977996826,
+      "learning_rate": 9.433186414527713e-05,
+      "loss": 0.0882,
+      "step": 35170
+    },
+    {
+      "epoch": 2.301602878639189,
+      "grad_norm": 0.793401837348938,
+      "learning_rate": 9.432761521532123e-05,
+      "loss": 0.0769,
+      "step": 35180
+    },
+    {
+      "epoch": 2.3022571148184494,
+      "grad_norm": 0.86360102891922,
+      "learning_rate": 9.432336478919206e-05,
+      "loss": 0.0851,
+      "step": 35190
+    },
+    {
+      "epoch": 2.30291135099771,
+      "grad_norm": 0.860308825969696,
+      "learning_rate": 9.43191128670331e-05,
+      "loss": 0.0844,
+      "step": 35200
+    },
+    {
+      "epoch": 2.303565587176971,
+      "grad_norm": 0.8408114314079285,
+      "learning_rate": 9.431485944898784e-05,
+      "loss": 0.0892,
+      "step": 35210
+    },
+    {
+      "epoch": 2.3042198233562314,
+      "grad_norm": 0.8600111603736877,
+      "learning_rate": 9.431060453519986e-05,
+      "loss": 0.0838,
+      "step": 35220
+    },
+    {
+      "epoch": 2.3048740595354924,
+      "grad_norm": 0.9598779082298279,
+      "learning_rate": 9.430634812581276e-05,
+      "loss": 0.0891,
+      "step": 35230
+    },
+    {
+      "epoch": 2.305528295714753,
+      "grad_norm": 0.9931259155273438,
+      "learning_rate": 9.430209022097023e-05,
+      "loss": 0.08,
+      "step": 35240
+    },
+    {
+      "epoch": 2.306182531894014,
+      "grad_norm": 0.687402069568634,
+      "learning_rate": 9.429783082081596e-05,
+      "loss": 0.0882,
+      "step": 35250
+    },
+    {
+      "epoch": 2.3068367680732744,
+      "grad_norm": 0.8022744059562683,
+      "learning_rate": 9.429356992549372e-05,
+      "loss": 0.0828,
+      "step": 35260
+    },
+    {
+      "epoch": 2.307491004252535,
+      "grad_norm": 0.9008263945579529,
+      "learning_rate": 9.428930753514734e-05,
+      "loss": 0.0775,
+      "step": 35270
+    },
+    {
+      "epoch": 2.308145240431796,
+      "grad_norm": 0.8905544877052307,
+      "learning_rate": 9.428504364992066e-05,
+      "loss": 0.0863,
+      "step": 35280
+    },
+    {
+      "epoch": 2.3087994766110564,
+      "grad_norm": 0.9496570825576782,
+      "learning_rate": 9.428077826995762e-05,
+      "loss": 0.0863,
+      "step": 35290
+    },
+    {
+      "epoch": 2.3094537127903174,
+      "grad_norm": 0.8719865679740906,
+      "learning_rate": 9.427651139540218e-05,
+      "loss": 0.0777,
+      "step": 35300
+    },
+    {
+      "epoch": 2.310107948969578,
+      "grad_norm": 1.0145576000213623,
+      "learning_rate": 9.427224302639837e-05,
+      "loss": 0.089,
+      "step": 35310
+    },
+    {
+      "epoch": 2.310762185148839,
+      "grad_norm": 0.8923918604850769,
+      "learning_rate": 9.426797316309026e-05,
+      "loss": 0.0766,
+      "step": 35320
+    },
+    {
+      "epoch": 2.3114164213280994,
+      "grad_norm": 0.7120718955993652,
+      "learning_rate": 9.426370180562195e-05,
+      "loss": 0.0789,
+      "step": 35330
+    },
+    {
+      "epoch": 2.31207065750736,
+      "grad_norm": 0.9214240312576294,
+      "learning_rate": 9.425942895413761e-05,
+      "loss": 0.082,
+      "step": 35340
+    },
+    {
+      "epoch": 2.312724893686621,
+      "grad_norm": 0.7993736863136292,
+      "learning_rate": 9.425515460878148e-05,
+      "loss": 0.078,
+      "step": 35350
+    },
+    {
+      "epoch": 2.3133791298658815,
+      "grad_norm": 0.7409219741821289,
+      "learning_rate": 9.42508787696978e-05,
+      "loss": 0.0859,
+      "step": 35360
+    },
+    {
+      "epoch": 2.3140333660451424,
+      "grad_norm": 0.8568753600120544,
+      "learning_rate": 9.424660143703092e-05,
+      "loss": 0.0974,
+      "step": 35370
+    },
+    {
+      "epoch": 2.314687602224403,
+      "grad_norm": 0.8471707701683044,
+      "learning_rate": 9.424232261092521e-05,
+      "loss": 0.0926,
+      "step": 35380
+    },
+    {
+      "epoch": 2.315341838403664,
+      "grad_norm": 0.9042620658874512,
+      "learning_rate": 9.423804229152507e-05,
+      "loss": 0.084,
+      "step": 35390
+    },
+    {
+      "epoch": 2.3159960745829244,
+      "grad_norm": 0.9078444838523865,
+      "learning_rate": 9.423376047897499e-05,
+      "loss": 0.0903,
+      "step": 35400
+    },
+    {
+      "epoch": 2.316650310762185,
+      "grad_norm": 0.8419270515441895,
+      "learning_rate": 9.422947717341948e-05,
+      "loss": 0.0791,
+      "step": 35410
+    },
+    {
+      "epoch": 2.317304546941446,
+      "grad_norm": 0.8305777907371521,
+      "learning_rate": 9.422519237500313e-05,
+      "loss": 0.092,
+      "step": 35420
+    },
+    {
+      "epoch": 2.3179587831207065,
+      "grad_norm": 0.7356106638908386,
+      "learning_rate": 9.422090608387055e-05,
+      "loss": 0.0809,
+      "step": 35430
+    },
+    {
+      "epoch": 2.3186130192999674,
+      "grad_norm": 0.9605696201324463,
+      "learning_rate": 9.421661830016642e-05,
+      "loss": 0.0746,
+      "step": 35440
+    },
+    {
+      "epoch": 2.319267255479228,
+      "grad_norm": 1.005142092704773,
+      "learning_rate": 9.421232902403545e-05,
+      "loss": 0.0917,
+      "step": 35450
+    },
+    {
+      "epoch": 2.319921491658489,
+      "grad_norm": 0.965522289276123,
+      "learning_rate": 9.420803825562243e-05,
+      "loss": 0.0977,
+      "step": 35460
+    },
+    {
+      "epoch": 2.3205757278377495,
+      "grad_norm": 0.860283613204956,
+      "learning_rate": 9.420374599507217e-05,
+      "loss": 0.0872,
+      "step": 35470
+    },
+    {
+      "epoch": 2.32122996401701,
+      "grad_norm": 0.834602952003479,
+      "learning_rate": 9.419945224252955e-05,
+      "loss": 0.1007,
+      "step": 35480
+    },
+    {
+      "epoch": 2.321884200196271,
+      "grad_norm": 0.8806980848312378,
+      "learning_rate": 9.419515699813952e-05,
+      "loss": 0.0795,
+      "step": 35490
+    },
+    {
+      "epoch": 2.3225384363755315,
+      "grad_norm": 1.0480656623840332,
+      "learning_rate": 9.419086026204703e-05,
+      "loss": 0.0898,
+      "step": 35500
+    },
+    {
+      "epoch": 2.3231926725547924,
+      "grad_norm": 0.8428975343704224,
+      "learning_rate": 9.41865620343971e-05,
+      "loss": 0.0792,
+      "step": 35510
+    },
+    {
+      "epoch": 2.323846908734053,
+      "grad_norm": 0.9427582621574402,
+      "learning_rate": 9.418226231533482e-05,
+      "loss": 0.0796,
+      "step": 35520
+    },
+    {
+      "epoch": 2.324501144913314,
+      "grad_norm": 0.7207076549530029,
+      "learning_rate": 9.417796110500532e-05,
+      "loss": 0.0826,
+      "step": 35530
+    },
+    {
+      "epoch": 2.3251553810925745,
+      "grad_norm": 0.9341986179351807,
+      "learning_rate": 9.417365840355377e-05,
+      "loss": 0.0965,
+      "step": 35540
+    },
+    {
+      "epoch": 2.325809617271835,
+      "grad_norm": 0.8259122371673584,
+      "learning_rate": 9.416935421112541e-05,
+      "loss": 0.0909,
+      "step": 35550
+    },
+    {
+      "epoch": 2.326463853451096,
+      "grad_norm": 0.8562160134315491,
+      "learning_rate": 9.41650485278655e-05,
+      "loss": 0.0918,
+      "step": 35560
+    },
+    {
+      "epoch": 2.3271180896303565,
+      "grad_norm": 0.9982649683952332,
+      "learning_rate": 9.416074135391937e-05,
+      "loss": 0.0832,
+      "step": 35570
+    },
+    {
+      "epoch": 2.3277723258096175,
+      "grad_norm": 0.9082480072975159,
+      "learning_rate": 9.415643268943239e-05,
+      "loss": 0.0969,
+      "step": 35580
+    },
+    {
+      "epoch": 2.328426561988878,
+      "grad_norm": 0.8344404101371765,
+      "learning_rate": 9.415212253455004e-05,
+      "loss": 0.0815,
+      "step": 35590
+    },
+    {
+      "epoch": 2.3290807981681385,
+      "grad_norm": 0.8839426040649414,
+      "learning_rate": 9.414781088941772e-05,
+      "loss": 0.0899,
+      "step": 35600
+    },
+    {
+      "epoch": 2.3297350343473995,
+      "grad_norm": 0.7677926421165466,
+      "learning_rate": 9.414349775418104e-05,
+      "loss": 0.0774,
+      "step": 35610
+    },
+    {
+      "epoch": 2.33038927052666,
+      "grad_norm": 1.086987853050232,
+      "learning_rate": 9.413918312898551e-05,
+      "loss": 0.0909,
+      "step": 35620
+    },
+    {
+      "epoch": 2.331043506705921,
+      "grad_norm": 0.8547496795654297,
+      "learning_rate": 9.41348670139768e-05,
+      "loss": 0.0781,
+      "step": 35630
+    },
+    {
+      "epoch": 2.3316977428851815,
+      "grad_norm": 0.8789479732513428,
+      "learning_rate": 9.413054940930057e-05,
+      "loss": 0.0926,
+      "step": 35640
+    },
+    {
+      "epoch": 2.332351979064442,
+      "grad_norm": 0.8583962917327881,
+      "learning_rate": 9.412623031510257e-05,
+      "loss": 0.0854,
+      "step": 35650
+    },
+    {
+      "epoch": 2.333006215243703,
+      "grad_norm": 0.8400371670722961,
+      "learning_rate": 9.412190973152858e-05,
+      "loss": 0.0908,
+      "step": 35660
+    },
+    {
+      "epoch": 2.3336604514229635,
+      "grad_norm": 0.8149643540382385,
+      "learning_rate": 9.411758765872441e-05,
+      "loss": 0.0837,
+      "step": 35670
+    },
+    {
+      "epoch": 2.3343146876022245,
+      "grad_norm": 0.7172325849533081,
+      "learning_rate": 9.411326409683596e-05,
+      "loss": 0.0828,
+      "step": 35680
+    },
+    {
+      "epoch": 2.334968923781485,
+      "grad_norm": 0.9276403188705444,
+      "learning_rate": 9.410893904600917e-05,
+      "loss": 0.083,
+      "step": 35690
+    },
+    {
+      "epoch": 2.335623159960746,
+      "grad_norm": 0.7993019223213196,
+      "learning_rate": 9.410461250638997e-05,
+      "loss": 0.0831,
+      "step": 35700
+    },
+    {
+      "epoch": 2.3362773961400065,
+      "grad_norm": 0.7336340546607971,
+      "learning_rate": 9.410028447812447e-05,
+      "loss": 0.0866,
+      "step": 35710
+    },
+    {
+      "epoch": 2.336931632319267,
+      "grad_norm": 0.9019221067428589,
+      "learning_rate": 9.409595496135869e-05,
+      "loss": 0.0892,
+      "step": 35720
+    },
+    {
+      "epoch": 2.337585868498528,
+      "grad_norm": 0.8154904842376709,
+      "learning_rate": 9.409162395623879e-05,
+      "loss": 0.0865,
+      "step": 35730
+    },
+    {
+      "epoch": 2.3382401046777885,
+      "grad_norm": 0.8226253986358643,
+      "learning_rate": 9.408729146291093e-05,
+      "loss": 0.0901,
+      "step": 35740
+    },
+    {
+      "epoch": 2.3388943408570495,
+      "grad_norm": 0.7311269640922546,
+      "learning_rate": 9.408295748152138e-05,
+      "loss": 0.0889,
+      "step": 35750
+    },
+    {
+      "epoch": 2.33954857703631,
+      "grad_norm": 0.8451585173606873,
+      "learning_rate": 9.40786220122164e-05,
+      "loss": 0.0848,
+      "step": 35760
+    },
+    {
+      "epoch": 2.340202813215571,
+      "grad_norm": 1.0941405296325684,
+      "learning_rate": 9.407428505514233e-05,
+      "loss": 0.0871,
+      "step": 35770
+    },
+    {
+      "epoch": 2.3408570493948315,
+      "grad_norm": 0.9233847260475159,
+      "learning_rate": 9.406994661044554e-05,
+      "loss": 0.0976,
+      "step": 35780
+    },
+    {
+      "epoch": 2.341511285574092,
+      "grad_norm": 0.8936124444007874,
+      "learning_rate": 9.406560667827248e-05,
+      "loss": 0.0839,
+      "step": 35790
+    },
+    {
+      "epoch": 2.342165521753353,
+      "grad_norm": 0.9002397060394287,
+      "learning_rate": 9.406126525876963e-05,
+      "loss": 0.0781,
+      "step": 35800
+    },
+    {
+      "epoch": 2.3428197579326135,
+      "grad_norm": 0.7002133131027222,
+      "learning_rate": 9.405692235208353e-05,
+      "loss": 0.0845,
+      "step": 35810
+    },
+    {
+      "epoch": 2.3434739941118745,
+      "grad_norm": 0.8459717035293579,
+      "learning_rate": 9.405257795836074e-05,
+      "loss": 0.0769,
+      "step": 35820
+    },
+    {
+      "epoch": 2.344128230291135,
+      "grad_norm": 1.1364301443099976,
+      "learning_rate": 9.404823207774791e-05,
+      "loss": 0.0878,
+      "step": 35830
+    },
+    {
+      "epoch": 2.344782466470396,
+      "grad_norm": 0.8352745771408081,
+      "learning_rate": 9.404388471039173e-05,
+      "loss": 0.0828,
+      "step": 35840
+    },
+    {
+      "epoch": 2.3454367026496565,
+      "grad_norm": 0.8302096128463745,
+      "learning_rate": 9.403953585643895e-05,
+      "loss": 0.0885,
+      "step": 35850
+    },
+    {
+      "epoch": 2.346090938828917,
+      "grad_norm": 0.7666718363761902,
+      "learning_rate": 9.403518551603632e-05,
+      "loss": 0.0813,
+      "step": 35860
+    },
+    {
+      "epoch": 2.346745175008178,
+      "grad_norm": 0.9522207975387573,
+      "learning_rate": 9.40308336893307e-05,
+      "loss": 0.0909,
+      "step": 35870
+    },
+    {
+      "epoch": 2.3473994111874386,
+      "grad_norm": 0.7051165699958801,
+      "learning_rate": 9.402648037646895e-05,
+      "loss": 0.0871,
+      "step": 35880
+    },
+    {
+      "epoch": 2.3480536473666995,
+      "grad_norm": 0.7525814771652222,
+      "learning_rate": 9.402212557759805e-05,
+      "loss": 0.09,
+      "step": 35890
+    },
+    {
+      "epoch": 2.34870788354596,
+      "grad_norm": 1.1107085943222046,
+      "learning_rate": 9.401776929286494e-05,
+      "loss": 0.083,
+      "step": 35900
+    },
+    {
+      "epoch": 2.349362119725221,
+      "grad_norm": 0.7754766941070557,
+      "learning_rate": 9.401341152241668e-05,
+      "loss": 0.0819,
+      "step": 35910
+    },
+    {
+      "epoch": 2.3500163559044815,
+      "grad_norm": 0.9919909834861755,
+      "learning_rate": 9.400905226640036e-05,
+      "loss": 0.0859,
+      "step": 35920
+    },
+    {
+      "epoch": 2.350670592083742,
+      "grad_norm": 0.7626148462295532,
+      "learning_rate": 9.40046915249631e-05,
+      "loss": 0.0889,
+      "step": 35930
+    },
+    {
+      "epoch": 2.351324828263003,
+      "grad_norm": 0.8603907823562622,
+      "learning_rate": 9.40003292982521e-05,
+      "loss": 0.0847,
+      "step": 35940
+    },
+    {
+      "epoch": 2.3519790644422636,
+      "grad_norm": 0.8849925994873047,
+      "learning_rate": 9.399596558641459e-05,
+      "loss": 0.0782,
+      "step": 35950
+    },
+    {
+      "epoch": 2.3526333006215245,
+      "grad_norm": 0.9044029712677002,
+      "learning_rate": 9.399160038959785e-05,
+      "loss": 0.0841,
+      "step": 35960
+    },
+    {
+      "epoch": 2.353287536800785,
+      "grad_norm": 0.7774613499641418,
+      "learning_rate": 9.398723370794923e-05,
+      "loss": 0.0725,
+      "step": 35970
+    },
+    {
+      "epoch": 2.353941772980046,
+      "grad_norm": 0.7712434530258179,
+      "learning_rate": 9.398286554161612e-05,
+      "loss": 0.0855,
+      "step": 35980
+    },
+    {
+      "epoch": 2.3545960091593066,
+      "grad_norm": 0.8731211423873901,
+      "learning_rate": 9.397849589074593e-05,
+      "loss": 0.0941,
+      "step": 35990
+    },
+    {
+      "epoch": 2.355250245338567,
+      "grad_norm": 0.8799338340759277,
+      "learning_rate": 9.397412475548618e-05,
+      "loss": 0.0835,
+      "step": 36000
+    },
+    {
+      "epoch": 2.355904481517828,
+      "grad_norm": 0.8053388595581055,
+      "learning_rate": 9.396975213598439e-05,
+      "loss": 0.0872,
+      "step": 36010
+    },
+    {
+      "epoch": 2.3565587176970886,
+      "grad_norm": 0.8215650320053101,
+      "learning_rate": 9.396537803238815e-05,
+      "loss": 0.0894,
+      "step": 36020
+    },
+    {
+      "epoch": 2.3572129538763495,
+      "grad_norm": 0.7709128856658936,
+      "learning_rate": 9.39610024448451e-05,
+      "loss": 0.0852,
+      "step": 36030
+    },
+    {
+      "epoch": 2.35786719005561,
+      "grad_norm": 0.8839540481567383,
+      "learning_rate": 9.395662537350292e-05,
+      "loss": 0.0874,
+      "step": 36040
+    },
+    {
+      "epoch": 2.3585214262348706,
+      "grad_norm": 0.8985713124275208,
+      "learning_rate": 9.395224681850935e-05,
+      "loss": 0.0774,
+      "step": 36050
+    },
+    {
+      "epoch": 2.3591756624141316,
+      "grad_norm": 0.7541700601577759,
+      "learning_rate": 9.39478667800122e-05,
+      "loss": 0.0782,
+      "step": 36060
+    },
+    {
+      "epoch": 2.359829898593392,
+      "grad_norm": 0.6992083191871643,
+      "learning_rate": 9.394348525815928e-05,
+      "loss": 0.0821,
+      "step": 36070
+    },
+    {
+      "epoch": 2.360484134772653,
+      "grad_norm": 0.9433765411376953,
+      "learning_rate": 9.393910225309848e-05,
+      "loss": 0.0825,
+      "step": 36080
+    },
+    {
+      "epoch": 2.3611383709519136,
+      "grad_norm": 0.892088770866394,
+      "learning_rate": 9.393471776497776e-05,
+      "loss": 0.083,
+      "step": 36090
+    },
+    {
+      "epoch": 2.361792607131174,
+      "grad_norm": 0.7801177501678467,
+      "learning_rate": 9.393033179394506e-05,
+      "loss": 0.0998,
+      "step": 36100
+    },
+    {
+      "epoch": 2.362446843310435,
+      "grad_norm": 0.8416787981987,
+      "learning_rate": 9.392594434014847e-05,
+      "loss": 0.0978,
+      "step": 36110
+    },
+    {
+      "epoch": 2.3631010794896956,
+      "grad_norm": 0.8340370059013367,
+      "learning_rate": 9.392155540373606e-05,
+      "loss": 0.0899,
+      "step": 36120
+    },
+    {
+      "epoch": 2.3637553156689566,
+      "grad_norm": 0.8223138451576233,
+      "learning_rate": 9.391716498485597e-05,
+      "loss": 0.0878,
+      "step": 36130
+    },
+    {
+      "epoch": 2.364409551848217,
+      "grad_norm": 1.106229305267334,
+      "learning_rate": 9.391277308365638e-05,
+      "loss": 0.0879,
+      "step": 36140
+    },
+    {
+      "epoch": 2.365063788027478,
+      "grad_norm": 0.8265957832336426,
+      "learning_rate": 9.390837970028553e-05,
+      "loss": 0.0796,
+      "step": 36150
+    },
+    {
+      "epoch": 2.3657180242067386,
+      "grad_norm": 0.8237302303314209,
+      "learning_rate": 9.390398483489171e-05,
+      "loss": 0.0843,
+      "step": 36160
+    },
+    {
+      "epoch": 2.366372260385999,
+      "grad_norm": 0.819557249546051,
+      "learning_rate": 9.389958848762327e-05,
+      "loss": 0.0811,
+      "step": 36170
+    },
+    {
+      "epoch": 2.36702649656526,
+      "grad_norm": 0.9804088473320007,
+      "learning_rate": 9.389519065862858e-05,
+      "loss": 0.0887,
+      "step": 36180
+    },
+    {
+      "epoch": 2.3676807327445206,
+      "grad_norm": 1.014117956161499,
+      "learning_rate": 9.389079134805609e-05,
+      "loss": 0.0732,
+      "step": 36190
+    },
+    {
+      "epoch": 2.3683349689237816,
+      "grad_norm": 0.8419691920280457,
+      "learning_rate": 9.388639055605428e-05,
+      "loss": 0.0909,
+      "step": 36200
+    },
+    {
+      "epoch": 2.368989205103042,
+      "grad_norm": 0.9921837449073792,
+      "learning_rate": 9.388198828277169e-05,
+      "loss": 0.0783,
+      "step": 36210
+    },
+    {
+      "epoch": 2.369643441282303,
+      "grad_norm": 0.9146479964256287,
+      "learning_rate": 9.387758452835692e-05,
+      "loss": 0.078,
+      "step": 36220
+    },
+    {
+      "epoch": 2.3702976774615636,
+      "grad_norm": 0.7768714427947998,
+      "learning_rate": 9.387317929295859e-05,
+      "loss": 0.084,
+      "step": 36230
+    },
+    {
+      "epoch": 2.370951913640824,
+      "grad_norm": 0.8944301009178162,
+      "learning_rate": 9.38687725767254e-05,
+      "loss": 0.0873,
+      "step": 36240
+    },
+    {
+      "epoch": 2.371606149820085,
+      "grad_norm": 0.7859957814216614,
+      "learning_rate": 9.38643643798061e-05,
+      "loss": 0.0849,
+      "step": 36250
+    },
+    {
+      "epoch": 2.3722603859993456,
+      "grad_norm": 0.8967961668968201,
+      "learning_rate": 9.385995470234944e-05,
+      "loss": 0.075,
+      "step": 36260
+    },
+    {
+      "epoch": 2.3729146221786066,
+      "grad_norm": 0.7501564621925354,
+      "learning_rate": 9.38555435445043e-05,
+      "loss": 0.093,
+      "step": 36270
+    },
+    {
+      "epoch": 2.373568858357867,
+      "grad_norm": 0.7810158729553223,
+      "learning_rate": 9.385113090641953e-05,
+      "loss": 0.0823,
+      "step": 36280
+    },
+    {
+      "epoch": 2.374223094537128,
+      "grad_norm": 0.8977556228637695,
+      "learning_rate": 9.38467167882441e-05,
+      "loss": 0.0823,
+      "step": 36290
+    },
+    {
+      "epoch": 2.3748773307163886,
+      "grad_norm": 0.9205357432365417,
+      "learning_rate": 9.384230119012698e-05,
+      "loss": 0.0869,
+      "step": 36300
+    },
+    {
+      "epoch": 2.375531566895649,
+      "grad_norm": 0.7959444522857666,
+      "learning_rate": 9.383788411221724e-05,
+      "loss": 0.0865,
+      "step": 36310
+    },
+    {
+      "epoch": 2.37618580307491,
+      "grad_norm": 0.7368982434272766,
+      "learning_rate": 9.383346555466392e-05,
+      "loss": 0.0879,
+      "step": 36320
+    },
+    {
+      "epoch": 2.3768400392541706,
+      "grad_norm": 1.0566291809082031,
+      "learning_rate": 9.382904551761618e-05,
+      "loss": 0.0976,
+      "step": 36330
+    },
+    {
+      "epoch": 2.3774942754334316,
+      "grad_norm": 0.7619825601577759,
+      "learning_rate": 9.38246240012232e-05,
+      "loss": 0.0847,
+      "step": 36340
+    },
+    {
+      "epoch": 2.378148511612692,
+      "grad_norm": 0.7798367738723755,
+      "learning_rate": 9.382020100563425e-05,
+      "loss": 0.0897,
+      "step": 36350
+    },
+    {
+      "epoch": 2.378802747791953,
+      "grad_norm": 0.8497381210327148,
+      "learning_rate": 9.381577653099858e-05,
+      "loss": 0.0823,
+      "step": 36360
+    },
+    {
+      "epoch": 2.3794569839712136,
+      "grad_norm": 0.8219572901725769,
+      "learning_rate": 9.381135057746552e-05,
+      "loss": 0.0889,
+      "step": 36370
+    },
+    {
+      "epoch": 2.380111220150474,
+      "grad_norm": 0.8166351914405823,
+      "learning_rate": 9.380692314518451e-05,
+      "loss": 0.0907,
+      "step": 36380
+    },
+    {
+      "epoch": 2.380765456329735,
+      "grad_norm": 1.0758252143859863,
+      "learning_rate": 9.380249423430494e-05,
+      "loss": 0.096,
+      "step": 36390
+    },
+    {
+      "epoch": 2.3814196925089957,
+      "grad_norm": 0.8256650567054749,
+      "learning_rate": 9.379806384497633e-05,
+      "loss": 0.0718,
+      "step": 36400
+    },
+    {
+      "epoch": 2.3820739286882566,
+      "grad_norm": 0.8822137117385864,
+      "learning_rate": 9.379363197734818e-05,
+      "loss": 0.095,
+      "step": 36410
+    },
+    {
+      "epoch": 2.382728164867517,
+      "grad_norm": 0.9527086615562439,
+      "learning_rate": 9.37891986315701e-05,
+      "loss": 0.0912,
+      "step": 36420
+    },
+    {
+      "epoch": 2.383382401046778,
+      "grad_norm": 0.8427805304527283,
+      "learning_rate": 9.378476380779174e-05,
+      "loss": 0.0898,
+      "step": 36430
+    },
+    {
+      "epoch": 2.3840366372260386,
+      "grad_norm": 0.8125025629997253,
+      "learning_rate": 9.378032750616277e-05,
+      "loss": 0.0719,
+      "step": 36440
+    },
+    {
+      "epoch": 2.384690873405299,
+      "grad_norm": 0.7970302700996399,
+      "learning_rate": 9.377588972683292e-05,
+      "loss": 0.0855,
+      "step": 36450
+    },
+    {
+      "epoch": 2.38534510958456,
+      "grad_norm": 0.8656693696975708,
+      "learning_rate": 9.377145046995198e-05,
+      "loss": 0.0804,
+      "step": 36460
+    },
+    {
+      "epoch": 2.3859993457638207,
+      "grad_norm": 0.7493268251419067,
+      "learning_rate": 9.37670097356698e-05,
+      "loss": 0.0845,
+      "step": 36470
+    },
+    {
+      "epoch": 2.3866535819430816,
+      "grad_norm": 0.8390613794326782,
+      "learning_rate": 9.376256752413626e-05,
+      "loss": 0.0772,
+      "step": 36480
+    },
+    {
+      "epoch": 2.387307818122342,
+      "grad_norm": 0.8210413455963135,
+      "learning_rate": 9.37581238355013e-05,
+      "loss": 0.0785,
+      "step": 36490
+    },
+    {
+      "epoch": 2.3879620543016027,
+      "grad_norm": 0.8246368765830994,
+      "learning_rate": 9.375367866991488e-05,
+      "loss": 0.0812,
+      "step": 36500
+    },
+    {
+      "epoch": 2.3886162904808637,
+      "grad_norm": 0.7549490332603455,
+      "learning_rate": 9.374923202752707e-05,
+      "loss": 0.0878,
+      "step": 36510
+    },
+    {
+      "epoch": 2.389270526660124,
+      "grad_norm": 0.8099923729896545,
+      "learning_rate": 9.374478390848794e-05,
+      "loss": 0.087,
+      "step": 36520
+    },
+    {
+      "epoch": 2.389924762839385,
+      "grad_norm": 0.8852878212928772,
+      "learning_rate": 9.374033431294763e-05,
+      "loss": 0.0897,
+      "step": 36530
+    },
+    {
+      "epoch": 2.3905789990186457,
+      "grad_norm": 0.9920669198036194,
+      "learning_rate": 9.373588324105634e-05,
+      "loss": 0.0871,
+      "step": 36540
+    },
+    {
+      "epoch": 2.391233235197906,
+      "grad_norm": 0.8708438873291016,
+      "learning_rate": 9.373143069296426e-05,
+      "loss": 0.0796,
+      "step": 36550
+    },
+    {
+      "epoch": 2.391887471377167,
+      "grad_norm": 0.751469612121582,
+      "learning_rate": 9.372697666882171e-05,
+      "loss": 0.0851,
+      "step": 36560
+    },
+    {
+      "epoch": 2.3925417075564277,
+      "grad_norm": 0.8722557425498962,
+      "learning_rate": 9.372252116877903e-05,
+      "loss": 0.0951,
+      "step": 36570
+    },
+    {
+      "epoch": 2.3931959437356887,
+      "grad_norm": 0.844551146030426,
+      "learning_rate": 9.371806419298659e-05,
+      "loss": 0.08,
+      "step": 36580
+    },
+    {
+      "epoch": 2.393850179914949,
+      "grad_norm": 0.8585530519485474,
+      "learning_rate": 9.371360574159483e-05,
+      "loss": 0.086,
+      "step": 36590
+    },
+    {
+      "epoch": 2.39450441609421,
+      "grad_norm": 0.9330928921699524,
+      "learning_rate": 9.370914581475423e-05,
+      "loss": 0.0793,
+      "step": 36600
+    },
+    {
+      "epoch": 2.3951586522734707,
+      "grad_norm": 1.0671617984771729,
+      "learning_rate": 9.370468441261532e-05,
+      "loss": 0.0737,
+      "step": 36610
+    },
+    {
+      "epoch": 2.395812888452731,
+      "grad_norm": 0.8291199803352356,
+      "learning_rate": 9.370022153532871e-05,
+      "loss": 0.0971,
+      "step": 36620
+    },
+    {
+      "epoch": 2.396467124631992,
+      "grad_norm": 1.0896260738372803,
+      "learning_rate": 9.3695757183045e-05,
+      "loss": 0.0821,
+      "step": 36630
+    },
+    {
+      "epoch": 2.3971213608112527,
+      "grad_norm": 0.7931303977966309,
+      "learning_rate": 9.369129135591491e-05,
+      "loss": 0.0857,
+      "step": 36640
+    },
+    {
+      "epoch": 2.3977755969905137,
+      "grad_norm": 0.9220706224441528,
+      "learning_rate": 9.368682405408912e-05,
+      "loss": 0.0844,
+      "step": 36650
+    },
+    {
+      "epoch": 2.398429833169774,
+      "grad_norm": 0.9693981409072876,
+      "learning_rate": 9.368235527771847e-05,
+      "loss": 0.091,
+      "step": 36660
+    },
+    {
+      "epoch": 2.399084069349035,
+      "grad_norm": 0.8742493987083435,
+      "learning_rate": 9.367788502695376e-05,
+      "loss": 0.0868,
+      "step": 36670
+    },
+    {
+      "epoch": 2.3997383055282957,
+      "grad_norm": 0.8460095524787903,
+      "learning_rate": 9.367341330194587e-05,
+      "loss": 0.0846,
+      "step": 36680
+    },
+    {
+      "epoch": 2.4003925417075562,
+      "grad_norm": 0.883783757686615,
+      "learning_rate": 9.366894010284576e-05,
+      "loss": 0.0819,
+      "step": 36690
+    },
+    {
+      "epoch": 2.401046777886817,
+      "grad_norm": 0.8907153606414795,
+      "learning_rate": 9.366446542980439e-05,
+      "loss": 0.0831,
+      "step": 36700
+    },
+    {
+      "epoch": 2.4017010140660777,
+      "grad_norm": 0.8098263144493103,
+      "learning_rate": 9.36599892829728e-05,
+      "loss": 0.0865,
+      "step": 36710
+    },
+    {
+      "epoch": 2.4023552502453387,
+      "grad_norm": 0.9012731909751892,
+      "learning_rate": 9.365551166250206e-05,
+      "loss": 0.0835,
+      "step": 36720
+    },
+    {
+      "epoch": 2.403009486424599,
+      "grad_norm": 0.817130446434021,
+      "learning_rate": 9.365103256854332e-05,
+      "loss": 0.0877,
+      "step": 36730
+    },
+    {
+      "epoch": 2.40366372260386,
+      "grad_norm": 1.0334715843200684,
+      "learning_rate": 9.364655200124775e-05,
+      "loss": 0.0905,
+      "step": 36740
+    },
+    {
+      "epoch": 2.4043179587831207,
+      "grad_norm": 0.6475959420204163,
+      "learning_rate": 9.364206996076659e-05,
+      "loss": 0.0826,
+      "step": 36750
+    },
+    {
+      "epoch": 2.4049721949623812,
+      "grad_norm": 1.0449169874191284,
+      "learning_rate": 9.36375864472511e-05,
+      "loss": 0.0842,
+      "step": 36760
+    },
+    {
+      "epoch": 2.405626431141642,
+      "grad_norm": 0.8349801898002625,
+      "learning_rate": 9.363310146085262e-05,
+      "loss": 0.089,
+      "step": 36770
+    },
+    {
+      "epoch": 2.4062806673209027,
+      "grad_norm": 0.9147889614105225,
+      "learning_rate": 9.362861500172255e-05,
+      "loss": 0.082,
+      "step": 36780
+    },
+    {
+      "epoch": 2.4069349035001637,
+      "grad_norm": 0.7441397905349731,
+      "learning_rate": 9.362412707001229e-05,
+      "loss": 0.0753,
+      "step": 36790
+    },
+    {
+      "epoch": 2.4075891396794242,
+      "grad_norm": 1.0056695938110352,
+      "learning_rate": 9.361963766587334e-05,
+      "loss": 0.08,
+      "step": 36800
+    },
+    {
+      "epoch": 2.408243375858685,
+      "grad_norm": 0.7822602987289429,
+      "learning_rate": 9.361514678945722e-05,
+      "loss": 0.0866,
+      "step": 36810
+    },
+    {
+      "epoch": 2.4088976120379457,
+      "grad_norm": 0.8155549764633179,
+      "learning_rate": 9.36106544409155e-05,
+      "loss": 0.0787,
+      "step": 36820
+    },
+    {
+      "epoch": 2.4095518482172062,
+      "grad_norm": 0.7978348135948181,
+      "learning_rate": 9.360616062039985e-05,
+      "loss": 0.0847,
+      "step": 36830
+    },
+    {
+      "epoch": 2.410206084396467,
+      "grad_norm": 0.867654025554657,
+      "learning_rate": 9.360166532806189e-05,
+      "loss": 0.0844,
+      "step": 36840
+    },
+    {
+      "epoch": 2.4108603205757277,
+      "grad_norm": 0.975135087966919,
+      "learning_rate": 9.359716856405339e-05,
+      "loss": 0.0915,
+      "step": 36850
+    },
+    {
+      "epoch": 2.4115145567549887,
+      "grad_norm": 0.7564871907234192,
+      "learning_rate": 9.359267032852609e-05,
+      "loss": 0.0968,
+      "step": 36860
+    },
+    {
+      "epoch": 2.4121687929342492,
+      "grad_norm": 0.8742430210113525,
+      "learning_rate": 9.358817062163188e-05,
+      "loss": 0.0793,
+      "step": 36870
+    },
+    {
+      "epoch": 2.41282302911351,
+      "grad_norm": 0.8599095344543457,
+      "learning_rate": 9.358366944352258e-05,
+      "loss": 0.0825,
+      "step": 36880
+    },
+    {
+      "epoch": 2.4134772652927707,
+      "grad_norm": 0.6971985101699829,
+      "learning_rate": 9.357916679435012e-05,
+      "loss": 0.0813,
+      "step": 36890
+    },
+    {
+      "epoch": 2.4141315014720313,
+      "grad_norm": 0.9535905122756958,
+      "learning_rate": 9.357466267426649e-05,
+      "loss": 0.0847,
+      "step": 36900
+    },
+    {
+      "epoch": 2.4147857376512922,
+      "grad_norm": 0.9580764174461365,
+      "learning_rate": 9.357015708342373e-05,
+      "loss": 0.0974,
+      "step": 36910
+    },
+    {
+      "epoch": 2.4154399738305528,
+      "grad_norm": 0.9554745554924011,
+      "learning_rate": 9.35656500219739e-05,
+      "loss": 0.08,
+      "step": 36920
+    },
+    {
+      "epoch": 2.4160942100098137,
+      "grad_norm": 0.7604560852050781,
+      "learning_rate": 9.356114149006911e-05,
+      "loss": 0.0854,
+      "step": 36930
+    },
+    {
+      "epoch": 2.4167484461890743,
+      "grad_norm": 0.876243531703949,
+      "learning_rate": 9.355663148786158e-05,
+      "loss": 0.0835,
+      "step": 36940
+    },
+    {
+      "epoch": 2.4174026823683348,
+      "grad_norm": 0.9993807673454285,
+      "learning_rate": 9.355212001550349e-05,
+      "loss": 0.087,
+      "step": 36950
+    },
+    {
+      "epoch": 2.4180569185475957,
+      "grad_norm": 0.934245228767395,
+      "learning_rate": 9.354760707314713e-05,
+      "loss": 0.0764,
+      "step": 36960
+    },
+    {
+      "epoch": 2.4187111547268563,
+      "grad_norm": 0.8030011057853699,
+      "learning_rate": 9.354309266094482e-05,
+      "loss": 0.0891,
+      "step": 36970
+    },
+    {
+      "epoch": 2.4193653909061172,
+      "grad_norm": 0.7659550309181213,
+      "learning_rate": 9.353857677904893e-05,
+      "loss": 0.0852,
+      "step": 36980
+    },
+    {
+      "epoch": 2.4200196270853778,
+      "grad_norm": 0.8483705520629883,
+      "learning_rate": 9.353405942761191e-05,
+      "loss": 0.0777,
+      "step": 36990
+    },
+    {
+      "epoch": 2.4206738632646383,
+      "grad_norm": 0.7422915101051331,
+      "learning_rate": 9.35295406067862e-05,
+      "loss": 0.0827,
+      "step": 37000
+    },
+    {
+      "epoch": 2.4213280994438993,
+      "grad_norm": 0.8917139172554016,
+      "learning_rate": 9.352502031672435e-05,
+      "loss": 0.0917,
+      "step": 37010
+    },
+    {
+      "epoch": 2.42198233562316,
+      "grad_norm": 0.973841667175293,
+      "learning_rate": 9.35204985575789e-05,
+      "loss": 0.0856,
+      "step": 37020
+    },
+    {
+      "epoch": 2.4226365718024208,
+      "grad_norm": 0.8658559918403625,
+      "learning_rate": 9.351597532950247e-05,
+      "loss": 0.0911,
+      "step": 37030
+    },
+    {
+      "epoch": 2.4232908079816813,
+      "grad_norm": 1.0100637674331665,
+      "learning_rate": 9.351145063264778e-05,
+      "loss": 0.0832,
+      "step": 37040
+    },
+    {
+      "epoch": 2.4239450441609423,
+      "grad_norm": 0.8611866235733032,
+      "learning_rate": 9.35069244671675e-05,
+      "loss": 0.0859,
+      "step": 37050
+    },
+    {
+      "epoch": 2.424599280340203,
+      "grad_norm": 0.8808668255805969,
+      "learning_rate": 9.350239683321443e-05,
+      "loss": 0.0876,
+      "step": 37060
+    },
+    {
+      "epoch": 2.4252535165194633,
+      "grad_norm": 0.767738938331604,
+      "learning_rate": 9.349786773094137e-05,
+      "loss": 0.0824,
+      "step": 37070
+    },
+    {
+      "epoch": 2.4259077526987243,
+      "grad_norm": 0.8980047106742859,
+      "learning_rate": 9.34933371605012e-05,
+      "loss": 0.0844,
+      "step": 37080
+    },
+    {
+      "epoch": 2.426561988877985,
+      "grad_norm": 0.6932862401008606,
+      "learning_rate": 9.348880512204683e-05,
+      "loss": 0.0765,
+      "step": 37090
+    },
+    {
+      "epoch": 2.4272162250572458,
+      "grad_norm": 0.8698898553848267,
+      "learning_rate": 9.348427161573124e-05,
+      "loss": 0.0769,
+      "step": 37100
+    },
+    {
+      "epoch": 2.4278704612365063,
+      "grad_norm": 0.8262230157852173,
+      "learning_rate": 9.347973664170744e-05,
+      "loss": 0.0876,
+      "step": 37110
+    },
+    {
+      "epoch": 2.4285246974157673,
+      "grad_norm": 0.690406322479248,
+      "learning_rate": 9.347520020012848e-05,
+      "loss": 0.086,
+      "step": 37120
+    },
+    {
+      "epoch": 2.429178933595028,
+      "grad_norm": 0.7355995178222656,
+      "learning_rate": 9.347066229114751e-05,
+      "loss": 0.0938,
+      "step": 37130
+    },
+    {
+      "epoch": 2.4298331697742883,
+      "grad_norm": 0.8815680742263794,
+      "learning_rate": 9.34661229149177e-05,
+      "loss": 0.0923,
+      "step": 37140
+    },
+    {
+      "epoch": 2.4304874059535493,
+      "grad_norm": 0.8908131122589111,
+      "learning_rate": 9.346158207159222e-05,
+      "loss": 0.0841,
+      "step": 37150
+    },
+    {
+      "epoch": 2.43114164213281,
+      "grad_norm": 0.7907306551933289,
+      "learning_rate": 9.345703976132438e-05,
+      "loss": 0.0847,
+      "step": 37160
+    },
+    {
+      "epoch": 2.431795878312071,
+      "grad_norm": 0.9056040048599243,
+      "learning_rate": 9.345249598426746e-05,
+      "loss": 0.0879,
+      "step": 37170
+    },
+    {
+      "epoch": 2.4324501144913313,
+      "grad_norm": 0.7010656595230103,
+      "learning_rate": 9.344795074057487e-05,
+      "loss": 0.0732,
+      "step": 37180
+    },
+    {
+      "epoch": 2.4331043506705923,
+      "grad_norm": 0.7615019083023071,
+      "learning_rate": 9.344340403039998e-05,
+      "loss": 0.0846,
+      "step": 37190
+    },
+    {
+      "epoch": 2.433758586849853,
+      "grad_norm": 0.9144221544265747,
+      "learning_rate": 9.343885585389627e-05,
+      "loss": 0.0982,
+      "step": 37200
+    },
+    {
+      "epoch": 2.4344128230291133,
+      "grad_norm": 1.0988235473632812,
+      "learning_rate": 9.343430621121724e-05,
+      "loss": 0.0877,
+      "step": 37210
+    },
+    {
+      "epoch": 2.4350670592083743,
+      "grad_norm": 0.8255583643913269,
+      "learning_rate": 9.342975510251649e-05,
+      "loss": 0.0832,
+      "step": 37220
+    },
+    {
+      "epoch": 2.435721295387635,
+      "grad_norm": 0.7476474642753601,
+      "learning_rate": 9.342520252794759e-05,
+      "loss": 0.0843,
+      "step": 37230
+    },
+    {
+      "epoch": 2.436375531566896,
+      "grad_norm": 0.7885528802871704,
+      "learning_rate": 9.342064848766423e-05,
+      "loss": 0.0896,
+      "step": 37240
+    },
+    {
+      "epoch": 2.4370297677461563,
+      "grad_norm": 0.8484315872192383,
+      "learning_rate": 9.341609298182008e-05,
+      "loss": 0.0814,
+      "step": 37250
+    },
+    {
+      "epoch": 2.4376840039254173,
+      "grad_norm": 1.143760085105896,
+      "learning_rate": 9.341153601056896e-05,
+      "loss": 0.0843,
+      "step": 37260
+    },
+    {
+      "epoch": 2.438338240104678,
+      "grad_norm": 0.7596016526222229,
+      "learning_rate": 9.340697757406462e-05,
+      "loss": 0.0813,
+      "step": 37270
+    },
+    {
+      "epoch": 2.4389924762839383,
+      "grad_norm": 0.9794483184814453,
+      "learning_rate": 9.340241767246099e-05,
+      "loss": 0.0755,
+      "step": 37280
+    },
+    {
+      "epoch": 2.4396467124631993,
+      "grad_norm": 0.7860924005508423,
+      "learning_rate": 9.33978563059119e-05,
+      "loss": 0.0782,
+      "step": 37290
+    },
+    {
+      "epoch": 2.44030094864246,
+      "grad_norm": 0.8889253735542297,
+      "learning_rate": 9.339329347457135e-05,
+      "loss": 0.0896,
+      "step": 37300
+    },
+    {
+      "epoch": 2.440955184821721,
+      "grad_norm": 0.6699261665344238,
+      "learning_rate": 9.338872917859335e-05,
+      "loss": 0.0847,
+      "step": 37310
+    },
+    {
+      "epoch": 2.4416094210009813,
+      "grad_norm": 0.8872549533843994,
+      "learning_rate": 9.338416341813196e-05,
+      "loss": 0.081,
+      "step": 37320
+    },
+    {
+      "epoch": 2.4422636571802423,
+      "grad_norm": 0.8817291259765625,
+      "learning_rate": 9.337959619334125e-05,
+      "loss": 0.0812,
+      "step": 37330
+    },
+    {
+      "epoch": 2.442917893359503,
+      "grad_norm": 0.9922673106193542,
+      "learning_rate": 9.337502750437542e-05,
+      "loss": 0.0871,
+      "step": 37340
+    },
+    {
+      "epoch": 2.4435721295387633,
+      "grad_norm": 0.7725539803504944,
+      "learning_rate": 9.337045735138865e-05,
+      "loss": 0.0787,
+      "step": 37350
+    },
+    {
+      "epoch": 2.4442263657180243,
+      "grad_norm": 0.9328073859214783,
+      "learning_rate": 9.336588573453521e-05,
+      "loss": 0.0799,
+      "step": 37360
+    },
+    {
+      "epoch": 2.444880601897285,
+      "grad_norm": 0.7553234696388245,
+      "learning_rate": 9.33613126539694e-05,
+      "loss": 0.0799,
+      "step": 37370
+    },
+    {
+      "epoch": 2.445534838076546,
+      "grad_norm": 0.8015692830085754,
+      "learning_rate": 9.335673810984553e-05,
+      "loss": 0.0894,
+      "step": 37380
+    },
+    {
+      "epoch": 2.4461890742558063,
+      "grad_norm": 0.7665364146232605,
+      "learning_rate": 9.335216210231807e-05,
+      "loss": 0.0815,
+      "step": 37390
+    },
+    {
+      "epoch": 2.446843310435067,
+      "grad_norm": 0.7971277236938477,
+      "learning_rate": 9.334758463154145e-05,
+      "loss": 0.0827,
+      "step": 37400
+    },
+    {
+      "epoch": 2.447497546614328,
+      "grad_norm": 0.7936835885047913,
+      "learning_rate": 9.334300569767016e-05,
+      "loss": 0.0867,
+      "step": 37410
+    },
+    {
+      "epoch": 2.4481517827935884,
+      "grad_norm": 0.9368561506271362,
+      "learning_rate": 9.333842530085875e-05,
+      "loss": 0.0768,
+      "step": 37420
+    },
+    {
+      "epoch": 2.4488060189728493,
+      "grad_norm": 0.947975218296051,
+      "learning_rate": 9.333384344126184e-05,
+      "loss": 0.0807,
+      "step": 37430
+    },
+    {
+      "epoch": 2.44946025515211,
+      "grad_norm": 0.7947817444801331,
+      "learning_rate": 9.332926011903405e-05,
+      "loss": 0.0909,
+      "step": 37440
+    },
+    {
+      "epoch": 2.4501144913313704,
+      "grad_norm": 0.8594073057174683,
+      "learning_rate": 9.33246753343301e-05,
+      "loss": 0.082,
+      "step": 37450
+    },
+    {
+      "epoch": 2.4507687275106314,
+      "grad_norm": 1.020121693611145,
+      "learning_rate": 9.332008908730473e-05,
+      "loss": 0.078,
+      "step": 37460
+    },
+    {
+      "epoch": 2.451422963689892,
+      "grad_norm": 0.7262875437736511,
+      "learning_rate": 9.331550137811276e-05,
+      "loss": 0.0789,
+      "step": 37470
+    },
+    {
+      "epoch": 2.452077199869153,
+      "grad_norm": 0.7914772033691406,
+      "learning_rate": 9.331091220690902e-05,
+      "loss": 0.083,
+      "step": 37480
+    },
+    {
+      "epoch": 2.4527314360484134,
+      "grad_norm": 0.8740646243095398,
+      "learning_rate": 9.330632157384838e-05,
+      "loss": 0.0757,
+      "step": 37490
+    },
+    {
+      "epoch": 2.4533856722276743,
+      "grad_norm": 0.9303426742553711,
+      "learning_rate": 9.330172947908583e-05,
+      "loss": 0.0835,
+      "step": 37500
+    },
+    {
+      "epoch": 2.454039908406935,
+      "grad_norm": 0.7235417366027832,
+      "learning_rate": 9.329713592277634e-05,
+      "loss": 0.0836,
+      "step": 37510
+    },
+    {
+      "epoch": 2.4546941445861954,
+      "grad_norm": 0.7460285425186157,
+      "learning_rate": 9.329254090507498e-05,
+      "loss": 0.0807,
+      "step": 37520
+    },
+    {
+      "epoch": 2.4553483807654564,
+      "grad_norm": 0.6465203166007996,
+      "learning_rate": 9.32879444261368e-05,
+      "loss": 0.0784,
+      "step": 37530
+    },
+    {
+      "epoch": 2.456002616944717,
+      "grad_norm": 0.80832839012146,
+      "learning_rate": 9.328334648611699e-05,
+      "loss": 0.0777,
+      "step": 37540
+    },
+    {
+      "epoch": 2.456656853123978,
+      "grad_norm": 0.7820184230804443,
+      "learning_rate": 9.32787470851707e-05,
+      "loss": 0.0819,
+      "step": 37550
+    },
+    {
+      "epoch": 2.4573110893032384,
+      "grad_norm": 0.934441864490509,
+      "learning_rate": 9.32741462234532e-05,
+      "loss": 0.08,
+      "step": 37560
+    },
+    {
+      "epoch": 2.4579653254824994,
+      "grad_norm": 0.9643844962120056,
+      "learning_rate": 9.32695439011198e-05,
+      "loss": 0.0737,
+      "step": 37570
+    },
+    {
+      "epoch": 2.45861956166176,
+      "grad_norm": 1.1645561456680298,
+      "learning_rate": 9.326494011832578e-05,
+      "loss": 0.0802,
+      "step": 37580
+    },
+    {
+      "epoch": 2.4592737978410204,
+      "grad_norm": 0.8493544459342957,
+      "learning_rate": 9.326033487522659e-05,
+      "loss": 0.0881,
+      "step": 37590
+    },
+    {
+      "epoch": 2.4599280340202814,
+      "grad_norm": 0.7914831042289734,
+      "learning_rate": 9.325572817197763e-05,
+      "loss": 0.088,
+      "step": 37600
+    },
+    {
+      "epoch": 2.460582270199542,
+      "grad_norm": 0.7926651239395142,
+      "learning_rate": 9.325112000873439e-05,
+      "loss": 0.0873,
+      "step": 37610
+    },
+    {
+      "epoch": 2.461236506378803,
+      "grad_norm": 1.0331966876983643,
+      "learning_rate": 9.324651038565244e-05,
+      "loss": 0.0813,
+      "step": 37620
+    },
+    {
+      "epoch": 2.4618907425580634,
+      "grad_norm": 1.0331898927688599,
+      "learning_rate": 9.324189930288734e-05,
+      "loss": 0.0896,
+      "step": 37630
+    },
+    {
+      "epoch": 2.4625449787373244,
+      "grad_norm": 0.877951443195343,
+      "learning_rate": 9.323728676059474e-05,
+      "loss": 0.0802,
+      "step": 37640
+    },
+    {
+      "epoch": 2.463199214916585,
+      "grad_norm": 0.9235687255859375,
+      "learning_rate": 9.32326727589303e-05,
+      "loss": 0.0785,
+      "step": 37650
+    },
+    {
+      "epoch": 2.4638534510958454,
+      "grad_norm": 0.8438425660133362,
+      "learning_rate": 9.322805729804979e-05,
+      "loss": 0.0837,
+      "step": 37660
+    },
+    {
+      "epoch": 2.4645076872751064,
+      "grad_norm": 0.9116701483726501,
+      "learning_rate": 9.322344037810898e-05,
+      "loss": 0.0815,
+      "step": 37670
+    },
+    {
+      "epoch": 2.465161923454367,
+      "grad_norm": 0.7631514072418213,
+      "learning_rate": 9.321882199926369e-05,
+      "loss": 0.0812,
+      "step": 37680
+    },
+    {
+      "epoch": 2.465816159633628,
+      "grad_norm": 0.6126758456230164,
+      "learning_rate": 9.321420216166979e-05,
+      "loss": 0.0787,
+      "step": 37690
+    },
+    {
+      "epoch": 2.4664703958128884,
+      "grad_norm": 0.8149334192276001,
+      "learning_rate": 9.320958086548326e-05,
+      "loss": 0.0836,
+      "step": 37700
+    },
+    {
+      "epoch": 2.4671246319921494,
+      "grad_norm": 0.9959266781806946,
+      "learning_rate": 9.320495811086006e-05,
+      "loss": 0.0878,
+      "step": 37710
+    },
+    {
+      "epoch": 2.46777886817141,
+      "grad_norm": 0.7576195597648621,
+      "learning_rate": 9.320033389795619e-05,
+      "loss": 0.0739,
+      "step": 37720
+    },
+    {
+      "epoch": 2.4684331043506704,
+      "grad_norm": 0.8124213814735413,
+      "learning_rate": 9.319570822692778e-05,
+      "loss": 0.0902,
+      "step": 37730
+    },
+    {
+      "epoch": 2.4690873405299314,
+      "grad_norm": 0.8176689743995667,
+      "learning_rate": 9.319108109793091e-05,
+      "loss": 0.0825,
+      "step": 37740
+    },
+    {
+      "epoch": 2.469741576709192,
+      "grad_norm": 0.7739967703819275,
+      "learning_rate": 9.318645251112179e-05,
+      "loss": 0.0901,
+      "step": 37750
+    },
+    {
+      "epoch": 2.470395812888453,
+      "grad_norm": 0.7479947209358215,
+      "learning_rate": 9.318182246665663e-05,
+      "loss": 0.0813,
+      "step": 37760
+    },
+    {
+      "epoch": 2.4710500490677134,
+      "grad_norm": 0.8923253417015076,
+      "learning_rate": 9.317719096469172e-05,
+      "loss": 0.0796,
+      "step": 37770
+    },
+    {
+      "epoch": 2.4717042852469744,
+      "grad_norm": 0.9640159010887146,
+      "learning_rate": 9.317255800538339e-05,
+      "loss": 0.0797,
+      "step": 37780
+    },
+    {
+      "epoch": 2.472358521426235,
+      "grad_norm": 0.9136691689491272,
+      "learning_rate": 9.3167923588888e-05,
+      "loss": 0.0888,
+      "step": 37790
+    },
+    {
+      "epoch": 2.4730127576054954,
+      "grad_norm": 0.7730658650398254,
+      "learning_rate": 9.316328771536195e-05,
+      "loss": 0.0832,
+      "step": 37800
+    },
+    {
+      "epoch": 2.4736669937847564,
+      "grad_norm": 0.7457364201545715,
+      "learning_rate": 9.315865038496177e-05,
+      "loss": 0.0875,
+      "step": 37810
+    },
+    {
+      "epoch": 2.474321229964017,
+      "grad_norm": 0.7927015423774719,
+      "learning_rate": 9.315401159784394e-05,
+      "loss": 0.0857,
+      "step": 37820
+    },
+    {
+      "epoch": 2.474975466143278,
+      "grad_norm": 0.861659586429596,
+      "learning_rate": 9.314937135416506e-05,
+      "loss": 0.079,
+      "step": 37830
+    },
+    {
+      "epoch": 2.4756297023225384,
+      "grad_norm": 0.9620111584663391,
+      "learning_rate": 9.31447296540817e-05,
+      "loss": 0.0918,
+      "step": 37840
+    },
+    {
+      "epoch": 2.4762839385017994,
+      "grad_norm": 0.7795575857162476,
+      "learning_rate": 9.314008649775059e-05,
+      "loss": 0.0752,
+      "step": 37850
+    },
+    {
+      "epoch": 2.47693817468106,
+      "grad_norm": 0.9564193487167358,
+      "learning_rate": 9.313544188532841e-05,
+      "loss": 0.0756,
+      "step": 37860
+    },
+    {
+      "epoch": 2.4775924108603204,
+      "grad_norm": 0.7260618805885315,
+      "learning_rate": 9.313079581697194e-05,
+      "loss": 0.0773,
+      "step": 37870
+    },
+    {
+      "epoch": 2.4782466470395814,
+      "grad_norm": 1.0667495727539062,
+      "learning_rate": 9.312614829283799e-05,
+      "loss": 0.0941,
+      "step": 37880
+    },
+    {
+      "epoch": 2.478900883218842,
+      "grad_norm": 0.7811123728752136,
+      "learning_rate": 9.312149931308345e-05,
+      "loss": 0.0791,
+      "step": 37890
+    },
+    {
+      "epoch": 2.4795551193981025,
+      "grad_norm": 0.8247928619384766,
+      "learning_rate": 9.31168488778652e-05,
+      "loss": 0.0873,
+      "step": 37900
+    },
+    {
+      "epoch": 2.4802093555773634,
+      "grad_norm": 0.8802565932273865,
+      "learning_rate": 9.311219698734024e-05,
+      "loss": 0.0797,
+      "step": 37910
+    },
+    {
+      "epoch": 2.480863591756624,
+      "grad_norm": 0.9362874031066895,
+      "learning_rate": 9.310754364166554e-05,
+      "loss": 0.0814,
+      "step": 37920
+    },
+    {
+      "epoch": 2.481517827935885,
+      "grad_norm": 0.9797009825706482,
+      "learning_rate": 9.310288884099822e-05,
+      "loss": 0.0892,
+      "step": 37930
+    },
+    {
+      "epoch": 2.4821720641151455,
+      "grad_norm": 0.7946538925170898,
+      "learning_rate": 9.309823258549535e-05,
+      "loss": 0.0809,
+      "step": 37940
+    },
+    {
+      "epoch": 2.4828263002944064,
+      "grad_norm": 0.8862201571464539,
+      "learning_rate": 9.30935748753141e-05,
+      "loss": 0.084,
+      "step": 37950
+    },
+    {
+      "epoch": 2.483480536473667,
+      "grad_norm": 0.8820653557777405,
+      "learning_rate": 9.308891571061167e-05,
+      "loss": 0.0921,
+      "step": 37960
+    },
+    {
+      "epoch": 2.4841347726529275,
+      "grad_norm": 0.8164055347442627,
+      "learning_rate": 9.308425509154533e-05,
+      "loss": 0.0819,
+      "step": 37970
+    },
+    {
+      "epoch": 2.4847890088321885,
+      "grad_norm": 0.8694265484809875,
+      "learning_rate": 9.307959301827241e-05,
+      "loss": 0.0782,
+      "step": 37980
+    },
+    {
+      "epoch": 2.485443245011449,
+      "grad_norm": 1.1511107683181763,
+      "learning_rate": 9.307492949095021e-05,
+      "loss": 0.0913,
+      "step": 37990
+    },
+    {
+      "epoch": 2.48609748119071,
+      "grad_norm": 0.8150752186775208,
+      "learning_rate": 9.307026450973619e-05,
+      "loss": 0.0899,
+      "step": 38000
+    },
+    {
+      "epoch": 2.4867517173699705,
+      "grad_norm": 0.9316548109054565,
+      "learning_rate": 9.306559807478779e-05,
+      "loss": 0.0798,
+      "step": 38010
+    },
+    {
+      "epoch": 2.4874059535492314,
+      "grad_norm": 0.9588937759399414,
+      "learning_rate": 9.306093018626252e-05,
+      "loss": 0.082,
+      "step": 38020
+    },
+    {
+      "epoch": 2.488060189728492,
+      "grad_norm": 1.0078554153442383,
+      "learning_rate": 9.30562608443179e-05,
+      "loss": 0.078,
+      "step": 38030
+    },
+    {
+      "epoch": 2.4887144259077525,
+      "grad_norm": 1.0711711645126343,
+      "learning_rate": 9.30515900491116e-05,
+      "loss": 0.0909,
+      "step": 38040
+    },
+    {
+      "epoch": 2.4893686620870135,
+      "grad_norm": 0.867173433303833,
+      "learning_rate": 9.30469178008012e-05,
+      "loss": 0.0822,
+      "step": 38050
+    },
+    {
+      "epoch": 2.490022898266274,
+      "grad_norm": 0.7762174606323242,
+      "learning_rate": 9.304224409954442e-05,
+      "loss": 0.0857,
+      "step": 38060
+    },
+    {
+      "epoch": 2.490677134445535,
+      "grad_norm": 1.0067529678344727,
+      "learning_rate": 9.303756894549903e-05,
+      "loss": 0.0851,
+      "step": 38070
+    },
+    {
+      "epoch": 2.4913313706247955,
+      "grad_norm": 0.9543056488037109,
+      "learning_rate": 9.303289233882281e-05,
+      "loss": 0.0754,
+      "step": 38080
+    },
+    {
+      "epoch": 2.4919856068040565,
+      "grad_norm": 1.024979591369629,
+      "learning_rate": 9.302821427967363e-05,
+      "loss": 0.0859,
+      "step": 38090
+    },
+    {
+      "epoch": 2.492639842983317,
+      "grad_norm": 0.7905625700950623,
+      "learning_rate": 9.302353476820936e-05,
+      "loss": 0.0789,
+      "step": 38100
+    },
+    {
+      "epoch": 2.4932940791625775,
+      "grad_norm": 0.8276375532150269,
+      "learning_rate": 9.301885380458797e-05,
+      "loss": 0.0807,
+      "step": 38110
+    },
+    {
+      "epoch": 2.4939483153418385,
+      "grad_norm": 0.7747846245765686,
+      "learning_rate": 9.301417138896743e-05,
+      "loss": 0.0798,
+      "step": 38120
+    },
+    {
+      "epoch": 2.494602551521099,
+      "grad_norm": 0.9668558835983276,
+      "learning_rate": 9.30094875215058e-05,
+      "loss": 0.0816,
+      "step": 38130
+    },
+    {
+      "epoch": 2.49525678770036,
+      "grad_norm": 1.070252537727356,
+      "learning_rate": 9.300480220236119e-05,
+      "loss": 0.0888,
+      "step": 38140
+    },
+    {
+      "epoch": 2.4959110238796205,
+      "grad_norm": 0.9471240043640137,
+      "learning_rate": 9.30001154316917e-05,
+      "loss": 0.0879,
+      "step": 38150
+    },
+    {
+      "epoch": 2.4965652600588815,
+      "grad_norm": 0.7727608680725098,
+      "learning_rate": 9.299542720965554e-05,
+      "loss": 0.0879,
+      "step": 38160
+    },
+    {
+      "epoch": 2.497219496238142,
+      "grad_norm": 1.2870181798934937,
+      "learning_rate": 9.299073753641096e-05,
+      "loss": 0.0907,
+      "step": 38170
+    },
+    {
+      "epoch": 2.4978737324174025,
+      "grad_norm": 1.0817999839782715,
+      "learning_rate": 9.298604641211624e-05,
+      "loss": 0.1045,
+      "step": 38180
+    },
+    {
+      "epoch": 2.4985279685966635,
+      "grad_norm": 0.7632635831832886,
+      "learning_rate": 9.298135383692972e-05,
+      "loss": 0.0793,
+      "step": 38190
+    },
+    {
+      "epoch": 2.499182204775924,
+      "grad_norm": 0.9093852639198303,
+      "learning_rate": 9.297665981100978e-05,
+      "loss": 0.0774,
+      "step": 38200
+    },
+    {
+      "epoch": 2.499836440955185,
+      "grad_norm": 1.0639636516571045,
+      "learning_rate": 9.297196433451487e-05,
+      "loss": 0.0728,
+      "step": 38210
+    },
+    {
+      "epoch": 2.5004906771344455,
+      "grad_norm": 0.9197503924369812,
+      "learning_rate": 9.296726740760346e-05,
+      "loss": 0.0849,
+      "step": 38220
+    },
+    {
+      "epoch": 2.5011449133137065,
+      "grad_norm": 0.8370291590690613,
+      "learning_rate": 9.296256903043408e-05,
+      "loss": 0.0764,
+      "step": 38230
+    },
+    {
+      "epoch": 2.501799149492967,
+      "grad_norm": 0.7800946831703186,
+      "learning_rate": 9.295786920316533e-05,
+      "loss": 0.0802,
+      "step": 38240
+    },
+    {
+      "epoch": 2.5024533856722275,
+      "grad_norm": 1.0258721113204956,
+      "learning_rate": 9.295316792595586e-05,
+      "loss": 0.0874,
+      "step": 38250
+    },
+    {
+      "epoch": 2.5031076218514885,
+      "grad_norm": 0.7372403144836426,
+      "learning_rate": 9.294846519896429e-05,
+      "loss": 0.0774,
+      "step": 38260
+    },
+    {
+      "epoch": 2.503761858030749,
+      "grad_norm": 0.8242309093475342,
+      "learning_rate": 9.294376102234938e-05,
+      "loss": 0.0814,
+      "step": 38270
+    },
+    {
+      "epoch": 2.5044160942100095,
+      "grad_norm": 0.8311209678649902,
+      "learning_rate": 9.293905539626993e-05,
+      "loss": 0.079,
+      "step": 38280
+    },
+    {
+      "epoch": 2.5050703303892705,
+      "grad_norm": 0.7952736616134644,
+      "learning_rate": 9.293434832088475e-05,
+      "loss": 0.0779,
+      "step": 38290
+    },
+    {
+      "epoch": 2.5057245665685315,
+      "grad_norm": 0.9289050102233887,
+      "learning_rate": 9.29296397963527e-05,
+      "loss": 0.0824,
+      "step": 38300
+    },
+    {
+      "epoch": 2.506378802747792,
+      "grad_norm": 0.9673059582710266,
+      "learning_rate": 9.292492982283272e-05,
+      "loss": 0.0874,
+      "step": 38310
+    },
+    {
+      "epoch": 2.5070330389270525,
+      "grad_norm": 0.8958445191383362,
+      "learning_rate": 9.29202184004838e-05,
+      "loss": 0.0877,
+      "step": 38320
+    },
+    {
+      "epoch": 2.5076872751063135,
+      "grad_norm": 0.8611701726913452,
+      "learning_rate": 9.291550552946493e-05,
+      "loss": 0.0834,
+      "step": 38330
+    },
+    {
+      "epoch": 2.508341511285574,
+      "grad_norm": 0.7747957706451416,
+      "learning_rate": 9.29107912099352e-05,
+      "loss": 0.077,
+      "step": 38340
+    },
+    {
+      "epoch": 2.5089957474648346,
+      "grad_norm": 1.0127016305923462,
+      "learning_rate": 9.290607544205374e-05,
+      "loss": 0.0961,
+      "step": 38350
+    },
+    {
+      "epoch": 2.5096499836440955,
+      "grad_norm": 0.829646110534668,
+      "learning_rate": 9.290135822597969e-05,
+      "loss": 0.0863,
+      "step": 38360
+    },
+    {
+      "epoch": 2.5103042198233565,
+      "grad_norm": 0.9763094782829285,
+      "learning_rate": 9.28966395618723e-05,
+      "loss": 0.08,
+      "step": 38370
+    },
+    {
+      "epoch": 2.510958456002617,
+      "grad_norm": 0.8551107048988342,
+      "learning_rate": 9.289191944989083e-05,
+      "loss": 0.079,
+      "step": 38380
+    },
+    {
+      "epoch": 2.5116126921818775,
+      "grad_norm": 0.9324933886528015,
+      "learning_rate": 9.288719789019458e-05,
+      "loss": 0.0947,
+      "step": 38390
+    },
+    {
+      "epoch": 2.5122669283611385,
+      "grad_norm": 0.8589342832565308,
+      "learning_rate": 9.288247488294293e-05,
+      "loss": 0.0805,
+      "step": 38400
+    },
+    {
+      "epoch": 2.512921164540399,
+      "grad_norm": 0.7855199575424194,
+      "learning_rate": 9.28777504282953e-05,
+      "loss": 0.0841,
+      "step": 38410
+    },
+    {
+      "epoch": 2.5135754007196596,
+      "grad_norm": 0.9341410994529724,
+      "learning_rate": 9.287302452641112e-05,
+      "loss": 0.0938,
+      "step": 38420
+    },
+    {
+      "epoch": 2.5142296368989205,
+      "grad_norm": 1.012837290763855,
+      "learning_rate": 9.286829717744993e-05,
+      "loss": 0.076,
+      "step": 38430
+    },
+    {
+      "epoch": 2.514883873078181,
+      "grad_norm": 0.8277202844619751,
+      "learning_rate": 9.286356838157128e-05,
+      "loss": 0.0847,
+      "step": 38440
+    },
+    {
+      "epoch": 2.515538109257442,
+      "grad_norm": 0.7765419483184814,
+      "learning_rate": 9.28588381389348e-05,
+      "loss": 0.0896,
+      "step": 38450
+    },
+    {
+      "epoch": 2.5161923454367026,
+      "grad_norm": 1.0105395317077637,
+      "learning_rate": 9.285410644970013e-05,
+      "loss": 0.0918,
+      "step": 38460
+    },
+    {
+      "epoch": 2.5168465816159635,
+      "grad_norm": 0.7408658266067505,
+      "learning_rate": 9.284937331402697e-05,
+      "loss": 0.0808,
+      "step": 38470
+    },
+    {
+      "epoch": 2.517500817795224,
+      "grad_norm": 0.782345712184906,
+      "learning_rate": 9.284463873207508e-05,
+      "loss": 0.0748,
+      "step": 38480
+    },
+    {
+      "epoch": 2.5181550539744846,
+      "grad_norm": 0.7079125046730042,
+      "learning_rate": 9.283990270400428e-05,
+      "loss": 0.0831,
+      "step": 38490
+    },
+    {
+      "epoch": 2.5188092901537456,
+      "grad_norm": 0.920768678188324,
+      "learning_rate": 9.28351652299744e-05,
+      "loss": 0.0907,
+      "step": 38500
+    },
+    {
+      "epoch": 2.519463526333006,
+      "grad_norm": 0.9361595511436462,
+      "learning_rate": 9.283042631014535e-05,
+      "loss": 0.0848,
+      "step": 38510
+    },
+    {
+      "epoch": 2.520117762512267,
+      "grad_norm": 0.9975919723510742,
+      "learning_rate": 9.282568594467711e-05,
+      "loss": 0.0841,
+      "step": 38520
+    },
+    {
+      "epoch": 2.5207719986915276,
+      "grad_norm": 0.7760083675384521,
+      "learning_rate": 9.282094413372963e-05,
+      "loss": 0.0785,
+      "step": 38530
+    },
+    {
+      "epoch": 2.5214262348707885,
+      "grad_norm": 0.8321374654769897,
+      "learning_rate": 9.2816200877463e-05,
+      "loss": 0.0842,
+      "step": 38540
+    },
+    {
+      "epoch": 2.522080471050049,
+      "grad_norm": 1.0121804475784302,
+      "learning_rate": 9.28114561760373e-05,
+      "loss": 0.08,
+      "step": 38550
+    },
+    {
+      "epoch": 2.5227347072293096,
+      "grad_norm": 0.8037693500518799,
+      "learning_rate": 9.280671002961267e-05,
+      "loss": 0.0767,
+      "step": 38560
+    },
+    {
+      "epoch": 2.5233889434085706,
+      "grad_norm": 0.9284687042236328,
+      "learning_rate": 9.280196243834931e-05,
+      "loss": 0.0862,
+      "step": 38570
+    },
+    {
+      "epoch": 2.524043179587831,
+      "grad_norm": 0.7956443428993225,
+      "learning_rate": 9.279721340240745e-05,
+      "loss": 0.0783,
+      "step": 38580
+    },
+    {
+      "epoch": 2.524697415767092,
+      "grad_norm": 0.7840268015861511,
+      "learning_rate": 9.279246292194743e-05,
+      "loss": 0.0842,
+      "step": 38590
+    },
+    {
+      "epoch": 2.5253516519463526,
+      "grad_norm": 0.9286373257637024,
+      "learning_rate": 9.278771099712956e-05,
+      "loss": 0.0921,
+      "step": 38600
+    },
+    {
+      "epoch": 2.5260058881256136,
+      "grad_norm": 0.8080535531044006,
+      "learning_rate": 9.27829576281142e-05,
+      "loss": 0.0879,
+      "step": 38610
+    },
+    {
+      "epoch": 2.526660124304874,
+      "grad_norm": 0.9005104899406433,
+      "learning_rate": 9.277820281506184e-05,
+      "loss": 0.0819,
+      "step": 38620
+    },
+    {
+      "epoch": 2.5273143604841346,
+      "grad_norm": 0.7505617141723633,
+      "learning_rate": 9.277344655813292e-05,
+      "loss": 0.0816,
+      "step": 38630
+    },
+    {
+      "epoch": 2.5279685966633956,
+      "grad_norm": 0.6954260468482971,
+      "learning_rate": 9.276868885748802e-05,
+      "loss": 0.0822,
+      "step": 38640
+    },
+    {
+      "epoch": 2.528622832842656,
+      "grad_norm": 0.7404685616493225,
+      "learning_rate": 9.276392971328771e-05,
+      "loss": 0.071,
+      "step": 38650
+    },
+    {
+      "epoch": 2.529277069021917,
+      "grad_norm": 0.9218646287918091,
+      "learning_rate": 9.27591691256926e-05,
+      "loss": 0.0841,
+      "step": 38660
+    },
+    {
+      "epoch": 2.5299313052011776,
+      "grad_norm": 0.894137978553772,
+      "learning_rate": 9.275440709486342e-05,
+      "loss": 0.0914,
+      "step": 38670
+    },
+    {
+      "epoch": 2.5305855413804386,
+      "grad_norm": 0.7454158067703247,
+      "learning_rate": 9.274964362096085e-05,
+      "loss": 0.0858,
+      "step": 38680
+    },
+    {
+      "epoch": 2.531239777559699,
+      "grad_norm": 0.8861522674560547,
+      "learning_rate": 9.274487870414569e-05,
+      "loss": 0.0809,
+      "step": 38690
+    },
+    {
+      "epoch": 2.5318940137389596,
+      "grad_norm": 0.653781533241272,
+      "learning_rate": 9.27401123445788e-05,
+      "loss": 0.0796,
+      "step": 38700
+    },
+    {
+      "epoch": 2.5325482499182206,
+      "grad_norm": 0.8634945154190063,
+      "learning_rate": 9.273534454242101e-05,
+      "loss": 0.0744,
+      "step": 38710
+    },
+    {
+      "epoch": 2.533202486097481,
+      "grad_norm": 0.7516806125640869,
+      "learning_rate": 9.273057529783327e-05,
+      "loss": 0.0827,
+      "step": 38720
+    },
+    {
+      "epoch": 2.5338567222767416,
+      "grad_norm": 0.7837287783622742,
+      "learning_rate": 9.272580461097654e-05,
+      "loss": 0.0785,
+      "step": 38730
+    },
+    {
+      "epoch": 2.5345109584560026,
+      "grad_norm": 0.8319099545478821,
+      "learning_rate": 9.272103248201185e-05,
+      "loss": 0.0801,
+      "step": 38740
+    },
+    {
+      "epoch": 2.5351651946352636,
+      "grad_norm": 0.9561896324157715,
+      "learning_rate": 9.271625891110028e-05,
+      "loss": 0.0834,
+      "step": 38750
+    },
+    {
+      "epoch": 2.535819430814524,
+      "grad_norm": 0.8369835615158081,
+      "learning_rate": 9.271148389840294e-05,
+      "loss": 0.0829,
+      "step": 38760
+    },
+    {
+      "epoch": 2.5364736669937846,
+      "grad_norm": 0.8404773473739624,
+      "learning_rate": 9.270670744408101e-05,
+      "loss": 0.0795,
+      "step": 38770
+    },
+    {
+      "epoch": 2.5371279031730456,
+      "grad_norm": 0.7207024693489075,
+      "learning_rate": 9.270192954829571e-05,
+      "loss": 0.0775,
+      "step": 38780
+    },
+    {
+      "epoch": 2.537782139352306,
+      "grad_norm": 0.7661088109016418,
+      "learning_rate": 9.269715021120827e-05,
+      "loss": 0.0861,
+      "step": 38790
+    },
+    {
+      "epoch": 2.5384363755315666,
+      "grad_norm": 1.0687098503112793,
+      "learning_rate": 9.269236943298006e-05,
+      "loss": 0.0865,
+      "step": 38800
+    },
+    {
+      "epoch": 2.5390906117108276,
+      "grad_norm": 0.8831005096435547,
+      "learning_rate": 9.26875872137724e-05,
+      "loss": 0.0743,
+      "step": 38810
+    },
+    {
+      "epoch": 2.5397448478900886,
+      "grad_norm": 0.9108031988143921,
+      "learning_rate": 9.268280355374673e-05,
+      "loss": 0.0824,
+      "step": 38820
+    },
+    {
+      "epoch": 2.540399084069349,
+      "grad_norm": 0.8143110871315002,
+      "learning_rate": 9.26780184530645e-05,
+      "loss": 0.0764,
+      "step": 38830
+    },
+    {
+      "epoch": 2.5410533202486096,
+      "grad_norm": 0.7931360006332397,
+      "learning_rate": 9.267323191188721e-05,
+      "loss": 0.0811,
+      "step": 38840
+    },
+    {
+      "epoch": 2.5417075564278706,
+      "grad_norm": 0.9138084053993225,
+      "learning_rate": 9.266844393037644e-05,
+      "loss": 0.0804,
+      "step": 38850
+    },
+    {
+      "epoch": 2.542361792607131,
+      "grad_norm": 0.8782002329826355,
+      "learning_rate": 9.266365450869376e-05,
+      "loss": 0.081,
+      "step": 38860
+    },
+    {
+      "epoch": 2.5430160287863917,
+      "grad_norm": 0.8105145692825317,
+      "learning_rate": 9.265886364700089e-05,
+      "loss": 0.0772,
+      "step": 38870
+    },
+    {
+      "epoch": 2.5436702649656526,
+      "grad_norm": 1.0995359420776367,
+      "learning_rate": 9.265407134545947e-05,
+      "loss": 0.0805,
+      "step": 38880
+    },
+    {
+      "epoch": 2.544324501144913,
+      "grad_norm": 0.8194625377655029,
+      "learning_rate": 9.264927760423128e-05,
+      "loss": 0.0827,
+      "step": 38890
+    },
+    {
+      "epoch": 2.544978737324174,
+      "grad_norm": 0.9476547837257385,
+      "learning_rate": 9.264448242347812e-05,
+      "loss": 0.0885,
+      "step": 38900
+    },
+    {
+      "epoch": 2.5456329735034346,
+      "grad_norm": 0.9267840385437012,
+      "learning_rate": 9.263968580336185e-05,
+      "loss": 0.0826,
+      "step": 38910
+    },
+    {
+      "epoch": 2.5462872096826956,
+      "grad_norm": 0.8388037085533142,
+      "learning_rate": 9.263488774404434e-05,
+      "loss": 0.0834,
+      "step": 38920
+    },
+    {
+      "epoch": 2.546941445861956,
+      "grad_norm": 1.0958263874053955,
+      "learning_rate": 9.263008824568756e-05,
+      "loss": 0.0903,
+      "step": 38930
+    },
+    {
+      "epoch": 2.5475956820412167,
+      "grad_norm": 0.7269095778465271,
+      "learning_rate": 9.26252873084535e-05,
+      "loss": 0.0842,
+      "step": 38940
+    },
+    {
+      "epoch": 2.5482499182204776,
+      "grad_norm": 0.782738447189331,
+      "learning_rate": 9.262048493250422e-05,
+      "loss": 0.0848,
+      "step": 38950
+    },
+    {
+      "epoch": 2.548904154399738,
+      "grad_norm": 0.8937769532203674,
+      "learning_rate": 9.261568111800177e-05,
+      "loss": 0.0772,
+      "step": 38960
+    },
+    {
+      "epoch": 2.549558390578999,
+      "grad_norm": 0.7501261234283447,
+      "learning_rate": 9.261087586510834e-05,
+      "loss": 0.0836,
+      "step": 38970
+    },
+    {
+      "epoch": 2.5502126267582597,
+      "grad_norm": 0.8389232754707336,
+      "learning_rate": 9.260606917398609e-05,
+      "loss": 0.0939,
+      "step": 38980
+    },
+    {
+      "epoch": 2.5508668629375206,
+      "grad_norm": 0.7912153005599976,
+      "learning_rate": 9.260126104479727e-05,
+      "loss": 0.093,
+      "step": 38990
+    },
+    {
+      "epoch": 2.551521099116781,
+      "grad_norm": 0.8847309947013855,
+      "learning_rate": 9.259645147770415e-05,
+      "loss": 0.0745,
+      "step": 39000
+    },
+    {
+      "epoch": 2.5521753352960417,
+      "grad_norm": 0.7888216376304626,
+      "learning_rate": 9.25916404728691e-05,
+      "loss": 0.0829,
+      "step": 39010
+    },
+    {
+      "epoch": 2.5528295714753027,
+      "grad_norm": 1.0451523065567017,
+      "learning_rate": 9.258682803045448e-05,
+      "loss": 0.0973,
+      "step": 39020
+    },
+    {
+      "epoch": 2.553483807654563,
+      "grad_norm": 0.8907498717308044,
+      "learning_rate": 9.25820141506227e-05,
+      "loss": 0.0958,
+      "step": 39030
+    },
+    {
+      "epoch": 2.554138043833824,
+      "grad_norm": 0.9830831289291382,
+      "learning_rate": 9.257719883353631e-05,
+      "loss": 0.0944,
+      "step": 39040
+    },
+    {
+      "epoch": 2.5547922800130847,
+      "grad_norm": 0.8762632608413696,
+      "learning_rate": 9.257238207935777e-05,
+      "loss": 0.0822,
+      "step": 39050
+    },
+    {
+      "epoch": 2.5554465161923456,
+      "grad_norm": 0.8764397501945496,
+      "learning_rate": 9.256756388824968e-05,
+      "loss": 0.0811,
+      "step": 39060
+    },
+    {
+      "epoch": 2.556100752371606,
+      "grad_norm": 0.8016869425773621,
+      "learning_rate": 9.256274426037468e-05,
+      "loss": 0.0797,
+      "step": 39070
+    },
+    {
+      "epoch": 2.5567549885508667,
+      "grad_norm": 0.9790674448013306,
+      "learning_rate": 9.255792319589544e-05,
+      "loss": 0.0853,
+      "step": 39080
+    },
+    {
+      "epoch": 2.5574092247301277,
+      "grad_norm": 0.8641859889030457,
+      "learning_rate": 9.255310069497468e-05,
+      "loss": 0.0833,
+      "step": 39090
+    },
+    {
+      "epoch": 2.558063460909388,
+      "grad_norm": 0.8329319357872009,
+      "learning_rate": 9.254827675777517e-05,
+      "loss": 0.0812,
+      "step": 39100
+    },
+    {
+      "epoch": 2.558717697088649,
+      "grad_norm": 0.9464491009712219,
+      "learning_rate": 9.254345138445973e-05,
+      "loss": 0.0826,
+      "step": 39110
+    },
+    {
+      "epoch": 2.5593719332679097,
+      "grad_norm": 0.8477292656898499,
+      "learning_rate": 9.253862457519122e-05,
+      "loss": 0.0755,
+      "step": 39120
+    },
+    {
+      "epoch": 2.5600261694471707,
+      "grad_norm": 1.1739881038665771,
+      "learning_rate": 9.253379633013259e-05,
+      "loss": 0.0791,
+      "step": 39130
+    },
+    {
+      "epoch": 2.560680405626431,
+      "grad_norm": 0.8935783505439758,
+      "learning_rate": 9.252896664944677e-05,
+      "loss": 0.0834,
+      "step": 39140
+    },
+    {
+      "epoch": 2.5613346418056917,
+      "grad_norm": 0.8971335291862488,
+      "learning_rate": 9.25241355332968e-05,
+      "loss": 0.0781,
+      "step": 39150
+    },
+    {
+      "epoch": 2.5619888779849527,
+      "grad_norm": 0.8568238019943237,
+      "learning_rate": 9.251930298184572e-05,
+      "loss": 0.0795,
+      "step": 39160
+    },
+    {
+      "epoch": 2.562643114164213,
+      "grad_norm": 0.8689826726913452,
+      "learning_rate": 9.251446899525667e-05,
+      "loss": 0.0755,
+      "step": 39170
+    },
+    {
+      "epoch": 2.5632973503434737,
+      "grad_norm": 0.8828352093696594,
+      "learning_rate": 9.250963357369278e-05,
+      "loss": 0.0883,
+      "step": 39180
+    },
+    {
+      "epoch": 2.5639515865227347,
+      "grad_norm": 0.7668461799621582,
+      "learning_rate": 9.250479671731726e-05,
+      "loss": 0.0745,
+      "step": 39190
+    },
+    {
+      "epoch": 2.5646058227019957,
+      "grad_norm": 0.7046298980712891,
+      "learning_rate": 9.24999584262934e-05,
+      "loss": 0.0733,
+      "step": 39200
+    },
+    {
+      "epoch": 2.565260058881256,
+      "grad_norm": 0.77901691198349,
+      "learning_rate": 9.249511870078449e-05,
+      "loss": 0.0894,
+      "step": 39210
+    },
+    {
+      "epoch": 2.5659142950605167,
+      "grad_norm": 0.7388069033622742,
+      "learning_rate": 9.249027754095385e-05,
+      "loss": 0.0842,
+      "step": 39220
+    },
+    {
+      "epoch": 2.5665685312397777,
+      "grad_norm": 0.841890275478363,
+      "learning_rate": 9.248543494696494e-05,
+      "loss": 0.0827,
+      "step": 39230
+    },
+    {
+      "epoch": 2.567222767419038,
+      "grad_norm": 0.9349285364151001,
+      "learning_rate": 9.248059091898114e-05,
+      "loss": 0.0818,
+      "step": 39240
+    },
+    {
+      "epoch": 2.5678770035982987,
+      "grad_norm": 0.9265575408935547,
+      "learning_rate": 9.2475745457166e-05,
+      "loss": 0.0782,
+      "step": 39250
+    },
+    {
+      "epoch": 2.5685312397775597,
+      "grad_norm": 0.9001551270484924,
+      "learning_rate": 9.247089856168307e-05,
+      "loss": 0.0785,
+      "step": 39260
+    },
+    {
+      "epoch": 2.5691854759568207,
+      "grad_norm": 0.8538965582847595,
+      "learning_rate": 9.246605023269592e-05,
+      "loss": 0.0813,
+      "step": 39270
+    },
+    {
+      "epoch": 2.569839712136081,
+      "grad_norm": 0.9209931492805481,
+      "learning_rate": 9.24612004703682e-05,
+      "loss": 0.0824,
+      "step": 39280
+    },
+    {
+      "epoch": 2.5704939483153417,
+      "grad_norm": 0.6929624676704407,
+      "learning_rate": 9.245634927486361e-05,
+      "loss": 0.0781,
+      "step": 39290
+    },
+    {
+      "epoch": 2.5711481844946027,
+      "grad_norm": 0.7202306985855103,
+      "learning_rate": 9.245149664634589e-05,
+      "loss": 0.0899,
+      "step": 39300
+    },
+    {
+      "epoch": 2.571802420673863,
+      "grad_norm": 1.034266710281372,
+      "learning_rate": 9.244664258497881e-05,
+      "loss": 0.0793,
+      "step": 39310
+    },
+    {
+      "epoch": 2.5724566568531237,
+      "grad_norm": 0.7991123795509338,
+      "learning_rate": 9.244178709092624e-05,
+      "loss": 0.0764,
+      "step": 39320
+    },
+    {
+      "epoch": 2.5731108930323847,
+      "grad_norm": 0.7285871505737305,
+      "learning_rate": 9.243693016435204e-05,
+      "loss": 0.0873,
+      "step": 39330
+    },
+    {
+      "epoch": 2.5737651292116452,
+      "grad_norm": 0.7833579182624817,
+      "learning_rate": 9.243207180542016e-05,
+      "loss": 0.0819,
+      "step": 39340
+    },
+    {
+      "epoch": 2.574419365390906,
+      "grad_norm": 1.0560925006866455,
+      "learning_rate": 9.242721201429456e-05,
+      "loss": 0.0778,
+      "step": 39350
+    },
+    {
+      "epoch": 2.5750736015701667,
+      "grad_norm": 0.9079683423042297,
+      "learning_rate": 9.242235079113928e-05,
+      "loss": 0.0715,
+      "step": 39360
+    },
+    {
+      "epoch": 2.5757278377494277,
+      "grad_norm": 1.1182979345321655,
+      "learning_rate": 9.241748813611842e-05,
+      "loss": 0.0802,
+      "step": 39370
+    },
+    {
+      "epoch": 2.5763820739286882,
+      "grad_norm": 0.8554732799530029,
+      "learning_rate": 9.241262404939608e-05,
+      "loss": 0.0885,
+      "step": 39380
+    },
+    {
+      "epoch": 2.5770363101079488,
+      "grad_norm": 0.9766889214515686,
+      "learning_rate": 9.240775853113646e-05,
+      "loss": 0.0796,
+      "step": 39390
+    },
+    {
+      "epoch": 2.5776905462872097,
+      "grad_norm": 0.7499867081642151,
+      "learning_rate": 9.240289158150378e-05,
+      "loss": 0.0782,
+      "step": 39400
+    },
+    {
+      "epoch": 2.5783447824664703,
+      "grad_norm": 0.8955393433570862,
+      "learning_rate": 9.239802320066228e-05,
+      "loss": 0.0791,
+      "step": 39410
+    },
+    {
+      "epoch": 2.5789990186457312,
+      "grad_norm": 0.8181598782539368,
+      "learning_rate": 9.239315338877631e-05,
+      "loss": 0.0776,
+      "step": 39420
+    },
+    {
+      "epoch": 2.5796532548249917,
+      "grad_norm": 1.1022766828536987,
+      "learning_rate": 9.238828214601023e-05,
+      "loss": 0.0855,
+      "step": 39430
+    },
+    {
+      "epoch": 2.5803074910042527,
+      "grad_norm": 0.7724325656890869,
+      "learning_rate": 9.238340947252847e-05,
+      "loss": 0.0685,
+      "step": 39440
+    },
+    {
+      "epoch": 2.5809617271835132,
+      "grad_norm": 0.8263006806373596,
+      "learning_rate": 9.237853536849548e-05,
+      "loss": 0.0796,
+      "step": 39450
+    },
+    {
+      "epoch": 2.5816159633627738,
+      "grad_norm": 0.7736819386482239,
+      "learning_rate": 9.237365983407578e-05,
+      "loss": 0.0894,
+      "step": 39460
+    },
+    {
+      "epoch": 2.5822701995420347,
+      "grad_norm": 0.9027025699615479,
+      "learning_rate": 9.236878286943393e-05,
+      "loss": 0.0799,
+      "step": 39470
+    },
+    {
+      "epoch": 2.5829244357212953,
+      "grad_norm": 0.7817458510398865,
+      "learning_rate": 9.236390447473455e-05,
+      "loss": 0.073,
+      "step": 39480
+    },
+    {
+      "epoch": 2.5835786719005562,
+      "grad_norm": 1.0722339153289795,
+      "learning_rate": 9.23590246501423e-05,
+      "loss": 0.0872,
+      "step": 39490
+    },
+    {
+      "epoch": 2.5842329080798168,
+      "grad_norm": 1.1376055479049683,
+      "learning_rate": 9.235414339582185e-05,
+      "loss": 0.0818,
+      "step": 39500
+    },
+    {
+      "epoch": 2.5848871442590777,
+      "grad_norm": 0.907859206199646,
+      "learning_rate": 9.234926071193799e-05,
+      "loss": 0.0813,
+      "step": 39510
+    },
+    {
+      "epoch": 2.5855413804383383,
+      "grad_norm": 0.9271683096885681,
+      "learning_rate": 9.234437659865554e-05,
+      "loss": 0.0776,
+      "step": 39520
+    },
+    {
+      "epoch": 2.586195616617599,
+      "grad_norm": 1.029358148574829,
+      "learning_rate": 9.23394910561393e-05,
+      "loss": 0.0809,
+      "step": 39530
+    },
+    {
+      "epoch": 2.5868498527968598,
+      "grad_norm": 0.8963791131973267,
+      "learning_rate": 9.233460408455422e-05,
+      "loss": 0.0724,
+      "step": 39540
+    },
+    {
+      "epoch": 2.5875040889761203,
+      "grad_norm": 0.7677753567695618,
+      "learning_rate": 9.232971568406518e-05,
+      "loss": 0.0768,
+      "step": 39550
+    },
+    {
+      "epoch": 2.5881583251553812,
+      "grad_norm": 0.8181271553039551,
+      "learning_rate": 9.232482585483725e-05,
+      "loss": 0.0746,
+      "step": 39560
+    },
+    {
+      "epoch": 2.5888125613346418,
+      "grad_norm": 0.853274405002594,
+      "learning_rate": 9.231993459703547e-05,
+      "loss": 0.0705,
+      "step": 39570
+    },
+    {
+      "epoch": 2.5894667975139027,
+      "grad_norm": 0.9550169110298157,
+      "learning_rate": 9.23150419108249e-05,
+      "loss": 0.0884,
+      "step": 39580
+    },
+    {
+      "epoch": 2.5901210336931633,
+      "grad_norm": 0.9100688099861145,
+      "learning_rate": 9.231014779637067e-05,
+      "loss": 0.0702,
+      "step": 39590
+    },
+    {
+      "epoch": 2.590775269872424,
+      "grad_norm": 1.1120498180389404,
+      "learning_rate": 9.2305252253838e-05,
+      "loss": 0.0797,
+      "step": 39600
+    },
+    {
+      "epoch": 2.5914295060516848,
+      "grad_norm": 0.8531731963157654,
+      "learning_rate": 9.230035528339211e-05,
+      "loss": 0.0931,
+      "step": 39610
+    },
+    {
+      "epoch": 2.5920837422309453,
+      "grad_norm": 0.7823407649993896,
+      "learning_rate": 9.229545688519829e-05,
+      "loss": 0.0795,
+      "step": 39620
+    },
+    {
+      "epoch": 2.592737978410206,
+      "grad_norm": 0.9250206351280212,
+      "learning_rate": 9.229055705942189e-05,
+      "loss": 0.0858,
+      "step": 39630
+    },
+    {
+      "epoch": 2.593392214589467,
+      "grad_norm": 0.7410492897033691,
+      "learning_rate": 9.228565580622828e-05,
+      "loss": 0.0803,
+      "step": 39640
+    },
+    {
+      "epoch": 2.5940464507687278,
+      "grad_norm": 0.6830748915672302,
+      "learning_rate": 9.228075312578288e-05,
+      "loss": 0.0849,
+      "step": 39650
+    },
+    {
+      "epoch": 2.5947006869479883,
+      "grad_norm": 1.0575082302093506,
+      "learning_rate": 9.227584901825116e-05,
+      "loss": 0.0926,
+      "step": 39660
+    },
+    {
+      "epoch": 2.595354923127249,
+      "grad_norm": 0.950671911239624,
+      "learning_rate": 9.22709434837987e-05,
+      "loss": 0.079,
+      "step": 39670
+    },
+    {
+      "epoch": 2.5960091593065098,
+      "grad_norm": 0.7701423168182373,
+      "learning_rate": 9.226603652259102e-05,
+      "loss": 0.0928,
+      "step": 39680
+    },
+    {
+      "epoch": 2.5966633954857703,
+      "grad_norm": 0.9989359378814697,
+      "learning_rate": 9.226112813479377e-05,
+      "loss": 0.0818,
+      "step": 39690
+    },
+    {
+      "epoch": 2.597317631665031,
+      "grad_norm": 0.8732580542564392,
+      "learning_rate": 9.225621832057259e-05,
+      "loss": 0.0892,
+      "step": 39700
+    },
+    {
+      "epoch": 2.597971867844292,
+      "grad_norm": 0.894294261932373,
+      "learning_rate": 9.225130708009323e-05,
+      "loss": 0.0831,
+      "step": 39710
+    },
+    {
+      "epoch": 2.5986261040235528,
+      "grad_norm": 0.771438479423523,
+      "learning_rate": 9.224639441352145e-05,
+      "loss": 0.0812,
+      "step": 39720
+    },
+    {
+      "epoch": 2.5992803402028133,
+      "grad_norm": 1.0812944173812866,
+      "learning_rate": 9.224148032102307e-05,
+      "loss": 0.0876,
+      "step": 39730
+    },
+    {
+      "epoch": 2.599934576382074,
+      "grad_norm": 0.8493326902389526,
+      "learning_rate": 9.223656480276394e-05,
+      "loss": 0.0823,
+      "step": 39740
+    },
+    {
+      "epoch": 2.600588812561335,
+      "grad_norm": 0.8835618495941162,
+      "learning_rate": 9.223164785890997e-05,
+      "loss": 0.0766,
+      "step": 39750
+    },
+    {
+      "epoch": 2.6012430487405953,
+      "grad_norm": 0.8064636588096619,
+      "learning_rate": 9.222672948962713e-05,
+      "loss": 0.0913,
+      "step": 39760
+    },
+    {
+      "epoch": 2.601897284919856,
+      "grad_norm": 1.0927833318710327,
+      "learning_rate": 9.222180969508145e-05,
+      "loss": 0.0847,
+      "step": 39770
+    },
+    {
+      "epoch": 2.602551521099117,
+      "grad_norm": 0.9413180351257324,
+      "learning_rate": 9.221688847543894e-05,
+      "loss": 0.0852,
+      "step": 39780
+    },
+    {
+      "epoch": 2.6032057572783773,
+      "grad_norm": 0.8473165035247803,
+      "learning_rate": 9.221196583086573e-05,
+      "loss": 0.0854,
+      "step": 39790
+    },
+    {
+      "epoch": 2.6038599934576383,
+      "grad_norm": 0.8541494011878967,
+      "learning_rate": 9.220704176152797e-05,
+      "loss": 0.0803,
+      "step": 39800
+    },
+    {
+      "epoch": 2.604514229636899,
+      "grad_norm": 0.7438517808914185,
+      "learning_rate": 9.220211626759185e-05,
+      "loss": 0.0776,
+      "step": 39810
+    },
+    {
+      "epoch": 2.60516846581616,
+      "grad_norm": 0.6814586520195007,
+      "learning_rate": 9.219718934922364e-05,
+      "loss": 0.0854,
+      "step": 39820
+    },
+    {
+      "epoch": 2.6058227019954203,
+      "grad_norm": 0.7921211123466492,
+      "learning_rate": 9.219226100658962e-05,
+      "loss": 0.0748,
+      "step": 39830
+    },
+    {
+      "epoch": 2.606476938174681,
+      "grad_norm": 0.9955710768699646,
+      "learning_rate": 9.218733123985613e-05,
+      "loss": 0.0926,
+      "step": 39840
+    },
+    {
+      "epoch": 2.607131174353942,
+      "grad_norm": 0.7756005525588989,
+      "learning_rate": 9.218240004918958e-05,
+      "loss": 0.0792,
+      "step": 39850
+    },
+    {
+      "epoch": 2.6077854105332023,
+      "grad_norm": 0.7657015323638916,
+      "learning_rate": 9.21774674347564e-05,
+      "loss": 0.077,
+      "step": 39860
+    },
+    {
+      "epoch": 2.6084396467124633,
+      "grad_norm": 0.8866888880729675,
+      "learning_rate": 9.217253339672307e-05,
+      "loss": 0.0916,
+      "step": 39870
+    },
+    {
+      "epoch": 2.609093882891724,
+      "grad_norm": 1.0825079679489136,
+      "learning_rate": 9.216759793525615e-05,
+      "loss": 0.0777,
+      "step": 39880
+    },
+    {
+      "epoch": 2.609748119070985,
+      "grad_norm": 0.8313194513320923,
+      "learning_rate": 9.21626610505222e-05,
+      "loss": 0.0753,
+      "step": 39890
+    },
+    {
+      "epoch": 2.6104023552502453,
+      "grad_norm": 0.8642783164978027,
+      "learning_rate": 9.215772274268787e-05,
+      "loss": 0.0903,
+      "step": 39900
+    },
+    {
+      "epoch": 2.611056591429506,
+      "grad_norm": 0.8965114951133728,
+      "learning_rate": 9.215278301191982e-05,
+      "loss": 0.078,
+      "step": 39910
+    },
+    {
+      "epoch": 2.611710827608767,
+      "grad_norm": 0.7599003911018372,
+      "learning_rate": 9.214784185838483e-05,
+      "loss": 0.0928,
+      "step": 39920
+    },
+    {
+      "epoch": 2.6123650637880274,
+      "grad_norm": 1.0140457153320312,
+      "learning_rate": 9.21428992822496e-05,
+      "loss": 0.0797,
+      "step": 39930
+    },
+    {
+      "epoch": 2.6130192999672883,
+      "grad_norm": 0.9935689568519592,
+      "learning_rate": 9.213795528368102e-05,
+      "loss": 0.0964,
+      "step": 39940
+    },
+    {
+      "epoch": 2.613673536146549,
+      "grad_norm": 0.7364537715911865,
+      "learning_rate": 9.213300986284593e-05,
+      "loss": 0.0837,
+      "step": 39950
+    },
+    {
+      "epoch": 2.61432777232581,
+      "grad_norm": 0.7998794317245483,
+      "learning_rate": 9.212806301991125e-05,
+      "loss": 0.0843,
+      "step": 39960
+    },
+    {
+      "epoch": 2.6149820085050703,
+      "grad_norm": 0.8715366125106812,
+      "learning_rate": 9.212311475504398e-05,
+      "loss": 0.0864,
+      "step": 39970
+    },
+    {
+      "epoch": 2.615636244684331,
+      "grad_norm": 0.9843313694000244,
+      "learning_rate": 9.21181650684111e-05,
+      "loss": 0.0816,
+      "step": 39980
+    },
+    {
+      "epoch": 2.616290480863592,
+      "grad_norm": 0.8258795738220215,
+      "learning_rate": 9.21132139601797e-05,
+      "loss": 0.0795,
+      "step": 39990
+    },
+    {
+      "epoch": 2.6169447170428524,
+      "grad_norm": 0.8979169130325317,
+      "learning_rate": 9.210826143051688e-05,
+      "loss": 0.0815,
+      "step": 40000
+    },
+    {
+      "epoch": 2.6175989532221133,
+      "grad_norm": 0.8749635815620422,
+      "learning_rate": 9.210330747958979e-05,
+      "loss": 0.0771,
+      "step": 40010
+    },
+    {
+      "epoch": 2.618253189401374,
+      "grad_norm": 0.8296936750411987,
+      "learning_rate": 9.209835210756565e-05,
+      "loss": 0.0793,
+      "step": 40020
+    },
+    {
+      "epoch": 2.618907425580635,
+      "grad_norm": 0.7305333614349365,
+      "learning_rate": 9.209339531461173e-05,
+      "loss": 0.0747,
+      "step": 40030
+    },
+    {
+      "epoch": 2.6195616617598954,
+      "grad_norm": 0.888248860836029,
+      "learning_rate": 9.208843710089534e-05,
+      "loss": 0.0855,
+      "step": 40040
+    },
+    {
+      "epoch": 2.620215897939156,
+      "grad_norm": 0.7976614236831665,
+      "learning_rate": 9.20834774665838e-05,
+      "loss": 0.0876,
+      "step": 40050
+    },
+    {
+      "epoch": 2.620870134118417,
+      "grad_norm": 1.2806037664413452,
+      "learning_rate": 9.207851641184453e-05,
+      "loss": 0.0805,
+      "step": 40060
+    },
+    {
+      "epoch": 2.6215243702976774,
+      "grad_norm": 0.7971034049987793,
+      "learning_rate": 9.207355393684499e-05,
+      "loss": 0.0706,
+      "step": 40070
+    },
+    {
+      "epoch": 2.622178606476938,
+      "grad_norm": 0.7905651330947876,
+      "learning_rate": 9.206859004175264e-05,
+      "loss": 0.0784,
+      "step": 40080
+    },
+    {
+      "epoch": 2.622832842656199,
+      "grad_norm": 0.7352070808410645,
+      "learning_rate": 9.206362472673505e-05,
+      "loss": 0.0723,
+      "step": 40090
+    },
+    {
+      "epoch": 2.62348707883546,
+      "grad_norm": 0.9722936153411865,
+      "learning_rate": 9.205865799195982e-05,
+      "loss": 0.081,
+      "step": 40100
+    },
+    {
+      "epoch": 2.6241413150147204,
+      "grad_norm": 0.849582850933075,
+      "learning_rate": 9.205368983759457e-05,
+      "loss": 0.0933,
+      "step": 40110
+    },
+    {
+      "epoch": 2.624795551193981,
+      "grad_norm": 0.7181478142738342,
+      "learning_rate": 9.2048720263807e-05,
+      "loss": 0.0824,
+      "step": 40120
+    },
+    {
+      "epoch": 2.625449787373242,
+      "grad_norm": 0.8331068158149719,
+      "learning_rate": 9.204374927076486e-05,
+      "loss": 0.0774,
+      "step": 40130
+    },
+    {
+      "epoch": 2.6261040235525024,
+      "grad_norm": 0.7740342617034912,
+      "learning_rate": 9.20387768586359e-05,
+      "loss": 0.0791,
+      "step": 40140
+    },
+    {
+      "epoch": 2.626758259731763,
+      "grad_norm": 0.7920116186141968,
+      "learning_rate": 9.203380302758797e-05,
+      "loss": 0.0765,
+      "step": 40150
+    },
+    {
+      "epoch": 2.627412495911024,
+      "grad_norm": 0.8771019577980042,
+      "learning_rate": 9.202882777778896e-05,
+      "loss": 0.08,
+      "step": 40160
+    },
+    {
+      "epoch": 2.628066732090285,
+      "grad_norm": 0.9749835729598999,
+      "learning_rate": 9.202385110940678e-05,
+      "loss": 0.0779,
+      "step": 40170
+    },
+    {
+      "epoch": 2.6287209682695454,
+      "grad_norm": 0.8514975309371948,
+      "learning_rate": 9.201887302260943e-05,
+      "loss": 0.0887,
+      "step": 40180
+    },
+    {
+      "epoch": 2.629375204448806,
+      "grad_norm": 0.820841372013092,
+      "learning_rate": 9.201389351756491e-05,
+      "loss": 0.0747,
+      "step": 40190
+    },
+    {
+      "epoch": 2.630029440628067,
+      "grad_norm": 0.9230198860168457,
+      "learning_rate": 9.200891259444129e-05,
+      "loss": 0.0867,
+      "step": 40200
+    },
+    {
+      "epoch": 2.6306836768073274,
+      "grad_norm": 0.7532345056533813,
+      "learning_rate": 9.20039302534067e-05,
+      "loss": 0.0861,
+      "step": 40210
+    },
+    {
+      "epoch": 2.631337912986588,
+      "grad_norm": 0.7638508677482605,
+      "learning_rate": 9.19989464946293e-05,
+      "loss": 0.087,
+      "step": 40220
+    },
+    {
+      "epoch": 2.631992149165849,
+      "grad_norm": 0.7697664499282837,
+      "learning_rate": 9.199396131827731e-05,
+      "loss": 0.0725,
+      "step": 40230
+    },
+    {
+      "epoch": 2.6326463853451094,
+      "grad_norm": 0.7672895193099976,
+      "learning_rate": 9.198897472451901e-05,
+      "loss": 0.086,
+      "step": 40240
+    },
+    {
+      "epoch": 2.6333006215243704,
+      "grad_norm": 0.830986499786377,
+      "learning_rate": 9.198398671352267e-05,
+      "loss": 0.0793,
+      "step": 40250
+    },
+    {
+      "epoch": 2.633954857703631,
+      "grad_norm": 0.8693823218345642,
+      "learning_rate": 9.19789972854567e-05,
+      "loss": 0.0868,
+      "step": 40260
+    },
+    {
+      "epoch": 2.634609093882892,
+      "grad_norm": 0.8019348978996277,
+      "learning_rate": 9.197400644048944e-05,
+      "loss": 0.0741,
+      "step": 40270
+    },
+    {
+      "epoch": 2.6352633300621524,
+      "grad_norm": 0.9302101731300354,
+      "learning_rate": 9.196901417878941e-05,
+      "loss": 0.0779,
+      "step": 40280
+    },
+    {
+      "epoch": 2.635917566241413,
+      "grad_norm": 0.9065577387809753,
+      "learning_rate": 9.196402050052507e-05,
+      "loss": 0.0985,
+      "step": 40290
+    },
+    {
+      "epoch": 2.636571802420674,
+      "grad_norm": 0.8283711671829224,
+      "learning_rate": 9.195902540586498e-05,
+      "loss": 0.0797,
+      "step": 40300
+    },
+    {
+      "epoch": 2.6372260385999344,
+      "grad_norm": 0.9816059470176697,
+      "learning_rate": 9.195402889497777e-05,
+      "loss": 0.0925,
+      "step": 40310
+    },
+    {
+      "epoch": 2.6378802747791954,
+      "grad_norm": 0.7747599482536316,
+      "learning_rate": 9.194903096803202e-05,
+      "loss": 0.0841,
+      "step": 40320
+    },
+    {
+      "epoch": 2.638534510958456,
+      "grad_norm": 0.9950234889984131,
+      "learning_rate": 9.194403162519648e-05,
+      "loss": 0.0904,
+      "step": 40330
+    },
+    {
+      "epoch": 2.639188747137717,
+      "grad_norm": 1.1040619611740112,
+      "learning_rate": 9.193903086663987e-05,
+      "loss": 0.0822,
+      "step": 40340
+    },
+    {
+      "epoch": 2.6398429833169774,
+      "grad_norm": 1.090773582458496,
+      "learning_rate": 9.193402869253097e-05,
+      "loss": 0.0944,
+      "step": 40350
+    },
+    {
+      "epoch": 2.640497219496238,
+      "grad_norm": 1.1542222499847412,
+      "learning_rate": 9.192902510303862e-05,
+      "loss": 0.0864,
+      "step": 40360
+    },
+    {
+      "epoch": 2.641151455675499,
+      "grad_norm": 0.9892890453338623,
+      "learning_rate": 9.192402009833173e-05,
+      "loss": 0.0861,
+      "step": 40370
+    },
+    {
+      "epoch": 2.6418056918547594,
+      "grad_norm": 1.2127673625946045,
+      "learning_rate": 9.19190136785792e-05,
+      "loss": 0.0837,
+      "step": 40380
+    },
+    {
+      "epoch": 2.6424599280340204,
+      "grad_norm": 0.8235146999359131,
+      "learning_rate": 9.191400584395003e-05,
+      "loss": 0.0825,
+      "step": 40390
+    },
+    {
+      "epoch": 2.643114164213281,
+      "grad_norm": 0.7639740109443665,
+      "learning_rate": 9.190899659461323e-05,
+      "loss": 0.0767,
+      "step": 40400
+    },
+    {
+      "epoch": 2.643768400392542,
+      "grad_norm": 0.8715619444847107,
+      "learning_rate": 9.19039859307379e-05,
+      "loss": 0.0795,
+      "step": 40410
+    },
+    {
+      "epoch": 2.6444226365718024,
+      "grad_norm": 0.8103946447372437,
+      "learning_rate": 9.189897385249313e-05,
+      "loss": 0.0898,
+      "step": 40420
+    },
+    {
+      "epoch": 2.645076872751063,
+      "grad_norm": 0.9815798997879028,
+      "learning_rate": 9.189396036004811e-05,
+      "loss": 0.0782,
+      "step": 40430
+    },
+    {
+      "epoch": 2.645731108930324,
+      "grad_norm": 0.884543776512146,
+      "learning_rate": 9.188894545357207e-05,
+      "loss": 0.0775,
+      "step": 40440
+    },
+    {
+      "epoch": 2.6463853451095845,
+      "grad_norm": 0.8589003682136536,
+      "learning_rate": 9.188392913323423e-05,
+      "loss": 0.0867,
+      "step": 40450
+    },
+    {
+      "epoch": 2.6470395812888454,
+      "grad_norm": 0.8187326192855835,
+      "learning_rate": 9.187891139920397e-05,
+      "loss": 0.081,
+      "step": 40460
+    },
+    {
+      "epoch": 2.647693817468106,
+      "grad_norm": 0.8445239663124084,
+      "learning_rate": 9.187389225165062e-05,
+      "loss": 0.0812,
+      "step": 40470
+    },
+    {
+      "epoch": 2.648348053647367,
+      "grad_norm": 0.7465826869010925,
+      "learning_rate": 9.186887169074356e-05,
+      "loss": 0.08,
+      "step": 40480
+    },
+    {
+      "epoch": 2.6490022898266274,
+      "grad_norm": 0.8137046098709106,
+      "learning_rate": 9.186384971665229e-05,
+      "loss": 0.0792,
+      "step": 40490
+    },
+    {
+      "epoch": 2.649656526005888,
+      "grad_norm": 0.912079393863678,
+      "learning_rate": 9.185882632954632e-05,
+      "loss": 0.0809,
+      "step": 40500
+    },
+    {
+      "epoch": 2.650310762185149,
+      "grad_norm": 0.9054073691368103,
+      "learning_rate": 9.185380152959515e-05,
+      "loss": 0.0777,
+      "step": 40510
+    },
+    {
+      "epoch": 2.6509649983644095,
+      "grad_norm": 0.9537298083305359,
+      "learning_rate": 9.184877531696844e-05,
+      "loss": 0.0822,
+      "step": 40520
+    },
+    {
+      "epoch": 2.65161923454367,
+      "grad_norm": 0.8448399901390076,
+      "learning_rate": 9.18437476918358e-05,
+      "loss": 0.0788,
+      "step": 40530
+    },
+    {
+      "epoch": 2.652273470722931,
+      "grad_norm": 0.8022094368934631,
+      "learning_rate": 9.183871865436693e-05,
+      "loss": 0.0927,
+      "step": 40540
+    },
+    {
+      "epoch": 2.652927706902192,
+      "grad_norm": 0.7613011598587036,
+      "learning_rate": 9.183368820473159e-05,
+      "loss": 0.0743,
+      "step": 40550
+    },
+    {
+      "epoch": 2.6535819430814525,
+      "grad_norm": 0.8020244836807251,
+      "learning_rate": 9.182865634309956e-05,
+      "loss": 0.0729,
+      "step": 40560
+    },
+    {
+      "epoch": 2.654236179260713,
+      "grad_norm": 0.7869201898574829,
+      "learning_rate": 9.182362306964067e-05,
+      "loss": 0.0818,
+      "step": 40570
+    },
+    {
+      "epoch": 2.654890415439974,
+      "grad_norm": 0.859542965888977,
+      "learning_rate": 9.181858838452481e-05,
+      "loss": 0.0909,
+      "step": 40580
+    },
+    {
+      "epoch": 2.6555446516192345,
+      "grad_norm": 0.8157853484153748,
+      "learning_rate": 9.181355228792194e-05,
+      "loss": 0.0727,
+      "step": 40590
+    },
+    {
+      "epoch": 2.656198887798495,
+      "grad_norm": 0.8959032297134399,
+      "learning_rate": 9.180851478000199e-05,
+      "loss": 0.0762,
+      "step": 40600
+    },
+    {
+      "epoch": 2.656853123977756,
+      "grad_norm": 0.759432315826416,
+      "learning_rate": 9.180347586093505e-05,
+      "loss": 0.0805,
+      "step": 40610
+    },
+    {
+      "epoch": 2.657507360157017,
+      "grad_norm": 0.842648983001709,
+      "learning_rate": 9.179843553089114e-05,
+      "loss": 0.0766,
+      "step": 40620
+    },
+    {
+      "epoch": 2.6581615963362775,
+      "grad_norm": 0.8938133120536804,
+      "learning_rate": 9.179339379004043e-05,
+      "loss": 0.0839,
+      "step": 40630
+    },
+    {
+      "epoch": 2.658815832515538,
+      "grad_norm": 1.022493600845337,
+      "learning_rate": 9.178835063855306e-05,
+      "loss": 0.0812,
+      "step": 40640
+    },
+    {
+      "epoch": 2.659470068694799,
+      "grad_norm": 0.8210968375205994,
+      "learning_rate": 9.178330607659927e-05,
+      "loss": 0.0804,
+      "step": 40650
+    },
+    {
+      "epoch": 2.6601243048740595,
+      "grad_norm": 0.7599284648895264,
+      "learning_rate": 9.177826010434931e-05,
+      "loss": 0.0731,
+      "step": 40660
+    },
+    {
+      "epoch": 2.66077854105332,
+      "grad_norm": 0.8126060962677002,
+      "learning_rate": 9.177321272197352e-05,
+      "loss": 0.079,
+      "step": 40670
+    },
+    {
+      "epoch": 2.661432777232581,
+      "grad_norm": 0.9725340604782104,
+      "learning_rate": 9.176816392964223e-05,
+      "loss": 0.0808,
+      "step": 40680
+    },
+    {
+      "epoch": 2.662087013411842,
+      "grad_norm": 0.7665135264396667,
+      "learning_rate": 9.176311372752589e-05,
+      "loss": 0.0781,
+      "step": 40690
+    },
+    {
+      "epoch": 2.6627412495911025,
+      "grad_norm": 0.9257283806800842,
+      "learning_rate": 9.175806211579491e-05,
+      "loss": 0.088,
+      "step": 40700
+    },
+    {
+      "epoch": 2.663395485770363,
+      "grad_norm": 0.9089275598526001,
+      "learning_rate": 9.175300909461982e-05,
+      "loss": 0.0823,
+      "step": 40710
+    },
+    {
+      "epoch": 2.664049721949624,
+      "grad_norm": 0.8126506209373474,
+      "learning_rate": 9.174795466417119e-05,
+      "loss": 0.0721,
+      "step": 40720
+    },
+    {
+      "epoch": 2.6647039581288845,
+      "grad_norm": 1.0364658832550049,
+      "learning_rate": 9.17428988246196e-05,
+      "loss": 0.0827,
+      "step": 40730
+    },
+    {
+      "epoch": 2.665358194308145,
+      "grad_norm": 0.8109714984893799,
+      "learning_rate": 9.173784157613568e-05,
+      "loss": 0.0815,
+      "step": 40740
+    },
+    {
+      "epoch": 2.666012430487406,
+      "grad_norm": 0.8420194387435913,
+      "learning_rate": 9.173278291889015e-05,
+      "loss": 0.0824,
+      "step": 40750
+    },
+    {
+      "epoch": 2.6666666666666665,
+      "grad_norm": 0.7918026447296143,
+      "learning_rate": 9.172772285305375e-05,
+      "loss": 0.0894,
+      "step": 40760
+    },
+    {
+      "epoch": 2.6673209028459275,
+      "grad_norm": 0.9323230385780334,
+      "learning_rate": 9.172266137879728e-05,
+      "loss": 0.0864,
+      "step": 40770
+    },
+    {
+      "epoch": 2.667975139025188,
+      "grad_norm": 0.8225651383399963,
+      "learning_rate": 9.171759849629155e-05,
+      "loss": 0.0806,
+      "step": 40780
+    },
+    {
+      "epoch": 2.668629375204449,
+      "grad_norm": 0.9506831765174866,
+      "learning_rate": 9.171253420570748e-05,
+      "loss": 0.0861,
+      "step": 40790
+    },
+    {
+      "epoch": 2.6692836113837095,
+      "grad_norm": 0.8589584827423096,
+      "learning_rate": 9.170746850721598e-05,
+      "loss": 0.0791,
+      "step": 40800
+    },
+    {
+      "epoch": 2.66993784756297,
+      "grad_norm": 0.9063448309898376,
+      "learning_rate": 9.170240140098802e-05,
+      "loss": 0.0776,
+      "step": 40810
+    },
+    {
+      "epoch": 2.670592083742231,
+      "grad_norm": 0.7764676809310913,
+      "learning_rate": 9.169733288719467e-05,
+      "loss": 0.0784,
+      "step": 40820
+    },
+    {
+      "epoch": 2.6712463199214915,
+      "grad_norm": 0.8123670816421509,
+      "learning_rate": 9.169226296600696e-05,
+      "loss": 0.0658,
+      "step": 40830
+    },
+    {
+      "epoch": 2.6719005561007525,
+      "grad_norm": 0.8324966430664062,
+      "learning_rate": 9.168719163759604e-05,
+      "loss": 0.0741,
+      "step": 40840
+    },
+    {
+      "epoch": 2.672554792280013,
+      "grad_norm": 0.925599217414856,
+      "learning_rate": 9.168211890213307e-05,
+      "loss": 0.0765,
+      "step": 40850
+    },
+    {
+      "epoch": 2.673209028459274,
+      "grad_norm": 1.0423486232757568,
+      "learning_rate": 9.167704475978928e-05,
+      "loss": 0.0788,
+      "step": 40860
+    },
+    {
+      "epoch": 2.6738632646385345,
+      "grad_norm": 0.8119267821311951,
+      "learning_rate": 9.167196921073593e-05,
+      "loss": 0.0855,
+      "step": 40870
+    },
+    {
+      "epoch": 2.674517500817795,
+      "grad_norm": 0.9077923893928528,
+      "learning_rate": 9.166689225514431e-05,
+      "loss": 0.0715,
+      "step": 40880
+    },
+    {
+      "epoch": 2.675171736997056,
+      "grad_norm": 0.9296697378158569,
+      "learning_rate": 9.166181389318583e-05,
+      "loss": 0.0768,
+      "step": 40890
+    },
+    {
+      "epoch": 2.6758259731763165,
+      "grad_norm": 0.7002086043357849,
+      "learning_rate": 9.165673412503183e-05,
+      "loss": 0.0896,
+      "step": 40900
+    },
+    {
+      "epoch": 2.6764802093555775,
+      "grad_norm": 0.7667686939239502,
+      "learning_rate": 9.165165295085385e-05,
+      "loss": 0.0858,
+      "step": 40910
+    },
+    {
+      "epoch": 2.677134445534838,
+      "grad_norm": 1.128864049911499,
+      "learning_rate": 9.164657037082331e-05,
+      "loss": 0.0809,
+      "step": 40920
+    },
+    {
+      "epoch": 2.677788681714099,
+      "grad_norm": 0.8613152503967285,
+      "learning_rate": 9.164148638511182e-05,
+      "loss": 0.0882,
+      "step": 40930
+    },
+    {
+      "epoch": 2.6784429178933595,
+      "grad_norm": 0.7017923593521118,
+      "learning_rate": 9.163640099389095e-05,
+      "loss": 0.0807,
+      "step": 40940
+    },
+    {
+      "epoch": 2.67909715407262,
+      "grad_norm": 0.8696222305297852,
+      "learning_rate": 9.163131419733235e-05,
+      "loss": 0.0817,
+      "step": 40950
+    },
+    {
+      "epoch": 2.679751390251881,
+      "grad_norm": 0.950098991394043,
+      "learning_rate": 9.16262259956077e-05,
+      "loss": 0.0775,
+      "step": 40960
+    },
+    {
+      "epoch": 2.6804056264311416,
+      "grad_norm": 1.0551055669784546,
+      "learning_rate": 9.162113638888879e-05,
+      "loss": 0.0838,
+      "step": 40970
+    },
+    {
+      "epoch": 2.681059862610402,
+      "grad_norm": 0.8011052012443542,
+      "learning_rate": 9.161604537734733e-05,
+      "loss": 0.0891,
+      "step": 40980
+    },
+    {
+      "epoch": 2.681714098789663,
+      "grad_norm": 0.9275655746459961,
+      "learning_rate": 9.161095296115523e-05,
+      "loss": 0.0828,
+      "step": 40990
+    },
+    {
+      "epoch": 2.682368334968924,
+      "grad_norm": 1.1314018964767456,
+      "learning_rate": 9.160585914048432e-05,
+      "loss": 0.0815,
+      "step": 41000
+    },
+    {
+      "epoch": 2.6830225711481845,
+      "grad_norm": 0.9840704202651978,
+      "learning_rate": 9.160076391550654e-05,
+      "loss": 0.0804,
+      "step": 41010
+    },
+    {
+      "epoch": 2.683676807327445,
+      "grad_norm": 0.8175294399261475,
+      "learning_rate": 9.15956672863939e-05,
+      "loss": 0.0729,
+      "step": 41020
+    },
+    {
+      "epoch": 2.684331043506706,
+      "grad_norm": 0.7896432280540466,
+      "learning_rate": 9.159056925331837e-05,
+      "loss": 0.079,
+      "step": 41030
+    },
+    {
+      "epoch": 2.6849852796859666,
+      "grad_norm": 0.7831984758377075,
+      "learning_rate": 9.158546981645208e-05,
+      "loss": 0.0845,
+      "step": 41040
+    },
+    {
+      "epoch": 2.685639515865227,
+      "grad_norm": 0.8683128952980042,
+      "learning_rate": 9.158036897596712e-05,
+      "loss": 0.0869,
+      "step": 41050
+    },
+    {
+      "epoch": 2.686293752044488,
+      "grad_norm": 0.8791904449462891,
+      "learning_rate": 9.157526673203565e-05,
+      "loss": 0.0835,
+      "step": 41060
+    },
+    {
+      "epoch": 2.686947988223749,
+      "grad_norm": 0.7771376371383667,
+      "learning_rate": 9.15701630848299e-05,
+      "loss": 0.0825,
+      "step": 41070
+    },
+    {
+      "epoch": 2.6876022244030096,
+      "grad_norm": 0.6314225792884827,
+      "learning_rate": 9.156505803452212e-05,
+      "loss": 0.0699,
+      "step": 41080
+    },
+    {
+      "epoch": 2.68825646058227,
+      "grad_norm": 1.073372483253479,
+      "learning_rate": 9.15599515812846e-05,
+      "loss": 0.0892,
+      "step": 41090
+    },
+    {
+      "epoch": 2.688910696761531,
+      "grad_norm": 0.9210318922996521,
+      "learning_rate": 9.155484372528975e-05,
+      "loss": 0.0779,
+      "step": 41100
+    },
+    {
+      "epoch": 2.6895649329407916,
+      "grad_norm": 0.8791244626045227,
+      "learning_rate": 9.154973446670993e-05,
+      "loss": 0.0981,
+      "step": 41110
+    },
+    {
+      "epoch": 2.690219169120052,
+      "grad_norm": 0.8100093007087708,
+      "learning_rate": 9.154462380571761e-05,
+      "loss": 0.0867,
+      "step": 41120
+    },
+    {
+      "epoch": 2.690873405299313,
+      "grad_norm": 1.020846962928772,
+      "learning_rate": 9.153951174248528e-05,
+      "loss": 0.0945,
+      "step": 41130
+    },
+    {
+      "epoch": 2.691527641478574,
+      "grad_norm": 0.929047167301178,
+      "learning_rate": 9.153439827718549e-05,
+      "loss": 0.0766,
+      "step": 41140
+    },
+    {
+      "epoch": 2.6921818776578346,
+      "grad_norm": 0.8004492521286011,
+      "learning_rate": 9.152928340999083e-05,
+      "loss": 0.0754,
+      "step": 41150
+    },
+    {
+      "epoch": 2.692836113837095,
+      "grad_norm": 0.7606692910194397,
+      "learning_rate": 9.152416714107393e-05,
+      "loss": 0.0761,
+      "step": 41160
+    },
+    {
+      "epoch": 2.693490350016356,
+      "grad_norm": 0.9273959994316101,
+      "learning_rate": 9.15190494706075e-05,
+      "loss": 0.0838,
+      "step": 41170
+    },
+    {
+      "epoch": 2.6941445861956166,
+      "grad_norm": 0.8332855701446533,
+      "learning_rate": 9.151393039876425e-05,
+      "loss": 0.0781,
+      "step": 41180
+    },
+    {
+      "epoch": 2.694798822374877,
+      "grad_norm": 0.6218408942222595,
+      "learning_rate": 9.150880992571698e-05,
+      "loss": 0.0824,
+      "step": 41190
+    },
+    {
+      "epoch": 2.695453058554138,
+      "grad_norm": 0.8714901804924011,
+      "learning_rate": 9.150368805163851e-05,
+      "loss": 0.0875,
+      "step": 41200
+    },
+    {
+      "epoch": 2.6961072947333986,
+      "grad_norm": 0.7702926397323608,
+      "learning_rate": 9.149856477670173e-05,
+      "loss": 0.0842,
+      "step": 41210
+    },
+    {
+      "epoch": 2.6967615309126596,
+      "grad_norm": 0.8732930421829224,
+      "learning_rate": 9.149344010107955e-05,
+      "loss": 0.086,
+      "step": 41220
+    },
+    {
+      "epoch": 2.69741576709192,
+      "grad_norm": 0.8141332268714905,
+      "learning_rate": 9.148831402494495e-05,
+      "loss": 0.0917,
+      "step": 41230
+    },
+    {
+      "epoch": 2.698070003271181,
+      "grad_norm": 0.9626291990280151,
+      "learning_rate": 9.148318654847094e-05,
+      "loss": 0.0798,
+      "step": 41240
+    },
+    {
+      "epoch": 2.6987242394504416,
+      "grad_norm": 0.8662136793136597,
+      "learning_rate": 9.14780576718306e-05,
+      "loss": 0.0837,
+      "step": 41250
+    },
+    {
+      "epoch": 2.699378475629702,
+      "grad_norm": 0.842068076133728,
+      "learning_rate": 9.147292739519702e-05,
+      "loss": 0.0848,
+      "step": 41260
+    },
+    {
+      "epoch": 2.700032711808963,
+      "grad_norm": 0.8036767244338989,
+      "learning_rate": 9.146779571874337e-05,
+      "loss": 0.071,
+      "step": 41270
+    },
+    {
+      "epoch": 2.7006869479882236,
+      "grad_norm": 0.8152961134910583,
+      "learning_rate": 9.146266264264288e-05,
+      "loss": 0.0795,
+      "step": 41280
+    },
+    {
+      "epoch": 2.7013411841674846,
+      "grad_norm": 0.9094104766845703,
+      "learning_rate": 9.145752816706878e-05,
+      "loss": 0.0715,
+      "step": 41290
+    },
+    {
+      "epoch": 2.701995420346745,
+      "grad_norm": 0.8534713983535767,
+      "learning_rate": 9.145239229219438e-05,
+      "loss": 0.0788,
+      "step": 41300
+    },
+    {
+      "epoch": 2.702649656526006,
+      "grad_norm": 0.947786271572113,
+      "learning_rate": 9.144725501819303e-05,
+      "loss": 0.0888,
+      "step": 41310
+    },
+    {
+      "epoch": 2.7033038927052666,
+      "grad_norm": 0.9250187277793884,
+      "learning_rate": 9.14421163452381e-05,
+      "loss": 0.0849,
+      "step": 41320
+    },
+    {
+      "epoch": 2.703958128884527,
+      "grad_norm": 0.9689621925354004,
+      "learning_rate": 9.143697627350308e-05,
+      "loss": 0.0792,
+      "step": 41330
+    },
+    {
+      "epoch": 2.704612365063788,
+      "grad_norm": 0.9268077611923218,
+      "learning_rate": 9.143183480316143e-05,
+      "loss": 0.079,
+      "step": 41340
+    },
+    {
+      "epoch": 2.7052666012430486,
+      "grad_norm": 0.7102203369140625,
+      "learning_rate": 9.142669193438669e-05,
+      "loss": 0.0812,
+      "step": 41350
+    },
+    {
+      "epoch": 2.7059208374223096,
+      "grad_norm": 0.8151041269302368,
+      "learning_rate": 9.142154766735247e-05,
+      "loss": 0.0776,
+      "step": 41360
+    },
+    {
+      "epoch": 2.70657507360157,
+      "grad_norm": 1.0687074661254883,
+      "learning_rate": 9.141640200223236e-05,
+      "loss": 0.0804,
+      "step": 41370
+    },
+    {
+      "epoch": 2.707229309780831,
+      "grad_norm": 0.9339028596878052,
+      "learning_rate": 9.141125493920009e-05,
+      "loss": 0.0873,
+      "step": 41380
+    },
+    {
+      "epoch": 2.7078835459600916,
+      "grad_norm": 0.8078384399414062,
+      "learning_rate": 9.140610647842934e-05,
+      "loss": 0.0742,
+      "step": 41390
+    },
+    {
+      "epoch": 2.708537782139352,
+      "grad_norm": 1.0456902980804443,
+      "learning_rate": 9.140095662009392e-05,
+      "loss": 0.0764,
+      "step": 41400
+    },
+    {
+      "epoch": 2.709192018318613,
+      "grad_norm": 0.9626041650772095,
+      "learning_rate": 9.139580536436763e-05,
+      "loss": 0.075,
+      "step": 41410
+    },
+    {
+      "epoch": 2.7098462544978736,
+      "grad_norm": 1.0236674547195435,
+      "learning_rate": 9.139065271142434e-05,
+      "loss": 0.0861,
+      "step": 41420
+    },
+    {
+      "epoch": 2.7105004906771346,
+      "grad_norm": 0.8825036287307739,
+      "learning_rate": 9.138549866143797e-05,
+      "loss": 0.0693,
+      "step": 41430
+    },
+    {
+      "epoch": 2.711154726856395,
+      "grad_norm": 1.041279911994934,
+      "learning_rate": 9.138034321458248e-05,
+      "loss": 0.0874,
+      "step": 41440
+    },
+    {
+      "epoch": 2.711808963035656,
+      "grad_norm": 0.7983617782592773,
+      "learning_rate": 9.13751863710319e-05,
+      "loss": 0.082,
+      "step": 41450
+    },
+    {
+      "epoch": 2.7124631992149166,
+      "grad_norm": 0.7135721445083618,
+      "learning_rate": 9.137002813096026e-05,
+      "loss": 0.0747,
+      "step": 41460
+    },
+    {
+      "epoch": 2.713117435394177,
+      "grad_norm": 0.9826575517654419,
+      "learning_rate": 9.136486849454167e-05,
+      "loss": 0.0825,
+      "step": 41470
+    },
+    {
+      "epoch": 2.713771671573438,
+      "grad_norm": 0.9839840531349182,
+      "learning_rate": 9.135970746195029e-05,
+      "loss": 0.0923,
+      "step": 41480
+    },
+    {
+      "epoch": 2.7144259077526987,
+      "grad_norm": 0.9663382172584534,
+      "learning_rate": 9.135454503336032e-05,
+      "loss": 0.0676,
+      "step": 41490
+    },
+    {
+      "epoch": 2.715080143931959,
+      "grad_norm": 0.8061355352401733,
+      "learning_rate": 9.134938120894598e-05,
+      "loss": 0.0809,
+      "step": 41500
+    },
+    {
+      "epoch": 2.71573438011122,
+      "grad_norm": 0.8093533515930176,
+      "learning_rate": 9.134421598888159e-05,
+      "loss": 0.0772,
+      "step": 41510
+    },
+    {
+      "epoch": 2.716388616290481,
+      "grad_norm": 0.7529310584068298,
+      "learning_rate": 9.133904937334148e-05,
+      "loss": 0.0766,
+      "step": 41520
+    },
+    {
+      "epoch": 2.7170428524697416,
+      "grad_norm": 0.7155537605285645,
+      "learning_rate": 9.133388136250005e-05,
+      "loss": 0.0707,
+      "step": 41530
+    },
+    {
+      "epoch": 2.717697088649002,
+      "grad_norm": 0.8347744345664978,
+      "learning_rate": 9.13287119565317e-05,
+      "loss": 0.0737,
+      "step": 41540
+    },
+    {
+      "epoch": 2.718351324828263,
+      "grad_norm": 0.7836735844612122,
+      "learning_rate": 9.132354115561094e-05,
+      "loss": 0.0874,
+      "step": 41550
+    },
+    {
+      "epoch": 2.7190055610075237,
+      "grad_norm": 0.8764668107032776,
+      "learning_rate": 9.13183689599123e-05,
+      "loss": 0.0857,
+      "step": 41560
+    },
+    {
+      "epoch": 2.719659797186784,
+      "grad_norm": 0.7778729200363159,
+      "learning_rate": 9.131319536961035e-05,
+      "loss": 0.0865,
+      "step": 41570
+    },
+    {
+      "epoch": 2.720314033366045,
+      "grad_norm": 0.8253190517425537,
+      "learning_rate": 9.13080203848797e-05,
+      "loss": 0.0805,
+      "step": 41580
+    },
+    {
+      "epoch": 2.720968269545306,
+      "grad_norm": 0.7873772382736206,
+      "learning_rate": 9.130284400589503e-05,
+      "loss": 0.0815,
+      "step": 41590
+    },
+    {
+      "epoch": 2.7216225057245667,
+      "grad_norm": 0.7975731492042542,
+      "learning_rate": 9.129766623283105e-05,
+      "loss": 0.0859,
+      "step": 41600
+    },
+    {
+      "epoch": 2.722276741903827,
+      "grad_norm": 0.8026714324951172,
+      "learning_rate": 9.129248706586253e-05,
+      "loss": 0.0773,
+      "step": 41610
+    },
+    {
+      "epoch": 2.722930978083088,
+      "grad_norm": 0.8112697005271912,
+      "learning_rate": 9.128730650516429e-05,
+      "loss": 0.0743,
+      "step": 41620
+    },
+    {
+      "epoch": 2.7235852142623487,
+      "grad_norm": 0.8520511388778687,
+      "learning_rate": 9.128212455091115e-05,
+      "loss": 0.0847,
+      "step": 41630
+    },
+    {
+      "epoch": 2.724239450441609,
+      "grad_norm": 0.8587812185287476,
+      "learning_rate": 9.127694120327806e-05,
+      "loss": 0.0889,
+      "step": 41640
+    },
+    {
+      "epoch": 2.72489368662087,
+      "grad_norm": 1.0693318843841553,
+      "learning_rate": 9.127175646243994e-05,
+      "loss": 0.0834,
+      "step": 41650
+    },
+    {
+      "epoch": 2.7255479228001307,
+      "grad_norm": 0.8526062369346619,
+      "learning_rate": 9.126657032857181e-05,
+      "loss": 0.0772,
+      "step": 41660
+    },
+    {
+      "epoch": 2.7262021589793917,
+      "grad_norm": 1.0225043296813965,
+      "learning_rate": 9.126138280184871e-05,
+      "loss": 0.0982,
+      "step": 41670
+    },
+    {
+      "epoch": 2.726856395158652,
+      "grad_norm": 0.7854475378990173,
+      "learning_rate": 9.125619388244571e-05,
+      "loss": 0.0768,
+      "step": 41680
+    },
+    {
+      "epoch": 2.727510631337913,
+      "grad_norm": 0.929229736328125,
+      "learning_rate": 9.125100357053797e-05,
+      "loss": 0.0778,
+      "step": 41690
+    },
+    {
+      "epoch": 2.7281648675171737,
+      "grad_norm": 0.7836943864822388,
+      "learning_rate": 9.124581186630071e-05,
+      "loss": 0.0752,
+      "step": 41700
+    },
+    {
+      "epoch": 2.728819103696434,
+      "grad_norm": 0.9992619156837463,
+      "learning_rate": 9.124061876990908e-05,
+      "loss": 0.0795,
+      "step": 41710
+    },
+    {
+      "epoch": 2.729473339875695,
+      "grad_norm": 0.8057607412338257,
+      "learning_rate": 9.123542428153844e-05,
+      "loss": 0.072,
+      "step": 41720
+    },
+    {
+      "epoch": 2.7301275760549557,
+      "grad_norm": 0.8987215161323547,
+      "learning_rate": 9.123022840136407e-05,
+      "loss": 0.0805,
+      "step": 41730
+    },
+    {
+      "epoch": 2.7307818122342167,
+      "grad_norm": 0.8965569734573364,
+      "learning_rate": 9.122503112956138e-05,
+      "loss": 0.0797,
+      "step": 41740
+    },
+    {
+      "epoch": 2.731436048413477,
+      "grad_norm": 0.7639127969741821,
+      "learning_rate": 9.121983246630575e-05,
+      "loss": 0.0801,
+      "step": 41750
+    },
+    {
+      "epoch": 2.732090284592738,
+      "grad_norm": 0.83674156665802,
+      "learning_rate": 9.121463241177269e-05,
+      "loss": 0.0802,
+      "step": 41760
+    },
+    {
+      "epoch": 2.7327445207719987,
+      "grad_norm": 1.0172474384307861,
+      "learning_rate": 9.120943096613768e-05,
+      "loss": 0.0752,
+      "step": 41770
+    },
+    {
+      "epoch": 2.7333987569512592,
+      "grad_norm": 0.9789382815361023,
+      "learning_rate": 9.120422812957632e-05,
+      "loss": 0.0932,
+      "step": 41780
+    },
+    {
+      "epoch": 2.73405299313052,
+      "grad_norm": 0.934238612651825,
+      "learning_rate": 9.119902390226418e-05,
+      "loss": 0.0801,
+      "step": 41790
+    },
+    {
+      "epoch": 2.7347072293097807,
+      "grad_norm": 0.7868991494178772,
+      "learning_rate": 9.119381828437694e-05,
+      "loss": 0.0683,
+      "step": 41800
+    },
+    {
+      "epoch": 2.7353614654890417,
+      "grad_norm": 0.7209054827690125,
+      "learning_rate": 9.118861127609031e-05,
+      "loss": 0.0724,
+      "step": 41810
+    },
+    {
+      "epoch": 2.736015701668302,
+      "grad_norm": 0.8778037428855896,
+      "learning_rate": 9.118340287758001e-05,
+      "loss": 0.0798,
+      "step": 41820
+    },
+    {
+      "epoch": 2.736669937847563,
+      "grad_norm": 0.9034239053726196,
+      "learning_rate": 9.117819308902186e-05,
+      "loss": 0.0832,
+      "step": 41830
+    },
+    {
+      "epoch": 2.7373241740268237,
+      "grad_norm": 0.9231058359146118,
+      "learning_rate": 9.117298191059171e-05,
+      "loss": 0.0784,
+      "step": 41840
+    },
+    {
+      "epoch": 2.7379784102060842,
+      "grad_norm": 0.8640102744102478,
+      "learning_rate": 9.116776934246543e-05,
+      "loss": 0.0845,
+      "step": 41850
+    },
+    {
+      "epoch": 2.738632646385345,
+      "grad_norm": 0.9153121709823608,
+      "learning_rate": 9.116255538481896e-05,
+      "loss": 0.0832,
+      "step": 41860
+    },
+    {
+      "epoch": 2.7392868825646057,
+      "grad_norm": 0.7354351282119751,
+      "learning_rate": 9.115734003782832e-05,
+      "loss": 0.0759,
+      "step": 41870
+    },
+    {
+      "epoch": 2.7399411187438667,
+      "grad_norm": 0.9227734804153442,
+      "learning_rate": 9.115212330166949e-05,
+      "loss": 0.0839,
+      "step": 41880
+    },
+    {
+      "epoch": 2.7405953549231272,
+      "grad_norm": 0.8186841011047363,
+      "learning_rate": 9.114690517651859e-05,
+      "loss": 0.0802,
+      "step": 41890
+    },
+    {
+      "epoch": 2.741249591102388,
+      "grad_norm": 0.8185954093933105,
+      "learning_rate": 9.114168566255172e-05,
+      "loss": 0.0824,
+      "step": 41900
+    },
+    {
+      "epoch": 2.7419038272816487,
+      "grad_norm": 1.1469035148620605,
+      "learning_rate": 9.113646475994506e-05,
+      "loss": 0.0783,
+      "step": 41910
+    },
+    {
+      "epoch": 2.7425580634609092,
+      "grad_norm": 0.8847156763076782,
+      "learning_rate": 9.113124246887483e-05,
+      "loss": 0.0723,
+      "step": 41920
+    },
+    {
+      "epoch": 2.74321229964017,
+      "grad_norm": 0.870272696018219,
+      "learning_rate": 9.11260187895173e-05,
+      "loss": 0.0812,
+      "step": 41930
+    },
+    {
+      "epoch": 2.7438665358194307,
+      "grad_norm": 1.0734666585922241,
+      "learning_rate": 9.112079372204877e-05,
+      "loss": 0.0863,
+      "step": 41940
+    },
+    {
+      "epoch": 2.7445207719986913,
+      "grad_norm": 0.6786089539527893,
+      "learning_rate": 9.111556726664563e-05,
+      "loss": 0.0796,
+      "step": 41950
+    },
+    {
+      "epoch": 2.7451750081779522,
+      "grad_norm": 0.8662351369857788,
+      "learning_rate": 9.111033942348426e-05,
+      "loss": 0.0884,
+      "step": 41960
+    },
+    {
+      "epoch": 2.745829244357213,
+      "grad_norm": 0.7827929854393005,
+      "learning_rate": 9.110511019274111e-05,
+      "loss": 0.0764,
+      "step": 41970
+    },
+    {
+      "epoch": 2.7464834805364737,
+      "grad_norm": 1.1550228595733643,
+      "learning_rate": 9.109987957459267e-05,
+      "loss": 0.0845,
+      "step": 41980
+    },
+    {
+      "epoch": 2.7471377167157343,
+      "grad_norm": 0.8338910937309265,
+      "learning_rate": 9.109464756921553e-05,
+      "loss": 0.0824,
+      "step": 41990
+    },
+    {
+      "epoch": 2.7477919528949952,
+      "grad_norm": 0.9103622436523438,
+      "learning_rate": 9.108941417678626e-05,
+      "loss": 0.0896,
+      "step": 42000
+    },
+    {
+      "epoch": 2.7484461890742558,
+      "grad_norm": 0.8045547008514404,
+      "learning_rate": 9.108417939748149e-05,
+      "loss": 0.0805,
+      "step": 42010
+    },
+    {
+      "epoch": 2.7491004252535163,
+      "grad_norm": 0.8297812342643738,
+      "learning_rate": 9.107894323147792e-05,
+      "loss": 0.0819,
+      "step": 42020
+    },
+    {
+      "epoch": 2.7497546614327772,
+      "grad_norm": 0.8854258060455322,
+      "learning_rate": 9.107370567895229e-05,
+      "loss": 0.0837,
+      "step": 42030
+    },
+    {
+      "epoch": 2.750408897612038,
+      "grad_norm": 0.8766602277755737,
+      "learning_rate": 9.106846674008137e-05,
+      "loss": 0.0776,
+      "step": 42040
+    },
+    {
+      "epoch": 2.7510631337912987,
+      "grad_norm": 1.0775314569473267,
+      "learning_rate": 9.1063226415042e-05,
+      "loss": 0.0909,
+      "step": 42050
+    },
+    {
+      "epoch": 2.7517173699705593,
+      "grad_norm": 0.7865322828292847,
+      "learning_rate": 9.105798470401103e-05,
+      "loss": 0.0914,
+      "step": 42060
+    },
+    {
+      "epoch": 2.7523716061498202,
+      "grad_norm": 0.9470481872558594,
+      "learning_rate": 9.10527416071654e-05,
+      "loss": 0.0841,
+      "step": 42070
+    },
+    {
+      "epoch": 2.7530258423290808,
+      "grad_norm": 0.6476520299911499,
+      "learning_rate": 9.104749712468207e-05,
+      "loss": 0.0813,
+      "step": 42080
+    },
+    {
+      "epoch": 2.7536800785083413,
+      "grad_norm": 0.8619006276130676,
+      "learning_rate": 9.104225125673809e-05,
+      "loss": 0.08,
+      "step": 42090
+    },
+    {
+      "epoch": 2.7543343146876023,
+      "grad_norm": 0.7804111242294312,
+      "learning_rate": 9.103700400351047e-05,
+      "loss": 0.0889,
+      "step": 42100
+    },
+    {
+      "epoch": 2.754988550866863,
+      "grad_norm": 0.8227022290229797,
+      "learning_rate": 9.103175536517634e-05,
+      "loss": 0.08,
+      "step": 42110
+    },
+    {
+      "epoch": 2.7556427870461238,
+      "grad_norm": 0.8172056674957275,
+      "learning_rate": 9.102650534191287e-05,
+      "loss": 0.0863,
+      "step": 42120
+    },
+    {
+      "epoch": 2.7562970232253843,
+      "grad_norm": 0.8061167001724243,
+      "learning_rate": 9.102125393389725e-05,
+      "loss": 0.0707,
+      "step": 42130
+    },
+    {
+      "epoch": 2.7569512594046452,
+      "grad_norm": 0.9391945004463196,
+      "learning_rate": 9.10160011413067e-05,
+      "loss": 0.0777,
+      "step": 42140
+    },
+    {
+      "epoch": 2.7576054955839058,
+      "grad_norm": 1.0496068000793457,
+      "learning_rate": 9.101074696431858e-05,
+      "loss": 0.0854,
+      "step": 42150
+    },
+    {
+      "epoch": 2.7582597317631663,
+      "grad_norm": 0.8995298743247986,
+      "learning_rate": 9.100549140311017e-05,
+      "loss": 0.0802,
+      "step": 42160
+    },
+    {
+      "epoch": 2.7589139679424273,
+      "grad_norm": 0.8396767973899841,
+      "learning_rate": 9.100023445785889e-05,
+      "loss": 0.0736,
+      "step": 42170
+    },
+    {
+      "epoch": 2.759568204121688,
+      "grad_norm": 1.035605549812317,
+      "learning_rate": 9.099497612874217e-05,
+      "loss": 0.079,
+      "step": 42180
+    },
+    {
+      "epoch": 2.7602224403009488,
+      "grad_norm": 0.8803637027740479,
+      "learning_rate": 9.09897164159375e-05,
+      "loss": 0.0805,
+      "step": 42190
+    },
+    {
+      "epoch": 2.7608766764802093,
+      "grad_norm": 0.6873325109481812,
+      "learning_rate": 9.09844553196224e-05,
+      "loss": 0.0776,
+      "step": 42200
+    },
+    {
+      "epoch": 2.7615309126594703,
+      "grad_norm": 0.7511876225471497,
+      "learning_rate": 9.097919283997444e-05,
+      "loss": 0.0746,
+      "step": 42210
+    },
+    {
+      "epoch": 2.762185148838731,
+      "grad_norm": 0.7745641469955444,
+      "learning_rate": 9.097392897717126e-05,
+      "loss": 0.0829,
+      "step": 42220
+    },
+    {
+      "epoch": 2.7628393850179913,
+      "grad_norm": 0.8654570579528809,
+      "learning_rate": 9.09686637313905e-05,
+      "loss": 0.0783,
+      "step": 42230
+    },
+    {
+      "epoch": 2.7634936211972523,
+      "grad_norm": 0.8436213731765747,
+      "learning_rate": 9.09633971028099e-05,
+      "loss": 0.0826,
+      "step": 42240
+    },
+    {
+      "epoch": 2.764147857376513,
+      "grad_norm": 0.8295747637748718,
+      "learning_rate": 9.095812909160724e-05,
+      "loss": 0.0761,
+      "step": 42250
+    },
+    {
+      "epoch": 2.7648020935557738,
+      "grad_norm": 0.8413954377174377,
+      "learning_rate": 9.095285969796027e-05,
+      "loss": 0.0794,
+      "step": 42260
+    },
+    {
+      "epoch": 2.7654563297350343,
+      "grad_norm": 0.8630293607711792,
+      "learning_rate": 9.094758892204691e-05,
+      "loss": 0.0864,
+      "step": 42270
+    },
+    {
+      "epoch": 2.7661105659142953,
+      "grad_norm": 1.0896077156066895,
+      "learning_rate": 9.094231676404503e-05,
+      "loss": 0.0755,
+      "step": 42280
+    },
+    {
+      "epoch": 2.766764802093556,
+      "grad_norm": 0.9205356240272522,
+      "learning_rate": 9.093704322413259e-05,
+      "loss": 0.0888,
+      "step": 42290
+    },
+    {
+      "epoch": 2.7674190382728163,
+      "grad_norm": 0.8817057013511658,
+      "learning_rate": 9.093176830248756e-05,
+      "loss": 0.0755,
+      "step": 42300
+    },
+    {
+      "epoch": 2.7680732744520773,
+      "grad_norm": 0.9554494619369507,
+      "learning_rate": 9.092649199928802e-05,
+      "loss": 0.0742,
+      "step": 42310
+    },
+    {
+      "epoch": 2.768727510631338,
+      "grad_norm": 1.0500036478042603,
+      "learning_rate": 9.092121431471203e-05,
+      "loss": 0.0775,
+      "step": 42320
+    },
+    {
+      "epoch": 2.769381746810599,
+      "grad_norm": 1.2276242971420288,
+      "learning_rate": 9.091593524893776e-05,
+      "loss": 0.0775,
+      "step": 42330
+    },
+    {
+      "epoch": 2.7700359829898593,
+      "grad_norm": 0.8922539353370667,
+      "learning_rate": 9.091065480214333e-05,
+      "loss": 0.0841,
+      "step": 42340
+    },
+    {
+      "epoch": 2.7706902191691203,
+      "grad_norm": 0.9390623569488525,
+      "learning_rate": 9.090537297450703e-05,
+      "loss": 0.0816,
+      "step": 42350
+    },
+    {
+      "epoch": 2.771344455348381,
+      "grad_norm": 0.7786974310874939,
+      "learning_rate": 9.090008976620712e-05,
+      "loss": 0.0739,
+      "step": 42360
+    },
+    {
+      "epoch": 2.7719986915276413,
+      "grad_norm": 0.7234275341033936,
+      "learning_rate": 9.08948051774219e-05,
+      "loss": 0.077,
+      "step": 42370
+    },
+    {
+      "epoch": 2.7726529277069023,
+      "grad_norm": 0.8942136764526367,
+      "learning_rate": 9.088951920832978e-05,
+      "loss": 0.0838,
+      "step": 42380
+    },
+    {
+      "epoch": 2.773307163886163,
+      "grad_norm": 0.9873566031455994,
+      "learning_rate": 9.088423185910912e-05,
+      "loss": 0.0831,
+      "step": 42390
+    },
+    {
+      "epoch": 2.7739614000654234,
+      "grad_norm": 0.9224919080734253,
+      "learning_rate": 9.087894312993844e-05,
+      "loss": 0.0798,
+      "step": 42400
+    },
+    {
+      "epoch": 2.7746156362446843,
+      "grad_norm": 0.9617442488670349,
+      "learning_rate": 9.08736530209962e-05,
+      "loss": 0.082,
+      "step": 42410
+    },
+    {
+      "epoch": 2.7752698724239453,
+      "grad_norm": 0.9834263324737549,
+      "learning_rate": 9.086836153246099e-05,
+      "loss": 0.0846,
+      "step": 42420
+    },
+    {
+      "epoch": 2.775924108603206,
+      "grad_norm": 0.8556174635887146,
+      "learning_rate": 9.086306866451139e-05,
+      "loss": 0.0863,
+      "step": 42430
+    },
+    {
+      "epoch": 2.7765783447824663,
+      "grad_norm": 0.8653731346130371,
+      "learning_rate": 9.085777441732606e-05,
+      "loss": 0.0901,
+      "step": 42440
+    },
+    {
+      "epoch": 2.7772325809617273,
+      "grad_norm": 1.0214253664016724,
+      "learning_rate": 9.085247879108367e-05,
+      "loss": 0.0748,
+      "step": 42450
+    },
+    {
+      "epoch": 2.777886817140988,
+      "grad_norm": 0.7947637438774109,
+      "learning_rate": 9.084718178596301e-05,
+      "loss": 0.0846,
+      "step": 42460
+    },
+    {
+      "epoch": 2.7785410533202484,
+      "grad_norm": 0.9689314961433411,
+      "learning_rate": 9.084188340214281e-05,
+      "loss": 0.0747,
+      "step": 42470
+    },
+    {
+      "epoch": 2.7791952894995093,
+      "grad_norm": 0.9164670705795288,
+      "learning_rate": 9.083658363980196e-05,
+      "loss": 0.0836,
+      "step": 42480
+    },
+    {
+      "epoch": 2.7798495256787703,
+      "grad_norm": 1.0349498987197876,
+      "learning_rate": 9.08312824991193e-05,
+      "loss": 0.0858,
+      "step": 42490
+    },
+    {
+      "epoch": 2.780503761858031,
+      "grad_norm": 0.8005316257476807,
+      "learning_rate": 9.082597998027377e-05,
+      "loss": 0.0791,
+      "step": 42500
+    },
+    {
+      "epoch": 2.7811579980372914,
+      "grad_norm": 0.7661714553833008,
+      "learning_rate": 9.082067608344436e-05,
+      "loss": 0.0844,
+      "step": 42510
+    },
+    {
+      "epoch": 2.7818122342165523,
+      "grad_norm": 0.9406249523162842,
+      "learning_rate": 9.081537080881007e-05,
+      "loss": 0.0822,
+      "step": 42520
+    },
+    {
+      "epoch": 2.782466470395813,
+      "grad_norm": 0.8644420504570007,
+      "learning_rate": 9.081006415654995e-05,
+      "loss": 0.0803,
+      "step": 42530
+    },
+    {
+      "epoch": 2.7831207065750734,
+      "grad_norm": 0.9211861491203308,
+      "learning_rate": 9.080475612684316e-05,
+      "loss": 0.0808,
+      "step": 42540
+    },
+    {
+      "epoch": 2.7837749427543343,
+      "grad_norm": 1.1344019174575806,
+      "learning_rate": 9.079944671986883e-05,
+      "loss": 0.0834,
+      "step": 42550
+    },
+    {
+      "epoch": 2.784429178933595,
+      "grad_norm": 0.8112305998802185,
+      "learning_rate": 9.079413593580616e-05,
+      "loss": 0.0936,
+      "step": 42560
+    },
+    {
+      "epoch": 2.785083415112856,
+      "grad_norm": 0.973592221736908,
+      "learning_rate": 9.078882377483444e-05,
+      "loss": 0.0733,
+      "step": 42570
+    },
+    {
+      "epoch": 2.7857376512921164,
+      "grad_norm": 0.7716719508171082,
+      "learning_rate": 9.078351023713294e-05,
+      "loss": 0.0764,
+      "step": 42580
+    },
+    {
+      "epoch": 2.7863918874713773,
+      "grad_norm": 0.7506532073020935,
+      "learning_rate": 9.077819532288102e-05,
+      "loss": 0.0697,
+      "step": 42590
+    },
+    {
+      "epoch": 2.787046123650638,
+      "grad_norm": 0.7960049510002136,
+      "learning_rate": 9.077287903225804e-05,
+      "loss": 0.0835,
+      "step": 42600
+    },
+    {
+      "epoch": 2.7877003598298984,
+      "grad_norm": 0.873766303062439,
+      "learning_rate": 9.076756136544346e-05,
+      "loss": 0.076,
+      "step": 42610
+    },
+    {
+      "epoch": 2.7883545960091594,
+      "grad_norm": 0.680327296257019,
+      "learning_rate": 9.076224232261679e-05,
+      "loss": 0.0727,
+      "step": 42620
+    },
+    {
+      "epoch": 2.78900883218842,
+      "grad_norm": 0.9833268523216248,
+      "learning_rate": 9.075692190395752e-05,
+      "loss": 0.0864,
+      "step": 42630
+    },
+    {
+      "epoch": 2.789663068367681,
+      "grad_norm": 0.9118520617485046,
+      "learning_rate": 9.075160010964526e-05,
+      "loss": 0.079,
+      "step": 42640
+    },
+    {
+      "epoch": 2.7903173045469414,
+      "grad_norm": 0.990631639957428,
+      "learning_rate": 9.074627693985961e-05,
+      "loss": 0.0777,
+      "step": 42650
+    },
+    {
+      "epoch": 2.7909715407262023,
+      "grad_norm": 0.91771399974823,
+      "learning_rate": 9.074095239478026e-05,
+      "loss": 0.0775,
+      "step": 42660
+    },
+    {
+      "epoch": 2.791625776905463,
+      "grad_norm": 0.832398533821106,
+      "learning_rate": 9.07356264745869e-05,
+      "loss": 0.0769,
+      "step": 42670
+    },
+    {
+      "epoch": 2.7922800130847234,
+      "grad_norm": 0.7836474180221558,
+      "learning_rate": 9.073029917945934e-05,
+      "loss": 0.087,
+      "step": 42680
+    },
+    {
+      "epoch": 2.7929342492639844,
+      "grad_norm": 0.8258154988288879,
+      "learning_rate": 9.072497050957736e-05,
+      "loss": 0.0893,
+      "step": 42690
+    },
+    {
+      "epoch": 2.793588485443245,
+      "grad_norm": 0.8014414310455322,
+      "learning_rate": 9.071964046512081e-05,
+      "loss": 0.0776,
+      "step": 42700
+    },
+    {
+      "epoch": 2.794242721622506,
+      "grad_norm": 0.9521371126174927,
+      "learning_rate": 9.071430904626961e-05,
+      "loss": 0.0834,
+      "step": 42710
+    },
+    {
+      "epoch": 2.7948969578017664,
+      "grad_norm": 1.008519172668457,
+      "learning_rate": 9.07089762532037e-05,
+      "loss": 0.0725,
+      "step": 42720
+    },
+    {
+      "epoch": 2.7955511939810274,
+      "grad_norm": 1.2083194255828857,
+      "learning_rate": 9.070364208610307e-05,
+      "loss": 0.0781,
+      "step": 42730
+    },
+    {
+      "epoch": 2.796205430160288,
+      "grad_norm": 1.0434963703155518,
+      "learning_rate": 9.069830654514778e-05,
+      "loss": 0.0841,
+      "step": 42740
+    },
+    {
+      "epoch": 2.7968596663395484,
+      "grad_norm": 1.085412859916687,
+      "learning_rate": 9.06929696305179e-05,
+      "loss": 0.0919,
+      "step": 42750
+    },
+    {
+      "epoch": 2.7975139025188094,
+      "grad_norm": 0.9131748676300049,
+      "learning_rate": 9.06876313423936e-05,
+      "loss": 0.0799,
+      "step": 42760
+    },
+    {
+      "epoch": 2.79816813869807,
+      "grad_norm": 0.977798581123352,
+      "learning_rate": 9.068229168095501e-05,
+      "loss": 0.0895,
+      "step": 42770
+    },
+    {
+      "epoch": 2.798822374877331,
+      "grad_norm": 0.891613245010376,
+      "learning_rate": 9.067695064638237e-05,
+      "loss": 0.0834,
+      "step": 42780
+    },
+    {
+      "epoch": 2.7994766110565914,
+      "grad_norm": 0.8886749148368835,
+      "learning_rate": 9.0671608238856e-05,
+      "loss": 0.0821,
+      "step": 42790
+    },
+    {
+      "epoch": 2.8001308472358524,
+      "grad_norm": 0.9983289241790771,
+      "learning_rate": 9.066626445855617e-05,
+      "loss": 0.0828,
+      "step": 42800
+    },
+    {
+      "epoch": 2.800785083415113,
+      "grad_norm": 0.7890607118606567,
+      "learning_rate": 9.066091930566327e-05,
+      "loss": 0.0857,
+      "step": 42810
+    },
+    {
+      "epoch": 2.8014393195943734,
+      "grad_norm": 0.7882013320922852,
+      "learning_rate": 9.06555727803577e-05,
+      "loss": 0.0775,
+      "step": 42820
+    },
+    {
+      "epoch": 2.8020935557736344,
+      "grad_norm": 0.8657622337341309,
+      "learning_rate": 9.065022488281992e-05,
+      "loss": 0.0844,
+      "step": 42830
+    },
+    {
+      "epoch": 2.802747791952895,
+      "grad_norm": 1.038025975227356,
+      "learning_rate": 9.064487561323045e-05,
+      "loss": 0.0796,
+      "step": 42840
+    },
+    {
+      "epoch": 2.8034020281321554,
+      "grad_norm": 0.8643264174461365,
+      "learning_rate": 9.063952497176983e-05,
+      "loss": 0.0823,
+      "step": 42850
+    },
+    {
+      "epoch": 2.8040562643114164,
+      "grad_norm": 0.7769449949264526,
+      "learning_rate": 9.063417295861866e-05,
+      "loss": 0.07,
+      "step": 42860
+    },
+    {
+      "epoch": 2.8047105004906774,
+      "grad_norm": 0.988023579120636,
+      "learning_rate": 9.06288195739576e-05,
+      "loss": 0.0752,
+      "step": 42870
+    },
+    {
+      "epoch": 2.805364736669938,
+      "grad_norm": 0.7536209225654602,
+      "learning_rate": 9.06234648179673e-05,
+      "loss": 0.0876,
+      "step": 42880
+    },
+    {
+      "epoch": 2.8060189728491984,
+      "grad_norm": 0.9088309407234192,
+      "learning_rate": 9.061810869082855e-05,
+      "loss": 0.0749,
+      "step": 42890
+    },
+    {
+      "epoch": 2.8066732090284594,
+      "grad_norm": 0.937651515007019,
+      "learning_rate": 9.061275119272207e-05,
+      "loss": 0.0868,
+      "step": 42900
+    },
+    {
+      "epoch": 2.80732744520772,
+      "grad_norm": 0.8864895701408386,
+      "learning_rate": 9.060739232382876e-05,
+      "loss": 0.0844,
+      "step": 42910
+    },
+    {
+      "epoch": 2.8079816813869805,
+      "grad_norm": 0.8680556416511536,
+      "learning_rate": 9.060203208432945e-05,
+      "loss": 0.0741,
+      "step": 42920
+    },
+    {
+      "epoch": 2.8086359175662414,
+      "grad_norm": 0.7065665125846863,
+      "learning_rate": 9.059667047440508e-05,
+      "loss": 0.0766,
+      "step": 42930
+    },
+    {
+      "epoch": 2.8092901537455024,
+      "grad_norm": 1.2820072174072266,
+      "learning_rate": 9.059130749423662e-05,
+      "loss": 0.0857,
+      "step": 42940
+    },
+    {
+      "epoch": 2.809944389924763,
+      "grad_norm": 0.8697900772094727,
+      "learning_rate": 9.058594314400506e-05,
+      "loss": 0.0755,
+      "step": 42950
+    },
+    {
+      "epoch": 2.8105986261040234,
+      "grad_norm": 0.9628492593765259,
+      "learning_rate": 9.058057742389147e-05,
+      "loss": 0.0819,
+      "step": 42960
+    },
+    {
+      "epoch": 2.8112528622832844,
+      "grad_norm": 1.09038507938385,
+      "learning_rate": 9.057521033407698e-05,
+      "loss": 0.0815,
+      "step": 42970
+    },
+    {
+      "epoch": 2.811907098462545,
+      "grad_norm": 0.9920393824577332,
+      "learning_rate": 9.056984187474275e-05,
+      "loss": 0.081,
+      "step": 42980
+    },
+    {
+      "epoch": 2.8125613346418055,
+      "grad_norm": 0.7963695526123047,
+      "learning_rate": 9.056447204606993e-05,
+      "loss": 0.0774,
+      "step": 42990
+    },
+    {
+      "epoch": 2.8132155708210664,
+      "grad_norm": 1.1177492141723633,
+      "learning_rate": 9.055910084823979e-05,
+      "loss": 0.0786,
+      "step": 43000
+    },
+    {
+      "epoch": 2.813869807000327,
+      "grad_norm": 0.9136730432510376,
+      "learning_rate": 9.055372828143365e-05,
+      "loss": 0.0744,
+      "step": 43010
+    },
+    {
+      "epoch": 2.814524043179588,
+      "grad_norm": 0.7779806852340698,
+      "learning_rate": 9.054835434583281e-05,
+      "loss": 0.0788,
+      "step": 43020
+    },
+    {
+      "epoch": 2.8151782793588485,
+      "grad_norm": 1.0529919862747192,
+      "learning_rate": 9.054297904161868e-05,
+      "loss": 0.0788,
+      "step": 43030
+    },
+    {
+      "epoch": 2.8158325155381094,
+      "grad_norm": 0.7755911946296692,
+      "learning_rate": 9.053760236897266e-05,
+      "loss": 0.08,
+      "step": 43040
+    },
+    {
+      "epoch": 2.81648675171737,
+      "grad_norm": 0.9372928738594055,
+      "learning_rate": 9.053222432807626e-05,
+      "loss": 0.0827,
+      "step": 43050
+    },
+    {
+      "epoch": 2.8171409878966305,
+      "grad_norm": 0.91545170545578,
+      "learning_rate": 9.052684491911099e-05,
+      "loss": 0.0858,
+      "step": 43060
+    },
+    {
+      "epoch": 2.8177952240758914,
+      "grad_norm": 0.7959693670272827,
+      "learning_rate": 9.052146414225841e-05,
+      "loss": 0.0906,
+      "step": 43070
+    },
+    {
+      "epoch": 2.818449460255152,
+      "grad_norm": 0.9491099119186401,
+      "learning_rate": 9.051608199770016e-05,
+      "loss": 0.0794,
+      "step": 43080
+    },
+    {
+      "epoch": 2.819103696434413,
+      "grad_norm": 0.7243504524230957,
+      "learning_rate": 9.051069848561787e-05,
+      "loss": 0.0793,
+      "step": 43090
+    },
+    {
+      "epoch": 2.8197579326136735,
+      "grad_norm": 1.0417028665542603,
+      "learning_rate": 9.050531360619328e-05,
+      "loss": 0.0791,
+      "step": 43100
+    },
+    {
+      "epoch": 2.8204121687929344,
+      "grad_norm": 0.7411209344863892,
+      "learning_rate": 9.04999273596081e-05,
+      "loss": 0.0815,
+      "step": 43110
+    },
+    {
+      "epoch": 2.821066404972195,
+      "grad_norm": 0.7973235249519348,
+      "learning_rate": 9.049453974604418e-05,
+      "loss": 0.0752,
+      "step": 43120
+    },
+    {
+      "epoch": 2.8217206411514555,
+      "grad_norm": 0.9462906718254089,
+      "learning_rate": 9.048915076568334e-05,
+      "loss": 0.0808,
+      "step": 43130
+    },
+    {
+      "epoch": 2.8223748773307165,
+      "grad_norm": 1.0456901788711548,
+      "learning_rate": 9.048376041870745e-05,
+      "loss": 0.0969,
+      "step": 43140
+    },
+    {
+      "epoch": 2.823029113509977,
+      "grad_norm": 1.11058509349823,
+      "learning_rate": 9.04783687052985e-05,
+      "loss": 0.0767,
+      "step": 43150
+    },
+    {
+      "epoch": 2.823683349689238,
+      "grad_norm": 0.7125511765480042,
+      "learning_rate": 9.047297562563843e-05,
+      "loss": 0.0781,
+      "step": 43160
+    },
+    {
+      "epoch": 2.8243375858684985,
+      "grad_norm": 0.7025366425514221,
+      "learning_rate": 9.04675811799093e-05,
+      "loss": 0.0664,
+      "step": 43170
+    },
+    {
+      "epoch": 2.8249918220477594,
+      "grad_norm": 0.9043560028076172,
+      "learning_rate": 9.046218536829319e-05,
+      "loss": 0.077,
+      "step": 43180
+    },
+    {
+      "epoch": 2.82564605822702,
+      "grad_norm": 0.7520603537559509,
+      "learning_rate": 9.045678819097218e-05,
+      "loss": 0.0734,
+      "step": 43190
+    },
+    {
+      "epoch": 2.8263002944062805,
+      "grad_norm": 0.903668224811554,
+      "learning_rate": 9.045138964812848e-05,
+      "loss": 0.0824,
+      "step": 43200
+    },
+    {
+      "epoch": 2.8269545305855415,
+      "grad_norm": 0.8586761355400085,
+      "learning_rate": 9.044598973994429e-05,
+      "loss": 0.0787,
+      "step": 43210
+    },
+    {
+      "epoch": 2.827608766764802,
+      "grad_norm": 0.8279673457145691,
+      "learning_rate": 9.044058846660187e-05,
+      "loss": 0.0903,
+      "step": 43220
+    },
+    {
+      "epoch": 2.828263002944063,
+      "grad_norm": 0.7114253640174866,
+      "learning_rate": 9.043518582828354e-05,
+      "loss": 0.0761,
+      "step": 43230
+    },
+    {
+      "epoch": 2.8289172391233235,
+      "grad_norm": 0.8194983601570129,
+      "learning_rate": 9.042978182517163e-05,
+      "loss": 0.0902,
+      "step": 43240
+    },
+    {
+      "epoch": 2.8295714753025845,
+      "grad_norm": 0.7969952821731567,
+      "learning_rate": 9.042437645744856e-05,
+      "loss": 0.0691,
+      "step": 43250
+    },
+    {
+      "epoch": 2.830225711481845,
+      "grad_norm": 0.7822580337524414,
+      "learning_rate": 9.041896972529677e-05,
+      "loss": 0.0752,
+      "step": 43260
+    },
+    {
+      "epoch": 2.8308799476611055,
+      "grad_norm": 0.7220231294631958,
+      "learning_rate": 9.041356162889873e-05,
+      "loss": 0.0728,
+      "step": 43270
+    },
+    {
+      "epoch": 2.8315341838403665,
+      "grad_norm": 0.7740190029144287,
+      "learning_rate": 9.040815216843702e-05,
+      "loss": 0.0816,
+      "step": 43280
+    },
+    {
+      "epoch": 2.832188420019627,
+      "grad_norm": 0.8419371247291565,
+      "learning_rate": 9.040274134409419e-05,
+      "loss": 0.0792,
+      "step": 43290
+    },
+    {
+      "epoch": 2.8328426561988875,
+      "grad_norm": 0.950294554233551,
+      "learning_rate": 9.039732915605287e-05,
+      "loss": 0.0856,
+      "step": 43300
+    },
+    {
+      "epoch": 2.8334968923781485,
+      "grad_norm": 0.9349603056907654,
+      "learning_rate": 9.039191560449575e-05,
+      "loss": 0.0815,
+      "step": 43310
+    },
+    {
+      "epoch": 2.8341511285574095,
+      "grad_norm": 0.8912672996520996,
+      "learning_rate": 9.038650068960556e-05,
+      "loss": 0.0821,
+      "step": 43320
+    },
+    {
+      "epoch": 2.83480536473667,
+      "grad_norm": 0.9765848517417908,
+      "learning_rate": 9.038108441156506e-05,
+      "loss": 0.0797,
+      "step": 43330
+    },
+    {
+      "epoch": 2.8354596009159305,
+      "grad_norm": 0.8395116329193115,
+      "learning_rate": 9.037566677055705e-05,
+      "loss": 0.0738,
+      "step": 43340
+    },
+    {
+      "epoch": 2.8361138370951915,
+      "grad_norm": 0.7863714098930359,
+      "learning_rate": 9.037024776676438e-05,
+      "loss": 0.087,
+      "step": 43350
+    },
+    {
+      "epoch": 2.836768073274452,
+      "grad_norm": 0.9483030438423157,
+      "learning_rate": 9.036482740036998e-05,
+      "loss": 0.0786,
+      "step": 43360
+    },
+    {
+      "epoch": 2.8374223094537125,
+      "grad_norm": 0.8250223994255066,
+      "learning_rate": 9.03594056715568e-05,
+      "loss": 0.0719,
+      "step": 43370
+    },
+    {
+      "epoch": 2.8380765456329735,
+      "grad_norm": 0.8315423727035522,
+      "learning_rate": 9.035398258050784e-05,
+      "loss": 0.0811,
+      "step": 43380
+    },
+    {
+      "epoch": 2.8387307818122345,
+      "grad_norm": 0.8926718235015869,
+      "learning_rate": 9.034855812740612e-05,
+      "loss": 0.0749,
+      "step": 43390
+    },
+    {
+      "epoch": 2.839385017991495,
+      "grad_norm": 0.8397383689880371,
+      "learning_rate": 9.034313231243477e-05,
+      "loss": 0.0738,
+      "step": 43400
+    },
+    {
+      "epoch": 2.8400392541707555,
+      "grad_norm": 0.900603711605072,
+      "learning_rate": 9.033770513577688e-05,
+      "loss": 0.0747,
+      "step": 43410
+    },
+    {
+      "epoch": 2.8406934903500165,
+      "grad_norm": 0.8045016527175903,
+      "learning_rate": 9.033227659761565e-05,
+      "loss": 0.0794,
+      "step": 43420
+    },
+    {
+      "epoch": 2.841347726529277,
+      "grad_norm": 0.861449122428894,
+      "learning_rate": 9.032684669813431e-05,
+      "loss": 0.0744,
+      "step": 43430
+    },
+    {
+      "epoch": 2.8420019627085376,
+      "grad_norm": 0.8809171319007874,
+      "learning_rate": 9.032141543751614e-05,
+      "loss": 0.0807,
+      "step": 43440
+    },
+    {
+      "epoch": 2.8426561988877985,
+      "grad_norm": 0.8424410820007324,
+      "learning_rate": 9.031598281594445e-05,
+      "loss": 0.0771,
+      "step": 43450
+    },
+    {
+      "epoch": 2.843310435067059,
+      "grad_norm": 0.788705587387085,
+      "learning_rate": 9.031054883360261e-05,
+      "loss": 0.0815,
+      "step": 43460
+    },
+    {
+      "epoch": 2.84396467124632,
+      "grad_norm": 0.7892941832542419,
+      "learning_rate": 9.030511349067404e-05,
+      "loss": 0.0707,
+      "step": 43470
+    },
+    {
+      "epoch": 2.8446189074255805,
+      "grad_norm": 0.8143117427825928,
+      "learning_rate": 9.029967678734216e-05,
+      "loss": 0.0738,
+      "step": 43480
+    },
+    {
+      "epoch": 2.8452731436048415,
+      "grad_norm": 0.9846142530441284,
+      "learning_rate": 9.02942387237905e-05,
+      "loss": 0.0776,
+      "step": 43490
+    },
+    {
+      "epoch": 2.845927379784102,
+      "grad_norm": 0.9837608933448792,
+      "learning_rate": 9.028879930020262e-05,
+      "loss": 0.0807,
+      "step": 43500
+    },
+    {
+      "epoch": 2.8465816159633626,
+      "grad_norm": 1.0797531604766846,
+      "learning_rate": 9.028335851676211e-05,
+      "loss": 0.0818,
+      "step": 43510
+    },
+    {
+      "epoch": 2.8472358521426235,
+      "grad_norm": 0.9254376292228699,
+      "learning_rate": 9.02779163736526e-05,
+      "loss": 0.0763,
+      "step": 43520
+    },
+    {
+      "epoch": 2.847890088321884,
+      "grad_norm": 0.9471592903137207,
+      "learning_rate": 9.027247287105776e-05,
+      "loss": 0.0708,
+      "step": 43530
+    },
+    {
+      "epoch": 2.848544324501145,
+      "grad_norm": 0.9006256461143494,
+      "learning_rate": 9.026702800916136e-05,
+      "loss": 0.0736,
+      "step": 43540
+    },
+    {
+      "epoch": 2.8491985606804056,
+      "grad_norm": 0.888878345489502,
+      "learning_rate": 9.026158178814715e-05,
+      "loss": 0.0706,
+      "step": 43550
+    },
+    {
+      "epoch": 2.8498527968596665,
+      "grad_norm": 0.9179434180259705,
+      "learning_rate": 9.025613420819897e-05,
+      "loss": 0.076,
+      "step": 43560
+    },
+    {
+      "epoch": 2.850507033038927,
+      "grad_norm": 0.9074091911315918,
+      "learning_rate": 9.025068526950069e-05,
+      "loss": 0.0775,
+      "step": 43570
+    },
+    {
+      "epoch": 2.8511612692181876,
+      "grad_norm": 0.7732144594192505,
+      "learning_rate": 9.024523497223622e-05,
+      "loss": 0.077,
+      "step": 43580
+    },
+    {
+      "epoch": 2.8518155053974485,
+      "grad_norm": 0.8621454834938049,
+      "learning_rate": 9.02397833165895e-05,
+      "loss": 0.0852,
+      "step": 43590
+    },
+    {
+      "epoch": 2.852469741576709,
+      "grad_norm": 0.9191235303878784,
+      "learning_rate": 9.023433030274459e-05,
+      "loss": 0.0815,
+      "step": 43600
+    },
+    {
+      "epoch": 2.85312397775597,
+      "grad_norm": 0.7941545248031616,
+      "learning_rate": 9.022887593088551e-05,
+      "loss": 0.0837,
+      "step": 43610
+    },
+    {
+      "epoch": 2.8537782139352306,
+      "grad_norm": 1.0019004344940186,
+      "learning_rate": 9.022342020119637e-05,
+      "loss": 0.0839,
+      "step": 43620
+    },
+    {
+      "epoch": 2.8544324501144915,
+      "grad_norm": 1.2812579870224,
+      "learning_rate": 9.021796311386128e-05,
+      "loss": 0.0911,
+      "step": 43630
+    },
+    {
+      "epoch": 2.855086686293752,
+      "grad_norm": 0.9368396997451782,
+      "learning_rate": 9.021250466906448e-05,
+      "loss": 0.0748,
+      "step": 43640
+    },
+    {
+      "epoch": 2.8557409224730126,
+      "grad_norm": 0.8556075096130371,
+      "learning_rate": 9.020704486699018e-05,
+      "loss": 0.0798,
+      "step": 43650
+    },
+    {
+      "epoch": 2.8563951586522736,
+      "grad_norm": 0.6791936755180359,
+      "learning_rate": 9.020158370782266e-05,
+      "loss": 0.0834,
+      "step": 43660
+    },
+    {
+      "epoch": 2.857049394831534,
+      "grad_norm": 0.7594597339630127,
+      "learning_rate": 9.019612119174627e-05,
+      "loss": 0.0826,
+      "step": 43670
+    },
+    {
+      "epoch": 2.857703631010795,
+      "grad_norm": 0.8290628790855408,
+      "learning_rate": 9.019065731894537e-05,
+      "loss": 0.0792,
+      "step": 43680
+    },
+    {
+      "epoch": 2.8583578671900556,
+      "grad_norm": 0.8027207255363464,
+      "learning_rate": 9.018519208960439e-05,
+      "loss": 0.0695,
+      "step": 43690
+    },
+    {
+      "epoch": 2.8590121033693165,
+      "grad_norm": 0.8073681592941284,
+      "learning_rate": 9.017972550390777e-05,
+      "loss": 0.0823,
+      "step": 43700
+    },
+    {
+      "epoch": 2.859666339548577,
+      "grad_norm": 0.971709668636322,
+      "learning_rate": 9.017425756204005e-05,
+      "loss": 0.0919,
+      "step": 43710
+    },
+    {
+      "epoch": 2.8603205757278376,
+      "grad_norm": 0.8581392765045166,
+      "learning_rate": 9.016878826418577e-05,
+      "loss": 0.0817,
+      "step": 43720
+    },
+    {
+      "epoch": 2.8609748119070986,
+      "grad_norm": 0.8417885303497314,
+      "learning_rate": 9.016331761052954e-05,
+      "loss": 0.0865,
+      "step": 43730
+    },
+    {
+      "epoch": 2.861629048086359,
+      "grad_norm": 1.0083153247833252,
+      "learning_rate": 9.015784560125602e-05,
+      "loss": 0.077,
+      "step": 43740
+    },
+    {
+      "epoch": 2.8622832842656196,
+      "grad_norm": 0.8471361398696899,
+      "learning_rate": 9.015237223654988e-05,
+      "loss": 0.0737,
+      "step": 43750
+    },
+    {
+      "epoch": 2.8629375204448806,
+      "grad_norm": 0.8287156224250793,
+      "learning_rate": 9.014689751659589e-05,
+      "loss": 0.0779,
+      "step": 43760
+    },
+    {
+      "epoch": 2.8635917566241416,
+      "grad_norm": 0.8068413734436035,
+      "learning_rate": 9.01414214415788e-05,
+      "loss": 0.0885,
+      "step": 43770
+    },
+    {
+      "epoch": 2.864245992803402,
+      "grad_norm": 0.8625628352165222,
+      "learning_rate": 9.013594401168346e-05,
+      "loss": 0.0933,
+      "step": 43780
+    },
+    {
+      "epoch": 2.8649002289826626,
+      "grad_norm": 0.9650261998176575,
+      "learning_rate": 9.013046522709477e-05,
+      "loss": 0.0869,
+      "step": 43790
+    },
+    {
+      "epoch": 2.8655544651619236,
+      "grad_norm": 0.7974777817726135,
+      "learning_rate": 9.01249850879976e-05,
+      "loss": 0.0733,
+      "step": 43800
+    },
+    {
+      "epoch": 2.866208701341184,
+      "grad_norm": 1.0560264587402344,
+      "learning_rate": 9.011950359457699e-05,
+      "loss": 0.0885,
+      "step": 43810
+    },
+    {
+      "epoch": 2.8668629375204446,
+      "grad_norm": 0.8955875635147095,
+      "learning_rate": 9.011402074701788e-05,
+      "loss": 0.0829,
+      "step": 43820
+    },
+    {
+      "epoch": 2.8675171736997056,
+      "grad_norm": 0.9224720001220703,
+      "learning_rate": 9.01085365455054e-05,
+      "loss": 0.077,
+      "step": 43830
+    },
+    {
+      "epoch": 2.8681714098789666,
+      "grad_norm": 0.9114225506782532,
+      "learning_rate": 9.010305099022462e-05,
+      "loss": 0.0843,
+      "step": 43840
+    },
+    {
+      "epoch": 2.868825646058227,
+      "grad_norm": 0.8845502138137817,
+      "learning_rate": 9.009756408136067e-05,
+      "loss": 0.0749,
+      "step": 43850
+    },
+    {
+      "epoch": 2.8694798822374876,
+      "grad_norm": 0.8369585275650024,
+      "learning_rate": 9.009207581909879e-05,
+      "loss": 0.0707,
+      "step": 43860
+    },
+    {
+      "epoch": 2.8701341184167486,
+      "grad_norm": 0.8531777262687683,
+      "learning_rate": 9.008658620362421e-05,
+      "loss": 0.0816,
+      "step": 43870
+    },
+    {
+      "epoch": 2.870788354596009,
+      "grad_norm": 0.9684780836105347,
+      "learning_rate": 9.00810952351222e-05,
+      "loss": 0.0852,
+      "step": 43880
+    },
+    {
+      "epoch": 2.8714425907752696,
+      "grad_norm": 1.1896883249282837,
+      "learning_rate": 9.007560291377813e-05,
+      "loss": 0.0782,
+      "step": 43890
+    },
+    {
+      "epoch": 2.8720968269545306,
+      "grad_norm": 0.8664726614952087,
+      "learning_rate": 9.007010923977732e-05,
+      "loss": 0.0761,
+      "step": 43900
+    },
+    {
+      "epoch": 2.872751063133791,
+      "grad_norm": 0.7793802618980408,
+      "learning_rate": 9.006461421330528e-05,
+      "loss": 0.0801,
+      "step": 43910
+    },
+    {
+      "epoch": 2.873405299313052,
+      "grad_norm": 0.8414815664291382,
+      "learning_rate": 9.005911783454742e-05,
+      "loss": 0.0774,
+      "step": 43920
+    },
+    {
+      "epoch": 2.8740595354923126,
+      "grad_norm": 0.9556751251220703,
+      "learning_rate": 9.005362010368926e-05,
+      "loss": 0.0771,
+      "step": 43930
+    },
+    {
+      "epoch": 2.8747137716715736,
+      "grad_norm": 0.720085620880127,
+      "learning_rate": 9.00481210209164e-05,
+      "loss": 0.077,
+      "step": 43940
+    },
+    {
+      "epoch": 2.875368007850834,
+      "grad_norm": 0.7043220400810242,
+      "learning_rate": 9.004262058641441e-05,
+      "loss": 0.0704,
+      "step": 43950
+    },
+    {
+      "epoch": 2.8760222440300947,
+      "grad_norm": 0.9286543726921082,
+      "learning_rate": 9.003711880036896e-05,
+      "loss": 0.0778,
+      "step": 43960
+    },
+    {
+      "epoch": 2.8766764802093556,
+      "grad_norm": 0.8943794369697571,
+      "learning_rate": 9.003161566296576e-05,
+      "loss": 0.0684,
+      "step": 43970
+    },
+    {
+      "epoch": 2.877330716388616,
+      "grad_norm": 0.7501460313796997,
+      "learning_rate": 9.002611117439054e-05,
+      "loss": 0.0767,
+      "step": 43980
+    },
+    {
+      "epoch": 2.877984952567877,
+      "grad_norm": 0.9430428743362427,
+      "learning_rate": 9.002060533482909e-05,
+      "loss": 0.0751,
+      "step": 43990
+    },
+    {
+      "epoch": 2.8786391887471376,
+      "grad_norm": 0.986933171749115,
+      "learning_rate": 9.001509814446726e-05,
+      "loss": 0.0842,
+      "step": 44000
+    },
+    {
+      "epoch": 2.8792934249263986,
+      "grad_norm": 0.7722887992858887,
+      "learning_rate": 9.000958960349092e-05,
+      "loss": 0.0782,
+      "step": 44010
+    },
+    {
+      "epoch": 2.879947661105659,
+      "grad_norm": 0.8584244251251221,
+      "learning_rate": 9.0004079712086e-05,
+      "loss": 0.0796,
+      "step": 44020
+    },
+    {
+      "epoch": 2.8806018972849197,
+      "grad_norm": 0.8732749819755554,
+      "learning_rate": 8.999856847043849e-05,
+      "loss": 0.0848,
+      "step": 44030
+    },
+    {
+      "epoch": 2.8812561334641806,
+      "grad_norm": 0.8751747012138367,
+      "learning_rate": 8.999305587873437e-05,
+      "loss": 0.0776,
+      "step": 44040
+    },
+    {
+      "epoch": 2.881910369643441,
+      "grad_norm": 0.9108805060386658,
+      "learning_rate": 8.998754193715974e-05,
+      "loss": 0.0745,
+      "step": 44050
+    },
+    {
+      "epoch": 2.882564605822702,
+      "grad_norm": 0.9251412749290466,
+      "learning_rate": 8.99820266459007e-05,
+      "loss": 0.0819,
+      "step": 44060
+    },
+    {
+      "epoch": 2.8832188420019627,
+      "grad_norm": 1.0415140390396118,
+      "learning_rate": 8.99765100051434e-05,
+      "loss": 0.072,
+      "step": 44070
+    },
+    {
+      "epoch": 2.8838730781812236,
+      "grad_norm": 0.8129945397377014,
+      "learning_rate": 8.997099201507406e-05,
+      "loss": 0.0788,
+      "step": 44080
+    },
+    {
+      "epoch": 2.884527314360484,
+      "grad_norm": 0.760057270526886,
+      "learning_rate": 8.996547267587889e-05,
+      "loss": 0.0876,
+      "step": 44090
+    },
+    {
+      "epoch": 2.8851815505397447,
+      "grad_norm": 0.9552372694015503,
+      "learning_rate": 8.995995198774421e-05,
+      "loss": 0.0968,
+      "step": 44100
+    },
+    {
+      "epoch": 2.8858357867190056,
+      "grad_norm": 0.7973808646202087,
+      "learning_rate": 8.995442995085636e-05,
+      "loss": 0.0824,
+      "step": 44110
+    },
+    {
+      "epoch": 2.886490022898266,
+      "grad_norm": 0.7297688722610474,
+      "learning_rate": 8.99489065654017e-05,
+      "loss": 0.0783,
+      "step": 44120
+    },
+    {
+      "epoch": 2.887144259077527,
+      "grad_norm": 0.9278412461280823,
+      "learning_rate": 8.994338183156669e-05,
+      "loss": 0.083,
+      "step": 44130
+    },
+    {
+      "epoch": 2.8877984952567877,
+      "grad_norm": 0.9200074672698975,
+      "learning_rate": 8.993785574953778e-05,
+      "loss": 0.0781,
+      "step": 44140
+    },
+    {
+      "epoch": 2.8884527314360486,
+      "grad_norm": 0.8482991456985474,
+      "learning_rate": 8.99323283195015e-05,
+      "loss": 0.0826,
+      "step": 44150
+    },
+    {
+      "epoch": 2.889106967615309,
+      "grad_norm": 0.7914182543754578,
+      "learning_rate": 8.992679954164442e-05,
+      "loss": 0.0751,
+      "step": 44160
+    },
+    {
+      "epoch": 2.8897612037945697,
+      "grad_norm": 0.8093661069869995,
+      "learning_rate": 8.992126941615313e-05,
+      "loss": 0.0754,
+      "step": 44170
+    },
+    {
+      "epoch": 2.8904154399738307,
+      "grad_norm": 0.8084908127784729,
+      "learning_rate": 8.99157379432143e-05,
+      "loss": 0.0813,
+      "step": 44180
+    },
+    {
+      "epoch": 2.891069676153091,
+      "grad_norm": 0.7260185480117798,
+      "learning_rate": 8.991020512301464e-05,
+      "loss": 0.0744,
+      "step": 44190
+    },
+    {
+      "epoch": 2.8917239123323517,
+      "grad_norm": 0.7913504838943481,
+      "learning_rate": 8.990467095574089e-05,
+      "loss": 0.0689,
+      "step": 44200
+    },
+    {
+      "epoch": 2.8923781485116127,
+      "grad_norm": 0.6776607632637024,
+      "learning_rate": 8.989913544157983e-05,
+      "loss": 0.0707,
+      "step": 44210
+    },
+    {
+      "epoch": 2.8930323846908736,
+      "grad_norm": 0.9057397246360779,
+      "learning_rate": 8.98935985807183e-05,
+      "loss": 0.0873,
+      "step": 44220
+    },
+    {
+      "epoch": 2.893686620870134,
+      "grad_norm": 0.9263217449188232,
+      "learning_rate": 8.988806037334322e-05,
+      "loss": 0.0689,
+      "step": 44230
+    },
+    {
+      "epoch": 2.8943408570493947,
+      "grad_norm": 0.9351694583892822,
+      "learning_rate": 8.988252081964147e-05,
+      "loss": 0.0806,
+      "step": 44240
+    },
+    {
+      "epoch": 2.8949950932286557,
+      "grad_norm": 0.75326007604599,
+      "learning_rate": 8.987697991980007e-05,
+      "loss": 0.0745,
+      "step": 44250
+    },
+    {
+      "epoch": 2.895649329407916,
+      "grad_norm": 1.0243571996688843,
+      "learning_rate": 8.987143767400601e-05,
+      "loss": 0.0782,
+      "step": 44260
+    },
+    {
+      "epoch": 2.8963035655871767,
+      "grad_norm": 0.8010449409484863,
+      "learning_rate": 8.986589408244634e-05,
+      "loss": 0.0745,
+      "step": 44270
+    },
+    {
+      "epoch": 2.8969578017664377,
+      "grad_norm": 0.9779613614082336,
+      "learning_rate": 8.986034914530823e-05,
+      "loss": 0.0723,
+      "step": 44280
+    },
+    {
+      "epoch": 2.8976120379456987,
+      "grad_norm": 0.8159413933753967,
+      "learning_rate": 8.985480286277877e-05,
+      "loss": 0.0781,
+      "step": 44290
+    },
+    {
+      "epoch": 2.898266274124959,
+      "grad_norm": 0.8815252184867859,
+      "learning_rate": 8.984925523504519e-05,
+      "loss": 0.0839,
+      "step": 44300
+    },
+    {
+      "epoch": 2.8989205103042197,
+      "grad_norm": 0.7953386306762695,
+      "learning_rate": 8.984370626229474e-05,
+      "loss": 0.0898,
+      "step": 44310
+    },
+    {
+      "epoch": 2.8995747464834807,
+      "grad_norm": 0.8342095017433167,
+      "learning_rate": 8.983815594471472e-05,
+      "loss": 0.0854,
+      "step": 44320
+    },
+    {
+      "epoch": 2.900228982662741,
+      "grad_norm": 0.864380955696106,
+      "learning_rate": 8.983260428249246e-05,
+      "loss": 0.0869,
+      "step": 44330
+    },
+    {
+      "epoch": 2.9008832188420017,
+      "grad_norm": 0.9503182172775269,
+      "learning_rate": 8.982705127581533e-05,
+      "loss": 0.0722,
+      "step": 44340
+    },
+    {
+      "epoch": 2.9015374550212627,
+      "grad_norm": 0.8214649558067322,
+      "learning_rate": 8.982149692487078e-05,
+      "loss": 0.0772,
+      "step": 44350
+    },
+    {
+      "epoch": 2.9021916912005232,
+      "grad_norm": 0.8064943552017212,
+      "learning_rate": 8.981594122984627e-05,
+      "loss": 0.0757,
+      "step": 44360
+    },
+    {
+      "epoch": 2.902845927379784,
+      "grad_norm": 0.7172738909721375,
+      "learning_rate": 8.981038419092931e-05,
+      "loss": 0.0714,
+      "step": 44370
+    },
+    {
+      "epoch": 2.9035001635590447,
+      "grad_norm": 0.9183410406112671,
+      "learning_rate": 8.980482580830747e-05,
+      "loss": 0.0785,
+      "step": 44380
+    },
+    {
+      "epoch": 2.9041543997383057,
+      "grad_norm": 0.8381460905075073,
+      "learning_rate": 8.97992660821684e-05,
+      "loss": 0.0714,
+      "step": 44390
+    },
+    {
+      "epoch": 2.904808635917566,
+      "grad_norm": 0.8850300908088684,
+      "learning_rate": 8.979370501269971e-05,
+      "loss": 0.0808,
+      "step": 44400
+    },
+    {
+      "epoch": 2.9054628720968267,
+      "grad_norm": 0.848743736743927,
+      "learning_rate": 8.97881426000891e-05,
+      "loss": 0.0867,
+      "step": 44410
+    },
+    {
+      "epoch": 2.9061171082760877,
+      "grad_norm": 0.7968313097953796,
+      "learning_rate": 8.978257884452433e-05,
+      "loss": 0.0775,
+      "step": 44420
+    },
+    {
+      "epoch": 2.9067713444553482,
+      "grad_norm": 0.9209117293357849,
+      "learning_rate": 8.977701374619321e-05,
+      "loss": 0.0702,
+      "step": 44430
+    },
+    {
+      "epoch": 2.907425580634609,
+      "grad_norm": 0.7473788857460022,
+      "learning_rate": 8.977144730528353e-05,
+      "loss": 0.0786,
+      "step": 44440
+    },
+    {
+      "epoch": 2.9080798168138697,
+      "grad_norm": 0.7625455856323242,
+      "learning_rate": 8.97658795219832e-05,
+      "loss": 0.0802,
+      "step": 44450
+    },
+    {
+      "epoch": 2.9087340529931307,
+      "grad_norm": 1.0252678394317627,
+      "learning_rate": 8.976031039648017e-05,
+      "loss": 0.0686,
+      "step": 44460
+    },
+    {
+      "epoch": 2.9093882891723912,
+      "grad_norm": 0.7554779052734375,
+      "learning_rate": 8.975473992896239e-05,
+      "loss": 0.0744,
+      "step": 44470
+    },
+    {
+      "epoch": 2.9100425253516518,
+      "grad_norm": 0.8992921710014343,
+      "learning_rate": 8.974916811961786e-05,
+      "loss": 0.0847,
+      "step": 44480
+    },
+    {
+      "epoch": 2.9106967615309127,
+      "grad_norm": 0.7527303695678711,
+      "learning_rate": 8.974359496863466e-05,
+      "loss": 0.0777,
+      "step": 44490
+    },
+    {
+      "epoch": 2.9113509977101732,
+      "grad_norm": 0.8266311883926392,
+      "learning_rate": 8.97380204762009e-05,
+      "loss": 0.076,
+      "step": 44500
+    },
+    {
+      "epoch": 2.912005233889434,
+      "grad_norm": 0.8097569942474365,
+      "learning_rate": 8.973244464250474e-05,
+      "loss": 0.0814,
+      "step": 44510
+    },
+    {
+      "epoch": 2.9126594700686947,
+      "grad_norm": 0.8458988070487976,
+      "learning_rate": 8.972686746773436e-05,
+      "loss": 0.0836,
+      "step": 44520
+    },
+    {
+      "epoch": 2.9133137062479557,
+      "grad_norm": 0.8754256963729858,
+      "learning_rate": 8.972128895207803e-05,
+      "loss": 0.0954,
+      "step": 44530
+    },
+    {
+      "epoch": 2.9139679424272162,
+      "grad_norm": 0.7934485673904419,
+      "learning_rate": 8.971570909572401e-05,
+      "loss": 0.0784,
+      "step": 44540
+    },
+    {
+      "epoch": 2.9146221786064768,
+      "grad_norm": 0.9816693067550659,
+      "learning_rate": 8.971012789886066e-05,
+      "loss": 0.0724,
+      "step": 44550
+    },
+    {
+      "epoch": 2.9152764147857377,
+      "grad_norm": 0.8291686773300171,
+      "learning_rate": 8.970454536167634e-05,
+      "loss": 0.0774,
+      "step": 44560
+    },
+    {
+      "epoch": 2.9159306509649983,
+      "grad_norm": 0.7562156915664673,
+      "learning_rate": 8.96989614843595e-05,
+      "loss": 0.0802,
+      "step": 44570
+    },
+    {
+      "epoch": 2.9165848871442592,
+      "grad_norm": 0.680461049079895,
+      "learning_rate": 8.969337626709858e-05,
+      "loss": 0.0795,
+      "step": 44580
+    },
+    {
+      "epoch": 2.9172391233235198,
+      "grad_norm": 0.8741086721420288,
+      "learning_rate": 8.968778971008211e-05,
+      "loss": 0.0826,
+      "step": 44590
+    },
+    {
+      "epoch": 2.9178933595027807,
+      "grad_norm": 0.79926598072052,
+      "learning_rate": 8.968220181349866e-05,
+      "loss": 0.0677,
+      "step": 44600
+    },
+    {
+      "epoch": 2.9185475956820413,
+      "grad_norm": 0.7387502193450928,
+      "learning_rate": 8.967661257753683e-05,
+      "loss": 0.0756,
+      "step": 44610
+    },
+    {
+      "epoch": 2.9192018318613018,
+      "grad_norm": 0.8162443041801453,
+      "learning_rate": 8.967102200238527e-05,
+      "loss": 0.0716,
+      "step": 44620
+    },
+    {
+      "epoch": 2.9198560680405627,
+      "grad_norm": 0.8832371830940247,
+      "learning_rate": 8.966543008823268e-05,
+      "loss": 0.0711,
+      "step": 44630
+    },
+    {
+      "epoch": 2.9205103042198233,
+      "grad_norm": 0.8890617489814758,
+      "learning_rate": 8.965983683526779e-05,
+      "loss": 0.0754,
+      "step": 44640
+    },
+    {
+      "epoch": 2.921164540399084,
+      "grad_norm": 0.8303614854812622,
+      "learning_rate": 8.96542422436794e-05,
+      "loss": 0.0728,
+      "step": 44650
+    },
+    {
+      "epoch": 2.9218187765783448,
+      "grad_norm": 0.7883252501487732,
+      "learning_rate": 8.964864631365634e-05,
+      "loss": 0.0793,
+      "step": 44660
+    },
+    {
+      "epoch": 2.9224730127576057,
+      "grad_norm": 1.1344659328460693,
+      "learning_rate": 8.964304904538747e-05,
+      "loss": 0.0856,
+      "step": 44670
+    },
+    {
+      "epoch": 2.9231272489368663,
+      "grad_norm": 0.7178243398666382,
+      "learning_rate": 8.963745043906174e-05,
+      "loss": 0.0784,
+      "step": 44680
+    },
+    {
+      "epoch": 2.923781485116127,
+      "grad_norm": 0.9424094557762146,
+      "learning_rate": 8.96318504948681e-05,
+      "loss": 0.0805,
+      "step": 44690
+    },
+    {
+      "epoch": 2.9244357212953878,
+      "grad_norm": 0.6671366691589355,
+      "learning_rate": 8.962624921299558e-05,
+      "loss": 0.0822,
+      "step": 44700
+    },
+    {
+      "epoch": 2.9250899574746483,
+      "grad_norm": 0.7286959290504456,
+      "learning_rate": 8.962064659363321e-05,
+      "loss": 0.0875,
+      "step": 44710
+    },
+    {
+      "epoch": 2.925744193653909,
+      "grad_norm": 0.8117363452911377,
+      "learning_rate": 8.961504263697011e-05,
+      "loss": 0.0746,
+      "step": 44720
+    },
+    {
+      "epoch": 2.92639842983317,
+      "grad_norm": 0.8143247365951538,
+      "learning_rate": 8.960943734319542e-05,
+      "loss": 0.083,
+      "step": 44730
+    },
+    {
+      "epoch": 2.9270526660124307,
+      "grad_norm": 0.8705511689186096,
+      "learning_rate": 8.960383071249836e-05,
+      "loss": 0.0858,
+      "step": 44740
+    },
+    {
+      "epoch": 2.9277069021916913,
+      "grad_norm": 1.2389694452285767,
+      "learning_rate": 8.959822274506812e-05,
+      "loss": 0.0911,
+      "step": 44750
+    },
+    {
+      "epoch": 2.928361138370952,
+      "grad_norm": 0.7589144110679626,
+      "learning_rate": 8.959261344109404e-05,
+      "loss": 0.0772,
+      "step": 44760
+    },
+    {
+      "epoch": 2.9290153745502128,
+      "grad_norm": 0.8595485687255859,
+      "learning_rate": 8.958700280076542e-05,
+      "loss": 0.0727,
+      "step": 44770
+    },
+    {
+      "epoch": 2.9296696107294733,
+      "grad_norm": 1.0086075067520142,
+      "learning_rate": 8.958139082427162e-05,
+      "loss": 0.0796,
+      "step": 44780
+    },
+    {
+      "epoch": 2.930323846908734,
+      "grad_norm": 0.928106963634491,
+      "learning_rate": 8.957577751180209e-05,
+      "loss": 0.0692,
+      "step": 44790
+    },
+    {
+      "epoch": 2.930978083087995,
+      "grad_norm": 0.8739972114562988,
+      "learning_rate": 8.957016286354626e-05,
+      "loss": 0.0811,
+      "step": 44800
+    },
+    {
+      "epoch": 2.9316323192672553,
+      "grad_norm": 0.870585024356842,
+      "learning_rate": 8.956454687969367e-05,
+      "loss": 0.0709,
+      "step": 44810
+    },
+    {
+      "epoch": 2.9322865554465163,
+      "grad_norm": 1.0133564472198486,
+      "learning_rate": 8.955892956043385e-05,
+      "loss": 0.0771,
+      "step": 44820
+    },
+    {
+      "epoch": 2.932940791625777,
+      "grad_norm": 0.8007000684738159,
+      "learning_rate": 8.955331090595642e-05,
+      "loss": 0.0804,
+      "step": 44830
+    },
+    {
+      "epoch": 2.933595027805038,
+      "grad_norm": 0.8323869109153748,
+      "learning_rate": 8.9547690916451e-05,
+      "loss": 0.0738,
+      "step": 44840
+    },
+    {
+      "epoch": 2.9342492639842983,
+      "grad_norm": 0.6912589073181152,
+      "learning_rate": 8.954206959210731e-05,
+      "loss": 0.0738,
+      "step": 44850
+    },
+    {
+      "epoch": 2.934903500163559,
+      "grad_norm": 0.9906479716300964,
+      "learning_rate": 8.953644693311506e-05,
+      "loss": 0.0778,
+      "step": 44860
+    },
+    {
+      "epoch": 2.93555773634282,
+      "grad_norm": 0.6855130791664124,
+      "learning_rate": 8.953082293966404e-05,
+      "loss": 0.0713,
+      "step": 44870
+    },
+    {
+      "epoch": 2.9362119725220803,
+      "grad_norm": 0.7666285634040833,
+      "learning_rate": 8.952519761194407e-05,
+      "loss": 0.0767,
+      "step": 44880
+    },
+    {
+      "epoch": 2.9368662087013413,
+      "grad_norm": 0.8128482103347778,
+      "learning_rate": 8.951957095014503e-05,
+      "loss": 0.0734,
+      "step": 44890
+    },
+    {
+      "epoch": 2.937520444880602,
+      "grad_norm": 1.090126872062683,
+      "learning_rate": 8.951394295445681e-05,
+      "loss": 0.0873,
+      "step": 44900
+    },
+    {
+      "epoch": 2.938174681059863,
+      "grad_norm": 0.8404224514961243,
+      "learning_rate": 8.950831362506941e-05,
+      "loss": 0.073,
+      "step": 44910
+    },
+    {
+      "epoch": 2.9388289172391233,
+      "grad_norm": 0.8833885192871094,
+      "learning_rate": 8.950268296217279e-05,
+      "loss": 0.0803,
+      "step": 44920
+    },
+    {
+      "epoch": 2.939483153418384,
+      "grad_norm": 0.8460028767585754,
+      "learning_rate": 8.949705096595703e-05,
+      "loss": 0.0759,
+      "step": 44930
+    },
+    {
+      "epoch": 2.940137389597645,
+      "grad_norm": 0.7936156988143921,
+      "learning_rate": 8.949141763661222e-05,
+      "loss": 0.0911,
+      "step": 44940
+    },
+    {
+      "epoch": 2.9407916257769053,
+      "grad_norm": 0.8652473092079163,
+      "learning_rate": 8.948578297432848e-05,
+      "loss": 0.0829,
+      "step": 44950
+    },
+    {
+      "epoch": 2.9414458619561663,
+      "grad_norm": 0.9275278449058533,
+      "learning_rate": 8.948014697929603e-05,
+      "loss": 0.0766,
+      "step": 44960
+    },
+    {
+      "epoch": 2.942100098135427,
+      "grad_norm": 0.6803890466690063,
+      "learning_rate": 8.947450965170505e-05,
+      "loss": 0.0729,
+      "step": 44970
+    },
+    {
+      "epoch": 2.942754334314688,
+      "grad_norm": 0.8319152593612671,
+      "learning_rate": 8.946887099174587e-05,
+      "loss": 0.0726,
+      "step": 44980
+    },
+    {
+      "epoch": 2.9434085704939483,
+      "grad_norm": 0.9895662069320679,
+      "learning_rate": 8.946323099960877e-05,
+      "loss": 0.0735,
+      "step": 44990
+    },
+    {
+      "epoch": 2.944062806673209,
+      "grad_norm": 0.899389386177063,
+      "learning_rate": 8.945758967548415e-05,
+      "loss": 0.0678,
+      "step": 45000
+    },
+    {
+      "epoch": 2.94471704285247,
+      "grad_norm": 0.9055632948875427,
+      "learning_rate": 8.945194701956236e-05,
+      "loss": 0.0743,
+      "step": 45010
+    },
+    {
+      "epoch": 2.9453712790317303,
+      "grad_norm": 0.8592407703399658,
+      "learning_rate": 8.944630303203391e-05,
+      "loss": 0.0785,
+      "step": 45020
+    },
+    {
+      "epoch": 2.9460255152109913,
+      "grad_norm": 0.8779101967811584,
+      "learning_rate": 8.944065771308928e-05,
+      "loss": 0.0785,
+      "step": 45030
+    },
+    {
+      "epoch": 2.946679751390252,
+      "grad_norm": 0.8470450639724731,
+      "learning_rate": 8.943501106291901e-05,
+      "loss": 0.0746,
+      "step": 45040
+    },
+    {
+      "epoch": 2.947333987569513,
+      "grad_norm": 0.8635161519050598,
+      "learning_rate": 8.94293630817137e-05,
+      "loss": 0.0732,
+      "step": 45050
+    },
+    {
+      "epoch": 2.9479882237487733,
+      "grad_norm": 0.892550528049469,
+      "learning_rate": 8.942371376966398e-05,
+      "loss": 0.0829,
+      "step": 45060
+    },
+    {
+      "epoch": 2.948642459928034,
+      "grad_norm": 0.7665706276893616,
+      "learning_rate": 8.941806312696054e-05,
+      "loss": 0.0738,
+      "step": 45070
+    },
+    {
+      "epoch": 2.949296696107295,
+      "grad_norm": 0.8420522212982178,
+      "learning_rate": 8.941241115379408e-05,
+      "loss": 0.0801,
+      "step": 45080
+    },
+    {
+      "epoch": 2.9499509322865554,
+      "grad_norm": 0.8662616610527039,
+      "learning_rate": 8.940675785035538e-05,
+      "loss": 0.0738,
+      "step": 45090
+    },
+    {
+      "epoch": 2.950605168465816,
+      "grad_norm": 0.7357196807861328,
+      "learning_rate": 8.940110321683525e-05,
+      "loss": 0.0687,
+      "step": 45100
+    },
+    {
+      "epoch": 2.951259404645077,
+      "grad_norm": 0.8650752305984497,
+      "learning_rate": 8.939544725342454e-05,
+      "loss": 0.0807,
+      "step": 45110
+    },
+    {
+      "epoch": 2.951913640824338,
+      "grad_norm": 1.0531340837478638,
+      "learning_rate": 8.93897899603142e-05,
+      "loss": 0.09,
+      "step": 45120
+    },
+    {
+      "epoch": 2.9525678770035984,
+      "grad_norm": 0.7418067455291748,
+      "learning_rate": 8.938413133769514e-05,
+      "loss": 0.0777,
+      "step": 45130
+    },
+    {
+      "epoch": 2.953222113182859,
+      "grad_norm": 1.0505800247192383,
+      "learning_rate": 8.937847138575833e-05,
+      "loss": 0.0821,
+      "step": 45140
+    },
+    {
+      "epoch": 2.95387634936212,
+      "grad_norm": 0.9159501194953918,
+      "learning_rate": 8.937281010469486e-05,
+      "loss": 0.0742,
+      "step": 45150
+    },
+    {
+      "epoch": 2.9545305855413804,
+      "grad_norm": 0.8090870976448059,
+      "learning_rate": 8.936714749469579e-05,
+      "loss": 0.0825,
+      "step": 45160
+    },
+    {
+      "epoch": 2.955184821720641,
+      "grad_norm": 0.8931045532226562,
+      "learning_rate": 8.936148355595224e-05,
+      "loss": 0.0765,
+      "step": 45170
+    },
+    {
+      "epoch": 2.955839057899902,
+      "grad_norm": 0.8412348628044128,
+      "learning_rate": 8.93558182886554e-05,
+      "loss": 0.0762,
+      "step": 45180
+    },
+    {
+      "epoch": 2.956493294079163,
+      "grad_norm": 0.832350492477417,
+      "learning_rate": 8.935015169299646e-05,
+      "loss": 0.086,
+      "step": 45190
+    },
+    {
+      "epoch": 2.9571475302584234,
+      "grad_norm": 0.8142176866531372,
+      "learning_rate": 8.934448376916672e-05,
+      "loss": 0.0752,
+      "step": 45200
+    },
+    {
+      "epoch": 2.957801766437684,
+      "grad_norm": 0.7122102379798889,
+      "learning_rate": 8.933881451735746e-05,
+      "loss": 0.0754,
+      "step": 45210
+    },
+    {
+      "epoch": 2.958456002616945,
+      "grad_norm": 0.9692149758338928,
+      "learning_rate": 8.933314393776005e-05,
+      "loss": 0.0744,
+      "step": 45220
+    },
+    {
+      "epoch": 2.9591102387962054,
+      "grad_norm": 0.8291851878166199,
+      "learning_rate": 8.932747203056586e-05,
+      "loss": 0.0773,
+      "step": 45230
+    },
+    {
+      "epoch": 2.959764474975466,
+      "grad_norm": 0.7037114500999451,
+      "learning_rate": 8.932179879596636e-05,
+      "loss": 0.0838,
+      "step": 45240
+    },
+    {
+      "epoch": 2.960418711154727,
+      "grad_norm": 0.7591250538825989,
+      "learning_rate": 8.9316124234153e-05,
+      "loss": 0.0729,
+      "step": 45250
+    },
+    {
+      "epoch": 2.9610729473339874,
+      "grad_norm": 0.8546152114868164,
+      "learning_rate": 8.931044834531737e-05,
+      "loss": 0.074,
+      "step": 45260
+    },
+    {
+      "epoch": 2.9617271835132484,
+      "grad_norm": 0.8192047476768494,
+      "learning_rate": 8.930477112965102e-05,
+      "loss": 0.0773,
+      "step": 45270
+    },
+    {
+      "epoch": 2.962381419692509,
+      "grad_norm": 0.7500526905059814,
+      "learning_rate": 8.929909258734553e-05,
+      "loss": 0.0756,
+      "step": 45280
+    },
+    {
+      "epoch": 2.96303565587177,
+      "grad_norm": 0.8506186008453369,
+      "learning_rate": 8.929341271859262e-05,
+      "loss": 0.0812,
+      "step": 45290
+    },
+    {
+      "epoch": 2.9636898920510304,
+      "grad_norm": 0.7039439678192139,
+      "learning_rate": 8.928773152358398e-05,
+      "loss": 0.0697,
+      "step": 45300
+    },
+    {
+      "epoch": 2.964344128230291,
+      "grad_norm": 1.1579927206039429,
+      "learning_rate": 8.928204900251136e-05,
+      "loss": 0.0707,
+      "step": 45310
+    },
+    {
+      "epoch": 2.964998364409552,
+      "grad_norm": 0.8826687932014465,
+      "learning_rate": 8.927636515556657e-05,
+      "loss": 0.0853,
+      "step": 45320
+    },
+    {
+      "epoch": 2.9656526005888124,
+      "grad_norm": 0.7980294823646545,
+      "learning_rate": 8.927067998294145e-05,
+      "loss": 0.0846,
+      "step": 45330
+    },
+    {
+      "epoch": 2.9663068367680734,
+      "grad_norm": 0.7435978651046753,
+      "learning_rate": 8.926499348482787e-05,
+      "loss": 0.067,
+      "step": 45340
+    },
+    {
+      "epoch": 2.966961072947334,
+      "grad_norm": 0.715040922164917,
+      "learning_rate": 8.92593056614178e-05,
+      "loss": 0.0712,
+      "step": 45350
+    },
+    {
+      "epoch": 2.967615309126595,
+      "grad_norm": 0.9521136283874512,
+      "learning_rate": 8.925361651290321e-05,
+      "loss": 0.0809,
+      "step": 45360
+    },
+    {
+      "epoch": 2.9682695453058554,
+      "grad_norm": 0.7266244888305664,
+      "learning_rate": 8.924792603947611e-05,
+      "loss": 0.0802,
+      "step": 45370
+    },
+    {
+      "epoch": 2.968923781485116,
+      "grad_norm": 0.9576558470726013,
+      "learning_rate": 8.924223424132856e-05,
+      "loss": 0.0811,
+      "step": 45380
+    },
+    {
+      "epoch": 2.969578017664377,
+      "grad_norm": 0.8989412784576416,
+      "learning_rate": 8.92365411186527e-05,
+      "loss": 0.0844,
+      "step": 45390
+    },
+    {
+      "epoch": 2.9702322538436374,
+      "grad_norm": 0.9169609546661377,
+      "learning_rate": 8.923084667164067e-05,
+      "loss": 0.0728,
+      "step": 45400
+    },
+    {
+      "epoch": 2.9708864900228984,
+      "grad_norm": 0.9715309739112854,
+      "learning_rate": 8.92251509004847e-05,
+      "loss": 0.0868,
+      "step": 45410
+    },
+    {
+      "epoch": 2.971540726202159,
+      "grad_norm": 0.8212486505508423,
+      "learning_rate": 8.9219453805377e-05,
+      "loss": 0.0794,
+      "step": 45420
+    },
+    {
+      "epoch": 2.97219496238142,
+      "grad_norm": 0.8119533658027649,
+      "learning_rate": 8.921375538650987e-05,
+      "loss": 0.0831,
+      "step": 45430
+    },
+    {
+      "epoch": 2.9728491985606804,
+      "grad_norm": 0.8991037011146545,
+      "learning_rate": 8.920805564407565e-05,
+      "loss": 0.081,
+      "step": 45440
+    },
+    {
+      "epoch": 2.973503434739941,
+      "grad_norm": 0.6757112145423889,
+      "learning_rate": 8.920235457826675e-05,
+      "loss": 0.0706,
+      "step": 45450
+    },
+    {
+      "epoch": 2.974157670919202,
+      "grad_norm": 0.8806977272033691,
+      "learning_rate": 8.919665218927556e-05,
+      "loss": 0.068,
+      "step": 45460
+    },
+    {
+      "epoch": 2.9748119070984624,
+      "grad_norm": 1.013301968574524,
+      "learning_rate": 8.919094847729455e-05,
+      "loss": 0.0812,
+      "step": 45470
+    },
+    {
+      "epoch": 2.9754661432777234,
+      "grad_norm": 0.9366437792778015,
+      "learning_rate": 8.918524344251626e-05,
+      "loss": 0.0773,
+      "step": 45480
+    },
+    {
+      "epoch": 2.976120379456984,
+      "grad_norm": 0.8117071986198425,
+      "learning_rate": 8.917953708513324e-05,
+      "loss": 0.0676,
+      "step": 45490
+    },
+    {
+      "epoch": 2.976774615636245,
+      "grad_norm": 1.0920495986938477,
+      "learning_rate": 8.917382940533808e-05,
+      "loss": 0.0787,
+      "step": 45500
+    },
+    {
+      "epoch": 2.9774288518155054,
+      "grad_norm": 0.7851189970970154,
+      "learning_rate": 8.916812040332344e-05,
+      "loss": 0.0745,
+      "step": 45510
+    },
+    {
+      "epoch": 2.978083087994766,
+      "grad_norm": 0.8046877384185791,
+      "learning_rate": 8.916241007928203e-05,
+      "loss": 0.0787,
+      "step": 45520
+    },
+    {
+      "epoch": 2.978737324174027,
+      "grad_norm": 0.8224988579750061,
+      "learning_rate": 8.915669843340655e-05,
+      "loss": 0.0895,
+      "step": 45530
+    },
+    {
+      "epoch": 2.9793915603532874,
+      "grad_norm": 0.9415143132209778,
+      "learning_rate": 8.915098546588983e-05,
+      "loss": 0.0723,
+      "step": 45540
+    },
+    {
+      "epoch": 2.980045796532548,
+      "grad_norm": 0.9764366149902344,
+      "learning_rate": 8.914527117692465e-05,
+      "loss": 0.084,
+      "step": 45550
+    },
+    {
+      "epoch": 2.980700032711809,
+      "grad_norm": 0.774044930934906,
+      "learning_rate": 8.913955556670392e-05,
+      "loss": 0.0786,
+      "step": 45560
+    },
+    {
+      "epoch": 2.98135426889107,
+      "grad_norm": 0.8664652705192566,
+      "learning_rate": 8.913383863542054e-05,
+      "loss": 0.0765,
+      "step": 45570
+    },
+    {
+      "epoch": 2.9820085050703304,
+      "grad_norm": 0.8627799153327942,
+      "learning_rate": 8.912812038326746e-05,
+      "loss": 0.0839,
+      "step": 45580
+    },
+    {
+      "epoch": 2.982662741249591,
+      "grad_norm": 0.8925349712371826,
+      "learning_rate": 8.912240081043773e-05,
+      "loss": 0.0813,
+      "step": 45590
+    },
+    {
+      "epoch": 2.983316977428852,
+      "grad_norm": 0.8013269305229187,
+      "learning_rate": 8.911667991712433e-05,
+      "loss": 0.0722,
+      "step": 45600
+    },
+    {
+      "epoch": 2.9839712136081125,
+      "grad_norm": 0.6747923493385315,
+      "learning_rate": 8.911095770352043e-05,
+      "loss": 0.0757,
+      "step": 45610
+    },
+    {
+      "epoch": 2.984625449787373,
+      "grad_norm": 0.8834814429283142,
+      "learning_rate": 8.910523416981911e-05,
+      "loss": 0.081,
+      "step": 45620
+    },
+    {
+      "epoch": 2.985279685966634,
+      "grad_norm": 0.8894004225730896,
+      "learning_rate": 8.90995093162136e-05,
+      "loss": 0.0748,
+      "step": 45630
+    },
+    {
+      "epoch": 2.985933922145895,
+      "grad_norm": 0.7748491168022156,
+      "learning_rate": 8.909378314289708e-05,
+      "loss": 0.0718,
+      "step": 45640
+    },
+    {
+      "epoch": 2.9865881583251555,
+      "grad_norm": 0.8032260537147522,
+      "learning_rate": 8.908805565006288e-05,
+      "loss": 0.0714,
+      "step": 45650
+    },
+    {
+      "epoch": 2.987242394504416,
+      "grad_norm": 0.7444846630096436,
+      "learning_rate": 8.90823268379043e-05,
+      "loss": 0.0705,
+      "step": 45660
+    },
+    {
+      "epoch": 2.987896630683677,
+      "grad_norm": 0.8824648261070251,
+      "learning_rate": 8.907659670661467e-05,
+      "loss": 0.087,
+      "step": 45670
+    },
+    {
+      "epoch": 2.9885508668629375,
+      "grad_norm": 0.7613914608955383,
+      "learning_rate": 8.907086525638741e-05,
+      "loss": 0.0791,
+      "step": 45680
+    },
+    {
+      "epoch": 2.989205103042198,
+      "grad_norm": 1.0592745542526245,
+      "learning_rate": 8.9065132487416e-05,
+      "loss": 0.0757,
+      "step": 45690
+    },
+    {
+      "epoch": 2.989859339221459,
+      "grad_norm": 0.8600308299064636,
+      "learning_rate": 8.905939839989391e-05,
+      "loss": 0.0767,
+      "step": 45700
+    },
+    {
+      "epoch": 2.99051357540072,
+      "grad_norm": 0.9262317419052124,
+      "learning_rate": 8.90536629940147e-05,
+      "loss": 0.0879,
+      "step": 45710
+    },
+    {
+      "epoch": 2.9911678115799805,
+      "grad_norm": 0.7141725420951843,
+      "learning_rate": 8.904792626997191e-05,
+      "loss": 0.0699,
+      "step": 45720
+    },
+    {
+      "epoch": 2.991822047759241,
+      "grad_norm": 0.8366238474845886,
+      "learning_rate": 8.904218822795923e-05,
+      "loss": 0.0717,
+      "step": 45730
+    },
+    {
+      "epoch": 2.992476283938502,
+      "grad_norm": 0.937675416469574,
+      "learning_rate": 8.903644886817029e-05,
+      "loss": 0.0782,
+      "step": 45740
+    },
+    {
+      "epoch": 2.9931305201177625,
+      "grad_norm": 0.7373892068862915,
+      "learning_rate": 8.903070819079884e-05,
+      "loss": 0.0793,
+      "step": 45750
+    },
+    {
+      "epoch": 2.993784756297023,
+      "grad_norm": 0.9344786405563354,
+      "learning_rate": 8.902496619603862e-05,
+      "loss": 0.0852,
+      "step": 45760
+    },
+    {
+      "epoch": 2.994438992476284,
+      "grad_norm": 0.7279993295669556,
+      "learning_rate": 8.901922288408343e-05,
+      "loss": 0.0699,
+      "step": 45770
+    },
+    {
+      "epoch": 2.9950932286555445,
+      "grad_norm": 1.427483081817627,
+      "learning_rate": 8.901347825512715e-05,
+      "loss": 0.0969,
+      "step": 45780
+    },
+    {
+      "epoch": 2.9957474648348055,
+      "grad_norm": 0.707801103591919,
+      "learning_rate": 8.900773230936366e-05,
+      "loss": 0.0845,
+      "step": 45790
+    },
+    {
+      "epoch": 2.996401701014066,
+      "grad_norm": 0.7957580089569092,
+      "learning_rate": 8.900198504698689e-05,
+      "loss": 0.0735,
+      "step": 45800
+    },
+    {
+      "epoch": 2.997055937193327,
+      "grad_norm": 1.1683952808380127,
+      "learning_rate": 8.899623646819087e-05,
+      "loss": 0.0686,
+      "step": 45810
+    },
+    {
+      "epoch": 2.9977101733725875,
+      "grad_norm": 0.8779153227806091,
+      "learning_rate": 8.899048657316956e-05,
+      "loss": 0.0689,
+      "step": 45820
+    },
+    {
+      "epoch": 2.998364409551848,
+      "grad_norm": 0.8528333306312561,
+      "learning_rate": 8.89847353621171e-05,
+      "loss": 0.0696,
+      "step": 45830
+    },
+    {
+      "epoch": 2.999018645731109,
+      "grad_norm": 0.9389132261276245,
+      "learning_rate": 8.897898283522756e-05,
+      "loss": 0.0808,
+      "step": 45840
+    },
+    {
+      "epoch": 2.9996728819103695,
+      "grad_norm": 0.8795191049575806,
+      "learning_rate": 8.897322899269513e-05,
+      "loss": 0.0823,
+      "step": 45850
+    },
+    {
+      "epoch": 3.0003271180896305,
+      "grad_norm": 0.8083428740501404,
+      "learning_rate": 8.896747383471402e-05,
+      "loss": 0.0703,
+      "step": 45860
+    },
+    {
+      "epoch": 3.000981354268891,
+      "grad_norm": 1.016609787940979,
+      "learning_rate": 8.896171736147846e-05,
+      "loss": 0.0926,
+      "step": 45870
+    },
+    {
+      "epoch": 3.001635590448152,
+      "grad_norm": 0.9670884609222412,
+      "learning_rate": 8.895595957318277e-05,
+      "loss": 0.0756,
+      "step": 45880
+    },
+    {
+      "epoch": 3.0022898266274125,
+      "grad_norm": 0.9883562922477722,
+      "learning_rate": 8.895020047002127e-05,
+      "loss": 0.0797,
+      "step": 45890
+    },
+    {
+      "epoch": 3.002944062806673,
+      "grad_norm": 0.7922661304473877,
+      "learning_rate": 8.894444005218835e-05,
+      "loss": 0.0877,
+      "step": 45900
+    },
+    {
+      "epoch": 3.003598298985934,
+      "grad_norm": 0.8021111488342285,
+      "learning_rate": 8.893867831987845e-05,
+      "loss": 0.0731,
+      "step": 45910
+    },
+    {
+      "epoch": 3.0042525351651945,
+      "grad_norm": 0.8403140902519226,
+      "learning_rate": 8.893291527328604e-05,
+      "loss": 0.0819,
+      "step": 45920
+    },
+    {
+      "epoch": 3.0049067713444555,
+      "grad_norm": 1.021638035774231,
+      "learning_rate": 8.892715091260564e-05,
+      "loss": 0.077,
+      "step": 45930
+    },
+    {
+      "epoch": 3.005561007523716,
+      "grad_norm": 0.723585307598114,
+      "learning_rate": 8.89213852380318e-05,
+      "loss": 0.0711,
+      "step": 45940
+    },
+    {
+      "epoch": 3.006215243702977,
+      "grad_norm": 0.807429313659668,
+      "learning_rate": 8.891561824975911e-05,
+      "loss": 0.0756,
+      "step": 45950
+    },
+    {
+      "epoch": 3.0068694798822375,
+      "grad_norm": 0.9832606911659241,
+      "learning_rate": 8.890984994798229e-05,
+      "loss": 0.0752,
+      "step": 45960
+    },
+    {
+      "epoch": 3.007523716061498,
+      "grad_norm": 0.9934665560722351,
+      "learning_rate": 8.890408033289595e-05,
+      "loss": 0.0826,
+      "step": 45970
+    },
+    {
+      "epoch": 3.008177952240759,
+      "grad_norm": 0.8675631880760193,
+      "learning_rate": 8.889830940469487e-05,
+      "loss": 0.0748,
+      "step": 45980
+    },
+    {
+      "epoch": 3.0088321884200195,
+      "grad_norm": 0.8953355550765991,
+      "learning_rate": 8.889253716357385e-05,
+      "loss": 0.0859,
+      "step": 45990
+    },
+    {
+      "epoch": 3.0094864245992805,
+      "grad_norm": 0.9766877889633179,
+      "learning_rate": 8.88867636097277e-05,
+      "loss": 0.0766,
+      "step": 46000
+    },
+    {
+      "epoch": 3.010140660778541,
+      "grad_norm": 0.8925821185112,
+      "learning_rate": 8.888098874335129e-05,
+      "loss": 0.0754,
+      "step": 46010
+    },
+    {
+      "epoch": 3.0107948969578016,
+      "grad_norm": 0.9773553013801575,
+      "learning_rate": 8.887521256463953e-05,
+      "loss": 0.0707,
+      "step": 46020
+    },
+    {
+      "epoch": 3.0114491331370625,
+      "grad_norm": 1.1067787408828735,
+      "learning_rate": 8.886943507378741e-05,
+      "loss": 0.0846,
+      "step": 46030
+    },
+    {
+      "epoch": 3.012103369316323,
+      "grad_norm": 0.7693238854408264,
+      "learning_rate": 8.886365627098991e-05,
+      "loss": 0.0787,
+      "step": 46040
+    },
+    {
+      "epoch": 3.012757605495584,
+      "grad_norm": 0.9626716375350952,
+      "learning_rate": 8.885787615644208e-05,
+      "loss": 0.0769,
+      "step": 46050
+    },
+    {
+      "epoch": 3.0134118416748445,
+      "grad_norm": 1.0409643650054932,
+      "learning_rate": 8.885209473033905e-05,
+      "loss": 0.076,
+      "step": 46060
+    },
+    {
+      "epoch": 3.0140660778541055,
+      "grad_norm": 0.8555409908294678,
+      "learning_rate": 8.88463119928759e-05,
+      "loss": 0.0866,
+      "step": 46070
+    },
+    {
+      "epoch": 3.014720314033366,
+      "grad_norm": 0.8579902052879333,
+      "learning_rate": 8.884052794424785e-05,
+      "loss": 0.0643,
+      "step": 46080
+    },
+    {
+      "epoch": 3.0153745502126266,
+      "grad_norm": 1.0672752857208252,
+      "learning_rate": 8.883474258465013e-05,
+      "loss": 0.084,
+      "step": 46090
+    },
+    {
+      "epoch": 3.0160287863918875,
+      "grad_norm": 0.9049206972122192,
+      "learning_rate": 8.882895591427797e-05,
+      "loss": 0.0775,
+      "step": 46100
+    },
+    {
+      "epoch": 3.016683022571148,
+      "grad_norm": 0.9947949647903442,
+      "learning_rate": 8.882316793332674e-05,
+      "loss": 0.0693,
+      "step": 46110
+    },
+    {
+      "epoch": 3.017337258750409,
+      "grad_norm": 0.6652380228042603,
+      "learning_rate": 8.881737864199177e-05,
+      "loss": 0.0727,
+      "step": 46120
+    },
+    {
+      "epoch": 3.0179914949296696,
+      "grad_norm": 0.7874847650527954,
+      "learning_rate": 8.881158804046847e-05,
+      "loss": 0.0773,
+      "step": 46130
+    },
+    {
+      "epoch": 3.0186457311089305,
+      "grad_norm": 0.8908420205116272,
+      "learning_rate": 8.880579612895227e-05,
+      "loss": 0.0756,
+      "step": 46140
+    },
+    {
+      "epoch": 3.019299967288191,
+      "grad_norm": 0.7234876751899719,
+      "learning_rate": 8.88000029076387e-05,
+      "loss": 0.0749,
+      "step": 46150
+    },
+    {
+      "epoch": 3.0199542034674516,
+      "grad_norm": 0.6146116256713867,
+      "learning_rate": 8.879420837672327e-05,
+      "loss": 0.073,
+      "step": 46160
+    },
+    {
+      "epoch": 3.0206084396467126,
+      "grad_norm": 0.8692243099212646,
+      "learning_rate": 8.878841253640156e-05,
+      "loss": 0.0831,
+      "step": 46170
+    },
+    {
+      "epoch": 3.021262675825973,
+      "grad_norm": 0.9530091881752014,
+      "learning_rate": 8.878261538686921e-05,
+      "loss": 0.0667,
+      "step": 46180
+    },
+    {
+      "epoch": 3.021916912005234,
+      "grad_norm": 0.8727622032165527,
+      "learning_rate": 8.877681692832187e-05,
+      "loss": 0.075,
+      "step": 46190
+    },
+    {
+      "epoch": 3.0225711481844946,
+      "grad_norm": 0.7857223749160767,
+      "learning_rate": 8.877101716095525e-05,
+      "loss": 0.0768,
+      "step": 46200
+    },
+    {
+      "epoch": 3.023225384363755,
+      "grad_norm": 0.929732084274292,
+      "learning_rate": 8.876521608496516e-05,
+      "loss": 0.0851,
+      "step": 46210
+    },
+    {
+      "epoch": 3.023879620543016,
+      "grad_norm": 0.8777803182601929,
+      "learning_rate": 8.875941370054733e-05,
+      "loss": 0.0718,
+      "step": 46220
+    },
+    {
+      "epoch": 3.0245338567222766,
+      "grad_norm": 0.99214106798172,
+      "learning_rate": 8.875361000789764e-05,
+      "loss": 0.0747,
+      "step": 46230
+    },
+    {
+      "epoch": 3.0251880929015376,
+      "grad_norm": 1.0649714469909668,
+      "learning_rate": 8.874780500721198e-05,
+      "loss": 0.0842,
+      "step": 46240
+    },
+    {
+      "epoch": 3.025842329080798,
+      "grad_norm": 0.7679323554039001,
+      "learning_rate": 8.874199869868629e-05,
+      "loss": 0.0746,
+      "step": 46250
+    },
+    {
+      "epoch": 3.026496565260059,
+      "grad_norm": 0.8521194458007812,
+      "learning_rate": 8.873619108251653e-05,
+      "loss": 0.0822,
+      "step": 46260
+    },
+    {
+      "epoch": 3.0271508014393196,
+      "grad_norm": 1.05009126663208,
+      "learning_rate": 8.873038215889872e-05,
+      "loss": 0.0786,
+      "step": 46270
+    },
+    {
+      "epoch": 3.02780503761858,
+      "grad_norm": 0.8912391662597656,
+      "learning_rate": 8.872457192802897e-05,
+      "loss": 0.0807,
+      "step": 46280
+    },
+    {
+      "epoch": 3.028459273797841,
+      "grad_norm": 0.7786313891410828,
+      "learning_rate": 8.871876039010334e-05,
+      "loss": 0.0881,
+      "step": 46290
+    },
+    {
+      "epoch": 3.0291135099771016,
+      "grad_norm": 0.6473716497421265,
+      "learning_rate": 8.8712947545318e-05,
+      "loss": 0.0724,
+      "step": 46300
+    },
+    {
+      "epoch": 3.0297677461563626,
+      "grad_norm": 0.8231253027915955,
+      "learning_rate": 8.870713339386916e-05,
+      "loss": 0.068,
+      "step": 46310
+    },
+    {
+      "epoch": 3.030421982335623,
+      "grad_norm": 0.8397748470306396,
+      "learning_rate": 8.870131793595304e-05,
+      "loss": 0.0842,
+      "step": 46320
+    },
+    {
+      "epoch": 3.031076218514884,
+      "grad_norm": 1.0040104389190674,
+      "learning_rate": 8.869550117176597e-05,
+      "loss": 0.0821,
+      "step": 46330
+    },
+    {
+      "epoch": 3.0317304546941446,
+      "grad_norm": 0.7324417233467102,
+      "learning_rate": 8.868968310150423e-05,
+      "loss": 0.073,
+      "step": 46340
+    },
+    {
+      "epoch": 3.032384690873405,
+      "grad_norm": 0.8197325468063354,
+      "learning_rate": 8.868386372536423e-05,
+      "loss": 0.0722,
+      "step": 46350
+    },
+    {
+      "epoch": 3.033038927052666,
+      "grad_norm": 0.8666032552719116,
+      "learning_rate": 8.867804304354237e-05,
+      "loss": 0.0721,
+      "step": 46360
+    },
+    {
+      "epoch": 3.0336931632319266,
+      "grad_norm": 0.9289610981941223,
+      "learning_rate": 8.867222105623512e-05,
+      "loss": 0.0715,
+      "step": 46370
+    },
+    {
+      "epoch": 3.0343473994111876,
+      "grad_norm": 0.851219892501831,
+      "learning_rate": 8.866639776363898e-05,
+      "loss": 0.078,
+      "step": 46380
+    },
+    {
+      "epoch": 3.035001635590448,
+      "grad_norm": 0.9999982118606567,
+      "learning_rate": 8.866057316595053e-05,
+      "loss": 0.081,
+      "step": 46390
+    },
+    {
+      "epoch": 3.035655871769709,
+      "grad_norm": 0.8318057656288147,
+      "learning_rate": 8.865474726336632e-05,
+      "loss": 0.0796,
+      "step": 46400
+    },
+    {
+      "epoch": 3.0363101079489696,
+      "grad_norm": 0.7948132157325745,
+      "learning_rate": 8.864892005608303e-05,
+      "loss": 0.0706,
+      "step": 46410
+    },
+    {
+      "epoch": 3.03696434412823,
+      "grad_norm": 0.9667806625366211,
+      "learning_rate": 8.864309154429733e-05,
+      "loss": 0.0714,
+      "step": 46420
+    },
+    {
+      "epoch": 3.037618580307491,
+      "grad_norm": 0.821867823600769,
+      "learning_rate": 8.863726172820593e-05,
+      "loss": 0.0669,
+      "step": 46430
+    },
+    {
+      "epoch": 3.0382728164867516,
+      "grad_norm": 0.8645114302635193,
+      "learning_rate": 8.863143060800563e-05,
+      "loss": 0.0738,
+      "step": 46440
+    },
+    {
+      "epoch": 3.0389270526660126,
+      "grad_norm": 1.0619080066680908,
+      "learning_rate": 8.862559818389322e-05,
+      "loss": 0.0883,
+      "step": 46450
+    },
+    {
+      "epoch": 3.039581288845273,
+      "grad_norm": 0.9746699333190918,
+      "learning_rate": 8.861976445606559e-05,
+      "loss": 0.0819,
+      "step": 46460
+    },
+    {
+      "epoch": 3.0402355250245336,
+      "grad_norm": 0.8296138644218445,
+      "learning_rate": 8.861392942471961e-05,
+      "loss": 0.073,
+      "step": 46470
+    },
+    {
+      "epoch": 3.0408897612037946,
+      "grad_norm": 0.7673683762550354,
+      "learning_rate": 8.860809309005224e-05,
+      "loss": 0.0815,
+      "step": 46480
+    },
+    {
+      "epoch": 3.041543997383055,
+      "grad_norm": 0.9790584444999695,
+      "learning_rate": 8.860225545226049e-05,
+      "loss": 0.084,
+      "step": 46490
+    },
+    {
+      "epoch": 3.042198233562316,
+      "grad_norm": 0.7433209419250488,
+      "learning_rate": 8.859641651154138e-05,
+      "loss": 0.0741,
+      "step": 46500
+    },
+    {
+      "epoch": 3.0428524697415766,
+      "grad_norm": 0.8458490371704102,
+      "learning_rate": 8.859057626809198e-05,
+      "loss": 0.0691,
+      "step": 46510
+    },
+    {
+      "epoch": 3.0435067059208376,
+      "grad_norm": 0.8308846950531006,
+      "learning_rate": 8.858473472210944e-05,
+      "loss": 0.0762,
+      "step": 46520
+    },
+    {
+      "epoch": 3.044160942100098,
+      "grad_norm": 0.8369269967079163,
+      "learning_rate": 8.857889187379089e-05,
+      "loss": 0.0701,
+      "step": 46530
+    },
+    {
+      "epoch": 3.0448151782793587,
+      "grad_norm": 0.8432846665382385,
+      "learning_rate": 8.857304772333357e-05,
+      "loss": 0.0691,
+      "step": 46540
+    },
+    {
+      "epoch": 3.0454694144586196,
+      "grad_norm": 0.7842530608177185,
+      "learning_rate": 8.856720227093474e-05,
+      "loss": 0.085,
+      "step": 46550
+    },
+    {
+      "epoch": 3.04612365063788,
+      "grad_norm": 0.7288837432861328,
+      "learning_rate": 8.856135551679166e-05,
+      "loss": 0.0784,
+      "step": 46560
+    },
+    {
+      "epoch": 3.046777886817141,
+      "grad_norm": 0.7969174981117249,
+      "learning_rate": 8.855550746110171e-05,
+      "loss": 0.0849,
+      "step": 46570
+    },
+    {
+      "epoch": 3.0474321229964016,
+      "grad_norm": 0.9206051826477051,
+      "learning_rate": 8.854965810406229e-05,
+      "loss": 0.083,
+      "step": 46580
+    },
+    {
+      "epoch": 3.0480863591756626,
+      "grad_norm": 0.8779579401016235,
+      "learning_rate": 8.854380744587078e-05,
+      "loss": 0.0749,
+      "step": 46590
+    },
+    {
+      "epoch": 3.048740595354923,
+      "grad_norm": 0.9052606225013733,
+      "learning_rate": 8.85379554867247e-05,
+      "loss": 0.0712,
+      "step": 46600
+    },
+    {
+      "epoch": 3.0493948315341837,
+      "grad_norm": 0.8449774980545044,
+      "learning_rate": 8.853210222682156e-05,
+      "loss": 0.0723,
+      "step": 46610
+    },
+    {
+      "epoch": 3.0500490677134446,
+      "grad_norm": 0.8391105532646179,
+      "learning_rate": 8.85262476663589e-05,
+      "loss": 0.0764,
+      "step": 46620
+    },
+    {
+      "epoch": 3.050703303892705,
+      "grad_norm": 0.8949602842330933,
+      "learning_rate": 8.852039180553436e-05,
+      "loss": 0.0765,
+      "step": 46630
+    },
+    {
+      "epoch": 3.051357540071966,
+      "grad_norm": 0.7765173316001892,
+      "learning_rate": 8.851453464454554e-05,
+      "loss": 0.081,
+      "step": 46640
+    },
+    {
+      "epoch": 3.0520117762512267,
+      "grad_norm": 0.7792592644691467,
+      "learning_rate": 8.85086761835902e-05,
+      "loss": 0.0815,
+      "step": 46650
+    },
+    {
+      "epoch": 3.052666012430487,
+      "grad_norm": 0.8847215175628662,
+      "learning_rate": 8.850281642286603e-05,
+      "loss": 0.071,
+      "step": 46660
+    },
+    {
+      "epoch": 3.053320248609748,
+      "grad_norm": 0.9311257004737854,
+      "learning_rate": 8.849695536257083e-05,
+      "loss": 0.0749,
+      "step": 46670
+    },
+    {
+      "epoch": 3.0539744847890087,
+      "grad_norm": 0.7684807181358337,
+      "learning_rate": 8.849109300290242e-05,
+      "loss": 0.0735,
+      "step": 46680
+    },
+    {
+      "epoch": 3.0546287209682697,
+      "grad_norm": 0.864181399345398,
+      "learning_rate": 8.84852293440587e-05,
+      "loss": 0.0766,
+      "step": 46690
+    },
+    {
+      "epoch": 3.05528295714753,
+      "grad_norm": 0.8090917468070984,
+      "learning_rate": 8.847936438623754e-05,
+      "loss": 0.0857,
+      "step": 46700
+    },
+    {
+      "epoch": 3.055937193326791,
+      "grad_norm": 0.8556300401687622,
+      "learning_rate": 8.84734981296369e-05,
+      "loss": 0.0884,
+      "step": 46710
+    },
+    {
+      "epoch": 3.0565914295060517,
+      "grad_norm": 0.860185444355011,
+      "learning_rate": 8.846763057445481e-05,
+      "loss": 0.0749,
+      "step": 46720
+    },
+    {
+      "epoch": 3.057245665685312,
+      "grad_norm": 0.983341634273529,
+      "learning_rate": 8.846176172088932e-05,
+      "loss": 0.078,
+      "step": 46730
+    },
+    {
+      "epoch": 3.057899901864573,
+      "grad_norm": 0.7510449886322021,
+      "learning_rate": 8.84558915691385e-05,
+      "loss": 0.0709,
+      "step": 46740
+    },
+    {
+      "epoch": 3.0585541380438337,
+      "grad_norm": 0.8257303237915039,
+      "learning_rate": 8.845002011940047e-05,
+      "loss": 0.0808,
+      "step": 46750
+    },
+    {
+      "epoch": 3.0592083742230947,
+      "grad_norm": 0.8801044821739197,
+      "learning_rate": 8.844414737187342e-05,
+      "loss": 0.0736,
+      "step": 46760
+    },
+    {
+      "epoch": 3.059862610402355,
+      "grad_norm": 0.8823607563972473,
+      "learning_rate": 8.843827332675558e-05,
+      "loss": 0.0812,
+      "step": 46770
+    },
+    {
+      "epoch": 3.060516846581616,
+      "grad_norm": 1.0760579109191895,
+      "learning_rate": 8.84323979842452e-05,
+      "loss": 0.0779,
+      "step": 46780
+    },
+    {
+      "epoch": 3.0611710827608767,
+      "grad_norm": 0.8162868022918701,
+      "learning_rate": 8.842652134454061e-05,
+      "loss": 0.0686,
+      "step": 46790
+    },
+    {
+      "epoch": 3.061825318940137,
+      "grad_norm": 0.8627833724021912,
+      "learning_rate": 8.842064340784013e-05,
+      "loss": 0.0785,
+      "step": 46800
+    },
+    {
+      "epoch": 3.062479555119398,
+      "grad_norm": 0.6602069735527039,
+      "learning_rate": 8.84147641743422e-05,
+      "loss": 0.0682,
+      "step": 46810
+    },
+    {
+      "epoch": 3.0631337912986587,
+      "grad_norm": 0.8161008954048157,
+      "learning_rate": 8.84088836442452e-05,
+      "loss": 0.0776,
+      "step": 46820
+    },
+    {
+      "epoch": 3.0637880274779197,
+      "grad_norm": 0.9480443000793457,
+      "learning_rate": 8.840300181774767e-05,
+      "loss": 0.0783,
+      "step": 46830
+    },
+    {
+      "epoch": 3.06444226365718,
+      "grad_norm": 0.92530757188797,
+      "learning_rate": 8.83971186950481e-05,
+      "loss": 0.0822,
+      "step": 46840
+    },
+    {
+      "epoch": 3.065096499836441,
+      "grad_norm": 1.08670973777771,
+      "learning_rate": 8.839123427634508e-05,
+      "loss": 0.0785,
+      "step": 46850
+    },
+    {
+      "epoch": 3.0657507360157017,
+      "grad_norm": 0.8924565315246582,
+      "learning_rate": 8.838534856183722e-05,
+      "loss": 0.0715,
+      "step": 46860
+    },
+    {
+      "epoch": 3.066404972194962,
+      "grad_norm": 0.6959300637245178,
+      "learning_rate": 8.837946155172318e-05,
+      "loss": 0.0748,
+      "step": 46870
+    },
+    {
+      "epoch": 3.067059208374223,
+      "grad_norm": 0.7405787110328674,
+      "learning_rate": 8.837357324620164e-05,
+      "loss": 0.0748,
+      "step": 46880
+    },
+    {
+      "epoch": 3.0677134445534837,
+      "grad_norm": 0.8853090405464172,
+      "learning_rate": 8.836768364547139e-05,
+      "loss": 0.0854,
+      "step": 46890
+    },
+    {
+      "epoch": 3.0683676807327447,
+      "grad_norm": 0.9329870939254761,
+      "learning_rate": 8.836179274973118e-05,
+      "loss": 0.0803,
+      "step": 46900
+    },
+    {
+      "epoch": 3.069021916912005,
+      "grad_norm": 0.9848145246505737,
+      "learning_rate": 8.835590055917984e-05,
+      "loss": 0.0866,
+      "step": 46910
+    },
+    {
+      "epoch": 3.0696761530912657,
+      "grad_norm": 0.8550765514373779,
+      "learning_rate": 8.835000707401628e-05,
+      "loss": 0.0753,
+      "step": 46920
+    },
+    {
+      "epoch": 3.0703303892705267,
+      "grad_norm": 0.7886782288551331,
+      "learning_rate": 8.834411229443942e-05,
+      "loss": 0.0732,
+      "step": 46930
+    },
+    {
+      "epoch": 3.0709846254497872,
+      "grad_norm": 0.7411202788352966,
+      "learning_rate": 8.833821622064819e-05,
+      "loss": 0.0774,
+      "step": 46940
+    },
+    {
+      "epoch": 3.071638861629048,
+      "grad_norm": 1.315242052078247,
+      "learning_rate": 8.833231885284164e-05,
+      "loss": 0.0744,
+      "step": 46950
+    },
+    {
+      "epoch": 3.0722930978083087,
+      "grad_norm": 0.9081234335899353,
+      "learning_rate": 8.832642019121877e-05,
+      "loss": 0.0826,
+      "step": 46960
+    },
+    {
+      "epoch": 3.0729473339875697,
+      "grad_norm": 0.8793578147888184,
+      "learning_rate": 8.832052023597872e-05,
+      "loss": 0.0802,
+      "step": 46970
+    },
+    {
+      "epoch": 3.07360157016683,
+      "grad_norm": 0.9718412756919861,
+      "learning_rate": 8.831461898732061e-05,
+      "loss": 0.0784,
+      "step": 46980
+    },
+    {
+      "epoch": 3.0742558063460907,
+      "grad_norm": 0.7683785557746887,
+      "learning_rate": 8.83087164454436e-05,
+      "loss": 0.0773,
+      "step": 46990
+    },
+    {
+      "epoch": 3.0749100425253517,
+      "grad_norm": 0.8692295551300049,
+      "learning_rate": 8.830281261054698e-05,
+      "loss": 0.0702,
+      "step": 47000
+    },
+    {
+      "epoch": 3.0755642787046122,
+      "grad_norm": 0.8362782597541809,
+      "learning_rate": 8.829690748282997e-05,
+      "loss": 0.0676,
+      "step": 47010
+    },
+    {
+      "epoch": 3.076218514883873,
+      "grad_norm": 1.0380651950836182,
+      "learning_rate": 8.829100106249189e-05,
+      "loss": 0.0862,
+      "step": 47020
+    },
+    {
+      "epoch": 3.0768727510631337,
+      "grad_norm": 0.9438252449035645,
+      "learning_rate": 8.828509334973209e-05,
+      "loss": 0.0737,
+      "step": 47030
+    },
+    {
+      "epoch": 3.0775269872423947,
+      "grad_norm": 0.8855924606323242,
+      "learning_rate": 8.827918434475001e-05,
+      "loss": 0.0872,
+      "step": 47040
+    },
+    {
+      "epoch": 3.0781812234216552,
+      "grad_norm": 0.8621238470077515,
+      "learning_rate": 8.827327404774505e-05,
+      "loss": 0.0671,
+      "step": 47050
+    },
+    {
+      "epoch": 3.0788354596009158,
+      "grad_norm": 0.7865177392959595,
+      "learning_rate": 8.826736245891672e-05,
+      "loss": 0.0722,
+      "step": 47060
+    },
+    {
+      "epoch": 3.0794896957801767,
+      "grad_norm": 1.0064588785171509,
+      "learning_rate": 8.826144957846455e-05,
+      "loss": 0.069,
+      "step": 47070
+    },
+    {
+      "epoch": 3.0801439319594373,
+      "grad_norm": 0.8774595856666565,
+      "learning_rate": 8.825553540658811e-05,
+      "loss": 0.0753,
+      "step": 47080
+    },
+    {
+      "epoch": 3.0807981681386982,
+      "grad_norm": 0.9509215950965881,
+      "learning_rate": 8.824961994348701e-05,
+      "loss": 0.0741,
+      "step": 47090
+    },
+    {
+      "epoch": 3.0814524043179587,
+      "grad_norm": 0.8457993865013123,
+      "learning_rate": 8.824370318936095e-05,
+      "loss": 0.0712,
+      "step": 47100
+    },
+    {
+      "epoch": 3.0821066404972193,
+      "grad_norm": 0.8608481884002686,
+      "learning_rate": 8.823778514440959e-05,
+      "loss": 0.0801,
+      "step": 47110
+    },
+    {
+      "epoch": 3.0827608766764802,
+      "grad_norm": 0.9374340772628784,
+      "learning_rate": 8.82318658088327e-05,
+      "loss": 0.0866,
+      "step": 47120
+    },
+    {
+      "epoch": 3.0834151128557408,
+      "grad_norm": 0.9799492955207825,
+      "learning_rate": 8.822594518283008e-05,
+      "loss": 0.0875,
+      "step": 47130
+    },
+    {
+      "epoch": 3.0840693490350017,
+      "grad_norm": 0.8801363706588745,
+      "learning_rate": 8.822002326660154e-05,
+      "loss": 0.0756,
+      "step": 47140
+    },
+    {
+      "epoch": 3.0847235852142623,
+      "grad_norm": 0.8965553641319275,
+      "learning_rate": 8.8214100060347e-05,
+      "loss": 0.0792,
+      "step": 47150
+    },
+    {
+      "epoch": 3.0853778213935232,
+      "grad_norm": 0.8476041555404663,
+      "learning_rate": 8.820817556426636e-05,
+      "loss": 0.075,
+      "step": 47160
+    },
+    {
+      "epoch": 3.0860320575727838,
+      "grad_norm": 0.9920099973678589,
+      "learning_rate": 8.82022497785596e-05,
+      "loss": 0.0683,
+      "step": 47170
+    },
+    {
+      "epoch": 3.0866862937520443,
+      "grad_norm": 0.8582835793495178,
+      "learning_rate": 8.81963227034267e-05,
+      "loss": 0.0713,
+      "step": 47180
+    },
+    {
+      "epoch": 3.0873405299313053,
+      "grad_norm": 1.1154992580413818,
+      "learning_rate": 8.819039433906774e-05,
+      "loss": 0.0937,
+      "step": 47190
+    },
+    {
+      "epoch": 3.087994766110566,
+      "grad_norm": 1.022351861000061,
+      "learning_rate": 8.818446468568282e-05,
+      "loss": 0.0752,
+      "step": 47200
+    },
+    {
+      "epoch": 3.0886490022898268,
+      "grad_norm": 0.8366675972938538,
+      "learning_rate": 8.817853374347206e-05,
+      "loss": 0.0817,
+      "step": 47210
+    },
+    {
+      "epoch": 3.0893032384690873,
+      "grad_norm": 0.9166560769081116,
+      "learning_rate": 8.817260151263568e-05,
+      "loss": 0.0757,
+      "step": 47220
+    },
+    {
+      "epoch": 3.0899574746483482,
+      "grad_norm": 1.0350792407989502,
+      "learning_rate": 8.816666799337388e-05,
+      "loss": 0.0726,
+      "step": 47230
+    },
+    {
+      "epoch": 3.0906117108276088,
+      "grad_norm": 1.0422755479812622,
+      "learning_rate": 8.816073318588693e-05,
+      "loss": 0.0991,
+      "step": 47240
+    },
+    {
+      "epoch": 3.0912659470068693,
+      "grad_norm": 1.030333161354065,
+      "learning_rate": 8.815479709037515e-05,
+      "loss": 0.0694,
+      "step": 47250
+    },
+    {
+      "epoch": 3.0919201831861303,
+      "grad_norm": 1.0390321016311646,
+      "learning_rate": 8.814885970703893e-05,
+      "loss": 0.0855,
+      "step": 47260
+    },
+    {
+      "epoch": 3.092574419365391,
+      "grad_norm": 0.8490506410598755,
+      "learning_rate": 8.814292103607862e-05,
+      "loss": 0.0812,
+      "step": 47270
+    },
+    {
+      "epoch": 3.0932286555446518,
+      "grad_norm": 0.8992865681648254,
+      "learning_rate": 8.813698107769471e-05,
+      "loss": 0.0766,
+      "step": 47280
+    },
+    {
+      "epoch": 3.0938828917239123,
+      "grad_norm": 0.9719147086143494,
+      "learning_rate": 8.813103983208766e-05,
+      "loss": 0.0738,
+      "step": 47290
+    },
+    {
+      "epoch": 3.0945371279031733,
+      "grad_norm": 0.7177313566207886,
+      "learning_rate": 8.812509729945802e-05,
+      "loss": 0.0762,
+      "step": 47300
+    },
+    {
+      "epoch": 3.095191364082434,
+      "grad_norm": 0.8135479092597961,
+      "learning_rate": 8.811915348000635e-05,
+      "loss": 0.0761,
+      "step": 47310
+    },
+    {
+      "epoch": 3.0958456002616943,
+      "grad_norm": 0.8107147216796875,
+      "learning_rate": 8.811320837393329e-05,
+      "loss": 0.069,
+      "step": 47320
+    },
+    {
+      "epoch": 3.0964998364409553,
+      "grad_norm": 0.881219744682312,
+      "learning_rate": 8.810726198143949e-05,
+      "loss": 0.0791,
+      "step": 47330
+    },
+    {
+      "epoch": 3.097154072620216,
+      "grad_norm": 0.9612583518028259,
+      "learning_rate": 8.810131430272564e-05,
+      "loss": 0.0726,
+      "step": 47340
+    },
+    {
+      "epoch": 3.0978083087994768,
+      "grad_norm": 1.0407850742340088,
+      "learning_rate": 8.809536533799253e-05,
+      "loss": 0.0722,
+      "step": 47350
+    },
+    {
+      "epoch": 3.0984625449787373,
+      "grad_norm": 0.861651599407196,
+      "learning_rate": 8.808941508744093e-05,
+      "loss": 0.0808,
+      "step": 47360
+    },
+    {
+      "epoch": 3.099116781157998,
+      "grad_norm": 0.8344012498855591,
+      "learning_rate": 8.808346355127166e-05,
+      "loss": 0.0697,
+      "step": 47370
+    },
+    {
+      "epoch": 3.099771017337259,
+      "grad_norm": 0.8500309586524963,
+      "learning_rate": 8.807751072968563e-05,
+      "loss": 0.0738,
+      "step": 47380
+    },
+    {
+      "epoch": 3.1004252535165193,
+      "grad_norm": 0.8990835547447205,
+      "learning_rate": 8.807155662288375e-05,
+      "loss": 0.0724,
+      "step": 47390
+    },
+    {
+      "epoch": 3.1010794896957803,
+      "grad_norm": 0.8191524744033813,
+      "learning_rate": 8.8065601231067e-05,
+      "loss": 0.0699,
+      "step": 47400
+    },
+    {
+      "epoch": 3.101733725875041,
+      "grad_norm": 0.9893473982810974,
+      "learning_rate": 8.805964455443636e-05,
+      "loss": 0.0728,
+      "step": 47410
+    },
+    {
+      "epoch": 3.102387962054302,
+      "grad_norm": 0.8191789984703064,
+      "learning_rate": 8.805368659319291e-05,
+      "loss": 0.0682,
+      "step": 47420
+    },
+    {
+      "epoch": 3.1030421982335623,
+      "grad_norm": 0.8927891850471497,
+      "learning_rate": 8.804772734753773e-05,
+      "loss": 0.0692,
+      "step": 47430
+    },
+    {
+      "epoch": 3.103696434412823,
+      "grad_norm": 0.8646766543388367,
+      "learning_rate": 8.804176681767196e-05,
+      "loss": 0.0781,
+      "step": 47440
+    },
+    {
+      "epoch": 3.104350670592084,
+      "grad_norm": 0.9096436500549316,
+      "learning_rate": 8.803580500379681e-05,
+      "loss": 0.0854,
+      "step": 47450
+    },
+    {
+      "epoch": 3.1050049067713443,
+      "grad_norm": 0.8713791370391846,
+      "learning_rate": 8.802984190611349e-05,
+      "loss": 0.0833,
+      "step": 47460
+    },
+    {
+      "epoch": 3.1056591429506053,
+      "grad_norm": 0.9313731789588928,
+      "learning_rate": 8.802387752482327e-05,
+      "loss": 0.0796,
+      "step": 47470
+    },
+    {
+      "epoch": 3.106313379129866,
+      "grad_norm": 0.8143359422683716,
+      "learning_rate": 8.801791186012744e-05,
+      "loss": 0.0747,
+      "step": 47480
+    },
+    {
+      "epoch": 3.106967615309127,
+      "grad_norm": 0.8441165685653687,
+      "learning_rate": 8.80119449122274e-05,
+      "loss": 0.0727,
+      "step": 47490
+    },
+    {
+      "epoch": 3.1076218514883873,
+      "grad_norm": 0.9759175777435303,
+      "learning_rate": 8.800597668132452e-05,
+      "loss": 0.0743,
+      "step": 47500
+    },
+    {
+      "epoch": 3.108276087667648,
+      "grad_norm": 0.8443938493728638,
+      "learning_rate": 8.800000716762024e-05,
+      "loss": 0.0739,
+      "step": 47510
+    },
+    {
+      "epoch": 3.108930323846909,
+      "grad_norm": 0.9399232864379883,
+      "learning_rate": 8.799403637131609e-05,
+      "loss": 0.0766,
+      "step": 47520
+    },
+    {
+      "epoch": 3.1095845600261693,
+      "grad_norm": 0.8907540440559387,
+      "learning_rate": 8.798806429261355e-05,
+      "loss": 0.0757,
+      "step": 47530
+    },
+    {
+      "epoch": 3.1102387962054303,
+      "grad_norm": 0.9284667372703552,
+      "learning_rate": 8.798209093171421e-05,
+      "loss": 0.0811,
+      "step": 47540
+    },
+    {
+      "epoch": 3.110893032384691,
+      "grad_norm": 0.9799740314483643,
+      "learning_rate": 8.79761162888197e-05,
+      "loss": 0.0741,
+      "step": 47550
+    },
+    {
+      "epoch": 3.1115472685639514,
+      "grad_norm": 0.9789896011352539,
+      "learning_rate": 8.797014036413167e-05,
+      "loss": 0.0732,
+      "step": 47560
+    },
+    {
+      "epoch": 3.1122015047432123,
+      "grad_norm": 0.7399867177009583,
+      "learning_rate": 8.796416315785181e-05,
+      "loss": 0.0726,
+      "step": 47570
+    },
+    {
+      "epoch": 3.112855740922473,
+      "grad_norm": 0.9549564719200134,
+      "learning_rate": 8.795818467018188e-05,
+      "loss": 0.0777,
+      "step": 47580
+    },
+    {
+      "epoch": 3.113509977101734,
+      "grad_norm": 0.9988911151885986,
+      "learning_rate": 8.795220490132369e-05,
+      "loss": 0.0775,
+      "step": 47590
+    },
+    {
+      "epoch": 3.1141642132809944,
+      "grad_norm": 0.9814069271087646,
+      "learning_rate": 8.794622385147903e-05,
+      "loss": 0.0853,
+      "step": 47600
+    },
+    {
+      "epoch": 3.1148184494602553,
+      "grad_norm": 0.7243561148643494,
+      "learning_rate": 8.79402415208498e-05,
+      "loss": 0.0778,
+      "step": 47610
+    },
+    {
+      "epoch": 3.115472685639516,
+      "grad_norm": 0.8936898708343506,
+      "learning_rate": 8.793425790963792e-05,
+      "loss": 0.0838,
+      "step": 47620
+    },
+    {
+      "epoch": 3.1161269218187764,
+      "grad_norm": 0.9839106202125549,
+      "learning_rate": 8.792827301804536e-05,
+      "loss": 0.0809,
+      "step": 47630
+    },
+    {
+      "epoch": 3.1167811579980373,
+      "grad_norm": 0.9321417212486267,
+      "learning_rate": 8.79222868462741e-05,
+      "loss": 0.0699,
+      "step": 47640
+    },
+    {
+      "epoch": 3.117435394177298,
+      "grad_norm": 0.9388688206672668,
+      "learning_rate": 8.791629939452621e-05,
+      "loss": 0.0872,
+      "step": 47650
+    },
+    {
+      "epoch": 3.118089630356559,
+      "grad_norm": 0.7993639707565308,
+      "learning_rate": 8.791031066300378e-05,
+      "loss": 0.0793,
+      "step": 47660
+    },
+    {
+      "epoch": 3.1187438665358194,
+      "grad_norm": 0.7926256060600281,
+      "learning_rate": 8.790432065190892e-05,
+      "loss": 0.0762,
+      "step": 47670
+    },
+    {
+      "epoch": 3.1193981027150803,
+      "grad_norm": 0.8904451131820679,
+      "learning_rate": 8.789832936144386e-05,
+      "loss": 0.0817,
+      "step": 47680
+    },
+    {
+      "epoch": 3.120052338894341,
+      "grad_norm": 1.026097059249878,
+      "learning_rate": 8.789233679181077e-05,
+      "loss": 0.0745,
+      "step": 47690
+    },
+    {
+      "epoch": 3.1207065750736014,
+      "grad_norm": 0.8326067328453064,
+      "learning_rate": 8.788634294321195e-05,
+      "loss": 0.0662,
+      "step": 47700
+    },
+    {
+      "epoch": 3.1213608112528624,
+      "grad_norm": 0.9273573160171509,
+      "learning_rate": 8.788034781584968e-05,
+      "loss": 0.0782,
+      "step": 47710
+    },
+    {
+      "epoch": 3.122015047432123,
+      "grad_norm": 0.7090617418289185,
+      "learning_rate": 8.787435140992635e-05,
+      "loss": 0.0721,
+      "step": 47720
+    },
+    {
+      "epoch": 3.122669283611384,
+      "grad_norm": 0.823361873626709,
+      "learning_rate": 8.786835372564431e-05,
+      "loss": 0.0743,
+      "step": 47730
+    },
+    {
+      "epoch": 3.1233235197906444,
+      "grad_norm": 0.8674502372741699,
+      "learning_rate": 8.786235476320603e-05,
+      "loss": 0.064,
+      "step": 47740
+    },
+    {
+      "epoch": 3.1239777559699053,
+      "grad_norm": 0.8977984189987183,
+      "learning_rate": 8.785635452281397e-05,
+      "loss": 0.0797,
+      "step": 47750
+    },
+    {
+      "epoch": 3.124631992149166,
+      "grad_norm": 0.7301965355873108,
+      "learning_rate": 8.785035300467068e-05,
+      "loss": 0.0726,
+      "step": 47760
+    },
+    {
+      "epoch": 3.1252862283284264,
+      "grad_norm": 0.8135896921157837,
+      "learning_rate": 8.78443502089787e-05,
+      "loss": 0.0799,
+      "step": 47770
+    },
+    {
+      "epoch": 3.1259404645076874,
+      "grad_norm": 0.8342558145523071,
+      "learning_rate": 8.783834613594064e-05,
+      "loss": 0.0738,
+      "step": 47780
+    },
+    {
+      "epoch": 3.126594700686948,
+      "grad_norm": 0.7653647065162659,
+      "learning_rate": 8.783234078575917e-05,
+      "loss": 0.0768,
+      "step": 47790
+    },
+    {
+      "epoch": 3.127248936866209,
+      "grad_norm": 0.7020546793937683,
+      "learning_rate": 8.7826334158637e-05,
+      "loss": 0.0703,
+      "step": 47800
+    },
+    {
+      "epoch": 3.1279031730454694,
+      "grad_norm": 0.9205434918403625,
+      "learning_rate": 8.782032625477681e-05,
+      "loss": 0.0709,
+      "step": 47810
+    },
+    {
+      "epoch": 3.1285574092247304,
+      "grad_norm": 0.830418586730957,
+      "learning_rate": 8.781431707438145e-05,
+      "loss": 0.0749,
+      "step": 47820
+    },
+    {
+      "epoch": 3.129211645403991,
+      "grad_norm": 1.1934880018234253,
+      "learning_rate": 8.780830661765371e-05,
+      "loss": 0.072,
+      "step": 47830
+    },
+    {
+      "epoch": 3.1298658815832514,
+      "grad_norm": 0.9115713834762573,
+      "learning_rate": 8.780229488479646e-05,
+      "loss": 0.0744,
+      "step": 47840
+    },
+    {
+      "epoch": 3.1305201177625124,
+      "grad_norm": 0.963766872882843,
+      "learning_rate": 8.779628187601261e-05,
+      "loss": 0.0832,
+      "step": 47850
+    },
+    {
+      "epoch": 3.131174353941773,
+      "grad_norm": 0.7308657765388489,
+      "learning_rate": 8.779026759150515e-05,
+      "loss": 0.0826,
+      "step": 47860
+    },
+    {
+      "epoch": 3.131828590121034,
+      "grad_norm": 0.8448659777641296,
+      "learning_rate": 8.778425203147703e-05,
+      "loss": 0.0819,
+      "step": 47870
+    },
+    {
+      "epoch": 3.1324828263002944,
+      "grad_norm": 1.0432631969451904,
+      "learning_rate": 8.777823519613131e-05,
+      "loss": 0.0763,
+      "step": 47880
+    },
+    {
+      "epoch": 3.133137062479555,
+      "grad_norm": 0.9087404608726501,
+      "learning_rate": 8.77722170856711e-05,
+      "loss": 0.0852,
+      "step": 47890
+    },
+    {
+      "epoch": 3.133791298658816,
+      "grad_norm": 0.8869801163673401,
+      "learning_rate": 8.776619770029946e-05,
+      "loss": 0.0716,
+      "step": 47900
+    },
+    {
+      "epoch": 3.1344455348380764,
+      "grad_norm": 0.7883999943733215,
+      "learning_rate": 8.776017704021964e-05,
+      "loss": 0.0694,
+      "step": 47910
+    },
+    {
+      "epoch": 3.1350997710173374,
+      "grad_norm": 1.1051850318908691,
+      "learning_rate": 8.77541551056348e-05,
+      "loss": 0.083,
+      "step": 47920
+    },
+    {
+      "epoch": 3.135754007196598,
+      "grad_norm": 0.7399940490722656,
+      "learning_rate": 8.77481318967482e-05,
+      "loss": 0.0688,
+      "step": 47930
+    },
+    {
+      "epoch": 3.136408243375859,
+      "grad_norm": 0.8819198608398438,
+      "learning_rate": 8.774210741376316e-05,
+      "loss": 0.0813,
+      "step": 47940
+    },
+    {
+      "epoch": 3.1370624795551194,
+      "grad_norm": 0.8618602156639099,
+      "learning_rate": 8.773608165688303e-05,
+      "loss": 0.081,
+      "step": 47950
+    },
+    {
+      "epoch": 3.13771671573438,
+      "grad_norm": 0.9459993243217468,
+      "learning_rate": 8.773005462631115e-05,
+      "loss": 0.072,
+      "step": 47960
+    },
+    {
+      "epoch": 3.138370951913641,
+      "grad_norm": 0.8384934067726135,
+      "learning_rate": 8.772402632225098e-05,
+      "loss": 0.0748,
+      "step": 47970
+    },
+    {
+      "epoch": 3.1390251880929014,
+      "grad_norm": 0.803860604763031,
+      "learning_rate": 8.7717996744906e-05,
+      "loss": 0.0765,
+      "step": 47980
+    },
+    {
+      "epoch": 3.1396794242721624,
+      "grad_norm": 1.1907633543014526,
+      "learning_rate": 8.77119658944797e-05,
+      "loss": 0.0909,
+      "step": 47990
+    },
+    {
+      "epoch": 3.140333660451423,
+      "grad_norm": 0.8750865459442139,
+      "learning_rate": 8.770593377117566e-05,
+      "loss": 0.067,
+      "step": 48000
+    },
+    {
+      "epoch": 3.1409878966306835,
+      "grad_norm": 0.7471033930778503,
+      "learning_rate": 8.769990037519747e-05,
+      "loss": 0.0719,
+      "step": 48010
+    },
+    {
+      "epoch": 3.1416421328099444,
+      "grad_norm": 1.0200992822647095,
+      "learning_rate": 8.769386570674876e-05,
+      "loss": 0.0804,
+      "step": 48020
+    },
+    {
+      "epoch": 3.142296368989205,
+      "grad_norm": 0.8135073184967041,
+      "learning_rate": 8.768782976603323e-05,
+      "loss": 0.0677,
+      "step": 48030
+    },
+    {
+      "epoch": 3.142950605168466,
+      "grad_norm": 0.8066560626029968,
+      "learning_rate": 8.76817925532546e-05,
+      "loss": 0.0816,
+      "step": 48040
+    },
+    {
+      "epoch": 3.1436048413477264,
+      "grad_norm": 0.9547781348228455,
+      "learning_rate": 8.767575406861665e-05,
+      "loss": 0.0732,
+      "step": 48050
+    },
+    {
+      "epoch": 3.1442590775269874,
+      "grad_norm": 0.9590566754341125,
+      "learning_rate": 8.766971431232318e-05,
+      "loss": 0.0797,
+      "step": 48060
+    },
+    {
+      "epoch": 3.144913313706248,
+      "grad_norm": 0.8913863897323608,
+      "learning_rate": 8.766367328457808e-05,
+      "loss": 0.08,
+      "step": 48070
+    },
+    {
+      "epoch": 3.1455675498855085,
+      "grad_norm": 0.9507361650466919,
+      "learning_rate": 8.765763098558521e-05,
+      "loss": 0.0722,
+      "step": 48080
+    },
+    {
+      "epoch": 3.1462217860647694,
+      "grad_norm": 1.0220569372177124,
+      "learning_rate": 8.765158741554855e-05,
+      "loss": 0.083,
+      "step": 48090
+    },
+    {
+      "epoch": 3.14687602224403,
+      "grad_norm": 0.9265881776809692,
+      "learning_rate": 8.764554257467207e-05,
+      "loss": 0.0781,
+      "step": 48100
+    },
+    {
+      "epoch": 3.147530258423291,
+      "grad_norm": 0.765208899974823,
+      "learning_rate": 8.763949646315979e-05,
+      "loss": 0.0691,
+      "step": 48110
+    },
+    {
+      "epoch": 3.1481844946025515,
+      "grad_norm": 0.7979570627212524,
+      "learning_rate": 8.76334490812158e-05,
+      "loss": 0.0834,
+      "step": 48120
+    },
+    {
+      "epoch": 3.1488387307818124,
+      "grad_norm": 0.6835886240005493,
+      "learning_rate": 8.76274004290442e-05,
+      "loss": 0.0679,
+      "step": 48130
+    },
+    {
+      "epoch": 3.149492966961073,
+      "grad_norm": 0.8190062642097473,
+      "learning_rate": 8.762135050684915e-05,
+      "loss": 0.0679,
+      "step": 48140
+    },
+    {
+      "epoch": 3.1501472031403335,
+      "grad_norm": 0.946603536605835,
+      "learning_rate": 8.761529931483487e-05,
+      "loss": 0.0689,
+      "step": 48150
+    },
+    {
+      "epoch": 3.1508014393195944,
+      "grad_norm": 0.8142273426055908,
+      "learning_rate": 8.760924685320557e-05,
+      "loss": 0.0722,
+      "step": 48160
+    },
+    {
+      "epoch": 3.151455675498855,
+      "grad_norm": 0.7121886610984802,
+      "learning_rate": 8.760319312216557e-05,
+      "loss": 0.0683,
+      "step": 48170
+    },
+    {
+      "epoch": 3.152109911678116,
+      "grad_norm": 0.9496175646781921,
+      "learning_rate": 8.759713812191917e-05,
+      "loss": 0.066,
+      "step": 48180
+    },
+    {
+      "epoch": 3.1527641478573765,
+      "grad_norm": 0.7960578799247742,
+      "learning_rate": 8.759108185267078e-05,
+      "loss": 0.0642,
+      "step": 48190
+    },
+    {
+      "epoch": 3.1534183840366374,
+      "grad_norm": 1.0279724597930908,
+      "learning_rate": 8.758502431462476e-05,
+      "loss": 0.0795,
+      "step": 48200
+    },
+    {
+      "epoch": 3.154072620215898,
+      "grad_norm": 0.8479052186012268,
+      "learning_rate": 8.757896550798562e-05,
+      "loss": 0.0718,
+      "step": 48210
+    },
+    {
+      "epoch": 3.1547268563951585,
+      "grad_norm": 0.7294306755065918,
+      "learning_rate": 8.757290543295784e-05,
+      "loss": 0.0758,
+      "step": 48220
+    },
+    {
+      "epoch": 3.1553810925744195,
+      "grad_norm": 0.7830544710159302,
+      "learning_rate": 8.756684408974596e-05,
+      "loss": 0.0777,
+      "step": 48230
+    },
+    {
+      "epoch": 3.15603532875368,
+      "grad_norm": 0.8489717841148376,
+      "learning_rate": 8.756078147855455e-05,
+      "loss": 0.0847,
+      "step": 48240
+    },
+    {
+      "epoch": 3.156689564932941,
+      "grad_norm": 0.8141524791717529,
+      "learning_rate": 8.755471759958828e-05,
+      "loss": 0.074,
+      "step": 48250
+    },
+    {
+      "epoch": 3.1573438011122015,
+      "grad_norm": 1.0648870468139648,
+      "learning_rate": 8.754865245305179e-05,
+      "loss": 0.085,
+      "step": 48260
+    },
+    {
+      "epoch": 3.1579980372914624,
+      "grad_norm": 0.9387083649635315,
+      "learning_rate": 8.754258603914982e-05,
+      "loss": 0.0737,
+      "step": 48270
+    },
+    {
+      "epoch": 3.158652273470723,
+      "grad_norm": 0.906022846698761,
+      "learning_rate": 8.75365183580871e-05,
+      "loss": 0.0797,
+      "step": 48280
+    },
+    {
+      "epoch": 3.1593065096499835,
+      "grad_norm": 0.9713001847267151,
+      "learning_rate": 8.753044941006846e-05,
+      "loss": 0.0766,
+      "step": 48290
+    },
+    {
+      "epoch": 3.1599607458292445,
+      "grad_norm": 1.0321050882339478,
+      "learning_rate": 8.75243791952987e-05,
+      "loss": 0.0814,
+      "step": 48300
+    },
+    {
+      "epoch": 3.160614982008505,
+      "grad_norm": 0.7994502186775208,
+      "learning_rate": 8.751830771398272e-05,
+      "loss": 0.0746,
+      "step": 48310
+    },
+    {
+      "epoch": 3.161269218187766,
+      "grad_norm": 1.0904173851013184,
+      "learning_rate": 8.75122349663255e-05,
+      "loss": 0.0742,
+      "step": 48320
+    },
+    {
+      "epoch": 3.1619234543670265,
+      "grad_norm": 1.043516755104065,
+      "learning_rate": 8.750616095253194e-05,
+      "loss": 0.085,
+      "step": 48330
+    },
+    {
+      "epoch": 3.162577690546287,
+      "grad_norm": 0.9570246934890747,
+      "learning_rate": 8.750008567280709e-05,
+      "loss": 0.0795,
+      "step": 48340
+    },
+    {
+      "epoch": 3.163231926725548,
+      "grad_norm": 0.9861446022987366,
+      "learning_rate": 8.749400912735602e-05,
+      "loss": 0.0799,
+      "step": 48350
+    },
+    {
+      "epoch": 3.1638861629048085,
+      "grad_norm": 1.004128336906433,
+      "learning_rate": 8.748793131638379e-05,
+      "loss": 0.0798,
+      "step": 48360
+    },
+    {
+      "epoch": 3.1645403990840695,
+      "grad_norm": 0.8075050115585327,
+      "learning_rate": 8.748185224009558e-05,
+      "loss": 0.0796,
+      "step": 48370
+    },
+    {
+      "epoch": 3.16519463526333,
+      "grad_norm": 0.7377058863639832,
+      "learning_rate": 8.747577189869653e-05,
+      "loss": 0.0672,
+      "step": 48380
+    },
+    {
+      "epoch": 3.165848871442591,
+      "grad_norm": 0.8605074882507324,
+      "learning_rate": 8.746969029239192e-05,
+      "loss": 0.071,
+      "step": 48390
+    },
+    {
+      "epoch": 3.1665031076218515,
+      "grad_norm": 0.8945094347000122,
+      "learning_rate": 8.746360742138698e-05,
+      "loss": 0.0756,
+      "step": 48400
+    },
+    {
+      "epoch": 3.167157343801112,
+      "grad_norm": 0.9711698293685913,
+      "learning_rate": 8.745752328588703e-05,
+      "loss": 0.0706,
+      "step": 48410
+    },
+    {
+      "epoch": 3.167811579980373,
+      "grad_norm": 1.2028074264526367,
+      "learning_rate": 8.745143788609744e-05,
+      "loss": 0.0807,
+      "step": 48420
+    },
+    {
+      "epoch": 3.1684658161596335,
+      "grad_norm": 0.8748884201049805,
+      "learning_rate": 8.744535122222361e-05,
+      "loss": 0.0808,
+      "step": 48430
+    },
+    {
+      "epoch": 3.1691200523388945,
+      "grad_norm": 0.8327896595001221,
+      "learning_rate": 8.743926329447097e-05,
+      "loss": 0.0734,
+      "step": 48440
+    },
+    {
+      "epoch": 3.169774288518155,
+      "grad_norm": 0.7776886224746704,
+      "learning_rate": 8.743317410304501e-05,
+      "loss": 0.0674,
+      "step": 48450
+    },
+    {
+      "epoch": 3.1704285246974155,
+      "grad_norm": 0.8404861688613892,
+      "learning_rate": 8.742708364815125e-05,
+      "loss": 0.0654,
+      "step": 48460
+    },
+    {
+      "epoch": 3.1710827608766765,
+      "grad_norm": 0.9241907596588135,
+      "learning_rate": 8.742099192999525e-05,
+      "loss": 0.075,
+      "step": 48470
+    },
+    {
+      "epoch": 3.171736997055937,
+      "grad_norm": 0.7482475638389587,
+      "learning_rate": 8.741489894878264e-05,
+      "loss": 0.0702,
+      "step": 48480
+    },
+    {
+      "epoch": 3.172391233235198,
+      "grad_norm": 0.9275398254394531,
+      "learning_rate": 8.740880470471907e-05,
+      "loss": 0.072,
+      "step": 48490
+    },
+    {
+      "epoch": 3.1730454694144585,
+      "grad_norm": 0.873641848564148,
+      "learning_rate": 8.740270919801023e-05,
+      "loss": 0.0783,
+      "step": 48500
+    },
+    {
+      "epoch": 3.1736997055937195,
+      "grad_norm": 0.8717873692512512,
+      "learning_rate": 8.739661242886186e-05,
+      "loss": 0.0775,
+      "step": 48510
+    },
+    {
+      "epoch": 3.17435394177298,
+      "grad_norm": 0.8635610938072205,
+      "learning_rate": 8.739051439747973e-05,
+      "loss": 0.0736,
+      "step": 48520
+    },
+    {
+      "epoch": 3.1750081779522406,
+      "grad_norm": 1.115180253982544,
+      "learning_rate": 8.73844151040697e-05,
+      "loss": 0.0772,
+      "step": 48530
+    },
+    {
+      "epoch": 3.1756624141315015,
+      "grad_norm": 0.6833996772766113,
+      "learning_rate": 8.737831454883761e-05,
+      "loss": 0.0643,
+      "step": 48540
+    },
+    {
+      "epoch": 3.176316650310762,
+      "grad_norm": 0.8258626461029053,
+      "learning_rate": 8.737221273198939e-05,
+      "loss": 0.0809,
+      "step": 48550
+    },
+    {
+      "epoch": 3.176970886490023,
+      "grad_norm": 0.9197008609771729,
+      "learning_rate": 8.736610965373095e-05,
+      "loss": 0.0705,
+      "step": 48560
+    },
+    {
+      "epoch": 3.1776251226692835,
+      "grad_norm": 0.9148569703102112,
+      "learning_rate": 8.736000531426833e-05,
+      "loss": 0.0733,
+      "step": 48570
+    },
+    {
+      "epoch": 3.1782793588485445,
+      "grad_norm": 0.9025515913963318,
+      "learning_rate": 8.735389971380755e-05,
+      "loss": 0.0835,
+      "step": 48580
+    },
+    {
+      "epoch": 3.178933595027805,
+      "grad_norm": 0.9588202238082886,
+      "learning_rate": 8.734779285255469e-05,
+      "loss": 0.0648,
+      "step": 48590
+    },
+    {
+      "epoch": 3.1795878312070656,
+      "grad_norm": 0.7988367676734924,
+      "learning_rate": 8.734168473071587e-05,
+      "loss": 0.083,
+      "step": 48600
+    },
+    {
+      "epoch": 3.1802420673863265,
+      "grad_norm": 0.7926768064498901,
+      "learning_rate": 8.733557534849726e-05,
+      "loss": 0.0718,
+      "step": 48610
+    },
+    {
+      "epoch": 3.180896303565587,
+      "grad_norm": 1.0557847023010254,
+      "learning_rate": 8.732946470610506e-05,
+      "loss": 0.0718,
+      "step": 48620
+    },
+    {
+      "epoch": 3.181550539744848,
+      "grad_norm": 0.996004045009613,
+      "learning_rate": 8.732335280374555e-05,
+      "loss": 0.081,
+      "step": 48630
+    },
+    {
+      "epoch": 3.1822047759241086,
+      "grad_norm": 0.9401144981384277,
+      "learning_rate": 8.731723964162498e-05,
+      "loss": 0.0697,
+      "step": 48640
+    },
+    {
+      "epoch": 3.1828590121033695,
+      "grad_norm": 0.7425965666770935,
+      "learning_rate": 8.731112521994969e-05,
+      "loss": 0.0752,
+      "step": 48650
+    },
+    {
+      "epoch": 3.18351324828263,
+      "grad_norm": 0.9412823915481567,
+      "learning_rate": 8.730500953892609e-05,
+      "loss": 0.07,
+      "step": 48660
+    },
+    {
+      "epoch": 3.1841674844618906,
+      "grad_norm": 0.7894400358200073,
+      "learning_rate": 8.729889259876057e-05,
+      "loss": 0.0686,
+      "step": 48670
+    },
+    {
+      "epoch": 3.1848217206411515,
+      "grad_norm": 0.8867034912109375,
+      "learning_rate": 8.729277439965962e-05,
+      "loss": 0.081,
+      "step": 48680
+    },
+    {
+      "epoch": 3.185475956820412,
+      "grad_norm": 0.8769906163215637,
+      "learning_rate": 8.728665494182971e-05,
+      "loss": 0.0688,
+      "step": 48690
+    },
+    {
+      "epoch": 3.186130192999673,
+      "grad_norm": 0.8141859173774719,
+      "learning_rate": 8.728053422547743e-05,
+      "loss": 0.0803,
+      "step": 48700
+    },
+    {
+      "epoch": 3.1867844291789336,
+      "grad_norm": 0.8036234974861145,
+      "learning_rate": 8.727441225080934e-05,
+      "loss": 0.0813,
+      "step": 48710
+    },
+    {
+      "epoch": 3.1874386653581945,
+      "grad_norm": 0.7681657075881958,
+      "learning_rate": 8.726828901803207e-05,
+      "loss": 0.0797,
+      "step": 48720
+    },
+    {
+      "epoch": 3.188092901537455,
+      "grad_norm": 0.9178977608680725,
+      "learning_rate": 8.726216452735232e-05,
+      "loss": 0.0743,
+      "step": 48730
+    },
+    {
+      "epoch": 3.1887471377167156,
+      "grad_norm": 0.8207674026489258,
+      "learning_rate": 8.72560387789768e-05,
+      "loss": 0.0692,
+      "step": 48740
+    },
+    {
+      "epoch": 3.1894013738959766,
+      "grad_norm": 0.7519068121910095,
+      "learning_rate": 8.724991177311224e-05,
+      "loss": 0.0789,
+      "step": 48750
+    },
+    {
+      "epoch": 3.190055610075237,
+      "grad_norm": 0.8407567143440247,
+      "learning_rate": 8.724378350996549e-05,
+      "loss": 0.0686,
+      "step": 48760
+    },
+    {
+      "epoch": 3.190709846254498,
+      "grad_norm": 0.6936128735542297,
+      "learning_rate": 8.723765398974335e-05,
+      "loss": 0.076,
+      "step": 48770
+    },
+    {
+      "epoch": 3.1913640824337586,
+      "grad_norm": 0.7411690354347229,
+      "learning_rate": 8.723152321265275e-05,
+      "loss": 0.077,
+      "step": 48780
+    },
+    {
+      "epoch": 3.192018318613019,
+      "grad_norm": 0.9644950032234192,
+      "learning_rate": 8.722539117890058e-05,
+      "loss": 0.0803,
+      "step": 48790
+    },
+    {
+      "epoch": 3.19267255479228,
+      "grad_norm": 0.9315986633300781,
+      "learning_rate": 8.721925788869383e-05,
+      "loss": 0.0792,
+      "step": 48800
+    },
+    {
+      "epoch": 3.1933267909715406,
+      "grad_norm": 0.8755950331687927,
+      "learning_rate": 8.721312334223952e-05,
+      "loss": 0.0743,
+      "step": 48810
+    },
+    {
+      "epoch": 3.1939810271508016,
+      "grad_norm": 0.7426033616065979,
+      "learning_rate": 8.720698753974473e-05,
+      "loss": 0.0711,
+      "step": 48820
+    },
+    {
+      "epoch": 3.194635263330062,
+      "grad_norm": 0.8731258511543274,
+      "learning_rate": 8.720085048141649e-05,
+      "loss": 0.0696,
+      "step": 48830
+    },
+    {
+      "epoch": 3.195289499509323,
+      "grad_norm": 0.9991781711578369,
+      "learning_rate": 8.7194712167462e-05,
+      "loss": 0.0847,
+      "step": 48840
+    },
+    {
+      "epoch": 3.1959437356885836,
+      "grad_norm": 0.78664630651474,
+      "learning_rate": 8.718857259808843e-05,
+      "loss": 0.0788,
+      "step": 48850
+    },
+    {
+      "epoch": 3.196597971867844,
+      "grad_norm": 0.8035576939582825,
+      "learning_rate": 8.7182431773503e-05,
+      "loss": 0.0652,
+      "step": 48860
+    },
+    {
+      "epoch": 3.197252208047105,
+      "grad_norm": 0.9748409390449524,
+      "learning_rate": 8.717628969391298e-05,
+      "loss": 0.0878,
+      "step": 48870
+    },
+    {
+      "epoch": 3.1979064442263656,
+      "grad_norm": 0.9272085428237915,
+      "learning_rate": 8.717014635952569e-05,
+      "loss": 0.0736,
+      "step": 48880
+    },
+    {
+      "epoch": 3.1985606804056266,
+      "grad_norm": 0.7096579670906067,
+      "learning_rate": 8.716400177054849e-05,
+      "loss": 0.0637,
+      "step": 48890
+    },
+    {
+      "epoch": 3.199214916584887,
+      "grad_norm": 0.9682911038398743,
+      "learning_rate": 8.715785592718875e-05,
+      "loss": 0.0681,
+      "step": 48900
+    },
+    {
+      "epoch": 3.1998691527641476,
+      "grad_norm": 0.7501932382583618,
+      "learning_rate": 8.715170882965391e-05,
+      "loss": 0.0706,
+      "step": 48910
+    },
+    {
+      "epoch": 3.2005233889434086,
+      "grad_norm": 0.8990116715431213,
+      "learning_rate": 8.714556047815147e-05,
+      "loss": 0.0719,
+      "step": 48920
+    },
+    {
+      "epoch": 3.201177625122669,
+      "grad_norm": 0.8213568925857544,
+      "learning_rate": 8.713941087288897e-05,
+      "loss": 0.0646,
+      "step": 48930
+    },
+    {
+      "epoch": 3.20183186130193,
+      "grad_norm": 0.9337528944015503,
+      "learning_rate": 8.713326001407393e-05,
+      "loss": 0.0844,
+      "step": 48940
+    },
+    {
+      "epoch": 3.2024860974811906,
+      "grad_norm": 0.7658262848854065,
+      "learning_rate": 8.712710790191399e-05,
+      "loss": 0.0723,
+      "step": 48950
+    },
+    {
+      "epoch": 3.2031403336604516,
+      "grad_norm": 0.803835391998291,
+      "learning_rate": 8.712095453661677e-05,
+      "loss": 0.0714,
+      "step": 48960
+    },
+    {
+      "epoch": 3.203794569839712,
+      "grad_norm": 0.8985826373100281,
+      "learning_rate": 8.711479991839e-05,
+      "loss": 0.0761,
+      "step": 48970
+    },
+    {
+      "epoch": 3.2044488060189726,
+      "grad_norm": 0.8591262698173523,
+      "learning_rate": 8.710864404744139e-05,
+      "loss": 0.0807,
+      "step": 48980
+    },
+    {
+      "epoch": 3.2051030421982336,
+      "grad_norm": 0.8583003878593445,
+      "learning_rate": 8.710248692397872e-05,
+      "loss": 0.0725,
+      "step": 48990
+    },
+    {
+      "epoch": 3.205757278377494,
+      "grad_norm": 0.8409457802772522,
+      "learning_rate": 8.709632854820982e-05,
+      "loss": 0.0805,
+      "step": 49000
+    },
+    {
+      "epoch": 3.206411514556755,
+      "grad_norm": 0.8898982405662537,
+      "learning_rate": 8.709016892034252e-05,
+      "loss": 0.0737,
+      "step": 49010
+    },
+    {
+      "epoch": 3.2070657507360156,
+      "grad_norm": 0.8796902298927307,
+      "learning_rate": 8.708400804058478e-05,
+      "loss": 0.0786,
+      "step": 49020
+    },
+    {
+      "epoch": 3.2077199869152766,
+      "grad_norm": 0.9222922325134277,
+      "learning_rate": 8.70778459091445e-05,
+      "loss": 0.0753,
+      "step": 49030
+    },
+    {
+      "epoch": 3.208374223094537,
+      "grad_norm": 0.9596317410469055,
+      "learning_rate": 8.707168252622966e-05,
+      "loss": 0.0713,
+      "step": 49040
+    },
+    {
+      "epoch": 3.2090284592737977,
+      "grad_norm": 0.9624839425086975,
+      "learning_rate": 8.706551789204833e-05,
+      "loss": 0.0772,
+      "step": 49050
+    },
+    {
+      "epoch": 3.2096826954530586,
+      "grad_norm": 0.9401121139526367,
+      "learning_rate": 8.705935200680854e-05,
+      "loss": 0.0724,
+      "step": 49060
+    },
+    {
+      "epoch": 3.210336931632319,
+      "grad_norm": 0.7766807675361633,
+      "learning_rate": 8.705318487071846e-05,
+      "loss": 0.0691,
+      "step": 49070
+    },
+    {
+      "epoch": 3.21099116781158,
+      "grad_norm": 0.8511127233505249,
+      "learning_rate": 8.704701648398621e-05,
+      "loss": 0.0736,
+      "step": 49080
+    },
+    {
+      "epoch": 3.2116454039908406,
+      "grad_norm": 0.9326490163803101,
+      "learning_rate": 8.704084684681998e-05,
+      "loss": 0.0704,
+      "step": 49090
+    },
+    {
+      "epoch": 3.2122996401701016,
+      "grad_norm": 1.0042059421539307,
+      "learning_rate": 8.703467595942803e-05,
+      "loss": 0.0755,
+      "step": 49100
+    },
+    {
+      "epoch": 3.212953876349362,
+      "grad_norm": 0.7557867169380188,
+      "learning_rate": 8.702850382201863e-05,
+      "loss": 0.0707,
+      "step": 49110
+    },
+    {
+      "epoch": 3.2136081125286227,
+      "grad_norm": 1.0037438869476318,
+      "learning_rate": 8.702233043480015e-05,
+      "loss": 0.0776,
+      "step": 49120
+    },
+    {
+      "epoch": 3.2142623487078836,
+      "grad_norm": 0.7470929026603699,
+      "learning_rate": 8.701615579798089e-05,
+      "loss": 0.0707,
+      "step": 49130
+    },
+    {
+      "epoch": 3.214916584887144,
+      "grad_norm": 0.8272511959075928,
+      "learning_rate": 8.70099799117693e-05,
+      "loss": 0.0767,
+      "step": 49140
+    },
+    {
+      "epoch": 3.215570821066405,
+      "grad_norm": 1.0610016584396362,
+      "learning_rate": 8.700380277637384e-05,
+      "loss": 0.0749,
+      "step": 49150
+    },
+    {
+      "epoch": 3.2162250572456657,
+      "grad_norm": 0.7501885890960693,
+      "learning_rate": 8.699762439200298e-05,
+      "loss": 0.0777,
+      "step": 49160
+    },
+    {
+      "epoch": 3.2168792934249266,
+      "grad_norm": 0.8301616311073303,
+      "learning_rate": 8.699144475886526e-05,
+      "loss": 0.0798,
+      "step": 49170
+    },
+    {
+      "epoch": 3.217533529604187,
+      "grad_norm": 0.8588626384735107,
+      "learning_rate": 8.698526387716928e-05,
+      "loss": 0.0745,
+      "step": 49180
+    },
+    {
+      "epoch": 3.2181877657834477,
+      "grad_norm": 1.1801152229309082,
+      "learning_rate": 8.697908174712363e-05,
+      "loss": 0.0882,
+      "step": 49190
+    },
+    {
+      "epoch": 3.2188420019627086,
+      "grad_norm": 0.9209977984428406,
+      "learning_rate": 8.6972898368937e-05,
+      "loss": 0.0757,
+      "step": 49200
+    },
+    {
+      "epoch": 3.219496238141969,
+      "grad_norm": 1.0483894348144531,
+      "learning_rate": 8.696671374281808e-05,
+      "loss": 0.0722,
+      "step": 49210
+    },
+    {
+      "epoch": 3.22015047432123,
+      "grad_norm": 0.9337593913078308,
+      "learning_rate": 8.696052786897563e-05,
+      "loss": 0.075,
+      "step": 49220
+    },
+    {
+      "epoch": 3.2208047105004907,
+      "grad_norm": 0.8857582807540894,
+      "learning_rate": 8.695434074761843e-05,
+      "loss": 0.0744,
+      "step": 49230
+    },
+    {
+      "epoch": 3.221458946679751,
+      "grad_norm": 0.8099195957183838,
+      "learning_rate": 8.69481523789553e-05,
+      "loss": 0.072,
+      "step": 49240
+    },
+    {
+      "epoch": 3.222113182859012,
+      "grad_norm": 0.8427746295928955,
+      "learning_rate": 8.694196276319514e-05,
+      "loss": 0.0747,
+      "step": 49250
+    },
+    {
+      "epoch": 3.2227674190382727,
+      "grad_norm": 0.8717663884162903,
+      "learning_rate": 8.693577190054685e-05,
+      "loss": 0.0792,
+      "step": 49260
+    },
+    {
+      "epoch": 3.2234216552175337,
+      "grad_norm": 0.8552377820014954,
+      "learning_rate": 8.69295797912194e-05,
+      "loss": 0.0777,
+      "step": 49270
+    },
+    {
+      "epoch": 3.224075891396794,
+      "grad_norm": 0.8015262484550476,
+      "learning_rate": 8.692338643542177e-05,
+      "loss": 0.0846,
+      "step": 49280
+    },
+    {
+      "epoch": 3.224730127576055,
+      "grad_norm": 0.7176341414451599,
+      "learning_rate": 8.691719183336302e-05,
+      "loss": 0.0658,
+      "step": 49290
+    },
+    {
+      "epoch": 3.2253843637553157,
+      "grad_norm": 0.8882294297218323,
+      "learning_rate": 8.69109959852522e-05,
+      "loss": 0.0737,
+      "step": 49300
+    },
+    {
+      "epoch": 3.226038599934576,
+      "grad_norm": 0.9010905623435974,
+      "learning_rate": 8.69047988912985e-05,
+      "loss": 0.0838,
+      "step": 49310
+    },
+    {
+      "epoch": 3.226692836113837,
+      "grad_norm": 0.8391245007514954,
+      "learning_rate": 8.689860055171104e-05,
+      "loss": 0.0727,
+      "step": 49320
+    },
+    {
+      "epoch": 3.2273470722930977,
+      "grad_norm": 1.213536024093628,
+      "learning_rate": 8.689240096669903e-05,
+      "loss": 0.0761,
+      "step": 49330
+    },
+    {
+      "epoch": 3.2280013084723587,
+      "grad_norm": 0.9631514549255371,
+      "learning_rate": 8.688620013647175e-05,
+      "loss": 0.0703,
+      "step": 49340
+    },
+    {
+      "epoch": 3.228655544651619,
+      "grad_norm": 0.8843837380409241,
+      "learning_rate": 8.687999806123847e-05,
+      "loss": 0.0657,
+      "step": 49350
+    },
+    {
+      "epoch": 3.2293097808308797,
+      "grad_norm": 0.8727587461471558,
+      "learning_rate": 8.687379474120852e-05,
+      "loss": 0.085,
+      "step": 49360
+    },
+    {
+      "epoch": 3.2299640170101407,
+      "grad_norm": 0.8236467838287354,
+      "learning_rate": 8.686759017659132e-05,
+      "loss": 0.0652,
+      "step": 49370
+    },
+    {
+      "epoch": 3.230618253189401,
+      "grad_norm": 0.876015841960907,
+      "learning_rate": 8.686138436759623e-05,
+      "loss": 0.0741,
+      "step": 49380
+    },
+    {
+      "epoch": 3.231272489368662,
+      "grad_norm": 0.8100466132164001,
+      "learning_rate": 8.685517731443278e-05,
+      "loss": 0.0748,
+      "step": 49390
+    },
+    {
+      "epoch": 3.2319267255479227,
+      "grad_norm": 0.8338639140129089,
+      "learning_rate": 8.684896901731041e-05,
+      "loss": 0.0731,
+      "step": 49400
+    },
+    {
+      "epoch": 3.2325809617271837,
+      "grad_norm": 0.8085967302322388,
+      "learning_rate": 8.684275947643872e-05,
+      "loss": 0.075,
+      "step": 49410
+    },
+    {
+      "epoch": 3.233235197906444,
+      "grad_norm": 0.8683192729949951,
+      "learning_rate": 8.683654869202726e-05,
+      "loss": 0.072,
+      "step": 49420
+    },
+    {
+      "epoch": 3.2338894340857047,
+      "grad_norm": 1.1093887090682983,
+      "learning_rate": 8.683033666428568e-05,
+      "loss": 0.0715,
+      "step": 49430
+    },
+    {
+      "epoch": 3.2345436702649657,
+      "grad_norm": 1.2019805908203125,
+      "learning_rate": 8.682412339342363e-05,
+      "loss": 0.0825,
+      "step": 49440
+    },
+    {
+      "epoch": 3.2351979064442262,
+      "grad_norm": 0.8940714001655579,
+      "learning_rate": 8.681790887965087e-05,
+      "loss": 0.0761,
+      "step": 49450
+    },
+    {
+      "epoch": 3.235852142623487,
+      "grad_norm": 0.9425767660140991,
+      "learning_rate": 8.681169312317709e-05,
+      "loss": 0.0733,
+      "step": 49460
+    },
+    {
+      "epoch": 3.2365063788027477,
+      "grad_norm": 0.9423227310180664,
+      "learning_rate": 8.680547612421215e-05,
+      "loss": 0.0888,
+      "step": 49470
+    },
+    {
+      "epoch": 3.2371606149820087,
+      "grad_norm": 1.0191290378570557,
+      "learning_rate": 8.679925788296586e-05,
+      "loss": 0.0749,
+      "step": 49480
+    },
+    {
+      "epoch": 3.237814851161269,
+      "grad_norm": 0.8777630925178528,
+      "learning_rate": 8.67930383996481e-05,
+      "loss": 0.083,
+      "step": 49490
+    },
+    {
+      "epoch": 3.2384690873405297,
+      "grad_norm": 1.0613380670547485,
+      "learning_rate": 8.678681767446882e-05,
+      "loss": 0.0801,
+      "step": 49500
+    },
+    {
+      "epoch": 3.2391233235197907,
+      "grad_norm": 0.8227213621139526,
+      "learning_rate": 8.678059570763794e-05,
+      "loss": 0.0692,
+      "step": 49510
+    },
+    {
+      "epoch": 3.2397775596990512,
+      "grad_norm": 0.8650736212730408,
+      "learning_rate": 8.677437249936552e-05,
+      "loss": 0.0787,
+      "step": 49520
+    },
+    {
+      "epoch": 3.240431795878312,
+      "grad_norm": 0.8299550414085388,
+      "learning_rate": 8.676814804986158e-05,
+      "loss": 0.0721,
+      "step": 49530
+    },
+    {
+      "epoch": 3.2410860320575727,
+      "grad_norm": 1.0536143779754639,
+      "learning_rate": 8.67619223593362e-05,
+      "loss": 0.074,
+      "step": 49540
+    },
+    {
+      "epoch": 3.2417402682368337,
+      "grad_norm": 0.9724963903427124,
+      "learning_rate": 8.675569542799953e-05,
+      "loss": 0.0775,
+      "step": 49550
+    },
+    {
+      "epoch": 3.2423945044160942,
+      "grad_norm": 1.0580730438232422,
+      "learning_rate": 8.674946725606176e-05,
+      "loss": 0.0856,
+      "step": 49560
+    },
+    {
+      "epoch": 3.2430487405953548,
+      "grad_norm": 0.9586568474769592,
+      "learning_rate": 8.674323784373308e-05,
+      "loss": 0.0783,
+      "step": 49570
+    },
+    {
+      "epoch": 3.2437029767746157,
+      "grad_norm": 0.8590267300605774,
+      "learning_rate": 8.673700719122375e-05,
+      "loss": 0.0748,
+      "step": 49580
+    },
+    {
+      "epoch": 3.2443572129538762,
+      "grad_norm": 0.9745254516601562,
+      "learning_rate": 8.673077529874409e-05,
+      "loss": 0.0857,
+      "step": 49590
+    },
+    {
+      "epoch": 3.245011449133137,
+      "grad_norm": 0.7989119291305542,
+      "learning_rate": 8.672454216650445e-05,
+      "loss": 0.0696,
+      "step": 49600
+    },
+    {
+      "epoch": 3.2456656853123977,
+      "grad_norm": 0.9528154730796814,
+      "learning_rate": 8.671830779471518e-05,
+      "loss": 0.082,
+      "step": 49610
+    },
+    {
+      "epoch": 3.2463199214916587,
+      "grad_norm": 0.8685365319252014,
+      "learning_rate": 8.671207218358672e-05,
+      "loss": 0.0709,
+      "step": 49620
+    },
+    {
+      "epoch": 3.2469741576709192,
+      "grad_norm": 0.7714129090309143,
+      "learning_rate": 8.670583533332957e-05,
+      "loss": 0.078,
+      "step": 49630
+    },
+    {
+      "epoch": 3.2476283938501798,
+      "grad_norm": 0.865516722202301,
+      "learning_rate": 8.669959724415419e-05,
+      "loss": 0.0676,
+      "step": 49640
+    },
+    {
+      "epoch": 3.2482826300294407,
+      "grad_norm": 0.8643177151679993,
+      "learning_rate": 8.669335791627116e-05,
+      "loss": 0.0699,
+      "step": 49650
+    },
+    {
+      "epoch": 3.2489368662087013,
+      "grad_norm": 0.9267588257789612,
+      "learning_rate": 8.668711734989105e-05,
+      "loss": 0.0746,
+      "step": 49660
+    },
+    {
+      "epoch": 3.2495911023879622,
+      "grad_norm": 1.2003788948059082,
+      "learning_rate": 8.668087554522455e-05,
+      "loss": 0.0673,
+      "step": 49670
+    },
+    {
+      "epoch": 3.2502453385672228,
+      "grad_norm": 0.869156002998352,
+      "learning_rate": 8.667463250248228e-05,
+      "loss": 0.0709,
+      "step": 49680
+    },
+    {
+      "epoch": 3.2508995747464837,
+      "grad_norm": 1.0042074918746948,
+      "learning_rate": 8.666838822187498e-05,
+      "loss": 0.0782,
+      "step": 49690
+    },
+    {
+      "epoch": 3.2515538109257442,
+      "grad_norm": 0.9066942930221558,
+      "learning_rate": 8.666214270361342e-05,
+      "loss": 0.0717,
+      "step": 49700
+    },
+    {
+      "epoch": 3.2522080471050048,
+      "grad_norm": 0.8262027502059937,
+      "learning_rate": 8.665589594790838e-05,
+      "loss": 0.0818,
+      "step": 49710
+    },
+    {
+      "epoch": 3.2528622832842657,
+      "grad_norm": 0.7925301194190979,
+      "learning_rate": 8.664964795497073e-05,
+      "loss": 0.0704,
+      "step": 49720
+    },
+    {
+      "epoch": 3.2535165194635263,
+      "grad_norm": 0.8457410335540771,
+      "learning_rate": 8.664339872501133e-05,
+      "loss": 0.0743,
+      "step": 49730
+    },
+    {
+      "epoch": 3.2541707556427872,
+      "grad_norm": 1.0192575454711914,
+      "learning_rate": 8.663714825824114e-05,
+      "loss": 0.0792,
+      "step": 49740
+    },
+    {
+      "epoch": 3.2548249918220478,
+      "grad_norm": 0.887332558631897,
+      "learning_rate": 8.66308965548711e-05,
+      "loss": 0.0744,
+      "step": 49750
+    },
+    {
+      "epoch": 3.2554792280013083,
+      "grad_norm": 0.9287620782852173,
+      "learning_rate": 8.662464361511224e-05,
+      "loss": 0.0757,
+      "step": 49760
+    },
+    {
+      "epoch": 3.2561334641805693,
+      "grad_norm": 0.9490354657173157,
+      "learning_rate": 8.661838943917561e-05,
+      "loss": 0.0802,
+      "step": 49770
+    },
+    {
+      "epoch": 3.25678770035983,
+      "grad_norm": 0.7195919752120972,
+      "learning_rate": 8.661213402727229e-05,
+      "loss": 0.0737,
+      "step": 49780
+    },
+    {
+      "epoch": 3.2574419365390908,
+      "grad_norm": 0.7042505145072937,
+      "learning_rate": 8.660587737961344e-05,
+      "loss": 0.0644,
+      "step": 49790
+    },
+    {
+      "epoch": 3.2580961727183513,
+      "grad_norm": 0.7407015562057495,
+      "learning_rate": 8.659961949641023e-05,
+      "loss": 0.067,
+      "step": 49800
+    },
+    {
+      "epoch": 3.258750408897612,
+      "grad_norm": 0.7478853464126587,
+      "learning_rate": 8.659336037787384e-05,
+      "loss": 0.0737,
+      "step": 49810
+    },
+    {
+      "epoch": 3.2594046450768728,
+      "grad_norm": 0.8127663135528564,
+      "learning_rate": 8.658710002421561e-05,
+      "loss": 0.0697,
+      "step": 49820
+    },
+    {
+      "epoch": 3.2600588812561333,
+      "grad_norm": 1.1231528520584106,
+      "learning_rate": 8.658083843564677e-05,
+      "loss": 0.0831,
+      "step": 49830
+    },
+    {
+      "epoch": 3.2607131174353943,
+      "grad_norm": 0.6892096400260925,
+      "learning_rate": 8.657457561237871e-05,
+      "loss": 0.0696,
+      "step": 49840
+    },
+    {
+      "epoch": 3.261367353614655,
+      "grad_norm": 0.8850003480911255,
+      "learning_rate": 8.656831155462281e-05,
+      "loss": 0.0721,
+      "step": 49850
+    },
+    {
+      "epoch": 3.2620215897939158,
+      "grad_norm": 0.7798917889595032,
+      "learning_rate": 8.656204626259048e-05,
+      "loss": 0.0768,
+      "step": 49860
+    },
+    {
+      "epoch": 3.2626758259731763,
+      "grad_norm": 0.934593915939331,
+      "learning_rate": 8.655577973649321e-05,
+      "loss": 0.0803,
+      "step": 49870
+    },
+    {
+      "epoch": 3.263330062152437,
+      "grad_norm": 0.9441463947296143,
+      "learning_rate": 8.65495119765425e-05,
+      "loss": 0.0735,
+      "step": 49880
+    },
+    {
+      "epoch": 3.263984298331698,
+      "grad_norm": 0.6787394881248474,
+      "learning_rate": 8.65432429829499e-05,
+      "loss": 0.0688,
+      "step": 49890
+    },
+    {
+      "epoch": 3.2646385345109583,
+      "grad_norm": 0.8261919021606445,
+      "learning_rate": 8.653697275592702e-05,
+      "loss": 0.0669,
+      "step": 49900
+    },
+    {
+      "epoch": 3.2652927706902193,
+      "grad_norm": 1.0204743146896362,
+      "learning_rate": 8.653070129568548e-05,
+      "loss": 0.0691,
+      "step": 49910
+    },
+    {
+      "epoch": 3.26594700686948,
+      "grad_norm": 1.0907118320465088,
+      "learning_rate": 8.652442860243698e-05,
+      "loss": 0.0808,
+      "step": 49920
+    },
+    {
+      "epoch": 3.2666012430487408,
+      "grad_norm": 1.026333212852478,
+      "learning_rate": 8.651815467639321e-05,
+      "loss": 0.0761,
+      "step": 49930
+    },
+    {
+      "epoch": 3.2672554792280013,
+      "grad_norm": 0.9486099481582642,
+      "learning_rate": 8.651187951776593e-05,
+      "loss": 0.0708,
+      "step": 49940
+    },
+    {
+      "epoch": 3.267909715407262,
+      "grad_norm": 0.9680077433586121,
+      "learning_rate": 8.6505603126767e-05,
+      "loss": 0.0742,
+      "step": 49950
+    },
+    {
+      "epoch": 3.268563951586523,
+      "grad_norm": 0.789247989654541,
+      "learning_rate": 8.649932550360821e-05,
+      "loss": 0.0803,
+      "step": 49960
+    },
+    {
+      "epoch": 3.2692181877657833,
+      "grad_norm": 0.725689709186554,
+      "learning_rate": 8.649304664850145e-05,
+      "loss": 0.0762,
+      "step": 49970
+    },
+    {
+      "epoch": 3.2698724239450443,
+      "grad_norm": 0.7912468314170837,
+      "learning_rate": 8.648676656165867e-05,
+      "loss": 0.0754,
+      "step": 49980
+    },
+    {
+      "epoch": 3.270526660124305,
+      "grad_norm": 0.6831502318382263,
+      "learning_rate": 8.648048524329182e-05,
+      "loss": 0.0736,
+      "step": 49990
+    },
+    {
+      "epoch": 3.271180896303566,
+      "grad_norm": 0.8818649053573608,
+      "learning_rate": 8.647420269361294e-05,
+      "loss": 0.0841,
+      "step": 50000
+    },
+    {
+      "epoch": 3.2718351324828263,
+      "grad_norm": 0.8740116953849792,
+      "learning_rate": 8.646791891283403e-05,
+      "loss": 0.07,
+      "step": 50010
+    },
+    {
+      "epoch": 3.272489368662087,
+      "grad_norm": 0.8219292759895325,
+      "learning_rate": 8.646163390116723e-05,
+      "loss": 0.0731,
+      "step": 50020
+    },
+    {
+      "epoch": 3.273143604841348,
+      "grad_norm": 0.7342281341552734,
+      "learning_rate": 8.645534765882469e-05,
+      "loss": 0.0729,
+      "step": 50030
+    },
+    {
+      "epoch": 3.2737978410206083,
+      "grad_norm": 0.6889681220054626,
+      "learning_rate": 8.644906018601852e-05,
+      "loss": 0.0697,
+      "step": 50040
+    },
+    {
+      "epoch": 3.2744520771998693,
+      "grad_norm": 0.7851772904396057,
+      "learning_rate": 8.6442771482961e-05,
+      "loss": 0.0777,
+      "step": 50050
+    },
+    {
+      "epoch": 3.27510631337913,
+      "grad_norm": 0.9841508865356445,
+      "learning_rate": 8.643648154986435e-05,
+      "loss": 0.0747,
+      "step": 50060
+    },
+    {
+      "epoch": 3.275760549558391,
+      "grad_norm": 0.8496854305267334,
+      "learning_rate": 8.64301903869409e-05,
+      "loss": 0.0708,
+      "step": 50070
+    },
+    {
+      "epoch": 3.2764147857376513,
+      "grad_norm": 0.8300930261611938,
+      "learning_rate": 8.642389799440298e-05,
+      "loss": 0.0692,
+      "step": 50080
+    },
+    {
+      "epoch": 3.277069021916912,
+      "grad_norm": 0.7653340697288513,
+      "learning_rate": 8.641760437246297e-05,
+      "loss": 0.0751,
+      "step": 50090
+    },
+    {
+      "epoch": 3.277723258096173,
+      "grad_norm": 0.7286269068717957,
+      "learning_rate": 8.641130952133332e-05,
+      "loss": 0.0665,
+      "step": 50100
+    },
+    {
+      "epoch": 3.2783774942754333,
+      "grad_norm": 0.9306923151016235,
+      "learning_rate": 8.64050134412265e-05,
+      "loss": 0.0827,
+      "step": 50110
+    },
+    {
+      "epoch": 3.2790317304546943,
+      "grad_norm": 0.7376532554626465,
+      "learning_rate": 8.639871613235495e-05,
+      "loss": 0.072,
+      "step": 50120
+    },
+    {
+      "epoch": 3.279685966633955,
+      "grad_norm": 0.8583968877792358,
+      "learning_rate": 8.639241759493131e-05,
+      "loss": 0.0756,
+      "step": 50130
+    },
+    {
+      "epoch": 3.280340202813216,
+      "grad_norm": 0.9432209134101868,
+      "learning_rate": 8.638611782916812e-05,
+      "loss": 0.0757,
+      "step": 50140
+    },
+    {
+      "epoch": 3.2809944389924763,
+      "grad_norm": 0.8525569438934326,
+      "learning_rate": 8.637981683527803e-05,
+      "loss": 0.0743,
+      "step": 50150
+    },
+    {
+      "epoch": 3.281648675171737,
+      "grad_norm": 1.0892137289047241,
+      "learning_rate": 8.637351461347371e-05,
+      "loss": 0.0712,
+      "step": 50160
+    },
+    {
+      "epoch": 3.282302911350998,
+      "grad_norm": 0.8420467376708984,
+      "learning_rate": 8.636721116396787e-05,
+      "loss": 0.0643,
+      "step": 50170
+    },
+    {
+      "epoch": 3.2829571475302584,
+      "grad_norm": 0.7865369319915771,
+      "learning_rate": 8.636090648697329e-05,
+      "loss": 0.0713,
+      "step": 50180
+    },
+    {
+      "epoch": 3.2836113837095193,
+      "grad_norm": 0.8081366419792175,
+      "learning_rate": 8.635460058270274e-05,
+      "loss": 0.0859,
+      "step": 50190
+    },
+    {
+      "epoch": 3.28426561988878,
+      "grad_norm": 0.8444221019744873,
+      "learning_rate": 8.63482934513691e-05,
+      "loss": 0.0721,
+      "step": 50200
+    },
+    {
+      "epoch": 3.2849198560680404,
+      "grad_norm": 0.9457107782363892,
+      "learning_rate": 8.634198509318521e-05,
+      "loss": 0.0678,
+      "step": 50210
+    },
+    {
+      "epoch": 3.2855740922473013,
+      "grad_norm": 0.908445417881012,
+      "learning_rate": 8.633567550836403e-05,
+      "loss": 0.0757,
+      "step": 50220
+    },
+    {
+      "epoch": 3.286228328426562,
+      "grad_norm": 0.9582734107971191,
+      "learning_rate": 8.63293646971185e-05,
+      "loss": 0.0725,
+      "step": 50230
+    },
+    {
+      "epoch": 3.286882564605823,
+      "grad_norm": 1.1066676378250122,
+      "learning_rate": 8.632305265966163e-05,
+      "loss": 0.0838,
+      "step": 50240
+    },
+    {
+      "epoch": 3.2875368007850834,
+      "grad_norm": 1.1428422927856445,
+      "learning_rate": 8.631673939620646e-05,
+      "loss": 0.0725,
+      "step": 50250
+    },
+    {
+      "epoch": 3.288191036964344,
+      "grad_norm": 0.8173267841339111,
+      "learning_rate": 8.631042490696612e-05,
+      "loss": 0.0736,
+      "step": 50260
+    },
+    {
+      "epoch": 3.288845273143605,
+      "grad_norm": 0.7522175908088684,
+      "learning_rate": 8.63041091921537e-05,
+      "loss": 0.067,
+      "step": 50270
+    },
+    {
+      "epoch": 3.2894995093228654,
+      "grad_norm": 0.7062971591949463,
+      "learning_rate": 8.629779225198238e-05,
+      "loss": 0.0683,
+      "step": 50280
+    },
+    {
+      "epoch": 3.2901537455021264,
+      "grad_norm": 0.9391506314277649,
+      "learning_rate": 8.629147408666537e-05,
+      "loss": 0.0705,
+      "step": 50290
+    },
+    {
+      "epoch": 3.290807981681387,
+      "grad_norm": 0.7575945854187012,
+      "learning_rate": 8.628515469641593e-05,
+      "loss": 0.0696,
+      "step": 50300
+    },
+    {
+      "epoch": 3.291462217860648,
+      "grad_norm": 0.7844806909561157,
+      "learning_rate": 8.627883408144737e-05,
+      "loss": 0.0746,
+      "step": 50310
+    },
+    {
+      "epoch": 3.2921164540399084,
+      "grad_norm": 0.7494459748268127,
+      "learning_rate": 8.627251224197302e-05,
+      "loss": 0.0704,
+      "step": 50320
+    },
+    {
+      "epoch": 3.292770690219169,
+      "grad_norm": 0.7377359867095947,
+      "learning_rate": 8.626618917820624e-05,
+      "loss": 0.0679,
+      "step": 50330
+    },
+    {
+      "epoch": 3.29342492639843,
+      "grad_norm": 0.7377490401268005,
+      "learning_rate": 8.625986489036048e-05,
+      "loss": 0.076,
+      "step": 50340
+    },
+    {
+      "epoch": 3.2940791625776904,
+      "grad_norm": 0.7678962349891663,
+      "learning_rate": 8.625353937864917e-05,
+      "loss": 0.071,
+      "step": 50350
+    },
+    {
+      "epoch": 3.2947333987569514,
+      "grad_norm": 0.9204226136207581,
+      "learning_rate": 8.624721264328584e-05,
+      "loss": 0.0675,
+      "step": 50360
+    },
+    {
+      "epoch": 3.295387634936212,
+      "grad_norm": 0.8501763343811035,
+      "learning_rate": 8.624088468448401e-05,
+      "loss": 0.0716,
+      "step": 50370
+    },
+    {
+      "epoch": 3.296041871115473,
+      "grad_norm": 1.0551090240478516,
+      "learning_rate": 8.623455550245727e-05,
+      "loss": 0.0702,
+      "step": 50380
+    },
+    {
+      "epoch": 3.2966961072947334,
+      "grad_norm": 0.9518636465072632,
+      "learning_rate": 8.622822509741928e-05,
+      "loss": 0.0756,
+      "step": 50390
+    },
+    {
+      "epoch": 3.297350343473994,
+      "grad_norm": 0.971017062664032,
+      "learning_rate": 8.622189346958365e-05,
+      "loss": 0.069,
+      "step": 50400
+    },
+    {
+      "epoch": 3.298004579653255,
+      "grad_norm": 0.8658537268638611,
+      "learning_rate": 8.621556061916414e-05,
+      "loss": 0.0779,
+      "step": 50410
+    },
+    {
+      "epoch": 3.2986588158325154,
+      "grad_norm": 0.9639972448348999,
+      "learning_rate": 8.620922654637446e-05,
+      "loss": 0.0717,
+      "step": 50420
+    },
+    {
+      "epoch": 3.2993130520117764,
+      "grad_norm": 0.9584673643112183,
+      "learning_rate": 8.620289125142845e-05,
+      "loss": 0.0691,
+      "step": 50430
+    },
+    {
+      "epoch": 3.299967288191037,
+      "grad_norm": 0.7982688546180725,
+      "learning_rate": 8.61965547345399e-05,
+      "loss": 0.0863,
+      "step": 50440
+    },
+    {
+      "epoch": 3.300621524370298,
+      "grad_norm": 1.0012849569320679,
+      "learning_rate": 8.619021699592271e-05,
+      "loss": 0.0813,
+      "step": 50450
+    },
+    {
+      "epoch": 3.3012757605495584,
+      "grad_norm": 0.9669886231422424,
+      "learning_rate": 8.618387803579076e-05,
+      "loss": 0.0789,
+      "step": 50460
+    },
+    {
+      "epoch": 3.301929996728819,
+      "grad_norm": 0.7313616871833801,
+      "learning_rate": 8.617753785435804e-05,
+      "loss": 0.0641,
+      "step": 50470
+    },
+    {
+      "epoch": 3.30258423290808,
+      "grad_norm": 0.7947266697883606,
+      "learning_rate": 8.617119645183856e-05,
+      "loss": 0.0751,
+      "step": 50480
+    },
+    {
+      "epoch": 3.3032384690873404,
+      "grad_norm": 0.8369613289833069,
+      "learning_rate": 8.616485382844631e-05,
+      "loss": 0.0857,
+      "step": 50490
+    },
+    {
+      "epoch": 3.3038927052666014,
+      "grad_norm": 1.0216795206069946,
+      "learning_rate": 8.615850998439542e-05,
+      "loss": 0.0753,
+      "step": 50500
+    },
+    {
+      "epoch": 3.304546941445862,
+      "grad_norm": 0.9088335633277893,
+      "learning_rate": 8.615216491989997e-05,
+      "loss": 0.0789,
+      "step": 50510
+    },
+    {
+      "epoch": 3.305201177625123,
+      "grad_norm": 0.9669744372367859,
+      "learning_rate": 8.614581863517414e-05,
+      "loss": 0.0717,
+      "step": 50520
+    },
+    {
+      "epoch": 3.3058554138043834,
+      "grad_norm": 0.8303861021995544,
+      "learning_rate": 8.613947113043215e-05,
+      "loss": 0.0739,
+      "step": 50530
+    },
+    {
+      "epoch": 3.306509649983644,
+      "grad_norm": 0.8657481074333191,
+      "learning_rate": 8.613312240588822e-05,
+      "loss": 0.0677,
+      "step": 50540
+    },
+    {
+      "epoch": 3.307163886162905,
+      "grad_norm": 0.8606223464012146,
+      "learning_rate": 8.612677246175665e-05,
+      "loss": 0.0825,
+      "step": 50550
+    },
+    {
+      "epoch": 3.3078181223421654,
+      "grad_norm": 0.9812949895858765,
+      "learning_rate": 8.612042129825177e-05,
+      "loss": 0.0715,
+      "step": 50560
+    },
+    {
+      "epoch": 3.3084723585214264,
+      "grad_norm": 0.8385498523712158,
+      "learning_rate": 8.611406891558793e-05,
+      "loss": 0.0658,
+      "step": 50570
+    },
+    {
+      "epoch": 3.309126594700687,
+      "grad_norm": 0.820543110370636,
+      "learning_rate": 8.610771531397957e-05,
+      "loss": 0.068,
+      "step": 50580
+    },
+    {
+      "epoch": 3.309780830879948,
+      "grad_norm": 0.8643449544906616,
+      "learning_rate": 8.61013604936411e-05,
+      "loss": 0.0748,
+      "step": 50590
+    },
+    {
+      "epoch": 3.3104350670592084,
+      "grad_norm": 0.8752486705780029,
+      "learning_rate": 8.609500445478704e-05,
+      "loss": 0.081,
+      "step": 50600
+    },
+    {
+      "epoch": 3.311089303238469,
+      "grad_norm": 0.8542597889900208,
+      "learning_rate": 8.608864719763192e-05,
+      "loss": 0.0713,
+      "step": 50610
+    },
+    {
+      "epoch": 3.31174353941773,
+      "grad_norm": 0.7838973999023438,
+      "learning_rate": 8.608228872239031e-05,
+      "loss": 0.0758,
+      "step": 50620
+    },
+    {
+      "epoch": 3.3123977755969904,
+      "grad_norm": 0.9226830005645752,
+      "learning_rate": 8.607592902927684e-05,
+      "loss": 0.0815,
+      "step": 50630
+    },
+    {
+      "epoch": 3.3130520117762514,
+      "grad_norm": 0.9006392955780029,
+      "learning_rate": 8.606956811850613e-05,
+      "loss": 0.0697,
+      "step": 50640
+    },
+    {
+      "epoch": 3.313706247955512,
+      "grad_norm": 0.9188740253448486,
+      "learning_rate": 8.606320599029292e-05,
+      "loss": 0.0697,
+      "step": 50650
+    },
+    {
+      "epoch": 3.3143604841347725,
+      "grad_norm": 0.8508504033088684,
+      "learning_rate": 8.605684264485192e-05,
+      "loss": 0.0752,
+      "step": 50660
+    },
+    {
+      "epoch": 3.3150147203140334,
+      "grad_norm": 0.8384444713592529,
+      "learning_rate": 8.605047808239791e-05,
+      "loss": 0.0719,
+      "step": 50670
+    },
+    {
+      "epoch": 3.315668956493294,
+      "grad_norm": 0.8773384094238281,
+      "learning_rate": 8.604411230314572e-05,
+      "loss": 0.0824,
+      "step": 50680
+    },
+    {
+      "epoch": 3.316323192672555,
+      "grad_norm": 0.7789053320884705,
+      "learning_rate": 8.603774530731023e-05,
+      "loss": 0.0652,
+      "step": 50690
+    },
+    {
+      "epoch": 3.3169774288518155,
+      "grad_norm": 0.7404478788375854,
+      "learning_rate": 8.60313770951063e-05,
+      "loss": 0.0704,
+      "step": 50700
+    },
+    {
+      "epoch": 3.317631665031076,
+      "grad_norm": 1.0024648904800415,
+      "learning_rate": 8.60250076667489e-05,
+      "loss": 0.0829,
+      "step": 50710
+    },
+    {
+      "epoch": 3.318285901210337,
+      "grad_norm": 0.7322053909301758,
+      "learning_rate": 8.601863702245303e-05,
+      "loss": 0.0716,
+      "step": 50720
+    },
+    {
+      "epoch": 3.3189401373895975,
+      "grad_norm": 1.0306543111801147,
+      "learning_rate": 8.601226516243368e-05,
+      "loss": 0.0719,
+      "step": 50730
+    },
+    {
+      "epoch": 3.3195943735688584,
+      "grad_norm": 0.8243963718414307,
+      "learning_rate": 8.600589208690595e-05,
+      "loss": 0.0726,
+      "step": 50740
+    },
+    {
+      "epoch": 3.320248609748119,
+      "grad_norm": 0.7126163244247437,
+      "learning_rate": 8.599951779608493e-05,
+      "loss": 0.072,
+      "step": 50750
+    },
+    {
+      "epoch": 3.32090284592738,
+      "grad_norm": 0.9243032932281494,
+      "learning_rate": 8.599314229018575e-05,
+      "loss": 0.0765,
+      "step": 50760
+    },
+    {
+      "epoch": 3.3215570821066405,
+      "grad_norm": 0.8897315859794617,
+      "learning_rate": 8.598676556942365e-05,
+      "loss": 0.0771,
+      "step": 50770
+    },
+    {
+      "epoch": 3.322211318285901,
+      "grad_norm": 0.8531565070152283,
+      "learning_rate": 8.598038763401382e-05,
+      "loss": 0.076,
+      "step": 50780
+    },
+    {
+      "epoch": 3.322865554465162,
+      "grad_norm": 0.7867346405982971,
+      "learning_rate": 8.597400848417156e-05,
+      "loss": 0.0847,
+      "step": 50790
+    },
+    {
+      "epoch": 3.3235197906444225,
+      "grad_norm": 0.8421074748039246,
+      "learning_rate": 8.596762812011216e-05,
+      "loss": 0.0792,
+      "step": 50800
+    },
+    {
+      "epoch": 3.3241740268236835,
+      "grad_norm": 0.8326323628425598,
+      "learning_rate": 8.596124654205097e-05,
+      "loss": 0.0737,
+      "step": 50810
+    },
+    {
+      "epoch": 3.324828263002944,
+      "grad_norm": 1.0936020612716675,
+      "learning_rate": 8.595486375020341e-05,
+      "loss": 0.0802,
+      "step": 50820
+    },
+    {
+      "epoch": 3.325482499182205,
+      "grad_norm": 1.0565087795257568,
+      "learning_rate": 8.59484797447849e-05,
+      "loss": 0.0741,
+      "step": 50830
+    },
+    {
+      "epoch": 3.3261367353614655,
+      "grad_norm": 0.8278268575668335,
+      "learning_rate": 8.594209452601092e-05,
+      "loss": 0.0665,
+      "step": 50840
+    },
+    {
+      "epoch": 3.326790971540726,
+      "grad_norm": 0.720306932926178,
+      "learning_rate": 8.593570809409698e-05,
+      "loss": 0.0753,
+      "step": 50850
+    },
+    {
+      "epoch": 3.327445207719987,
+      "grad_norm": 0.9338958859443665,
+      "learning_rate": 8.592932044925866e-05,
+      "loss": 0.0729,
+      "step": 50860
+    },
+    {
+      "epoch": 3.3280994438992475,
+      "grad_norm": 0.8932592868804932,
+      "learning_rate": 8.592293159171155e-05,
+      "loss": 0.0749,
+      "step": 50870
+    },
+    {
+      "epoch": 3.3287536800785085,
+      "grad_norm": 0.7628698945045471,
+      "learning_rate": 8.591654152167128e-05,
+      "loss": 0.0762,
+      "step": 50880
+    },
+    {
+      "epoch": 3.329407916257769,
+      "grad_norm": 0.855076789855957,
+      "learning_rate": 8.591015023935353e-05,
+      "loss": 0.0792,
+      "step": 50890
+    },
+    {
+      "epoch": 3.33006215243703,
+      "grad_norm": 0.8656195998191833,
+      "learning_rate": 8.590375774497406e-05,
+      "loss": 0.0678,
+      "step": 50900
+    },
+    {
+      "epoch": 3.3307163886162905,
+      "grad_norm": 1.1251041889190674,
+      "learning_rate": 8.589736403874858e-05,
+      "loss": 0.0817,
+      "step": 50910
+    },
+    {
+      "epoch": 3.331370624795551,
+      "grad_norm": 0.815820038318634,
+      "learning_rate": 8.589096912089292e-05,
+      "loss": 0.0795,
+      "step": 50920
+    },
+    {
+      "epoch": 3.332024860974812,
+      "grad_norm": 0.8184264302253723,
+      "learning_rate": 8.588457299162293e-05,
+      "loss": 0.0679,
+      "step": 50930
+    },
+    {
+      "epoch": 3.3326790971540725,
+      "grad_norm": 0.6833990216255188,
+      "learning_rate": 8.587817565115449e-05,
+      "loss": 0.063,
+      "step": 50940
+    },
+    {
+      "epoch": 3.3333333333333335,
+      "grad_norm": 0.7902950644493103,
+      "learning_rate": 8.587177709970353e-05,
+      "loss": 0.0683,
+      "step": 50950
+    },
+    {
+      "epoch": 3.333987569512594,
+      "grad_norm": 0.8151212334632874,
+      "learning_rate": 8.586537733748601e-05,
+      "loss": 0.0757,
+      "step": 50960
+    },
+    {
+      "epoch": 3.334641805691855,
+      "grad_norm": 0.8784273266792297,
+      "learning_rate": 8.585897636471796e-05,
+      "loss": 0.071,
+      "step": 50970
+    },
+    {
+      "epoch": 3.3352960418711155,
+      "grad_norm": 0.8391785621643066,
+      "learning_rate": 8.585257418161538e-05,
+      "loss": 0.0744,
+      "step": 50980
+    },
+    {
+      "epoch": 3.335950278050376,
+      "grad_norm": 1.0281047821044922,
+      "learning_rate": 8.584617078839443e-05,
+      "loss": 0.0855,
+      "step": 50990
+    },
+    {
+      "epoch": 3.336604514229637,
+      "grad_norm": 0.838431715965271,
+      "learning_rate": 8.58397661852712e-05,
+      "loss": 0.0777,
+      "step": 51000
+    },
+    {
+      "epoch": 3.3372587504088975,
+      "grad_norm": 0.7739060521125793,
+      "learning_rate": 8.583336037246186e-05,
+      "loss": 0.067,
+      "step": 51010
+    },
+    {
+      "epoch": 3.3379129865881585,
+      "grad_norm": 0.9541481733322144,
+      "learning_rate": 8.582695335018263e-05,
+      "loss": 0.0777,
+      "step": 51020
+    },
+    {
+      "epoch": 3.338567222767419,
+      "grad_norm": 0.9411327242851257,
+      "learning_rate": 8.582054511864977e-05,
+      "loss": 0.0823,
+      "step": 51030
+    },
+    {
+      "epoch": 3.33922145894668,
+      "grad_norm": 0.785683274269104,
+      "learning_rate": 8.581413567807956e-05,
+      "loss": 0.0685,
+      "step": 51040
+    },
+    {
+      "epoch": 3.3398756951259405,
+      "grad_norm": 0.7861221432685852,
+      "learning_rate": 8.580772502868835e-05,
+      "loss": 0.0741,
+      "step": 51050
+    },
+    {
+      "epoch": 3.340529931305201,
+      "grad_norm": 0.9560049176216125,
+      "learning_rate": 8.58013131706925e-05,
+      "loss": 0.0704,
+      "step": 51060
+    },
+    {
+      "epoch": 3.341184167484462,
+      "grad_norm": 0.7720146775245667,
+      "learning_rate": 8.579490010430846e-05,
+      "loss": 0.069,
+      "step": 51070
+    },
+    {
+      "epoch": 3.3418384036637225,
+      "grad_norm": 0.7846336364746094,
+      "learning_rate": 8.578848582975266e-05,
+      "loss": 0.074,
+      "step": 51080
+    },
+    {
+      "epoch": 3.3424926398429835,
+      "grad_norm": 0.9714375734329224,
+      "learning_rate": 8.57820703472416e-05,
+      "loss": 0.0714,
+      "step": 51090
+    },
+    {
+      "epoch": 3.343146876022244,
+      "grad_norm": 0.7882207036018372,
+      "learning_rate": 8.577565365699183e-05,
+      "loss": 0.0806,
+      "step": 51100
+    },
+    {
+      "epoch": 3.3438011122015046,
+      "grad_norm": 0.8583884835243225,
+      "learning_rate": 8.576923575921991e-05,
+      "loss": 0.076,
+      "step": 51110
+    },
+    {
+      "epoch": 3.3444553483807655,
+      "grad_norm": 1.0091041326522827,
+      "learning_rate": 8.576281665414249e-05,
+      "loss": 0.0838,
+      "step": 51120
+    },
+    {
+      "epoch": 3.345109584560026,
+      "grad_norm": 0.9896878004074097,
+      "learning_rate": 8.57563963419762e-05,
+      "loss": 0.0801,
+      "step": 51130
+    },
+    {
+      "epoch": 3.345763820739287,
+      "grad_norm": 0.9059678912162781,
+      "learning_rate": 8.574997482293778e-05,
+      "loss": 0.0756,
+      "step": 51140
+    },
+    {
+      "epoch": 3.3464180569185475,
+      "grad_norm": 0.8360334038734436,
+      "learning_rate": 8.574355209724393e-05,
+      "loss": 0.0782,
+      "step": 51150
+    },
+    {
+      "epoch": 3.347072293097808,
+      "grad_norm": 1.1059879064559937,
+      "learning_rate": 8.573712816511148e-05,
+      "loss": 0.0771,
+      "step": 51160
+    },
+    {
+      "epoch": 3.347726529277069,
+      "grad_norm": 0.9057727456092834,
+      "learning_rate": 8.57307030267572e-05,
+      "loss": 0.0817,
+      "step": 51170
+    },
+    {
+      "epoch": 3.3483807654563296,
+      "grad_norm": 0.7811506390571594,
+      "learning_rate": 8.572427668239802e-05,
+      "loss": 0.0817,
+      "step": 51180
+    },
+    {
+      "epoch": 3.3490350016355905,
+      "grad_norm": 0.927189290523529,
+      "learning_rate": 8.571784913225078e-05,
+      "loss": 0.0707,
+      "step": 51190
+    },
+    {
+      "epoch": 3.349689237814851,
+      "grad_norm": 0.7805935740470886,
+      "learning_rate": 8.571142037653249e-05,
+      "loss": 0.0763,
+      "step": 51200
+    },
+    {
+      "epoch": 3.350343473994112,
+      "grad_norm": 0.9339170455932617,
+      "learning_rate": 8.570499041546007e-05,
+      "loss": 0.0666,
+      "step": 51210
+    },
+    {
+      "epoch": 3.3509977101733726,
+      "grad_norm": 0.8374119997024536,
+      "learning_rate": 8.56985592492506e-05,
+      "loss": 0.0721,
+      "step": 51220
+    },
+    {
+      "epoch": 3.351651946352633,
+      "grad_norm": 1.1060223579406738,
+      "learning_rate": 8.569212687812113e-05,
+      "loss": 0.0828,
+      "step": 51230
+    },
+    {
+      "epoch": 3.352306182531894,
+      "grad_norm": 0.7476975917816162,
+      "learning_rate": 8.568569330228879e-05,
+      "loss": 0.0794,
+      "step": 51240
+    },
+    {
+      "epoch": 3.3529604187111546,
+      "grad_norm": 0.9365630149841309,
+      "learning_rate": 8.56792585219707e-05,
+      "loss": 0.0771,
+      "step": 51250
+    },
+    {
+      "epoch": 3.3536146548904155,
+      "grad_norm": 0.8895155787467957,
+      "learning_rate": 8.567282253738407e-05,
+      "loss": 0.0798,
+      "step": 51260
+    },
+    {
+      "epoch": 3.354268891069676,
+      "grad_norm": 0.9719933271408081,
+      "learning_rate": 8.566638534874612e-05,
+      "loss": 0.0788,
+      "step": 51270
+    },
+    {
+      "epoch": 3.354923127248937,
+      "grad_norm": 0.9939890503883362,
+      "learning_rate": 8.565994695627411e-05,
+      "loss": 0.0669,
+      "step": 51280
+    },
+    {
+      "epoch": 3.3555773634281976,
+      "grad_norm": 0.8980028033256531,
+      "learning_rate": 8.565350736018539e-05,
+      "loss": 0.0667,
+      "step": 51290
+    },
+    {
+      "epoch": 3.356231599607458,
+      "grad_norm": 0.8218081593513489,
+      "learning_rate": 8.564706656069726e-05,
+      "loss": 0.0762,
+      "step": 51300
+    },
+    {
+      "epoch": 3.356885835786719,
+      "grad_norm": 0.9311951398849487,
+      "learning_rate": 8.564062455802718e-05,
+      "loss": 0.077,
+      "step": 51310
+    },
+    {
+      "epoch": 3.3575400719659796,
+      "grad_norm": 0.7359833121299744,
+      "learning_rate": 8.563418135239254e-05,
+      "loss": 0.0766,
+      "step": 51320
+    },
+    {
+      "epoch": 3.3581943081452406,
+      "grad_norm": 0.6619185209274292,
+      "learning_rate": 8.56277369440108e-05,
+      "loss": 0.0756,
+      "step": 51330
+    },
+    {
+      "epoch": 3.358848544324501,
+      "grad_norm": 0.9243825078010559,
+      "learning_rate": 8.562129133309953e-05,
+      "loss": 0.0796,
+      "step": 51340
+    },
+    {
+      "epoch": 3.359502780503762,
+      "grad_norm": 0.7941439747810364,
+      "learning_rate": 8.561484451987626e-05,
+      "loss": 0.0657,
+      "step": 51350
+    },
+    {
+      "epoch": 3.3601570166830226,
+      "grad_norm": 1.0141104459762573,
+      "learning_rate": 8.560839650455857e-05,
+      "loss": 0.073,
+      "step": 51360
+    },
+    {
+      "epoch": 3.360811252862283,
+      "grad_norm": 0.7483166456222534,
+      "learning_rate": 8.560194728736412e-05,
+      "loss": 0.0696,
+      "step": 51370
+    },
+    {
+      "epoch": 3.361465489041544,
+      "grad_norm": 0.6753225922584534,
+      "learning_rate": 8.559549686851057e-05,
+      "loss": 0.0641,
+      "step": 51380
+    },
+    {
+      "epoch": 3.3621197252208046,
+      "grad_norm": 0.84939044713974,
+      "learning_rate": 8.558904524821565e-05,
+      "loss": 0.0747,
+      "step": 51390
+    },
+    {
+      "epoch": 3.3627739614000656,
+      "grad_norm": 0.9355828166007996,
+      "learning_rate": 8.558259242669713e-05,
+      "loss": 0.0762,
+      "step": 51400
+    },
+    {
+      "epoch": 3.363428197579326,
+      "grad_norm": 1.0123666524887085,
+      "learning_rate": 8.557613840417277e-05,
+      "loss": 0.0796,
+      "step": 51410
+    },
+    {
+      "epoch": 3.364082433758587,
+      "grad_norm": 0.8886246085166931,
+      "learning_rate": 8.556968318086047e-05,
+      "loss": 0.0722,
+      "step": 51420
+    },
+    {
+      "epoch": 3.3647366699378476,
+      "grad_norm": 0.7857908606529236,
+      "learning_rate": 8.556322675697806e-05,
+      "loss": 0.0708,
+      "step": 51430
+    },
+    {
+      "epoch": 3.365390906117108,
+      "grad_norm": 0.9863922595977783,
+      "learning_rate": 8.555676913274349e-05,
+      "loss": 0.0667,
+      "step": 51440
+    },
+    {
+      "epoch": 3.366045142296369,
+      "grad_norm": 1.3981025218963623,
+      "learning_rate": 8.55503103083747e-05,
+      "loss": 0.0731,
+      "step": 51450
+    },
+    {
+      "epoch": 3.3666993784756296,
+      "grad_norm": 0.8313714265823364,
+      "learning_rate": 8.55438502840897e-05,
+      "loss": 0.0696,
+      "step": 51460
+    },
+    {
+      "epoch": 3.3673536146548906,
+      "grad_norm": 0.7925471067428589,
+      "learning_rate": 8.553738906010654e-05,
+      "loss": 0.0685,
+      "step": 51470
+    },
+    {
+      "epoch": 3.368007850834151,
+      "grad_norm": 0.9922884106636047,
+      "learning_rate": 8.55309266366433e-05,
+      "loss": 0.0907,
+      "step": 51480
+    },
+    {
+      "epoch": 3.368662087013412,
+      "grad_norm": 0.8721675276756287,
+      "learning_rate": 8.55244630139181e-05,
+      "loss": 0.0703,
+      "step": 51490
+    },
+    {
+      "epoch": 3.3693163231926726,
+      "grad_norm": 0.9497895240783691,
+      "learning_rate": 8.551799819214912e-05,
+      "loss": 0.0769,
+      "step": 51500
+    },
+    {
+      "epoch": 3.369970559371933,
+      "grad_norm": 0.9085605144500732,
+      "learning_rate": 8.551153217155453e-05,
+      "loss": 0.0754,
+      "step": 51510
+    },
+    {
+      "epoch": 3.370624795551194,
+      "grad_norm": 0.8060374855995178,
+      "learning_rate": 8.550506495235262e-05,
+      "loss": 0.0747,
+      "step": 51520
+    },
+    {
+      "epoch": 3.3712790317304546,
+      "grad_norm": 0.9739006161689758,
+      "learning_rate": 8.549859653476164e-05,
+      "loss": 0.0746,
+      "step": 51530
+    },
+    {
+      "epoch": 3.3719332679097156,
+      "grad_norm": 0.8732509613037109,
+      "learning_rate": 8.549212691899993e-05,
+      "loss": 0.0795,
+      "step": 51540
+    },
+    {
+      "epoch": 3.372587504088976,
+      "grad_norm": 0.7843785285949707,
+      "learning_rate": 8.548565610528585e-05,
+      "loss": 0.0675,
+      "step": 51550
+    },
+    {
+      "epoch": 3.3732417402682366,
+      "grad_norm": 0.7273816466331482,
+      "learning_rate": 8.547918409383782e-05,
+      "loss": 0.0759,
+      "step": 51560
+    },
+    {
+      "epoch": 3.3738959764474976,
+      "grad_norm": 0.7831087708473206,
+      "learning_rate": 8.547271088487427e-05,
+      "loss": 0.0703,
+      "step": 51570
+    },
+    {
+      "epoch": 3.374550212626758,
+      "grad_norm": 0.7511657476425171,
+      "learning_rate": 8.54662364786137e-05,
+      "loss": 0.0706,
+      "step": 51580
+    },
+    {
+      "epoch": 3.375204448806019,
+      "grad_norm": 0.9097614288330078,
+      "learning_rate": 8.545976087527463e-05,
+      "loss": 0.068,
+      "step": 51590
+    },
+    {
+      "epoch": 3.3758586849852796,
+      "grad_norm": 0.7851161956787109,
+      "learning_rate": 8.545328407507565e-05,
+      "loss": 0.074,
+      "step": 51600
+    },
+    {
+      "epoch": 3.37651292116454,
+      "grad_norm": 0.9306683540344238,
+      "learning_rate": 8.544680607823534e-05,
+      "loss": 0.0856,
+      "step": 51610
+    },
+    {
+      "epoch": 3.377167157343801,
+      "grad_norm": 0.841254472732544,
+      "learning_rate": 8.544032688497236e-05,
+      "loss": 0.0771,
+      "step": 51620
+    },
+    {
+      "epoch": 3.3778213935230617,
+      "grad_norm": 1.2638838291168213,
+      "learning_rate": 8.543384649550543e-05,
+      "loss": 0.0771,
+      "step": 51630
+    },
+    {
+      "epoch": 3.3784756297023226,
+      "grad_norm": 0.802693784236908,
+      "learning_rate": 8.542736491005322e-05,
+      "loss": 0.0809,
+      "step": 51640
+    },
+    {
+      "epoch": 3.379129865881583,
+      "grad_norm": 0.7496511340141296,
+      "learning_rate": 8.542088212883454e-05,
+      "loss": 0.0827,
+      "step": 51650
+    },
+    {
+      "epoch": 3.379784102060844,
+      "grad_norm": 0.8439844250679016,
+      "learning_rate": 8.541439815206819e-05,
+      "loss": 0.0753,
+      "step": 51660
+    },
+    {
+      "epoch": 3.3804383382401046,
+      "grad_norm": 0.7389265894889832,
+      "learning_rate": 8.540791297997304e-05,
+      "loss": 0.0754,
+      "step": 51670
+    },
+    {
+      "epoch": 3.381092574419365,
+      "grad_norm": 0.983440101146698,
+      "learning_rate": 8.540142661276796e-05,
+      "loss": 0.067,
+      "step": 51680
+    },
+    {
+      "epoch": 3.381746810598626,
+      "grad_norm": 0.8338766694068909,
+      "learning_rate": 8.539493905067189e-05,
+      "loss": 0.0833,
+      "step": 51690
+    },
+    {
+      "epoch": 3.3824010467778867,
+      "grad_norm": 0.9599545001983643,
+      "learning_rate": 8.538845029390378e-05,
+      "loss": 0.0757,
+      "step": 51700
+    },
+    {
+      "epoch": 3.3830552829571476,
+      "grad_norm": 0.7370075583457947,
+      "learning_rate": 8.538196034268268e-05,
+      "loss": 0.07,
+      "step": 51710
+    },
+    {
+      "epoch": 3.383709519136408,
+      "grad_norm": 0.8070911765098572,
+      "learning_rate": 8.537546919722764e-05,
+      "loss": 0.0717,
+      "step": 51720
+    },
+    {
+      "epoch": 3.384363755315669,
+      "grad_norm": 0.8971413373947144,
+      "learning_rate": 8.536897685775772e-05,
+      "loss": 0.0659,
+      "step": 51730
+    },
+    {
+      "epoch": 3.3850179914949297,
+      "grad_norm": 0.9960509538650513,
+      "learning_rate": 8.536248332449207e-05,
+      "loss": 0.0804,
+      "step": 51740
+    },
+    {
+      "epoch": 3.38567222767419,
+      "grad_norm": 0.7966985106468201,
+      "learning_rate": 8.535598859764987e-05,
+      "loss": 0.0661,
+      "step": 51750
+    },
+    {
+      "epoch": 3.386326463853451,
+      "grad_norm": 0.9922680854797363,
+      "learning_rate": 8.534949267745034e-05,
+      "loss": 0.0726,
+      "step": 51760
+    },
+    {
+      "epoch": 3.3869807000327117,
+      "grad_norm": 0.8376561403274536,
+      "learning_rate": 8.534299556411271e-05,
+      "loss": 0.0679,
+      "step": 51770
+    },
+    {
+      "epoch": 3.3876349362119726,
+      "grad_norm": 0.915437638759613,
+      "learning_rate": 8.53364972578563e-05,
+      "loss": 0.0787,
+      "step": 51780
+    },
+    {
+      "epoch": 3.388289172391233,
+      "grad_norm": 0.7827186584472656,
+      "learning_rate": 8.532999775890043e-05,
+      "loss": 0.0735,
+      "step": 51790
+    },
+    {
+      "epoch": 3.388943408570494,
+      "grad_norm": 0.7581836581230164,
+      "learning_rate": 8.532349706746447e-05,
+      "loss": 0.0699,
+      "step": 51800
+    },
+    {
+      "epoch": 3.3895976447497547,
+      "grad_norm": 0.7785212397575378,
+      "learning_rate": 8.531699518376787e-05,
+      "loss": 0.0668,
+      "step": 51810
+    },
+    {
+      "epoch": 3.390251880929015,
+      "grad_norm": 0.9209142327308655,
+      "learning_rate": 8.531049210803003e-05,
+      "loss": 0.0723,
+      "step": 51820
+    },
+    {
+      "epoch": 3.390906117108276,
+      "grad_norm": 0.8338945508003235,
+      "learning_rate": 8.530398784047051e-05,
+      "loss": 0.081,
+      "step": 51830
+    },
+    {
+      "epoch": 3.3915603532875367,
+      "grad_norm": 0.8179078102111816,
+      "learning_rate": 8.529748238130879e-05,
+      "loss": 0.0715,
+      "step": 51840
+    },
+    {
+      "epoch": 3.3922145894667977,
+      "grad_norm": 0.7469951510429382,
+      "learning_rate": 8.529097573076447e-05,
+      "loss": 0.0812,
+      "step": 51850
+    },
+    {
+      "epoch": 3.392868825646058,
+      "grad_norm": 0.8803715705871582,
+      "learning_rate": 8.528446788905718e-05,
+      "loss": 0.0612,
+      "step": 51860
+    },
+    {
+      "epoch": 3.393523061825319,
+      "grad_norm": 0.7448040843009949,
+      "learning_rate": 8.527795885640655e-05,
+      "loss": 0.0772,
+      "step": 51870
+    },
+    {
+      "epoch": 3.3941772980045797,
+      "grad_norm": 0.9333431124687195,
+      "learning_rate": 8.527144863303227e-05,
+      "loss": 0.0727,
+      "step": 51880
+    },
+    {
+      "epoch": 3.39483153418384,
+      "grad_norm": 0.8752960562705994,
+      "learning_rate": 8.526493721915412e-05,
+      "loss": 0.0793,
+      "step": 51890
+    },
+    {
+      "epoch": 3.395485770363101,
+      "grad_norm": 0.9031022191047668,
+      "learning_rate": 8.525842461499185e-05,
+      "loss": 0.0699,
+      "step": 51900
+    },
+    {
+      "epoch": 3.3961400065423617,
+      "grad_norm": 0.8795397281646729,
+      "learning_rate": 8.525191082076527e-05,
+      "loss": 0.085,
+      "step": 51910
+    },
+    {
+      "epoch": 3.3967942427216227,
+      "grad_norm": 0.88941890001297,
+      "learning_rate": 8.524539583669426e-05,
+      "loss": 0.0742,
+      "step": 51920
+    },
+    {
+      "epoch": 3.397448478900883,
+      "grad_norm": 0.9229047894477844,
+      "learning_rate": 8.52388796629987e-05,
+      "loss": 0.0754,
+      "step": 51930
+    },
+    {
+      "epoch": 3.398102715080144,
+      "grad_norm": 0.9691248536109924,
+      "learning_rate": 8.523236229989855e-05,
+      "loss": 0.0729,
+      "step": 51940
+    },
+    {
+      "epoch": 3.3987569512594047,
+      "grad_norm": 0.8035882711410522,
+      "learning_rate": 8.522584374761375e-05,
+      "loss": 0.0768,
+      "step": 51950
+    },
+    {
+      "epoch": 3.399411187438665,
+      "grad_norm": 0.7715499997138977,
+      "learning_rate": 8.521932400636434e-05,
+      "loss": 0.075,
+      "step": 51960
+    },
+    {
+      "epoch": 3.400065423617926,
+      "grad_norm": 0.75547194480896,
+      "learning_rate": 8.52128030763704e-05,
+      "loss": 0.0622,
+      "step": 51970
+    },
+    {
+      "epoch": 3.4007196597971867,
+      "grad_norm": 1.1841288805007935,
+      "learning_rate": 8.520628095785199e-05,
+      "loss": 0.0721,
+      "step": 51980
+    },
+    {
+      "epoch": 3.4013738959764477,
+      "grad_norm": 0.8948665261268616,
+      "learning_rate": 8.519975765102927e-05,
+      "loss": 0.0674,
+      "step": 51990
+    },
+    {
+      "epoch": 3.402028132155708,
+      "grad_norm": 0.9164159297943115,
+      "learning_rate": 8.519323315612242e-05,
+      "loss": 0.0703,
+      "step": 52000
+    },
+    {
+      "epoch": 3.4026823683349687,
+      "grad_norm": 0.8155785799026489,
+      "learning_rate": 8.518670747335165e-05,
+      "loss": 0.0767,
+      "step": 52010
+    },
+    {
+      "epoch": 3.4033366045142297,
+      "grad_norm": 0.8073797821998596,
+      "learning_rate": 8.518018060293722e-05,
+      "loss": 0.0724,
+      "step": 52020
+    },
+    {
+      "epoch": 3.4039908406934902,
+      "grad_norm": 0.816716194152832,
+      "learning_rate": 8.517365254509942e-05,
+      "loss": 0.0661,
+      "step": 52030
+    },
+    {
+      "epoch": 3.404645076872751,
+      "grad_norm": 0.8661830425262451,
+      "learning_rate": 8.516712330005862e-05,
+      "loss": 0.0731,
+      "step": 52040
+    },
+    {
+      "epoch": 3.4052993130520117,
+      "grad_norm": 0.8978433012962341,
+      "learning_rate": 8.516059286803517e-05,
+      "loss": 0.0668,
+      "step": 52050
+    },
+    {
+      "epoch": 3.4059535492312722,
+      "grad_norm": 0.892529308795929,
+      "learning_rate": 8.515406124924949e-05,
+      "loss": 0.082,
+      "step": 52060
+    },
+    {
+      "epoch": 3.406607785410533,
+      "grad_norm": 0.6987180709838867,
+      "learning_rate": 8.514752844392206e-05,
+      "loss": 0.0676,
+      "step": 52070
+    },
+    {
+      "epoch": 3.4072620215897937,
+      "grad_norm": 0.8603876829147339,
+      "learning_rate": 8.514099445227336e-05,
+      "loss": 0.0658,
+      "step": 52080
+    },
+    {
+      "epoch": 3.4079162577690547,
+      "grad_norm": 0.7895892858505249,
+      "learning_rate": 8.513445927452396e-05,
+      "loss": 0.0637,
+      "step": 52090
+    },
+    {
+      "epoch": 3.4085704939483152,
+      "grad_norm": 0.8536374568939209,
+      "learning_rate": 8.51279229108944e-05,
+      "loss": 0.0765,
+      "step": 52100
+    },
+    {
+      "epoch": 3.409224730127576,
+      "grad_norm": 1.0041425228118896,
+      "learning_rate": 8.51213853616053e-05,
+      "loss": 0.0794,
+      "step": 52110
+    },
+    {
+      "epoch": 3.4098789663068367,
+      "grad_norm": 0.9329466223716736,
+      "learning_rate": 8.511484662687737e-05,
+      "loss": 0.0693,
+      "step": 52120
+    },
+    {
+      "epoch": 3.4105332024860973,
+      "grad_norm": 0.9463549852371216,
+      "learning_rate": 8.510830670693124e-05,
+      "loss": 0.073,
+      "step": 52130
+    },
+    {
+      "epoch": 3.4111874386653582,
+      "grad_norm": 0.8371836543083191,
+      "learning_rate": 8.51017656019877e-05,
+      "loss": 0.0707,
+      "step": 52140
+    },
+    {
+      "epoch": 3.4118416748446188,
+      "grad_norm": 1.1609052419662476,
+      "learning_rate": 8.50952233122675e-05,
+      "loss": 0.0705,
+      "step": 52150
+    },
+    {
+      "epoch": 3.4124959110238797,
+      "grad_norm": 0.9662759900093079,
+      "learning_rate": 8.50886798379915e-05,
+      "loss": 0.0732,
+      "step": 52160
+    },
+    {
+      "epoch": 3.4131501472031402,
+      "grad_norm": 0.9604242444038391,
+      "learning_rate": 8.50821351793805e-05,
+      "loss": 0.0683,
+      "step": 52170
+    },
+    {
+      "epoch": 3.413804383382401,
+      "grad_norm": 0.8330959677696228,
+      "learning_rate": 8.507558933665545e-05,
+      "loss": 0.0712,
+      "step": 52180
+    },
+    {
+      "epoch": 3.4144586195616617,
+      "grad_norm": 0.7776506543159485,
+      "learning_rate": 8.506904231003726e-05,
+      "loss": 0.0555,
+      "step": 52190
+    },
+    {
+      "epoch": 3.4151128557409223,
+      "grad_norm": 0.8300237059593201,
+      "learning_rate": 8.506249409974694e-05,
+      "loss": 0.0837,
+      "step": 52200
+    },
+    {
+      "epoch": 3.4157670919201832,
+      "grad_norm": 0.8520910143852234,
+      "learning_rate": 8.505594470600546e-05,
+      "loss": 0.0696,
+      "step": 52210
+    },
+    {
+      "epoch": 3.4164213280994438,
+      "grad_norm": 0.8499581217765808,
+      "learning_rate": 8.504939412903394e-05,
+      "loss": 0.0756,
+      "step": 52220
+    },
+    {
+      "epoch": 3.4170755642787047,
+      "grad_norm": 1.0196528434753418,
+      "learning_rate": 8.504284236905342e-05,
+      "loss": 0.079,
+      "step": 52230
+    },
+    {
+      "epoch": 3.4177298004579653,
+      "grad_norm": 1.0071736574172974,
+      "learning_rate": 8.503628942628508e-05,
+      "loss": 0.0836,
+      "step": 52240
+    },
+    {
+      "epoch": 3.4183840366372262,
+      "grad_norm": 0.8221054673194885,
+      "learning_rate": 8.502973530095008e-05,
+      "loss": 0.0666,
+      "step": 52250
+    },
+    {
+      "epoch": 3.4190382728164868,
+      "grad_norm": 0.9382101893424988,
+      "learning_rate": 8.502317999326965e-05,
+      "loss": 0.0742,
+      "step": 52260
+    },
+    {
+      "epoch": 3.4196925089957473,
+      "grad_norm": 0.9391849637031555,
+      "learning_rate": 8.501662350346505e-05,
+      "loss": 0.0725,
+      "step": 52270
+    },
+    {
+      "epoch": 3.4203467451750083,
+      "grad_norm": 0.7886925339698792,
+      "learning_rate": 8.501006583175757e-05,
+      "loss": 0.0861,
+      "step": 52280
+    },
+    {
+      "epoch": 3.4210009813542688,
+      "grad_norm": 0.7476629018783569,
+      "learning_rate": 8.500350697836855e-05,
+      "loss": 0.0649,
+      "step": 52290
+    },
+    {
+      "epoch": 3.4216552175335297,
+      "grad_norm": 1.02727472782135,
+      "learning_rate": 8.499694694351936e-05,
+      "loss": 0.0923,
+      "step": 52300
+    },
+    {
+      "epoch": 3.4223094537127903,
+      "grad_norm": 1.176798939704895,
+      "learning_rate": 8.499038572743144e-05,
+      "loss": 0.0644,
+      "step": 52310
+    },
+    {
+      "epoch": 3.4229636898920512,
+      "grad_norm": 0.8501316905021667,
+      "learning_rate": 8.498382333032622e-05,
+      "loss": 0.0703,
+      "step": 52320
+    },
+    {
+      "epoch": 3.4236179260713118,
+      "grad_norm": 0.7579331994056702,
+      "learning_rate": 8.497725975242523e-05,
+      "loss": 0.0661,
+      "step": 52330
+    },
+    {
+      "epoch": 3.4242721622505723,
+      "grad_norm": 0.8716086149215698,
+      "learning_rate": 8.497069499394998e-05,
+      "loss": 0.0781,
+      "step": 52340
+    },
+    {
+      "epoch": 3.4249263984298333,
+      "grad_norm": 1.071758508682251,
+      "learning_rate": 8.496412905512207e-05,
+      "loss": 0.0681,
+      "step": 52350
+    },
+    {
+      "epoch": 3.425580634609094,
+      "grad_norm": 0.8886385560035706,
+      "learning_rate": 8.49575619361631e-05,
+      "loss": 0.0763,
+      "step": 52360
+    },
+    {
+      "epoch": 3.4262348707883548,
+      "grad_norm": 0.9664177298545837,
+      "learning_rate": 8.495099363729472e-05,
+      "loss": 0.0768,
+      "step": 52370
+    },
+    {
+      "epoch": 3.4268891069676153,
+      "grad_norm": 0.7250670194625854,
+      "learning_rate": 8.494442415873868e-05,
+      "loss": 0.0724,
+      "step": 52380
+    },
+    {
+      "epoch": 3.4275433431468763,
+      "grad_norm": 0.9024300575256348,
+      "learning_rate": 8.493785350071665e-05,
+      "loss": 0.0842,
+      "step": 52390
+    },
+    {
+      "epoch": 3.428197579326137,
+      "grad_norm": 0.9288867712020874,
+      "learning_rate": 8.493128166345046e-05,
+      "loss": 0.0703,
+      "step": 52400
+    },
+    {
+      "epoch": 3.4288518155053973,
+      "grad_norm": 0.7886371612548828,
+      "learning_rate": 8.492470864716188e-05,
+      "loss": 0.0677,
+      "step": 52410
+    },
+    {
+      "epoch": 3.4295060516846583,
+      "grad_norm": 0.9110251665115356,
+      "learning_rate": 8.491813445207282e-05,
+      "loss": 0.0747,
+      "step": 52420
+    },
+    {
+      "epoch": 3.430160287863919,
+      "grad_norm": 0.7221360802650452,
+      "learning_rate": 8.491155907840511e-05,
+      "loss": 0.067,
+      "step": 52430
+    },
+    {
+      "epoch": 3.4308145240431798,
+      "grad_norm": 0.6658129096031189,
+      "learning_rate": 8.490498252638074e-05,
+      "loss": 0.0795,
+      "step": 52440
+    },
+    {
+      "epoch": 3.4314687602224403,
+      "grad_norm": 0.8011776208877563,
+      "learning_rate": 8.489840479622166e-05,
+      "loss": 0.068,
+      "step": 52450
+    },
+    {
+      "epoch": 3.432122996401701,
+      "grad_norm": 0.8690958619117737,
+      "learning_rate": 8.48918258881499e-05,
+      "loss": 0.0713,
+      "step": 52460
+    },
+    {
+      "epoch": 3.432777232580962,
+      "grad_norm": 0.8399226069450378,
+      "learning_rate": 8.488524580238752e-05,
+      "loss": 0.0726,
+      "step": 52470
+    },
+    {
+      "epoch": 3.4334314687602223,
+      "grad_norm": 0.7656417489051819,
+      "learning_rate": 8.487866453915658e-05,
+      "loss": 0.0731,
+      "step": 52480
+    },
+    {
+      "epoch": 3.4340857049394833,
+      "grad_norm": 0.8498892784118652,
+      "learning_rate": 8.487208209867928e-05,
+      "loss": 0.0818,
+      "step": 52490
+    },
+    {
+      "epoch": 3.434739941118744,
+      "grad_norm": 0.9708839058876038,
+      "learning_rate": 8.48654984811777e-05,
+      "loss": 0.0719,
+      "step": 52500
+    },
+    {
+      "epoch": 3.4353941772980043,
+      "grad_norm": 0.9287952184677124,
+      "learning_rate": 8.485891368687415e-05,
+      "loss": 0.0764,
+      "step": 52510
+    },
+    {
+      "epoch": 3.4360484134772653,
+      "grad_norm": 0.7302289605140686,
+      "learning_rate": 8.485232771599081e-05,
+      "loss": 0.0717,
+      "step": 52520
+    },
+    {
+      "epoch": 3.436702649656526,
+      "grad_norm": 0.9905624985694885,
+      "learning_rate": 8.484574056875003e-05,
+      "loss": 0.0685,
+      "step": 52530
+    },
+    {
+      "epoch": 3.437356885835787,
+      "grad_norm": 0.8764323592185974,
+      "learning_rate": 8.483915224537411e-05,
+      "loss": 0.0694,
+      "step": 52540
+    },
+    {
+      "epoch": 3.4380111220150473,
+      "grad_norm": 0.85755854845047,
+      "learning_rate": 8.483256274608544e-05,
+      "loss": 0.0723,
+      "step": 52550
+    },
+    {
+      "epoch": 3.4386653581943083,
+      "grad_norm": 0.8264038562774658,
+      "learning_rate": 8.482597207110642e-05,
+      "loss": 0.0667,
+      "step": 52560
+    },
+    {
+      "epoch": 3.439319594373569,
+      "grad_norm": 0.8556039333343506,
+      "learning_rate": 8.481938022065951e-05,
+      "loss": 0.0814,
+      "step": 52570
+    },
+    {
+      "epoch": 3.4399738305528293,
+      "grad_norm": 0.8458713293075562,
+      "learning_rate": 8.48127871949672e-05,
+      "loss": 0.0724,
+      "step": 52580
+    },
+    {
+      "epoch": 3.4406280667320903,
+      "grad_norm": 0.9362344145774841,
+      "learning_rate": 8.480619299425202e-05,
+      "loss": 0.0773,
+      "step": 52590
+    },
+    {
+      "epoch": 3.441282302911351,
+      "grad_norm": 1.080499291419983,
+      "learning_rate": 8.479959761873655e-05,
+      "loss": 0.0651,
+      "step": 52600
+    },
+    {
+      "epoch": 3.441936539090612,
+      "grad_norm": 1.0057145357131958,
+      "learning_rate": 8.479300106864338e-05,
+      "loss": 0.0717,
+      "step": 52610
+    },
+    {
+      "epoch": 3.4425907752698723,
+      "grad_norm": 0.8773415684700012,
+      "learning_rate": 8.478640334419519e-05,
+      "loss": 0.0668,
+      "step": 52620
+    },
+    {
+      "epoch": 3.4432450114491333,
+      "grad_norm": 0.7790507674217224,
+      "learning_rate": 8.477980444561465e-05,
+      "loss": 0.0702,
+      "step": 52630
+    },
+    {
+      "epoch": 3.443899247628394,
+      "grad_norm": 0.7044256925582886,
+      "learning_rate": 8.47732043731245e-05,
+      "loss": 0.0746,
+      "step": 52640
+    },
+    {
+      "epoch": 3.4445534838076544,
+      "grad_norm": 0.7038286924362183,
+      "learning_rate": 8.476660312694751e-05,
+      "loss": 0.0774,
+      "step": 52650
+    },
+    {
+      "epoch": 3.4452077199869153,
+      "grad_norm": 0.8981152772903442,
+      "learning_rate": 8.476000070730647e-05,
+      "loss": 0.0802,
+      "step": 52660
+    },
+    {
+      "epoch": 3.445861956166176,
+      "grad_norm": 0.8281787633895874,
+      "learning_rate": 8.475339711442428e-05,
+      "loss": 0.0639,
+      "step": 52670
+    },
+    {
+      "epoch": 3.446516192345437,
+      "grad_norm": 0.7848443388938904,
+      "learning_rate": 8.474679234852377e-05,
+      "loss": 0.0688,
+      "step": 52680
+    },
+    {
+      "epoch": 3.4471704285246973,
+      "grad_norm": 0.9246609210968018,
+      "learning_rate": 8.474018640982789e-05,
+      "loss": 0.0724,
+      "step": 52690
+    },
+    {
+      "epoch": 3.4478246647039583,
+      "grad_norm": 0.9001889228820801,
+      "learning_rate": 8.473357929855958e-05,
+      "loss": 0.0872,
+      "step": 52700
+    },
+    {
+      "epoch": 3.448478900883219,
+      "grad_norm": 0.7608310580253601,
+      "learning_rate": 8.472697101494192e-05,
+      "loss": 0.076,
+      "step": 52710
+    },
+    {
+      "epoch": 3.4491331370624794,
+      "grad_norm": 0.9559541940689087,
+      "learning_rate": 8.472036155919791e-05,
+      "loss": 0.071,
+      "step": 52720
+    },
+    {
+      "epoch": 3.4497873732417403,
+      "grad_norm": 0.8626623153686523,
+      "learning_rate": 8.471375093155061e-05,
+      "loss": 0.0722,
+      "step": 52730
+    },
+    {
+      "epoch": 3.450441609421001,
+      "grad_norm": 0.6806735992431641,
+      "learning_rate": 8.470713913222321e-05,
+      "loss": 0.0697,
+      "step": 52740
+    },
+    {
+      "epoch": 3.451095845600262,
+      "grad_norm": 0.9566436409950256,
+      "learning_rate": 8.470052616143883e-05,
+      "loss": 0.0716,
+      "step": 52750
+    },
+    {
+      "epoch": 3.4517500817795224,
+      "grad_norm": 0.8389673829078674,
+      "learning_rate": 8.469391201942068e-05,
+      "loss": 0.0661,
+      "step": 52760
+    },
+    {
+      "epoch": 3.4524043179587833,
+      "grad_norm": 0.8601690530776978,
+      "learning_rate": 8.468729670639201e-05,
+      "loss": 0.0764,
+      "step": 52770
+    },
+    {
+      "epoch": 3.453058554138044,
+      "grad_norm": 0.8300301432609558,
+      "learning_rate": 8.468068022257611e-05,
+      "loss": 0.0793,
+      "step": 52780
+    },
+    {
+      "epoch": 3.4537127903173044,
+      "grad_norm": 0.8958084583282471,
+      "learning_rate": 8.46740625681963e-05,
+      "loss": 0.0737,
+      "step": 52790
+    },
+    {
+      "epoch": 3.4543670264965654,
+      "grad_norm": 0.9533824920654297,
+      "learning_rate": 8.466744374347593e-05,
+      "loss": 0.0647,
+      "step": 52800
+    },
+    {
+      "epoch": 3.455021262675826,
+      "grad_norm": 0.7993425130844116,
+      "learning_rate": 8.466082374863844e-05,
+      "loss": 0.0792,
+      "step": 52810
+    },
+    {
+      "epoch": 3.455675498855087,
+      "grad_norm": 0.8701884746551514,
+      "learning_rate": 8.465420258390723e-05,
+      "loss": 0.0681,
+      "step": 52820
+    },
+    {
+      "epoch": 3.4563297350343474,
+      "grad_norm": 0.8877395987510681,
+      "learning_rate": 8.464758024950581e-05,
+      "loss": 0.0706,
+      "step": 52830
+    },
+    {
+      "epoch": 3.4569839712136083,
+      "grad_norm": 0.9851657152175903,
+      "learning_rate": 8.464095674565769e-05,
+      "loss": 0.0668,
+      "step": 52840
+    },
+    {
+      "epoch": 3.457638207392869,
+      "grad_norm": 1.3369970321655273,
+      "learning_rate": 8.46343320725864e-05,
+      "loss": 0.0743,
+      "step": 52850
+    },
+    {
+      "epoch": 3.4582924435721294,
+      "grad_norm": 1.114278793334961,
+      "learning_rate": 8.462770623051561e-05,
+      "loss": 0.0723,
+      "step": 52860
+    },
+    {
+      "epoch": 3.4589466797513904,
+      "grad_norm": 0.8872862458229065,
+      "learning_rate": 8.46210792196689e-05,
+      "loss": 0.0695,
+      "step": 52870
+    },
+    {
+      "epoch": 3.459600915930651,
+      "grad_norm": 0.9526658654212952,
+      "learning_rate": 8.461445104026997e-05,
+      "loss": 0.076,
+      "step": 52880
+    },
+    {
+      "epoch": 3.460255152109912,
+      "grad_norm": 0.7298262715339661,
+      "learning_rate": 8.460782169254254e-05,
+      "loss": 0.0723,
+      "step": 52890
+    },
+    {
+      "epoch": 3.4609093882891724,
+      "grad_norm": 0.845991313457489,
+      "learning_rate": 8.460119117671037e-05,
+      "loss": 0.0694,
+      "step": 52900
+    },
+    {
+      "epoch": 3.4615636244684334,
+      "grad_norm": 0.8948342800140381,
+      "learning_rate": 8.459455949299724e-05,
+      "loss": 0.0685,
+      "step": 52910
+    },
+    {
+      "epoch": 3.462217860647694,
+      "grad_norm": 0.7942617535591125,
+      "learning_rate": 8.458792664162702e-05,
+      "loss": 0.0754,
+      "step": 52920
+    },
+    {
+      "epoch": 3.4628720968269544,
+      "grad_norm": 0.959294319152832,
+      "learning_rate": 8.458129262282355e-05,
+      "loss": 0.0776,
+      "step": 52930
+    },
+    {
+      "epoch": 3.4635263330062154,
+      "grad_norm": 0.7469987273216248,
+      "learning_rate": 8.457465743681077e-05,
+      "loss": 0.0724,
+      "step": 52940
+    },
+    {
+      "epoch": 3.464180569185476,
+      "grad_norm": 0.8524947166442871,
+      "learning_rate": 8.456802108381261e-05,
+      "loss": 0.0758,
+      "step": 52950
+    },
+    {
+      "epoch": 3.4648348053647364,
+      "grad_norm": 0.8579651713371277,
+      "learning_rate": 8.45613835640531e-05,
+      "loss": 0.0835,
+      "step": 52960
+    },
+    {
+      "epoch": 3.4654890415439974,
+      "grad_norm": 0.9005305767059326,
+      "learning_rate": 8.455474487775625e-05,
+      "loss": 0.0741,
+      "step": 52970
+    },
+    {
+      "epoch": 3.466143277723258,
+      "grad_norm": 0.8470668792724609,
+      "learning_rate": 8.454810502514614e-05,
+      "loss": 0.0657,
+      "step": 52980
+    },
+    {
+      "epoch": 3.466797513902519,
+      "grad_norm": 0.7774618864059448,
+      "learning_rate": 8.454146400644687e-05,
+      "loss": 0.0734,
+      "step": 52990
+    },
+    {
+      "epoch": 3.4674517500817794,
+      "grad_norm": 0.7881149649620056,
+      "learning_rate": 8.453482182188259e-05,
+      "loss": 0.0763,
+      "step": 53000
+    },
+    {
+      "epoch": 3.4681059862610404,
+      "grad_norm": 0.9058419466018677,
+      "learning_rate": 8.452817847167753e-05,
+      "loss": 0.0701,
+      "step": 53010
+    },
+    {
+      "epoch": 3.468760222440301,
+      "grad_norm": 0.927578866481781,
+      "learning_rate": 8.452153395605587e-05,
+      "loss": 0.0646,
+      "step": 53020
+    },
+    {
+      "epoch": 3.4694144586195614,
+      "grad_norm": 0.8536058664321899,
+      "learning_rate": 8.451488827524192e-05,
+      "loss": 0.0701,
+      "step": 53030
+    },
+    {
+      "epoch": 3.4700686947988224,
+      "grad_norm": 0.7442119717597961,
+      "learning_rate": 8.450824142945997e-05,
+      "loss": 0.0696,
+      "step": 53040
+    },
+    {
+      "epoch": 3.470722930978083,
+      "grad_norm": 0.7530878186225891,
+      "learning_rate": 8.450159341893436e-05,
+      "loss": 0.0736,
+      "step": 53050
+    },
+    {
+      "epoch": 3.471377167157344,
+      "grad_norm": 0.8248975276947021,
+      "learning_rate": 8.449494424388951e-05,
+      "loss": 0.0658,
+      "step": 53060
+    },
+    {
+      "epoch": 3.4720314033366044,
+      "grad_norm": 0.973678708076477,
+      "learning_rate": 8.44882939045498e-05,
+      "loss": 0.0659,
+      "step": 53070
+    },
+    {
+      "epoch": 3.4726856395158654,
+      "grad_norm": 0.950872004032135,
+      "learning_rate": 8.448164240113972e-05,
+      "loss": 0.0722,
+      "step": 53080
+    },
+    {
+      "epoch": 3.473339875695126,
+      "grad_norm": 0.7618328332901001,
+      "learning_rate": 8.447498973388379e-05,
+      "loss": 0.0724,
+      "step": 53090
+    },
+    {
+      "epoch": 3.4739941118743864,
+      "grad_norm": 0.7460528612136841,
+      "learning_rate": 8.446833590300656e-05,
+      "loss": 0.0718,
+      "step": 53100
+    },
+    {
+      "epoch": 3.4746483480536474,
+      "grad_norm": 1.07584810256958,
+      "learning_rate": 8.446168090873257e-05,
+      "loss": 0.0778,
+      "step": 53110
+    },
+    {
+      "epoch": 3.475302584232908,
+      "grad_norm": 0.7845546007156372,
+      "learning_rate": 8.445502475128649e-05,
+      "loss": 0.0703,
+      "step": 53120
+    },
+    {
+      "epoch": 3.475956820412169,
+      "grad_norm": 0.9104862809181213,
+      "learning_rate": 8.444836743089294e-05,
+      "loss": 0.0711,
+      "step": 53130
+    },
+    {
+      "epoch": 3.4766110565914294,
+      "grad_norm": 0.8521603941917419,
+      "learning_rate": 8.444170894777665e-05,
+      "loss": 0.0751,
+      "step": 53140
+    },
+    {
+      "epoch": 3.4772652927706904,
+      "grad_norm": 1.0778313875198364,
+      "learning_rate": 8.443504930216237e-05,
+      "loss": 0.0772,
+      "step": 53150
+    },
+    {
+      "epoch": 3.477919528949951,
+      "grad_norm": 1.079459309577942,
+      "learning_rate": 8.442838849427486e-05,
+      "loss": 0.086,
+      "step": 53160
+    },
+    {
+      "epoch": 3.4785737651292115,
+      "grad_norm": 0.8837111592292786,
+      "learning_rate": 8.442172652433895e-05,
+      "loss": 0.0755,
+      "step": 53170
+    },
+    {
+      "epoch": 3.4792280013084724,
+      "grad_norm": 0.906924843788147,
+      "learning_rate": 8.441506339257949e-05,
+      "loss": 0.0708,
+      "step": 53180
+    },
+    {
+      "epoch": 3.479882237487733,
+      "grad_norm": 0.8300266861915588,
+      "learning_rate": 8.440839909922139e-05,
+      "loss": 0.0653,
+      "step": 53190
+    },
+    {
+      "epoch": 3.480536473666994,
+      "grad_norm": 0.9430700540542603,
+      "learning_rate": 8.440173364448958e-05,
+      "loss": 0.0642,
+      "step": 53200
+    },
+    {
+      "epoch": 3.4811907098462544,
+      "grad_norm": 0.9260400533676147,
+      "learning_rate": 8.439506702860902e-05,
+      "loss": 0.0705,
+      "step": 53210
+    },
+    {
+      "epoch": 3.4818449460255154,
+      "grad_norm": 0.8879568576812744,
+      "learning_rate": 8.438839925180476e-05,
+      "loss": 0.0817,
+      "step": 53220
+    },
+    {
+      "epoch": 3.482499182204776,
+      "grad_norm": 0.8787881731987,
+      "learning_rate": 8.438173031430185e-05,
+      "loss": 0.072,
+      "step": 53230
+    },
+    {
+      "epoch": 3.4831534183840365,
+      "grad_norm": 0.9385049343109131,
+      "learning_rate": 8.437506021632535e-05,
+      "loss": 0.0777,
+      "step": 53240
+    },
+    {
+      "epoch": 3.4838076545632974,
+      "grad_norm": 0.8981115818023682,
+      "learning_rate": 8.436838895810042e-05,
+      "loss": 0.0648,
+      "step": 53250
+    },
+    {
+      "epoch": 3.484461890742558,
+      "grad_norm": 0.8167151808738708,
+      "learning_rate": 8.436171653985223e-05,
+      "loss": 0.0831,
+      "step": 53260
+    },
+    {
+      "epoch": 3.485116126921819,
+      "grad_norm": 1.022867202758789,
+      "learning_rate": 8.4355042961806e-05,
+      "loss": 0.0701,
+      "step": 53270
+    },
+    {
+      "epoch": 3.4857703631010795,
+      "grad_norm": 0.6986026167869568,
+      "learning_rate": 8.434836822418697e-05,
+      "loss": 0.0672,
+      "step": 53280
+    },
+    {
+      "epoch": 3.4864245992803404,
+      "grad_norm": 0.8744747042655945,
+      "learning_rate": 8.434169232722043e-05,
+      "loss": 0.0703,
+      "step": 53290
+    },
+    {
+      "epoch": 3.487078835459601,
+      "grad_norm": 1.0003246068954468,
+      "learning_rate": 8.433501527113169e-05,
+      "loss": 0.0639,
+      "step": 53300
+    },
+    {
+      "epoch": 3.4877330716388615,
+      "grad_norm": 0.8735449314117432,
+      "learning_rate": 8.432833705614616e-05,
+      "loss": 0.0683,
+      "step": 53310
+    },
+    {
+      "epoch": 3.4883873078181225,
+      "grad_norm": 0.8451821804046631,
+      "learning_rate": 8.43216576824892e-05,
+      "loss": 0.0666,
+      "step": 53320
+    },
+    {
+      "epoch": 3.489041543997383,
+      "grad_norm": 0.8688971996307373,
+      "learning_rate": 8.43149771503863e-05,
+      "loss": 0.0644,
+      "step": 53330
+    },
+    {
+      "epoch": 3.489695780176644,
+      "grad_norm": 0.7318007349967957,
+      "learning_rate": 8.430829546006293e-05,
+      "loss": 0.0689,
+      "step": 53340
+    },
+    {
+      "epoch": 3.4903500163559045,
+      "grad_norm": 0.8060617446899414,
+      "learning_rate": 8.430161261174461e-05,
+      "loss": 0.0713,
+      "step": 53350
+    },
+    {
+      "epoch": 3.4910042525351654,
+      "grad_norm": 0.8988872170448303,
+      "learning_rate": 8.42949286056569e-05,
+      "loss": 0.0741,
+      "step": 53360
+    },
+    {
+      "epoch": 3.491658488714426,
+      "grad_norm": 1.0134598016738892,
+      "learning_rate": 8.42882434420254e-05,
+      "loss": 0.0698,
+      "step": 53370
+    },
+    {
+      "epoch": 3.4923127248936865,
+      "grad_norm": 0.8395692706108093,
+      "learning_rate": 8.428155712107577e-05,
+      "loss": 0.064,
+      "step": 53380
+    },
+    {
+      "epoch": 3.4929669610729475,
+      "grad_norm": 0.8911391496658325,
+      "learning_rate": 8.427486964303368e-05,
+      "loss": 0.0741,
+      "step": 53390
+    },
+    {
+      "epoch": 3.493621197252208,
+      "grad_norm": 0.7914614081382751,
+      "learning_rate": 8.426818100812486e-05,
+      "loss": 0.073,
+      "step": 53400
+    },
+    {
+      "epoch": 3.4942754334314685,
+      "grad_norm": 0.8505756855010986,
+      "learning_rate": 8.426149121657504e-05,
+      "loss": 0.0766,
+      "step": 53410
+    },
+    {
+      "epoch": 3.4949296696107295,
+      "grad_norm": 0.6521218419075012,
+      "learning_rate": 8.425480026861006e-05,
+      "loss": 0.0668,
+      "step": 53420
+    },
+    {
+      "epoch": 3.49558390578999,
+      "grad_norm": 0.7567178010940552,
+      "learning_rate": 8.424810816445571e-05,
+      "loss": 0.0648,
+      "step": 53430
+    },
+    {
+      "epoch": 3.496238141969251,
+      "grad_norm": 0.8685494065284729,
+      "learning_rate": 8.42414149043379e-05,
+      "loss": 0.0749,
+      "step": 53440
+    },
+    {
+      "epoch": 3.4968923781485115,
+      "grad_norm": 0.8376366496086121,
+      "learning_rate": 8.423472048848254e-05,
+      "loss": 0.0722,
+      "step": 53450
+    },
+    {
+      "epoch": 3.4975466143277725,
+      "grad_norm": 0.8271428346633911,
+      "learning_rate": 8.422802491711557e-05,
+      "loss": 0.0799,
+      "step": 53460
+    },
+    {
+      "epoch": 3.498200850507033,
+      "grad_norm": 0.9023700952529907,
+      "learning_rate": 8.4221328190463e-05,
+      "loss": 0.0633,
+      "step": 53470
+    },
+    {
+      "epoch": 3.4988550866862935,
+      "grad_norm": 0.8496003150939941,
+      "learning_rate": 8.421463030875085e-05,
+      "loss": 0.0744,
+      "step": 53480
+    },
+    {
+      "epoch": 3.4995093228655545,
+      "grad_norm": 0.7680093050003052,
+      "learning_rate": 8.420793127220521e-05,
+      "loss": 0.0715,
+      "step": 53490
+    },
+    {
+      "epoch": 3.500163559044815,
+      "grad_norm": 0.992078423500061,
+      "learning_rate": 8.420123108105215e-05,
+      "loss": 0.0807,
+      "step": 53500
+    },
+    {
+      "epoch": 3.500817795224076,
+      "grad_norm": 0.7784678936004639,
+      "learning_rate": 8.419452973551786e-05,
+      "loss": 0.0692,
+      "step": 53510
+    },
+    {
+      "epoch": 3.5014720314033365,
+      "grad_norm": 0.8641963601112366,
+      "learning_rate": 8.418782723582852e-05,
+      "loss": 0.0731,
+      "step": 53520
+    },
+    {
+      "epoch": 3.5021262675825975,
+      "grad_norm": 0.8115648627281189,
+      "learning_rate": 8.418112358221036e-05,
+      "loss": 0.076,
+      "step": 53530
+    },
+    {
+      "epoch": 3.502780503761858,
+      "grad_norm": 0.9229583144187927,
+      "learning_rate": 8.417441877488961e-05,
+      "loss": 0.0721,
+      "step": 53540
+    },
+    {
+      "epoch": 3.5034347399411185,
+      "grad_norm": 0.9231205582618713,
+      "learning_rate": 8.416771281409262e-05,
+      "loss": 0.0679,
+      "step": 53550
+    },
+    {
+      "epoch": 3.5040889761203795,
+      "grad_norm": 0.8276332020759583,
+      "learning_rate": 8.41610057000457e-05,
+      "loss": 0.0737,
+      "step": 53560
+    },
+    {
+      "epoch": 3.50474321229964,
+      "grad_norm": 0.75644451379776,
+      "learning_rate": 8.415429743297524e-05,
+      "loss": 0.0627,
+      "step": 53570
+    },
+    {
+      "epoch": 3.505397448478901,
+      "grad_norm": 0.7527329325675964,
+      "learning_rate": 8.41475880131077e-05,
+      "loss": 0.0754,
+      "step": 53580
+    },
+    {
+      "epoch": 3.5060516846581615,
+      "grad_norm": 0.8161507248878479,
+      "learning_rate": 8.414087744066947e-05,
+      "loss": 0.065,
+      "step": 53590
+    },
+    {
+      "epoch": 3.5067059208374225,
+      "grad_norm": 0.851140558719635,
+      "learning_rate": 8.413416571588713e-05,
+      "loss": 0.0655,
+      "step": 53600
+    },
+    {
+      "epoch": 3.507360157016683,
+      "grad_norm": 0.8948183059692383,
+      "learning_rate": 8.412745283898714e-05,
+      "loss": 0.0698,
+      "step": 53610
+    },
+    {
+      "epoch": 3.5080143931959435,
+      "grad_norm": 1.1371095180511475,
+      "learning_rate": 8.412073881019613e-05,
+      "loss": 0.0762,
+      "step": 53620
+    },
+    {
+      "epoch": 3.5086686293752045,
+      "grad_norm": 0.9154216647148132,
+      "learning_rate": 8.41140236297407e-05,
+      "loss": 0.0722,
+      "step": 53630
+    },
+    {
+      "epoch": 3.509322865554465,
+      "grad_norm": 0.8594346046447754,
+      "learning_rate": 8.41073072978475e-05,
+      "loss": 0.0718,
+      "step": 53640
+    },
+    {
+      "epoch": 3.509977101733726,
+      "grad_norm": 0.9689813852310181,
+      "learning_rate": 8.410058981474324e-05,
+      "loss": 0.0693,
+      "step": 53650
+    },
+    {
+      "epoch": 3.5106313379129865,
+      "grad_norm": 0.80068039894104,
+      "learning_rate": 8.409387118065464e-05,
+      "loss": 0.0758,
+      "step": 53660
+    },
+    {
+      "epoch": 3.5112855740922475,
+      "grad_norm": 0.8130351901054382,
+      "learning_rate": 8.408715139580846e-05,
+      "loss": 0.0785,
+      "step": 53670
+    },
+    {
+      "epoch": 3.511939810271508,
+      "grad_norm": 0.758581817150116,
+      "learning_rate": 8.408043046043154e-05,
+      "loss": 0.0636,
+      "step": 53680
+    },
+    {
+      "epoch": 3.5125940464507686,
+      "grad_norm": 0.955032467842102,
+      "learning_rate": 8.407370837475071e-05,
+      "loss": 0.0794,
+      "step": 53690
+    },
+    {
+      "epoch": 3.5132482826300295,
+      "grad_norm": 0.7809516191482544,
+      "learning_rate": 8.406698513899285e-05,
+      "loss": 0.0668,
+      "step": 53700
+    },
+    {
+      "epoch": 3.51390251880929,
+      "grad_norm": 0.8768904805183411,
+      "learning_rate": 8.406026075338489e-05,
+      "loss": 0.077,
+      "step": 53710
+    },
+    {
+      "epoch": 3.514556754988551,
+      "grad_norm": 1.029979944229126,
+      "learning_rate": 8.405353521815382e-05,
+      "loss": 0.0761,
+      "step": 53720
+    },
+    {
+      "epoch": 3.5152109911678115,
+      "grad_norm": 0.7512595057487488,
+      "learning_rate": 8.404680853352662e-05,
+      "loss": 0.0804,
+      "step": 53730
+    },
+    {
+      "epoch": 3.5158652273470725,
+      "grad_norm": 0.8705400824546814,
+      "learning_rate": 8.404008069973035e-05,
+      "loss": 0.0709,
+      "step": 53740
+    },
+    {
+      "epoch": 3.516519463526333,
+      "grad_norm": 0.9716213941574097,
+      "learning_rate": 8.403335171699209e-05,
+      "loss": 0.0821,
+      "step": 53750
+    },
+    {
+      "epoch": 3.5171736997055936,
+      "grad_norm": 0.7278077006340027,
+      "learning_rate": 8.402662158553894e-05,
+      "loss": 0.0783,
+      "step": 53760
+    },
+    {
+      "epoch": 3.5178279358848545,
+      "grad_norm": 0.8177416324615479,
+      "learning_rate": 8.401989030559807e-05,
+      "loss": 0.0697,
+      "step": 53770
+    },
+    {
+      "epoch": 3.518482172064115,
+      "grad_norm": 0.7998149991035461,
+      "learning_rate": 8.401315787739667e-05,
+      "loss": 0.0764,
+      "step": 53780
+    },
+    {
+      "epoch": 3.5191364082433756,
+      "grad_norm": 0.9496130347251892,
+      "learning_rate": 8.400642430116203e-05,
+      "loss": 0.0741,
+      "step": 53790
+    },
+    {
+      "epoch": 3.5197906444226366,
+      "grad_norm": 0.8604884743690491,
+      "learning_rate": 8.399968957712135e-05,
+      "loss": 0.0614,
+      "step": 53800
+    },
+    {
+      "epoch": 3.5204448806018975,
+      "grad_norm": 1.4663547277450562,
+      "learning_rate": 8.3992953705502e-05,
+      "loss": 0.0646,
+      "step": 53810
+    },
+    {
+      "epoch": 3.521099116781158,
+      "grad_norm": 0.8717848658561707,
+      "learning_rate": 8.39862166865313e-05,
+      "loss": 0.0814,
+      "step": 53820
+    },
+    {
+      "epoch": 3.5217533529604186,
+      "grad_norm": 0.8972777724266052,
+      "learning_rate": 8.397947852043666e-05,
+      "loss": 0.0727,
+      "step": 53830
+    },
+    {
+      "epoch": 3.5224075891396796,
+      "grad_norm": 1.0742758512496948,
+      "learning_rate": 8.39727392074455e-05,
+      "loss": 0.0723,
+      "step": 53840
+    },
+    {
+      "epoch": 3.52306182531894,
+      "grad_norm": 0.9516122937202454,
+      "learning_rate": 8.396599874778531e-05,
+      "loss": 0.0652,
+      "step": 53850
+    },
+    {
+      "epoch": 3.5237160614982006,
+      "grad_norm": 0.8731452226638794,
+      "learning_rate": 8.395925714168356e-05,
+      "loss": 0.0755,
+      "step": 53860
+    },
+    {
+      "epoch": 3.5243702976774616,
+      "grad_norm": 0.7560011148452759,
+      "learning_rate": 8.395251438936784e-05,
+      "loss": 0.0697,
+      "step": 53870
+    },
+    {
+      "epoch": 3.5250245338567225,
+      "grad_norm": 0.9268012642860413,
+      "learning_rate": 8.39457704910657e-05,
+      "loss": 0.0666,
+      "step": 53880
+    },
+    {
+      "epoch": 3.525678770035983,
+      "grad_norm": 0.8395716547966003,
+      "learning_rate": 8.393902544700478e-05,
+      "loss": 0.0705,
+      "step": 53890
+    },
+    {
+      "epoch": 3.5263330062152436,
+      "grad_norm": 0.9042119979858398,
+      "learning_rate": 8.393227925741276e-05,
+      "loss": 0.0776,
+      "step": 53900
+    },
+    {
+      "epoch": 3.5269872423945046,
+      "grad_norm": 0.8316322565078735,
+      "learning_rate": 8.392553192251731e-05,
+      "loss": 0.0697,
+      "step": 53910
+    },
+    {
+      "epoch": 3.527641478573765,
+      "grad_norm": 0.9326664209365845,
+      "learning_rate": 8.391878344254618e-05,
+      "loss": 0.0822,
+      "step": 53920
+    },
+    {
+      "epoch": 3.5282957147530256,
+      "grad_norm": 0.7888161540031433,
+      "learning_rate": 8.391203381772716e-05,
+      "loss": 0.071,
+      "step": 53930
+    },
+    {
+      "epoch": 3.5289499509322866,
+      "grad_norm": 1.0052454471588135,
+      "learning_rate": 8.390528304828807e-05,
+      "loss": 0.0663,
+      "step": 53940
+    },
+    {
+      "epoch": 3.529604187111547,
+      "grad_norm": 0.800828754901886,
+      "learning_rate": 8.389853113445676e-05,
+      "loss": 0.069,
+      "step": 53950
+    },
+    {
+      "epoch": 3.530258423290808,
+      "grad_norm": 0.9678196310997009,
+      "learning_rate": 8.38917780764611e-05,
+      "loss": 0.0734,
+      "step": 53960
+    },
+    {
+      "epoch": 3.5309126594700686,
+      "grad_norm": 0.9162442088127136,
+      "learning_rate": 8.388502387452906e-05,
+      "loss": 0.0679,
+      "step": 53970
+    },
+    {
+      "epoch": 3.5315668956493296,
+      "grad_norm": 0.7903563380241394,
+      "learning_rate": 8.38782685288886e-05,
+      "loss": 0.0633,
+      "step": 53980
+    },
+    {
+      "epoch": 3.53222113182859,
+      "grad_norm": 0.8106311559677124,
+      "learning_rate": 8.387151203976772e-05,
+      "loss": 0.0761,
+      "step": 53990
+    },
+    {
+      "epoch": 3.5328753680078506,
+      "grad_norm": 0.8316931128501892,
+      "learning_rate": 8.386475440739447e-05,
+      "loss": 0.0641,
+      "step": 54000
+    },
+    {
+      "epoch": 3.5335296041871116,
+      "grad_norm": 0.8786323666572571,
+      "learning_rate": 8.385799563199697e-05,
+      "loss": 0.0891,
+      "step": 54010
+    },
+    {
+      "epoch": 3.534183840366372,
+      "grad_norm": 0.9065492749214172,
+      "learning_rate": 8.385123571380331e-05,
+      "loss": 0.0743,
+      "step": 54020
+    },
+    {
+      "epoch": 3.534838076545633,
+      "grad_norm": 0.744517982006073,
+      "learning_rate": 8.384447465304166e-05,
+      "loss": 0.0679,
+      "step": 54030
+    },
+    {
+      "epoch": 3.5354923127248936,
+      "grad_norm": 0.8627418875694275,
+      "learning_rate": 8.383771244994023e-05,
+      "loss": 0.0755,
+      "step": 54040
+    },
+    {
+      "epoch": 3.5361465489041546,
+      "grad_norm": 0.8611176609992981,
+      "learning_rate": 8.383094910472728e-05,
+      "loss": 0.0785,
+      "step": 54050
+    },
+    {
+      "epoch": 3.536800785083415,
+      "grad_norm": 0.9356468319892883,
+      "learning_rate": 8.382418461763105e-05,
+      "loss": 0.0814,
+      "step": 54060
+    },
+    {
+      "epoch": 3.5374550212626756,
+      "grad_norm": 0.8152367472648621,
+      "learning_rate": 8.381741898887989e-05,
+      "loss": 0.0728,
+      "step": 54070
+    },
+    {
+      "epoch": 3.5381092574419366,
+      "grad_norm": 0.937050998210907,
+      "learning_rate": 8.381065221870214e-05,
+      "loss": 0.0698,
+      "step": 54080
+    },
+    {
+      "epoch": 3.538763493621197,
+      "grad_norm": 0.8761103749275208,
+      "learning_rate": 8.380388430732623e-05,
+      "loss": 0.0731,
+      "step": 54090
+    },
+    {
+      "epoch": 3.539417729800458,
+      "grad_norm": 0.7785167694091797,
+      "learning_rate": 8.379711525498055e-05,
+      "loss": 0.0663,
+      "step": 54100
+    },
+    {
+      "epoch": 3.5400719659797186,
+      "grad_norm": 0.7450352907180786,
+      "learning_rate": 8.37903450618936e-05,
+      "loss": 0.0678,
+      "step": 54110
+    },
+    {
+      "epoch": 3.5407262021589796,
+      "grad_norm": 0.8352435231208801,
+      "learning_rate": 8.378357372829391e-05,
+      "loss": 0.0692,
+      "step": 54120
+    },
+    {
+      "epoch": 3.54138043833824,
+      "grad_norm": 1.016081690788269,
+      "learning_rate": 8.377680125440997e-05,
+      "loss": 0.0641,
+      "step": 54130
+    },
+    {
+      "epoch": 3.5420346745175006,
+      "grad_norm": 0.9595414400100708,
+      "learning_rate": 8.377002764047042e-05,
+      "loss": 0.0721,
+      "step": 54140
+    },
+    {
+      "epoch": 3.5426889106967616,
+      "grad_norm": 1.0154024362564087,
+      "learning_rate": 8.376325288670386e-05,
+      "loss": 0.0712,
+      "step": 54150
+    },
+    {
+      "epoch": 3.543343146876022,
+      "grad_norm": 0.8880914449691772,
+      "learning_rate": 8.3756476993339e-05,
+      "loss": 0.0839,
+      "step": 54160
+    },
+    {
+      "epoch": 3.543997383055283,
+      "grad_norm": 1.0185810327529907,
+      "learning_rate": 8.374969996060447e-05,
+      "loss": 0.077,
+      "step": 54170
+    },
+    {
+      "epoch": 3.5446516192345436,
+      "grad_norm": 0.900940477848053,
+      "learning_rate": 8.374292178872907e-05,
+      "loss": 0.0718,
+      "step": 54180
+    },
+    {
+      "epoch": 3.5453058554138046,
+      "grad_norm": 0.9907526969909668,
+      "learning_rate": 8.373614247794157e-05,
+      "loss": 0.0685,
+      "step": 54190
+    },
+    {
+      "epoch": 3.545960091593065,
+      "grad_norm": 0.9706956148147583,
+      "learning_rate": 8.37293620284708e-05,
+      "loss": 0.0699,
+      "step": 54200
+    },
+    {
+      "epoch": 3.5466143277723257,
+      "grad_norm": 0.8409591913223267,
+      "learning_rate": 8.372258044054559e-05,
+      "loss": 0.0759,
+      "step": 54210
+    },
+    {
+      "epoch": 3.5472685639515866,
+      "grad_norm": 0.7982200980186462,
+      "learning_rate": 8.371579771439483e-05,
+      "loss": 0.0678,
+      "step": 54220
+    },
+    {
+      "epoch": 3.547922800130847,
+      "grad_norm": 0.8349514603614807,
+      "learning_rate": 8.37090138502475e-05,
+      "loss": 0.0714,
+      "step": 54230
+    },
+    {
+      "epoch": 3.5485770363101077,
+      "grad_norm": 0.7618018984794617,
+      "learning_rate": 8.370222884833254e-05,
+      "loss": 0.0592,
+      "step": 54240
+    },
+    {
+      "epoch": 3.5492312724893686,
+      "grad_norm": 1.4026459455490112,
+      "learning_rate": 8.369544270887897e-05,
+      "loss": 0.0804,
+      "step": 54250
+    },
+    {
+      "epoch": 3.5498855086686296,
+      "grad_norm": 0.8770639896392822,
+      "learning_rate": 8.368865543211584e-05,
+      "loss": 0.0781,
+      "step": 54260
+    },
+    {
+      "epoch": 3.55053974484789,
+      "grad_norm": 1.1316838264465332,
+      "learning_rate": 8.368186701827223e-05,
+      "loss": 0.0718,
+      "step": 54270
+    },
+    {
+      "epoch": 3.5511939810271507,
+      "grad_norm": 0.9863431453704834,
+      "learning_rate": 8.367507746757728e-05,
+      "loss": 0.0842,
+      "step": 54280
+    },
+    {
+      "epoch": 3.5518482172064116,
+      "grad_norm": 0.7806870341300964,
+      "learning_rate": 8.366828678026016e-05,
+      "loss": 0.0717,
+      "step": 54290
+    },
+    {
+      "epoch": 3.552502453385672,
+      "grad_norm": 0.8638490438461304,
+      "learning_rate": 8.366149495655004e-05,
+      "loss": 0.073,
+      "step": 54300
+    },
+    {
+      "epoch": 3.5531566895649327,
+      "grad_norm": 0.9356234669685364,
+      "learning_rate": 8.36547019966762e-05,
+      "loss": 0.0662,
+      "step": 54310
+    },
+    {
+      "epoch": 3.5538109257441937,
+      "grad_norm": 0.7376917600631714,
+      "learning_rate": 8.36479079008679e-05,
+      "loss": 0.0667,
+      "step": 54320
+    },
+    {
+      "epoch": 3.5544651619234546,
+      "grad_norm": 0.7717798948287964,
+      "learning_rate": 8.364111266935446e-05,
+      "loss": 0.0645,
+      "step": 54330
+    },
+    {
+      "epoch": 3.555119398102715,
+      "grad_norm": 0.8551239371299744,
+      "learning_rate": 8.363431630236525e-05,
+      "loss": 0.0726,
+      "step": 54340
+    },
+    {
+      "epoch": 3.5557736342819757,
+      "grad_norm": 0.7663902640342712,
+      "learning_rate": 8.362751880012965e-05,
+      "loss": 0.076,
+      "step": 54350
+    },
+    {
+      "epoch": 3.5564278704612367,
+      "grad_norm": 1.0010058879852295,
+      "learning_rate": 8.362072016287709e-05,
+      "loss": 0.0705,
+      "step": 54360
+    },
+    {
+      "epoch": 3.557082106640497,
+      "grad_norm": 0.7997660040855408,
+      "learning_rate": 8.361392039083706e-05,
+      "loss": 0.0669,
+      "step": 54370
+    },
+    {
+      "epoch": 3.5577363428197577,
+      "grad_norm": 0.6869263648986816,
+      "learning_rate": 8.360711948423906e-05,
+      "loss": 0.0673,
+      "step": 54380
+    },
+    {
+      "epoch": 3.5583905789990187,
+      "grad_norm": 1.1054303646087646,
+      "learning_rate": 8.360031744331264e-05,
+      "loss": 0.0772,
+      "step": 54390
+    },
+    {
+      "epoch": 3.559044815178279,
+      "grad_norm": 0.9529674649238586,
+      "learning_rate": 8.359351426828739e-05,
+      "loss": 0.067,
+      "step": 54400
+    },
+    {
+      "epoch": 3.55969905135754,
+      "grad_norm": 0.6143543124198914,
+      "learning_rate": 8.358670995939293e-05,
+      "loss": 0.0758,
+      "step": 54410
+    },
+    {
+      "epoch": 3.5603532875368007,
+      "grad_norm": 0.9302859306335449,
+      "learning_rate": 8.357990451685892e-05,
+      "loss": 0.0653,
+      "step": 54420
+    },
+    {
+      "epoch": 3.5610075237160617,
+      "grad_norm": 0.8283950090408325,
+      "learning_rate": 8.357309794091507e-05,
+      "loss": 0.0701,
+      "step": 54430
+    },
+    {
+      "epoch": 3.561661759895322,
+      "grad_norm": 0.7890059947967529,
+      "learning_rate": 8.356629023179111e-05,
+      "loss": 0.0711,
+      "step": 54440
+    },
+    {
+      "epoch": 3.5623159960745827,
+      "grad_norm": 0.8997254371643066,
+      "learning_rate": 8.355948138971683e-05,
+      "loss": 0.0659,
+      "step": 54450
+    },
+    {
+      "epoch": 3.5629702322538437,
+      "grad_norm": 1.2762157917022705,
+      "learning_rate": 8.355267141492205e-05,
+      "loss": 0.0692,
+      "step": 54460
+    },
+    {
+      "epoch": 3.563624468433104,
+      "grad_norm": 0.8240480422973633,
+      "learning_rate": 8.354586030763659e-05,
+      "loss": 0.064,
+      "step": 54470
+    },
+    {
+      "epoch": 3.564278704612365,
+      "grad_norm": 0.8595862984657288,
+      "learning_rate": 8.353904806809039e-05,
+      "loss": 0.0612,
+      "step": 54480
+    },
+    {
+      "epoch": 3.5649329407916257,
+      "grad_norm": 1.0054336786270142,
+      "learning_rate": 8.353223469651335e-05,
+      "loss": 0.075,
+      "step": 54490
+    },
+    {
+      "epoch": 3.5655871769708867,
+      "grad_norm": 0.9695059061050415,
+      "learning_rate": 8.352542019313544e-05,
+      "loss": 0.0723,
+      "step": 54500
+    },
+    {
+      "epoch": 3.566241413150147,
+      "grad_norm": 0.8478056192398071,
+      "learning_rate": 8.351860455818667e-05,
+      "loss": 0.0661,
+      "step": 54510
+    },
+    {
+      "epoch": 3.5668956493294077,
+      "grad_norm": 0.7672955989837646,
+      "learning_rate": 8.35117877918971e-05,
+      "loss": 0.0677,
+      "step": 54520
+    },
+    {
+      "epoch": 3.5675498855086687,
+      "grad_norm": 0.8970052003860474,
+      "learning_rate": 8.350496989449681e-05,
+      "loss": 0.0616,
+      "step": 54530
+    },
+    {
+      "epoch": 3.568204121687929,
+      "grad_norm": 0.7992386221885681,
+      "learning_rate": 8.34981508662159e-05,
+      "loss": 0.0715,
+      "step": 54540
+    },
+    {
+      "epoch": 3.56885835786719,
+      "grad_norm": 0.8448268175125122,
+      "learning_rate": 8.349133070728456e-05,
+      "loss": 0.0611,
+      "step": 54550
+    },
+    {
+      "epoch": 3.5695125940464507,
+      "grad_norm": 0.821503758430481,
+      "learning_rate": 8.348450941793298e-05,
+      "loss": 0.0661,
+      "step": 54560
+    },
+    {
+      "epoch": 3.5701668302257117,
+      "grad_norm": 0.9001857042312622,
+      "learning_rate": 8.347768699839139e-05,
+      "loss": 0.0688,
+      "step": 54570
+    },
+    {
+      "epoch": 3.570821066404972,
+      "grad_norm": 0.891309380531311,
+      "learning_rate": 8.347086344889006e-05,
+      "loss": 0.0628,
+      "step": 54580
+    },
+    {
+      "epoch": 3.5714753025842327,
+      "grad_norm": 0.8133125305175781,
+      "learning_rate": 8.34640387696593e-05,
+      "loss": 0.0648,
+      "step": 54590
+    },
+    {
+      "epoch": 3.5721295387634937,
+      "grad_norm": 1.0169897079467773,
+      "learning_rate": 8.345721296092947e-05,
+      "loss": 0.0794,
+      "step": 54600
+    },
+    {
+      "epoch": 3.5727837749427542,
+      "grad_norm": 0.8700054883956909,
+      "learning_rate": 8.345038602293097e-05,
+      "loss": 0.0629,
+      "step": 54610
+    },
+    {
+      "epoch": 3.573438011122015,
+      "grad_norm": 0.9666095972061157,
+      "learning_rate": 8.344355795589421e-05,
+      "loss": 0.0728,
+      "step": 54620
+    },
+    {
+      "epoch": 3.5740922473012757,
+      "grad_norm": 0.785966157913208,
+      "learning_rate": 8.343672876004965e-05,
+      "loss": 0.0687,
+      "step": 54630
+    },
+    {
+      "epoch": 3.5747464834805367,
+      "grad_norm": 0.988416314125061,
+      "learning_rate": 8.342989843562782e-05,
+      "loss": 0.0689,
+      "step": 54640
+    },
+    {
+      "epoch": 3.575400719659797,
+      "grad_norm": 0.9573351740837097,
+      "learning_rate": 8.342306698285923e-05,
+      "loss": 0.0716,
+      "step": 54650
+    },
+    {
+      "epoch": 3.5760549558390577,
+      "grad_norm": 0.8525318503379822,
+      "learning_rate": 8.341623440197448e-05,
+      "loss": 0.0748,
+      "step": 54660
+    },
+    {
+      "epoch": 3.5767091920183187,
+      "grad_norm": 1.0283527374267578,
+      "learning_rate": 8.340940069320418e-05,
+      "loss": 0.0686,
+      "step": 54670
+    },
+    {
+      "epoch": 3.5773634281975792,
+      "grad_norm": 0.908446192741394,
+      "learning_rate": 8.3402565856779e-05,
+      "loss": 0.0638,
+      "step": 54680
+    },
+    {
+      "epoch": 3.5780176643768398,
+      "grad_norm": 0.984634518623352,
+      "learning_rate": 8.339572989292961e-05,
+      "loss": 0.0744,
+      "step": 54690
+    },
+    {
+      "epoch": 3.5786719005561007,
+      "grad_norm": 1.0757420063018799,
+      "learning_rate": 8.338889280188674e-05,
+      "loss": 0.0748,
+      "step": 54700
+    },
+    {
+      "epoch": 3.5793261367353617,
+      "grad_norm": 0.6891494393348694,
+      "learning_rate": 8.338205458388118e-05,
+      "loss": 0.0737,
+      "step": 54710
+    },
+    {
+      "epoch": 3.5799803729146222,
+      "grad_norm": 0.9240961074829102,
+      "learning_rate": 8.337521523914375e-05,
+      "loss": 0.0848,
+      "step": 54720
+    },
+    {
+      "epoch": 3.5806346090938828,
+      "grad_norm": 0.8011013269424438,
+      "learning_rate": 8.336837476790526e-05,
+      "loss": 0.063,
+      "step": 54730
+    },
+    {
+      "epoch": 3.5812888452731437,
+      "grad_norm": 0.9099512696266174,
+      "learning_rate": 8.336153317039662e-05,
+      "loss": 0.0673,
+      "step": 54740
+    },
+    {
+      "epoch": 3.5819430814524043,
+      "grad_norm": 0.8403385281562805,
+      "learning_rate": 8.335469044684872e-05,
+      "loss": 0.072,
+      "step": 54750
+    },
+    {
+      "epoch": 3.582597317631665,
+      "grad_norm": 0.8340772986412048,
+      "learning_rate": 8.334784659749255e-05,
+      "loss": 0.0665,
+      "step": 54760
+    },
+    {
+      "epoch": 3.5832515538109257,
+      "grad_norm": 0.8868764042854309,
+      "learning_rate": 8.334100162255912e-05,
+      "loss": 0.0762,
+      "step": 54770
+    },
+    {
+      "epoch": 3.5839057899901867,
+      "grad_norm": 0.8480395078659058,
+      "learning_rate": 8.33341555222794e-05,
+      "loss": 0.0677,
+      "step": 54780
+    },
+    {
+      "epoch": 3.5845600261694472,
+      "grad_norm": 1.0152699947357178,
+      "learning_rate": 8.332730829688456e-05,
+      "loss": 0.0769,
+      "step": 54790
+    },
+    {
+      "epoch": 3.5852142623487078,
+      "grad_norm": 1.082312822341919,
+      "learning_rate": 8.332045994660563e-05,
+      "loss": 0.0708,
+      "step": 54800
+    },
+    {
+      "epoch": 3.5858684985279687,
+      "grad_norm": 0.8725629448890686,
+      "learning_rate": 8.33136104716738e-05,
+      "loss": 0.0608,
+      "step": 54810
+    },
+    {
+      "epoch": 3.5865227347072293,
+      "grad_norm": 0.7891134023666382,
+      "learning_rate": 8.330675987232024e-05,
+      "loss": 0.0615,
+      "step": 54820
+    },
+    {
+      "epoch": 3.58717697088649,
+      "grad_norm": 1.0689077377319336,
+      "learning_rate": 8.32999081487762e-05,
+      "loss": 0.0745,
+      "step": 54830
+    },
+    {
+      "epoch": 3.5878312070657508,
+      "grad_norm": 0.77810138463974,
+      "learning_rate": 8.329305530127291e-05,
+      "loss": 0.0666,
+      "step": 54840
+    },
+    {
+      "epoch": 3.5884854432450113,
+      "grad_norm": 1.0314850807189941,
+      "learning_rate": 8.32862013300417e-05,
+      "loss": 0.0775,
+      "step": 54850
+    },
+    {
+      "epoch": 3.5891396794242723,
+      "grad_norm": 0.9134453535079956,
+      "learning_rate": 8.32793462353139e-05,
+      "loss": 0.0732,
+      "step": 54860
+    },
+    {
+      "epoch": 3.589793915603533,
+      "grad_norm": 0.786392331123352,
+      "learning_rate": 8.32724900173209e-05,
+      "loss": 0.0733,
+      "step": 54870
+    },
+    {
+      "epoch": 3.5904481517827938,
+      "grad_norm": 0.7125436663627625,
+      "learning_rate": 8.326563267629408e-05,
+      "loss": 0.0712,
+      "step": 54880
+    },
+    {
+      "epoch": 3.5911023879620543,
+      "grad_norm": 0.9112322330474854,
+      "learning_rate": 8.325877421246491e-05,
+      "loss": 0.0807,
+      "step": 54890
+    },
+    {
+      "epoch": 3.591756624141315,
+      "grad_norm": 0.8013171553611755,
+      "learning_rate": 8.325191462606491e-05,
+      "loss": 0.0722,
+      "step": 54900
+    },
+    {
+      "epoch": 3.5924108603205758,
+      "grad_norm": 0.7827738523483276,
+      "learning_rate": 8.324505391732557e-05,
+      "loss": 0.0701,
+      "step": 54910
+    },
+    {
+      "epoch": 3.5930650964998363,
+      "grad_norm": 0.8412222266197205,
+      "learning_rate": 8.323819208647847e-05,
+      "loss": 0.0771,
+      "step": 54920
+    },
+    {
+      "epoch": 3.5937193326790973,
+      "grad_norm": 0.9452791213989258,
+      "learning_rate": 8.323132913375522e-05,
+      "loss": 0.0665,
+      "step": 54930
+    },
+    {
+      "epoch": 3.594373568858358,
+      "grad_norm": 0.7313148975372314,
+      "learning_rate": 8.322446505938746e-05,
+      "loss": 0.0686,
+      "step": 54940
+    },
+    {
+      "epoch": 3.5950278050376188,
+      "grad_norm": 0.7917100787162781,
+      "learning_rate": 8.321759986360687e-05,
+      "loss": 0.0706,
+      "step": 54950
+    },
+    {
+      "epoch": 3.5956820412168793,
+      "grad_norm": 1.0193794965744019,
+      "learning_rate": 8.321073354664516e-05,
+      "loss": 0.0816,
+      "step": 54960
+    },
+    {
+      "epoch": 3.59633627739614,
+      "grad_norm": 0.9962078332901001,
+      "learning_rate": 8.32038661087341e-05,
+      "loss": 0.0657,
+      "step": 54970
+    },
+    {
+      "epoch": 3.596990513575401,
+      "grad_norm": 0.8794001340866089,
+      "learning_rate": 8.319699755010549e-05,
+      "loss": 0.076,
+      "step": 54980
+    },
+    {
+      "epoch": 3.5976447497546613,
+      "grad_norm": 0.8249982595443726,
+      "learning_rate": 8.319012787099115e-05,
+      "loss": 0.0815,
+      "step": 54990
+    },
+    {
+      "epoch": 3.5982989859339223,
+      "grad_norm": 0.9822542667388916,
+      "learning_rate": 8.318325707162293e-05,
+      "loss": 0.0676,
+      "step": 55000
+    },
+    {
+      "epoch": 3.598953222113183,
+      "grad_norm": 0.8483203053474426,
+      "learning_rate": 8.317638515223277e-05,
+      "loss": 0.0686,
+      "step": 55010
+    },
+    {
+      "epoch": 3.5996074582924438,
+      "grad_norm": 0.8871238827705383,
+      "learning_rate": 8.31695121130526e-05,
+      "loss": 0.0722,
+      "step": 55020
+    },
+    {
+      "epoch": 3.6002616944717043,
+      "grad_norm": 1.047977089881897,
+      "learning_rate": 8.31626379543144e-05,
+      "loss": 0.0836,
+      "step": 55030
+    },
+    {
+      "epoch": 3.600915930650965,
+      "grad_norm": 0.9134694933891296,
+      "learning_rate": 8.31557626762502e-05,
+      "loss": 0.0733,
+      "step": 55040
+    },
+    {
+      "epoch": 3.601570166830226,
+      "grad_norm": 0.9001691341400146,
+      "learning_rate": 8.314888627909208e-05,
+      "loss": 0.0687,
+      "step": 55050
+    },
+    {
+      "epoch": 3.6022244030094863,
+      "grad_norm": 0.8247069716453552,
+      "learning_rate": 8.31420087630721e-05,
+      "loss": 0.0676,
+      "step": 55060
+    },
+    {
+      "epoch": 3.6028786391887473,
+      "grad_norm": 0.8471789360046387,
+      "learning_rate": 8.313513012842238e-05,
+      "loss": 0.0694,
+      "step": 55070
+    },
+    {
+      "epoch": 3.603532875368008,
+      "grad_norm": 0.7431899905204773,
+      "learning_rate": 8.312825037537513e-05,
+      "loss": 0.0741,
+      "step": 55080
+    },
+    {
+      "epoch": 3.604187111547269,
+      "grad_norm": 0.9894888997077942,
+      "learning_rate": 8.312136950416256e-05,
+      "loss": 0.0754,
+      "step": 55090
+    },
+    {
+      "epoch": 3.6048413477265293,
+      "grad_norm": 0.8959627747535706,
+      "learning_rate": 8.311448751501689e-05,
+      "loss": 0.0798,
+      "step": 55100
+    },
+    {
+      "epoch": 3.60549558390579,
+      "grad_norm": 0.7383559942245483,
+      "learning_rate": 8.310760440817043e-05,
+      "loss": 0.0726,
+      "step": 55110
+    },
+    {
+      "epoch": 3.606149820085051,
+      "grad_norm": 0.8835324645042419,
+      "learning_rate": 8.31007201838555e-05,
+      "loss": 0.0751,
+      "step": 55120
+    },
+    {
+      "epoch": 3.6068040562643113,
+      "grad_norm": 0.8275827169418335,
+      "learning_rate": 8.309383484230446e-05,
+      "loss": 0.064,
+      "step": 55130
+    },
+    {
+      "epoch": 3.607458292443572,
+      "grad_norm": 0.8732197880744934,
+      "learning_rate": 8.308694838374969e-05,
+      "loss": 0.0656,
+      "step": 55140
+    },
+    {
+      "epoch": 3.608112528622833,
+      "grad_norm": 1.2900716066360474,
+      "learning_rate": 8.308006080842362e-05,
+      "loss": 0.0748,
+      "step": 55150
+    },
+    {
+      "epoch": 3.608766764802094,
+      "grad_norm": 0.8810641169548035,
+      "learning_rate": 8.307317211655877e-05,
+      "loss": 0.07,
+      "step": 55160
+    },
+    {
+      "epoch": 3.6094210009813543,
+      "grad_norm": 0.9505282640457153,
+      "learning_rate": 8.30662823083876e-05,
+      "loss": 0.0701,
+      "step": 55170
+    },
+    {
+      "epoch": 3.610075237160615,
+      "grad_norm": 1.0817748308181763,
+      "learning_rate": 8.30593913841427e-05,
+      "loss": 0.0761,
+      "step": 55180
+    },
+    {
+      "epoch": 3.610729473339876,
+      "grad_norm": 0.8153062462806702,
+      "learning_rate": 8.305249934405664e-05,
+      "loss": 0.0643,
+      "step": 55190
+    },
+    {
+      "epoch": 3.6113837095191363,
+      "grad_norm": 0.8734539747238159,
+      "learning_rate": 8.304560618836204e-05,
+      "loss": 0.0721,
+      "step": 55200
+    },
+    {
+      "epoch": 3.612037945698397,
+      "grad_norm": 0.8349213004112244,
+      "learning_rate": 8.303871191729156e-05,
+      "loss": 0.0701,
+      "step": 55210
+    },
+    {
+      "epoch": 3.612692181877658,
+      "grad_norm": 0.8556681871414185,
+      "learning_rate": 8.303181653107791e-05,
+      "loss": 0.064,
+      "step": 55220
+    },
+    {
+      "epoch": 3.613346418056919,
+      "grad_norm": 1.0858343839645386,
+      "learning_rate": 8.302492002995383e-05,
+      "loss": 0.0683,
+      "step": 55230
+    },
+    {
+      "epoch": 3.6140006542361793,
+      "grad_norm": 0.6986839175224304,
+      "learning_rate": 8.301802241415209e-05,
+      "loss": 0.0651,
+      "step": 55240
+    },
+    {
+      "epoch": 3.61465489041544,
+      "grad_norm": 0.9676486849784851,
+      "learning_rate": 8.301112368390548e-05,
+      "loss": 0.067,
+      "step": 55250
+    },
+    {
+      "epoch": 3.615309126594701,
+      "grad_norm": 1.1534702777862549,
+      "learning_rate": 8.300422383944688e-05,
+      "loss": 0.0665,
+      "step": 55260
+    },
+    {
+      "epoch": 3.6159633627739614,
+      "grad_norm": 0.9202556014060974,
+      "learning_rate": 8.299732288100918e-05,
+      "loss": 0.0648,
+      "step": 55270
+    },
+    {
+      "epoch": 3.616617598953222,
+      "grad_norm": 0.9827196598052979,
+      "learning_rate": 8.299042080882528e-05,
+      "loss": 0.0777,
+      "step": 55280
+    },
+    {
+      "epoch": 3.617271835132483,
+      "grad_norm": 0.9134148359298706,
+      "learning_rate": 8.298351762312816e-05,
+      "loss": 0.0693,
+      "step": 55290
+    },
+    {
+      "epoch": 3.6179260713117434,
+      "grad_norm": 0.7244698405265808,
+      "learning_rate": 8.297661332415083e-05,
+      "loss": 0.0672,
+      "step": 55300
+    },
+    {
+      "epoch": 3.6185803074910043,
+      "grad_norm": 0.9606476426124573,
+      "learning_rate": 8.296970791212631e-05,
+      "loss": 0.0731,
+      "step": 55310
+    },
+    {
+      "epoch": 3.619234543670265,
+      "grad_norm": 0.9693164825439453,
+      "learning_rate": 8.296280138728768e-05,
+      "loss": 0.07,
+      "step": 55320
+    },
+    {
+      "epoch": 3.619888779849526,
+      "grad_norm": 0.8021335005760193,
+      "learning_rate": 8.295589374986804e-05,
+      "loss": 0.07,
+      "step": 55330
+    },
+    {
+      "epoch": 3.6205430160287864,
+      "grad_norm": 1.0578341484069824,
+      "learning_rate": 8.294898500010056e-05,
+      "loss": 0.0635,
+      "step": 55340
+    },
+    {
+      "epoch": 3.621197252208047,
+      "grad_norm": 0.8854939341545105,
+      "learning_rate": 8.294207513821845e-05,
+      "loss": 0.0651,
+      "step": 55350
+    },
+    {
+      "epoch": 3.621851488387308,
+      "grad_norm": 0.8481228351593018,
+      "learning_rate": 8.293516416445488e-05,
+      "loss": 0.0743,
+      "step": 55360
+    },
+    {
+      "epoch": 3.6225057245665684,
+      "grad_norm": 0.8567721247673035,
+      "learning_rate": 8.292825207904316e-05,
+      "loss": 0.0671,
+      "step": 55370
+    },
+    {
+      "epoch": 3.6231599607458294,
+      "grad_norm": 0.7755191922187805,
+      "learning_rate": 8.292133888221659e-05,
+      "loss": 0.0683,
+      "step": 55380
+    },
+    {
+      "epoch": 3.62381419692509,
+      "grad_norm": 1.062867522239685,
+      "learning_rate": 8.291442457420846e-05,
+      "loss": 0.0778,
+      "step": 55390
+    },
+    {
+      "epoch": 3.624468433104351,
+      "grad_norm": 0.8446351885795593,
+      "learning_rate": 8.290750915525219e-05,
+      "loss": 0.0645,
+      "step": 55400
+    },
+    {
+      "epoch": 3.6251226692836114,
+      "grad_norm": 0.8222762942314148,
+      "learning_rate": 8.290059262558119e-05,
+      "loss": 0.0755,
+      "step": 55410
+    },
+    {
+      "epoch": 3.625776905462872,
+      "grad_norm": 0.9399179220199585,
+      "learning_rate": 8.28936749854289e-05,
+      "loss": 0.0672,
+      "step": 55420
+    },
+    {
+      "epoch": 3.626431141642133,
+      "grad_norm": 0.7565219402313232,
+      "learning_rate": 8.288675623502881e-05,
+      "loss": 0.065,
+      "step": 55430
+    },
+    {
+      "epoch": 3.6270853778213934,
+      "grad_norm": 0.8849060535430908,
+      "learning_rate": 8.287983637461447e-05,
+      "loss": 0.0748,
+      "step": 55440
+    },
+    {
+      "epoch": 3.6277396140006544,
+      "grad_norm": 0.8676406145095825,
+      "learning_rate": 8.28729154044194e-05,
+      "loss": 0.0742,
+      "step": 55450
+    },
+    {
+      "epoch": 3.628393850179915,
+      "grad_norm": 0.7463305592536926,
+      "learning_rate": 8.286599332467722e-05,
+      "loss": 0.0722,
+      "step": 55460
+    },
+    {
+      "epoch": 3.629048086359176,
+      "grad_norm": 0.9079608917236328,
+      "learning_rate": 8.285907013562158e-05,
+      "loss": 0.0723,
+      "step": 55470
+    },
+    {
+      "epoch": 3.6297023225384364,
+      "grad_norm": 0.9543100595474243,
+      "learning_rate": 8.285214583748616e-05,
+      "loss": 0.0726,
+      "step": 55480
+    },
+    {
+      "epoch": 3.630356558717697,
+      "grad_norm": 0.9939130544662476,
+      "learning_rate": 8.284522043050463e-05,
+      "loss": 0.0679,
+      "step": 55490
+    },
+    {
+      "epoch": 3.631010794896958,
+      "grad_norm": 0.9271246194839478,
+      "learning_rate": 8.28382939149108e-05,
+      "loss": 0.0642,
+      "step": 55500
+    },
+    {
+      "epoch": 3.6316650310762184,
+      "grad_norm": 0.9283300042152405,
+      "learning_rate": 8.283136629093841e-05,
+      "loss": 0.067,
+      "step": 55510
+    },
+    {
+      "epoch": 3.6323192672554794,
+      "grad_norm": 0.9301807284355164,
+      "learning_rate": 8.28244375588213e-05,
+      "loss": 0.07,
+      "step": 55520
+    },
+    {
+      "epoch": 3.63297350343474,
+      "grad_norm": 0.9366194605827332,
+      "learning_rate": 8.281750771879335e-05,
+      "loss": 0.0707,
+      "step": 55530
+    },
+    {
+      "epoch": 3.633627739614001,
+      "grad_norm": 0.9586436748504639,
+      "learning_rate": 8.281057677108844e-05,
+      "loss": 0.0641,
+      "step": 55540
+    },
+    {
+      "epoch": 3.6342819757932614,
+      "grad_norm": 0.8174638748168945,
+      "learning_rate": 8.280364471594052e-05,
+      "loss": 0.0686,
+      "step": 55550
+    },
+    {
+      "epoch": 3.634936211972522,
+      "grad_norm": 0.9204336404800415,
+      "learning_rate": 8.279671155358355e-05,
+      "loss": 0.066,
+      "step": 55560
+    },
+    {
+      "epoch": 3.635590448151783,
+      "grad_norm": 0.7173357009887695,
+      "learning_rate": 8.278977728425157e-05,
+      "loss": 0.0668,
+      "step": 55570
+    },
+    {
+      "epoch": 3.6362446843310434,
+      "grad_norm": 0.8251804113388062,
+      "learning_rate": 8.27828419081786e-05,
+      "loss": 0.0689,
+      "step": 55580
+    },
+    {
+      "epoch": 3.636898920510304,
+      "grad_norm": 0.796325147151947,
+      "learning_rate": 8.277590542559875e-05,
+      "loss": 0.0657,
+      "step": 55590
+    },
+    {
+      "epoch": 3.637553156689565,
+      "grad_norm": 0.8071014881134033,
+      "learning_rate": 8.276896783674612e-05,
+      "loss": 0.0692,
+      "step": 55600
+    },
+    {
+      "epoch": 3.638207392868826,
+      "grad_norm": 0.8185661435127258,
+      "learning_rate": 8.27620291418549e-05,
+      "loss": 0.0657,
+      "step": 55610
+    },
+    {
+      "epoch": 3.6388616290480864,
+      "grad_norm": 0.800986111164093,
+      "learning_rate": 8.275508934115927e-05,
+      "loss": 0.0719,
+      "step": 55620
+    },
+    {
+      "epoch": 3.639515865227347,
+      "grad_norm": 0.9986090660095215,
+      "learning_rate": 8.274814843489346e-05,
+      "loss": 0.0764,
+      "step": 55630
+    },
+    {
+      "epoch": 3.640170101406608,
+      "grad_norm": 0.688541829586029,
+      "learning_rate": 8.274120642329178e-05,
+      "loss": 0.0721,
+      "step": 55640
+    },
+    {
+      "epoch": 3.6408243375858684,
+      "grad_norm": 1.0835293531417847,
+      "learning_rate": 8.273426330658849e-05,
+      "loss": 0.0819,
+      "step": 55650
+    },
+    {
+      "epoch": 3.641478573765129,
+      "grad_norm": 0.7599365711212158,
+      "learning_rate": 8.272731908501798e-05,
+      "loss": 0.075,
+      "step": 55660
+    },
+    {
+      "epoch": 3.64213280994439,
+      "grad_norm": 1.0994549989700317,
+      "learning_rate": 8.272037375881461e-05,
+      "loss": 0.0601,
+      "step": 55670
+    },
+    {
+      "epoch": 3.642787046123651,
+      "grad_norm": 0.8615381121635437,
+      "learning_rate": 8.27134273282128e-05,
+      "loss": 0.0733,
+      "step": 55680
+    },
+    {
+      "epoch": 3.6434412823029114,
+      "grad_norm": 1.0048024654388428,
+      "learning_rate": 8.270647979344706e-05,
+      "loss": 0.0808,
+      "step": 55690
+    },
+    {
+      "epoch": 3.644095518482172,
+      "grad_norm": 1.1392261981964111,
+      "learning_rate": 8.269953115475183e-05,
+      "loss": 0.0863,
+      "step": 55700
+    },
+    {
+      "epoch": 3.644749754661433,
+      "grad_norm": 1.387350082397461,
+      "learning_rate": 8.269258141236167e-05,
+      "loss": 0.083,
+      "step": 55710
+    },
+    {
+      "epoch": 3.6454039908406934,
+      "grad_norm": 0.8560718894004822,
+      "learning_rate": 8.268563056651115e-05,
+      "loss": 0.0751,
+      "step": 55720
+    },
+    {
+      "epoch": 3.646058227019954,
+      "grad_norm": 0.892548143863678,
+      "learning_rate": 8.267867861743488e-05,
+      "loss": 0.0709,
+      "step": 55730
+    },
+    {
+      "epoch": 3.646712463199215,
+      "grad_norm": 0.8326752185821533,
+      "learning_rate": 8.267172556536748e-05,
+      "loss": 0.0713,
+      "step": 55740
+    },
+    {
+      "epoch": 3.647366699378476,
+      "grad_norm": 0.7968476414680481,
+      "learning_rate": 8.26647714105437e-05,
+      "loss": 0.0821,
+      "step": 55750
+    },
+    {
+      "epoch": 3.6480209355577364,
+      "grad_norm": 0.8511175513267517,
+      "learning_rate": 8.265781615319818e-05,
+      "loss": 0.0745,
+      "step": 55760
+    },
+    {
+      "epoch": 3.648675171736997,
+      "grad_norm": 0.9187440872192383,
+      "learning_rate": 8.265085979356573e-05,
+      "loss": 0.0699,
+      "step": 55770
+    },
+    {
+      "epoch": 3.649329407916258,
+      "grad_norm": 0.8875986933708191,
+      "learning_rate": 8.264390233188113e-05,
+      "loss": 0.0783,
+      "step": 55780
+    },
+    {
+      "epoch": 3.6499836440955185,
+      "grad_norm": 0.9916920065879822,
+      "learning_rate": 8.263694376837923e-05,
+      "loss": 0.0637,
+      "step": 55790
+    },
+    {
+      "epoch": 3.650637880274779,
+      "grad_norm": 0.9589375853538513,
+      "learning_rate": 8.262998410329486e-05,
+      "loss": 0.0643,
+      "step": 55800
+    },
+    {
+      "epoch": 3.65129211645404,
+      "grad_norm": 0.9402400851249695,
+      "learning_rate": 8.262302333686296e-05,
+      "loss": 0.0665,
+      "step": 55810
+    },
+    {
+      "epoch": 3.6519463526333005,
+      "grad_norm": 0.9499921798706055,
+      "learning_rate": 8.261606146931846e-05,
+      "loss": 0.08,
+      "step": 55820
+    },
+    {
+      "epoch": 3.6526005888125614,
+      "grad_norm": 1.1458688974380493,
+      "learning_rate": 8.260909850089636e-05,
+      "loss": 0.0697,
+      "step": 55830
+    },
+    {
+      "epoch": 3.653254824991822,
+      "grad_norm": 0.8103237152099609,
+      "learning_rate": 8.260213443183167e-05,
+      "loss": 0.0725,
+      "step": 55840
+    },
+    {
+      "epoch": 3.653909061171083,
+      "grad_norm": 0.7244288921356201,
+      "learning_rate": 8.259516926235942e-05,
+      "loss": 0.0653,
+      "step": 55850
+    },
+    {
+      "epoch": 3.6545632973503435,
+      "grad_norm": 0.8501254320144653,
+      "learning_rate": 8.258820299271475e-05,
+      "loss": 0.0662,
+      "step": 55860
+    },
+    {
+      "epoch": 3.655217533529604,
+      "grad_norm": 0.9807097911834717,
+      "learning_rate": 8.258123562313274e-05,
+      "loss": 0.0734,
+      "step": 55870
+    },
+    {
+      "epoch": 3.655871769708865,
+      "grad_norm": 0.9171924591064453,
+      "learning_rate": 8.257426715384859e-05,
+      "loss": 0.0711,
+      "step": 55880
+    },
+    {
+      "epoch": 3.6565260058881255,
+      "grad_norm": 0.7962984442710876,
+      "learning_rate": 8.256729758509748e-05,
+      "loss": 0.0753,
+      "step": 55890
+    },
+    {
+      "epoch": 3.6571802420673865,
+      "grad_norm": 0.7744097709655762,
+      "learning_rate": 8.256032691711469e-05,
+      "loss": 0.075,
+      "step": 55900
+    },
+    {
+      "epoch": 3.657834478246647,
+      "grad_norm": 0.8903494477272034,
+      "learning_rate": 8.255335515013545e-05,
+      "loss": 0.0681,
+      "step": 55910
+    },
+    {
+      "epoch": 3.658488714425908,
+      "grad_norm": 0.8612160682678223,
+      "learning_rate": 8.25463822843951e-05,
+      "loss": 0.0739,
+      "step": 55920
+    },
+    {
+      "epoch": 3.6591429506051685,
+      "grad_norm": 1.115964651107788,
+      "learning_rate": 8.253940832012901e-05,
+      "loss": 0.0776,
+      "step": 55930
+    },
+    {
+      "epoch": 3.659797186784429,
+      "grad_norm": 0.90339195728302,
+      "learning_rate": 8.253243325757255e-05,
+      "loss": 0.0724,
+      "step": 55940
+    },
+    {
+      "epoch": 3.66045142296369,
+      "grad_norm": 0.8702420592308044,
+      "learning_rate": 8.252545709696114e-05,
+      "loss": 0.0617,
+      "step": 55950
+    },
+    {
+      "epoch": 3.6611056591429505,
+      "grad_norm": 0.7411156892776489,
+      "learning_rate": 8.251847983853025e-05,
+      "loss": 0.0724,
+      "step": 55960
+    },
+    {
+      "epoch": 3.6617598953222115,
+      "grad_norm": 0.856332540512085,
+      "learning_rate": 8.251150148251538e-05,
+      "loss": 0.0616,
+      "step": 55970
+    },
+    {
+      "epoch": 3.662414131501472,
+      "grad_norm": 0.808392345905304,
+      "learning_rate": 8.250452202915209e-05,
+      "loss": 0.0727,
+      "step": 55980
+    },
+    {
+      "epoch": 3.663068367680733,
+      "grad_norm": 0.7766302227973938,
+      "learning_rate": 8.249754147867592e-05,
+      "loss": 0.0691,
+      "step": 55990
+    },
+    {
+      "epoch": 3.6637226038599935,
+      "grad_norm": 0.782347559928894,
+      "learning_rate": 8.24905598313225e-05,
+      "loss": 0.0603,
+      "step": 56000
+    },
+    {
+      "epoch": 3.664376840039254,
+      "grad_norm": 0.721568763256073,
+      "learning_rate": 8.248357708732749e-05,
+      "loss": 0.0681,
+      "step": 56010
+    },
+    {
+      "epoch": 3.665031076218515,
+      "grad_norm": 0.8713940382003784,
+      "learning_rate": 8.247659324692653e-05,
+      "loss": 0.0746,
+      "step": 56020
+    },
+    {
+      "epoch": 3.6656853123977755,
+      "grad_norm": 0.958366334438324,
+      "learning_rate": 8.246960831035539e-05,
+      "loss": 0.0689,
+      "step": 56030
+    },
+    {
+      "epoch": 3.666339548577036,
+      "grad_norm": 1.0349668264389038,
+      "learning_rate": 8.246262227784982e-05,
+      "loss": 0.074,
+      "step": 56040
+    },
+    {
+      "epoch": 3.666993784756297,
+      "grad_norm": 0.8951663374900818,
+      "learning_rate": 8.245563514964562e-05,
+      "loss": 0.0762,
+      "step": 56050
+    },
+    {
+      "epoch": 3.667648020935558,
+      "grad_norm": 0.993411660194397,
+      "learning_rate": 8.24486469259786e-05,
+      "loss": 0.0689,
+      "step": 56060
+    },
+    {
+      "epoch": 3.6683022571148185,
+      "grad_norm": 0.9579418897628784,
+      "learning_rate": 8.244165760708464e-05,
+      "loss": 0.0658,
+      "step": 56070
+    },
+    {
+      "epoch": 3.668956493294079,
+      "grad_norm": 0.8194859623908997,
+      "learning_rate": 8.243466719319967e-05,
+      "loss": 0.0703,
+      "step": 56080
+    },
+    {
+      "epoch": 3.66961072947334,
+      "grad_norm": 0.8599490523338318,
+      "learning_rate": 8.242767568455963e-05,
+      "loss": 0.0615,
+      "step": 56090
+    },
+    {
+      "epoch": 3.6702649656526005,
+      "grad_norm": 0.7916556596755981,
+      "learning_rate": 8.242068308140047e-05,
+      "loss": 0.0769,
+      "step": 56100
+    },
+    {
+      "epoch": 3.670919201831861,
+      "grad_norm": 0.8496348857879639,
+      "learning_rate": 8.241368938395824e-05,
+      "loss": 0.066,
+      "step": 56110
+    },
+    {
+      "epoch": 3.671573438011122,
+      "grad_norm": 0.7955387234687805,
+      "learning_rate": 8.240669459246897e-05,
+      "loss": 0.0701,
+      "step": 56120
+    },
+    {
+      "epoch": 3.672227674190383,
+      "grad_norm": 0.7705776691436768,
+      "learning_rate": 8.239969870716878e-05,
+      "loss": 0.0636,
+      "step": 56130
+    },
+    {
+      "epoch": 3.6728819103696435,
+      "grad_norm": 1.0588468313217163,
+      "learning_rate": 8.239270172829379e-05,
+      "loss": 0.0748,
+      "step": 56140
+    },
+    {
+      "epoch": 3.673536146548904,
+      "grad_norm": 0.9020249247550964,
+      "learning_rate": 8.238570365608016e-05,
+      "loss": 0.0689,
+      "step": 56150
+    },
+    {
+      "epoch": 3.674190382728165,
+      "grad_norm": 0.7228168845176697,
+      "learning_rate": 8.237870449076411e-05,
+      "loss": 0.0632,
+      "step": 56160
+    },
+    {
+      "epoch": 3.6748446189074255,
+      "grad_norm": 0.729026198387146,
+      "learning_rate": 8.237170423258184e-05,
+      "loss": 0.0761,
+      "step": 56170
+    },
+    {
+      "epoch": 3.675498855086686,
+      "grad_norm": 0.7430927157402039,
+      "learning_rate": 8.236470288176966e-05,
+      "loss": 0.0644,
+      "step": 56180
+    },
+    {
+      "epoch": 3.676153091265947,
+      "grad_norm": 0.8580201268196106,
+      "learning_rate": 8.235770043856389e-05,
+      "loss": 0.0615,
+      "step": 56190
+    },
+    {
+      "epoch": 3.676807327445208,
+      "grad_norm": 1.0346200466156006,
+      "learning_rate": 8.235069690320087e-05,
+      "loss": 0.079,
+      "step": 56200
+    },
+    {
+      "epoch": 3.6774615636244685,
+      "grad_norm": 1.0010615587234497,
+      "learning_rate": 8.234369227591698e-05,
+      "loss": 0.0673,
+      "step": 56210
+    },
+    {
+      "epoch": 3.678115799803729,
+      "grad_norm": 0.8165103793144226,
+      "learning_rate": 8.233668655694865e-05,
+      "loss": 0.0738,
+      "step": 56220
+    },
+    {
+      "epoch": 3.67877003598299,
+      "grad_norm": 0.8358020186424255,
+      "learning_rate": 8.232967974653235e-05,
+      "loss": 0.0637,
+      "step": 56230
+    },
+    {
+      "epoch": 3.6794242721622505,
+      "grad_norm": 0.8464856743812561,
+      "learning_rate": 8.232267184490457e-05,
+      "loss": 0.0696,
+      "step": 56240
+    },
+    {
+      "epoch": 3.680078508341511,
+      "grad_norm": 0.8467486500740051,
+      "learning_rate": 8.231566285230183e-05,
+      "loss": 0.0697,
+      "step": 56250
+    },
+    {
+      "epoch": 3.680732744520772,
+      "grad_norm": 0.9639089703559875,
+      "learning_rate": 8.230865276896074e-05,
+      "loss": 0.0811,
+      "step": 56260
+    },
+    {
+      "epoch": 3.6813869807000326,
+      "grad_norm": 0.7469546794891357,
+      "learning_rate": 8.230164159511788e-05,
+      "loss": 0.072,
+      "step": 56270
+    },
+    {
+      "epoch": 3.6820412168792935,
+      "grad_norm": 0.9208206534385681,
+      "learning_rate": 8.22946293310099e-05,
+      "loss": 0.0836,
+      "step": 56280
+    },
+    {
+      "epoch": 3.682695453058554,
+      "grad_norm": 0.7601810693740845,
+      "learning_rate": 8.228761597687348e-05,
+      "loss": 0.071,
+      "step": 56290
+    },
+    {
+      "epoch": 3.683349689237815,
+      "grad_norm": 0.8234832286834717,
+      "learning_rate": 8.228060153294534e-05,
+      "loss": 0.0734,
+      "step": 56300
+    },
+    {
+      "epoch": 3.6840039254170756,
+      "grad_norm": 0.8467697501182556,
+      "learning_rate": 8.227358599946224e-05,
+      "loss": 0.0743,
+      "step": 56310
+    },
+    {
+      "epoch": 3.684658161596336,
+      "grad_norm": 0.6788933873176575,
+      "learning_rate": 8.2266569376661e-05,
+      "loss": 0.0639,
+      "step": 56320
+    },
+    {
+      "epoch": 3.685312397775597,
+      "grad_norm": 0.9612674713134766,
+      "learning_rate": 8.225955166477839e-05,
+      "loss": 0.0667,
+      "step": 56330
+    },
+    {
+      "epoch": 3.6859666339548576,
+      "grad_norm": 0.8955544233322144,
+      "learning_rate": 8.22525328640513e-05,
+      "loss": 0.0671,
+      "step": 56340
+    },
+    {
+      "epoch": 3.6866208701341185,
+      "grad_norm": 0.9106106758117676,
+      "learning_rate": 8.224551297471668e-05,
+      "loss": 0.0695,
+      "step": 56350
+    },
+    {
+      "epoch": 3.687275106313379,
+      "grad_norm": 0.8671728372573853,
+      "learning_rate": 8.223849199701139e-05,
+      "loss": 0.06,
+      "step": 56360
+    },
+    {
+      "epoch": 3.68792934249264,
+      "grad_norm": 0.9457067251205444,
+      "learning_rate": 8.223146993117247e-05,
+      "loss": 0.0762,
+      "step": 56370
+    },
+    {
+      "epoch": 3.6885835786719006,
+      "grad_norm": 1.0926858186721802,
+      "learning_rate": 8.222444677743691e-05,
+      "loss": 0.068,
+      "step": 56380
+    },
+    {
+      "epoch": 3.689237814851161,
+      "grad_norm": 0.964023768901825,
+      "learning_rate": 8.221742253604175e-05,
+      "loss": 0.0635,
+      "step": 56390
+    },
+    {
+      "epoch": 3.689892051030422,
+      "grad_norm": 0.8525315523147583,
+      "learning_rate": 8.221039720722409e-05,
+      "loss": 0.0806,
+      "step": 56400
+    },
+    {
+      "epoch": 3.6905462872096826,
+      "grad_norm": 0.989080548286438,
+      "learning_rate": 8.220337079122105e-05,
+      "loss": 0.0713,
+      "step": 56410
+    },
+    {
+      "epoch": 3.6912005233889436,
+      "grad_norm": 0.7577918171882629,
+      "learning_rate": 8.21963432882698e-05,
+      "loss": 0.0716,
+      "step": 56420
+    },
+    {
+      "epoch": 3.691854759568204,
+      "grad_norm": 0.8016083240509033,
+      "learning_rate": 8.21893146986075e-05,
+      "loss": 0.071,
+      "step": 56430
+    },
+    {
+      "epoch": 3.692508995747465,
+      "grad_norm": 0.7300130128860474,
+      "learning_rate": 8.218228502247141e-05,
+      "loss": 0.073,
+      "step": 56440
+    },
+    {
+      "epoch": 3.6931632319267256,
+      "grad_norm": 0.8202305436134338,
+      "learning_rate": 8.217525426009882e-05,
+      "loss": 0.0685,
+      "step": 56450
+    },
+    {
+      "epoch": 3.693817468105986,
+      "grad_norm": 0.9385225176811218,
+      "learning_rate": 8.216822241172702e-05,
+      "loss": 0.0751,
+      "step": 56460
+    },
+    {
+      "epoch": 3.694471704285247,
+      "grad_norm": 1.1209096908569336,
+      "learning_rate": 8.216118947759333e-05,
+      "loss": 0.0709,
+      "step": 56470
+    },
+    {
+      "epoch": 3.6951259404645076,
+      "grad_norm": 1.0361253023147583,
+      "learning_rate": 8.215415545793515e-05,
+      "loss": 0.0765,
+      "step": 56480
+    },
+    {
+      "epoch": 3.6957801766437686,
+      "grad_norm": 0.8918325901031494,
+      "learning_rate": 8.214712035298991e-05,
+      "loss": 0.0622,
+      "step": 56490
+    },
+    {
+      "epoch": 3.696434412823029,
+      "grad_norm": 0.7726563811302185,
+      "learning_rate": 8.214008416299505e-05,
+      "loss": 0.0717,
+      "step": 56500
+    },
+    {
+      "epoch": 3.69708864900229,
+      "grad_norm": 0.9196798801422119,
+      "learning_rate": 8.213304688818804e-05,
+      "loss": 0.081,
+      "step": 56510
+    },
+    {
+      "epoch": 3.6977428851815506,
+      "grad_norm": 1.0043820142745972,
+      "learning_rate": 8.212600852880644e-05,
+      "loss": 0.0756,
+      "step": 56520
+    },
+    {
+      "epoch": 3.698397121360811,
+      "grad_norm": 0.926852285861969,
+      "learning_rate": 8.21189690850878e-05,
+      "loss": 0.0677,
+      "step": 56530
+    },
+    {
+      "epoch": 3.699051357540072,
+      "grad_norm": 0.8300206661224365,
+      "learning_rate": 8.211192855726972e-05,
+      "loss": 0.074,
+      "step": 56540
+    },
+    {
+      "epoch": 3.6997055937193326,
+      "grad_norm": 1.1823694705963135,
+      "learning_rate": 8.210488694558982e-05,
+      "loss": 0.0653,
+      "step": 56550
+    },
+    {
+      "epoch": 3.700359829898593,
+      "grad_norm": 0.7642592787742615,
+      "learning_rate": 8.20978442502858e-05,
+      "loss": 0.0634,
+      "step": 56560
+    },
+    {
+      "epoch": 3.701014066077854,
+      "grad_norm": 0.9200793504714966,
+      "learning_rate": 8.209080047159535e-05,
+      "loss": 0.0862,
+      "step": 56570
+    },
+    {
+      "epoch": 3.701668302257115,
+      "grad_norm": 0.8010613918304443,
+      "learning_rate": 8.208375560975624e-05,
+      "loss": 0.0636,
+      "step": 56580
+    },
+    {
+      "epoch": 3.7023225384363756,
+      "grad_norm": 0.7096419930458069,
+      "learning_rate": 8.207670966500621e-05,
+      "loss": 0.065,
+      "step": 56590
+    },
+    {
+      "epoch": 3.702976774615636,
+      "grad_norm": 1.8889129161834717,
+      "learning_rate": 8.206966263758311e-05,
+      "loss": 0.0617,
+      "step": 56600
+    },
+    {
+      "epoch": 3.703631010794897,
+      "grad_norm": 0.7090383172035217,
+      "learning_rate": 8.20626145277248e-05,
+      "loss": 0.0718,
+      "step": 56610
+    },
+    {
+      "epoch": 3.7042852469741576,
+      "grad_norm": 0.8787170648574829,
+      "learning_rate": 8.205556533566916e-05,
+      "loss": 0.0734,
+      "step": 56620
+    },
+    {
+      "epoch": 3.704939483153418,
+      "grad_norm": 0.8815798759460449,
+      "learning_rate": 8.204851506165412e-05,
+      "loss": 0.0692,
+      "step": 56630
+    },
+    {
+      "epoch": 3.705593719332679,
+      "grad_norm": 0.6900599002838135,
+      "learning_rate": 8.204146370591765e-05,
+      "loss": 0.0632,
+      "step": 56640
+    },
+    {
+      "epoch": 3.70624795551194,
+      "grad_norm": 0.9211147427558899,
+      "learning_rate": 8.203441126869773e-05,
+      "loss": 0.0785,
+      "step": 56650
+    },
+    {
+      "epoch": 3.7069021916912006,
+      "grad_norm": 0.6548269987106323,
+      "learning_rate": 8.202735775023244e-05,
+      "loss": 0.0694,
+      "step": 56660
+    },
+    {
+      "epoch": 3.707556427870461,
+      "grad_norm": 0.7686384916305542,
+      "learning_rate": 8.202030315075982e-05,
+      "loss": 0.0664,
+      "step": 56670
+    },
+    {
+      "epoch": 3.708210664049722,
+      "grad_norm": 0.7667417526245117,
+      "learning_rate": 8.2013247470518e-05,
+      "loss": 0.06,
+      "step": 56680
+    },
+    {
+      "epoch": 3.7088649002289826,
+      "grad_norm": 0.70942622423172,
+      "learning_rate": 8.200619070974512e-05,
+      "loss": 0.0621,
+      "step": 56690
+    },
+    {
+      "epoch": 3.709519136408243,
+      "grad_norm": 0.9921366572380066,
+      "learning_rate": 8.199913286867934e-05,
+      "loss": 0.0727,
+      "step": 56700
+    },
+    {
+      "epoch": 3.710173372587504,
+      "grad_norm": 0.9375357031822205,
+      "learning_rate": 8.199207394755893e-05,
+      "loss": 0.0697,
+      "step": 56710
+    },
+    {
+      "epoch": 3.7108276087667647,
+      "grad_norm": 0.9243108034133911,
+      "learning_rate": 8.198501394662212e-05,
+      "loss": 0.0649,
+      "step": 56720
+    },
+    {
+      "epoch": 3.7114818449460256,
+      "grad_norm": 0.9306524395942688,
+      "learning_rate": 8.197795286610719e-05,
+      "loss": 0.0714,
+      "step": 56730
+    },
+    {
+      "epoch": 3.712136081125286,
+      "grad_norm": 1.01764976978302,
+      "learning_rate": 8.19708907062525e-05,
+      "loss": 0.0695,
+      "step": 56740
+    },
+    {
+      "epoch": 3.712790317304547,
+      "grad_norm": 0.7517834901809692,
+      "learning_rate": 8.196382746729641e-05,
+      "loss": 0.0747,
+      "step": 56750
+    },
+    {
+      "epoch": 3.7134445534838076,
+      "grad_norm": 0.7786497473716736,
+      "learning_rate": 8.195676314947729e-05,
+      "loss": 0.0817,
+      "step": 56760
+    },
+    {
+      "epoch": 3.714098789663068,
+      "grad_norm": 0.939345121383667,
+      "learning_rate": 8.194969775303361e-05,
+      "loss": 0.0732,
+      "step": 56770
+    },
+    {
+      "epoch": 3.714753025842329,
+      "grad_norm": 1.078481674194336,
+      "learning_rate": 8.194263127820385e-05,
+      "loss": 0.0724,
+      "step": 56780
+    },
+    {
+      "epoch": 3.7154072620215897,
+      "grad_norm": 0.9831727147102356,
+      "learning_rate": 8.19355637252265e-05,
+      "loss": 0.0663,
+      "step": 56790
+    },
+    {
+      "epoch": 3.7160614982008506,
+      "grad_norm": 0.8869754672050476,
+      "learning_rate": 8.192849509434014e-05,
+      "loss": 0.074,
+      "step": 56800
+    },
+    {
+      "epoch": 3.716715734380111,
+      "grad_norm": 0.7121245861053467,
+      "learning_rate": 8.192142538578331e-05,
+      "loss": 0.0716,
+      "step": 56810
+    },
+    {
+      "epoch": 3.717369970559372,
+      "grad_norm": 0.8707069754600525,
+      "learning_rate": 8.191435459979468e-05,
+      "loss": 0.0698,
+      "step": 56820
+    },
+    {
+      "epoch": 3.7180242067386327,
+      "grad_norm": 1.0101736783981323,
+      "learning_rate": 8.190728273661288e-05,
+      "loss": 0.0761,
+      "step": 56830
+    },
+    {
+      "epoch": 3.718678442917893,
+      "grad_norm": 0.8770877122879028,
+      "learning_rate": 8.190020979647659e-05,
+      "loss": 0.0802,
+      "step": 56840
+    },
+    {
+      "epoch": 3.719332679097154,
+      "grad_norm": 0.9432161450386047,
+      "learning_rate": 8.189313577962457e-05,
+      "loss": 0.0711,
+      "step": 56850
+    },
+    {
+      "epoch": 3.7199869152764147,
+      "grad_norm": 0.8004513382911682,
+      "learning_rate": 8.188606068629558e-05,
+      "loss": 0.0676,
+      "step": 56860
+    },
+    {
+      "epoch": 3.7206411514556756,
+      "grad_norm": 0.9776605367660522,
+      "learning_rate": 8.187898451672841e-05,
+      "loss": 0.0655,
+      "step": 56870
+    },
+    {
+      "epoch": 3.721295387634936,
+      "grad_norm": 0.774079442024231,
+      "learning_rate": 8.18719072711619e-05,
+      "loss": 0.0632,
+      "step": 56880
+    },
+    {
+      "epoch": 3.721949623814197,
+      "grad_norm": 0.9294254183769226,
+      "learning_rate": 8.186482894983494e-05,
+      "loss": 0.0656,
+      "step": 56890
+    },
+    {
+      "epoch": 3.7226038599934577,
+      "grad_norm": 0.9034769535064697,
+      "learning_rate": 8.185774955298644e-05,
+      "loss": 0.0775,
+      "step": 56900
+    },
+    {
+      "epoch": 3.723258096172718,
+      "grad_norm": 1.0133552551269531,
+      "learning_rate": 8.185066908085535e-05,
+      "loss": 0.0684,
+      "step": 56910
+    },
+    {
+      "epoch": 3.723912332351979,
+      "grad_norm": 1.045736312866211,
+      "learning_rate": 8.184358753368062e-05,
+      "loss": 0.0627,
+      "step": 56920
+    },
+    {
+      "epoch": 3.7245665685312397,
+      "grad_norm": 0.8517866730690002,
+      "learning_rate": 8.183650491170132e-05,
+      "loss": 0.0639,
+      "step": 56930
+    },
+    {
+      "epoch": 3.7252208047105007,
+      "grad_norm": 0.6794936060905457,
+      "learning_rate": 8.182942121515648e-05,
+      "loss": 0.0624,
+      "step": 56940
+    },
+    {
+      "epoch": 3.725875040889761,
+      "grad_norm": 0.8077151775360107,
+      "learning_rate": 8.182233644428518e-05,
+      "loss": 0.0797,
+      "step": 56950
+    },
+    {
+      "epoch": 3.726529277069022,
+      "grad_norm": 0.8589492440223694,
+      "learning_rate": 8.18152505993266e-05,
+      "loss": 0.0772,
+      "step": 56960
+    },
+    {
+      "epoch": 3.7271835132482827,
+      "grad_norm": 0.8781915307044983,
+      "learning_rate": 8.180816368051985e-05,
+      "loss": 0.0741,
+      "step": 56970
+    },
+    {
+      "epoch": 3.727837749427543,
+      "grad_norm": 0.91343092918396,
+      "learning_rate": 8.180107568810417e-05,
+      "loss": 0.0756,
+      "step": 56980
+    },
+    {
+      "epoch": 3.728491985606804,
+      "grad_norm": 0.8857640624046326,
+      "learning_rate": 8.179398662231876e-05,
+      "loss": 0.0808,
+      "step": 56990
+    },
+    {
+      "epoch": 3.7291462217860647,
+      "grad_norm": 0.807052493095398,
+      "learning_rate": 8.178689648340294e-05,
+      "loss": 0.0756,
+      "step": 57000
+    },
+    {
+      "epoch": 3.7298004579653252,
+      "grad_norm": 0.7484919428825378,
+      "learning_rate": 8.1779805271596e-05,
+      "loss": 0.066,
+      "step": 57010
+    },
+    {
+      "epoch": 3.730454694144586,
+      "grad_norm": 0.8383365869522095,
+      "learning_rate": 8.17727129871373e-05,
+      "loss": 0.0684,
+      "step": 57020
+    },
+    {
+      "epoch": 3.731108930323847,
+      "grad_norm": 0.9385350346565247,
+      "learning_rate": 8.176561963026618e-05,
+      "loss": 0.0805,
+      "step": 57030
+    },
+    {
+      "epoch": 3.7317631665031077,
+      "grad_norm": 0.7031242847442627,
+      "learning_rate": 8.17585252012221e-05,
+      "loss": 0.0684,
+      "step": 57040
+    },
+    {
+      "epoch": 3.732417402682368,
+      "grad_norm": 0.9507877826690674,
+      "learning_rate": 8.175142970024451e-05,
+      "loss": 0.0738,
+      "step": 57050
+    },
+    {
+      "epoch": 3.733071638861629,
+      "grad_norm": 0.9051758646965027,
+      "learning_rate": 8.174433312757289e-05,
+      "loss": 0.0642,
+      "step": 57060
+    },
+    {
+      "epoch": 3.7337258750408897,
+      "grad_norm": 0.8826471567153931,
+      "learning_rate": 8.173723548344675e-05,
+      "loss": 0.0764,
+      "step": 57070
+    },
+    {
+      "epoch": 3.7343801112201502,
+      "grad_norm": 0.886605978012085,
+      "learning_rate": 8.173013676810573e-05,
+      "loss": 0.0695,
+      "step": 57080
+    },
+    {
+      "epoch": 3.735034347399411,
+      "grad_norm": 0.7051117420196533,
+      "learning_rate": 8.172303698178935e-05,
+      "loss": 0.069,
+      "step": 57090
+    },
+    {
+      "epoch": 3.735688583578672,
+      "grad_norm": 0.8651570677757263,
+      "learning_rate": 8.171593612473727e-05,
+      "loss": 0.0696,
+      "step": 57100
+    },
+    {
+      "epoch": 3.7363428197579327,
+      "grad_norm": 0.9195097088813782,
+      "learning_rate": 8.170883419718917e-05,
+      "loss": 0.0791,
+      "step": 57110
+    },
+    {
+      "epoch": 3.7369970559371932,
+      "grad_norm": 0.835708498954773,
+      "learning_rate": 8.170173119938478e-05,
+      "loss": 0.0733,
+      "step": 57120
+    },
+    {
+      "epoch": 3.737651292116454,
+      "grad_norm": 0.8207480311393738,
+      "learning_rate": 8.16946271315638e-05,
+      "loss": 0.073,
+      "step": 57130
+    },
+    {
+      "epoch": 3.7383055282957147,
+      "grad_norm": 1.0654559135437012,
+      "learning_rate": 8.168752199396603e-05,
+      "loss": 0.0728,
+      "step": 57140
+    },
+    {
+      "epoch": 3.7389597644749752,
+      "grad_norm": 0.7882975339889526,
+      "learning_rate": 8.168041578683129e-05,
+      "loss": 0.0645,
+      "step": 57150
+    },
+    {
+      "epoch": 3.739614000654236,
+      "grad_norm": 0.7596569061279297,
+      "learning_rate": 8.167330851039945e-05,
+      "loss": 0.0689,
+      "step": 57160
+    },
+    {
+      "epoch": 3.7402682368334967,
+      "grad_norm": 1.1013258695602417,
+      "learning_rate": 8.166620016491038e-05,
+      "loss": 0.0647,
+      "step": 57170
+    },
+    {
+      "epoch": 3.7409224730127577,
+      "grad_norm": 0.8467496037483215,
+      "learning_rate": 8.165909075060398e-05,
+      "loss": 0.0658,
+      "step": 57180
+    },
+    {
+      "epoch": 3.7415767091920182,
+      "grad_norm": 0.7248758673667908,
+      "learning_rate": 8.165198026772025e-05,
+      "loss": 0.0705,
+      "step": 57190
+    },
+    {
+      "epoch": 3.742230945371279,
+      "grad_norm": 0.8981897234916687,
+      "learning_rate": 8.164486871649919e-05,
+      "loss": 0.074,
+      "step": 57200
+    },
+    {
+      "epoch": 3.7428851815505397,
+      "grad_norm": 0.8243358135223389,
+      "learning_rate": 8.16377560971808e-05,
+      "loss": 0.0758,
+      "step": 57210
+    },
+    {
+      "epoch": 3.7435394177298003,
+      "grad_norm": 0.7456525564193726,
+      "learning_rate": 8.163064241000517e-05,
+      "loss": 0.0586,
+      "step": 57220
+    },
+    {
+      "epoch": 3.7441936539090612,
+      "grad_norm": 0.9510387182235718,
+      "learning_rate": 8.162352765521243e-05,
+      "loss": 0.0635,
+      "step": 57230
+    },
+    {
+      "epoch": 3.7448478900883218,
+      "grad_norm": 0.9684381484985352,
+      "learning_rate": 8.161641183304267e-05,
+      "loss": 0.0621,
+      "step": 57240
+    },
+    {
+      "epoch": 3.7455021262675827,
+      "grad_norm": 0.8673816919326782,
+      "learning_rate": 8.16092949437361e-05,
+      "loss": 0.0764,
+      "step": 57250
+    },
+    {
+      "epoch": 3.7461563624468432,
+      "grad_norm": 0.7953469157218933,
+      "learning_rate": 8.160217698753291e-05,
+      "loss": 0.0687,
+      "step": 57260
+    },
+    {
+      "epoch": 3.746810598626104,
+      "grad_norm": 0.7955390214920044,
+      "learning_rate": 8.159505796467342e-05,
+      "loss": 0.0706,
+      "step": 57270
+    },
+    {
+      "epoch": 3.7474648348053647,
+      "grad_norm": 0.8372431397438049,
+      "learning_rate": 8.158793787539782e-05,
+      "loss": 0.0737,
+      "step": 57280
+    },
+    {
+      "epoch": 3.7481190709846253,
+      "grad_norm": 1.1341184377670288,
+      "learning_rate": 8.158081671994648e-05,
+      "loss": 0.0737,
+      "step": 57290
+    },
+    {
+      "epoch": 3.7487733071638862,
+      "grad_norm": 0.8782973289489746,
+      "learning_rate": 8.157369449855974e-05,
+      "loss": 0.0708,
+      "step": 57300
+    },
+    {
+      "epoch": 3.7494275433431468,
+      "grad_norm": 0.9155938625335693,
+      "learning_rate": 8.156657121147803e-05,
+      "loss": 0.0695,
+      "step": 57310
+    },
+    {
+      "epoch": 3.7500817795224077,
+      "grad_norm": 0.7682180404663086,
+      "learning_rate": 8.155944685894175e-05,
+      "loss": 0.0709,
+      "step": 57320
+    },
+    {
+      "epoch": 3.7507360157016683,
+      "grad_norm": 0.9652989506721497,
+      "learning_rate": 8.155232144119135e-05,
+      "loss": 0.0628,
+      "step": 57330
+    },
+    {
+      "epoch": 3.7513902518809292,
+      "grad_norm": 1.136732578277588,
+      "learning_rate": 8.154519495846737e-05,
+      "loss": 0.0689,
+      "step": 57340
+    },
+    {
+      "epoch": 3.7520444880601898,
+      "grad_norm": 0.9039020538330078,
+      "learning_rate": 8.153806741101033e-05,
+      "loss": 0.0792,
+      "step": 57350
+    },
+    {
+      "epoch": 3.7526987242394503,
+      "grad_norm": 0.8383570313453674,
+      "learning_rate": 8.15309387990608e-05,
+      "loss": 0.0765,
+      "step": 57360
+    },
+    {
+      "epoch": 3.7533529604187112,
+      "grad_norm": 0.9474335312843323,
+      "learning_rate": 8.15238091228594e-05,
+      "loss": 0.0641,
+      "step": 57370
+    },
+    {
+      "epoch": 3.7540071965979718,
+      "grad_norm": 0.7904636859893799,
+      "learning_rate": 8.151667838264677e-05,
+      "loss": 0.0684,
+      "step": 57380
+    },
+    {
+      "epoch": 3.7546614327772327,
+      "grad_norm": 0.861436128616333,
+      "learning_rate": 8.150954657866356e-05,
+      "loss": 0.0669,
+      "step": 57390
+    },
+    {
+      "epoch": 3.7553156689564933,
+      "grad_norm": 0.7418460249900818,
+      "learning_rate": 8.150241371115055e-05,
+      "loss": 0.0683,
+      "step": 57400
+    },
+    {
+      "epoch": 3.7559699051357542,
+      "grad_norm": 0.7864526510238647,
+      "learning_rate": 8.149527978034844e-05,
+      "loss": 0.0677,
+      "step": 57410
+    },
+    {
+      "epoch": 3.7566241413150148,
+      "grad_norm": 0.8774205446243286,
+      "learning_rate": 8.148814478649805e-05,
+      "loss": 0.0747,
+      "step": 57420
+    },
+    {
+      "epoch": 3.7572783774942753,
+      "grad_norm": 0.7475970387458801,
+      "learning_rate": 8.148100872984019e-05,
+      "loss": 0.0685,
+      "step": 57430
+    },
+    {
+      "epoch": 3.7579326136735363,
+      "grad_norm": 1.0144479274749756,
+      "learning_rate": 8.14738716106157e-05,
+      "loss": 0.0736,
+      "step": 57440
+    },
+    {
+      "epoch": 3.758586849852797,
+      "grad_norm": 1.0739880800247192,
+      "learning_rate": 8.146673342906552e-05,
+      "loss": 0.0676,
+      "step": 57450
+    },
+    {
+      "epoch": 3.7592410860320573,
+      "grad_norm": 0.9529927968978882,
+      "learning_rate": 8.145959418543057e-05,
+      "loss": 0.0645,
+      "step": 57460
+    },
+    {
+      "epoch": 3.7598953222113183,
+      "grad_norm": 0.8140749931335449,
+      "learning_rate": 8.14524538799518e-05,
+      "loss": 0.0721,
+      "step": 57470
+    },
+    {
+      "epoch": 3.7605495583905793,
+      "grad_norm": 0.7363076210021973,
+      "learning_rate": 8.144531251287024e-05,
+      "loss": 0.0655,
+      "step": 57480
+    },
+    {
+      "epoch": 3.7612037945698398,
+      "grad_norm": 0.9342272281646729,
+      "learning_rate": 8.14381700844269e-05,
+      "loss": 0.0688,
+      "step": 57490
+    },
+    {
+      "epoch": 3.7618580307491003,
+      "grad_norm": 0.8387247920036316,
+      "learning_rate": 8.143102659486287e-05,
+      "loss": 0.0605,
+      "step": 57500
+    },
+    {
+      "epoch": 3.7625122669283613,
+      "grad_norm": 1.0524139404296875,
+      "learning_rate": 8.142388204441927e-05,
+      "loss": 0.0688,
+      "step": 57510
+    },
+    {
+      "epoch": 3.763166503107622,
+      "grad_norm": 1.1220943927764893,
+      "learning_rate": 8.141673643333723e-05,
+      "loss": 0.064,
+      "step": 57520
+    },
+    {
+      "epoch": 3.7638207392868823,
+      "grad_norm": 0.7932508587837219,
+      "learning_rate": 8.140958976185794e-05,
+      "loss": 0.0703,
+      "step": 57530
+    },
+    {
+      "epoch": 3.7644749754661433,
+      "grad_norm": 0.8882383704185486,
+      "learning_rate": 8.140244203022262e-05,
+      "loss": 0.0643,
+      "step": 57540
+    },
+    {
+      "epoch": 3.7651292116454043,
+      "grad_norm": 0.8746048212051392,
+      "learning_rate": 8.139529323867254e-05,
+      "loss": 0.0687,
+      "step": 57550
+    },
+    {
+      "epoch": 3.765783447824665,
+      "grad_norm": 1.058364987373352,
+      "learning_rate": 8.138814338744896e-05,
+      "loss": 0.0711,
+      "step": 57560
+    },
+    {
+      "epoch": 3.7664376840039253,
+      "grad_norm": 0.7067134976387024,
+      "learning_rate": 8.138099247679322e-05,
+      "loss": 0.0648,
+      "step": 57570
+    },
+    {
+      "epoch": 3.7670919201831863,
+      "grad_norm": 0.7847724556922913,
+      "learning_rate": 8.13738405069467e-05,
+      "loss": 0.065,
+      "step": 57580
+    },
+    {
+      "epoch": 3.767746156362447,
+      "grad_norm": 0.9245442748069763,
+      "learning_rate": 8.136668747815078e-05,
+      "loss": 0.0655,
+      "step": 57590
+    },
+    {
+      "epoch": 3.7684003925417073,
+      "grad_norm": 0.940708339214325,
+      "learning_rate": 8.135953339064688e-05,
+      "loss": 0.0632,
+      "step": 57600
+    },
+    {
+      "epoch": 3.7690546287209683,
+      "grad_norm": 1.0202513933181763,
+      "learning_rate": 8.13523782446765e-05,
+      "loss": 0.0688,
+      "step": 57610
+    },
+    {
+      "epoch": 3.769708864900229,
+      "grad_norm": 1.0422993898391724,
+      "learning_rate": 8.134522204048112e-05,
+      "loss": 0.0748,
+      "step": 57620
+    },
+    {
+      "epoch": 3.77036310107949,
+      "grad_norm": 0.8074596524238586,
+      "learning_rate": 8.13380647783023e-05,
+      "loss": 0.0715,
+      "step": 57630
+    },
+    {
+      "epoch": 3.7710173372587503,
+      "grad_norm": 0.8613049983978271,
+      "learning_rate": 8.13309064583816e-05,
+      "loss": 0.0843,
+      "step": 57640
+    },
+    {
+      "epoch": 3.7716715734380113,
+      "grad_norm": 0.8982073664665222,
+      "learning_rate": 8.132374708096065e-05,
+      "loss": 0.0653,
+      "step": 57650
+    },
+    {
+      "epoch": 3.772325809617272,
+      "grad_norm": 0.8297070860862732,
+      "learning_rate": 8.131658664628107e-05,
+      "loss": 0.0653,
+      "step": 57660
+    },
+    {
+      "epoch": 3.7729800457965323,
+      "grad_norm": 0.9720892906188965,
+      "learning_rate": 8.130942515458456e-05,
+      "loss": 0.0639,
+      "step": 57670
+    },
+    {
+      "epoch": 3.7736342819757933,
+      "grad_norm": 0.9780619144439697,
+      "learning_rate": 8.130226260611284e-05,
+      "loss": 0.0744,
+      "step": 57680
+    },
+    {
+      "epoch": 3.774288518155054,
+      "grad_norm": 0.7452018857002258,
+      "learning_rate": 8.129509900110767e-05,
+      "loss": 0.0618,
+      "step": 57690
+    },
+    {
+      "epoch": 3.774942754334315,
+      "grad_norm": 0.9250910878181458,
+      "learning_rate": 8.128793433981084e-05,
+      "loss": 0.0693,
+      "step": 57700
+    },
+    {
+      "epoch": 3.7755969905135753,
+      "grad_norm": 1.0864530801773071,
+      "learning_rate": 8.128076862246416e-05,
+      "loss": 0.0729,
+      "step": 57710
+    },
+    {
+      "epoch": 3.7762512266928363,
+      "grad_norm": 0.7634531259536743,
+      "learning_rate": 8.127360184930952e-05,
+      "loss": 0.0699,
+      "step": 57720
+    },
+    {
+      "epoch": 3.776905462872097,
+      "grad_norm": 0.8827683925628662,
+      "learning_rate": 8.126643402058877e-05,
+      "loss": 0.0723,
+      "step": 57730
+    },
+    {
+      "epoch": 3.7775596990513574,
+      "grad_norm": 0.6894218325614929,
+      "learning_rate": 8.12592651365439e-05,
+      "loss": 0.0621,
+      "step": 57740
+    },
+    {
+      "epoch": 3.7782139352306183,
+      "grad_norm": 0.8839403986930847,
+      "learning_rate": 8.125209519741683e-05,
+      "loss": 0.0728,
+      "step": 57750
+    },
+    {
+      "epoch": 3.778868171409879,
+      "grad_norm": 0.7568190097808838,
+      "learning_rate": 8.124492420344961e-05,
+      "loss": 0.0659,
+      "step": 57760
+    },
+    {
+      "epoch": 3.77952240758914,
+      "grad_norm": 0.9448241591453552,
+      "learning_rate": 8.123775215488423e-05,
+      "loss": 0.0676,
+      "step": 57770
+    },
+    {
+      "epoch": 3.7801766437684003,
+      "grad_norm": 0.7295737862586975,
+      "learning_rate": 8.123057905196281e-05,
+      "loss": 0.0656,
+      "step": 57780
+    },
+    {
+      "epoch": 3.7808308799476613,
+      "grad_norm": 1.049312710762024,
+      "learning_rate": 8.122340489492743e-05,
+      "loss": 0.0662,
+      "step": 57790
+    },
+    {
+      "epoch": 3.781485116126922,
+      "grad_norm": 0.6659790277481079,
+      "learning_rate": 8.121622968402025e-05,
+      "loss": 0.063,
+      "step": 57800
+    },
+    {
+      "epoch": 3.7821393523061824,
+      "grad_norm": 1.10063636302948,
+      "learning_rate": 8.120905341948345e-05,
+      "loss": 0.0679,
+      "step": 57810
+    },
+    {
+      "epoch": 3.7827935884854433,
+      "grad_norm": 1.0684069395065308,
+      "learning_rate": 8.120187610155924e-05,
+      "loss": 0.0664,
+      "step": 57820
+    },
+    {
+      "epoch": 3.783447824664704,
+      "grad_norm": 0.7764497995376587,
+      "learning_rate": 8.119469773048992e-05,
+      "loss": 0.08,
+      "step": 57830
+    },
+    {
+      "epoch": 3.784102060843965,
+      "grad_norm": 0.8615806698799133,
+      "learning_rate": 8.11875183065177e-05,
+      "loss": 0.0738,
+      "step": 57840
+    },
+    {
+      "epoch": 3.7847562970232254,
+      "grad_norm": 0.8051279783248901,
+      "learning_rate": 8.118033782988496e-05,
+      "loss": 0.0642,
+      "step": 57850
+    },
+    {
+      "epoch": 3.7854105332024863,
+      "grad_norm": 0.7473447322845459,
+      "learning_rate": 8.117315630083404e-05,
+      "loss": 0.0814,
+      "step": 57860
+    },
+    {
+      "epoch": 3.786064769381747,
+      "grad_norm": 0.7426646947860718,
+      "learning_rate": 8.116597371960734e-05,
+      "loss": 0.0726,
+      "step": 57870
+    },
+    {
+      "epoch": 3.7867190055610074,
+      "grad_norm": 0.7152010798454285,
+      "learning_rate": 8.115879008644729e-05,
+      "loss": 0.0693,
+      "step": 57880
+    },
+    {
+      "epoch": 3.7873732417402683,
+      "grad_norm": 1.007025957107544,
+      "learning_rate": 8.115160540159636e-05,
+      "loss": 0.0714,
+      "step": 57890
+    },
+    {
+      "epoch": 3.788027477919529,
+      "grad_norm": 0.8072939515113831,
+      "learning_rate": 8.114441966529707e-05,
+      "loss": 0.0619,
+      "step": 57900
+    },
+    {
+      "epoch": 3.7886817140987894,
+      "grad_norm": 0.8661327362060547,
+      "learning_rate": 8.11372328777919e-05,
+      "loss": 0.0662,
+      "step": 57910
+    },
+    {
+      "epoch": 3.7893359502780504,
+      "grad_norm": 0.8400626182556152,
+      "learning_rate": 8.113004503932348e-05,
+      "loss": 0.0643,
+      "step": 57920
+    },
+    {
+      "epoch": 3.7899901864573113,
+      "grad_norm": 0.8952879309654236,
+      "learning_rate": 8.11228561501344e-05,
+      "loss": 0.0638,
+      "step": 57930
+    },
+    {
+      "epoch": 3.790644422636572,
+      "grad_norm": 1.0547457933425903,
+      "learning_rate": 8.11156662104673e-05,
+      "loss": 0.0639,
+      "step": 57940
+    },
+    {
+      "epoch": 3.7912986588158324,
+      "grad_norm": 1.0903702974319458,
+      "learning_rate": 8.110847522056485e-05,
+      "loss": 0.0619,
+      "step": 57950
+    },
+    {
+      "epoch": 3.7919528949950934,
+      "grad_norm": 0.9032535552978516,
+      "learning_rate": 8.11012831806698e-05,
+      "loss": 0.0725,
+      "step": 57960
+    },
+    {
+      "epoch": 3.792607131174354,
+      "grad_norm": 1.0580861568450928,
+      "learning_rate": 8.109409009102486e-05,
+      "loss": 0.0721,
+      "step": 57970
+    },
+    {
+      "epoch": 3.7932613673536144,
+      "grad_norm": 1.0803550481796265,
+      "learning_rate": 8.108689595187285e-05,
+      "loss": 0.0703,
+      "step": 57980
+    },
+    {
+      "epoch": 3.7939156035328754,
+      "grad_norm": 0.8773861527442932,
+      "learning_rate": 8.107970076345654e-05,
+      "loss": 0.067,
+      "step": 57990
+    },
+    {
+      "epoch": 3.7945698397121363,
+      "grad_norm": 0.7425010204315186,
+      "learning_rate": 8.107250452601885e-05,
+      "loss": 0.0721,
+      "step": 58000
+    },
+    {
+      "epoch": 3.795224075891397,
+      "grad_norm": 0.7278298139572144,
+      "learning_rate": 8.106530723980261e-05,
+      "loss": 0.0675,
+      "step": 58010
+    },
+    {
+      "epoch": 3.7958783120706574,
+      "grad_norm": 0.9992676973342896,
+      "learning_rate": 8.105810890505081e-05,
+      "loss": 0.0685,
+      "step": 58020
+    },
+    {
+      "epoch": 3.7965325482499184,
+      "grad_norm": 0.8017996549606323,
+      "learning_rate": 8.105090952200637e-05,
+      "loss": 0.0699,
+      "step": 58030
+    },
+    {
+      "epoch": 3.797186784429179,
+      "grad_norm": 1.016209363937378,
+      "learning_rate": 8.10437090909123e-05,
+      "loss": 0.0703,
+      "step": 58040
+    },
+    {
+      "epoch": 3.7978410206084394,
+      "grad_norm": 0.8442249894142151,
+      "learning_rate": 8.103650761201163e-05,
+      "loss": 0.0719,
+      "step": 58050
+    },
+    {
+      "epoch": 3.7984952567877004,
+      "grad_norm": 0.8751046061515808,
+      "learning_rate": 8.102930508554744e-05,
+      "loss": 0.07,
+      "step": 58060
+    },
+    {
+      "epoch": 3.799149492966961,
+      "grad_norm": 0.6956860423088074,
+      "learning_rate": 8.102210151176282e-05,
+      "loss": 0.0631,
+      "step": 58070
+    },
+    {
+      "epoch": 3.799803729146222,
+      "grad_norm": 0.9162025451660156,
+      "learning_rate": 8.101489689090091e-05,
+      "loss": 0.0699,
+      "step": 58080
+    },
+    {
+      "epoch": 3.8004579653254824,
+      "grad_norm": 0.9595727920532227,
+      "learning_rate": 8.10076912232049e-05,
+      "loss": 0.0604,
+      "step": 58090
+    },
+    {
+      "epoch": 3.8011122015047434,
+      "grad_norm": 0.9500911235809326,
+      "learning_rate": 8.100048450891799e-05,
+      "loss": 0.0625,
+      "step": 58100
+    },
+    {
+      "epoch": 3.801766437684004,
+      "grad_norm": 0.6980655193328857,
+      "learning_rate": 8.099327674828342e-05,
+      "loss": 0.0588,
+      "step": 58110
+    },
+    {
+      "epoch": 3.8024206738632644,
+      "grad_norm": 0.9078855514526367,
+      "learning_rate": 8.098606794154448e-05,
+      "loss": 0.0599,
+      "step": 58120
+    },
+    {
+      "epoch": 3.8030749100425254,
+      "grad_norm": 0.8954933881759644,
+      "learning_rate": 8.097885808894449e-05,
+      "loss": 0.0646,
+      "step": 58130
+    },
+    {
+      "epoch": 3.803729146221786,
+      "grad_norm": 0.8276561498641968,
+      "learning_rate": 8.097164719072679e-05,
+      "loss": 0.0717,
+      "step": 58140
+    },
+    {
+      "epoch": 3.804383382401047,
+      "grad_norm": 0.8717600703239441,
+      "learning_rate": 8.096443524713477e-05,
+      "loss": 0.069,
+      "step": 58150
+    },
+    {
+      "epoch": 3.8050376185803074,
+      "grad_norm": 0.9472079277038574,
+      "learning_rate": 8.095722225841185e-05,
+      "loss": 0.0702,
+      "step": 58160
+    },
+    {
+      "epoch": 3.8056918547595684,
+      "grad_norm": 0.8693017959594727,
+      "learning_rate": 8.09500082248015e-05,
+      "loss": 0.0697,
+      "step": 58170
+    },
+    {
+      "epoch": 3.806346090938829,
+      "grad_norm": 1.0408918857574463,
+      "learning_rate": 8.094279314654718e-05,
+      "loss": 0.0752,
+      "step": 58180
+    },
+    {
+      "epoch": 3.8070003271180894,
+      "grad_norm": 0.827529788017273,
+      "learning_rate": 8.093557702389246e-05,
+      "loss": 0.0712,
+      "step": 58190
+    },
+    {
+      "epoch": 3.8076545632973504,
+      "grad_norm": 0.6943844556808472,
+      "learning_rate": 8.092835985708088e-05,
+      "loss": 0.0648,
+      "step": 58200
+    },
+    {
+      "epoch": 3.808308799476611,
+      "grad_norm": 0.8800909519195557,
+      "learning_rate": 8.092114164635604e-05,
+      "loss": 0.0684,
+      "step": 58210
+    },
+    {
+      "epoch": 3.808963035655872,
+      "grad_norm": 0.7584820985794067,
+      "learning_rate": 8.091392239196159e-05,
+      "loss": 0.072,
+      "step": 58220
+    },
+    {
+      "epoch": 3.8096172718351324,
+      "grad_norm": 0.9473629593849182,
+      "learning_rate": 8.090670209414117e-05,
+      "loss": 0.0664,
+      "step": 58230
+    },
+    {
+      "epoch": 3.8102715080143934,
+      "grad_norm": 1.0267720222473145,
+      "learning_rate": 8.08994807531385e-05,
+      "loss": 0.0638,
+      "step": 58240
+    },
+    {
+      "epoch": 3.810925744193654,
+      "grad_norm": 0.7905268669128418,
+      "learning_rate": 8.089225836919732e-05,
+      "loss": 0.0636,
+      "step": 58250
+    },
+    {
+      "epoch": 3.8115799803729145,
+      "grad_norm": 0.8722211718559265,
+      "learning_rate": 8.08850349425614e-05,
+      "loss": 0.0767,
+      "step": 58260
+    },
+    {
+      "epoch": 3.8122342165521754,
+      "grad_norm": 0.8511031866073608,
+      "learning_rate": 8.087781047347455e-05,
+      "loss": 0.0705,
+      "step": 58270
+    },
+    {
+      "epoch": 3.812888452731436,
+      "grad_norm": 0.8996208310127258,
+      "learning_rate": 8.087058496218063e-05,
+      "loss": 0.0645,
+      "step": 58280
+    },
+    {
+      "epoch": 3.813542688910697,
+      "grad_norm": 0.975518524646759,
+      "learning_rate": 8.08633584089235e-05,
+      "loss": 0.0808,
+      "step": 58290
+    },
+    {
+      "epoch": 3.8141969250899574,
+      "grad_norm": 0.9929527044296265,
+      "learning_rate": 8.085613081394708e-05,
+      "loss": 0.0718,
+      "step": 58300
+    },
+    {
+      "epoch": 3.8148511612692184,
+      "grad_norm": 0.8056535124778748,
+      "learning_rate": 8.084890217749532e-05,
+      "loss": 0.0671,
+      "step": 58310
+    },
+    {
+      "epoch": 3.815505397448479,
+      "grad_norm": 0.766545832157135,
+      "learning_rate": 8.084167249981219e-05,
+      "loss": 0.0711,
+      "step": 58320
+    },
+    {
+      "epoch": 3.8161596336277395,
+      "grad_norm": 1.0355461835861206,
+      "learning_rate": 8.083444178114174e-05,
+      "loss": 0.0759,
+      "step": 58330
+    },
+    {
+      "epoch": 3.8168138698070004,
+      "grad_norm": 0.8508825898170471,
+      "learning_rate": 8.082721002172801e-05,
+      "loss": 0.0696,
+      "step": 58340
+    },
+    {
+      "epoch": 3.817468105986261,
+      "grad_norm": 0.863624095916748,
+      "learning_rate": 8.081997722181512e-05,
+      "loss": 0.073,
+      "step": 58350
+    },
+    {
+      "epoch": 3.8181223421655215,
+      "grad_norm": 0.8428578972816467,
+      "learning_rate": 8.081274338164714e-05,
+      "loss": 0.0697,
+      "step": 58360
+    },
+    {
+      "epoch": 3.8187765783447825,
+      "grad_norm": 0.8792449235916138,
+      "learning_rate": 8.080550850146829e-05,
+      "loss": 0.0665,
+      "step": 58370
+    },
+    {
+      "epoch": 3.8194308145240434,
+      "grad_norm": 0.7911483645439148,
+      "learning_rate": 8.079827258152272e-05,
+      "loss": 0.0738,
+      "step": 58380
+    },
+    {
+      "epoch": 3.820085050703304,
+      "grad_norm": 0.9824073314666748,
+      "learning_rate": 8.079103562205468e-05,
+      "loss": 0.0693,
+      "step": 58390
+    },
+    {
+      "epoch": 3.8207392868825645,
+      "grad_norm": 0.77364581823349,
+      "learning_rate": 8.078379762330843e-05,
+      "loss": 0.0666,
+      "step": 58400
+    },
+    {
+      "epoch": 3.8213935230618254,
+      "grad_norm": 0.8866234421730042,
+      "learning_rate": 8.07765585855283e-05,
+      "loss": 0.0706,
+      "step": 58410
+    },
+    {
+      "epoch": 3.822047759241086,
+      "grad_norm": 0.7510147094726562,
+      "learning_rate": 8.076931850895859e-05,
+      "loss": 0.0679,
+      "step": 58420
+    },
+    {
+      "epoch": 3.8227019954203465,
+      "grad_norm": 0.9007807374000549,
+      "learning_rate": 8.076207739384368e-05,
+      "loss": 0.073,
+      "step": 58430
+    },
+    {
+      "epoch": 3.8233562315996075,
+      "grad_norm": 0.8261735439300537,
+      "learning_rate": 8.075483524042797e-05,
+      "loss": 0.0709,
+      "step": 58440
+    },
+    {
+      "epoch": 3.8240104677788684,
+      "grad_norm": 0.7386045455932617,
+      "learning_rate": 8.074759204895593e-05,
+      "loss": 0.0606,
+      "step": 58450
+    },
+    {
+      "epoch": 3.824664703958129,
+      "grad_norm": 1.0054618120193481,
+      "learning_rate": 8.0740347819672e-05,
+      "loss": 0.0652,
+      "step": 58460
+    },
+    {
+      "epoch": 3.8253189401373895,
+      "grad_norm": 0.8960070610046387,
+      "learning_rate": 8.073310255282074e-05,
+      "loss": 0.0707,
+      "step": 58470
+    },
+    {
+      "epoch": 3.8259731763166505,
+      "grad_norm": 0.8370187282562256,
+      "learning_rate": 8.072585624864666e-05,
+      "loss": 0.0668,
+      "step": 58480
+    },
+    {
+      "epoch": 3.826627412495911,
+      "grad_norm": 0.9938347935676575,
+      "learning_rate": 8.071860890739435e-05,
+      "loss": 0.0786,
+      "step": 58490
+    },
+    {
+      "epoch": 3.8272816486751715,
+      "grad_norm": 1.007035732269287,
+      "learning_rate": 8.071136052930843e-05,
+      "loss": 0.0697,
+      "step": 58500
+    },
+    {
+      "epoch": 3.8279358848544325,
+      "grad_norm": 0.86492520570755,
+      "learning_rate": 8.070411111463353e-05,
+      "loss": 0.0749,
+      "step": 58510
+    },
+    {
+      "epoch": 3.828590121033693,
+      "grad_norm": 0.7350853085517883,
+      "learning_rate": 8.069686066361437e-05,
+      "loss": 0.0648,
+      "step": 58520
+    },
+    {
+      "epoch": 3.829244357212954,
+      "grad_norm": 1.0213762521743774,
+      "learning_rate": 8.068960917649566e-05,
+      "loss": 0.0731,
+      "step": 58530
+    },
+    {
+      "epoch": 3.8298985933922145,
+      "grad_norm": 0.8180403113365173,
+      "learning_rate": 8.068235665352214e-05,
+      "loss": 0.0651,
+      "step": 58540
+    },
+    {
+      "epoch": 3.8305528295714755,
+      "grad_norm": 0.8237452507019043,
+      "learning_rate": 8.067510309493861e-05,
+      "loss": 0.0702,
+      "step": 58550
+    },
+    {
+      "epoch": 3.831207065750736,
+      "grad_norm": 0.7584102153778076,
+      "learning_rate": 8.066784850098992e-05,
+      "loss": 0.062,
+      "step": 58560
+    },
+    {
+      "epoch": 3.8318613019299965,
+      "grad_norm": 1.0070369243621826,
+      "learning_rate": 8.066059287192091e-05,
+      "loss": 0.0657,
+      "step": 58570
+    },
+    {
+      "epoch": 3.8325155381092575,
+      "grad_norm": 1.0255883932113647,
+      "learning_rate": 8.065333620797649e-05,
+      "loss": 0.0699,
+      "step": 58580
+    },
+    {
+      "epoch": 3.833169774288518,
+      "grad_norm": 0.8956732153892517,
+      "learning_rate": 8.064607850940156e-05,
+      "loss": 0.0733,
+      "step": 58590
+    },
+    {
+      "epoch": 3.833824010467779,
+      "grad_norm": 0.8609063625335693,
+      "learning_rate": 8.063881977644112e-05,
+      "loss": 0.0692,
+      "step": 58600
+    },
+    {
+      "epoch": 3.8344782466470395,
+      "grad_norm": 1.1151846647262573,
+      "learning_rate": 8.063156000934016e-05,
+      "loss": 0.0674,
+      "step": 58610
+    },
+    {
+      "epoch": 3.8351324828263005,
+      "grad_norm": 0.8186620473861694,
+      "learning_rate": 8.062429920834372e-05,
+      "loss": 0.0711,
+      "step": 58620
+    },
+    {
+      "epoch": 3.835786719005561,
+      "grad_norm": 0.8908758759498596,
+      "learning_rate": 8.061703737369686e-05,
+      "loss": 0.0635,
+      "step": 58630
+    },
+    {
+      "epoch": 3.8364409551848215,
+      "grad_norm": 1.0307601690292358,
+      "learning_rate": 8.060977450564469e-05,
+      "loss": 0.0708,
+      "step": 58640
+    },
+    {
+      "epoch": 3.8370951913640825,
+      "grad_norm": 0.8627874851226807,
+      "learning_rate": 8.060251060443236e-05,
+      "loss": 0.0646,
+      "step": 58650
+    },
+    {
+      "epoch": 3.837749427543343,
+      "grad_norm": 0.9624312520027161,
+      "learning_rate": 8.059524567030503e-05,
+      "loss": 0.0685,
+      "step": 58660
+    },
+    {
+      "epoch": 3.838403663722604,
+      "grad_norm": 0.8009755611419678,
+      "learning_rate": 8.058797970350793e-05,
+      "loss": 0.064,
+      "step": 58670
+    },
+    {
+      "epoch": 3.8390578999018645,
+      "grad_norm": 0.9126613140106201,
+      "learning_rate": 8.058071270428628e-05,
+      "loss": 0.0725,
+      "step": 58680
+    },
+    {
+      "epoch": 3.8397121360811255,
+      "grad_norm": 0.7582046389579773,
+      "learning_rate": 8.057344467288539e-05,
+      "loss": 0.0622,
+      "step": 58690
+    },
+    {
+      "epoch": 3.840366372260386,
+      "grad_norm": 0.7236906886100769,
+      "learning_rate": 8.056617560955056e-05,
+      "loss": 0.0658,
+      "step": 58700
+    },
+    {
+      "epoch": 3.8410206084396465,
+      "grad_norm": 0.7379127144813538,
+      "learning_rate": 8.055890551452714e-05,
+      "loss": 0.0655,
+      "step": 58710
+    },
+    {
+      "epoch": 3.8416748446189075,
+      "grad_norm": 0.8637368679046631,
+      "learning_rate": 8.055163438806051e-05,
+      "loss": 0.0803,
+      "step": 58720
+    },
+    {
+      "epoch": 3.842329080798168,
+      "grad_norm": 0.9424805045127869,
+      "learning_rate": 8.05443622303961e-05,
+      "loss": 0.0723,
+      "step": 58730
+    },
+    {
+      "epoch": 3.842983316977429,
+      "grad_norm": 0.9135764241218567,
+      "learning_rate": 8.053708904177934e-05,
+      "loss": 0.0717,
+      "step": 58740
+    },
+    {
+      "epoch": 3.8436375531566895,
+      "grad_norm": 0.7317261695861816,
+      "learning_rate": 8.052981482245577e-05,
+      "loss": 0.0679,
+      "step": 58750
+    },
+    {
+      "epoch": 3.8442917893359505,
+      "grad_norm": 1.1053870916366577,
+      "learning_rate": 8.052253957267086e-05,
+      "loss": 0.0699,
+      "step": 58760
+    },
+    {
+      "epoch": 3.844946025515211,
+      "grad_norm": 1.033487319946289,
+      "learning_rate": 8.05152632926702e-05,
+      "loss": 0.0776,
+      "step": 58770
+    },
+    {
+      "epoch": 3.8456002616944716,
+      "grad_norm": 1.0361863374710083,
+      "learning_rate": 8.050798598269937e-05,
+      "loss": 0.0778,
+      "step": 58780
+    },
+    {
+      "epoch": 3.8462544978737325,
+      "grad_norm": 0.8072547912597656,
+      "learning_rate": 8.050070764300401e-05,
+      "loss": 0.0695,
+      "step": 58790
+    },
+    {
+      "epoch": 3.846908734052993,
+      "grad_norm": 0.7639461159706116,
+      "learning_rate": 8.049342827382977e-05,
+      "loss": 0.0718,
+      "step": 58800
+    },
+    {
+      "epoch": 3.8475629702322536,
+      "grad_norm": 0.9850027561187744,
+      "learning_rate": 8.048614787542234e-05,
+      "loss": 0.0647,
+      "step": 58810
+    },
+    {
+      "epoch": 3.8482172064115145,
+      "grad_norm": 0.8612027168273926,
+      "learning_rate": 8.047886644802749e-05,
+      "loss": 0.0667,
+      "step": 58820
+    },
+    {
+      "epoch": 3.8488714425907755,
+      "grad_norm": 0.7390583157539368,
+      "learning_rate": 8.047158399189096e-05,
+      "loss": 0.0606,
+      "step": 58830
+    },
+    {
+      "epoch": 3.849525678770036,
+      "grad_norm": 0.8510991930961609,
+      "learning_rate": 8.046430050725854e-05,
+      "loss": 0.0688,
+      "step": 58840
+    },
+    {
+      "epoch": 3.8501799149492966,
+      "grad_norm": 0.8368331789970398,
+      "learning_rate": 8.045701599437609e-05,
+      "loss": 0.0683,
+      "step": 58850
+    },
+    {
+      "epoch": 3.8508341511285575,
+      "grad_norm": 1.0364656448364258,
+      "learning_rate": 8.044973045348949e-05,
+      "loss": 0.0755,
+      "step": 58860
+    },
+    {
+      "epoch": 3.851488387307818,
+      "grad_norm": 0.7408995628356934,
+      "learning_rate": 8.044244388484463e-05,
+      "loss": 0.0695,
+      "step": 58870
+    },
+    {
+      "epoch": 3.8521426234870786,
+      "grad_norm": 0.7264641523361206,
+      "learning_rate": 8.043515628868743e-05,
+      "loss": 0.0771,
+      "step": 58880
+    },
+    {
+      "epoch": 3.8527968596663396,
+      "grad_norm": 0.8603324294090271,
+      "learning_rate": 8.042786766526389e-05,
+      "loss": 0.0619,
+      "step": 58890
+    },
+    {
+      "epoch": 3.8534510958456005,
+      "grad_norm": 1.0376695394515991,
+      "learning_rate": 8.042057801482001e-05,
+      "loss": 0.0681,
+      "step": 58900
+    },
+    {
+      "epoch": 3.854105332024861,
+      "grad_norm": 0.7530226111412048,
+      "learning_rate": 8.041328733760185e-05,
+      "loss": 0.0709,
+      "step": 58910
+    },
+    {
+      "epoch": 3.8547595682041216,
+      "grad_norm": 0.8410870432853699,
+      "learning_rate": 8.040599563385548e-05,
+      "loss": 0.0628,
+      "step": 58920
+    },
+    {
+      "epoch": 3.8554138043833825,
+      "grad_norm": 1.0103882551193237,
+      "learning_rate": 8.039870290382703e-05,
+      "loss": 0.0743,
+      "step": 58930
+    },
+    {
+      "epoch": 3.856068040562643,
+      "grad_norm": 0.925134539604187,
+      "learning_rate": 8.039140914776262e-05,
+      "loss": 0.0767,
+      "step": 58940
+    },
+    {
+      "epoch": 3.8567222767419036,
+      "grad_norm": 0.9100943803787231,
+      "learning_rate": 8.038411436590845e-05,
+      "loss": 0.0697,
+      "step": 58950
+    },
+    {
+      "epoch": 3.8573765129211646,
+      "grad_norm": 0.9153822660446167,
+      "learning_rate": 8.037681855851072e-05,
+      "loss": 0.0622,
+      "step": 58960
+    },
+    {
+      "epoch": 3.858030749100425,
+      "grad_norm": 0.9018275737762451,
+      "learning_rate": 8.036952172581571e-05,
+      "loss": 0.0669,
+      "step": 58970
+    },
+    {
+      "epoch": 3.858684985279686,
+      "grad_norm": 0.9553847908973694,
+      "learning_rate": 8.03622238680697e-05,
+      "loss": 0.0738,
+      "step": 58980
+    },
+    {
+      "epoch": 3.8593392214589466,
+      "grad_norm": 0.8879307508468628,
+      "learning_rate": 8.0354924985519e-05,
+      "loss": 0.0707,
+      "step": 58990
+    },
+    {
+      "epoch": 3.8599934576382076,
+      "grad_norm": 0.8077122569084167,
+      "learning_rate": 8.034762507840997e-05,
+      "loss": 0.0819,
+      "step": 59000
+    },
+    {
+      "epoch": 3.860647693817468,
+      "grad_norm": 0.7566487193107605,
+      "learning_rate": 8.034032414698901e-05,
+      "loss": 0.0752,
+      "step": 59010
+    },
+    {
+      "epoch": 3.8613019299967286,
+      "grad_norm": 0.8122360706329346,
+      "learning_rate": 8.033302219150253e-05,
+      "loss": 0.0715,
+      "step": 59020
+    },
+    {
+      "epoch": 3.8619561661759896,
+      "grad_norm": 1.0947107076644897,
+      "learning_rate": 8.0325719212197e-05,
+      "loss": 0.0641,
+      "step": 59030
+    },
+    {
+      "epoch": 3.86261040235525,
+      "grad_norm": 0.782038688659668,
+      "learning_rate": 8.031841520931893e-05,
+      "loss": 0.0608,
+      "step": 59040
+    },
+    {
+      "epoch": 3.863264638534511,
+      "grad_norm": 0.7737072706222534,
+      "learning_rate": 8.031111018311483e-05,
+      "loss": 0.0664,
+      "step": 59050
+    },
+    {
+      "epoch": 3.8639188747137716,
+      "grad_norm": 0.8657538890838623,
+      "learning_rate": 8.030380413383125e-05,
+      "loss": 0.0669,
+      "step": 59060
+    },
+    {
+      "epoch": 3.8645731108930326,
+      "grad_norm": 0.8043084144592285,
+      "learning_rate": 8.029649706171483e-05,
+      "loss": 0.0719,
+      "step": 59070
+    },
+    {
+      "epoch": 3.865227347072293,
+      "grad_norm": 0.7678789496421814,
+      "learning_rate": 8.028918896701217e-05,
+      "loss": 0.0679,
+      "step": 59080
+    },
+    {
+      "epoch": 3.8658815832515536,
+      "grad_norm": 0.935185432434082,
+      "learning_rate": 8.028187984996993e-05,
+      "loss": 0.0701,
+      "step": 59090
+    },
+    {
+      "epoch": 3.8665358194308146,
+      "grad_norm": 0.9353100061416626,
+      "learning_rate": 8.027456971083485e-05,
+      "loss": 0.0636,
+      "step": 59100
+    },
+    {
+      "epoch": 3.867190055610075,
+      "grad_norm": 0.8760929107666016,
+      "learning_rate": 8.026725854985363e-05,
+      "loss": 0.0776,
+      "step": 59110
+    },
+    {
+      "epoch": 3.867844291789336,
+      "grad_norm": 0.9680326581001282,
+      "learning_rate": 8.025994636727306e-05,
+      "loss": 0.0749,
+      "step": 59120
+    },
+    {
+      "epoch": 3.8684985279685966,
+      "grad_norm": 0.6920056343078613,
+      "learning_rate": 8.025263316333994e-05,
+      "loss": 0.0715,
+      "step": 59130
+    },
+    {
+      "epoch": 3.8691527641478576,
+      "grad_norm": 0.9525841474533081,
+      "learning_rate": 8.024531893830112e-05,
+      "loss": 0.0673,
+      "step": 59140
+    },
+    {
+      "epoch": 3.869807000327118,
+      "grad_norm": 0.8676741719245911,
+      "learning_rate": 8.023800369240344e-05,
+      "loss": 0.0744,
+      "step": 59150
+    },
+    {
+      "epoch": 3.8704612365063786,
+      "grad_norm": 0.8752907514572144,
+      "learning_rate": 8.023068742589386e-05,
+      "loss": 0.0646,
+      "step": 59160
+    },
+    {
+      "epoch": 3.8711154726856396,
+      "grad_norm": 0.9634379744529724,
+      "learning_rate": 8.022337013901928e-05,
+      "loss": 0.0594,
+      "step": 59170
+    },
+    {
+      "epoch": 3.8717697088649,
+      "grad_norm": 0.7896409034729004,
+      "learning_rate": 8.02160518320267e-05,
+      "loss": 0.072,
+      "step": 59180
+    },
+    {
+      "epoch": 3.872423945044161,
+      "grad_norm": 0.8738341927528381,
+      "learning_rate": 8.020873250516312e-05,
+      "loss": 0.0677,
+      "step": 59190
+    },
+    {
+      "epoch": 3.8730781812234216,
+      "grad_norm": 0.8021124601364136,
+      "learning_rate": 8.02014121586756e-05,
+      "loss": 0.0841,
+      "step": 59200
+    },
+    {
+      "epoch": 3.8737324174026826,
+      "grad_norm": 0.9408907890319824,
+      "learning_rate": 8.019409079281122e-05,
+      "loss": 0.0625,
+      "step": 59210
+    },
+    {
+      "epoch": 3.874386653581943,
+      "grad_norm": 0.8680176138877869,
+      "learning_rate": 8.018676840781707e-05,
+      "loss": 0.0702,
+      "step": 59220
+    },
+    {
+      "epoch": 3.8750408897612036,
+      "grad_norm": 0.6961979269981384,
+      "learning_rate": 8.017944500394033e-05,
+      "loss": 0.0578,
+      "step": 59230
+    },
+    {
+      "epoch": 3.8756951259404646,
+      "grad_norm": 0.8254890441894531,
+      "learning_rate": 8.017212058142817e-05,
+      "loss": 0.0717,
+      "step": 59240
+    },
+    {
+      "epoch": 3.876349362119725,
+      "grad_norm": 1.0662809610366821,
+      "learning_rate": 8.016479514052783e-05,
+      "loss": 0.0675,
+      "step": 59250
+    },
+    {
+      "epoch": 3.8770035982989857,
+      "grad_norm": 0.8766838908195496,
+      "learning_rate": 8.015746868148651e-05,
+      "loss": 0.0688,
+      "step": 59260
+    },
+    {
+      "epoch": 3.8776578344782466,
+      "grad_norm": 0.8927856683731079,
+      "learning_rate": 8.015014120455156e-05,
+      "loss": 0.0628,
+      "step": 59270
+    },
+    {
+      "epoch": 3.8783120706575076,
+      "grad_norm": 0.8246244192123413,
+      "learning_rate": 8.014281270997026e-05,
+      "loss": 0.0691,
+      "step": 59280
+    },
+    {
+      "epoch": 3.878966306836768,
+      "grad_norm": 0.9650526642799377,
+      "learning_rate": 8.013548319798998e-05,
+      "loss": 0.0771,
+      "step": 59290
+    },
+    {
+      "epoch": 3.8796205430160287,
+      "grad_norm": 0.6766818165779114,
+      "learning_rate": 8.012815266885811e-05,
+      "loss": 0.0567,
+      "step": 59300
+    },
+    {
+      "epoch": 3.8802747791952896,
+      "grad_norm": 0.8025507926940918,
+      "learning_rate": 8.012082112282207e-05,
+      "loss": 0.0614,
+      "step": 59310
+    },
+    {
+      "epoch": 3.88092901537455,
+      "grad_norm": 0.8738349080085754,
+      "learning_rate": 8.011348856012932e-05,
+      "loss": 0.0754,
+      "step": 59320
+    },
+    {
+      "epoch": 3.8815832515538107,
+      "grad_norm": 0.819463312625885,
+      "learning_rate": 8.010615498102736e-05,
+      "loss": 0.0634,
+      "step": 59330
+    },
+    {
+      "epoch": 3.8822374877330716,
+      "grad_norm": 0.7592009902000427,
+      "learning_rate": 8.009882038576371e-05,
+      "loss": 0.0649,
+      "step": 59340
+    },
+    {
+      "epoch": 3.8828917239123326,
+      "grad_norm": 1.0910615921020508,
+      "learning_rate": 8.009148477458594e-05,
+      "loss": 0.0777,
+      "step": 59350
+    },
+    {
+      "epoch": 3.883545960091593,
+      "grad_norm": 0.8218048214912415,
+      "learning_rate": 8.008414814774163e-05,
+      "loss": 0.064,
+      "step": 59360
+    },
+    {
+      "epoch": 3.8842001962708537,
+      "grad_norm": 0.8546844720840454,
+      "learning_rate": 8.007681050547844e-05,
+      "loss": 0.0787,
+      "step": 59370
+    },
+    {
+      "epoch": 3.8848544324501146,
+      "grad_norm": 0.8277244567871094,
+      "learning_rate": 8.0069471848044e-05,
+      "loss": 0.0666,
+      "step": 59380
+    },
+    {
+      "epoch": 3.885508668629375,
+      "grad_norm": 0.7746628522872925,
+      "learning_rate": 8.006213217568604e-05,
+      "loss": 0.0641,
+      "step": 59390
+    },
+    {
+      "epoch": 3.8861629048086357,
+      "grad_norm": 1.0281668901443481,
+      "learning_rate": 8.005479148865226e-05,
+      "loss": 0.0729,
+      "step": 59400
+    },
+    {
+      "epoch": 3.8868171409878967,
+      "grad_norm": 0.7792064547538757,
+      "learning_rate": 8.004744978719046e-05,
+      "loss": 0.0722,
+      "step": 59410
+    },
+    {
+      "epoch": 3.887471377167157,
+      "grad_norm": 0.7201108336448669,
+      "learning_rate": 8.004010707154843e-05,
+      "loss": 0.0694,
+      "step": 59420
+    },
+    {
+      "epoch": 3.888125613346418,
+      "grad_norm": 0.8858181834220886,
+      "learning_rate": 8.003276334197399e-05,
+      "loss": 0.0624,
+      "step": 59430
+    },
+    {
+      "epoch": 3.8887798495256787,
+      "grad_norm": 0.8069655895233154,
+      "learning_rate": 8.002541859871502e-05,
+      "loss": 0.0636,
+      "step": 59440
+    },
+    {
+      "epoch": 3.8894340857049396,
+      "grad_norm": 0.812025249004364,
+      "learning_rate": 8.001807284201944e-05,
+      "loss": 0.0638,
+      "step": 59450
+    },
+    {
+      "epoch": 3.8900883218842,
+      "grad_norm": 0.8210043907165527,
+      "learning_rate": 8.001072607213518e-05,
+      "loss": 0.0755,
+      "step": 59460
+    },
+    {
+      "epoch": 3.8907425580634607,
+      "grad_norm": 0.8998602032661438,
+      "learning_rate": 8.000337828931021e-05,
+      "loss": 0.0702,
+      "step": 59470
+    },
+    {
+      "epoch": 3.8913967942427217,
+      "grad_norm": 0.7417557239532471,
+      "learning_rate": 7.999602949379252e-05,
+      "loss": 0.0666,
+      "step": 59480
+    },
+    {
+      "epoch": 3.892051030421982,
+      "grad_norm": 0.778843104839325,
+      "learning_rate": 7.998867968583018e-05,
+      "loss": 0.0689,
+      "step": 59490
+    },
+    {
+      "epoch": 3.892705266601243,
+      "grad_norm": 0.7285590767860413,
+      "learning_rate": 7.998132886567125e-05,
+      "loss": 0.0667,
+      "step": 59500
+    },
+    {
+      "epoch": 3.8933595027805037,
+      "grad_norm": 0.8084492683410645,
+      "learning_rate": 7.997397703356384e-05,
+      "loss": 0.0588,
+      "step": 59510
+    },
+    {
+      "epoch": 3.8940137389597647,
+      "grad_norm": 0.8009549379348755,
+      "learning_rate": 7.996662418975609e-05,
+      "loss": 0.0634,
+      "step": 59520
+    },
+    {
+      "epoch": 3.894667975139025,
+      "grad_norm": 0.8678660988807678,
+      "learning_rate": 7.99592703344962e-05,
+      "loss": 0.0692,
+      "step": 59530
+    },
+    {
+      "epoch": 3.8953222113182857,
+      "grad_norm": 1.06533682346344,
+      "learning_rate": 7.995191546803235e-05,
+      "loss": 0.0712,
+      "step": 59540
+    },
+    {
+      "epoch": 3.8959764474975467,
+      "grad_norm": 1.0511845350265503,
+      "learning_rate": 7.99445595906128e-05,
+      "loss": 0.0764,
+      "step": 59550
+    },
+    {
+      "epoch": 3.896630683676807,
+      "grad_norm": 0.9604825973510742,
+      "learning_rate": 7.993720270248584e-05,
+      "loss": 0.0743,
+      "step": 59560
+    },
+    {
+      "epoch": 3.897284919856068,
+      "grad_norm": 0.9758427739143372,
+      "learning_rate": 7.992984480389977e-05,
+      "loss": 0.0586,
+      "step": 59570
+    },
+    {
+      "epoch": 3.8979391560353287,
+      "grad_norm": 0.862629771232605,
+      "learning_rate": 7.992248589510293e-05,
+      "loss": 0.0597,
+      "step": 59580
+    },
+    {
+      "epoch": 3.8985933922145897,
+      "grad_norm": 0.876379132270813,
+      "learning_rate": 7.991512597634375e-05,
+      "loss": 0.0651,
+      "step": 59590
+    },
+    {
+      "epoch": 3.89924762839385,
+      "grad_norm": 1.1144251823425293,
+      "learning_rate": 7.990776504787059e-05,
+      "loss": 0.0755,
+      "step": 59600
+    },
+    {
+      "epoch": 3.8999018645731107,
+      "grad_norm": 0.9200652837753296,
+      "learning_rate": 7.990040310993193e-05,
+      "loss": 0.058,
+      "step": 59610
+    },
+    {
+      "epoch": 3.9005561007523717,
+      "grad_norm": 0.8671579360961914,
+      "learning_rate": 7.989304016277625e-05,
+      "loss": 0.0728,
+      "step": 59620
+    },
+    {
+      "epoch": 3.901210336931632,
+      "grad_norm": 0.8644618391990662,
+      "learning_rate": 7.988567620665206e-05,
+      "loss": 0.0712,
+      "step": 59630
+    },
+    {
+      "epoch": 3.901864573110893,
+      "grad_norm": 1.062548279762268,
+      "learning_rate": 7.987831124180792e-05,
+      "loss": 0.0605,
+      "step": 59640
+    },
+    {
+      "epoch": 3.9025188092901537,
+      "grad_norm": 0.7029675841331482,
+      "learning_rate": 7.987094526849242e-05,
+      "loss": 0.0605,
+      "step": 59650
+    },
+    {
+      "epoch": 3.9031730454694147,
+      "grad_norm": 0.7304003834724426,
+      "learning_rate": 7.986357828695419e-05,
+      "loss": 0.0624,
+      "step": 59660
+    },
+    {
+      "epoch": 3.903827281648675,
+      "grad_norm": 0.8633514046669006,
+      "learning_rate": 7.985621029744186e-05,
+      "loss": 0.074,
+      "step": 59670
+    },
+    {
+      "epoch": 3.9044815178279357,
+      "grad_norm": 1.0137802362442017,
+      "learning_rate": 7.984884130020414e-05,
+      "loss": 0.0597,
+      "step": 59680
+    },
+    {
+      "epoch": 3.9051357540071967,
+      "grad_norm": 1.002747893333435,
+      "learning_rate": 7.984147129548973e-05,
+      "loss": 0.0765,
+      "step": 59690
+    },
+    {
+      "epoch": 3.9057899901864572,
+      "grad_norm": 0.7557197213172913,
+      "learning_rate": 7.983410028354741e-05,
+      "loss": 0.0758,
+      "step": 59700
+    },
+    {
+      "epoch": 3.9064442263657178,
+      "grad_norm": 1.0572227239608765,
+      "learning_rate": 7.982672826462595e-05,
+      "loss": 0.0619,
+      "step": 59710
+    },
+    {
+      "epoch": 3.9070984625449787,
+      "grad_norm": 0.8649874925613403,
+      "learning_rate": 7.981935523897421e-05,
+      "loss": 0.0617,
+      "step": 59720
+    },
+    {
+      "epoch": 3.9077526987242397,
+      "grad_norm": 0.711691677570343,
+      "learning_rate": 7.981198120684101e-05,
+      "loss": 0.066,
+      "step": 59730
+    },
+    {
+      "epoch": 3.9084069349035,
+      "grad_norm": 0.9565288424491882,
+      "learning_rate": 7.980460616847527e-05,
+      "loss": 0.0708,
+      "step": 59740
+    },
+    {
+      "epoch": 3.9090611710827607,
+      "grad_norm": 0.9531120657920837,
+      "learning_rate": 7.97972301241259e-05,
+      "loss": 0.0662,
+      "step": 59750
+    },
+    {
+      "epoch": 3.9097154072620217,
+      "grad_norm": 0.8800640106201172,
+      "learning_rate": 7.978985307404187e-05,
+      "loss": 0.0723,
+      "step": 59760
+    },
+    {
+      "epoch": 3.9103696434412822,
+      "grad_norm": 0.7907775044441223,
+      "learning_rate": 7.978247501847216e-05,
+      "loss": 0.0713,
+      "step": 59770
+    },
+    {
+      "epoch": 3.9110238796205428,
+      "grad_norm": 0.8069142699241638,
+      "learning_rate": 7.977509595766583e-05,
+      "loss": 0.0609,
+      "step": 59780
+    },
+    {
+      "epoch": 3.9116781157998037,
+      "grad_norm": 0.9594196677207947,
+      "learning_rate": 7.976771589187193e-05,
+      "loss": 0.0742,
+      "step": 59790
+    },
+    {
+      "epoch": 3.9123323519790647,
+      "grad_norm": 0.9948168992996216,
+      "learning_rate": 7.976033482133953e-05,
+      "loss": 0.0713,
+      "step": 59800
+    },
+    {
+      "epoch": 3.9129865881583252,
+      "grad_norm": 1.1436748504638672,
+      "learning_rate": 7.975295274631777e-05,
+      "loss": 0.0675,
+      "step": 59810
+    },
+    {
+      "epoch": 3.9136408243375858,
+      "grad_norm": 0.8440989255905151,
+      "learning_rate": 7.974556966705584e-05,
+      "loss": 0.0783,
+      "step": 59820
+    },
+    {
+      "epoch": 3.9142950605168467,
+      "grad_norm": 0.7184342741966248,
+      "learning_rate": 7.973818558380294e-05,
+      "loss": 0.0633,
+      "step": 59830
+    },
+    {
+      "epoch": 3.9149492966961073,
+      "grad_norm": 0.8069654703140259,
+      "learning_rate": 7.973080049680825e-05,
+      "loss": 0.0722,
+      "step": 59840
+    },
+    {
+      "epoch": 3.9156035328753678,
+      "grad_norm": 1.0971922874450684,
+      "learning_rate": 7.972341440632109e-05,
+      "loss": 0.0677,
+      "step": 59850
+    },
+    {
+      "epoch": 3.9162577690546287,
+      "grad_norm": 0.946134626865387,
+      "learning_rate": 7.971602731259075e-05,
+      "loss": 0.0721,
+      "step": 59860
+    },
+    {
+      "epoch": 3.9169120052338893,
+      "grad_norm": 0.9672790765762329,
+      "learning_rate": 7.970863921586655e-05,
+      "loss": 0.0707,
+      "step": 59870
+    },
+    {
+      "epoch": 3.9175662414131502,
+      "grad_norm": 1.017909288406372,
+      "learning_rate": 7.970125011639786e-05,
+      "loss": 0.0661,
+      "step": 59880
+    },
+    {
+      "epoch": 3.9182204775924108,
+      "grad_norm": 0.9773590564727783,
+      "learning_rate": 7.969386001443408e-05,
+      "loss": 0.0664,
+      "step": 59890
+    },
+    {
+      "epoch": 3.9188747137716717,
+      "grad_norm": 1.026491403579712,
+      "learning_rate": 7.968646891022466e-05,
+      "loss": 0.0707,
+      "step": 59900
+    },
+    {
+      "epoch": 3.9195289499509323,
+      "grad_norm": 0.9397926926612854,
+      "learning_rate": 7.967907680401904e-05,
+      "loss": 0.0628,
+      "step": 59910
+    },
+    {
+      "epoch": 3.920183186130193,
+      "grad_norm": 0.814552366733551,
+      "learning_rate": 7.967168369606676e-05,
+      "loss": 0.0702,
+      "step": 59920
+    },
+    {
+      "epoch": 3.9208374223094538,
+      "grad_norm": 1.042052149772644,
+      "learning_rate": 7.966428958661734e-05,
+      "loss": 0.0678,
+      "step": 59930
+    },
+    {
+      "epoch": 3.9214916584887143,
+      "grad_norm": 0.8458235859870911,
+      "learning_rate": 7.965689447592035e-05,
+      "loss": 0.0752,
+      "step": 59940
+    },
+    {
+      "epoch": 3.9221458946679753,
+      "grad_norm": 0.748357355594635,
+      "learning_rate": 7.964949836422537e-05,
+      "loss": 0.0692,
+      "step": 59950
+    },
+    {
+      "epoch": 3.9228001308472358,
+      "grad_norm": 0.9684262275695801,
+      "learning_rate": 7.964210125178209e-05,
+      "loss": 0.0646,
+      "step": 59960
+    },
+    {
+      "epoch": 3.9234543670264967,
+      "grad_norm": 1.0138685703277588,
+      "learning_rate": 7.963470313884011e-05,
+      "loss": 0.0704,
+      "step": 59970
+    },
+    {
+      "epoch": 3.9241086032057573,
+      "grad_norm": 0.9098789095878601,
+      "learning_rate": 7.962730402564924e-05,
+      "loss": 0.0663,
+      "step": 59980
+    },
+    {
+      "epoch": 3.924762839385018,
+      "grad_norm": 0.9759104251861572,
+      "learning_rate": 7.961990391245911e-05,
+      "loss": 0.0704,
+      "step": 59990
+    },
+    {
+      "epoch": 3.9254170755642788,
+      "grad_norm": 0.8418338894844055,
+      "learning_rate": 7.961250279951956e-05,
+      "loss": 0.066,
+      "step": 60000
+    },
+    {
+      "epoch": 3.9260713117435393,
+      "grad_norm": 0.8269692659378052,
+      "learning_rate": 7.960510068708039e-05,
+      "loss": 0.0734,
+      "step": 60010
+    },
+    {
+      "epoch": 3.9267255479228003,
+      "grad_norm": 1.259716510772705,
+      "learning_rate": 7.959769757539142e-05,
+      "loss": 0.0765,
+      "step": 60020
+    },
+    {
+      "epoch": 3.927379784102061,
+      "grad_norm": 0.7313357591629028,
+      "learning_rate": 7.959029346470252e-05,
+      "loss": 0.0689,
+      "step": 60030
+    },
+    {
+      "epoch": 3.9280340202813218,
+      "grad_norm": 1.043642282485962,
+      "learning_rate": 7.958288835526362e-05,
+      "loss": 0.0626,
+      "step": 60040
+    },
+    {
+      "epoch": 3.9286882564605823,
+      "grad_norm": 1.0208543539047241,
+      "learning_rate": 7.957548224732467e-05,
+      "loss": 0.0666,
+      "step": 60050
+    },
+    {
+      "epoch": 3.929342492639843,
+      "grad_norm": 0.9676765203475952,
+      "learning_rate": 7.956807514113562e-05,
+      "loss": 0.0722,
+      "step": 60060
+    },
+    {
+      "epoch": 3.929996728819104,
+      "grad_norm": 0.8631986975669861,
+      "learning_rate": 7.956066703694647e-05,
+      "loss": 0.0663,
+      "step": 60070
+    },
+    {
+      "epoch": 3.9306509649983643,
+      "grad_norm": 0.9647482633590698,
+      "learning_rate": 7.95532579350073e-05,
+      "loss": 0.064,
+      "step": 60080
+    },
+    {
+      "epoch": 3.9313052011776253,
+      "grad_norm": 0.7296038866043091,
+      "learning_rate": 7.954584783556818e-05,
+      "loss": 0.0616,
+      "step": 60090
+    },
+    {
+      "epoch": 3.931959437356886,
+      "grad_norm": 0.8985334038734436,
+      "learning_rate": 7.953843673887919e-05,
+      "loss": 0.0698,
+      "step": 60100
+    },
+    {
+      "epoch": 3.9326136735361468,
+      "grad_norm": 0.88104248046875,
+      "learning_rate": 7.953102464519049e-05,
+      "loss": 0.0714,
+      "step": 60110
+    },
+    {
+      "epoch": 3.9332679097154073,
+      "grad_norm": 0.8251603245735168,
+      "learning_rate": 7.952361155475228e-05,
+      "loss": 0.0693,
+      "step": 60120
+    },
+    {
+      "epoch": 3.933922145894668,
+      "grad_norm": 0.8818602561950684,
+      "learning_rate": 7.951619746781474e-05,
+      "loss": 0.0675,
+      "step": 60130
+    },
+    {
+      "epoch": 3.934576382073929,
+      "grad_norm": 0.7289047837257385,
+      "learning_rate": 7.950878238462812e-05,
+      "loss": 0.0592,
+      "step": 60140
+    },
+    {
+      "epoch": 3.9352306182531893,
+      "grad_norm": 1.0739281177520752,
+      "learning_rate": 7.950136630544272e-05,
+      "loss": 0.0726,
+      "step": 60150
+    },
+    {
+      "epoch": 3.93588485443245,
+      "grad_norm": 0.9551029801368713,
+      "learning_rate": 7.949394923050882e-05,
+      "loss": 0.0606,
+      "step": 60160
+    },
+    {
+      "epoch": 3.936539090611711,
+      "grad_norm": 0.7134016752243042,
+      "learning_rate": 7.94865311600768e-05,
+      "loss": 0.0611,
+      "step": 60170
+    },
+    {
+      "epoch": 3.937193326790972,
+      "grad_norm": 0.8816996216773987,
+      "learning_rate": 7.9479112094397e-05,
+      "loss": 0.0675,
+      "step": 60180
+    },
+    {
+      "epoch": 3.9378475629702323,
+      "grad_norm": 1.0640590190887451,
+      "learning_rate": 7.947169203371986e-05,
+      "loss": 0.0742,
+      "step": 60190
+    },
+    {
+      "epoch": 3.938501799149493,
+      "grad_norm": 0.8332981467247009,
+      "learning_rate": 7.946427097829584e-05,
+      "loss": 0.0756,
+      "step": 60200
+    },
+    {
+      "epoch": 3.939156035328754,
+      "grad_norm": 0.9748693704605103,
+      "learning_rate": 7.94568489283754e-05,
+      "loss": 0.066,
+      "step": 60210
+    },
+    {
+      "epoch": 3.9398102715080143,
+      "grad_norm": 0.9146521687507629,
+      "learning_rate": 7.944942588420903e-05,
+      "loss": 0.0652,
+      "step": 60220
+    },
+    {
+      "epoch": 3.940464507687275,
+      "grad_norm": 0.869734525680542,
+      "learning_rate": 7.944200184604732e-05,
+      "loss": 0.0743,
+      "step": 60230
+    },
+    {
+      "epoch": 3.941118743866536,
+      "grad_norm": 1.1516258716583252,
+      "learning_rate": 7.943457681414084e-05,
+      "loss": 0.0722,
+      "step": 60240
+    },
+    {
+      "epoch": 3.941772980045797,
+      "grad_norm": 0.8145464658737183,
+      "learning_rate": 7.942715078874019e-05,
+      "loss": 0.0613,
+      "step": 60250
+    },
+    {
+      "epoch": 3.9424272162250573,
+      "grad_norm": 1.0588626861572266,
+      "learning_rate": 7.941972377009601e-05,
+      "loss": 0.0701,
+      "step": 60260
+    },
+    {
+      "epoch": 3.943081452404318,
+      "grad_norm": 0.8844826221466064,
+      "learning_rate": 7.941229575845903e-05,
+      "loss": 0.0676,
+      "step": 60270
+    },
+    {
+      "epoch": 3.943735688583579,
+      "grad_norm": 0.7616518139839172,
+      "learning_rate": 7.94048667540799e-05,
+      "loss": 0.0708,
+      "step": 60280
+    },
+    {
+      "epoch": 3.9443899247628393,
+      "grad_norm": 1.0263601541519165,
+      "learning_rate": 7.939743675720942e-05,
+      "loss": 0.0709,
+      "step": 60290
+    },
+    {
+      "epoch": 3.9450441609421,
+      "grad_norm": 0.725226640701294,
+      "learning_rate": 7.939000576809834e-05,
+      "loss": 0.0628,
+      "step": 60300
+    },
+    {
+      "epoch": 3.945698397121361,
+      "grad_norm": 0.8320260047912598,
+      "learning_rate": 7.93825737869975e-05,
+      "loss": 0.0657,
+      "step": 60310
+    },
+    {
+      "epoch": 3.9463526333006214,
+      "grad_norm": 0.9669265747070312,
+      "learning_rate": 7.937514081415773e-05,
+      "loss": 0.0732,
+      "step": 60320
+    },
+    {
+      "epoch": 3.9470068694798823,
+      "grad_norm": 1.055615782737732,
+      "learning_rate": 7.936770684982992e-05,
+      "loss": 0.0672,
+      "step": 60330
+    },
+    {
+      "epoch": 3.947661105659143,
+      "grad_norm": 0.8739088177680969,
+      "learning_rate": 7.936027189426497e-05,
+      "loss": 0.069,
+      "step": 60340
+    },
+    {
+      "epoch": 3.948315341838404,
+      "grad_norm": 1.0369230508804321,
+      "learning_rate": 7.935283594771385e-05,
+      "loss": 0.0676,
+      "step": 60350
+    },
+    {
+      "epoch": 3.9489695780176644,
+      "grad_norm": 0.9298377633094788,
+      "learning_rate": 7.934539901042754e-05,
+      "loss": 0.0865,
+      "step": 60360
+    },
+    {
+      "epoch": 3.949623814196925,
+      "grad_norm": 1.1504807472229004,
+      "learning_rate": 7.933796108265705e-05,
+      "loss": 0.06,
+      "step": 60370
+    },
+    {
+      "epoch": 3.950278050376186,
+      "grad_norm": 0.8967733383178711,
+      "learning_rate": 7.933052216465345e-05,
+      "loss": 0.0642,
+      "step": 60380
+    },
+    {
+      "epoch": 3.9509322865554464,
+      "grad_norm": 0.9686173796653748,
+      "learning_rate": 7.932308225666779e-05,
+      "loss": 0.071,
+      "step": 60390
+    },
+    {
+      "epoch": 3.9515865227347073,
+      "grad_norm": 1.0625016689300537,
+      "learning_rate": 7.93156413589512e-05,
+      "loss": 0.065,
+      "step": 60400
+    },
+    {
+      "epoch": 3.952240758913968,
+      "grad_norm": 0.830640971660614,
+      "learning_rate": 7.930819947175484e-05,
+      "loss": 0.0665,
+      "step": 60410
+    },
+    {
+      "epoch": 3.952894995093229,
+      "grad_norm": 0.9542146921157837,
+      "learning_rate": 7.930075659532987e-05,
+      "loss": 0.0593,
+      "step": 60420
+    },
+    {
+      "epoch": 3.9535492312724894,
+      "grad_norm": 0.8056615591049194,
+      "learning_rate": 7.929331272992753e-05,
+      "loss": 0.0647,
+      "step": 60430
+    },
+    {
+      "epoch": 3.95420346745175,
+      "grad_norm": 0.9513580203056335,
+      "learning_rate": 7.928586787579904e-05,
+      "loss": 0.0679,
+      "step": 60440
+    },
+    {
+      "epoch": 3.954857703631011,
+      "grad_norm": 0.8633571267127991,
+      "learning_rate": 7.927842203319573e-05,
+      "loss": 0.0692,
+      "step": 60450
+    },
+    {
+      "epoch": 3.9555119398102714,
+      "grad_norm": 1.0166418552398682,
+      "learning_rate": 7.927097520236888e-05,
+      "loss": 0.0779,
+      "step": 60460
+    },
+    {
+      "epoch": 3.9561661759895324,
+      "grad_norm": 0.9727340340614319,
+      "learning_rate": 7.926352738356986e-05,
+      "loss": 0.0689,
+      "step": 60470
+    },
+    {
+      "epoch": 3.956820412168793,
+      "grad_norm": 1.02824866771698,
+      "learning_rate": 7.925607857705003e-05,
+      "loss": 0.0735,
+      "step": 60480
+    },
+    {
+      "epoch": 3.957474648348054,
+      "grad_norm": 0.790596604347229,
+      "learning_rate": 7.924862878306083e-05,
+      "loss": 0.0734,
+      "step": 60490
+    },
+    {
+      "epoch": 3.9581288845273144,
+      "grad_norm": 1.0034263134002686,
+      "learning_rate": 7.924117800185372e-05,
+      "loss": 0.0716,
+      "step": 60500
+    },
+    {
+      "epoch": 3.958783120706575,
+      "grad_norm": 0.8736487627029419,
+      "learning_rate": 7.923372623368013e-05,
+      "loss": 0.0647,
+      "step": 60510
+    },
+    {
+      "epoch": 3.959437356885836,
+      "grad_norm": 0.8763427734375,
+      "learning_rate": 7.922627347879162e-05,
+      "loss": 0.0682,
+      "step": 60520
+    },
+    {
+      "epoch": 3.9600915930650964,
+      "grad_norm": 0.7622628808021545,
+      "learning_rate": 7.921881973743974e-05,
+      "loss": 0.0625,
+      "step": 60530
+    },
+    {
+      "epoch": 3.9607458292443574,
+      "grad_norm": 0.9626938104629517,
+      "learning_rate": 7.921136500987607e-05,
+      "loss": 0.0804,
+      "step": 60540
+    },
+    {
+      "epoch": 3.961400065423618,
+      "grad_norm": 0.8307819962501526,
+      "learning_rate": 7.920390929635221e-05,
+      "loss": 0.0618,
+      "step": 60550
+    },
+    {
+      "epoch": 3.962054301602879,
+      "grad_norm": 1.078370213508606,
+      "learning_rate": 7.919645259711982e-05,
+      "loss": 0.0683,
+      "step": 60560
+    },
+    {
+      "epoch": 3.9627085377821394,
+      "grad_norm": 0.7358745336532593,
+      "learning_rate": 7.918899491243059e-05,
+      "loss": 0.0674,
+      "step": 60570
+    },
+    {
+      "epoch": 3.9633627739614,
+      "grad_norm": 0.9152573347091675,
+      "learning_rate": 7.918153624253624e-05,
+      "loss": 0.0735,
+      "step": 60580
+    },
+    {
+      "epoch": 3.964017010140661,
+      "grad_norm": 1.0013970136642456,
+      "learning_rate": 7.91740765876885e-05,
+      "loss": 0.0639,
+      "step": 60590
+    },
+    {
+      "epoch": 3.9646712463199214,
+      "grad_norm": 0.7813799381256104,
+      "learning_rate": 7.916661594813915e-05,
+      "loss": 0.0686,
+      "step": 60600
+    },
+    {
+      "epoch": 3.965325482499182,
+      "grad_norm": 1.0708343982696533,
+      "learning_rate": 7.915915432414005e-05,
+      "loss": 0.0627,
+      "step": 60610
+    },
+    {
+      "epoch": 3.965979718678443,
+      "grad_norm": 0.8936020731925964,
+      "learning_rate": 7.915169171594299e-05,
+      "loss": 0.0745,
+      "step": 60620
+    },
+    {
+      "epoch": 3.966633954857704,
+      "grad_norm": 1.0120549201965332,
+      "learning_rate": 7.914422812379989e-05,
+      "loss": 0.0654,
+      "step": 60630
+    },
+    {
+      "epoch": 3.9672881910369644,
+      "grad_norm": 0.7676621079444885,
+      "learning_rate": 7.913676354796267e-05,
+      "loss": 0.0603,
+      "step": 60640
+    },
+    {
+      "epoch": 3.967942427216225,
+      "grad_norm": 0.7795009613037109,
+      "learning_rate": 7.912929798868324e-05,
+      "loss": 0.0684,
+      "step": 60650
+    },
+    {
+      "epoch": 3.968596663395486,
+      "grad_norm": 0.9069650769233704,
+      "learning_rate": 7.912183144621364e-05,
+      "loss": 0.0792,
+      "step": 60660
+    },
+    {
+      "epoch": 3.9692508995747464,
+      "grad_norm": 0.9909687638282776,
+      "learning_rate": 7.911436392080585e-05,
+      "loss": 0.0599,
+      "step": 60670
+    },
+    {
+      "epoch": 3.969905135754007,
+      "grad_norm": 0.9287083745002747,
+      "learning_rate": 7.91068954127119e-05,
+      "loss": 0.066,
+      "step": 60680
+    },
+    {
+      "epoch": 3.970559371933268,
+      "grad_norm": 0.9105084538459778,
+      "learning_rate": 7.909942592218391e-05,
+      "loss": 0.069,
+      "step": 60690
+    },
+    {
+      "epoch": 3.971213608112529,
+      "grad_norm": 0.8097362518310547,
+      "learning_rate": 7.909195544947398e-05,
+      "loss": 0.0583,
+      "step": 60700
+    },
+    {
+      "epoch": 3.9718678442917894,
+      "grad_norm": 0.8813385367393494,
+      "learning_rate": 7.908448399483423e-05,
+      "loss": 0.0726,
+      "step": 60710
+    },
+    {
+      "epoch": 3.97252208047105,
+      "grad_norm": 0.7956545352935791,
+      "learning_rate": 7.907701155851691e-05,
+      "loss": 0.0611,
+      "step": 60720
+    },
+    {
+      "epoch": 3.973176316650311,
+      "grad_norm": 0.8885143995285034,
+      "learning_rate": 7.906953814077417e-05,
+      "loss": 0.0635,
+      "step": 60730
+    },
+    {
+      "epoch": 3.9738305528295714,
+      "grad_norm": 1.125998854637146,
+      "learning_rate": 7.906206374185828e-05,
+      "loss": 0.0778,
+      "step": 60740
+    },
+    {
+      "epoch": 3.974484789008832,
+      "grad_norm": 0.6007951498031616,
+      "learning_rate": 7.905458836202153e-05,
+      "loss": 0.0584,
+      "step": 60750
+    },
+    {
+      "epoch": 3.975139025188093,
+      "grad_norm": 1.0695325136184692,
+      "learning_rate": 7.904711200151622e-05,
+      "loss": 0.0726,
+      "step": 60760
+    },
+    {
+      "epoch": 3.9757932613673534,
+      "grad_norm": 0.9608235359191895,
+      "learning_rate": 7.90396346605947e-05,
+      "loss": 0.068,
+      "step": 60770
+    },
+    {
+      "epoch": 3.9764474975466144,
+      "grad_norm": 0.7435832023620605,
+      "learning_rate": 7.903215633950934e-05,
+      "loss": 0.076,
+      "step": 60780
+    },
+    {
+      "epoch": 3.977101733725875,
+      "grad_norm": 0.7819331288337708,
+      "learning_rate": 7.902467703851258e-05,
+      "loss": 0.0699,
+      "step": 60790
+    },
+    {
+      "epoch": 3.977755969905136,
+      "grad_norm": 0.7933452129364014,
+      "learning_rate": 7.901719675785685e-05,
+      "loss": 0.0679,
+      "step": 60800
+    },
+    {
+      "epoch": 3.9784102060843964,
+      "grad_norm": 0.8844022154808044,
+      "learning_rate": 7.900971549779461e-05,
+      "loss": 0.0693,
+      "step": 60810
+    },
+    {
+      "epoch": 3.979064442263657,
+      "grad_norm": 1.0027797222137451,
+      "learning_rate": 7.90022332585784e-05,
+      "loss": 0.0745,
+      "step": 60820
+    },
+    {
+      "epoch": 3.979718678442918,
+      "grad_norm": 0.8335436582565308,
+      "learning_rate": 7.899475004046078e-05,
+      "loss": 0.062,
+      "step": 60830
+    },
+    {
+      "epoch": 3.9803729146221785,
+      "grad_norm": 1.0376518964767456,
+      "learning_rate": 7.898726584369427e-05,
+      "loss": 0.0655,
+      "step": 60840
+    },
+    {
+      "epoch": 3.9810271508014394,
+      "grad_norm": 0.9191878437995911,
+      "learning_rate": 7.897978066853155e-05,
+      "loss": 0.075,
+      "step": 60850
+    },
+    {
+      "epoch": 3.9816813869807,
+      "grad_norm": 0.7158858180046082,
+      "learning_rate": 7.897229451522521e-05,
+      "loss": 0.0654,
+      "step": 60860
+    },
+    {
+      "epoch": 3.982335623159961,
+      "grad_norm": 0.9076741337776184,
+      "learning_rate": 7.896480738402795e-05,
+      "loss": 0.067,
+      "step": 60870
+    },
+    {
+      "epoch": 3.9829898593392215,
+      "grad_norm": 0.9363234043121338,
+      "learning_rate": 7.895731927519248e-05,
+      "loss": 0.0638,
+      "step": 60880
+    },
+    {
+      "epoch": 3.983644095518482,
+      "grad_norm": 0.9065988659858704,
+      "learning_rate": 7.894983018897153e-05,
+      "loss": 0.0649,
+      "step": 60890
+    },
+    {
+      "epoch": 3.984298331697743,
+      "grad_norm": 0.8118206858634949,
+      "learning_rate": 7.89423401256179e-05,
+      "loss": 0.074,
+      "step": 60900
+    },
+    {
+      "epoch": 3.9849525678770035,
+      "grad_norm": 1.0652772188186646,
+      "learning_rate": 7.893484908538437e-05,
+      "loss": 0.0632,
+      "step": 60910
+    },
+    {
+      "epoch": 3.9856068040562644,
+      "grad_norm": 0.8931100964546204,
+      "learning_rate": 7.892735706852381e-05,
+      "loss": 0.0689,
+      "step": 60920
+    },
+    {
+      "epoch": 3.986261040235525,
+      "grad_norm": 0.8381471633911133,
+      "learning_rate": 7.891986407528908e-05,
+      "loss": 0.0657,
+      "step": 60930
+    },
+    {
+      "epoch": 3.986915276414786,
+      "grad_norm": 1.1571853160858154,
+      "learning_rate": 7.89123701059331e-05,
+      "loss": 0.0668,
+      "step": 60940
+    },
+    {
+      "epoch": 3.9875695125940465,
+      "grad_norm": 0.904621422290802,
+      "learning_rate": 7.890487516070881e-05,
+      "loss": 0.0755,
+      "step": 60950
+    },
+    {
+      "epoch": 3.988223748773307,
+      "grad_norm": 0.9490391612052917,
+      "learning_rate": 7.889737923986918e-05,
+      "loss": 0.0705,
+      "step": 60960
+    },
+    {
+      "epoch": 3.988877984952568,
+      "grad_norm": 1.0775212049484253,
+      "learning_rate": 7.888988234366719e-05,
+      "loss": 0.0643,
+      "step": 60970
+    },
+    {
+      "epoch": 3.9895322211318285,
+      "grad_norm": 0.9176508784294128,
+      "learning_rate": 7.888238447235592e-05,
+      "loss": 0.0724,
+      "step": 60980
+    },
+    {
+      "epoch": 3.9901864573110895,
+      "grad_norm": 1.2155365943908691,
+      "learning_rate": 7.887488562618844e-05,
+      "loss": 0.0776,
+      "step": 60990
+    },
+    {
+      "epoch": 3.99084069349035,
+      "grad_norm": 1.0456100702285767,
+      "learning_rate": 7.886738580541782e-05,
+      "loss": 0.065,
+      "step": 61000
+    },
+    {
+      "epoch": 3.991494929669611,
+      "grad_norm": 0.962174117565155,
+      "learning_rate": 7.885988501029724e-05,
+      "loss": 0.0642,
+      "step": 61010
+    },
+    {
+      "epoch": 3.9921491658488715,
+      "grad_norm": 0.8093088269233704,
+      "learning_rate": 7.885238324107982e-05,
+      "loss": 0.0724,
+      "step": 61020
+    },
+    {
+      "epoch": 3.992803402028132,
+      "grad_norm": 0.9736089706420898,
+      "learning_rate": 7.884488049801882e-05,
+      "loss": 0.0679,
+      "step": 61030
+    },
+    {
+      "epoch": 3.993457638207393,
+      "grad_norm": 0.9136343002319336,
+      "learning_rate": 7.883737678136746e-05,
+      "loss": 0.0578,
+      "step": 61040
+    },
+    {
+      "epoch": 3.9941118743866535,
+      "grad_norm": 0.998466968536377,
+      "learning_rate": 7.8829872091379e-05,
+      "loss": 0.0724,
+      "step": 61050
+    },
+    {
+      "epoch": 3.994766110565914,
+      "grad_norm": 0.8486067652702332,
+      "learning_rate": 7.882236642830675e-05,
+      "loss": 0.0747,
+      "step": 61060
+    },
+    {
+      "epoch": 3.995420346745175,
+      "grad_norm": 0.8077092170715332,
+      "learning_rate": 7.881485979240404e-05,
+      "loss": 0.0668,
+      "step": 61070
+    },
+    {
+      "epoch": 3.996074582924436,
+      "grad_norm": 1.0443347692489624,
+      "learning_rate": 7.880735218392423e-05,
+      "loss": 0.068,
+      "step": 61080
+    },
+    {
+      "epoch": 3.9967288191036965,
+      "grad_norm": 0.664110541343689,
+      "learning_rate": 7.879984360312077e-05,
+      "loss": 0.0608,
+      "step": 61090
+    },
+    {
+      "epoch": 3.997383055282957,
+      "grad_norm": 0.8030071258544922,
+      "learning_rate": 7.879233405024702e-05,
+      "loss": 0.0696,
+      "step": 61100
+    },
+    {
+      "epoch": 3.998037291462218,
+      "grad_norm": 0.9400510191917419,
+      "learning_rate": 7.87848235255565e-05,
+      "loss": 0.0756,
+      "step": 61110
+    },
+    {
+      "epoch": 3.9986915276414785,
+      "grad_norm": 0.8629300594329834,
+      "learning_rate": 7.87773120293027e-05,
+      "loss": 0.0779,
+      "step": 61120
+    },
+    {
+      "epoch": 3.999345763820739,
+      "grad_norm": 0.8903558850288391,
+      "learning_rate": 7.876979956173914e-05,
+      "loss": 0.0825,
+      "step": 61130
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 1.0015478134155273,
+      "learning_rate": 7.87622861231194e-05,
+      "loss": 0.0693,
+      "step": 61140
+    },
+    {
+      "epoch": 4.000654236179261,
+      "grad_norm": 0.8407610654830933,
+      "learning_rate": 7.875477171369707e-05,
+      "loss": 0.0606,
+      "step": 61150
+    },
+    {
+      "epoch": 4.001308472358521,
+      "grad_norm": 1.2632551193237305,
+      "learning_rate": 7.874725633372577e-05,
+      "loss": 0.0688,
+      "step": 61160
+    },
+    {
+      "epoch": 4.001962708537782,
+      "grad_norm": 0.812816858291626,
+      "learning_rate": 7.87397399834592e-05,
+      "loss": 0.067,
+      "step": 61170
+    },
+    {
+      "epoch": 4.002616944717043,
+      "grad_norm": 0.790305495262146,
+      "learning_rate": 7.873222266315101e-05,
+      "loss": 0.0604,
+      "step": 61180
+    },
+    {
+      "epoch": 4.003271180896304,
+      "grad_norm": 0.8392609357833862,
+      "learning_rate": 7.872470437305496e-05,
+      "loss": 0.0706,
+      "step": 61190
+    },
+    {
+      "epoch": 4.003925417075564,
+      "grad_norm": 0.8045458793640137,
+      "learning_rate": 7.87171851134248e-05,
+      "loss": 0.0696,
+      "step": 61200
+    },
+    {
+      "epoch": 4.004579653254825,
+      "grad_norm": 0.9475613832473755,
+      "learning_rate": 7.870966488451434e-05,
+      "loss": 0.0672,
+      "step": 61210
+    },
+    {
+      "epoch": 4.005233889434086,
+      "grad_norm": 0.8142966032028198,
+      "learning_rate": 7.87021436865774e-05,
+      "loss": 0.0612,
+      "step": 61220
+    },
+    {
+      "epoch": 4.005888125613346,
+      "grad_norm": 0.8628994822502136,
+      "learning_rate": 7.869462151986781e-05,
+      "loss": 0.0615,
+      "step": 61230
+    },
+    {
+      "epoch": 4.006542361792607,
+      "grad_norm": 0.8791362643241882,
+      "learning_rate": 7.868709838463952e-05,
+      "loss": 0.0711,
+      "step": 61240
+    },
+    {
+      "epoch": 4.007196597971868,
+      "grad_norm": 1.1582494974136353,
+      "learning_rate": 7.867957428114641e-05,
+      "loss": 0.0688,
+      "step": 61250
+    },
+    {
+      "epoch": 4.007850834151129,
+      "grad_norm": 0.9910046458244324,
+      "learning_rate": 7.867204920964245e-05,
+      "loss": 0.0659,
+      "step": 61260
+    },
+    {
+      "epoch": 4.008505070330389,
+      "grad_norm": 0.832694411277771,
+      "learning_rate": 7.866452317038164e-05,
+      "loss": 0.0706,
+      "step": 61270
+    },
+    {
+      "epoch": 4.00915930650965,
+      "grad_norm": 0.8842899203300476,
+      "learning_rate": 7.865699616361798e-05,
+      "loss": 0.0744,
+      "step": 61280
+    },
+    {
+      "epoch": 4.009813542688911,
+      "grad_norm": 0.9337363839149475,
+      "learning_rate": 7.864946818960557e-05,
+      "loss": 0.0715,
+      "step": 61290
+    },
+    {
+      "epoch": 4.010467778868171,
+      "grad_norm": 0.8401015400886536,
+      "learning_rate": 7.864193924859846e-05,
+      "loss": 0.0635,
+      "step": 61300
+    },
+    {
+      "epoch": 4.011122015047432,
+      "grad_norm": 0.9122945666313171,
+      "learning_rate": 7.86344093408508e-05,
+      "loss": 0.0663,
+      "step": 61310
+    },
+    {
+      "epoch": 4.011776251226693,
+      "grad_norm": 0.809889018535614,
+      "learning_rate": 7.862687846661671e-05,
+      "loss": 0.0704,
+      "step": 61320
+    },
+    {
+      "epoch": 4.012430487405954,
+      "grad_norm": 0.9183257818222046,
+      "learning_rate": 7.86193466261504e-05,
+      "loss": 0.0702,
+      "step": 61330
+    },
+    {
+      "epoch": 4.013084723585214,
+      "grad_norm": 0.874580979347229,
+      "learning_rate": 7.861181381970608e-05,
+      "loss": 0.072,
+      "step": 61340
+    },
+    {
+      "epoch": 4.013738959764475,
+      "grad_norm": 0.7762249708175659,
+      "learning_rate": 7.860428004753801e-05,
+      "loss": 0.0557,
+      "step": 61350
+    },
+    {
+      "epoch": 4.014393195943736,
+      "grad_norm": 0.7530931830406189,
+      "learning_rate": 7.859674530990047e-05,
+      "loss": 0.0666,
+      "step": 61360
+    },
+    {
+      "epoch": 4.015047432122996,
+      "grad_norm": 0.9463986158370972,
+      "learning_rate": 7.858920960704779e-05,
+      "loss": 0.0644,
+      "step": 61370
+    },
+    {
+      "epoch": 4.015701668302257,
+      "grad_norm": 0.828858494758606,
+      "learning_rate": 7.85816729392343e-05,
+      "loss": 0.0718,
+      "step": 61380
+    },
+    {
+      "epoch": 4.016355904481518,
+      "grad_norm": 0.7672846913337708,
+      "learning_rate": 7.857413530671438e-05,
+      "loss": 0.0615,
+      "step": 61390
+    },
+    {
+      "epoch": 4.017010140660778,
+      "grad_norm": 1.026138186454773,
+      "learning_rate": 7.856659670974246e-05,
+      "loss": 0.0762,
+      "step": 61400
+    },
+    {
+      "epoch": 4.017664376840039,
+      "grad_norm": 0.8137958645820618,
+      "learning_rate": 7.855905714857299e-05,
+      "loss": 0.0633,
+      "step": 61410
+    },
+    {
+      "epoch": 4.0183186130193,
+      "grad_norm": 0.9455887079238892,
+      "learning_rate": 7.855151662346043e-05,
+      "loss": 0.0701,
+      "step": 61420
+    },
+    {
+      "epoch": 4.018972849198561,
+      "grad_norm": 0.8994813561439514,
+      "learning_rate": 7.854397513465932e-05,
+      "loss": 0.0601,
+      "step": 61430
+    },
+    {
+      "epoch": 4.019627085377821,
+      "grad_norm": 0.7181370854377747,
+      "learning_rate": 7.853643268242417e-05,
+      "loss": 0.062,
+      "step": 61440
+    },
+    {
+      "epoch": 4.020281321557082,
+      "grad_norm": 0.7311094403266907,
+      "learning_rate": 7.852888926700959e-05,
+      "loss": 0.0645,
+      "step": 61450
+    },
+    {
+      "epoch": 4.020935557736343,
+      "grad_norm": 0.8224121928215027,
+      "learning_rate": 7.852134488867018e-05,
+      "loss": 0.0667,
+      "step": 61460
+    },
+    {
+      "epoch": 4.021589793915603,
+      "grad_norm": 1.1656960248947144,
+      "learning_rate": 7.851379954766058e-05,
+      "loss": 0.0815,
+      "step": 61470
+    },
+    {
+      "epoch": 4.022244030094864,
+      "grad_norm": 0.7743502259254456,
+      "learning_rate": 7.850625324423546e-05,
+      "loss": 0.0586,
+      "step": 61480
+    },
+    {
+      "epoch": 4.022898266274125,
+      "grad_norm": 0.8907047510147095,
+      "learning_rate": 7.849870597864953e-05,
+      "loss": 0.0665,
+      "step": 61490
+    },
+    {
+      "epoch": 4.023552502453386,
+      "grad_norm": 1.395200490951538,
+      "learning_rate": 7.849115775115755e-05,
+      "loss": 0.0752,
+      "step": 61500
+    },
+    {
+      "epoch": 4.024206738632646,
+      "grad_norm": 0.8665695190429688,
+      "learning_rate": 7.848360856201425e-05,
+      "loss": 0.066,
+      "step": 61510
+    },
+    {
+      "epoch": 4.024860974811907,
+      "grad_norm": 1.0678719282150269,
+      "learning_rate": 7.847605841147447e-05,
+      "loss": 0.0763,
+      "step": 61520
+    },
+    {
+      "epoch": 4.025515210991168,
+      "grad_norm": 0.961864173412323,
+      "learning_rate": 7.846850729979304e-05,
+      "loss": 0.0756,
+      "step": 61530
+    },
+    {
+      "epoch": 4.026169447170428,
+      "grad_norm": 0.9026308655738831,
+      "learning_rate": 7.846095522722482e-05,
+      "loss": 0.0656,
+      "step": 61540
+    },
+    {
+      "epoch": 4.026823683349689,
+      "grad_norm": 0.8607982397079468,
+      "learning_rate": 7.845340219402472e-05,
+      "loss": 0.0614,
+      "step": 61550
+    },
+    {
+      "epoch": 4.02747791952895,
+      "grad_norm": 0.8916060924530029,
+      "learning_rate": 7.844584820044769e-05,
+      "loss": 0.0714,
+      "step": 61560
+    },
+    {
+      "epoch": 4.028132155708211,
+      "grad_norm": 0.7897719144821167,
+      "learning_rate": 7.843829324674867e-05,
+      "loss": 0.0649,
+      "step": 61570
+    },
+    {
+      "epoch": 4.028786391887471,
+      "grad_norm": 0.8527302742004395,
+      "learning_rate": 7.843073733318268e-05,
+      "loss": 0.0782,
+      "step": 61580
+    },
+    {
+      "epoch": 4.029440628066732,
+      "grad_norm": 0.8170759677886963,
+      "learning_rate": 7.842318046000475e-05,
+      "loss": 0.0603,
+      "step": 61590
+    },
+    {
+      "epoch": 4.030094864245993,
+      "grad_norm": 0.9056137204170227,
+      "learning_rate": 7.841562262746991e-05,
+      "loss": 0.064,
+      "step": 61600
+    },
+    {
+      "epoch": 4.030749100425253,
+      "grad_norm": 0.9502385258674622,
+      "learning_rate": 7.84080638358333e-05,
+      "loss": 0.0672,
+      "step": 61610
+    },
+    {
+      "epoch": 4.031403336604514,
+      "grad_norm": 0.964516818523407,
+      "learning_rate": 7.840050408535002e-05,
+      "loss": 0.0765,
+      "step": 61620
+    },
+    {
+      "epoch": 4.032057572783775,
+      "grad_norm": 1.0085734128952026,
+      "learning_rate": 7.839294337627525e-05,
+      "loss": 0.0691,
+      "step": 61630
+    },
+    {
+      "epoch": 4.032711808963036,
+      "grad_norm": 0.9676178097724915,
+      "learning_rate": 7.838538170886419e-05,
+      "loss": 0.0619,
+      "step": 61640
+    },
+    {
+      "epoch": 4.033366045142296,
+      "grad_norm": 1.0016093254089355,
+      "learning_rate": 7.837781908337204e-05,
+      "loss": 0.0653,
+      "step": 61650
+    },
+    {
+      "epoch": 4.034020281321557,
+      "grad_norm": 0.9340269565582275,
+      "learning_rate": 7.837025550005408e-05,
+      "loss": 0.0682,
+      "step": 61660
+    },
+    {
+      "epoch": 4.034674517500818,
+      "grad_norm": 0.9540444612503052,
+      "learning_rate": 7.836269095916557e-05,
+      "loss": 0.0739,
+      "step": 61670
+    },
+    {
+      "epoch": 4.035328753680078,
+      "grad_norm": 0.7741513848304749,
+      "learning_rate": 7.835512546096188e-05,
+      "loss": 0.0704,
+      "step": 61680
+    },
+    {
+      "epoch": 4.035982989859339,
+      "grad_norm": 0.8669870495796204,
+      "learning_rate": 7.834755900569834e-05,
+      "loss": 0.0625,
+      "step": 61690
+    },
+    {
+      "epoch": 4.0366372260386,
+      "grad_norm": 0.8938657641410828,
+      "learning_rate": 7.833999159363035e-05,
+      "loss": 0.0677,
+      "step": 61700
+    },
+    {
+      "epoch": 4.037291462217861,
+      "grad_norm": 0.7298368215560913,
+      "learning_rate": 7.83324232250133e-05,
+      "loss": 0.0606,
+      "step": 61710
+    },
+    {
+      "epoch": 4.037945698397121,
+      "grad_norm": 0.8773128986358643,
+      "learning_rate": 7.832485390010266e-05,
+      "loss": 0.0631,
+      "step": 61720
+    },
+    {
+      "epoch": 4.038599934576382,
+      "grad_norm": 0.8813597559928894,
+      "learning_rate": 7.831728361915394e-05,
+      "loss": 0.0735,
+      "step": 61730
+    },
+    {
+      "epoch": 4.039254170755643,
+      "grad_norm": 0.7509139776229858,
+      "learning_rate": 7.830971238242261e-05,
+      "loss": 0.0666,
+      "step": 61740
+    },
+    {
+      "epoch": 4.039908406934903,
+      "grad_norm": 0.8333325386047363,
+      "learning_rate": 7.830214019016426e-05,
+      "loss": 0.0649,
+      "step": 61750
+    },
+    {
+      "epoch": 4.040562643114164,
+      "grad_norm": 0.9280720353126526,
+      "learning_rate": 7.829456704263442e-05,
+      "loss": 0.06,
+      "step": 61760
+    },
+    {
+      "epoch": 4.041216879293425,
+      "grad_norm": 0.6866008639335632,
+      "learning_rate": 7.828699294008877e-05,
+      "loss": 0.066,
+      "step": 61770
+    },
+    {
+      "epoch": 4.041871115472686,
+      "grad_norm": 0.8785973787307739,
+      "learning_rate": 7.827941788278292e-05,
+      "loss": 0.0582,
+      "step": 61780
+    },
+    {
+      "epoch": 4.042525351651946,
+      "grad_norm": 0.9217095971107483,
+      "learning_rate": 7.827184187097253e-05,
+      "loss": 0.0737,
+      "step": 61790
+    },
+    {
+      "epoch": 4.043179587831207,
+      "grad_norm": 0.7473239302635193,
+      "learning_rate": 7.826426490491335e-05,
+      "loss": 0.0639,
+      "step": 61800
+    },
+    {
+      "epoch": 4.043833824010468,
+      "grad_norm": 0.8309671878814697,
+      "learning_rate": 7.82566869848611e-05,
+      "loss": 0.0633,
+      "step": 61810
+    },
+    {
+      "epoch": 4.044488060189728,
+      "grad_norm": 1.0445795059204102,
+      "learning_rate": 7.824910811107156e-05,
+      "loss": 0.0683,
+      "step": 61820
+    },
+    {
+      "epoch": 4.045142296368989,
+      "grad_norm": 0.8129024505615234,
+      "learning_rate": 7.824152828380053e-05,
+      "loss": 0.0628,
+      "step": 61830
+    },
+    {
+      "epoch": 4.04579653254825,
+      "grad_norm": 0.8035739660263062,
+      "learning_rate": 7.823394750330387e-05,
+      "loss": 0.0664,
+      "step": 61840
+    },
+    {
+      "epoch": 4.04645076872751,
+      "grad_norm": 1.1228084564208984,
+      "learning_rate": 7.822636576983741e-05,
+      "loss": 0.0663,
+      "step": 61850
+    },
+    {
+      "epoch": 4.047105004906771,
+      "grad_norm": 0.9300438761711121,
+      "learning_rate": 7.821878308365708e-05,
+      "loss": 0.0679,
+      "step": 61860
+    },
+    {
+      "epoch": 4.047759241086032,
+      "grad_norm": 0.8193831443786621,
+      "learning_rate": 7.821119944501885e-05,
+      "loss": 0.063,
+      "step": 61870
+    },
+    {
+      "epoch": 4.048413477265293,
+      "grad_norm": 0.8168728351593018,
+      "learning_rate": 7.820361485417862e-05,
+      "loss": 0.0709,
+      "step": 61880
+    },
+    {
+      "epoch": 4.049067713444553,
+      "grad_norm": 1.0577526092529297,
+      "learning_rate": 7.819602931139243e-05,
+      "loss": 0.0641,
+      "step": 61890
+    },
+    {
+      "epoch": 4.049721949623814,
+      "grad_norm": 0.8535493612289429,
+      "learning_rate": 7.81884428169163e-05,
+      "loss": 0.057,
+      "step": 61900
+    },
+    {
+      "epoch": 4.050376185803075,
+      "grad_norm": 0.8765014410018921,
+      "learning_rate": 7.81808553710063e-05,
+      "loss": 0.0662,
+      "step": 61910
+    },
+    {
+      "epoch": 4.051030421982335,
+      "grad_norm": 0.9272754788398743,
+      "learning_rate": 7.817326697391853e-05,
+      "loss": 0.0657,
+      "step": 61920
+    },
+    {
+      "epoch": 4.051684658161596,
+      "grad_norm": 0.7854313254356384,
+      "learning_rate": 7.81656776259091e-05,
+      "loss": 0.068,
+      "step": 61930
+    },
+    {
+      "epoch": 4.052338894340857,
+      "grad_norm": 0.753021776676178,
+      "learning_rate": 7.81580873272342e-05,
+      "loss": 0.0648,
+      "step": 61940
+    },
+    {
+      "epoch": 4.052993130520118,
+      "grad_norm": 0.939609944820404,
+      "learning_rate": 7.815049607815e-05,
+      "loss": 0.0682,
+      "step": 61950
+    },
+    {
+      "epoch": 4.053647366699378,
+      "grad_norm": 0.9001626372337341,
+      "learning_rate": 7.814290387891271e-05,
+      "loss": 0.0608,
+      "step": 61960
+    },
+    {
+      "epoch": 4.054301602878639,
+      "grad_norm": 0.9824931621551514,
+      "learning_rate": 7.813531072977863e-05,
+      "loss": 0.069,
+      "step": 61970
+    },
+    {
+      "epoch": 4.0549558390579,
+      "grad_norm": 0.9642009139060974,
+      "learning_rate": 7.812771663100402e-05,
+      "loss": 0.0602,
+      "step": 61980
+    },
+    {
+      "epoch": 4.05561007523716,
+      "grad_norm": 0.7149823307991028,
+      "learning_rate": 7.812012158284521e-05,
+      "loss": 0.0667,
+      "step": 61990
+    },
+    {
+      "epoch": 4.056264311416421,
+      "grad_norm": 0.8310834765434265,
+      "learning_rate": 7.811252558555854e-05,
+      "loss": 0.0746,
+      "step": 62000
+    },
+    {
+      "epoch": 4.056918547595682,
+      "grad_norm": 0.7982217073440552,
+      "learning_rate": 7.810492863940041e-05,
+      "loss": 0.0852,
+      "step": 62010
+    },
+    {
+      "epoch": 4.057572783774943,
+      "grad_norm": 0.9088074564933777,
+      "learning_rate": 7.809733074462722e-05,
+      "loss": 0.064,
+      "step": 62020
+    },
+    {
+      "epoch": 4.058227019954203,
+      "grad_norm": 0.8191305994987488,
+      "learning_rate": 7.808973190149544e-05,
+      "loss": 0.0684,
+      "step": 62030
+    },
+    {
+      "epoch": 4.058881256133464,
+      "grad_norm": 0.8682653307914734,
+      "learning_rate": 7.808213211026153e-05,
+      "loss": 0.0623,
+      "step": 62040
+    },
+    {
+      "epoch": 4.059535492312725,
+      "grad_norm": 1.0228744745254517,
+      "learning_rate": 7.807453137118204e-05,
+      "loss": 0.0726,
+      "step": 62050
+    },
+    {
+      "epoch": 4.060189728491985,
+      "grad_norm": 0.8825092315673828,
+      "learning_rate": 7.806692968451346e-05,
+      "loss": 0.0669,
+      "step": 62060
+    },
+    {
+      "epoch": 4.060843964671246,
+      "grad_norm": 0.8798748254776001,
+      "learning_rate": 7.80593270505124e-05,
+      "loss": 0.0669,
+      "step": 62070
+    },
+    {
+      "epoch": 4.061498200850507,
+      "grad_norm": 0.8253244161605835,
+      "learning_rate": 7.805172346943547e-05,
+      "loss": 0.058,
+      "step": 62080
+    },
+    {
+      "epoch": 4.062152437029768,
+      "grad_norm": 0.9651762247085571,
+      "learning_rate": 7.804411894153932e-05,
+      "loss": 0.0709,
+      "step": 62090
+    },
+    {
+      "epoch": 4.062806673209028,
+      "grad_norm": 1.1336904764175415,
+      "learning_rate": 7.803651346708056e-05,
+      "loss": 0.072,
+      "step": 62100
+    },
+    {
+      "epoch": 4.063460909388289,
+      "grad_norm": 1.0221117734909058,
+      "learning_rate": 7.802890704631598e-05,
+      "loss": 0.0734,
+      "step": 62110
+    },
+    {
+      "epoch": 4.06411514556755,
+      "grad_norm": 0.8365046381950378,
+      "learning_rate": 7.802129967950227e-05,
+      "loss": 0.0662,
+      "step": 62120
+    },
+    {
+      "epoch": 4.06476938174681,
+      "grad_norm": 0.8922891020774841,
+      "learning_rate": 7.801369136689621e-05,
+      "loss": 0.065,
+      "step": 62130
+    },
+    {
+      "epoch": 4.065423617926071,
+      "grad_norm": 0.689914345741272,
+      "learning_rate": 7.80060821087546e-05,
+      "loss": 0.0674,
+      "step": 62140
+    },
+    {
+      "epoch": 4.066077854105332,
+      "grad_norm": 0.8463851809501648,
+      "learning_rate": 7.799847190533428e-05,
+      "loss": 0.0628,
+      "step": 62150
+    },
+    {
+      "epoch": 4.066732090284593,
+      "grad_norm": 0.7686126232147217,
+      "learning_rate": 7.799086075689208e-05,
+      "loss": 0.0665,
+      "step": 62160
+    },
+    {
+      "epoch": 4.067386326463853,
+      "grad_norm": 0.9086697697639465,
+      "learning_rate": 7.798324866368493e-05,
+      "loss": 0.0655,
+      "step": 62170
+    },
+    {
+      "epoch": 4.068040562643114,
+      "grad_norm": 0.8965178728103638,
+      "learning_rate": 7.797563562596974e-05,
+      "loss": 0.0722,
+      "step": 62180
+    },
+    {
+      "epoch": 4.068694798822375,
+      "grad_norm": 0.8882538676261902,
+      "learning_rate": 7.796802164400348e-05,
+      "loss": 0.0615,
+      "step": 62190
+    },
+    {
+      "epoch": 4.069349035001635,
+      "grad_norm": 0.8429491519927979,
+      "learning_rate": 7.796040671804316e-05,
+      "loss": 0.0628,
+      "step": 62200
+    },
+    {
+      "epoch": 4.070003271180896,
+      "grad_norm": 0.8130377531051636,
+      "learning_rate": 7.795279084834577e-05,
+      "loss": 0.0657,
+      "step": 62210
+    },
+    {
+      "epoch": 4.070657507360157,
+      "grad_norm": 0.9137871861457825,
+      "learning_rate": 7.794517403516838e-05,
+      "loss": 0.0606,
+      "step": 62220
+    },
+    {
+      "epoch": 4.071311743539418,
+      "grad_norm": 0.8658760190010071,
+      "learning_rate": 7.793755627876808e-05,
+      "loss": 0.0676,
+      "step": 62230
+    },
+    {
+      "epoch": 4.071965979718678,
+      "grad_norm": 0.9319032430648804,
+      "learning_rate": 7.7929937579402e-05,
+      "loss": 0.0693,
+      "step": 62240
+    },
+    {
+      "epoch": 4.072620215897939,
+      "grad_norm": 0.8796036839485168,
+      "learning_rate": 7.792231793732727e-05,
+      "loss": 0.0669,
+      "step": 62250
+    },
+    {
+      "epoch": 4.0732744520772,
+      "grad_norm": 0.7122048139572144,
+      "learning_rate": 7.791469735280106e-05,
+      "loss": 0.0639,
+      "step": 62260
+    },
+    {
+      "epoch": 4.07392868825646,
+      "grad_norm": 1.0055348873138428,
+      "learning_rate": 7.790707582608063e-05,
+      "loss": 0.0808,
+      "step": 62270
+    },
+    {
+      "epoch": 4.074582924435721,
+      "grad_norm": 0.8951319456100464,
+      "learning_rate": 7.78994533574232e-05,
+      "loss": 0.0677,
+      "step": 62280
+    },
+    {
+      "epoch": 4.075237160614982,
+      "grad_norm": 0.9522911906242371,
+      "learning_rate": 7.789182994708604e-05,
+      "loss": 0.0545,
+      "step": 62290
+    },
+    {
+      "epoch": 4.075891396794242,
+      "grad_norm": 0.8506065607070923,
+      "learning_rate": 7.788420559532646e-05,
+      "loss": 0.0595,
+      "step": 62300
+    },
+    {
+      "epoch": 4.076545632973503,
+      "grad_norm": 0.959356427192688,
+      "learning_rate": 7.787658030240183e-05,
+      "loss": 0.0713,
+      "step": 62310
+    },
+    {
+      "epoch": 4.077199869152764,
+      "grad_norm": 0.7769831418991089,
+      "learning_rate": 7.786895406856952e-05,
+      "loss": 0.0627,
+      "step": 62320
+    },
+    {
+      "epoch": 4.077854105332025,
+      "grad_norm": 0.6790773868560791,
+      "learning_rate": 7.786132689408688e-05,
+      "loss": 0.068,
+      "step": 62330
+    },
+    {
+      "epoch": 4.078508341511285,
+      "grad_norm": 0.7900551557540894,
+      "learning_rate": 7.78536987792114e-05,
+      "loss": 0.0749,
+      "step": 62340
+    },
+    {
+      "epoch": 4.079162577690546,
+      "grad_norm": 0.8323625922203064,
+      "learning_rate": 7.784606972420056e-05,
+      "loss": 0.0585,
+      "step": 62350
+    },
+    {
+      "epoch": 4.079816813869807,
+      "grad_norm": 1.0454132556915283,
+      "learning_rate": 7.783843972931184e-05,
+      "loss": 0.0613,
+      "step": 62360
+    },
+    {
+      "epoch": 4.080471050049067,
+      "grad_norm": 0.9331693053245544,
+      "learning_rate": 7.783080879480274e-05,
+      "loss": 0.0635,
+      "step": 62370
+    },
+    {
+      "epoch": 4.081125286228328,
+      "grad_norm": 0.9045494198799133,
+      "learning_rate": 7.782317692093088e-05,
+      "loss": 0.0703,
+      "step": 62380
+    },
+    {
+      "epoch": 4.081779522407589,
+      "grad_norm": 0.9631790518760681,
+      "learning_rate": 7.781554410795381e-05,
+      "loss": 0.0675,
+      "step": 62390
+    },
+    {
+      "epoch": 4.08243375858685,
+      "grad_norm": 0.9105682373046875,
+      "learning_rate": 7.78079103561292e-05,
+      "loss": 0.0779,
+      "step": 62400
+    },
+    {
+      "epoch": 4.08308799476611,
+      "grad_norm": 0.8074235320091248,
+      "learning_rate": 7.780027566571465e-05,
+      "loss": 0.07,
+      "step": 62410
+    },
+    {
+      "epoch": 4.083742230945371,
+      "grad_norm": 0.8759629726409912,
+      "learning_rate": 7.779264003696794e-05,
+      "loss": 0.0637,
+      "step": 62420
+    },
+    {
+      "epoch": 4.084396467124632,
+      "grad_norm": 0.8032673001289368,
+      "learning_rate": 7.77850034701467e-05,
+      "loss": 0.064,
+      "step": 62430
+    },
+    {
+      "epoch": 4.085050703303892,
+      "grad_norm": 0.8850411176681519,
+      "learning_rate": 7.777736596550874e-05,
+      "loss": 0.0641,
+      "step": 62440
+    },
+    {
+      "epoch": 4.085704939483153,
+      "grad_norm": 1.1167175769805908,
+      "learning_rate": 7.776972752331182e-05,
+      "loss": 0.0689,
+      "step": 62450
+    },
+    {
+      "epoch": 4.086359175662414,
+      "grad_norm": 0.9858927726745605,
+      "learning_rate": 7.776208814381379e-05,
+      "loss": 0.0725,
+      "step": 62460
+    },
+    {
+      "epoch": 4.087013411841675,
+      "grad_norm": 0.9899501204490662,
+      "learning_rate": 7.775444782727245e-05,
+      "loss": 0.0684,
+      "step": 62470
+    },
+    {
+      "epoch": 4.087667648020935,
+      "grad_norm": 0.7681850790977478,
+      "learning_rate": 7.77468065739457e-05,
+      "loss": 0.0633,
+      "step": 62480
+    },
+    {
+      "epoch": 4.088321884200196,
+      "grad_norm": 0.6893104314804077,
+      "learning_rate": 7.773916438409149e-05,
+      "loss": 0.0572,
+      "step": 62490
+    },
+    {
+      "epoch": 4.088976120379457,
+      "grad_norm": 0.7790807485580444,
+      "learning_rate": 7.773152125796772e-05,
+      "loss": 0.0634,
+      "step": 62500
+    },
+    {
+      "epoch": 4.089630356558717,
+      "grad_norm": 1.0509587526321411,
+      "learning_rate": 7.772387719583238e-05,
+      "loss": 0.063,
+      "step": 62510
+    },
+    {
+      "epoch": 4.090284592737978,
+      "grad_norm": 1.0617860555648804,
+      "learning_rate": 7.771623219794346e-05,
+      "loss": 0.061,
+      "step": 62520
+    },
+    {
+      "epoch": 4.090938828917239,
+      "grad_norm": 0.7485182881355286,
+      "learning_rate": 7.770858626455903e-05,
+      "loss": 0.0698,
+      "step": 62530
+    },
+    {
+      "epoch": 4.0915930650965,
+      "grad_norm": 0.7513982653617859,
+      "learning_rate": 7.770093939593716e-05,
+      "loss": 0.0646,
+      "step": 62540
+    },
+    {
+      "epoch": 4.09224730127576,
+      "grad_norm": 0.9014798998832703,
+      "learning_rate": 7.769329159233592e-05,
+      "loss": 0.0747,
+      "step": 62550
+    },
+    {
+      "epoch": 4.092901537455021,
+      "grad_norm": 1.199467658996582,
+      "learning_rate": 7.768564285401346e-05,
+      "loss": 0.0664,
+      "step": 62560
+    },
+    {
+      "epoch": 4.093555773634282,
+      "grad_norm": 0.9420590996742249,
+      "learning_rate": 7.767799318122794e-05,
+      "loss": 0.0718,
+      "step": 62570
+    },
+    {
+      "epoch": 4.094210009813542,
+      "grad_norm": 0.9552091360092163,
+      "learning_rate": 7.767034257423758e-05,
+      "loss": 0.0804,
+      "step": 62580
+    },
+    {
+      "epoch": 4.094864245992803,
+      "grad_norm": 1.0671602487564087,
+      "learning_rate": 7.766269103330057e-05,
+      "loss": 0.0742,
+      "step": 62590
+    },
+    {
+      "epoch": 4.095518482172064,
+      "grad_norm": 0.961463451385498,
+      "learning_rate": 7.76550385586752e-05,
+      "loss": 0.0706,
+      "step": 62600
+    },
+    {
+      "epoch": 4.096172718351325,
+      "grad_norm": 0.858124315738678,
+      "learning_rate": 7.764738515061975e-05,
+      "loss": 0.057,
+      "step": 62610
+    },
+    {
+      "epoch": 4.096826954530585,
+      "grad_norm": 0.830001950263977,
+      "learning_rate": 7.763973080939254e-05,
+      "loss": 0.0554,
+      "step": 62620
+    },
+    {
+      "epoch": 4.097481190709846,
+      "grad_norm": 0.9393772482872009,
+      "learning_rate": 7.763207553525193e-05,
+      "loss": 0.0708,
+      "step": 62630
+    },
+    {
+      "epoch": 4.098135426889107,
+      "grad_norm": 0.8235999941825867,
+      "learning_rate": 7.76244193284563e-05,
+      "loss": 0.0615,
+      "step": 62640
+    },
+    {
+      "epoch": 4.098789663068367,
+      "grad_norm": 1.2535772323608398,
+      "learning_rate": 7.761676218926408e-05,
+      "loss": 0.0678,
+      "step": 62650
+    },
+    {
+      "epoch": 4.099443899247628,
+      "grad_norm": 0.9056881070137024,
+      "learning_rate": 7.76091041179337e-05,
+      "loss": 0.0637,
+      "step": 62660
+    },
+    {
+      "epoch": 4.100098135426889,
+      "grad_norm": 1.1135672330856323,
+      "learning_rate": 7.760144511472365e-05,
+      "loss": 0.0627,
+      "step": 62670
+    },
+    {
+      "epoch": 4.10075237160615,
+      "grad_norm": 1.0625379085540771,
+      "learning_rate": 7.759378517989245e-05,
+      "loss": 0.0658,
+      "step": 62680
+    },
+    {
+      "epoch": 4.10140660778541,
+      "grad_norm": 0.947235107421875,
+      "learning_rate": 7.75861243136986e-05,
+      "loss": 0.0638,
+      "step": 62690
+    },
+    {
+      "epoch": 4.102060843964671,
+      "grad_norm": 0.715565025806427,
+      "learning_rate": 7.757846251640074e-05,
+      "loss": 0.0647,
+      "step": 62700
+    },
+    {
+      "epoch": 4.102715080143932,
+      "grad_norm": 0.9149342775344849,
+      "learning_rate": 7.757079978825744e-05,
+      "loss": 0.0605,
+      "step": 62710
+    },
+    {
+      "epoch": 4.103369316323192,
+      "grad_norm": 0.8739985227584839,
+      "learning_rate": 7.756313612952733e-05,
+      "loss": 0.0592,
+      "step": 62720
+    },
+    {
+      "epoch": 4.104023552502453,
+      "grad_norm": 1.1399884223937988,
+      "learning_rate": 7.755547154046908e-05,
+      "loss": 0.0712,
+      "step": 62730
+    },
+    {
+      "epoch": 4.104677788681714,
+      "grad_norm": 0.9350805878639221,
+      "learning_rate": 7.754780602134142e-05,
+      "loss": 0.063,
+      "step": 62740
+    },
+    {
+      "epoch": 4.105332024860974,
+      "grad_norm": 0.8487550616264343,
+      "learning_rate": 7.754013957240305e-05,
+      "loss": 0.0741,
+      "step": 62750
+    },
+    {
+      "epoch": 4.105986261040235,
+      "grad_norm": 0.7867029905319214,
+      "learning_rate": 7.753247219391273e-05,
+      "loss": 0.0604,
+      "step": 62760
+    },
+    {
+      "epoch": 4.106640497219496,
+      "grad_norm": 0.9640946388244629,
+      "learning_rate": 7.752480388612928e-05,
+      "loss": 0.063,
+      "step": 62770
+    },
+    {
+      "epoch": 4.107294733398757,
+      "grad_norm": 1.0482327938079834,
+      "learning_rate": 7.751713464931151e-05,
+      "loss": 0.0709,
+      "step": 62780
+    },
+    {
+      "epoch": 4.107948969578017,
+      "grad_norm": 0.8405677080154419,
+      "learning_rate": 7.75094644837183e-05,
+      "loss": 0.0572,
+      "step": 62790
+    },
+    {
+      "epoch": 4.108603205757278,
+      "grad_norm": 0.8008997440338135,
+      "learning_rate": 7.750179338960849e-05,
+      "loss": 0.0732,
+      "step": 62800
+    },
+    {
+      "epoch": 4.109257441936539,
+      "grad_norm": 0.8923006057739258,
+      "learning_rate": 7.749412136724103e-05,
+      "loss": 0.0666,
+      "step": 62810
+    },
+    {
+      "epoch": 4.109911678115799,
+      "grad_norm": 1.0257177352905273,
+      "learning_rate": 7.748644841687486e-05,
+      "loss": 0.0722,
+      "step": 62820
+    },
+    {
+      "epoch": 4.11056591429506,
+      "grad_norm": 0.7566149234771729,
+      "learning_rate": 7.747877453876901e-05,
+      "loss": 0.0667,
+      "step": 62830
+    },
+    {
+      "epoch": 4.111220150474321,
+      "grad_norm": 0.9833534955978394,
+      "learning_rate": 7.747109973318242e-05,
+      "loss": 0.0588,
+      "step": 62840
+    },
+    {
+      "epoch": 4.111874386653582,
+      "grad_norm": 0.9134366512298584,
+      "learning_rate": 7.746342400037417e-05,
+      "loss": 0.0701,
+      "step": 62850
+    },
+    {
+      "epoch": 4.112528622832842,
+      "grad_norm": 0.8957915902137756,
+      "learning_rate": 7.745574734060335e-05,
+      "loss": 0.0588,
+      "step": 62860
+    },
+    {
+      "epoch": 4.113182859012103,
+      "grad_norm": 0.7879095077514648,
+      "learning_rate": 7.744806975412904e-05,
+      "loss": 0.0669,
+      "step": 62870
+    },
+    {
+      "epoch": 4.113837095191364,
+      "grad_norm": 0.9840167760848999,
+      "learning_rate": 7.744039124121039e-05,
+      "loss": 0.0612,
+      "step": 62880
+    },
+    {
+      "epoch": 4.114491331370624,
+      "grad_norm": 0.656774640083313,
+      "learning_rate": 7.743271180210657e-05,
+      "loss": 0.0671,
+      "step": 62890
+    },
+    {
+      "epoch": 4.115145567549885,
+      "grad_norm": 0.6617786884307861,
+      "learning_rate": 7.742503143707679e-05,
+      "loss": 0.0633,
+      "step": 62900
+    },
+    {
+      "epoch": 4.115799803729146,
+      "grad_norm": 0.9550051093101501,
+      "learning_rate": 7.741735014638027e-05,
+      "loss": 0.0704,
+      "step": 62910
+    },
+    {
+      "epoch": 4.116454039908407,
+      "grad_norm": 0.7846367359161377,
+      "learning_rate": 7.740966793027626e-05,
+      "loss": 0.0664,
+      "step": 62920
+    },
+    {
+      "epoch": 4.117108276087667,
+      "grad_norm": 1.068386435508728,
+      "learning_rate": 7.740198478902409e-05,
+      "loss": 0.068,
+      "step": 62930
+    },
+    {
+      "epoch": 4.117762512266928,
+      "grad_norm": 1.1259160041809082,
+      "learning_rate": 7.739430072288309e-05,
+      "loss": 0.0667,
+      "step": 62940
+    },
+    {
+      "epoch": 4.118416748446189,
+      "grad_norm": 1.1424773931503296,
+      "learning_rate": 7.738661573211256e-05,
+      "loss": 0.0688,
+      "step": 62950
+    },
+    {
+      "epoch": 4.119070984625449,
+      "grad_norm": 0.8278184533119202,
+      "learning_rate": 7.737892981697194e-05,
+      "loss": 0.0603,
+      "step": 62960
+    },
+    {
+      "epoch": 4.11972522080471,
+      "grad_norm": 0.7246772646903992,
+      "learning_rate": 7.737124297772065e-05,
+      "loss": 0.0647,
+      "step": 62970
+    },
+    {
+      "epoch": 4.120379456983971,
+      "grad_norm": 0.9246161580085754,
+      "learning_rate": 7.736355521461811e-05,
+      "loss": 0.0655,
+      "step": 62980
+    },
+    {
+      "epoch": 4.121033693163232,
+      "grad_norm": 0.9616913199424744,
+      "learning_rate": 7.735586652792382e-05,
+      "loss": 0.0727,
+      "step": 62990
+    },
+    {
+      "epoch": 4.121687929342492,
+      "grad_norm": 0.8557519912719727,
+      "learning_rate": 7.734817691789729e-05,
+      "loss": 0.066,
+      "step": 63000
+    },
+    {
+      "epoch": 4.122342165521753,
+      "grad_norm": 0.8217513561248779,
+      "learning_rate": 7.734048638479807e-05,
+      "loss": 0.0749,
+      "step": 63010
+    },
+    {
+      "epoch": 4.122996401701014,
+      "grad_norm": 1.1255154609680176,
+      "learning_rate": 7.733279492888572e-05,
+      "loss": 0.0753,
+      "step": 63020
+    },
+    {
+      "epoch": 4.123650637880274,
+      "grad_norm": 0.9445163607597351,
+      "learning_rate": 7.732510255041985e-05,
+      "loss": 0.0624,
+      "step": 63030
+    },
+    {
+      "epoch": 4.124304874059535,
+      "grad_norm": 0.9506664276123047,
+      "learning_rate": 7.731740924966014e-05,
+      "loss": 0.067,
+      "step": 63040
+    },
+    {
+      "epoch": 4.124959110238796,
+      "grad_norm": 0.9376375079154968,
+      "learning_rate": 7.730971502686621e-05,
+      "loss": 0.0673,
+      "step": 63050
+    },
+    {
+      "epoch": 4.125613346418057,
+      "grad_norm": 0.8614791035652161,
+      "learning_rate": 7.730201988229777e-05,
+      "loss": 0.0703,
+      "step": 63060
+    },
+    {
+      "epoch": 4.126267582597317,
+      "grad_norm": 0.8429438471794128,
+      "learning_rate": 7.729432381621455e-05,
+      "loss": 0.0619,
+      "step": 63070
+    },
+    {
+      "epoch": 4.126921818776578,
+      "grad_norm": 1.0551830530166626,
+      "learning_rate": 7.728662682887633e-05,
+      "loss": 0.0709,
+      "step": 63080
+    },
+    {
+      "epoch": 4.127576054955839,
+      "grad_norm": 0.9198732376098633,
+      "learning_rate": 7.727892892054289e-05,
+      "loss": 0.0647,
+      "step": 63090
+    },
+    {
+      "epoch": 4.128230291135099,
+      "grad_norm": 0.8567238450050354,
+      "learning_rate": 7.727123009147406e-05,
+      "loss": 0.0637,
+      "step": 63100
+    },
+    {
+      "epoch": 4.12888452731436,
+      "grad_norm": 0.947485089302063,
+      "learning_rate": 7.72635303419297e-05,
+      "loss": 0.0737,
+      "step": 63110
+    },
+    {
+      "epoch": 4.129538763493621,
+      "grad_norm": 0.8637893199920654,
+      "learning_rate": 7.725582967216966e-05,
+      "loss": 0.0712,
+      "step": 63120
+    },
+    {
+      "epoch": 4.130192999672882,
+      "grad_norm": 1.1613445281982422,
+      "learning_rate": 7.724812808245392e-05,
+      "loss": 0.0756,
+      "step": 63130
+    },
+    {
+      "epoch": 4.130847235852142,
+      "grad_norm": 0.8602867722511292,
+      "learning_rate": 7.724042557304238e-05,
+      "loss": 0.067,
+      "step": 63140
+    },
+    {
+      "epoch": 4.131501472031403,
+      "grad_norm": 0.8593533635139465,
+      "learning_rate": 7.723272214419506e-05,
+      "loss": 0.0615,
+      "step": 63150
+    },
+    {
+      "epoch": 4.132155708210664,
+      "grad_norm": 0.9966337084770203,
+      "learning_rate": 7.722501779617193e-05,
+      "loss": 0.0593,
+      "step": 63160
+    },
+    {
+      "epoch": 4.132809944389924,
+      "grad_norm": 0.997571587562561,
+      "learning_rate": 7.721731252923305e-05,
+      "loss": 0.0681,
+      "step": 63170
+    },
+    {
+      "epoch": 4.133464180569185,
+      "grad_norm": 0.94794100522995,
+      "learning_rate": 7.720960634363848e-05,
+      "loss": 0.0818,
+      "step": 63180
+    },
+    {
+      "epoch": 4.134118416748446,
+      "grad_norm": 0.8688943982124329,
+      "learning_rate": 7.720189923964833e-05,
+      "loss": 0.0653,
+      "step": 63190
+    },
+    {
+      "epoch": 4.1347726529277065,
+      "grad_norm": 0.8499542474746704,
+      "learning_rate": 7.719419121752277e-05,
+      "loss": 0.0734,
+      "step": 63200
+    },
+    {
+      "epoch": 4.135426889106967,
+      "grad_norm": 0.8392613530158997,
+      "learning_rate": 7.718648227752192e-05,
+      "loss": 0.0738,
+      "step": 63210
+    },
+    {
+      "epoch": 4.136081125286228,
+      "grad_norm": 1.165153980255127,
+      "learning_rate": 7.7178772419906e-05,
+      "loss": 0.0717,
+      "step": 63220
+    },
+    {
+      "epoch": 4.136735361465489,
+      "grad_norm": 0.7674516439437866,
+      "learning_rate": 7.717106164493523e-05,
+      "loss": 0.0682,
+      "step": 63230
+    },
+    {
+      "epoch": 4.1373895976447495,
+      "grad_norm": 0.8776054978370667,
+      "learning_rate": 7.716334995286988e-05,
+      "loss": 0.0712,
+      "step": 63240
+    },
+    {
+      "epoch": 4.13804383382401,
+      "grad_norm": 1.0000853538513184,
+      "learning_rate": 7.715563734397022e-05,
+      "loss": 0.0649,
+      "step": 63250
+    },
+    {
+      "epoch": 4.138698070003271,
+      "grad_norm": 0.9027912616729736,
+      "learning_rate": 7.714792381849658e-05,
+      "loss": 0.0746,
+      "step": 63260
+    },
+    {
+      "epoch": 4.1393523061825315,
+      "grad_norm": 1.001630425453186,
+      "learning_rate": 7.714020937670931e-05,
+      "loss": 0.0703,
+      "step": 63270
+    },
+    {
+      "epoch": 4.140006542361792,
+      "grad_norm": 0.9271620512008667,
+      "learning_rate": 7.713249401886882e-05,
+      "loss": 0.0685,
+      "step": 63280
+    },
+    {
+      "epoch": 4.140660778541053,
+      "grad_norm": 0.795330822467804,
+      "learning_rate": 7.712477774523547e-05,
+      "loss": 0.0696,
+      "step": 63290
+    },
+    {
+      "epoch": 4.141315014720314,
+      "grad_norm": 0.7591599225997925,
+      "learning_rate": 7.711706055606975e-05,
+      "loss": 0.063,
+      "step": 63300
+    },
+    {
+      "epoch": 4.1419692508995745,
+      "grad_norm": 0.9598954319953918,
+      "learning_rate": 7.710934245163211e-05,
+      "loss": 0.0658,
+      "step": 63310
+    },
+    {
+      "epoch": 4.142623487078835,
+      "grad_norm": 0.6453430652618408,
+      "learning_rate": 7.710162343218307e-05,
+      "loss": 0.0666,
+      "step": 63320
+    },
+    {
+      "epoch": 4.143277723258096,
+      "grad_norm": 0.8312489986419678,
+      "learning_rate": 7.709390349798315e-05,
+      "loss": 0.0651,
+      "step": 63330
+    },
+    {
+      "epoch": 4.1439319594373565,
+      "grad_norm": 0.7793470025062561,
+      "learning_rate": 7.708618264929295e-05,
+      "loss": 0.0713,
+      "step": 63340
+    },
+    {
+      "epoch": 4.1445861956166175,
+      "grad_norm": 0.827540397644043,
+      "learning_rate": 7.707846088637305e-05,
+      "loss": 0.0641,
+      "step": 63350
+    },
+    {
+      "epoch": 4.145240431795878,
+      "grad_norm": 0.8500876426696777,
+      "learning_rate": 7.707073820948407e-05,
+      "loss": 0.0698,
+      "step": 63360
+    },
+    {
+      "epoch": 4.145894667975139,
+      "grad_norm": 0.7591197490692139,
+      "learning_rate": 7.706301461888667e-05,
+      "loss": 0.0623,
+      "step": 63370
+    },
+    {
+      "epoch": 4.1465489041543995,
+      "grad_norm": 0.7708221077919006,
+      "learning_rate": 7.705529011484159e-05,
+      "loss": 0.0595,
+      "step": 63380
+    },
+    {
+      "epoch": 4.14720314033366,
+      "grad_norm": 0.8088019490242004,
+      "learning_rate": 7.704756469760947e-05,
+      "loss": 0.0609,
+      "step": 63390
+    },
+    {
+      "epoch": 4.147857376512921,
+      "grad_norm": 0.8570855855941772,
+      "learning_rate": 7.703983836745112e-05,
+      "loss": 0.0717,
+      "step": 63400
+    },
+    {
+      "epoch": 4.1485116126921815,
+      "grad_norm": 0.7673255801200867,
+      "learning_rate": 7.703211112462731e-05,
+      "loss": 0.06,
+      "step": 63410
+    },
+    {
+      "epoch": 4.1491658488714425,
+      "grad_norm": 0.9685210585594177,
+      "learning_rate": 7.702438296939887e-05,
+      "loss": 0.0609,
+      "step": 63420
+    },
+    {
+      "epoch": 4.149820085050703,
+      "grad_norm": 0.7904552817344666,
+      "learning_rate": 7.701665390202661e-05,
+      "loss": 0.0636,
+      "step": 63430
+    },
+    {
+      "epoch": 4.150474321229964,
+      "grad_norm": 0.8309029936790466,
+      "learning_rate": 7.700892392277144e-05,
+      "loss": 0.0626,
+      "step": 63440
+    },
+    {
+      "epoch": 4.1511285574092245,
+      "grad_norm": 0.7211902141571045,
+      "learning_rate": 7.700119303189424e-05,
+      "loss": 0.0626,
+      "step": 63450
+    },
+    {
+      "epoch": 4.1517827935884855,
+      "grad_norm": 0.7313457131385803,
+      "learning_rate": 7.699346122965599e-05,
+      "loss": 0.0544,
+      "step": 63460
+    },
+    {
+      "epoch": 4.152437029767746,
+      "grad_norm": 0.9150106906890869,
+      "learning_rate": 7.698572851631761e-05,
+      "loss": 0.0679,
+      "step": 63470
+    },
+    {
+      "epoch": 4.1530912659470065,
+      "grad_norm": 1.2330702543258667,
+      "learning_rate": 7.69779948921401e-05,
+      "loss": 0.0696,
+      "step": 63480
+    },
+    {
+      "epoch": 4.1537455021262675,
+      "grad_norm": 0.9700013995170593,
+      "learning_rate": 7.697026035738454e-05,
+      "loss": 0.0679,
+      "step": 63490
+    },
+    {
+      "epoch": 4.1543997383055284,
+      "grad_norm": 0.8161741495132446,
+      "learning_rate": 7.696252491231197e-05,
+      "loss": 0.0617,
+      "step": 63500
+    },
+    {
+      "epoch": 4.155053974484789,
+      "grad_norm": 0.9676251411437988,
+      "learning_rate": 7.695478855718344e-05,
+      "loss": 0.0713,
+      "step": 63510
+    },
+    {
+      "epoch": 4.1557082106640495,
+      "grad_norm": 0.9095385074615479,
+      "learning_rate": 7.694705129226012e-05,
+      "loss": 0.063,
+      "step": 63520
+    },
+    {
+      "epoch": 4.1563624468433105,
+      "grad_norm": 0.7652897238731384,
+      "learning_rate": 7.693931311780315e-05,
+      "loss": 0.0679,
+      "step": 63530
+    },
+    {
+      "epoch": 4.157016683022571,
+      "grad_norm": 0.8652570247650146,
+      "learning_rate": 7.693157403407372e-05,
+      "loss": 0.0624,
+      "step": 63540
+    },
+    {
+      "epoch": 4.1576709192018315,
+      "grad_norm": 0.7441179752349854,
+      "learning_rate": 7.692383404133301e-05,
+      "loss": 0.0568,
+      "step": 63550
+    },
+    {
+      "epoch": 4.1583251553810925,
+      "grad_norm": 0.8349494338035583,
+      "learning_rate": 7.691609313984232e-05,
+      "loss": 0.0621,
+      "step": 63560
+    },
+    {
+      "epoch": 4.1589793915603535,
+      "grad_norm": 0.8901420831680298,
+      "learning_rate": 7.690835132986287e-05,
+      "loss": 0.0687,
+      "step": 63570
+    },
+    {
+      "epoch": 4.159633627739614,
+      "grad_norm": 0.8408117890357971,
+      "learning_rate": 7.690060861165601e-05,
+      "loss": 0.0641,
+      "step": 63580
+    },
+    {
+      "epoch": 4.1602878639188745,
+      "grad_norm": 0.8637434244155884,
+      "learning_rate": 7.689286498548304e-05,
+      "loss": 0.0681,
+      "step": 63590
+    },
+    {
+      "epoch": 4.1609421000981355,
+      "grad_norm": 0.8616120219230652,
+      "learning_rate": 7.688512045160538e-05,
+      "loss": 0.0649,
+      "step": 63600
+    },
+    {
+      "epoch": 4.1615963362773964,
+      "grad_norm": 0.8039658069610596,
+      "learning_rate": 7.687737501028438e-05,
+      "loss": 0.0656,
+      "step": 63610
+    },
+    {
+      "epoch": 4.1622505724566565,
+      "grad_norm": 0.9429184198379517,
+      "learning_rate": 7.686962866178147e-05,
+      "loss": 0.0668,
+      "step": 63620
+    },
+    {
+      "epoch": 4.1629048086359175,
+      "grad_norm": 0.9747768044471741,
+      "learning_rate": 7.686188140635815e-05,
+      "loss": 0.0672,
+      "step": 63630
+    },
+    {
+      "epoch": 4.1635590448151785,
+      "grad_norm": 1.0797864198684692,
+      "learning_rate": 7.685413324427588e-05,
+      "loss": 0.0729,
+      "step": 63640
+    },
+    {
+      "epoch": 4.1642132809944385,
+      "grad_norm": 1.0528181791305542,
+      "learning_rate": 7.684638417579617e-05,
+      "loss": 0.0679,
+      "step": 63650
+    },
+    {
+      "epoch": 4.1648675171736995,
+      "grad_norm": 1.1309136152267456,
+      "learning_rate": 7.68386342011806e-05,
+      "loss": 0.0744,
+      "step": 63660
+    },
+    {
+      "epoch": 4.1655217533529605,
+      "grad_norm": 1.1569007635116577,
+      "learning_rate": 7.683088332069073e-05,
+      "loss": 0.0766,
+      "step": 63670
+    },
+    {
+      "epoch": 4.1661759895322215,
+      "grad_norm": 0.8142968416213989,
+      "learning_rate": 7.682313153458817e-05,
+      "loss": 0.0634,
+      "step": 63680
+    },
+    {
+      "epoch": 4.1668302257114815,
+      "grad_norm": 1.0308458805084229,
+      "learning_rate": 7.68153788431346e-05,
+      "loss": 0.0671,
+      "step": 63690
+    },
+    {
+      "epoch": 4.1674844618907425,
+      "grad_norm": 0.8902773857116699,
+      "learning_rate": 7.680762524659167e-05,
+      "loss": 0.0575,
+      "step": 63700
+    },
+    {
+      "epoch": 4.1681386980700035,
+      "grad_norm": 1.0067930221557617,
+      "learning_rate": 7.679987074522107e-05,
+      "loss": 0.0612,
+      "step": 63710
+    },
+    {
+      "epoch": 4.168792934249264,
+      "grad_norm": 0.8824333548545837,
+      "learning_rate": 7.679211533928454e-05,
+      "loss": 0.0631,
+      "step": 63720
+    },
+    {
+      "epoch": 4.1694471704285245,
+      "grad_norm": 0.8123067021369934,
+      "learning_rate": 7.678435902904386e-05,
+      "loss": 0.0651,
+      "step": 63730
+    },
+    {
+      "epoch": 4.1701014066077855,
+      "grad_norm": 1.0830172300338745,
+      "learning_rate": 7.677660181476081e-05,
+      "loss": 0.0693,
+      "step": 63740
+    },
+    {
+      "epoch": 4.1707556427870465,
+      "grad_norm": 0.9839672446250916,
+      "learning_rate": 7.676884369669723e-05,
+      "loss": 0.0662,
+      "step": 63750
+    },
+    {
+      "epoch": 4.1714098789663066,
+      "grad_norm": 0.941388726234436,
+      "learning_rate": 7.676108467511498e-05,
+      "loss": 0.066,
+      "step": 63760
+    },
+    {
+      "epoch": 4.1720641151455675,
+      "grad_norm": 0.8638181090354919,
+      "learning_rate": 7.675332475027593e-05,
+      "loss": 0.0714,
+      "step": 63770
+    },
+    {
+      "epoch": 4.1727183513248285,
+      "grad_norm": 0.8280830979347229,
+      "learning_rate": 7.674556392244201e-05,
+      "loss": 0.0676,
+      "step": 63780
+    },
+    {
+      "epoch": 4.173372587504089,
+      "grad_norm": 0.7582976222038269,
+      "learning_rate": 7.673780219187518e-05,
+      "loss": 0.0591,
+      "step": 63790
+    },
+    {
+      "epoch": 4.1740268236833495,
+      "grad_norm": 0.7644749283790588,
+      "learning_rate": 7.673003955883737e-05,
+      "loss": 0.058,
+      "step": 63800
+    },
+    {
+      "epoch": 4.1746810598626105,
+      "grad_norm": 0.88578200340271,
+      "learning_rate": 7.672227602359064e-05,
+      "loss": 0.069,
+      "step": 63810
+    },
+    {
+      "epoch": 4.1753352960418715,
+      "grad_norm": 0.8269320726394653,
+      "learning_rate": 7.671451158639702e-05,
+      "loss": 0.0553,
+      "step": 63820
+    },
+    {
+      "epoch": 4.175989532221132,
+      "grad_norm": 0.791187584400177,
+      "learning_rate": 7.670674624751857e-05,
+      "loss": 0.0666,
+      "step": 63830
+    },
+    {
+      "epoch": 4.1766437684003925,
+      "grad_norm": 0.8734759092330933,
+      "learning_rate": 7.669898000721738e-05,
+      "loss": 0.0613,
+      "step": 63840
+    },
+    {
+      "epoch": 4.1772980045796535,
+      "grad_norm": 0.9316699504852295,
+      "learning_rate": 7.66912128657556e-05,
+      "loss": 0.0632,
+      "step": 63850
+    },
+    {
+      "epoch": 4.177952240758914,
+      "grad_norm": 0.9909458160400391,
+      "learning_rate": 7.668344482339539e-05,
+      "loss": 0.0668,
+      "step": 63860
+    },
+    {
+      "epoch": 4.1786064769381746,
+      "grad_norm": 1.1051509380340576,
+      "learning_rate": 7.667567588039895e-05,
+      "loss": 0.0661,
+      "step": 63870
+    },
+    {
+      "epoch": 4.1792607131174355,
+      "grad_norm": 0.8877078890800476,
+      "learning_rate": 7.666790603702846e-05,
+      "loss": 0.0608,
+      "step": 63880
+    },
+    {
+      "epoch": 4.1799149492966965,
+      "grad_norm": 0.7553628087043762,
+      "learning_rate": 7.666013529354621e-05,
+      "loss": 0.0662,
+      "step": 63890
+    },
+    {
+      "epoch": 4.180569185475957,
+      "grad_norm": 0.8676695823669434,
+      "learning_rate": 7.665236365021448e-05,
+      "loss": 0.0669,
+      "step": 63900
+    },
+    {
+      "epoch": 4.1812234216552175,
+      "grad_norm": 0.7713704109191895,
+      "learning_rate": 7.664459110729558e-05,
+      "loss": 0.0642,
+      "step": 63910
+    },
+    {
+      "epoch": 4.1818776578344785,
+      "grad_norm": 1.018075704574585,
+      "learning_rate": 7.663681766505187e-05,
+      "loss": 0.0665,
+      "step": 63920
+    },
+    {
+      "epoch": 4.182531894013739,
+      "grad_norm": 0.94621342420578,
+      "learning_rate": 7.662904332374567e-05,
+      "loss": 0.0613,
+      "step": 63930
+    },
+    {
+      "epoch": 4.183186130193,
+      "grad_norm": 0.9044004678726196,
+      "learning_rate": 7.662126808363946e-05,
+      "loss": 0.0602,
+      "step": 63940
+    },
+    {
+      "epoch": 4.1838403663722605,
+      "grad_norm": 0.8055490255355835,
+      "learning_rate": 7.661349194499561e-05,
+      "loss": 0.0643,
+      "step": 63950
+    },
+    {
+      "epoch": 4.1844946025515215,
+      "grad_norm": 0.9511623382568359,
+      "learning_rate": 7.660571490807662e-05,
+      "loss": 0.0748,
+      "step": 63960
+    },
+    {
+      "epoch": 4.185148838730782,
+      "grad_norm": 0.8675695061683655,
+      "learning_rate": 7.659793697314496e-05,
+      "loss": 0.0609,
+      "step": 63970
+    },
+    {
+      "epoch": 4.1858030749100426,
+      "grad_norm": 0.9198430180549622,
+      "learning_rate": 7.659015814046318e-05,
+      "loss": 0.0718,
+      "step": 63980
+    },
+    {
+      "epoch": 4.1864573110893035,
+      "grad_norm": 0.8027588129043579,
+      "learning_rate": 7.658237841029383e-05,
+      "loss": 0.0562,
+      "step": 63990
+    },
+    {
+      "epoch": 4.187111547268564,
+      "grad_norm": 0.7004244327545166,
+      "learning_rate": 7.657459778289949e-05,
+      "loss": 0.0602,
+      "step": 64000
+    },
+    {
+      "epoch": 4.187765783447825,
+      "grad_norm": 0.8982917070388794,
+      "learning_rate": 7.656681625854278e-05,
+      "loss": 0.0646,
+      "step": 64010
+    },
+    {
+      "epoch": 4.1884200196270855,
+      "grad_norm": 0.9718723297119141,
+      "learning_rate": 7.655903383748637e-05,
+      "loss": 0.0811,
+      "step": 64020
+    },
+    {
+      "epoch": 4.1890742558063465,
+      "grad_norm": 0.9556106925010681,
+      "learning_rate": 7.655125051999289e-05,
+      "loss": 0.0666,
+      "step": 64030
+    },
+    {
+      "epoch": 4.189728491985607,
+      "grad_norm": 0.8579035997390747,
+      "learning_rate": 7.654346630632507e-05,
+      "loss": 0.0681,
+      "step": 64040
+    },
+    {
+      "epoch": 4.190382728164868,
+      "grad_norm": 0.8351026177406311,
+      "learning_rate": 7.653568119674567e-05,
+      "loss": 0.0722,
+      "step": 64050
+    },
+    {
+      "epoch": 4.1910369643441285,
+      "grad_norm": 0.8565570116043091,
+      "learning_rate": 7.652789519151741e-05,
+      "loss": 0.0584,
+      "step": 64060
+    },
+    {
+      "epoch": 4.191691200523389,
+      "grad_norm": 0.9618107080459595,
+      "learning_rate": 7.652010829090312e-05,
+      "loss": 0.0592,
+      "step": 64070
+    },
+    {
+      "epoch": 4.19234543670265,
+      "grad_norm": 0.6823198199272156,
+      "learning_rate": 7.651232049516566e-05,
+      "loss": 0.0549,
+      "step": 64080
+    },
+    {
+      "epoch": 4.1929996728819106,
+      "grad_norm": 0.9959099292755127,
+      "learning_rate": 7.650453180456783e-05,
+      "loss": 0.068,
+      "step": 64090
+    },
+    {
+      "epoch": 4.193653909061171,
+      "grad_norm": 0.767525315284729,
+      "learning_rate": 7.649674221937252e-05,
+      "loss": 0.0582,
+      "step": 64100
+    },
+    {
+      "epoch": 4.194308145240432,
+      "grad_norm": 1.036657452583313,
+      "learning_rate": 7.64889517398427e-05,
+      "loss": 0.073,
+      "step": 64110
+    },
+    {
+      "epoch": 4.194962381419693,
+      "grad_norm": 1.0274145603179932,
+      "learning_rate": 7.648116036624126e-05,
+      "loss": 0.0697,
+      "step": 64120
+    },
+    {
+      "epoch": 4.1956166175989535,
+      "grad_norm": 0.8063427209854126,
+      "learning_rate": 7.647336809883124e-05,
+      "loss": 0.0559,
+      "step": 64130
+    },
+    {
+      "epoch": 4.196270853778214,
+      "grad_norm": 0.8468816876411438,
+      "learning_rate": 7.646557493787558e-05,
+      "loss": 0.0635,
+      "step": 64140
+    },
+    {
+      "epoch": 4.196925089957475,
+      "grad_norm": 0.9064197540283203,
+      "learning_rate": 7.645778088363738e-05,
+      "loss": 0.0647,
+      "step": 64150
+    },
+    {
+      "epoch": 4.197579326136736,
+      "grad_norm": 0.9117610454559326,
+      "learning_rate": 7.644998593637968e-05,
+      "loss": 0.0668,
+      "step": 64160
+    },
+    {
+      "epoch": 4.198233562315996,
+      "grad_norm": 0.6964724063873291,
+      "learning_rate": 7.64421900963656e-05,
+      "loss": 0.064,
+      "step": 64170
+    },
+    {
+      "epoch": 4.198887798495257,
+      "grad_norm": 0.8209680318832397,
+      "learning_rate": 7.643439336385824e-05,
+      "loss": 0.0636,
+      "step": 64180
+    },
+    {
+      "epoch": 4.199542034674518,
+      "grad_norm": 0.8586738705635071,
+      "learning_rate": 7.642659573912078e-05,
+      "loss": 0.0644,
+      "step": 64190
+    },
+    {
+      "epoch": 4.200196270853779,
+      "grad_norm": 0.8765914440155029,
+      "learning_rate": 7.641879722241643e-05,
+      "loss": 0.0718,
+      "step": 64200
+    },
+    {
+      "epoch": 4.200850507033039,
+      "grad_norm": 0.9443649649620056,
+      "learning_rate": 7.641099781400838e-05,
+      "loss": 0.0835,
+      "step": 64210
+    },
+    {
+      "epoch": 4.2015047432123,
+      "grad_norm": 1.0230027437210083,
+      "learning_rate": 7.640319751415987e-05,
+      "loss": 0.0697,
+      "step": 64220
+    },
+    {
+      "epoch": 4.202158979391561,
+      "grad_norm": 1.0767507553100586,
+      "learning_rate": 7.63953963231342e-05,
+      "loss": 0.0671,
+      "step": 64230
+    },
+    {
+      "epoch": 4.202813215570821,
+      "grad_norm": 0.9476938247680664,
+      "learning_rate": 7.63875942411947e-05,
+      "loss": 0.0624,
+      "step": 64240
+    },
+    {
+      "epoch": 4.203467451750082,
+      "grad_norm": 0.9079118967056274,
+      "learning_rate": 7.637979126860468e-05,
+      "loss": 0.0665,
+      "step": 64250
+    },
+    {
+      "epoch": 4.204121687929343,
+      "grad_norm": 0.7623441219329834,
+      "learning_rate": 7.637198740562752e-05,
+      "loss": 0.0564,
+      "step": 64260
+    },
+    {
+      "epoch": 4.204775924108604,
+      "grad_norm": 0.7987393736839294,
+      "learning_rate": 7.636418265252662e-05,
+      "loss": 0.0642,
+      "step": 64270
+    },
+    {
+      "epoch": 4.205430160287864,
+      "grad_norm": 0.7768983840942383,
+      "learning_rate": 7.635637700956542e-05,
+      "loss": 0.0689,
+      "step": 64280
+    },
+    {
+      "epoch": 4.206084396467125,
+      "grad_norm": 0.9263744354248047,
+      "learning_rate": 7.634857047700737e-05,
+      "loss": 0.0591,
+      "step": 64290
+    },
+    {
+      "epoch": 4.206738632646386,
+      "grad_norm": 0.9548198580741882,
+      "learning_rate": 7.634076305511598e-05,
+      "loss": 0.0626,
+      "step": 64300
+    },
+    {
+      "epoch": 4.207392868825646,
+      "grad_norm": 0.9324436783790588,
+      "learning_rate": 7.633295474415473e-05,
+      "loss": 0.062,
+      "step": 64310
+    },
+    {
+      "epoch": 4.208047105004907,
+      "grad_norm": 0.9058679342269897,
+      "learning_rate": 7.63251455443872e-05,
+      "loss": 0.0726,
+      "step": 64320
+    },
+    {
+      "epoch": 4.208701341184168,
+      "grad_norm": 0.6921377778053284,
+      "learning_rate": 7.631733545607697e-05,
+      "loss": 0.0582,
+      "step": 64330
+    },
+    {
+      "epoch": 4.209355577363429,
+      "grad_norm": 0.7973593473434448,
+      "learning_rate": 7.630952447948765e-05,
+      "loss": 0.0667,
+      "step": 64340
+    },
+    {
+      "epoch": 4.210009813542689,
+      "grad_norm": 0.8190951943397522,
+      "learning_rate": 7.630171261488289e-05,
+      "loss": 0.0559,
+      "step": 64350
+    },
+    {
+      "epoch": 4.21066404972195,
+      "grad_norm": 0.9391740560531616,
+      "learning_rate": 7.629389986252634e-05,
+      "loss": 0.0589,
+      "step": 64360
+    },
+    {
+      "epoch": 4.211318285901211,
+      "grad_norm": 0.8340182304382324,
+      "learning_rate": 7.628608622268171e-05,
+      "loss": 0.0663,
+      "step": 64370
+    },
+    {
+      "epoch": 4.211972522080471,
+      "grad_norm": 0.9111093878746033,
+      "learning_rate": 7.627827169561275e-05,
+      "loss": 0.0664,
+      "step": 64380
+    },
+    {
+      "epoch": 4.212626758259732,
+      "grad_norm": 0.8605757355690002,
+      "learning_rate": 7.627045628158318e-05,
+      "loss": 0.0681,
+      "step": 64390
+    },
+    {
+      "epoch": 4.213280994438993,
+      "grad_norm": 0.9816390872001648,
+      "learning_rate": 7.626263998085683e-05,
+      "loss": 0.0607,
+      "step": 64400
+    },
+    {
+      "epoch": 4.213935230618254,
+      "grad_norm": 0.8772796988487244,
+      "learning_rate": 7.625482279369749e-05,
+      "loss": 0.0548,
+      "step": 64410
+    },
+    {
+      "epoch": 4.214589466797514,
+      "grad_norm": 0.8973631858825684,
+      "learning_rate": 7.624700472036904e-05,
+      "loss": 0.0561,
+      "step": 64420
+    },
+    {
+      "epoch": 4.215243702976775,
+      "grad_norm": 1.0034244060516357,
+      "learning_rate": 7.623918576113533e-05,
+      "loss": 0.0648,
+      "step": 64430
+    },
+    {
+      "epoch": 4.215897939156036,
+      "grad_norm": 0.8082411885261536,
+      "learning_rate": 7.62313659162603e-05,
+      "loss": 0.0651,
+      "step": 64440
+    },
+    {
+      "epoch": 4.216552175335296,
+      "grad_norm": 0.8607401847839355,
+      "learning_rate": 7.622354518600786e-05,
+      "loss": 0.063,
+      "step": 64450
+    },
+    {
+      "epoch": 4.217206411514557,
+      "grad_norm": 1.0780029296875,
+      "learning_rate": 7.621572357064202e-05,
+      "loss": 0.0556,
+      "step": 64460
+    },
+    {
+      "epoch": 4.217860647693818,
+      "grad_norm": 0.8647873997688293,
+      "learning_rate": 7.620790107042674e-05,
+      "loss": 0.0654,
+      "step": 64470
+    },
+    {
+      "epoch": 4.218514883873079,
+      "grad_norm": 0.9308034181594849,
+      "learning_rate": 7.620007768562606e-05,
+      "loss": 0.0719,
+      "step": 64480
+    },
+    {
+      "epoch": 4.219169120052339,
+      "grad_norm": 1.1081721782684326,
+      "learning_rate": 7.619225341650404e-05,
+      "loss": 0.0765,
+      "step": 64490
+    },
+    {
+      "epoch": 4.2198233562316,
+      "grad_norm": 0.7883846163749695,
+      "learning_rate": 7.618442826332482e-05,
+      "loss": 0.0683,
+      "step": 64500
+    },
+    {
+      "epoch": 4.220477592410861,
+      "grad_norm": 0.6897490620613098,
+      "learning_rate": 7.617660222635243e-05,
+      "loss": 0.0579,
+      "step": 64510
+    },
+    {
+      "epoch": 4.221131828590121,
+      "grad_norm": 0.7844201326370239,
+      "learning_rate": 7.616877530585107e-05,
+      "loss": 0.0584,
+      "step": 64520
+    },
+    {
+      "epoch": 4.221786064769382,
+      "grad_norm": 0.8688981533050537,
+      "learning_rate": 7.616094750208493e-05,
+      "loss": 0.0738,
+      "step": 64530
+    },
+    {
+      "epoch": 4.222440300948643,
+      "grad_norm": 0.886189341545105,
+      "learning_rate": 7.61531188153182e-05,
+      "loss": 0.0587,
+      "step": 64540
+    },
+    {
+      "epoch": 4.223094537127903,
+      "grad_norm": 1.0094472169876099,
+      "learning_rate": 7.61452892458151e-05,
+      "loss": 0.0697,
+      "step": 64550
+    },
+    {
+      "epoch": 4.223748773307164,
+      "grad_norm": 1.0703370571136475,
+      "learning_rate": 7.613745879383995e-05,
+      "loss": 0.0695,
+      "step": 64560
+    },
+    {
+      "epoch": 4.224403009486425,
+      "grad_norm": 0.8577084541320801,
+      "learning_rate": 7.612962745965699e-05,
+      "loss": 0.0613,
+      "step": 64570
+    },
+    {
+      "epoch": 4.225057245665686,
+      "grad_norm": 0.6969327330589294,
+      "learning_rate": 7.612179524353058e-05,
+      "loss": 0.0734,
+      "step": 64580
+    },
+    {
+      "epoch": 4.225711481844946,
+      "grad_norm": 0.8476243019104004,
+      "learning_rate": 7.611396214572508e-05,
+      "loss": 0.0715,
+      "step": 64590
+    },
+    {
+      "epoch": 4.226365718024207,
+      "grad_norm": 0.8455610871315002,
+      "learning_rate": 7.610612816650488e-05,
+      "loss": 0.0605,
+      "step": 64600
+    },
+    {
+      "epoch": 4.227019954203468,
+      "grad_norm": 1.1364946365356445,
+      "learning_rate": 7.609829330613439e-05,
+      "loss": 0.0741,
+      "step": 64610
+    },
+    {
+      "epoch": 4.227674190382728,
+      "grad_norm": 1.1362121105194092,
+      "learning_rate": 7.609045756487805e-05,
+      "loss": 0.0667,
+      "step": 64620
+    },
+    {
+      "epoch": 4.228328426561989,
+      "grad_norm": 1.0451109409332275,
+      "learning_rate": 7.608262094300034e-05,
+      "loss": 0.0725,
+      "step": 64630
+    },
+    {
+      "epoch": 4.22898266274125,
+      "grad_norm": 0.9875683188438416,
+      "learning_rate": 7.607478344076577e-05,
+      "loss": 0.057,
+      "step": 64640
+    },
+    {
+      "epoch": 4.229636898920511,
+      "grad_norm": 1.0684120655059814,
+      "learning_rate": 7.606694505843887e-05,
+      "loss": 0.0645,
+      "step": 64650
+    },
+    {
+      "epoch": 4.230291135099771,
+      "grad_norm": 1.1099648475646973,
+      "learning_rate": 7.605910579628421e-05,
+      "loss": 0.0743,
+      "step": 64660
+    },
+    {
+      "epoch": 4.230945371279032,
+      "grad_norm": 1.0001264810562134,
+      "learning_rate": 7.60512656545664e-05,
+      "loss": 0.0664,
+      "step": 64670
+    },
+    {
+      "epoch": 4.231599607458293,
+      "grad_norm": 0.8117485046386719,
+      "learning_rate": 7.604342463355003e-05,
+      "loss": 0.0688,
+      "step": 64680
+    },
+    {
+      "epoch": 4.232253843637553,
+      "grad_norm": 0.9693712592124939,
+      "learning_rate": 7.60355827334998e-05,
+      "loss": 0.0747,
+      "step": 64690
+    },
+    {
+      "epoch": 4.232908079816814,
+      "grad_norm": 0.9077291488647461,
+      "learning_rate": 7.602773995468036e-05,
+      "loss": 0.0593,
+      "step": 64700
+    },
+    {
+      "epoch": 4.233562315996075,
+      "grad_norm": 1.1760419607162476,
+      "learning_rate": 7.601989629735643e-05,
+      "loss": 0.076,
+      "step": 64710
+    },
+    {
+      "epoch": 4.234216552175336,
+      "grad_norm": 0.9444860219955444,
+      "learning_rate": 7.601205176179279e-05,
+      "loss": 0.0658,
+      "step": 64720
+    },
+    {
+      "epoch": 4.234870788354596,
+      "grad_norm": 0.9393828511238098,
+      "learning_rate": 7.600420634825416e-05,
+      "loss": 0.0571,
+      "step": 64730
+    },
+    {
+      "epoch": 4.235525024533857,
+      "grad_norm": 1.003543496131897,
+      "learning_rate": 7.599636005700537e-05,
+      "loss": 0.0668,
+      "step": 64740
+    },
+    {
+      "epoch": 4.236179260713118,
+      "grad_norm": 0.8727512955665588,
+      "learning_rate": 7.598851288831124e-05,
+      "loss": 0.0624,
+      "step": 64750
+    },
+    {
+      "epoch": 4.236833496892378,
+      "grad_norm": 1.0206762552261353,
+      "learning_rate": 7.598066484243667e-05,
+      "loss": 0.0789,
+      "step": 64760
+    },
+    {
+      "epoch": 4.237487733071639,
+      "grad_norm": 1.1758549213409424,
+      "learning_rate": 7.597281591964649e-05,
+      "loss": 0.0747,
+      "step": 64770
+    },
+    {
+      "epoch": 4.2381419692509,
+      "grad_norm": 0.7732208371162415,
+      "learning_rate": 7.596496612020567e-05,
+      "loss": 0.0596,
+      "step": 64780
+    },
+    {
+      "epoch": 4.238796205430161,
+      "grad_norm": 0.8588206768035889,
+      "learning_rate": 7.595711544437917e-05,
+      "loss": 0.063,
+      "step": 64790
+    },
+    {
+      "epoch": 4.239450441609421,
+      "grad_norm": 0.811578094959259,
+      "learning_rate": 7.594926389243193e-05,
+      "loss": 0.0732,
+      "step": 64800
+    },
+    {
+      "epoch": 4.240104677788682,
+      "grad_norm": 0.7321622371673584,
+      "learning_rate": 7.594141146462897e-05,
+      "loss": 0.0645,
+      "step": 64810
+    },
+    {
+      "epoch": 4.240758913967943,
+      "grad_norm": 0.690522313117981,
+      "learning_rate": 7.593355816123535e-05,
+      "loss": 0.0696,
+      "step": 64820
+    },
+    {
+      "epoch": 4.241413150147203,
+      "grad_norm": 0.8633084297180176,
+      "learning_rate": 7.592570398251614e-05,
+      "loss": 0.0669,
+      "step": 64830
+    },
+    {
+      "epoch": 4.242067386326464,
+      "grad_norm": 0.8941366076469421,
+      "learning_rate": 7.591784892873642e-05,
+      "loss": 0.0604,
+      "step": 64840
+    },
+    {
+      "epoch": 4.242721622505725,
+      "grad_norm": 0.9090977311134338,
+      "learning_rate": 7.590999300016131e-05,
+      "loss": 0.0656,
+      "step": 64850
+    },
+    {
+      "epoch": 4.243375858684986,
+      "grad_norm": 0.734219491481781,
+      "learning_rate": 7.5902136197056e-05,
+      "loss": 0.0589,
+      "step": 64860
+    },
+    {
+      "epoch": 4.244030094864246,
+      "grad_norm": 0.9445002675056458,
+      "learning_rate": 7.589427851968567e-05,
+      "loss": 0.0696,
+      "step": 64870
+    },
+    {
+      "epoch": 4.244684331043507,
+      "grad_norm": 0.8005701899528503,
+      "learning_rate": 7.58864199683155e-05,
+      "loss": 0.0658,
+      "step": 64880
+    },
+    {
+      "epoch": 4.245338567222768,
+      "grad_norm": 1.0320810079574585,
+      "learning_rate": 7.58785605432108e-05,
+      "loss": 0.0689,
+      "step": 64890
+    },
+    {
+      "epoch": 4.245992803402028,
+      "grad_norm": 0.820569634437561,
+      "learning_rate": 7.58707002446368e-05,
+      "loss": 0.0665,
+      "step": 64900
+    },
+    {
+      "epoch": 4.246647039581289,
+      "grad_norm": 0.8624978065490723,
+      "learning_rate": 7.58628390728588e-05,
+      "loss": 0.0674,
+      "step": 64910
+    },
+    {
+      "epoch": 4.24730127576055,
+      "grad_norm": 0.9039542078971863,
+      "learning_rate": 7.58549770281422e-05,
+      "loss": 0.0871,
+      "step": 64920
+    },
+    {
+      "epoch": 4.247955511939811,
+      "grad_norm": 0.8915390968322754,
+      "learning_rate": 7.584711411075227e-05,
+      "loss": 0.0585,
+      "step": 64930
+    },
+    {
+      "epoch": 4.248609748119071,
+      "grad_norm": 0.9667330980300903,
+      "learning_rate": 7.583925032095447e-05,
+      "loss": 0.0598,
+      "step": 64940
+    },
+    {
+      "epoch": 4.249263984298332,
+      "grad_norm": 0.9445711970329285,
+      "learning_rate": 7.583138565901422e-05,
+      "loss": 0.0579,
+      "step": 64950
+    },
+    {
+      "epoch": 4.249918220477593,
+      "grad_norm": 1.0453894138336182,
+      "learning_rate": 7.582352012519694e-05,
+      "loss": 0.0719,
+      "step": 64960
+    },
+    {
+      "epoch": 4.250572456656853,
+      "grad_norm": 0.9736664891242981,
+      "learning_rate": 7.581565371976813e-05,
+      "loss": 0.063,
+      "step": 64970
+    },
+    {
+      "epoch": 4.251226692836114,
+      "grad_norm": 0.7672659158706665,
+      "learning_rate": 7.580778644299332e-05,
+      "loss": 0.0648,
+      "step": 64980
+    },
+    {
+      "epoch": 4.251880929015375,
+      "grad_norm": 0.7975163459777832,
+      "learning_rate": 7.579991829513802e-05,
+      "loss": 0.0604,
+      "step": 64990
+    },
+    {
+      "epoch": 4.252535165194635,
+      "grad_norm": 0.8647359013557434,
+      "learning_rate": 7.579204927646782e-05,
+      "loss": 0.0739,
+      "step": 65000
+    },
+    {
+      "epoch": 4.253189401373896,
+      "grad_norm": 0.7860682606697083,
+      "learning_rate": 7.57841793872483e-05,
+      "loss": 0.0621,
+      "step": 65010
+    },
+    {
+      "epoch": 4.253843637553157,
+      "grad_norm": 0.9216861128807068,
+      "learning_rate": 7.577630862774515e-05,
+      "loss": 0.0593,
+      "step": 65020
+    },
+    {
+      "epoch": 4.254497873732418,
+      "grad_norm": 0.9444064497947693,
+      "learning_rate": 7.576843699822394e-05,
+      "loss": 0.0583,
+      "step": 65030
+    },
+    {
+      "epoch": 4.255152109911678,
+      "grad_norm": 0.833625853061676,
+      "learning_rate": 7.57605644989504e-05,
+      "loss": 0.0653,
+      "step": 65040
+    },
+    {
+      "epoch": 4.255806346090939,
+      "grad_norm": 0.8485620617866516,
+      "learning_rate": 7.575269113019027e-05,
+      "loss": 0.0723,
+      "step": 65050
+    },
+    {
+      "epoch": 4.2564605822702,
+      "grad_norm": 0.9002768397331238,
+      "learning_rate": 7.574481689220926e-05,
+      "loss": 0.0667,
+      "step": 65060
+    },
+    {
+      "epoch": 4.257114818449461,
+      "grad_norm": 0.7625710964202881,
+      "learning_rate": 7.573694178527316e-05,
+      "loss": 0.0671,
+      "step": 65070
+    },
+    {
+      "epoch": 4.257769054628721,
+      "grad_norm": 0.912322998046875,
+      "learning_rate": 7.572906580964779e-05,
+      "loss": 0.0707,
+      "step": 65080
+    },
+    {
+      "epoch": 4.258423290807982,
+      "grad_norm": 0.8781670331954956,
+      "learning_rate": 7.572118896559896e-05,
+      "loss": 0.0731,
+      "step": 65090
+    },
+    {
+      "epoch": 4.259077526987243,
+      "grad_norm": 1.0701106786727905,
+      "learning_rate": 7.571331125339256e-05,
+      "loss": 0.062,
+      "step": 65100
+    },
+    {
+      "epoch": 4.259731763166503,
+      "grad_norm": 1.057981014251709,
+      "learning_rate": 7.570543267329446e-05,
+      "loss": 0.0619,
+      "step": 65110
+    },
+    {
+      "epoch": 4.260385999345764,
+      "grad_norm": 0.6521860361099243,
+      "learning_rate": 7.56975532255706e-05,
+      "loss": 0.0532,
+      "step": 65120
+    },
+    {
+      "epoch": 4.261040235525025,
+      "grad_norm": 0.7353628277778625,
+      "learning_rate": 7.568967291048692e-05,
+      "loss": 0.0675,
+      "step": 65130
+    },
+    {
+      "epoch": 4.261694471704285,
+      "grad_norm": 0.8143835067749023,
+      "learning_rate": 7.56817917283094e-05,
+      "loss": 0.0615,
+      "step": 65140
+    },
+    {
+      "epoch": 4.262348707883546,
+      "grad_norm": 0.7694095373153687,
+      "learning_rate": 7.567390967930406e-05,
+      "loss": 0.065,
+      "step": 65150
+    },
+    {
+      "epoch": 4.263002944062807,
+      "grad_norm": 1.1253188848495483,
+      "learning_rate": 7.566602676373694e-05,
+      "loss": 0.0745,
+      "step": 65160
+    },
+    {
+      "epoch": 4.263657180242068,
+      "grad_norm": 0.760852575302124,
+      "learning_rate": 7.56581429818741e-05,
+      "loss": 0.064,
+      "step": 65170
+    },
+    {
+      "epoch": 4.264311416421328,
+      "grad_norm": 0.808857262134552,
+      "learning_rate": 7.565025833398164e-05,
+      "loss": 0.0654,
+      "step": 65180
+    },
+    {
+      "epoch": 4.264965652600589,
+      "grad_norm": 0.7772423028945923,
+      "learning_rate": 7.56423728203257e-05,
+      "loss": 0.0645,
+      "step": 65190
+    },
+    {
+      "epoch": 4.26561988877985,
+      "grad_norm": 0.8490001559257507,
+      "learning_rate": 7.563448644117242e-05,
+      "loss": 0.0618,
+      "step": 65200
+    },
+    {
+      "epoch": 4.26627412495911,
+      "grad_norm": 0.776184618473053,
+      "learning_rate": 7.562659919678801e-05,
+      "loss": 0.0724,
+      "step": 65210
+    },
+    {
+      "epoch": 4.266928361138371,
+      "grad_norm": 0.765455424785614,
+      "learning_rate": 7.561871108743865e-05,
+      "loss": 0.0681,
+      "step": 65220
+    },
+    {
+      "epoch": 4.267582597317632,
+      "grad_norm": 0.8536463975906372,
+      "learning_rate": 7.561082211339062e-05,
+      "loss": 0.0693,
+      "step": 65230
+    },
+    {
+      "epoch": 4.268236833496893,
+      "grad_norm": 0.7764464616775513,
+      "learning_rate": 7.560293227491017e-05,
+      "loss": 0.0532,
+      "step": 65240
+    },
+    {
+      "epoch": 4.268891069676153,
+      "grad_norm": 0.8680564165115356,
+      "learning_rate": 7.55950415722636e-05,
+      "loss": 0.0577,
+      "step": 65250
+    },
+    {
+      "epoch": 4.269545305855414,
+      "grad_norm": 0.8063961267471313,
+      "learning_rate": 7.558715000571726e-05,
+      "loss": 0.0709,
+      "step": 65260
+    },
+    {
+      "epoch": 4.270199542034675,
+      "grad_norm": 0.8812915086746216,
+      "learning_rate": 7.55792575755375e-05,
+      "loss": 0.0586,
+      "step": 65270
+    },
+    {
+      "epoch": 4.270853778213935,
+      "grad_norm": 0.7072269320487976,
+      "learning_rate": 7.55713642819907e-05,
+      "loss": 0.071,
+      "step": 65280
+    },
+    {
+      "epoch": 4.271508014393196,
+      "grad_norm": 0.9915341138839722,
+      "learning_rate": 7.55634701253433e-05,
+      "loss": 0.0548,
+      "step": 65290
+    },
+    {
+      "epoch": 4.272162250572457,
+      "grad_norm": 0.9940981864929199,
+      "learning_rate": 7.555557510586175e-05,
+      "loss": 0.0664,
+      "step": 65300
+    },
+    {
+      "epoch": 4.272816486751718,
+      "grad_norm": 1.065885305404663,
+      "learning_rate": 7.554767922381253e-05,
+      "loss": 0.0663,
+      "step": 65310
+    },
+    {
+      "epoch": 4.273470722930978,
+      "grad_norm": 0.7399426102638245,
+      "learning_rate": 7.553978247946212e-05,
+      "loss": 0.0616,
+      "step": 65320
+    },
+    {
+      "epoch": 4.274124959110239,
+      "grad_norm": 0.7760506868362427,
+      "learning_rate": 7.553188487307705e-05,
+      "loss": 0.0592,
+      "step": 65330
+    },
+    {
+      "epoch": 4.2747791952895,
+      "grad_norm": 0.9584298133850098,
+      "learning_rate": 7.552398640492393e-05,
+      "loss": 0.0646,
+      "step": 65340
+    },
+    {
+      "epoch": 4.27543343146876,
+      "grad_norm": 0.807236909866333,
+      "learning_rate": 7.551608707526933e-05,
+      "loss": 0.0694,
+      "step": 65350
+    },
+    {
+      "epoch": 4.276087667648021,
+      "grad_norm": 0.9472173452377319,
+      "learning_rate": 7.550818688437986e-05,
+      "loss": 0.0608,
+      "step": 65360
+    },
+    {
+      "epoch": 4.276741903827282,
+      "grad_norm": 0.9590891003608704,
+      "learning_rate": 7.55002858325222e-05,
+      "loss": 0.0615,
+      "step": 65370
+    },
+    {
+      "epoch": 4.277396140006543,
+      "grad_norm": 0.8383166193962097,
+      "learning_rate": 7.549238391996302e-05,
+      "loss": 0.0632,
+      "step": 65380
+    },
+    {
+      "epoch": 4.278050376185803,
+      "grad_norm": 0.7799770832061768,
+      "learning_rate": 7.5484481146969e-05,
+      "loss": 0.0631,
+      "step": 65390
+    },
+    {
+      "epoch": 4.278704612365064,
+      "grad_norm": 1.1362422704696655,
+      "learning_rate": 7.547657751380694e-05,
+      "loss": 0.0738,
+      "step": 65400
+    },
+    {
+      "epoch": 4.279358848544325,
+      "grad_norm": 0.9604383111000061,
+      "learning_rate": 7.546867302074354e-05,
+      "loss": 0.0579,
+      "step": 65410
+    },
+    {
+      "epoch": 4.280013084723585,
+      "grad_norm": 0.9436408877372742,
+      "learning_rate": 7.546076766804567e-05,
+      "loss": 0.0669,
+      "step": 65420
+    },
+    {
+      "epoch": 4.280667320902846,
+      "grad_norm": 0.8412542939186096,
+      "learning_rate": 7.545286145598007e-05,
+      "loss": 0.0744,
+      "step": 65430
+    },
+    {
+      "epoch": 4.281321557082107,
+      "grad_norm": 0.7971706986427307,
+      "learning_rate": 7.544495438481367e-05,
+      "loss": 0.0613,
+      "step": 65440
+    },
+    {
+      "epoch": 4.281975793261367,
+      "grad_norm": 0.8335549235343933,
+      "learning_rate": 7.543704645481333e-05,
+      "loss": 0.0696,
+      "step": 65450
+    },
+    {
+      "epoch": 4.282630029440628,
+      "grad_norm": 0.7970999479293823,
+      "learning_rate": 7.542913766624596e-05,
+      "loss": 0.0631,
+      "step": 65460
+    },
+    {
+      "epoch": 4.283284265619889,
+      "grad_norm": 0.990964949131012,
+      "learning_rate": 7.542122801937849e-05,
+      "loss": 0.0677,
+      "step": 65470
+    },
+    {
+      "epoch": 4.28393850179915,
+      "grad_norm": 1.0644537210464478,
+      "learning_rate": 7.541331751447792e-05,
+      "loss": 0.0689,
+      "step": 65480
+    },
+    {
+      "epoch": 4.28459273797841,
+      "grad_norm": 0.9389525055885315,
+      "learning_rate": 7.540540615181123e-05,
+      "loss": 0.0643,
+      "step": 65490
+    },
+    {
+      "epoch": 4.285246974157671,
+      "grad_norm": 0.8526713252067566,
+      "learning_rate": 7.539749393164546e-05,
+      "loss": 0.0662,
+      "step": 65500
+    },
+    {
+      "epoch": 4.285901210336932,
+      "grad_norm": 0.9678772687911987,
+      "learning_rate": 7.538958085424765e-05,
+      "loss": 0.065,
+      "step": 65510
+    },
+    {
+      "epoch": 4.286555446516193,
+      "grad_norm": 0.8349334001541138,
+      "learning_rate": 7.53816669198849e-05,
+      "loss": 0.0737,
+      "step": 65520
+    },
+    {
+      "epoch": 4.287209682695453,
+      "grad_norm": 0.8528868556022644,
+      "learning_rate": 7.537375212882433e-05,
+      "loss": 0.0585,
+      "step": 65530
+    },
+    {
+      "epoch": 4.287863918874714,
+      "grad_norm": 0.7921110987663269,
+      "learning_rate": 7.536583648133311e-05,
+      "loss": 0.0621,
+      "step": 65540
+    },
+    {
+      "epoch": 4.288518155053975,
+      "grad_norm": 1.04022216796875,
+      "learning_rate": 7.535791997767834e-05,
+      "loss": 0.061,
+      "step": 65550
+    },
+    {
+      "epoch": 4.289172391233235,
+      "grad_norm": 0.7660422325134277,
+      "learning_rate": 7.535000261812729e-05,
+      "loss": 0.0618,
+      "step": 65560
+    },
+    {
+      "epoch": 4.289826627412496,
+      "grad_norm": 1.0147818326950073,
+      "learning_rate": 7.534208440294717e-05,
+      "loss": 0.0584,
+      "step": 65570
+    },
+    {
+      "epoch": 4.290480863591757,
+      "grad_norm": 0.7676869630813599,
+      "learning_rate": 7.533416533240523e-05,
+      "loss": 0.0664,
+      "step": 65580
+    },
+    {
+      "epoch": 4.291135099771017,
+      "grad_norm": 0.7164015173912048,
+      "learning_rate": 7.532624540676876e-05,
+      "loss": 0.0643,
+      "step": 65590
+    },
+    {
+      "epoch": 4.291789335950278,
+      "grad_norm": 0.7891098856925964,
+      "learning_rate": 7.53183246263051e-05,
+      "loss": 0.0799,
+      "step": 65600
+    },
+    {
+      "epoch": 4.292443572129539,
+      "grad_norm": 0.8349225521087646,
+      "learning_rate": 7.531040299128158e-05,
+      "loss": 0.0689,
+      "step": 65610
+    },
+    {
+      "epoch": 4.2930978083088,
+      "grad_norm": 0.9937372803688049,
+      "learning_rate": 7.530248050196557e-05,
+      "loss": 0.0657,
+      "step": 65620
+    },
+    {
+      "epoch": 4.29375204448806,
+      "grad_norm": 0.9106799960136414,
+      "learning_rate": 7.529455715862452e-05,
+      "loss": 0.0684,
+      "step": 65630
+    },
+    {
+      "epoch": 4.294406280667321,
+      "grad_norm": 0.9614897966384888,
+      "learning_rate": 7.52866329615258e-05,
+      "loss": 0.0642,
+      "step": 65640
+    },
+    {
+      "epoch": 4.295060516846582,
+      "grad_norm": 0.914214015007019,
+      "learning_rate": 7.527870791093691e-05,
+      "loss": 0.0648,
+      "step": 65650
+    },
+    {
+      "epoch": 4.295714753025842,
+      "grad_norm": 0.8048765063285828,
+      "learning_rate": 7.527078200712533e-05,
+      "loss": 0.0655,
+      "step": 65660
+    },
+    {
+      "epoch": 4.296368989205103,
+      "grad_norm": 0.9100319743156433,
+      "learning_rate": 7.526285525035858e-05,
+      "loss": 0.0678,
+      "step": 65670
+    },
+    {
+      "epoch": 4.297023225384364,
+      "grad_norm": 0.9633187055587769,
+      "learning_rate": 7.52549276409042e-05,
+      "loss": 0.0681,
+      "step": 65680
+    },
+    {
+      "epoch": 4.297677461563625,
+      "grad_norm": 0.7984033823013306,
+      "learning_rate": 7.52469991790298e-05,
+      "loss": 0.0609,
+      "step": 65690
+    },
+    {
+      "epoch": 4.298331697742885,
+      "grad_norm": 0.9760509729385376,
+      "learning_rate": 7.523906986500296e-05,
+      "loss": 0.0739,
+      "step": 65700
+    },
+    {
+      "epoch": 4.298985933922146,
+      "grad_norm": 1.10667085647583,
+      "learning_rate": 7.52311396990913e-05,
+      "loss": 0.0653,
+      "step": 65710
+    },
+    {
+      "epoch": 4.299640170101407,
+      "grad_norm": 1.0350356101989746,
+      "learning_rate": 7.522320868156253e-05,
+      "loss": 0.0656,
+      "step": 65720
+    },
+    {
+      "epoch": 4.300294406280667,
+      "grad_norm": 1.0190367698669434,
+      "learning_rate": 7.521527681268431e-05,
+      "loss": 0.0714,
+      "step": 65730
+    },
+    {
+      "epoch": 4.300948642459928,
+      "grad_norm": 0.7813534140586853,
+      "learning_rate": 7.520734409272437e-05,
+      "loss": 0.0592,
+      "step": 65740
+    },
+    {
+      "epoch": 4.301602878639189,
+      "grad_norm": 0.9878474473953247,
+      "learning_rate": 7.519941052195045e-05,
+      "loss": 0.0642,
+      "step": 65750
+    },
+    {
+      "epoch": 4.30225711481845,
+      "grad_norm": 0.7413721680641174,
+      "learning_rate": 7.519147610063035e-05,
+      "loss": 0.0593,
+      "step": 65760
+    },
+    {
+      "epoch": 4.30291135099771,
+      "grad_norm": 0.8811038732528687,
+      "learning_rate": 7.518354082903184e-05,
+      "loss": 0.0664,
+      "step": 65770
+    },
+    {
+      "epoch": 4.303565587176971,
+      "grad_norm": 0.9010792970657349,
+      "learning_rate": 7.517560470742279e-05,
+      "loss": 0.0669,
+      "step": 65780
+    },
+    {
+      "epoch": 4.304219823356232,
+      "grad_norm": 0.8902906775474548,
+      "learning_rate": 7.516766773607107e-05,
+      "loss": 0.068,
+      "step": 65790
+    },
+    {
+      "epoch": 4.304874059535492,
+      "grad_norm": 0.9187259674072266,
+      "learning_rate": 7.515972991524454e-05,
+      "loss": 0.0681,
+      "step": 65800
+    },
+    {
+      "epoch": 4.305528295714753,
+      "grad_norm": 0.7792700529098511,
+      "learning_rate": 7.515179124521116e-05,
+      "loss": 0.0599,
+      "step": 65810
+    },
+    {
+      "epoch": 4.306182531894014,
+      "grad_norm": 0.7929428815841675,
+      "learning_rate": 7.514385172623886e-05,
+      "loss": 0.0596,
+      "step": 65820
+    },
+    {
+      "epoch": 4.306836768073275,
+      "grad_norm": 0.9037280678749084,
+      "learning_rate": 7.513591135859561e-05,
+      "loss": 0.0637,
+      "step": 65830
+    },
+    {
+      "epoch": 4.307491004252535,
+      "grad_norm": 0.8752690553665161,
+      "learning_rate": 7.512797014254944e-05,
+      "loss": 0.0617,
+      "step": 65840
+    },
+    {
+      "epoch": 4.308145240431796,
+      "grad_norm": 0.9982262253761292,
+      "learning_rate": 7.512002807836838e-05,
+      "loss": 0.0617,
+      "step": 65850
+    },
+    {
+      "epoch": 4.308799476611057,
+      "grad_norm": 1.1322745084762573,
+      "learning_rate": 7.511208516632047e-05,
+      "loss": 0.0665,
+      "step": 65860
+    },
+    {
+      "epoch": 4.309453712790317,
+      "grad_norm": 0.8582587242126465,
+      "learning_rate": 7.510414140667385e-05,
+      "loss": 0.0717,
+      "step": 65870
+    },
+    {
+      "epoch": 4.310107948969578,
+      "grad_norm": 1.0602346658706665,
+      "learning_rate": 7.50961967996966e-05,
+      "loss": 0.0662,
+      "step": 65880
+    },
+    {
+      "epoch": 4.310762185148839,
+      "grad_norm": 0.9373438954353333,
+      "learning_rate": 7.508825134565692e-05,
+      "loss": 0.0596,
+      "step": 65890
+    },
+    {
+      "epoch": 4.311416421328099,
+      "grad_norm": 0.8777552843093872,
+      "learning_rate": 7.508030504482296e-05,
+      "loss": 0.0611,
+      "step": 65900
+    },
+    {
+      "epoch": 4.31207065750736,
+      "grad_norm": 0.8960667848587036,
+      "learning_rate": 7.50723578974629e-05,
+      "loss": 0.0631,
+      "step": 65910
+    },
+    {
+      "epoch": 4.312724893686621,
+      "grad_norm": 1.027182698249817,
+      "learning_rate": 7.506440990384502e-05,
+      "loss": 0.0668,
+      "step": 65920
+    },
+    {
+      "epoch": 4.313379129865882,
+      "grad_norm": 0.7670314311981201,
+      "learning_rate": 7.505646106423756e-05,
+      "loss": 0.0557,
+      "step": 65930
+    },
+    {
+      "epoch": 4.314033366045142,
+      "grad_norm": 0.9890926480293274,
+      "learning_rate": 7.504851137890885e-05,
+      "loss": 0.0718,
+      "step": 65940
+    },
+    {
+      "epoch": 4.314687602224403,
+      "grad_norm": 0.7140153050422668,
+      "learning_rate": 7.504056084812718e-05,
+      "loss": 0.0695,
+      "step": 65950
+    },
+    {
+      "epoch": 4.315341838403664,
+      "grad_norm": 0.8227149844169617,
+      "learning_rate": 7.50326094721609e-05,
+      "loss": 0.0566,
+      "step": 65960
+    },
+    {
+      "epoch": 4.315996074582925,
+      "grad_norm": 1.0063236951828003,
+      "learning_rate": 7.502465725127839e-05,
+      "loss": 0.0601,
+      "step": 65970
+    },
+    {
+      "epoch": 4.316650310762185,
+      "grad_norm": 0.8858514428138733,
+      "learning_rate": 7.501670418574808e-05,
+      "loss": 0.0561,
+      "step": 65980
+    },
+    {
+      "epoch": 4.317304546941446,
+      "grad_norm": 0.9521268606185913,
+      "learning_rate": 7.500875027583843e-05,
+      "loss": 0.0743,
+      "step": 65990
+    },
+    {
+      "epoch": 4.317958783120707,
+      "grad_norm": 0.860778272151947,
+      "learning_rate": 7.500079552181782e-05,
+      "loss": 0.056,
+      "step": 66000
+    },
+    {
+      "epoch": 4.318613019299967,
+      "grad_norm": 0.6561046838760376,
+      "learning_rate": 7.499283992395483e-05,
+      "loss": 0.0579,
+      "step": 66010
+    },
+    {
+      "epoch": 4.319267255479228,
+      "grad_norm": 0.9803190231323242,
+      "learning_rate": 7.498488348251794e-05,
+      "loss": 0.0664,
+      "step": 66020
+    },
+    {
+      "epoch": 4.319921491658489,
+      "grad_norm": 0.782440185546875,
+      "learning_rate": 7.497692619777568e-05,
+      "loss": 0.067,
+      "step": 66030
+    },
+    {
+      "epoch": 4.320575727837749,
+      "grad_norm": 0.900099515914917,
+      "learning_rate": 7.496896806999667e-05,
+      "loss": 0.0708,
+      "step": 66040
+    },
+    {
+      "epoch": 4.32122996401701,
+      "grad_norm": 0.7509415745735168,
+      "learning_rate": 7.496100909944952e-05,
+      "loss": 0.0639,
+      "step": 66050
+    },
+    {
+      "epoch": 4.321884200196271,
+      "grad_norm": 0.8425498008728027,
+      "learning_rate": 7.495304928640284e-05,
+      "loss": 0.0697,
+      "step": 66060
+    },
+    {
+      "epoch": 4.322538436375532,
+      "grad_norm": 0.8149582147598267,
+      "learning_rate": 7.494508863112529e-05,
+      "loss": 0.0615,
+      "step": 66070
+    },
+    {
+      "epoch": 4.323192672554792,
+      "grad_norm": 0.8617916703224182,
+      "learning_rate": 7.49371271338856e-05,
+      "loss": 0.0663,
+      "step": 66080
+    },
+    {
+      "epoch": 4.323846908734053,
+      "grad_norm": 0.7951449155807495,
+      "learning_rate": 7.492916479495246e-05,
+      "loss": 0.061,
+      "step": 66090
+    },
+    {
+      "epoch": 4.324501144913314,
+      "grad_norm": 0.9237793684005737,
+      "learning_rate": 7.492120161459463e-05,
+      "loss": 0.0655,
+      "step": 66100
+    },
+    {
+      "epoch": 4.325155381092574,
+      "grad_norm": 0.8202549815177917,
+      "learning_rate": 7.491323759308089e-05,
+      "loss": 0.0588,
+      "step": 66110
+    },
+    {
+      "epoch": 4.325809617271835,
+      "grad_norm": 0.8115940690040588,
+      "learning_rate": 7.490527273068003e-05,
+      "loss": 0.0609,
+      "step": 66120
+    },
+    {
+      "epoch": 4.326463853451096,
+      "grad_norm": 0.8623285889625549,
+      "learning_rate": 7.489730702766092e-05,
+      "loss": 0.0694,
+      "step": 66130
+    },
+    {
+      "epoch": 4.327118089630357,
+      "grad_norm": 0.9182894825935364,
+      "learning_rate": 7.488934048429239e-05,
+      "loss": 0.0687,
+      "step": 66140
+    },
+    {
+      "epoch": 4.327772325809617,
+      "grad_norm": 0.8562094569206238,
+      "learning_rate": 7.488137310084334e-05,
+      "loss": 0.0567,
+      "step": 66150
+    },
+    {
+      "epoch": 4.328426561988878,
+      "grad_norm": 0.6775592565536499,
+      "learning_rate": 7.487340487758271e-05,
+      "loss": 0.0616,
+      "step": 66160
+    },
+    {
+      "epoch": 4.329080798168139,
+      "grad_norm": 0.9210445880889893,
+      "learning_rate": 7.486543581477942e-05,
+      "loss": 0.066,
+      "step": 66170
+    },
+    {
+      "epoch": 4.329735034347399,
+      "grad_norm": 0.8640944957733154,
+      "learning_rate": 7.485746591270247e-05,
+      "loss": 0.0679,
+      "step": 66180
+    },
+    {
+      "epoch": 4.33038927052666,
+      "grad_norm": 0.9306260943412781,
+      "learning_rate": 7.484949517162083e-05,
+      "loss": 0.0677,
+      "step": 66190
+    },
+    {
+      "epoch": 4.331043506705921,
+      "grad_norm": 0.9359939098358154,
+      "learning_rate": 7.484152359180358e-05,
+      "loss": 0.063,
+      "step": 66200
+    },
+    {
+      "epoch": 4.331697742885182,
+      "grad_norm": 0.6757792830467224,
+      "learning_rate": 7.483355117351975e-05,
+      "loss": 0.0645,
+      "step": 66210
+    },
+    {
+      "epoch": 4.332351979064442,
+      "grad_norm": 1.1225383281707764,
+      "learning_rate": 7.482557791703843e-05,
+      "loss": 0.0551,
+      "step": 66220
+    },
+    {
+      "epoch": 4.333006215243703,
+      "grad_norm": 0.8238813877105713,
+      "learning_rate": 7.481760382262876e-05,
+      "loss": 0.0648,
+      "step": 66230
+    },
+    {
+      "epoch": 4.333660451422964,
+      "grad_norm": 0.8821508288383484,
+      "learning_rate": 7.480962889055989e-05,
+      "loss": 0.064,
+      "step": 66240
+    },
+    {
+      "epoch": 4.334314687602224,
+      "grad_norm": 0.8771295547485352,
+      "learning_rate": 7.480165312110096e-05,
+      "loss": 0.0546,
+      "step": 66250
+    },
+    {
+      "epoch": 4.334968923781485,
+      "grad_norm": 0.9171880483627319,
+      "learning_rate": 7.479367651452119e-05,
+      "loss": 0.0613,
+      "step": 66260
+    },
+    {
+      "epoch": 4.335623159960746,
+      "grad_norm": 0.8392135500907898,
+      "learning_rate": 7.478569907108983e-05,
+      "loss": 0.0664,
+      "step": 66270
+    },
+    {
+      "epoch": 4.336277396140007,
+      "grad_norm": 0.8099778890609741,
+      "learning_rate": 7.477772079107612e-05,
+      "loss": 0.0602,
+      "step": 66280
+    },
+    {
+      "epoch": 4.336931632319267,
+      "grad_norm": 0.8127948641777039,
+      "learning_rate": 7.476974167474934e-05,
+      "loss": 0.0604,
+      "step": 66290
+    },
+    {
+      "epoch": 4.337585868498528,
+      "grad_norm": 0.9666029214859009,
+      "learning_rate": 7.476176172237883e-05,
+      "loss": 0.061,
+      "step": 66300
+    },
+    {
+      "epoch": 4.338240104677789,
+      "grad_norm": 0.9391130208969116,
+      "learning_rate": 7.475378093423391e-05,
+      "loss": 0.0647,
+      "step": 66310
+    },
+    {
+      "epoch": 4.338894340857049,
+      "grad_norm": 0.7683467864990234,
+      "learning_rate": 7.474579931058397e-05,
+      "loss": 0.0622,
+      "step": 66320
+    },
+    {
+      "epoch": 4.33954857703631,
+      "grad_norm": 0.8079342246055603,
+      "learning_rate": 7.47378168516984e-05,
+      "loss": 0.0572,
+      "step": 66330
+    },
+    {
+      "epoch": 4.340202813215571,
+      "grad_norm": 1.2922800779342651,
+      "learning_rate": 7.472983355784664e-05,
+      "loss": 0.0781,
+      "step": 66340
+    },
+    {
+      "epoch": 4.340857049394831,
+      "grad_norm": 0.7486710548400879,
+      "learning_rate": 7.472184942929815e-05,
+      "loss": 0.0631,
+      "step": 66350
+    },
+    {
+      "epoch": 4.341511285574092,
+      "grad_norm": 0.7846554517745972,
+      "learning_rate": 7.471386446632238e-05,
+      "loss": 0.063,
+      "step": 66360
+    },
+    {
+      "epoch": 4.342165521753353,
+      "grad_norm": 0.6966447234153748,
+      "learning_rate": 7.470587866918889e-05,
+      "loss": 0.0614,
+      "step": 66370
+    },
+    {
+      "epoch": 4.342819757932614,
+      "grad_norm": 0.8768433928489685,
+      "learning_rate": 7.469789203816719e-05,
+      "loss": 0.0598,
+      "step": 66380
+    },
+    {
+      "epoch": 4.343473994111874,
+      "grad_norm": 1.1041473150253296,
+      "learning_rate": 7.468990457352687e-05,
+      "loss": 0.0684,
+      "step": 66390
+    },
+    {
+      "epoch": 4.344128230291135,
+      "grad_norm": 0.9726035594940186,
+      "learning_rate": 7.468191627553753e-05,
+      "loss": 0.0635,
+      "step": 66400
+    },
+    {
+      "epoch": 4.344782466470396,
+      "grad_norm": 0.9272595643997192,
+      "learning_rate": 7.467392714446876e-05,
+      "loss": 0.0695,
+      "step": 66410
+    },
+    {
+      "epoch": 4.345436702649657,
+      "grad_norm": 0.8459830284118652,
+      "learning_rate": 7.466593718059026e-05,
+      "loss": 0.0527,
+      "step": 66420
+    },
+    {
+      "epoch": 4.346090938828917,
+      "grad_norm": 0.9491352438926697,
+      "learning_rate": 7.465794638417167e-05,
+      "loss": 0.0611,
+      "step": 66430
+    },
+    {
+      "epoch": 4.346745175008178,
+      "grad_norm": 0.9533461928367615,
+      "learning_rate": 7.464995475548275e-05,
+      "loss": 0.0687,
+      "step": 66440
+    },
+    {
+      "epoch": 4.347399411187439,
+      "grad_norm": 0.8215408325195312,
+      "learning_rate": 7.464196229479317e-05,
+      "loss": 0.0617,
+      "step": 66450
+    },
+    {
+      "epoch": 4.348053647366699,
+      "grad_norm": 0.7235555052757263,
+      "learning_rate": 7.463396900237277e-05,
+      "loss": 0.058,
+      "step": 66460
+    },
+    {
+      "epoch": 4.34870788354596,
+      "grad_norm": 0.8705189824104309,
+      "learning_rate": 7.462597487849131e-05,
+      "loss": 0.0595,
+      "step": 66470
+    },
+    {
+      "epoch": 4.349362119725221,
+      "grad_norm": 0.906433641910553,
+      "learning_rate": 7.461797992341861e-05,
+      "loss": 0.0626,
+      "step": 66480
+    },
+    {
+      "epoch": 4.350016355904481,
+      "grad_norm": 0.7737306356430054,
+      "learning_rate": 7.460998413742451e-05,
+      "loss": 0.0704,
+      "step": 66490
+    },
+    {
+      "epoch": 4.350670592083742,
+      "grad_norm": 0.782434344291687,
+      "learning_rate": 7.460198752077892e-05,
+      "loss": 0.0603,
+      "step": 66500
+    },
+    {
+      "epoch": 4.351324828263003,
+      "grad_norm": 0.7962082028388977,
+      "learning_rate": 7.459399007375172e-05,
+      "loss": 0.0656,
+      "step": 66510
+    },
+    {
+      "epoch": 4.351979064442264,
+      "grad_norm": 0.8921648263931274,
+      "learning_rate": 7.458599179661286e-05,
+      "loss": 0.0634,
+      "step": 66520
+    },
+    {
+      "epoch": 4.352633300621524,
+      "grad_norm": 0.868022084236145,
+      "learning_rate": 7.45779926896323e-05,
+      "loss": 0.0594,
+      "step": 66530
+    },
+    {
+      "epoch": 4.353287536800785,
+      "grad_norm": 1.0638127326965332,
+      "learning_rate": 7.456999275308002e-05,
+      "loss": 0.0655,
+      "step": 66540
+    },
+    {
+      "epoch": 4.353941772980046,
+      "grad_norm": 0.8913347721099854,
+      "learning_rate": 7.456199198722604e-05,
+      "loss": 0.056,
+      "step": 66550
+    },
+    {
+      "epoch": 4.354596009159306,
+      "grad_norm": 0.9430878162384033,
+      "learning_rate": 7.455399039234043e-05,
+      "loss": 0.0648,
+      "step": 66560
+    },
+    {
+      "epoch": 4.355250245338567,
+      "grad_norm": 1.181059718132019,
+      "learning_rate": 7.454598796869325e-05,
+      "loss": 0.0701,
+      "step": 66570
+    },
+    {
+      "epoch": 4.355904481517828,
+      "grad_norm": 0.9202277064323425,
+      "learning_rate": 7.45379847165546e-05,
+      "loss": 0.0633,
+      "step": 66580
+    },
+    {
+      "epoch": 4.356558717697089,
+      "grad_norm": 0.9049434661865234,
+      "learning_rate": 7.45299806361946e-05,
+      "loss": 0.0625,
+      "step": 66590
+    },
+    {
+      "epoch": 4.357212953876349,
+      "grad_norm": 1.0947158336639404,
+      "learning_rate": 7.452197572788345e-05,
+      "loss": 0.0639,
+      "step": 66600
+    },
+    {
+      "epoch": 4.35786719005561,
+      "grad_norm": 0.9060896635055542,
+      "learning_rate": 7.451396999189129e-05,
+      "loss": 0.0599,
+      "step": 66610
+    },
+    {
+      "epoch": 4.358521426234871,
+      "grad_norm": 0.9136892557144165,
+      "learning_rate": 7.450596342848835e-05,
+      "loss": 0.0671,
+      "step": 66620
+    },
+    {
+      "epoch": 4.359175662414131,
+      "grad_norm": 0.7498853802680969,
+      "learning_rate": 7.449795603794487e-05,
+      "loss": 0.0585,
+      "step": 66630
+    },
+    {
+      "epoch": 4.359829898593392,
+      "grad_norm": 0.9020270705223083,
+      "learning_rate": 7.448994782053114e-05,
+      "loss": 0.0678,
+      "step": 66640
+    },
+    {
+      "epoch": 4.360484134772653,
+      "grad_norm": 0.9565746188163757,
+      "learning_rate": 7.448193877651743e-05,
+      "loss": 0.0662,
+      "step": 66650
+    },
+    {
+      "epoch": 4.361138370951914,
+      "grad_norm": 0.9026892185211182,
+      "learning_rate": 7.447392890617408e-05,
+      "loss": 0.0647,
+      "step": 66660
+    },
+    {
+      "epoch": 4.361792607131174,
+      "grad_norm": 1.156490683555603,
+      "learning_rate": 7.446591820977144e-05,
+      "loss": 0.0648,
+      "step": 66670
+    },
+    {
+      "epoch": 4.362446843310435,
+      "grad_norm": 0.8447253704071045,
+      "learning_rate": 7.445790668757992e-05,
+      "loss": 0.0662,
+      "step": 66680
+    },
+    {
+      "epoch": 4.363101079489696,
+      "grad_norm": 0.8401336073875427,
+      "learning_rate": 7.44498943398699e-05,
+      "loss": 0.0621,
+      "step": 66690
+    },
+    {
+      "epoch": 4.363755315668956,
+      "grad_norm": 0.88148033618927,
+      "learning_rate": 7.44418811669118e-05,
+      "loss": 0.0688,
+      "step": 66700
+    },
+    {
+      "epoch": 4.364409551848217,
+      "grad_norm": 1.0117405652999878,
+      "learning_rate": 7.443386716897614e-05,
+      "loss": 0.0547,
+      "step": 66710
+    },
+    {
+      "epoch": 4.365063788027478,
+      "grad_norm": 0.9008330702781677,
+      "learning_rate": 7.442585234633337e-05,
+      "loss": 0.0676,
+      "step": 66720
+    },
+    {
+      "epoch": 4.365718024206739,
+      "grad_norm": 0.8661233186721802,
+      "learning_rate": 7.441783669925402e-05,
+      "loss": 0.0613,
+      "step": 66730
+    },
+    {
+      "epoch": 4.366372260385999,
+      "grad_norm": 0.849505603313446,
+      "learning_rate": 7.440982022800864e-05,
+      "loss": 0.0652,
+      "step": 66740
+    },
+    {
+      "epoch": 4.36702649656526,
+      "grad_norm": 0.7633360028266907,
+      "learning_rate": 7.440180293286783e-05,
+      "loss": 0.0601,
+      "step": 66750
+    },
+    {
+      "epoch": 4.367680732744521,
+      "grad_norm": 0.9463858008384705,
+      "learning_rate": 7.439378481410215e-05,
+      "loss": 0.0651,
+      "step": 66760
+    },
+    {
+      "epoch": 4.368334968923781,
+      "grad_norm": 0.9335405826568604,
+      "learning_rate": 7.438576587198228e-05,
+      "loss": 0.0639,
+      "step": 66770
+    },
+    {
+      "epoch": 4.368989205103042,
+      "grad_norm": 0.8489439487457275,
+      "learning_rate": 7.437774610677884e-05,
+      "loss": 0.057,
+      "step": 66780
+    },
+    {
+      "epoch": 4.369643441282303,
+      "grad_norm": 0.9657695889472961,
+      "learning_rate": 7.436972551876255e-05,
+      "loss": 0.0632,
+      "step": 66790
+    },
+    {
+      "epoch": 4.370297677461563,
+      "grad_norm": 0.8305515050888062,
+      "learning_rate": 7.43617041082041e-05,
+      "loss": 0.0664,
+      "step": 66800
+    },
+    {
+      "epoch": 4.370951913640824,
+      "grad_norm": 1.0466861724853516,
+      "learning_rate": 7.435368187537424e-05,
+      "loss": 0.057,
+      "step": 66810
+    },
+    {
+      "epoch": 4.371606149820085,
+      "grad_norm": 1.0870552062988281,
+      "learning_rate": 7.434565882054377e-05,
+      "loss": 0.0614,
+      "step": 66820
+    },
+    {
+      "epoch": 4.372260385999346,
+      "grad_norm": 0.9726265072822571,
+      "learning_rate": 7.433763494398345e-05,
+      "loss": 0.0579,
+      "step": 66830
+    },
+    {
+      "epoch": 4.372914622178606,
+      "grad_norm": 0.8406111001968384,
+      "learning_rate": 7.432961024596413e-05,
+      "loss": 0.0625,
+      "step": 66840
+    },
+    {
+      "epoch": 4.373568858357867,
+      "grad_norm": 0.9154638648033142,
+      "learning_rate": 7.432158472675665e-05,
+      "loss": 0.0695,
+      "step": 66850
+    },
+    {
+      "epoch": 4.374223094537128,
+      "grad_norm": 0.8148864507675171,
+      "learning_rate": 7.43135583866319e-05,
+      "loss": 0.0657,
+      "step": 66860
+    },
+    {
+      "epoch": 4.374877330716389,
+      "grad_norm": 1.0831327438354492,
+      "learning_rate": 7.430553122586079e-05,
+      "loss": 0.0759,
+      "step": 66870
+    },
+    {
+      "epoch": 4.375531566895649,
+      "grad_norm": 0.7323132157325745,
+      "learning_rate": 7.429750324471425e-05,
+      "loss": 0.0575,
+      "step": 66880
+    },
+    {
+      "epoch": 4.37618580307491,
+      "grad_norm": 0.7715146541595459,
+      "learning_rate": 7.428947444346327e-05,
+      "loss": 0.0591,
+      "step": 66890
+    },
+    {
+      "epoch": 4.376840039254171,
+      "grad_norm": 0.7162055969238281,
+      "learning_rate": 7.428144482237882e-05,
+      "loss": 0.0633,
+      "step": 66900
+    },
+    {
+      "epoch": 4.377494275433431,
+      "grad_norm": 0.9126048684120178,
+      "learning_rate": 7.427341438173192e-05,
+      "loss": 0.0738,
+      "step": 66910
+    },
+    {
+      "epoch": 4.378148511612692,
+      "grad_norm": 0.8502604365348816,
+      "learning_rate": 7.426538312179364e-05,
+      "loss": 0.0591,
+      "step": 66920
+    },
+    {
+      "epoch": 4.378802747791953,
+      "grad_norm": 1.0491002798080444,
+      "learning_rate": 7.425735104283502e-05,
+      "loss": 0.0642,
+      "step": 66930
+    },
+    {
+      "epoch": 4.379456983971213,
+      "grad_norm": 1.0379705429077148,
+      "learning_rate": 7.424931814512721e-05,
+      "loss": 0.0692,
+      "step": 66940
+    },
+    {
+      "epoch": 4.380111220150474,
+      "grad_norm": 0.9952113032341003,
+      "learning_rate": 7.42412844289413e-05,
+      "loss": 0.075,
+      "step": 66950
+    },
+    {
+      "epoch": 4.380765456329735,
+      "grad_norm": 0.9746718406677246,
+      "learning_rate": 7.423324989454847e-05,
+      "loss": 0.0598,
+      "step": 66960
+    },
+    {
+      "epoch": 4.381419692508996,
+      "grad_norm": 0.8205549716949463,
+      "learning_rate": 7.42252145422199e-05,
+      "loss": 0.0565,
+      "step": 66970
+    },
+    {
+      "epoch": 4.382073928688256,
+      "grad_norm": 0.9923220276832581,
+      "learning_rate": 7.421717837222682e-05,
+      "loss": 0.0717,
+      "step": 66980
+    },
+    {
+      "epoch": 4.382728164867517,
+      "grad_norm": 0.804681658744812,
+      "learning_rate": 7.420914138484045e-05,
+      "loss": 0.0645,
+      "step": 66990
+    },
+    {
+      "epoch": 4.383382401046778,
+      "grad_norm": 0.7509680390357971,
+      "learning_rate": 7.420110358033205e-05,
+      "loss": 0.0627,
+      "step": 67000
+    },
+    {
+      "epoch": 4.384036637226038,
+      "grad_norm": 0.7687981128692627,
+      "learning_rate": 7.419306495897295e-05,
+      "loss": 0.069,
+      "step": 67010
+    },
+    {
+      "epoch": 4.384690873405299,
+      "grad_norm": 0.8498753309249878,
+      "learning_rate": 7.418502552103446e-05,
+      "loss": 0.0653,
+      "step": 67020
+    },
+    {
+      "epoch": 4.38534510958456,
+      "grad_norm": 0.9675426483154297,
+      "learning_rate": 7.417698526678792e-05,
+      "loss": 0.0728,
+      "step": 67030
+    },
+    {
+      "epoch": 4.385999345763821,
+      "grad_norm": 0.9091965556144714,
+      "learning_rate": 7.416894419650473e-05,
+      "loss": 0.0641,
+      "step": 67040
+    },
+    {
+      "epoch": 4.386653581943081,
+      "grad_norm": 0.8700924515724182,
+      "learning_rate": 7.416090231045629e-05,
+      "loss": 0.057,
+      "step": 67050
+    },
+    {
+      "epoch": 4.387307818122342,
+      "grad_norm": 0.7336810231208801,
+      "learning_rate": 7.415285960891403e-05,
+      "loss": 0.0683,
+      "step": 67060
+    },
+    {
+      "epoch": 4.387962054301603,
+      "grad_norm": 0.9728383421897888,
+      "learning_rate": 7.414481609214941e-05,
+      "loss": 0.0596,
+      "step": 67070
+    },
+    {
+      "epoch": 4.388616290480863,
+      "grad_norm": 0.6819276809692383,
+      "learning_rate": 7.413677176043393e-05,
+      "loss": 0.0561,
+      "step": 67080
+    },
+    {
+      "epoch": 4.389270526660124,
+      "grad_norm": 0.9143689870834351,
+      "learning_rate": 7.41287266140391e-05,
+      "loss": 0.0592,
+      "step": 67090
+    },
+    {
+      "epoch": 4.389924762839385,
+      "grad_norm": 0.8440172672271729,
+      "learning_rate": 7.412068065323648e-05,
+      "loss": 0.0643,
+      "step": 67100
+    },
+    {
+      "epoch": 4.390578999018646,
+      "grad_norm": 0.8751739859580994,
+      "learning_rate": 7.411263387829761e-05,
+      "loss": 0.0652,
+      "step": 67110
+    },
+    {
+      "epoch": 4.391233235197906,
+      "grad_norm": 1.0761669874191284,
+      "learning_rate": 7.41045862894941e-05,
+      "loss": 0.0728,
+      "step": 67120
+    },
+    {
+      "epoch": 4.391887471377167,
+      "grad_norm": 0.7881830334663391,
+      "learning_rate": 7.409653788709762e-05,
+      "loss": 0.0655,
+      "step": 67130
+    },
+    {
+      "epoch": 4.392541707556428,
+      "grad_norm": 0.9639447331428528,
+      "learning_rate": 7.408848867137977e-05,
+      "loss": 0.0584,
+      "step": 67140
+    },
+    {
+      "epoch": 4.393195943735688,
+      "grad_norm": 0.8875806331634521,
+      "learning_rate": 7.408043864261225e-05,
+      "loss": 0.0745,
+      "step": 67150
+    },
+    {
+      "epoch": 4.393850179914949,
+      "grad_norm": 0.7383980751037598,
+      "learning_rate": 7.407238780106679e-05,
+      "loss": 0.0572,
+      "step": 67160
+    },
+    {
+      "epoch": 4.39450441609421,
+      "grad_norm": 0.848702609539032,
+      "learning_rate": 7.406433614701509e-05,
+      "loss": 0.0632,
+      "step": 67170
+    },
+    {
+      "epoch": 4.395158652273471,
+      "grad_norm": 0.7714468240737915,
+      "learning_rate": 7.405628368072894e-05,
+      "loss": 0.0595,
+      "step": 67180
+    },
+    {
+      "epoch": 4.395812888452731,
+      "grad_norm": 0.775048017501831,
+      "learning_rate": 7.404823040248013e-05,
+      "loss": 0.0612,
+      "step": 67190
+    },
+    {
+      "epoch": 4.396467124631992,
+      "grad_norm": 0.6783040165901184,
+      "learning_rate": 7.404017631254047e-05,
+      "loss": 0.0576,
+      "step": 67200
+    },
+    {
+      "epoch": 4.397121360811253,
+      "grad_norm": 0.9950321912765503,
+      "learning_rate": 7.403212141118182e-05,
+      "loss": 0.0645,
+      "step": 67210
+    },
+    {
+      "epoch": 4.397775596990513,
+      "grad_norm": 0.9404067993164062,
+      "learning_rate": 7.402406569867604e-05,
+      "loss": 0.0632,
+      "step": 67220
+    },
+    {
+      "epoch": 4.398429833169774,
+      "grad_norm": 0.8246940970420837,
+      "learning_rate": 7.401600917529504e-05,
+      "loss": 0.0574,
+      "step": 67230
+    },
+    {
+      "epoch": 4.399084069349035,
+      "grad_norm": 0.8735194802284241,
+      "learning_rate": 7.400795184131075e-05,
+      "loss": 0.0571,
+      "step": 67240
+    },
+    {
+      "epoch": 4.399738305528295,
+      "grad_norm": 0.8854332566261292,
+      "learning_rate": 7.399989369699512e-05,
+      "loss": 0.0599,
+      "step": 67250
+    },
+    {
+      "epoch": 4.400392541707556,
+      "grad_norm": 1.0150554180145264,
+      "learning_rate": 7.399183474262011e-05,
+      "loss": 0.0703,
+      "step": 67260
+    },
+    {
+      "epoch": 4.401046777886817,
+      "grad_norm": 1.0136165618896484,
+      "learning_rate": 7.398377497845779e-05,
+      "loss": 0.0587,
+      "step": 67270
+    },
+    {
+      "epoch": 4.401701014066078,
+      "grad_norm": 0.7560604214668274,
+      "learning_rate": 7.397571440478015e-05,
+      "loss": 0.0743,
+      "step": 67280
+    },
+    {
+      "epoch": 4.402355250245338,
+      "grad_norm": 0.8288975358009338,
+      "learning_rate": 7.396765302185928e-05,
+      "loss": 0.0663,
+      "step": 67290
+    },
+    {
+      "epoch": 4.403009486424599,
+      "grad_norm": 1.1244462728500366,
+      "learning_rate": 7.395959082996725e-05,
+      "loss": 0.0719,
+      "step": 67300
+    },
+    {
+      "epoch": 4.40366372260386,
+      "grad_norm": 0.8826910257339478,
+      "learning_rate": 7.39515278293762e-05,
+      "loss": 0.0693,
+      "step": 67310
+    },
+    {
+      "epoch": 4.404317958783121,
+      "grad_norm": 0.8879522681236267,
+      "learning_rate": 7.394346402035828e-05,
+      "loss": 0.0762,
+      "step": 67320
+    },
+    {
+      "epoch": 4.404972194962381,
+      "grad_norm": 0.8239995241165161,
+      "learning_rate": 7.393539940318563e-05,
+      "loss": 0.0548,
+      "step": 67330
+    },
+    {
+      "epoch": 4.405626431141642,
+      "grad_norm": 1.0602549314498901,
+      "learning_rate": 7.39273339781305e-05,
+      "loss": 0.0598,
+      "step": 67340
+    },
+    {
+      "epoch": 4.406280667320903,
+      "grad_norm": 1.0400123596191406,
+      "learning_rate": 7.391926774546509e-05,
+      "loss": 0.0723,
+      "step": 67350
+    },
+    {
+      "epoch": 4.406934903500163,
+      "grad_norm": 0.9202961325645447,
+      "learning_rate": 7.391120070546165e-05,
+      "loss": 0.0636,
+      "step": 67360
+    },
+    {
+      "epoch": 4.407589139679424,
+      "grad_norm": 0.8202659487724304,
+      "learning_rate": 7.39031328583925e-05,
+      "loss": 0.0651,
+      "step": 67370
+    },
+    {
+      "epoch": 4.408243375858685,
+      "grad_norm": 0.8003469109535217,
+      "learning_rate": 7.389506420452991e-05,
+      "loss": 0.0572,
+      "step": 67380
+    },
+    {
+      "epoch": 4.408897612037945,
+      "grad_norm": 0.8167177438735962,
+      "learning_rate": 7.388699474414624e-05,
+      "loss": 0.0746,
+      "step": 67390
+    },
+    {
+      "epoch": 4.409551848217206,
+      "grad_norm": 0.949032187461853,
+      "learning_rate": 7.387892447751387e-05,
+      "loss": 0.0654,
+      "step": 67400
+    },
+    {
+      "epoch": 4.410206084396467,
+      "grad_norm": 0.9453916549682617,
+      "learning_rate": 7.387085340490514e-05,
+      "loss": 0.0617,
+      "step": 67410
+    },
+    {
+      "epoch": 4.410860320575728,
+      "grad_norm": 1.0809295177459717,
+      "learning_rate": 7.386278152659254e-05,
+      "loss": 0.0699,
+      "step": 67420
+    },
+    {
+      "epoch": 4.411514556754988,
+      "grad_norm": 1.0627373456954956,
+      "learning_rate": 7.385470884284845e-05,
+      "loss": 0.0636,
+      "step": 67430
+    },
+    {
+      "epoch": 4.412168792934249,
+      "grad_norm": 0.8991785645484924,
+      "learning_rate": 7.384663535394541e-05,
+      "loss": 0.064,
+      "step": 67440
+    },
+    {
+      "epoch": 4.41282302911351,
+      "grad_norm": 0.9950599074363708,
+      "learning_rate": 7.383856106015585e-05,
+      "loss": 0.0611,
+      "step": 67450
+    },
+    {
+      "epoch": 4.41347726529277,
+      "grad_norm": 0.7597075700759888,
+      "learning_rate": 7.383048596175236e-05,
+      "loss": 0.0645,
+      "step": 67460
+    },
+    {
+      "epoch": 4.414131501472031,
+      "grad_norm": 0.9785023927688599,
+      "learning_rate": 7.382241005900745e-05,
+      "loss": 0.062,
+      "step": 67470
+    },
+    {
+      "epoch": 4.414785737651292,
+      "grad_norm": 0.8805351257324219,
+      "learning_rate": 7.381433335219374e-05,
+      "loss": 0.0663,
+      "step": 67480
+    },
+    {
+      "epoch": 4.415439973830553,
+      "grad_norm": 0.7586386203765869,
+      "learning_rate": 7.38062558415838e-05,
+      "loss": 0.0644,
+      "step": 67490
+    },
+    {
+      "epoch": 4.416094210009813,
+      "grad_norm": 0.8569243550300598,
+      "learning_rate": 7.379817752745033e-05,
+      "loss": 0.0602,
+      "step": 67500
+    },
+    {
+      "epoch": 4.416748446189074,
+      "grad_norm": 0.8760682344436646,
+      "learning_rate": 7.379009841006593e-05,
+      "loss": 0.0596,
+      "step": 67510
+    },
+    {
+      "epoch": 4.417402682368335,
+      "grad_norm": 0.9443554878234863,
+      "learning_rate": 7.378201848970332e-05,
+      "loss": 0.0648,
+      "step": 67520
+    },
+    {
+      "epoch": 4.418056918547595,
+      "grad_norm": 0.9215699434280396,
+      "learning_rate": 7.377393776663523e-05,
+      "loss": 0.0693,
+      "step": 67530
+    },
+    {
+      "epoch": 4.418711154726856,
+      "grad_norm": 1.1914721727371216,
+      "learning_rate": 7.376585624113437e-05,
+      "loss": 0.0702,
+      "step": 67540
+    },
+    {
+      "epoch": 4.419365390906117,
+      "grad_norm": 0.8695497512817383,
+      "learning_rate": 7.375777391347355e-05,
+      "loss": 0.0529,
+      "step": 67550
+    },
+    {
+      "epoch": 4.420019627085378,
+      "grad_norm": 0.8049673438072205,
+      "learning_rate": 7.374969078392555e-05,
+      "loss": 0.0646,
+      "step": 67560
+    },
+    {
+      "epoch": 4.420673863264638,
+      "grad_norm": 1.0911378860473633,
+      "learning_rate": 7.37416068527632e-05,
+      "loss": 0.0677,
+      "step": 67570
+    },
+    {
+      "epoch": 4.421328099443899,
+      "grad_norm": 0.8425750136375427,
+      "learning_rate": 7.373352212025935e-05,
+      "loss": 0.0604,
+      "step": 67580
+    },
+    {
+      "epoch": 4.42198233562316,
+      "grad_norm": 1.0606712102890015,
+      "learning_rate": 7.372543658668688e-05,
+      "loss": 0.0569,
+      "step": 67590
+    },
+    {
+      "epoch": 4.42263657180242,
+      "grad_norm": 0.8173533082008362,
+      "learning_rate": 7.371735025231871e-05,
+      "loss": 0.0595,
+      "step": 67600
+    },
+    {
+      "epoch": 4.423290807981681,
+      "grad_norm": 1.046075701713562,
+      "learning_rate": 7.370926311742776e-05,
+      "loss": 0.0632,
+      "step": 67610
+    },
+    {
+      "epoch": 4.423945044160942,
+      "grad_norm": 0.6317776441574097,
+      "learning_rate": 7.3701175182287e-05,
+      "loss": 0.0688,
+      "step": 67620
+    },
+    {
+      "epoch": 4.424599280340203,
+      "grad_norm": 0.8393598198890686,
+      "learning_rate": 7.369308644716944e-05,
+      "loss": 0.0633,
+      "step": 67630
+    },
+    {
+      "epoch": 4.425253516519463,
+      "grad_norm": 0.9631373286247253,
+      "learning_rate": 7.368499691234806e-05,
+      "loss": 0.0578,
+      "step": 67640
+    },
+    {
+      "epoch": 4.425907752698724,
+      "grad_norm": 0.829046905040741,
+      "learning_rate": 7.367690657809592e-05,
+      "loss": 0.0647,
+      "step": 67650
+    },
+    {
+      "epoch": 4.426561988877985,
+      "grad_norm": 0.9944875836372375,
+      "learning_rate": 7.366881544468609e-05,
+      "loss": 0.0598,
+      "step": 67660
+    },
+    {
+      "epoch": 4.427216225057245,
+      "grad_norm": 0.9320645928382874,
+      "learning_rate": 7.366072351239165e-05,
+      "loss": 0.0583,
+      "step": 67670
+    },
+    {
+      "epoch": 4.427870461236506,
+      "grad_norm": 1.0392950773239136,
+      "learning_rate": 7.365263078148575e-05,
+      "loss": 0.0623,
+      "step": 67680
+    },
+    {
+      "epoch": 4.428524697415767,
+      "grad_norm": 0.7760281562805176,
+      "learning_rate": 7.364453725224154e-05,
+      "loss": 0.0646,
+      "step": 67690
+    },
+    {
+      "epoch": 4.429178933595027,
+      "grad_norm": 0.8868238925933838,
+      "learning_rate": 7.363644292493218e-05,
+      "loss": 0.0664,
+      "step": 67700
+    },
+    {
+      "epoch": 4.429833169774288,
+      "grad_norm": 0.9134417772293091,
+      "learning_rate": 7.362834779983087e-05,
+      "loss": 0.0597,
+      "step": 67710
+    },
+    {
+      "epoch": 4.430487405953549,
+      "grad_norm": 0.7500336170196533,
+      "learning_rate": 7.362025187721086e-05,
+      "loss": 0.0573,
+      "step": 67720
+    },
+    {
+      "epoch": 4.43114164213281,
+      "grad_norm": 0.777053713798523,
+      "learning_rate": 7.361215515734541e-05,
+      "loss": 0.0696,
+      "step": 67730
+    },
+    {
+      "epoch": 4.43179587831207,
+      "grad_norm": 0.8513823747634888,
+      "learning_rate": 7.36040576405078e-05,
+      "loss": 0.0611,
+      "step": 67740
+    },
+    {
+      "epoch": 4.432450114491331,
+      "grad_norm": 0.7951275706291199,
+      "learning_rate": 7.359595932697134e-05,
+      "loss": 0.06,
+      "step": 67750
+    },
+    {
+      "epoch": 4.433104350670592,
+      "grad_norm": 0.9466954469680786,
+      "learning_rate": 7.358786021700936e-05,
+      "loss": 0.0645,
+      "step": 67760
+    },
+    {
+      "epoch": 4.433758586849853,
+      "grad_norm": 0.8108925819396973,
+      "learning_rate": 7.357976031089524e-05,
+      "loss": 0.0666,
+      "step": 67770
+    },
+    {
+      "epoch": 4.434412823029113,
+      "grad_norm": 0.8179001212120056,
+      "learning_rate": 7.357165960890237e-05,
+      "loss": 0.069,
+      "step": 67780
+    },
+    {
+      "epoch": 4.435067059208374,
+      "grad_norm": 0.8625308275222778,
+      "learning_rate": 7.356355811130419e-05,
+      "loss": 0.0601,
+      "step": 67790
+    },
+    {
+      "epoch": 4.435721295387635,
+      "grad_norm": 1.066897988319397,
+      "learning_rate": 7.35554558183741e-05,
+      "loss": 0.069,
+      "step": 67800
+    },
+    {
+      "epoch": 4.436375531566895,
+      "grad_norm": 0.8988897800445557,
+      "learning_rate": 7.35473527303856e-05,
+      "loss": 0.0809,
+      "step": 67810
+    },
+    {
+      "epoch": 4.437029767746156,
+      "grad_norm": 0.8011250495910645,
+      "learning_rate": 7.35392488476122e-05,
+      "loss": 0.0633,
+      "step": 67820
+    },
+    {
+      "epoch": 4.437684003925417,
+      "grad_norm": 0.9169591665267944,
+      "learning_rate": 7.353114417032742e-05,
+      "loss": 0.057,
+      "step": 67830
+    },
+    {
+      "epoch": 4.438338240104677,
+      "grad_norm": 0.9640666246414185,
+      "learning_rate": 7.35230386988048e-05,
+      "loss": 0.0624,
+      "step": 67840
+    },
+    {
+      "epoch": 4.438992476283938,
+      "grad_norm": 0.7338112592697144,
+      "learning_rate": 7.351493243331794e-05,
+      "loss": 0.0615,
+      "step": 67850
+    },
+    {
+      "epoch": 4.439646712463199,
+      "grad_norm": 0.8081340789794922,
+      "learning_rate": 7.350682537414044e-05,
+      "loss": 0.0662,
+      "step": 67860
+    },
+    {
+      "epoch": 4.44030094864246,
+      "grad_norm": 1.210657000541687,
+      "learning_rate": 7.349871752154593e-05,
+      "loss": 0.0784,
+      "step": 67870
+    },
+    {
+      "epoch": 4.44095518482172,
+      "grad_norm": 0.9144163727760315,
+      "learning_rate": 7.349060887580808e-05,
+      "loss": 0.0635,
+      "step": 67880
+    },
+    {
+      "epoch": 4.441609421000981,
+      "grad_norm": 0.9125617742538452,
+      "learning_rate": 7.348249943720058e-05,
+      "loss": 0.0627,
+      "step": 67890
+    },
+    {
+      "epoch": 4.442263657180242,
+      "grad_norm": 0.8639421463012695,
+      "learning_rate": 7.347438920599712e-05,
+      "loss": 0.0649,
+      "step": 67900
+    },
+    {
+      "epoch": 4.442917893359502,
+      "grad_norm": 0.795691728591919,
+      "learning_rate": 7.346627818247149e-05,
+      "loss": 0.0674,
+      "step": 67910
+    },
+    {
+      "epoch": 4.443572129538763,
+      "grad_norm": 0.9141677618026733,
+      "learning_rate": 7.34581663668974e-05,
+      "loss": 0.0657,
+      "step": 67920
+    },
+    {
+      "epoch": 4.444226365718024,
+      "grad_norm": 0.9653276205062866,
+      "learning_rate": 7.345005375954869e-05,
+      "loss": 0.0506,
+      "step": 67930
+    },
+    {
+      "epoch": 4.444880601897285,
+      "grad_norm": 1.0340138673782349,
+      "learning_rate": 7.344194036069916e-05,
+      "loss": 0.0768,
+      "step": 67940
+    },
+    {
+      "epoch": 4.445534838076545,
+      "grad_norm": 0.7893163561820984,
+      "learning_rate": 7.343382617062266e-05,
+      "loss": 0.0591,
+      "step": 67950
+    },
+    {
+      "epoch": 4.446189074255806,
+      "grad_norm": 0.840461015701294,
+      "learning_rate": 7.342571118959307e-05,
+      "loss": 0.059,
+      "step": 67960
+    },
+    {
+      "epoch": 4.446843310435067,
+      "grad_norm": 1.0385587215423584,
+      "learning_rate": 7.34175954178843e-05,
+      "loss": 0.0784,
+      "step": 67970
+    },
+    {
+      "epoch": 4.447497546614327,
+      "grad_norm": 0.9671325087547302,
+      "learning_rate": 7.340947885577028e-05,
+      "loss": 0.0583,
+      "step": 67980
+    },
+    {
+      "epoch": 4.448151782793588,
+      "grad_norm": 0.8433911800384521,
+      "learning_rate": 7.340136150352492e-05,
+      "loss": 0.0611,
+      "step": 67990
+    },
+    {
+      "epoch": 4.448806018972849,
+      "grad_norm": 0.8698466420173645,
+      "learning_rate": 7.339324336142226e-05,
+      "loss": 0.0668,
+      "step": 68000
+    },
+    {
+      "epoch": 4.44946025515211,
+      "grad_norm": 0.8934153318405151,
+      "learning_rate": 7.338512442973628e-05,
+      "loss": 0.0628,
+      "step": 68010
+    },
+    {
+      "epoch": 4.45011449133137,
+      "grad_norm": 0.9508333206176758,
+      "learning_rate": 7.337700470874103e-05,
+      "loss": 0.0582,
+      "step": 68020
+    },
+    {
+      "epoch": 4.450768727510631,
+      "grad_norm": 0.8118358254432678,
+      "learning_rate": 7.336888419871055e-05,
+      "loss": 0.0571,
+      "step": 68030
+    },
+    {
+      "epoch": 4.451422963689892,
+      "grad_norm": 0.7559012174606323,
+      "learning_rate": 7.336076289991895e-05,
+      "loss": 0.0645,
+      "step": 68040
+    },
+    {
+      "epoch": 4.452077199869152,
+      "grad_norm": 1.2024433612823486,
+      "learning_rate": 7.335264081264035e-05,
+      "loss": 0.0652,
+      "step": 68050
+    },
+    {
+      "epoch": 4.452731436048413,
+      "grad_norm": 0.8405249714851379,
+      "learning_rate": 7.334451793714885e-05,
+      "loss": 0.0679,
+      "step": 68060
+    },
+    {
+      "epoch": 4.453385672227674,
+      "grad_norm": 0.9205291271209717,
+      "learning_rate": 7.333639427371866e-05,
+      "loss": 0.0597,
+      "step": 68070
+    },
+    {
+      "epoch": 4.454039908406935,
+      "grad_norm": 1.0684325695037842,
+      "learning_rate": 7.332826982262395e-05,
+      "loss": 0.0664,
+      "step": 68080
+    },
+    {
+      "epoch": 4.454694144586195,
+      "grad_norm": 0.9805346727371216,
+      "learning_rate": 7.332014458413897e-05,
+      "loss": 0.0621,
+      "step": 68090
+    },
+    {
+      "epoch": 4.455348380765456,
+      "grad_norm": 0.7872211337089539,
+      "learning_rate": 7.331201855853794e-05,
+      "loss": 0.073,
+      "step": 68100
+    },
+    {
+      "epoch": 4.456002616944717,
+      "grad_norm": 0.7325934767723083,
+      "learning_rate": 7.330389174609515e-05,
+      "loss": 0.0638,
+      "step": 68110
+    },
+    {
+      "epoch": 4.456656853123977,
+      "grad_norm": 0.7671561241149902,
+      "learning_rate": 7.32957641470849e-05,
+      "loss": 0.0569,
+      "step": 68120
+    },
+    {
+      "epoch": 4.457311089303238,
+      "grad_norm": 0.9188522100448608,
+      "learning_rate": 7.328763576178151e-05,
+      "loss": 0.0645,
+      "step": 68130
+    },
+    {
+      "epoch": 4.457965325482499,
+      "grad_norm": 1.0198373794555664,
+      "learning_rate": 7.327950659045935e-05,
+      "loss": 0.0603,
+      "step": 68140
+    },
+    {
+      "epoch": 4.458619561661759,
+      "grad_norm": 0.9594274759292603,
+      "learning_rate": 7.327137663339276e-05,
+      "loss": 0.0574,
+      "step": 68150
+    },
+    {
+      "epoch": 4.45927379784102,
+      "grad_norm": 1.0565496683120728,
+      "learning_rate": 7.32632458908562e-05,
+      "loss": 0.0609,
+      "step": 68160
+    },
+    {
+      "epoch": 4.459928034020281,
+      "grad_norm": 0.9333295226097107,
+      "learning_rate": 7.325511436312408e-05,
+      "loss": 0.0769,
+      "step": 68170
+    },
+    {
+      "epoch": 4.460582270199542,
+      "grad_norm": 0.9997411966323853,
+      "learning_rate": 7.324698205047087e-05,
+      "loss": 0.0607,
+      "step": 68180
+    },
+    {
+      "epoch": 4.461236506378802,
+      "grad_norm": 1.1432136297225952,
+      "learning_rate": 7.323884895317102e-05,
+      "loss": 0.0573,
+      "step": 68190
+    },
+    {
+      "epoch": 4.461890742558063,
+      "grad_norm": 0.8411798477172852,
+      "learning_rate": 7.32307150714991e-05,
+      "loss": 0.0577,
+      "step": 68200
+    },
+    {
+      "epoch": 4.462544978737324,
+      "grad_norm": 0.8566927313804626,
+      "learning_rate": 7.32225804057296e-05,
+      "loss": 0.0617,
+      "step": 68210
+    },
+    {
+      "epoch": 4.463199214916585,
+      "grad_norm": 0.9064366221427917,
+      "learning_rate": 7.321444495613712e-05,
+      "loss": 0.0702,
+      "step": 68220
+    },
+    {
+      "epoch": 4.463853451095845,
+      "grad_norm": 0.6811468005180359,
+      "learning_rate": 7.320630872299624e-05,
+      "loss": 0.0588,
+      "step": 68230
+    },
+    {
+      "epoch": 4.464507687275106,
+      "grad_norm": 0.8559892177581787,
+      "learning_rate": 7.319817170658158e-05,
+      "loss": 0.0576,
+      "step": 68240
+    },
+    {
+      "epoch": 4.465161923454367,
+      "grad_norm": 0.7884718775749207,
+      "learning_rate": 7.319003390716779e-05,
+      "loss": 0.0656,
+      "step": 68250
+    },
+    {
+      "epoch": 4.465816159633627,
+      "grad_norm": 0.9382063150405884,
+      "learning_rate": 7.318189532502953e-05,
+      "loss": 0.0734,
+      "step": 68260
+    },
+    {
+      "epoch": 4.466470395812888,
+      "grad_norm": 0.9593857526779175,
+      "learning_rate": 7.317375596044152e-05,
+      "loss": 0.0686,
+      "step": 68270
+    },
+    {
+      "epoch": 4.467124631992149,
+      "grad_norm": 0.887768030166626,
+      "learning_rate": 7.316561581367845e-05,
+      "loss": 0.0603,
+      "step": 68280
+    },
+    {
+      "epoch": 4.4677788681714095,
+      "grad_norm": 0.8870118260383606,
+      "learning_rate": 7.315747488501509e-05,
+      "loss": 0.0611,
+      "step": 68290
+    },
+    {
+      "epoch": 4.46843310435067,
+      "grad_norm": 0.7766203284263611,
+      "learning_rate": 7.314933317472624e-05,
+      "loss": 0.0577,
+      "step": 68300
+    },
+    {
+      "epoch": 4.469087340529931,
+      "grad_norm": 0.6949660181999207,
+      "learning_rate": 7.314119068308668e-05,
+      "loss": 0.0602,
+      "step": 68310
+    },
+    {
+      "epoch": 4.469741576709192,
+      "grad_norm": 0.8689270615577698,
+      "learning_rate": 7.313304741037124e-05,
+      "loss": 0.0587,
+      "step": 68320
+    },
+    {
+      "epoch": 4.4703958128884524,
+      "grad_norm": 0.9906384348869324,
+      "learning_rate": 7.312490335685477e-05,
+      "loss": 0.0614,
+      "step": 68330
+    },
+    {
+      "epoch": 4.471050049067713,
+      "grad_norm": 0.8280085325241089,
+      "learning_rate": 7.311675852281218e-05,
+      "loss": 0.0673,
+      "step": 68340
+    },
+    {
+      "epoch": 4.471704285246974,
+      "grad_norm": 0.6465350985527039,
+      "learning_rate": 7.310861290851836e-05,
+      "loss": 0.0626,
+      "step": 68350
+    },
+    {
+      "epoch": 4.4723585214262345,
+      "grad_norm": 0.8401268720626831,
+      "learning_rate": 7.310046651424824e-05,
+      "loss": 0.0602,
+      "step": 68360
+    },
+    {
+      "epoch": 4.473012757605495,
+      "grad_norm": 0.9460334777832031,
+      "learning_rate": 7.309231934027681e-05,
+      "loss": 0.0631,
+      "step": 68370
+    },
+    {
+      "epoch": 4.473666993784756,
+      "grad_norm": 0.826092004776001,
+      "learning_rate": 7.308417138687902e-05,
+      "loss": 0.0529,
+      "step": 68380
+    },
+    {
+      "epoch": 4.474321229964017,
+      "grad_norm": 0.9380882978439331,
+      "learning_rate": 7.307602265432993e-05,
+      "loss": 0.0649,
+      "step": 68390
+    },
+    {
+      "epoch": 4.4749754661432775,
+      "grad_norm": 0.720436155796051,
+      "learning_rate": 7.306787314290455e-05,
+      "loss": 0.0636,
+      "step": 68400
+    },
+    {
+      "epoch": 4.475629702322538,
+      "grad_norm": 0.8422783017158508,
+      "learning_rate": 7.305972285287793e-05,
+      "loss": 0.0531,
+      "step": 68410
+    },
+    {
+      "epoch": 4.476283938501799,
+      "grad_norm": 0.8465076684951782,
+      "learning_rate": 7.30515717845252e-05,
+      "loss": 0.0689,
+      "step": 68420
+    },
+    {
+      "epoch": 4.4769381746810595,
+      "grad_norm": 0.7343236804008484,
+      "learning_rate": 7.304341993812149e-05,
+      "loss": 0.0649,
+      "step": 68430
+    },
+    {
+      "epoch": 4.4775924108603204,
+      "grad_norm": 1.0546813011169434,
+      "learning_rate": 7.30352673139419e-05,
+      "loss": 0.0678,
+      "step": 68440
+    },
+    {
+      "epoch": 4.478246647039581,
+      "grad_norm": 0.815166711807251,
+      "learning_rate": 7.302711391226163e-05,
+      "loss": 0.0632,
+      "step": 68450
+    },
+    {
+      "epoch": 4.478900883218842,
+      "grad_norm": 0.7304500341415405,
+      "learning_rate": 7.301895973335587e-05,
+      "loss": 0.0643,
+      "step": 68460
+    },
+    {
+      "epoch": 4.4795551193981025,
+      "grad_norm": 0.8881242871284485,
+      "learning_rate": 7.301080477749987e-05,
+      "loss": 0.0757,
+      "step": 68470
+    },
+    {
+      "epoch": 4.480209355577363,
+      "grad_norm": 0.8304605484008789,
+      "learning_rate": 7.300264904496883e-05,
+      "loss": 0.0634,
+      "step": 68480
+    },
+    {
+      "epoch": 4.480863591756624,
+      "grad_norm": 0.691174328327179,
+      "learning_rate": 7.299449253603808e-05,
+      "loss": 0.0649,
+      "step": 68490
+    },
+    {
+      "epoch": 4.4815178279358845,
+      "grad_norm": 0.9316902756690979,
+      "learning_rate": 7.29863352509829e-05,
+      "loss": 0.0646,
+      "step": 68500
+    },
+    {
+      "epoch": 4.4821720641151455,
+      "grad_norm": 0.9124321937561035,
+      "learning_rate": 7.297817719007861e-05,
+      "loss": 0.067,
+      "step": 68510
+    },
+    {
+      "epoch": 4.482826300294406,
+      "grad_norm": 1.0439180135726929,
+      "learning_rate": 7.297001835360058e-05,
+      "loss": 0.0721,
+      "step": 68520
+    },
+    {
+      "epoch": 4.483480536473667,
+      "grad_norm": 0.8428361415863037,
+      "learning_rate": 7.296185874182421e-05,
+      "loss": 0.0645,
+      "step": 68530
+    },
+    {
+      "epoch": 4.4841347726529275,
+      "grad_norm": 0.8357476592063904,
+      "learning_rate": 7.295369835502485e-05,
+      "loss": 0.0679,
+      "step": 68540
+    },
+    {
+      "epoch": 4.4847890088321885,
+      "grad_norm": 0.942311704158783,
+      "learning_rate": 7.2945537193478e-05,
+      "loss": 0.0604,
+      "step": 68550
+    },
+    {
+      "epoch": 4.485443245011449,
+      "grad_norm": 0.8047347068786621,
+      "learning_rate": 7.293737525745908e-05,
+      "loss": 0.0733,
+      "step": 68560
+    },
+    {
+      "epoch": 4.4860974811907095,
+      "grad_norm": 0.9718091487884521,
+      "learning_rate": 7.29292125472436e-05,
+      "loss": 0.0591,
+      "step": 68570
+    },
+    {
+      "epoch": 4.4867517173699705,
+      "grad_norm": 0.843663215637207,
+      "learning_rate": 7.292104906310707e-05,
+      "loss": 0.0692,
+      "step": 68580
+    },
+    {
+      "epoch": 4.487405953549231,
+      "grad_norm": 0.8390454649925232,
+      "learning_rate": 7.2912884805325e-05,
+      "loss": 0.0613,
+      "step": 68590
+    },
+    {
+      "epoch": 4.4880601897284915,
+      "grad_norm": 0.7371008396148682,
+      "learning_rate": 7.2904719774173e-05,
+      "loss": 0.057,
+      "step": 68600
+    },
+    {
+      "epoch": 4.4887144259077525,
+      "grad_norm": 1.1576838493347168,
+      "learning_rate": 7.289655396992661e-05,
+      "loss": 0.0655,
+      "step": 68610
+    },
+    {
+      "epoch": 4.4893686620870135,
+      "grad_norm": 1.0340911149978638,
+      "learning_rate": 7.28883873928615e-05,
+      "loss": 0.0626,
+      "step": 68620
+    },
+    {
+      "epoch": 4.490022898266274,
+      "grad_norm": 0.9780303835868835,
+      "learning_rate": 7.288022004325327e-05,
+      "loss": 0.0646,
+      "step": 68630
+    },
+    {
+      "epoch": 4.4906771344455345,
+      "grad_norm": 1.1134583950042725,
+      "learning_rate": 7.287205192137763e-05,
+      "loss": 0.0665,
+      "step": 68640
+    },
+    {
+      "epoch": 4.4913313706247955,
+      "grad_norm": 0.9058104157447815,
+      "learning_rate": 7.286388302751023e-05,
+      "loss": 0.0627,
+      "step": 68650
+    },
+    {
+      "epoch": 4.4919856068040565,
+      "grad_norm": 0.9129322171211243,
+      "learning_rate": 7.285571336192683e-05,
+      "loss": 0.0526,
+      "step": 68660
+    },
+    {
+      "epoch": 4.492639842983317,
+      "grad_norm": 0.7618893384933472,
+      "learning_rate": 7.284754292490314e-05,
+      "loss": 0.0584,
+      "step": 68670
+    },
+    {
+      "epoch": 4.4932940791625775,
+      "grad_norm": 0.9606322050094604,
+      "learning_rate": 7.283937171671498e-05,
+      "loss": 0.0634,
+      "step": 68680
+    },
+    {
+      "epoch": 4.4939483153418385,
+      "grad_norm": 0.9516724348068237,
+      "learning_rate": 7.283119973763813e-05,
+      "loss": 0.0665,
+      "step": 68690
+    },
+    {
+      "epoch": 4.494602551521099,
+      "grad_norm": 0.8869428038597107,
+      "learning_rate": 7.282302698794838e-05,
+      "loss": 0.0613,
+      "step": 68700
+    },
+    {
+      "epoch": 4.4952567877003595,
+      "grad_norm": 0.7321935892105103,
+      "learning_rate": 7.281485346792165e-05,
+      "loss": 0.0611,
+      "step": 68710
+    },
+    {
+      "epoch": 4.4959110238796205,
+      "grad_norm": 0.9069811701774597,
+      "learning_rate": 7.280667917783376e-05,
+      "loss": 0.0624,
+      "step": 68720
+    },
+    {
+      "epoch": 4.4965652600588815,
+      "grad_norm": 0.7614008188247681,
+      "learning_rate": 7.279850411796065e-05,
+      "loss": 0.0561,
+      "step": 68730
+    },
+    {
+      "epoch": 4.4972194962381415,
+      "grad_norm": 1.1257866621017456,
+      "learning_rate": 7.279032828857822e-05,
+      "loss": 0.0602,
+      "step": 68740
+    },
+    {
+      "epoch": 4.4978737324174025,
+      "grad_norm": 0.8876655101776123,
+      "learning_rate": 7.278215168996245e-05,
+      "loss": 0.0628,
+      "step": 68750
+    },
+    {
+      "epoch": 4.4985279685966635,
+      "grad_norm": 1.048601746559143,
+      "learning_rate": 7.27739743223893e-05,
+      "loss": 0.0645,
+      "step": 68760
+    },
+    {
+      "epoch": 4.4991822047759245,
+      "grad_norm": 0.7537106871604919,
+      "learning_rate": 7.27657961861348e-05,
+      "loss": 0.0627,
+      "step": 68770
+    },
+    {
+      "epoch": 4.4998364409551845,
+      "grad_norm": 0.9030201435089111,
+      "learning_rate": 7.275761728147497e-05,
+      "loss": 0.0707,
+      "step": 68780
+    },
+    {
+      "epoch": 4.5004906771344455,
+      "grad_norm": 0.799105703830719,
+      "learning_rate": 7.274943760868589e-05,
+      "loss": 0.0571,
+      "step": 68790
+    },
+    {
+      "epoch": 4.5011449133137065,
+      "grad_norm": 0.8551787734031677,
+      "learning_rate": 7.27412571680436e-05,
+      "loss": 0.0625,
+      "step": 68800
+    },
+    {
+      "epoch": 4.5017991494929674,
+      "grad_norm": 0.8719650506973267,
+      "learning_rate": 7.273307595982424e-05,
+      "loss": 0.0596,
+      "step": 68810
+    },
+    {
+      "epoch": 4.5024533856722275,
+      "grad_norm": 0.7927690148353577,
+      "learning_rate": 7.272489398430397e-05,
+      "loss": 0.0626,
+      "step": 68820
+    },
+    {
+      "epoch": 4.5031076218514885,
+      "grad_norm": 0.9607967138290405,
+      "learning_rate": 7.271671124175893e-05,
+      "loss": 0.0632,
+      "step": 68830
+    },
+    {
+      "epoch": 4.5037618580307495,
+      "grad_norm": 0.8309448957443237,
+      "learning_rate": 7.270852773246528e-05,
+      "loss": 0.0614,
+      "step": 68840
+    },
+    {
+      "epoch": 4.5044160942100095,
+      "grad_norm": 0.9089359045028687,
+      "learning_rate": 7.27003434566993e-05,
+      "loss": 0.0693,
+      "step": 68850
+    },
+    {
+      "epoch": 4.5050703303892705,
+      "grad_norm": 0.9338735342025757,
+      "learning_rate": 7.269215841473717e-05,
+      "loss": 0.0612,
+      "step": 68860
+    },
+    {
+      "epoch": 4.5057245665685315,
+      "grad_norm": 0.8226944804191589,
+      "learning_rate": 7.268397260685518e-05,
+      "loss": 0.068,
+      "step": 68870
+    },
+    {
+      "epoch": 4.506378802747792,
+      "grad_norm": 0.8187761902809143,
+      "learning_rate": 7.267578603332963e-05,
+      "loss": 0.0563,
+      "step": 68880
+    },
+    {
+      "epoch": 4.5070330389270525,
+      "grad_norm": 0.8850693702697754,
+      "learning_rate": 7.266759869443683e-05,
+      "loss": 0.0683,
+      "step": 68890
+    },
+    {
+      "epoch": 4.5076872751063135,
+      "grad_norm": 0.8132352232933044,
+      "learning_rate": 7.265941059045314e-05,
+      "loss": 0.0595,
+      "step": 68900
+    },
+    {
+      "epoch": 4.5083415112855745,
+      "grad_norm": 0.8258399367332458,
+      "learning_rate": 7.265122172165489e-05,
+      "loss": 0.0637,
+      "step": 68910
+    },
+    {
+      "epoch": 4.508995747464835,
+      "grad_norm": 0.7450704574584961,
+      "learning_rate": 7.264303208831854e-05,
+      "loss": 0.0644,
+      "step": 68920
+    },
+    {
+      "epoch": 4.5096499836440955,
+      "grad_norm": 0.7500464916229248,
+      "learning_rate": 7.263484169072044e-05,
+      "loss": 0.0666,
+      "step": 68930
+    },
+    {
+      "epoch": 4.5103042198233565,
+      "grad_norm": 0.8899851441383362,
+      "learning_rate": 7.262665052913707e-05,
+      "loss": 0.0559,
+      "step": 68940
+    },
+    {
+      "epoch": 4.510958456002617,
+      "grad_norm": 0.9518905282020569,
+      "learning_rate": 7.261845860384492e-05,
+      "loss": 0.0637,
+      "step": 68950
+    },
+    {
+      "epoch": 4.5116126921818775,
+      "grad_norm": 1.0154138803482056,
+      "learning_rate": 7.261026591512047e-05,
+      "loss": 0.0598,
+      "step": 68960
+    },
+    {
+      "epoch": 4.5122669283611385,
+      "grad_norm": 0.7621309757232666,
+      "learning_rate": 7.260207246324024e-05,
+      "loss": 0.0729,
+      "step": 68970
+    },
+    {
+      "epoch": 4.5129211645403995,
+      "grad_norm": 0.8984786868095398,
+      "learning_rate": 7.25938782484808e-05,
+      "loss": 0.0629,
+      "step": 68980
+    },
+    {
+      "epoch": 4.51357540071966,
+      "grad_norm": 0.9162548780441284,
+      "learning_rate": 7.25856832711187e-05,
+      "loss": 0.0639,
+      "step": 68990
+    },
+    {
+      "epoch": 4.5142296368989205,
+      "grad_norm": 0.844033420085907,
+      "learning_rate": 7.257748753143057e-05,
+      "loss": 0.0614,
+      "step": 69000
+    },
+    {
+      "epoch": 4.5148838730781815,
+      "grad_norm": 0.8666905760765076,
+      "learning_rate": 7.256929102969302e-05,
+      "loss": 0.0638,
+      "step": 69010
+    },
+    {
+      "epoch": 4.515538109257442,
+      "grad_norm": 0.9945257902145386,
+      "learning_rate": 7.256109376618271e-05,
+      "loss": 0.0602,
+      "step": 69020
+    },
+    {
+      "epoch": 4.516192345436703,
+      "grad_norm": 1.0383166074752808,
+      "learning_rate": 7.25528957411763e-05,
+      "loss": 0.0587,
+      "step": 69030
+    },
+    {
+      "epoch": 4.5168465816159635,
+      "grad_norm": 0.9124101400375366,
+      "learning_rate": 7.254469695495054e-05,
+      "loss": 0.0667,
+      "step": 69040
+    },
+    {
+      "epoch": 4.517500817795224,
+      "grad_norm": 0.7469764947891235,
+      "learning_rate": 7.253649740778212e-05,
+      "loss": 0.0641,
+      "step": 69050
+    },
+    {
+      "epoch": 4.518155053974485,
+      "grad_norm": 0.8720923662185669,
+      "learning_rate": 7.25282970999478e-05,
+      "loss": 0.0733,
+      "step": 69060
+    },
+    {
+      "epoch": 4.5188092901537456,
+      "grad_norm": 0.8793687224388123,
+      "learning_rate": 7.252009603172436e-05,
+      "loss": 0.0716,
+      "step": 69070
+    },
+    {
+      "epoch": 4.5194635263330065,
+      "grad_norm": 0.8005584478378296,
+      "learning_rate": 7.251189420338865e-05,
+      "loss": 0.0605,
+      "step": 69080
+    },
+    {
+      "epoch": 4.520117762512267,
+      "grad_norm": 0.868381142616272,
+      "learning_rate": 7.250369161521746e-05,
+      "loss": 0.0644,
+      "step": 69090
+    },
+    {
+      "epoch": 4.520771998691528,
+      "grad_norm": 0.8739510774612427,
+      "learning_rate": 7.249548826748764e-05,
+      "loss": 0.067,
+      "step": 69100
+    },
+    {
+      "epoch": 4.5214262348707885,
+      "grad_norm": 0.9072023034095764,
+      "learning_rate": 7.248728416047611e-05,
+      "loss": 0.0612,
+      "step": 69110
+    },
+    {
+      "epoch": 4.5220804710500495,
+      "grad_norm": 0.9014410972595215,
+      "learning_rate": 7.247907929445976e-05,
+      "loss": 0.0686,
+      "step": 69120
+    },
+    {
+      "epoch": 4.52273470722931,
+      "grad_norm": 0.883929967880249,
+      "learning_rate": 7.247087366971554e-05,
+      "loss": 0.054,
+      "step": 69130
+    },
+    {
+      "epoch": 4.523388943408571,
+      "grad_norm": 1.133589744567871,
+      "learning_rate": 7.24626672865204e-05,
+      "loss": 0.0698,
+      "step": 69140
+    },
+    {
+      "epoch": 4.5240431795878315,
+      "grad_norm": 0.7994188666343689,
+      "learning_rate": 7.245446014515132e-05,
+      "loss": 0.0606,
+      "step": 69150
+    },
+    {
+      "epoch": 4.524697415767092,
+      "grad_norm": 1.077908992767334,
+      "learning_rate": 7.244625224588533e-05,
+      "loss": 0.0638,
+      "step": 69160
+    },
+    {
+      "epoch": 4.525351651946353,
+      "grad_norm": 0.7507327795028687,
+      "learning_rate": 7.243804358899943e-05,
+      "loss": 0.0579,
+      "step": 69170
+    },
+    {
+      "epoch": 4.5260058881256136,
+      "grad_norm": 0.8101299405097961,
+      "learning_rate": 7.242983417477076e-05,
+      "loss": 0.0637,
+      "step": 69180
+    },
+    {
+      "epoch": 4.526660124304874,
+      "grad_norm": 0.8456845879554749,
+      "learning_rate": 7.242162400347634e-05,
+      "loss": 0.0586,
+      "step": 69190
+    },
+    {
+      "epoch": 4.527314360484135,
+      "grad_norm": 0.812188982963562,
+      "learning_rate": 7.24134130753933e-05,
+      "loss": 0.0627,
+      "step": 69200
+    },
+    {
+      "epoch": 4.527968596663396,
+      "grad_norm": 1.0019590854644775,
+      "learning_rate": 7.24052013907988e-05,
+      "loss": 0.0598,
+      "step": 69210
+    },
+    {
+      "epoch": 4.5286228328426565,
+      "grad_norm": 1.0545828342437744,
+      "learning_rate": 7.239698894996997e-05,
+      "loss": 0.062,
+      "step": 69220
+    },
+    {
+      "epoch": 4.529277069021917,
+      "grad_norm": 0.7775808572769165,
+      "learning_rate": 7.238877575318405e-05,
+      "loss": 0.0578,
+      "step": 69230
+    },
+    {
+      "epoch": 4.529931305201178,
+      "grad_norm": 0.9661263227462769,
+      "learning_rate": 7.238056180071823e-05,
+      "loss": 0.0685,
+      "step": 69240
+    },
+    {
+      "epoch": 4.530585541380439,
+      "grad_norm": 0.8217857480049133,
+      "learning_rate": 7.237234709284975e-05,
+      "loss": 0.0544,
+      "step": 69250
+    },
+    {
+      "epoch": 4.5312397775596995,
+      "grad_norm": 0.7686980962753296,
+      "learning_rate": 7.236413162985587e-05,
+      "loss": 0.06,
+      "step": 69260
+    },
+    {
+      "epoch": 4.53189401373896,
+      "grad_norm": 0.7001199722290039,
+      "learning_rate": 7.235591541201391e-05,
+      "loss": 0.0599,
+      "step": 69270
+    },
+    {
+      "epoch": 4.532548249918221,
+      "grad_norm": 0.6666156649589539,
+      "learning_rate": 7.234769843960116e-05,
+      "loss": 0.0668,
+      "step": 69280
+    },
+    {
+      "epoch": 4.5332024860974816,
+      "grad_norm": 0.9138648509979248,
+      "learning_rate": 7.233948071289499e-05,
+      "loss": 0.06,
+      "step": 69290
+    },
+    {
+      "epoch": 4.533856722276742,
+      "grad_norm": 1.0828460454940796,
+      "learning_rate": 7.233126223217275e-05,
+      "loss": 0.0742,
+      "step": 69300
+    },
+    {
+      "epoch": 4.534510958456003,
+      "grad_norm": 0.9059333801269531,
+      "learning_rate": 7.232304299771187e-05,
+      "loss": 0.0619,
+      "step": 69310
+    },
+    {
+      "epoch": 4.535165194635264,
+      "grad_norm": 0.7670040130615234,
+      "learning_rate": 7.231482300978971e-05,
+      "loss": 0.0635,
+      "step": 69320
+    },
+    {
+      "epoch": 4.535819430814524,
+      "grad_norm": 0.7467994689941406,
+      "learning_rate": 7.230660226868376e-05,
+      "loss": 0.0586,
+      "step": 69330
+    },
+    {
+      "epoch": 4.536473666993785,
+      "grad_norm": 0.8463179469108582,
+      "learning_rate": 7.22983807746715e-05,
+      "loss": 0.0564,
+      "step": 69340
+    },
+    {
+      "epoch": 4.537127903173046,
+      "grad_norm": 0.6910958290100098,
+      "learning_rate": 7.22901585280304e-05,
+      "loss": 0.0637,
+      "step": 69350
+    },
+    {
+      "epoch": 4.537782139352307,
+      "grad_norm": 0.78616863489151,
+      "learning_rate": 7.228193552903798e-05,
+      "loss": 0.059,
+      "step": 69360
+    },
+    {
+      "epoch": 4.538436375531567,
+      "grad_norm": 0.934597373008728,
+      "learning_rate": 7.227371177797181e-05,
+      "loss": 0.0633,
+      "step": 69370
+    },
+    {
+      "epoch": 4.539090611710828,
+      "grad_norm": 0.9315169453620911,
+      "learning_rate": 7.226548727510945e-05,
+      "loss": 0.0573,
+      "step": 69380
+    },
+    {
+      "epoch": 4.539744847890089,
+      "grad_norm": 1.0546493530273438,
+      "learning_rate": 7.22572620207285e-05,
+      "loss": 0.069,
+      "step": 69390
+    },
+    {
+      "epoch": 4.540399084069349,
+      "grad_norm": 0.7650620937347412,
+      "learning_rate": 7.224903601510658e-05,
+      "loss": 0.0564,
+      "step": 69400
+    },
+    {
+      "epoch": 4.54105332024861,
+      "grad_norm": 0.7123848795890808,
+      "learning_rate": 7.224080925852136e-05,
+      "loss": 0.065,
+      "step": 69410
+    },
+    {
+      "epoch": 4.541707556427871,
+      "grad_norm": 0.8295034766197205,
+      "learning_rate": 7.22325817512505e-05,
+      "loss": 0.0546,
+      "step": 69420
+    },
+    {
+      "epoch": 4.542361792607132,
+      "grad_norm": 0.9228529930114746,
+      "learning_rate": 7.222435349357169e-05,
+      "loss": 0.0676,
+      "step": 69430
+    },
+    {
+      "epoch": 4.543016028786392,
+      "grad_norm": 0.7449051141738892,
+      "learning_rate": 7.221612448576266e-05,
+      "loss": 0.0575,
+      "step": 69440
+    },
+    {
+      "epoch": 4.543670264965653,
+      "grad_norm": 0.832432210445404,
+      "learning_rate": 7.220789472810115e-05,
+      "loss": 0.0595,
+      "step": 69450
+    },
+    {
+      "epoch": 4.544324501144914,
+      "grad_norm": 1.0171115398406982,
+      "learning_rate": 7.219966422086497e-05,
+      "loss": 0.0561,
+      "step": 69460
+    },
+    {
+      "epoch": 4.544978737324174,
+      "grad_norm": 0.8047521710395813,
+      "learning_rate": 7.219143296433191e-05,
+      "loss": 0.0636,
+      "step": 69470
+    },
+    {
+      "epoch": 4.545632973503435,
+      "grad_norm": 0.7108994126319885,
+      "learning_rate": 7.218320095877976e-05,
+      "loss": 0.0626,
+      "step": 69480
+    },
+    {
+      "epoch": 4.546287209682696,
+      "grad_norm": 0.9195765256881714,
+      "learning_rate": 7.217496820448642e-05,
+      "loss": 0.0578,
+      "step": 69490
+    },
+    {
+      "epoch": 4.546941445861956,
+      "grad_norm": 1.0596858263015747,
+      "learning_rate": 7.216673470172975e-05,
+      "loss": 0.0603,
+      "step": 69500
+    },
+    {
+      "epoch": 4.547595682041217,
+      "grad_norm": 0.7993507385253906,
+      "learning_rate": 7.215850045078765e-05,
+      "loss": 0.0695,
+      "step": 69510
+    },
+    {
+      "epoch": 4.548249918220478,
+      "grad_norm": 1.125396728515625,
+      "learning_rate": 7.215026545193802e-05,
+      "loss": 0.0666,
+      "step": 69520
+    },
+    {
+      "epoch": 4.548904154399739,
+      "grad_norm": 0.8783840537071228,
+      "learning_rate": 7.214202970545888e-05,
+      "loss": 0.0567,
+      "step": 69530
+    },
+    {
+      "epoch": 4.549558390578999,
+      "grad_norm": 0.7901387214660645,
+      "learning_rate": 7.213379321162814e-05,
+      "loss": 0.0585,
+      "step": 69540
+    },
+    {
+      "epoch": 4.55021262675826,
+      "grad_norm": 0.8331219553947449,
+      "learning_rate": 7.212555597072384e-05,
+      "loss": 0.0596,
+      "step": 69550
+    },
+    {
+      "epoch": 4.550866862937521,
+      "grad_norm": 0.8178802132606506,
+      "learning_rate": 7.2117317983024e-05,
+      "loss": 0.0582,
+      "step": 69560
+    },
+    {
+      "epoch": 4.551521099116782,
+      "grad_norm": 0.7966588139533997,
+      "learning_rate": 7.210907924880668e-05,
+      "loss": 0.0586,
+      "step": 69570
+    },
+    {
+      "epoch": 4.552175335296042,
+      "grad_norm": 0.7240191698074341,
+      "learning_rate": 7.210083976834994e-05,
+      "loss": 0.0554,
+      "step": 69580
+    },
+    {
+      "epoch": 4.552829571475303,
+      "grad_norm": 0.6991345882415771,
+      "learning_rate": 7.20925995419319e-05,
+      "loss": 0.0614,
+      "step": 69590
+    },
+    {
+      "epoch": 4.553483807654564,
+      "grad_norm": 0.9733833074569702,
+      "learning_rate": 7.208435856983068e-05,
+      "loss": 0.0631,
+      "step": 69600
+    },
+    {
+      "epoch": 4.554138043833824,
+      "grad_norm": 0.8843761086463928,
+      "learning_rate": 7.207611685232447e-05,
+      "loss": 0.0647,
+      "step": 69610
+    },
+    {
+      "epoch": 4.554792280013085,
+      "grad_norm": 1.0091403722763062,
+      "learning_rate": 7.206787438969138e-05,
+      "loss": 0.0581,
+      "step": 69620
+    },
+    {
+      "epoch": 4.555446516192346,
+      "grad_norm": 1.0898354053497314,
+      "learning_rate": 7.205963118220967e-05,
+      "loss": 0.0744,
+      "step": 69630
+    },
+    {
+      "epoch": 4.556100752371606,
+      "grad_norm": 0.7941951155662537,
+      "learning_rate": 7.205138723015756e-05,
+      "loss": 0.0728,
+      "step": 69640
+    },
+    {
+      "epoch": 4.556754988550867,
+      "grad_norm": 1.1205252408981323,
+      "learning_rate": 7.204314253381329e-05,
+      "loss": 0.0603,
+      "step": 69650
+    },
+    {
+      "epoch": 4.557409224730128,
+      "grad_norm": 0.9061139225959778,
+      "learning_rate": 7.203489709345515e-05,
+      "loss": 0.0613,
+      "step": 69660
+    },
+    {
+      "epoch": 4.558063460909389,
+      "grad_norm": 0.958292543888092,
+      "learning_rate": 7.202665090936145e-05,
+      "loss": 0.0667,
+      "step": 69670
+    },
+    {
+      "epoch": 4.558717697088649,
+      "grad_norm": 0.6560031771659851,
+      "learning_rate": 7.201840398181052e-05,
+      "loss": 0.0674,
+      "step": 69680
+    },
+    {
+      "epoch": 4.55937193326791,
+      "grad_norm": 0.8264176845550537,
+      "learning_rate": 7.201015631108071e-05,
+      "loss": 0.0682,
+      "step": 69690
+    },
+    {
+      "epoch": 4.560026169447171,
+      "grad_norm": 1.0350055694580078,
+      "learning_rate": 7.200190789745038e-05,
+      "loss": 0.061,
+      "step": 69700
+    },
+    {
+      "epoch": 4.560680405626432,
+      "grad_norm": 0.8044286370277405,
+      "learning_rate": 7.199365874119796e-05,
+      "loss": 0.0655,
+      "step": 69710
+    },
+    {
+      "epoch": 4.561334641805692,
+      "grad_norm": 0.9388670325279236,
+      "learning_rate": 7.198540884260189e-05,
+      "loss": 0.06,
+      "step": 69720
+    },
+    {
+      "epoch": 4.561988877984953,
+      "grad_norm": 0.7581605315208435,
+      "learning_rate": 7.197715820194062e-05,
+      "loss": 0.0535,
+      "step": 69730
+    },
+    {
+      "epoch": 4.562643114164214,
+      "grad_norm": 1.0746386051177979,
+      "learning_rate": 7.19689068194926e-05,
+      "loss": 0.0575,
+      "step": 69740
+    },
+    {
+      "epoch": 4.563297350343474,
+      "grad_norm": 0.7626873254776001,
+      "learning_rate": 7.196065469553637e-05,
+      "loss": 0.0616,
+      "step": 69750
+    },
+    {
+      "epoch": 4.563951586522735,
+      "grad_norm": 0.9289671182632446,
+      "learning_rate": 7.195240183035045e-05,
+      "loss": 0.0586,
+      "step": 69760
+    },
+    {
+      "epoch": 4.564605822701996,
+      "grad_norm": 0.8049091100692749,
+      "learning_rate": 7.194414822421341e-05,
+      "loss": 0.0527,
+      "step": 69770
+    },
+    {
+      "epoch": 4.565260058881256,
+      "grad_norm": 0.7008090615272522,
+      "learning_rate": 7.19358938774038e-05,
+      "loss": 0.0763,
+      "step": 69780
+    },
+    {
+      "epoch": 4.565914295060517,
+      "grad_norm": 1.1009507179260254,
+      "learning_rate": 7.192763879020025e-05,
+      "loss": 0.0732,
+      "step": 69790
+    },
+    {
+      "epoch": 4.566568531239778,
+      "grad_norm": 0.9115166068077087,
+      "learning_rate": 7.191938296288138e-05,
+      "loss": 0.0635,
+      "step": 69800
+    },
+    {
+      "epoch": 4.567222767419039,
+      "grad_norm": 0.9855472445487976,
+      "learning_rate": 7.191112639572585e-05,
+      "loss": 0.0594,
+      "step": 69810
+    },
+    {
+      "epoch": 4.567877003598299,
+      "grad_norm": 1.0158358812332153,
+      "learning_rate": 7.190286908901234e-05,
+      "loss": 0.0598,
+      "step": 69820
+    },
+    {
+      "epoch": 4.56853123977756,
+      "grad_norm": 0.7461121082305908,
+      "learning_rate": 7.189461104301955e-05,
+      "loss": 0.0578,
+      "step": 69830
+    },
+    {
+      "epoch": 4.569185475956821,
+      "grad_norm": 0.8417521119117737,
+      "learning_rate": 7.188635225802622e-05,
+      "loss": 0.0617,
+      "step": 69840
+    },
+    {
+      "epoch": 4.569839712136081,
+      "grad_norm": 0.9156501293182373,
+      "learning_rate": 7.187809273431112e-05,
+      "loss": 0.0567,
+      "step": 69850
+    },
+    {
+      "epoch": 4.570493948315342,
+      "grad_norm": 1.0383937358856201,
+      "learning_rate": 7.186983247215299e-05,
+      "loss": 0.0659,
+      "step": 69860
+    },
+    {
+      "epoch": 4.571148184494603,
+      "grad_norm": 0.6928499937057495,
+      "learning_rate": 7.186157147183067e-05,
+      "loss": 0.066,
+      "step": 69870
+    },
+    {
+      "epoch": 4.571802420673864,
+      "grad_norm": 0.7794655561447144,
+      "learning_rate": 7.185330973362298e-05,
+      "loss": 0.0602,
+      "step": 69880
+    },
+    {
+      "epoch": 4.572456656853124,
+      "grad_norm": 0.9522107839584351,
+      "learning_rate": 7.184504725780876e-05,
+      "loss": 0.0602,
+      "step": 69890
+    },
+    {
+      "epoch": 4.573110893032385,
+      "grad_norm": 0.9727455377578735,
+      "learning_rate": 7.183678404466691e-05,
+      "loss": 0.061,
+      "step": 69900
+    },
+    {
+      "epoch": 4.573765129211646,
+      "grad_norm": 0.8245759606361389,
+      "learning_rate": 7.182852009447633e-05,
+      "loss": 0.07,
+      "step": 69910
+    },
+    {
+      "epoch": 4.574419365390906,
+      "grad_norm": 0.9080676436424255,
+      "learning_rate": 7.182025540751595e-05,
+      "loss": 0.0647,
+      "step": 69920
+    },
+    {
+      "epoch": 4.575073601570167,
+      "grad_norm": 0.8655606508255005,
+      "learning_rate": 7.181198998406473e-05,
+      "loss": 0.0543,
+      "step": 69930
+    },
+    {
+      "epoch": 4.575727837749428,
+      "grad_norm": 1.0954169034957886,
+      "learning_rate": 7.180372382440164e-05,
+      "loss": 0.064,
+      "step": 69940
+    },
+    {
+      "epoch": 4.576382073928688,
+      "grad_norm": 0.817754328250885,
+      "learning_rate": 7.179545692880569e-05,
+      "loss": 0.0651,
+      "step": 69950
+    },
+    {
+      "epoch": 4.577036310107949,
+      "grad_norm": 0.6452881693840027,
+      "learning_rate": 7.17871892975559e-05,
+      "loss": 0.0624,
+      "step": 69960
+    },
+    {
+      "epoch": 4.57769054628721,
+      "grad_norm": 0.8822810053825378,
+      "learning_rate": 7.177892093093134e-05,
+      "loss": 0.0512,
+      "step": 69970
+    },
+    {
+      "epoch": 4.578344782466471,
+      "grad_norm": 0.7924016714096069,
+      "learning_rate": 7.17706518292111e-05,
+      "loss": 0.0663,
+      "step": 69980
+    },
+    {
+      "epoch": 4.578999018645731,
+      "grad_norm": 1.0157910585403442,
+      "learning_rate": 7.176238199267424e-05,
+      "loss": 0.0615,
+      "step": 69990
+    },
+    {
+      "epoch": 4.579653254824992,
+      "grad_norm": 0.6557690501213074,
+      "learning_rate": 7.175411142159991e-05,
+      "loss": 0.0521,
+      "step": 70000
+    },
+    {
+      "epoch": 4.580307491004253,
+      "grad_norm": 0.7803752422332764,
+      "learning_rate": 7.174584011626728e-05,
+      "loss": 0.0611,
+      "step": 70010
+    },
+    {
+      "epoch": 4.580961727183514,
+      "grad_norm": 0.889769971370697,
+      "learning_rate": 7.17375680769555e-05,
+      "loss": 0.0615,
+      "step": 70020
+    },
+    {
+      "epoch": 4.581615963362774,
+      "grad_norm": 0.706065833568573,
+      "learning_rate": 7.17292953039438e-05,
+      "loss": 0.0589,
+      "step": 70030
+    },
+    {
+      "epoch": 4.582270199542035,
+      "grad_norm": 0.9106086492538452,
+      "learning_rate": 7.172102179751141e-05,
+      "loss": 0.0638,
+      "step": 70040
+    },
+    {
+      "epoch": 4.582924435721296,
+      "grad_norm": 0.7716435790061951,
+      "learning_rate": 7.171274755793756e-05,
+      "loss": 0.0571,
+      "step": 70050
+    },
+    {
+      "epoch": 4.583578671900556,
+      "grad_norm": 0.8700406551361084,
+      "learning_rate": 7.170447258550152e-05,
+      "loss": 0.0659,
+      "step": 70060
+    },
+    {
+      "epoch": 4.584232908079817,
+      "grad_norm": 0.902850329875946,
+      "learning_rate": 7.169619688048262e-05,
+      "loss": 0.06,
+      "step": 70070
+    },
+    {
+      "epoch": 4.584887144259078,
+      "grad_norm": 1.30494225025177,
+      "learning_rate": 7.168792044316017e-05,
+      "loss": 0.0595,
+      "step": 70080
+    },
+    {
+      "epoch": 4.585541380438338,
+      "grad_norm": 0.878594160079956,
+      "learning_rate": 7.167964327381355e-05,
+      "loss": 0.0602,
+      "step": 70090
+    },
+    {
+      "epoch": 4.586195616617599,
+      "grad_norm": 1.009494662284851,
+      "learning_rate": 7.167136537272208e-05,
+      "loss": 0.0622,
+      "step": 70100
+    },
+    {
+      "epoch": 4.58684985279686,
+      "grad_norm": 0.858677864074707,
+      "learning_rate": 7.166308674016522e-05,
+      "loss": 0.0642,
+      "step": 70110
+    },
+    {
+      "epoch": 4.587504088976121,
+      "grad_norm": 0.8348826169967651,
+      "learning_rate": 7.165480737642234e-05,
+      "loss": 0.0596,
+      "step": 70120
+    },
+    {
+      "epoch": 4.588158325155381,
+      "grad_norm": 0.9198652505874634,
+      "learning_rate": 7.164652728177294e-05,
+      "loss": 0.0584,
+      "step": 70130
+    },
+    {
+      "epoch": 4.588812561334642,
+      "grad_norm": 0.8578175902366638,
+      "learning_rate": 7.163824645649648e-05,
+      "loss": 0.0597,
+      "step": 70140
+    },
+    {
+      "epoch": 4.589466797513903,
+      "grad_norm": 0.8815035223960876,
+      "learning_rate": 7.162996490087243e-05,
+      "loss": 0.0548,
+      "step": 70150
+    },
+    {
+      "epoch": 4.590121033693164,
+      "grad_norm": 0.8675861954689026,
+      "learning_rate": 7.162168261518036e-05,
+      "loss": 0.0571,
+      "step": 70160
+    },
+    {
+      "epoch": 4.590775269872424,
+      "grad_norm": 0.9907076358795166,
+      "learning_rate": 7.161339959969979e-05,
+      "loss": 0.0618,
+      "step": 70170
+    },
+    {
+      "epoch": 4.591429506051685,
+      "grad_norm": 0.8220086693763733,
+      "learning_rate": 7.160511585471031e-05,
+      "loss": 0.0592,
+      "step": 70180
+    },
+    {
+      "epoch": 4.592083742230946,
+      "grad_norm": 0.9038228988647461,
+      "learning_rate": 7.159683138049148e-05,
+      "loss": 0.0595,
+      "step": 70190
+    },
+    {
+      "epoch": 4.592737978410206,
+      "grad_norm": 0.9181106686592102,
+      "learning_rate": 7.158854617732297e-05,
+      "loss": 0.0556,
+      "step": 70200
+    },
+    {
+      "epoch": 4.593392214589467,
+      "grad_norm": 0.8551499843597412,
+      "learning_rate": 7.158026024548441e-05,
+      "loss": 0.0594,
+      "step": 70210
+    },
+    {
+      "epoch": 4.594046450768728,
+      "grad_norm": 0.9677573442459106,
+      "learning_rate": 7.157197358525546e-05,
+      "loss": 0.06,
+      "step": 70220
+    },
+    {
+      "epoch": 4.594700686947988,
+      "grad_norm": 0.8928607702255249,
+      "learning_rate": 7.156368619691582e-05,
+      "loss": 0.0647,
+      "step": 70230
+    },
+    {
+      "epoch": 4.595354923127249,
+      "grad_norm": 0.8647032976150513,
+      "learning_rate": 7.155539808074525e-05,
+      "loss": 0.0641,
+      "step": 70240
+    },
+    {
+      "epoch": 4.59600915930651,
+      "grad_norm": 0.9325335621833801,
+      "learning_rate": 7.154710923702345e-05,
+      "loss": 0.0613,
+      "step": 70250
+    },
+    {
+      "epoch": 4.596663395485771,
+      "grad_norm": 0.8028609156608582,
+      "learning_rate": 7.153881966603019e-05,
+      "loss": 0.0595,
+      "step": 70260
+    },
+    {
+      "epoch": 4.597317631665031,
+      "grad_norm": 0.9629660248756409,
+      "learning_rate": 7.153052936804529e-05,
+      "loss": 0.0555,
+      "step": 70270
+    },
+    {
+      "epoch": 4.597971867844292,
+      "grad_norm": 0.9992688298225403,
+      "learning_rate": 7.152223834334855e-05,
+      "loss": 0.0691,
+      "step": 70280
+    },
+    {
+      "epoch": 4.598626104023553,
+      "grad_norm": 0.7620879411697388,
+      "learning_rate": 7.151394659221984e-05,
+      "loss": 0.0601,
+      "step": 70290
+    },
+    {
+      "epoch": 4.599280340202813,
+      "grad_norm": 0.883135199546814,
+      "learning_rate": 7.150565411493899e-05,
+      "loss": 0.0554,
+      "step": 70300
+    },
+    {
+      "epoch": 4.599934576382074,
+      "grad_norm": 0.8866992592811584,
+      "learning_rate": 7.149736091178593e-05,
+      "loss": 0.0674,
+      "step": 70310
+    },
+    {
+      "epoch": 4.600588812561335,
+      "grad_norm": 1.0410959720611572,
+      "learning_rate": 7.148906698304054e-05,
+      "loss": 0.0634,
+      "step": 70320
+    },
+    {
+      "epoch": 4.601243048740596,
+      "grad_norm": 0.7747849225997925,
+      "learning_rate": 7.14807723289828e-05,
+      "loss": 0.0543,
+      "step": 70330
+    },
+    {
+      "epoch": 4.601897284919856,
+      "grad_norm": 0.9914830923080444,
+      "learning_rate": 7.147247694989265e-05,
+      "loss": 0.058,
+      "step": 70340
+    },
+    {
+      "epoch": 4.602551521099117,
+      "grad_norm": 0.8402534127235413,
+      "learning_rate": 7.146418084605008e-05,
+      "loss": 0.0634,
+      "step": 70350
+    },
+    {
+      "epoch": 4.603205757278378,
+      "grad_norm": 0.840990424156189,
+      "learning_rate": 7.145588401773513e-05,
+      "loss": 0.0559,
+      "step": 70360
+    },
+    {
+      "epoch": 4.603859993457638,
+      "grad_norm": 0.8721504807472229,
+      "learning_rate": 7.144758646522782e-05,
+      "loss": 0.0638,
+      "step": 70370
+    },
+    {
+      "epoch": 4.604514229636899,
+      "grad_norm": 0.8527781963348389,
+      "learning_rate": 7.143928818880823e-05,
+      "loss": 0.0532,
+      "step": 70380
+    },
+    {
+      "epoch": 4.60516846581616,
+      "grad_norm": 1.0726641416549683,
+      "learning_rate": 7.143098918875643e-05,
+      "loss": 0.0602,
+      "step": 70390
+    },
+    {
+      "epoch": 4.60582270199542,
+      "grad_norm": 1.0144989490509033,
+      "learning_rate": 7.142268946535254e-05,
+      "loss": 0.0642,
+      "step": 70400
+    },
+    {
+      "epoch": 4.606476938174681,
+      "grad_norm": 1.0396263599395752,
+      "learning_rate": 7.141438901887669e-05,
+      "loss": 0.0704,
+      "step": 70410
+    },
+    {
+      "epoch": 4.607131174353942,
+      "grad_norm": 0.7926657199859619,
+      "learning_rate": 7.140608784960904e-05,
+      "loss": 0.0692,
+      "step": 70420
+    },
+    {
+      "epoch": 4.607785410533203,
+      "grad_norm": 1.1654900312423706,
+      "learning_rate": 7.13977859578298e-05,
+      "loss": 0.0588,
+      "step": 70430
+    },
+    {
+      "epoch": 4.608439646712463,
+      "grad_norm": 0.9347232580184937,
+      "learning_rate": 7.138948334381917e-05,
+      "loss": 0.0604,
+      "step": 70440
+    },
+    {
+      "epoch": 4.609093882891724,
+      "grad_norm": 1.0925201177597046,
+      "learning_rate": 7.138118000785736e-05,
+      "loss": 0.0708,
+      "step": 70450
+    },
+    {
+      "epoch": 4.609748119070985,
+      "grad_norm": 0.9445613026618958,
+      "learning_rate": 7.137287595022467e-05,
+      "loss": 0.0523,
+      "step": 70460
+    },
+    {
+      "epoch": 4.610402355250246,
+      "grad_norm": 0.795276939868927,
+      "learning_rate": 7.136457117120136e-05,
+      "loss": 0.0655,
+      "step": 70470
+    },
+    {
+      "epoch": 4.611056591429506,
+      "grad_norm": 0.9454329609870911,
+      "learning_rate": 7.135626567106775e-05,
+      "loss": 0.0694,
+      "step": 70480
+    },
+    {
+      "epoch": 4.611710827608767,
+      "grad_norm": 0.8667651414871216,
+      "learning_rate": 7.134795945010416e-05,
+      "loss": 0.0673,
+      "step": 70490
+    },
+    {
+      "epoch": 4.612365063788028,
+      "grad_norm": 0.7597100734710693,
+      "learning_rate": 7.133965250859094e-05,
+      "loss": 0.0647,
+      "step": 70500
+    },
+    {
+      "epoch": 4.613019299967288,
+      "grad_norm": 1.0633947849273682,
+      "learning_rate": 7.13313448468085e-05,
+      "loss": 0.0676,
+      "step": 70510
+    },
+    {
+      "epoch": 4.613673536146549,
+      "grad_norm": 0.9230862259864807,
+      "learning_rate": 7.13230364650372e-05,
+      "loss": 0.0718,
+      "step": 70520
+    },
+    {
+      "epoch": 4.61432777232581,
+      "grad_norm": 0.8642282485961914,
+      "learning_rate": 7.131472736355754e-05,
+      "loss": 0.0678,
+      "step": 70530
+    },
+    {
+      "epoch": 4.61498200850507,
+      "grad_norm": 0.8637971878051758,
+      "learning_rate": 7.130641754264991e-05,
+      "loss": 0.0718,
+      "step": 70540
+    },
+    {
+      "epoch": 4.615636244684331,
+      "grad_norm": 0.943467915058136,
+      "learning_rate": 7.12981070025948e-05,
+      "loss": 0.0527,
+      "step": 70550
+    },
+    {
+      "epoch": 4.616290480863592,
+      "grad_norm": 0.843879222869873,
+      "learning_rate": 7.128979574367272e-05,
+      "loss": 0.0649,
+      "step": 70560
+    },
+    {
+      "epoch": 4.616944717042853,
+      "grad_norm": 0.9107741713523865,
+      "learning_rate": 7.128148376616422e-05,
+      "loss": 0.0617,
+      "step": 70570
+    },
+    {
+      "epoch": 4.617598953222113,
+      "grad_norm": 0.8617720007896423,
+      "learning_rate": 7.127317107034981e-05,
+      "loss": 0.061,
+      "step": 70580
+    },
+    {
+      "epoch": 4.618253189401374,
+      "grad_norm": 0.8260530829429626,
+      "learning_rate": 7.12648576565101e-05,
+      "loss": 0.0536,
+      "step": 70590
+    },
+    {
+      "epoch": 4.618907425580635,
+      "grad_norm": 0.9773645401000977,
+      "learning_rate": 7.125654352492567e-05,
+      "loss": 0.0531,
+      "step": 70600
+    },
+    {
+      "epoch": 4.619561661759896,
+      "grad_norm": 1.027761459350586,
+      "learning_rate": 7.124822867587715e-05,
+      "loss": 0.0566,
+      "step": 70610
+    },
+    {
+      "epoch": 4.620215897939156,
+      "grad_norm": 1.0088636875152588,
+      "learning_rate": 7.123991310964519e-05,
+      "loss": 0.0577,
+      "step": 70620
+    },
+    {
+      "epoch": 4.620870134118417,
+      "grad_norm": 0.9433162212371826,
+      "learning_rate": 7.123159682651045e-05,
+      "loss": 0.06,
+      "step": 70630
+    },
+    {
+      "epoch": 4.621524370297678,
+      "grad_norm": 1.0811787843704224,
+      "learning_rate": 7.122327982675363e-05,
+      "loss": 0.0597,
+      "step": 70640
+    },
+    {
+      "epoch": 4.622178606476938,
+      "grad_norm": 0.8905490636825562,
+      "learning_rate": 7.121496211065547e-05,
+      "loss": 0.0581,
+      "step": 70650
+    },
+    {
+      "epoch": 4.622832842656199,
+      "grad_norm": 0.801537811756134,
+      "learning_rate": 7.12066436784967e-05,
+      "loss": 0.0565,
+      "step": 70660
+    },
+    {
+      "epoch": 4.62348707883546,
+      "grad_norm": 0.9896419048309326,
+      "learning_rate": 7.119832453055809e-05,
+      "loss": 0.0645,
+      "step": 70670
+    },
+    {
+      "epoch": 4.62414131501472,
+      "grad_norm": 0.9551672339439392,
+      "learning_rate": 7.119000466712042e-05,
+      "loss": 0.058,
+      "step": 70680
+    },
+    {
+      "epoch": 4.624795551193981,
+      "grad_norm": 0.8658273220062256,
+      "learning_rate": 7.118168408846454e-05,
+      "loss": 0.0601,
+      "step": 70690
+    },
+    {
+      "epoch": 4.625449787373242,
+      "grad_norm": 0.7730889916419983,
+      "learning_rate": 7.117336279487124e-05,
+      "loss": 0.0589,
+      "step": 70700
+    },
+    {
+      "epoch": 4.626104023552503,
+      "grad_norm": 1.2071642875671387,
+      "learning_rate": 7.116504078662144e-05,
+      "loss": 0.0607,
+      "step": 70710
+    },
+    {
+      "epoch": 4.626758259731763,
+      "grad_norm": 0.6655055284500122,
+      "learning_rate": 7.1156718063996e-05,
+      "loss": 0.0672,
+      "step": 70720
+    },
+    {
+      "epoch": 4.627412495911024,
+      "grad_norm": 0.765287458896637,
+      "learning_rate": 7.114839462727585e-05,
+      "loss": 0.0657,
+      "step": 70730
+    },
+    {
+      "epoch": 4.628066732090285,
+      "grad_norm": 0.9906727075576782,
+      "learning_rate": 7.114007047674189e-05,
+      "loss": 0.0612,
+      "step": 70740
+    },
+    {
+      "epoch": 4.628720968269545,
+      "grad_norm": 0.8197884559631348,
+      "learning_rate": 7.113174561267514e-05,
+      "loss": 0.0638,
+      "step": 70750
+    },
+    {
+      "epoch": 4.629375204448806,
+      "grad_norm": 0.8256412744522095,
+      "learning_rate": 7.112342003535654e-05,
+      "loss": 0.062,
+      "step": 70760
+    },
+    {
+      "epoch": 4.630029440628067,
+      "grad_norm": 1.0937832593917847,
+      "learning_rate": 7.111509374506711e-05,
+      "loss": 0.0612,
+      "step": 70770
+    },
+    {
+      "epoch": 4.630683676807328,
+      "grad_norm": 0.8414881825447083,
+      "learning_rate": 7.11067667420879e-05,
+      "loss": 0.0588,
+      "step": 70780
+    },
+    {
+      "epoch": 4.631337912986588,
+      "grad_norm": 1.1070261001586914,
+      "learning_rate": 7.109843902669997e-05,
+      "loss": 0.0754,
+      "step": 70790
+    },
+    {
+      "epoch": 4.631992149165849,
+      "grad_norm": 1.002337098121643,
+      "learning_rate": 7.109011059918438e-05,
+      "loss": 0.062,
+      "step": 70800
+    },
+    {
+      "epoch": 4.63264638534511,
+      "grad_norm": 0.7594942450523376,
+      "learning_rate": 7.108178145982223e-05,
+      "loss": 0.06,
+      "step": 70810
+    },
+    {
+      "epoch": 4.63330062152437,
+      "grad_norm": 0.8567973375320435,
+      "learning_rate": 7.107345160889469e-05,
+      "loss": 0.0627,
+      "step": 70820
+    },
+    {
+      "epoch": 4.633954857703631,
+      "grad_norm": 0.919387936592102,
+      "learning_rate": 7.106512104668287e-05,
+      "loss": 0.0654,
+      "step": 70830
+    },
+    {
+      "epoch": 4.634609093882892,
+      "grad_norm": 0.9777571558952332,
+      "learning_rate": 7.1056789773468e-05,
+      "loss": 0.0661,
+      "step": 70840
+    },
+    {
+      "epoch": 4.635263330062152,
+      "grad_norm": 0.7844616174697876,
+      "learning_rate": 7.104845778953122e-05,
+      "loss": 0.0597,
+      "step": 70850
+    },
+    {
+      "epoch": 4.635917566241413,
+      "grad_norm": 0.8564948439598083,
+      "learning_rate": 7.10401250951538e-05,
+      "loss": 0.0679,
+      "step": 70860
+    },
+    {
+      "epoch": 4.636571802420674,
+      "grad_norm": 0.8882778286933899,
+      "learning_rate": 7.103179169061697e-05,
+      "loss": 0.0637,
+      "step": 70870
+    },
+    {
+      "epoch": 4.637226038599935,
+      "grad_norm": 0.8899109363555908,
+      "learning_rate": 7.102345757620204e-05,
+      "loss": 0.0639,
+      "step": 70880
+    },
+    {
+      "epoch": 4.637880274779195,
+      "grad_norm": 0.7381958961486816,
+      "learning_rate": 7.101512275219026e-05,
+      "loss": 0.0618,
+      "step": 70890
+    },
+    {
+      "epoch": 4.638534510958456,
+      "grad_norm": 1.1825381517410278,
+      "learning_rate": 7.100678721886296e-05,
+      "loss": 0.069,
+      "step": 70900
+    },
+    {
+      "epoch": 4.639188747137717,
+      "grad_norm": 0.8245072960853577,
+      "learning_rate": 7.099845097650152e-05,
+      "loss": 0.0648,
+      "step": 70910
+    },
+    {
+      "epoch": 4.639842983316978,
+      "grad_norm": 1.0069905519485474,
+      "learning_rate": 7.099011402538729e-05,
+      "loss": 0.0565,
+      "step": 70920
+    },
+    {
+      "epoch": 4.640497219496238,
+      "grad_norm": 0.9455946087837219,
+      "learning_rate": 7.098177636580165e-05,
+      "loss": 0.0644,
+      "step": 70930
+    },
+    {
+      "epoch": 4.641151455675499,
+      "grad_norm": 1.0115948915481567,
+      "learning_rate": 7.097343799802603e-05,
+      "loss": 0.0598,
+      "step": 70940
+    },
+    {
+      "epoch": 4.64180569185476,
+      "grad_norm": 0.910192608833313,
+      "learning_rate": 7.096509892234188e-05,
+      "loss": 0.062,
+      "step": 70950
+    },
+    {
+      "epoch": 4.64245992803402,
+      "grad_norm": 1.005014419555664,
+      "learning_rate": 7.095675913903067e-05,
+      "loss": 0.0573,
+      "step": 70960
+    },
+    {
+      "epoch": 4.643114164213281,
+      "grad_norm": 0.8233394622802734,
+      "learning_rate": 7.094841864837385e-05,
+      "loss": 0.0607,
+      "step": 70970
+    },
+    {
+      "epoch": 4.643768400392542,
+      "grad_norm": 0.84469074010849,
+      "learning_rate": 7.094007745065298e-05,
+      "loss": 0.0547,
+      "step": 70980
+    },
+    {
+      "epoch": 4.644422636571802,
+      "grad_norm": 0.846990168094635,
+      "learning_rate": 7.093173554614958e-05,
+      "loss": 0.0596,
+      "step": 70990
+    },
+    {
+      "epoch": 4.645076872751063,
+      "grad_norm": 1.1320626735687256,
+      "learning_rate": 7.092339293514521e-05,
+      "loss": 0.0595,
+      "step": 71000
+    },
+    {
+      "epoch": 4.645731108930324,
+      "grad_norm": 0.9012898206710815,
+      "learning_rate": 7.091504961792145e-05,
+      "loss": 0.0565,
+      "step": 71010
+    },
+    {
+      "epoch": 4.646385345109585,
+      "grad_norm": 0.9785043001174927,
+      "learning_rate": 7.090670559475991e-05,
+      "loss": 0.0636,
+      "step": 71020
+    },
+    {
+      "epoch": 4.647039581288845,
+      "grad_norm": 0.9479057788848877,
+      "learning_rate": 7.089836086594223e-05,
+      "loss": 0.063,
+      "step": 71030
+    },
+    {
+      "epoch": 4.647693817468106,
+      "grad_norm": 0.8791054487228394,
+      "learning_rate": 7.089001543175007e-05,
+      "loss": 0.0666,
+      "step": 71040
+    },
+    {
+      "epoch": 4.648348053647367,
+      "grad_norm": 1.0434389114379883,
+      "learning_rate": 7.088166929246509e-05,
+      "loss": 0.068,
+      "step": 71050
+    },
+    {
+      "epoch": 4.649002289826628,
+      "grad_norm": 0.8566122651100159,
+      "learning_rate": 7.087332244836901e-05,
+      "loss": 0.0675,
+      "step": 71060
+    },
+    {
+      "epoch": 4.649656526005888,
+      "grad_norm": 0.9243730902671814,
+      "learning_rate": 7.086497489974355e-05,
+      "loss": 0.0639,
+      "step": 71070
+    },
+    {
+      "epoch": 4.650310762185149,
+      "grad_norm": 0.8859608173370361,
+      "learning_rate": 7.085662664687049e-05,
+      "loss": 0.06,
+      "step": 71080
+    },
+    {
+      "epoch": 4.65096499836441,
+      "grad_norm": 1.0583763122558594,
+      "learning_rate": 7.084827769003157e-05,
+      "loss": 0.0573,
+      "step": 71090
+    },
+    {
+      "epoch": 4.65161923454367,
+      "grad_norm": 0.8830341100692749,
+      "learning_rate": 7.083992802950859e-05,
+      "loss": 0.0557,
+      "step": 71100
+    },
+    {
+      "epoch": 4.652273470722931,
+      "grad_norm": 0.8435945510864258,
+      "learning_rate": 7.08315776655834e-05,
+      "loss": 0.0535,
+      "step": 71110
+    },
+    {
+      "epoch": 4.652927706902192,
+      "grad_norm": 0.9333560466766357,
+      "learning_rate": 7.082322659853782e-05,
+      "loss": 0.065,
+      "step": 71120
+    },
+    {
+      "epoch": 4.653581943081452,
+      "grad_norm": 0.8376295566558838,
+      "learning_rate": 7.081487482865375e-05,
+      "loss": 0.0518,
+      "step": 71130
+    },
+    {
+      "epoch": 4.654236179260713,
+      "grad_norm": 0.7499420642852783,
+      "learning_rate": 7.080652235621304e-05,
+      "loss": 0.0536,
+      "step": 71140
+    },
+    {
+      "epoch": 4.654890415439974,
+      "grad_norm": 0.7663356065750122,
+      "learning_rate": 7.079816918149764e-05,
+      "loss": 0.0598,
+      "step": 71150
+    },
+    {
+      "epoch": 4.655544651619235,
+      "grad_norm": 1.0369300842285156,
+      "learning_rate": 7.07898153047895e-05,
+      "loss": 0.0584,
+      "step": 71160
+    },
+    {
+      "epoch": 4.656198887798495,
+      "grad_norm": 0.8538703322410583,
+      "learning_rate": 7.078146072637055e-05,
+      "loss": 0.0608,
+      "step": 71170
+    },
+    {
+      "epoch": 4.656853123977756,
+      "grad_norm": 0.9272082448005676,
+      "learning_rate": 7.077310544652282e-05,
+      "loss": 0.0666,
+      "step": 71180
+    },
+    {
+      "epoch": 4.657507360157017,
+      "grad_norm": 0.8066675066947937,
+      "learning_rate": 7.076474946552828e-05,
+      "loss": 0.0551,
+      "step": 71190
+    },
+    {
+      "epoch": 4.658161596336277,
+      "grad_norm": 0.9657180905342102,
+      "learning_rate": 7.0756392783669e-05,
+      "loss": 0.067,
+      "step": 71200
+    },
+    {
+      "epoch": 4.658815832515538,
+      "grad_norm": 0.9008907675743103,
+      "learning_rate": 7.074803540122703e-05,
+      "loss": 0.0601,
+      "step": 71210
+    },
+    {
+      "epoch": 4.659470068694799,
+      "grad_norm": 0.7514058947563171,
+      "learning_rate": 7.073967731848445e-05,
+      "loss": 0.0573,
+      "step": 71220
+    },
+    {
+      "epoch": 4.66012430487406,
+      "grad_norm": 0.8084816336631775,
+      "learning_rate": 7.073131853572335e-05,
+      "loss": 0.0641,
+      "step": 71230
+    },
+    {
+      "epoch": 4.66077854105332,
+      "grad_norm": 0.7935675978660583,
+      "learning_rate": 7.072295905322592e-05,
+      "loss": 0.0567,
+      "step": 71240
+    },
+    {
+      "epoch": 4.661432777232581,
+      "grad_norm": 0.9092238545417786,
+      "learning_rate": 7.071459887127424e-05,
+      "loss": 0.0674,
+      "step": 71250
+    },
+    {
+      "epoch": 4.662087013411842,
+      "grad_norm": 0.9651187062263489,
+      "learning_rate": 7.070623799015052e-05,
+      "loss": 0.0626,
+      "step": 71260
+    },
+    {
+      "epoch": 4.662741249591102,
+      "grad_norm": 0.7449125051498413,
+      "learning_rate": 7.069787641013699e-05,
+      "loss": 0.0598,
+      "step": 71270
+    },
+    {
+      "epoch": 4.663395485770363,
+      "grad_norm": 0.8150971531867981,
+      "learning_rate": 7.068951413151583e-05,
+      "loss": 0.0704,
+      "step": 71280
+    },
+    {
+      "epoch": 4.664049721949624,
+      "grad_norm": 1.1093558073043823,
+      "learning_rate": 7.06811511545693e-05,
+      "loss": 0.0529,
+      "step": 71290
+    },
+    {
+      "epoch": 4.664703958128884,
+      "grad_norm": 0.9424728155136108,
+      "learning_rate": 7.06727874795797e-05,
+      "loss": 0.0607,
+      "step": 71300
+    },
+    {
+      "epoch": 4.665358194308145,
+      "grad_norm": 0.9972444772720337,
+      "learning_rate": 7.06644231068293e-05,
+      "loss": 0.0585,
+      "step": 71310
+    },
+    {
+      "epoch": 4.666012430487406,
+      "grad_norm": 0.8003326654434204,
+      "learning_rate": 7.065605803660042e-05,
+      "loss": 0.0623,
+      "step": 71320
+    },
+    {
+      "epoch": 4.666666666666667,
+      "grad_norm": 0.8432341814041138,
+      "learning_rate": 7.064769226917543e-05,
+      "loss": 0.0526,
+      "step": 71330
+    },
+    {
+      "epoch": 4.667320902845927,
+      "grad_norm": 0.828727126121521,
+      "learning_rate": 7.063932580483665e-05,
+      "loss": 0.0643,
+      "step": 71340
+    },
+    {
+      "epoch": 4.667975139025188,
+      "grad_norm": 0.9466299414634705,
+      "learning_rate": 7.063095864386651e-05,
+      "loss": 0.057,
+      "step": 71350
+    },
+    {
+      "epoch": 4.668629375204449,
+      "grad_norm": 0.6953588128089905,
+      "learning_rate": 7.06225907865474e-05,
+      "loss": 0.0522,
+      "step": 71360
+    },
+    {
+      "epoch": 4.66928361138371,
+      "grad_norm": 0.8251481056213379,
+      "learning_rate": 7.061422223316176e-05,
+      "loss": 0.06,
+      "step": 71370
+    },
+    {
+      "epoch": 4.66993784756297,
+      "grad_norm": 1.085668683052063,
+      "learning_rate": 7.060585298399207e-05,
+      "loss": 0.0653,
+      "step": 71380
+    },
+    {
+      "epoch": 4.670592083742231,
+      "grad_norm": 1.0009715557098389,
+      "learning_rate": 7.05974830393208e-05,
+      "loss": 0.065,
+      "step": 71390
+    },
+    {
+      "epoch": 4.671246319921492,
+      "grad_norm": 0.8641042709350586,
+      "learning_rate": 7.058911239943046e-05,
+      "loss": 0.0589,
+      "step": 71400
+    },
+    {
+      "epoch": 4.671900556100752,
+      "grad_norm": 0.8176802396774292,
+      "learning_rate": 7.058074106460357e-05,
+      "loss": 0.058,
+      "step": 71410
+    },
+    {
+      "epoch": 4.672554792280013,
+      "grad_norm": 0.8100340366363525,
+      "learning_rate": 7.057236903512269e-05,
+      "loss": 0.0636,
+      "step": 71420
+    },
+    {
+      "epoch": 4.673209028459274,
+      "grad_norm": 0.9662539958953857,
+      "learning_rate": 7.056399631127043e-05,
+      "loss": 0.0568,
+      "step": 71430
+    },
+    {
+      "epoch": 4.673863264638534,
+      "grad_norm": 1.049534559249878,
+      "learning_rate": 7.055562289332934e-05,
+      "loss": 0.0701,
+      "step": 71440
+    },
+    {
+      "epoch": 4.674517500817795,
+      "grad_norm": 0.670726478099823,
+      "learning_rate": 7.054724878158206e-05,
+      "loss": 0.0551,
+      "step": 71450
+    },
+    {
+      "epoch": 4.675171736997056,
+      "grad_norm": 1.3592555522918701,
+      "learning_rate": 7.053887397631127e-05,
+      "loss": 0.0645,
+      "step": 71460
+    },
+    {
+      "epoch": 4.675825973176317,
+      "grad_norm": 0.7808588147163391,
+      "learning_rate": 7.053049847779961e-05,
+      "loss": 0.0649,
+      "step": 71470
+    },
+    {
+      "epoch": 4.676480209355577,
+      "grad_norm": 0.8169131875038147,
+      "learning_rate": 7.052212228632977e-05,
+      "loss": 0.0547,
+      "step": 71480
+    },
+    {
+      "epoch": 4.677134445534838,
+      "grad_norm": 0.8290903568267822,
+      "learning_rate": 7.051374540218451e-05,
+      "loss": 0.065,
+      "step": 71490
+    },
+    {
+      "epoch": 4.677788681714099,
+      "grad_norm": 1.1689528226852417,
+      "learning_rate": 7.050536782564653e-05,
+      "loss": 0.0707,
+      "step": 71500
+    },
+    {
+      "epoch": 4.67844291789336,
+      "grad_norm": 0.7976114749908447,
+      "learning_rate": 7.04969895569986e-05,
+      "loss": 0.0652,
+      "step": 71510
+    },
+    {
+      "epoch": 4.67909715407262,
+      "grad_norm": 0.9616780281066895,
+      "learning_rate": 7.04886105965235e-05,
+      "loss": 0.061,
+      "step": 71520
+    },
+    {
+      "epoch": 4.679751390251881,
+      "grad_norm": 0.9938467741012573,
+      "learning_rate": 7.048023094450411e-05,
+      "loss": 0.062,
+      "step": 71530
+    },
+    {
+      "epoch": 4.680405626431142,
+      "grad_norm": 0.8165993690490723,
+      "learning_rate": 7.047185060122317e-05,
+      "loss": 0.0564,
+      "step": 71540
+    },
+    {
+      "epoch": 4.681059862610402,
+      "grad_norm": 0.8592141270637512,
+      "learning_rate": 7.046346956696359e-05,
+      "loss": 0.0621,
+      "step": 71550
+    },
+    {
+      "epoch": 4.681714098789663,
+      "grad_norm": 0.957463800907135,
+      "learning_rate": 7.045508784200826e-05,
+      "loss": 0.0575,
+      "step": 71560
+    },
+    {
+      "epoch": 4.682368334968924,
+      "grad_norm": 0.7497161626815796,
+      "learning_rate": 7.044670542664007e-05,
+      "loss": 0.0598,
+      "step": 71570
+    },
+    {
+      "epoch": 4.683022571148184,
+      "grad_norm": 0.8132786750793457,
+      "learning_rate": 7.043832232114194e-05,
+      "loss": 0.0562,
+      "step": 71580
+    },
+    {
+      "epoch": 4.683676807327445,
+      "grad_norm": 0.8646707534790039,
+      "learning_rate": 7.042993852579683e-05,
+      "loss": 0.056,
+      "step": 71590
+    },
+    {
+      "epoch": 4.684331043506706,
+      "grad_norm": 0.82687908411026,
+      "learning_rate": 7.042155404088772e-05,
+      "loss": 0.0574,
+      "step": 71600
+    },
+    {
+      "epoch": 4.684985279685967,
+      "grad_norm": 0.9083144664764404,
+      "learning_rate": 7.04131688666976e-05,
+      "loss": 0.0564,
+      "step": 71610
+    },
+    {
+      "epoch": 4.685639515865227,
+      "grad_norm": 0.7439766526222229,
+      "learning_rate": 7.04047830035095e-05,
+      "loss": 0.06,
+      "step": 71620
+    },
+    {
+      "epoch": 4.686293752044488,
+      "grad_norm": 0.8073269128799438,
+      "learning_rate": 7.039639645160646e-05,
+      "loss": 0.064,
+      "step": 71630
+    },
+    {
+      "epoch": 4.686947988223749,
+      "grad_norm": 1.0741745233535767,
+      "learning_rate": 7.038800921127152e-05,
+      "loss": 0.0638,
+      "step": 71640
+    },
+    {
+      "epoch": 4.687602224403009,
+      "grad_norm": 0.9227171540260315,
+      "learning_rate": 7.037962128278783e-05,
+      "loss": 0.0577,
+      "step": 71650
+    },
+    {
+      "epoch": 4.68825646058227,
+      "grad_norm": 1.138973355293274,
+      "learning_rate": 7.037123266643846e-05,
+      "loss": 0.0701,
+      "step": 71660
+    },
+    {
+      "epoch": 4.688910696761531,
+      "grad_norm": 1.039119839668274,
+      "learning_rate": 7.036284336250658e-05,
+      "loss": 0.0623,
+      "step": 71670
+    },
+    {
+      "epoch": 4.689564932940792,
+      "grad_norm": 1.0590921640396118,
+      "learning_rate": 7.035445337127532e-05,
+      "loss": 0.0638,
+      "step": 71680
+    },
+    {
+      "epoch": 4.690219169120052,
+      "grad_norm": 0.8854411244392395,
+      "learning_rate": 7.034606269302789e-05,
+      "loss": 0.0663,
+      "step": 71690
+    },
+    {
+      "epoch": 4.690873405299313,
+      "grad_norm": 0.7692389488220215,
+      "learning_rate": 7.033767132804747e-05,
+      "loss": 0.0529,
+      "step": 71700
+    },
+    {
+      "epoch": 4.691527641478574,
+      "grad_norm": 0.6598808765411377,
+      "learning_rate": 7.03292792766173e-05,
+      "loss": 0.0584,
+      "step": 71710
+    },
+    {
+      "epoch": 4.692181877657834,
+      "grad_norm": 0.8510991334915161,
+      "learning_rate": 7.032088653902067e-05,
+      "loss": 0.0621,
+      "step": 71720
+    },
+    {
+      "epoch": 4.692836113837095,
+      "grad_norm": 0.9420493245124817,
+      "learning_rate": 7.031249311554079e-05,
+      "loss": 0.0635,
+      "step": 71730
+    },
+    {
+      "epoch": 4.693490350016356,
+      "grad_norm": 0.8198224306106567,
+      "learning_rate": 7.0304099006461e-05,
+      "loss": 0.0673,
+      "step": 71740
+    },
+    {
+      "epoch": 4.694144586195616,
+      "grad_norm": 1.035709023475647,
+      "learning_rate": 7.029570421206464e-05,
+      "loss": 0.064,
+      "step": 71750
+    },
+    {
+      "epoch": 4.694798822374877,
+      "grad_norm": 0.8997196555137634,
+      "learning_rate": 7.028730873263502e-05,
+      "loss": 0.0555,
+      "step": 71760
+    },
+    {
+      "epoch": 4.695453058554138,
+      "grad_norm": 0.9662396907806396,
+      "learning_rate": 7.027891256845553e-05,
+      "loss": 0.0634,
+      "step": 71770
+    },
+    {
+      "epoch": 4.696107294733399,
+      "grad_norm": 0.7257537245750427,
+      "learning_rate": 7.027051571980957e-05,
+      "loss": 0.0614,
+      "step": 71780
+    },
+    {
+      "epoch": 4.696761530912659,
+      "grad_norm": 0.9215990900993347,
+      "learning_rate": 7.026211818698053e-05,
+      "loss": 0.0698,
+      "step": 71790
+    },
+    {
+      "epoch": 4.69741576709192,
+      "grad_norm": 0.7259654402732849,
+      "learning_rate": 7.025371997025185e-05,
+      "loss": 0.0678,
+      "step": 71800
+    },
+    {
+      "epoch": 4.698070003271181,
+      "grad_norm": 0.9841349124908447,
+      "learning_rate": 7.0245321069907e-05,
+      "loss": 0.0576,
+      "step": 71810
+    },
+    {
+      "epoch": 4.698724239450442,
+      "grad_norm": 0.9428102374076843,
+      "learning_rate": 7.02369214862295e-05,
+      "loss": 0.0601,
+      "step": 71820
+    },
+    {
+      "epoch": 4.699378475629702,
+      "grad_norm": 0.7501752376556396,
+      "learning_rate": 7.022852121950281e-05,
+      "loss": 0.0564,
+      "step": 71830
+    },
+    {
+      "epoch": 4.700032711808963,
+      "grad_norm": 0.705059289932251,
+      "learning_rate": 7.022012027001048e-05,
+      "loss": 0.0499,
+      "step": 71840
+    },
+    {
+      "epoch": 4.700686947988224,
+      "grad_norm": 0.7624207735061646,
+      "learning_rate": 7.021171863803606e-05,
+      "loss": 0.072,
+      "step": 71850
+    },
+    {
+      "epoch": 4.701341184167484,
+      "grad_norm": 0.7773285508155823,
+      "learning_rate": 7.020331632386312e-05,
+      "loss": 0.0545,
+      "step": 71860
+    },
+    {
+      "epoch": 4.701995420346745,
+      "grad_norm": 1.0297824144363403,
+      "learning_rate": 7.019491332777528e-05,
+      "loss": 0.0549,
+      "step": 71870
+    },
+    {
+      "epoch": 4.702649656526006,
+      "grad_norm": 0.9911094307899475,
+      "learning_rate": 7.018650965005616e-05,
+      "loss": 0.0671,
+      "step": 71880
+    },
+    {
+      "epoch": 4.703303892705266,
+      "grad_norm": 0.960195779800415,
+      "learning_rate": 7.017810529098938e-05,
+      "loss": 0.0664,
+      "step": 71890
+    },
+    {
+      "epoch": 4.703958128884527,
+      "grad_norm": 0.7779977321624756,
+      "learning_rate": 7.016970025085864e-05,
+      "loss": 0.0593,
+      "step": 71900
+    },
+    {
+      "epoch": 4.704612365063788,
+      "grad_norm": 0.8631134629249573,
+      "learning_rate": 7.016129452994761e-05,
+      "loss": 0.06,
+      "step": 71910
+    },
+    {
+      "epoch": 4.705266601243049,
+      "grad_norm": 0.854272186756134,
+      "learning_rate": 7.015288812854003e-05,
+      "loss": 0.0593,
+      "step": 71920
+    },
+    {
+      "epoch": 4.705920837422309,
+      "grad_norm": 1.105404257774353,
+      "learning_rate": 7.01444810469196e-05,
+      "loss": 0.0675,
+      "step": 71930
+    },
+    {
+      "epoch": 4.70657507360157,
+      "grad_norm": 0.8611233234405518,
+      "learning_rate": 7.013607328537012e-05,
+      "loss": 0.0567,
+      "step": 71940
+    },
+    {
+      "epoch": 4.707229309780831,
+      "grad_norm": 0.8065164089202881,
+      "learning_rate": 7.012766484417536e-05,
+      "loss": 0.065,
+      "step": 71950
+    },
+    {
+      "epoch": 4.707883545960092,
+      "grad_norm": 0.853921115398407,
+      "learning_rate": 7.011925572361912e-05,
+      "loss": 0.0626,
+      "step": 71960
+    },
+    {
+      "epoch": 4.708537782139352,
+      "grad_norm": 0.850261926651001,
+      "learning_rate": 7.011084592398523e-05,
+      "loss": 0.0757,
+      "step": 71970
+    },
+    {
+      "epoch": 4.709192018318613,
+      "grad_norm": 1.0401116609573364,
+      "learning_rate": 7.010243544555756e-05,
+      "loss": 0.0597,
+      "step": 71980
+    },
+    {
+      "epoch": 4.709846254497874,
+      "grad_norm": 1.0012506246566772,
+      "learning_rate": 7.009402428861995e-05,
+      "loss": 0.0637,
+      "step": 71990
+    },
+    {
+      "epoch": 4.710500490677134,
+      "grad_norm": 0.854714572429657,
+      "learning_rate": 7.008561245345634e-05,
+      "loss": 0.0605,
+      "step": 72000
+    },
+    {
+      "epoch": 4.711154726856395,
+      "grad_norm": 1.0572654008865356,
+      "learning_rate": 7.007719994035063e-05,
+      "loss": 0.0718,
+      "step": 72010
+    },
+    {
+      "epoch": 4.711808963035656,
+      "grad_norm": 0.9308390021324158,
+      "learning_rate": 7.006878674958676e-05,
+      "loss": 0.0705,
+      "step": 72020
+    },
+    {
+      "epoch": 4.712463199214916,
+      "grad_norm": 0.945894181728363,
+      "learning_rate": 7.006037288144872e-05,
+      "loss": 0.0676,
+      "step": 72030
+    },
+    {
+      "epoch": 4.713117435394177,
+      "grad_norm": 1.6244274377822876,
+      "learning_rate": 7.005195833622048e-05,
+      "loss": 0.059,
+      "step": 72040
+    },
+    {
+      "epoch": 4.713771671573438,
+      "grad_norm": 0.9291044473648071,
+      "learning_rate": 7.004354311418606e-05,
+      "loss": 0.0635,
+      "step": 72050
+    },
+    {
+      "epoch": 4.714425907752699,
+      "grad_norm": 0.9611480832099915,
+      "learning_rate": 7.003512721562949e-05,
+      "loss": 0.0566,
+      "step": 72060
+    },
+    {
+      "epoch": 4.715080143931959,
+      "grad_norm": 0.7825486660003662,
+      "learning_rate": 7.002671064083482e-05,
+      "loss": 0.0575,
+      "step": 72070
+    },
+    {
+      "epoch": 4.71573438011122,
+      "grad_norm": 0.8755865693092346,
+      "learning_rate": 7.001829339008616e-05,
+      "loss": 0.065,
+      "step": 72080
+    },
+    {
+      "epoch": 4.716388616290481,
+      "grad_norm": 0.8769670724868774,
+      "learning_rate": 7.000987546366758e-05,
+      "loss": 0.068,
+      "step": 72090
+    },
+    {
+      "epoch": 4.717042852469741,
+      "grad_norm": 0.7498029470443726,
+      "learning_rate": 7.000145686186324e-05,
+      "loss": 0.0537,
+      "step": 72100
+    },
+    {
+      "epoch": 4.717697088649002,
+      "grad_norm": 1.0412774085998535,
+      "learning_rate": 6.999303758495727e-05,
+      "loss": 0.0519,
+      "step": 72110
+    },
+    {
+      "epoch": 4.718351324828263,
+      "grad_norm": 0.8896167278289795,
+      "learning_rate": 6.998461763323385e-05,
+      "loss": 0.0576,
+      "step": 72120
+    },
+    {
+      "epoch": 4.719005561007524,
+      "grad_norm": 1.0666289329528809,
+      "learning_rate": 6.997619700697719e-05,
+      "loss": 0.0625,
+      "step": 72130
+    },
+    {
+      "epoch": 4.719659797186784,
+      "grad_norm": 1.330458164215088,
+      "learning_rate": 6.996777570647147e-05,
+      "loss": 0.07,
+      "step": 72140
+    },
+    {
+      "epoch": 4.720314033366045,
+      "grad_norm": 1.0409564971923828,
+      "learning_rate": 6.995935373200095e-05,
+      "loss": 0.0543,
+      "step": 72150
+    },
+    {
+      "epoch": 4.720968269545306,
+      "grad_norm": 0.8398634195327759,
+      "learning_rate": 6.995093108384992e-05,
+      "loss": 0.0582,
+      "step": 72160
+    },
+    {
+      "epoch": 4.721622505724566,
+      "grad_norm": 0.9149483442306519,
+      "learning_rate": 6.994250776230262e-05,
+      "loss": 0.0605,
+      "step": 72170
+    },
+    {
+      "epoch": 4.722276741903827,
+      "grad_norm": 0.8937086462974548,
+      "learning_rate": 6.993408376764339e-05,
+      "loss": 0.0579,
+      "step": 72180
+    },
+    {
+      "epoch": 4.722930978083088,
+      "grad_norm": 0.8356661796569824,
+      "learning_rate": 6.992565910015655e-05,
+      "loss": 0.0569,
+      "step": 72190
+    },
+    {
+      "epoch": 4.723585214262348,
+      "grad_norm": 1.0383862257003784,
+      "learning_rate": 6.991723376012646e-05,
+      "loss": 0.0665,
+      "step": 72200
+    },
+    {
+      "epoch": 4.724239450441609,
+      "grad_norm": 0.9304238557815552,
+      "learning_rate": 6.99088077478375e-05,
+      "loss": 0.054,
+      "step": 72210
+    },
+    {
+      "epoch": 4.72489368662087,
+      "grad_norm": 1.0787382125854492,
+      "learning_rate": 6.990038106357407e-05,
+      "loss": 0.059,
+      "step": 72220
+    },
+    {
+      "epoch": 4.725547922800131,
+      "grad_norm": 0.9750184416770935,
+      "learning_rate": 6.989195370762057e-05,
+      "loss": 0.0621,
+      "step": 72230
+    },
+    {
+      "epoch": 4.726202158979391,
+      "grad_norm": 0.9108086824417114,
+      "learning_rate": 6.988352568026148e-05,
+      "loss": 0.0576,
+      "step": 72240
+    },
+    {
+      "epoch": 4.726856395158652,
+      "grad_norm": 1.0215849876403809,
+      "learning_rate": 6.987509698178125e-05,
+      "loss": 0.0619,
+      "step": 72250
+    },
+    {
+      "epoch": 4.727510631337913,
+      "grad_norm": 1.3526380062103271,
+      "learning_rate": 6.986666761246436e-05,
+      "loss": 0.0573,
+      "step": 72260
+    },
+    {
+      "epoch": 4.728164867517174,
+      "grad_norm": 0.9454095363616943,
+      "learning_rate": 6.985823757259535e-05,
+      "loss": 0.0601,
+      "step": 72270
+    },
+    {
+      "epoch": 4.728819103696434,
+      "grad_norm": 1.032045841217041,
+      "learning_rate": 6.984980686245874e-05,
+      "loss": 0.0577,
+      "step": 72280
+    },
+    {
+      "epoch": 4.729473339875695,
+      "grad_norm": 0.8855615854263306,
+      "learning_rate": 6.98413754823391e-05,
+      "loss": 0.0683,
+      "step": 72290
+    },
+    {
+      "epoch": 4.730127576054956,
+      "grad_norm": 0.9899473786354065,
+      "learning_rate": 6.983294343252098e-05,
+      "loss": 0.0705,
+      "step": 72300
+    },
+    {
+      "epoch": 4.730781812234216,
+      "grad_norm": 0.9037162065505981,
+      "learning_rate": 6.982451071328902e-05,
+      "loss": 0.0659,
+      "step": 72310
+    },
+    {
+      "epoch": 4.731436048413477,
+      "grad_norm": 0.9520037174224854,
+      "learning_rate": 6.98160773249278e-05,
+      "loss": 0.0558,
+      "step": 72320
+    },
+    {
+      "epoch": 4.732090284592738,
+      "grad_norm": 0.6964289546012878,
+      "learning_rate": 6.980764326772204e-05,
+      "loss": 0.054,
+      "step": 72330
+    },
+    {
+      "epoch": 4.732744520771998,
+      "grad_norm": 1.047905683517456,
+      "learning_rate": 6.979920854195637e-05,
+      "loss": 0.0623,
+      "step": 72340
+    },
+    {
+      "epoch": 4.733398756951259,
+      "grad_norm": 1.0188045501708984,
+      "learning_rate": 6.979077314791546e-05,
+      "loss": 0.0634,
+      "step": 72350
+    },
+    {
+      "epoch": 4.73405299313052,
+      "grad_norm": 0.7242242097854614,
+      "learning_rate": 6.978233708588407e-05,
+      "loss": 0.0592,
+      "step": 72360
+    },
+    {
+      "epoch": 4.734707229309781,
+      "grad_norm": 1.0020005702972412,
+      "learning_rate": 6.977390035614692e-05,
+      "loss": 0.0608,
+      "step": 72370
+    },
+    {
+      "epoch": 4.735361465489041,
+      "grad_norm": 0.9756394028663635,
+      "learning_rate": 6.976546295898878e-05,
+      "loss": 0.0702,
+      "step": 72380
+    },
+    {
+      "epoch": 4.736015701668302,
+      "grad_norm": 0.9397419095039368,
+      "learning_rate": 6.975702489469442e-05,
+      "loss": 0.0573,
+      "step": 72390
+    },
+    {
+      "epoch": 4.736669937847563,
+      "grad_norm": 0.9236851334571838,
+      "learning_rate": 6.974858616354867e-05,
+      "loss": 0.0593,
+      "step": 72400
+    },
+    {
+      "epoch": 4.737324174026824,
+      "grad_norm": 0.7359132170677185,
+      "learning_rate": 6.974014676583632e-05,
+      "loss": 0.0565,
+      "step": 72410
+    },
+    {
+      "epoch": 4.737978410206084,
+      "grad_norm": 1.030502438545227,
+      "learning_rate": 6.973170670184226e-05,
+      "loss": 0.0568,
+      "step": 72420
+    },
+    {
+      "epoch": 4.738632646385345,
+      "grad_norm": 0.7706143856048584,
+      "learning_rate": 6.972326597185136e-05,
+      "loss": 0.0571,
+      "step": 72430
+    },
+    {
+      "epoch": 4.739286882564606,
+      "grad_norm": 0.8049507737159729,
+      "learning_rate": 6.971482457614848e-05,
+      "loss": 0.0619,
+      "step": 72440
+    },
+    {
+      "epoch": 4.739941118743866,
+      "grad_norm": 0.9160942435264587,
+      "learning_rate": 6.970638251501859e-05,
+      "loss": 0.0699,
+      "step": 72450
+    },
+    {
+      "epoch": 4.740595354923127,
+      "grad_norm": 0.8005819916725159,
+      "learning_rate": 6.96979397887466e-05,
+      "loss": 0.0572,
+      "step": 72460
+    },
+    {
+      "epoch": 4.741249591102388,
+      "grad_norm": 0.9010353088378906,
+      "learning_rate": 6.96894963976175e-05,
+      "loss": 0.06,
+      "step": 72470
+    },
+    {
+      "epoch": 4.741903827281648,
+      "grad_norm": 0.8157947659492493,
+      "learning_rate": 6.968105234191623e-05,
+      "loss": 0.0609,
+      "step": 72480
+    },
+    {
+      "epoch": 4.742558063460909,
+      "grad_norm": 0.968302309513092,
+      "learning_rate": 6.967260762192785e-05,
+      "loss": 0.0542,
+      "step": 72490
+    },
+    {
+      "epoch": 4.74321229964017,
+      "grad_norm": 1.0909911394119263,
+      "learning_rate": 6.966416223793736e-05,
+      "loss": 0.0624,
+      "step": 72500
+    },
+    {
+      "epoch": 4.743866535819431,
+      "grad_norm": 0.7629793286323547,
+      "learning_rate": 6.965571619022981e-05,
+      "loss": 0.0592,
+      "step": 72510
+    },
+    {
+      "epoch": 4.744520771998691,
+      "grad_norm": 0.7732744216918945,
+      "learning_rate": 6.964726947909031e-05,
+      "loss": 0.0632,
+      "step": 72520
+    },
+    {
+      "epoch": 4.745175008177952,
+      "grad_norm": 0.8113254904747009,
+      "learning_rate": 6.963882210480394e-05,
+      "loss": 0.0632,
+      "step": 72530
+    },
+    {
+      "epoch": 4.745829244357213,
+      "grad_norm": 1.129671573638916,
+      "learning_rate": 6.963037406765581e-05,
+      "loss": 0.0577,
+      "step": 72540
+    },
+    {
+      "epoch": 4.746483480536473,
+      "grad_norm": 0.8630444407463074,
+      "learning_rate": 6.962192536793106e-05,
+      "loss": 0.0709,
+      "step": 72550
+    },
+    {
+      "epoch": 4.747137716715734,
+      "grad_norm": 0.8324365019798279,
+      "learning_rate": 6.961347600591489e-05,
+      "loss": 0.057,
+      "step": 72560
+    },
+    {
+      "epoch": 4.747791952894995,
+      "grad_norm": 0.919032633304596,
+      "learning_rate": 6.960502598189245e-05,
+      "loss": 0.0581,
+      "step": 72570
+    },
+    {
+      "epoch": 4.748446189074256,
+      "grad_norm": 1.0946227312088013,
+      "learning_rate": 6.959657529614898e-05,
+      "loss": 0.0643,
+      "step": 72580
+    },
+    {
+      "epoch": 4.749100425253516,
+      "grad_norm": 0.8766310214996338,
+      "learning_rate": 6.95881239489697e-05,
+      "loss": 0.0577,
+      "step": 72590
+    },
+    {
+      "epoch": 4.749754661432777,
+      "grad_norm": 0.9453266859054565,
+      "learning_rate": 6.957967194063985e-05,
+      "loss": 0.0631,
+      "step": 72600
+    },
+    {
+      "epoch": 4.750408897612038,
+      "grad_norm": 0.74550861120224,
+      "learning_rate": 6.957121927144474e-05,
+      "loss": 0.0627,
+      "step": 72610
+    },
+    {
+      "epoch": 4.751063133791298,
+      "grad_norm": 0.7826007008552551,
+      "learning_rate": 6.956276594166964e-05,
+      "loss": 0.0572,
+      "step": 72620
+    },
+    {
+      "epoch": 4.751717369970559,
+      "grad_norm": 0.9839434027671814,
+      "learning_rate": 6.955431195159989e-05,
+      "loss": 0.0724,
+      "step": 72630
+    },
+    {
+      "epoch": 4.75237160614982,
+      "grad_norm": 0.808029294013977,
+      "learning_rate": 6.954585730152083e-05,
+      "loss": 0.0634,
+      "step": 72640
+    },
+    {
+      "epoch": 4.75302584232908,
+      "grad_norm": 0.935702383518219,
+      "learning_rate": 6.953740199171782e-05,
+      "loss": 0.0734,
+      "step": 72650
+    },
+    {
+      "epoch": 4.753680078508341,
+      "grad_norm": 1.2413403987884521,
+      "learning_rate": 6.952894602247626e-05,
+      "loss": 0.0655,
+      "step": 72660
+    },
+    {
+      "epoch": 4.754334314687602,
+      "grad_norm": 1.0404640436172485,
+      "learning_rate": 6.952048939408156e-05,
+      "loss": 0.0731,
+      "step": 72670
+    },
+    {
+      "epoch": 4.754988550866863,
+      "grad_norm": 0.900425136089325,
+      "learning_rate": 6.951203210681914e-05,
+      "loss": 0.0559,
+      "step": 72680
+    },
+    {
+      "epoch": 4.755642787046123,
+      "grad_norm": 0.9717649817466736,
+      "learning_rate": 6.950357416097446e-05,
+      "loss": 0.0594,
+      "step": 72690
+    },
+    {
+      "epoch": 4.756297023225384,
+      "grad_norm": 1.1071817874908447,
+      "learning_rate": 6.949511555683301e-05,
+      "loss": 0.0638,
+      "step": 72700
+    },
+    {
+      "epoch": 4.756951259404645,
+      "grad_norm": 0.9194473028182983,
+      "learning_rate": 6.948665629468027e-05,
+      "loss": 0.0655,
+      "step": 72710
+    },
+    {
+      "epoch": 4.757605495583906,
+      "grad_norm": 0.7610183358192444,
+      "learning_rate": 6.94781963748018e-05,
+      "loss": 0.0632,
+      "step": 72720
+    },
+    {
+      "epoch": 4.758259731763166,
+      "grad_norm": 0.8666313886642456,
+      "learning_rate": 6.946973579748309e-05,
+      "loss": 0.0659,
+      "step": 72730
+    },
+    {
+      "epoch": 4.758913967942427,
+      "grad_norm": 0.8880037665367126,
+      "learning_rate": 6.946127456300974e-05,
+      "loss": 0.062,
+      "step": 72740
+    },
+    {
+      "epoch": 4.759568204121688,
+      "grad_norm": 0.8450115323066711,
+      "learning_rate": 6.945281267166736e-05,
+      "loss": 0.072,
+      "step": 72750
+    },
+    {
+      "epoch": 4.760222440300948,
+      "grad_norm": 1.100077509880066,
+      "learning_rate": 6.94443501237415e-05,
+      "loss": 0.0661,
+      "step": 72760
+    },
+    {
+      "epoch": 4.760876676480209,
+      "grad_norm": 0.956047534942627,
+      "learning_rate": 6.943588691951785e-05,
+      "loss": 0.0607,
+      "step": 72770
+    },
+    {
+      "epoch": 4.76153091265947,
+      "grad_norm": 0.7662896513938904,
+      "learning_rate": 6.942742305928205e-05,
+      "loss": 0.0627,
+      "step": 72780
+    },
+    {
+      "epoch": 4.76218514883873,
+      "grad_norm": 0.9403966069221497,
+      "learning_rate": 6.941895854331977e-05,
+      "loss": 0.0646,
+      "step": 72790
+    },
+    {
+      "epoch": 4.762839385017991,
+      "grad_norm": 0.918593168258667,
+      "learning_rate": 6.94104933719167e-05,
+      "loss": 0.0521,
+      "step": 72800
+    },
+    {
+      "epoch": 4.763493621197252,
+      "grad_norm": 0.8933160901069641,
+      "learning_rate": 6.940202754535856e-05,
+      "loss": 0.0629,
+      "step": 72810
+    },
+    {
+      "epoch": 4.764147857376513,
+      "grad_norm": 0.8830978274345398,
+      "learning_rate": 6.939356106393113e-05,
+      "loss": 0.059,
+      "step": 72820
+    },
+    {
+      "epoch": 4.764802093555773,
+      "grad_norm": 0.9082246422767639,
+      "learning_rate": 6.938509392792016e-05,
+      "loss": 0.057,
+      "step": 72830
+    },
+    {
+      "epoch": 4.765456329735034,
+      "grad_norm": 1.0356862545013428,
+      "learning_rate": 6.93766261376114e-05,
+      "loss": 0.0589,
+      "step": 72840
+    },
+    {
+      "epoch": 4.766110565914295,
+      "grad_norm": 0.8933234810829163,
+      "learning_rate": 6.936815769329071e-05,
+      "loss": 0.0596,
+      "step": 72850
+    },
+    {
+      "epoch": 4.766764802093556,
+      "grad_norm": 0.8427624106407166,
+      "learning_rate": 6.935968859524389e-05,
+      "loss": 0.0563,
+      "step": 72860
+    },
+    {
+      "epoch": 4.767419038272816,
+      "grad_norm": 0.8006119728088379,
+      "learning_rate": 6.935121884375683e-05,
+      "loss": 0.0591,
+      "step": 72870
+    },
+    {
+      "epoch": 4.768073274452077,
+      "grad_norm": 0.7767458558082581,
+      "learning_rate": 6.934274843911537e-05,
+      "loss": 0.0603,
+      "step": 72880
+    },
+    {
+      "epoch": 4.768727510631338,
+      "grad_norm": 0.8365658521652222,
+      "learning_rate": 6.933427738160542e-05,
+      "loss": 0.0648,
+      "step": 72890
+    },
+    {
+      "epoch": 4.769381746810598,
+      "grad_norm": 0.9282712340354919,
+      "learning_rate": 6.93258056715129e-05,
+      "loss": 0.0722,
+      "step": 72900
+    },
+    {
+      "epoch": 4.770035982989859,
+      "grad_norm": 0.7623592615127563,
+      "learning_rate": 6.931733330912375e-05,
+      "loss": 0.0618,
+      "step": 72910
+    },
+    {
+      "epoch": 4.77069021916912,
+      "grad_norm": 0.8280691504478455,
+      "learning_rate": 6.930886029472396e-05,
+      "loss": 0.0585,
+      "step": 72920
+    },
+    {
+      "epoch": 4.77134445534838,
+      "grad_norm": 0.8656930923461914,
+      "learning_rate": 6.930038662859947e-05,
+      "loss": 0.06,
+      "step": 72930
+    },
+    {
+      "epoch": 4.771998691527641,
+      "grad_norm": 0.8957453370094299,
+      "learning_rate": 6.929191231103634e-05,
+      "loss": 0.0626,
+      "step": 72940
+    },
+    {
+      "epoch": 4.772652927706902,
+      "grad_norm": 0.8083595037460327,
+      "learning_rate": 6.928343734232057e-05,
+      "loss": 0.0679,
+      "step": 72950
+    },
+    {
+      "epoch": 4.773307163886163,
+      "grad_norm": 0.9139804244041443,
+      "learning_rate": 6.92749617227382e-05,
+      "loss": 0.0612,
+      "step": 72960
+    },
+    {
+      "epoch": 4.773961400065423,
+      "grad_norm": 0.9249205589294434,
+      "learning_rate": 6.926648545257534e-05,
+      "loss": 0.0599,
+      "step": 72970
+    },
+    {
+      "epoch": 4.774615636244684,
+      "grad_norm": 0.9592366218566895,
+      "learning_rate": 6.925800853211807e-05,
+      "loss": 0.054,
+      "step": 72980
+    },
+    {
+      "epoch": 4.775269872423945,
+      "grad_norm": 0.9399287700653076,
+      "learning_rate": 6.924953096165248e-05,
+      "loss": 0.0585,
+      "step": 72990
+    },
+    {
+      "epoch": 4.775924108603205,
+      "grad_norm": 0.8287541270256042,
+      "learning_rate": 6.924105274146476e-05,
+      "loss": 0.0568,
+      "step": 73000
+    },
+    {
+      "epoch": 4.776578344782466,
+      "grad_norm": 0.6379443407058716,
+      "learning_rate": 6.923257387184103e-05,
+      "loss": 0.0574,
+      "step": 73010
+    },
+    {
+      "epoch": 4.777232580961727,
+      "grad_norm": 0.7155212163925171,
+      "learning_rate": 6.922409435306751e-05,
+      "loss": 0.0646,
+      "step": 73020
+    },
+    {
+      "epoch": 4.777886817140988,
+      "grad_norm": 0.8226625919342041,
+      "learning_rate": 6.921561418543037e-05,
+      "loss": 0.0576,
+      "step": 73030
+    },
+    {
+      "epoch": 4.778541053320248,
+      "grad_norm": 0.8837935328483582,
+      "learning_rate": 6.920713336921588e-05,
+      "loss": 0.0607,
+      "step": 73040
+    },
+    {
+      "epoch": 4.779195289499509,
+      "grad_norm": 0.8143121004104614,
+      "learning_rate": 6.919865190471026e-05,
+      "loss": 0.0559,
+      "step": 73050
+    },
+    {
+      "epoch": 4.77984952567877,
+      "grad_norm": 0.7172751426696777,
+      "learning_rate": 6.919016979219978e-05,
+      "loss": 0.052,
+      "step": 73060
+    },
+    {
+      "epoch": 4.78050376185803,
+      "grad_norm": 0.9022568464279175,
+      "learning_rate": 6.918168703197073e-05,
+      "loss": 0.0589,
+      "step": 73070
+    },
+    {
+      "epoch": 4.781157998037291,
+      "grad_norm": 0.9577611684799194,
+      "learning_rate": 6.917320362430945e-05,
+      "loss": 0.0654,
+      "step": 73080
+    },
+    {
+      "epoch": 4.781812234216552,
+      "grad_norm": 0.7758936285972595,
+      "learning_rate": 6.916471956950228e-05,
+      "loss": 0.0634,
+      "step": 73090
+    },
+    {
+      "epoch": 4.782466470395812,
+      "grad_norm": 0.9259306192398071,
+      "learning_rate": 6.915623486783555e-05,
+      "loss": 0.064,
+      "step": 73100
+    },
+    {
+      "epoch": 4.783120706575073,
+      "grad_norm": 0.9896383285522461,
+      "learning_rate": 6.914774951959565e-05,
+      "loss": 0.0581,
+      "step": 73110
+    },
+    {
+      "epoch": 4.783774942754334,
+      "grad_norm": 0.8976383209228516,
+      "learning_rate": 6.913926352506898e-05,
+      "loss": 0.0583,
+      "step": 73120
+    },
+    {
+      "epoch": 4.784429178933595,
+      "grad_norm": 0.9846516251564026,
+      "learning_rate": 6.913077688454198e-05,
+      "loss": 0.0639,
+      "step": 73130
+    },
+    {
+      "epoch": 4.785083415112855,
+      "grad_norm": 0.7982523441314697,
+      "learning_rate": 6.912228959830109e-05,
+      "loss": 0.062,
+      "step": 73140
+    },
+    {
+      "epoch": 4.785737651292116,
+      "grad_norm": 0.7241307497024536,
+      "learning_rate": 6.911380166663278e-05,
+      "loss": 0.0607,
+      "step": 73150
+    },
+    {
+      "epoch": 4.786391887471377,
+      "grad_norm": 0.7322345972061157,
+      "learning_rate": 6.910531308982353e-05,
+      "loss": 0.0645,
+      "step": 73160
+    },
+    {
+      "epoch": 4.787046123650638,
+      "grad_norm": 0.8883731961250305,
+      "learning_rate": 6.909682386815987e-05,
+      "loss": 0.0642,
+      "step": 73170
+    },
+    {
+      "epoch": 4.787700359829898,
+      "grad_norm": 0.8451051115989685,
+      "learning_rate": 6.908833400192829e-05,
+      "loss": 0.0635,
+      "step": 73180
+    },
+    {
+      "epoch": 4.788354596009159,
+      "grad_norm": 1.008467435836792,
+      "learning_rate": 6.90798434914154e-05,
+      "loss": 0.0643,
+      "step": 73190
+    },
+    {
+      "epoch": 4.78900883218842,
+      "grad_norm": 0.964712381362915,
+      "learning_rate": 6.907135233690774e-05,
+      "loss": 0.0602,
+      "step": 73200
+    },
+    {
+      "epoch": 4.78966306836768,
+      "grad_norm": 0.9574010968208313,
+      "learning_rate": 6.906286053869194e-05,
+      "loss": 0.0586,
+      "step": 73210
+    },
+    {
+      "epoch": 4.790317304546941,
+      "grad_norm": 0.9692860245704651,
+      "learning_rate": 6.90543680970546e-05,
+      "loss": 0.0646,
+      "step": 73220
+    },
+    {
+      "epoch": 4.790971540726202,
+      "grad_norm": 0.8585093021392822,
+      "learning_rate": 6.904587501228236e-05,
+      "loss": 0.0654,
+      "step": 73230
+    },
+    {
+      "epoch": 4.791625776905462,
+      "grad_norm": 0.9351073503494263,
+      "learning_rate": 6.903738128466188e-05,
+      "loss": 0.0553,
+      "step": 73240
+    },
+    {
+      "epoch": 4.792280013084723,
+      "grad_norm": 0.7714135050773621,
+      "learning_rate": 6.902888691447986e-05,
+      "loss": 0.0557,
+      "step": 73250
+    },
+    {
+      "epoch": 4.792934249263984,
+      "grad_norm": 0.977641224861145,
+      "learning_rate": 6.9020391902023e-05,
+      "loss": 0.0522,
+      "step": 73260
+    },
+    {
+      "epoch": 4.793588485443245,
+      "grad_norm": 1.0779730081558228,
+      "learning_rate": 6.901189624757803e-05,
+      "loss": 0.0606,
+      "step": 73270
+    },
+    {
+      "epoch": 4.794242721622505,
+      "grad_norm": 0.9080124497413635,
+      "learning_rate": 6.900339995143172e-05,
+      "loss": 0.0553,
+      "step": 73280
+    },
+    {
+      "epoch": 4.794896957801766,
+      "grad_norm": 0.8090987801551819,
+      "learning_rate": 6.899490301387079e-05,
+      "loss": 0.0565,
+      "step": 73290
+    },
+    {
+      "epoch": 4.795551193981027,
+      "grad_norm": 1.0443229675292969,
+      "learning_rate": 6.89864054351821e-05,
+      "loss": 0.0588,
+      "step": 73300
+    },
+    {
+      "epoch": 4.796205430160288,
+      "grad_norm": 0.7326402068138123,
+      "learning_rate": 6.897790721565243e-05,
+      "loss": 0.0569,
+      "step": 73310
+    },
+    {
+      "epoch": 4.796859666339548,
+      "grad_norm": 0.8527650237083435,
+      "learning_rate": 6.89694083555686e-05,
+      "loss": 0.0687,
+      "step": 73320
+    },
+    {
+      "epoch": 4.797513902518809,
+      "grad_norm": 1.0699238777160645,
+      "learning_rate": 6.896090885521749e-05,
+      "loss": 0.0616,
+      "step": 73330
+    },
+    {
+      "epoch": 4.79816813869807,
+      "grad_norm": 0.9271957874298096,
+      "learning_rate": 6.895240871488599e-05,
+      "loss": 0.062,
+      "step": 73340
+    },
+    {
+      "epoch": 4.79882237487733,
+      "grad_norm": 0.6596797704696655,
+      "learning_rate": 6.894390793486098e-05,
+      "loss": 0.0538,
+      "step": 73350
+    },
+    {
+      "epoch": 4.799476611056591,
+      "grad_norm": 0.8873812556266785,
+      "learning_rate": 6.89354065154294e-05,
+      "loss": 0.0678,
+      "step": 73360
+    },
+    {
+      "epoch": 4.800130847235852,
+      "grad_norm": 1.0261344909667969,
+      "learning_rate": 6.89269044568782e-05,
+      "loss": 0.0636,
+      "step": 73370
+    },
+    {
+      "epoch": 4.8007850834151125,
+      "grad_norm": 0.9206944108009338,
+      "learning_rate": 6.891840175949432e-05,
+      "loss": 0.0535,
+      "step": 73380
+    },
+    {
+      "epoch": 4.801439319594373,
+      "grad_norm": 0.9742897152900696,
+      "learning_rate": 6.890989842356479e-05,
+      "loss": 0.0599,
+      "step": 73390
+    },
+    {
+      "epoch": 4.802093555773634,
+      "grad_norm": 0.8803852796554565,
+      "learning_rate": 6.890139444937657e-05,
+      "loss": 0.066,
+      "step": 73400
+    },
+    {
+      "epoch": 4.802747791952895,
+      "grad_norm": 1.0795328617095947,
+      "learning_rate": 6.889288983721673e-05,
+      "loss": 0.0587,
+      "step": 73410
+    },
+    {
+      "epoch": 4.803402028132155,
+      "grad_norm": 0.8851546049118042,
+      "learning_rate": 6.888438458737232e-05,
+      "loss": 0.0598,
+      "step": 73420
+    },
+    {
+      "epoch": 4.804056264311416,
+      "grad_norm": 0.9637795686721802,
+      "learning_rate": 6.887587870013039e-05,
+      "loss": 0.0534,
+      "step": 73430
+    },
+    {
+      "epoch": 4.804710500490677,
+      "grad_norm": 1.036117434501648,
+      "learning_rate": 6.886737217577805e-05,
+      "loss": 0.0589,
+      "step": 73440
+    },
+    {
+      "epoch": 4.8053647366699375,
+      "grad_norm": 0.8892022967338562,
+      "learning_rate": 6.88588650146024e-05,
+      "loss": 0.058,
+      "step": 73450
+    },
+    {
+      "epoch": 4.806018972849198,
+      "grad_norm": 0.7642195224761963,
+      "learning_rate": 6.885035721689062e-05,
+      "loss": 0.0707,
+      "step": 73460
+    },
+    {
+      "epoch": 4.806673209028459,
+      "grad_norm": 0.8952467441558838,
+      "learning_rate": 6.884184878292985e-05,
+      "loss": 0.0618,
+      "step": 73470
+    },
+    {
+      "epoch": 4.80732744520772,
+      "grad_norm": 0.7745544910430908,
+      "learning_rate": 6.883333971300725e-05,
+      "loss": 0.0601,
+      "step": 73480
+    },
+    {
+      "epoch": 4.8079816813869805,
+      "grad_norm": 0.8764887452125549,
+      "learning_rate": 6.882483000741007e-05,
+      "loss": 0.0614,
+      "step": 73490
+    },
+    {
+      "epoch": 4.808635917566241,
+      "grad_norm": 0.8772285580635071,
+      "learning_rate": 6.881631966642549e-05,
+      "loss": 0.0532,
+      "step": 73500
+    },
+    {
+      "epoch": 4.809290153745502,
+      "grad_norm": 1.0648410320281982,
+      "learning_rate": 6.880780869034077e-05,
+      "loss": 0.0655,
+      "step": 73510
+    },
+    {
+      "epoch": 4.8099443899247625,
+      "grad_norm": 1.125880479812622,
+      "learning_rate": 6.879929707944317e-05,
+      "loss": 0.0671,
+      "step": 73520
+    },
+    {
+      "epoch": 4.8105986261040234,
+      "grad_norm": 0.867889404296875,
+      "learning_rate": 6.879078483402002e-05,
+      "loss": 0.0512,
+      "step": 73530
+    },
+    {
+      "epoch": 4.811252862283284,
+      "grad_norm": 0.8720964789390564,
+      "learning_rate": 6.878227195435859e-05,
+      "loss": 0.0601,
+      "step": 73540
+    },
+    {
+      "epoch": 4.8119070984625445,
+      "grad_norm": 0.8965845108032227,
+      "learning_rate": 6.877375844074622e-05,
+      "loss": 0.0582,
+      "step": 73550
+    },
+    {
+      "epoch": 4.8125613346418055,
+      "grad_norm": 0.7062650322914124,
+      "learning_rate": 6.876524429347027e-05,
+      "loss": 0.0674,
+      "step": 73560
+    },
+    {
+      "epoch": 4.813215570821066,
+      "grad_norm": 1.128409743309021,
+      "learning_rate": 6.87567295128181e-05,
+      "loss": 0.0603,
+      "step": 73570
+    },
+    {
+      "epoch": 4.813869807000327,
+      "grad_norm": 0.952961802482605,
+      "learning_rate": 6.874821409907713e-05,
+      "loss": 0.0563,
+      "step": 73580
+    },
+    {
+      "epoch": 4.8145240431795875,
+      "grad_norm": 0.8573753833770752,
+      "learning_rate": 6.873969805253477e-05,
+      "loss": 0.06,
+      "step": 73590
+    },
+    {
+      "epoch": 4.8151782793588485,
+      "grad_norm": 0.8490943312644958,
+      "learning_rate": 6.873118137347844e-05,
+      "loss": 0.0584,
+      "step": 73600
+    },
+    {
+      "epoch": 4.815832515538109,
+      "grad_norm": 0.9143913388252258,
+      "learning_rate": 6.872266406219562e-05,
+      "loss": 0.0739,
+      "step": 73610
+    },
+    {
+      "epoch": 4.81648675171737,
+      "grad_norm": 0.973491370677948,
+      "learning_rate": 6.871414611897379e-05,
+      "loss": 0.0572,
+      "step": 73620
+    },
+    {
+      "epoch": 4.8171409878966305,
+      "grad_norm": 0.913781464099884,
+      "learning_rate": 6.870562754410044e-05,
+      "loss": 0.0646,
+      "step": 73630
+    },
+    {
+      "epoch": 4.8177952240758914,
+      "grad_norm": 1.1292165517807007,
+      "learning_rate": 6.86971083378631e-05,
+      "loss": 0.0538,
+      "step": 73640
+    },
+    {
+      "epoch": 4.818449460255152,
+      "grad_norm": 0.815707802772522,
+      "learning_rate": 6.868858850054933e-05,
+      "loss": 0.0577,
+      "step": 73650
+    },
+    {
+      "epoch": 4.8191036964344125,
+      "grad_norm": 1.1895596981048584,
+      "learning_rate": 6.868006803244669e-05,
+      "loss": 0.0534,
+      "step": 73660
+    },
+    {
+      "epoch": 4.8197579326136735,
+      "grad_norm": 0.672910749912262,
+      "learning_rate": 6.867154693384275e-05,
+      "loss": 0.053,
+      "step": 73670
+    },
+    {
+      "epoch": 4.820412168792934,
+      "grad_norm": 1.2378745079040527,
+      "learning_rate": 6.866302520502515e-05,
+      "loss": 0.0745,
+      "step": 73680
+    },
+    {
+      "epoch": 4.8210664049721945,
+      "grad_norm": 1.0225582122802734,
+      "learning_rate": 6.865450284628148e-05,
+      "loss": 0.0637,
+      "step": 73690
+    },
+    {
+      "epoch": 4.8217206411514555,
+      "grad_norm": 0.8776040077209473,
+      "learning_rate": 6.864597985789944e-05,
+      "loss": 0.0664,
+      "step": 73700
+    },
+    {
+      "epoch": 4.8223748773307165,
+      "grad_norm": 0.7907946705818176,
+      "learning_rate": 6.863745624016666e-05,
+      "loss": 0.078,
+      "step": 73710
+    },
+    {
+      "epoch": 4.823029113509977,
+      "grad_norm": 0.7620204091072083,
+      "learning_rate": 6.862893199337087e-05,
+      "loss": 0.0517,
+      "step": 73720
+    },
+    {
+      "epoch": 4.8236833496892375,
+      "grad_norm": 0.7737089991569519,
+      "learning_rate": 6.862040711779976e-05,
+      "loss": 0.0583,
+      "step": 73730
+    },
+    {
+      "epoch": 4.8243375858684985,
+      "grad_norm": 0.7591310739517212,
+      "learning_rate": 6.861188161374106e-05,
+      "loss": 0.0602,
+      "step": 73740
+    },
+    {
+      "epoch": 4.8249918220477594,
+      "grad_norm": 0.8480481505393982,
+      "learning_rate": 6.860335548148257e-05,
+      "loss": 0.0603,
+      "step": 73750
+    },
+    {
+      "epoch": 4.82564605822702,
+      "grad_norm": 0.8201214671134949,
+      "learning_rate": 6.859482872131203e-05,
+      "loss": 0.0617,
+      "step": 73760
+    },
+    {
+      "epoch": 4.8263002944062805,
+      "grad_norm": 1.0084785223007202,
+      "learning_rate": 6.858630133351726e-05,
+      "loss": 0.0647,
+      "step": 73770
+    },
+    {
+      "epoch": 4.8269545305855415,
+      "grad_norm": 0.8590355515480042,
+      "learning_rate": 6.857777331838607e-05,
+      "loss": 0.0621,
+      "step": 73780
+    },
+    {
+      "epoch": 4.827608766764802,
+      "grad_norm": 1.0885320901870728,
+      "learning_rate": 6.856924467620631e-05,
+      "loss": 0.0556,
+      "step": 73790
+    },
+    {
+      "epoch": 4.8282630029440625,
+      "grad_norm": 0.7727965116500854,
+      "learning_rate": 6.856071540726585e-05,
+      "loss": 0.0552,
+      "step": 73800
+    },
+    {
+      "epoch": 4.8289172391233235,
+      "grad_norm": 0.9133251309394836,
+      "learning_rate": 6.855218551185255e-05,
+      "loss": 0.0601,
+      "step": 73810
+    },
+    {
+      "epoch": 4.8295714753025845,
+      "grad_norm": 0.782835841178894,
+      "learning_rate": 6.854365499025435e-05,
+      "loss": 0.0563,
+      "step": 73820
+    },
+    {
+      "epoch": 4.8302257114818445,
+      "grad_norm": 1.1449490785598755,
+      "learning_rate": 6.853512384275916e-05,
+      "loss": 0.0622,
+      "step": 73830
+    },
+    {
+      "epoch": 4.8308799476611055,
+      "grad_norm": 0.7554023265838623,
+      "learning_rate": 6.852659206965493e-05,
+      "loss": 0.0579,
+      "step": 73840
+    },
+    {
+      "epoch": 4.8315341838403665,
+      "grad_norm": 1.0752381086349487,
+      "learning_rate": 6.851805967122962e-05,
+      "loss": 0.057,
+      "step": 73850
+    },
+    {
+      "epoch": 4.8321884200196275,
+      "grad_norm": 0.9013246893882751,
+      "learning_rate": 6.850952664777124e-05,
+      "loss": 0.0545,
+      "step": 73860
+    },
+    {
+      "epoch": 4.8328426561988875,
+      "grad_norm": 0.8916118741035461,
+      "learning_rate": 6.85009929995678e-05,
+      "loss": 0.0527,
+      "step": 73870
+    },
+    {
+      "epoch": 4.8334968923781485,
+      "grad_norm": 0.8229614496231079,
+      "learning_rate": 6.849245872690731e-05,
+      "loss": 0.0666,
+      "step": 73880
+    },
+    {
+      "epoch": 4.8341511285574095,
+      "grad_norm": 0.7096636891365051,
+      "learning_rate": 6.848392383007784e-05,
+      "loss": 0.0589,
+      "step": 73890
+    },
+    {
+      "epoch": 4.8348053647366696,
+      "grad_norm": 0.8488548398017883,
+      "learning_rate": 6.847538830936746e-05,
+      "loss": 0.0672,
+      "step": 73900
+    },
+    {
+      "epoch": 4.8354596009159305,
+      "grad_norm": 0.8226892948150635,
+      "learning_rate": 6.84668521650643e-05,
+      "loss": 0.0587,
+      "step": 73910
+    },
+    {
+      "epoch": 4.8361138370951915,
+      "grad_norm": 0.7988223433494568,
+      "learning_rate": 6.845831539745643e-05,
+      "loss": 0.0618,
+      "step": 73920
+    },
+    {
+      "epoch": 4.8367680732744525,
+      "grad_norm": 0.9851675033569336,
+      "learning_rate": 6.8449778006832e-05,
+      "loss": 0.057,
+      "step": 73930
+    },
+    {
+      "epoch": 4.8374223094537125,
+      "grad_norm": 1.0461887121200562,
+      "learning_rate": 6.84412399934792e-05,
+      "loss": 0.0602,
+      "step": 73940
+    },
+    {
+      "epoch": 4.8380765456329735,
+      "grad_norm": 1.085433006286621,
+      "learning_rate": 6.843270135768616e-05,
+      "loss": 0.0609,
+      "step": 73950
+    },
+    {
+      "epoch": 4.8387307818122345,
+      "grad_norm": 0.9901970028877258,
+      "learning_rate": 6.842416209974111e-05,
+      "loss": 0.0498,
+      "step": 73960
+    },
+    {
+      "epoch": 4.839385017991495,
+      "grad_norm": 0.7033235430717468,
+      "learning_rate": 6.841562221993228e-05,
+      "loss": 0.0648,
+      "step": 73970
+    },
+    {
+      "epoch": 4.8400392541707555,
+      "grad_norm": 0.8223658800125122,
+      "learning_rate": 6.84070817185479e-05,
+      "loss": 0.0568,
+      "step": 73980
+    },
+    {
+      "epoch": 4.8406934903500165,
+      "grad_norm": 0.8278485536575317,
+      "learning_rate": 6.839854059587624e-05,
+      "loss": 0.062,
+      "step": 73990
+    },
+    {
+      "epoch": 4.841347726529277,
+      "grad_norm": 0.8352931141853333,
+      "learning_rate": 6.838999885220558e-05,
+      "loss": 0.0645,
+      "step": 74000
+    },
+    {
+      "epoch": 4.8420019627085376,
+      "grad_norm": 1.1042554378509521,
+      "learning_rate": 6.838145648782422e-05,
+      "loss": 0.0612,
+      "step": 74010
+    },
+    {
+      "epoch": 4.8426561988877985,
+      "grad_norm": 0.8824191689491272,
+      "learning_rate": 6.837291350302052e-05,
+      "loss": 0.0611,
+      "step": 74020
+    },
+    {
+      "epoch": 4.8433104350670595,
+      "grad_norm": 0.9088771939277649,
+      "learning_rate": 6.836436989808278e-05,
+      "loss": 0.0644,
+      "step": 74030
+    },
+    {
+      "epoch": 4.84396467124632,
+      "grad_norm": 0.8344402313232422,
+      "learning_rate": 6.835582567329942e-05,
+      "loss": 0.0568,
+      "step": 74040
+    },
+    {
+      "epoch": 4.8446189074255805,
+      "grad_norm": 0.8632908463478088,
+      "learning_rate": 6.834728082895878e-05,
+      "loss": 0.0596,
+      "step": 74050
+    },
+    {
+      "epoch": 4.8452731436048415,
+      "grad_norm": 0.7749559879302979,
+      "learning_rate": 6.83387353653493e-05,
+      "loss": 0.062,
+      "step": 74060
+    },
+    {
+      "epoch": 4.8459273797841025,
+      "grad_norm": 0.9213570952415466,
+      "learning_rate": 6.83301892827594e-05,
+      "loss": 0.0571,
+      "step": 74070
+    },
+    {
+      "epoch": 4.846581615963363,
+      "grad_norm": 0.7469812035560608,
+      "learning_rate": 6.832164258147756e-05,
+      "loss": 0.0528,
+      "step": 74080
+    },
+    {
+      "epoch": 4.8472358521426235,
+      "grad_norm": 0.6800960898399353,
+      "learning_rate": 6.831309526179221e-05,
+      "loss": 0.063,
+      "step": 74090
+    },
+    {
+      "epoch": 4.8478900883218845,
+      "grad_norm": 0.8719648122787476,
+      "learning_rate": 6.830454732399188e-05,
+      "loss": 0.0532,
+      "step": 74100
+    },
+    {
+      "epoch": 4.848544324501145,
+      "grad_norm": 0.9584552645683289,
+      "learning_rate": 6.829599876836507e-05,
+      "loss": 0.0636,
+      "step": 74110
+    },
+    {
+      "epoch": 4.849198560680406,
+      "grad_norm": 0.9922688603401184,
+      "learning_rate": 6.828744959520031e-05,
+      "loss": 0.0601,
+      "step": 74120
+    },
+    {
+      "epoch": 4.8498527968596665,
+      "grad_norm": 0.8676353693008423,
+      "learning_rate": 6.82788998047862e-05,
+      "loss": 0.0577,
+      "step": 74130
+    },
+    {
+      "epoch": 4.850507033038927,
+      "grad_norm": 0.9001226425170898,
+      "learning_rate": 6.827034939741125e-05,
+      "loss": 0.0597,
+      "step": 74140
+    },
+    {
+      "epoch": 4.851161269218188,
+      "grad_norm": 0.8655483722686768,
+      "learning_rate": 6.82617983733641e-05,
+      "loss": 0.0614,
+      "step": 74150
+    },
+    {
+      "epoch": 4.8518155053974485,
+      "grad_norm": 0.9211030602455139,
+      "learning_rate": 6.825324673293336e-05,
+      "loss": 0.0545,
+      "step": 74160
+    },
+    {
+      "epoch": 4.8524697415767095,
+      "grad_norm": 0.8237171173095703,
+      "learning_rate": 6.824469447640766e-05,
+      "loss": 0.0544,
+      "step": 74170
+    },
+    {
+      "epoch": 4.85312397775597,
+      "grad_norm": 1.0644043684005737,
+      "learning_rate": 6.82361416040757e-05,
+      "loss": 0.0702,
+      "step": 74180
+    },
+    {
+      "epoch": 4.853778213935231,
+      "grad_norm": 0.8750390410423279,
+      "learning_rate": 6.82275881162261e-05,
+      "loss": 0.0582,
+      "step": 74190
+    },
+    {
+      "epoch": 4.8544324501144915,
+      "grad_norm": 0.9315994381904602,
+      "learning_rate": 6.821903401314764e-05,
+      "loss": 0.0555,
+      "step": 74200
+    },
+    {
+      "epoch": 4.8550866862937525,
+      "grad_norm": 1.0597717761993408,
+      "learning_rate": 6.821047929512898e-05,
+      "loss": 0.0574,
+      "step": 74210
+    },
+    {
+      "epoch": 4.855740922473013,
+      "grad_norm": 0.8364295959472656,
+      "learning_rate": 6.820192396245886e-05,
+      "loss": 0.0573,
+      "step": 74220
+    },
+    {
+      "epoch": 4.856395158652274,
+      "grad_norm": 0.7751799821853638,
+      "learning_rate": 6.81933680154261e-05,
+      "loss": 0.0636,
+      "step": 74230
+    },
+    {
+      "epoch": 4.8570493948315345,
+      "grad_norm": 0.9654813408851624,
+      "learning_rate": 6.818481145431947e-05,
+      "loss": 0.0612,
+      "step": 74240
+    },
+    {
+      "epoch": 4.857703631010795,
+      "grad_norm": 0.922402024269104,
+      "learning_rate": 6.817625427942773e-05,
+      "loss": 0.058,
+      "step": 74250
+    },
+    {
+      "epoch": 4.858357867190056,
+      "grad_norm": 0.663686990737915,
+      "learning_rate": 6.816769649103974e-05,
+      "loss": 0.0621,
+      "step": 74260
+    },
+    {
+      "epoch": 4.8590121033693165,
+      "grad_norm": 1.038624882698059,
+      "learning_rate": 6.815913808944436e-05,
+      "loss": 0.0532,
+      "step": 74270
+    },
+    {
+      "epoch": 4.859666339548577,
+      "grad_norm": 0.7460023760795593,
+      "learning_rate": 6.815057907493045e-05,
+      "loss": 0.066,
+      "step": 74280
+    },
+    {
+      "epoch": 4.860320575727838,
+      "grad_norm": 0.8232451677322388,
+      "learning_rate": 6.814201944778689e-05,
+      "loss": 0.0635,
+      "step": 74290
+    },
+    {
+      "epoch": 4.860974811907099,
+      "grad_norm": 0.7617900967597961,
+      "learning_rate": 6.81334592083026e-05,
+      "loss": 0.0578,
+      "step": 74300
+    },
+    {
+      "epoch": 4.8616290480863595,
+      "grad_norm": 0.8007754683494568,
+      "learning_rate": 6.81248983567665e-05,
+      "loss": 0.0553,
+      "step": 74310
+    },
+    {
+      "epoch": 4.86228328426562,
+      "grad_norm": 0.8569623231887817,
+      "learning_rate": 6.811633689346752e-05,
+      "loss": 0.0563,
+      "step": 74320
+    },
+    {
+      "epoch": 4.862937520444881,
+      "grad_norm": 0.979371964931488,
+      "learning_rate": 6.810777481869471e-05,
+      "loss": 0.059,
+      "step": 74330
+    },
+    {
+      "epoch": 4.863591756624142,
+      "grad_norm": 0.7974584102630615,
+      "learning_rate": 6.809921213273697e-05,
+      "loss": 0.0644,
+      "step": 74340
+    },
+    {
+      "epoch": 4.864245992803402,
+      "grad_norm": 0.9594705104827881,
+      "learning_rate": 6.809064883588336e-05,
+      "loss": 0.0618,
+      "step": 74350
+    },
+    {
+      "epoch": 4.864900228982663,
+      "grad_norm": 0.8043680191040039,
+      "learning_rate": 6.808208492842291e-05,
+      "loss": 0.0581,
+      "step": 74360
+    },
+    {
+      "epoch": 4.865554465161924,
+      "grad_norm": 0.8218439221382141,
+      "learning_rate": 6.807352041064467e-05,
+      "loss": 0.0607,
+      "step": 74370
+    },
+    {
+      "epoch": 4.8662087013411846,
+      "grad_norm": 1.0335355997085571,
+      "learning_rate": 6.806495528283771e-05,
+      "loss": 0.0679,
+      "step": 74380
+    },
+    {
+      "epoch": 4.866862937520445,
+      "grad_norm": 1.074435830116272,
+      "learning_rate": 6.805638954529117e-05,
+      "loss": 0.0628,
+      "step": 74390
+    },
+    {
+      "epoch": 4.867517173699706,
+      "grad_norm": 1.0719654560089111,
+      "learning_rate": 6.80478231982941e-05,
+      "loss": 0.0632,
+      "step": 74400
+    },
+    {
+      "epoch": 4.868171409878967,
+      "grad_norm": 0.9550548791885376,
+      "learning_rate": 6.803925624213565e-05,
+      "loss": 0.0554,
+      "step": 74410
+    },
+    {
+      "epoch": 4.868825646058227,
+      "grad_norm": 0.9627480506896973,
+      "learning_rate": 6.803068867710503e-05,
+      "loss": 0.0676,
+      "step": 74420
+    },
+    {
+      "epoch": 4.869479882237488,
+      "grad_norm": 1.0203245878219604,
+      "learning_rate": 6.802212050349135e-05,
+      "loss": 0.0535,
+      "step": 74430
+    },
+    {
+      "epoch": 4.870134118416749,
+      "grad_norm": 0.996152937412262,
+      "learning_rate": 6.801355172158385e-05,
+      "loss": 0.0572,
+      "step": 74440
+    },
+    {
+      "epoch": 4.870788354596009,
+      "grad_norm": 0.8001197576522827,
+      "learning_rate": 6.800498233167172e-05,
+      "loss": 0.06,
+      "step": 74450
+    },
+    {
+      "epoch": 4.87144259077527,
+      "grad_norm": 1.0456775426864624,
+      "learning_rate": 6.799641233404423e-05,
+      "loss": 0.0618,
+      "step": 74460
+    },
+    {
+      "epoch": 4.872096826954531,
+      "grad_norm": 0.9173516631126404,
+      "learning_rate": 6.798784172899064e-05,
+      "loss": 0.0584,
+      "step": 74470
+    },
+    {
+      "epoch": 4.872751063133792,
+      "grad_norm": 1.2566989660263062,
+      "learning_rate": 6.79792705168002e-05,
+      "loss": 0.0665,
+      "step": 74480
+    },
+    {
+      "epoch": 4.873405299313052,
+      "grad_norm": 0.9484615921974182,
+      "learning_rate": 6.797069869776222e-05,
+      "loss": 0.0643,
+      "step": 74490
+    },
+    {
+      "epoch": 4.874059535492313,
+      "grad_norm": 0.8985061049461365,
+      "learning_rate": 6.796212627216605e-05,
+      "loss": 0.0664,
+      "step": 74500
+    },
+    {
+      "epoch": 4.874713771671574,
+      "grad_norm": 0.8476536273956299,
+      "learning_rate": 6.795355324030099e-05,
+      "loss": 0.0543,
+      "step": 74510
+    },
+    {
+      "epoch": 4.875368007850835,
+      "grad_norm": 1.0744060277938843,
+      "learning_rate": 6.794497960245644e-05,
+      "loss": 0.0656,
+      "step": 74520
+    },
+    {
+      "epoch": 4.876022244030095,
+      "grad_norm": 1.001219391822815,
+      "learning_rate": 6.793640535892176e-05,
+      "loss": 0.063,
+      "step": 74530
+    },
+    {
+      "epoch": 4.876676480209356,
+      "grad_norm": 1.04497492313385,
+      "learning_rate": 6.792783050998637e-05,
+      "loss": 0.0629,
+      "step": 74540
+    },
+    {
+      "epoch": 4.877330716388617,
+      "grad_norm": 1.1128970384597778,
+      "learning_rate": 6.791925505593965e-05,
+      "loss": 0.0598,
+      "step": 74550
+    },
+    {
+      "epoch": 4.877984952567877,
+      "grad_norm": 0.9689093828201294,
+      "learning_rate": 6.791067899707113e-05,
+      "loss": 0.0577,
+      "step": 74560
+    },
+    {
+      "epoch": 4.878639188747138,
+      "grad_norm": 1.0436021089553833,
+      "learning_rate": 6.790210233367019e-05,
+      "loss": 0.0557,
+      "step": 74570
+    },
+    {
+      "epoch": 4.879293424926399,
+      "grad_norm": 0.6721805334091187,
+      "learning_rate": 6.789352506602632e-05,
+      "loss": 0.0665,
+      "step": 74580
+    },
+    {
+      "epoch": 4.879947661105659,
+      "grad_norm": 0.8921915888786316,
+      "learning_rate": 6.78849471944291e-05,
+      "loss": 0.0541,
+      "step": 74590
+    },
+    {
+      "epoch": 4.88060189728492,
+      "grad_norm": 0.7612684965133667,
+      "learning_rate": 6.787636871916798e-05,
+      "loss": 0.0611,
+      "step": 74600
+    },
+    {
+      "epoch": 4.881256133464181,
+      "grad_norm": 0.7255061268806458,
+      "learning_rate": 6.786778964053253e-05,
+      "loss": 0.068,
+      "step": 74610
+    },
+    {
+      "epoch": 4.881910369643442,
+      "grad_norm": 0.7865713834762573,
+      "learning_rate": 6.785920995881234e-05,
+      "loss": 0.0665,
+      "step": 74620
+    },
+    {
+      "epoch": 4.882564605822702,
+      "grad_norm": 0.8941336274147034,
+      "learning_rate": 6.785062967429697e-05,
+      "loss": 0.0531,
+      "step": 74630
+    },
+    {
+      "epoch": 4.883218842001963,
+      "grad_norm": 0.8453198671340942,
+      "learning_rate": 6.784204878727601e-05,
+      "loss": 0.0552,
+      "step": 74640
+    },
+    {
+      "epoch": 4.883873078181224,
+      "grad_norm": 0.9211252927780151,
+      "learning_rate": 6.783346729803913e-05,
+      "loss": 0.064,
+      "step": 74650
+    },
+    {
+      "epoch": 4.884527314360485,
+      "grad_norm": 0.7787407040596008,
+      "learning_rate": 6.782488520687596e-05,
+      "loss": 0.064,
+      "step": 74660
+    },
+    {
+      "epoch": 4.885181550539745,
+      "grad_norm": 0.8281296491622925,
+      "learning_rate": 6.781630251407617e-05,
+      "loss": 0.0531,
+      "step": 74670
+    },
+    {
+      "epoch": 4.885835786719006,
+      "grad_norm": 0.9956787824630737,
+      "learning_rate": 6.780771921992945e-05,
+      "loss": 0.0556,
+      "step": 74680
+    },
+    {
+      "epoch": 4.886490022898267,
+      "grad_norm": 1.0824602842330933,
+      "learning_rate": 6.779913532472548e-05,
+      "loss": 0.0639,
+      "step": 74690
+    },
+    {
+      "epoch": 4.887144259077527,
+      "grad_norm": 0.9609588980674744,
+      "learning_rate": 6.779055082875403e-05,
+      "loss": 0.0633,
+      "step": 74700
+    },
+    {
+      "epoch": 4.887798495256788,
+      "grad_norm": 0.9233350157737732,
+      "learning_rate": 6.778196573230481e-05,
+      "loss": 0.0671,
+      "step": 74710
+    },
+    {
+      "epoch": 4.888452731436049,
+      "grad_norm": 0.7840463519096375,
+      "learning_rate": 6.777338003566765e-05,
+      "loss": 0.0555,
+      "step": 74720
+    },
+    {
+      "epoch": 4.889106967615309,
+      "grad_norm": 0.8873052000999451,
+      "learning_rate": 6.776479373913228e-05,
+      "loss": 0.0642,
+      "step": 74730
+    },
+    {
+      "epoch": 4.88976120379457,
+      "grad_norm": 1.742559790611267,
+      "learning_rate": 6.775620684298853e-05,
+      "loss": 0.0636,
+      "step": 74740
+    },
+    {
+      "epoch": 4.890415439973831,
+      "grad_norm": 1.021909475326538,
+      "learning_rate": 6.774761934752624e-05,
+      "loss": 0.0568,
+      "step": 74750
+    },
+    {
+      "epoch": 4.891069676153092,
+      "grad_norm": 0.7874606847763062,
+      "learning_rate": 6.773903125303524e-05,
+      "loss": 0.0537,
+      "step": 74760
+    },
+    {
+      "epoch": 4.891723912332352,
+      "grad_norm": 0.8984439969062805,
+      "learning_rate": 6.773044255980543e-05,
+      "loss": 0.071,
+      "step": 74770
+    },
+    {
+      "epoch": 4.892378148511613,
+      "grad_norm": 0.8923892378807068,
+      "learning_rate": 6.772185326812668e-05,
+      "loss": 0.0552,
+      "step": 74780
+    },
+    {
+      "epoch": 4.893032384690874,
+      "grad_norm": 1.0217084884643555,
+      "learning_rate": 6.77132633782889e-05,
+      "loss": 0.0557,
+      "step": 74790
+    },
+    {
+      "epoch": 4.893686620870134,
+      "grad_norm": 0.9023916721343994,
+      "learning_rate": 6.770467289058203e-05,
+      "loss": 0.065,
+      "step": 74800
+    },
+    {
+      "epoch": 4.894340857049395,
+      "grad_norm": 1.1047594547271729,
+      "learning_rate": 6.7696081805296e-05,
+      "loss": 0.0521,
+      "step": 74810
+    },
+    {
+      "epoch": 4.894995093228656,
+      "grad_norm": 1.103213906288147,
+      "learning_rate": 6.768749012272081e-05,
+      "loss": 0.0691,
+      "step": 74820
+    },
+    {
+      "epoch": 4.895649329407917,
+      "grad_norm": 0.9325263500213623,
+      "learning_rate": 6.767889784314645e-05,
+      "loss": 0.0625,
+      "step": 74830
+    },
+    {
+      "epoch": 4.896303565587177,
+      "grad_norm": 1.620137333869934,
+      "learning_rate": 6.76703049668629e-05,
+      "loss": 0.0641,
+      "step": 74840
+    },
+    {
+      "epoch": 4.896957801766438,
+      "grad_norm": 1.0535722970962524,
+      "learning_rate": 6.766171149416023e-05,
+      "loss": 0.052,
+      "step": 74850
+    },
+    {
+      "epoch": 4.897612037945699,
+      "grad_norm": 0.9076412320137024,
+      "learning_rate": 6.765311742532849e-05,
+      "loss": 0.0637,
+      "step": 74860
+    },
+    {
+      "epoch": 4.898266274124959,
+      "grad_norm": 1.7605758905410767,
+      "learning_rate": 6.764452276065774e-05,
+      "loss": 0.0574,
+      "step": 74870
+    },
+    {
+      "epoch": 4.89892051030422,
+      "grad_norm": 0.7411386966705322,
+      "learning_rate": 6.763592750043805e-05,
+      "loss": 0.0568,
+      "step": 74880
+    },
+    {
+      "epoch": 4.899574746483481,
+      "grad_norm": 0.8387102484703064,
+      "learning_rate": 6.762733164495956e-05,
+      "loss": 0.0567,
+      "step": 74890
+    },
+    {
+      "epoch": 4.900228982662741,
+      "grad_norm": 0.9322872161865234,
+      "learning_rate": 6.761873519451241e-05,
+      "loss": 0.0554,
+      "step": 74900
+    },
+    {
+      "epoch": 4.900883218842002,
+      "grad_norm": 0.9480436444282532,
+      "learning_rate": 6.761013814938673e-05,
+      "loss": 0.0575,
+      "step": 74910
+    },
+    {
+      "epoch": 4.901537455021263,
+      "grad_norm": 0.8049683570861816,
+      "learning_rate": 6.760154050987272e-05,
+      "loss": 0.0543,
+      "step": 74920
+    },
+    {
+      "epoch": 4.902191691200524,
+      "grad_norm": 1.016456961631775,
+      "learning_rate": 6.759294227626054e-05,
+      "loss": 0.0538,
+      "step": 74930
+    },
+    {
+      "epoch": 4.902845927379784,
+      "grad_norm": 0.9667950868606567,
+      "learning_rate": 6.758434344884042e-05,
+      "loss": 0.0569,
+      "step": 74940
+    },
+    {
+      "epoch": 4.903500163559045,
+      "grad_norm": 0.902862548828125,
+      "learning_rate": 6.75757440279026e-05,
+      "loss": 0.0586,
+      "step": 74950
+    },
+    {
+      "epoch": 4.904154399738306,
+      "grad_norm": 0.7040520906448364,
+      "learning_rate": 6.756714401373732e-05,
+      "loss": 0.0593,
+      "step": 74960
+    },
+    {
+      "epoch": 4.904808635917567,
+      "grad_norm": 0.824975848197937,
+      "learning_rate": 6.755854340663484e-05,
+      "loss": 0.0565,
+      "step": 74970
+    },
+    {
+      "epoch": 4.905462872096827,
+      "grad_norm": 0.8577737212181091,
+      "learning_rate": 6.754994220688551e-05,
+      "loss": 0.0581,
+      "step": 74980
+    },
+    {
+      "epoch": 4.906117108276088,
+      "grad_norm": 0.8069767951965332,
+      "learning_rate": 6.754134041477957e-05,
+      "loss": 0.0481,
+      "step": 74990
+    },
+    {
+      "epoch": 4.906771344455349,
+      "grad_norm": 0.7826264500617981,
+      "learning_rate": 6.75327380306074e-05,
+      "loss": 0.0642,
+      "step": 75000
+    },
+    {
+      "epoch": 4.907425580634609,
+      "grad_norm": 0.895269513130188,
+      "learning_rate": 6.752413505465935e-05,
+      "loss": 0.0549,
+      "step": 75010
+    },
+    {
+      "epoch": 4.90807981681387,
+      "grad_norm": 0.6955884099006653,
+      "learning_rate": 6.751553148722576e-05,
+      "loss": 0.0546,
+      "step": 75020
+    },
+    {
+      "epoch": 4.908734052993131,
+      "grad_norm": 0.8068095445632935,
+      "learning_rate": 6.750692732859706e-05,
+      "loss": 0.0569,
+      "step": 75030
+    },
+    {
+      "epoch": 4.909388289172391,
+      "grad_norm": 0.6808393001556396,
+      "learning_rate": 6.749832257906365e-05,
+      "loss": 0.0557,
+      "step": 75040
+    },
+    {
+      "epoch": 4.910042525351652,
+      "grad_norm": 0.9451243281364441,
+      "learning_rate": 6.748971723891597e-05,
+      "loss": 0.0634,
+      "step": 75050
+    },
+    {
+      "epoch": 4.910696761530913,
+      "grad_norm": 0.7461559772491455,
+      "learning_rate": 6.748111130844445e-05,
+      "loss": 0.062,
+      "step": 75060
+    },
+    {
+      "epoch": 4.911350997710174,
+      "grad_norm": 0.8373140096664429,
+      "learning_rate": 6.747250478793959e-05,
+      "loss": 0.0538,
+      "step": 75070
+    },
+    {
+      "epoch": 4.912005233889434,
+      "grad_norm": 0.9267625212669373,
+      "learning_rate": 6.746389767769185e-05,
+      "loss": 0.0657,
+      "step": 75080
+    },
+    {
+      "epoch": 4.912659470068695,
+      "grad_norm": 0.8543652892112732,
+      "learning_rate": 6.745528997799178e-05,
+      "loss": 0.0612,
+      "step": 75090
+    },
+    {
+      "epoch": 4.913313706247956,
+      "grad_norm": 0.9111425280570984,
+      "learning_rate": 6.744668168912989e-05,
+      "loss": 0.0604,
+      "step": 75100
+    },
+    {
+      "epoch": 4.913967942427217,
+      "grad_norm": 1.040820598602295,
+      "learning_rate": 6.743807281139675e-05,
+      "loss": 0.0614,
+      "step": 75110
+    },
+    {
+      "epoch": 4.914622178606477,
+      "grad_norm": 0.8948336839675903,
+      "learning_rate": 6.74294633450829e-05,
+      "loss": 0.0527,
+      "step": 75120
+    },
+    {
+      "epoch": 4.915276414785738,
+      "grad_norm": 0.923996090888977,
+      "learning_rate": 6.742085329047895e-05,
+      "loss": 0.0636,
+      "step": 75130
+    },
+    {
+      "epoch": 4.915930650964999,
+      "grad_norm": 0.8746695518493652,
+      "learning_rate": 6.741224264787553e-05,
+      "loss": 0.0587,
+      "step": 75140
+    },
+    {
+      "epoch": 4.916584887144259,
+      "grad_norm": 0.7787982821464539,
+      "learning_rate": 6.740363141756325e-05,
+      "loss": 0.0542,
+      "step": 75150
+    },
+    {
+      "epoch": 4.91723912332352,
+      "grad_norm": 0.9488137364387512,
+      "learning_rate": 6.739501959983277e-05,
+      "loss": 0.0602,
+      "step": 75160
+    },
+    {
+      "epoch": 4.917893359502781,
+      "grad_norm": 0.7842972874641418,
+      "learning_rate": 6.738640719497475e-05,
+      "loss": 0.0538,
+      "step": 75170
+    },
+    {
+      "epoch": 4.918547595682041,
+      "grad_norm": 0.9717864394187927,
+      "learning_rate": 6.73777942032799e-05,
+      "loss": 0.0614,
+      "step": 75180
+    },
+    {
+      "epoch": 4.919201831861302,
+      "grad_norm": 0.7789155840873718,
+      "learning_rate": 6.736918062503889e-05,
+      "loss": 0.068,
+      "step": 75190
+    },
+    {
+      "epoch": 4.919856068040563,
+      "grad_norm": 0.8811984062194824,
+      "learning_rate": 6.736056646054251e-05,
+      "loss": 0.0531,
+      "step": 75200
+    },
+    {
+      "epoch": 4.920510304219824,
+      "grad_norm": 0.9033105969429016,
+      "learning_rate": 6.73519517100815e-05,
+      "loss": 0.0591,
+      "step": 75210
+    },
+    {
+      "epoch": 4.921164540399084,
+      "grad_norm": 0.7949878573417664,
+      "learning_rate": 6.734333637394657e-05,
+      "loss": 0.0538,
+      "step": 75220
+    },
+    {
+      "epoch": 4.921818776578345,
+      "grad_norm": 0.9282236099243164,
+      "learning_rate": 6.73347204524286e-05,
+      "loss": 0.0672,
+      "step": 75230
+    },
+    {
+      "epoch": 4.922473012757606,
+      "grad_norm": 0.8523067831993103,
+      "learning_rate": 6.732610394581831e-05,
+      "loss": 0.0512,
+      "step": 75240
+    },
+    {
+      "epoch": 4.923127248936867,
+      "grad_norm": 0.9058271646499634,
+      "learning_rate": 6.731748685440658e-05,
+      "loss": 0.064,
+      "step": 75250
+    },
+    {
+      "epoch": 4.923781485116127,
+      "grad_norm": 0.7754740715026855,
+      "learning_rate": 6.730886917848426e-05,
+      "loss": 0.06,
+      "step": 75260
+    },
+    {
+      "epoch": 4.924435721295388,
+      "grad_norm": 0.9523494839668274,
+      "learning_rate": 6.730025091834223e-05,
+      "loss": 0.062,
+      "step": 75270
+    },
+    {
+      "epoch": 4.925089957474649,
+      "grad_norm": 0.825276255607605,
+      "learning_rate": 6.729163207427134e-05,
+      "loss": 0.0568,
+      "step": 75280
+    },
+    {
+      "epoch": 4.925744193653909,
+      "grad_norm": 1.0787001848220825,
+      "learning_rate": 6.728301264656251e-05,
+      "loss": 0.0591,
+      "step": 75290
+    },
+    {
+      "epoch": 4.92639842983317,
+      "grad_norm": 0.859107494354248,
+      "learning_rate": 6.727439263550669e-05,
+      "loss": 0.0568,
+      "step": 75300
+    },
+    {
+      "epoch": 4.927052666012431,
+      "grad_norm": 0.897099494934082,
+      "learning_rate": 6.726577204139482e-05,
+      "loss": 0.063,
+      "step": 75310
+    },
+    {
+      "epoch": 4.927706902191691,
+      "grad_norm": 0.7519966959953308,
+      "learning_rate": 6.725715086451784e-05,
+      "loss": 0.0582,
+      "step": 75320
+    },
+    {
+      "epoch": 4.928361138370952,
+      "grad_norm": 0.7944568991661072,
+      "learning_rate": 6.724852910516677e-05,
+      "loss": 0.0605,
+      "step": 75330
+    },
+    {
+      "epoch": 4.929015374550213,
+      "grad_norm": 1.1941640377044678,
+      "learning_rate": 6.723990676363262e-05,
+      "loss": 0.065,
+      "step": 75340
+    },
+    {
+      "epoch": 4.929669610729473,
+      "grad_norm": 0.7790656089782715,
+      "learning_rate": 6.723128384020638e-05,
+      "loss": 0.0591,
+      "step": 75350
+    },
+    {
+      "epoch": 4.930323846908734,
+      "grad_norm": 0.9245325326919556,
+      "learning_rate": 6.722266033517913e-05,
+      "loss": 0.0511,
+      "step": 75360
+    },
+    {
+      "epoch": 4.930978083087995,
+      "grad_norm": 0.9800746440887451,
+      "learning_rate": 6.721403624884194e-05,
+      "loss": 0.0634,
+      "step": 75370
+    },
+    {
+      "epoch": 4.931632319267256,
+      "grad_norm": 0.7326732277870178,
+      "learning_rate": 6.720541158148587e-05,
+      "loss": 0.0555,
+      "step": 75380
+    },
+    {
+      "epoch": 4.932286555446516,
+      "grad_norm": 0.8972048759460449,
+      "learning_rate": 6.719678633340202e-05,
+      "loss": 0.0515,
+      "step": 75390
+    },
+    {
+      "epoch": 4.932940791625777,
+      "grad_norm": 0.9582744240760803,
+      "learning_rate": 6.718816050488157e-05,
+      "loss": 0.0626,
+      "step": 75400
+    },
+    {
+      "epoch": 4.933595027805038,
+      "grad_norm": 0.7547270655632019,
+      "learning_rate": 6.717953409621559e-05,
+      "loss": 0.0576,
+      "step": 75410
+    },
+    {
+      "epoch": 4.934249263984299,
+      "grad_norm": 1.1005964279174805,
+      "learning_rate": 6.71709071076953e-05,
+      "loss": 0.0601,
+      "step": 75420
+    },
+    {
+      "epoch": 4.934903500163559,
+      "grad_norm": 0.8800162672996521,
+      "learning_rate": 6.716227953961185e-05,
+      "loss": 0.0592,
+      "step": 75430
+    },
+    {
+      "epoch": 4.93555773634282,
+      "grad_norm": 0.7670087814331055,
+      "learning_rate": 6.715365139225647e-05,
+      "loss": 0.06,
+      "step": 75440
+    },
+    {
+      "epoch": 4.936211972522081,
+      "grad_norm": 0.9640793204307556,
+      "learning_rate": 6.714502266592034e-05,
+      "loss": 0.0634,
+      "step": 75450
+    },
+    {
+      "epoch": 4.936866208701341,
+      "grad_norm": 0.8268434405326843,
+      "learning_rate": 6.713639336089476e-05,
+      "loss": 0.0591,
+      "step": 75460
+    },
+    {
+      "epoch": 4.937520444880602,
+      "grad_norm": 1.175800085067749,
+      "learning_rate": 6.712776347747096e-05,
+      "loss": 0.0675,
+      "step": 75470
+    },
+    {
+      "epoch": 4.938174681059863,
+      "grad_norm": 0.8210011720657349,
+      "learning_rate": 6.71191330159402e-05,
+      "loss": 0.0622,
+      "step": 75480
+    },
+    {
+      "epoch": 4.938828917239123,
+      "grad_norm": 0.8271769881248474,
+      "learning_rate": 6.711050197659384e-05,
+      "loss": 0.0598,
+      "step": 75490
+    },
+    {
+      "epoch": 4.939483153418384,
+      "grad_norm": 0.8794811964035034,
+      "learning_rate": 6.710187035972314e-05,
+      "loss": 0.0586,
+      "step": 75500
+    },
+    {
+      "epoch": 4.940137389597645,
+      "grad_norm": 0.8313082456588745,
+      "learning_rate": 6.709323816561946e-05,
+      "loss": 0.061,
+      "step": 75510
+    },
+    {
+      "epoch": 4.940791625776906,
+      "grad_norm": 0.9724137783050537,
+      "learning_rate": 6.708460539457418e-05,
+      "loss": 0.0639,
+      "step": 75520
+    },
+    {
+      "epoch": 4.941445861956166,
+      "grad_norm": 0.9716455936431885,
+      "learning_rate": 6.707597204687865e-05,
+      "loss": 0.0561,
+      "step": 75530
+    },
+    {
+      "epoch": 4.942100098135427,
+      "grad_norm": 0.6842145323753357,
+      "learning_rate": 6.706733812282428e-05,
+      "loss": 0.0518,
+      "step": 75540
+    },
+    {
+      "epoch": 4.942754334314688,
+      "grad_norm": 0.767361581325531,
+      "learning_rate": 6.705870362270248e-05,
+      "loss": 0.0609,
+      "step": 75550
+    },
+    {
+      "epoch": 4.943408570493949,
+      "grad_norm": 0.9804648756980896,
+      "learning_rate": 6.705006854680471e-05,
+      "loss": 0.066,
+      "step": 75560
+    },
+    {
+      "epoch": 4.944062806673209,
+      "grad_norm": 0.6798174977302551,
+      "learning_rate": 6.704143289542241e-05,
+      "loss": 0.0534,
+      "step": 75570
+    },
+    {
+      "epoch": 4.94471704285247,
+      "grad_norm": 0.8720900416374207,
+      "learning_rate": 6.703279666884705e-05,
+      "loss": 0.0508,
+      "step": 75580
+    },
+    {
+      "epoch": 4.945371279031731,
+      "grad_norm": 1.0530959367752075,
+      "learning_rate": 6.702415986737014e-05,
+      "loss": 0.0507,
+      "step": 75590
+    },
+    {
+      "epoch": 4.946025515210991,
+      "grad_norm": 0.9512234330177307,
+      "learning_rate": 6.701552249128318e-05,
+      "loss": 0.0578,
+      "step": 75600
+    },
+    {
+      "epoch": 4.946679751390252,
+      "grad_norm": 0.7706218361854553,
+      "learning_rate": 6.70068845408777e-05,
+      "loss": 0.0596,
+      "step": 75610
+    },
+    {
+      "epoch": 4.947333987569513,
+      "grad_norm": 0.896196722984314,
+      "learning_rate": 6.69982460164453e-05,
+      "loss": 0.0652,
+      "step": 75620
+    },
+    {
+      "epoch": 4.947988223748773,
+      "grad_norm": 0.8798102140426636,
+      "learning_rate": 6.69896069182775e-05,
+      "loss": 0.0601,
+      "step": 75630
+    },
+    {
+      "epoch": 4.948642459928034,
+      "grad_norm": 0.9904597997665405,
+      "learning_rate": 6.69809672466659e-05,
+      "loss": 0.0522,
+      "step": 75640
+    },
+    {
+      "epoch": 4.949296696107295,
+      "grad_norm": 0.8591594696044922,
+      "learning_rate": 6.697232700190213e-05,
+      "loss": 0.0663,
+      "step": 75650
+    },
+    {
+      "epoch": 4.949950932286556,
+      "grad_norm": 0.8061467409133911,
+      "learning_rate": 6.696368618427779e-05,
+      "loss": 0.0626,
+      "step": 75660
+    },
+    {
+      "epoch": 4.950605168465816,
+      "grad_norm": 0.7854329943656921,
+      "learning_rate": 6.695504479408458e-05,
+      "loss": 0.0545,
+      "step": 75670
+    },
+    {
+      "epoch": 4.951259404645077,
+      "grad_norm": 0.7829649448394775,
+      "learning_rate": 6.694640283161413e-05,
+      "loss": 0.0505,
+      "step": 75680
+    },
+    {
+      "epoch": 4.951913640824338,
+      "grad_norm": 0.6959307193756104,
+      "learning_rate": 6.693776029715814e-05,
+      "loss": 0.0623,
+      "step": 75690
+    },
+    {
+      "epoch": 4.952567877003599,
+      "grad_norm": 0.7448030710220337,
+      "learning_rate": 6.692911719100833e-05,
+      "loss": 0.0537,
+      "step": 75700
+    },
+    {
+      "epoch": 4.953222113182859,
+      "grad_norm": 0.686690628528595,
+      "learning_rate": 6.692047351345641e-05,
+      "loss": 0.0572,
+      "step": 75710
+    },
+    {
+      "epoch": 4.95387634936212,
+      "grad_norm": 1.1191184520721436,
+      "learning_rate": 6.691182926479413e-05,
+      "loss": 0.0699,
+      "step": 75720
+    },
+    {
+      "epoch": 4.954530585541381,
+      "grad_norm": 0.7493934631347656,
+      "learning_rate": 6.690318444531328e-05,
+      "loss": 0.0585,
+      "step": 75730
+    },
+    {
+      "epoch": 4.955184821720641,
+      "grad_norm": 1.0006502866744995,
+      "learning_rate": 6.689453905530559e-05,
+      "loss": 0.0562,
+      "step": 75740
+    },
+    {
+      "epoch": 4.955839057899902,
+      "grad_norm": 0.9580645561218262,
+      "learning_rate": 6.688589309506292e-05,
+      "loss": 0.0582,
+      "step": 75750
+    },
+    {
+      "epoch": 4.956493294079163,
+      "grad_norm": 1.0518616437911987,
+      "learning_rate": 6.687724656487707e-05,
+      "loss": 0.0699,
+      "step": 75760
+    },
+    {
+      "epoch": 4.957147530258423,
+      "grad_norm": 1.082767128944397,
+      "learning_rate": 6.686859946503989e-05,
+      "loss": 0.055,
+      "step": 75770
+    },
+    {
+      "epoch": 4.957801766437684,
+      "grad_norm": 0.933538556098938,
+      "learning_rate": 6.685995179584324e-05,
+      "loss": 0.0637,
+      "step": 75780
+    },
+    {
+      "epoch": 4.958456002616945,
+      "grad_norm": 0.955034077167511,
+      "learning_rate": 6.685130355757899e-05,
+      "loss": 0.0596,
+      "step": 75790
+    },
+    {
+      "epoch": 4.959110238796205,
+      "grad_norm": 0.9905108213424683,
+      "learning_rate": 6.684265475053905e-05,
+      "loss": 0.068,
+      "step": 75800
+    },
+    {
+      "epoch": 4.959764474975466,
+      "grad_norm": 0.7588913440704346,
+      "learning_rate": 6.683400537501534e-05,
+      "loss": 0.054,
+      "step": 75810
+    },
+    {
+      "epoch": 4.960418711154727,
+      "grad_norm": 0.8682058453559875,
+      "learning_rate": 6.68253554312998e-05,
+      "loss": 0.0569,
+      "step": 75820
+    },
+    {
+      "epoch": 4.961072947333988,
+      "grad_norm": 0.8857309818267822,
+      "learning_rate": 6.68167049196844e-05,
+      "loss": 0.0706,
+      "step": 75830
+    },
+    {
+      "epoch": 4.961727183513248,
+      "grad_norm": 0.8646987676620483,
+      "learning_rate": 6.680805384046109e-05,
+      "loss": 0.0552,
+      "step": 75840
+    },
+    {
+      "epoch": 4.962381419692509,
+      "grad_norm": 0.9026349782943726,
+      "learning_rate": 6.67994021939219e-05,
+      "loss": 0.0506,
+      "step": 75850
+    },
+    {
+      "epoch": 4.96303565587177,
+      "grad_norm": 0.9483004808425903,
+      "learning_rate": 6.679074998035881e-05,
+      "loss": 0.0557,
+      "step": 75860
+    },
+    {
+      "epoch": 4.963689892051031,
+      "grad_norm": 0.9569962024688721,
+      "learning_rate": 6.67820972000639e-05,
+      "loss": 0.0614,
+      "step": 75870
+    },
+    {
+      "epoch": 4.964344128230291,
+      "grad_norm": 0.9677038192749023,
+      "learning_rate": 6.677344385332918e-05,
+      "loss": 0.063,
+      "step": 75880
+    },
+    {
+      "epoch": 4.964998364409552,
+      "grad_norm": 0.8951756358146667,
+      "learning_rate": 6.676478994044673e-05,
+      "loss": 0.0585,
+      "step": 75890
+    },
+    {
+      "epoch": 4.965652600588813,
+      "grad_norm": 0.9659004211425781,
+      "learning_rate": 6.675613546170866e-05,
+      "loss": 0.0542,
+      "step": 75900
+    },
+    {
+      "epoch": 4.966306836768073,
+      "grad_norm": 0.7111480236053467,
+      "learning_rate": 6.674748041740707e-05,
+      "loss": 0.0635,
+      "step": 75910
+    },
+    {
+      "epoch": 4.966961072947334,
+      "grad_norm": 0.9708228707313538,
+      "learning_rate": 6.673882480783412e-05,
+      "loss": 0.0607,
+      "step": 75920
+    },
+    {
+      "epoch": 4.967615309126595,
+      "grad_norm": 0.7749221324920654,
+      "learning_rate": 6.673016863328189e-05,
+      "loss": 0.0577,
+      "step": 75930
+    },
+    {
+      "epoch": 4.968269545305855,
+      "grad_norm": 0.9135736227035522,
+      "learning_rate": 6.672151189404262e-05,
+      "loss": 0.0608,
+      "step": 75940
+    },
+    {
+      "epoch": 4.968923781485116,
+      "grad_norm": 0.9059275388717651,
+      "learning_rate": 6.671285459040847e-05,
+      "loss": 0.0552,
+      "step": 75950
+    },
+    {
+      "epoch": 4.969578017664377,
+      "grad_norm": 0.874383270740509,
+      "learning_rate": 6.670419672267163e-05,
+      "loss": 0.066,
+      "step": 75960
+    },
+    {
+      "epoch": 4.970232253843638,
+      "grad_norm": 0.8657017946243286,
+      "learning_rate": 6.669553829112435e-05,
+      "loss": 0.0529,
+      "step": 75970
+    },
+    {
+      "epoch": 4.970886490022898,
+      "grad_norm": 0.919597864151001,
+      "learning_rate": 6.668687929605889e-05,
+      "loss": 0.0594,
+      "step": 75980
+    },
+    {
+      "epoch": 4.971540726202159,
+      "grad_norm": 0.8647698163986206,
+      "learning_rate": 6.667821973776747e-05,
+      "loss": 0.0531,
+      "step": 75990
+    },
+    {
+      "epoch": 4.97219496238142,
+      "grad_norm": 0.6683645248413086,
+      "learning_rate": 6.666955961654238e-05,
+      "loss": 0.058,
+      "step": 76000
+    },
+    {
+      "epoch": 4.972849198560681,
+      "grad_norm": 0.7765533924102783,
+      "learning_rate": 6.666089893267595e-05,
+      "loss": 0.0542,
+      "step": 76010
+    },
+    {
+      "epoch": 4.973503434739941,
+      "grad_norm": 0.8284790515899658,
+      "learning_rate": 6.665223768646049e-05,
+      "loss": 0.0565,
+      "step": 76020
+    },
+    {
+      "epoch": 4.974157670919202,
+      "grad_norm": 0.869004487991333,
+      "learning_rate": 6.664357587818832e-05,
+      "loss": 0.0616,
+      "step": 76030
+    },
+    {
+      "epoch": 4.974811907098463,
+      "grad_norm": 0.99537593126297,
+      "learning_rate": 6.663491350815184e-05,
+      "loss": 0.0506,
+      "step": 76040
+    },
+    {
+      "epoch": 4.975466143277723,
+      "grad_norm": 0.8469506502151489,
+      "learning_rate": 6.66262505766434e-05,
+      "loss": 0.0676,
+      "step": 76050
+    },
+    {
+      "epoch": 4.976120379456984,
+      "grad_norm": 0.9716391563415527,
+      "learning_rate": 6.661758708395537e-05,
+      "loss": 0.0547,
+      "step": 76060
+    },
+    {
+      "epoch": 4.976774615636245,
+      "grad_norm": 1.0621095895767212,
+      "learning_rate": 6.660892303038022e-05,
+      "loss": 0.0573,
+      "step": 76070
+    },
+    {
+      "epoch": 4.977428851815505,
+      "grad_norm": 0.9522149562835693,
+      "learning_rate": 6.660025841621035e-05,
+      "loss": 0.056,
+      "step": 76080
+    },
+    {
+      "epoch": 4.978083087994766,
+      "grad_norm": 1.1043899059295654,
+      "learning_rate": 6.659159324173823e-05,
+      "loss": 0.0607,
+      "step": 76090
+    },
+    {
+      "epoch": 4.978737324174027,
+      "grad_norm": 0.8276578783988953,
+      "learning_rate": 6.658292750725632e-05,
+      "loss": 0.0537,
+      "step": 76100
+    },
+    {
+      "epoch": 4.979391560353288,
+      "grad_norm": 0.8061304688453674,
+      "learning_rate": 6.657426121305711e-05,
+      "loss": 0.0655,
+      "step": 76110
+    },
+    {
+      "epoch": 4.980045796532548,
+      "grad_norm": 0.778200626373291,
+      "learning_rate": 6.656559435943313e-05,
+      "loss": 0.0538,
+      "step": 76120
+    },
+    {
+      "epoch": 4.980700032711809,
+      "grad_norm": 0.8858152031898499,
+      "learning_rate": 6.655692694667688e-05,
+      "loss": 0.0532,
+      "step": 76130
+    },
+    {
+      "epoch": 4.98135426889107,
+      "grad_norm": 0.631655216217041,
+      "learning_rate": 6.654825897508095e-05,
+      "loss": 0.0556,
+      "step": 76140
+    },
+    {
+      "epoch": 4.982008505070331,
+      "grad_norm": 0.7844308018684387,
+      "learning_rate": 6.653959044493785e-05,
+      "loss": 0.048,
+      "step": 76150
+    },
+    {
+      "epoch": 4.982662741249591,
+      "grad_norm": 1.1140578985214233,
+      "learning_rate": 6.65309213565402e-05,
+      "loss": 0.0558,
+      "step": 76160
+    },
+    {
+      "epoch": 4.983316977428852,
+      "grad_norm": 1.0043301582336426,
+      "learning_rate": 6.652225171018061e-05,
+      "loss": 0.0553,
+      "step": 76170
+    },
+    {
+      "epoch": 4.983971213608113,
+      "grad_norm": 0.8583059310913086,
+      "learning_rate": 6.65135815061517e-05,
+      "loss": 0.0594,
+      "step": 76180
+    },
+    {
+      "epoch": 4.984625449787373,
+      "grad_norm": 0.8297145366668701,
+      "learning_rate": 6.650491074474608e-05,
+      "loss": 0.0639,
+      "step": 76190
+    },
+    {
+      "epoch": 4.985279685966634,
+      "grad_norm": 0.7165802121162415,
+      "learning_rate": 6.649623942625647e-05,
+      "loss": 0.0607,
+      "step": 76200
+    },
+    {
+      "epoch": 4.985933922145895,
+      "grad_norm": 0.9171334505081177,
+      "learning_rate": 6.64875675509755e-05,
+      "loss": 0.0548,
+      "step": 76210
+    },
+    {
+      "epoch": 4.986588158325155,
+      "grad_norm": 1.0761557817459106,
+      "learning_rate": 6.647889511919588e-05,
+      "loss": 0.061,
+      "step": 76220
+    },
+    {
+      "epoch": 4.987242394504416,
+      "grad_norm": 0.9394139051437378,
+      "learning_rate": 6.647022213121035e-05,
+      "loss": 0.0599,
+      "step": 76230
+    },
+    {
+      "epoch": 4.987896630683677,
+      "grad_norm": 0.7724557518959045,
+      "learning_rate": 6.646154858731162e-05,
+      "loss": 0.0599,
+      "step": 76240
+    },
+    {
+      "epoch": 4.988550866862937,
+      "grad_norm": 0.7974488139152527,
+      "learning_rate": 6.645287448779243e-05,
+      "loss": 0.0499,
+      "step": 76250
+    },
+    {
+      "epoch": 4.989205103042198,
+      "grad_norm": 0.9333834648132324,
+      "learning_rate": 6.64441998329456e-05,
+      "loss": 0.0697,
+      "step": 76260
+    },
+    {
+      "epoch": 4.989859339221459,
+      "grad_norm": 0.9652996063232422,
+      "learning_rate": 6.64355246230639e-05,
+      "loss": 0.0572,
+      "step": 76270
+    },
+    {
+      "epoch": 4.99051357540072,
+      "grad_norm": 1.0794624090194702,
+      "learning_rate": 6.642684885844013e-05,
+      "loss": 0.0615,
+      "step": 76280
+    },
+    {
+      "epoch": 4.99116781157998,
+      "grad_norm": 0.8401366472244263,
+      "learning_rate": 6.641817253936713e-05,
+      "loss": 0.0558,
+      "step": 76290
+    },
+    {
+      "epoch": 4.991822047759241,
+      "grad_norm": 0.6911501288414001,
+      "learning_rate": 6.640949566613777e-05,
+      "loss": 0.0558,
+      "step": 76300
+    },
+    {
+      "epoch": 4.992476283938502,
+      "grad_norm": 0.7828950881958008,
+      "learning_rate": 6.640081823904487e-05,
+      "loss": 0.0589,
+      "step": 76310
+    },
+    {
+      "epoch": 4.993130520117763,
+      "grad_norm": 0.6949774622917175,
+      "learning_rate": 6.639214025838135e-05,
+      "loss": 0.0613,
+      "step": 76320
+    },
+    {
+      "epoch": 4.993784756297023,
+      "grad_norm": 0.9022195339202881,
+      "learning_rate": 6.638346172444011e-05,
+      "loss": 0.0609,
+      "step": 76330
+    },
+    {
+      "epoch": 4.994438992476284,
+      "grad_norm": 0.824991762638092,
+      "learning_rate": 6.637478263751407e-05,
+      "loss": 0.0599,
+      "step": 76340
+    },
+    {
+      "epoch": 4.995093228655545,
+      "grad_norm": 1.041867733001709,
+      "learning_rate": 6.636610299789616e-05,
+      "loss": 0.0673,
+      "step": 76350
+    },
+    {
+      "epoch": 4.995747464834805,
+      "grad_norm": 1.073553204536438,
+      "learning_rate": 6.635742280587935e-05,
+      "loss": 0.0762,
+      "step": 76360
+    },
+    {
+      "epoch": 4.996401701014066,
+      "grad_norm": 0.8955064415931702,
+      "learning_rate": 6.634874206175666e-05,
+      "loss": 0.0659,
+      "step": 76370
+    },
+    {
+      "epoch": 4.997055937193327,
+      "grad_norm": 1.0121138095855713,
+      "learning_rate": 6.6340060765821e-05,
+      "loss": 0.0524,
+      "step": 76380
+    },
+    {
+      "epoch": 4.997710173372587,
+      "grad_norm": 1.0602036714553833,
+      "learning_rate": 6.633137891836546e-05,
+      "loss": 0.0647,
+      "step": 76390
+    },
+    {
+      "epoch": 4.998364409551848,
+      "grad_norm": 0.947830319404602,
+      "learning_rate": 6.632269651968306e-05,
+      "loss": 0.0552,
+      "step": 76400
+    },
+    {
+      "epoch": 4.999018645731109,
+      "grad_norm": 0.9607699513435364,
+      "learning_rate": 6.631401357006683e-05,
+      "loss": 0.073,
+      "step": 76410
+    },
+    {
+      "epoch": 4.99967288191037,
+      "grad_norm": 0.7737681269645691,
+      "learning_rate": 6.630533006980986e-05,
+      "loss": 0.0596,
+      "step": 76420
+    },
+    {
+      "epoch": 5.00032711808963,
+      "grad_norm": 0.8846662044525146,
+      "learning_rate": 6.629664601920524e-05,
+      "loss": 0.057,
+      "step": 76430
+    },
+    {
+      "epoch": 5.000981354268891,
+      "grad_norm": 1.1258676052093506,
+      "learning_rate": 6.628796141854608e-05,
+      "loss": 0.0602,
+      "step": 76440
+    },
+    {
+      "epoch": 5.001635590448152,
+      "grad_norm": 0.844662070274353,
+      "learning_rate": 6.627927626812548e-05,
+      "loss": 0.0529,
+      "step": 76450
+    },
+    {
+      "epoch": 5.002289826627412,
+      "grad_norm": 0.9069094657897949,
+      "learning_rate": 6.627059056823665e-05,
+      "loss": 0.056,
+      "step": 76460
+    },
+    {
+      "epoch": 5.002944062806673,
+      "grad_norm": 0.8290544152259827,
+      "learning_rate": 6.62619043191727e-05,
+      "loss": 0.0576,
+      "step": 76470
+    },
+    {
+      "epoch": 5.003598298985934,
+      "grad_norm": 1.026877522468567,
+      "learning_rate": 6.625321752122682e-05,
+      "loss": 0.0594,
+      "step": 76480
+    },
+    {
+      "epoch": 5.004252535165195,
+      "grad_norm": 1.0628007650375366,
+      "learning_rate": 6.624453017469223e-05,
+      "loss": 0.0548,
+      "step": 76490
+    },
+    {
+      "epoch": 5.004906771344455,
+      "grad_norm": 0.9122229218482971,
+      "learning_rate": 6.623584227986215e-05,
+      "loss": 0.0582,
+      "step": 76500
+    },
+    {
+      "epoch": 5.005561007523716,
+      "grad_norm": 0.9323796629905701,
+      "learning_rate": 6.622715383702981e-05,
+      "loss": 0.0652,
+      "step": 76510
+    },
+    {
+      "epoch": 5.006215243702977,
+      "grad_norm": 0.9565145969390869,
+      "learning_rate": 6.621846484648849e-05,
+      "loss": 0.0614,
+      "step": 76520
+    },
+    {
+      "epoch": 5.006869479882237,
+      "grad_norm": 0.9995282292366028,
+      "learning_rate": 6.620977530853141e-05,
+      "loss": 0.0626,
+      "step": 76530
+    },
+    {
+      "epoch": 5.007523716061498,
+      "grad_norm": 1.09110689163208,
+      "learning_rate": 6.620108522345192e-05,
+      "loss": 0.0655,
+      "step": 76540
+    },
+    {
+      "epoch": 5.008177952240759,
+      "grad_norm": 0.8959344625473022,
+      "learning_rate": 6.619239459154331e-05,
+      "loss": 0.0594,
+      "step": 76550
+    },
+    {
+      "epoch": 5.00883218842002,
+      "grad_norm": 0.7433099150657654,
+      "learning_rate": 6.618370341309891e-05,
+      "loss": 0.0563,
+      "step": 76560
+    },
+    {
+      "epoch": 5.00948642459928,
+      "grad_norm": 0.8508242964744568,
+      "learning_rate": 6.61750116884121e-05,
+      "loss": 0.0621,
+      "step": 76570
+    },
+    {
+      "epoch": 5.010140660778541,
+      "grad_norm": 1.0113868713378906,
+      "learning_rate": 6.616631941777621e-05,
+      "loss": 0.0628,
+      "step": 76580
+    },
+    {
+      "epoch": 5.010794896957802,
+      "grad_norm": 0.8707073330879211,
+      "learning_rate": 6.615762660148464e-05,
+      "loss": 0.062,
+      "step": 76590
+    },
+    {
+      "epoch": 5.011449133137062,
+      "grad_norm": 0.8594477772712708,
+      "learning_rate": 6.61489332398308e-05,
+      "loss": 0.0603,
+      "step": 76600
+    },
+    {
+      "epoch": 5.012103369316323,
+      "grad_norm": 0.7275976538658142,
+      "learning_rate": 6.614023933310813e-05,
+      "loss": 0.0524,
+      "step": 76610
+    },
+    {
+      "epoch": 5.012757605495584,
+      "grad_norm": 0.8984965682029724,
+      "learning_rate": 6.613154488161003e-05,
+      "loss": 0.0653,
+      "step": 76620
+    },
+    {
+      "epoch": 5.013411841674845,
+      "grad_norm": 0.8797445893287659,
+      "learning_rate": 6.612284988562997e-05,
+      "loss": 0.0669,
+      "step": 76630
+    },
+    {
+      "epoch": 5.014066077854105,
+      "grad_norm": 0.9222575426101685,
+      "learning_rate": 6.611415434546147e-05,
+      "loss": 0.0512,
+      "step": 76640
+    },
+    {
+      "epoch": 5.014720314033366,
+      "grad_norm": 0.7390499114990234,
+      "learning_rate": 6.6105458261398e-05,
+      "loss": 0.0504,
+      "step": 76650
+    },
+    {
+      "epoch": 5.015374550212627,
+      "grad_norm": 0.9179421663284302,
+      "learning_rate": 6.609676163373306e-05,
+      "loss": 0.0605,
+      "step": 76660
+    },
+    {
+      "epoch": 5.016028786391887,
+      "grad_norm": 0.7605708837509155,
+      "learning_rate": 6.608806446276021e-05,
+      "loss": 0.0527,
+      "step": 76670
+    },
+    {
+      "epoch": 5.016683022571148,
+      "grad_norm": 0.8641387820243835,
+      "learning_rate": 6.6079366748773e-05,
+      "loss": 0.0592,
+      "step": 76680
+    },
+    {
+      "epoch": 5.017337258750409,
+      "grad_norm": 0.8649716377258301,
+      "learning_rate": 6.607066849206498e-05,
+      "loss": 0.0731,
+      "step": 76690
+    },
+    {
+      "epoch": 5.01799149492967,
+      "grad_norm": 0.8245428800582886,
+      "learning_rate": 6.606196969292974e-05,
+      "loss": 0.0621,
+      "step": 76700
+    },
+    {
+      "epoch": 5.01864573110893,
+      "grad_norm": 1.014141321182251,
+      "learning_rate": 6.605327035166091e-05,
+      "loss": 0.0639,
+      "step": 76710
+    },
+    {
+      "epoch": 5.019299967288191,
+      "grad_norm": 0.9020232558250427,
+      "learning_rate": 6.604457046855212e-05,
+      "loss": 0.0535,
+      "step": 76720
+    },
+    {
+      "epoch": 5.019954203467452,
+      "grad_norm": 0.9599230885505676,
+      "learning_rate": 6.603587004389697e-05,
+      "loss": 0.0541,
+      "step": 76730
+    },
+    {
+      "epoch": 5.020608439646712,
+      "grad_norm": 0.9951395988464355,
+      "learning_rate": 6.602716907798917e-05,
+      "loss": 0.0624,
+      "step": 76740
+    },
+    {
+      "epoch": 5.021262675825973,
+      "grad_norm": 0.8179054260253906,
+      "learning_rate": 6.601846757112238e-05,
+      "loss": 0.053,
+      "step": 76750
+    },
+    {
+      "epoch": 5.021916912005234,
+      "grad_norm": 0.8312829732894897,
+      "learning_rate": 6.600976552359029e-05,
+      "loss": 0.0579,
+      "step": 76760
+    },
+    {
+      "epoch": 5.022571148184495,
+      "grad_norm": 0.8234156370162964,
+      "learning_rate": 6.600106293568663e-05,
+      "loss": 0.0565,
+      "step": 76770
+    },
+    {
+      "epoch": 5.023225384363755,
+      "grad_norm": 0.7905632853507996,
+      "learning_rate": 6.599235980770514e-05,
+      "loss": 0.0578,
+      "step": 76780
+    },
+    {
+      "epoch": 5.023879620543016,
+      "grad_norm": 0.83710116147995,
+      "learning_rate": 6.598365613993956e-05,
+      "loss": 0.058,
+      "step": 76790
+    },
+    {
+      "epoch": 5.024533856722277,
+      "grad_norm": 1.039841890335083,
+      "learning_rate": 6.597495193268366e-05,
+      "loss": 0.0648,
+      "step": 76800
+    },
+    {
+      "epoch": 5.025188092901537,
+      "grad_norm": 1.0208362340927124,
+      "learning_rate": 6.596624718623124e-05,
+      "loss": 0.063,
+      "step": 76810
+    },
+    {
+      "epoch": 5.025842329080798,
+      "grad_norm": 0.7998032569885254,
+      "learning_rate": 6.59575419008761e-05,
+      "loss": 0.0638,
+      "step": 76820
+    },
+    {
+      "epoch": 5.026496565260059,
+      "grad_norm": 0.8299915790557861,
+      "learning_rate": 6.594883607691209e-05,
+      "loss": 0.0595,
+      "step": 76830
+    },
+    {
+      "epoch": 5.02715080143932,
+      "grad_norm": 0.8311706185340881,
+      "learning_rate": 6.594012971463302e-05,
+      "loss": 0.0569,
+      "step": 76840
+    },
+    {
+      "epoch": 5.02780503761858,
+      "grad_norm": 1.0256335735321045,
+      "learning_rate": 6.593142281433277e-05,
+      "loss": 0.0599,
+      "step": 76850
+    },
+    {
+      "epoch": 5.028459273797841,
+      "grad_norm": 0.8700340390205383,
+      "learning_rate": 6.592271537630521e-05,
+      "loss": 0.0632,
+      "step": 76860
+    },
+    {
+      "epoch": 5.029113509977102,
+      "grad_norm": 0.7327423095703125,
+      "learning_rate": 6.591400740084425e-05,
+      "loss": 0.0572,
+      "step": 76870
+    },
+    {
+      "epoch": 5.029767746156362,
+      "grad_norm": 0.907089352607727,
+      "learning_rate": 6.590529888824381e-05,
+      "loss": 0.0511,
+      "step": 76880
+    },
+    {
+      "epoch": 5.030421982335623,
+      "grad_norm": 1.0830025672912598,
+      "learning_rate": 6.589658983879782e-05,
+      "loss": 0.0629,
+      "step": 76890
+    },
+    {
+      "epoch": 5.031076218514884,
+      "grad_norm": 0.7621312737464905,
+      "learning_rate": 6.588788025280022e-05,
+      "loss": 0.0578,
+      "step": 76900
+    },
+    {
+      "epoch": 5.031730454694144,
+      "grad_norm": 0.8991169929504395,
+      "learning_rate": 6.587917013054503e-05,
+      "loss": 0.0497,
+      "step": 76910
+    },
+    {
+      "epoch": 5.032384690873405,
+      "grad_norm": 1.5194605588912964,
+      "learning_rate": 6.587045947232616e-05,
+      "loss": 0.0594,
+      "step": 76920
+    },
+    {
+      "epoch": 5.033038927052666,
+      "grad_norm": 1.2414143085479736,
+      "learning_rate": 6.586174827843768e-05,
+      "loss": 0.0645,
+      "step": 76930
+    },
+    {
+      "epoch": 5.033693163231927,
+      "grad_norm": 1.0320972204208374,
+      "learning_rate": 6.58530365491736e-05,
+      "loss": 0.0553,
+      "step": 76940
+    },
+    {
+      "epoch": 5.034347399411187,
+      "grad_norm": 0.9476144313812256,
+      "learning_rate": 6.584432428482797e-05,
+      "loss": 0.0631,
+      "step": 76950
+    },
+    {
+      "epoch": 5.035001635590448,
+      "grad_norm": 1.1414525508880615,
+      "learning_rate": 6.583561148569481e-05,
+      "loss": 0.0699,
+      "step": 76960
+    },
+    {
+      "epoch": 5.035655871769709,
+      "grad_norm": 0.8667239546775818,
+      "learning_rate": 6.582689815206825e-05,
+      "loss": 0.0637,
+      "step": 76970
+    },
+    {
+      "epoch": 5.036310107948969,
+      "grad_norm": 0.6989625692367554,
+      "learning_rate": 6.581818428424238e-05,
+      "loss": 0.0546,
+      "step": 76980
+    },
+    {
+      "epoch": 5.03696434412823,
+      "grad_norm": 0.7704555988311768,
+      "learning_rate": 6.580946988251128e-05,
+      "loss": 0.0492,
+      "step": 76990
+    },
+    {
+      "epoch": 5.037618580307491,
+      "grad_norm": 1.1614646911621094,
+      "learning_rate": 6.580075494716912e-05,
+      "loss": 0.0721,
+      "step": 77000
+    },
+    {
+      "epoch": 5.038272816486752,
+      "grad_norm": 1.0777459144592285,
+      "learning_rate": 6.579203947851006e-05,
+      "loss": 0.0586,
+      "step": 77010
+    },
+    {
+      "epoch": 5.038927052666012,
+      "grad_norm": 0.9171910285949707,
+      "learning_rate": 6.578332347682824e-05,
+      "loss": 0.0592,
+      "step": 77020
+    },
+    {
+      "epoch": 5.039581288845273,
+      "grad_norm": 0.9307588934898376,
+      "learning_rate": 6.577460694241784e-05,
+      "loss": 0.0551,
+      "step": 77030
+    },
+    {
+      "epoch": 5.040235525024534,
+      "grad_norm": 0.7374948263168335,
+      "learning_rate": 6.576588987557312e-05,
+      "loss": 0.0557,
+      "step": 77040
+    },
+    {
+      "epoch": 5.040889761203794,
+      "grad_norm": 0.7271601557731628,
+      "learning_rate": 6.575717227658825e-05,
+      "loss": 0.057,
+      "step": 77050
+    },
+    {
+      "epoch": 5.041543997383055,
+      "grad_norm": 0.8527136445045471,
+      "learning_rate": 6.57484541457575e-05,
+      "loss": 0.0569,
+      "step": 77060
+    },
+    {
+      "epoch": 5.042198233562316,
+      "grad_norm": 0.8172820806503296,
+      "learning_rate": 6.57397354833751e-05,
+      "loss": 0.059,
+      "step": 77070
+    },
+    {
+      "epoch": 5.042852469741577,
+      "grad_norm": 0.8320928812026978,
+      "learning_rate": 6.573101628973537e-05,
+      "loss": 0.0557,
+      "step": 77080
+    },
+    {
+      "epoch": 5.043506705920837,
+      "grad_norm": 0.8559585213661194,
+      "learning_rate": 6.572229656513258e-05,
+      "loss": 0.0589,
+      "step": 77090
+    },
+    {
+      "epoch": 5.044160942100098,
+      "grad_norm": 0.797950267791748,
+      "learning_rate": 6.571357630986104e-05,
+      "loss": 0.0525,
+      "step": 77100
+    },
+    {
+      "epoch": 5.044815178279359,
+      "grad_norm": 0.9195685982704163,
+      "learning_rate": 6.570485552421509e-05,
+      "loss": 0.0576,
+      "step": 77110
+    },
+    {
+      "epoch": 5.045469414458619,
+      "grad_norm": 0.6936109066009521,
+      "learning_rate": 6.569613420848908e-05,
+      "loss": 0.0578,
+      "step": 77120
+    },
+    {
+      "epoch": 5.04612365063788,
+      "grad_norm": 0.9056085348129272,
+      "learning_rate": 6.568741236297738e-05,
+      "loss": 0.0642,
+      "step": 77130
+    },
+    {
+      "epoch": 5.046777886817141,
+      "grad_norm": 0.9640796184539795,
+      "learning_rate": 6.567868998797438e-05,
+      "loss": 0.0651,
+      "step": 77140
+    },
+    {
+      "epoch": 5.047432122996402,
+      "grad_norm": 0.8185075521469116,
+      "learning_rate": 6.566996708377444e-05,
+      "loss": 0.0585,
+      "step": 77150
+    },
+    {
+      "epoch": 5.048086359175662,
+      "grad_norm": 0.7742406129837036,
+      "learning_rate": 6.566124365067203e-05,
+      "loss": 0.0685,
+      "step": 77160
+    },
+    {
+      "epoch": 5.048740595354923,
+      "grad_norm": 0.9033831357955933,
+      "learning_rate": 6.56525196889616e-05,
+      "loss": 0.0545,
+      "step": 77170
+    },
+    {
+      "epoch": 5.049394831534184,
+      "grad_norm": 1.0280879735946655,
+      "learning_rate": 6.564379519893756e-05,
+      "loss": 0.0575,
+      "step": 77180
+    },
+    {
+      "epoch": 5.050049067713444,
+      "grad_norm": 0.8477870225906372,
+      "learning_rate": 6.56350701808944e-05,
+      "loss": 0.0596,
+      "step": 77190
+    },
+    {
+      "epoch": 5.050703303892705,
+      "grad_norm": 0.7754467129707336,
+      "learning_rate": 6.562634463512663e-05,
+      "loss": 0.057,
+      "step": 77200
+    },
+    {
+      "epoch": 5.051357540071966,
+      "grad_norm": 0.9043933749198914,
+      "learning_rate": 6.561761856192873e-05,
+      "loss": 0.0679,
+      "step": 77210
+    },
+    {
+      "epoch": 5.052011776251227,
+      "grad_norm": 0.8377389311790466,
+      "learning_rate": 6.560889196159525e-05,
+      "loss": 0.0571,
+      "step": 77220
+    },
+    {
+      "epoch": 5.052666012430487,
+      "grad_norm": 0.8205317854881287,
+      "learning_rate": 6.560016483442075e-05,
+      "loss": 0.0512,
+      "step": 77230
+    },
+    {
+      "epoch": 5.053320248609748,
+      "grad_norm": 0.6298996806144714,
+      "learning_rate": 6.559143718069977e-05,
+      "loss": 0.0653,
+      "step": 77240
+    },
+    {
+      "epoch": 5.053974484789009,
+      "grad_norm": 1.1496926546096802,
+      "learning_rate": 6.558270900072687e-05,
+      "loss": 0.0769,
+      "step": 77250
+    },
+    {
+      "epoch": 5.054628720968269,
+      "grad_norm": 1.5552153587341309,
+      "learning_rate": 6.557398029479669e-05,
+      "loss": 0.0614,
+      "step": 77260
+    },
+    {
+      "epoch": 5.05528295714753,
+      "grad_norm": 0.9075239300727844,
+      "learning_rate": 6.556525106320382e-05,
+      "loss": 0.0637,
+      "step": 77270
+    },
+    {
+      "epoch": 5.055937193326791,
+      "grad_norm": 0.866381049156189,
+      "learning_rate": 6.555652130624292e-05,
+      "loss": 0.0659,
+      "step": 77280
+    },
+    {
+      "epoch": 5.056591429506052,
+      "grad_norm": 0.8625958561897278,
+      "learning_rate": 6.554779102420863e-05,
+      "loss": 0.0547,
+      "step": 77290
+    },
+    {
+      "epoch": 5.057245665685312,
+      "grad_norm": 0.7577462196350098,
+      "learning_rate": 6.55390602173956e-05,
+      "loss": 0.0527,
+      "step": 77300
+    },
+    {
+      "epoch": 5.057899901864573,
+      "grad_norm": 0.868981659412384,
+      "learning_rate": 6.553032888609856e-05,
+      "loss": 0.0541,
+      "step": 77310
+    },
+    {
+      "epoch": 5.058554138043834,
+      "grad_norm": 1.144863486289978,
+      "learning_rate": 6.552159703061216e-05,
+      "loss": 0.0578,
+      "step": 77320
+    },
+    {
+      "epoch": 5.059208374223094,
+      "grad_norm": 0.9633135795593262,
+      "learning_rate": 6.551286465123118e-05,
+      "loss": 0.0614,
+      "step": 77330
+    },
+    {
+      "epoch": 5.059862610402355,
+      "grad_norm": 0.9470215439796448,
+      "learning_rate": 6.55041317482503e-05,
+      "loss": 0.0568,
+      "step": 77340
+    },
+    {
+      "epoch": 5.060516846581616,
+      "grad_norm": 1.0022468566894531,
+      "learning_rate": 6.549539832196436e-05,
+      "loss": 0.0624,
+      "step": 77350
+    },
+    {
+      "epoch": 5.061171082760876,
+      "grad_norm": 0.7574188113212585,
+      "learning_rate": 6.548666437266806e-05,
+      "loss": 0.0468,
+      "step": 77360
+    },
+    {
+      "epoch": 5.061825318940137,
+      "grad_norm": 0.8999849557876587,
+      "learning_rate": 6.547792990065622e-05,
+      "loss": 0.0618,
+      "step": 77370
+    },
+    {
+      "epoch": 5.062479555119398,
+      "grad_norm": 0.6741798520088196,
+      "learning_rate": 6.546919490622365e-05,
+      "loss": 0.0581,
+      "step": 77380
+    },
+    {
+      "epoch": 5.063133791298659,
+      "grad_norm": 0.9915578365325928,
+      "learning_rate": 6.546045938966518e-05,
+      "loss": 0.0584,
+      "step": 77390
+    },
+    {
+      "epoch": 5.063788027477919,
+      "grad_norm": 0.8000921607017517,
+      "learning_rate": 6.545172335127568e-05,
+      "loss": 0.0561,
+      "step": 77400
+    },
+    {
+      "epoch": 5.06444226365718,
+      "grad_norm": 1.028486728668213,
+      "learning_rate": 6.544298679134998e-05,
+      "loss": 0.0495,
+      "step": 77410
+    },
+    {
+      "epoch": 5.065096499836441,
+      "grad_norm": 0.7525604367256165,
+      "learning_rate": 6.543424971018298e-05,
+      "loss": 0.0567,
+      "step": 77420
+    },
+    {
+      "epoch": 5.065750736015701,
+      "grad_norm": 1.2675774097442627,
+      "learning_rate": 6.542551210806959e-05,
+      "loss": 0.0497,
+      "step": 77430
+    },
+    {
+      "epoch": 5.066404972194962,
+      "grad_norm": 0.8450837731361389,
+      "learning_rate": 6.541677398530468e-05,
+      "loss": 0.0589,
+      "step": 77440
+    },
+    {
+      "epoch": 5.067059208374223,
+      "grad_norm": 0.6929242610931396,
+      "learning_rate": 6.540803534218322e-05,
+      "loss": 0.0615,
+      "step": 77450
+    },
+    {
+      "epoch": 5.067713444553484,
+      "grad_norm": 0.9417129158973694,
+      "learning_rate": 6.539929617900019e-05,
+      "loss": 0.0595,
+      "step": 77460
+    },
+    {
+      "epoch": 5.068367680732744,
+      "grad_norm": 0.944251537322998,
+      "learning_rate": 6.53905564960505e-05,
+      "loss": 0.0538,
+      "step": 77470
+    },
+    {
+      "epoch": 5.069021916912005,
+      "grad_norm": 0.9781545400619507,
+      "learning_rate": 6.538181629362916e-05,
+      "loss": 0.0592,
+      "step": 77480
+    },
+    {
+      "epoch": 5.069676153091266,
+      "grad_norm": 0.9492000937461853,
+      "learning_rate": 6.537307557203119e-05,
+      "loss": 0.0644,
+      "step": 77490
+    },
+    {
+      "epoch": 5.070330389270526,
+      "grad_norm": 0.7973672151565552,
+      "learning_rate": 6.536433433155161e-05,
+      "loss": 0.0495,
+      "step": 77500
+    },
+    {
+      "epoch": 5.070984625449787,
+      "grad_norm": 0.8644813895225525,
+      "learning_rate": 6.535559257248545e-05,
+      "loss": 0.0611,
+      "step": 77510
+    },
+    {
+      "epoch": 5.071638861629048,
+      "grad_norm": 0.9546951651573181,
+      "learning_rate": 6.534685029512777e-05,
+      "loss": 0.061,
+      "step": 77520
+    },
+    {
+      "epoch": 5.072293097808309,
+      "grad_norm": 0.9337143898010254,
+      "learning_rate": 6.533810749977363e-05,
+      "loss": 0.0528,
+      "step": 77530
+    },
+    {
+      "epoch": 5.072947333987569,
+      "grad_norm": 0.8706703186035156,
+      "learning_rate": 6.532936418671815e-05,
+      "loss": 0.0517,
+      "step": 77540
+    },
+    {
+      "epoch": 5.07360157016683,
+      "grad_norm": 0.788536548614502,
+      "learning_rate": 6.532062035625641e-05,
+      "loss": 0.0638,
+      "step": 77550
+    },
+    {
+      "epoch": 5.074255806346091,
+      "grad_norm": 0.8205485343933105,
+      "learning_rate": 6.531187600868357e-05,
+      "loss": 0.0554,
+      "step": 77560
+    },
+    {
+      "epoch": 5.074910042525351,
+      "grad_norm": 0.8340345621109009,
+      "learning_rate": 6.530313114429475e-05,
+      "loss": 0.0518,
+      "step": 77570
+    },
+    {
+      "epoch": 5.075564278704612,
+      "grad_norm": 0.793782651424408,
+      "learning_rate": 6.529438576338512e-05,
+      "loss": 0.052,
+      "step": 77580
+    },
+    {
+      "epoch": 5.076218514883873,
+      "grad_norm": 0.9006875157356262,
+      "learning_rate": 6.528563986624987e-05,
+      "loss": 0.0585,
+      "step": 77590
+    },
+    {
+      "epoch": 5.076872751063134,
+      "grad_norm": 0.9196233749389648,
+      "learning_rate": 6.527689345318416e-05,
+      "loss": 0.0536,
+      "step": 77600
+    },
+    {
+      "epoch": 5.077526987242394,
+      "grad_norm": 0.987467885017395,
+      "learning_rate": 6.526814652448325e-05,
+      "loss": 0.0572,
+      "step": 77610
+    },
+    {
+      "epoch": 5.078181223421655,
+      "grad_norm": 0.994082510471344,
+      "learning_rate": 6.525939908044236e-05,
+      "loss": 0.0542,
+      "step": 77620
+    },
+    {
+      "epoch": 5.078835459600916,
+      "grad_norm": 1.2005406618118286,
+      "learning_rate": 6.525065112135672e-05,
+      "loss": 0.0593,
+      "step": 77630
+    },
+    {
+      "epoch": 5.079489695780176,
+      "grad_norm": 1.0194610357284546,
+      "learning_rate": 6.52419026475216e-05,
+      "loss": 0.0574,
+      "step": 77640
+    },
+    {
+      "epoch": 5.080143931959437,
+      "grad_norm": 0.9802126884460449,
+      "learning_rate": 6.52331536592323e-05,
+      "loss": 0.0615,
+      "step": 77650
+    },
+    {
+      "epoch": 5.080798168138698,
+      "grad_norm": 0.977064311504364,
+      "learning_rate": 6.522440415678413e-05,
+      "loss": 0.0562,
+      "step": 77660
+    },
+    {
+      "epoch": 5.081452404317959,
+      "grad_norm": 0.8457438945770264,
+      "learning_rate": 6.521565414047237e-05,
+      "loss": 0.0578,
+      "step": 77670
+    },
+    {
+      "epoch": 5.082106640497219,
+      "grad_norm": 0.8083365559577942,
+      "learning_rate": 6.52069036105924e-05,
+      "loss": 0.0578,
+      "step": 77680
+    },
+    {
+      "epoch": 5.08276087667648,
+      "grad_norm": 1.0686533451080322,
+      "learning_rate": 6.519815256743954e-05,
+      "loss": 0.0556,
+      "step": 77690
+    },
+    {
+      "epoch": 5.083415112855741,
+      "grad_norm": 0.9897651672363281,
+      "learning_rate": 6.518940101130916e-05,
+      "loss": 0.0556,
+      "step": 77700
+    },
+    {
+      "epoch": 5.084069349035001,
+      "grad_norm": 0.7918063998222351,
+      "learning_rate": 6.518064894249667e-05,
+      "loss": 0.0558,
+      "step": 77710
+    },
+    {
+      "epoch": 5.084723585214262,
+      "grad_norm": 0.7735630869865417,
+      "learning_rate": 6.517189636129749e-05,
+      "loss": 0.0557,
+      "step": 77720
+    },
+    {
+      "epoch": 5.085377821393523,
+      "grad_norm": 0.947536289691925,
+      "learning_rate": 6.516314326800698e-05,
+      "loss": 0.0578,
+      "step": 77730
+    },
+    {
+      "epoch": 5.086032057572784,
+      "grad_norm": 0.74627685546875,
+      "learning_rate": 6.515438966292062e-05,
+      "loss": 0.0529,
+      "step": 77740
+    },
+    {
+      "epoch": 5.086686293752044,
+      "grad_norm": 1.1361474990844727,
+      "learning_rate": 6.514563554633388e-05,
+      "loss": 0.0678,
+      "step": 77750
+    },
+    {
+      "epoch": 5.087340529931305,
+      "grad_norm": 1.06615149974823,
+      "learning_rate": 6.513688091854224e-05,
+      "loss": 0.0656,
+      "step": 77760
+    },
+    {
+      "epoch": 5.087994766110566,
+      "grad_norm": 0.8696459531784058,
+      "learning_rate": 6.512812577984114e-05,
+      "loss": 0.0606,
+      "step": 77770
+    },
+    {
+      "epoch": 5.088649002289826,
+      "grad_norm": 0.7366641163825989,
+      "learning_rate": 6.511937013052612e-05,
+      "loss": 0.0488,
+      "step": 77780
+    },
+    {
+      "epoch": 5.089303238469087,
+      "grad_norm": 0.7504761815071106,
+      "learning_rate": 6.511061397089271e-05,
+      "loss": 0.061,
+      "step": 77790
+    },
+    {
+      "epoch": 5.089957474648348,
+      "grad_norm": 0.8598735332489014,
+      "learning_rate": 6.510185730123646e-05,
+      "loss": 0.0606,
+      "step": 77800
+    },
+    {
+      "epoch": 5.090611710827609,
+      "grad_norm": 0.8309386968612671,
+      "learning_rate": 6.50931001218529e-05,
+      "loss": 0.069,
+      "step": 77810
+    },
+    {
+      "epoch": 5.091265947006869,
+      "grad_norm": 0.9889697432518005,
+      "learning_rate": 6.508434243303764e-05,
+      "loss": 0.0667,
+      "step": 77820
+    },
+    {
+      "epoch": 5.09192018318613,
+      "grad_norm": 0.9573984742164612,
+      "learning_rate": 6.507558423508629e-05,
+      "loss": 0.0594,
+      "step": 77830
+    },
+    {
+      "epoch": 5.092574419365391,
+      "grad_norm": 1.049492359161377,
+      "learning_rate": 6.50668255282944e-05,
+      "loss": 0.0538,
+      "step": 77840
+    },
+    {
+      "epoch": 5.093228655544651,
+      "grad_norm": 0.8158976435661316,
+      "learning_rate": 6.505806631295765e-05,
+      "loss": 0.0592,
+      "step": 77850
+    },
+    {
+      "epoch": 5.093882891723912,
+      "grad_norm": 0.7729289531707764,
+      "learning_rate": 6.504930658937165e-05,
+      "loss": 0.0601,
+      "step": 77860
+    },
+    {
+      "epoch": 5.094537127903173,
+      "grad_norm": 0.7943597435951233,
+      "learning_rate": 6.50405463578321e-05,
+      "loss": 0.0674,
+      "step": 77870
+    },
+    {
+      "epoch": 5.095191364082433,
+      "grad_norm": 1.0894060134887695,
+      "learning_rate": 6.503178561863466e-05,
+      "loss": 0.0594,
+      "step": 77880
+    },
+    {
+      "epoch": 5.095845600261694,
+      "grad_norm": 0.8474149703979492,
+      "learning_rate": 6.502302437207504e-05,
+      "loss": 0.059,
+      "step": 77890
+    },
+    {
+      "epoch": 5.096499836440955,
+      "grad_norm": 0.9915545582771301,
+      "learning_rate": 6.501426261844894e-05,
+      "loss": 0.0529,
+      "step": 77900
+    },
+    {
+      "epoch": 5.097154072620216,
+      "grad_norm": 1.004508137702942,
+      "learning_rate": 6.500550035805212e-05,
+      "loss": 0.0504,
+      "step": 77910
+    },
+    {
+      "epoch": 5.097808308799476,
+      "grad_norm": 0.861196756362915,
+      "learning_rate": 6.499673759118028e-05,
+      "loss": 0.0581,
+      "step": 77920
+    },
+    {
+      "epoch": 5.098462544978737,
+      "grad_norm": 0.7143073081970215,
+      "learning_rate": 6.498797431812923e-05,
+      "loss": 0.057,
+      "step": 77930
+    },
+    {
+      "epoch": 5.099116781157998,
+      "grad_norm": 0.7340794205665588,
+      "learning_rate": 6.497921053919475e-05,
+      "loss": 0.0698,
+      "step": 77940
+    },
+    {
+      "epoch": 5.099771017337258,
+      "grad_norm": 0.691623866558075,
+      "learning_rate": 6.497044625467263e-05,
+      "loss": 0.0555,
+      "step": 77950
+    },
+    {
+      "epoch": 5.100425253516519,
+      "grad_norm": 1.062201738357544,
+      "learning_rate": 6.496168146485865e-05,
+      "loss": 0.0621,
+      "step": 77960
+    },
+    {
+      "epoch": 5.10107948969578,
+      "grad_norm": 0.9496244192123413,
+      "learning_rate": 6.495291617004873e-05,
+      "loss": 0.0566,
+      "step": 77970
+    },
+    {
+      "epoch": 5.101733725875041,
+      "grad_norm": 0.9271469116210938,
+      "learning_rate": 6.494415037053865e-05,
+      "loss": 0.0504,
+      "step": 77980
+    },
+    {
+      "epoch": 5.102387962054301,
+      "grad_norm": 0.8734414577484131,
+      "learning_rate": 6.493538406662429e-05,
+      "loss": 0.0591,
+      "step": 77990
+    },
+    {
+      "epoch": 5.103042198233562,
+      "grad_norm": 0.9724219441413879,
+      "learning_rate": 6.492661725860157e-05,
+      "loss": 0.0653,
+      "step": 78000
+    },
+    {
+      "epoch": 5.103696434412823,
+      "grad_norm": 0.709010124206543,
+      "learning_rate": 6.491784994676637e-05,
+      "loss": 0.0516,
+      "step": 78010
+    },
+    {
+      "epoch": 5.104350670592083,
+      "grad_norm": 0.7753589749336243,
+      "learning_rate": 6.490908213141461e-05,
+      "loss": 0.0519,
+      "step": 78020
+    },
+    {
+      "epoch": 5.105004906771344,
+      "grad_norm": 0.8125455379486084,
+      "learning_rate": 6.490031381284221e-05,
+      "loss": 0.0581,
+      "step": 78030
+    },
+    {
+      "epoch": 5.105659142950605,
+      "grad_norm": 1.1419060230255127,
+      "learning_rate": 6.489154499134517e-05,
+      "loss": 0.0563,
+      "step": 78040
+    },
+    {
+      "epoch": 5.106313379129866,
+      "grad_norm": 0.945682942867279,
+      "learning_rate": 6.488277566721941e-05,
+      "loss": 0.0652,
+      "step": 78050
+    },
+    {
+      "epoch": 5.106967615309126,
+      "grad_norm": 0.9057697653770447,
+      "learning_rate": 6.487400584076094e-05,
+      "loss": 0.0589,
+      "step": 78060
+    },
+    {
+      "epoch": 5.107621851488387,
+      "grad_norm": 0.9359369874000549,
+      "learning_rate": 6.486523551226577e-05,
+      "loss": 0.0612,
+      "step": 78070
+    },
+    {
+      "epoch": 5.108276087667648,
+      "grad_norm": 0.8022433519363403,
+      "learning_rate": 6.485646468202993e-05,
+      "loss": 0.0674,
+      "step": 78080
+    },
+    {
+      "epoch": 5.108930323846908,
+      "grad_norm": 0.8566710948944092,
+      "learning_rate": 6.484769335034942e-05,
+      "loss": 0.0553,
+      "step": 78090
+    },
+    {
+      "epoch": 5.109584560026169,
+      "grad_norm": 0.9492458701133728,
+      "learning_rate": 6.483892151752034e-05,
+      "loss": 0.0629,
+      "step": 78100
+    },
+    {
+      "epoch": 5.11023879620543,
+      "grad_norm": 0.8819661736488342,
+      "learning_rate": 6.483014918383873e-05,
+      "loss": 0.0587,
+      "step": 78110
+    },
+    {
+      "epoch": 5.110893032384691,
+      "grad_norm": 0.9766054749488831,
+      "learning_rate": 6.482137634960068e-05,
+      "loss": 0.0615,
+      "step": 78120
+    },
+    {
+      "epoch": 5.111547268563951,
+      "grad_norm": 0.8624576926231384,
+      "learning_rate": 6.481260301510233e-05,
+      "loss": 0.0602,
+      "step": 78130
+    },
+    {
+      "epoch": 5.112201504743212,
+      "grad_norm": 0.9257481098175049,
+      "learning_rate": 6.480382918063978e-05,
+      "loss": 0.0551,
+      "step": 78140
+    },
+    {
+      "epoch": 5.112855740922473,
+      "grad_norm": 0.9216415882110596,
+      "learning_rate": 6.479505484650916e-05,
+      "loss": 0.0538,
+      "step": 78150
+    },
+    {
+      "epoch": 5.113509977101733,
+      "grad_norm": 0.7884585857391357,
+      "learning_rate": 6.478628001300664e-05,
+      "loss": 0.0556,
+      "step": 78160
+    },
+    {
+      "epoch": 5.114164213280994,
+      "grad_norm": 0.9612773060798645,
+      "learning_rate": 6.477750468042841e-05,
+      "loss": 0.0561,
+      "step": 78170
+    },
+    {
+      "epoch": 5.114818449460255,
+      "grad_norm": 0.7759575247764587,
+      "learning_rate": 6.476872884907062e-05,
+      "loss": 0.0633,
+      "step": 78180
+    },
+    {
+      "epoch": 5.115472685639516,
+      "grad_norm": 0.9810332655906677,
+      "learning_rate": 6.475995251922949e-05,
+      "loss": 0.0571,
+      "step": 78190
+    },
+    {
+      "epoch": 5.116126921818776,
+      "grad_norm": 0.7260820865631104,
+      "learning_rate": 6.475117569120127e-05,
+      "loss": 0.0624,
+      "step": 78200
+    },
+    {
+      "epoch": 5.116781157998037,
+      "grad_norm": 0.998258113861084,
+      "learning_rate": 6.474239836528219e-05,
+      "loss": 0.055,
+      "step": 78210
+    },
+    {
+      "epoch": 5.117435394177298,
+      "grad_norm": 0.7317970395088196,
+      "learning_rate": 6.473362054176847e-05,
+      "loss": 0.0594,
+      "step": 78220
+    },
+    {
+      "epoch": 5.118089630356558,
+      "grad_norm": 1.0319292545318604,
+      "learning_rate": 6.472484222095645e-05,
+      "loss": 0.0558,
+      "step": 78230
+    },
+    {
+      "epoch": 5.118743866535819,
+      "grad_norm": 0.6902409195899963,
+      "learning_rate": 6.471606340314238e-05,
+      "loss": 0.0631,
+      "step": 78240
+    },
+    {
+      "epoch": 5.11939810271508,
+      "grad_norm": 1.0602803230285645,
+      "learning_rate": 6.470728408862257e-05,
+      "loss": 0.053,
+      "step": 78250
+    },
+    {
+      "epoch": 5.120052338894341,
+      "grad_norm": 0.9925687313079834,
+      "learning_rate": 6.469850427769336e-05,
+      "loss": 0.0685,
+      "step": 78260
+    },
+    {
+      "epoch": 5.120706575073601,
+      "grad_norm": 0.8387590050697327,
+      "learning_rate": 6.468972397065108e-05,
+      "loss": 0.0516,
+      "step": 78270
+    },
+    {
+      "epoch": 5.121360811252862,
+      "grad_norm": 0.93202805519104,
+      "learning_rate": 6.468094316779207e-05,
+      "loss": 0.0657,
+      "step": 78280
+    },
+    {
+      "epoch": 5.122015047432123,
+      "grad_norm": 0.8118649125099182,
+      "learning_rate": 6.467216186941274e-05,
+      "loss": 0.0504,
+      "step": 78290
+    },
+    {
+      "epoch": 5.122669283611383,
+      "grad_norm": 0.8819132447242737,
+      "learning_rate": 6.466338007580948e-05,
+      "loss": 0.0667,
+      "step": 78300
+    },
+    {
+      "epoch": 5.123323519790644,
+      "grad_norm": 1.2045010328292847,
+      "learning_rate": 6.465459778727867e-05,
+      "loss": 0.068,
+      "step": 78310
+    },
+    {
+      "epoch": 5.123977755969905,
+      "grad_norm": 0.8089714646339417,
+      "learning_rate": 6.464581500411675e-05,
+      "loss": 0.0599,
+      "step": 78320
+    },
+    {
+      "epoch": 5.124631992149165,
+      "grad_norm": 0.8597386479377747,
+      "learning_rate": 6.463703172662019e-05,
+      "loss": 0.0561,
+      "step": 78330
+    },
+    {
+      "epoch": 5.125286228328426,
+      "grad_norm": 0.7467419505119324,
+      "learning_rate": 6.46282479550854e-05,
+      "loss": 0.0557,
+      "step": 78340
+    },
+    {
+      "epoch": 5.125940464507687,
+      "grad_norm": 0.8646554946899414,
+      "learning_rate": 6.461946368980888e-05,
+      "loss": 0.0532,
+      "step": 78350
+    },
+    {
+      "epoch": 5.126594700686948,
+      "grad_norm": 0.681583821773529,
+      "learning_rate": 6.461067893108712e-05,
+      "loss": 0.0543,
+      "step": 78360
+    },
+    {
+      "epoch": 5.127248936866208,
+      "grad_norm": 0.9374173879623413,
+      "learning_rate": 6.460189367921663e-05,
+      "loss": 0.0547,
+      "step": 78370
+    },
+    {
+      "epoch": 5.127903173045469,
+      "grad_norm": 0.9243113398551941,
+      "learning_rate": 6.459310793449391e-05,
+      "loss": 0.0538,
+      "step": 78380
+    },
+    {
+      "epoch": 5.12855740922473,
+      "grad_norm": 0.8543265461921692,
+      "learning_rate": 6.458432169721556e-05,
+      "loss": 0.0603,
+      "step": 78390
+    },
+    {
+      "epoch": 5.12921164540399,
+      "grad_norm": 0.811305046081543,
+      "learning_rate": 6.457553496767809e-05,
+      "loss": 0.0606,
+      "step": 78400
+    },
+    {
+      "epoch": 5.129865881583251,
+      "grad_norm": 0.9650918245315552,
+      "learning_rate": 6.456674774617809e-05,
+      "loss": 0.0528,
+      "step": 78410
+    },
+    {
+      "epoch": 5.130520117762512,
+      "grad_norm": 0.989138126373291,
+      "learning_rate": 6.455796003301215e-05,
+      "loss": 0.0625,
+      "step": 78420
+    },
+    {
+      "epoch": 5.131174353941773,
+      "grad_norm": 0.9064369201660156,
+      "learning_rate": 6.45491718284769e-05,
+      "loss": 0.0566,
+      "step": 78430
+    },
+    {
+      "epoch": 5.131828590121033,
+      "grad_norm": 0.7815073728561401,
+      "learning_rate": 6.454038313286891e-05,
+      "loss": 0.0654,
+      "step": 78440
+    },
+    {
+      "epoch": 5.132482826300294,
+      "grad_norm": 0.9094436168670654,
+      "learning_rate": 6.453159394648487e-05,
+      "loss": 0.0573,
+      "step": 78450
+    },
+    {
+      "epoch": 5.133137062479555,
+      "grad_norm": 0.8054584860801697,
+      "learning_rate": 6.452280426962143e-05,
+      "loss": 0.054,
+      "step": 78460
+    },
+    {
+      "epoch": 5.1337912986588154,
+      "grad_norm": 0.8743946552276611,
+      "learning_rate": 6.451401410257525e-05,
+      "loss": 0.0594,
+      "step": 78470
+    },
+    {
+      "epoch": 5.134445534838076,
+      "grad_norm": 0.9614243507385254,
+      "learning_rate": 6.450522344564303e-05,
+      "loss": 0.0619,
+      "step": 78480
+    },
+    {
+      "epoch": 5.135099771017337,
+      "grad_norm": 0.9176281094551086,
+      "learning_rate": 6.449643229912148e-05,
+      "loss": 0.0582,
+      "step": 78490
+    },
+    {
+      "epoch": 5.135754007196598,
+      "grad_norm": 0.9308127164840698,
+      "learning_rate": 6.448764066330733e-05,
+      "loss": 0.0574,
+      "step": 78500
+    },
+    {
+      "epoch": 5.136408243375858,
+      "grad_norm": 0.8228217363357544,
+      "learning_rate": 6.44788485384973e-05,
+      "loss": 0.0693,
+      "step": 78510
+    },
+    {
+      "epoch": 5.137062479555119,
+      "grad_norm": 0.9158545136451721,
+      "learning_rate": 6.447005592498816e-05,
+      "loss": 0.0499,
+      "step": 78520
+    },
+    {
+      "epoch": 5.13771671573438,
+      "grad_norm": 0.8017662763595581,
+      "learning_rate": 6.446126282307669e-05,
+      "loss": 0.0561,
+      "step": 78530
+    },
+    {
+      "epoch": 5.1383709519136405,
+      "grad_norm": 0.8243115544319153,
+      "learning_rate": 6.445246923305966e-05,
+      "loss": 0.056,
+      "step": 78540
+    },
+    {
+      "epoch": 5.139025188092901,
+      "grad_norm": 0.7880877256393433,
+      "learning_rate": 6.44436751552339e-05,
+      "loss": 0.0569,
+      "step": 78550
+    },
+    {
+      "epoch": 5.139679424272162,
+      "grad_norm": 0.9965237379074097,
+      "learning_rate": 6.443488058989624e-05,
+      "loss": 0.0646,
+      "step": 78560
+    },
+    {
+      "epoch": 5.140333660451423,
+      "grad_norm": 0.9175546765327454,
+      "learning_rate": 6.442608553734348e-05,
+      "loss": 0.0679,
+      "step": 78570
+    },
+    {
+      "epoch": 5.1409878966306835,
+      "grad_norm": 0.8487790822982788,
+      "learning_rate": 6.441728999787251e-05,
+      "loss": 0.0511,
+      "step": 78580
+    },
+    {
+      "epoch": 5.141642132809944,
+      "grad_norm": 0.9188631772994995,
+      "learning_rate": 6.44084939717802e-05,
+      "loss": 0.0572,
+      "step": 78590
+    },
+    {
+      "epoch": 5.142296368989205,
+      "grad_norm": 0.8880249857902527,
+      "learning_rate": 6.439969745936341e-05,
+      "loss": 0.0729,
+      "step": 78600
+    },
+    {
+      "epoch": 5.1429506051684655,
+      "grad_norm": 0.7750711441040039,
+      "learning_rate": 6.439090046091907e-05,
+      "loss": 0.062,
+      "step": 78610
+    },
+    {
+      "epoch": 5.143604841347726,
+      "grad_norm": 0.8902852535247803,
+      "learning_rate": 6.438210297674411e-05,
+      "loss": 0.0582,
+      "step": 78620
+    },
+    {
+      "epoch": 5.144259077526987,
+      "grad_norm": 0.9351129531860352,
+      "learning_rate": 6.437330500713545e-05,
+      "loss": 0.0535,
+      "step": 78630
+    },
+    {
+      "epoch": 5.144913313706248,
+      "grad_norm": 0.8082217574119568,
+      "learning_rate": 6.436450655239004e-05,
+      "loss": 0.0601,
+      "step": 78640
+    },
+    {
+      "epoch": 5.1455675498855085,
+      "grad_norm": 0.9158061146736145,
+      "learning_rate": 6.435570761280487e-05,
+      "loss": 0.0562,
+      "step": 78650
+    },
+    {
+      "epoch": 5.146221786064769,
+      "grad_norm": 0.8845149278640747,
+      "learning_rate": 6.434690818867693e-05,
+      "loss": 0.0515,
+      "step": 78660
+    },
+    {
+      "epoch": 5.14687602224403,
+      "grad_norm": 0.9246596693992615,
+      "learning_rate": 6.43381082803032e-05,
+      "loss": 0.0547,
+      "step": 78670
+    },
+    {
+      "epoch": 5.1475302584232905,
+      "grad_norm": 0.8602831363677979,
+      "learning_rate": 6.432930788798072e-05,
+      "loss": 0.0475,
+      "step": 78680
+    },
+    {
+      "epoch": 5.1481844946025515,
+      "grad_norm": 0.7768710255622864,
+      "learning_rate": 6.432050701200651e-05,
+      "loss": 0.0519,
+      "step": 78690
+    },
+    {
+      "epoch": 5.148838730781812,
+      "grad_norm": 0.8790880441665649,
+      "learning_rate": 6.431170565267764e-05,
+      "loss": 0.0609,
+      "step": 78700
+    },
+    {
+      "epoch": 5.149492966961073,
+      "grad_norm": 1.016045331954956,
+      "learning_rate": 6.430290381029116e-05,
+      "loss": 0.0583,
+      "step": 78710
+    },
+    {
+      "epoch": 5.1501472031403335,
+      "grad_norm": 1.148781418800354,
+      "learning_rate": 6.429410148514419e-05,
+      "loss": 0.0667,
+      "step": 78720
+    },
+    {
+      "epoch": 5.150801439319594,
+      "grad_norm": 0.873008668422699,
+      "learning_rate": 6.428529867753377e-05,
+      "loss": 0.0617,
+      "step": 78730
+    },
+    {
+      "epoch": 5.151455675498855,
+      "grad_norm": 0.7208111882209778,
+      "learning_rate": 6.427649538775708e-05,
+      "loss": 0.0633,
+      "step": 78740
+    },
+    {
+      "epoch": 5.1521099116781155,
+      "grad_norm": 1.1463489532470703,
+      "learning_rate": 6.426769161611121e-05,
+      "loss": 0.056,
+      "step": 78750
+    },
+    {
+      "epoch": 5.1527641478573765,
+      "grad_norm": 0.7651934623718262,
+      "learning_rate": 6.425888736289336e-05,
+      "loss": 0.0522,
+      "step": 78760
+    },
+    {
+      "epoch": 5.153418384036637,
+      "grad_norm": 0.9302027821540833,
+      "learning_rate": 6.425008262840064e-05,
+      "loss": 0.063,
+      "step": 78770
+    },
+    {
+      "epoch": 5.1540726202158975,
+      "grad_norm": 0.8309484720230103,
+      "learning_rate": 6.424127741293027e-05,
+      "loss": 0.0663,
+      "step": 78780
+    },
+    {
+      "epoch": 5.1547268563951585,
+      "grad_norm": 0.8712273240089417,
+      "learning_rate": 6.423247171677943e-05,
+      "loss": 0.0648,
+      "step": 78790
+    },
+    {
+      "epoch": 5.1553810925744195,
+      "grad_norm": 0.9975765943527222,
+      "learning_rate": 6.422366554024536e-05,
+      "loss": 0.0636,
+      "step": 78800
+    },
+    {
+      "epoch": 5.15603532875368,
+      "grad_norm": 1.0808351039886475,
+      "learning_rate": 6.421485888362526e-05,
+      "loss": 0.0615,
+      "step": 78810
+    },
+    {
+      "epoch": 5.1566895649329405,
+      "grad_norm": 0.7145622968673706,
+      "learning_rate": 6.42060517472164e-05,
+      "loss": 0.0601,
+      "step": 78820
+    },
+    {
+      "epoch": 5.1573438011122015,
+      "grad_norm": 0.8157432675361633,
+      "learning_rate": 6.419724413131604e-05,
+      "loss": 0.0544,
+      "step": 78830
+    },
+    {
+      "epoch": 5.1579980372914624,
+      "grad_norm": 0.8889109492301941,
+      "learning_rate": 6.418843603622144e-05,
+      "loss": 0.0558,
+      "step": 78840
+    },
+    {
+      "epoch": 5.1586522734707225,
+      "grad_norm": 0.8644919991493225,
+      "learning_rate": 6.417962746222993e-05,
+      "loss": 0.057,
+      "step": 78850
+    },
+    {
+      "epoch": 5.1593065096499835,
+      "grad_norm": 0.9862198233604431,
+      "learning_rate": 6.41708184096388e-05,
+      "loss": 0.0577,
+      "step": 78860
+    },
+    {
+      "epoch": 5.1599607458292445,
+      "grad_norm": 0.806100606918335,
+      "learning_rate": 6.41620088787454e-05,
+      "loss": 0.0588,
+      "step": 78870
+    },
+    {
+      "epoch": 5.160614982008505,
+      "grad_norm": 0.9563096165657043,
+      "learning_rate": 6.415319886984703e-05,
+      "loss": 0.0603,
+      "step": 78880
+    },
+    {
+      "epoch": 5.1612692181877655,
+      "grad_norm": 0.9669555425643921,
+      "learning_rate": 6.41443883832411e-05,
+      "loss": 0.0509,
+      "step": 78890
+    },
+    {
+      "epoch": 5.1619234543670265,
+      "grad_norm": 0.8312886357307434,
+      "learning_rate": 6.413557741922495e-05,
+      "loss": 0.056,
+      "step": 78900
+    },
+    {
+      "epoch": 5.1625776905462875,
+      "grad_norm": 0.8751150369644165,
+      "learning_rate": 6.412676597809602e-05,
+      "loss": 0.0561,
+      "step": 78910
+    },
+    {
+      "epoch": 5.1632319267255475,
+      "grad_norm": 0.9386425018310547,
+      "learning_rate": 6.411795406015166e-05,
+      "loss": 0.0561,
+      "step": 78920
+    },
+    {
+      "epoch": 5.1638861629048085,
+      "grad_norm": 1.336055874824524,
+      "learning_rate": 6.410914166568933e-05,
+      "loss": 0.0501,
+      "step": 78930
+    },
+    {
+      "epoch": 5.1645403990840695,
+      "grad_norm": 0.7663227915763855,
+      "learning_rate": 6.410032879500647e-05,
+      "loss": 0.0583,
+      "step": 78940
+    },
+    {
+      "epoch": 5.1651946352633304,
+      "grad_norm": 0.8238208889961243,
+      "learning_rate": 6.409151544840055e-05,
+      "loss": 0.0637,
+      "step": 78950
+    },
+    {
+      "epoch": 5.1658488714425905,
+      "grad_norm": 0.8930804133415222,
+      "learning_rate": 6.4082701626169e-05,
+      "loss": 0.0519,
+      "step": 78960
+    },
+    {
+      "epoch": 5.1665031076218515,
+      "grad_norm": 0.7918708324432373,
+      "learning_rate": 6.407388732860935e-05,
+      "loss": 0.0553,
+      "step": 78970
+    },
+    {
+      "epoch": 5.1671573438011125,
+      "grad_norm": 0.7444891333580017,
+      "learning_rate": 6.40650725560191e-05,
+      "loss": 0.0473,
+      "step": 78980
+    },
+    {
+      "epoch": 5.1678115799803725,
+      "grad_norm": 0.9502220749855042,
+      "learning_rate": 6.405625730869575e-05,
+      "loss": 0.057,
+      "step": 78990
+    },
+    {
+      "epoch": 5.1684658161596335,
+      "grad_norm": 0.8537470698356628,
+      "learning_rate": 6.404744158693685e-05,
+      "loss": 0.0563,
+      "step": 79000
+    },
+    {
+      "epoch": 5.1691200523388945,
+      "grad_norm": 0.902565062046051,
+      "learning_rate": 6.403862539103998e-05,
+      "loss": 0.0536,
+      "step": 79010
+    },
+    {
+      "epoch": 5.1697742885181555,
+      "grad_norm": 0.9885287284851074,
+      "learning_rate": 6.402980872130266e-05,
+      "loss": 0.0589,
+      "step": 79020
+    },
+    {
+      "epoch": 5.1704285246974155,
+      "grad_norm": 0.8051989674568176,
+      "learning_rate": 6.402099157802252e-05,
+      "loss": 0.0661,
+      "step": 79030
+    },
+    {
+      "epoch": 5.1710827608766765,
+      "grad_norm": 1.0724586248397827,
+      "learning_rate": 6.401217396149713e-05,
+      "loss": 0.0528,
+      "step": 79040
+    },
+    {
+      "epoch": 5.1717369970559375,
+      "grad_norm": 0.9905614256858826,
+      "learning_rate": 6.400335587202413e-05,
+      "loss": 0.0511,
+      "step": 79050
+    },
+    {
+      "epoch": 5.172391233235198,
+      "grad_norm": 0.8915742635726929,
+      "learning_rate": 6.399453730990113e-05,
+      "loss": 0.0595,
+      "step": 79060
+    },
+    {
+      "epoch": 5.1730454694144585,
+      "grad_norm": 0.8688280582427979,
+      "learning_rate": 6.398571827542581e-05,
+      "loss": 0.0592,
+      "step": 79070
+    },
+    {
+      "epoch": 5.1736997055937195,
+      "grad_norm": 0.9881342053413391,
+      "learning_rate": 6.39768987688958e-05,
+      "loss": 0.0612,
+      "step": 79080
+    },
+    {
+      "epoch": 5.1743539417729805,
+      "grad_norm": 0.8909708261489868,
+      "learning_rate": 6.396807879060882e-05,
+      "loss": 0.0552,
+      "step": 79090
+    },
+    {
+      "epoch": 5.1750081779522406,
+      "grad_norm": 1.0886934995651245,
+      "learning_rate": 6.395925834086254e-05,
+      "loss": 0.0531,
+      "step": 79100
+    },
+    {
+      "epoch": 5.1756624141315015,
+      "grad_norm": 0.8756330609321594,
+      "learning_rate": 6.395043741995468e-05,
+      "loss": 0.0536,
+      "step": 79110
+    },
+    {
+      "epoch": 5.1763166503107625,
+      "grad_norm": 0.8764585852622986,
+      "learning_rate": 6.394161602818296e-05,
+      "loss": 0.0525,
+      "step": 79120
+    },
+    {
+      "epoch": 5.176970886490023,
+      "grad_norm": 0.8468575477600098,
+      "learning_rate": 6.393279416584516e-05,
+      "loss": 0.0644,
+      "step": 79130
+    },
+    {
+      "epoch": 5.1776251226692835,
+      "grad_norm": 0.982122540473938,
+      "learning_rate": 6.392397183323901e-05,
+      "loss": 0.0528,
+      "step": 79140
+    },
+    {
+      "epoch": 5.1782793588485445,
+      "grad_norm": 0.9616347551345825,
+      "learning_rate": 6.391514903066228e-05,
+      "loss": 0.0526,
+      "step": 79150
+    },
+    {
+      "epoch": 5.1789335950278055,
+      "grad_norm": 0.8660449981689453,
+      "learning_rate": 6.390632575841278e-05,
+      "loss": 0.0643,
+      "step": 79160
+    },
+    {
+      "epoch": 5.179587831207066,
+      "grad_norm": 0.7441914081573486,
+      "learning_rate": 6.389750201678833e-05,
+      "loss": 0.0664,
+      "step": 79170
+    },
+    {
+      "epoch": 5.1802420673863265,
+      "grad_norm": 0.796704888343811,
+      "learning_rate": 6.388867780608672e-05,
+      "loss": 0.0534,
+      "step": 79180
+    },
+    {
+      "epoch": 5.1808963035655875,
+      "grad_norm": 0.6812049746513367,
+      "learning_rate": 6.387985312660582e-05,
+      "loss": 0.0643,
+      "step": 79190
+    },
+    {
+      "epoch": 5.181550539744848,
+      "grad_norm": 0.8211969137191772,
+      "learning_rate": 6.387102797864349e-05,
+      "loss": 0.0599,
+      "step": 79200
+    },
+    {
+      "epoch": 5.1822047759241086,
+      "grad_norm": 1.1284834146499634,
+      "learning_rate": 6.386220236249757e-05,
+      "loss": 0.0679,
+      "step": 79210
+    },
+    {
+      "epoch": 5.1828590121033695,
+      "grad_norm": 0.8889693021774292,
+      "learning_rate": 6.385337627846596e-05,
+      "loss": 0.0503,
+      "step": 79220
+    },
+    {
+      "epoch": 5.18351324828263,
+      "grad_norm": 0.8759040236473083,
+      "learning_rate": 6.384454972684658e-05,
+      "loss": 0.0576,
+      "step": 79230
+    },
+    {
+      "epoch": 5.184167484461891,
+      "grad_norm": 0.808881938457489,
+      "learning_rate": 6.383572270793733e-05,
+      "loss": 0.0624,
+      "step": 79240
+    },
+    {
+      "epoch": 5.1848217206411515,
+      "grad_norm": 0.8137565851211548,
+      "learning_rate": 6.382689522203616e-05,
+      "loss": 0.0621,
+      "step": 79250
+    },
+    {
+      "epoch": 5.1854759568204125,
+      "grad_norm": 1.0435824394226074,
+      "learning_rate": 6.381806726944101e-05,
+      "loss": 0.0673,
+      "step": 79260
+    },
+    {
+      "epoch": 5.186130192999673,
+      "grad_norm": 0.6479164361953735,
+      "learning_rate": 6.380923885044985e-05,
+      "loss": 0.0537,
+      "step": 79270
+    },
+    {
+      "epoch": 5.186784429178934,
+      "grad_norm": 0.784147322177887,
+      "learning_rate": 6.380040996536067e-05,
+      "loss": 0.0504,
+      "step": 79280
+    },
+    {
+      "epoch": 5.1874386653581945,
+      "grad_norm": 0.8886092901229858,
+      "learning_rate": 6.379158061447145e-05,
+      "loss": 0.0666,
+      "step": 79290
+    },
+    {
+      "epoch": 5.188092901537455,
+      "grad_norm": 0.8956160545349121,
+      "learning_rate": 6.378275079808022e-05,
+      "loss": 0.0645,
+      "step": 79300
+    },
+    {
+      "epoch": 5.188747137716716,
+      "grad_norm": 0.7196115851402283,
+      "learning_rate": 6.3773920516485e-05,
+      "loss": 0.0659,
+      "step": 79310
+    },
+    {
+      "epoch": 5.1894013738959766,
+      "grad_norm": 0.9113817811012268,
+      "learning_rate": 6.376508976998386e-05,
+      "loss": 0.0635,
+      "step": 79320
+    },
+    {
+      "epoch": 5.1900556100752375,
+      "grad_norm": 1.1962367296218872,
+      "learning_rate": 6.375625855887481e-05,
+      "loss": 0.0569,
+      "step": 79330
+    },
+    {
+      "epoch": 5.190709846254498,
+      "grad_norm": 0.8185920715332031,
+      "learning_rate": 6.374742688345598e-05,
+      "loss": 0.0622,
+      "step": 79340
+    },
+    {
+      "epoch": 5.191364082433759,
+      "grad_norm": 0.9433519244194031,
+      "learning_rate": 6.373859474402542e-05,
+      "loss": 0.0622,
+      "step": 79350
+    },
+    {
+      "epoch": 5.1920183186130195,
+      "grad_norm": 0.8798799514770508,
+      "learning_rate": 6.372976214088127e-05,
+      "loss": 0.0532,
+      "step": 79360
+    },
+    {
+      "epoch": 5.19267255479228,
+      "grad_norm": 0.9233197569847107,
+      "learning_rate": 6.372092907432163e-05,
+      "loss": 0.0551,
+      "step": 79370
+    },
+    {
+      "epoch": 5.193326790971541,
+      "grad_norm": 1.0525169372558594,
+      "learning_rate": 6.371209554464465e-05,
+      "loss": 0.0567,
+      "step": 79380
+    },
+    {
+      "epoch": 5.193981027150802,
+      "grad_norm": 0.9267981648445129,
+      "learning_rate": 6.37032615521485e-05,
+      "loss": 0.0595,
+      "step": 79390
+    },
+    {
+      "epoch": 5.1946352633300625,
+      "grad_norm": 0.8152212500572205,
+      "learning_rate": 6.369442709713132e-05,
+      "loss": 0.0625,
+      "step": 79400
+    },
+    {
+      "epoch": 5.195289499509323,
+      "grad_norm": 0.7746557593345642,
+      "learning_rate": 6.36855921798913e-05,
+      "loss": 0.0582,
+      "step": 79410
+    },
+    {
+      "epoch": 5.195943735688584,
+      "grad_norm": 0.8332052826881409,
+      "learning_rate": 6.367675680072668e-05,
+      "loss": 0.061,
+      "step": 79420
+    },
+    {
+      "epoch": 5.196597971867845,
+      "grad_norm": 0.7936050295829773,
+      "learning_rate": 6.366792095993563e-05,
+      "loss": 0.0552,
+      "step": 79430
+    },
+    {
+      "epoch": 5.197252208047105,
+      "grad_norm": 0.8946577906608582,
+      "learning_rate": 6.365908465781641e-05,
+      "loss": 0.063,
+      "step": 79440
+    },
+    {
+      "epoch": 5.197906444226366,
+      "grad_norm": 1.0016376972198486,
+      "learning_rate": 6.365024789466725e-05,
+      "loss": 0.0655,
+      "step": 79450
+    },
+    {
+      "epoch": 5.198560680405627,
+      "grad_norm": 0.8920219540596008,
+      "learning_rate": 6.364141067078645e-05,
+      "loss": 0.0636,
+      "step": 79460
+    },
+    {
+      "epoch": 5.1992149165848875,
+      "grad_norm": 0.9238928556442261,
+      "learning_rate": 6.363257298647224e-05,
+      "loss": 0.0572,
+      "step": 79470
+    },
+    {
+      "epoch": 5.199869152764148,
+      "grad_norm": 0.8631906509399414,
+      "learning_rate": 6.362373484202294e-05,
+      "loss": 0.0568,
+      "step": 79480
+    },
+    {
+      "epoch": 5.200523388943409,
+      "grad_norm": 0.8432829976081848,
+      "learning_rate": 6.361489623773686e-05,
+      "loss": 0.0616,
+      "step": 79490
+    },
+    {
+      "epoch": 5.20117762512267,
+      "grad_norm": 0.8368816375732422,
+      "learning_rate": 6.360605717391235e-05,
+      "loss": 0.0518,
+      "step": 79500
+    },
+    {
+      "epoch": 5.20183186130193,
+      "grad_norm": 0.8626534342765808,
+      "learning_rate": 6.35972176508477e-05,
+      "loss": 0.0623,
+      "step": 79510
+    },
+    {
+      "epoch": 5.202486097481191,
+      "grad_norm": 0.7735602259635925,
+      "learning_rate": 6.35883776688413e-05,
+      "loss": 0.0494,
+      "step": 79520
+    },
+    {
+      "epoch": 5.203140333660452,
+      "grad_norm": 0.8893035650253296,
+      "learning_rate": 6.357953722819151e-05,
+      "loss": 0.0573,
+      "step": 79530
+    },
+    {
+      "epoch": 5.203794569839713,
+      "grad_norm": 0.8206644058227539,
+      "learning_rate": 6.357069632919674e-05,
+      "loss": 0.0615,
+      "step": 79540
+    },
+    {
+      "epoch": 5.204448806018973,
+      "grad_norm": 1.0286612510681152,
+      "learning_rate": 6.356185497215537e-05,
+      "loss": 0.0694,
+      "step": 79550
+    },
+    {
+      "epoch": 5.205103042198234,
+      "grad_norm": 1.0252892971038818,
+      "learning_rate": 6.355301315736583e-05,
+      "loss": 0.0615,
+      "step": 79560
+    },
+    {
+      "epoch": 5.205757278377495,
+      "grad_norm": 0.8636437058448792,
+      "learning_rate": 6.354417088512655e-05,
+      "loss": 0.0534,
+      "step": 79570
+    },
+    {
+      "epoch": 5.206411514556755,
+      "grad_norm": 0.9313245415687561,
+      "learning_rate": 6.353532815573599e-05,
+      "loss": 0.054,
+      "step": 79580
+    },
+    {
+      "epoch": 5.207065750736016,
+      "grad_norm": 0.9524334073066711,
+      "learning_rate": 6.35264849694926e-05,
+      "loss": 0.0564,
+      "step": 79590
+    },
+    {
+      "epoch": 5.207719986915277,
+      "grad_norm": 0.7033629417419434,
+      "learning_rate": 6.351764132669486e-05,
+      "loss": 0.0526,
+      "step": 79600
+    },
+    {
+      "epoch": 5.208374223094538,
+      "grad_norm": 0.8188180923461914,
+      "learning_rate": 6.350879722764129e-05,
+      "loss": 0.0498,
+      "step": 79610
+    },
+    {
+      "epoch": 5.209028459273798,
+      "grad_norm": 0.8515734672546387,
+      "learning_rate": 6.349995267263038e-05,
+      "loss": 0.0608,
+      "step": 79620
+    },
+    {
+      "epoch": 5.209682695453059,
+      "grad_norm": 0.7757560014724731,
+      "learning_rate": 6.349110766196065e-05,
+      "loss": 0.0531,
+      "step": 79630
+    },
+    {
+      "epoch": 5.21033693163232,
+      "grad_norm": 0.9493059515953064,
+      "learning_rate": 6.348226219593066e-05,
+      "loss": 0.0522,
+      "step": 79640
+    },
+    {
+      "epoch": 5.21099116781158,
+      "grad_norm": 0.7367783188819885,
+      "learning_rate": 6.347341627483897e-05,
+      "loss": 0.0652,
+      "step": 79650
+    },
+    {
+      "epoch": 5.211645403990841,
+      "grad_norm": 0.8916527628898621,
+      "learning_rate": 6.346456989898415e-05,
+      "loss": 0.0641,
+      "step": 79660
+    },
+    {
+      "epoch": 5.212299640170102,
+      "grad_norm": 0.9219633936882019,
+      "learning_rate": 6.345572306866477e-05,
+      "loss": 0.0557,
+      "step": 79670
+    },
+    {
+      "epoch": 5.212953876349362,
+      "grad_norm": 0.934571385383606,
+      "learning_rate": 6.344687578417945e-05,
+      "loss": 0.0571,
+      "step": 79680
+    },
+    {
+      "epoch": 5.213608112528623,
+      "grad_norm": 0.8010873198509216,
+      "learning_rate": 6.343802804582681e-05,
+      "loss": 0.0538,
+      "step": 79690
+    },
+    {
+      "epoch": 5.214262348707884,
+      "grad_norm": 1.0014433860778809,
+      "learning_rate": 6.342917985390548e-05,
+      "loss": 0.0588,
+      "step": 79700
+    },
+    {
+      "epoch": 5.214916584887145,
+      "grad_norm": 0.7398951649665833,
+      "learning_rate": 6.342033120871411e-05,
+      "loss": 0.0599,
+      "step": 79710
+    },
+    {
+      "epoch": 5.215570821066405,
+      "grad_norm": 0.9867905974388123,
+      "learning_rate": 6.341148211055138e-05,
+      "loss": 0.0536,
+      "step": 79720
+    },
+    {
+      "epoch": 5.216225057245666,
+      "grad_norm": 0.8380576372146606,
+      "learning_rate": 6.340263255971594e-05,
+      "loss": 0.0582,
+      "step": 79730
+    },
+    {
+      "epoch": 5.216879293424927,
+      "grad_norm": 0.7339340448379517,
+      "learning_rate": 6.33937825565065e-05,
+      "loss": 0.0551,
+      "step": 79740
+    },
+    {
+      "epoch": 5.217533529604187,
+      "grad_norm": 0.9456587433815002,
+      "learning_rate": 6.338493210122177e-05,
+      "loss": 0.0566,
+      "step": 79750
+    },
+    {
+      "epoch": 5.218187765783448,
+      "grad_norm": 0.8973842859268188,
+      "learning_rate": 6.33760811941605e-05,
+      "loss": 0.0538,
+      "step": 79760
+    },
+    {
+      "epoch": 5.218842001962709,
+      "grad_norm": 1.1991941928863525,
+      "learning_rate": 6.336722983562138e-05,
+      "loss": 0.0595,
+      "step": 79770
+    },
+    {
+      "epoch": 5.21949623814197,
+      "grad_norm": 0.8048141598701477,
+      "learning_rate": 6.335837802590322e-05,
+      "loss": 0.0571,
+      "step": 79780
+    },
+    {
+      "epoch": 5.22015047432123,
+      "grad_norm": 0.7296687960624695,
+      "learning_rate": 6.334952576530475e-05,
+      "loss": 0.0605,
+      "step": 79790
+    },
+    {
+      "epoch": 5.220804710500491,
+      "grad_norm": 1.0618641376495361,
+      "learning_rate": 6.334067305412479e-05,
+      "loss": 0.0621,
+      "step": 79800
+    },
+    {
+      "epoch": 5.221458946679752,
+      "grad_norm": 1.0293445587158203,
+      "learning_rate": 6.333181989266213e-05,
+      "loss": 0.0624,
+      "step": 79810
+    },
+    {
+      "epoch": 5.222113182859012,
+      "grad_norm": 0.7226553559303284,
+      "learning_rate": 6.332296628121557e-05,
+      "loss": 0.0597,
+      "step": 79820
+    },
+    {
+      "epoch": 5.222767419038273,
+      "grad_norm": 0.8370351195335388,
+      "learning_rate": 6.331411222008397e-05,
+      "loss": 0.0536,
+      "step": 79830
+    },
+    {
+      "epoch": 5.223421655217534,
+      "grad_norm": 0.97150719165802,
+      "learning_rate": 6.330525770956615e-05,
+      "loss": 0.0544,
+      "step": 79840
+    },
+    {
+      "epoch": 5.224075891396795,
+      "grad_norm": 0.9659034013748169,
+      "learning_rate": 6.3296402749961e-05,
+      "loss": 0.0608,
+      "step": 79850
+    },
+    {
+      "epoch": 5.224730127576055,
+      "grad_norm": 0.9073391556739807,
+      "learning_rate": 6.328754734156737e-05,
+      "loss": 0.0587,
+      "step": 79860
+    },
+    {
+      "epoch": 5.225384363755316,
+      "grad_norm": 0.9242889881134033,
+      "learning_rate": 6.327869148468418e-05,
+      "loss": 0.0547,
+      "step": 79870
+    },
+    {
+      "epoch": 5.226038599934577,
+      "grad_norm": 0.8243046998977661,
+      "learning_rate": 6.326983517961033e-05,
+      "loss": 0.06,
+      "step": 79880
+    },
+    {
+      "epoch": 5.226692836113837,
+      "grad_norm": 0.8463775515556335,
+      "learning_rate": 6.326097842664473e-05,
+      "loss": 0.0539,
+      "step": 79890
+    },
+    {
+      "epoch": 5.227347072293098,
+      "grad_norm": 0.8504155278205872,
+      "learning_rate": 6.325212122608635e-05,
+      "loss": 0.0629,
+      "step": 79900
+    },
+    {
+      "epoch": 5.228001308472359,
+      "grad_norm": 0.6780961751937866,
+      "learning_rate": 6.324326357823413e-05,
+      "loss": 0.0494,
+      "step": 79910
+    },
+    {
+      "epoch": 5.22865554465162,
+      "grad_norm": 0.8905619978904724,
+      "learning_rate": 6.3234405483387e-05,
+      "loss": 0.0537,
+      "step": 79920
+    },
+    {
+      "epoch": 5.22930978083088,
+      "grad_norm": 0.8399018049240112,
+      "learning_rate": 6.3225546941844e-05,
+      "loss": 0.0542,
+      "step": 79930
+    },
+    {
+      "epoch": 5.229964017010141,
+      "grad_norm": 1.064653992652893,
+      "learning_rate": 6.32166879539041e-05,
+      "loss": 0.0607,
+      "step": 79940
+    },
+    {
+      "epoch": 5.230618253189402,
+      "grad_norm": 0.7720884084701538,
+      "learning_rate": 6.320782851986631e-05,
+      "loss": 0.047,
+      "step": 79950
+    },
+    {
+      "epoch": 5.231272489368662,
+      "grad_norm": 0.7760630249977112,
+      "learning_rate": 6.319896864002968e-05,
+      "loss": 0.0611,
+      "step": 79960
+    },
+    {
+      "epoch": 5.231926725547923,
+      "grad_norm": 0.9401276111602783,
+      "learning_rate": 6.319010831469324e-05,
+      "loss": 0.0509,
+      "step": 79970
+    },
+    {
+      "epoch": 5.232580961727184,
+      "grad_norm": 0.865756630897522,
+      "learning_rate": 6.318124754415605e-05,
+      "loss": 0.0533,
+      "step": 79980
+    },
+    {
+      "epoch": 5.233235197906445,
+      "grad_norm": 0.7279103994369507,
+      "learning_rate": 6.317238632871718e-05,
+      "loss": 0.0492,
+      "step": 79990
+    },
+    {
+      "epoch": 5.233889434085705,
+      "grad_norm": 0.7956836223602295,
+      "learning_rate": 6.316352466867574e-05,
+      "loss": 0.0549,
+      "step": 80000
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 180000,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 12,
+  "save_steps": 10000,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 0.0,
+  "train_batch_size": 16,
+  "trial_name": null,
+  "trial_params": null
+}