diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,33440 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.0,
+  "eval_steps": 500,
+  "global_step": 4771,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.00020959966464053657,
+      "grad_norm": 0.22152887284755707,
+      "learning_rate": 0.0,
+      "loss": 0.1548,
+      "step": 1
+    },
+    {
+      "epoch": 0.00041919932928107315,
+      "grad_norm": 0.1861487776041031,
+      "learning_rate": 6.944444444444444e-08,
+      "loss": 0.1418,
+      "step": 2
+    },
+    {
+      "epoch": 0.0006287989939216097,
+      "grad_norm": 0.20890986919403076,
+      "learning_rate": 1.3888888888888888e-07,
+      "loss": 0.1602,
+      "step": 3
+    },
+    {
+      "epoch": 0.0008383986585621463,
+      "grad_norm": 0.20199677348136902,
+      "learning_rate": 2.0833333333333333e-07,
+      "loss": 0.1467,
+      "step": 4
+    },
+    {
+      "epoch": 0.0010479983232026828,
+      "grad_norm": 0.18860988318920135,
+      "learning_rate": 2.7777777777777776e-07,
+      "loss": 0.1357,
+      "step": 5
+    },
+    {
+      "epoch": 0.0012575979878432194,
+      "grad_norm": 0.1798183023929596,
+      "learning_rate": 3.472222222222223e-07,
+      "loss": 0.1343,
+      "step": 6
+    },
+    {
+      "epoch": 0.001467197652483756,
+      "grad_norm": 0.12569160759449005,
+      "learning_rate": 4.1666666666666667e-07,
+      "loss": 0.149,
+      "step": 7
+    },
+    {
+      "epoch": 0.0016767973171242926,
+      "grad_norm": 0.12502586841583252,
+      "learning_rate": 4.861111111111112e-07,
+      "loss": 0.1465,
+      "step": 8
+    },
+    {
+      "epoch": 0.0018863969817648292,
+      "grad_norm": 0.09551655501127243,
+      "learning_rate": 5.555555555555555e-07,
+      "loss": 0.1407,
+      "step": 9
+    },
+    {
+      "epoch": 0.0020959966464053656,
+      "grad_norm": 0.27150362730026245,
+      "learning_rate": 6.25e-07,
+      "loss": 0.141,
+      "step": 10
+    },
+    {
+      "epoch": 0.002305596311045902,
+      "grad_norm": 0.2170093059539795,
+      "learning_rate": 6.944444444444446e-07,
+      "loss": 0.1529,
+      "step": 11
+    },
+    {
+      "epoch": 0.0025151959756864388,
+      "grad_norm": 0.1994446963071823,
+      "learning_rate": 7.63888888888889e-07,
+      "loss": 0.1387,
+      "step": 12
+    },
+    {
+      "epoch": 0.0027247956403269754,
+      "grad_norm": 0.16523687541484833,
+      "learning_rate": 8.333333333333333e-07,
+      "loss": 0.1401,
+      "step": 13
+    },
+    {
+      "epoch": 0.002934395304967512,
+      "grad_norm": 0.22484736144542694,
+      "learning_rate": 9.027777777777779e-07,
+      "loss": 0.1438,
+      "step": 14
+    },
+    {
+      "epoch": 0.0031439949696080486,
+      "grad_norm": 0.14437100291252136,
+      "learning_rate": 9.722222222222224e-07,
+      "loss": 0.1485,
+      "step": 15
+    },
+    {
+      "epoch": 0.003353594634248585,
+      "grad_norm": 0.12035862356424332,
+      "learning_rate": 1.0416666666666667e-06,
+      "loss": 0.1463,
+      "step": 16
+    },
+    {
+      "epoch": 0.0035631942988891218,
+      "grad_norm": 0.19523362815380096,
+      "learning_rate": 1.111111111111111e-06,
+      "loss": 0.1401,
+      "step": 17
+    },
+    {
+      "epoch": 0.0037727939635296584,
+      "grad_norm": 0.2469567507505417,
+      "learning_rate": 1.1805555555555556e-06,
+      "loss": 0.1442,
+      "step": 18
+    },
+    {
+      "epoch": 0.0039823936281701946,
+      "grad_norm": 0.21258766949176788,
+      "learning_rate": 1.25e-06,
+      "loss": 0.1348,
+      "step": 19
+    },
+    {
+      "epoch": 0.004191993292810731,
+      "grad_norm": 0.13423168659210205,
+      "learning_rate": 1.3194444444444446e-06,
+      "loss": 0.1419,
+      "step": 20
+    },
+    {
+      "epoch": 0.004401592957451268,
+      "grad_norm": 0.15541696548461914,
+      "learning_rate": 1.3888888888888892e-06,
+      "loss": 0.1387,
+      "step": 21
+    },
+    {
+      "epoch": 0.004611192622091804,
+      "grad_norm": 0.1762888878583908,
+      "learning_rate": 1.4583333333333335e-06,
+      "loss": 0.1372,
+      "step": 22
+    },
+    {
+      "epoch": 0.004820792286732341,
+      "grad_norm": 0.1527898907661438,
+      "learning_rate": 1.527777777777778e-06,
+      "loss": 0.1325,
+      "step": 23
+    },
+    {
+      "epoch": 0.0050303919513728776,
+      "grad_norm": 0.11752472072839737,
+      "learning_rate": 1.5972222222222221e-06,
+      "loss": 0.1354,
+      "step": 24
+    },
+    {
+      "epoch": 0.005239991616013414,
+      "grad_norm": 0.1533581018447876,
+      "learning_rate": 1.6666666666666667e-06,
+      "loss": 0.1342,
+      "step": 25
+    },
+    {
+      "epoch": 0.005449591280653951,
+      "grad_norm": 0.14854849874973297,
+      "learning_rate": 1.7361111111111112e-06,
+      "loss": 0.1397,
+      "step": 26
+    },
+    {
+      "epoch": 0.005659190945294487,
+      "grad_norm": 0.1156805008649826,
+      "learning_rate": 1.8055555555555557e-06,
+      "loss": 0.1357,
+      "step": 27
+    },
+    {
+      "epoch": 0.005868790609935024,
+      "grad_norm": 0.11853114515542984,
+      "learning_rate": 1.8750000000000003e-06,
+      "loss": 0.1258,
+      "step": 28
+    },
+    {
+      "epoch": 0.0060783902745755606,
+      "grad_norm": 0.14961375296115875,
+      "learning_rate": 1.944444444444445e-06,
+      "loss": 0.1292,
+      "step": 29
+    },
+    {
+      "epoch": 0.006287989939216097,
+      "grad_norm": 0.10973469167947769,
+      "learning_rate": 2.0138888888888893e-06,
+      "loss": 0.1403,
+      "step": 30
+    },
+    {
+      "epoch": 0.006497589603856634,
+      "grad_norm": 0.10710575431585312,
+      "learning_rate": 2.0833333333333334e-06,
+      "loss": 0.1435,
+      "step": 31
+    },
+    {
+      "epoch": 0.00670718926849717,
+      "grad_norm": 0.12166962772607803,
+      "learning_rate": 2.152777777777778e-06,
+      "loss": 0.1351,
+      "step": 32
+    },
+    {
+      "epoch": 0.006916788933137707,
+      "grad_norm": 0.10747282952070236,
+      "learning_rate": 2.222222222222222e-06,
+      "loss": 0.1308,
+      "step": 33
+    },
+    {
+      "epoch": 0.0071263885977782436,
+      "grad_norm": 0.10978581756353378,
+      "learning_rate": 2.2916666666666666e-06,
+      "loss": 0.1303,
+      "step": 34
+    },
+    {
+      "epoch": 0.00733598826241878,
+      "grad_norm": 0.1297036111354828,
+      "learning_rate": 2.361111111111111e-06,
+      "loss": 0.1341,
+      "step": 35
+    },
+    {
+      "epoch": 0.007545587927059317,
+      "grad_norm": 0.10622533410787582,
+      "learning_rate": 2.4305555555555557e-06,
+      "loss": 0.1305,
+      "step": 36
+    },
+    {
+      "epoch": 0.007755187591699853,
+      "grad_norm": 0.12965598702430725,
+      "learning_rate": 2.5e-06,
+      "loss": 0.1179,
+      "step": 37
+    },
+    {
+      "epoch": 0.007964787256340389,
+      "grad_norm": 0.08929996192455292,
+      "learning_rate": 2.5694444444444443e-06,
+      "loss": 0.1246,
+      "step": 38
+    },
+    {
+      "epoch": 0.008174386920980926,
+      "grad_norm": 0.1213381290435791,
+      "learning_rate": 2.6388888888888893e-06,
+      "loss": 0.135,
+      "step": 39
+    },
+    {
+      "epoch": 0.008383986585621462,
+      "grad_norm": 0.13696981966495514,
+      "learning_rate": 2.7083333333333334e-06,
+      "loss": 0.1359,
+      "step": 40
+    },
+    {
+      "epoch": 0.008593586250261999,
+      "grad_norm": 0.10106691718101501,
+      "learning_rate": 2.7777777777777783e-06,
+      "loss": 0.1308,
+      "step": 41
+    },
+    {
+      "epoch": 0.008803185914902536,
+      "grad_norm": 0.16903014481067657,
+      "learning_rate": 2.8472222222222224e-06,
+      "loss": 0.1243,
+      "step": 42
+    },
+    {
+      "epoch": 0.009012785579543072,
+      "grad_norm": 0.26333266496658325,
+      "learning_rate": 2.916666666666667e-06,
+      "loss": 0.1269,
+      "step": 43
+    },
+    {
+      "epoch": 0.009222385244183609,
+      "grad_norm": 0.5060687065124512,
+      "learning_rate": 2.986111111111111e-06,
+      "loss": 0.1267,
+      "step": 44
+    },
+    {
+      "epoch": 0.009431984908824145,
+      "grad_norm": 0.37180081009864807,
+      "learning_rate": 3.055555555555556e-06,
+      "loss": 0.1359,
+      "step": 45
+    },
+    {
+      "epoch": 0.009641584573464682,
+      "grad_norm": 0.37329721450805664,
+      "learning_rate": 3.125e-06,
+      "loss": 0.1266,
+      "step": 46
+    },
+    {
+      "epoch": 0.009851184238105219,
+      "grad_norm": 0.9444931745529175,
+      "learning_rate": 3.1944444444444443e-06,
+      "loss": 0.1265,
+      "step": 47
+    },
+    {
+      "epoch": 0.010060783902745755,
+      "grad_norm": 0.3645467460155487,
+      "learning_rate": 3.2638888888888892e-06,
+      "loss": 0.1252,
+      "step": 48
+    },
+    {
+      "epoch": 0.010270383567386292,
+      "grad_norm": 0.35541367530822754,
+      "learning_rate": 3.3333333333333333e-06,
+      "loss": 0.1315,
+      "step": 49
+    },
+    {
+      "epoch": 0.010479983232026828,
+      "grad_norm": 0.26033562421798706,
+      "learning_rate": 3.4027777777777783e-06,
+      "loss": 0.1319,
+      "step": 50
+    },
+    {
+      "epoch": 0.010689582896667365,
+      "grad_norm": 0.43073058128356934,
+      "learning_rate": 3.4722222222222224e-06,
+      "loss": 0.1278,
+      "step": 51
+    },
+    {
+      "epoch": 0.010899182561307902,
+      "grad_norm": 0.3041709363460541,
+      "learning_rate": 3.5416666666666673e-06,
+      "loss": 0.1283,
+      "step": 52
+    },
+    {
+      "epoch": 0.011108782225948438,
+      "grad_norm": 0.3577912747859955,
+      "learning_rate": 3.6111111111111115e-06,
+      "loss": 0.1314,
+      "step": 53
+    },
+    {
+      "epoch": 0.011318381890588975,
+      "grad_norm": 0.25406020879745483,
+      "learning_rate": 3.680555555555556e-06,
+      "loss": 0.1303,
+      "step": 54
+    },
+    {
+      "epoch": 0.011527981555229511,
+      "grad_norm": 0.32610198855400085,
+      "learning_rate": 3.7500000000000005e-06,
+      "loss": 0.1254,
+      "step": 55
+    },
+    {
+      "epoch": 0.011737581219870048,
+      "grad_norm": 0.2049138844013214,
+      "learning_rate": 3.819444444444444e-06,
+      "loss": 0.1319,
+      "step": 56
+    },
+    {
+      "epoch": 0.011947180884510585,
+      "grad_norm": 0.29326915740966797,
+      "learning_rate": 3.88888888888889e-06,
+      "loss": 0.1192,
+      "step": 57
+    },
+    {
+      "epoch": 0.012156780549151121,
+      "grad_norm": 0.2198316603899002,
+      "learning_rate": 3.958333333333333e-06,
+      "loss": 0.1177,
+      "step": 58
+    },
+    {
+      "epoch": 0.012366380213791658,
+      "grad_norm": 0.20488092303276062,
+      "learning_rate": 4.027777777777779e-06,
+      "loss": 0.1266,
+      "step": 59
+    },
+    {
+      "epoch": 0.012575979878432194,
+      "grad_norm": 0.22538422048091888,
+      "learning_rate": 4.097222222222222e-06,
+      "loss": 0.1168,
+      "step": 60
+    },
+    {
+      "epoch": 0.012785579543072731,
+      "grad_norm": 0.2462712526321411,
+      "learning_rate": 4.166666666666667e-06,
+      "loss": 0.1159,
+      "step": 61
+    },
+    {
+      "epoch": 0.012995179207713268,
+      "grad_norm": 0.17663376033306122,
+      "learning_rate": 4.236111111111111e-06,
+      "loss": 0.1273,
+      "step": 62
+    },
+    {
+      "epoch": 0.013204778872353804,
+      "grad_norm": 0.16204069554805756,
+      "learning_rate": 4.305555555555556e-06,
+      "loss": 0.1134,
+      "step": 63
+    },
+    {
+      "epoch": 0.01341437853699434,
+      "grad_norm": 0.2351599931716919,
+      "learning_rate": 4.3750000000000005e-06,
+      "loss": 0.1272,
+      "step": 64
+    },
+    {
+      "epoch": 0.013623978201634877,
+      "grad_norm": 0.22874589264392853,
+      "learning_rate": 4.444444444444444e-06,
+      "loss": 0.1286,
+      "step": 65
+    },
+    {
+      "epoch": 0.013833577866275414,
+      "grad_norm": 0.25400885939598083,
+      "learning_rate": 4.5138888888888895e-06,
+      "loss": 0.1205,
+      "step": 66
+    },
+    {
+      "epoch": 0.01404317753091595,
+      "grad_norm": 0.07518032938241959,
+      "learning_rate": 4.583333333333333e-06,
+      "loss": 0.1247,
+      "step": 67
+    },
+    {
+      "epoch": 0.014252777195556487,
+      "grad_norm": 0.2462567836046219,
+      "learning_rate": 4.652777777777779e-06,
+      "loss": 0.1212,
+      "step": 68
+    },
+    {
+      "epoch": 0.014462376860197024,
+      "grad_norm": 0.3761221468448639,
+      "learning_rate": 4.722222222222222e-06,
+      "loss": 0.1161,
+      "step": 69
+    },
+    {
+      "epoch": 0.01467197652483756,
+      "grad_norm": 0.3552791476249695,
+      "learning_rate": 4.791666666666668e-06,
+      "loss": 0.1275,
+      "step": 70
+    },
+    {
+      "epoch": 0.014881576189478097,
+      "grad_norm": 0.2262704074382782,
+      "learning_rate": 4.861111111111111e-06,
+      "loss": 0.1217,
+      "step": 71
+    },
+    {
+      "epoch": 0.015091175854118634,
+      "grad_norm": 1.2746901512145996,
+      "learning_rate": 4.930555555555556e-06,
+      "loss": 0.1272,
+      "step": 72
+    },
+    {
+      "epoch": 0.01530077551875917,
+      "grad_norm": 0.47996070981025696,
+      "learning_rate": 5e-06,
+      "loss": 0.1326,
+      "step": 73
+    },
+    {
+      "epoch": 0.015510375183399707,
+      "grad_norm": 0.591677725315094,
+      "learning_rate": 5.069444444444445e-06,
+      "loss": 0.1359,
+      "step": 74
+    },
+    {
+      "epoch": 0.015719974848040243,
+      "grad_norm": 0.4169331192970276,
+      "learning_rate": 5.138888888888889e-06,
+      "loss": 0.1353,
+      "step": 75
+    },
+    {
+      "epoch": 0.015929574512680778,
+      "grad_norm": 0.2262561172246933,
+      "learning_rate": 5.208333333333334e-06,
+      "loss": 0.1318,
+      "step": 76
+    },
+    {
+      "epoch": 0.016139174177321317,
+      "grad_norm": 0.42291682958602905,
+      "learning_rate": 5.2777777777777785e-06,
+      "loss": 0.1287,
+      "step": 77
+    },
+    {
+      "epoch": 0.01634877384196185,
+      "grad_norm": 0.36955782771110535,
+      "learning_rate": 5.347222222222222e-06,
+      "loss": 0.1348,
+      "step": 78
+    },
+    {
+      "epoch": 0.01655837350660239,
+      "grad_norm": 0.286143034696579,
+      "learning_rate": 5.416666666666667e-06,
+      "loss": 0.1284,
+      "step": 79
+    },
+    {
+      "epoch": 0.016767973171242925,
+      "grad_norm": 0.3029952347278595,
+      "learning_rate": 5.486111111111112e-06,
+      "loss": 0.1357,
+      "step": 80
+    },
+    {
+      "epoch": 0.016977572835883463,
+      "grad_norm": 0.24525821208953857,
+      "learning_rate": 5.555555555555557e-06,
+      "loss": 0.1263,
+      "step": 81
+    },
+    {
+      "epoch": 0.017187172500523998,
+      "grad_norm": 0.3474200665950775,
+      "learning_rate": 5.625e-06,
+      "loss": 0.133,
+      "step": 82
+    },
+    {
+      "epoch": 0.017396772165164536,
+      "grad_norm": 0.2455555945634842,
+      "learning_rate": 5.694444444444445e-06,
+      "loss": 0.1292,
+      "step": 83
+    },
+    {
+      "epoch": 0.01760637182980507,
+      "grad_norm": 0.26876527070999146,
+      "learning_rate": 5.7638888888888886e-06,
+      "loss": 0.1327,
+      "step": 84
+    },
+    {
+      "epoch": 0.01781597149444561,
+      "grad_norm": 0.27033689618110657,
+      "learning_rate": 5.833333333333334e-06,
+      "loss": 0.1282,
+      "step": 85
+    },
+    {
+      "epoch": 0.018025571159086144,
+      "grad_norm": 0.19568723440170288,
+      "learning_rate": 5.9027777777777785e-06,
+      "loss": 0.1258,
+      "step": 86
+    },
+    {
+      "epoch": 0.018235170823726683,
+      "grad_norm": 0.30962151288986206,
+      "learning_rate": 5.972222222222222e-06,
+      "loss": 0.1259,
+      "step": 87
+    },
+    {
+      "epoch": 0.018444770488367217,
+      "grad_norm": 0.22522631287574768,
+      "learning_rate": 6.041666666666667e-06,
+      "loss": 0.1231,
+      "step": 88
+    },
+    {
+      "epoch": 0.018654370153007756,
+      "grad_norm": 0.3179832398891449,
+      "learning_rate": 6.111111111111112e-06,
+      "loss": 0.118,
+      "step": 89
+    },
+    {
+      "epoch": 0.01886396981764829,
+      "grad_norm": 0.34084218740463257,
+      "learning_rate": 6.180555555555557e-06,
+      "loss": 0.1219,
+      "step": 90
+    },
+    {
+      "epoch": 0.01907356948228883,
+      "grad_norm": 0.32963189482688904,
+      "learning_rate": 6.25e-06,
+      "loss": 0.1245,
+      "step": 91
+    },
+    {
+      "epoch": 0.019283169146929364,
+      "grad_norm": 0.45660048723220825,
+      "learning_rate": 6.319444444444445e-06,
+      "loss": 0.1265,
+      "step": 92
+    },
+    {
+      "epoch": 0.019492768811569902,
+      "grad_norm": 0.6942883133888245,
+      "learning_rate": 6.3888888888888885e-06,
+      "loss": 0.1288,
+      "step": 93
+    },
+    {
+      "epoch": 0.019702368476210437,
+      "grad_norm": 0.7902718782424927,
+      "learning_rate": 6.458333333333334e-06,
+      "loss": 0.129,
+      "step": 94
+    },
+    {
+      "epoch": 0.019911968140850975,
+      "grad_norm": 0.4735928773880005,
+      "learning_rate": 6.5277777777777784e-06,
+      "loss": 0.1187,
+      "step": 95
+    },
+    {
+      "epoch": 0.02012156780549151,
+      "grad_norm": 0.22121404111385345,
+      "learning_rate": 6.597222222222223e-06,
+      "loss": 0.1236,
+      "step": 96
+    },
+    {
+      "epoch": 0.02033116747013205,
+      "grad_norm": 0.515511691570282,
+      "learning_rate": 6.666666666666667e-06,
+      "loss": 0.1235,
+      "step": 97
+    },
+    {
+      "epoch": 0.020540767134772583,
+      "grad_norm": 0.4277538061141968,
+      "learning_rate": 6.736111111111112e-06,
+      "loss": 0.1242,
+      "step": 98
+    },
+    {
+      "epoch": 0.020750366799413122,
+      "grad_norm": 0.17368359863758087,
+      "learning_rate": 6.8055555555555566e-06,
+      "loss": 0.1263,
+      "step": 99
+    },
+    {
+      "epoch": 0.020959966464053657,
+      "grad_norm": 0.37945282459259033,
+      "learning_rate": 6.875e-06,
+      "loss": 0.1212,
+      "step": 100
+    },
+    {
+      "epoch": 0.021169566128694195,
+      "grad_norm": 0.366905152797699,
+      "learning_rate": 6.944444444444445e-06,
+      "loss": 0.1252,
+      "step": 101
+    },
+    {
+      "epoch": 0.02137916579333473,
+      "grad_norm": 0.17307527363300323,
+      "learning_rate": 7.013888888888889e-06,
+      "loss": 0.1171,
+      "step": 102
+    },
+    {
+      "epoch": 0.021588765457975268,
+      "grad_norm": 0.374055415391922,
+      "learning_rate": 7.083333333333335e-06,
+      "loss": 0.1113,
+      "step": 103
+    },
+    {
+      "epoch": 0.021798365122615803,
+      "grad_norm": 0.3751891255378723,
+      "learning_rate": 7.152777777777778e-06,
+      "loss": 0.1286,
+      "step": 104
+    },
+    {
+      "epoch": 0.02200796478725634,
+      "grad_norm": 0.20103880763053894,
+      "learning_rate": 7.222222222222223e-06,
+      "loss": 0.1132,
+      "step": 105
+    },
+    {
+      "epoch": 0.022217564451896876,
+      "grad_norm": 0.20718024671077728,
+      "learning_rate": 7.291666666666667e-06,
+      "loss": 0.1152,
+      "step": 106
+    },
+    {
+      "epoch": 0.022427164116537415,
+      "grad_norm": 0.30478349328041077,
+      "learning_rate": 7.361111111111112e-06,
+      "loss": 0.1166,
+      "step": 107
+    },
+    {
+      "epoch": 0.02263676378117795,
+      "grad_norm": 0.2688567042350769,
+      "learning_rate": 7.4305555555555565e-06,
+      "loss": 0.115,
+      "step": 108
+    },
+    {
+      "epoch": 0.022846363445818488,
+      "grad_norm": 0.19921305775642395,
+      "learning_rate": 7.500000000000001e-06,
+      "loss": 0.1181,
+      "step": 109
+    },
+    {
+      "epoch": 0.023055963110459023,
+      "grad_norm": 0.34289443492889404,
+      "learning_rate": 7.569444444444445e-06,
+      "loss": 0.1153,
+      "step": 110
+    },
+    {
+      "epoch": 0.02326556277509956,
+      "grad_norm": 0.4600578546524048,
+      "learning_rate": 7.638888888888888e-06,
+      "loss": 0.1192,
+      "step": 111
+    },
+    {
+      "epoch": 0.023475162439740096,
+      "grad_norm": 0.29753392934799194,
+      "learning_rate": 7.708333333333334e-06,
+      "loss": 0.1153,
+      "step": 112
+    },
+    {
+      "epoch": 0.023684762104380634,
+      "grad_norm": 0.19611868262290955,
+      "learning_rate": 7.77777777777778e-06,
+      "loss": 0.1139,
+      "step": 113
+    },
+    {
+      "epoch": 0.02389436176902117,
+      "grad_norm": 0.3326936960220337,
+      "learning_rate": 7.847222222222223e-06,
+      "loss": 0.1125,
+      "step": 114
+    },
+    {
+      "epoch": 0.024103961433661707,
+      "grad_norm": 0.43471208214759827,
+      "learning_rate": 7.916666666666667e-06,
+      "loss": 0.1144,
+      "step": 115
+    },
+    {
+      "epoch": 0.024313561098302242,
+      "grad_norm": 0.419888436794281,
+      "learning_rate": 7.986111111111112e-06,
+      "loss": 0.1153,
+      "step": 116
+    },
+    {
+      "epoch": 0.02452316076294278,
+      "grad_norm": 0.1664547622203827,
+      "learning_rate": 8.055555555555557e-06,
+      "loss": 0.1098,
+      "step": 117
+    },
+    {
+      "epoch": 0.024732760427583315,
+      "grad_norm": 0.23118549585342407,
+      "learning_rate": 8.125000000000001e-06,
+      "loss": 0.1162,
+      "step": 118
+    },
+    {
+      "epoch": 0.024942360092223854,
+      "grad_norm": 0.33760204911231995,
+      "learning_rate": 8.194444444444445e-06,
+      "loss": 0.1115,
+      "step": 119
+    },
+    {
+      "epoch": 0.02515195975686439,
+      "grad_norm": 0.25965481996536255,
+      "learning_rate": 8.263888888888888e-06,
+      "loss": 0.1102,
+      "step": 120
+    },
+    {
+      "epoch": 0.025361559421504927,
+      "grad_norm": 0.2375420331954956,
+      "learning_rate": 8.333333333333334e-06,
+      "loss": 0.1186,
+      "step": 121
+    },
+    {
+      "epoch": 0.025571159086145462,
+      "grad_norm": 0.23151437938213348,
+      "learning_rate": 8.402777777777779e-06,
+      "loss": 0.1088,
+      "step": 122
+    },
+    {
+      "epoch": 0.025780758750786,
+      "grad_norm": 0.34959205985069275,
+      "learning_rate": 8.472222222222223e-06,
+      "loss": 0.1084,
+      "step": 123
+    },
+    {
+      "epoch": 0.025990358415426535,
+      "grad_norm": 0.3837231695652008,
+      "learning_rate": 8.541666666666666e-06,
+      "loss": 0.1142,
+      "step": 124
+    },
+    {
+      "epoch": 0.026199958080067073,
+      "grad_norm": 0.317340612411499,
+      "learning_rate": 8.611111111111112e-06,
+      "loss": 0.1201,
+      "step": 125
+    },
+    {
+      "epoch": 0.026409557744707608,
+      "grad_norm": 0.3754269778728485,
+      "learning_rate": 8.680555555555557e-06,
+      "loss": 0.1122,
+      "step": 126
+    },
+    {
+      "epoch": 0.026619157409348147,
+      "grad_norm": 0.39836880564689636,
+      "learning_rate": 8.750000000000001e-06,
+      "loss": 0.1105,
+      "step": 127
+    },
+    {
+      "epoch": 0.02682875707398868,
+      "grad_norm": 0.458625465631485,
+      "learning_rate": 8.819444444444445e-06,
+      "loss": 0.1169,
+      "step": 128
+    },
+    {
+      "epoch": 0.02703835673862922,
+      "grad_norm": 0.34019047021865845,
+      "learning_rate": 8.888888888888888e-06,
+      "loss": 0.1127,
+      "step": 129
+    },
+    {
+      "epoch": 0.027247956403269755,
+      "grad_norm": 0.3323642313480377,
+      "learning_rate": 8.958333333333334e-06,
+      "loss": 0.1098,
+      "step": 130
+    },
+    {
+      "epoch": 0.027457556067910293,
+      "grad_norm": 0.2268533557653427,
+      "learning_rate": 9.027777777777779e-06,
+      "loss": 0.1095,
+      "step": 131
+    },
+    {
+      "epoch": 0.027667155732550828,
+      "grad_norm": 0.4628676176071167,
+      "learning_rate": 9.097222222222223e-06,
+      "loss": 0.1077,
+      "step": 132
+    },
+    {
+      "epoch": 0.027876755397191366,
+      "grad_norm": 0.5547027587890625,
+      "learning_rate": 9.166666666666666e-06,
+      "loss": 0.1156,
+      "step": 133
+    },
+    {
+      "epoch": 0.0280863550618319,
+      "grad_norm": 0.5809503197669983,
+      "learning_rate": 9.236111111111112e-06,
+      "loss": 0.1168,
+      "step": 134
+    },
+    {
+      "epoch": 0.028295954726472436,
+      "grad_norm": 0.497976154088974,
+      "learning_rate": 9.305555555555557e-06,
+      "loss": 0.1161,
+      "step": 135
+    },
+    {
+      "epoch": 0.028505554391112974,
+      "grad_norm": 0.67780601978302,
+      "learning_rate": 9.375000000000001e-06,
+      "loss": 0.1201,
+      "step": 136
+    },
+    {
+      "epoch": 0.02871515405575351,
+      "grad_norm": 0.6441759467124939,
+      "learning_rate": 9.444444444444445e-06,
+      "loss": 0.1192,
+      "step": 137
+    },
+    {
+      "epoch": 0.028924753720394047,
+      "grad_norm": 0.3711428642272949,
+      "learning_rate": 9.51388888888889e-06,
+      "loss": 0.1195,
+      "step": 138
+    },
+    {
+      "epoch": 0.029134353385034582,
+      "grad_norm": 0.3569689393043518,
+      "learning_rate": 9.583333333333335e-06,
+      "loss": 0.1118,
+      "step": 139
+    },
+    {
+      "epoch": 0.02934395304967512,
+      "grad_norm": 0.3563655912876129,
+      "learning_rate": 9.652777777777779e-06,
+      "loss": 0.1176,
+      "step": 140
+    },
+    {
+      "epoch": 0.029553552714315656,
+      "grad_norm": 0.2493436187505722,
+      "learning_rate": 9.722222222222223e-06,
+      "loss": 0.1147,
+      "step": 141
+    },
+    {
+      "epoch": 0.029763152378956194,
+      "grad_norm": 0.29483145475387573,
+      "learning_rate": 9.791666666666666e-06,
+      "loss": 0.1153,
+      "step": 142
+    },
+    {
+      "epoch": 0.02997275204359673,
+      "grad_norm": 0.2700521647930145,
+      "learning_rate": 9.861111111111112e-06,
+      "loss": 0.1111,
+      "step": 143
+    },
+    {
+      "epoch": 0.030182351708237267,
+      "grad_norm": 0.26881495118141174,
+      "learning_rate": 9.930555555555557e-06,
+      "loss": 0.1104,
+      "step": 144
+    },
+    {
+      "epoch": 0.030391951372877802,
+      "grad_norm": 0.30803802609443665,
+      "learning_rate": 1e-05,
+      "loss": 0.1122,
+      "step": 145
+    },
+    {
+      "epoch": 0.03060155103751834,
+      "grad_norm": 0.31767815351486206,
+      "learning_rate": 9.99999884750052e-06,
+      "loss": 0.1096,
+      "step": 146
+    },
+    {
+      "epoch": 0.030811150702158875,
+      "grad_norm": 0.22586488723754883,
+      "learning_rate": 9.99999539000261e-06,
+      "loss": 0.1082,
+      "step": 147
+    },
+    {
+      "epoch": 0.031020750366799413,
+      "grad_norm": 0.2010418325662613,
+      "learning_rate": 9.999989627507863e-06,
+      "loss": 0.1123,
+      "step": 148
+    },
+    {
+      "epoch": 0.03123035003143995,
+      "grad_norm": 0.2666708528995514,
+      "learning_rate": 9.999981560018935e-06,
+      "loss": 0.1035,
+      "step": 149
+    },
+    {
+      "epoch": 0.03143994969608049,
+      "grad_norm": 0.34085071086883545,
+      "learning_rate": 9.999971187539547e-06,
+      "loss": 0.1063,
+      "step": 150
+    },
+    {
+      "epoch": 0.03164954936072102,
+      "grad_norm": 0.31965869665145874,
+      "learning_rate": 9.999958510074482e-06,
+      "loss": 0.1095,
+      "step": 151
+    },
+    {
+      "epoch": 0.031859149025361556,
+      "grad_norm": 0.26403307914733887,
+      "learning_rate": 9.99994352762958e-06,
+      "loss": 0.1017,
+      "step": 152
+    },
+    {
+      "epoch": 0.0320687486900021,
+      "grad_norm": 0.17258979380130768,
+      "learning_rate": 9.999926240211752e-06,
+      "loss": 0.1045,
+      "step": 153
+    },
+    {
+      "epoch": 0.03227834835464263,
+      "grad_norm": 0.2114904820919037,
+      "learning_rate": 9.999906647828966e-06,
+      "loss": 0.1017,
+      "step": 154
+    },
+    {
+      "epoch": 0.03248794801928317,
+      "grad_norm": 0.2613145112991333,
+      "learning_rate": 9.999884750490255e-06,
+      "loss": 0.1064,
+      "step": 155
+    },
+    {
+      "epoch": 0.0326975476839237,
+      "grad_norm": 0.23586612939834595,
+      "learning_rate": 9.99986054820571e-06,
+      "loss": 0.1097,
+      "step": 156
+    },
+    {
+      "epoch": 0.032907147348564245,
+      "grad_norm": 0.24500718712806702,
+      "learning_rate": 9.999834040986491e-06,
+      "loss": 0.102,
+      "step": 157
+    },
+    {
+      "epoch": 0.03311674701320478,
+      "grad_norm": 0.36189910769462585,
+      "learning_rate": 9.999805228844818e-06,
+      "loss": 0.1048,
+      "step": 158
+    },
+    {
+      "epoch": 0.033326346677845314,
+      "grad_norm": 0.43638429045677185,
+      "learning_rate": 9.999774111793974e-06,
+      "loss": 0.1153,
+      "step": 159
+    },
+    {
+      "epoch": 0.03353594634248585,
+      "grad_norm": 0.3592098355293274,
+      "learning_rate": 9.999740689848302e-06,
+      "loss": 0.115,
+      "step": 160
+    },
+    {
+      "epoch": 0.03374554600712639,
+      "grad_norm": 0.23697715997695923,
+      "learning_rate": 9.999704963023213e-06,
+      "loss": 0.0969,
+      "step": 161
+    },
+    {
+      "epoch": 0.033955145671766926,
+      "grad_norm": 0.3014034628868103,
+      "learning_rate": 9.999666931335172e-06,
+      "loss": 0.102,
+      "step": 162
+    },
+    {
+      "epoch": 0.03416474533640746,
+      "grad_norm": 0.35838088393211365,
+      "learning_rate": 9.999626594801714e-06,
+      "loss": 0.1007,
+      "step": 163
+    },
+    {
+      "epoch": 0.034374345001047996,
+      "grad_norm": 0.2900119125843048,
+      "learning_rate": 9.999583953441436e-06,
+      "loss": 0.0993,
+      "step": 164
+    },
+    {
+      "epoch": 0.03458394466568854,
+      "grad_norm": 0.21882066130638123,
+      "learning_rate": 9.999539007273993e-06,
+      "loss": 0.1039,
+      "step": 165
+    },
+    {
+      "epoch": 0.03479354433032907,
+      "grad_norm": 0.1922779232263565,
+      "learning_rate": 9.999491756320105e-06,
+      "loss": 0.1049,
+      "step": 166
+    },
+    {
+      "epoch": 0.03500314399496961,
+      "grad_norm": 0.2402738481760025,
+      "learning_rate": 9.999442200601559e-06,
+      "loss": 0.1003,
+      "step": 167
+    },
+    {
+      "epoch": 0.03521274365961014,
+      "grad_norm": 0.25979745388031006,
+      "learning_rate": 9.999390340141195e-06,
+      "loss": 0.1041,
+      "step": 168
+    },
+    {
+      "epoch": 0.035422343324250684,
+      "grad_norm": 0.2431039661169052,
+      "learning_rate": 9.999336174962922e-06,
+      "loss": 0.1008,
+      "step": 169
+    },
+    {
+      "epoch": 0.03563194298889122,
+      "grad_norm": 0.22206492722034454,
+      "learning_rate": 9.999279705091711e-06,
+      "loss": 0.1052,
+      "step": 170
+    },
+    {
+      "epoch": 0.035841542653531754,
+      "grad_norm": 0.2511206567287445,
+      "learning_rate": 9.999220930553595e-06,
+      "loss": 0.1107,
+      "step": 171
+    },
+    {
+      "epoch": 0.03605114231817229,
+      "grad_norm": 0.3174304664134979,
+      "learning_rate": 9.999159851375668e-06,
+      "loss": 0.1022,
+      "step": 172
+    },
+    {
+      "epoch": 0.03626074198281283,
+      "grad_norm": 0.4061136841773987,
+      "learning_rate": 9.99909646758609e-06,
+      "loss": 0.101,
+      "step": 173
+    },
+    {
+      "epoch": 0.036470341647453365,
+      "grad_norm": 0.47826582193374634,
+      "learning_rate": 9.999030779214076e-06,
+      "loss": 0.1052,
+      "step": 174
+    },
+    {
+      "epoch": 0.0366799413120939,
+      "grad_norm": 0.4733898639678955,
+      "learning_rate": 9.998962786289912e-06,
+      "loss": 0.1049,
+      "step": 175
+    },
+    {
+      "epoch": 0.036889540976734435,
+      "grad_norm": 0.3916158974170685,
+      "learning_rate": 9.998892488844942e-06,
+      "loss": 0.1058,
+      "step": 176
+    },
+    {
+      "epoch": 0.03709914064137498,
+      "grad_norm": 0.25847071409225464,
+      "learning_rate": 9.998819886911574e-06,
+      "loss": 0.0942,
+      "step": 177
+    },
+    {
+      "epoch": 0.03730874030601551,
+      "grad_norm": 0.2088022381067276,
+      "learning_rate": 9.998744980523276e-06,
+      "loss": 0.1051,
+      "step": 178
+    },
+    {
+      "epoch": 0.037518339970656046,
+      "grad_norm": 0.34360599517822266,
+      "learning_rate": 9.99866776971458e-06,
+      "loss": 0.0934,
+      "step": 179
+    },
+    {
+      "epoch": 0.03772793963529658,
+      "grad_norm": 0.4142899513244629,
+      "learning_rate": 9.99858825452108e-06,
+      "loss": 0.0996,
+      "step": 180
+    },
+    {
+      "epoch": 0.03793753929993712,
+      "grad_norm": 0.32127901911735535,
+      "learning_rate": 9.998506434979434e-06,
+      "loss": 0.0974,
+      "step": 181
+    },
+    {
+      "epoch": 0.03814713896457766,
+      "grad_norm": 0.16188944876194,
+      "learning_rate": 9.998422311127357e-06,
+      "loss": 0.0959,
+      "step": 182
+    },
+    {
+      "epoch": 0.03835673862921819,
+      "grad_norm": 0.20421625673770905,
+      "learning_rate": 9.998335883003636e-06,
+      "loss": 0.0971,
+      "step": 183
+    },
+    {
+      "epoch": 0.03856633829385873,
+      "grad_norm": 0.29221752285957336,
+      "learning_rate": 9.99824715064811e-06,
+      "loss": 0.1037,
+      "step": 184
+    },
+    {
+      "epoch": 0.03877593795849927,
+      "grad_norm": 0.3054715096950531,
+      "learning_rate": 9.998156114101687e-06,
+      "loss": 0.0977,
+      "step": 185
+    },
+    {
+      "epoch": 0.038985537623139804,
+      "grad_norm": 0.22702643275260925,
+      "learning_rate": 9.998062773406332e-06,
+      "loss": 0.0999,
+      "step": 186
+    },
+    {
+      "epoch": 0.03919513728778034,
+      "grad_norm": 0.1409185230731964,
+      "learning_rate": 9.997967128605078e-06,
+      "loss": 0.0935,
+      "step": 187
+    },
+    {
+      "epoch": 0.039404736952420874,
+      "grad_norm": 0.21621251106262207,
+      "learning_rate": 9.997869179742014e-06,
+      "loss": 0.0987,
+      "step": 188
+    },
+    {
+      "epoch": 0.039614336617061416,
+      "grad_norm": 0.3209380507469177,
+      "learning_rate": 9.9977689268623e-06,
+      "loss": 0.0975,
+      "step": 189
+    },
+    {
+      "epoch": 0.03982393628170195,
+      "grad_norm": 0.3941796123981476,
+      "learning_rate": 9.997666370012145e-06,
+      "loss": 0.0999,
+      "step": 190
+    },
+    {
+      "epoch": 0.040033535946342486,
+      "grad_norm": 0.40109798312187195,
+      "learning_rate": 9.997561509238833e-06,
+      "loss": 0.0977,
+      "step": 191
+    },
+    {
+      "epoch": 0.04024313561098302,
+      "grad_norm": 0.31651192903518677,
+      "learning_rate": 9.997454344590702e-06,
+      "loss": 0.1025,
+      "step": 192
+    },
+    {
+      "epoch": 0.04045273527562356,
+      "grad_norm": 0.20569577813148499,
+      "learning_rate": 9.997344876117157e-06,
+      "loss": 0.0967,
+      "step": 193
+    },
+    {
+      "epoch": 0.0406623349402641,
+      "grad_norm": 0.2173171490430832,
+      "learning_rate": 9.997233103868664e-06,
+      "loss": 0.098,
+      "step": 194
+    },
+    {
+      "epoch": 0.04087193460490463,
+      "grad_norm": 0.26545464992523193,
+      "learning_rate": 9.997119027896745e-06,
+      "loss": 0.0997,
+      "step": 195
+    },
+    {
+      "epoch": 0.04108153426954517,
+      "grad_norm": 0.25196799635887146,
+      "learning_rate": 9.997002648253994e-06,
+      "loss": 0.0936,
+      "step": 196
+    },
+    {
+      "epoch": 0.04129113393418571,
+      "grad_norm": 0.18679989874362946,
+      "learning_rate": 9.99688396499406e-06,
+      "loss": 0.0958,
+      "step": 197
+    },
+    {
+      "epoch": 0.041500733598826244,
+      "grad_norm": 0.16168518364429474,
+      "learning_rate": 9.996762978171657e-06,
+      "loss": 0.0906,
+      "step": 198
+    },
+    {
+      "epoch": 0.04171033326346678,
+      "grad_norm": 0.19161826372146606,
+      "learning_rate": 9.996639687842558e-06,
+      "loss": 0.0974,
+      "step": 199
+    },
+    {
+      "epoch": 0.04191993292810731,
+      "grad_norm": 0.18548256158828735,
+      "learning_rate": 9.9965140940636e-06,
+      "loss": 0.0976,
+      "step": 200
+    },
+    {
+      "epoch": 0.042129532592747855,
+      "grad_norm": 0.14844392240047455,
+      "learning_rate": 9.996386196892683e-06,
+      "loss": 0.0948,
+      "step": 201
+    },
+    {
+      "epoch": 0.04233913225738839,
+      "grad_norm": 0.10816401243209839,
+      "learning_rate": 9.996255996388767e-06,
+      "loss": 0.0927,
+      "step": 202
+    },
+    {
+      "epoch": 0.042548731922028925,
+      "grad_norm": 0.14506785571575165,
+      "learning_rate": 9.996123492611875e-06,
+      "loss": 0.1019,
+      "step": 203
+    },
+    {
+      "epoch": 0.04275833158666946,
+      "grad_norm": 0.18897125124931335,
+      "learning_rate": 9.99598868562309e-06,
+      "loss": 0.1029,
+      "step": 204
+    },
+    {
+      "epoch": 0.042967931251309995,
+      "grad_norm": 0.24414150416851044,
+      "learning_rate": 9.99585157548456e-06,
+      "loss": 0.0895,
+      "step": 205
+    },
+    {
+      "epoch": 0.043177530915950536,
+      "grad_norm": 0.34987929463386536,
+      "learning_rate": 9.995712162259489e-06,
+      "loss": 0.0933,
+      "step": 206
+    },
+    {
+      "epoch": 0.04338713058059107,
+      "grad_norm": 0.4623091518878937,
+      "learning_rate": 9.995570446012152e-06,
+      "loss": 0.106,
+      "step": 207
+    },
+    {
+      "epoch": 0.043596730245231606,
+      "grad_norm": 0.42472946643829346,
+      "learning_rate": 9.995426426807875e-06,
+      "loss": 0.1011,
+      "step": 208
+    },
+    {
+      "epoch": 0.04380632990987214,
+      "grad_norm": 0.20523960888385773,
+      "learning_rate": 9.995280104713055e-06,
+      "loss": 0.096,
+      "step": 209
+    },
+    {
+      "epoch": 0.04401592957451268,
+      "grad_norm": 0.23212192952632904,
+      "learning_rate": 9.995131479795142e-06,
+      "loss": 0.0945,
+      "step": 210
+    },
+    {
+      "epoch": 0.04422552923915322,
+      "grad_norm": 0.40809500217437744,
+      "learning_rate": 9.994980552122655e-06,
+      "loss": 0.0935,
+      "step": 211
+    },
+    {
+      "epoch": 0.04443512890379375,
+      "grad_norm": 0.3801667094230652,
+      "learning_rate": 9.99482732176517e-06,
+      "loss": 0.0985,
+      "step": 212
+    },
+    {
+      "epoch": 0.04464472856843429,
+      "grad_norm": 0.17014552652835846,
+      "learning_rate": 9.994671788793328e-06,
+      "loss": 0.0951,
+      "step": 213
+    },
+    {
+      "epoch": 0.04485432823307483,
+      "grad_norm": 0.31415608525276184,
+      "learning_rate": 9.99451395327883e-06,
+      "loss": 0.0986,
+      "step": 214
+    },
+    {
+      "epoch": 0.045063927897715364,
+      "grad_norm": 0.3655521273612976,
+      "learning_rate": 9.994353815294438e-06,
+      "loss": 0.0979,
+      "step": 215
+    },
+    {
+      "epoch": 0.0452735275623559,
+      "grad_norm": 0.22098667919635773,
+      "learning_rate": 9.99419137491397e-06,
+      "loss": 0.0965,
+      "step": 216
+    },
+    {
+      "epoch": 0.045483127226996434,
+      "grad_norm": 0.2817780673503876,
+      "learning_rate": 9.99402663221232e-06,
+      "loss": 0.0976,
+      "step": 217
+    },
+    {
+      "epoch": 0.045692726891636976,
+      "grad_norm": 0.27700185775756836,
+      "learning_rate": 9.993859587265429e-06,
+      "loss": 0.1014,
+      "step": 218
+    },
+    {
+      "epoch": 0.04590232655627751,
+      "grad_norm": 0.18445956707000732,
+      "learning_rate": 9.993690240150305e-06,
+      "loss": 0.0954,
+      "step": 219
+    },
+    {
+      "epoch": 0.046111926220918045,
+      "grad_norm": 0.2870791554450989,
+      "learning_rate": 9.993518590945017e-06,
+      "loss": 0.0995,
+      "step": 220
+    },
+    {
+      "epoch": 0.04632152588555858,
+      "grad_norm": 0.20852920413017273,
+      "learning_rate": 9.993344639728694e-06,
+      "loss": 0.0923,
+      "step": 221
+    },
+    {
+      "epoch": 0.04653112555019912,
+      "grad_norm": 0.2063411921262741,
+      "learning_rate": 9.993168386581533e-06,
+      "loss": 0.0903,
+      "step": 222
+    },
+    {
+      "epoch": 0.04674072521483966,
+      "grad_norm": 0.2189558446407318,
+      "learning_rate": 9.992989831584781e-06,
+      "loss": 0.0934,
+      "step": 223
+    },
+    {
+      "epoch": 0.04695032487948019,
+      "grad_norm": 0.2275790423154831,
+      "learning_rate": 9.992808974820755e-06,
+      "loss": 0.0944,
+      "step": 224
+    },
+    {
+      "epoch": 0.04715992454412073,
+      "grad_norm": 0.29324543476104736,
+      "learning_rate": 9.992625816372828e-06,
+      "loss": 0.0966,
+      "step": 225
+    },
+    {
+      "epoch": 0.04736952420876127,
+      "grad_norm": 0.24611949920654297,
+      "learning_rate": 9.992440356325437e-06,
+      "loss": 0.0948,
+      "step": 226
+    },
+    {
+      "epoch": 0.0475791238734018,
+      "grad_norm": 0.25625893473625183,
+      "learning_rate": 9.992252594764079e-06,
+      "loss": 0.0896,
+      "step": 227
+    },
+    {
+      "epoch": 0.04778872353804234,
+      "grad_norm": 0.22875913977622986,
+      "learning_rate": 9.99206253177531e-06,
+      "loss": 0.0887,
+      "step": 228
+    },
+    {
+      "epoch": 0.04799832320268287,
+      "grad_norm": 0.2700249254703522,
+      "learning_rate": 9.991870167446751e-06,
+      "loss": 0.0935,
+      "step": 229
+    },
+    {
+      "epoch": 0.048207922867323415,
+      "grad_norm": 0.2741994857788086,
+      "learning_rate": 9.991675501867083e-06,
+      "loss": 0.0945,
+      "step": 230
+    },
+    {
+      "epoch": 0.04841752253196395,
+      "grad_norm": 0.2627350389957428,
+      "learning_rate": 9.991478535126045e-06,
+      "loss": 0.0937,
+      "step": 231
+    },
+    {
+      "epoch": 0.048627122196604484,
+      "grad_norm": 0.2349289357662201,
+      "learning_rate": 9.99127926731444e-06,
+      "loss": 0.091,
+      "step": 232
+    },
+    {
+      "epoch": 0.04883672186124502,
+      "grad_norm": 0.19692082703113556,
+      "learning_rate": 9.991077698524128e-06,
+      "loss": 0.0925,
+      "step": 233
+    },
+    {
+      "epoch": 0.04904632152588556,
+      "grad_norm": 0.21425145864486694,
+      "learning_rate": 9.990873828848035e-06,
+      "loss": 0.092,
+      "step": 234
+    },
+    {
+      "epoch": 0.049255921190526096,
+      "grad_norm": 0.22980082035064697,
+      "learning_rate": 9.990667658380145e-06,
+      "loss": 0.0891,
+      "step": 235
+    },
+    {
+      "epoch": 0.04946552085516663,
+      "grad_norm": 0.3201009929180145,
+      "learning_rate": 9.990459187215498e-06,
+      "loss": 0.0925,
+      "step": 236
+    },
+    {
+      "epoch": 0.049675120519807166,
+      "grad_norm": 0.366510272026062,
+      "learning_rate": 9.990248415450204e-06,
+      "loss": 0.0991,
+      "step": 237
+    },
+    {
+      "epoch": 0.04988472018444771,
+      "grad_norm": 0.3486904203891754,
+      "learning_rate": 9.990035343181426e-06,
+      "loss": 0.0938,
+      "step": 238
+    },
+    {
+      "epoch": 0.05009431984908824,
+      "grad_norm": 0.2655981779098511,
+      "learning_rate": 9.989819970507392e-06,
+      "loss": 0.0972,
+      "step": 239
+    },
+    {
+      "epoch": 0.05030391951372878,
+      "grad_norm": 0.19759038090705872,
+      "learning_rate": 9.989602297527387e-06,
+      "loss": 0.0894,
+      "step": 240
+    },
+    {
+      "epoch": 0.05051351917836931,
+      "grad_norm": 0.2099776715040207,
+      "learning_rate": 9.98938232434176e-06,
+      "loss": 0.0906,
+      "step": 241
+    },
+    {
+      "epoch": 0.050723118843009854,
+      "grad_norm": 0.22497335076332092,
+      "learning_rate": 9.98916005105192e-06,
+      "loss": 0.0953,
+      "step": 242
+    },
+    {
+      "epoch": 0.05093271850765039,
+      "grad_norm": 0.19926096498966217,
+      "learning_rate": 9.98893547776033e-06,
+      "loss": 0.0873,
+      "step": 243
+    },
+    {
+      "epoch": 0.051142318172290924,
+      "grad_norm": 0.19615811109542847,
+      "learning_rate": 9.988708604570523e-06,
+      "loss": 0.0868,
+      "step": 244
+    },
+    {
+      "epoch": 0.05135191783693146,
+      "grad_norm": 0.2284836620092392,
+      "learning_rate": 9.988479431587085e-06,
+      "loss": 0.0891,
+      "step": 245
+    },
+    {
+      "epoch": 0.051561517501572,
+      "grad_norm": 0.24722371995449066,
+      "learning_rate": 9.988247958915665e-06,
+      "loss": 0.091,
+      "step": 246
+    },
+    {
+      "epoch": 0.051771117166212535,
+      "grad_norm": 0.2157488912343979,
+      "learning_rate": 9.988014186662971e-06,
+      "loss": 0.0939,
+      "step": 247
+    },
+    {
+      "epoch": 0.05198071683085307,
+      "grad_norm": 0.19561699032783508,
+      "learning_rate": 9.987778114936775e-06,
+      "loss": 0.0886,
+      "step": 248
+    },
+    {
+      "epoch": 0.052190316495493605,
+      "grad_norm": 0.20849882066249847,
+      "learning_rate": 9.987539743845902e-06,
+      "loss": 0.09,
+      "step": 249
+    },
+    {
+      "epoch": 0.05239991616013415,
+      "grad_norm": 0.23204147815704346,
+      "learning_rate": 9.987299073500245e-06,
+      "loss": 0.0882,
+      "step": 250
+    },
+    {
+      "epoch": 0.05260951582477468,
+      "grad_norm": 0.2583141028881073,
+      "learning_rate": 9.98705610401075e-06,
+      "loss": 0.0949,
+      "step": 251
+    },
+    {
+      "epoch": 0.052819115489415217,
+      "grad_norm": 0.27133336663246155,
+      "learning_rate": 9.986810835489426e-06,
+      "loss": 0.0914,
+      "step": 252
+    },
+    {
+      "epoch": 0.05302871515405575,
+      "grad_norm": 0.2617621123790741,
+      "learning_rate": 9.986563268049345e-06,
+      "loss": 0.0862,
+      "step": 253
+    },
+    {
+      "epoch": 0.05323831481869629,
+      "grad_norm": 0.2514387369155884,
+      "learning_rate": 9.98631340180463e-06,
+      "loss": 0.093,
+      "step": 254
+    },
+    {
+      "epoch": 0.05344791448333683,
+      "grad_norm": 0.22442054748535156,
+      "learning_rate": 9.986061236870478e-06,
+      "loss": 0.0963,
+      "step": 255
+    },
+    {
+      "epoch": 0.05365751414797736,
+      "grad_norm": 0.16423439979553223,
+      "learning_rate": 9.985806773363127e-06,
+      "loss": 0.0916,
+      "step": 256
+    },
+    {
+      "epoch": 0.0538671138126179,
+      "grad_norm": 0.19374483823776245,
+      "learning_rate": 9.985550011399889e-06,
+      "loss": 0.088,
+      "step": 257
+    },
+    {
+      "epoch": 0.05407671347725844,
+      "grad_norm": 0.28776028752326965,
+      "learning_rate": 9.985290951099134e-06,
+      "loss": 0.0912,
+      "step": 258
+    },
+    {
+      "epoch": 0.054286313141898974,
+      "grad_norm": 0.3693079948425293,
+      "learning_rate": 9.985029592580284e-06,
+      "loss": 0.0912,
+      "step": 259
+    },
+    {
+      "epoch": 0.05449591280653951,
+      "grad_norm": 0.4039154052734375,
+      "learning_rate": 9.984765935963826e-06,
+      "loss": 0.0935,
+      "step": 260
+    },
+    {
+      "epoch": 0.054705512471180044,
+      "grad_norm": 0.3492220640182495,
+      "learning_rate": 9.98449998137131e-06,
+      "loss": 0.0939,
+      "step": 261
+    },
+    {
+      "epoch": 0.054915112135820586,
+      "grad_norm": 0.21770621836185455,
+      "learning_rate": 9.984231728925338e-06,
+      "loss": 0.0911,
+      "step": 262
+    },
+    {
+      "epoch": 0.05512471180046112,
+      "grad_norm": 0.17481771111488342,
+      "learning_rate": 9.983961178749573e-06,
+      "loss": 0.0894,
+      "step": 263
+    },
+    {
+      "epoch": 0.055334311465101656,
+      "grad_norm": 0.23308181762695312,
+      "learning_rate": 9.98368833096874e-06,
+      "loss": 0.0929,
+      "step": 264
+    },
+    {
+      "epoch": 0.05554391112974219,
+      "grad_norm": 0.2207728773355484,
+      "learning_rate": 9.983413185708622e-06,
+      "loss": 0.0912,
+      "step": 265
+    },
+    {
+      "epoch": 0.05575351079438273,
+      "grad_norm": 0.2060740441083908,
+      "learning_rate": 9.98313574309606e-06,
+      "loss": 0.0877,
+      "step": 266
+    },
+    {
+      "epoch": 0.05596311045902327,
+      "grad_norm": 0.22956304252147675,
+      "learning_rate": 9.982856003258954e-06,
+      "loss": 0.0892,
+      "step": 267
+    },
+    {
+      "epoch": 0.0561727101236638,
+      "grad_norm": 0.22290349006652832,
+      "learning_rate": 9.982573966326268e-06,
+      "loss": 0.0882,
+      "step": 268
+    },
+    {
+      "epoch": 0.05638230978830434,
+      "grad_norm": 0.20171046257019043,
+      "learning_rate": 9.982289632428017e-06,
+      "loss": 0.0891,
+      "step": 269
+    },
+    {
+      "epoch": 0.05659190945294487,
+      "grad_norm": 0.2069445252418518,
+      "learning_rate": 9.982003001695282e-06,
+      "loss": 0.0881,
+      "step": 270
+    },
+    {
+      "epoch": 0.056801509117585414,
+      "grad_norm": 0.206475630402565,
+      "learning_rate": 9.981714074260196e-06,
+      "loss": 0.0902,
+      "step": 271
+    },
+    {
+      "epoch": 0.05701110878222595,
+      "grad_norm": 0.19828340411186218,
+      "learning_rate": 9.98142285025596e-06,
+      "loss": 0.0878,
+      "step": 272
+    },
+    {
+      "epoch": 0.05722070844686648,
+      "grad_norm": 0.18767406046390533,
+      "learning_rate": 9.981129329816821e-06,
+      "loss": 0.0905,
+      "step": 273
+    },
+    {
+      "epoch": 0.05743030811150702,
+      "grad_norm": 0.1480463147163391,
+      "learning_rate": 9.980833513078097e-06,
+      "loss": 0.089,
+      "step": 274
+    },
+    {
+      "epoch": 0.05763990777614756,
+      "grad_norm": 0.12146848440170288,
+      "learning_rate": 9.980535400176158e-06,
+      "loss": 0.0853,
+      "step": 275
+    },
+    {
+      "epoch": 0.057849507440788095,
+      "grad_norm": 0.17080923914909363,
+      "learning_rate": 9.980234991248434e-06,
+      "loss": 0.0851,
+      "step": 276
+    },
+    {
+      "epoch": 0.05805910710542863,
+      "grad_norm": 0.20135587453842163,
+      "learning_rate": 9.979932286433414e-06,
+      "loss": 0.0851,
+      "step": 277
+    },
+    {
+      "epoch": 0.058268706770069165,
+      "grad_norm": 0.193786159157753,
+      "learning_rate": 9.979627285870644e-06,
+      "loss": 0.0844,
+      "step": 278
+    },
+    {
+      "epoch": 0.058478306434709706,
+      "grad_norm": 0.17661581933498383,
+      "learning_rate": 9.979319989700729e-06,
+      "loss": 0.0899,
+      "step": 279
+    },
+    {
+      "epoch": 0.05868790609935024,
+      "grad_norm": 0.1813191920518875,
+      "learning_rate": 9.979010398065334e-06,
+      "loss": 0.0852,
+      "step": 280
+    },
+    {
+      "epoch": 0.058897505763990776,
+      "grad_norm": 0.20303796231746674,
+      "learning_rate": 9.97869851110718e-06,
+      "loss": 0.0912,
+      "step": 281
+    },
+    {
+      "epoch": 0.05910710542863131,
+      "grad_norm": 0.1993362307548523,
+      "learning_rate": 9.978384328970045e-06,
+      "loss": 0.0903,
+      "step": 282
+    },
+    {
+      "epoch": 0.05931670509327185,
+      "grad_norm": 0.18841809034347534,
+      "learning_rate": 9.978067851798771e-06,
+      "loss": 0.0918,
+      "step": 283
+    },
+    {
+      "epoch": 0.05952630475791239,
+      "grad_norm": 0.22483417391777039,
+      "learning_rate": 9.97774907973925e-06,
+      "loss": 0.0929,
+      "step": 284
+    },
+    {
+      "epoch": 0.05973590442255292,
+      "grad_norm": 0.3062545657157898,
+      "learning_rate": 9.977428012938437e-06,
+      "loss": 0.0889,
+      "step": 285
+    },
+    {
+      "epoch": 0.05994550408719346,
+      "grad_norm": 0.3955790400505066,
+      "learning_rate": 9.977104651544342e-06,
+      "loss": 0.0917,
+      "step": 286
+    },
+    {
+      "epoch": 0.060155103751834,
+      "grad_norm": 0.4466722309589386,
+      "learning_rate": 9.97677899570604e-06,
+      "loss": 0.091,
+      "step": 287
+    },
+    {
+      "epoch": 0.060364703416474534,
+      "grad_norm": 0.3777875304222107,
+      "learning_rate": 9.976451045573653e-06,
+      "loss": 0.0872,
+      "step": 288
+    },
+    {
+      "epoch": 0.06057430308111507,
+      "grad_norm": 0.21344004571437836,
+      "learning_rate": 9.976120801298368e-06,
+      "loss": 0.0874,
+      "step": 289
+    },
+    {
+      "epoch": 0.060783902745755604,
+      "grad_norm": 0.2490052580833435,
+      "learning_rate": 9.975788263032427e-06,
+      "loss": 0.0873,
+      "step": 290
+    },
+    {
+      "epoch": 0.060993502410396146,
+      "grad_norm": 0.3740801513195038,
+      "learning_rate": 9.97545343092913e-06,
+      "loss": 0.0905,
+      "step": 291
+    },
+    {
+      "epoch": 0.06120310207503668,
+      "grad_norm": 0.356458842754364,
+      "learning_rate": 9.975116305142836e-06,
+      "loss": 0.0872,
+      "step": 292
+    },
+    {
+      "epoch": 0.061412701739677215,
+      "grad_norm": 0.2755817472934723,
+      "learning_rate": 9.974776885828958e-06,
+      "loss": 0.0902,
+      "step": 293
+    },
+    {
+      "epoch": 0.06162230140431775,
+      "grad_norm": 0.222996786236763,
+      "learning_rate": 9.974435173143968e-06,
+      "loss": 0.0848,
+      "step": 294
+    },
+    {
+      "epoch": 0.06183190106895829,
+      "grad_norm": 0.24334965646266937,
+      "learning_rate": 9.974091167245397e-06,
+      "loss": 0.0915,
+      "step": 295
+    },
+    {
+      "epoch": 0.06204150073359883,
+      "grad_norm": 0.25788915157318115,
+      "learning_rate": 9.973744868291832e-06,
+      "loss": 0.0894,
+      "step": 296
+    },
+    {
+      "epoch": 0.06225110039823936,
+      "grad_norm": 0.22377200424671173,
+      "learning_rate": 9.973396276442917e-06,
+      "loss": 0.0896,
+      "step": 297
+    },
+    {
+      "epoch": 0.0624607000628799,
+      "grad_norm": 0.23889151215553284,
+      "learning_rate": 9.973045391859348e-06,
+      "loss": 0.0859,
+      "step": 298
+    },
+    {
+      "epoch": 0.06267029972752043,
+      "grad_norm": 0.18626217544078827,
+      "learning_rate": 9.97269221470289e-06,
+      "loss": 0.0922,
+      "step": 299
+    },
+    {
+      "epoch": 0.06287989939216097,
+      "grad_norm": 0.14671076834201813,
+      "learning_rate": 9.97233674513635e-06,
+      "loss": 0.0922,
+      "step": 300
+    },
+    {
+      "epoch": 0.06308949905680152,
+      "grad_norm": 0.21669505536556244,
+      "learning_rate": 9.971978983323606e-06,
+      "loss": 0.084,
+      "step": 301
+    },
+    {
+      "epoch": 0.06329909872144204,
+      "grad_norm": 0.2512351870536804,
+      "learning_rate": 9.971618929429584e-06,
+      "loss": 0.0917,
+      "step": 302
+    },
+    {
+      "epoch": 0.06350869838608258,
+      "grad_norm": 0.28498363494873047,
+      "learning_rate": 9.971256583620268e-06,
+      "loss": 0.0854,
+      "step": 303
+    },
+    {
+      "epoch": 0.06371829805072311,
+      "grad_norm": 0.2733207643032074,
+      "learning_rate": 9.970891946062698e-06,
+      "loss": 0.0867,
+      "step": 304
+    },
+    {
+      "epoch": 0.06392789771536365,
+      "grad_norm": 0.19943512976169586,
+      "learning_rate": 9.970525016924974e-06,
+      "loss": 0.0913,
+      "step": 305
+    },
+    {
+      "epoch": 0.0641374973800042,
+      "grad_norm": 0.16102051734924316,
+      "learning_rate": 9.97015579637625e-06,
+      "loss": 0.0878,
+      "step": 306
+    },
+    {
+      "epoch": 0.06434709704464472,
+      "grad_norm": 0.19645822048187256,
+      "learning_rate": 9.969784284586736e-06,
+      "loss": 0.0856,
+      "step": 307
+    },
+    {
+      "epoch": 0.06455669670928527,
+      "grad_norm": 0.2531105875968933,
+      "learning_rate": 9.9694104817277e-06,
+      "loss": 0.0897,
+      "step": 308
+    },
+    {
+      "epoch": 0.06476629637392581,
+      "grad_norm": 0.24819032847881317,
+      "learning_rate": 9.969034387971463e-06,
+      "loss": 0.0854,
+      "step": 309
+    },
+    {
+      "epoch": 0.06497589603856634,
+      "grad_norm": 0.23713769018650055,
+      "learning_rate": 9.968656003491407e-06,
+      "loss": 0.0861,
+      "step": 310
+    },
+    {
+      "epoch": 0.06518549570320688,
+      "grad_norm": 0.21117182075977325,
+      "learning_rate": 9.968275328461964e-06,
+      "loss": 0.085,
+      "step": 311
+    },
+    {
+      "epoch": 0.0653950953678474,
+      "grad_norm": 0.16627727448940277,
+      "learning_rate": 9.967892363058626e-06,
+      "loss": 0.0854,
+      "step": 312
+    },
+    {
+      "epoch": 0.06560469503248795,
+      "grad_norm": 0.22320255637168884,
+      "learning_rate": 9.967507107457942e-06,
+      "loss": 0.085,
+      "step": 313
+    },
+    {
+      "epoch": 0.06581429469712849,
+      "grad_norm": 0.30917251110076904,
+      "learning_rate": 9.967119561837513e-06,
+      "loss": 0.0909,
+      "step": 314
+    },
+    {
+      "epoch": 0.06602389436176902,
+      "grad_norm": 0.3397315740585327,
+      "learning_rate": 9.966729726375997e-06,
+      "loss": 0.0938,
+      "step": 315
+    },
+    {
+      "epoch": 0.06623349402640956,
+      "grad_norm": 0.3161088228225708,
+      "learning_rate": 9.96633760125311e-06,
+      "loss": 0.0945,
+      "step": 316
+    },
+    {
+      "epoch": 0.0664430936910501,
+      "grad_norm": 0.2754218876361847,
+      "learning_rate": 9.965943186649619e-06,
+      "loss": 0.088,
+      "step": 317
+    },
+    {
+      "epoch": 0.06665269335569063,
+      "grad_norm": 0.2540167272090912,
+      "learning_rate": 9.965546482747352e-06,
+      "loss": 0.0835,
+      "step": 318
+    },
+    {
+      "epoch": 0.06686229302033117,
+      "grad_norm": 0.24167752265930176,
+      "learning_rate": 9.965147489729187e-06,
+      "loss": 0.0872,
+      "step": 319
+    },
+    {
+      "epoch": 0.0670718926849717,
+      "grad_norm": 0.19365692138671875,
+      "learning_rate": 9.96474620777906e-06,
+      "loss": 0.0865,
+      "step": 320
+    },
+    {
+      "epoch": 0.06728149234961224,
+      "grad_norm": 0.13701936602592468,
+      "learning_rate": 9.964342637081962e-06,
+      "loss": 0.0913,
+      "step": 321
+    },
+    {
+      "epoch": 0.06749109201425278,
+      "grad_norm": 0.2027350515127182,
+      "learning_rate": 9.963936777823941e-06,
+      "loss": 0.0852,
+      "step": 322
+    },
+    {
+      "epoch": 0.06770069167889331,
+      "grad_norm": 0.26346316933631897,
+      "learning_rate": 9.963528630192098e-06,
+      "loss": 0.086,
+      "step": 323
+    },
+    {
+      "epoch": 0.06791029134353385,
+      "grad_norm": 0.24423374235630035,
+      "learning_rate": 9.963118194374585e-06,
+      "loss": 0.0885,
+      "step": 324
+    },
+    {
+      "epoch": 0.0681198910081744,
+      "grad_norm": 0.21973365545272827,
+      "learning_rate": 9.962705470560616e-06,
+      "loss": 0.0876,
+      "step": 325
+    },
+    {
+      "epoch": 0.06832949067281492,
+      "grad_norm": 0.23790043592453003,
+      "learning_rate": 9.962290458940456e-06,
+      "loss": 0.0858,
+      "step": 326
+    },
+    {
+      "epoch": 0.06853909033745546,
+      "grad_norm": 0.23737956583499908,
+      "learning_rate": 9.961873159705426e-06,
+      "loss": 0.0813,
+      "step": 327
+    },
+    {
+      "epoch": 0.06874869000209599,
+      "grad_norm": 0.1772824376821518,
+      "learning_rate": 9.961453573047898e-06,
+      "loss": 0.0837,
+      "step": 328
+    },
+    {
+      "epoch": 0.06895828966673653,
+      "grad_norm": 0.14443804323673248,
+      "learning_rate": 9.961031699161305e-06,
+      "loss": 0.0847,
+      "step": 329
+    },
+    {
+      "epoch": 0.06916788933137707,
+      "grad_norm": 0.16732539236545563,
+      "learning_rate": 9.960607538240129e-06,
+      "loss": 0.0901,
+      "step": 330
+    },
+    {
+      "epoch": 0.0693774889960176,
+      "grad_norm": 0.16969183087348938,
+      "learning_rate": 9.960181090479908e-06,
+      "loss": 0.0868,
+      "step": 331
+    },
+    {
+      "epoch": 0.06958708866065814,
+      "grad_norm": 0.17372508347034454,
+      "learning_rate": 9.959752356077234e-06,
+      "loss": 0.085,
+      "step": 332
+    },
+    {
+      "epoch": 0.06979668832529869,
+      "grad_norm": 0.18569670617580414,
+      "learning_rate": 9.959321335229754e-06,
+      "loss": 0.0894,
+      "step": 333
+    },
+    {
+      "epoch": 0.07000628798993921,
+      "grad_norm": 0.2023860216140747,
+      "learning_rate": 9.95888802813617e-06,
+      "loss": 0.0892,
+      "step": 334
+    },
+    {
+      "epoch": 0.07021588765457976,
+      "grad_norm": 0.20857451856136322,
+      "learning_rate": 9.958452434996235e-06,
+      "loss": 0.0826,
+      "step": 335
+    },
+    {
+      "epoch": 0.07042548731922028,
+      "grad_norm": 0.23407702147960663,
+      "learning_rate": 9.958014556010757e-06,
+      "loss": 0.0795,
+      "step": 336
+    },
+    {
+      "epoch": 0.07063508698386083,
+      "grad_norm": 0.3349778950214386,
+      "learning_rate": 9.957574391381597e-06,
+      "loss": 0.0886,
+      "step": 337
+    },
+    {
+      "epoch": 0.07084468664850137,
+      "grad_norm": 0.36891162395477295,
+      "learning_rate": 9.957131941311675e-06,
+      "loss": 0.0923,
+      "step": 338
+    },
+    {
+      "epoch": 0.0710542863131419,
+      "grad_norm": 0.2576613128185272,
+      "learning_rate": 9.956687206004955e-06,
+      "loss": 0.0866,
+      "step": 339
+    },
+    {
+      "epoch": 0.07126388597778244,
+      "grad_norm": 0.2293458729982376,
+      "learning_rate": 9.956240185666465e-06,
+      "loss": 0.0864,
+      "step": 340
+    },
+    {
+      "epoch": 0.07147348564242297,
+      "grad_norm": 0.266307532787323,
+      "learning_rate": 9.955790880502278e-06,
+      "loss": 0.0923,
+      "step": 341
+    },
+    {
+      "epoch": 0.07168308530706351,
+      "grad_norm": 0.22257249057292938,
+      "learning_rate": 9.955339290719525e-06,
+      "loss": 0.0858,
+      "step": 342
+    },
+    {
+      "epoch": 0.07189268497170405,
+      "grad_norm": 0.25660571455955505,
+      "learning_rate": 9.954885416526388e-06,
+      "loss": 0.09,
+      "step": 343
+    },
+    {
+      "epoch": 0.07210228463634458,
+      "grad_norm": 0.24506278336048126,
+      "learning_rate": 9.954429258132102e-06,
+      "loss": 0.0849,
+      "step": 344
+    },
+    {
+      "epoch": 0.07231188430098512,
+      "grad_norm": 0.1874029040336609,
+      "learning_rate": 9.953970815746958e-06,
+      "loss": 0.0879,
+      "step": 345
+    },
+    {
+      "epoch": 0.07252148396562566,
+      "grad_norm": 0.2142496556043625,
+      "learning_rate": 9.953510089582297e-06,
+      "loss": 0.083,
+      "step": 346
+    },
+    {
+      "epoch": 0.07273108363026619,
+      "grad_norm": 0.200823113322258,
+      "learning_rate": 9.953047079850514e-06,
+      "loss": 0.0836,
+      "step": 347
+    },
+    {
+      "epoch": 0.07294068329490673,
+      "grad_norm": 0.22651934623718262,
+      "learning_rate": 9.952581786765057e-06,
+      "loss": 0.0872,
+      "step": 348
+    },
+    {
+      "epoch": 0.07315028295954726,
+      "grad_norm": 0.2763821482658386,
+      "learning_rate": 9.952114210540423e-06,
+      "loss": 0.0924,
+      "step": 349
+    },
+    {
+      "epoch": 0.0733598826241878,
+      "grad_norm": 0.26317474246025085,
+      "learning_rate": 9.951644351392167e-06,
+      "loss": 0.0901,
+      "step": 350
+    },
+    {
+      "epoch": 0.07356948228882834,
+      "grad_norm": 0.22677338123321533,
+      "learning_rate": 9.951172209536895e-06,
+      "loss": 0.0854,
+      "step": 351
+    },
+    {
+      "epoch": 0.07377908195346887,
+      "grad_norm": 0.23431210219860077,
+      "learning_rate": 9.95069778519226e-06,
+      "loss": 0.087,
+      "step": 352
+    },
+    {
+      "epoch": 0.07398868161810941,
+      "grad_norm": 0.23959867656230927,
+      "learning_rate": 9.950221078576977e-06,
+      "loss": 0.0839,
+      "step": 353
+    },
+    {
+      "epoch": 0.07419828128274995,
+      "grad_norm": 0.21534642577171326,
+      "learning_rate": 9.949742089910805e-06,
+      "loss": 0.0893,
+      "step": 354
+    },
+    {
+      "epoch": 0.07440788094739048,
+      "grad_norm": 0.2186158299446106,
+      "learning_rate": 9.949260819414557e-06,
+      "loss": 0.0807,
+      "step": 355
+    },
+    {
+      "epoch": 0.07461748061203102,
+      "grad_norm": 0.23446103930473328,
+      "learning_rate": 9.948777267310099e-06,
+      "loss": 0.0836,
+      "step": 356
+    },
+    {
+      "epoch": 0.07482708027667155,
+      "grad_norm": 0.20559126138687134,
+      "learning_rate": 9.948291433820348e-06,
+      "loss": 0.0866,
+      "step": 357
+    },
+    {
+      "epoch": 0.07503667994131209,
+      "grad_norm": 0.1898106038570404,
+      "learning_rate": 9.947803319169275e-06,
+      "loss": 0.0869,
+      "step": 358
+    },
+    {
+      "epoch": 0.07524627960595263,
+      "grad_norm": 0.2029874622821808,
+      "learning_rate": 9.9473129235819e-06,
+      "loss": 0.0866,
+      "step": 359
+    },
+    {
+      "epoch": 0.07545587927059316,
+      "grad_norm": 0.20538607239723206,
+      "learning_rate": 9.946820247284295e-06,
+      "loss": 0.0795,
+      "step": 360
+    },
+    {
+      "epoch": 0.0756654789352337,
+      "grad_norm": 0.24543048441410065,
+      "learning_rate": 9.946325290503583e-06,
+      "loss": 0.079,
+      "step": 361
+    },
+    {
+      "epoch": 0.07587507859987425,
+      "grad_norm": 0.2604290246963501,
+      "learning_rate": 9.945828053467939e-06,
+      "loss": 0.0871,
+      "step": 362
+    },
+    {
+      "epoch": 0.07608467826451477,
+      "grad_norm": 0.249879851937294,
+      "learning_rate": 9.945328536406588e-06,
+      "loss": 0.0826,
+      "step": 363
+    },
+    {
+      "epoch": 0.07629427792915532,
+      "grad_norm": 0.23733551800251007,
+      "learning_rate": 9.944826739549812e-06,
+      "loss": 0.0852,
+      "step": 364
+    },
+    {
+      "epoch": 0.07650387759379584,
+      "grad_norm": 0.18714891374111176,
+      "learning_rate": 9.944322663128936e-06,
+      "loss": 0.0845,
+      "step": 365
+    },
+    {
+      "epoch": 0.07671347725843639,
+      "grad_norm": 0.17555269598960876,
+      "learning_rate": 9.943816307376337e-06,
+      "loss": 0.0801,
+      "step": 366
+    },
+    {
+      "epoch": 0.07692307692307693,
+      "grad_norm": 0.20943832397460938,
+      "learning_rate": 9.943307672525448e-06,
+      "loss": 0.0853,
+      "step": 367
+    },
+    {
+      "epoch": 0.07713267658771746,
+      "grad_norm": 0.21346049010753632,
+      "learning_rate": 9.94279675881075e-06,
+      "loss": 0.0841,
+      "step": 368
+    },
+    {
+      "epoch": 0.077342276252358,
+      "grad_norm": 0.2275484949350357,
+      "learning_rate": 9.942283566467773e-06,
+      "loss": 0.0872,
+      "step": 369
+    },
+    {
+      "epoch": 0.07755187591699854,
+      "grad_norm": 0.21712201833724976,
+      "learning_rate": 9.941768095733098e-06,
+      "loss": 0.0833,
+      "step": 370
+    },
+    {
+      "epoch": 0.07776147558163907,
+      "grad_norm": 0.19494830071926117,
+      "learning_rate": 9.941250346844358e-06,
+      "loss": 0.0855,
+      "step": 371
+    },
+    {
+      "epoch": 0.07797107524627961,
+      "grad_norm": 0.19557230174541473,
+      "learning_rate": 9.940730320040233e-06,
+      "loss": 0.0827,
+      "step": 372
+    },
+    {
+      "epoch": 0.07818067491092014,
+      "grad_norm": 0.1738850325345993,
+      "learning_rate": 9.940208015560458e-06,
+      "loss": 0.0811,
+      "step": 373
+    },
+    {
+      "epoch": 0.07839027457556068,
+      "grad_norm": 0.19140446186065674,
+      "learning_rate": 9.939683433645813e-06,
+      "loss": 0.0806,
+      "step": 374
+    },
+    {
+      "epoch": 0.07859987424020122,
+      "grad_norm": 0.23054589331150055,
+      "learning_rate": 9.939156574538131e-06,
+      "loss": 0.0817,
+      "step": 375
+    },
+    {
+      "epoch": 0.07880947390484175,
+      "grad_norm": 0.25932809710502625,
+      "learning_rate": 9.938627438480295e-06,
+      "loss": 0.0801,
+      "step": 376
+    },
+    {
+      "epoch": 0.07901907356948229,
+      "grad_norm": 0.24759088456630707,
+      "learning_rate": 9.938096025716235e-06,
+      "loss": 0.0894,
+      "step": 377
+    },
+    {
+      "epoch": 0.07922867323412283,
+      "grad_norm": 0.15014545619487762,
+      "learning_rate": 9.937562336490934e-06,
+      "loss": 0.0855,
+      "step": 378
+    },
+    {
+      "epoch": 0.07943827289876336,
+      "grad_norm": 0.10033906251192093,
+      "learning_rate": 9.93702637105042e-06,
+      "loss": 0.0854,
+      "step": 379
+    },
+    {
+      "epoch": 0.0796478725634039,
+      "grad_norm": 0.1658346801996231,
+      "learning_rate": 9.936488129641777e-06,
+      "loss": 0.0812,
+      "step": 380
+    },
+    {
+      "epoch": 0.07985747222804443,
+      "grad_norm": 0.2382059395313263,
+      "learning_rate": 9.935947612513129e-06,
+      "loss": 0.0819,
+      "step": 381
+    },
+    {
+      "epoch": 0.08006707189268497,
+      "grad_norm": 0.2596674859523773,
+      "learning_rate": 9.935404819913658e-06,
+      "loss": 0.0849,
+      "step": 382
+    },
+    {
+      "epoch": 0.08027667155732551,
+      "grad_norm": 0.19236969947814941,
+      "learning_rate": 9.93485975209359e-06,
+      "loss": 0.0805,
+      "step": 383
+    },
+    {
+      "epoch": 0.08048627122196604,
+      "grad_norm": 0.12010635435581207,
+      "learning_rate": 9.934312409304201e-06,
+      "loss": 0.0858,
+      "step": 384
+    },
+    {
+      "epoch": 0.08069587088660658,
+      "grad_norm": 0.1619727462530136,
+      "learning_rate": 9.933762791797816e-06,
+      "loss": 0.0829,
+      "step": 385
+    },
+    {
+      "epoch": 0.08090547055124712,
+      "grad_norm": 0.256807804107666,
+      "learning_rate": 9.93321089982781e-06,
+      "loss": 0.0805,
+      "step": 386
+    },
+    {
+      "epoch": 0.08111507021588765,
+      "grad_norm": 0.3097319006919861,
+      "learning_rate": 9.932656733648602e-06,
+      "loss": 0.0834,
+      "step": 387
+    },
+    {
+      "epoch": 0.0813246698805282,
+      "grad_norm": 0.3272354006767273,
+      "learning_rate": 9.932100293515667e-06,
+      "loss": 0.0817,
+      "step": 388
+    },
+    {
+      "epoch": 0.08153426954516872,
+      "grad_norm": 0.36321696639060974,
+      "learning_rate": 9.931541579685519e-06,
+      "loss": 0.0859,
+      "step": 389
+    },
+    {
+      "epoch": 0.08174386920980926,
+      "grad_norm": 0.40557923913002014,
+      "learning_rate": 9.930980592415728e-06,
+      "loss": 0.0852,
+      "step": 390
+    },
+    {
+      "epoch": 0.0819534688744498,
+      "grad_norm": 0.3595048785209656,
+      "learning_rate": 9.93041733196491e-06,
+      "loss": 0.0873,
+      "step": 391
+    },
+    {
+      "epoch": 0.08216306853909033,
+      "grad_norm": 0.18180204927921295,
+      "learning_rate": 9.929851798592723e-06,
+      "loss": 0.0863,
+      "step": 392
+    },
+    {
+      "epoch": 0.08237266820373088,
+      "grad_norm": 0.3407440781593323,
+      "learning_rate": 9.929283992559882e-06,
+      "loss": 0.0889,
+      "step": 393
+    },
+    {
+      "epoch": 0.08258226786837142,
+      "grad_norm": 0.39048606157302856,
+      "learning_rate": 9.928713914128146e-06,
+      "loss": 0.0907,
+      "step": 394
+    },
+    {
+      "epoch": 0.08279186753301195,
+      "grad_norm": 0.2129063457250595,
+      "learning_rate": 9.928141563560316e-06,
+      "loss": 0.0911,
+      "step": 395
+    },
+    {
+      "epoch": 0.08300146719765249,
+      "grad_norm": 0.3368994891643524,
+      "learning_rate": 9.92756694112025e-06,
+      "loss": 0.09,
+      "step": 396
+    },
+    {
+      "epoch": 0.08321106686229301,
+      "grad_norm": 0.3096039891242981,
+      "learning_rate": 9.926990047072849e-06,
+      "loss": 0.0918,
+      "step": 397
+    },
+    {
+      "epoch": 0.08342066652693356,
+      "grad_norm": 0.21943970024585724,
+      "learning_rate": 9.92641088168406e-06,
+      "loss": 0.0892,
+      "step": 398
+    },
+    {
+      "epoch": 0.0836302661915741,
+      "grad_norm": 0.23184233903884888,
+      "learning_rate": 9.925829445220876e-06,
+      "loss": 0.0848,
+      "step": 399
+    },
+    {
+      "epoch": 0.08383986585621463,
+      "grad_norm": 0.22483260929584503,
+      "learning_rate": 9.925245737951342e-06,
+      "loss": 0.0806,
+      "step": 400
+    },
+    {
+      "epoch": 0.08404946552085517,
+      "grad_norm": 0.20455621182918549,
+      "learning_rate": 9.924659760144546e-06,
+      "loss": 0.0873,
+      "step": 401
+    },
+    {
+      "epoch": 0.08425906518549571,
+      "grad_norm": 0.19038546085357666,
+      "learning_rate": 9.924071512070623e-06,
+      "loss": 0.0832,
+      "step": 402
+    },
+    {
+      "epoch": 0.08446866485013624,
+      "grad_norm": 0.1766170859336853,
+      "learning_rate": 9.923480994000756e-06,
+      "loss": 0.0841,
+      "step": 403
+    },
+    {
+      "epoch": 0.08467826451477678,
+      "grad_norm": 0.14354343712329865,
+      "learning_rate": 9.922888206207174e-06,
+      "loss": 0.0889,
+      "step": 404
+    },
+    {
+      "epoch": 0.08488786417941731,
+      "grad_norm": 0.17357495427131653,
+      "learning_rate": 9.922293148963152e-06,
+      "loss": 0.0795,
+      "step": 405
+    },
+    {
+      "epoch": 0.08509746384405785,
+      "grad_norm": 0.16406401991844177,
+      "learning_rate": 9.921695822543009e-06,
+      "loss": 0.0837,
+      "step": 406
+    },
+    {
+      "epoch": 0.08530706350869839,
+      "grad_norm": 0.16059234738349915,
+      "learning_rate": 9.921096227222115e-06,
+      "loss": 0.0823,
+      "step": 407
+    },
+    {
+      "epoch": 0.08551666317333892,
+      "grad_norm": 0.16859528422355652,
+      "learning_rate": 9.920494363276882e-06,
+      "loss": 0.0842,
+      "step": 408
+    },
+    {
+      "epoch": 0.08572626283797946,
+      "grad_norm": 0.11957108229398727,
+      "learning_rate": 9.91989023098477e-06,
+      "loss": 0.0817,
+      "step": 409
+    },
+    {
+      "epoch": 0.08593586250261999,
+      "grad_norm": 0.1288835108280182,
+      "learning_rate": 9.919283830624283e-06,
+      "loss": 0.0845,
+      "step": 410
+    },
+    {
+      "epoch": 0.08614546216726053,
+      "grad_norm": 0.14702066779136658,
+      "learning_rate": 9.918675162474974e-06,
+      "loss": 0.0813,
+      "step": 411
+    },
+    {
+      "epoch": 0.08635506183190107,
+      "grad_norm": 0.13575831055641174,
+      "learning_rate": 9.918064226817435e-06,
+      "loss": 0.0801,
+      "step": 412
+    },
+    {
+      "epoch": 0.0865646614965416,
+      "grad_norm": 0.16246341168880463,
+      "learning_rate": 9.91745102393331e-06,
+      "loss": 0.0843,
+      "step": 413
+    },
+    {
+      "epoch": 0.08677426116118214,
+      "grad_norm": 0.1807781308889389,
+      "learning_rate": 9.916835554105282e-06,
+      "loss": 0.0809,
+      "step": 414
+    },
+    {
+      "epoch": 0.08698386082582268,
+      "grad_norm": 0.18538668751716614,
+      "learning_rate": 9.916217817617087e-06,
+      "loss": 0.083,
+      "step": 415
+    },
+    {
+      "epoch": 0.08719346049046321,
+      "grad_norm": 0.19886340200901031,
+      "learning_rate": 9.915597814753498e-06,
+      "loss": 0.078,
+      "step": 416
+    },
+    {
+      "epoch": 0.08740306015510375,
+      "grad_norm": 0.20253928005695343,
+      "learning_rate": 9.914975545800339e-06,
+      "loss": 0.0817,
+      "step": 417
+    },
+    {
+      "epoch": 0.08761265981974428,
+      "grad_norm": 0.1704034060239792,
+      "learning_rate": 9.914351011044472e-06,
+      "loss": 0.0832,
+      "step": 418
+    },
+    {
+      "epoch": 0.08782225948438482,
+      "grad_norm": 0.14112436771392822,
+      "learning_rate": 9.913724210773812e-06,
+      "loss": 0.0819,
+      "step": 419
+    },
+    {
+      "epoch": 0.08803185914902537,
+      "grad_norm": 0.11341799050569534,
+      "learning_rate": 9.91309514527731e-06,
+      "loss": 0.0804,
+      "step": 420
+    },
+    {
+      "epoch": 0.0882414588136659,
+      "grad_norm": 0.12630800902843475,
+      "learning_rate": 9.912463814844968e-06,
+      "loss": 0.0773,
+      "step": 421
+    },
+    {
+      "epoch": 0.08845105847830644,
+      "grad_norm": 0.1604209989309311,
+      "learning_rate": 9.911830219767827e-06,
+      "loss": 0.0821,
+      "step": 422
+    },
+    {
+      "epoch": 0.08866065814294698,
+      "grad_norm": 0.1671822965145111,
+      "learning_rate": 9.911194360337975e-06,
+      "loss": 0.083,
+      "step": 423
+    },
+    {
+      "epoch": 0.0888702578075875,
+      "grad_norm": 0.14921946823596954,
+      "learning_rate": 9.910556236848543e-06,
+      "loss": 0.0817,
+      "step": 424
+    },
+    {
+      "epoch": 0.08907985747222805,
+      "grad_norm": 0.13964901864528656,
+      "learning_rate": 9.909915849593705e-06,
+      "loss": 0.0815,
+      "step": 425
+    },
+    {
+      "epoch": 0.08928945713686857,
+      "grad_norm": 0.1505957692861557,
+      "learning_rate": 9.909273198868682e-06,
+      "loss": 0.0837,
+      "step": 426
+    },
+    {
+      "epoch": 0.08949905680150912,
+      "grad_norm": 0.15062043070793152,
+      "learning_rate": 9.908628284969732e-06,
+      "loss": 0.08,
+      "step": 427
+    },
+    {
+      "epoch": 0.08970865646614966,
+      "grad_norm": 0.15089496970176697,
+      "learning_rate": 9.907981108194165e-06,
+      "loss": 0.0811,
+      "step": 428
+    },
+    {
+      "epoch": 0.08991825613079019,
+      "grad_norm": 0.1628749966621399,
+      "learning_rate": 9.907331668840325e-06,
+      "loss": 0.0797,
+      "step": 429
+    },
+    {
+      "epoch": 0.09012785579543073,
+      "grad_norm": 0.1800432950258255,
+      "learning_rate": 9.906679967207604e-06,
+      "loss": 0.08,
+      "step": 430
+    },
+    {
+      "epoch": 0.09033745546007127,
+      "grad_norm": 0.1852189302444458,
+      "learning_rate": 9.906026003596438e-06,
+      "loss": 0.0866,
+      "step": 431
+    },
+    {
+      "epoch": 0.0905470551247118,
+      "grad_norm": 0.16221773624420166,
+      "learning_rate": 9.905369778308304e-06,
+      "loss": 0.0785,
+      "step": 432
+    },
+    {
+      "epoch": 0.09075665478935234,
+      "grad_norm": 0.13077011704444885,
+      "learning_rate": 9.904711291645721e-06,
+      "loss": 0.0793,
+      "step": 433
+    },
+    {
+      "epoch": 0.09096625445399287,
+      "grad_norm": 0.14190125465393066,
+      "learning_rate": 9.904050543912252e-06,
+      "loss": 0.0774,
+      "step": 434
+    },
+    {
+      "epoch": 0.09117585411863341,
+      "grad_norm": 0.1918221116065979,
+      "learning_rate": 9.9033875354125e-06,
+      "loss": 0.086,
+      "step": 435
+    },
+    {
+      "epoch": 0.09138545378327395,
+      "grad_norm": 0.24989114701747894,
+      "learning_rate": 9.902722266452112e-06,
+      "loss": 0.0751,
+      "step": 436
+    },
+    {
+      "epoch": 0.09159505344791448,
+      "grad_norm": 0.2872444987297058,
+      "learning_rate": 9.902054737337778e-06,
+      "loss": 0.0794,
+      "step": 437
+    },
+    {
+      "epoch": 0.09180465311255502,
+      "grad_norm": 0.26268768310546875,
+      "learning_rate": 9.90138494837723e-06,
+      "loss": 0.0783,
+      "step": 438
+    },
+    {
+      "epoch": 0.09201425277719556,
+      "grad_norm": 0.19035717844963074,
+      "learning_rate": 9.900712899879237e-06,
+      "loss": 0.0777,
+      "step": 439
+    },
+    {
+      "epoch": 0.09222385244183609,
+      "grad_norm": 0.14101530611515045,
+      "learning_rate": 9.900038592153616e-06,
+      "loss": 0.0783,
+      "step": 440
+    },
+    {
+      "epoch": 0.09243345210647663,
+      "grad_norm": 0.21867899596691132,
+      "learning_rate": 9.899362025511221e-06,
+      "loss": 0.0799,
+      "step": 441
+    },
+    {
+      "epoch": 0.09264305177111716,
+      "grad_norm": 0.3020075857639313,
+      "learning_rate": 9.898683200263951e-06,
+      "loss": 0.0903,
+      "step": 442
+    },
+    {
+      "epoch": 0.0928526514357577,
+      "grad_norm": 0.29326069355010986,
+      "learning_rate": 9.898002116724743e-06,
+      "loss": 0.081,
+      "step": 443
+    },
+    {
+      "epoch": 0.09306225110039824,
+      "grad_norm": 0.2977989912033081,
+      "learning_rate": 9.897318775207576e-06,
+      "loss": 0.0794,
+      "step": 444
+    },
+    {
+      "epoch": 0.09327185076503877,
+      "grad_norm": 0.3779714107513428,
+      "learning_rate": 9.89663317602747e-06,
+      "loss": 0.0853,
+      "step": 445
+    },
+    {
+      "epoch": 0.09348145042967931,
+      "grad_norm": 0.3647499680519104,
+      "learning_rate": 9.895945319500488e-06,
+      "loss": 0.0878,
+      "step": 446
+    },
+    {
+      "epoch": 0.09369105009431986,
+      "grad_norm": 0.18845055997371674,
+      "learning_rate": 9.895255205943732e-06,
+      "loss": 0.0776,
+      "step": 447
+    },
+    {
+      "epoch": 0.09390064975896038,
+      "grad_norm": 0.23743781447410583,
+      "learning_rate": 9.894562835675343e-06,
+      "loss": 0.0822,
+      "step": 448
+    },
+    {
+      "epoch": 0.09411024942360093,
+      "grad_norm": 0.32447198033332825,
+      "learning_rate": 9.893868209014502e-06,
+      "loss": 0.0813,
+      "step": 449
+    },
+    {
+      "epoch": 0.09431984908824145,
+      "grad_norm": 0.22521555423736572,
+      "learning_rate": 9.893171326281433e-06,
+      "loss": 0.0779,
+      "step": 450
+    },
+    {
+      "epoch": 0.094529448752882,
+      "grad_norm": 0.2047656625509262,
+      "learning_rate": 9.8924721877974e-06,
+      "loss": 0.0787,
+      "step": 451
+    },
+    {
+      "epoch": 0.09473904841752254,
+      "grad_norm": 0.241439089179039,
+      "learning_rate": 9.891770793884703e-06,
+      "loss": 0.0838,
+      "step": 452
+    },
+    {
+      "epoch": 0.09494864808216306,
+      "grad_norm": 0.18499858677387238,
+      "learning_rate": 9.891067144866687e-06,
+      "loss": 0.0806,
+      "step": 453
+    },
+    {
+      "epoch": 0.0951582477468036,
+      "grad_norm": 0.19206255674362183,
+      "learning_rate": 9.890361241067734e-06,
+      "loss": 0.084,
+      "step": 454
+    },
+    {
+      "epoch": 0.09536784741144415,
+      "grad_norm": 0.21595366299152374,
+      "learning_rate": 9.889653082813264e-06,
+      "loss": 0.0796,
+      "step": 455
+    },
+    {
+      "epoch": 0.09557744707608468,
+      "grad_norm": 0.19207443296909332,
+      "learning_rate": 9.888942670429738e-06,
+      "loss": 0.0791,
+      "step": 456
+    },
+    {
+      "epoch": 0.09578704674072522,
+      "grad_norm": 0.16599828004837036,
+      "learning_rate": 9.888230004244657e-06,
+      "loss": 0.0806,
+      "step": 457
+    },
+    {
+      "epoch": 0.09599664640536575,
+      "grad_norm": 0.1895645707845688,
+      "learning_rate": 9.88751508458656e-06,
+      "loss": 0.0784,
+      "step": 458
+    },
+    {
+      "epoch": 0.09620624607000629,
+      "grad_norm": 0.18604148924350739,
+      "learning_rate": 9.886797911785023e-06,
+      "loss": 0.0858,
+      "step": 459
+    },
+    {
+      "epoch": 0.09641584573464683,
+      "grad_norm": 0.12998424470424652,
+      "learning_rate": 9.886078486170665e-06,
+      "loss": 0.0792,
+      "step": 460
+    },
+    {
+      "epoch": 0.09662544539928736,
+      "grad_norm": 0.16682370007038116,
+      "learning_rate": 9.885356808075139e-06,
+      "loss": 0.0798,
+      "step": 461
+    },
+    {
+      "epoch": 0.0968350450639279,
+      "grad_norm": 0.17146462202072144,
+      "learning_rate": 9.884632877831139e-06,
+      "loss": 0.0797,
+      "step": 462
+    },
+    {
+      "epoch": 0.09704464472856844,
+      "grad_norm": 0.11582599580287933,
+      "learning_rate": 9.883906695772399e-06,
+      "loss": 0.0812,
+      "step": 463
+    },
+    {
+      "epoch": 0.09725424439320897,
+      "grad_norm": 0.1586909145116806,
+      "learning_rate": 9.883178262233684e-06,
+      "loss": 0.0817,
+      "step": 464
+    },
+    {
+      "epoch": 0.09746384405784951,
+      "grad_norm": 0.1509367823600769,
+      "learning_rate": 9.882447577550805e-06,
+      "loss": 0.0845,
+      "step": 465
+    },
+    {
+      "epoch": 0.09767344372249004,
+      "grad_norm": 0.1255314201116562,
+      "learning_rate": 9.881714642060609e-06,
+      "loss": 0.0789,
+      "step": 466
+    },
+    {
+      "epoch": 0.09788304338713058,
+      "grad_norm": 0.17706221342086792,
+      "learning_rate": 9.880979456100974e-06,
+      "loss": 0.0821,
+      "step": 467
+    },
+    {
+      "epoch": 0.09809264305177112,
+      "grad_norm": 0.23085959255695343,
+      "learning_rate": 9.880242020010827e-06,
+      "loss": 0.0801,
+      "step": 468
+    },
+    {
+      "epoch": 0.09830224271641165,
+      "grad_norm": 0.354623407125473,
+      "learning_rate": 9.87950233413012e-06,
+      "loss": 0.0826,
+      "step": 469
+    },
+    {
+      "epoch": 0.09851184238105219,
+      "grad_norm": 0.4140135943889618,
+      "learning_rate": 9.87876039879985e-06,
+      "loss": 0.0901,
+      "step": 470
+    },
+    {
+      "epoch": 0.09872144204569272,
+      "grad_norm": 0.289201021194458,
+      "learning_rate": 9.878016214362051e-06,
+      "loss": 0.0836,
+      "step": 471
+    },
+    {
+      "epoch": 0.09893104171033326,
+      "grad_norm": 0.16922371089458466,
+      "learning_rate": 9.877269781159791e-06,
+      "loss": 0.0786,
+      "step": 472
+    },
+    {
+      "epoch": 0.0991406413749738,
+      "grad_norm": 0.2799324095249176,
+      "learning_rate": 9.876521099537173e-06,
+      "loss": 0.0827,
+      "step": 473
+    },
+    {
+      "epoch": 0.09935024103961433,
+      "grad_norm": 0.3039519190788269,
+      "learning_rate": 9.875770169839343e-06,
+      "loss": 0.0852,
+      "step": 474
+    },
+    {
+      "epoch": 0.09955984070425487,
+      "grad_norm": 0.19314952194690704,
+      "learning_rate": 9.875016992412476e-06,
+      "loss": 0.0777,
+      "step": 475
+    },
+    {
+      "epoch": 0.09976944036889542,
+      "grad_norm": 0.1906445324420929,
+      "learning_rate": 9.87426156760379e-06,
+      "loss": 0.0788,
+      "step": 476
+    },
+    {
+      "epoch": 0.09997904003353594,
+      "grad_norm": 0.19903278350830078,
+      "learning_rate": 9.87350389576153e-06,
+      "loss": 0.0836,
+      "step": 477
+    },
+    {
+      "epoch": 0.10018863969817648,
+      "grad_norm": 0.2028076946735382,
+      "learning_rate": 9.872743977234992e-06,
+      "loss": 0.0797,
+      "step": 478
+    },
+    {
+      "epoch": 0.10039823936281701,
+      "grad_norm": 0.1903727948665619,
+      "learning_rate": 9.871981812374488e-06,
+      "loss": 0.0791,
+      "step": 479
+    },
+    {
+      "epoch": 0.10060783902745755,
+      "grad_norm": 0.16171634197235107,
+      "learning_rate": 9.871217401531382e-06,
+      "loss": 0.0819,
+      "step": 480
+    },
+    {
+      "epoch": 0.1008174386920981,
+      "grad_norm": 0.18417572975158691,
+      "learning_rate": 9.870450745058066e-06,
+      "loss": 0.0836,
+      "step": 481
+    },
+    {
+      "epoch": 0.10102703835673862,
+      "grad_norm": 0.18208323419094086,
+      "learning_rate": 9.869681843307968e-06,
+      "loss": 0.0834,
+      "step": 482
+    },
+    {
+      "epoch": 0.10123663802137917,
+      "grad_norm": 0.17092318832874298,
+      "learning_rate": 9.868910696635551e-06,
+      "loss": 0.0791,
+      "step": 483
+    },
+    {
+      "epoch": 0.10144623768601971,
+      "grad_norm": 0.13848985731601715,
+      "learning_rate": 9.868137305396317e-06,
+      "loss": 0.0766,
+      "step": 484
+    },
+    {
+      "epoch": 0.10165583735066024,
+      "grad_norm": 0.17035254836082458,
+      "learning_rate": 9.867361669946793e-06,
+      "loss": 0.0812,
+      "step": 485
+    },
+    {
+      "epoch": 0.10186543701530078,
+      "grad_norm": 0.1731637865304947,
+      "learning_rate": 9.866583790644553e-06,
+      "loss": 0.078,
+      "step": 486
+    },
+    {
+      "epoch": 0.1020750366799413,
+      "grad_norm": 0.16337701678276062,
+      "learning_rate": 9.865803667848195e-06,
+      "loss": 0.0807,
+      "step": 487
+    },
+    {
+      "epoch": 0.10228463634458185,
+      "grad_norm": 0.16541989147663116,
+      "learning_rate": 9.865021301917358e-06,
+      "loss": 0.0792,
+      "step": 488
+    },
+    {
+      "epoch": 0.10249423600922239,
+      "grad_norm": 0.13967937231063843,
+      "learning_rate": 9.864236693212709e-06,
+      "loss": 0.0748,
+      "step": 489
+    },
+    {
+      "epoch": 0.10270383567386292,
+      "grad_norm": 0.15016354620456696,
+      "learning_rate": 9.863449842095955e-06,
+      "loss": 0.0794,
+      "step": 490
+    },
+    {
+      "epoch": 0.10291343533850346,
+      "grad_norm": 0.18421977758407593,
+      "learning_rate": 9.862660748929835e-06,
+      "loss": 0.0783,
+      "step": 491
+    },
+    {
+      "epoch": 0.103123035003144,
+      "grad_norm": 0.21068471670150757,
+      "learning_rate": 9.86186941407812e-06,
+      "loss": 0.0785,
+      "step": 492
+    },
+    {
+      "epoch": 0.10333263466778453,
+      "grad_norm": 0.23309852182865143,
+      "learning_rate": 9.861075837905616e-06,
+      "loss": 0.0783,
+      "step": 493
+    },
+    {
+      "epoch": 0.10354223433242507,
+      "grad_norm": 0.23478467762470245,
+      "learning_rate": 9.860280020778158e-06,
+      "loss": 0.0806,
+      "step": 494
+    },
+    {
+      "epoch": 0.1037518339970656,
+      "grad_norm": 0.19462421536445618,
+      "learning_rate": 9.859481963062623e-06,
+      "loss": 0.0747,
+      "step": 495
+    },
+    {
+      "epoch": 0.10396143366170614,
+      "grad_norm": 0.15075665712356567,
+      "learning_rate": 9.85868166512691e-06,
+      "loss": 0.0748,
+      "step": 496
+    },
+    {
+      "epoch": 0.10417103332634668,
+      "grad_norm": 0.13068079948425293,
+      "learning_rate": 9.85787912733996e-06,
+      "loss": 0.079,
+      "step": 497
+    },
+    {
+      "epoch": 0.10438063299098721,
+      "grad_norm": 0.14314256608486176,
+      "learning_rate": 9.85707435007174e-06,
+      "loss": 0.0805,
+      "step": 498
+    },
+    {
+      "epoch": 0.10459023265562775,
+      "grad_norm": 0.18346239626407623,
+      "learning_rate": 9.856267333693255e-06,
+      "loss": 0.0804,
+      "step": 499
+    },
+    {
+      "epoch": 0.1047998323202683,
+      "grad_norm": 0.2016962468624115,
+      "learning_rate": 9.855458078576537e-06,
+      "loss": 0.077,
+      "step": 500
+    },
+    {
+      "epoch": 0.10500943198490882,
+      "grad_norm": 0.18932455778121948,
+      "learning_rate": 9.854646585094654e-06,
+      "loss": 0.0811,
+      "step": 501
+    },
+    {
+      "epoch": 0.10521903164954936,
+      "grad_norm": 0.1751541942358017,
+      "learning_rate": 9.853832853621703e-06,
+      "loss": 0.0776,
+      "step": 502
+    },
+    {
+      "epoch": 0.10542863131418989,
+      "grad_norm": 0.18486347794532776,
+      "learning_rate": 9.853016884532814e-06,
+      "loss": 0.077,
+      "step": 503
+    },
+    {
+      "epoch": 0.10563823097883043,
+      "grad_norm": 0.20413736999034882,
+      "learning_rate": 9.85219867820415e-06,
+      "loss": 0.0793,
+      "step": 504
+    },
+    {
+      "epoch": 0.10584783064347097,
+      "grad_norm": 0.19140803813934326,
+      "learning_rate": 9.851378235012905e-06,
+      "loss": 0.0792,
+      "step": 505
+    },
+    {
+      "epoch": 0.1060574303081115,
+      "grad_norm": 0.14764432609081268,
+      "learning_rate": 9.8505555553373e-06,
+      "loss": 0.0826,
+      "step": 506
+    },
+    {
+      "epoch": 0.10626702997275204,
+      "grad_norm": 0.1288614273071289,
+      "learning_rate": 9.84973063955659e-06,
+      "loss": 0.0768,
+      "step": 507
+    },
+    {
+      "epoch": 0.10647662963739259,
+      "grad_norm": 0.14826104044914246,
+      "learning_rate": 9.848903488051065e-06,
+      "loss": 0.0772,
+      "step": 508
+    },
+    {
+      "epoch": 0.10668622930203311,
+      "grad_norm": 0.14957068860530853,
+      "learning_rate": 9.848074101202037e-06,
+      "loss": 0.0755,
+      "step": 509
+    },
+    {
+      "epoch": 0.10689582896667366,
+      "grad_norm": 0.11332403123378754,
+      "learning_rate": 9.847242479391857e-06,
+      "loss": 0.0733,
+      "step": 510
+    },
+    {
+      "epoch": 0.10710542863131418,
+      "grad_norm": 0.10584530979394913,
+      "learning_rate": 9.8464086230039e-06,
+      "loss": 0.0752,
+      "step": 511
+    },
+    {
+      "epoch": 0.10731502829595473,
+      "grad_norm": 0.1642168015241623,
+      "learning_rate": 9.845572532422576e-06,
+      "loss": 0.0752,
+      "step": 512
+    },
+    {
+      "epoch": 0.10752462796059527,
+      "grad_norm": 0.22667261958122253,
+      "learning_rate": 9.844734208033318e-06,
+      "loss": 0.0826,
+      "step": 513
+    },
+    {
+      "epoch": 0.1077342276252358,
+      "grad_norm": 0.2658557593822479,
+      "learning_rate": 9.843893650222599e-06,
+      "loss": 0.0838,
+      "step": 514
+    },
+    {
+      "epoch": 0.10794382728987634,
+      "grad_norm": 0.27352774143218994,
+      "learning_rate": 9.843050859377913e-06,
+      "loss": 0.0832,
+      "step": 515
+    },
+    {
+      "epoch": 0.10815342695451688,
+      "grad_norm": 0.24916625022888184,
+      "learning_rate": 9.842205835887785e-06,
+      "loss": 0.0777,
+      "step": 516
+    },
+    {
+      "epoch": 0.10836302661915741,
+      "grad_norm": 0.2406039983034134,
+      "learning_rate": 9.841358580141775e-06,
+      "loss": 0.0757,
+      "step": 517
+    },
+    {
+      "epoch": 0.10857262628379795,
+      "grad_norm": 0.23990845680236816,
+      "learning_rate": 9.840509092530465e-06,
+      "loss": 0.0778,
+      "step": 518
+    },
+    {
+      "epoch": 0.10878222594843848,
+      "grad_norm": 0.21896076202392578,
+      "learning_rate": 9.839657373445468e-06,
+      "loss": 0.0787,
+      "step": 519
+    },
+    {
+      "epoch": 0.10899182561307902,
+      "grad_norm": 0.2108394056558609,
+      "learning_rate": 9.838803423279428e-06,
+      "loss": 0.0782,
+      "step": 520
+    },
+    {
+      "epoch": 0.10920142527771956,
+      "grad_norm": 0.20889486372470856,
+      "learning_rate": 9.837947242426013e-06,
+      "loss": 0.078,
+      "step": 521
+    },
+    {
+      "epoch": 0.10941102494236009,
+      "grad_norm": 0.20450422167778015,
+      "learning_rate": 9.837088831279927e-06,
+      "loss": 0.0837,
+      "step": 522
+    },
+    {
+      "epoch": 0.10962062460700063,
+      "grad_norm": 0.19421492516994476,
+      "learning_rate": 9.836228190236892e-06,
+      "loss": 0.0744,
+      "step": 523
+    },
+    {
+      "epoch": 0.10983022427164117,
+      "grad_norm": 0.19942732155323029,
+      "learning_rate": 9.835365319693667e-06,
+      "loss": 0.0774,
+      "step": 524
+    },
+    {
+      "epoch": 0.1100398239362817,
+      "grad_norm": 0.2001711130142212,
+      "learning_rate": 9.834500220048034e-06,
+      "loss": 0.0787,
+      "step": 525
+    },
+    {
+      "epoch": 0.11024942360092224,
+      "grad_norm": 0.1794813573360443,
+      "learning_rate": 9.833632891698801e-06,
+      "loss": 0.0814,
+      "step": 526
+    },
+    {
+      "epoch": 0.11045902326556277,
+      "grad_norm": 0.16165196895599365,
+      "learning_rate": 9.832763335045812e-06,
+      "loss": 0.0761,
+      "step": 527
+    },
+    {
+      "epoch": 0.11066862293020331,
+      "grad_norm": 0.18171708285808563,
+      "learning_rate": 9.83189155048993e-06,
+      "loss": 0.0809,
+      "step": 528
+    },
+    {
+      "epoch": 0.11087822259484385,
+      "grad_norm": 0.1997329592704773,
+      "learning_rate": 9.831017538433045e-06,
+      "loss": 0.0764,
+      "step": 529
+    },
+    {
+      "epoch": 0.11108782225948438,
+      "grad_norm": 0.19061146676540375,
+      "learning_rate": 9.83014129927808e-06,
+      "loss": 0.0783,
+      "step": 530
+    },
+    {
+      "epoch": 0.11129742192412492,
+      "grad_norm": 0.1783018410205841,
+      "learning_rate": 9.829262833428978e-06,
+      "loss": 0.079,
+      "step": 531
+    },
+    {
+      "epoch": 0.11150702158876546,
+      "grad_norm": 0.18371321260929108,
+      "learning_rate": 9.828382141290713e-06,
+      "loss": 0.0769,
+      "step": 532
+    },
+    {
+      "epoch": 0.11171662125340599,
+      "grad_norm": 0.16765272617340088,
+      "learning_rate": 9.827499223269285e-06,
+      "loss": 0.0772,
+      "step": 533
+    },
+    {
+      "epoch": 0.11192622091804653,
+      "grad_norm": 0.12363309413194656,
+      "learning_rate": 9.826614079771716e-06,
+      "loss": 0.0765,
+      "step": 534
+    },
+    {
+      "epoch": 0.11213582058268706,
+      "grad_norm": 0.10953837633132935,
+      "learning_rate": 9.82572671120606e-06,
+      "loss": 0.0817,
+      "step": 535
+    },
+    {
+      "epoch": 0.1123454202473276,
+      "grad_norm": 0.14569856226444244,
+      "learning_rate": 9.824837117981392e-06,
+      "loss": 0.0754,
+      "step": 536
+    },
+    {
+      "epoch": 0.11255501991196815,
+      "grad_norm": 0.16665959358215332,
+      "learning_rate": 9.823945300507815e-06,
+      "loss": 0.0772,
+      "step": 537
+    },
+    {
+      "epoch": 0.11276461957660867,
+      "grad_norm": 0.167064368724823,
+      "learning_rate": 9.823051259196456e-06,
+      "loss": 0.0795,
+      "step": 538
+    },
+    {
+      "epoch": 0.11297421924124922,
+      "grad_norm": 0.1980743259191513,
+      "learning_rate": 9.82215499445947e-06,
+      "loss": 0.0751,
+      "step": 539
+    },
+    {
+      "epoch": 0.11318381890588974,
+      "grad_norm": 0.2310371845960617,
+      "learning_rate": 9.821256506710032e-06,
+      "loss": 0.0757,
+      "step": 540
+    },
+    {
+      "epoch": 0.11339341857053029,
+      "grad_norm": 0.2004477083683014,
+      "learning_rate": 9.820355796362346e-06,
+      "loss": 0.0774,
+      "step": 541
+    },
+    {
+      "epoch": 0.11360301823517083,
+      "grad_norm": 0.16151970624923706,
+      "learning_rate": 9.81945286383164e-06,
+      "loss": 0.0766,
+      "step": 542
+    },
+    {
+      "epoch": 0.11381261789981136,
+      "grad_norm": 0.2251511514186859,
+      "learning_rate": 9.818547709534163e-06,
+      "loss": 0.0783,
+      "step": 543
+    },
+    {
+      "epoch": 0.1140222175644519,
+      "grad_norm": 0.2668745219707489,
+      "learning_rate": 9.817640333887194e-06,
+      "loss": 0.0803,
+      "step": 544
+    },
+    {
+      "epoch": 0.11423181722909244,
+      "grad_norm": 0.24963510036468506,
+      "learning_rate": 9.816730737309032e-06,
+      "loss": 0.0782,
+      "step": 545
+    },
+    {
+      "epoch": 0.11444141689373297,
+      "grad_norm": 0.2355589121580124,
+      "learning_rate": 9.815818920219e-06,
+      "loss": 0.0758,
+      "step": 546
+    },
+    {
+      "epoch": 0.11465101655837351,
+      "grad_norm": 0.20662012696266174,
+      "learning_rate": 9.814904883037445e-06,
+      "loss": 0.0769,
+      "step": 547
+    },
+    {
+      "epoch": 0.11486061622301404,
+      "grad_norm": 0.15815681219100952,
+      "learning_rate": 9.813988626185742e-06,
+      "loss": 0.0782,
+      "step": 548
+    },
+    {
+      "epoch": 0.11507021588765458,
+      "grad_norm": 0.16817304491996765,
+      "learning_rate": 9.81307015008628e-06,
+      "loss": 0.0769,
+      "step": 549
+    },
+    {
+      "epoch": 0.11527981555229512,
+      "grad_norm": 0.19368018209934235,
+      "learning_rate": 9.81214945516248e-06,
+      "loss": 0.0796,
+      "step": 550
+    },
+    {
+      "epoch": 0.11548941521693565,
+      "grad_norm": 0.2023620903491974,
+      "learning_rate": 9.81122654183878e-06,
+      "loss": 0.0763,
+      "step": 551
+    },
+    {
+      "epoch": 0.11569901488157619,
+      "grad_norm": 0.2068459391593933,
+      "learning_rate": 9.810301410540643e-06,
+      "loss": 0.074,
+      "step": 552
+    },
+    {
+      "epoch": 0.11590861454621673,
+      "grad_norm": 0.1989372819662094,
+      "learning_rate": 9.809374061694555e-06,
+      "loss": 0.0758,
+      "step": 553
+    },
+    {
+      "epoch": 0.11611821421085726,
+      "grad_norm": 0.17017100751399994,
+      "learning_rate": 9.808444495728024e-06,
+      "loss": 0.0743,
+      "step": 554
+    },
+    {
+      "epoch": 0.1163278138754978,
+      "grad_norm": 0.16604629158973694,
+      "learning_rate": 9.80751271306958e-06,
+      "loss": 0.0794,
+      "step": 555
+    },
+    {
+      "epoch": 0.11653741354013833,
+      "grad_norm": 0.1727730929851532,
+      "learning_rate": 9.806578714148774e-06,
+      "loss": 0.0776,
+      "step": 556
+    },
+    {
+      "epoch": 0.11674701320477887,
+      "grad_norm": 0.18742415308952332,
+      "learning_rate": 9.805642499396177e-06,
+      "loss": 0.0755,
+      "step": 557
+    },
+    {
+      "epoch": 0.11695661286941941,
+      "grad_norm": 0.22661146521568298,
+      "learning_rate": 9.804704069243389e-06,
+      "loss": 0.0803,
+      "step": 558
+    },
+    {
+      "epoch": 0.11716621253405994,
+      "grad_norm": 0.25402432680130005,
+      "learning_rate": 9.80376342412302e-06,
+      "loss": 0.0761,
+      "step": 559
+    },
+    {
+      "epoch": 0.11737581219870048,
+      "grad_norm": 0.2397623509168625,
+      "learning_rate": 9.802820564468712e-06,
+      "loss": 0.0736,
+      "step": 560
+    },
+    {
+      "epoch": 0.11758541186334102,
+      "grad_norm": 0.20311333239078522,
+      "learning_rate": 9.801875490715123e-06,
+      "loss": 0.076,
+      "step": 561
+    },
+    {
+      "epoch": 0.11779501152798155,
+      "grad_norm": 0.16478881239891052,
+      "learning_rate": 9.800928203297927e-06,
+      "loss": 0.0755,
+      "step": 562
+    },
+    {
+      "epoch": 0.1180046111926221,
+      "grad_norm": 0.14853514730930328,
+      "learning_rate": 9.79997870265383e-06,
+      "loss": 0.0707,
+      "step": 563
+    },
+    {
+      "epoch": 0.11821421085726262,
+      "grad_norm": 0.17599347233772278,
+      "learning_rate": 9.799026989220543e-06,
+      "loss": 0.0779,
+      "step": 564
+    },
+    {
+      "epoch": 0.11842381052190316,
+      "grad_norm": 0.2129105031490326,
+      "learning_rate": 9.798073063436815e-06,
+      "loss": 0.0767,
+      "step": 565
+    },
+    {
+      "epoch": 0.1186334101865437,
+      "grad_norm": 0.23926801979541779,
+      "learning_rate": 9.7971169257424e-06,
+      "loss": 0.0771,
+      "step": 566
+    },
+    {
+      "epoch": 0.11884300985118423,
+      "grad_norm": 0.2424420565366745,
+      "learning_rate": 9.79615857657808e-06,
+      "loss": 0.0823,
+      "step": 567
+    },
+    {
+      "epoch": 0.11905260951582478,
+      "grad_norm": 0.21663707494735718,
+      "learning_rate": 9.795198016385651e-06,
+      "loss": 0.0778,
+      "step": 568
+    },
+    {
+      "epoch": 0.11926220918046532,
+      "grad_norm": 0.21969793736934662,
+      "learning_rate": 9.794235245607933e-06,
+      "loss": 0.0756,
+      "step": 569
+    },
+    {
+      "epoch": 0.11947180884510585,
+      "grad_norm": 0.26351258158683777,
+      "learning_rate": 9.793270264688763e-06,
+      "loss": 0.0776,
+      "step": 570
+    },
+    {
+      "epoch": 0.11968140850974639,
+      "grad_norm": 0.3025349974632263,
+      "learning_rate": 9.792303074072995e-06,
+      "loss": 0.0747,
+      "step": 571
+    },
+    {
+      "epoch": 0.11989100817438691,
+      "grad_norm": 0.30396711826324463,
+      "learning_rate": 9.791333674206507e-06,
+      "loss": 0.0796,
+      "step": 572
+    },
+    {
+      "epoch": 0.12010060783902746,
+      "grad_norm": 0.23939970135688782,
+      "learning_rate": 9.790362065536189e-06,
+      "loss": 0.0797,
+      "step": 573
+    },
+    {
+      "epoch": 0.120310207503668,
+      "grad_norm": 0.1715303361415863,
+      "learning_rate": 9.789388248509957e-06,
+      "loss": 0.075,
+      "step": 574
+    },
+    {
+      "epoch": 0.12051980716830853,
+      "grad_norm": 0.16927644610404968,
+      "learning_rate": 9.788412223576734e-06,
+      "loss": 0.0821,
+      "step": 575
+    },
+    {
+      "epoch": 0.12072940683294907,
+      "grad_norm": 0.18329590559005737,
+      "learning_rate": 9.787433991186472e-06,
+      "loss": 0.0732,
+      "step": 576
+    },
+    {
+      "epoch": 0.12093900649758961,
+      "grad_norm": 0.18802891671657562,
+      "learning_rate": 9.786453551790133e-06,
+      "loss": 0.0783,
+      "step": 577
+    },
+    {
+      "epoch": 0.12114860616223014,
+      "grad_norm": 0.1802418977022171,
+      "learning_rate": 9.785470905839703e-06,
+      "loss": 0.0733,
+      "step": 578
+    },
+    {
+      "epoch": 0.12135820582687068,
+      "grad_norm": 0.19172324240207672,
+      "learning_rate": 9.784486053788179e-06,
+      "loss": 0.0771,
+      "step": 579
+    },
+    {
+      "epoch": 0.12156780549151121,
+      "grad_norm": 0.1751202493906021,
+      "learning_rate": 9.783498996089577e-06,
+      "loss": 0.0773,
+      "step": 580
+    },
+    {
+      "epoch": 0.12177740515615175,
+      "grad_norm": 0.12133464962244034,
+      "learning_rate": 9.782509733198932e-06,
+      "loss": 0.0751,
+      "step": 581
+    },
+    {
+      "epoch": 0.12198700482079229,
+      "grad_norm": 0.14133097231388092,
+      "learning_rate": 9.781518265572294e-06,
+      "loss": 0.0748,
+      "step": 582
+    },
+    {
+      "epoch": 0.12219660448543282,
+      "grad_norm": 0.18668805062770844,
+      "learning_rate": 9.780524593666727e-06,
+      "loss": 0.0736,
+      "step": 583
+    },
+    {
+      "epoch": 0.12240620415007336,
+      "grad_norm": 0.18670430779457092,
+      "learning_rate": 9.779528717940317e-06,
+      "loss": 0.0739,
+      "step": 584
+    },
+    {
+      "epoch": 0.1226158038147139,
+      "grad_norm": 0.19386443495750427,
+      "learning_rate": 9.77853063885216e-06,
+      "loss": 0.079,
+      "step": 585
+    },
+    {
+      "epoch": 0.12282540347935443,
+      "grad_norm": 0.19840072095394135,
+      "learning_rate": 9.77753035686237e-06,
+      "loss": 0.0821,
+      "step": 586
+    },
+    {
+      "epoch": 0.12303500314399497,
+      "grad_norm": 0.2034991979598999,
+      "learning_rate": 9.77652787243208e-06,
+      "loss": 0.0744,
+      "step": 587
+    },
+    {
+      "epoch": 0.1232446028086355,
+      "grad_norm": 0.22970953583717346,
+      "learning_rate": 9.775523186023432e-06,
+      "loss": 0.0718,
+      "step": 588
+    },
+    {
+      "epoch": 0.12345420247327604,
+      "grad_norm": 0.20958104729652405,
+      "learning_rate": 9.774516298099588e-06,
+      "loss": 0.0769,
+      "step": 589
+    },
+    {
+      "epoch": 0.12366380213791658,
+      "grad_norm": 0.1604912430047989,
+      "learning_rate": 9.773507209124721e-06,
+      "loss": 0.0713,
+      "step": 590
+    },
+    {
+      "epoch": 0.12387340180255711,
+      "grad_norm": 0.16623252630233765,
+      "learning_rate": 9.772495919564022e-06,
+      "loss": 0.0728,
+      "step": 591
+    },
+    {
+      "epoch": 0.12408300146719765,
+      "grad_norm": 0.20733430981636047,
+      "learning_rate": 9.771482429883697e-06,
+      "loss": 0.0747,
+      "step": 592
+    },
+    {
+      "epoch": 0.1242926011318382,
+      "grad_norm": 0.2416054755449295,
+      "learning_rate": 9.770466740550963e-06,
+      "loss": 0.0795,
+      "step": 593
+    },
+    {
+      "epoch": 0.12450220079647872,
+      "grad_norm": 0.21449747681617737,
+      "learning_rate": 9.769448852034051e-06,
+      "loss": 0.0782,
+      "step": 594
+    },
+    {
+      "epoch": 0.12471180046111927,
+      "grad_norm": 0.15167087316513062,
+      "learning_rate": 9.768428764802209e-06,
+      "loss": 0.0818,
+      "step": 595
+    },
+    {
+      "epoch": 0.1249214001257598,
+      "grad_norm": 0.15075545012950897,
+      "learning_rate": 9.767406479325698e-06,
+      "loss": 0.0739,
+      "step": 596
+    },
+    {
+      "epoch": 0.12513099979040035,
+      "grad_norm": 0.1983119398355484,
+      "learning_rate": 9.76638199607579e-06,
+      "loss": 0.076,
+      "step": 597
+    },
+    {
+      "epoch": 0.12534059945504086,
+      "grad_norm": 0.18619777262210846,
+      "learning_rate": 9.765355315524772e-06,
+      "loss": 0.0787,
+      "step": 598
+    },
+    {
+      "epoch": 0.1255501991196814,
+      "grad_norm": 0.13842901587486267,
+      "learning_rate": 9.764326438145944e-06,
+      "loss": 0.078,
+      "step": 599
+    },
+    {
+      "epoch": 0.12575979878432195,
+      "grad_norm": 0.16321855783462524,
+      "learning_rate": 9.763295364413616e-06,
+      "loss": 0.0754,
+      "step": 600
+    },
+    {
+      "epoch": 0.1259693984489625,
+      "grad_norm": 0.18347276747226715,
+      "learning_rate": 9.762262094803115e-06,
+      "loss": 0.0757,
+      "step": 601
+    },
+    {
+      "epoch": 0.12617899811360303,
+      "grad_norm": 0.17401129007339478,
+      "learning_rate": 9.761226629790777e-06,
+      "loss": 0.0757,
+      "step": 602
+    },
+    {
+      "epoch": 0.12638859777824354,
+      "grad_norm": 0.19704696536064148,
+      "learning_rate": 9.760188969853953e-06,
+      "loss": 0.0786,
+      "step": 603
+    },
+    {
+      "epoch": 0.1265981974428841,
+      "grad_norm": 0.21989691257476807,
+      "learning_rate": 9.759149115471001e-06,
+      "loss": 0.0737,
+      "step": 604
+    },
+    {
+      "epoch": 0.12680779710752463,
+      "grad_norm": 0.19509382545948029,
+      "learning_rate": 9.758107067121298e-06,
+      "loss": 0.0775,
+      "step": 605
+    },
+    {
+      "epoch": 0.12701739677216517,
+      "grad_norm": 0.16452129185199738,
+      "learning_rate": 9.757062825285223e-06,
+      "loss": 0.0751,
+      "step": 606
+    },
+    {
+      "epoch": 0.1272269964368057,
+      "grad_norm": 0.16775386035442352,
+      "learning_rate": 9.756016390444174e-06,
+      "loss": 0.0785,
+      "step": 607
+    },
+    {
+      "epoch": 0.12743659610144623,
+      "grad_norm": 0.1647682636976242,
+      "learning_rate": 9.754967763080558e-06,
+      "loss": 0.0779,
+      "step": 608
+    },
+    {
+      "epoch": 0.12764619576608677,
+      "grad_norm": 0.15964777767658234,
+      "learning_rate": 9.75391694367779e-06,
+      "loss": 0.0757,
+      "step": 609
+    },
+    {
+      "epoch": 0.1278557954307273,
+      "grad_norm": 0.1809694468975067,
+      "learning_rate": 9.7528639327203e-06,
+      "loss": 0.0745,
+      "step": 610
+    },
+    {
+      "epoch": 0.12806539509536785,
+      "grad_norm": 0.21518062055110931,
+      "learning_rate": 9.751808730693521e-06,
+      "loss": 0.0721,
+      "step": 611
+    },
+    {
+      "epoch": 0.1282749947600084,
+      "grad_norm": 0.23137721419334412,
+      "learning_rate": 9.750751338083906e-06,
+      "loss": 0.0773,
+      "step": 612
+    },
+    {
+      "epoch": 0.1284845944246489,
+      "grad_norm": 0.25104039907455444,
+      "learning_rate": 9.749691755378912e-06,
+      "loss": 0.0789,
+      "step": 613
+    },
+    {
+      "epoch": 0.12869419408928945,
+      "grad_norm": 0.3227115869522095,
+      "learning_rate": 9.748629983067004e-06,
+      "loss": 0.0814,
+      "step": 614
+    },
+    {
+      "epoch": 0.12890379375393,
+      "grad_norm": 0.3414157032966614,
+      "learning_rate": 9.747566021637662e-06,
+      "loss": 0.0784,
+      "step": 615
+    },
+    {
+      "epoch": 0.12911339341857053,
+      "grad_norm": 0.2576848566532135,
+      "learning_rate": 9.746499871581368e-06,
+      "loss": 0.0821,
+      "step": 616
+    },
+    {
+      "epoch": 0.12932299308321107,
+      "grad_norm": 0.18533697724342346,
+      "learning_rate": 9.74543153338962e-06,
+      "loss": 0.0791,
+      "step": 617
+    },
+    {
+      "epoch": 0.12953259274785162,
+      "grad_norm": 0.21570216119289398,
+      "learning_rate": 9.744361007554922e-06,
+      "loss": 0.0762,
+      "step": 618
+    },
+    {
+      "epoch": 0.12974219241249213,
+      "grad_norm": 0.2508070170879364,
+      "learning_rate": 9.743288294570784e-06,
+      "loss": 0.0747,
+      "step": 619
+    },
+    {
+      "epoch": 0.12995179207713267,
+      "grad_norm": 0.23686811327934265,
+      "learning_rate": 9.742213394931726e-06,
+      "loss": 0.0766,
+      "step": 620
+    },
+    {
+      "epoch": 0.1301613917417732,
+      "grad_norm": 0.18374064564704895,
+      "learning_rate": 9.741136309133279e-06,
+      "loss": 0.0751,
+      "step": 621
+    },
+    {
+      "epoch": 0.13037099140641376,
+      "grad_norm": 0.21081578731536865,
+      "learning_rate": 9.74005703767198e-06,
+      "loss": 0.0769,
+      "step": 622
+    },
+    {
+      "epoch": 0.1305805910710543,
+      "grad_norm": 0.20336848497390747,
+      "learning_rate": 9.738975581045368e-06,
+      "loss": 0.0788,
+      "step": 623
+    },
+    {
+      "epoch": 0.1307901907356948,
+      "grad_norm": 0.1740516871213913,
+      "learning_rate": 9.737891939752e-06,
+      "loss": 0.0744,
+      "step": 624
+    },
+    {
+      "epoch": 0.13099979040033535,
+      "grad_norm": 0.19728244841098785,
+      "learning_rate": 9.73680611429143e-06,
+      "loss": 0.0737,
+      "step": 625
+    },
+    {
+      "epoch": 0.1312093900649759,
+      "grad_norm": 0.19369123876094818,
+      "learning_rate": 9.735718105164228e-06,
+      "loss": 0.0759,
+      "step": 626
+    },
+    {
+      "epoch": 0.13141898972961644,
+      "grad_norm": 0.1801711618900299,
+      "learning_rate": 9.734627912871962e-06,
+      "loss": 0.0759,
+      "step": 627
+    },
+    {
+      "epoch": 0.13162858939425698,
+      "grad_norm": 0.1812150627374649,
+      "learning_rate": 9.733535537917211e-06,
+      "loss": 0.078,
+      "step": 628
+    },
+    {
+      "epoch": 0.1318381890588975,
+      "grad_norm": 0.17281141877174377,
+      "learning_rate": 9.732440980803561e-06,
+      "loss": 0.0759,
+      "step": 629
+    },
+    {
+      "epoch": 0.13204778872353803,
+      "grad_norm": 0.18802779912948608,
+      "learning_rate": 9.7313442420356e-06,
+      "loss": 0.075,
+      "step": 630
+    },
+    {
+      "epoch": 0.13225738838817858,
+      "grad_norm": 0.17967858910560608,
+      "learning_rate": 9.730245322118929e-06,
+      "loss": 0.0737,
+      "step": 631
+    },
+    {
+      "epoch": 0.13246698805281912,
+      "grad_norm": 0.15727940201759338,
+      "learning_rate": 9.729144221560145e-06,
+      "loss": 0.0728,
+      "step": 632
+    },
+    {
+      "epoch": 0.13267658771745966,
+      "grad_norm": 0.1466912478208542,
+      "learning_rate": 9.72804094086686e-06,
+      "loss": 0.0765,
+      "step": 633
+    },
+    {
+      "epoch": 0.1328861873821002,
+      "grad_norm": 0.14428047835826874,
+      "learning_rate": 9.726935480547679e-06,
+      "loss": 0.0759,
+      "step": 634
+    },
+    {
+      "epoch": 0.13309578704674072,
+      "grad_norm": 0.14207249879837036,
+      "learning_rate": 9.725827841112226e-06,
+      "loss": 0.0763,
+      "step": 635
+    },
+    {
+      "epoch": 0.13330538671138126,
+      "grad_norm": 0.1447642594575882,
+      "learning_rate": 9.72471802307112e-06,
+      "loss": 0.0738,
+      "step": 636
+    },
+    {
+      "epoch": 0.1335149863760218,
+      "grad_norm": 0.15724752843379974,
+      "learning_rate": 9.723606026935986e-06,
+      "loss": 0.0746,
+      "step": 637
+    },
+    {
+      "epoch": 0.13372458604066234,
+      "grad_norm": 0.13662230968475342,
+      "learning_rate": 9.722491853219455e-06,
+      "loss": 0.0735,
+      "step": 638
+    },
+    {
+      "epoch": 0.13393418570530288,
+      "grad_norm": 0.14362376928329468,
+      "learning_rate": 9.72137550243516e-06,
+      "loss": 0.0738,
+      "step": 639
+    },
+    {
+      "epoch": 0.1341437853699434,
+      "grad_norm": 0.18275748193264008,
+      "learning_rate": 9.720256975097741e-06,
+      "loss": 0.0806,
+      "step": 640
+    },
+    {
+      "epoch": 0.13435338503458394,
+      "grad_norm": 0.17419646680355072,
+      "learning_rate": 9.719136271722835e-06,
+      "loss": 0.076,
+      "step": 641
+    },
+    {
+      "epoch": 0.13456298469922448,
+      "grad_norm": 0.15311293303966522,
+      "learning_rate": 9.718013392827087e-06,
+      "loss": 0.0743,
+      "step": 642
+    },
+    {
+      "epoch": 0.13477258436386502,
+      "grad_norm": 0.18459968268871307,
+      "learning_rate": 9.716888338928147e-06,
+      "loss": 0.0764,
+      "step": 643
+    },
+    {
+      "epoch": 0.13498218402850556,
+      "grad_norm": 0.2108069807291031,
+      "learning_rate": 9.715761110544663e-06,
+      "loss": 0.0752,
+      "step": 644
+    },
+    {
+      "epoch": 0.13519178369314608,
+      "grad_norm": 0.19579482078552246,
+      "learning_rate": 9.714631708196287e-06,
+      "loss": 0.0788,
+      "step": 645
+    },
+    {
+      "epoch": 0.13540138335778662,
+      "grad_norm": 0.18873129785060883,
+      "learning_rate": 9.71350013240367e-06,
+      "loss": 0.0733,
+      "step": 646
+    },
+    {
+      "epoch": 0.13561098302242716,
+      "grad_norm": 0.21143223345279694,
+      "learning_rate": 9.712366383688474e-06,
+      "loss": 0.0746,
+      "step": 647
+    },
+    {
+      "epoch": 0.1358205826870677,
+      "grad_norm": 0.23884397745132446,
+      "learning_rate": 9.71123046257335e-06,
+      "loss": 0.0744,
+      "step": 648
+    },
+    {
+      "epoch": 0.13603018235170825,
+      "grad_norm": 0.22789151966571808,
+      "learning_rate": 9.710092369581966e-06,
+      "loss": 0.0766,
+      "step": 649
+    },
+    {
+      "epoch": 0.1362397820163488,
+      "grad_norm": 0.18522526323795319,
+      "learning_rate": 9.708952105238975e-06,
+      "loss": 0.0743,
+      "step": 650
+    },
+    {
+      "epoch": 0.1364493816809893,
+      "grad_norm": 0.20992542803287506,
+      "learning_rate": 9.707809670070043e-06,
+      "loss": 0.073,
+      "step": 651
+    },
+    {
+      "epoch": 0.13665898134562984,
+      "grad_norm": 0.1938726305961609,
+      "learning_rate": 9.706665064601831e-06,
+      "loss": 0.0729,
+      "step": 652
+    },
+    {
+      "epoch": 0.13686858101027038,
+      "grad_norm": 0.12811291217803955,
+      "learning_rate": 9.705518289362001e-06,
+      "loss": 0.0764,
+      "step": 653
+    },
+    {
+      "epoch": 0.13707818067491093,
+      "grad_norm": 0.19632869958877563,
+      "learning_rate": 9.704369344879219e-06,
+      "loss": 0.0765,
+      "step": 654
+    },
+    {
+      "epoch": 0.13728778033955147,
+      "grad_norm": 0.21263957023620605,
+      "learning_rate": 9.703218231683143e-06,
+      "loss": 0.0765,
+      "step": 655
+    },
+    {
+      "epoch": 0.13749738000419198,
+      "grad_norm": 0.18144431710243225,
+      "learning_rate": 9.702064950304442e-06,
+      "loss": 0.0739,
+      "step": 656
+    },
+    {
+      "epoch": 0.13770697966883252,
+      "grad_norm": 0.21412257850170135,
+      "learning_rate": 9.700909501274773e-06,
+      "loss": 0.0771,
+      "step": 657
+    },
+    {
+      "epoch": 0.13791657933347307,
+      "grad_norm": 0.2056640088558197,
+      "learning_rate": 9.699751885126803e-06,
+      "loss": 0.0746,
+      "step": 658
+    },
+    {
+      "epoch": 0.1381261789981136,
+      "grad_norm": 0.17441654205322266,
+      "learning_rate": 9.698592102394188e-06,
+      "loss": 0.0806,
+      "step": 659
+    },
+    {
+      "epoch": 0.13833577866275415,
+      "grad_norm": 0.18476596474647522,
+      "learning_rate": 9.697430153611592e-06,
+      "loss": 0.0764,
+      "step": 660
+    },
+    {
+      "epoch": 0.13854537832739466,
+      "grad_norm": 0.15858349204063416,
+      "learning_rate": 9.69626603931467e-06,
+      "loss": 0.0748,
+      "step": 661
+    },
+    {
+      "epoch": 0.1387549779920352,
+      "grad_norm": 0.14706814289093018,
+      "learning_rate": 9.695099760040079e-06,
+      "loss": 0.0737,
+      "step": 662
+    },
+    {
+      "epoch": 0.13896457765667575,
+      "grad_norm": 0.18588264286518097,
+      "learning_rate": 9.693931316325473e-06,
+      "loss": 0.0706,
+      "step": 663
+    },
+    {
+      "epoch": 0.1391741773213163,
+      "grad_norm": 0.18025395274162292,
+      "learning_rate": 9.692760708709506e-06,
+      "loss": 0.0763,
+      "step": 664
+    },
+    {
+      "epoch": 0.13938377698595683,
+      "grad_norm": 0.18428079783916473,
+      "learning_rate": 9.691587937731827e-06,
+      "loss": 0.0772,
+      "step": 665
+    },
+    {
+      "epoch": 0.13959337665059737,
+      "grad_norm": 0.20305953919887543,
+      "learning_rate": 9.690413003933084e-06,
+      "loss": 0.0742,
+      "step": 666
+    },
+    {
+      "epoch": 0.1398029763152379,
+      "grad_norm": 0.21288353204727173,
+      "learning_rate": 9.68923590785492e-06,
+      "loss": 0.077,
+      "step": 667
+    },
+    {
+      "epoch": 0.14001257597987843,
+      "grad_norm": 0.2299814224243164,
+      "learning_rate": 9.688056650039976e-06,
+      "loss": 0.0754,
+      "step": 668
+    },
+    {
+      "epoch": 0.14022217564451897,
+      "grad_norm": 0.23318803310394287,
+      "learning_rate": 9.68687523103189e-06,
+      "loss": 0.0748,
+      "step": 669
+    },
+    {
+      "epoch": 0.1404317753091595,
+      "grad_norm": 0.21237865090370178,
+      "learning_rate": 9.685691651375297e-06,
+      "loss": 0.0719,
+      "step": 670
+    },
+    {
+      "epoch": 0.14064137497380005,
+      "grad_norm": 0.20599277317523956,
+      "learning_rate": 9.684505911615825e-06,
+      "loss": 0.0731,
+      "step": 671
+    },
+    {
+      "epoch": 0.14085097463844057,
+      "grad_norm": 0.2014642357826233,
+      "learning_rate": 9.683318012300103e-06,
+      "loss": 0.0739,
+      "step": 672
+    },
+    {
+      "epoch": 0.1410605743030811,
+      "grad_norm": 0.21525225043296814,
+      "learning_rate": 9.682127953975748e-06,
+      "loss": 0.0796,
+      "step": 673
+    },
+    {
+      "epoch": 0.14127017396772165,
+      "grad_norm": 0.20060989260673523,
+      "learning_rate": 9.68093573719138e-06,
+      "loss": 0.0717,
+      "step": 674
+    },
+    {
+      "epoch": 0.1414797736323622,
+      "grad_norm": 0.19899265468120575,
+      "learning_rate": 9.679741362496608e-06,
+      "loss": 0.0723,
+      "step": 675
+    },
+    {
+      "epoch": 0.14168937329700274,
+      "grad_norm": 0.1738089621067047,
+      "learning_rate": 9.678544830442041e-06,
+      "loss": 0.0776,
+      "step": 676
+    },
+    {
+      "epoch": 0.14189897296164325,
+      "grad_norm": 0.1539677530527115,
+      "learning_rate": 9.677346141579277e-06,
+      "loss": 0.0725,
+      "step": 677
+    },
+    {
+      "epoch": 0.1421085726262838,
+      "grad_norm": 0.16590017080307007,
+      "learning_rate": 9.676145296460917e-06,
+      "loss": 0.0733,
+      "step": 678
+    },
+    {
+      "epoch": 0.14231817229092433,
+      "grad_norm": 0.17394165694713593,
+      "learning_rate": 9.674942295640544e-06,
+      "loss": 0.0744,
+      "step": 679
+    },
+    {
+      "epoch": 0.14252777195556487,
+      "grad_norm": 0.17285530269145966,
+      "learning_rate": 9.673737139672746e-06,
+      "loss": 0.0729,
+      "step": 680
+    },
+    {
+      "epoch": 0.14273737162020542,
+      "grad_norm": 0.14739976823329926,
+      "learning_rate": 9.672529829113095e-06,
+      "loss": 0.073,
+      "step": 681
+    },
+    {
+      "epoch": 0.14294697128484593,
+      "grad_norm": 0.1775324046611786,
+      "learning_rate": 9.671320364518164e-06,
+      "loss": 0.0745,
+      "step": 682
+    },
+    {
+      "epoch": 0.14315657094948647,
+      "grad_norm": 0.1740763634443283,
+      "learning_rate": 9.670108746445514e-06,
+      "loss": 0.0734,
+      "step": 683
+    },
+    {
+      "epoch": 0.14336617061412701,
+      "grad_norm": 0.14297354221343994,
+      "learning_rate": 9.668894975453705e-06,
+      "loss": 0.0763,
+      "step": 684
+    },
+    {
+      "epoch": 0.14357577027876756,
+      "grad_norm": 0.1698678582906723,
+      "learning_rate": 9.66767905210228e-06,
+      "loss": 0.0727,
+      "step": 685
+    },
+    {
+      "epoch": 0.1437853699434081,
+      "grad_norm": 0.18393908441066742,
+      "learning_rate": 9.666460976951783e-06,
+      "loss": 0.0755,
+      "step": 686
+    },
+    {
+      "epoch": 0.14399496960804864,
+      "grad_norm": 0.16469605267047882,
+      "learning_rate": 9.665240750563743e-06,
+      "loss": 0.0718,
+      "step": 687
+    },
+    {
+      "epoch": 0.14420456927268915,
+      "grad_norm": 0.15286237001419067,
+      "learning_rate": 9.664018373500688e-06,
+      "loss": 0.0774,
+      "step": 688
+    },
+    {
+      "epoch": 0.1444141689373297,
+      "grad_norm": 0.15138964354991913,
+      "learning_rate": 9.662793846326131e-06,
+      "loss": 0.0738,
+      "step": 689
+    },
+    {
+      "epoch": 0.14462376860197024,
+      "grad_norm": 0.13426071405410767,
+      "learning_rate": 9.661567169604579e-06,
+      "loss": 0.0732,
+      "step": 690
+    },
+    {
+      "epoch": 0.14483336826661078,
+      "grad_norm": 0.14377839863300323,
+      "learning_rate": 9.66033834390153e-06,
+      "loss": 0.0743,
+      "step": 691
+    },
+    {
+      "epoch": 0.14504296793125132,
+      "grad_norm": 0.14970825612545013,
+      "learning_rate": 9.659107369783473e-06,
+      "loss": 0.0719,
+      "step": 692
+    },
+    {
+      "epoch": 0.14525256759589183,
+      "grad_norm": 0.12804554402828217,
+      "learning_rate": 9.657874247817886e-06,
+      "loss": 0.0715,
+      "step": 693
+    },
+    {
+      "epoch": 0.14546216726053238,
+      "grad_norm": 0.13127097487449646,
+      "learning_rate": 9.656638978573238e-06,
+      "loss": 0.0761,
+      "step": 694
+    },
+    {
+      "epoch": 0.14567176692517292,
+      "grad_norm": 0.16138702630996704,
+      "learning_rate": 9.655401562618989e-06,
+      "loss": 0.0727,
+      "step": 695
+    },
+    {
+      "epoch": 0.14588136658981346,
+      "grad_norm": 0.16640619933605194,
+      "learning_rate": 9.654162000525585e-06,
+      "loss": 0.0704,
+      "step": 696
+    },
+    {
+      "epoch": 0.146090966254454,
+      "grad_norm": 0.17615856230258942,
+      "learning_rate": 9.652920292864467e-06,
+      "loss": 0.0763,
+      "step": 697
+    },
+    {
+      "epoch": 0.14630056591909452,
+      "grad_norm": 0.20064422488212585,
+      "learning_rate": 9.65167644020806e-06,
+      "loss": 0.0709,
+      "step": 698
+    },
+    {
+      "epoch": 0.14651016558373506,
+      "grad_norm": 0.2409866899251938,
+      "learning_rate": 9.650430443129781e-06,
+      "loss": 0.0752,
+      "step": 699
+    },
+    {
+      "epoch": 0.1467197652483756,
+      "grad_norm": 0.28271573781967163,
+      "learning_rate": 9.649182302204034e-06,
+      "loss": 0.0732,
+      "step": 700
+    },
+    {
+      "epoch": 0.14692936491301614,
+      "grad_norm": 0.29788634181022644,
+      "learning_rate": 9.64793201800621e-06,
+      "loss": 0.0734,
+      "step": 701
+    },
+    {
+      "epoch": 0.14713896457765668,
+      "grad_norm": 0.25497421622276306,
+      "learning_rate": 9.64667959111269e-06,
+      "loss": 0.0739,
+      "step": 702
+    },
+    {
+      "epoch": 0.14734856424229723,
+      "grad_norm": 0.1792333573102951,
+      "learning_rate": 9.645425022100847e-06,
+      "loss": 0.0752,
+      "step": 703
+    },
+    {
+      "epoch": 0.14755816390693774,
+      "grad_norm": 0.16406488418579102,
+      "learning_rate": 9.644168311549032e-06,
+      "loss": 0.0736,
+      "step": 704
+    },
+    {
+      "epoch": 0.14776776357157828,
+      "grad_norm": 0.20219507813453674,
+      "learning_rate": 9.64290946003659e-06,
+      "loss": 0.0803,
+      "step": 705
+    },
+    {
+      "epoch": 0.14797736323621882,
+      "grad_norm": 0.20861254632472992,
+      "learning_rate": 9.641648468143852e-06,
+      "loss": 0.0759,
+      "step": 706
+    },
+    {
+      "epoch": 0.14818696290085936,
+      "grad_norm": 0.17586849629878998,
+      "learning_rate": 9.640385336452135e-06,
+      "loss": 0.0752,
+      "step": 707
+    },
+    {
+      "epoch": 0.1483965625654999,
+      "grad_norm": 0.16276384890079498,
+      "learning_rate": 9.639120065543738e-06,
+      "loss": 0.0714,
+      "step": 708
+    },
+    {
+      "epoch": 0.14860616223014042,
+      "grad_norm": 0.20247167348861694,
+      "learning_rate": 9.637852656001957e-06,
+      "loss": 0.0752,
+      "step": 709
+    },
+    {
+      "epoch": 0.14881576189478096,
+      "grad_norm": 0.18190069496631622,
+      "learning_rate": 9.636583108411066e-06,
+      "loss": 0.0753,
+      "step": 710
+    },
+    {
+      "epoch": 0.1490253615594215,
+      "grad_norm": 0.1566484421491623,
+      "learning_rate": 9.635311423356324e-06,
+      "loss": 0.072,
+      "step": 711
+    },
+    {
+      "epoch": 0.14923496122406205,
+      "grad_norm": 0.1933658868074417,
+      "learning_rate": 9.63403760142398e-06,
+      "loss": 0.0746,
+      "step": 712
+    },
+    {
+      "epoch": 0.1494445608887026,
+      "grad_norm": 0.1682639867067337,
+      "learning_rate": 9.632761643201262e-06,
+      "loss": 0.073,
+      "step": 713
+    },
+    {
+      "epoch": 0.1496541605533431,
+      "grad_norm": 0.13671089708805084,
+      "learning_rate": 9.63148354927639e-06,
+      "loss": 0.0719,
+      "step": 714
+    },
+    {
+      "epoch": 0.14986376021798364,
+      "grad_norm": 0.15228819847106934,
+      "learning_rate": 9.630203320238564e-06,
+      "loss": 0.0724,
+      "step": 715
+    },
+    {
+      "epoch": 0.15007335988262419,
+      "grad_norm": 0.15035517513751984,
+      "learning_rate": 9.628920956677969e-06,
+      "loss": 0.0737,
+      "step": 716
+    },
+    {
+      "epoch": 0.15028295954726473,
+      "grad_norm": 0.1775098294019699,
+      "learning_rate": 9.627636459185774e-06,
+      "loss": 0.0695,
+      "step": 717
+    },
+    {
+      "epoch": 0.15049255921190527,
+      "grad_norm": 0.20350755751132965,
+      "learning_rate": 9.626349828354133e-06,
+      "loss": 0.0755,
+      "step": 718
+    },
+    {
+      "epoch": 0.1507021588765458,
+      "grad_norm": 0.18517783284187317,
+      "learning_rate": 9.625061064776183e-06,
+      "loss": 0.0733,
+      "step": 719
+    },
+    {
+      "epoch": 0.15091175854118632,
+      "grad_norm": 0.19070328772068024,
+      "learning_rate": 9.623770169046042e-06,
+      "loss": 0.0725,
+      "step": 720
+    },
+    {
+      "epoch": 0.15112135820582687,
+      "grad_norm": 0.19574813544750214,
+      "learning_rate": 9.622477141758813e-06,
+      "loss": 0.0717,
+      "step": 721
+    },
+    {
+      "epoch": 0.1513309578704674,
+      "grad_norm": 0.18914958834648132,
+      "learning_rate": 9.621181983510582e-06,
+      "loss": 0.0761,
+      "step": 722
+    },
+    {
+      "epoch": 0.15154055753510795,
+      "grad_norm": 0.18750722706317902,
+      "learning_rate": 9.619884694898417e-06,
+      "loss": 0.0726,
+      "step": 723
+    },
+    {
+      "epoch": 0.1517501571997485,
+      "grad_norm": 0.23439928889274597,
+      "learning_rate": 9.618585276520367e-06,
+      "loss": 0.0723,
+      "step": 724
+    },
+    {
+      "epoch": 0.151959756864389,
+      "grad_norm": 0.27055707573890686,
+      "learning_rate": 9.617283728975464e-06,
+      "loss": 0.0748,
+      "step": 725
+    },
+    {
+      "epoch": 0.15216935652902955,
+      "grad_norm": 0.20538926124572754,
+      "learning_rate": 9.61598005286372e-06,
+      "loss": 0.0721,
+      "step": 726
+    },
+    {
+      "epoch": 0.1523789561936701,
+      "grad_norm": 0.15176212787628174,
+      "learning_rate": 9.614674248786131e-06,
+      "loss": 0.0724,
+      "step": 727
+    },
+    {
+      "epoch": 0.15258855585831063,
+      "grad_norm": 0.18770189583301544,
+      "learning_rate": 9.613366317344674e-06,
+      "loss": 0.0714,
+      "step": 728
+    },
+    {
+      "epoch": 0.15279815552295117,
+      "grad_norm": 0.14766095578670502,
+      "learning_rate": 9.6120562591423e-06,
+      "loss": 0.0722,
+      "step": 729
+    },
+    {
+      "epoch": 0.1530077551875917,
+      "grad_norm": 0.1400596648454666,
+      "learning_rate": 9.610744074782951e-06,
+      "loss": 0.0731,
+      "step": 730
+    },
+    {
+      "epoch": 0.15321735485223223,
+      "grad_norm": 0.16470056772232056,
+      "learning_rate": 9.60942976487154e-06,
+      "loss": 0.0709,
+      "step": 731
+    },
+    {
+      "epoch": 0.15342695451687277,
+      "grad_norm": 0.1465228945016861,
+      "learning_rate": 9.608113330013964e-06,
+      "loss": 0.0711,
+      "step": 732
+    },
+    {
+      "epoch": 0.1536365541815133,
+      "grad_norm": 0.14212250709533691,
+      "learning_rate": 9.606794770817102e-06,
+      "loss": 0.0772,
+      "step": 733
+    },
+    {
+      "epoch": 0.15384615384615385,
+      "grad_norm": 0.12148991227149963,
+      "learning_rate": 9.605474087888806e-06,
+      "loss": 0.0738,
+      "step": 734
+    },
+    {
+      "epoch": 0.1540557535107944,
+      "grad_norm": 0.1582231968641281,
+      "learning_rate": 9.604151281837912e-06,
+      "loss": 0.0723,
+      "step": 735
+    },
+    {
+      "epoch": 0.1542653531754349,
+      "grad_norm": 0.1933947652578354,
+      "learning_rate": 9.602826353274235e-06,
+      "loss": 0.073,
+      "step": 736
+    },
+    {
+      "epoch": 0.15447495284007545,
+      "grad_norm": 0.16174952685832977,
+      "learning_rate": 9.601499302808565e-06,
+      "loss": 0.0688,
+      "step": 737
+    },
+    {
+      "epoch": 0.154684552504716,
+      "grad_norm": 0.14676834642887115,
+      "learning_rate": 9.600170131052671e-06,
+      "loss": 0.0712,
+      "step": 738
+    },
+    {
+      "epoch": 0.15489415216935654,
+      "grad_norm": 0.1428784430027008,
+      "learning_rate": 9.598838838619302e-06,
+      "loss": 0.0703,
+      "step": 739
+    },
+    {
+      "epoch": 0.15510375183399708,
+      "grad_norm": 0.1900653839111328,
+      "learning_rate": 9.597505426122184e-06,
+      "loss": 0.0687,
+      "step": 740
+    },
+    {
+      "epoch": 0.1553133514986376,
+      "grad_norm": 0.2115175575017929,
+      "learning_rate": 9.596169894176021e-06,
+      "loss": 0.0719,
+      "step": 741
+    },
+    {
+      "epoch": 0.15552295116327813,
+      "grad_norm": 0.16086673736572266,
+      "learning_rate": 9.59483224339649e-06,
+      "loss": 0.0721,
+      "step": 742
+    },
+    {
+      "epoch": 0.15573255082791868,
+      "grad_norm": 0.16307683289051056,
+      "learning_rate": 9.59349247440025e-06,
+      "loss": 0.0734,
+      "step": 743
+    },
+    {
+      "epoch": 0.15594215049255922,
+      "grad_norm": 0.17130297422409058,
+      "learning_rate": 9.592150587804934e-06,
+      "loss": 0.0731,
+      "step": 744
+    },
+    {
+      "epoch": 0.15615175015719976,
+      "grad_norm": 0.1763344258069992,
+      "learning_rate": 9.590806584229149e-06,
+      "loss": 0.0758,
+      "step": 745
+    },
+    {
+      "epoch": 0.15636134982184027,
+      "grad_norm": 0.20289908349514008,
+      "learning_rate": 9.589460464292483e-06,
+      "loss": 0.0694,
+      "step": 746
+    },
+    {
+      "epoch": 0.15657094948648081,
+      "grad_norm": 0.2093002051115036,
+      "learning_rate": 9.588112228615495e-06,
+      "loss": 0.0719,
+      "step": 747
+    },
+    {
+      "epoch": 0.15678054915112136,
+      "grad_norm": 0.19119895994663239,
+      "learning_rate": 9.586761877819726e-06,
+      "loss": 0.0685,
+      "step": 748
+    },
+    {
+      "epoch": 0.1569901488157619,
+      "grad_norm": 0.17849349975585938,
+      "learning_rate": 9.585409412527682e-06,
+      "loss": 0.0719,
+      "step": 749
+    },
+    {
+      "epoch": 0.15719974848040244,
+      "grad_norm": 0.21115368604660034,
+      "learning_rate": 9.584054833362851e-06,
+      "loss": 0.0738,
+      "step": 750
+    },
+    {
+      "epoch": 0.15740934814504295,
+      "grad_norm": 0.19666269421577454,
+      "learning_rate": 9.582698140949696e-06,
+      "loss": 0.0716,
+      "step": 751
+    },
+    {
+      "epoch": 0.1576189478096835,
+      "grad_norm": 0.1261245310306549,
+      "learning_rate": 9.581339335913647e-06,
+      "loss": 0.0695,
+      "step": 752
+    },
+    {
+      "epoch": 0.15782854747432404,
+      "grad_norm": 0.16366009414196014,
+      "learning_rate": 9.579978418881118e-06,
+      "loss": 0.0715,
+      "step": 753
+    },
+    {
+      "epoch": 0.15803814713896458,
+      "grad_norm": 0.21418072283267975,
+      "learning_rate": 9.57861539047949e-06,
+      "loss": 0.0721,
+      "step": 754
+    },
+    {
+      "epoch": 0.15824774680360512,
+      "grad_norm": 0.16937729716300964,
+      "learning_rate": 9.577250251337114e-06,
+      "loss": 0.0725,
+      "step": 755
+    },
+    {
+      "epoch": 0.15845734646824566,
+      "grad_norm": 0.1247527152299881,
+      "learning_rate": 9.575883002083326e-06,
+      "loss": 0.0726,
+      "step": 756
+    },
+    {
+      "epoch": 0.15866694613288618,
+      "grad_norm": 0.1638677716255188,
+      "learning_rate": 9.574513643348424e-06,
+      "loss": 0.0717,
+      "step": 757
+    },
+    {
+      "epoch": 0.15887654579752672,
+      "grad_norm": 0.19583426415920258,
+      "learning_rate": 9.573142175763683e-06,
+      "loss": 0.0752,
+      "step": 758
+    },
+    {
+      "epoch": 0.15908614546216726,
+      "grad_norm": 0.19242113828659058,
+      "learning_rate": 9.571768599961349e-06,
+      "loss": 0.0718,
+      "step": 759
+    },
+    {
+      "epoch": 0.1592957451268078,
+      "grad_norm": 0.18468959629535675,
+      "learning_rate": 9.570392916574639e-06,
+      "loss": 0.0712,
+      "step": 760
+    },
+    {
+      "epoch": 0.15950534479144834,
+      "grad_norm": 0.18113423883914948,
+      "learning_rate": 9.569015126237744e-06,
+      "loss": 0.0728,
+      "step": 761
+    },
+    {
+      "epoch": 0.15971494445608886,
+      "grad_norm": 0.15867562592029572,
+      "learning_rate": 9.567635229585826e-06,
+      "loss": 0.0714,
+      "step": 762
+    },
+    {
+      "epoch": 0.1599245441207294,
+      "grad_norm": 0.14228221774101257,
+      "learning_rate": 9.566253227255015e-06,
+      "loss": 0.0751,
+      "step": 763
+    },
+    {
+      "epoch": 0.16013414378536994,
+      "grad_norm": 0.14296479523181915,
+      "learning_rate": 9.564869119882414e-06,
+      "loss": 0.0713,
+      "step": 764
+    },
+    {
+      "epoch": 0.16034374345001048,
+      "grad_norm": 0.15336044132709503,
+      "learning_rate": 9.563482908106098e-06,
+      "loss": 0.0734,
+      "step": 765
+    },
+    {
+      "epoch": 0.16055334311465103,
+      "grad_norm": 0.16982528567314148,
+      "learning_rate": 9.56209459256511e-06,
+      "loss": 0.0748,
+      "step": 766
+    },
+    {
+      "epoch": 0.16076294277929154,
+      "grad_norm": 0.20288409292697906,
+      "learning_rate": 9.560704173899461e-06,
+      "loss": 0.0717,
+      "step": 767
+    },
+    {
+      "epoch": 0.16097254244393208,
+      "grad_norm": 0.24066397547721863,
+      "learning_rate": 9.559311652750135e-06,
+      "loss": 0.0727,
+      "step": 768
+    },
+    {
+      "epoch": 0.16118214210857262,
+      "grad_norm": 0.24527938663959503,
+      "learning_rate": 9.557917029759087e-06,
+      "loss": 0.0755,
+      "step": 769
+    },
+    {
+      "epoch": 0.16139174177321317,
+      "grad_norm": 0.18823891878128052,
+      "learning_rate": 9.556520305569232e-06,
+      "loss": 0.0698,
+      "step": 770
+    },
+    {
+      "epoch": 0.1616013414378537,
+      "grad_norm": 0.1405959278345108,
+      "learning_rate": 9.555121480824463e-06,
+      "loss": 0.0722,
+      "step": 771
+    },
+    {
+      "epoch": 0.16181094110249425,
+      "grad_norm": 0.1578516960144043,
+      "learning_rate": 9.553720556169639e-06,
+      "loss": 0.0736,
+      "step": 772
+    },
+    {
+      "epoch": 0.16202054076713476,
+      "grad_norm": 0.15858958661556244,
+      "learning_rate": 9.552317532250584e-06,
+      "loss": 0.0727,
+      "step": 773
+    },
+    {
+      "epoch": 0.1622301404317753,
+      "grad_norm": 0.1704108715057373,
+      "learning_rate": 9.55091240971409e-06,
+      "loss": 0.0697,
+      "step": 774
+    },
+    {
+      "epoch": 0.16243974009641585,
+      "grad_norm": 0.18552088737487793,
+      "learning_rate": 9.549505189207924e-06,
+      "loss": 0.0715,
+      "step": 775
+    },
+    {
+      "epoch": 0.1626493397610564,
+      "grad_norm": 0.14782559871673584,
+      "learning_rate": 9.548095871380808e-06,
+      "loss": 0.0746,
+      "step": 776
+    },
+    {
+      "epoch": 0.16285893942569693,
+      "grad_norm": 0.13771621882915497,
+      "learning_rate": 9.54668445688244e-06,
+      "loss": 0.0711,
+      "step": 777
+    },
+    {
+      "epoch": 0.16306853909033744,
+      "grad_norm": 0.15888865292072296,
+      "learning_rate": 9.545270946363484e-06,
+      "loss": 0.0713,
+      "step": 778
+    },
+    {
+      "epoch": 0.163278138754978,
+      "grad_norm": 0.13176760077476501,
+      "learning_rate": 9.543855340475565e-06,
+      "loss": 0.0713,
+      "step": 779
+    },
+    {
+      "epoch": 0.16348773841961853,
+      "grad_norm": 0.1236625537276268,
+      "learning_rate": 9.542437639871279e-06,
+      "loss": 0.0736,
+      "step": 780
+    },
+    {
+      "epoch": 0.16369733808425907,
+      "grad_norm": 0.14519374072551727,
+      "learning_rate": 9.541017845204182e-06,
+      "loss": 0.0725,
+      "step": 781
+    },
+    {
+      "epoch": 0.1639069377488996,
+      "grad_norm": 0.13058701157569885,
+      "learning_rate": 9.539595957128803e-06,
+      "loss": 0.0672,
+      "step": 782
+    },
+    {
+      "epoch": 0.16411653741354013,
+      "grad_norm": 0.13305912911891937,
+      "learning_rate": 9.53817197630063e-06,
+      "loss": 0.0706,
+      "step": 783
+    },
+    {
+      "epoch": 0.16432613707818067,
+      "grad_norm": 0.15352047979831696,
+      "learning_rate": 9.53674590337612e-06,
+      "loss": 0.0721,
+      "step": 784
+    },
+    {
+      "epoch": 0.1645357367428212,
+      "grad_norm": 0.15192946791648865,
+      "learning_rate": 9.535317739012689e-06,
+      "loss": 0.075,
+      "step": 785
+    },
+    {
+      "epoch": 0.16474533640746175,
+      "grad_norm": 0.1807214766740799,
+      "learning_rate": 9.533887483868723e-06,
+      "loss": 0.0691,
+      "step": 786
+    },
+    {
+      "epoch": 0.1649549360721023,
+      "grad_norm": 0.22403304278850555,
+      "learning_rate": 9.53245513860357e-06,
+      "loss": 0.0723,
+      "step": 787
+    },
+    {
+      "epoch": 0.16516453573674283,
+      "grad_norm": 0.22124335169792175,
+      "learning_rate": 9.531020703877539e-06,
+      "loss": 0.0738,
+      "step": 788
+    },
+    {
+      "epoch": 0.16537413540138335,
+      "grad_norm": 0.197389617562294,
+      "learning_rate": 9.529584180351902e-06,
+      "loss": 0.0728,
+      "step": 789
+    },
+    {
+      "epoch": 0.1655837350660239,
+      "grad_norm": 0.19445890188217163,
+      "learning_rate": 9.528145568688902e-06,
+      "loss": 0.0709,
+      "step": 790
+    },
+    {
+      "epoch": 0.16579333473066443,
+      "grad_norm": 0.1834816336631775,
+      "learning_rate": 9.526704869551736e-06,
+      "loss": 0.0734,
+      "step": 791
+    },
+    {
+      "epoch": 0.16600293439530497,
+      "grad_norm": 0.15776723623275757,
+      "learning_rate": 9.525262083604562e-06,
+      "loss": 0.0735,
+      "step": 792
+    },
+    {
+      "epoch": 0.16621253405994552,
+      "grad_norm": 0.1593591719865799,
+      "learning_rate": 9.523817211512511e-06,
+      "loss": 0.0708,
+      "step": 793
+    },
+    {
+      "epoch": 0.16642213372458603,
+      "grad_norm": 0.19884854555130005,
+      "learning_rate": 9.522370253941664e-06,
+      "loss": 0.0729,
+      "step": 794
+    },
+    {
+      "epoch": 0.16663173338922657,
+      "grad_norm": 0.208924800157547,
+      "learning_rate": 9.520921211559067e-06,
+      "loss": 0.0696,
+      "step": 795
+    },
+    {
+      "epoch": 0.1668413330538671,
+      "grad_norm": 0.1883474439382553,
+      "learning_rate": 9.519470085032733e-06,
+      "loss": 0.0718,
+      "step": 796
+    },
+    {
+      "epoch": 0.16705093271850766,
+      "grad_norm": 0.14821593463420868,
+      "learning_rate": 9.518016875031628e-06,
+      "loss": 0.0723,
+      "step": 797
+    },
+    {
+      "epoch": 0.1672605323831482,
+      "grad_norm": 0.12627242505550385,
+      "learning_rate": 9.516561582225682e-06,
+      "loss": 0.0725,
+      "step": 798
+    },
+    {
+      "epoch": 0.1674701320477887,
+      "grad_norm": 0.18372437357902527,
+      "learning_rate": 9.515104207285785e-06,
+      "loss": 0.0732,
+      "step": 799
+    },
+    {
+      "epoch": 0.16767973171242925,
+      "grad_norm": 0.20521710813045502,
+      "learning_rate": 9.513644750883786e-06,
+      "loss": 0.0708,
+      "step": 800
+    },
+    {
+      "epoch": 0.1678893313770698,
+      "grad_norm": 0.1553443819284439,
+      "learning_rate": 9.512183213692494e-06,
+      "loss": 0.0744,
+      "step": 801
+    },
+    {
+      "epoch": 0.16809893104171034,
+      "grad_norm": 0.1297149360179901,
+      "learning_rate": 9.510719596385678e-06,
+      "loss": 0.0729,
+      "step": 802
+    },
+    {
+      "epoch": 0.16830853070635088,
+      "grad_norm": 0.17294561862945557,
+      "learning_rate": 9.509253899638066e-06,
+      "loss": 0.0716,
+      "step": 803
+    },
+    {
+      "epoch": 0.16851813037099142,
+      "grad_norm": 0.1920863389968872,
+      "learning_rate": 9.507786124125342e-06,
+      "loss": 0.0718,
+      "step": 804
+    },
+    {
+      "epoch": 0.16872773003563193,
+      "grad_norm": 0.17762252688407898,
+      "learning_rate": 9.506316270524152e-06,
+      "loss": 0.0725,
+      "step": 805
+    },
+    {
+      "epoch": 0.16893732970027248,
+      "grad_norm": 0.1815175563097,
+      "learning_rate": 9.504844339512096e-06,
+      "loss": 0.0707,
+      "step": 806
+    },
+    {
+      "epoch": 0.16914692936491302,
+      "grad_norm": 0.18983136117458344,
+      "learning_rate": 9.503370331767736e-06,
+      "loss": 0.0685,
+      "step": 807
+    },
+    {
+      "epoch": 0.16935652902955356,
+      "grad_norm": 0.16265596449375153,
+      "learning_rate": 9.50189424797059e-06,
+      "loss": 0.0699,
+      "step": 808
+    },
+    {
+      "epoch": 0.1695661286941941,
+      "grad_norm": 0.14986425638198853,
+      "learning_rate": 9.500416088801128e-06,
+      "loss": 0.0718,
+      "step": 809
+    },
+    {
+      "epoch": 0.16977572835883462,
+      "grad_norm": 0.16179025173187256,
+      "learning_rate": 9.498935854940785e-06,
+      "loss": 0.0711,
+      "step": 810
+    },
+    {
+      "epoch": 0.16998532802347516,
+      "grad_norm": 0.14040148258209229,
+      "learning_rate": 9.497453547071948e-06,
+      "loss": 0.0735,
+      "step": 811
+    },
+    {
+      "epoch": 0.1701949276881157,
+      "grad_norm": 0.14858156442642212,
+      "learning_rate": 9.495969165877959e-06,
+      "loss": 0.0714,
+      "step": 812
+    },
+    {
+      "epoch": 0.17040452735275624,
+      "grad_norm": 0.1716984659433365,
+      "learning_rate": 9.494482712043119e-06,
+      "loss": 0.0701,
+      "step": 813
+    },
+    {
+      "epoch": 0.17061412701739678,
+      "grad_norm": 0.1606951206922531,
+      "learning_rate": 9.492994186252681e-06,
+      "loss": 0.0727,
+      "step": 814
+    },
+    {
+      "epoch": 0.1708237266820373,
+      "grad_norm": 0.14578759670257568,
+      "learning_rate": 9.491503589192859e-06,
+      "loss": 0.0709,
+      "step": 815
+    },
+    {
+      "epoch": 0.17103332634667784,
+      "grad_norm": 0.16695253551006317,
+      "learning_rate": 9.490010921550814e-06,
+      "loss": 0.0683,
+      "step": 816
+    },
+    {
+      "epoch": 0.17124292601131838,
+      "grad_norm": 0.19223074615001678,
+      "learning_rate": 9.488516184014667e-06,
+      "loss": 0.0721,
+      "step": 817
+    },
+    {
+      "epoch": 0.17145252567595892,
+      "grad_norm": 0.2144256979227066,
+      "learning_rate": 9.48701937727349e-06,
+      "loss": 0.0729,
+      "step": 818
+    },
+    {
+      "epoch": 0.17166212534059946,
+      "grad_norm": 0.2188243865966797,
+      "learning_rate": 9.485520502017314e-06,
+      "loss": 0.0684,
+      "step": 819
+    },
+    {
+      "epoch": 0.17187172500523998,
+      "grad_norm": 0.1678440421819687,
+      "learning_rate": 9.48401955893712e-06,
+      "loss": 0.0722,
+      "step": 820
+    },
+    {
+      "epoch": 0.17208132466988052,
+      "grad_norm": 0.13523709774017334,
+      "learning_rate": 9.482516548724836e-06,
+      "loss": 0.0705,
+      "step": 821
+    },
+    {
+      "epoch": 0.17229092433452106,
+      "grad_norm": 0.1449403315782547,
+      "learning_rate": 9.481011472073359e-06,
+      "loss": 0.0697,
+      "step": 822
+    },
+    {
+      "epoch": 0.1725005239991616,
+      "grad_norm": 0.1349162459373474,
+      "learning_rate": 9.47950432967652e-06,
+      "loss": 0.0749,
+      "step": 823
+    },
+    {
+      "epoch": 0.17271012366380215,
+      "grad_norm": 0.1460842341184616,
+      "learning_rate": 9.477995122229117e-06,
+      "loss": 0.0669,
+      "step": 824
+    },
+    {
+      "epoch": 0.1729197233284427,
+      "grad_norm": 0.16330771148204803,
+      "learning_rate": 9.476483850426895e-06,
+      "loss": 0.0705,
+      "step": 825
+    },
+    {
+      "epoch": 0.1731293229930832,
+      "grad_norm": 0.1619548201560974,
+      "learning_rate": 9.474970514966545e-06,
+      "loss": 0.0695,
+      "step": 826
+    },
+    {
+      "epoch": 0.17333892265772374,
+      "grad_norm": 0.19208967685699463,
+      "learning_rate": 9.473455116545718e-06,
+      "loss": 0.0697,
+      "step": 827
+    },
+    {
+      "epoch": 0.17354852232236428,
+      "grad_norm": 0.20909924805164337,
+      "learning_rate": 9.471937655863011e-06,
+      "loss": 0.0732,
+      "step": 828
+    },
+    {
+      "epoch": 0.17375812198700483,
+      "grad_norm": 0.18117384612560272,
+      "learning_rate": 9.470418133617973e-06,
+      "loss": 0.0719,
+      "step": 829
+    },
+    {
+      "epoch": 0.17396772165164537,
+      "grad_norm": 0.1741568148136139,
+      "learning_rate": 9.468896550511106e-06,
+      "loss": 0.0695,
+      "step": 830
+    },
+    {
+      "epoch": 0.17417732131628588,
+      "grad_norm": 0.18191158771514893,
+      "learning_rate": 9.467372907243858e-06,
+      "loss": 0.0737,
+      "step": 831
+    },
+    {
+      "epoch": 0.17438692098092642,
+      "grad_norm": 0.1610174924135208,
+      "learning_rate": 9.465847204518626e-06,
+      "loss": 0.069,
+      "step": 832
+    },
+    {
+      "epoch": 0.17459652064556697,
+      "grad_norm": 0.1415458768606186,
+      "learning_rate": 9.464319443038759e-06,
+      "loss": 0.071,
+      "step": 833
+    },
+    {
+      "epoch": 0.1748061203102075,
+      "grad_norm": 0.12450738251209259,
+      "learning_rate": 9.462789623508559e-06,
+      "loss": 0.0735,
+      "step": 834
+    },
+    {
+      "epoch": 0.17501571997484805,
+      "grad_norm": 0.1258174180984497,
+      "learning_rate": 9.461257746633267e-06,
+      "loss": 0.0696,
+      "step": 835
+    },
+    {
+      "epoch": 0.17522531963948856,
+      "grad_norm": 0.16757184267044067,
+      "learning_rate": 9.459723813119081e-06,
+      "loss": 0.0707,
+      "step": 836
+    },
+    {
+      "epoch": 0.1754349193041291,
+      "grad_norm": 0.18003013730049133,
+      "learning_rate": 9.458187823673145e-06,
+      "loss": 0.0681,
+      "step": 837
+    },
+    {
+      "epoch": 0.17564451896876965,
+      "grad_norm": 0.15927360951900482,
+      "learning_rate": 9.456649779003548e-06,
+      "loss": 0.0696,
+      "step": 838
+    },
+    {
+      "epoch": 0.1758541186334102,
+      "grad_norm": 0.1543167531490326,
+      "learning_rate": 9.45510967981933e-06,
+      "loss": 0.0673,
+      "step": 839
+    },
+    {
+      "epoch": 0.17606371829805073,
+      "grad_norm": 0.1319018304347992,
+      "learning_rate": 9.453567526830471e-06,
+      "loss": 0.072,
+      "step": 840
+    },
+    {
+      "epoch": 0.17627331796269127,
+      "grad_norm": 0.1290387362241745,
+      "learning_rate": 9.452023320747909e-06,
+      "loss": 0.0718,
+      "step": 841
+    },
+    {
+      "epoch": 0.1764829176273318,
+      "grad_norm": 0.1876518577337265,
+      "learning_rate": 9.450477062283523e-06,
+      "loss": 0.0717,
+      "step": 842
+    },
+    {
+      "epoch": 0.17669251729197233,
+      "grad_norm": 0.2217532843351364,
+      "learning_rate": 9.448928752150134e-06,
+      "loss": 0.0691,
+      "step": 843
+    },
+    {
+      "epoch": 0.17690211695661287,
+      "grad_norm": 0.2288810908794403,
+      "learning_rate": 9.447378391061515e-06,
+      "loss": 0.0718,
+      "step": 844
+    },
+    {
+      "epoch": 0.1771117166212534,
+      "grad_norm": 0.23466289043426514,
+      "learning_rate": 9.44582597973238e-06,
+      "loss": 0.0716,
+      "step": 845
+    },
+    {
+      "epoch": 0.17732131628589395,
+      "grad_norm": 0.20989495515823364,
+      "learning_rate": 9.444271518878393e-06,
+      "loss": 0.0726,
+      "step": 846
+    },
+    {
+      "epoch": 0.17753091595053447,
+      "grad_norm": 0.1636846363544464,
+      "learning_rate": 9.442715009216159e-06,
+      "loss": 0.0689,
+      "step": 847
+    },
+    {
+      "epoch": 0.177740515615175,
+      "grad_norm": 0.13839955627918243,
+      "learning_rate": 9.441156451463228e-06,
+      "loss": 0.0693,
+      "step": 848
+    },
+    {
+      "epoch": 0.17795011527981555,
+      "grad_norm": 0.17058932781219482,
+      "learning_rate": 9.439595846338097e-06,
+      "loss": 0.0693,
+      "step": 849
+    },
+    {
+      "epoch": 0.1781597149444561,
+      "grad_norm": 0.20214615762233734,
+      "learning_rate": 9.4380331945602e-06,
+      "loss": 0.0727,
+      "step": 850
+    },
+    {
+      "epoch": 0.17836931460909664,
+      "grad_norm": 0.178322434425354,
+      "learning_rate": 9.436468496849924e-06,
+      "loss": 0.0766,
+      "step": 851
+    },
+    {
+      "epoch": 0.17857891427373715,
+      "grad_norm": 0.14965812861919403,
+      "learning_rate": 9.434901753928593e-06,
+      "loss": 0.0709,
+      "step": 852
+    },
+    {
+      "epoch": 0.1787885139383777,
+      "grad_norm": 0.16711200773715973,
+      "learning_rate": 9.433332966518473e-06,
+      "loss": 0.0673,
+      "step": 853
+    },
+    {
+      "epoch": 0.17899811360301823,
+      "grad_norm": 0.17375777661800385,
+      "learning_rate": 9.43176213534278e-06,
+      "loss": 0.0738,
+      "step": 854
+    },
+    {
+      "epoch": 0.17920771326765877,
+      "grad_norm": 0.1556476205587387,
+      "learning_rate": 9.43018926112566e-06,
+      "loss": 0.068,
+      "step": 855
+    },
+    {
+      "epoch": 0.17941731293229932,
+      "grad_norm": 0.1631685495376587,
+      "learning_rate": 9.42861434459221e-06,
+      "loss": 0.0718,
+      "step": 856
+    },
+    {
+      "epoch": 0.17962691259693986,
+      "grad_norm": 0.1664845198392868,
+      "learning_rate": 9.42703738646847e-06,
+      "loss": 0.0689,
+      "step": 857
+    },
+    {
+      "epoch": 0.17983651226158037,
+      "grad_norm": 0.15820707380771637,
+      "learning_rate": 9.425458387481412e-06,
+      "loss": 0.072,
+      "step": 858
+    },
+    {
+      "epoch": 0.18004611192622091,
+      "grad_norm": 0.15286661684513092,
+      "learning_rate": 9.423877348358956e-06,
+      "loss": 0.0743,
+      "step": 859
+    },
+    {
+      "epoch": 0.18025571159086146,
+      "grad_norm": 0.14925047755241394,
+      "learning_rate": 9.422294269829963e-06,
+      "loss": 0.0665,
+      "step": 860
+    },
+    {
+      "epoch": 0.180465311255502,
+      "grad_norm": 0.15763109922409058,
+      "learning_rate": 9.420709152624232e-06,
+      "loss": 0.0725,
+      "step": 861
+    },
+    {
+      "epoch": 0.18067491092014254,
+      "grad_norm": 0.1531532257795334,
+      "learning_rate": 9.419121997472497e-06,
+      "loss": 0.0699,
+      "step": 862
+    },
+    {
+      "epoch": 0.18088451058478305,
+      "grad_norm": 0.14012256264686584,
+      "learning_rate": 9.41753280510644e-06,
+      "loss": 0.0718,
+      "step": 863
+    },
+    {
+      "epoch": 0.1810941102494236,
+      "grad_norm": 0.14636793732643127,
+      "learning_rate": 9.415941576258679e-06,
+      "loss": 0.0683,
+      "step": 864
+    },
+    {
+      "epoch": 0.18130370991406414,
+      "grad_norm": 0.16276605427265167,
+      "learning_rate": 9.414348311662766e-06,
+      "loss": 0.0706,
+      "step": 865
+    },
+    {
+      "epoch": 0.18151330957870468,
+      "grad_norm": 0.1377471536397934,
+      "learning_rate": 9.4127530120532e-06,
+      "loss": 0.0727,
+      "step": 866
+    },
+    {
+      "epoch": 0.18172290924334522,
+      "grad_norm": 0.11346344649791718,
+      "learning_rate": 9.41115567816541e-06,
+      "loss": 0.0699,
+      "step": 867
+    },
+    {
+      "epoch": 0.18193250890798573,
+      "grad_norm": 0.1656496673822403,
+      "learning_rate": 9.40955631073577e-06,
+      "loss": 0.0673,
+      "step": 868
+    },
+    {
+      "epoch": 0.18214210857262628,
+      "grad_norm": 0.16461622714996338,
+      "learning_rate": 9.407954910501586e-06,
+      "loss": 0.0769,
+      "step": 869
+    },
+    {
+      "epoch": 0.18235170823726682,
+      "grad_norm": 0.12896548211574554,
+      "learning_rate": 9.406351478201105e-06,
+      "loss": 0.0699,
+      "step": 870
+    },
+    {
+      "epoch": 0.18256130790190736,
+      "grad_norm": 0.17316146194934845,
+      "learning_rate": 9.404746014573508e-06,
+      "loss": 0.0713,
+      "step": 871
+    },
+    {
+      "epoch": 0.1827709075665479,
+      "grad_norm": 0.17203350365161896,
+      "learning_rate": 9.403138520358911e-06,
+      "loss": 0.0718,
+      "step": 872
+    },
+    {
+      "epoch": 0.18298050723118844,
+      "grad_norm": 0.17705942690372467,
+      "learning_rate": 9.401528996298375e-06,
+      "loss": 0.074,
+      "step": 873
+    },
+    {
+      "epoch": 0.18319010689582896,
+      "grad_norm": 0.2316497564315796,
+      "learning_rate": 9.399917443133883e-06,
+      "loss": 0.0689,
+      "step": 874
+    },
+    {
+      "epoch": 0.1833997065604695,
+      "grad_norm": 0.23241883516311646,
+      "learning_rate": 9.398303861608363e-06,
+      "loss": 0.0707,
+      "step": 875
+    },
+    {
+      "epoch": 0.18360930622511004,
+      "grad_norm": 0.21609675884246826,
+      "learning_rate": 9.396688252465678e-06,
+      "loss": 0.0739,
+      "step": 876
+    },
+    {
+      "epoch": 0.18381890588975058,
+      "grad_norm": 0.21849660575389862,
+      "learning_rate": 9.395070616450622e-06,
+      "loss": 0.0729,
+      "step": 877
+    },
+    {
+      "epoch": 0.18402850555439113,
+      "grad_norm": 0.17387786507606506,
+      "learning_rate": 9.393450954308922e-06,
+      "loss": 0.0715,
+      "step": 878
+    },
+    {
+      "epoch": 0.18423810521903164,
+      "grad_norm": 0.1386607140302658,
+      "learning_rate": 9.391829266787248e-06,
+      "loss": 0.0708,
+      "step": 879
+    },
+    {
+      "epoch": 0.18444770488367218,
+      "grad_norm": 0.14751608669757843,
+      "learning_rate": 9.390205554633193e-06,
+      "loss": 0.0714,
+      "step": 880
+    },
+    {
+      "epoch": 0.18465730454831272,
+      "grad_norm": 0.14015577733516693,
+      "learning_rate": 9.38857981859529e-06,
+      "loss": 0.0699,
+      "step": 881
+    },
+    {
+      "epoch": 0.18486690421295326,
+      "grad_norm": 0.13030049204826355,
+      "learning_rate": 9.386952059423e-06,
+      "loss": 0.0732,
+      "step": 882
+    },
+    {
+      "epoch": 0.1850765038775938,
+      "grad_norm": 0.14583101868629456,
+      "learning_rate": 9.385322277866724e-06,
+      "loss": 0.0681,
+      "step": 883
+    },
+    {
+      "epoch": 0.18528610354223432,
+      "grad_norm": 0.1531733274459839,
+      "learning_rate": 9.38369047467779e-06,
+      "loss": 0.0691,
+      "step": 884
+    },
+    {
+      "epoch": 0.18549570320687486,
+      "grad_norm": 0.16215740144252777,
+      "learning_rate": 9.382056650608454e-06,
+      "loss": 0.0709,
+      "step": 885
+    },
+    {
+      "epoch": 0.1857053028715154,
+      "grad_norm": 0.1694212257862091,
+      "learning_rate": 9.380420806411914e-06,
+      "loss": 0.0694,
+      "step": 886
+    },
+    {
+      "epoch": 0.18591490253615595,
+      "grad_norm": 0.16096119582653046,
+      "learning_rate": 9.378782942842292e-06,
+      "loss": 0.0656,
+      "step": 887
+    },
+    {
+      "epoch": 0.1861245022007965,
+      "grad_norm": 0.13555142283439636,
+      "learning_rate": 9.377143060654645e-06,
+      "loss": 0.0741,
+      "step": 888
+    },
+    {
+      "epoch": 0.186334101865437,
+      "grad_norm": 0.11763742566108704,
+      "learning_rate": 9.375501160604955e-06,
+      "loss": 0.069,
+      "step": 889
+    },
+    {
+      "epoch": 0.18654370153007754,
+      "grad_norm": 0.14128689467906952,
+      "learning_rate": 9.373857243450138e-06,
+      "loss": 0.0691,
+      "step": 890
+    },
+    {
+      "epoch": 0.18675330119471809,
+      "grad_norm": 0.15519973635673523,
+      "learning_rate": 9.37221130994804e-06,
+      "loss": 0.0691,
+      "step": 891
+    },
+    {
+      "epoch": 0.18696290085935863,
+      "grad_norm": 0.1528463065624237,
+      "learning_rate": 9.370563360857437e-06,
+      "loss": 0.0708,
+      "step": 892
+    },
+    {
+      "epoch": 0.18717250052399917,
+      "grad_norm": 0.1430312991142273,
+      "learning_rate": 9.368913396938031e-06,
+      "loss": 0.0702,
+      "step": 893
+    },
+    {
+      "epoch": 0.1873821001886397,
+      "grad_norm": 0.13003818690776825,
+      "learning_rate": 9.367261418950459e-06,
+      "loss": 0.0729,
+      "step": 894
+    },
+    {
+      "epoch": 0.18759169985328022,
+      "grad_norm": 0.1460046023130417,
+      "learning_rate": 9.365607427656277e-06,
+      "loss": 0.0701,
+      "step": 895
+    },
+    {
+      "epoch": 0.18780129951792077,
+      "grad_norm": 0.1653718799352646,
+      "learning_rate": 9.363951423817982e-06,
+      "loss": 0.0713,
+      "step": 896
+    },
+    {
+      "epoch": 0.1880108991825613,
+      "grad_norm": 0.1808050125837326,
+      "learning_rate": 9.362293408198983e-06,
+      "loss": 0.0671,
+      "step": 897
+    },
+    {
+      "epoch": 0.18822049884720185,
+      "grad_norm": 0.17613857984542847,
+      "learning_rate": 9.360633381563631e-06,
+      "loss": 0.0684,
+      "step": 898
+    },
+    {
+      "epoch": 0.1884300985118424,
+      "grad_norm": 0.17661046981811523,
+      "learning_rate": 9.358971344677197e-06,
+      "loss": 0.068,
+      "step": 899
+    },
+    {
+      "epoch": 0.1886396981764829,
+      "grad_norm": 0.205166295170784,
+      "learning_rate": 9.357307298305875e-06,
+      "loss": 0.0695,
+      "step": 900
+    },
+    {
+      "epoch": 0.18884929784112345,
+      "grad_norm": 0.22776846587657928,
+      "learning_rate": 9.355641243216798e-06,
+      "loss": 0.0695,
+      "step": 901
+    },
+    {
+      "epoch": 0.189058897505764,
+      "grad_norm": 0.21361425518989563,
+      "learning_rate": 9.35397318017801e-06,
+      "loss": 0.0736,
+      "step": 902
+    },
+    {
+      "epoch": 0.18926849717040453,
+      "grad_norm": 0.18417485058307648,
+      "learning_rate": 9.35230310995849e-06,
+      "loss": 0.0719,
+      "step": 903
+    },
+    {
+      "epoch": 0.18947809683504507,
+      "grad_norm": 0.17140980064868927,
+      "learning_rate": 9.35063103332814e-06,
+      "loss": 0.0692,
+      "step": 904
+    },
+    {
+      "epoch": 0.1896876964996856,
+      "grad_norm": 0.17577317357063293,
+      "learning_rate": 9.348956951057788e-06,
+      "loss": 0.0702,
+      "step": 905
+    },
+    {
+      "epoch": 0.18989729616432613,
+      "grad_norm": 0.18869447708129883,
+      "learning_rate": 9.347280863919186e-06,
+      "loss": 0.0682,
+      "step": 906
+    },
+    {
+      "epoch": 0.19010689582896667,
+      "grad_norm": 0.15782123804092407,
+      "learning_rate": 9.345602772685008e-06,
+      "loss": 0.0674,
+      "step": 907
+    },
+    {
+      "epoch": 0.1903164954936072,
+      "grad_norm": 0.15781988203525543,
+      "learning_rate": 9.343922678128854e-06,
+      "loss": 0.0703,
+      "step": 908
+    },
+    {
+      "epoch": 0.19052609515824775,
+      "grad_norm": 0.17203202843666077,
+      "learning_rate": 9.342240581025248e-06,
+      "loss": 0.0672,
+      "step": 909
+    },
+    {
+      "epoch": 0.1907356948228883,
+      "grad_norm": 0.16652663052082062,
+      "learning_rate": 9.340556482149637e-06,
+      "loss": 0.0671,
+      "step": 910
+    },
+    {
+      "epoch": 0.1909452944875288,
+      "grad_norm": 0.15487754344940186,
+      "learning_rate": 9.33887038227839e-06,
+      "loss": 0.0699,
+      "step": 911
+    },
+    {
+      "epoch": 0.19115489415216935,
+      "grad_norm": 0.13165274262428284,
+      "learning_rate": 9.337182282188794e-06,
+      "loss": 0.0666,
+      "step": 912
+    },
+    {
+      "epoch": 0.1913644938168099,
+      "grad_norm": 0.1349601447582245,
+      "learning_rate": 9.335492182659071e-06,
+      "loss": 0.0664,
+      "step": 913
+    },
+    {
+      "epoch": 0.19157409348145044,
+      "grad_norm": 0.1551382839679718,
+      "learning_rate": 9.33380008446835e-06,
+      "loss": 0.0677,
+      "step": 914
+    },
+    {
+      "epoch": 0.19178369314609098,
+      "grad_norm": 0.16102422773838043,
+      "learning_rate": 9.332105988396692e-06,
+      "loss": 0.0693,
+      "step": 915
+    },
+    {
+      "epoch": 0.1919932928107315,
+      "grad_norm": 0.17208023369312286,
+      "learning_rate": 9.330409895225072e-06,
+      "loss": 0.0704,
+      "step": 916
+    },
+    {
+      "epoch": 0.19220289247537203,
+      "grad_norm": 0.16486892104148865,
+      "learning_rate": 9.32871180573539e-06,
+      "loss": 0.0695,
+      "step": 917
+    },
+    {
+      "epoch": 0.19241249214001258,
+      "grad_norm": 0.15366514027118683,
+      "learning_rate": 9.327011720710464e-06,
+      "loss": 0.0688,
+      "step": 918
+    },
+    {
+      "epoch": 0.19262209180465312,
+      "grad_norm": 0.17730504274368286,
+      "learning_rate": 9.325309640934036e-06,
+      "loss": 0.0678,
+      "step": 919
+    },
+    {
+      "epoch": 0.19283169146929366,
+      "grad_norm": 0.2048606127500534,
+      "learning_rate": 9.323605567190758e-06,
+      "loss": 0.0696,
+      "step": 920
+    },
+    {
+      "epoch": 0.19304129113393417,
+      "grad_norm": 0.1947096735239029,
+      "learning_rate": 9.321899500266216e-06,
+      "loss": 0.0694,
+      "step": 921
+    },
+    {
+      "epoch": 0.19325089079857471,
+      "grad_norm": 0.15000185370445251,
+      "learning_rate": 9.3201914409469e-06,
+      "loss": 0.0675,
+      "step": 922
+    },
+    {
+      "epoch": 0.19346049046321526,
+      "grad_norm": 0.15103282034397125,
+      "learning_rate": 9.318481390020226e-06,
+      "loss": 0.071,
+      "step": 923
+    },
+    {
+      "epoch": 0.1936700901278558,
+      "grad_norm": 0.16864833235740662,
+      "learning_rate": 9.31676934827453e-06,
+      "loss": 0.0695,
+      "step": 924
+    },
+    {
+      "epoch": 0.19387968979249634,
+      "grad_norm": 0.1579030156135559,
+      "learning_rate": 9.31505531649906e-06,
+      "loss": 0.0652,
+      "step": 925
+    },
+    {
+      "epoch": 0.19408928945713688,
+      "grad_norm": 0.13766202330589294,
+      "learning_rate": 9.313339295483985e-06,
+      "loss": 0.071,
+      "step": 926
+    },
+    {
+      "epoch": 0.1942988891217774,
+      "grad_norm": 0.1618116796016693,
+      "learning_rate": 9.311621286020393e-06,
+      "loss": 0.0692,
+      "step": 927
+    },
+    {
+      "epoch": 0.19450848878641794,
+      "grad_norm": 0.15055081248283386,
+      "learning_rate": 9.309901288900283e-06,
+      "loss": 0.0675,
+      "step": 928
+    },
+    {
+      "epoch": 0.19471808845105848,
+      "grad_norm": 0.13458986580371857,
+      "learning_rate": 9.308179304916573e-06,
+      "loss": 0.0717,
+      "step": 929
+    },
+    {
+      "epoch": 0.19492768811569902,
+      "grad_norm": 0.13572897017002106,
+      "learning_rate": 9.306455334863098e-06,
+      "loss": 0.0708,
+      "step": 930
+    },
+    {
+      "epoch": 0.19513728778033956,
+      "grad_norm": 0.13782450556755066,
+      "learning_rate": 9.30472937953461e-06,
+      "loss": 0.067,
+      "step": 931
+    },
+    {
+      "epoch": 0.19534688744498008,
+      "grad_norm": 0.1412908285856247,
+      "learning_rate": 9.303001439726772e-06,
+      "loss": 0.0667,
+      "step": 932
+    },
+    {
+      "epoch": 0.19555648710962062,
+      "grad_norm": 0.1327088475227356,
+      "learning_rate": 9.301271516236162e-06,
+      "loss": 0.0674,
+      "step": 933
+    },
+    {
+      "epoch": 0.19576608677426116,
+      "grad_norm": 0.13067220151424408,
+      "learning_rate": 9.299539609860278e-06,
+      "loss": 0.0683,
+      "step": 934
+    },
+    {
+      "epoch": 0.1959756864389017,
+      "grad_norm": 0.13276226818561554,
+      "learning_rate": 9.297805721397525e-06,
+      "loss": 0.0717,
+      "step": 935
+    },
+    {
+      "epoch": 0.19618528610354224,
+      "grad_norm": 0.13292694091796875,
+      "learning_rate": 9.29606985164723e-06,
+      "loss": 0.0682,
+      "step": 936
+    },
+    {
+      "epoch": 0.19639488576818276,
+      "grad_norm": 0.1614527702331543,
+      "learning_rate": 9.294332001409625e-06,
+      "loss": 0.0693,
+      "step": 937
+    },
+    {
+      "epoch": 0.1966044854328233,
+      "grad_norm": 0.20869992673397064,
+      "learning_rate": 9.292592171485858e-06,
+      "loss": 0.0687,
+      "step": 938
+    },
+    {
+      "epoch": 0.19681408509746384,
+      "grad_norm": 0.25103163719177246,
+      "learning_rate": 9.290850362677993e-06,
+      "loss": 0.0673,
+      "step": 939
+    },
+    {
+      "epoch": 0.19702368476210438,
+      "grad_norm": 0.2684808671474457,
+      "learning_rate": 9.289106575789001e-06,
+      "loss": 0.0711,
+      "step": 940
+    },
+    {
+      "epoch": 0.19723328442674493,
+      "grad_norm": 0.2292683720588684,
+      "learning_rate": 9.28736081162277e-06,
+      "loss": 0.075,
+      "step": 941
+    },
+    {
+      "epoch": 0.19744288409138544,
+      "grad_norm": 0.148003488779068,
+      "learning_rate": 9.285613070984094e-06,
+      "loss": 0.0736,
+      "step": 942
+    },
+    {
+      "epoch": 0.19765248375602598,
+      "grad_norm": 0.1527291238307953,
+      "learning_rate": 9.283863354678683e-06,
+      "loss": 0.0691,
+      "step": 943
+    },
+    {
+      "epoch": 0.19786208342066652,
+      "grad_norm": 0.17190136015415192,
+      "learning_rate": 9.282111663513156e-06,
+      "loss": 0.0696,
+      "step": 944
+    },
+    {
+      "epoch": 0.19807168308530707,
+      "grad_norm": 0.16085271537303925,
+      "learning_rate": 9.280357998295041e-06,
+      "loss": 0.0692,
+      "step": 945
+    },
+    {
+      "epoch": 0.1982812827499476,
+      "grad_norm": 0.16574037075042725,
+      "learning_rate": 9.278602359832778e-06,
+      "loss": 0.0695,
+      "step": 946
+    },
+    {
+      "epoch": 0.19849088241458815,
+      "grad_norm": 0.1433473825454712,
+      "learning_rate": 9.276844748935715e-06,
+      "loss": 0.066,
+      "step": 947
+    },
+    {
+      "epoch": 0.19870048207922866,
+      "grad_norm": 0.15432019531726837,
+      "learning_rate": 9.275085166414113e-06,
+      "loss": 0.0716,
+      "step": 948
+    },
+    {
+      "epoch": 0.1989100817438692,
+      "grad_norm": 0.155776709318161,
+      "learning_rate": 9.273323613079135e-06,
+      "loss": 0.0687,
+      "step": 949
+    },
+    {
+      "epoch": 0.19911968140850975,
+      "grad_norm": 0.14493581652641296,
+      "learning_rate": 9.27156008974286e-06,
+      "loss": 0.0688,
+      "step": 950
+    },
+    {
+      "epoch": 0.1993292810731503,
+      "grad_norm": 0.16887947916984558,
+      "learning_rate": 9.269794597218271e-06,
+      "loss": 0.0682,
+      "step": 951
+    },
+    {
+      "epoch": 0.19953888073779083,
+      "grad_norm": 0.15968920290470123,
+      "learning_rate": 9.26802713631926e-06,
+      "loss": 0.0652,
+      "step": 952
+    },
+    {
+      "epoch": 0.19974848040243134,
+      "grad_norm": 0.15079444646835327,
+      "learning_rate": 9.266257707860625e-06,
+      "loss": 0.0661,
+      "step": 953
+    },
+    {
+      "epoch": 0.19995808006707189,
+      "grad_norm": 0.15379679203033447,
+      "learning_rate": 9.264486312658073e-06,
+      "loss": 0.0723,
+      "step": 954
+    },
+    {
+      "epoch": 0.20016767973171243,
+      "grad_norm": 0.1347336769104004,
+      "learning_rate": 9.262712951528217e-06,
+      "loss": 0.0729,
+      "step": 955
+    },
+    {
+      "epoch": 0.20037727939635297,
+      "grad_norm": 0.13610079884529114,
+      "learning_rate": 9.260937625288576e-06,
+      "loss": 0.0693,
+      "step": 956
+    },
+    {
+      "epoch": 0.2005868790609935,
+      "grad_norm": 0.14802809059619904,
+      "learning_rate": 9.259160334757575e-06,
+      "loss": 0.0685,
+      "step": 957
+    },
+    {
+      "epoch": 0.20079647872563403,
+      "grad_norm": 0.15134483575820923,
+      "learning_rate": 9.257381080754544e-06,
+      "loss": 0.0655,
+      "step": 958
+    },
+    {
+      "epoch": 0.20100607839027457,
+      "grad_norm": 0.1678546667098999,
+      "learning_rate": 9.255599864099718e-06,
+      "loss": 0.0715,
+      "step": 959
+    },
+    {
+      "epoch": 0.2012156780549151,
+      "grad_norm": 0.15980494022369385,
+      "learning_rate": 9.25381668561424e-06,
+      "loss": 0.0695,
+      "step": 960
+    },
+    {
+      "epoch": 0.20142527771955565,
+      "grad_norm": 0.13359016180038452,
+      "learning_rate": 9.252031546120153e-06,
+      "loss": 0.0671,
+      "step": 961
+    },
+    {
+      "epoch": 0.2016348773841962,
+      "grad_norm": 0.154948428273201,
+      "learning_rate": 9.250244446440406e-06,
+      "loss": 0.0676,
+      "step": 962
+    },
+    {
+      "epoch": 0.20184447704883673,
+      "grad_norm": 0.16912510991096497,
+      "learning_rate": 9.248455387398853e-06,
+      "loss": 0.0687,
+      "step": 963
+    },
+    {
+      "epoch": 0.20205407671347725,
+      "grad_norm": 0.1481633484363556,
+      "learning_rate": 9.246664369820249e-06,
+      "loss": 0.0696,
+      "step": 964
+    },
+    {
+      "epoch": 0.2022636763781178,
+      "grad_norm": 0.1526942402124405,
+      "learning_rate": 9.244871394530252e-06,
+      "loss": 0.0669,
+      "step": 965
+    },
+    {
+      "epoch": 0.20247327604275833,
+      "grad_norm": 0.1594848930835724,
+      "learning_rate": 9.243076462355424e-06,
+      "loss": 0.0695,
+      "step": 966
+    },
+    {
+      "epoch": 0.20268287570739887,
+      "grad_norm": 0.15428997576236725,
+      "learning_rate": 9.241279574123228e-06,
+      "loss": 0.0676,
+      "step": 967
+    },
+    {
+      "epoch": 0.20289247537203942,
+      "grad_norm": 0.15648357570171356,
+      "learning_rate": 9.239480730662029e-06,
+      "loss": 0.0701,
+      "step": 968
+    },
+    {
+      "epoch": 0.20310207503667993,
+      "grad_norm": 0.14083035290241241,
+      "learning_rate": 9.237679932801095e-06,
+      "loss": 0.0707,
+      "step": 969
+    },
+    {
+      "epoch": 0.20331167470132047,
+      "grad_norm": 0.11451932042837143,
+      "learning_rate": 9.235877181370592e-06,
+      "loss": 0.0687,
+      "step": 970
+    },
+    {
+      "epoch": 0.203521274365961,
+      "grad_norm": 0.12343299388885498,
+      "learning_rate": 9.234072477201588e-06,
+      "loss": 0.0672,
+      "step": 971
+    },
+    {
+      "epoch": 0.20373087403060156,
+      "grad_norm": 0.1318097859621048,
+      "learning_rate": 9.23226582112605e-06,
+      "loss": 0.0692,
+      "step": 972
+    },
+    {
+      "epoch": 0.2039404736952421,
+      "grad_norm": 0.13317148387432098,
+      "learning_rate": 9.23045721397685e-06,
+      "loss": 0.0652,
+      "step": 973
+    },
+    {
+      "epoch": 0.2041500733598826,
+      "grad_norm": 0.14441275596618652,
+      "learning_rate": 9.228646656587751e-06,
+      "loss": 0.0669,
+      "step": 974
+    },
+    {
+      "epoch": 0.20435967302452315,
+      "grad_norm": 0.1513681560754776,
+      "learning_rate": 9.226834149793422e-06,
+      "loss": 0.0654,
+      "step": 975
+    },
+    {
+      "epoch": 0.2045692726891637,
+      "grad_norm": 0.17051391303539276,
+      "learning_rate": 9.225019694429429e-06,
+      "loss": 0.0664,
+      "step": 976
+    },
+    {
+      "epoch": 0.20477887235380424,
+      "grad_norm": 0.19565874338150024,
+      "learning_rate": 9.223203291332234e-06,
+      "loss": 0.0691,
+      "step": 977
+    },
+    {
+      "epoch": 0.20498847201844478,
+      "grad_norm": 0.2343725860118866,
+      "learning_rate": 9.2213849413392e-06,
+      "loss": 0.069,
+      "step": 978
+    },
+    {
+      "epoch": 0.20519807168308532,
+      "grad_norm": 0.27315953373908997,
+      "learning_rate": 9.219564645288583e-06,
+      "loss": 0.0675,
+      "step": 979
+    },
+    {
+      "epoch": 0.20540767134772583,
+      "grad_norm": 0.2724561393260956,
+      "learning_rate": 9.217742404019544e-06,
+      "loss": 0.0702,
+      "step": 980
+    },
+    {
+      "epoch": 0.20561727101236638,
+      "grad_norm": 0.23024438321590424,
+      "learning_rate": 9.215918218372128e-06,
+      "loss": 0.0723,
+      "step": 981
+    },
+    {
+      "epoch": 0.20582687067700692,
+      "grad_norm": 0.15981607139110565,
+      "learning_rate": 9.214092089187293e-06,
+      "loss": 0.0696,
+      "step": 982
+    },
+    {
+      "epoch": 0.20603647034164746,
+      "grad_norm": 0.165577694773674,
+      "learning_rate": 9.212264017306878e-06,
+      "loss": 0.0725,
+      "step": 983
+    },
+    {
+      "epoch": 0.206246070006288,
+      "grad_norm": 0.19346004724502563,
+      "learning_rate": 9.210434003573627e-06,
+      "loss": 0.0689,
+      "step": 984
+    },
+    {
+      "epoch": 0.20645566967092852,
+      "grad_norm": 0.17302066087722778,
+      "learning_rate": 9.208602048831176e-06,
+      "loss": 0.0695,
+      "step": 985
+    },
+    {
+      "epoch": 0.20666526933556906,
+      "grad_norm": 0.17835290729999542,
+      "learning_rate": 9.206768153924052e-06,
+      "loss": 0.0684,
+      "step": 986
+    },
+    {
+      "epoch": 0.2068748690002096,
+      "grad_norm": 0.1855718344449997,
+      "learning_rate": 9.204932319697686e-06,
+      "loss": 0.0694,
+      "step": 987
+    },
+    {
+      "epoch": 0.20708446866485014,
+      "grad_norm": 0.16161462664604187,
+      "learning_rate": 9.203094546998392e-06,
+      "loss": 0.0689,
+      "step": 988
+    },
+    {
+      "epoch": 0.20729406832949068,
+      "grad_norm": 0.1590941995382309,
+      "learning_rate": 9.201254836673386e-06,
+      "loss": 0.0676,
+      "step": 989
+    },
+    {
+      "epoch": 0.2075036679941312,
+      "grad_norm": 0.17780111730098724,
+      "learning_rate": 9.199413189570772e-06,
+      "loss": 0.0699,
+      "step": 990
+    },
+    {
+      "epoch": 0.20771326765877174,
+      "grad_norm": 0.15813374519348145,
+      "learning_rate": 9.197569606539551e-06,
+      "loss": 0.0678,
+      "step": 991
+    },
+    {
+      "epoch": 0.20792286732341228,
+      "grad_norm": 0.14259178936481476,
+      "learning_rate": 9.195724088429611e-06,
+      "loss": 0.0682,
+      "step": 992
+    },
+    {
+      "epoch": 0.20813246698805282,
+      "grad_norm": 0.15479755401611328,
+      "learning_rate": 9.193876636091741e-06,
+      "loss": 0.0703,
+      "step": 993
+    },
+    {
+      "epoch": 0.20834206665269336,
+      "grad_norm": 0.16291916370391846,
+      "learning_rate": 9.192027250377611e-06,
+      "loss": 0.0714,
+      "step": 994
+    },
+    {
+      "epoch": 0.2085516663173339,
+      "grad_norm": 0.16300472617149353,
+      "learning_rate": 9.19017593213979e-06,
+      "loss": 0.0686,
+      "step": 995
+    },
+    {
+      "epoch": 0.20876126598197442,
+      "grad_norm": 0.16274508833885193,
+      "learning_rate": 9.188322682231733e-06,
+      "loss": 0.0674,
+      "step": 996
+    },
+    {
+      "epoch": 0.20897086564661496,
+      "grad_norm": 0.18186375498771667,
+      "learning_rate": 9.186467501507792e-06,
+      "loss": 0.0696,
+      "step": 997
+    },
+    {
+      "epoch": 0.2091804653112555,
+      "grad_norm": 0.20312577486038208,
+      "learning_rate": 9.184610390823202e-06,
+      "loss": 0.0659,
+      "step": 998
+    },
+    {
+      "epoch": 0.20939006497589605,
+      "grad_norm": 0.21246276795864105,
+      "learning_rate": 9.18275135103409e-06,
+      "loss": 0.0714,
+      "step": 999
+    },
+    {
+      "epoch": 0.2095996646405366,
+      "grad_norm": 0.19625847041606903,
+      "learning_rate": 9.180890382997473e-06,
+      "loss": 0.0662,
+      "step": 1000
+    },
+    {
+      "epoch": 0.2098092643051771,
+      "grad_norm": 0.181460440158844,
+      "learning_rate": 9.179027487571258e-06,
+      "loss": 0.0662,
+      "step": 1001
+    },
+    {
+      "epoch": 0.21001886396981764,
+      "grad_norm": 0.17908546328544617,
+      "learning_rate": 9.177162665614242e-06,
+      "loss": 0.0685,
+      "step": 1002
+    },
+    {
+      "epoch": 0.21022846363445818,
+      "grad_norm": 0.19209085404872894,
+      "learning_rate": 9.175295917986103e-06,
+      "loss": 0.0682,
+      "step": 1003
+    },
+    {
+      "epoch": 0.21043806329909873,
+      "grad_norm": 0.21326391398906708,
+      "learning_rate": 9.173427245547414e-06,
+      "loss": 0.0691,
+      "step": 1004
+    },
+    {
+      "epoch": 0.21064766296373927,
+      "grad_norm": 0.19269311428070068,
+      "learning_rate": 9.17155664915963e-06,
+      "loss": 0.0727,
+      "step": 1005
+    },
+    {
+      "epoch": 0.21085726262837978,
+      "grad_norm": 0.19682708382606506,
+      "learning_rate": 9.169684129685099e-06,
+      "loss": 0.0713,
+      "step": 1006
+    },
+    {
+      "epoch": 0.21106686229302032,
+      "grad_norm": 0.18189750611782074,
+      "learning_rate": 9.16780968798705e-06,
+      "loss": 0.0714,
+      "step": 1007
+    },
+    {
+      "epoch": 0.21127646195766087,
+      "grad_norm": 0.169154092669487,
+      "learning_rate": 9.165933324929599e-06,
+      "loss": 0.0702,
+      "step": 1008
+    },
+    {
+      "epoch": 0.2114860616223014,
+      "grad_norm": 0.20186983048915863,
+      "learning_rate": 9.164055041377754e-06,
+      "loss": 0.0691,
+      "step": 1009
+    },
+    {
+      "epoch": 0.21169566128694195,
+      "grad_norm": 0.19826248288154602,
+      "learning_rate": 9.162174838197396e-06,
+      "loss": 0.0679,
+      "step": 1010
+    },
+    {
+      "epoch": 0.21190526095158246,
+      "grad_norm": 0.1933613270521164,
+      "learning_rate": 9.160292716255303e-06,
+      "loss": 0.0691,
+      "step": 1011
+    },
+    {
+      "epoch": 0.212114860616223,
+      "grad_norm": 0.1532355546951294,
+      "learning_rate": 9.158408676419133e-06,
+      "loss": 0.0712,
+      "step": 1012
+    },
+    {
+      "epoch": 0.21232446028086355,
+      "grad_norm": 0.13550697267055511,
+      "learning_rate": 9.156522719557428e-06,
+      "loss": 0.0688,
+      "step": 1013
+    },
+    {
+      "epoch": 0.2125340599455041,
+      "grad_norm": 0.19453023374080658,
+      "learning_rate": 9.15463484653961e-06,
+      "loss": 0.0646,
+      "step": 1014
+    },
+    {
+      "epoch": 0.21274365961014463,
+      "grad_norm": 0.16020941734313965,
+      "learning_rate": 9.152745058235993e-06,
+      "loss": 0.0692,
+      "step": 1015
+    },
+    {
+      "epoch": 0.21295325927478517,
+      "grad_norm": 0.17123140394687653,
+      "learning_rate": 9.150853355517765e-06,
+      "loss": 0.0689,
+      "step": 1016
+    },
+    {
+      "epoch": 0.2131628589394257,
+      "grad_norm": 0.17353491485118866,
+      "learning_rate": 9.148959739257005e-06,
+      "loss": 0.0687,
+      "step": 1017
+    },
+    {
+      "epoch": 0.21337245860406623,
+      "grad_norm": 0.15080298483371735,
+      "learning_rate": 9.147064210326664e-06,
+      "loss": 0.0664,
+      "step": 1018
+    },
+    {
+      "epoch": 0.21358205826870677,
+      "grad_norm": 0.1491314321756363,
+      "learning_rate": 9.145166769600584e-06,
+      "loss": 0.0697,
+      "step": 1019
+    },
+    {
+      "epoch": 0.2137916579333473,
+      "grad_norm": 0.10465126484632492,
+      "learning_rate": 9.143267417953486e-06,
+      "loss": 0.068,
+      "step": 1020
+    },
+    {
+      "epoch": 0.21400125759798785,
+      "grad_norm": 0.15090526640415192,
+      "learning_rate": 9.141366156260967e-06,
+      "loss": 0.0727,
+      "step": 1021
+    },
+    {
+      "epoch": 0.21421085726262837,
+      "grad_norm": 0.11032428592443466,
+      "learning_rate": 9.139462985399512e-06,
+      "loss": 0.0682,
+      "step": 1022
+    },
+    {
+      "epoch": 0.2144204569272689,
+      "grad_norm": 0.13925404846668243,
+      "learning_rate": 9.137557906246479e-06,
+      "loss": 0.0682,
+      "step": 1023
+    },
+    {
+      "epoch": 0.21463005659190945,
+      "grad_norm": 0.15173637866973877,
+      "learning_rate": 9.135650919680112e-06,
+      "loss": 0.0653,
+      "step": 1024
+    },
+    {
+      "epoch": 0.21483965625655,
+      "grad_norm": 0.1346423327922821,
+      "learning_rate": 9.133742026579528e-06,
+      "loss": 0.0679,
+      "step": 1025
+    },
+    {
+      "epoch": 0.21504925592119054,
+      "grad_norm": 0.1535898894071579,
+      "learning_rate": 9.13183122782473e-06,
+      "loss": 0.0667,
+      "step": 1026
+    },
+    {
+      "epoch": 0.21525885558583105,
+      "grad_norm": 0.14157144725322723,
+      "learning_rate": 9.129918524296596e-06,
+      "loss": 0.0707,
+      "step": 1027
+    },
+    {
+      "epoch": 0.2154684552504716,
+      "grad_norm": 0.14217810332775116,
+      "learning_rate": 9.128003916876878e-06,
+      "loss": 0.0708,
+      "step": 1028
+    },
+    {
+      "epoch": 0.21567805491511213,
+      "grad_norm": 0.1338115632534027,
+      "learning_rate": 9.126087406448211e-06,
+      "loss": 0.0648,
+      "step": 1029
+    },
+    {
+      "epoch": 0.21588765457975267,
+      "grad_norm": 0.12991099059581757,
+      "learning_rate": 9.124168993894107e-06,
+      "loss": 0.0677,
+      "step": 1030
+    },
+    {
+      "epoch": 0.21609725424439322,
+      "grad_norm": 0.1184862032532692,
+      "learning_rate": 9.122248680098956e-06,
+      "loss": 0.0673,
+      "step": 1031
+    },
+    {
+      "epoch": 0.21630685390903376,
+      "grad_norm": 0.1306854635477066,
+      "learning_rate": 9.120326465948016e-06,
+      "loss": 0.068,
+      "step": 1032
+    },
+    {
+      "epoch": 0.21651645357367427,
+      "grad_norm": 0.16064853966236115,
+      "learning_rate": 9.118402352327433e-06,
+      "loss": 0.0663,
+      "step": 1033
+    },
+    {
+      "epoch": 0.21672605323831481,
+      "grad_norm": 0.1781821846961975,
+      "learning_rate": 9.11647634012422e-06,
+      "loss": 0.0683,
+      "step": 1034
+    },
+    {
+      "epoch": 0.21693565290295536,
+      "grad_norm": 0.18965835869312286,
+      "learning_rate": 9.11454843022627e-06,
+      "loss": 0.0679,
+      "step": 1035
+    },
+    {
+      "epoch": 0.2171452525675959,
+      "grad_norm": 0.19845952093601227,
+      "learning_rate": 9.112618623522351e-06,
+      "loss": 0.0663,
+      "step": 1036
+    },
+    {
+      "epoch": 0.21735485223223644,
+      "grad_norm": 0.16511160135269165,
+      "learning_rate": 9.110686920902097e-06,
+      "loss": 0.0717,
+      "step": 1037
+    },
+    {
+      "epoch": 0.21756445189687695,
+      "grad_norm": 0.1469397395849228,
+      "learning_rate": 9.108753323256028e-06,
+      "loss": 0.0662,
+      "step": 1038
+    },
+    {
+      "epoch": 0.2177740515615175,
+      "grad_norm": 0.15417161583900452,
+      "learning_rate": 9.106817831475529e-06,
+      "loss": 0.0736,
+      "step": 1039
+    },
+    {
+      "epoch": 0.21798365122615804,
+      "grad_norm": 0.10822763293981552,
+      "learning_rate": 9.104880446452866e-06,
+      "loss": 0.0666,
+      "step": 1040
+    },
+    {
+      "epoch": 0.21819325089079858,
+      "grad_norm": 0.1328975260257721,
+      "learning_rate": 9.102941169081167e-06,
+      "loss": 0.0671,
+      "step": 1041
+    },
+    {
+      "epoch": 0.21840285055543912,
+      "grad_norm": 0.15660911798477173,
+      "learning_rate": 9.101000000254442e-06,
+      "loss": 0.0679,
+      "step": 1042
+    },
+    {
+      "epoch": 0.21861245022007963,
+      "grad_norm": 0.14654120802879333,
+      "learning_rate": 9.09905694086757e-06,
+      "loss": 0.0692,
+      "step": 1043
+    },
+    {
+      "epoch": 0.21882204988472018,
+      "grad_norm": 0.15261763334274292,
+      "learning_rate": 9.097111991816297e-06,
+      "loss": 0.0677,
+      "step": 1044
+    },
+    {
+      "epoch": 0.21903164954936072,
+      "grad_norm": 0.1555902063846588,
+      "learning_rate": 9.095165153997249e-06,
+      "loss": 0.0653,
+      "step": 1045
+    },
+    {
+      "epoch": 0.21924124921400126,
+      "grad_norm": 0.1320633888244629,
+      "learning_rate": 9.093216428307914e-06,
+      "loss": 0.0683,
+      "step": 1046
+    },
+    {
+      "epoch": 0.2194508488786418,
+      "grad_norm": 0.12523691356182098,
+      "learning_rate": 9.091265815646658e-06,
+      "loss": 0.0646,
+      "step": 1047
+    },
+    {
+      "epoch": 0.21966044854328234,
+      "grad_norm": 0.14748716354370117,
+      "learning_rate": 9.089313316912708e-06,
+      "loss": 0.067,
+      "step": 1048
+    },
+    {
+      "epoch": 0.21987004820792286,
+      "grad_norm": 0.13303905725479126,
+      "learning_rate": 9.08735893300617e-06,
+      "loss": 0.0677,
+      "step": 1049
+    },
+    {
+      "epoch": 0.2200796478725634,
+      "grad_norm": 0.1327192634344101,
+      "learning_rate": 9.085402664828013e-06,
+      "loss": 0.072,
+      "step": 1050
+    },
+    {
+      "epoch": 0.22028924753720394,
+      "grad_norm": 0.13275259733200073,
+      "learning_rate": 9.083444513280076e-06,
+      "loss": 0.0659,
+      "step": 1051
+    },
+    {
+      "epoch": 0.22049884720184448,
+      "grad_norm": 0.11244450509548187,
+      "learning_rate": 9.081484479265067e-06,
+      "loss": 0.0719,
+      "step": 1052
+    },
+    {
+      "epoch": 0.22070844686648503,
+      "grad_norm": 0.11435073614120483,
+      "learning_rate": 9.07952256368656e-06,
+      "loss": 0.0648,
+      "step": 1053
+    },
+    {
+      "epoch": 0.22091804653112554,
+      "grad_norm": 0.1512385904788971,
+      "learning_rate": 9.077558767448999e-06,
+      "loss": 0.0677,
+      "step": 1054
+    },
+    {
+      "epoch": 0.22112764619576608,
+      "grad_norm": 0.1435592770576477,
+      "learning_rate": 9.075593091457692e-06,
+      "loss": 0.064,
+      "step": 1055
+    },
+    {
+      "epoch": 0.22133724586040662,
+      "grad_norm": 0.14917264878749847,
+      "learning_rate": 9.073625536618819e-06,
+      "loss": 0.0661,
+      "step": 1056
+    },
+    {
+      "epoch": 0.22154684552504716,
+      "grad_norm": 0.17828944325447083,
+      "learning_rate": 9.071656103839419e-06,
+      "loss": 0.0683,
+      "step": 1057
+    },
+    {
+      "epoch": 0.2217564451896877,
+      "grad_norm": 0.13577091693878174,
+      "learning_rate": 9.069684794027401e-06,
+      "loss": 0.0659,
+      "step": 1058
+    },
+    {
+      "epoch": 0.22196604485432822,
+      "grad_norm": 0.1218876987695694,
+      "learning_rate": 9.067711608091536e-06,
+      "loss": 0.0655,
+      "step": 1059
+    },
+    {
+      "epoch": 0.22217564451896876,
+      "grad_norm": 0.14669421315193176,
+      "learning_rate": 9.065736546941467e-06,
+      "loss": 0.0669,
+      "step": 1060
+    },
+    {
+      "epoch": 0.2223852441836093,
+      "grad_norm": 0.12597538530826569,
+      "learning_rate": 9.063759611487693e-06,
+      "loss": 0.0673,
+      "step": 1061
+    },
+    {
+      "epoch": 0.22259484384824985,
+      "grad_norm": 0.13924571871757507,
+      "learning_rate": 9.061780802641582e-06,
+      "loss": 0.0628,
+      "step": 1062
+    },
+    {
+      "epoch": 0.2228044435128904,
+      "grad_norm": 0.14818163216114044,
+      "learning_rate": 9.059800121315365e-06,
+      "loss": 0.0656,
+      "step": 1063
+    },
+    {
+      "epoch": 0.22301404317753093,
+      "grad_norm": 0.13732635974884033,
+      "learning_rate": 9.057817568422135e-06,
+      "loss": 0.0666,
+      "step": 1064
+    },
+    {
+      "epoch": 0.22322364284217144,
+      "grad_norm": 0.1377662718296051,
+      "learning_rate": 9.05583314487585e-06,
+      "loss": 0.0695,
+      "step": 1065
+    },
+    {
+      "epoch": 0.22343324250681199,
+      "grad_norm": 0.14314699172973633,
+      "learning_rate": 9.053846851591328e-06,
+      "loss": 0.0689,
+      "step": 1066
+    },
+    {
+      "epoch": 0.22364284217145253,
+      "grad_norm": 0.16072338819503784,
+      "learning_rate": 9.05185868948425e-06,
+      "loss": 0.0672,
+      "step": 1067
+    },
+    {
+      "epoch": 0.22385244183609307,
+      "grad_norm": 0.1533009558916092,
+      "learning_rate": 9.049868659471156e-06,
+      "loss": 0.0703,
+      "step": 1068
+    },
+    {
+      "epoch": 0.2240620415007336,
+      "grad_norm": 0.16292600333690643,
+      "learning_rate": 9.047876762469451e-06,
+      "loss": 0.0701,
+      "step": 1069
+    },
+    {
+      "epoch": 0.22427164116537412,
+      "grad_norm": 0.19941456615924835,
+      "learning_rate": 9.0458829993974e-06,
+      "loss": 0.0683,
+      "step": 1070
+    },
+    {
+      "epoch": 0.22448124083001467,
+      "grad_norm": 0.22352056205272675,
+      "learning_rate": 9.043887371174128e-06,
+      "loss": 0.0661,
+      "step": 1071
+    },
+    {
+      "epoch": 0.2246908404946552,
+      "grad_norm": 0.24914206564426422,
+      "learning_rate": 9.041889878719617e-06,
+      "loss": 0.0674,
+      "step": 1072
+    },
+    {
+      "epoch": 0.22490044015929575,
+      "grad_norm": 0.21870644390583038,
+      "learning_rate": 9.03989052295471e-06,
+      "loss": 0.0669,
+      "step": 1073
+    },
+    {
+      "epoch": 0.2251100398239363,
+      "grad_norm": 0.16291193664073944,
+      "learning_rate": 9.037889304801112e-06,
+      "loss": 0.0678,
+      "step": 1074
+    },
+    {
+      "epoch": 0.2253196394885768,
+      "grad_norm": 0.12002480030059814,
+      "learning_rate": 9.035886225181384e-06,
+      "loss": 0.0663,
+      "step": 1075
+    },
+    {
+      "epoch": 0.22552923915321735,
+      "grad_norm": 0.12656140327453613,
+      "learning_rate": 9.033881285018945e-06,
+      "loss": 0.0649,
+      "step": 1076
+    },
+    {
+      "epoch": 0.2257388388178579,
+      "grad_norm": 0.17447209358215332,
+      "learning_rate": 9.031874485238068e-06,
+      "loss": 0.0653,
+      "step": 1077
+    },
+    {
+      "epoch": 0.22594843848249843,
+      "grad_norm": 0.1942858248949051,
+      "learning_rate": 9.029865826763895e-06,
+      "loss": 0.068,
+      "step": 1078
+    },
+    {
+      "epoch": 0.22615803814713897,
+      "grad_norm": 0.16547489166259766,
+      "learning_rate": 9.027855310522411e-06,
+      "loss": 0.0645,
+      "step": 1079
+    },
+    {
+      "epoch": 0.2263676378117795,
+      "grad_norm": 0.1284940242767334,
+      "learning_rate": 9.025842937440466e-06,
+      "loss": 0.0648,
+      "step": 1080
+    },
+    {
+      "epoch": 0.22657723747642003,
+      "grad_norm": 0.12198358029127121,
+      "learning_rate": 9.023828708445762e-06,
+      "loss": 0.0707,
+      "step": 1081
+    },
+    {
+      "epoch": 0.22678683714106057,
+      "grad_norm": 0.13251115381717682,
+      "learning_rate": 9.02181262446686e-06,
+      "loss": 0.069,
+      "step": 1082
+    },
+    {
+      "epoch": 0.2269964368057011,
+      "grad_norm": 0.13787177205085754,
+      "learning_rate": 9.019794686433174e-06,
+      "loss": 0.0696,
+      "step": 1083
+    },
+    {
+      "epoch": 0.22720603647034165,
+      "grad_norm": 0.12242651730775833,
+      "learning_rate": 9.017774895274971e-06,
+      "loss": 0.066,
+      "step": 1084
+    },
+    {
+      "epoch": 0.2274156361349822,
+      "grad_norm": 0.11319856345653534,
+      "learning_rate": 9.015753251923378e-06,
+      "loss": 0.0644,
+      "step": 1085
+    },
+    {
+      "epoch": 0.2276252357996227,
+      "grad_norm": 0.11255019158124924,
+      "learning_rate": 9.013729757310368e-06,
+      "loss": 0.0681,
+      "step": 1086
+    },
+    {
+      "epoch": 0.22783483546426325,
+      "grad_norm": 0.10680454969406128,
+      "learning_rate": 9.011704412368776e-06,
+      "loss": 0.0675,
+      "step": 1087
+    },
+    {
+      "epoch": 0.2280444351289038,
+      "grad_norm": 0.10743577778339386,
+      "learning_rate": 9.00967721803228e-06,
+      "loss": 0.0701,
+      "step": 1088
+    },
+    {
+      "epoch": 0.22825403479354434,
+      "grad_norm": 0.09966851025819778,
+      "learning_rate": 9.007648175235421e-06,
+      "loss": 0.0657,
+      "step": 1089
+    },
+    {
+      "epoch": 0.22846363445818488,
+      "grad_norm": 0.10842377692461014,
+      "learning_rate": 9.005617284913586e-06,
+      "loss": 0.0646,
+      "step": 1090
+    },
+    {
+      "epoch": 0.2286732341228254,
+      "grad_norm": 0.12540468573570251,
+      "learning_rate": 9.003584548003015e-06,
+      "loss": 0.0661,
+      "step": 1091
+    },
+    {
+      "epoch": 0.22888283378746593,
+      "grad_norm": 0.13884596526622772,
+      "learning_rate": 9.001549965440798e-06,
+      "loss": 0.0704,
+      "step": 1092
+    },
+    {
+      "epoch": 0.22909243345210648,
+      "grad_norm": 0.16879726946353912,
+      "learning_rate": 8.99951353816488e-06,
+      "loss": 0.0625,
+      "step": 1093
+    },
+    {
+      "epoch": 0.22930203311674702,
+      "grad_norm": 0.19034424424171448,
+      "learning_rate": 8.99747526711405e-06,
+      "loss": 0.0672,
+      "step": 1094
+    },
+    {
+      "epoch": 0.22951163278138756,
+      "grad_norm": 0.18537910282611847,
+      "learning_rate": 8.995435153227951e-06,
+      "loss": 0.0687,
+      "step": 1095
+    },
+    {
+      "epoch": 0.22972123244602807,
+      "grad_norm": 0.15541553497314453,
+      "learning_rate": 8.993393197447078e-06,
+      "loss": 0.0692,
+      "step": 1096
+    },
+    {
+      "epoch": 0.22993083211066861,
+      "grad_norm": 0.12911075353622437,
+      "learning_rate": 8.991349400712772e-06,
+      "loss": 0.0684,
+      "step": 1097
+    },
+    {
+      "epoch": 0.23014043177530916,
+      "grad_norm": 0.12497539818286896,
+      "learning_rate": 8.989303763967218e-06,
+      "loss": 0.0694,
+      "step": 1098
+    },
+    {
+      "epoch": 0.2303500314399497,
+      "grad_norm": 0.12200130522251129,
+      "learning_rate": 8.98725628815346e-06,
+      "loss": 0.0662,
+      "step": 1099
+    },
+    {
+      "epoch": 0.23055963110459024,
+      "grad_norm": 0.14163267612457275,
+      "learning_rate": 8.985206974215381e-06,
+      "loss": 0.0688,
+      "step": 1100
+    },
+    {
+      "epoch": 0.23076923076923078,
+      "grad_norm": 0.18513454496860504,
+      "learning_rate": 8.983155823097713e-06,
+      "loss": 0.0702,
+      "step": 1101
+    },
+    {
+      "epoch": 0.2309788304338713,
+      "grad_norm": 0.20042511820793152,
+      "learning_rate": 8.98110283574604e-06,
+      "loss": 0.0672,
+      "step": 1102
+    },
+    {
+      "epoch": 0.23118843009851184,
+      "grad_norm": 0.1693289577960968,
+      "learning_rate": 8.979048013106786e-06,
+      "loss": 0.0691,
+      "step": 1103
+    },
+    {
+      "epoch": 0.23139802976315238,
+      "grad_norm": 0.1325998604297638,
+      "learning_rate": 8.976991356127225e-06,
+      "loss": 0.0681,
+      "step": 1104
+    },
+    {
+      "epoch": 0.23160762942779292,
+      "grad_norm": 0.14559799432754517,
+      "learning_rate": 8.974932865755472e-06,
+      "loss": 0.069,
+      "step": 1105
+    },
+    {
+      "epoch": 0.23181722909243346,
+      "grad_norm": 0.1867041438817978,
+      "learning_rate": 8.972872542940496e-06,
+      "loss": 0.0691,
+      "step": 1106
+    },
+    {
+      "epoch": 0.23202682875707398,
+      "grad_norm": 0.18667063117027283,
+      "learning_rate": 8.970810388632102e-06,
+      "loss": 0.0693,
+      "step": 1107
+    },
+    {
+      "epoch": 0.23223642842171452,
+      "grad_norm": 0.149042010307312,
+      "learning_rate": 8.968746403780945e-06,
+      "loss": 0.0651,
+      "step": 1108
+    },
+    {
+      "epoch": 0.23244602808635506,
+      "grad_norm": 0.11694446206092834,
+      "learning_rate": 8.96668058933852e-06,
+      "loss": 0.0658,
+      "step": 1109
+    },
+    {
+      "epoch": 0.2326556277509956,
+      "grad_norm": 0.12780682742595673,
+      "learning_rate": 8.964612946257167e-06,
+      "loss": 0.0652,
+      "step": 1110
+    },
+    {
+      "epoch": 0.23286522741563614,
+      "grad_norm": 0.17143504321575165,
+      "learning_rate": 8.962543475490068e-06,
+      "loss": 0.0654,
+      "step": 1111
+    },
+    {
+      "epoch": 0.23307482708027666,
+      "grad_norm": 0.1847982108592987,
+      "learning_rate": 8.960472177991252e-06,
+      "loss": 0.0667,
+      "step": 1112
+    },
+    {
+      "epoch": 0.2332844267449172,
+      "grad_norm": 0.1703869104385376,
+      "learning_rate": 8.958399054715583e-06,
+      "loss": 0.0655,
+      "step": 1113
+    },
+    {
+      "epoch": 0.23349402640955774,
+      "grad_norm": 0.14903445541858673,
+      "learning_rate": 8.956324106618773e-06,
+      "loss": 0.066,
+      "step": 1114
+    },
+    {
+      "epoch": 0.23370362607419828,
+      "grad_norm": 0.13800345361232758,
+      "learning_rate": 8.954247334657371e-06,
+      "loss": 0.0662,
+      "step": 1115
+    },
+    {
+      "epoch": 0.23391322573883883,
+      "grad_norm": 0.14202521741390228,
+      "learning_rate": 8.952168739788769e-06,
+      "loss": 0.0662,
+      "step": 1116
+    },
+    {
+      "epoch": 0.23412282540347937,
+      "grad_norm": 0.1242147833108902,
+      "learning_rate": 8.9500883229712e-06,
+      "loss": 0.0659,
+      "step": 1117
+    },
+    {
+      "epoch": 0.23433242506811988,
+      "grad_norm": 0.13212472200393677,
+      "learning_rate": 8.948006085163735e-06,
+      "loss": 0.0659,
+      "step": 1118
+    },
+    {
+      "epoch": 0.23454202473276042,
+      "grad_norm": 0.13743017613887787,
+      "learning_rate": 8.945922027326283e-06,
+      "loss": 0.0653,
+      "step": 1119
+    },
+    {
+      "epoch": 0.23475162439740097,
+      "grad_norm": 0.12877334654331207,
+      "learning_rate": 8.943836150419596e-06,
+      "loss": 0.0652,
+      "step": 1120
+    },
+    {
+      "epoch": 0.2349612240620415,
+      "grad_norm": 0.13227277994155884,
+      "learning_rate": 8.941748455405264e-06,
+      "loss": 0.0677,
+      "step": 1121
+    },
+    {
+      "epoch": 0.23517082372668205,
+      "grad_norm": 0.1250903308391571,
+      "learning_rate": 8.939658943245712e-06,
+      "loss": 0.0655,
+      "step": 1122
+    },
+    {
+      "epoch": 0.23538042339132256,
+      "grad_norm": 0.10851810872554779,
+      "learning_rate": 8.937567614904205e-06,
+      "loss": 0.0631,
+      "step": 1123
+    },
+    {
+      "epoch": 0.2355900230559631,
+      "grad_norm": 0.13291695713996887,
+      "learning_rate": 8.935474471344848e-06,
+      "loss": 0.0655,
+      "step": 1124
+    },
+    {
+      "epoch": 0.23579962272060365,
+      "grad_norm": 0.16525736451148987,
+      "learning_rate": 8.933379513532575e-06,
+      "loss": 0.0637,
+      "step": 1125
+    },
+    {
+      "epoch": 0.2360092223852442,
+      "grad_norm": 0.20342209935188293,
+      "learning_rate": 8.931282742433163e-06,
+      "loss": 0.066,
+      "step": 1126
+    },
+    {
+      "epoch": 0.23621882204988473,
+      "grad_norm": 0.24097025394439697,
+      "learning_rate": 8.929184159013225e-06,
+      "loss": 0.0687,
+      "step": 1127
+    },
+    {
+      "epoch": 0.23642842171452524,
+      "grad_norm": 0.23975655436515808,
+      "learning_rate": 8.927083764240205e-06,
+      "loss": 0.0708,
+      "step": 1128
+    },
+    {
+      "epoch": 0.23663802137916579,
+      "grad_norm": 0.20488587021827698,
+      "learning_rate": 8.924981559082386e-06,
+      "loss": 0.0659,
+      "step": 1129
+    },
+    {
+      "epoch": 0.23684762104380633,
+      "grad_norm": 0.16722434759140015,
+      "learning_rate": 8.922877544508882e-06,
+      "loss": 0.0665,
+      "step": 1130
+    },
+    {
+      "epoch": 0.23705722070844687,
+      "grad_norm": 0.15572957694530487,
+      "learning_rate": 8.920771721489646e-06,
+      "loss": 0.0673,
+      "step": 1131
+    },
+    {
+      "epoch": 0.2372668203730874,
+      "grad_norm": 0.13812050223350525,
+      "learning_rate": 8.91866409099546e-06,
+      "loss": 0.0702,
+      "step": 1132
+    },
+    {
+      "epoch": 0.23747642003772795,
+      "grad_norm": 0.12852191925048828,
+      "learning_rate": 8.916554653997943e-06,
+      "loss": 0.0691,
+      "step": 1133
+    },
+    {
+      "epoch": 0.23768601970236847,
+      "grad_norm": 0.14980794489383698,
+      "learning_rate": 8.914443411469544e-06,
+      "loss": 0.0705,
+      "step": 1134
+    },
+    {
+      "epoch": 0.237895619367009,
+      "grad_norm": 0.15687119960784912,
+      "learning_rate": 8.912330364383546e-06,
+      "loss": 0.0716,
+      "step": 1135
+    },
+    {
+      "epoch": 0.23810521903164955,
+      "grad_norm": 0.13968421518802643,
+      "learning_rate": 8.91021551371406e-06,
+      "loss": 0.0679,
+      "step": 1136
+    },
+    {
+      "epoch": 0.2383148186962901,
+      "grad_norm": 0.1380040943622589,
+      "learning_rate": 8.908098860436036e-06,
+      "loss": 0.0624,
+      "step": 1137
+    },
+    {
+      "epoch": 0.23852441836093063,
+      "grad_norm": 0.13776816427707672,
+      "learning_rate": 8.90598040552525e-06,
+      "loss": 0.0655,
+      "step": 1138
+    },
+    {
+      "epoch": 0.23873401802557115,
+      "grad_norm": 0.11478232592344284,
+      "learning_rate": 8.903860149958308e-06,
+      "loss": 0.0681,
+      "step": 1139
+    },
+    {
+      "epoch": 0.2389436176902117,
+      "grad_norm": 0.1092422679066658,
+      "learning_rate": 8.901738094712648e-06,
+      "loss": 0.0639,
+      "step": 1140
+    },
+    {
+      "epoch": 0.23915321735485223,
+      "grad_norm": 0.11861050873994827,
+      "learning_rate": 8.899614240766537e-06,
+      "loss": 0.0646,
+      "step": 1141
+    },
+    {
+      "epoch": 0.23936281701949277,
+      "grad_norm": 0.11883353441953659,
+      "learning_rate": 8.89748858909907e-06,
+      "loss": 0.0654,
+      "step": 1142
+    },
+    {
+      "epoch": 0.23957241668413332,
+      "grad_norm": 0.11742465198040009,
+      "learning_rate": 8.895361140690173e-06,
+      "loss": 0.0655,
+      "step": 1143
+    },
+    {
+      "epoch": 0.23978201634877383,
+      "grad_norm": 0.1311412900686264,
+      "learning_rate": 8.8932318965206e-06,
+      "loss": 0.0668,
+      "step": 1144
+    },
+    {
+      "epoch": 0.23999161601341437,
+      "grad_norm": 0.16970883309841156,
+      "learning_rate": 8.89110085757193e-06,
+      "loss": 0.0681,
+      "step": 1145
+    },
+    {
+      "epoch": 0.2402012156780549,
+      "grad_norm": 0.18959131836891174,
+      "learning_rate": 8.888968024826575e-06,
+      "loss": 0.0645,
+      "step": 1146
+    },
+    {
+      "epoch": 0.24041081534269546,
+      "grad_norm": 0.17889679968357086,
+      "learning_rate": 8.886833399267767e-06,
+      "loss": 0.0666,
+      "step": 1147
+    },
+    {
+      "epoch": 0.240620415007336,
+      "grad_norm": 0.1588187962770462,
+      "learning_rate": 8.88469698187957e-06,
+      "loss": 0.0651,
+      "step": 1148
+    },
+    {
+      "epoch": 0.2408300146719765,
+      "grad_norm": 0.13770383596420288,
+      "learning_rate": 8.88255877364687e-06,
+      "loss": 0.0638,
+      "step": 1149
+    },
+    {
+      "epoch": 0.24103961433661705,
+      "grad_norm": 0.12436515837907791,
+      "learning_rate": 8.880418775555382e-06,
+      "loss": 0.0687,
+      "step": 1150
+    },
+    {
+      "epoch": 0.2412492140012576,
+      "grad_norm": 0.13518837094306946,
+      "learning_rate": 8.878276988591645e-06,
+      "loss": 0.071,
+      "step": 1151
+    },
+    {
+      "epoch": 0.24145881366589814,
+      "grad_norm": 0.14451143145561218,
+      "learning_rate": 8.876133413743023e-06,
+      "loss": 0.0668,
+      "step": 1152
+    },
+    {
+      "epoch": 0.24166841333053868,
+      "grad_norm": 0.13057959079742432,
+      "learning_rate": 8.873988051997702e-06,
+      "loss": 0.066,
+      "step": 1153
+    },
+    {
+      "epoch": 0.24187801299517922,
+      "grad_norm": 0.11131347715854645,
+      "learning_rate": 8.871840904344692e-06,
+      "loss": 0.0641,
+      "step": 1154
+    },
+    {
+      "epoch": 0.24208761265981973,
+      "grad_norm": 0.116900235414505,
+      "learning_rate": 8.86969197177383e-06,
+      "loss": 0.0639,
+      "step": 1155
+    },
+    {
+      "epoch": 0.24229721232446028,
+      "grad_norm": 0.12555626034736633,
+      "learning_rate": 8.867541255275774e-06,
+      "loss": 0.0681,
+      "step": 1156
+    },
+    {
+      "epoch": 0.24250681198910082,
+      "grad_norm": 0.12001737207174301,
+      "learning_rate": 8.865388755842002e-06,
+      "loss": 0.0634,
+      "step": 1157
+    },
+    {
+      "epoch": 0.24271641165374136,
+      "grad_norm": 0.12198159843683243,
+      "learning_rate": 8.863234474464817e-06,
+      "loss": 0.0657,
+      "step": 1158
+    },
+    {
+      "epoch": 0.2429260113183819,
+      "grad_norm": 0.14027060568332672,
+      "learning_rate": 8.86107841213734e-06,
+      "loss": 0.067,
+      "step": 1159
+    },
+    {
+      "epoch": 0.24313561098302242,
+      "grad_norm": 0.1360728144645691,
+      "learning_rate": 8.85892056985352e-06,
+      "loss": 0.0653,
+      "step": 1160
+    },
+    {
+      "epoch": 0.24334521064766296,
+      "grad_norm": 0.11649616807699203,
+      "learning_rate": 8.856760948608117e-06,
+      "loss": 0.0665,
+      "step": 1161
+    },
+    {
+      "epoch": 0.2435548103123035,
+      "grad_norm": 0.11606436967849731,
+      "learning_rate": 8.854599549396717e-06,
+      "loss": 0.0644,
+      "step": 1162
+    },
+    {
+      "epoch": 0.24376440997694404,
+      "grad_norm": 0.13204872608184814,
+      "learning_rate": 8.852436373215727e-06,
+      "loss": 0.0654,
+      "step": 1163
+    },
+    {
+      "epoch": 0.24397400964158458,
+      "grad_norm": 0.1417747139930725,
+      "learning_rate": 8.850271421062368e-06,
+      "loss": 0.0639,
+      "step": 1164
+    },
+    {
+      "epoch": 0.2441836093062251,
+      "grad_norm": 0.1318253129720688,
+      "learning_rate": 8.848104693934683e-06,
+      "loss": 0.065,
+      "step": 1165
+    },
+    {
+      "epoch": 0.24439320897086564,
+      "grad_norm": 0.11403854191303253,
+      "learning_rate": 8.845936192831536e-06,
+      "loss": 0.065,
+      "step": 1166
+    },
+    {
+      "epoch": 0.24460280863550618,
+      "grad_norm": 0.11362165957689285,
+      "learning_rate": 8.8437659187526e-06,
+      "loss": 0.063,
+      "step": 1167
+    },
+    {
+      "epoch": 0.24481240830014672,
+      "grad_norm": 0.11161333322525024,
+      "learning_rate": 8.841593872698377e-06,
+      "loss": 0.0663,
+      "step": 1168
+    },
+    {
+      "epoch": 0.24502200796478726,
+      "grad_norm": 0.11067195981740952,
+      "learning_rate": 8.839420055670175e-06,
+      "loss": 0.0646,
+      "step": 1169
+    },
+    {
+      "epoch": 0.2452316076294278,
+      "grad_norm": 0.13394485414028168,
+      "learning_rate": 8.837244468670126e-06,
+      "loss": 0.0637,
+      "step": 1170
+    },
+    {
+      "epoch": 0.24544120729406832,
+      "grad_norm": 0.1596195250749588,
+      "learning_rate": 8.835067112701172e-06,
+      "loss": 0.0679,
+      "step": 1171
+    },
+    {
+      "epoch": 0.24565080695870886,
+      "grad_norm": 0.1561581790447235,
+      "learning_rate": 8.832887988767076e-06,
+      "loss": 0.066,
+      "step": 1172
+    },
+    {
+      "epoch": 0.2458604066233494,
+      "grad_norm": 0.15718278288841248,
+      "learning_rate": 8.830707097872413e-06,
+      "loss": 0.0637,
+      "step": 1173
+    },
+    {
+      "epoch": 0.24607000628798995,
+      "grad_norm": 0.18930961191654205,
+      "learning_rate": 8.828524441022575e-06,
+      "loss": 0.0649,
+      "step": 1174
+    },
+    {
+      "epoch": 0.2462796059526305,
+      "grad_norm": 0.20854079723358154,
+      "learning_rate": 8.826340019223765e-06,
+      "loss": 0.0711,
+      "step": 1175
+    },
+    {
+      "epoch": 0.246489205617271,
+      "grad_norm": 0.17393162846565247,
+      "learning_rate": 8.824153833483001e-06,
+      "loss": 0.0639,
+      "step": 1176
+    },
+    {
+      "epoch": 0.24669880528191154,
+      "grad_norm": 0.12246709316968918,
+      "learning_rate": 8.821965884808112e-06,
+      "loss": 0.0641,
+      "step": 1177
+    },
+    {
+      "epoch": 0.24690840494655208,
+      "grad_norm": 0.11515267938375473,
+      "learning_rate": 8.819776174207746e-06,
+      "loss": 0.0629,
+      "step": 1178
+    },
+    {
+      "epoch": 0.24711800461119263,
+      "grad_norm": 0.14278475940227509,
+      "learning_rate": 8.817584702691358e-06,
+      "loss": 0.0679,
+      "step": 1179
+    },
+    {
+      "epoch": 0.24732760427583317,
+      "grad_norm": 0.17082667350769043,
+      "learning_rate": 8.815391471269212e-06,
+      "loss": 0.0655,
+      "step": 1180
+    },
+    {
+      "epoch": 0.24753720394047368,
+      "grad_norm": 0.17258518934249878,
+      "learning_rate": 8.813196480952393e-06,
+      "loss": 0.0674,
+      "step": 1181
+    },
+    {
+      "epoch": 0.24774680360511422,
+      "grad_norm": 0.15490064024925232,
+      "learning_rate": 8.810999732752788e-06,
+      "loss": 0.0649,
+      "step": 1182
+    },
+    {
+      "epoch": 0.24795640326975477,
+      "grad_norm": 0.1470610648393631,
+      "learning_rate": 8.808801227683095e-06,
+      "loss": 0.0647,
+      "step": 1183
+    },
+    {
+      "epoch": 0.2481660029343953,
+      "grad_norm": 0.14642740786075592,
+      "learning_rate": 8.80660096675683e-06,
+      "loss": 0.066,
+      "step": 1184
+    },
+    {
+      "epoch": 0.24837560259903585,
+      "grad_norm": 0.13396520912647247,
+      "learning_rate": 8.80439895098831e-06,
+      "loss": 0.0664,
+      "step": 1185
+    },
+    {
+      "epoch": 0.2485852022636764,
+      "grad_norm": 0.12936848402023315,
+      "learning_rate": 8.802195181392663e-06,
+      "loss": 0.0659,
+      "step": 1186
+    },
+    {
+      "epoch": 0.2487948019283169,
+      "grad_norm": 0.1439359188079834,
+      "learning_rate": 8.799989658985828e-06,
+      "loss": 0.0656,
+      "step": 1187
+    },
+    {
+      "epoch": 0.24900440159295745,
+      "grad_norm": 0.14836929738521576,
+      "learning_rate": 8.797782384784549e-06,
+      "loss": 0.0661,
+      "step": 1188
+    },
+    {
+      "epoch": 0.249214001257598,
+      "grad_norm": 0.12386839091777802,
+      "learning_rate": 8.79557335980638e-06,
+      "loss": 0.0694,
+      "step": 1189
+    },
+    {
+      "epoch": 0.24942360092223853,
+      "grad_norm": 0.1206989586353302,
+      "learning_rate": 8.793362585069677e-06,
+      "loss": 0.0644,
+      "step": 1190
+    },
+    {
+      "epoch": 0.24963320058687907,
+      "grad_norm": 0.13306036591529846,
+      "learning_rate": 8.791150061593615e-06,
+      "loss": 0.066,
+      "step": 1191
+    },
+    {
+      "epoch": 0.2498428002515196,
+      "grad_norm": 0.12486235052347183,
+      "learning_rate": 8.78893579039816e-06,
+      "loss": 0.0662,
+      "step": 1192
+    },
+    {
+      "epoch": 0.25005239991616013,
+      "grad_norm": 0.12596681714057922,
+      "learning_rate": 8.786719772504092e-06,
+      "loss": 0.0653,
+      "step": 1193
+    },
+    {
+      "epoch": 0.2502619995808007,
+      "grad_norm": 0.1434643715620041,
+      "learning_rate": 8.784502008932998e-06,
+      "loss": 0.0651,
+      "step": 1194
+    },
+    {
+      "epoch": 0.2504715992454412,
+      "grad_norm": 0.13905848562717438,
+      "learning_rate": 8.782282500707262e-06,
+      "loss": 0.0611,
+      "step": 1195
+    },
+    {
+      "epoch": 0.2506811989100817,
+      "grad_norm": 0.14193584024906158,
+      "learning_rate": 8.780061248850078e-06,
+      "loss": 0.064,
+      "step": 1196
+    },
+    {
+      "epoch": 0.2508907985747223,
+      "grad_norm": 0.15785372257232666,
+      "learning_rate": 8.777838254385444e-06,
+      "loss": 0.065,
+      "step": 1197
+    },
+    {
+      "epoch": 0.2511003982393628,
+      "grad_norm": 0.16986316442489624,
+      "learning_rate": 8.775613518338161e-06,
+      "loss": 0.0641,
+      "step": 1198
+    },
+    {
+      "epoch": 0.2513099979040034,
+      "grad_norm": 0.16595768928527832,
+      "learning_rate": 8.773387041733829e-06,
+      "loss": 0.0638,
+      "step": 1199
+    },
+    {
+      "epoch": 0.2515195975686439,
+      "grad_norm": 0.15224842727184296,
+      "learning_rate": 8.771158825598855e-06,
+      "loss": 0.0669,
+      "step": 1200
+    },
+    {
+      "epoch": 0.2517291972332844,
+      "grad_norm": 0.14295248687267303,
+      "learning_rate": 8.768928870960447e-06,
+      "loss": 0.0642,
+      "step": 1201
+    },
+    {
+      "epoch": 0.251938796897925,
+      "grad_norm": 0.1518334299325943,
+      "learning_rate": 8.766697178846611e-06,
+      "loss": 0.0665,
+      "step": 1202
+    },
+    {
+      "epoch": 0.2521483965625655,
+      "grad_norm": 0.15003719925880432,
+      "learning_rate": 8.764463750286158e-06,
+      "loss": 0.0665,
+      "step": 1203
+    },
+    {
+      "epoch": 0.25235799622720606,
+      "grad_norm": 0.13841992616653442,
+      "learning_rate": 8.762228586308697e-06,
+      "loss": 0.0639,
+      "step": 1204
+    },
+    {
+      "epoch": 0.2525675958918466,
+      "grad_norm": 0.13648587465286255,
+      "learning_rate": 8.75999168794464e-06,
+      "loss": 0.0656,
+      "step": 1205
+    },
+    {
+      "epoch": 0.2527771955564871,
+      "grad_norm": 0.1389138102531433,
+      "learning_rate": 8.757753056225197e-06,
+      "loss": 0.0641,
+      "step": 1206
+    },
+    {
+      "epoch": 0.25298679522112766,
+      "grad_norm": 0.15803200006484985,
+      "learning_rate": 8.755512692182376e-06,
+      "loss": 0.0641,
+      "step": 1207
+    },
+    {
+      "epoch": 0.2531963948857682,
+      "grad_norm": 0.17170608043670654,
+      "learning_rate": 8.753270596848982e-06,
+      "loss": 0.0642,
+      "step": 1208
+    },
+    {
+      "epoch": 0.25340599455040874,
+      "grad_norm": 0.1522701382637024,
+      "learning_rate": 8.751026771258622e-06,
+      "loss": 0.0679,
+      "step": 1209
+    },
+    {
+      "epoch": 0.25361559421504926,
+      "grad_norm": 0.137386754155159,
+      "learning_rate": 8.748781216445702e-06,
+      "loss": 0.0632,
+      "step": 1210
+    },
+    {
+      "epoch": 0.25382519387968977,
+      "grad_norm": 0.15736961364746094,
+      "learning_rate": 8.746533933445418e-06,
+      "loss": 0.0675,
+      "step": 1211
+    },
+    {
+      "epoch": 0.25403479354433034,
+      "grad_norm": 0.15517348051071167,
+      "learning_rate": 8.74428492329377e-06,
+      "loss": 0.0628,
+      "step": 1212
+    },
+    {
+      "epoch": 0.25424439320897085,
+      "grad_norm": 0.14377760887145996,
+      "learning_rate": 8.74203418702755e-06,
+      "loss": 0.0655,
+      "step": 1213
+    },
+    {
+      "epoch": 0.2544539928736114,
+      "grad_norm": 0.13686007261276245,
+      "learning_rate": 8.739781725684346e-06,
+      "loss": 0.0634,
+      "step": 1214
+    },
+    {
+      "epoch": 0.25466359253825194,
+      "grad_norm": 0.1474810093641281,
+      "learning_rate": 8.737527540302543e-06,
+      "loss": 0.0695,
+      "step": 1215
+    },
+    {
+      "epoch": 0.25487319220289245,
+      "grad_norm": 0.16937799751758575,
+      "learning_rate": 8.735271631921322e-06,
+      "loss": 0.0691,
+      "step": 1216
+    },
+    {
+      "epoch": 0.255082791867533,
+      "grad_norm": 0.1835523098707199,
+      "learning_rate": 8.733014001580656e-06,
+      "loss": 0.0642,
+      "step": 1217
+    },
+    {
+      "epoch": 0.25529239153217353,
+      "grad_norm": 0.1800583302974701,
+      "learning_rate": 8.730754650321307e-06,
+      "loss": 0.0675,
+      "step": 1218
+    },
+    {
+      "epoch": 0.2555019911968141,
+      "grad_norm": 0.14136351644992828,
+      "learning_rate": 8.728493579184841e-06,
+      "loss": 0.0669,
+      "step": 1219
+    },
+    {
+      "epoch": 0.2557115908614546,
+      "grad_norm": 0.1205543577671051,
+      "learning_rate": 8.72623078921361e-06,
+      "loss": 0.0643,
+      "step": 1220
+    },
+    {
+      "epoch": 0.25592119052609513,
+      "grad_norm": 0.13225074112415314,
+      "learning_rate": 8.723966281450758e-06,
+      "loss": 0.0618,
+      "step": 1221
+    },
+    {
+      "epoch": 0.2561307901907357,
+      "grad_norm": 0.18301644921302795,
+      "learning_rate": 8.721700056940224e-06,
+      "loss": 0.0663,
+      "step": 1222
+    },
+    {
+      "epoch": 0.2563403898553762,
+      "grad_norm": 0.1921045482158661,
+      "learning_rate": 8.719432116726738e-06,
+      "loss": 0.0687,
+      "step": 1223
+    },
+    {
+      "epoch": 0.2565499895200168,
+      "grad_norm": 0.15844593942165375,
+      "learning_rate": 8.717162461855817e-06,
+      "loss": 0.0646,
+      "step": 1224
+    },
+    {
+      "epoch": 0.2567595891846573,
+      "grad_norm": 0.16561459004878998,
+      "learning_rate": 8.714891093373774e-06,
+      "loss": 0.0662,
+      "step": 1225
+    },
+    {
+      "epoch": 0.2569691888492978,
+      "grad_norm": 0.1608954221010208,
+      "learning_rate": 8.712618012327709e-06,
+      "loss": 0.0681,
+      "step": 1226
+    },
+    {
+      "epoch": 0.2571787885139384,
+      "grad_norm": 0.1320895552635193,
+      "learning_rate": 8.710343219765512e-06,
+      "loss": 0.0708,
+      "step": 1227
+    },
+    {
+      "epoch": 0.2573883881785789,
+      "grad_norm": 0.12631681561470032,
+      "learning_rate": 8.70806671673586e-06,
+      "loss": 0.065,
+      "step": 1228
+    },
+    {
+      "epoch": 0.25759798784321947,
+      "grad_norm": 0.1299394816160202,
+      "learning_rate": 8.705788504288222e-06,
+      "loss": 0.0662,
+      "step": 1229
+    },
+    {
+      "epoch": 0.25780758750786,
+      "grad_norm": 0.14543578028678894,
+      "learning_rate": 8.703508583472855e-06,
+      "loss": 0.0628,
+      "step": 1230
+    },
+    {
+      "epoch": 0.25801718717250055,
+      "grad_norm": 0.1560242772102356,
+      "learning_rate": 8.701226955340797e-06,
+      "loss": 0.063,
+      "step": 1231
+    },
+    {
+      "epoch": 0.25822678683714106,
+      "grad_norm": 0.15224745869636536,
+      "learning_rate": 8.698943620943885e-06,
+      "loss": 0.0642,
+      "step": 1232
+    },
+    {
+      "epoch": 0.2584363865017816,
+      "grad_norm": 0.17773905396461487,
+      "learning_rate": 8.696658581334728e-06,
+      "loss": 0.0693,
+      "step": 1233
+    },
+    {
+      "epoch": 0.25864598616642215,
+      "grad_norm": 0.18498320877552032,
+      "learning_rate": 8.694371837566737e-06,
+      "loss": 0.0649,
+      "step": 1234
+    },
+    {
+      "epoch": 0.25885558583106266,
+      "grad_norm": 0.15482178330421448,
+      "learning_rate": 8.692083390694095e-06,
+      "loss": 0.0639,
+      "step": 1235
+    },
+    {
+      "epoch": 0.25906518549570323,
+      "grad_norm": 0.1407817006111145,
+      "learning_rate": 8.689793241771775e-06,
+      "loss": 0.0639,
+      "step": 1236
+    },
+    {
+      "epoch": 0.25927478516034375,
+      "grad_norm": 0.11920575052499771,
+      "learning_rate": 8.68750139185554e-06,
+      "loss": 0.0641,
+      "step": 1237
+    },
+    {
+      "epoch": 0.25948438482498426,
+      "grad_norm": 0.12093566358089447,
+      "learning_rate": 8.685207842001928e-06,
+      "loss": 0.0656,
+      "step": 1238
+    },
+    {
+      "epoch": 0.25969398448962483,
+      "grad_norm": 0.14436878263950348,
+      "learning_rate": 8.682912593268265e-06,
+      "loss": 0.0635,
+      "step": 1239
+    },
+    {
+      "epoch": 0.25990358415426534,
+      "grad_norm": 0.17225278913974762,
+      "learning_rate": 8.680615646712663e-06,
+      "loss": 0.0664,
+      "step": 1240
+    },
+    {
+      "epoch": 0.2601131838189059,
+      "grad_norm": 0.16033266484737396,
+      "learning_rate": 8.678317003394013e-06,
+      "loss": 0.0665,
+      "step": 1241
+    },
+    {
+      "epoch": 0.2603227834835464,
+      "grad_norm": 0.13092049956321716,
+      "learning_rate": 8.676016664371986e-06,
+      "loss": 0.0654,
+      "step": 1242
+    },
+    {
+      "epoch": 0.26053238314818694,
+      "grad_norm": 0.12013121694326401,
+      "learning_rate": 8.673714630707043e-06,
+      "loss": 0.0668,
+      "step": 1243
+    },
+    {
+      "epoch": 0.2607419828128275,
+      "grad_norm": 0.10177928954362869,
+      "learning_rate": 8.671410903460416e-06,
+      "loss": 0.0633,
+      "step": 1244
+    },
+    {
+      "epoch": 0.260951582477468,
+      "grad_norm": 0.11560986191034317,
+      "learning_rate": 8.669105483694126e-06,
+      "loss": 0.0653,
+      "step": 1245
+    },
+    {
+      "epoch": 0.2611611821421086,
+      "grad_norm": 0.1309051811695099,
+      "learning_rate": 8.666798372470971e-06,
+      "loss": 0.0649,
+      "step": 1246
+    },
+    {
+      "epoch": 0.2613707818067491,
+      "grad_norm": 0.13534289598464966,
+      "learning_rate": 8.664489570854526e-06,
+      "loss": 0.0646,
+      "step": 1247
+    },
+    {
+      "epoch": 0.2615803814713896,
+      "grad_norm": 0.13940168917179108,
+      "learning_rate": 8.662179079909152e-06,
+      "loss": 0.0652,
+      "step": 1248
+    },
+    {
+      "epoch": 0.2617899811360302,
+      "grad_norm": 0.16176725924015045,
+      "learning_rate": 8.659866900699983e-06,
+      "loss": 0.0652,
+      "step": 1249
+    },
+    {
+      "epoch": 0.2619995808006707,
+      "grad_norm": 0.18724969029426575,
+      "learning_rate": 8.657553034292932e-06,
+      "loss": 0.0638,
+      "step": 1250
+    },
+    {
+      "epoch": 0.2622091804653113,
+      "grad_norm": 0.198954775929451,
+      "learning_rate": 8.65523748175469e-06,
+      "loss": 0.0639,
+      "step": 1251
+    },
+    {
+      "epoch": 0.2624187801299518,
+      "grad_norm": 0.17343983054161072,
+      "learning_rate": 8.652920244152732e-06,
+      "loss": 0.0675,
+      "step": 1252
+    },
+    {
+      "epoch": 0.2626283797945923,
+      "grad_norm": 0.12653297185897827,
+      "learning_rate": 8.650601322555299e-06,
+      "loss": 0.0635,
+      "step": 1253
+    },
+    {
+      "epoch": 0.2628379794592329,
+      "grad_norm": 0.16065514087677002,
+      "learning_rate": 8.648280718031412e-06,
+      "loss": 0.0663,
+      "step": 1254
+    },
+    {
+      "epoch": 0.2630475791238734,
+      "grad_norm": 0.17681656777858734,
+      "learning_rate": 8.645958431650875e-06,
+      "loss": 0.0636,
+      "step": 1255
+    },
+    {
+      "epoch": 0.26325717878851396,
+      "grad_norm": 0.13592669367790222,
+      "learning_rate": 8.643634464484257e-06,
+      "loss": 0.0646,
+      "step": 1256
+    },
+    {
+      "epoch": 0.26346677845315447,
+      "grad_norm": 0.1315130740404129,
+      "learning_rate": 8.641308817602908e-06,
+      "loss": 0.0654,
+      "step": 1257
+    },
+    {
+      "epoch": 0.263676378117795,
+      "grad_norm": 0.1273263692855835,
+      "learning_rate": 8.63898149207895e-06,
+      "loss": 0.0675,
+      "step": 1258
+    },
+    {
+      "epoch": 0.26388597778243555,
+      "grad_norm": 0.10599275678396225,
+      "learning_rate": 8.636652488985282e-06,
+      "loss": 0.0647,
+      "step": 1259
+    },
+    {
+      "epoch": 0.26409557744707607,
+      "grad_norm": 0.11827687919139862,
+      "learning_rate": 8.634321809395569e-06,
+      "loss": 0.0642,
+      "step": 1260
+    },
+    {
+      "epoch": 0.26430517711171664,
+      "grad_norm": 0.14330358803272247,
+      "learning_rate": 8.631989454384258e-06,
+      "loss": 0.063,
+      "step": 1261
+    },
+    {
+      "epoch": 0.26451477677635715,
+      "grad_norm": 0.13524498045444489,
+      "learning_rate": 8.629655425026564e-06,
+      "loss": 0.0629,
+      "step": 1262
+    },
+    {
+      "epoch": 0.2647243764409977,
+      "grad_norm": 0.12704388797283173,
+      "learning_rate": 8.627319722398471e-06,
+      "loss": 0.062,
+      "step": 1263
+    },
+    {
+      "epoch": 0.26493397610563824,
+      "grad_norm": 0.13736125826835632,
+      "learning_rate": 8.624982347576741e-06,
+      "loss": 0.0634,
+      "step": 1264
+    },
+    {
+      "epoch": 0.26514357577027875,
+      "grad_norm": 0.12946844100952148,
+      "learning_rate": 8.622643301638902e-06,
+      "loss": 0.0662,
+      "step": 1265
+    },
+    {
+      "epoch": 0.2653531754349193,
+      "grad_norm": 0.1286906898021698,
+      "learning_rate": 8.620302585663252e-06,
+      "loss": 0.0663,
+      "step": 1266
+    },
+    {
+      "epoch": 0.26556277509955983,
+      "grad_norm": 0.12868361175060272,
+      "learning_rate": 8.617960200728863e-06,
+      "loss": 0.0593,
+      "step": 1267
+    },
+    {
+      "epoch": 0.2657723747642004,
+      "grad_norm": 0.12836746871471405,
+      "learning_rate": 8.615616147915573e-06,
+      "loss": 0.0656,
+      "step": 1268
+    },
+    {
+      "epoch": 0.2659819744288409,
+      "grad_norm": 0.14441126585006714,
+      "learning_rate": 8.613270428303991e-06,
+      "loss": 0.0686,
+      "step": 1269
+    },
+    {
+      "epoch": 0.26619157409348143,
+      "grad_norm": 0.14020347595214844,
+      "learning_rate": 8.61092304297549e-06,
+      "loss": 0.0651,
+      "step": 1270
+    },
+    {
+      "epoch": 0.266401173758122,
+      "grad_norm": 0.125614196062088,
+      "learning_rate": 8.608573993012217e-06,
+      "loss": 0.0661,
+      "step": 1271
+    },
+    {
+      "epoch": 0.2666107734227625,
+      "grad_norm": 0.14085769653320312,
+      "learning_rate": 8.606223279497081e-06,
+      "loss": 0.0643,
+      "step": 1272
+    },
+    {
+      "epoch": 0.2668203730874031,
+      "grad_norm": 0.14944879710674286,
+      "learning_rate": 8.603870903513765e-06,
+      "loss": 0.0641,
+      "step": 1273
+    },
+    {
+      "epoch": 0.2670299727520436,
+      "grad_norm": 0.13971124589443207,
+      "learning_rate": 8.601516866146711e-06,
+      "loss": 0.0644,
+      "step": 1274
+    },
+    {
+      "epoch": 0.2672395724166841,
+      "grad_norm": 0.1283096969127655,
+      "learning_rate": 8.599161168481127e-06,
+      "loss": 0.0622,
+      "step": 1275
+    },
+    {
+      "epoch": 0.2674491720813247,
+      "grad_norm": 0.13809433579444885,
+      "learning_rate": 8.596803811602994e-06,
+      "loss": 0.0658,
+      "step": 1276
+    },
+    {
+      "epoch": 0.2676587717459652,
+      "grad_norm": 0.13916701078414917,
+      "learning_rate": 8.594444796599051e-06,
+      "loss": 0.0636,
+      "step": 1277
+    },
+    {
+      "epoch": 0.26786837141060577,
+      "grad_norm": 0.1267821043729782,
+      "learning_rate": 8.592084124556803e-06,
+      "loss": 0.0632,
+      "step": 1278
+    },
+    {
+      "epoch": 0.2680779710752463,
+      "grad_norm": 0.14250284433364868,
+      "learning_rate": 8.589721796564521e-06,
+      "loss": 0.0617,
+      "step": 1279
+    },
+    {
+      "epoch": 0.2682875707398868,
+      "grad_norm": 0.14109215140342712,
+      "learning_rate": 8.587357813711234e-06,
+      "loss": 0.0637,
+      "step": 1280
+    },
+    {
+      "epoch": 0.26849717040452736,
+      "grad_norm": 0.14766204357147217,
+      "learning_rate": 8.584992177086742e-06,
+      "loss": 0.0653,
+      "step": 1281
+    },
+    {
+      "epoch": 0.2687067700691679,
+      "grad_norm": 0.1617199033498764,
+      "learning_rate": 8.5826248877816e-06,
+      "loss": 0.0659,
+      "step": 1282
+    },
+    {
+      "epoch": 0.26891636973380845,
+      "grad_norm": 0.17226214706897736,
+      "learning_rate": 8.580255946887129e-06,
+      "loss": 0.0618,
+      "step": 1283
+    },
+    {
+      "epoch": 0.26912596939844896,
+      "grad_norm": 0.1840105652809143,
+      "learning_rate": 8.577885355495412e-06,
+      "loss": 0.0641,
+      "step": 1284
+    },
+    {
+      "epoch": 0.2693355690630895,
+      "grad_norm": 0.1407589465379715,
+      "learning_rate": 8.575513114699288e-06,
+      "loss": 0.0621,
+      "step": 1285
+    },
+    {
+      "epoch": 0.26954516872773004,
+      "grad_norm": 0.14098970592021942,
+      "learning_rate": 8.57313922559236e-06,
+      "loss": 0.0632,
+      "step": 1286
+    },
+    {
+      "epoch": 0.26975476839237056,
+      "grad_norm": 0.14431780576705933,
+      "learning_rate": 8.57076368926899e-06,
+      "loss": 0.0636,
+      "step": 1287
+    },
+    {
+      "epoch": 0.26996436805701113,
+      "grad_norm": 0.11286340653896332,
+      "learning_rate": 8.568386506824304e-06,
+      "loss": 0.0634,
+      "step": 1288
+    },
+    {
+      "epoch": 0.27017396772165164,
+      "grad_norm": 0.13212832808494568,
+      "learning_rate": 8.566007679354178e-06,
+      "loss": 0.0621,
+      "step": 1289
+    },
+    {
+      "epoch": 0.27038356738629216,
+      "grad_norm": 0.1318579912185669,
+      "learning_rate": 8.563627207955255e-06,
+      "loss": 0.0629,
+      "step": 1290
+    },
+    {
+      "epoch": 0.2705931670509327,
+      "grad_norm": 0.12174471467733383,
+      "learning_rate": 8.561245093724926e-06,
+      "loss": 0.0654,
+      "step": 1291
+    },
+    {
+      "epoch": 0.27080276671557324,
+      "grad_norm": 0.12569588422775269,
+      "learning_rate": 8.558861337761349e-06,
+      "loss": 0.0644,
+      "step": 1292
+    },
+    {
+      "epoch": 0.2710123663802138,
+      "grad_norm": 0.09849688410758972,
+      "learning_rate": 8.556475941163436e-06,
+      "loss": 0.0645,
+      "step": 1293
+    },
+    {
+      "epoch": 0.2712219660448543,
+      "grad_norm": 0.11899732053279877,
+      "learning_rate": 8.554088905030852e-06,
+      "loss": 0.0658,
+      "step": 1294
+    },
+    {
+      "epoch": 0.27143156570949484,
+      "grad_norm": 0.12569987773895264,
+      "learning_rate": 8.551700230464022e-06,
+      "loss": 0.0658,
+      "step": 1295
+    },
+    {
+      "epoch": 0.2716411653741354,
+      "grad_norm": 0.13572193682193756,
+      "learning_rate": 8.549309918564122e-06,
+      "loss": 0.0599,
+      "step": 1296
+    },
+    {
+      "epoch": 0.2718507650387759,
+      "grad_norm": 0.16335296630859375,
+      "learning_rate": 8.546917970433087e-06,
+      "loss": 0.065,
+      "step": 1297
+    },
+    {
+      "epoch": 0.2720603647034165,
+      "grad_norm": 0.1542273610830307,
+      "learning_rate": 8.544524387173605e-06,
+      "loss": 0.0633,
+      "step": 1298
+    },
+    {
+      "epoch": 0.272269964368057,
+      "grad_norm": 0.18188713490962982,
+      "learning_rate": 8.542129169889117e-06,
+      "loss": 0.0615,
+      "step": 1299
+    },
+    {
+      "epoch": 0.2724795640326976,
+      "grad_norm": 0.18146541714668274,
+      "learning_rate": 8.539732319683817e-06,
+      "loss": 0.0629,
+      "step": 1300
+    },
+    {
+      "epoch": 0.2726891636973381,
+      "grad_norm": 0.16707837581634521,
+      "learning_rate": 8.537333837662653e-06,
+      "loss": 0.0646,
+      "step": 1301
+    },
+    {
+      "epoch": 0.2728987633619786,
+      "grad_norm": 0.143534317612648,
+      "learning_rate": 8.534933724931324e-06,
+      "loss": 0.0642,
+      "step": 1302
+    },
+    {
+      "epoch": 0.27310836302661917,
+      "grad_norm": 0.10924031585454941,
+      "learning_rate": 8.532531982596284e-06,
+      "loss": 0.0634,
+      "step": 1303
+    },
+    {
+      "epoch": 0.2733179626912597,
+      "grad_norm": 0.11168647557497025,
+      "learning_rate": 8.530128611764731e-06,
+      "loss": 0.0634,
+      "step": 1304
+    },
+    {
+      "epoch": 0.27352756235590026,
+      "grad_norm": 0.14308258891105652,
+      "learning_rate": 8.527723613544623e-06,
+      "loss": 0.0638,
+      "step": 1305
+    },
+    {
+      "epoch": 0.27373716202054077,
+      "grad_norm": 0.18562361598014832,
+      "learning_rate": 8.525316989044663e-06,
+      "loss": 0.0666,
+      "step": 1306
+    },
+    {
+      "epoch": 0.2739467616851813,
+      "grad_norm": 0.21448108553886414,
+      "learning_rate": 8.5229087393743e-06,
+      "loss": 0.0633,
+      "step": 1307
+    },
+    {
+      "epoch": 0.27415636134982185,
+      "grad_norm": 0.20579898357391357,
+      "learning_rate": 8.520498865643742e-06,
+      "loss": 0.0624,
+      "step": 1308
+    },
+    {
+      "epoch": 0.27436596101446237,
+      "grad_norm": 0.18763695657253265,
+      "learning_rate": 8.518087368963938e-06,
+      "loss": 0.066,
+      "step": 1309
+    },
+    {
+      "epoch": 0.27457556067910294,
+      "grad_norm": 0.15348191559314728,
+      "learning_rate": 8.515674250446588e-06,
+      "loss": 0.0644,
+      "step": 1310
+    },
+    {
+      "epoch": 0.27478516034374345,
+      "grad_norm": 0.11642424762248993,
+      "learning_rate": 8.51325951120414e-06,
+      "loss": 0.0639,
+      "step": 1311
+    },
+    {
+      "epoch": 0.27499476000838396,
+      "grad_norm": 0.15585097670555115,
+      "learning_rate": 8.510843152349786e-06,
+      "loss": 0.0624,
+      "step": 1312
+    },
+    {
+      "epoch": 0.27520435967302453,
+      "grad_norm": 0.18433383107185364,
+      "learning_rate": 8.508425174997467e-06,
+      "loss": 0.0648,
+      "step": 1313
+    },
+    {
+      "epoch": 0.27541395933766505,
+      "grad_norm": 0.1758815348148346,
+      "learning_rate": 8.506005580261872e-06,
+      "loss": 0.0672,
+      "step": 1314
+    },
+    {
+      "epoch": 0.2756235590023056,
+      "grad_norm": 0.11144900321960449,
+      "learning_rate": 8.503584369258434e-06,
+      "loss": 0.0606,
+      "step": 1315
+    },
+    {
+      "epoch": 0.27583315866694613,
+      "grad_norm": 0.11539231985807419,
+      "learning_rate": 8.501161543103327e-06,
+      "loss": 0.0634,
+      "step": 1316
+    },
+    {
+      "epoch": 0.27604275833158665,
+      "grad_norm": 0.15406975150108337,
+      "learning_rate": 8.498737102913476e-06,
+      "loss": 0.0628,
+      "step": 1317
+    },
+    {
+      "epoch": 0.2762523579962272,
+      "grad_norm": 0.16817119717597961,
+      "learning_rate": 8.496311049806549e-06,
+      "loss": 0.0657,
+      "step": 1318
+    },
+    {
+      "epoch": 0.27646195766086773,
+      "grad_norm": 0.1416136920452118,
+      "learning_rate": 8.493883384900953e-06,
+      "loss": 0.0644,
+      "step": 1319
+    },
+    {
+      "epoch": 0.2766715573255083,
+      "grad_norm": 0.10792358964681625,
+      "learning_rate": 8.491454109315844e-06,
+      "loss": 0.0654,
+      "step": 1320
+    },
+    {
+      "epoch": 0.2768811569901488,
+      "grad_norm": 0.1336362659931183,
+      "learning_rate": 8.489023224171114e-06,
+      "loss": 0.064,
+      "step": 1321
+    },
+    {
+      "epoch": 0.2770907566547893,
+      "grad_norm": 0.15568262338638306,
+      "learning_rate": 8.486590730587403e-06,
+      "loss": 0.0648,
+      "step": 1322
+    },
+    {
+      "epoch": 0.2773003563194299,
+      "grad_norm": 0.1655251383781433,
+      "learning_rate": 8.48415662968609e-06,
+      "loss": 0.0647,
+      "step": 1323
+    },
+    {
+      "epoch": 0.2775099559840704,
+      "grad_norm": 0.1461930274963379,
+      "learning_rate": 8.481720922589294e-06,
+      "loss": 0.0631,
+      "step": 1324
+    },
+    {
+      "epoch": 0.277719555648711,
+      "grad_norm": 0.13114267587661743,
+      "learning_rate": 8.479283610419876e-06,
+      "loss": 0.0632,
+      "step": 1325
+    },
+    {
+      "epoch": 0.2779291553133515,
+      "grad_norm": 0.12551239132881165,
+      "learning_rate": 8.476844694301437e-06,
+      "loss": 0.0617,
+      "step": 1326
+    },
+    {
+      "epoch": 0.278138754977992,
+      "grad_norm": 0.13701596856117249,
+      "learning_rate": 8.474404175358315e-06,
+      "loss": 0.063,
+      "step": 1327
+    },
+    {
+      "epoch": 0.2783483546426326,
+      "grad_norm": 0.140949547290802,
+      "learning_rate": 8.47196205471559e-06,
+      "loss": 0.0644,
+      "step": 1328
+    },
+    {
+      "epoch": 0.2785579543072731,
+      "grad_norm": 0.13582463562488556,
+      "learning_rate": 8.469518333499079e-06,
+      "loss": 0.064,
+      "step": 1329
+    },
+    {
+      "epoch": 0.27876755397191366,
+      "grad_norm": 0.14434821903705597,
+      "learning_rate": 8.467073012835338e-06,
+      "loss": 0.0583,
+      "step": 1330
+    },
+    {
+      "epoch": 0.2789771536365542,
+      "grad_norm": 0.14979779720306396,
+      "learning_rate": 8.464626093851657e-06,
+      "loss": 0.0643,
+      "step": 1331
+    },
+    {
+      "epoch": 0.27918675330119475,
+      "grad_norm": 0.15899671614170074,
+      "learning_rate": 8.462177577676066e-06,
+      "loss": 0.0639,
+      "step": 1332
+    },
+    {
+      "epoch": 0.27939635296583526,
+      "grad_norm": 0.16557057201862335,
+      "learning_rate": 8.459727465437332e-06,
+      "loss": 0.0674,
+      "step": 1333
+    },
+    {
+      "epoch": 0.2796059526304758,
+      "grad_norm": 0.16371293365955353,
+      "learning_rate": 8.457275758264956e-06,
+      "loss": 0.0625,
+      "step": 1334
+    },
+    {
+      "epoch": 0.27981555229511634,
+      "grad_norm": 0.16819259524345398,
+      "learning_rate": 8.45482245728917e-06,
+      "loss": 0.066,
+      "step": 1335
+    },
+    {
+      "epoch": 0.28002515195975686,
+      "grad_norm": 0.1529543697834015,
+      "learning_rate": 8.452367563640953e-06,
+      "loss": 0.0608,
+      "step": 1336
+    },
+    {
+      "epoch": 0.2802347516243974,
+      "grad_norm": 0.13105508685112,
+      "learning_rate": 8.449911078452004e-06,
+      "loss": 0.0615,
+      "step": 1337
+    },
+    {
+      "epoch": 0.28044435128903794,
+      "grad_norm": 0.13537244498729706,
+      "learning_rate": 8.447453002854763e-06,
+      "loss": 0.0608,
+      "step": 1338
+    },
+    {
+      "epoch": 0.28065395095367845,
+      "grad_norm": 0.13226556777954102,
+      "learning_rate": 8.444993337982408e-06,
+      "loss": 0.0627,
+      "step": 1339
+    },
+    {
+      "epoch": 0.280863550618319,
+      "grad_norm": 0.11894568800926208,
+      "learning_rate": 8.442532084968836e-06,
+      "loss": 0.0631,
+      "step": 1340
+    },
+    {
+      "epoch": 0.28107315028295954,
+      "grad_norm": 0.0991387814283371,
+      "learning_rate": 8.44006924494869e-06,
+      "loss": 0.0619,
+      "step": 1341
+    },
+    {
+      "epoch": 0.2812827499476001,
+      "grad_norm": 0.11959309130907059,
+      "learning_rate": 8.437604819057336e-06,
+      "loss": 0.0658,
+      "step": 1342
+    },
+    {
+      "epoch": 0.2814923496122406,
+      "grad_norm": 0.13812269270420074,
+      "learning_rate": 8.435138808430873e-06,
+      "loss": 0.0599,
+      "step": 1343
+    },
+    {
+      "epoch": 0.28170194927688114,
+      "grad_norm": 0.15743722021579742,
+      "learning_rate": 8.432671214206135e-06,
+      "loss": 0.0643,
+      "step": 1344
+    },
+    {
+      "epoch": 0.2819115489415217,
+      "grad_norm": 0.15632624924182892,
+      "learning_rate": 8.43020203752068e-06,
+      "loss": 0.0648,
+      "step": 1345
+    },
+    {
+      "epoch": 0.2821211486061622,
+      "grad_norm": 0.1336655467748642,
+      "learning_rate": 8.427731279512797e-06,
+      "loss": 0.0648,
+      "step": 1346
+    },
+    {
+      "epoch": 0.2823307482708028,
+      "grad_norm": 0.11537309736013412,
+      "learning_rate": 8.425258941321508e-06,
+      "loss": 0.0627,
+      "step": 1347
+    },
+    {
+      "epoch": 0.2825403479354433,
+      "grad_norm": 0.12339965254068375,
+      "learning_rate": 8.422785024086557e-06,
+      "loss": 0.0608,
+      "step": 1348
+    },
+    {
+      "epoch": 0.2827499476000838,
+      "grad_norm": 0.139318585395813,
+      "learning_rate": 8.420309528948422e-06,
+      "loss": 0.065,
+      "step": 1349
+    },
+    {
+      "epoch": 0.2829595472647244,
+      "grad_norm": 0.12725292146205902,
+      "learning_rate": 8.417832457048302e-06,
+      "loss": 0.0621,
+      "step": 1350
+    },
+    {
+      "epoch": 0.2831691469293649,
+      "grad_norm": 0.1099279448390007,
+      "learning_rate": 8.415353809528133e-06,
+      "loss": 0.0631,
+      "step": 1351
+    },
+    {
+      "epoch": 0.28337874659400547,
+      "grad_norm": 0.11662383377552032,
+      "learning_rate": 8.412873587530565e-06,
+      "loss": 0.0633,
+      "step": 1352
+    },
+    {
+      "epoch": 0.283588346258646,
+      "grad_norm": 0.13634788990020752,
+      "learning_rate": 8.410391792198982e-06,
+      "loss": 0.0673,
+      "step": 1353
+    },
+    {
+      "epoch": 0.2837979459232865,
+      "grad_norm": 0.1415703296661377,
+      "learning_rate": 8.407908424677493e-06,
+      "loss": 0.065,
+      "step": 1354
+    },
+    {
+      "epoch": 0.28400754558792707,
+      "grad_norm": 0.13151812553405762,
+      "learning_rate": 8.405423486110926e-06,
+      "loss": 0.0633,
+      "step": 1355
+    },
+    {
+      "epoch": 0.2842171452525676,
+      "grad_norm": 0.1383858174085617,
+      "learning_rate": 8.40293697764484e-06,
+      "loss": 0.0613,
+      "step": 1356
+    },
+    {
+      "epoch": 0.28442674491720815,
+      "grad_norm": 0.15346135199069977,
+      "learning_rate": 8.400448900425515e-06,
+      "loss": 0.0641,
+      "step": 1357
+    },
+    {
+      "epoch": 0.28463634458184867,
+      "grad_norm": 0.14821362495422363,
+      "learning_rate": 8.397959255599952e-06,
+      "loss": 0.0655,
+      "step": 1358
+    },
+    {
+      "epoch": 0.2848459442464892,
+      "grad_norm": 0.1301778107881546,
+      "learning_rate": 8.395468044315878e-06,
+      "loss": 0.063,
+      "step": 1359
+    },
+    {
+      "epoch": 0.28505554391112975,
+      "grad_norm": 0.13555464148521423,
+      "learning_rate": 8.392975267721742e-06,
+      "loss": 0.0635,
+      "step": 1360
+    },
+    {
+      "epoch": 0.28526514357577026,
+      "grad_norm": 0.1415833681821823,
+      "learning_rate": 8.39048092696671e-06,
+      "loss": 0.0608,
+      "step": 1361
+    },
+    {
+      "epoch": 0.28547474324041083,
+      "grad_norm": 0.1284692883491516,
+      "learning_rate": 8.387985023200677e-06,
+      "loss": 0.063,
+      "step": 1362
+    },
+    {
+      "epoch": 0.28568434290505135,
+      "grad_norm": 0.10931932926177979,
+      "learning_rate": 8.385487557574253e-06,
+      "loss": 0.0633,
+      "step": 1363
+    },
+    {
+      "epoch": 0.28589394256969186,
+      "grad_norm": 0.11410527676343918,
+      "learning_rate": 8.382988531238766e-06,
+      "loss": 0.0621,
+      "step": 1364
+    },
+    {
+      "epoch": 0.28610354223433243,
+      "grad_norm": 0.13259191811084747,
+      "learning_rate": 8.380487945346269e-06,
+      "loss": 0.062,
+      "step": 1365
+    },
+    {
+      "epoch": 0.28631314189897294,
+      "grad_norm": 0.1383572220802307,
+      "learning_rate": 8.377985801049533e-06,
+      "loss": 0.0611,
+      "step": 1366
+    },
+    {
+      "epoch": 0.2865227415636135,
+      "grad_norm": 0.13124366104602814,
+      "learning_rate": 8.375482099502043e-06,
+      "loss": 0.0654,
+      "step": 1367
+    },
+    {
+      "epoch": 0.28673234122825403,
+      "grad_norm": 0.1339644193649292,
+      "learning_rate": 8.372976841858007e-06,
+      "loss": 0.0616,
+      "step": 1368
+    },
+    {
+      "epoch": 0.2869419408928946,
+      "grad_norm": 0.14934468269348145,
+      "learning_rate": 8.370470029272348e-06,
+      "loss": 0.0616,
+      "step": 1369
+    },
+    {
+      "epoch": 0.2871515405575351,
+      "grad_norm": 0.15344421565532684,
+      "learning_rate": 8.367961662900704e-06,
+      "loss": 0.0625,
+      "step": 1370
+    },
+    {
+      "epoch": 0.2873611402221756,
+      "grad_norm": 0.14816948771476746,
+      "learning_rate": 8.365451743899433e-06,
+      "loss": 0.065,
+      "step": 1371
+    },
+    {
+      "epoch": 0.2875707398868162,
+      "grad_norm": 0.16822992265224457,
+      "learning_rate": 8.362940273425609e-06,
+      "loss": 0.0622,
+      "step": 1372
+    },
+    {
+      "epoch": 0.2877803395514567,
+      "grad_norm": 0.19093914330005646,
+      "learning_rate": 8.360427252637015e-06,
+      "loss": 0.0627,
+      "step": 1373
+    },
+    {
+      "epoch": 0.2879899392160973,
+      "grad_norm": 0.17542891204357147,
+      "learning_rate": 8.357912682692158e-06,
+      "loss": 0.0645,
+      "step": 1374
+    },
+    {
+      "epoch": 0.2881995388807378,
+      "grad_norm": 0.13324493169784546,
+      "learning_rate": 8.355396564750251e-06,
+      "loss": 0.0636,
+      "step": 1375
+    },
+    {
+      "epoch": 0.2884091385453783,
+      "grad_norm": 0.11579709500074387,
+      "learning_rate": 8.352878899971225e-06,
+      "loss": 0.064,
+      "step": 1376
+    },
+    {
+      "epoch": 0.2886187382100189,
+      "grad_norm": 0.1267646849155426,
+      "learning_rate": 8.35035968951572e-06,
+      "loss": 0.0654,
+      "step": 1377
+    },
+    {
+      "epoch": 0.2888283378746594,
+      "grad_norm": 0.13641488552093506,
+      "learning_rate": 8.347838934545097e-06,
+      "loss": 0.064,
+      "step": 1378
+    },
+    {
+      "epoch": 0.28903793753929996,
+      "grad_norm": 0.14326706528663635,
+      "learning_rate": 8.34531663622142e-06,
+      "loss": 0.0621,
+      "step": 1379
+    },
+    {
+      "epoch": 0.2892475372039405,
+      "grad_norm": 0.14493903517723083,
+      "learning_rate": 8.342792795707468e-06,
+      "loss": 0.0651,
+      "step": 1380
+    },
+    {
+      "epoch": 0.289457136868581,
+      "grad_norm": 0.1295025646686554,
+      "learning_rate": 8.340267414166731e-06,
+      "loss": 0.0628,
+      "step": 1381
+    },
+    {
+      "epoch": 0.28966673653322156,
+      "grad_norm": 0.11359550803899765,
+      "learning_rate": 8.337740492763412e-06,
+      "loss": 0.0608,
+      "step": 1382
+    },
+    {
+      "epoch": 0.2898763361978621,
+      "grad_norm": 0.1159205511212349,
+      "learning_rate": 8.33521203266242e-06,
+      "loss": 0.0629,
+      "step": 1383
+    },
+    {
+      "epoch": 0.29008593586250264,
+      "grad_norm": 0.1216607317328453,
+      "learning_rate": 8.33268203502937e-06,
+      "loss": 0.0619,
+      "step": 1384
+    },
+    {
+      "epoch": 0.29029553552714316,
+      "grad_norm": 0.12574604153633118,
+      "learning_rate": 8.330150501030597e-06,
+      "loss": 0.0611,
+      "step": 1385
+    },
+    {
+      "epoch": 0.29050513519178367,
+      "grad_norm": 0.13100169599056244,
+      "learning_rate": 8.327617431833132e-06,
+      "loss": 0.06,
+      "step": 1386
+    },
+    {
+      "epoch": 0.29071473485642424,
+      "grad_norm": 0.14436903595924377,
+      "learning_rate": 8.325082828604724e-06,
+      "loss": 0.0687,
+      "step": 1387
+    },
+    {
+      "epoch": 0.29092433452106475,
+      "grad_norm": 0.150770902633667,
+      "learning_rate": 8.322546692513822e-06,
+      "loss": 0.0639,
+      "step": 1388
+    },
+    {
+      "epoch": 0.2911339341857053,
+      "grad_norm": 0.13927258551120758,
+      "learning_rate": 8.320009024729586e-06,
+      "loss": 0.0638,
+      "step": 1389
+    },
+    {
+      "epoch": 0.29134353385034584,
+      "grad_norm": 0.1205689087510109,
+      "learning_rate": 8.317469826421877e-06,
+      "loss": 0.0614,
+      "step": 1390
+    },
+    {
+      "epoch": 0.29155313351498635,
+      "grad_norm": 0.12410090863704681,
+      "learning_rate": 8.314929098761268e-06,
+      "loss": 0.064,
+      "step": 1391
+    },
+    {
+      "epoch": 0.2917627331796269,
+      "grad_norm": 0.1389668583869934,
+      "learning_rate": 8.312386842919034e-06,
+      "loss": 0.0623,
+      "step": 1392
+    },
+    {
+      "epoch": 0.29197233284426743,
+      "grad_norm": 0.1345934271812439,
+      "learning_rate": 8.309843060067152e-06,
+      "loss": 0.0655,
+      "step": 1393
+    },
+    {
+      "epoch": 0.292181932508908,
+      "grad_norm": 0.13228191435337067,
+      "learning_rate": 8.307297751378309e-06,
+      "loss": 0.0644,
+      "step": 1394
+    },
+    {
+      "epoch": 0.2923915321735485,
+      "grad_norm": 0.1374116837978363,
+      "learning_rate": 8.304750918025888e-06,
+      "loss": 0.0639,
+      "step": 1395
+    },
+    {
+      "epoch": 0.29260113183818903,
+      "grad_norm": 0.13251402974128723,
+      "learning_rate": 8.30220256118398e-06,
+      "loss": 0.067,
+      "step": 1396
+    },
+    {
+      "epoch": 0.2928107315028296,
+      "grad_norm": 0.124875508248806,
+      "learning_rate": 8.29965268202738e-06,
+      "loss": 0.0629,
+      "step": 1397
+    },
+    {
+      "epoch": 0.2930203311674701,
+      "grad_norm": 0.12808595597743988,
+      "learning_rate": 8.297101281731576e-06,
+      "loss": 0.0595,
+      "step": 1398
+    },
+    {
+      "epoch": 0.2932299308321107,
+      "grad_norm": 0.13613183796405792,
+      "learning_rate": 8.294548361472767e-06,
+      "loss": 0.064,
+      "step": 1399
+    },
+    {
+      "epoch": 0.2934395304967512,
+      "grad_norm": 0.1279173195362091,
+      "learning_rate": 8.291993922427848e-06,
+      "loss": 0.0584,
+      "step": 1400
+    },
+    {
+      "epoch": 0.29364913016139177,
+      "grad_norm": 0.11164448410272598,
+      "learning_rate": 8.289437965774414e-06,
+      "loss": 0.063,
+      "step": 1401
+    },
+    {
+      "epoch": 0.2938587298260323,
+      "grad_norm": 0.11260772496461868,
+      "learning_rate": 8.286880492690761e-06,
+      "loss": 0.0582,
+      "step": 1402
+    },
+    {
+      "epoch": 0.2940683294906728,
+      "grad_norm": 0.12324314564466476,
+      "learning_rate": 8.284321504355884e-06,
+      "loss": 0.0635,
+      "step": 1403
+    },
+    {
+      "epoch": 0.29427792915531337,
+      "grad_norm": 0.12230779230594635,
+      "learning_rate": 8.281761001949474e-06,
+      "loss": 0.063,
+      "step": 1404
+    },
+    {
+      "epoch": 0.2944875288199539,
+      "grad_norm": 0.11499473452568054,
+      "learning_rate": 8.279198986651925e-06,
+      "loss": 0.0613,
+      "step": 1405
+    },
+    {
+      "epoch": 0.29469712848459445,
+      "grad_norm": 0.12175990641117096,
+      "learning_rate": 8.276635459644327e-06,
+      "loss": 0.0623,
+      "step": 1406
+    },
+    {
+      "epoch": 0.29490672814923496,
+      "grad_norm": 0.14893771708011627,
+      "learning_rate": 8.274070422108458e-06,
+      "loss": 0.0619,
+      "step": 1407
+    },
+    {
+      "epoch": 0.2951163278138755,
+      "grad_norm": 0.17682506144046783,
+      "learning_rate": 8.271503875226807e-06,
+      "loss": 0.0652,
+      "step": 1408
+    },
+    {
+      "epoch": 0.29532592747851605,
+      "grad_norm": 0.1828548014163971,
+      "learning_rate": 8.268935820182549e-06,
+      "loss": 0.0639,
+      "step": 1409
+    },
+    {
+      "epoch": 0.29553552714315656,
+      "grad_norm": 0.16200965642929077,
+      "learning_rate": 8.266366258159557e-06,
+      "loss": 0.0657,
+      "step": 1410
+    },
+    {
+      "epoch": 0.29574512680779713,
+      "grad_norm": 0.12557782232761383,
+      "learning_rate": 8.263795190342398e-06,
+      "loss": 0.062,
+      "step": 1411
+    },
+    {
+      "epoch": 0.29595472647243765,
+      "grad_norm": 0.11403919011354446,
+      "learning_rate": 8.261222617916335e-06,
+      "loss": 0.064,
+      "step": 1412
+    },
+    {
+      "epoch": 0.29616432613707816,
+      "grad_norm": 0.14523234963417053,
+      "learning_rate": 8.258648542067322e-06,
+      "loss": 0.0627,
+      "step": 1413
+    },
+    {
+      "epoch": 0.29637392580171873,
+      "grad_norm": 0.16806814074516296,
+      "learning_rate": 8.256072963982008e-06,
+      "loss": 0.0636,
+      "step": 1414
+    },
+    {
+      "epoch": 0.29658352546635924,
+      "grad_norm": 0.16369985044002533,
+      "learning_rate": 8.253495884847735e-06,
+      "loss": 0.0627,
+      "step": 1415
+    },
+    {
+      "epoch": 0.2967931251309998,
+      "grad_norm": 0.14731934666633606,
+      "learning_rate": 8.250917305852532e-06,
+      "loss": 0.0622,
+      "step": 1416
+    },
+    {
+      "epoch": 0.2970027247956403,
+      "grad_norm": 0.14016751945018768,
+      "learning_rate": 8.248337228185128e-06,
+      "loss": 0.0643,
+      "step": 1417
+    },
+    {
+      "epoch": 0.29721232446028084,
+      "grad_norm": 0.12805722653865814,
+      "learning_rate": 8.245755653034938e-06,
+      "loss": 0.0641,
+      "step": 1418
+    },
+    {
+      "epoch": 0.2974219241249214,
+      "grad_norm": 0.13014602661132812,
+      "learning_rate": 8.243172581592066e-06,
+      "loss": 0.0627,
+      "step": 1419
+    },
+    {
+      "epoch": 0.2976315237895619,
+      "grad_norm": 0.12117176502943039,
+      "learning_rate": 8.240588015047306e-06,
+      "loss": 0.0617,
+      "step": 1420
+    },
+    {
+      "epoch": 0.2978411234542025,
+      "grad_norm": 0.11306478828191757,
+      "learning_rate": 8.238001954592143e-06,
+      "loss": 0.0598,
+      "step": 1421
+    },
+    {
+      "epoch": 0.298050723118843,
+      "grad_norm": 0.1199759840965271,
+      "learning_rate": 8.235414401418754e-06,
+      "loss": 0.0624,
+      "step": 1422
+    },
+    {
+      "epoch": 0.2982603227834835,
+      "grad_norm": 0.12165199220180511,
+      "learning_rate": 8.232825356719998e-06,
+      "loss": 0.0645,
+      "step": 1423
+    },
+    {
+      "epoch": 0.2984699224481241,
+      "grad_norm": 0.1166660338640213,
+      "learning_rate": 8.230234821689423e-06,
+      "loss": 0.0657,
+      "step": 1424
+    },
+    {
+      "epoch": 0.2986795221127646,
+      "grad_norm": 0.11805491894483566,
+      "learning_rate": 8.227642797521265e-06,
+      "loss": 0.0615,
+      "step": 1425
+    },
+    {
+      "epoch": 0.2988891217774052,
+      "grad_norm": 0.11944872885942459,
+      "learning_rate": 8.22504928541045e-06,
+      "loss": 0.062,
+      "step": 1426
+    },
+    {
+      "epoch": 0.2990987214420457,
+      "grad_norm": 0.1310216188430786,
+      "learning_rate": 8.222454286552583e-06,
+      "loss": 0.0641,
+      "step": 1427
+    },
+    {
+      "epoch": 0.2993083211066862,
+      "grad_norm": 0.15212152898311615,
+      "learning_rate": 8.21985780214396e-06,
+      "loss": 0.0626,
+      "step": 1428
+    },
+    {
+      "epoch": 0.2995179207713268,
+      "grad_norm": 0.14432914555072784,
+      "learning_rate": 8.217259833381559e-06,
+      "loss": 0.0609,
+      "step": 1429
+    },
+    {
+      "epoch": 0.2997275204359673,
+      "grad_norm": 0.11263640224933624,
+      "learning_rate": 8.214660381463043e-06,
+      "loss": 0.0616,
+      "step": 1430
+    },
+    {
+      "epoch": 0.29993712010060786,
+      "grad_norm": 0.12339860200881958,
+      "learning_rate": 8.212059447586758e-06,
+      "loss": 0.0627,
+      "step": 1431
+    },
+    {
+      "epoch": 0.30014671976524837,
+      "grad_norm": 0.15246935188770294,
+      "learning_rate": 8.209457032951735e-06,
+      "loss": 0.0607,
+      "step": 1432
+    },
+    {
+      "epoch": 0.3003563194298889,
+      "grad_norm": 0.16209474205970764,
+      "learning_rate": 8.206853138757687e-06,
+      "loss": 0.0666,
+      "step": 1433
+    },
+    {
+      "epoch": 0.30056591909452945,
+      "grad_norm": 0.15916895866394043,
+      "learning_rate": 8.204247766205008e-06,
+      "loss": 0.0616,
+      "step": 1434
+    },
+    {
+      "epoch": 0.30077551875916997,
+      "grad_norm": 0.16150227189064026,
+      "learning_rate": 8.201640916494776e-06,
+      "loss": 0.0629,
+      "step": 1435
+    },
+    {
+      "epoch": 0.30098511842381054,
+      "grad_norm": 0.15138226747512817,
+      "learning_rate": 8.199032590828745e-06,
+      "loss": 0.0625,
+      "step": 1436
+    },
+    {
+      "epoch": 0.30119471808845105,
+      "grad_norm": 0.1358492225408554,
+      "learning_rate": 8.196422790409352e-06,
+      "loss": 0.0648,
+      "step": 1437
+    },
+    {
+      "epoch": 0.3014043177530916,
+      "grad_norm": 0.11498319357633591,
+      "learning_rate": 8.19381151643972e-06,
+      "loss": 0.0634,
+      "step": 1438
+    },
+    {
+      "epoch": 0.30161391741773214,
+      "grad_norm": 0.09225708991289139,
+      "learning_rate": 8.191198770123643e-06,
+      "loss": 0.0627,
+      "step": 1439
+    },
+    {
+      "epoch": 0.30182351708237265,
+      "grad_norm": 0.10581572353839874,
+      "learning_rate": 8.188584552665592e-06,
+      "loss": 0.0615,
+      "step": 1440
+    },
+    {
+      "epoch": 0.3020331167470132,
+      "grad_norm": 0.1425763964653015,
+      "learning_rate": 8.185968865270729e-06,
+      "loss": 0.0624,
+      "step": 1441
+    },
+    {
+      "epoch": 0.30224271641165373,
+      "grad_norm": 0.14036968350410461,
+      "learning_rate": 8.183351709144877e-06,
+      "loss": 0.0617,
+      "step": 1442
+    },
+    {
+      "epoch": 0.3024523160762943,
+      "grad_norm": 0.11378060281276703,
+      "learning_rate": 8.18073308549455e-06,
+      "loss": 0.0581,
+      "step": 1443
+    },
+    {
+      "epoch": 0.3026619157409348,
+      "grad_norm": 0.12382522225379944,
+      "learning_rate": 8.178112995526932e-06,
+      "loss": 0.0589,
+      "step": 1444
+    },
+    {
+      "epoch": 0.30287151540557533,
+      "grad_norm": 0.1571829468011856,
+      "learning_rate": 8.17549144044988e-06,
+      "loss": 0.0616,
+      "step": 1445
+    },
+    {
+      "epoch": 0.3030811150702159,
+      "grad_norm": 0.1604911983013153,
+      "learning_rate": 8.172868421471936e-06,
+      "loss": 0.0628,
+      "step": 1446
+    },
+    {
+      "epoch": 0.3032907147348564,
+      "grad_norm": 0.13987775146961212,
+      "learning_rate": 8.17024393980231e-06,
+      "loss": 0.0592,
+      "step": 1447
+    },
+    {
+      "epoch": 0.303500314399497,
+      "grad_norm": 0.1492827981710434,
+      "learning_rate": 8.167617996650885e-06,
+      "loss": 0.0579,
+      "step": 1448
+    },
+    {
+      "epoch": 0.3037099140641375,
+      "grad_norm": 0.14875125885009766,
+      "learning_rate": 8.164990593228222e-06,
+      "loss": 0.0636,
+      "step": 1449
+    },
+    {
+      "epoch": 0.303919513728778,
+      "grad_norm": 0.12761233747005463,
+      "learning_rate": 8.16236173074555e-06,
+      "loss": 0.0606,
+      "step": 1450
+    },
+    {
+      "epoch": 0.3041291133934186,
+      "grad_norm": 0.13052557408809662,
+      "learning_rate": 8.15973141041478e-06,
+      "loss": 0.0608,
+      "step": 1451
+    },
+    {
+      "epoch": 0.3043387130580591,
+      "grad_norm": 0.14425970613956451,
+      "learning_rate": 8.157099633448486e-06,
+      "loss": 0.0626,
+      "step": 1452
+    },
+    {
+      "epoch": 0.30454831272269967,
+      "grad_norm": 0.13814501464366913,
+      "learning_rate": 8.154466401059916e-06,
+      "loss": 0.0627,
+      "step": 1453
+    },
+    {
+      "epoch": 0.3047579123873402,
+      "grad_norm": 0.160516157746315,
+      "learning_rate": 8.15183171446299e-06,
+      "loss": 0.0588,
+      "step": 1454
+    },
+    {
+      "epoch": 0.3049675120519807,
+      "grad_norm": 0.19766630232334137,
+      "learning_rate": 8.1491955748723e-06,
+      "loss": 0.0662,
+      "step": 1455
+    },
+    {
+      "epoch": 0.30517711171662126,
+      "grad_norm": 0.17199066281318665,
+      "learning_rate": 8.146557983503103e-06,
+      "loss": 0.0651,
+      "step": 1456
+    },
+    {
+      "epoch": 0.3053867113812618,
+      "grad_norm": 0.13344770669937134,
+      "learning_rate": 8.143918941571329e-06,
+      "loss": 0.0614,
+      "step": 1457
+    },
+    {
+      "epoch": 0.30559631104590235,
+      "grad_norm": 0.12617316842079163,
+      "learning_rate": 8.141278450293576e-06,
+      "loss": 0.0616,
+      "step": 1458
+    },
+    {
+      "epoch": 0.30580591071054286,
+      "grad_norm": 0.12648488581180573,
+      "learning_rate": 8.13863651088711e-06,
+      "loss": 0.0619,
+      "step": 1459
+    },
+    {
+      "epoch": 0.3060155103751834,
+      "grad_norm": 0.1468046009540558,
+      "learning_rate": 8.135993124569865e-06,
+      "loss": 0.0636,
+      "step": 1460
+    },
+    {
+      "epoch": 0.30622511003982394,
+      "grad_norm": 0.16080905497074127,
+      "learning_rate": 8.133348292560442e-06,
+      "loss": 0.0615,
+      "step": 1461
+    },
+    {
+      "epoch": 0.30643470970446446,
+      "grad_norm": 0.15127602219581604,
+      "learning_rate": 8.130702016078105e-06,
+      "loss": 0.0616,
+      "step": 1462
+    },
+    {
+      "epoch": 0.30664430936910503,
+      "grad_norm": 0.12708480656147003,
+      "learning_rate": 8.12805429634279e-06,
+      "loss": 0.063,
+      "step": 1463
+    },
+    {
+      "epoch": 0.30685390903374554,
+      "grad_norm": 0.12862573564052582,
+      "learning_rate": 8.125405134575093e-06,
+      "loss": 0.0598,
+      "step": 1464
+    },
+    {
+      "epoch": 0.30706350869838606,
+      "grad_norm": 0.13088904321193695,
+      "learning_rate": 8.122754531996278e-06,
+      "loss": 0.0642,
+      "step": 1465
+    },
+    {
+      "epoch": 0.3072731083630266,
+      "grad_norm": 0.1227186918258667,
+      "learning_rate": 8.120102489828273e-06,
+      "loss": 0.0607,
+      "step": 1466
+    },
+    {
+      "epoch": 0.30748270802766714,
+      "grad_norm": 0.12489646673202515,
+      "learning_rate": 8.117449009293668e-06,
+      "loss": 0.058,
+      "step": 1467
+    },
+    {
+      "epoch": 0.3076923076923077,
+      "grad_norm": 0.13636116683483124,
+      "learning_rate": 8.114794091615718e-06,
+      "loss": 0.0638,
+      "step": 1468
+    },
+    {
+      "epoch": 0.3079019073569482,
+      "grad_norm": 0.15484750270843506,
+      "learning_rate": 8.11213773801834e-06,
+      "loss": 0.0628,
+      "step": 1469
+    },
+    {
+      "epoch": 0.3081115070215888,
+      "grad_norm": 0.15130093693733215,
+      "learning_rate": 8.109479949726109e-06,
+      "loss": 0.0613,
+      "step": 1470
+    },
+    {
+      "epoch": 0.3083211066862293,
+      "grad_norm": 0.13728772103786469,
+      "learning_rate": 8.106820727964267e-06,
+      "loss": 0.0608,
+      "step": 1471
+    },
+    {
+      "epoch": 0.3085307063508698,
+      "grad_norm": 0.13158975541591644,
+      "learning_rate": 8.104160073958716e-06,
+      "loss": 0.0656,
+      "step": 1472
+    },
+    {
+      "epoch": 0.3087403060155104,
+      "grad_norm": 0.14283715188503265,
+      "learning_rate": 8.101497988936015e-06,
+      "loss": 0.0615,
+      "step": 1473
+    },
+    {
+      "epoch": 0.3089499056801509,
+      "grad_norm": 0.14496344327926636,
+      "learning_rate": 8.098834474123385e-06,
+      "loss": 0.0608,
+      "step": 1474
+    },
+    {
+      "epoch": 0.3091595053447915,
+      "grad_norm": 0.11485762149095535,
+      "learning_rate": 8.096169530748708e-06,
+      "loss": 0.0628,
+      "step": 1475
+    },
+    {
+      "epoch": 0.309369105009432,
+      "grad_norm": 0.10256731510162354,
+      "learning_rate": 8.093503160040517e-06,
+      "loss": 0.0632,
+      "step": 1476
+    },
+    {
+      "epoch": 0.3095787046740725,
+      "grad_norm": 0.12671735882759094,
+      "learning_rate": 8.090835363228016e-06,
+      "loss": 0.063,
+      "step": 1477
+    },
+    {
+      "epoch": 0.30978830433871307,
+      "grad_norm": 0.1311434656381607,
+      "learning_rate": 8.088166141541052e-06,
+      "loss": 0.0607,
+      "step": 1478
+    },
+    {
+      "epoch": 0.3099979040033536,
+      "grad_norm": 0.12401635199785233,
+      "learning_rate": 8.08549549621014e-06,
+      "loss": 0.0606,
+      "step": 1479
+    },
+    {
+      "epoch": 0.31020750366799416,
+      "grad_norm": 0.12939530611038208,
+      "learning_rate": 8.082823428466442e-06,
+      "loss": 0.0625,
+      "step": 1480
+    },
+    {
+      "epoch": 0.31041710333263467,
+      "grad_norm": 0.14530347287654877,
+      "learning_rate": 8.080149939541786e-06,
+      "loss": 0.0619,
+      "step": 1481
+    },
+    {
+      "epoch": 0.3106267029972752,
+      "grad_norm": 0.16024935245513916,
+      "learning_rate": 8.077475030668647e-06,
+      "loss": 0.057,
+      "step": 1482
+    },
+    {
+      "epoch": 0.31083630266191575,
+      "grad_norm": 0.16392071545124054,
+      "learning_rate": 8.074798703080158e-06,
+      "loss": 0.0627,
+      "step": 1483
+    },
+    {
+      "epoch": 0.31104590232655627,
+      "grad_norm": 0.16095848381519318,
+      "learning_rate": 8.072120958010106e-06,
+      "loss": 0.0598,
+      "step": 1484
+    },
+    {
+      "epoch": 0.31125550199119684,
+      "grad_norm": 0.13986064493656158,
+      "learning_rate": 8.069441796692932e-06,
+      "loss": 0.0633,
+      "step": 1485
+    },
+    {
+      "epoch": 0.31146510165583735,
+      "grad_norm": 0.11697458475828171,
+      "learning_rate": 8.066761220363724e-06,
+      "loss": 0.0608,
+      "step": 1486
+    },
+    {
+      "epoch": 0.31167470132047786,
+      "grad_norm": 0.1427479237318039,
+      "learning_rate": 8.064079230258233e-06,
+      "loss": 0.0633,
+      "step": 1487
+    },
+    {
+      "epoch": 0.31188430098511843,
+      "grad_norm": 0.1725568026304245,
+      "learning_rate": 8.061395827612854e-06,
+      "loss": 0.0622,
+      "step": 1488
+    },
+    {
+      "epoch": 0.31209390064975895,
+      "grad_norm": 0.15745431184768677,
+      "learning_rate": 8.058711013664633e-06,
+      "loss": 0.0627,
+      "step": 1489
+    },
+    {
+      "epoch": 0.3123035003143995,
+      "grad_norm": 0.13785965740680695,
+      "learning_rate": 8.056024789651269e-06,
+      "loss": 0.0605,
+      "step": 1490
+    },
+    {
+      "epoch": 0.31251309997904003,
+      "grad_norm": 0.12854214012622833,
+      "learning_rate": 8.053337156811112e-06,
+      "loss": 0.061,
+      "step": 1491
+    },
+    {
+      "epoch": 0.31272269964368055,
+      "grad_norm": 0.13071073591709137,
+      "learning_rate": 8.050648116383162e-06,
+      "loss": 0.0609,
+      "step": 1492
+    },
+    {
+      "epoch": 0.3129322993083211,
+      "grad_norm": 0.13808673620224,
+      "learning_rate": 8.047957669607062e-06,
+      "loss": 0.0584,
+      "step": 1493
+    },
+    {
+      "epoch": 0.31314189897296163,
+      "grad_norm": 0.1332629919052124,
+      "learning_rate": 8.04526581772311e-06,
+      "loss": 0.0627,
+      "step": 1494
+    },
+    {
+      "epoch": 0.3133514986376022,
+      "grad_norm": 0.11996429413557053,
+      "learning_rate": 8.042572561972249e-06,
+      "loss": 0.0623,
+      "step": 1495
+    },
+    {
+      "epoch": 0.3135610983022427,
+      "grad_norm": 0.1371801495552063,
+      "learning_rate": 8.039877903596069e-06,
+      "loss": 0.0623,
+      "step": 1496
+    },
+    {
+      "epoch": 0.3137706979668832,
+      "grad_norm": 0.1538374274969101,
+      "learning_rate": 8.037181843836807e-06,
+      "loss": 0.0595,
+      "step": 1497
+    },
+    {
+      "epoch": 0.3139802976315238,
+      "grad_norm": 0.09957927465438843,
+      "learning_rate": 8.034484383937345e-06,
+      "loss": 0.0629,
+      "step": 1498
+    },
+    {
+      "epoch": 0.3141898972961643,
+      "grad_norm": 0.12047518789768219,
+      "learning_rate": 8.031785525141214e-06,
+      "loss": 0.0626,
+      "step": 1499
+    },
+    {
+      "epoch": 0.3143994969608049,
+      "grad_norm": 0.17362119257450104,
+      "learning_rate": 8.029085268692584e-06,
+      "loss": 0.0599,
+      "step": 1500
+    },
+    {
+      "epoch": 0.3146090966254454,
+      "grad_norm": 0.12966394424438477,
+      "learning_rate": 8.026383615836273e-06,
+      "loss": 0.0631,
+      "step": 1501
+    },
+    {
+      "epoch": 0.3148186962900859,
+      "grad_norm": 0.16443343460559845,
+      "learning_rate": 8.023680567817746e-06,
+      "loss": 0.0623,
+      "step": 1502
+    },
+    {
+      "epoch": 0.3150282959547265,
+      "grad_norm": 0.18140558898448944,
+      "learning_rate": 8.020976125883105e-06,
+      "loss": 0.0622,
+      "step": 1503
+    },
+    {
+      "epoch": 0.315237895619367,
+      "grad_norm": 0.1121053621172905,
+      "learning_rate": 8.018270291279098e-06,
+      "loss": 0.0598,
+      "step": 1504
+    },
+    {
+      "epoch": 0.31544749528400756,
+      "grad_norm": 0.1866942197084427,
+      "learning_rate": 8.015563065253113e-06,
+      "loss": 0.0623,
+      "step": 1505
+    },
+    {
+      "epoch": 0.3156570949486481,
+      "grad_norm": 0.1441410928964615,
+      "learning_rate": 8.01285444905318e-06,
+      "loss": 0.0621,
+      "step": 1506
+    },
+    {
+      "epoch": 0.31586669461328865,
+      "grad_norm": 0.13300150632858276,
+      "learning_rate": 8.010144443927974e-06,
+      "loss": 0.0604,
+      "step": 1507
+    },
+    {
+      "epoch": 0.31607629427792916,
+      "grad_norm": 0.17902185022830963,
+      "learning_rate": 8.007433051126805e-06,
+      "loss": 0.0619,
+      "step": 1508
+    },
+    {
+      "epoch": 0.3162858939425697,
+      "grad_norm": 0.12487073242664337,
+      "learning_rate": 8.004720271899622e-06,
+      "loss": 0.063,
+      "step": 1509
+    },
+    {
+      "epoch": 0.31649549360721024,
+      "grad_norm": 0.14495615661144257,
+      "learning_rate": 8.002006107497018e-06,
+      "loss": 0.0617,
+      "step": 1510
+    },
+    {
+      "epoch": 0.31670509327185076,
+      "grad_norm": 0.14461486041545868,
+      "learning_rate": 7.999290559170222e-06,
+      "loss": 0.0651,
+      "step": 1511
+    },
+    {
+      "epoch": 0.3169146929364913,
+      "grad_norm": 0.16186927258968353,
+      "learning_rate": 7.996573628171103e-06,
+      "loss": 0.061,
+      "step": 1512
+    },
+    {
+      "epoch": 0.31712429260113184,
+      "grad_norm": 0.20015843212604523,
+      "learning_rate": 7.993855315752163e-06,
+      "loss": 0.063,
+      "step": 1513
+    },
+    {
+      "epoch": 0.31733389226577235,
+      "grad_norm": 0.1886589378118515,
+      "learning_rate": 7.991135623166543e-06,
+      "loss": 0.064,
+      "step": 1514
+    },
+    {
+      "epoch": 0.3175434919304129,
+      "grad_norm": 0.15665331482887268,
+      "learning_rate": 7.988414551668025e-06,
+      "loss": 0.0644,
+      "step": 1515
+    },
+    {
+      "epoch": 0.31775309159505344,
+      "grad_norm": 0.13003768026828766,
+      "learning_rate": 7.985692102511018e-06,
+      "loss": 0.0632,
+      "step": 1516
+    },
+    {
+      "epoch": 0.317962691259694,
+      "grad_norm": 0.14635911583900452,
+      "learning_rate": 7.982968276950568e-06,
+      "loss": 0.0614,
+      "step": 1517
+    },
+    {
+      "epoch": 0.3181722909243345,
+      "grad_norm": 0.13516655564308167,
+      "learning_rate": 7.980243076242367e-06,
+      "loss": 0.0647,
+      "step": 1518
+    },
+    {
+      "epoch": 0.31838189058897504,
+      "grad_norm": 0.15535247325897217,
+      "learning_rate": 7.977516501642725e-06,
+      "loss": 0.0612,
+      "step": 1519
+    },
+    {
+      "epoch": 0.3185914902536156,
+      "grad_norm": 0.16718165576457977,
+      "learning_rate": 7.974788554408594e-06,
+      "loss": 0.0611,
+      "step": 1520
+    },
+    {
+      "epoch": 0.3188010899182561,
+      "grad_norm": 0.14838221669197083,
+      "learning_rate": 7.97205923579756e-06,
+      "loss": 0.0602,
+      "step": 1521
+    },
+    {
+      "epoch": 0.3190106895828967,
+      "grad_norm": 0.15352647006511688,
+      "learning_rate": 7.969328547067832e-06,
+      "loss": 0.0625,
+      "step": 1522
+    },
+    {
+      "epoch": 0.3192202892475372,
+      "grad_norm": 0.1140708476305008,
+      "learning_rate": 7.966596489478261e-06,
+      "loss": 0.059,
+      "step": 1523
+    },
+    {
+      "epoch": 0.3194298889121777,
+      "grad_norm": 0.11593464761972427,
+      "learning_rate": 7.963863064288326e-06,
+      "loss": 0.0605,
+      "step": 1524
+    },
+    {
+      "epoch": 0.3196394885768183,
+      "grad_norm": 0.11878301203250885,
+      "learning_rate": 7.961128272758133e-06,
+      "loss": 0.0617,
+      "step": 1525
+    },
+    {
+      "epoch": 0.3198490882414588,
+      "grad_norm": 0.10293014347553253,
+      "learning_rate": 7.958392116148424e-06,
+      "loss": 0.0601,
+      "step": 1526
+    },
+    {
+      "epoch": 0.32005868790609937,
+      "grad_norm": 0.12255340069532394,
+      "learning_rate": 7.95565459572056e-06,
+      "loss": 0.0616,
+      "step": 1527
+    },
+    {
+      "epoch": 0.3202682875707399,
+      "grad_norm": 0.1113123744726181,
+      "learning_rate": 7.952915712736545e-06,
+      "loss": 0.0629,
+      "step": 1528
+    },
+    {
+      "epoch": 0.3204778872353804,
+      "grad_norm": 0.11932878196239471,
+      "learning_rate": 7.950175468458999e-06,
+      "loss": 0.0632,
+      "step": 1529
+    },
+    {
+      "epoch": 0.32068748690002097,
+      "grad_norm": 0.13146136701107025,
+      "learning_rate": 7.947433864151173e-06,
+      "loss": 0.0624,
+      "step": 1530
+    },
+    {
+      "epoch": 0.3208970865646615,
+      "grad_norm": 0.11554694920778275,
+      "learning_rate": 7.944690901076949e-06,
+      "loss": 0.0591,
+      "step": 1531
+    },
+    {
+      "epoch": 0.32110668622930205,
+      "grad_norm": 0.11757069826126099,
+      "learning_rate": 7.94194658050083e-06,
+      "loss": 0.0659,
+      "step": 1532
+    },
+    {
+      "epoch": 0.32131628589394257,
+      "grad_norm": 0.12100815027952194,
+      "learning_rate": 7.93920090368795e-06,
+      "loss": 0.0579,
+      "step": 1533
+    },
+    {
+      "epoch": 0.3215258855585831,
+      "grad_norm": 0.12287240475416183,
+      "learning_rate": 7.936453871904065e-06,
+      "loss": 0.0601,
+      "step": 1534
+    },
+    {
+      "epoch": 0.32173548522322365,
+      "grad_norm": 0.12194748222827911,
+      "learning_rate": 7.933705486415553e-06,
+      "loss": 0.06,
+      "step": 1535
+    },
+    {
+      "epoch": 0.32194508488786416,
+      "grad_norm": 0.11489801853895187,
+      "learning_rate": 7.93095574848942e-06,
+      "loss": 0.0594,
+      "step": 1536
+    },
+    {
+      "epoch": 0.32215468455250473,
+      "grad_norm": 0.11319497227668762,
+      "learning_rate": 7.928204659393297e-06,
+      "loss": 0.06,
+      "step": 1537
+    },
+    {
+      "epoch": 0.32236428421714525,
+      "grad_norm": 0.10197950154542923,
+      "learning_rate": 7.925452220395436e-06,
+      "loss": 0.064,
+      "step": 1538
+    },
+    {
+      "epoch": 0.3225738838817858,
+      "grad_norm": 0.10284919291734695,
+      "learning_rate": 7.922698432764709e-06,
+      "loss": 0.0616,
+      "step": 1539
+    },
+    {
+      "epoch": 0.32278348354642633,
+      "grad_norm": 0.10016728937625885,
+      "learning_rate": 7.919943297770609e-06,
+      "loss": 0.0588,
+      "step": 1540
+    },
+    {
+      "epoch": 0.32299308321106684,
+      "grad_norm": 0.10762354731559753,
+      "learning_rate": 7.917186816683256e-06,
+      "loss": 0.0587,
+      "step": 1541
+    },
+    {
+      "epoch": 0.3232026828757074,
+      "grad_norm": 0.13142406940460205,
+      "learning_rate": 7.914428990773388e-06,
+      "loss": 0.0594,
+      "step": 1542
+    },
+    {
+      "epoch": 0.32341228254034793,
+      "grad_norm": 0.15556885302066803,
+      "learning_rate": 7.91166982131236e-06,
+      "loss": 0.0594,
+      "step": 1543
+    },
+    {
+      "epoch": 0.3236218822049885,
+      "grad_norm": 0.18388265371322632,
+      "learning_rate": 7.908909309572147e-06,
+      "loss": 0.0606,
+      "step": 1544
+    },
+    {
+      "epoch": 0.323831481869629,
+      "grad_norm": 0.21343399584293365,
+      "learning_rate": 7.906147456825349e-06,
+      "loss": 0.0634,
+      "step": 1545
+    },
+    {
+      "epoch": 0.3240410815342695,
+      "grad_norm": 0.21992023289203644,
+      "learning_rate": 7.903384264345177e-06,
+      "loss": 0.0635,
+      "step": 1546
+    },
+    {
+      "epoch": 0.3242506811989101,
+      "grad_norm": 0.16700875759124756,
+      "learning_rate": 7.900619733405462e-06,
+      "loss": 0.0641,
+      "step": 1547
+    },
+    {
+      "epoch": 0.3244602808635506,
+      "grad_norm": 0.09242475032806396,
+      "learning_rate": 7.897853865280652e-06,
+      "loss": 0.0614,
+      "step": 1548
+    },
+    {
+      "epoch": 0.3246698805281912,
+      "grad_norm": 0.12979759275913239,
+      "learning_rate": 7.895086661245811e-06,
+      "loss": 0.061,
+      "step": 1549
+    },
+    {
+      "epoch": 0.3248794801928317,
+      "grad_norm": 0.19285419583320618,
+      "learning_rate": 7.892318122576623e-06,
+      "loss": 0.0602,
+      "step": 1550
+    },
+    {
+      "epoch": 0.3250890798574722,
+      "grad_norm": 0.1928529143333435,
+      "learning_rate": 7.889548250549379e-06,
+      "loss": 0.0613,
+      "step": 1551
+    },
+    {
+      "epoch": 0.3252986795221128,
+      "grad_norm": 0.13492649793624878,
+      "learning_rate": 7.886777046440993e-06,
+      "loss": 0.061,
+      "step": 1552
+    },
+    {
+      "epoch": 0.3255082791867533,
+      "grad_norm": 0.1183929294347763,
+      "learning_rate": 7.884004511528988e-06,
+      "loss": 0.0637,
+      "step": 1553
+    },
+    {
+      "epoch": 0.32571787885139386,
+      "grad_norm": 0.17243815958499908,
+      "learning_rate": 7.881230647091502e-06,
+      "loss": 0.0601,
+      "step": 1554
+    },
+    {
+      "epoch": 0.3259274785160344,
+      "grad_norm": 0.1707419902086258,
+      "learning_rate": 7.878455454407285e-06,
+      "loss": 0.0625,
+      "step": 1555
+    },
+    {
+      "epoch": 0.3261370781806749,
+      "grad_norm": 0.14786763489246368,
+      "learning_rate": 7.875678934755704e-06,
+      "loss": 0.0632,
+      "step": 1556
+    },
+    {
+      "epoch": 0.32634667784531546,
+      "grad_norm": 0.1466890424489975,
+      "learning_rate": 7.87290108941673e-06,
+      "loss": 0.0613,
+      "step": 1557
+    },
+    {
+      "epoch": 0.326556277509956,
+      "grad_norm": 0.10979172587394714,
+      "learning_rate": 7.87012191967095e-06,
+      "loss": 0.063,
+      "step": 1558
+    },
+    {
+      "epoch": 0.32676587717459654,
+      "grad_norm": 0.11911406368017197,
+      "learning_rate": 7.867341426799562e-06,
+      "loss": 0.0567,
+      "step": 1559
+    },
+    {
+      "epoch": 0.32697547683923706,
+      "grad_norm": 0.14850041270256042,
+      "learning_rate": 7.864559612084372e-06,
+      "loss": 0.0609,
+      "step": 1560
+    },
+    {
+      "epoch": 0.32718507650387757,
+      "grad_norm": 0.13936297595500946,
+      "learning_rate": 7.861776476807795e-06,
+      "loss": 0.0613,
+      "step": 1561
+    },
+    {
+      "epoch": 0.32739467616851814,
+      "grad_norm": 0.12747827172279358,
+      "learning_rate": 7.858992022252859e-06,
+      "loss": 0.0626,
+      "step": 1562
+    },
+    {
+      "epoch": 0.32760427583315865,
+      "grad_norm": 0.11281277984380722,
+      "learning_rate": 7.856206249703191e-06,
+      "loss": 0.0612,
+      "step": 1563
+    },
+    {
+      "epoch": 0.3278138754977992,
+      "grad_norm": 0.11081743985414505,
+      "learning_rate": 7.853419160443038e-06,
+      "loss": 0.0625,
+      "step": 1564
+    },
+    {
+      "epoch": 0.32802347516243974,
+      "grad_norm": 0.10656673461198807,
+      "learning_rate": 7.850630755757242e-06,
+      "loss": 0.0625,
+      "step": 1565
+    },
+    {
+      "epoch": 0.32823307482708025,
+      "grad_norm": 0.09804133325815201,
+      "learning_rate": 7.847841036931263e-06,
+      "loss": 0.0602,
+      "step": 1566
+    },
+    {
+      "epoch": 0.3284426744917208,
+      "grad_norm": 0.10339375585317612,
+      "learning_rate": 7.845050005251156e-06,
+      "loss": 0.06,
+      "step": 1567
+    },
+    {
+      "epoch": 0.32865227415636133,
+      "grad_norm": 0.09550289064645767,
+      "learning_rate": 7.842257662003587e-06,
+      "loss": 0.0571,
+      "step": 1568
+    },
+    {
+      "epoch": 0.3288618738210019,
+      "grad_norm": 0.09345971792936325,
+      "learning_rate": 7.839464008475825e-06,
+      "loss": 0.0604,
+      "step": 1569
+    },
+    {
+      "epoch": 0.3290714734856424,
+      "grad_norm": 0.10545200109481812,
+      "learning_rate": 7.836669045955746e-06,
+      "loss": 0.0598,
+      "step": 1570
+    },
+    {
+      "epoch": 0.32928107315028293,
+      "grad_norm": 0.11077643185853958,
+      "learning_rate": 7.83387277573183e-06,
+      "loss": 0.0586,
+      "step": 1571
+    },
+    {
+      "epoch": 0.3294906728149235,
+      "grad_norm": 0.12594513595104218,
+      "learning_rate": 7.831075199093148e-06,
+      "loss": 0.0598,
+      "step": 1572
+    },
+    {
+      "epoch": 0.329700272479564,
+      "grad_norm": 0.13642063736915588,
+      "learning_rate": 7.828276317329388e-06,
+      "loss": 0.0601,
+      "step": 1573
+    },
+    {
+      "epoch": 0.3299098721442046,
+      "grad_norm": 0.13808242976665497,
+      "learning_rate": 7.825476131730836e-06,
+      "loss": 0.0591,
+      "step": 1574
+    },
+    {
+      "epoch": 0.3301194718088451,
+      "grad_norm": 0.14252924919128418,
+      "learning_rate": 7.822674643588372e-06,
+      "loss": 0.0589,
+      "step": 1575
+    },
+    {
+      "epoch": 0.33032907147348567,
+      "grad_norm": 0.12825094163417816,
+      "learning_rate": 7.819871854193484e-06,
+      "loss": 0.0622,
+      "step": 1576
+    },
+    {
+      "epoch": 0.3305386711381262,
+      "grad_norm": 0.10428497940301895,
+      "learning_rate": 7.817067764838257e-06,
+      "loss": 0.0583,
+      "step": 1577
+    },
+    {
+      "epoch": 0.3307482708027667,
+      "grad_norm": 0.10690061748027802,
+      "learning_rate": 7.814262376815375e-06,
+      "loss": 0.0581,
+      "step": 1578
+    },
+    {
+      "epoch": 0.33095787046740727,
+      "grad_norm": 0.11123620718717575,
+      "learning_rate": 7.811455691418123e-06,
+      "loss": 0.0607,
+      "step": 1579
+    },
+    {
+      "epoch": 0.3311674701320478,
+      "grad_norm": 0.10212855786085129,
+      "learning_rate": 7.80864770994038e-06,
+      "loss": 0.0624,
+      "step": 1580
+    },
+    {
+      "epoch": 0.33137706979668835,
+      "grad_norm": 0.09154222905635834,
+      "learning_rate": 7.805838433676627e-06,
+      "loss": 0.0604,
+      "step": 1581
+    },
+    {
+      "epoch": 0.33158666946132886,
+      "grad_norm": 0.09208472818136215,
+      "learning_rate": 7.803027863921939e-06,
+      "loss": 0.06,
+      "step": 1582
+    },
+    {
+      "epoch": 0.3317962691259694,
+      "grad_norm": 0.09515499323606491,
+      "learning_rate": 7.800216001971988e-06,
+      "loss": 0.0654,
+      "step": 1583
+    },
+    {
+      "epoch": 0.33200586879060995,
+      "grad_norm": 0.09495844691991806,
+      "learning_rate": 7.79740284912304e-06,
+      "loss": 0.063,
+      "step": 1584
+    },
+    {
+      "epoch": 0.33221546845525046,
+      "grad_norm": 0.09243495017290115,
+      "learning_rate": 7.794588406671962e-06,
+      "loss": 0.0622,
+      "step": 1585
+    },
+    {
+      "epoch": 0.33242506811989103,
+      "grad_norm": 0.08537401258945465,
+      "learning_rate": 7.791772675916207e-06,
+      "loss": 0.0601,
+      "step": 1586
+    },
+    {
+      "epoch": 0.33263466778453155,
+      "grad_norm": 0.08266492187976837,
+      "learning_rate": 7.788955658153829e-06,
+      "loss": 0.0582,
+      "step": 1587
+    },
+    {
+      "epoch": 0.33284426744917206,
+      "grad_norm": 0.10280580073595047,
+      "learning_rate": 7.786137354683472e-06,
+      "loss": 0.0598,
+      "step": 1588
+    },
+    {
+      "epoch": 0.33305386711381263,
+      "grad_norm": 0.13502821326255798,
+      "learning_rate": 7.783317766804375e-06,
+      "loss": 0.0586,
+      "step": 1589
+    },
+    {
+      "epoch": 0.33326346677845314,
+      "grad_norm": 0.16894817352294922,
+      "learning_rate": 7.780496895816363e-06,
+      "loss": 0.0596,
+      "step": 1590
+    },
+    {
+      "epoch": 0.3334730664430937,
+      "grad_norm": 0.21050219237804413,
+      "learning_rate": 7.77767474301986e-06,
+      "loss": 0.0609,
+      "step": 1591
+    },
+    {
+      "epoch": 0.3336826661077342,
+      "grad_norm": 0.203968346118927,
+      "learning_rate": 7.774851309715878e-06,
+      "loss": 0.0625,
+      "step": 1592
+    },
+    {
+      "epoch": 0.33389226577237474,
+      "grad_norm": 0.13803130388259888,
+      "learning_rate": 7.77202659720602e-06,
+      "loss": 0.057,
+      "step": 1593
+    },
+    {
+      "epoch": 0.3341018654370153,
+      "grad_norm": 0.13375061750411987,
+      "learning_rate": 7.769200606792476e-06,
+      "loss": 0.057,
+      "step": 1594
+    },
+    {
+      "epoch": 0.3343114651016558,
+      "grad_norm": 0.12815599143505096,
+      "learning_rate": 7.766373339778026e-06,
+      "loss": 0.0587,
+      "step": 1595
+    },
+    {
+      "epoch": 0.3345210647662964,
+      "grad_norm": 0.14430293440818787,
+      "learning_rate": 7.763544797466041e-06,
+      "loss": 0.0586,
+      "step": 1596
+    },
+    {
+      "epoch": 0.3347306644309369,
+      "grad_norm": 0.15994097292423248,
+      "learning_rate": 7.76071498116048e-06,
+      "loss": 0.0569,
+      "step": 1597
+    },
+    {
+      "epoch": 0.3349402640955774,
+      "grad_norm": 0.12301892787218094,
+      "learning_rate": 7.757883892165886e-06,
+      "loss": 0.0596,
+      "step": 1598
+    },
+    {
+      "epoch": 0.335149863760218,
+      "grad_norm": 0.12102188915014267,
+      "learning_rate": 7.755051531787388e-06,
+      "loss": 0.0604,
+      "step": 1599
+    },
+    {
+      "epoch": 0.3353594634248585,
+      "grad_norm": 0.12119577080011368,
+      "learning_rate": 7.752217901330707e-06,
+      "loss": 0.0602,
+      "step": 1600
+    },
+    {
+      "epoch": 0.3355690630894991,
+      "grad_norm": 0.11957602202892303,
+      "learning_rate": 7.749383002102147e-06,
+      "loss": 0.0608,
+      "step": 1601
+    },
+    {
+      "epoch": 0.3357786627541396,
+      "grad_norm": 0.1391220986843109,
+      "learning_rate": 7.746546835408593e-06,
+      "loss": 0.0592,
+      "step": 1602
+    },
+    {
+      "epoch": 0.3359882624187801,
+      "grad_norm": 0.14518260955810547,
+      "learning_rate": 7.74370940255752e-06,
+      "loss": 0.0605,
+      "step": 1603
+    },
+    {
+      "epoch": 0.3361978620834207,
+      "grad_norm": 0.15618950128555298,
+      "learning_rate": 7.74087070485698e-06,
+      "loss": 0.0604,
+      "step": 1604
+    },
+    {
+      "epoch": 0.3364074617480612,
+      "grad_norm": 0.13505420088768005,
+      "learning_rate": 7.738030743615615e-06,
+      "loss": 0.0605,
+      "step": 1605
+    },
+    {
+      "epoch": 0.33661706141270176,
+      "grad_norm": 0.11345727741718292,
+      "learning_rate": 7.735189520142645e-06,
+      "loss": 0.0627,
+      "step": 1606
+    },
+    {
+      "epoch": 0.33682666107734227,
+      "grad_norm": 0.10767614841461182,
+      "learning_rate": 7.732347035747878e-06,
+      "loss": 0.0631,
+      "step": 1607
+    },
+    {
+      "epoch": 0.33703626074198284,
+      "grad_norm": 0.11709927022457123,
+      "learning_rate": 7.72950329174169e-06,
+      "loss": 0.0615,
+      "step": 1608
+    },
+    {
+      "epoch": 0.33724586040662335,
+      "grad_norm": 0.13534541428089142,
+      "learning_rate": 7.726658289435055e-06,
+      "loss": 0.0617,
+      "step": 1609
+    },
+    {
+      "epoch": 0.33745546007126387,
+      "grad_norm": 0.12872527539730072,
+      "learning_rate": 7.723812030139514e-06,
+      "loss": 0.0574,
+      "step": 1610
+    },
+    {
+      "epoch": 0.33766505973590444,
+      "grad_norm": 0.12165652960538864,
+      "learning_rate": 7.720964515167193e-06,
+      "loss": 0.0573,
+      "step": 1611
+    },
+    {
+      "epoch": 0.33787465940054495,
+      "grad_norm": 0.10215598344802856,
+      "learning_rate": 7.718115745830797e-06,
+      "loss": 0.0579,
+      "step": 1612
+    },
+    {
+      "epoch": 0.3380842590651855,
+      "grad_norm": 0.09414192289113998,
+      "learning_rate": 7.715265723443606e-06,
+      "loss": 0.0607,
+      "step": 1613
+    },
+    {
+      "epoch": 0.33829385872982604,
+      "grad_norm": 0.11345679312944412,
+      "learning_rate": 7.712414449319478e-06,
+      "loss": 0.0591,
+      "step": 1614
+    },
+    {
+      "epoch": 0.33850345839446655,
+      "grad_norm": 0.11734048277139664,
+      "learning_rate": 7.709561924772855e-06,
+      "loss": 0.0595,
+      "step": 1615
+    },
+    {
+      "epoch": 0.3387130580591071,
+      "grad_norm": 0.10740841180086136,
+      "learning_rate": 7.706708151118747e-06,
+      "loss": 0.0593,
+      "step": 1616
+    },
+    {
+      "epoch": 0.33892265772374763,
+      "grad_norm": 0.1052752360701561,
+      "learning_rate": 7.703853129672742e-06,
+      "loss": 0.0576,
+      "step": 1617
+    },
+    {
+      "epoch": 0.3391322573883882,
+      "grad_norm": 0.11419695615768433,
+      "learning_rate": 7.700996861751009e-06,
+      "loss": 0.0632,
+      "step": 1618
+    },
+    {
+      "epoch": 0.3393418570530287,
+      "grad_norm": 0.10033425688743591,
+      "learning_rate": 7.698139348670281e-06,
+      "loss": 0.061,
+      "step": 1619
+    },
+    {
+      "epoch": 0.33955145671766923,
+      "grad_norm": 0.10025940090417862,
+      "learning_rate": 7.695280591747875e-06,
+      "loss": 0.0609,
+      "step": 1620
+    },
+    {
+      "epoch": 0.3397610563823098,
+      "grad_norm": 0.09737148880958557,
+      "learning_rate": 7.692420592301675e-06,
+      "loss": 0.0609,
+      "step": 1621
+    },
+    {
+      "epoch": 0.3399706560469503,
+      "grad_norm": 0.09889432042837143,
+      "learning_rate": 7.689559351650142e-06,
+      "loss": 0.0598,
+      "step": 1622
+    },
+    {
+      "epoch": 0.3401802557115909,
+      "grad_norm": 0.10725081712007523,
+      "learning_rate": 7.686696871112306e-06,
+      "loss": 0.0581,
+      "step": 1623
+    },
+    {
+      "epoch": 0.3403898553762314,
+      "grad_norm": 0.10398346930742264,
+      "learning_rate": 7.683833152007772e-06,
+      "loss": 0.0577,
+      "step": 1624
+    },
+    {
+      "epoch": 0.3405994550408719,
+      "grad_norm": 0.11018167436122894,
+      "learning_rate": 7.68096819565671e-06,
+      "loss": 0.0603,
+      "step": 1625
+    },
+    {
+      "epoch": 0.3408090547055125,
+      "grad_norm": 0.10882928967475891,
+      "learning_rate": 7.67810200337987e-06,
+      "loss": 0.0595,
+      "step": 1626
+    },
+    {
+      "epoch": 0.341018654370153,
+      "grad_norm": 0.09781121462583542,
+      "learning_rate": 7.675234576498561e-06,
+      "loss": 0.0611,
+      "step": 1627
+    },
+    {
+      "epoch": 0.34122825403479357,
+      "grad_norm": 0.08528628945350647,
+      "learning_rate": 7.672365916334668e-06,
+      "loss": 0.0576,
+      "step": 1628
+    },
+    {
+      "epoch": 0.3414378536994341,
+      "grad_norm": 0.08956452459096909,
+      "learning_rate": 7.66949602421064e-06,
+      "loss": 0.0624,
+      "step": 1629
+    },
+    {
+      "epoch": 0.3416474533640746,
+      "grad_norm": 0.09840685874223709,
+      "learning_rate": 7.666624901449503e-06,
+      "loss": 0.0581,
+      "step": 1630
+    },
+    {
+      "epoch": 0.34185705302871516,
+      "grad_norm": 0.11417452245950699,
+      "learning_rate": 7.66375254937484e-06,
+      "loss": 0.0581,
+      "step": 1631
+    },
+    {
+      "epoch": 0.3420666526933557,
+      "grad_norm": 0.11850450932979584,
+      "learning_rate": 7.660878969310803e-06,
+      "loss": 0.0607,
+      "step": 1632
+    },
+    {
+      "epoch": 0.34227625235799625,
+      "grad_norm": 0.12879729270935059,
+      "learning_rate": 7.658004162582116e-06,
+      "loss": 0.0619,
+      "step": 1633
+    },
+    {
+      "epoch": 0.34248585202263676,
+      "grad_norm": 0.1304604858160019,
+      "learning_rate": 7.655128130514061e-06,
+      "loss": 0.0632,
+      "step": 1634
+    },
+    {
+      "epoch": 0.3426954516872773,
+      "grad_norm": 0.10286222398281097,
+      "learning_rate": 7.652250874432489e-06,
+      "loss": 0.0611,
+      "step": 1635
+    },
+    {
+      "epoch": 0.34290505135191784,
+      "grad_norm": 0.11092126369476318,
+      "learning_rate": 7.649372395663816e-06,
+      "loss": 0.0615,
+      "step": 1636
+    },
+    {
+      "epoch": 0.34311465101655836,
+      "grad_norm": 0.109388567507267,
+      "learning_rate": 7.646492695535018e-06,
+      "loss": 0.0641,
+      "step": 1637
+    },
+    {
+      "epoch": 0.34332425068119893,
+      "grad_norm": 0.09455049782991409,
+      "learning_rate": 7.643611775373637e-06,
+      "loss": 0.0613,
+      "step": 1638
+    },
+    {
+      "epoch": 0.34353385034583944,
+      "grad_norm": 0.09676145017147064,
+      "learning_rate": 7.640729636507778e-06,
+      "loss": 0.0597,
+      "step": 1639
+    },
+    {
+      "epoch": 0.34374345001047996,
+      "grad_norm": 0.09647035598754883,
+      "learning_rate": 7.637846280266103e-06,
+      "loss": 0.0593,
+      "step": 1640
+    },
+    {
+      "epoch": 0.3439530496751205,
+      "grad_norm": 0.10977063328027725,
+      "learning_rate": 7.634961707977843e-06,
+      "loss": 0.0596,
+      "step": 1641
+    },
+    {
+      "epoch": 0.34416264933976104,
+      "grad_norm": 0.1262187659740448,
+      "learning_rate": 7.632075920972782e-06,
+      "loss": 0.0604,
+      "step": 1642
+    },
+    {
+      "epoch": 0.3443722490044016,
+      "grad_norm": 0.1421433538198471,
+      "learning_rate": 7.629188920581267e-06,
+      "loss": 0.0571,
+      "step": 1643
+    },
+    {
+      "epoch": 0.3445818486690421,
+      "grad_norm": 0.14843688905239105,
+      "learning_rate": 7.626300708134207e-06,
+      "loss": 0.0601,
+      "step": 1644
+    },
+    {
+      "epoch": 0.3447914483336827,
+      "grad_norm": 0.14421787858009338,
+      "learning_rate": 7.623411284963066e-06,
+      "loss": 0.0608,
+      "step": 1645
+    },
+    {
+      "epoch": 0.3450010479983232,
+      "grad_norm": 0.14530403912067413,
+      "learning_rate": 7.620520652399867e-06,
+      "loss": 0.059,
+      "step": 1646
+    },
+    {
+      "epoch": 0.3452106476629637,
+      "grad_norm": 0.14376644790172577,
+      "learning_rate": 7.617628811777191e-06,
+      "loss": 0.0613,
+      "step": 1647
+    },
+    {
+      "epoch": 0.3454202473276043,
+      "grad_norm": 0.14656515419483185,
+      "learning_rate": 7.614735764428178e-06,
+      "loss": 0.0616,
+      "step": 1648
+    },
+    {
+      "epoch": 0.3456298469922448,
+      "grad_norm": 0.15034882724285126,
+      "learning_rate": 7.611841511686521e-06,
+      "loss": 0.0645,
+      "step": 1649
+    },
+    {
+      "epoch": 0.3458394466568854,
+      "grad_norm": 0.17289340496063232,
+      "learning_rate": 7.608946054886468e-06,
+      "loss": 0.0605,
+      "step": 1650
+    },
+    {
+      "epoch": 0.3460490463215259,
+      "grad_norm": 0.20710116624832153,
+      "learning_rate": 7.606049395362827e-06,
+      "loss": 0.0601,
+      "step": 1651
+    },
+    {
+      "epoch": 0.3462586459861664,
+      "grad_norm": 0.24352329969406128,
+      "learning_rate": 7.6031515344509545e-06,
+      "loss": 0.0619,
+      "step": 1652
+    },
+    {
+      "epoch": 0.34646824565080697,
+      "grad_norm": 0.24319349229335785,
+      "learning_rate": 7.6002524734867676e-06,
+      "loss": 0.0637,
+      "step": 1653
+    },
+    {
+      "epoch": 0.3466778453154475,
+      "grad_norm": 0.18296609818935394,
+      "learning_rate": 7.597352213806729e-06,
+      "loss": 0.0614,
+      "step": 1654
+    },
+    {
+      "epoch": 0.34688744498008806,
+      "grad_norm": 0.11361169070005417,
+      "learning_rate": 7.5944507567478585e-06,
+      "loss": 0.0596,
+      "step": 1655
+    },
+    {
+      "epoch": 0.34709704464472857,
+      "grad_norm": 0.14272648096084595,
+      "learning_rate": 7.59154810364773e-06,
+      "loss": 0.0609,
+      "step": 1656
+    },
+    {
+      "epoch": 0.3473066443093691,
+      "grad_norm": 0.18227772414684296,
+      "learning_rate": 7.588644255844464e-06,
+      "loss": 0.0609,
+      "step": 1657
+    },
+    {
+      "epoch": 0.34751624397400965,
+      "grad_norm": 0.15642352402210236,
+      "learning_rate": 7.585739214676731e-06,
+      "loss": 0.0611,
+      "step": 1658
+    },
+    {
+      "epoch": 0.34772584363865017,
+      "grad_norm": 0.13368570804595947,
+      "learning_rate": 7.582832981483761e-06,
+      "loss": 0.0607,
+      "step": 1659
+    },
+    {
+      "epoch": 0.34793544330329074,
+      "grad_norm": 0.12491488456726074,
+      "learning_rate": 7.57992555760532e-06,
+      "loss": 0.0628,
+      "step": 1660
+    },
+    {
+      "epoch": 0.34814504296793125,
+      "grad_norm": 0.12365151941776276,
+      "learning_rate": 7.577016944381734e-06,
+      "loss": 0.0582,
+      "step": 1661
+    },
+    {
+      "epoch": 0.34835464263257176,
+      "grad_norm": 0.14696553349494934,
+      "learning_rate": 7.574107143153872e-06,
+      "loss": 0.0611,
+      "step": 1662
+    },
+    {
+      "epoch": 0.34856424229721233,
+      "grad_norm": 0.14201070368289948,
+      "learning_rate": 7.571196155263152e-06,
+      "loss": 0.0617,
+      "step": 1663
+    },
+    {
+      "epoch": 0.34877384196185285,
+      "grad_norm": 0.13257716596126556,
+      "learning_rate": 7.568283982051538e-06,
+      "loss": 0.062,
+      "step": 1664
+    },
+    {
+      "epoch": 0.3489834416264934,
+      "grad_norm": 0.10744427889585495,
+      "learning_rate": 7.565370624861541e-06,
+      "loss": 0.0615,
+      "step": 1665
+    },
+    {
+      "epoch": 0.34919304129113393,
+      "grad_norm": 0.11801417917013168,
+      "learning_rate": 7.562456085036221e-06,
+      "loss": 0.0591,
+      "step": 1666
+    },
+    {
+      "epoch": 0.34940264095577445,
+      "grad_norm": 0.13146089017391205,
+      "learning_rate": 7.5595403639191775e-06,
+      "loss": 0.0618,
+      "step": 1667
+    },
+    {
+      "epoch": 0.349612240620415,
+      "grad_norm": 0.11520653963088989,
+      "learning_rate": 7.556623462854555e-06,
+      "loss": 0.0604,
+      "step": 1668
+    },
+    {
+      "epoch": 0.34982184028505553,
+      "grad_norm": 0.111871138215065,
+      "learning_rate": 7.553705383187051e-06,
+      "loss": 0.0567,
+      "step": 1669
+    },
+    {
+      "epoch": 0.3500314399496961,
+      "grad_norm": 0.10216851532459259,
+      "learning_rate": 7.550786126261893e-06,
+      "loss": 0.0601,
+      "step": 1670
+    },
+    {
+      "epoch": 0.3502410396143366,
+      "grad_norm": 0.11334113776683807,
+      "learning_rate": 7.5478656934248626e-06,
+      "loss": 0.0607,
+      "step": 1671
+    },
+    {
+      "epoch": 0.3504506392789771,
+      "grad_norm": 0.12396923452615738,
+      "learning_rate": 7.544944086022276e-06,
+      "loss": 0.0611,
+      "step": 1672
+    },
+    {
+      "epoch": 0.3506602389436177,
+      "grad_norm": 0.11316753178834915,
+      "learning_rate": 7.5420213054009935e-06,
+      "loss": 0.0623,
+      "step": 1673
+    },
+    {
+      "epoch": 0.3508698386082582,
+      "grad_norm": 0.12763440608978271,
+      "learning_rate": 7.539097352908419e-06,
+      "loss": 0.0587,
+      "step": 1674
+    },
+    {
+      "epoch": 0.3510794382728988,
+      "grad_norm": 0.12276781350374222,
+      "learning_rate": 7.536172229892491e-06,
+      "loss": 0.0599,
+      "step": 1675
+    },
+    {
+      "epoch": 0.3512890379375393,
+      "grad_norm": 0.10066503286361694,
+      "learning_rate": 7.533245937701692e-06,
+      "loss": 0.0587,
+      "step": 1676
+    },
+    {
+      "epoch": 0.35149863760217986,
+      "grad_norm": 0.11499958485364914,
+      "learning_rate": 7.530318477685043e-06,
+      "loss": 0.0578,
+      "step": 1677
+    },
+    {
+      "epoch": 0.3517082372668204,
+      "grad_norm": 0.10679785162210464,
+      "learning_rate": 7.527389851192099e-06,
+      "loss": 0.0618,
+      "step": 1678
+    },
+    {
+      "epoch": 0.3519178369314609,
+      "grad_norm": 0.11170266568660736,
+      "learning_rate": 7.52446005957296e-06,
+      "loss": 0.0567,
+      "step": 1679
+    },
+    {
+      "epoch": 0.35212743659610146,
+      "grad_norm": 0.13220727443695068,
+      "learning_rate": 7.521529104178258e-06,
+      "loss": 0.0624,
+      "step": 1680
+    },
+    {
+      "epoch": 0.352337036260742,
+      "grad_norm": 0.11351287364959717,
+      "learning_rate": 7.518596986359163e-06,
+      "loss": 0.0576,
+      "step": 1681
+    },
+    {
+      "epoch": 0.35254663592538255,
+      "grad_norm": 0.1113462969660759,
+      "learning_rate": 7.51566370746738e-06,
+      "loss": 0.0627,
+      "step": 1682
+    },
+    {
+      "epoch": 0.35275623559002306,
+      "grad_norm": 0.11083466559648514,
+      "learning_rate": 7.51272926885515e-06,
+      "loss": 0.0582,
+      "step": 1683
+    },
+    {
+      "epoch": 0.3529658352546636,
+      "grad_norm": 0.09305007755756378,
+      "learning_rate": 7.50979367187525e-06,
+      "loss": 0.0585,
+      "step": 1684
+    },
+    {
+      "epoch": 0.35317543491930414,
+      "grad_norm": 0.09011313319206238,
+      "learning_rate": 7.506856917880989e-06,
+      "loss": 0.0612,
+      "step": 1685
+    },
+    {
+      "epoch": 0.35338503458394466,
+      "grad_norm": 0.09725736081600189,
+      "learning_rate": 7.503919008226208e-06,
+      "loss": 0.0565,
+      "step": 1686
+    },
+    {
+      "epoch": 0.3535946342485852,
+      "grad_norm": 0.10482776910066605,
+      "learning_rate": 7.5009799442652856e-06,
+      "loss": 0.0598,
+      "step": 1687
+    },
+    {
+      "epoch": 0.35380423391322574,
+      "grad_norm": 0.12371653318405151,
+      "learning_rate": 7.498039727353127e-06,
+      "loss": 0.0615,
+      "step": 1688
+    },
+    {
+      "epoch": 0.35401383357786625,
+      "grad_norm": 0.13332313299179077,
+      "learning_rate": 7.495098358845174e-06,
+      "loss": 0.0587,
+      "step": 1689
+    },
+    {
+      "epoch": 0.3542234332425068,
+      "grad_norm": 0.11412235349416733,
+      "learning_rate": 7.492155840097396e-06,
+      "loss": 0.0608,
+      "step": 1690
+    },
+    {
+      "epoch": 0.35443303290714734,
+      "grad_norm": 0.09121813625097275,
+      "learning_rate": 7.4892121724662915e-06,
+      "loss": 0.0599,
+      "step": 1691
+    },
+    {
+      "epoch": 0.3546426325717879,
+      "grad_norm": 0.09299511462450027,
+      "learning_rate": 7.486267357308896e-06,
+      "loss": 0.0611,
+      "step": 1692
+    },
+    {
+      "epoch": 0.3548522322364284,
+      "grad_norm": 0.09584937989711761,
+      "learning_rate": 7.483321395982762e-06,
+      "loss": 0.0628,
+      "step": 1693
+    },
+    {
+      "epoch": 0.35506183190106894,
+      "grad_norm": 0.0966387689113617,
+      "learning_rate": 7.4803742898459816e-06,
+      "loss": 0.0595,
+      "step": 1694
+    },
+    {
+      "epoch": 0.3552714315657095,
+      "grad_norm": 0.09714031964540482,
+      "learning_rate": 7.4774260402571696e-06,
+      "loss": 0.0597,
+      "step": 1695
+    },
+    {
+      "epoch": 0.35548103123035,
+      "grad_norm": 0.09430839866399765,
+      "learning_rate": 7.474476648575466e-06,
+      "loss": 0.059,
+      "step": 1696
+    },
+    {
+      "epoch": 0.3556906308949906,
+      "grad_norm": 0.09601857513189316,
+      "learning_rate": 7.471526116160542e-06,
+      "loss": 0.0576,
+      "step": 1697
+    },
+    {
+      "epoch": 0.3559002305596311,
+      "grad_norm": 0.11064789444208145,
+      "learning_rate": 7.468574444372593e-06,
+      "loss": 0.0608,
+      "step": 1698
+    },
+    {
+      "epoch": 0.3561098302242716,
+      "grad_norm": 0.11107967793941498,
+      "learning_rate": 7.465621634572336e-06,
+      "loss": 0.0602,
+      "step": 1699
+    },
+    {
+      "epoch": 0.3563194298889122,
+      "grad_norm": 0.12140575051307678,
+      "learning_rate": 7.462667688121019e-06,
+      "loss": 0.0576,
+      "step": 1700
+    },
+    {
+      "epoch": 0.3565290295535527,
+      "grad_norm": 0.14458388090133667,
+      "learning_rate": 7.459712606380409e-06,
+      "loss": 0.0606,
+      "step": 1701
+    },
+    {
+      "epoch": 0.35673862921819327,
+      "grad_norm": 0.14192935824394226,
+      "learning_rate": 7.4567563907127985e-06,
+      "loss": 0.061,
+      "step": 1702
+    },
+    {
+      "epoch": 0.3569482288828338,
+      "grad_norm": 0.14503100514411926,
+      "learning_rate": 7.453799042481002e-06,
+      "loss": 0.059,
+      "step": 1703
+    },
+    {
+      "epoch": 0.3571578285474743,
+      "grad_norm": 0.13211016356945038,
+      "learning_rate": 7.450840563048356e-06,
+      "loss": 0.0612,
+      "step": 1704
+    },
+    {
+      "epoch": 0.35736742821211487,
+      "grad_norm": 0.10646926611661911,
+      "learning_rate": 7.44788095377872e-06,
+      "loss": 0.0621,
+      "step": 1705
+    },
+    {
+      "epoch": 0.3575770278767554,
+      "grad_norm": 0.09362011402845383,
+      "learning_rate": 7.444920216036473e-06,
+      "loss": 0.0574,
+      "step": 1706
+    },
+    {
+      "epoch": 0.35778662754139595,
+      "grad_norm": 0.11936776340007782,
+      "learning_rate": 7.441958351186514e-06,
+      "loss": 0.0604,
+      "step": 1707
+    },
+    {
+      "epoch": 0.35799622720603647,
+      "grad_norm": 0.13509775698184967,
+      "learning_rate": 7.4389953605942634e-06,
+      "loss": 0.0611,
+      "step": 1708
+    },
+    {
+      "epoch": 0.358205826870677,
+      "grad_norm": 0.11575739830732346,
+      "learning_rate": 7.436031245625657e-06,
+      "loss": 0.0607,
+      "step": 1709
+    },
+    {
+      "epoch": 0.35841542653531755,
+      "grad_norm": 0.10143768787384033,
+      "learning_rate": 7.433066007647152e-06,
+      "loss": 0.0576,
+      "step": 1710
+    },
+    {
+      "epoch": 0.35862502619995806,
+      "grad_norm": 0.09407970309257507,
+      "learning_rate": 7.430099648025723e-06,
+      "loss": 0.0597,
+      "step": 1711
+    },
+    {
+      "epoch": 0.35883462586459863,
+      "grad_norm": 0.11093775928020477,
+      "learning_rate": 7.427132168128862e-06,
+      "loss": 0.0594,
+      "step": 1712
+    },
+    {
+      "epoch": 0.35904422552923915,
+      "grad_norm": 0.13045847415924072,
+      "learning_rate": 7.4241635693245766e-06,
+      "loss": 0.0631,
+      "step": 1713
+    },
+    {
+      "epoch": 0.3592538251938797,
+      "grad_norm": 0.128498837351799,
+      "learning_rate": 7.421193852981386e-06,
+      "loss": 0.0617,
+      "step": 1714
+    },
+    {
+      "epoch": 0.35946342485852023,
+      "grad_norm": 0.12185259163379669,
+      "learning_rate": 7.418223020468335e-06,
+      "loss": 0.0578,
+      "step": 1715
+    },
+    {
+      "epoch": 0.35967302452316074,
+      "grad_norm": 0.13501186668872833,
+      "learning_rate": 7.415251073154972e-06,
+      "loss": 0.0589,
+      "step": 1716
+    },
+    {
+      "epoch": 0.3598826241878013,
+      "grad_norm": 0.12140613049268723,
+      "learning_rate": 7.412278012411368e-06,
+      "loss": 0.0621,
+      "step": 1717
+    },
+    {
+      "epoch": 0.36009222385244183,
+      "grad_norm": 0.10289863497018814,
+      "learning_rate": 7.4093038396081e-06,
+      "loss": 0.0575,
+      "step": 1718
+    },
+    {
+      "epoch": 0.3603018235170824,
+      "grad_norm": 0.11283747851848602,
+      "learning_rate": 7.4063285561162624e-06,
+      "loss": 0.0584,
+      "step": 1719
+    },
+    {
+      "epoch": 0.3605114231817229,
+      "grad_norm": 0.13699781894683838,
+      "learning_rate": 7.40335216330746e-06,
+      "loss": 0.0603,
+      "step": 1720
+    },
+    {
+      "epoch": 0.3607210228463634,
+      "grad_norm": 0.16813865303993225,
+      "learning_rate": 7.400374662553811e-06,
+      "loss": 0.0605,
+      "step": 1721
+    },
+    {
+      "epoch": 0.360930622511004,
+      "grad_norm": 0.1661754846572876,
+      "learning_rate": 7.39739605522794e-06,
+      "loss": 0.0595,
+      "step": 1722
+    },
+    {
+      "epoch": 0.3611402221756445,
+      "grad_norm": 0.13529466092586517,
+      "learning_rate": 7.394416342702986e-06,
+      "loss": 0.0588,
+      "step": 1723
+    },
+    {
+      "epoch": 0.3613498218402851,
+      "grad_norm": 0.1041502058506012,
+      "learning_rate": 7.391435526352594e-06,
+      "loss": 0.0619,
+      "step": 1724
+    },
+    {
+      "epoch": 0.3615594215049256,
+      "grad_norm": 0.08625433593988419,
+      "learning_rate": 7.3884536075509225e-06,
+      "loss": 0.0591,
+      "step": 1725
+    },
+    {
+      "epoch": 0.3617690211695661,
+      "grad_norm": 0.1282806396484375,
+      "learning_rate": 7.385470587672634e-06,
+      "loss": 0.0607,
+      "step": 1726
+    },
+    {
+      "epoch": 0.3619786208342067,
+      "grad_norm": 0.15391872823238373,
+      "learning_rate": 7.382486468092899e-06,
+      "loss": 0.0588,
+      "step": 1727
+    },
+    {
+      "epoch": 0.3621882204988472,
+      "grad_norm": 0.1479339599609375,
+      "learning_rate": 7.379501250187399e-06,
+      "loss": 0.0598,
+      "step": 1728
+    },
+    {
+      "epoch": 0.36239782016348776,
+      "grad_norm": 0.1250709593296051,
+      "learning_rate": 7.376514935332314e-06,
+      "loss": 0.0605,
+      "step": 1729
+    },
+    {
+      "epoch": 0.3626074198281283,
+      "grad_norm": 0.09578738361597061,
+      "learning_rate": 7.37352752490434e-06,
+      "loss": 0.0602,
+      "step": 1730
+    },
+    {
+      "epoch": 0.3628170194927688,
+      "grad_norm": 0.09304331988096237,
+      "learning_rate": 7.37053902028067e-06,
+      "loss": 0.0581,
+      "step": 1731
+    },
+    {
+      "epoch": 0.36302661915740936,
+      "grad_norm": 0.11090435832738876,
+      "learning_rate": 7.367549422839002e-06,
+      "loss": 0.0587,
+      "step": 1732
+    },
+    {
+      "epoch": 0.36323621882204987,
+      "grad_norm": 0.12240754812955856,
+      "learning_rate": 7.364558733957543e-06,
+      "loss": 0.0576,
+      "step": 1733
+    },
+    {
+      "epoch": 0.36344581848669044,
+      "grad_norm": 0.12205494195222855,
+      "learning_rate": 7.361566955014999e-06,
+      "loss": 0.0618,
+      "step": 1734
+    },
+    {
+      "epoch": 0.36365541815133096,
+      "grad_norm": 0.1121978685259819,
+      "learning_rate": 7.3585740873905795e-06,
+      "loss": 0.0586,
+      "step": 1735
+    },
+    {
+      "epoch": 0.36386501781597147,
+      "grad_norm": 0.09682106226682663,
+      "learning_rate": 7.355580132463998e-06,
+      "loss": 0.0604,
+      "step": 1736
+    },
+    {
+      "epoch": 0.36407461748061204,
+      "grad_norm": 0.08918868750333786,
+      "learning_rate": 7.352585091615461e-06,
+      "loss": 0.0609,
+      "step": 1737
+    },
+    {
+      "epoch": 0.36428421714525255,
+      "grad_norm": 0.08579257130622864,
+      "learning_rate": 7.349588966225686e-06,
+      "loss": 0.0565,
+      "step": 1738
+    },
+    {
+      "epoch": 0.3644938168098931,
+      "grad_norm": 0.08958027511835098,
+      "learning_rate": 7.346591757675886e-06,
+      "loss": 0.059,
+      "step": 1739
+    },
+    {
+      "epoch": 0.36470341647453364,
+      "grad_norm": 0.08919413387775421,
+      "learning_rate": 7.3435934673477735e-06,
+      "loss": 0.0588,
+      "step": 1740
+    },
+    {
+      "epoch": 0.36491301613917415,
+      "grad_norm": 0.08150623738765717,
+      "learning_rate": 7.340594096623559e-06,
+      "loss": 0.0612,
+      "step": 1741
+    },
+    {
+      "epoch": 0.3651226158038147,
+      "grad_norm": 0.08697859197854996,
+      "learning_rate": 7.337593646885952e-06,
+      "loss": 0.0579,
+      "step": 1742
+    },
+    {
+      "epoch": 0.36533221546845523,
+      "grad_norm": 0.09409506618976593,
+      "learning_rate": 7.3345921195181605e-06,
+      "loss": 0.0609,
+      "step": 1743
+    },
+    {
+      "epoch": 0.3655418151330958,
+      "grad_norm": 0.12264561653137207,
+      "learning_rate": 7.331589515903885e-06,
+      "loss": 0.0571,
+      "step": 1744
+    },
+    {
+      "epoch": 0.3657514147977363,
+      "grad_norm": 0.15233023464679718,
+      "learning_rate": 7.328585837427329e-06,
+      "loss": 0.0616,
+      "step": 1745
+    },
+    {
+      "epoch": 0.3659610144623769,
+      "grad_norm": 0.15437071025371552,
+      "learning_rate": 7.3255810854731845e-06,
+      "loss": 0.0583,
+      "step": 1746
+    },
+    {
+      "epoch": 0.3661706141270174,
+      "grad_norm": 0.14194343984127045,
+      "learning_rate": 7.322575261426643e-06,
+      "loss": 0.0579,
+      "step": 1747
+    },
+    {
+      "epoch": 0.3663802137916579,
+      "grad_norm": 0.14112666249275208,
+      "learning_rate": 7.319568366673389e-06,
+      "loss": 0.0576,
+      "step": 1748
+    },
+    {
+      "epoch": 0.3665898134562985,
+      "grad_norm": 0.15609388053417206,
+      "learning_rate": 7.316560402599598e-06,
+      "loss": 0.0629,
+      "step": 1749
+    },
+    {
+      "epoch": 0.366799413120939,
+      "grad_norm": 0.1522039920091629,
+      "learning_rate": 7.313551370591944e-06,
+      "loss": 0.0605,
+      "step": 1750
+    },
+    {
+      "epoch": 0.36700901278557957,
+      "grad_norm": 0.13480624556541443,
+      "learning_rate": 7.310541272037588e-06,
+      "loss": 0.0589,
+      "step": 1751
+    },
+    {
+      "epoch": 0.3672186124502201,
+      "grad_norm": 0.12041886895895004,
+      "learning_rate": 7.307530108324186e-06,
+      "loss": 0.0586,
+      "step": 1752
+    },
+    {
+      "epoch": 0.3674282121148606,
+      "grad_norm": 0.11600147187709808,
+      "learning_rate": 7.304517880839883e-06,
+      "loss": 0.058,
+      "step": 1753
+    },
+    {
+      "epoch": 0.36763781177950117,
+      "grad_norm": 0.11566165834665298,
+      "learning_rate": 7.3015045909733165e-06,
+      "loss": 0.0623,
+      "step": 1754
+    },
+    {
+      "epoch": 0.3678474114441417,
+      "grad_norm": 0.12034988403320312,
+      "learning_rate": 7.2984902401136115e-06,
+      "loss": 0.059,
+      "step": 1755
+    },
+    {
+      "epoch": 0.36805701110878225,
+      "grad_norm": 0.13276949524879456,
+      "learning_rate": 7.295474829650382e-06,
+      "loss": 0.0582,
+      "step": 1756
+    },
+    {
+      "epoch": 0.36826661077342276,
+      "grad_norm": 0.14941634237766266,
+      "learning_rate": 7.292458360973733e-06,
+      "loss": 0.059,
+      "step": 1757
+    },
+    {
+      "epoch": 0.3684762104380633,
+      "grad_norm": 0.14865432679653168,
+      "learning_rate": 7.289440835474257e-06,
+      "loss": 0.0625,
+      "step": 1758
+    },
+    {
+      "epoch": 0.36868581010270385,
+      "grad_norm": 0.11310829967260361,
+      "learning_rate": 7.286422254543031e-06,
+      "loss": 0.0582,
+      "step": 1759
+    },
+    {
+      "epoch": 0.36889540976734436,
+      "grad_norm": 0.0944291204214096,
+      "learning_rate": 7.28340261957162e-06,
+      "loss": 0.0588,
+      "step": 1760
+    },
+    {
+      "epoch": 0.36910500943198493,
+      "grad_norm": 0.11601667106151581,
+      "learning_rate": 7.2803819319520765e-06,
+      "loss": 0.0576,
+      "step": 1761
+    },
+    {
+      "epoch": 0.36931460909662545,
+      "grad_norm": 0.1313844472169876,
+      "learning_rate": 7.277360193076936e-06,
+      "loss": 0.062,
+      "step": 1762
+    },
+    {
+      "epoch": 0.36952420876126596,
+      "grad_norm": 0.12570670247077942,
+      "learning_rate": 7.274337404339218e-06,
+      "loss": 0.0607,
+      "step": 1763
+    },
+    {
+      "epoch": 0.36973380842590653,
+      "grad_norm": 0.10776617377996445,
+      "learning_rate": 7.271313567132431e-06,
+      "loss": 0.0618,
+      "step": 1764
+    },
+    {
+      "epoch": 0.36994340809054704,
+      "grad_norm": 0.09985363483428955,
+      "learning_rate": 7.26828868285056e-06,
+      "loss": 0.0583,
+      "step": 1765
+    },
+    {
+      "epoch": 0.3701530077551876,
+      "grad_norm": 0.10462814569473267,
+      "learning_rate": 7.265262752888078e-06,
+      "loss": 0.0603,
+      "step": 1766
+    },
+    {
+      "epoch": 0.3703626074198281,
+      "grad_norm": 0.10572189837694168,
+      "learning_rate": 7.262235778639938e-06,
+      "loss": 0.0587,
+      "step": 1767
+    },
+    {
+      "epoch": 0.37057220708446864,
+      "grad_norm": 0.12369339168071747,
+      "learning_rate": 7.259207761501572e-06,
+      "loss": 0.0567,
+      "step": 1768
+    },
+    {
+      "epoch": 0.3707818067491092,
+      "grad_norm": 0.14378707110881805,
+      "learning_rate": 7.256178702868899e-06,
+      "loss": 0.0586,
+      "step": 1769
+    },
+    {
+      "epoch": 0.3709914064137497,
+      "grad_norm": 0.13347645103931427,
+      "learning_rate": 7.253148604138312e-06,
+      "loss": 0.0572,
+      "step": 1770
+    },
+    {
+      "epoch": 0.3712010060783903,
+      "grad_norm": 0.12877966463565826,
+      "learning_rate": 7.250117466706686e-06,
+      "loss": 0.0575,
+      "step": 1771
+    },
+    {
+      "epoch": 0.3714106057430308,
+      "grad_norm": 0.15981978178024292,
+      "learning_rate": 7.247085291971377e-06,
+      "loss": 0.0589,
+      "step": 1772
+    },
+    {
+      "epoch": 0.3716202054076713,
+      "grad_norm": 0.17421603202819824,
+      "learning_rate": 7.244052081330214e-06,
+      "loss": 0.0602,
+      "step": 1773
+    },
+    {
+      "epoch": 0.3718298050723119,
+      "grad_norm": 0.166316419839859,
+      "learning_rate": 7.241017836181508e-06,
+      "loss": 0.059,
+      "step": 1774
+    },
+    {
+      "epoch": 0.3720394047369524,
+      "grad_norm": 0.14877326786518097,
+      "learning_rate": 7.237982557924044e-06,
+      "loss": 0.0604,
+      "step": 1775
+    },
+    {
+      "epoch": 0.372249004401593,
+      "grad_norm": 0.11550521105527878,
+      "learning_rate": 7.234946247957087e-06,
+      "loss": 0.058,
+      "step": 1776
+    },
+    {
+      "epoch": 0.3724586040662335,
+      "grad_norm": 0.12925846874713898,
+      "learning_rate": 7.231908907680373e-06,
+      "loss": 0.0607,
+      "step": 1777
+    },
+    {
+      "epoch": 0.372668203730874,
+      "grad_norm": 0.16908065974712372,
+      "learning_rate": 7.228870538494116e-06,
+      "loss": 0.0592,
+      "step": 1778
+    },
+    {
+      "epoch": 0.3728778033955146,
+      "grad_norm": 0.1643764078617096,
+      "learning_rate": 7.225831141799004e-06,
+      "loss": 0.0613,
+      "step": 1779
+    },
+    {
+      "epoch": 0.3730874030601551,
+      "grad_norm": 0.1174246221780777,
+      "learning_rate": 7.222790718996199e-06,
+      "loss": 0.058,
+      "step": 1780
+    },
+    {
+      "epoch": 0.37329700272479566,
+      "grad_norm": 0.11582423746585846,
+      "learning_rate": 7.219749271487333e-06,
+      "loss": 0.0582,
+      "step": 1781
+    },
+    {
+      "epoch": 0.37350660238943617,
+      "grad_norm": 0.1282123178243637,
+      "learning_rate": 7.216706800674514e-06,
+      "loss": 0.0604,
+      "step": 1782
+    },
+    {
+      "epoch": 0.37371620205407674,
+      "grad_norm": 0.12925361096858978,
+      "learning_rate": 7.213663307960321e-06,
+      "loss": 0.0586,
+      "step": 1783
+    },
+    {
+      "epoch": 0.37392580171871725,
+      "grad_norm": 0.12871502339839935,
+      "learning_rate": 7.210618794747803e-06,
+      "loss": 0.0582,
+      "step": 1784
+    },
+    {
+      "epoch": 0.37413540138335777,
+      "grad_norm": 0.11913754045963287,
+      "learning_rate": 7.207573262440479e-06,
+      "loss": 0.059,
+      "step": 1785
+    },
+    {
+      "epoch": 0.37434500104799834,
+      "grad_norm": 0.12632472813129425,
+      "learning_rate": 7.20452671244234e-06,
+      "loss": 0.0597,
+      "step": 1786
+    },
+    {
+      "epoch": 0.37455460071263885,
+      "grad_norm": 0.14089874923229218,
+      "learning_rate": 7.2014791461578445e-06,
+      "loss": 0.0598,
+      "step": 1787
+    },
+    {
+      "epoch": 0.3747642003772794,
+      "grad_norm": 0.1204831525683403,
+      "learning_rate": 7.1984305649919195e-06,
+      "loss": 0.0588,
+      "step": 1788
+    },
+    {
+      "epoch": 0.37497380004191994,
+      "grad_norm": 0.11883781105279922,
+      "learning_rate": 7.195380970349961e-06,
+      "loss": 0.0595,
+      "step": 1789
+    },
+    {
+      "epoch": 0.37518339970656045,
+      "grad_norm": 0.12562531232833862,
+      "learning_rate": 7.192330363637832e-06,
+      "loss": 0.0608,
+      "step": 1790
+    },
+    {
+      "epoch": 0.375392999371201,
+      "grad_norm": 0.1165882870554924,
+      "learning_rate": 7.18927874626186e-06,
+      "loss": 0.059,
+      "step": 1791
+    },
+    {
+      "epoch": 0.37560259903584153,
+      "grad_norm": 0.12378109246492386,
+      "learning_rate": 7.186226119628841e-06,
+      "loss": 0.0599,
+      "step": 1792
+    },
+    {
+      "epoch": 0.3758121987004821,
+      "grad_norm": 0.11985018849372864,
+      "learning_rate": 7.183172485146033e-06,
+      "loss": 0.0592,
+      "step": 1793
+    },
+    {
+      "epoch": 0.3760217983651226,
+      "grad_norm": 0.09804270416498184,
+      "learning_rate": 7.180117844221166e-06,
+      "loss": 0.0577,
+      "step": 1794
+    },
+    {
+      "epoch": 0.37623139802976313,
+      "grad_norm": 0.11297204345464706,
+      "learning_rate": 7.177062198262424e-06,
+      "loss": 0.0617,
+      "step": 1795
+    },
+    {
+      "epoch": 0.3764409976944037,
+      "grad_norm": 0.11378408223390579,
+      "learning_rate": 7.1740055486784595e-06,
+      "loss": 0.0601,
+      "step": 1796
+    },
+    {
+      "epoch": 0.3766505973590442,
+      "grad_norm": 0.10527677834033966,
+      "learning_rate": 7.170947896878392e-06,
+      "loss": 0.0609,
+      "step": 1797
+    },
+    {
+      "epoch": 0.3768601970236848,
+      "grad_norm": 0.11885930597782135,
+      "learning_rate": 7.167889244271792e-06,
+      "loss": 0.0564,
+      "step": 1798
+    },
+    {
+      "epoch": 0.3770697966883253,
+      "grad_norm": 0.11827868223190308,
+      "learning_rate": 7.164829592268702e-06,
+      "loss": 0.0605,
+      "step": 1799
+    },
+    {
+      "epoch": 0.3772793963529658,
+      "grad_norm": 0.09840241819620132,
+      "learning_rate": 7.16176894227962e-06,
+      "loss": 0.0598,
+      "step": 1800
+    },
+    {
+      "epoch": 0.3774889960176064,
+      "grad_norm": 0.11675871908664703,
+      "learning_rate": 7.158707295715504e-06,
+      "loss": 0.0584,
+      "step": 1801
+    },
+    {
+      "epoch": 0.3776985956822469,
+      "grad_norm": 0.1085144504904747,
+      "learning_rate": 7.155644653987772e-06,
+      "loss": 0.0565,
+      "step": 1802
+    },
+    {
+      "epoch": 0.37790819534688747,
+      "grad_norm": 0.0793166384100914,
+      "learning_rate": 7.152581018508305e-06,
+      "loss": 0.0586,
+      "step": 1803
+    },
+    {
+      "epoch": 0.378117795011528,
+      "grad_norm": 0.09427671879529953,
+      "learning_rate": 7.149516390689433e-06,
+      "loss": 0.0602,
+      "step": 1804
+    },
+    {
+      "epoch": 0.3783273946761685,
+      "grad_norm": 0.09693881869316101,
+      "learning_rate": 7.146450771943953e-06,
+      "loss": 0.0579,
+      "step": 1805
+    },
+    {
+      "epoch": 0.37853699434080906,
+      "grad_norm": 0.08774615824222565,
+      "learning_rate": 7.143384163685112e-06,
+      "loss": 0.0581,
+      "step": 1806
+    },
+    {
+      "epoch": 0.3787465940054496,
+      "grad_norm": 0.10799058526754379,
+      "learning_rate": 7.140316567326617e-06,
+      "loss": 0.0581,
+      "step": 1807
+    },
+    {
+      "epoch": 0.37895619367009015,
+      "grad_norm": 0.12135607749223709,
+      "learning_rate": 7.13724798428263e-06,
+      "loss": 0.0586,
+      "step": 1808
+    },
+    {
+      "epoch": 0.37916579333473066,
+      "grad_norm": 0.12671098113059998,
+      "learning_rate": 7.134178415967765e-06,
+      "loss": 0.0583,
+      "step": 1809
+    },
+    {
+      "epoch": 0.3793753929993712,
+      "grad_norm": 0.1515565663576126,
+      "learning_rate": 7.131107863797093e-06,
+      "loss": 0.0578,
+      "step": 1810
+    },
+    {
+      "epoch": 0.37958499266401174,
+      "grad_norm": 0.18114930391311646,
+      "learning_rate": 7.12803632918614e-06,
+      "loss": 0.0587,
+      "step": 1811
+    },
+    {
+      "epoch": 0.37979459232865226,
+      "grad_norm": 0.18494822084903717,
+      "learning_rate": 7.12496381355088e-06,
+      "loss": 0.0581,
+      "step": 1812
+    },
+    {
+      "epoch": 0.38000419199329283,
+      "grad_norm": 0.14751562476158142,
+      "learning_rate": 7.121890318307745e-06,
+      "loss": 0.0603,
+      "step": 1813
+    },
+    {
+      "epoch": 0.38021379165793334,
+      "grad_norm": 0.09215519577264786,
+      "learning_rate": 7.1188158448736135e-06,
+      "loss": 0.0607,
+      "step": 1814
+    },
+    {
+      "epoch": 0.38042339132257386,
+      "grad_norm": 0.10190481692552567,
+      "learning_rate": 7.115740394665816e-06,
+      "loss": 0.0595,
+      "step": 1815
+    },
+    {
+      "epoch": 0.3806329909872144,
+      "grad_norm": 0.10842154920101166,
+      "learning_rate": 7.112663969102138e-06,
+      "loss": 0.0562,
+      "step": 1816
+    },
+    {
+      "epoch": 0.38084259065185494,
+      "grad_norm": 0.11397344619035721,
+      "learning_rate": 7.1095865696008085e-06,
+      "loss": 0.0586,
+      "step": 1817
+    },
+    {
+      "epoch": 0.3810521903164955,
+      "grad_norm": 0.12245843559503555,
+      "learning_rate": 7.1065081975805086e-06,
+      "loss": 0.0613,
+      "step": 1818
+    },
+    {
+      "epoch": 0.381261789981136,
+      "grad_norm": 0.11635053157806396,
+      "learning_rate": 7.103428854460367e-06,
+      "loss": 0.0585,
+      "step": 1819
+    },
+    {
+      "epoch": 0.3814713896457766,
+      "grad_norm": 0.10383999347686768,
+      "learning_rate": 7.100348541659961e-06,
+      "loss": 0.0625,
+      "step": 1820
+    },
+    {
+      "epoch": 0.3816809893104171,
+      "grad_norm": 0.09212907403707504,
+      "learning_rate": 7.0972672605993106e-06,
+      "loss": 0.0634,
+      "step": 1821
+    },
+    {
+      "epoch": 0.3818905889750576,
+      "grad_norm": 0.10867664963006973,
+      "learning_rate": 7.094185012698893e-06,
+      "loss": 0.0571,
+      "step": 1822
+    },
+    {
+      "epoch": 0.3821001886396982,
+      "grad_norm": 0.10434751957654953,
+      "learning_rate": 7.091101799379617e-06,
+      "loss": 0.0615,
+      "step": 1823
+    },
+    {
+      "epoch": 0.3823097883043387,
+      "grad_norm": 0.09127280861139297,
+      "learning_rate": 7.088017622062847e-06,
+      "loss": 0.0572,
+      "step": 1824
+    },
+    {
+      "epoch": 0.3825193879689793,
+      "grad_norm": 0.11248623579740524,
+      "learning_rate": 7.084932482170385e-06,
+      "loss": 0.06,
+      "step": 1825
+    },
+    {
+      "epoch": 0.3827289876336198,
+      "grad_norm": 0.12546822428703308,
+      "learning_rate": 7.081846381124484e-06,
+      "loss": 0.0571,
+      "step": 1826
+    },
+    {
+      "epoch": 0.3829385872982603,
+      "grad_norm": 0.11402714997529984,
+      "learning_rate": 7.078759320347833e-06,
+      "loss": 0.0578,
+      "step": 1827
+    },
+    {
+      "epoch": 0.38314818696290087,
+      "grad_norm": 0.10201937705278397,
+      "learning_rate": 7.075671301263568e-06,
+      "loss": 0.0579,
+      "step": 1828
+    },
+    {
+      "epoch": 0.3833577866275414,
+      "grad_norm": 0.0954611748456955,
+      "learning_rate": 7.072582325295262e-06,
+      "loss": 0.0569,
+      "step": 1829
+    },
+    {
+      "epoch": 0.38356738629218196,
+      "grad_norm": 0.10198832303285599,
+      "learning_rate": 7.069492393866937e-06,
+      "loss": 0.0601,
+      "step": 1830
+    },
+    {
+      "epoch": 0.38377698595682247,
+      "grad_norm": 0.09932437539100647,
+      "learning_rate": 7.066401508403047e-06,
+      "loss": 0.0576,
+      "step": 1831
+    },
+    {
+      "epoch": 0.383986585621463,
+      "grad_norm": 0.08973968774080276,
+      "learning_rate": 7.063309670328491e-06,
+      "loss": 0.0588,
+      "step": 1832
+    },
+    {
+      "epoch": 0.38419618528610355,
+      "grad_norm": 0.10677745938301086,
+      "learning_rate": 7.060216881068607e-06,
+      "loss": 0.0569,
+      "step": 1833
+    },
+    {
+      "epoch": 0.38440578495074407,
+      "grad_norm": 0.1219630241394043,
+      "learning_rate": 7.057123142049166e-06,
+      "loss": 0.0576,
+      "step": 1834
+    },
+    {
+      "epoch": 0.38461538461538464,
+      "grad_norm": 0.14446859061717987,
+      "learning_rate": 7.0540284546963846e-06,
+      "loss": 0.0577,
+      "step": 1835
+    },
+    {
+      "epoch": 0.38482498428002515,
+      "grad_norm": 0.17997580766677856,
+      "learning_rate": 7.050932820436915e-06,
+      "loss": 0.0598,
+      "step": 1836
+    },
+    {
+      "epoch": 0.38503458394466566,
+      "grad_norm": 0.17284150421619415,
+      "learning_rate": 7.047836240697837e-06,
+      "loss": 0.0602,
+      "step": 1837
+    },
+    {
+      "epoch": 0.38524418360930623,
+      "grad_norm": 0.12842464447021484,
+      "learning_rate": 7.044738716906679e-06,
+      "loss": 0.0594,
+      "step": 1838
+    },
+    {
+      "epoch": 0.38545378327394675,
+      "grad_norm": 0.12296885251998901,
+      "learning_rate": 7.041640250491398e-06,
+      "loss": 0.0619,
+      "step": 1839
+    },
+    {
+      "epoch": 0.3856633829385873,
+      "grad_norm": 0.14263387024402618,
+      "learning_rate": 7.038540842880386e-06,
+      "loss": 0.06,
+      "step": 1840
+    },
+    {
+      "epoch": 0.38587298260322783,
+      "grad_norm": 0.14516735076904297,
+      "learning_rate": 7.035440495502469e-06,
+      "loss": 0.0623,
+      "step": 1841
+    },
+    {
+      "epoch": 0.38608258226786835,
+      "grad_norm": 0.12423399090766907,
+      "learning_rate": 7.0323392097869044e-06,
+      "loss": 0.0591,
+      "step": 1842
+    },
+    {
+      "epoch": 0.3862921819325089,
+      "grad_norm": 0.10325875133275986,
+      "learning_rate": 7.029236987163388e-06,
+      "loss": 0.0577,
+      "step": 1843
+    },
+    {
+      "epoch": 0.38650178159714943,
+      "grad_norm": 0.09640491008758545,
+      "learning_rate": 7.026133829062041e-06,
+      "loss": 0.0592,
+      "step": 1844
+    },
+    {
+      "epoch": 0.38671138126179,
+      "grad_norm": 0.09991448372602463,
+      "learning_rate": 7.023029736913421e-06,
+      "loss": 0.0593,
+      "step": 1845
+    },
+    {
+      "epoch": 0.3869209809264305,
+      "grad_norm": 0.10741396993398666,
+      "learning_rate": 7.019924712148511e-06,
+      "loss": 0.0589,
+      "step": 1846
+    },
+    {
+      "epoch": 0.387130580591071,
+      "grad_norm": 0.10864882171154022,
+      "learning_rate": 7.0168187561987285e-06,
+      "loss": 0.0581,
+      "step": 1847
+    },
+    {
+      "epoch": 0.3873401802557116,
+      "grad_norm": 0.10337284207344055,
+      "learning_rate": 7.013711870495919e-06,
+      "loss": 0.0576,
+      "step": 1848
+    },
+    {
+      "epoch": 0.3875497799203521,
+      "grad_norm": 0.1069745123386383,
+      "learning_rate": 7.010604056472355e-06,
+      "loss": 0.0572,
+      "step": 1849
+    },
+    {
+      "epoch": 0.3877593795849927,
+      "grad_norm": 0.09350664913654327,
+      "learning_rate": 7.0074953155607395e-06,
+      "loss": 0.0566,
+      "step": 1850
+    },
+    {
+      "epoch": 0.3879689792496332,
+      "grad_norm": 0.08867273479700089,
+      "learning_rate": 7.004385649194199e-06,
+      "loss": 0.0596,
+      "step": 1851
+    },
+    {
+      "epoch": 0.38817857891427376,
+      "grad_norm": 0.109881192445755,
+      "learning_rate": 7.001275058806292e-06,
+      "loss": 0.058,
+      "step": 1852
+    },
+    {
+      "epoch": 0.3883881785789143,
+      "grad_norm": 0.1241583600640297,
+      "learning_rate": 6.998163545830998e-06,
+      "loss": 0.0577,
+      "step": 1853
+    },
+    {
+      "epoch": 0.3885977782435548,
+      "grad_norm": 0.10716693848371506,
+      "learning_rate": 6.995051111702724e-06,
+      "loss": 0.0574,
+      "step": 1854
+    },
+    {
+      "epoch": 0.38880737790819536,
+      "grad_norm": 0.10287656635046005,
+      "learning_rate": 6.991937757856302e-06,
+      "loss": 0.0589,
+      "step": 1855
+    },
+    {
+      "epoch": 0.3890169775728359,
+      "grad_norm": 0.128378227353096,
+      "learning_rate": 6.9888234857269875e-06,
+      "loss": 0.0598,
+      "step": 1856
+    },
+    {
+      "epoch": 0.38922657723747645,
+      "grad_norm": 0.13415974378585815,
+      "learning_rate": 6.98570829675046e-06,
+      "loss": 0.0621,
+      "step": 1857
+    },
+    {
+      "epoch": 0.38943617690211696,
+      "grad_norm": 0.1083337590098381,
+      "learning_rate": 6.98259219236282e-06,
+      "loss": 0.0588,
+      "step": 1858
+    },
+    {
+      "epoch": 0.3896457765667575,
+      "grad_norm": 0.10138161480426788,
+      "learning_rate": 6.979475174000591e-06,
+      "loss": 0.0616,
+      "step": 1859
+    },
+    {
+      "epoch": 0.38985537623139804,
+      "grad_norm": 0.1118636503815651,
+      "learning_rate": 6.976357243100718e-06,
+      "loss": 0.0579,
+      "step": 1860
+    },
+    {
+      "epoch": 0.39006497589603856,
+      "grad_norm": 0.11336030066013336,
+      "learning_rate": 6.973238401100565e-06,
+      "loss": 0.0558,
+      "step": 1861
+    },
+    {
+      "epoch": 0.3902745755606791,
+      "grad_norm": 0.11859306693077087,
+      "learning_rate": 6.970118649437919e-06,
+      "loss": 0.0579,
+      "step": 1862
+    },
+    {
+      "epoch": 0.39048417522531964,
+      "grad_norm": 0.1387511044740677,
+      "learning_rate": 6.966997989550988e-06,
+      "loss": 0.0584,
+      "step": 1863
+    },
+    {
+      "epoch": 0.39069377488996015,
+      "grad_norm": 0.14954118430614471,
+      "learning_rate": 6.96387642287839e-06,
+      "loss": 0.0588,
+      "step": 1864
+    },
+    {
+      "epoch": 0.3909033745546007,
+      "grad_norm": 0.14520269632339478,
+      "learning_rate": 6.960753950859168e-06,
+      "loss": 0.0596,
+      "step": 1865
+    },
+    {
+      "epoch": 0.39111297421924124,
+      "grad_norm": 0.11415694653987885,
+      "learning_rate": 6.957630574932784e-06,
+      "loss": 0.0603,
+      "step": 1866
+    },
+    {
+      "epoch": 0.3913225738838818,
+      "grad_norm": 0.13128365576267242,
+      "learning_rate": 6.954506296539112e-06,
+      "loss": 0.0591,
+      "step": 1867
+    },
+    {
+      "epoch": 0.3915321735485223,
+      "grad_norm": 0.13788260519504547,
+      "learning_rate": 6.951381117118441e-06,
+      "loss": 0.059,
+      "step": 1868
+    },
+    {
+      "epoch": 0.39174177321316284,
+      "grad_norm": 0.1196761503815651,
+      "learning_rate": 6.948255038111482e-06,
+      "loss": 0.0558,
+      "step": 1869
+    },
+    {
+      "epoch": 0.3919513728778034,
+      "grad_norm": 0.13322733342647552,
+      "learning_rate": 6.945128060959354e-06,
+      "loss": 0.0576,
+      "step": 1870
+    },
+    {
+      "epoch": 0.3921609725424439,
+      "grad_norm": 0.12423675507307053,
+      "learning_rate": 6.942000187103594e-06,
+      "loss": 0.0593,
+      "step": 1871
+    },
+    {
+      "epoch": 0.3923705722070845,
+      "grad_norm": 0.1317249983549118,
+      "learning_rate": 6.938871417986153e-06,
+      "loss": 0.0581,
+      "step": 1872
+    },
+    {
+      "epoch": 0.392580171871725,
+      "grad_norm": 0.1220802292227745,
+      "learning_rate": 6.935741755049389e-06,
+      "loss": 0.0602,
+      "step": 1873
+    },
+    {
+      "epoch": 0.3927897715363655,
+      "grad_norm": 0.10902206599712372,
+      "learning_rate": 6.9326111997360775e-06,
+      "loss": 0.056,
+      "step": 1874
+    },
+    {
+      "epoch": 0.3929993712010061,
+      "grad_norm": 0.12212048470973969,
+      "learning_rate": 6.929479753489406e-06,
+      "loss": 0.0588,
+      "step": 1875
+    },
+    {
+      "epoch": 0.3932089708656466,
+      "grad_norm": 0.09955111145973206,
+      "learning_rate": 6.926347417752969e-06,
+      "loss": 0.0581,
+      "step": 1876
+    },
+    {
+      "epoch": 0.39341857053028717,
+      "grad_norm": 0.11181198060512543,
+      "learning_rate": 6.923214193970773e-06,
+      "loss": 0.0576,
+      "step": 1877
+    },
+    {
+      "epoch": 0.3936281701949277,
+      "grad_norm": 0.13282683491706848,
+      "learning_rate": 6.920080083587231e-06,
+      "loss": 0.0619,
+      "step": 1878
+    },
+    {
+      "epoch": 0.3938377698595682,
+      "grad_norm": 0.09793748706579208,
+      "learning_rate": 6.916945088047171e-06,
+      "loss": 0.0548,
+      "step": 1879
+    },
+    {
+      "epoch": 0.39404736952420877,
+      "grad_norm": 0.1649271547794342,
+      "learning_rate": 6.913809208795823e-06,
+      "loss": 0.06,
+      "step": 1880
+    },
+    {
+      "epoch": 0.3942569691888493,
+      "grad_norm": 0.1789517104625702,
+      "learning_rate": 6.910672447278827e-06,
+      "loss": 0.061,
+      "step": 1881
+    },
+    {
+      "epoch": 0.39446656885348985,
+      "grad_norm": 0.17387685179710388,
+      "learning_rate": 6.90753480494223e-06,
+      "loss": 0.0588,
+      "step": 1882
+    },
+    {
+      "epoch": 0.39467616851813037,
+      "grad_norm": 0.1808399260044098,
+      "learning_rate": 6.904396283232484e-06,
+      "loss": 0.0592,
+      "step": 1883
+    },
+    {
+      "epoch": 0.3948857681827709,
+      "grad_norm": 0.12922294437885284,
+      "learning_rate": 6.9012568835964474e-06,
+      "loss": 0.0611,
+      "step": 1884
+    },
+    {
+      "epoch": 0.39509536784741145,
+      "grad_norm": 0.14349722862243652,
+      "learning_rate": 6.898116607481382e-06,
+      "loss": 0.0583,
+      "step": 1885
+    },
+    {
+      "epoch": 0.39530496751205196,
+      "grad_norm": 0.14803367853164673,
+      "learning_rate": 6.894975456334956e-06,
+      "loss": 0.0587,
+      "step": 1886
+    },
+    {
+      "epoch": 0.39551456717669253,
+      "grad_norm": 0.15976615250110626,
+      "learning_rate": 6.891833431605237e-06,
+      "loss": 0.0583,
+      "step": 1887
+    },
+    {
+      "epoch": 0.39572416684133305,
+      "grad_norm": 0.1385822743177414,
+      "learning_rate": 6.8886905347406985e-06,
+      "loss": 0.0625,
+      "step": 1888
+    },
+    {
+      "epoch": 0.3959337665059736,
+      "grad_norm": 0.12182749807834625,
+      "learning_rate": 6.885546767190217e-06,
+      "loss": 0.0548,
+      "step": 1889
+    },
+    {
+      "epoch": 0.39614336617061413,
+      "grad_norm": 0.13791294395923615,
+      "learning_rate": 6.882402130403066e-06,
+      "loss": 0.0583,
+      "step": 1890
+    },
+    {
+      "epoch": 0.39635296583525464,
+      "grad_norm": 0.10044901072978973,
+      "learning_rate": 6.879256625828925e-06,
+      "loss": 0.057,
+      "step": 1891
+    },
+    {
+      "epoch": 0.3965625654998952,
+      "grad_norm": 0.13600683212280273,
+      "learning_rate": 6.8761102549178706e-06,
+      "loss": 0.0614,
+      "step": 1892
+    },
+    {
+      "epoch": 0.39677216516453573,
+      "grad_norm": 0.1270650029182434,
+      "learning_rate": 6.872963019120377e-06,
+      "loss": 0.0601,
+      "step": 1893
+    },
+    {
+      "epoch": 0.3969817648291763,
+      "grad_norm": 0.10988420993089676,
+      "learning_rate": 6.869814919887321e-06,
+      "loss": 0.0561,
+      "step": 1894
+    },
+    {
+      "epoch": 0.3971913644938168,
+      "grad_norm": 0.11587254703044891,
+      "learning_rate": 6.866665958669976e-06,
+      "loss": 0.0584,
+      "step": 1895
+    },
+    {
+      "epoch": 0.3974009641584573,
+      "grad_norm": 0.1132350042462349,
+      "learning_rate": 6.863516136920012e-06,
+      "loss": 0.0598,
+      "step": 1896
+    },
+    {
+      "epoch": 0.3976105638230979,
+      "grad_norm": 0.14506936073303223,
+      "learning_rate": 6.860365456089495e-06,
+      "loss": 0.058,
+      "step": 1897
+    },
+    {
+      "epoch": 0.3978201634877384,
+      "grad_norm": 0.12474244832992554,
+      "learning_rate": 6.8572139176308885e-06,
+      "loss": 0.0584,
+      "step": 1898
+    },
+    {
+      "epoch": 0.398029763152379,
+      "grad_norm": 0.1177903488278389,
+      "learning_rate": 6.854061522997053e-06,
+      "loss": 0.0563,
+      "step": 1899
+    },
+    {
+      "epoch": 0.3982393628170195,
+      "grad_norm": 0.11654134839773178,
+      "learning_rate": 6.85090827364124e-06,
+      "loss": 0.0556,
+      "step": 1900
+    },
+    {
+      "epoch": 0.39844896248166,
+      "grad_norm": 0.10216284543275833,
+      "learning_rate": 6.847754171017097e-06,
+      "loss": 0.0629,
+      "step": 1901
+    },
+    {
+      "epoch": 0.3986585621463006,
+      "grad_norm": 0.11993620544672012,
+      "learning_rate": 6.844599216578667e-06,
+      "loss": 0.0565,
+      "step": 1902
+    },
+    {
+      "epoch": 0.3988681618109411,
+      "grad_norm": 0.10797383636236191,
+      "learning_rate": 6.8414434117803785e-06,
+      "loss": 0.0574,
+      "step": 1903
+    },
+    {
+      "epoch": 0.39907776147558166,
+      "grad_norm": 0.111129529774189,
+      "learning_rate": 6.83828675807706e-06,
+      "loss": 0.0593,
+      "step": 1904
+    },
+    {
+      "epoch": 0.3992873611402222,
+      "grad_norm": 0.1041678711771965,
+      "learning_rate": 6.835129256923931e-06,
+      "loss": 0.0569,
+      "step": 1905
+    },
+    {
+      "epoch": 0.3994969608048627,
+      "grad_norm": 0.09219536185264587,
+      "learning_rate": 6.831970909776593e-06,
+      "loss": 0.0573,
+      "step": 1906
+    },
+    {
+      "epoch": 0.39970656046950326,
+      "grad_norm": 0.10893040150403976,
+      "learning_rate": 6.828811718091046e-06,
+      "loss": 0.0586,
+      "step": 1907
+    },
+    {
+      "epoch": 0.39991616013414377,
+      "grad_norm": 0.08363223820924759,
+      "learning_rate": 6.8256516833236774e-06,
+      "loss": 0.0584,
+      "step": 1908
+    },
+    {
+      "epoch": 0.40012575979878434,
+      "grad_norm": 0.1113632470369339,
+      "learning_rate": 6.822490806931262e-06,
+      "loss": 0.0603,
+      "step": 1909
+    },
+    {
+      "epoch": 0.40033535946342486,
+      "grad_norm": 0.13083884119987488,
+      "learning_rate": 6.819329090370964e-06,
+      "loss": 0.0586,
+      "step": 1910
+    },
+    {
+      "epoch": 0.40054495912806537,
+      "grad_norm": 0.12002018094062805,
+      "learning_rate": 6.816166535100332e-06,
+      "loss": 0.0589,
+      "step": 1911
+    },
+    {
+      "epoch": 0.40075455879270594,
+      "grad_norm": 0.13928495347499847,
+      "learning_rate": 6.813003142577306e-06,
+      "loss": 0.0586,
+      "step": 1912
+    },
+    {
+      "epoch": 0.40096415845734645,
+      "grad_norm": 0.14167679846286774,
+      "learning_rate": 6.809838914260208e-06,
+      "loss": 0.0567,
+      "step": 1913
+    },
+    {
+      "epoch": 0.401173758121987,
+      "grad_norm": 0.13765917718410492,
+      "learning_rate": 6.806673851607745e-06,
+      "loss": 0.0596,
+      "step": 1914
+    },
+    {
+      "epoch": 0.40138335778662754,
+      "grad_norm": 0.11585170775651932,
+      "learning_rate": 6.803507956079012e-06,
+      "loss": 0.0562,
+      "step": 1915
+    },
+    {
+      "epoch": 0.40159295745126805,
+      "grad_norm": 0.1145220398902893,
+      "learning_rate": 6.800341229133486e-06,
+      "loss": 0.0572,
+      "step": 1916
+    },
+    {
+      "epoch": 0.4018025571159086,
+      "grad_norm": 0.11287350952625275,
+      "learning_rate": 6.797173672231027e-06,
+      "loss": 0.0581,
+      "step": 1917
+    },
+    {
+      "epoch": 0.40201215678054913,
+      "grad_norm": 0.0974586084485054,
+      "learning_rate": 6.794005286831878e-06,
+      "loss": 0.0552,
+      "step": 1918
+    },
+    {
+      "epoch": 0.4022217564451897,
+      "grad_norm": 0.10374786704778671,
+      "learning_rate": 6.790836074396666e-06,
+      "loss": 0.0583,
+      "step": 1919
+    },
+    {
+      "epoch": 0.4024313561098302,
+      "grad_norm": 0.11808675527572632,
+      "learning_rate": 6.7876660363863955e-06,
+      "loss": 0.0586,
+      "step": 1920
+    },
+    {
+      "epoch": 0.4026409557744708,
+      "grad_norm": 0.11047716438770294,
+      "learning_rate": 6.784495174262452e-06,
+      "loss": 0.0589,
+      "step": 1921
+    },
+    {
+      "epoch": 0.4028505554391113,
+      "grad_norm": 0.09720434248447418,
+      "learning_rate": 6.781323489486606e-06,
+      "loss": 0.0586,
+      "step": 1922
+    },
+    {
+      "epoch": 0.4030601551037518,
+      "grad_norm": 0.11186391115188599,
+      "learning_rate": 6.778150983520999e-06,
+      "loss": 0.0589,
+      "step": 1923
+    },
+    {
+      "epoch": 0.4032697547683924,
+      "grad_norm": 0.11044489592313766,
+      "learning_rate": 6.774977657828159e-06,
+      "loss": 0.0598,
+      "step": 1924
+    },
+    {
+      "epoch": 0.4034793544330329,
+      "grad_norm": 0.08856026083230972,
+      "learning_rate": 6.771803513870988e-06,
+      "loss": 0.0551,
+      "step": 1925
+    },
+    {
+      "epoch": 0.40368895409767347,
+      "grad_norm": 0.11433659493923187,
+      "learning_rate": 6.768628553112763e-06,
+      "loss": 0.0583,
+      "step": 1926
+    },
+    {
+      "epoch": 0.403898553762314,
+      "grad_norm": 0.12080615013837814,
+      "learning_rate": 6.765452777017146e-06,
+      "loss": 0.0574,
+      "step": 1927
+    },
+    {
+      "epoch": 0.4041081534269545,
+      "grad_norm": 0.09817066043615341,
+      "learning_rate": 6.762276187048164e-06,
+      "loss": 0.0604,
+      "step": 1928
+    },
+    {
+      "epoch": 0.40431775309159507,
+      "grad_norm": 0.14070898294448853,
+      "learning_rate": 6.759098784670224e-06,
+      "loss": 0.058,
+      "step": 1929
+    },
+    {
+      "epoch": 0.4045273527562356,
+      "grad_norm": 0.16221067309379578,
+      "learning_rate": 6.755920571348111e-06,
+      "loss": 0.0593,
+      "step": 1930
+    },
+    {
+      "epoch": 0.40473695242087615,
+      "grad_norm": 0.14239110052585602,
+      "learning_rate": 6.75274154854698e-06,
+      "loss": 0.0559,
+      "step": 1931
+    },
+    {
+      "epoch": 0.40494655208551666,
+      "grad_norm": 0.1578344851732254,
+      "learning_rate": 6.749561717732359e-06,
+      "loss": 0.0568,
+      "step": 1932
+    },
+    {
+      "epoch": 0.4051561517501572,
+      "grad_norm": 0.11431273072957993,
+      "learning_rate": 6.7463810803701495e-06,
+      "loss": 0.0581,
+      "step": 1933
+    },
+    {
+      "epoch": 0.40536575141479775,
+      "grad_norm": 0.10527955740690231,
+      "learning_rate": 6.743199637926623e-06,
+      "loss": 0.0593,
+      "step": 1934
+    },
+    {
+      "epoch": 0.40557535107943826,
+      "grad_norm": 0.13100096583366394,
+      "learning_rate": 6.740017391868427e-06,
+      "loss": 0.059,
+      "step": 1935
+    },
+    {
+      "epoch": 0.40578495074407883,
+      "grad_norm": 0.11144158244132996,
+      "learning_rate": 6.736834343662576e-06,
+      "loss": 0.0579,
+      "step": 1936
+    },
+    {
+      "epoch": 0.40599455040871935,
+      "grad_norm": 0.13539659976959229,
+      "learning_rate": 6.73365049477645e-06,
+      "loss": 0.0581,
+      "step": 1937
+    },
+    {
+      "epoch": 0.40620415007335986,
+      "grad_norm": 0.13399067521095276,
+      "learning_rate": 6.7304658466778095e-06,
+      "loss": 0.0582,
+      "step": 1938
+    },
+    {
+      "epoch": 0.40641374973800043,
+      "grad_norm": 0.11665608733892441,
+      "learning_rate": 6.7272804008347705e-06,
+      "loss": 0.0557,
+      "step": 1939
+    },
+    {
+      "epoch": 0.40662334940264094,
+      "grad_norm": 0.12888075411319733,
+      "learning_rate": 6.7240941587158274e-06,
+      "loss": 0.0614,
+      "step": 1940
+    },
+    {
+      "epoch": 0.4068329490672815,
+      "grad_norm": 0.11636313796043396,
+      "learning_rate": 6.720907121789835e-06,
+      "loss": 0.0591,
+      "step": 1941
+    },
+    {
+      "epoch": 0.407042548731922,
+      "grad_norm": 0.11793368309736252,
+      "learning_rate": 6.717719291526016e-06,
+      "loss": 0.0597,
+      "step": 1942
+    },
+    {
+      "epoch": 0.40725214839656254,
+      "grad_norm": 0.13560403883457184,
+      "learning_rate": 6.71453066939396e-06,
+      "loss": 0.0574,
+      "step": 1943
+    },
+    {
+      "epoch": 0.4074617480612031,
+      "grad_norm": 0.09925508499145508,
+      "learning_rate": 6.711341256863623e-06,
+      "loss": 0.0591,
+      "step": 1944
+    },
+    {
+      "epoch": 0.4076713477258436,
+      "grad_norm": 0.1104302629828453,
+      "learning_rate": 6.708151055405321e-06,
+      "loss": 0.0595,
+      "step": 1945
+    },
+    {
+      "epoch": 0.4078809473904842,
+      "grad_norm": 0.1431284248828888,
+      "learning_rate": 6.704960066489738e-06,
+      "loss": 0.0606,
+      "step": 1946
+    },
+    {
+      "epoch": 0.4080905470551247,
+      "grad_norm": 0.1400553286075592,
+      "learning_rate": 6.701768291587918e-06,
+      "loss": 0.0602,
+      "step": 1947
+    },
+    {
+      "epoch": 0.4083001467197652,
+      "grad_norm": 0.16290639340877533,
+      "learning_rate": 6.698575732171269e-06,
+      "loss": 0.0591,
+      "step": 1948
+    },
+    {
+      "epoch": 0.4085097463844058,
+      "grad_norm": 0.14287681877613068,
+      "learning_rate": 6.695382389711561e-06,
+      "loss": 0.0571,
+      "step": 1949
+    },
+    {
+      "epoch": 0.4087193460490463,
+      "grad_norm": 0.13841015100479126,
+      "learning_rate": 6.692188265680923e-06,
+      "loss": 0.0579,
+      "step": 1950
+    },
+    {
+      "epoch": 0.4089289457136869,
+      "grad_norm": 0.1272697001695633,
+      "learning_rate": 6.688993361551847e-06,
+      "loss": 0.0557,
+      "step": 1951
+    },
+    {
+      "epoch": 0.4091385453783274,
+      "grad_norm": 0.16662158071994781,
+      "learning_rate": 6.6857976787971815e-06,
+      "loss": 0.0621,
+      "step": 1952
+    },
+    {
+      "epoch": 0.4093481450429679,
+      "grad_norm": 0.18397577106952667,
+      "learning_rate": 6.682601218890136e-06,
+      "loss": 0.06,
+      "step": 1953
+    },
+    {
+      "epoch": 0.4095577447076085,
+      "grad_norm": 0.14603066444396973,
+      "learning_rate": 6.679403983304278e-06,
+      "loss": 0.0582,
+      "step": 1954
+    },
+    {
+      "epoch": 0.409767344372249,
+      "grad_norm": 0.1586339771747589,
+      "learning_rate": 6.6762059735135325e-06,
+      "loss": 0.0604,
+      "step": 1955
+    },
+    {
+      "epoch": 0.40997694403688956,
+      "grad_norm": 0.1153809055685997,
+      "learning_rate": 6.673007190992181e-06,
+      "loss": 0.0626,
+      "step": 1956
+    },
+    {
+      "epoch": 0.41018654370153007,
+      "grad_norm": 0.15605716407299042,
+      "learning_rate": 6.669807637214862e-06,
+      "loss": 0.0584,
+      "step": 1957
+    },
+    {
+      "epoch": 0.41039614336617064,
+      "grad_norm": 0.17064720392227173,
+      "learning_rate": 6.66660731365657e-06,
+      "loss": 0.0572,
+      "step": 1958
+    },
+    {
+      "epoch": 0.41060574303081115,
+      "grad_norm": 0.141871839761734,
+      "learning_rate": 6.663406221792652e-06,
+      "loss": 0.0552,
+      "step": 1959
+    },
+    {
+      "epoch": 0.41081534269545167,
+      "grad_norm": 0.14091584086418152,
+      "learning_rate": 6.660204363098812e-06,
+      "loss": 0.06,
+      "step": 1960
+    },
+    {
+      "epoch": 0.41102494236009224,
+      "grad_norm": 0.10206598043441772,
+      "learning_rate": 6.657001739051105e-06,
+      "loss": 0.0599,
+      "step": 1961
+    },
+    {
+      "epoch": 0.41123454202473275,
+      "grad_norm": 0.1338677704334259,
+      "learning_rate": 6.65379835112594e-06,
+      "loss": 0.0609,
+      "step": 1962
+    },
+    {
+      "epoch": 0.4114441416893733,
+      "grad_norm": 0.13361436128616333,
+      "learning_rate": 6.65059420080008e-06,
+      "loss": 0.058,
+      "step": 1963
+    },
+    {
+      "epoch": 0.41165374135401384,
+      "grad_norm": 0.15927137434482574,
+      "learning_rate": 6.647389289550635e-06,
+      "loss": 0.061,
+      "step": 1964
+    },
+    {
+      "epoch": 0.41186334101865435,
+      "grad_norm": 0.11594053357839584,
+      "learning_rate": 6.64418361885507e-06,
+      "loss": 0.0563,
+      "step": 1965
+    },
+    {
+      "epoch": 0.4120729406832949,
+      "grad_norm": 0.10348820686340332,
+      "learning_rate": 6.640977190191198e-06,
+      "loss": 0.0613,
+      "step": 1966
+    },
+    {
+      "epoch": 0.41228254034793543,
+      "grad_norm": 0.14770418405532837,
+      "learning_rate": 6.637770005037182e-06,
+      "loss": 0.0593,
+      "step": 1967
+    },
+    {
+      "epoch": 0.412492140012576,
+      "grad_norm": 0.12195451557636261,
+      "learning_rate": 6.634562064871535e-06,
+      "loss": 0.0585,
+      "step": 1968
+    },
+    {
+      "epoch": 0.4127017396772165,
+      "grad_norm": 0.11958389729261398,
+      "learning_rate": 6.631353371173115e-06,
+      "loss": 0.058,
+      "step": 1969
+    },
+    {
+      "epoch": 0.41291133934185703,
+      "grad_norm": 0.09207571297883987,
+      "learning_rate": 6.628143925421129e-06,
+      "loss": 0.0556,
+      "step": 1970
+    },
+    {
+      "epoch": 0.4131209390064976,
+      "grad_norm": 0.10654924809932709,
+      "learning_rate": 6.624933729095133e-06,
+      "loss": 0.0564,
+      "step": 1971
+    },
+    {
+      "epoch": 0.4133305386711381,
+      "grad_norm": 0.12806440889835358,
+      "learning_rate": 6.621722783675024e-06,
+      "loss": 0.0573,
+      "step": 1972
+    },
+    {
+      "epoch": 0.4135401383357787,
+      "grad_norm": 0.15015651285648346,
+      "learning_rate": 6.6185110906410485e-06,
+      "loss": 0.0567,
+      "step": 1973
+    },
+    {
+      "epoch": 0.4137497380004192,
+      "grad_norm": 0.13939987123012543,
+      "learning_rate": 6.615298651473799e-06,
+      "loss": 0.058,
+      "step": 1974
+    },
+    {
+      "epoch": 0.4139593376650597,
+      "grad_norm": 0.10312556475400925,
+      "learning_rate": 6.612085467654204e-06,
+      "loss": 0.0589,
+      "step": 1975
+    },
+    {
+      "epoch": 0.4141689373297003,
+      "grad_norm": 0.11151403933763504,
+      "learning_rate": 6.608871540663544e-06,
+      "loss": 0.0598,
+      "step": 1976
+    },
+    {
+      "epoch": 0.4143785369943408,
+      "grad_norm": 0.10755238682031631,
+      "learning_rate": 6.605656871983439e-06,
+      "loss": 0.0598,
+      "step": 1977
+    },
+    {
+      "epoch": 0.41458813665898137,
+      "grad_norm": 0.09601911902427673,
+      "learning_rate": 6.6024414630958475e-06,
+      "loss": 0.0571,
+      "step": 1978
+    },
+    {
+      "epoch": 0.4147977363236219,
+      "grad_norm": 0.11368954926729202,
+      "learning_rate": 6.599225315483076e-06,
+      "loss": 0.0556,
+      "step": 1979
+    },
+    {
+      "epoch": 0.4150073359882624,
+      "grad_norm": 0.11148924380540848,
+      "learning_rate": 6.596008430627766e-06,
+      "loss": 0.0572,
+      "step": 1980
+    },
+    {
+      "epoch": 0.41521693565290296,
+      "grad_norm": 0.11703907698392868,
+      "learning_rate": 6.592790810012901e-06,
+      "loss": 0.0583,
+      "step": 1981
+    },
+    {
+      "epoch": 0.4154265353175435,
+      "grad_norm": 0.1259683072566986,
+      "learning_rate": 6.589572455121804e-06,
+      "loss": 0.0599,
+      "step": 1982
+    },
+    {
+      "epoch": 0.41563613498218405,
+      "grad_norm": 0.12415824085474014,
+      "learning_rate": 6.586353367438134e-06,
+      "loss": 0.0568,
+      "step": 1983
+    },
+    {
+      "epoch": 0.41584573464682456,
+      "grad_norm": 0.11688219755887985,
+      "learning_rate": 6.5831335484458915e-06,
+      "loss": 0.0564,
+      "step": 1984
+    },
+    {
+      "epoch": 0.4160553343114651,
+      "grad_norm": 0.10014292597770691,
+      "learning_rate": 6.579912999629412e-06,
+      "loss": 0.0595,
+      "step": 1985
+    },
+    {
+      "epoch": 0.41626493397610564,
+      "grad_norm": 0.09642448276281357,
+      "learning_rate": 6.576691722473368e-06,
+      "loss": 0.0587,
+      "step": 1986
+    },
+    {
+      "epoch": 0.41647453364074616,
+      "grad_norm": 0.10601100325584412,
+      "learning_rate": 6.573469718462768e-06,
+      "loss": 0.0572,
+      "step": 1987
+    },
+    {
+      "epoch": 0.41668413330538673,
+      "grad_norm": 0.10878970474004745,
+      "learning_rate": 6.570246989082954e-06,
+      "loss": 0.0598,
+      "step": 1988
+    },
+    {
+      "epoch": 0.41689373297002724,
+      "grad_norm": 0.11488792300224304,
+      "learning_rate": 6.567023535819605e-06,
+      "loss": 0.0606,
+      "step": 1989
+    },
+    {
+      "epoch": 0.4171033326346678,
+      "grad_norm": 0.1324160397052765,
+      "learning_rate": 6.5637993601587305e-06,
+      "loss": 0.058,
+      "step": 1990
+    },
+    {
+      "epoch": 0.4173129322993083,
+      "grad_norm": 0.1504766047000885,
+      "learning_rate": 6.560574463586677e-06,
+      "loss": 0.0565,
+      "step": 1991
+    },
+    {
+      "epoch": 0.41752253196394884,
+      "grad_norm": 0.1516587734222412,
+      "learning_rate": 6.557348847590118e-06,
+      "loss": 0.0564,
+      "step": 1992
+    },
+    {
+      "epoch": 0.4177321316285894,
+      "grad_norm": 0.14044304192066193,
+      "learning_rate": 6.554122513656065e-06,
+      "loss": 0.0586,
+      "step": 1993
+    },
+    {
+      "epoch": 0.4179417312932299,
+      "grad_norm": 0.11305470019578934,
+      "learning_rate": 6.550895463271856e-06,
+      "loss": 0.0593,
+      "step": 1994
+    },
+    {
+      "epoch": 0.4181513309578705,
+      "grad_norm": 0.11471327394247055,
+      "learning_rate": 6.547667697925161e-06,
+      "loss": 0.055,
+      "step": 1995
+    },
+    {
+      "epoch": 0.418360930622511,
+      "grad_norm": 0.15590576827526093,
+      "learning_rate": 6.544439219103977e-06,
+      "loss": 0.059,
+      "step": 1996
+    },
+    {
+      "epoch": 0.4185705302871515,
+      "grad_norm": 0.16013385355472565,
+      "learning_rate": 6.5412100282966366e-06,
+      "loss": 0.0576,
+      "step": 1997
+    },
+    {
+      "epoch": 0.4187801299517921,
+      "grad_norm": 0.13815461099147797,
+      "learning_rate": 6.53798012699179e-06,
+      "loss": 0.0577,
+      "step": 1998
+    },
+    {
+      "epoch": 0.4189897296164326,
+      "grad_norm": 0.11431432515382767,
+      "learning_rate": 6.534749516678427e-06,
+      "loss": 0.0585,
+      "step": 1999
+    },
+    {
+      "epoch": 0.4191993292810732,
+      "grad_norm": 0.1008606106042862,
+      "learning_rate": 6.531518198845854e-06,
+      "loss": 0.0561,
+      "step": 2000
+    },
+    {
+      "epoch": 0.4194089289457137,
+      "grad_norm": 0.10297921299934387,
+      "learning_rate": 6.52828617498371e-06,
+      "loss": 0.0571,
+      "step": 2001
+    },
+    {
+      "epoch": 0.4196185286103542,
+      "grad_norm": 0.10644082725048065,
+      "learning_rate": 6.525053446581957e-06,
+      "loss": 0.0551,
+      "step": 2002
+    },
+    {
+      "epoch": 0.41982812827499477,
+      "grad_norm": 0.10528367012739182,
+      "learning_rate": 6.52182001513088e-06,
+      "loss": 0.0583,
+      "step": 2003
+    },
+    {
+      "epoch": 0.4200377279396353,
+      "grad_norm": 0.09672725200653076,
+      "learning_rate": 6.518585882121096e-06,
+      "loss": 0.0558,
+      "step": 2004
+    },
+    {
+      "epoch": 0.42024732760427586,
+      "grad_norm": 0.09464428573846817,
+      "learning_rate": 6.515351049043533e-06,
+      "loss": 0.058,
+      "step": 2005
+    },
+    {
+      "epoch": 0.42045692726891637,
+      "grad_norm": 0.10385248810052872,
+      "learning_rate": 6.5121155173894515e-06,
+      "loss": 0.0584,
+      "step": 2006
+    },
+    {
+      "epoch": 0.4206665269335569,
+      "grad_norm": 0.10311248153448105,
+      "learning_rate": 6.508879288650431e-06,
+      "loss": 0.0564,
+      "step": 2007
+    },
+    {
+      "epoch": 0.42087612659819745,
+      "grad_norm": 0.08637700229883194,
+      "learning_rate": 6.505642364318372e-06,
+      "loss": 0.0596,
+      "step": 2008
+    },
+    {
+      "epoch": 0.42108572626283797,
+      "grad_norm": 0.10172644257545471,
+      "learning_rate": 6.502404745885495e-06,
+      "loss": 0.0594,
+      "step": 2009
+    },
+    {
+      "epoch": 0.42129532592747854,
+      "grad_norm": 0.12740617990493774,
+      "learning_rate": 6.499166434844344e-06,
+      "loss": 0.058,
+      "step": 2010
+    },
+    {
+      "epoch": 0.42150492559211905,
+      "grad_norm": 0.12112673372030258,
+      "learning_rate": 6.495927432687777e-06,
+      "loss": 0.0566,
+      "step": 2011
+    },
+    {
+      "epoch": 0.42171452525675956,
+      "grad_norm": 0.10495518893003464,
+      "learning_rate": 6.492687740908973e-06,
+      "loss": 0.0581,
+      "step": 2012
+    },
+    {
+      "epoch": 0.42192412492140013,
+      "grad_norm": 0.11961285024881363,
+      "learning_rate": 6.489447361001431e-06,
+      "loss": 0.0561,
+      "step": 2013
+    },
+    {
+      "epoch": 0.42213372458604065,
+      "grad_norm": 0.1219894215464592,
+      "learning_rate": 6.486206294458966e-06,
+      "loss": 0.0599,
+      "step": 2014
+    },
+    {
+      "epoch": 0.4223433242506812,
+      "grad_norm": 0.11397355794906616,
+      "learning_rate": 6.482964542775707e-06,
+      "loss": 0.0564,
+      "step": 2015
+    },
+    {
+      "epoch": 0.42255292391532173,
+      "grad_norm": 0.10773434489965439,
+      "learning_rate": 6.479722107446102e-06,
+      "loss": 0.0577,
+      "step": 2016
+    },
+    {
+      "epoch": 0.42276252357996225,
+      "grad_norm": 0.09557615965604782,
+      "learning_rate": 6.476478989964914e-06,
+      "loss": 0.0588,
+      "step": 2017
+    },
+    {
+      "epoch": 0.4229721232446028,
+      "grad_norm": 0.09783432632684708,
+      "learning_rate": 6.473235191827219e-06,
+      "loss": 0.056,
+      "step": 2018
+    },
+    {
+      "epoch": 0.42318172290924333,
+      "grad_norm": 0.10292766988277435,
+      "learning_rate": 6.469990714528403e-06,
+      "loss": 0.0593,
+      "step": 2019
+    },
+    {
+      "epoch": 0.4233913225738839,
+      "grad_norm": 0.09302861243486404,
+      "learning_rate": 6.466745559564175e-06,
+      "loss": 0.0565,
+      "step": 2020
+    },
+    {
+      "epoch": 0.4236009222385244,
+      "grad_norm": 0.1034521758556366,
+      "learning_rate": 6.463499728430549e-06,
+      "loss": 0.0567,
+      "step": 2021
+    },
+    {
+      "epoch": 0.4238105219031649,
+      "grad_norm": 0.11888016015291214,
+      "learning_rate": 6.460253222623851e-06,
+      "loss": 0.0578,
+      "step": 2022
+    },
+    {
+      "epoch": 0.4240201215678055,
+      "grad_norm": 0.11663860082626343,
+      "learning_rate": 6.457006043640722e-06,
+      "loss": 0.0627,
+      "step": 2023
+    },
+    {
+      "epoch": 0.424229721232446,
+      "grad_norm": 0.11908293515443802,
+      "learning_rate": 6.453758192978109e-06,
+      "loss": 0.0577,
+      "step": 2024
+    },
+    {
+      "epoch": 0.4244393208970866,
+      "grad_norm": 0.11339572072029114,
+      "learning_rate": 6.450509672133272e-06,
+      "loss": 0.0609,
+      "step": 2025
+    },
+    {
+      "epoch": 0.4246489205617271,
+      "grad_norm": 0.10132212191820145,
+      "learning_rate": 6.447260482603776e-06,
+      "loss": 0.0551,
+      "step": 2026
+    },
+    {
+      "epoch": 0.42485852022636766,
+      "grad_norm": 0.119560107588768,
+      "learning_rate": 6.444010625887498e-06,
+      "loss": 0.0589,
+      "step": 2027
+    },
+    {
+      "epoch": 0.4250681198910082,
+      "grad_norm": 0.12544332444667816,
+      "learning_rate": 6.4407601034826225e-06,
+      "loss": 0.0594,
+      "step": 2028
+    },
+    {
+      "epoch": 0.4252777195556487,
+      "grad_norm": 0.10760358721017838,
+      "learning_rate": 6.437508916887638e-06,
+      "loss": 0.0583,
+      "step": 2029
+    },
+    {
+      "epoch": 0.42548731922028926,
+      "grad_norm": 0.08328445255756378,
+      "learning_rate": 6.4342570676013415e-06,
+      "loss": 0.0586,
+      "step": 2030
+    },
+    {
+      "epoch": 0.4256969188849298,
+      "grad_norm": 0.09850215911865234,
+      "learning_rate": 6.4310045571228344e-06,
+      "loss": 0.0574,
+      "step": 2031
+    },
+    {
+      "epoch": 0.42590651854957035,
+      "grad_norm": 0.11578439176082611,
+      "learning_rate": 6.427751386951525e-06,
+      "loss": 0.0607,
+      "step": 2032
+    },
+    {
+      "epoch": 0.42611611821421086,
+      "grad_norm": 0.10482225567102432,
+      "learning_rate": 6.424497558587122e-06,
+      "loss": 0.0572,
+      "step": 2033
+    },
+    {
+      "epoch": 0.4263257178788514,
+      "grad_norm": 0.11043455451726913,
+      "learning_rate": 6.421243073529639e-06,
+      "loss": 0.0568,
+      "step": 2034
+    },
+    {
+      "epoch": 0.42653531754349194,
+      "grad_norm": 0.12734782695770264,
+      "learning_rate": 6.417987933279397e-06,
+      "loss": 0.0573,
+      "step": 2035
+    },
+    {
+      "epoch": 0.42674491720813246,
+      "grad_norm": 0.11755295097827911,
+      "learning_rate": 6.41473213933701e-06,
+      "loss": 0.0586,
+      "step": 2036
+    },
+    {
+      "epoch": 0.426954516872773,
+      "grad_norm": 0.10255907475948334,
+      "learning_rate": 6.411475693203402e-06,
+      "loss": 0.0627,
+      "step": 2037
+    },
+    {
+      "epoch": 0.42716411653741354,
+      "grad_norm": 0.07925237715244293,
+      "learning_rate": 6.408218596379792e-06,
+      "loss": 0.0562,
+      "step": 2038
+    },
+    {
+      "epoch": 0.42737371620205405,
+      "grad_norm": 0.07128457725048065,
+      "learning_rate": 6.404960850367701e-06,
+      "loss": 0.0545,
+      "step": 2039
+    },
+    {
+      "epoch": 0.4275833158666946,
+      "grad_norm": 0.08888307213783264,
+      "learning_rate": 6.4017024566689515e-06,
+      "loss": 0.0587,
+      "step": 2040
+    },
+    {
+      "epoch": 0.42779291553133514,
+      "grad_norm": 0.09860040247440338,
+      "learning_rate": 6.39844341678566e-06,
+      "loss": 0.0571,
+      "step": 2041
+    },
+    {
+      "epoch": 0.4280025151959757,
+      "grad_norm": 0.09772216528654099,
+      "learning_rate": 6.395183732220242e-06,
+      "loss": 0.0591,
+      "step": 2042
+    },
+    {
+      "epoch": 0.4282121148606162,
+      "grad_norm": 0.08198471367359161,
+      "learning_rate": 6.391923404475416e-06,
+      "loss": 0.0592,
+      "step": 2043
+    },
+    {
+      "epoch": 0.42842171452525674,
+      "grad_norm": 0.07226858288049698,
+      "learning_rate": 6.388662435054187e-06,
+      "loss": 0.0569,
+      "step": 2044
+    },
+    {
+      "epoch": 0.4286313141898973,
+      "grad_norm": 0.1091211661696434,
+      "learning_rate": 6.385400825459865e-06,
+      "loss": 0.0546,
+      "step": 2045
+    },
+    {
+      "epoch": 0.4288409138545378,
+      "grad_norm": 0.11992846429347992,
+      "learning_rate": 6.382138577196052e-06,
+      "loss": 0.0588,
+      "step": 2046
+    },
+    {
+      "epoch": 0.4290505135191784,
+      "grad_norm": 0.09196002781391144,
+      "learning_rate": 6.378875691766639e-06,
+      "loss": 0.0565,
+      "step": 2047
+    },
+    {
+      "epoch": 0.4292601131838189,
+      "grad_norm": 0.15287606418132782,
+      "learning_rate": 6.375612170675821e-06,
+      "loss": 0.0568,
+      "step": 2048
+    },
+    {
+      "epoch": 0.4294697128484594,
+      "grad_norm": 0.15699946880340576,
+      "learning_rate": 6.372348015428077e-06,
+      "loss": 0.0595,
+      "step": 2049
+    },
+    {
+      "epoch": 0.4296793125131,
+      "grad_norm": 0.12692134082317352,
+      "learning_rate": 6.3690832275281835e-06,
+      "loss": 0.059,
+      "step": 2050
+    },
+    {
+      "epoch": 0.4298889121777405,
+      "grad_norm": 0.14513273537158966,
+      "learning_rate": 6.3658178084812065e-06,
+      "loss": 0.0585,
+      "step": 2051
+    },
+    {
+      "epoch": 0.43009851184238107,
+      "grad_norm": 0.13621389865875244,
+      "learning_rate": 6.3625517597925025e-06,
+      "loss": 0.0557,
+      "step": 2052
+    },
+    {
+      "epoch": 0.4303081115070216,
+      "grad_norm": 0.14017432928085327,
+      "learning_rate": 6.359285082967721e-06,
+      "loss": 0.0575,
+      "step": 2053
+    },
+    {
+      "epoch": 0.4305177111716621,
+      "grad_norm": 0.11645541340112686,
+      "learning_rate": 6.356017779512799e-06,
+      "loss": 0.0586,
+      "step": 2054
+    },
+    {
+      "epoch": 0.43072731083630267,
+      "grad_norm": 0.12075808644294739,
+      "learning_rate": 6.352749850933961e-06,
+      "loss": 0.0586,
+      "step": 2055
+    },
+    {
+      "epoch": 0.4309369105009432,
+      "grad_norm": 0.1253325641155243,
+      "learning_rate": 6.349481298737723e-06,
+      "loss": 0.057,
+      "step": 2056
+    },
+    {
+      "epoch": 0.43114651016558375,
+      "grad_norm": 0.12212934345006943,
+      "learning_rate": 6.346212124430888e-06,
+      "loss": 0.0561,
+      "step": 2057
+    },
+    {
+      "epoch": 0.43135610983022427,
+      "grad_norm": 0.10774050652980804,
+      "learning_rate": 6.342942329520543e-06,
+      "loss": 0.0599,
+      "step": 2058
+    },
+    {
+      "epoch": 0.43156570949486484,
+      "grad_norm": 0.08897628635168076,
+      "learning_rate": 6.339671915514062e-06,
+      "loss": 0.0602,
+      "step": 2059
+    },
+    {
+      "epoch": 0.43177530915950535,
+      "grad_norm": 0.12079823017120361,
+      "learning_rate": 6.336400883919106e-06,
+      "loss": 0.0568,
+      "step": 2060
+    },
+    {
+      "epoch": 0.43198490882414586,
+      "grad_norm": 0.11661111563444138,
+      "learning_rate": 6.33312923624362e-06,
+      "loss": 0.0562,
+      "step": 2061
+    },
+    {
+      "epoch": 0.43219450848878643,
+      "grad_norm": 0.09649336338043213,
+      "learning_rate": 6.329856973995835e-06,
+      "loss": 0.0592,
+      "step": 2062
+    },
+    {
+      "epoch": 0.43240410815342695,
+      "grad_norm": 0.12174370884895325,
+      "learning_rate": 6.32658409868426e-06,
+      "loss": 0.0586,
+      "step": 2063
+    },
+    {
+      "epoch": 0.4326137078180675,
+      "grad_norm": 0.13494165241718292,
+      "learning_rate": 6.32331061181769e-06,
+      "loss": 0.0575,
+      "step": 2064
+    },
+    {
+      "epoch": 0.43282330748270803,
+      "grad_norm": 0.1156393364071846,
+      "learning_rate": 6.320036514905204e-06,
+      "loss": 0.0577,
+      "step": 2065
+    },
+    {
+      "epoch": 0.43303290714734854,
+      "grad_norm": 0.11403451859951019,
+      "learning_rate": 6.316761809456159e-06,
+      "loss": 0.0599,
+      "step": 2066
+    },
+    {
+      "epoch": 0.4332425068119891,
+      "grad_norm": 0.1350114345550537,
+      "learning_rate": 6.313486496980192e-06,
+      "loss": 0.0572,
+      "step": 2067
+    },
+    {
+      "epoch": 0.43345210647662963,
+      "grad_norm": 0.10050825029611588,
+      "learning_rate": 6.310210578987225e-06,
+      "loss": 0.0613,
+      "step": 2068
+    },
+    {
+      "epoch": 0.4336617061412702,
+      "grad_norm": 0.1249486580491066,
+      "learning_rate": 6.306934056987452e-06,
+      "loss": 0.0597,
+      "step": 2069
+    },
+    {
+      "epoch": 0.4338713058059107,
+      "grad_norm": 0.15121574699878693,
+      "learning_rate": 6.303656932491349e-06,
+      "loss": 0.0558,
+      "step": 2070
+    },
+    {
+      "epoch": 0.4340809054705512,
+      "grad_norm": 0.12158171832561493,
+      "learning_rate": 6.3003792070096735e-06,
+      "loss": 0.059,
+      "step": 2071
+    },
+    {
+      "epoch": 0.4342905051351918,
+      "grad_norm": 0.11560632288455963,
+      "learning_rate": 6.297100882053451e-06,
+      "loss": 0.0586,
+      "step": 2072
+    },
+    {
+      "epoch": 0.4345001047998323,
+      "grad_norm": 0.1341363489627838,
+      "learning_rate": 6.293821959133993e-06,
+      "loss": 0.0616,
+      "step": 2073
+    },
+    {
+      "epoch": 0.4347097044644729,
+      "grad_norm": 0.11600136011838913,
+      "learning_rate": 6.2905424397628816e-06,
+      "loss": 0.0593,
+      "step": 2074
+    },
+    {
+      "epoch": 0.4349193041291134,
+      "grad_norm": 0.10497263818979263,
+      "learning_rate": 6.28726232545197e-06,
+      "loss": 0.0561,
+      "step": 2075
+    },
+    {
+      "epoch": 0.4351289037937539,
+      "grad_norm": 0.10920844972133636,
+      "learning_rate": 6.283981617713397e-06,
+      "loss": 0.0577,
+      "step": 2076
+    },
+    {
+      "epoch": 0.4353385034583945,
+      "grad_norm": 0.11021441221237183,
+      "learning_rate": 6.280700318059563e-06,
+      "loss": 0.0574,
+      "step": 2077
+    },
+    {
+      "epoch": 0.435548103123035,
+      "grad_norm": 0.09715298563241959,
+      "learning_rate": 6.277418428003149e-06,
+      "loss": 0.0599,
+      "step": 2078
+    },
+    {
+      "epoch": 0.43575770278767556,
+      "grad_norm": 0.0873681902885437,
+      "learning_rate": 6.274135949057107e-06,
+      "loss": 0.0556,
+      "step": 2079
+    },
+    {
+      "epoch": 0.4359673024523161,
+      "grad_norm": 0.11400279402732849,
+      "learning_rate": 6.270852882734654e-06,
+      "loss": 0.0596,
+      "step": 2080
+    },
+    {
+      "epoch": 0.4361769021169566,
+      "grad_norm": 0.10710210353136063,
+      "learning_rate": 6.267569230549288e-06,
+      "loss": 0.0607,
+      "step": 2081
+    },
+    {
+      "epoch": 0.43638650178159716,
+      "grad_norm": 0.1295563280582428,
+      "learning_rate": 6.26428499401477e-06,
+      "loss": 0.0584,
+      "step": 2082
+    },
+    {
+      "epoch": 0.43659610144623767,
+      "grad_norm": 0.1621527224779129,
+      "learning_rate": 6.261000174645131e-06,
+      "loss": 0.0576,
+      "step": 2083
+    },
+    {
+      "epoch": 0.43680570111087824,
+      "grad_norm": 0.17567972838878632,
+      "learning_rate": 6.257714773954674e-06,
+      "loss": 0.057,
+      "step": 2084
+    },
+    {
+      "epoch": 0.43701530077551876,
+      "grad_norm": 0.16140656173229218,
+      "learning_rate": 6.254428793457967e-06,
+      "loss": 0.0583,
+      "step": 2085
+    },
+    {
+      "epoch": 0.43722490044015927,
+      "grad_norm": 0.13268983364105225,
+      "learning_rate": 6.251142234669848e-06,
+      "loss": 0.0565,
+      "step": 2086
+    },
+    {
+      "epoch": 0.43743450010479984,
+      "grad_norm": 0.11374874413013458,
+      "learning_rate": 6.24785509910542e-06,
+      "loss": 0.056,
+      "step": 2087
+    },
+    {
+      "epoch": 0.43764409976944035,
+      "grad_norm": 0.09471353888511658,
+      "learning_rate": 6.244567388280047e-06,
+      "loss": 0.0595,
+      "step": 2088
+    },
+    {
+      "epoch": 0.4378536994340809,
+      "grad_norm": 0.12923859059810638,
+      "learning_rate": 6.241279103709368e-06,
+      "loss": 0.0555,
+      "step": 2089
+    },
+    {
+      "epoch": 0.43806329909872144,
+      "grad_norm": 0.12571561336517334,
+      "learning_rate": 6.23799024690928e-06,
+      "loss": 0.0587,
+      "step": 2090
+    },
+    {
+      "epoch": 0.43827289876336195,
+      "grad_norm": 0.1770731657743454,
+      "learning_rate": 6.234700819395946e-06,
+      "loss": 0.0601,
+      "step": 2091
+    },
+    {
+      "epoch": 0.4384824984280025,
+      "grad_norm": 0.15990307927131653,
+      "learning_rate": 6.231410822685791e-06,
+      "loss": 0.0585,
+      "step": 2092
+    },
+    {
+      "epoch": 0.43869209809264303,
+      "grad_norm": 0.11964087188243866,
+      "learning_rate": 6.228120258295501e-06,
+      "loss": 0.0585,
+      "step": 2093
+    },
+    {
+      "epoch": 0.4389016977572836,
+      "grad_norm": 0.17662663757801056,
+      "learning_rate": 6.224829127742028e-06,
+      "loss": 0.0592,
+      "step": 2094
+    },
+    {
+      "epoch": 0.4391112974219241,
+      "grad_norm": 0.17629125714302063,
+      "learning_rate": 6.221537432542581e-06,
+      "loss": 0.0581,
+      "step": 2095
+    },
+    {
+      "epoch": 0.4393208970865647,
+      "grad_norm": 0.20414410531520844,
+      "learning_rate": 6.218245174214632e-06,
+      "loss": 0.0573,
+      "step": 2096
+    },
+    {
+      "epoch": 0.4395304967512052,
+      "grad_norm": 0.16704221069812775,
+      "learning_rate": 6.21495235427591e-06,
+      "loss": 0.0575,
+      "step": 2097
+    },
+    {
+      "epoch": 0.4397400964158457,
+      "grad_norm": 0.11289860308170319,
+      "learning_rate": 6.211658974244407e-06,
+      "loss": 0.0558,
+      "step": 2098
+    },
+    {
+      "epoch": 0.4399496960804863,
+      "grad_norm": 0.17604808509349823,
+      "learning_rate": 6.208365035638366e-06,
+      "loss": 0.0585,
+      "step": 2099
+    },
+    {
+      "epoch": 0.4401592957451268,
+      "grad_norm": 0.18010827898979187,
+      "learning_rate": 6.205070539976297e-06,
+      "loss": 0.0616,
+      "step": 2100
+    },
+    {
+      "epoch": 0.44036889540976737,
+      "grad_norm": 0.15220797061920166,
+      "learning_rate": 6.2017754887769576e-06,
+      "loss": 0.0565,
+      "step": 2101
+    },
+    {
+      "epoch": 0.4405784950744079,
+      "grad_norm": 0.12355081737041473,
+      "learning_rate": 6.198479883559367e-06,
+      "loss": 0.0581,
+      "step": 2102
+    },
+    {
+      "epoch": 0.4407880947390484,
+      "grad_norm": 0.16100534796714783,
+      "learning_rate": 6.195183725842799e-06,
+      "loss": 0.0599,
+      "step": 2103
+    },
+    {
+      "epoch": 0.44099769440368897,
+      "grad_norm": 0.1588708758354187,
+      "learning_rate": 6.191887017146784e-06,
+      "loss": 0.0561,
+      "step": 2104
+    },
+    {
+      "epoch": 0.4412072940683295,
+      "grad_norm": 0.10125992447137833,
+      "learning_rate": 6.1885897589911e-06,
+      "loss": 0.0616,
+      "step": 2105
+    },
+    {
+      "epoch": 0.44141689373297005,
+      "grad_norm": 0.16263973712921143,
+      "learning_rate": 6.185291952895784e-06,
+      "loss": 0.0601,
+      "step": 2106
+    },
+    {
+      "epoch": 0.44162649339761056,
+      "grad_norm": 0.15129660069942474,
+      "learning_rate": 6.181993600381126e-06,
+      "loss": 0.0571,
+      "step": 2107
+    },
+    {
+      "epoch": 0.4418360930622511,
+      "grad_norm": 0.13963650166988373,
+      "learning_rate": 6.17869470296766e-06,
+      "loss": 0.0569,
+      "step": 2108
+    },
+    {
+      "epoch": 0.44204569272689165,
+      "grad_norm": 0.13959255814552307,
+      "learning_rate": 6.175395262176184e-06,
+      "loss": 0.0612,
+      "step": 2109
+    },
+    {
+      "epoch": 0.44225529239153216,
+      "grad_norm": 0.13720089197158813,
+      "learning_rate": 6.172095279527735e-06,
+      "loss": 0.0559,
+      "step": 2110
+    },
+    {
+      "epoch": 0.44246489205617273,
+      "grad_norm": 0.1654689759016037,
+      "learning_rate": 6.168794756543605e-06,
+      "loss": 0.057,
+      "step": 2111
+    },
+    {
+      "epoch": 0.44267449172081325,
+      "grad_norm": 0.145694300532341,
+      "learning_rate": 6.1654936947453355e-06,
+      "loss": 0.0561,
+      "step": 2112
+    },
+    {
+      "epoch": 0.44288409138545376,
+      "grad_norm": 0.13383303582668304,
+      "learning_rate": 6.162192095654714e-06,
+      "loss": 0.0579,
+      "step": 2113
+    },
+    {
+      "epoch": 0.44309369105009433,
+      "grad_norm": 0.1771506667137146,
+      "learning_rate": 6.158889960793779e-06,
+      "loss": 0.0558,
+      "step": 2114
+    },
+    {
+      "epoch": 0.44330329071473484,
+      "grad_norm": 0.1819649338722229,
+      "learning_rate": 6.155587291684814e-06,
+      "loss": 0.0574,
+      "step": 2115
+    },
+    {
+      "epoch": 0.4435128903793754,
+      "grad_norm": 0.16896513104438782,
+      "learning_rate": 6.1522840898503446e-06,
+      "loss": 0.0588,
+      "step": 2116
+    },
+    {
+      "epoch": 0.4437224900440159,
+      "grad_norm": 0.135847270488739,
+      "learning_rate": 6.148980356813151e-06,
+      "loss": 0.0569,
+      "step": 2117
+    },
+    {
+      "epoch": 0.44393208970865644,
+      "grad_norm": 0.20298628509044647,
+      "learning_rate": 6.145676094096251e-06,
+      "loss": 0.0624,
+      "step": 2118
+    },
+    {
+      "epoch": 0.444141689373297,
+      "grad_norm": 0.15171730518341064,
+      "learning_rate": 6.142371303222909e-06,
+      "loss": 0.0567,
+      "step": 2119
+    },
+    {
+      "epoch": 0.4443512890379375,
+      "grad_norm": 0.1395554542541504,
+      "learning_rate": 6.139065985716635e-06,
+      "loss": 0.0578,
+      "step": 2120
+    },
+    {
+      "epoch": 0.4445608887025781,
+      "grad_norm": 0.14991877973079681,
+      "learning_rate": 6.135760143101177e-06,
+      "loss": 0.0587,
+      "step": 2121
+    },
+    {
+      "epoch": 0.4447704883672186,
+      "grad_norm": 0.16107945144176483,
+      "learning_rate": 6.13245377690053e-06,
+      "loss": 0.058,
+      "step": 2122
+    },
+    {
+      "epoch": 0.4449800880318591,
+      "grad_norm": 0.1380777806043625,
+      "learning_rate": 6.129146888638928e-06,
+      "loss": 0.0574,
+      "step": 2123
+    },
+    {
+      "epoch": 0.4451896876964997,
+      "grad_norm": 0.10554420948028564,
+      "learning_rate": 6.1258394798408424e-06,
+      "loss": 0.0586,
+      "step": 2124
+    },
+    {
+      "epoch": 0.4453992873611402,
+      "grad_norm": 0.1480768322944641,
+      "learning_rate": 6.122531552030992e-06,
+      "loss": 0.0602,
+      "step": 2125
+    },
+    {
+      "epoch": 0.4456088870257808,
+      "grad_norm": 0.1352657824754715,
+      "learning_rate": 6.119223106734328e-06,
+      "loss": 0.0581,
+      "step": 2126
+    },
+    {
+      "epoch": 0.4458184866904213,
+      "grad_norm": 0.1190410628914833,
+      "learning_rate": 6.115914145476045e-06,
+      "loss": 0.0586,
+      "step": 2127
+    },
+    {
+      "epoch": 0.44602808635506186,
+      "grad_norm": 0.13114245235919952,
+      "learning_rate": 6.112604669781572e-06,
+      "loss": 0.0564,
+      "step": 2128
+    },
+    {
+      "epoch": 0.4462376860197024,
+      "grad_norm": 0.1338747888803482,
+      "learning_rate": 6.109294681176578e-06,
+      "loss": 0.0603,
+      "step": 2129
+    },
+    {
+      "epoch": 0.4464472856843429,
+      "grad_norm": 0.12414805591106415,
+      "learning_rate": 6.105984181186968e-06,
+      "loss": 0.0564,
+      "step": 2130
+    },
+    {
+      "epoch": 0.44665688534898346,
+      "grad_norm": 0.09722696989774704,
+      "learning_rate": 6.102673171338878e-06,
+      "loss": 0.0565,
+      "step": 2131
+    },
+    {
+      "epoch": 0.44686648501362397,
+      "grad_norm": 0.13036277890205383,
+      "learning_rate": 6.099361653158687e-06,
+      "loss": 0.0595,
+      "step": 2132
+    },
+    {
+      "epoch": 0.44707608467826454,
+      "grad_norm": 0.12158705294132233,
+      "learning_rate": 6.0960496281729995e-06,
+      "loss": 0.0564,
+      "step": 2133
+    },
+    {
+      "epoch": 0.44728568434290505,
+      "grad_norm": 0.13005104660987854,
+      "learning_rate": 6.092737097908663e-06,
+      "loss": 0.0613,
+      "step": 2134
+    },
+    {
+      "epoch": 0.44749528400754557,
+      "grad_norm": 0.1160380020737648,
+      "learning_rate": 6.08942406389275e-06,
+      "loss": 0.0599,
+      "step": 2135
+    },
+    {
+      "epoch": 0.44770488367218614,
+      "grad_norm": 0.11387168616056442,
+      "learning_rate": 6.086110527652571e-06,
+      "loss": 0.0565,
+      "step": 2136
+    },
+    {
+      "epoch": 0.44791448333682665,
+      "grad_norm": 0.15158362686634064,
+      "learning_rate": 6.082796490715666e-06,
+      "loss": 0.0593,
+      "step": 2137
+    },
+    {
+      "epoch": 0.4481240830014672,
+      "grad_norm": 0.12060023099184036,
+      "learning_rate": 6.0794819546098006e-06,
+      "loss": 0.0565,
+      "step": 2138
+    },
+    {
+      "epoch": 0.44833368266610774,
+      "grad_norm": 0.12229592353105545,
+      "learning_rate": 6.076166920862979e-06,
+      "loss": 0.0584,
+      "step": 2139
+    },
+    {
+      "epoch": 0.44854328233074825,
+      "grad_norm": 0.10849461704492569,
+      "learning_rate": 6.072851391003432e-06,
+      "loss": 0.0565,
+      "step": 2140
+    },
+    {
+      "epoch": 0.4487528819953888,
+      "grad_norm": 0.1278999298810959,
+      "learning_rate": 6.069535366559615e-06,
+      "loss": 0.0572,
+      "step": 2141
+    },
+    {
+      "epoch": 0.44896248166002933,
+      "grad_norm": 0.1158517524600029,
+      "learning_rate": 6.066218849060217e-06,
+      "loss": 0.0586,
+      "step": 2142
+    },
+    {
+      "epoch": 0.4491720813246699,
+      "grad_norm": 0.11099664121866226,
+      "learning_rate": 6.0629018400341514e-06,
+      "loss": 0.0569,
+      "step": 2143
+    },
+    {
+      "epoch": 0.4493816809893104,
+      "grad_norm": 0.10765625536441803,
+      "learning_rate": 6.059584341010556e-06,
+      "loss": 0.0578,
+      "step": 2144
+    },
+    {
+      "epoch": 0.44959128065395093,
+      "grad_norm": 0.10618588328361511,
+      "learning_rate": 6.056266353518803e-06,
+      "loss": 0.0557,
+      "step": 2145
+    },
+    {
+      "epoch": 0.4498008803185915,
+      "grad_norm": 0.11303580552339554,
+      "learning_rate": 6.052947879088479e-06,
+      "loss": 0.0572,
+      "step": 2146
+    },
+    {
+      "epoch": 0.450010479983232,
+      "grad_norm": 0.09245932102203369,
+      "learning_rate": 6.0496289192494e-06,
+      "loss": 0.0542,
+      "step": 2147
+    },
+    {
+      "epoch": 0.4502200796478726,
+      "grad_norm": 0.09954038262367249,
+      "learning_rate": 6.046309475531609e-06,
+      "loss": 0.0579,
+      "step": 2148
+    },
+    {
+      "epoch": 0.4504296793125131,
+      "grad_norm": 0.08930417150259018,
+      "learning_rate": 6.0429895494653655e-06,
+      "loss": 0.0557,
+      "step": 2149
+    },
+    {
+      "epoch": 0.4506392789771536,
+      "grad_norm": 0.10569532960653305,
+      "learning_rate": 6.039669142581157e-06,
+      "loss": 0.0588,
+      "step": 2150
+    },
+    {
+      "epoch": 0.4508488786417942,
+      "grad_norm": 0.10851980745792389,
+      "learning_rate": 6.036348256409692e-06,
+      "loss": 0.0586,
+      "step": 2151
+    },
+    {
+      "epoch": 0.4510584783064347,
+      "grad_norm": 0.10847117751836777,
+      "learning_rate": 6.0330268924818925e-06,
+      "loss": 0.0563,
+      "step": 2152
+    },
+    {
+      "epoch": 0.45126807797107527,
+      "grad_norm": 0.10027307271957397,
+      "learning_rate": 6.029705052328912e-06,
+      "loss": 0.058,
+      "step": 2153
+    },
+    {
+      "epoch": 0.4514776776357158,
+      "grad_norm": 0.08025722205638885,
+      "learning_rate": 6.026382737482116e-06,
+      "loss": 0.0541,
+      "step": 2154
+    },
+    {
+      "epoch": 0.4516872773003563,
+      "grad_norm": 0.09462866187095642,
+      "learning_rate": 6.023059949473091e-06,
+      "loss": 0.0561,
+      "step": 2155
+    },
+    {
+      "epoch": 0.45189687696499686,
+      "grad_norm": 0.0991954430937767,
+      "learning_rate": 6.019736689833643e-06,
+      "loss": 0.0566,
+      "step": 2156
+    },
+    {
+      "epoch": 0.4521064766296374,
+      "grad_norm": 0.11048531532287598,
+      "learning_rate": 6.016412960095791e-06,
+      "loss": 0.057,
+      "step": 2157
+    },
+    {
+      "epoch": 0.45231607629427795,
+      "grad_norm": 0.09113509207963943,
+      "learning_rate": 6.013088761791776e-06,
+      "loss": 0.0562,
+      "step": 2158
+    },
+    {
+      "epoch": 0.45252567595891846,
+      "grad_norm": 0.07936461269855499,
+      "learning_rate": 6.0097640964540535e-06,
+      "loss": 0.0601,
+      "step": 2159
+    },
+    {
+      "epoch": 0.452735275623559,
+      "grad_norm": 0.10202574729919434,
+      "learning_rate": 6.006438965615291e-06,
+      "loss": 0.0573,
+      "step": 2160
+    },
+    {
+      "epoch": 0.45294487528819954,
+      "grad_norm": 0.10225759446620941,
+      "learning_rate": 6.003113370808375e-06,
+      "loss": 0.0587,
+      "step": 2161
+    },
+    {
+      "epoch": 0.45315447495284006,
+      "grad_norm": 0.09442250430583954,
+      "learning_rate": 5.999787313566403e-06,
+      "loss": 0.0609,
+      "step": 2162
+    },
+    {
+      "epoch": 0.45336407461748063,
+      "grad_norm": 0.11422485113143921,
+      "learning_rate": 5.996460795422688e-06,
+      "loss": 0.0597,
+      "step": 2163
+    },
+    {
+      "epoch": 0.45357367428212114,
+      "grad_norm": 0.08849114924669266,
+      "learning_rate": 5.993133817910752e-06,
+      "loss": 0.0596,
+      "step": 2164
+    },
+    {
+      "epoch": 0.4537832739467617,
+      "grad_norm": 0.1041831523180008,
+      "learning_rate": 5.9898063825643335e-06,
+      "loss": 0.0561,
+      "step": 2165
+    },
+    {
+      "epoch": 0.4539928736114022,
+      "grad_norm": 0.12931424379348755,
+      "learning_rate": 5.986478490917378e-06,
+      "loss": 0.0563,
+      "step": 2166
+    },
+    {
+      "epoch": 0.45420247327604274,
+      "grad_norm": 0.07999792695045471,
+      "learning_rate": 5.983150144504043e-06,
+      "loss": 0.0577,
+      "step": 2167
+    },
+    {
+      "epoch": 0.4544120729406833,
+      "grad_norm": 0.10252411663532257,
+      "learning_rate": 5.979821344858695e-06,
+      "loss": 0.0587,
+      "step": 2168
+    },
+    {
+      "epoch": 0.4546216726053238,
+      "grad_norm": 0.10504443198442459,
+      "learning_rate": 5.976492093515911e-06,
+      "loss": 0.0596,
+      "step": 2169
+    },
+    {
+      "epoch": 0.4548312722699644,
+      "grad_norm": 0.09300100803375244,
+      "learning_rate": 5.973162392010474e-06,
+      "loss": 0.0626,
+      "step": 2170
+    },
+    {
+      "epoch": 0.4550408719346049,
+      "grad_norm": 0.11920040845870972,
+      "learning_rate": 5.969832241877378e-06,
+      "loss": 0.0583,
+      "step": 2171
+    },
+    {
+      "epoch": 0.4552504715992454,
+      "grad_norm": 0.10893451422452927,
+      "learning_rate": 5.966501644651817e-06,
+      "loss": 0.055,
+      "step": 2172
+    },
+    {
+      "epoch": 0.455460071263886,
+      "grad_norm": 0.08888144791126251,
+      "learning_rate": 5.963170601869203e-06,
+      "loss": 0.0577,
+      "step": 2173
+    },
+    {
+      "epoch": 0.4556696709285265,
+      "grad_norm": 0.11214505136013031,
+      "learning_rate": 5.959839115065138e-06,
+      "loss": 0.057,
+      "step": 2174
+    },
+    {
+      "epoch": 0.4558792705931671,
+      "grad_norm": 0.10267043858766556,
+      "learning_rate": 5.956507185775441e-06,
+      "loss": 0.0584,
+      "step": 2175
+    },
+    {
+      "epoch": 0.4560888702578076,
+      "grad_norm": 0.10312189161777496,
+      "learning_rate": 5.953174815536131e-06,
+      "loss": 0.0589,
+      "step": 2176
+    },
+    {
+      "epoch": 0.4562984699224481,
+      "grad_norm": 0.11400537937879562,
+      "learning_rate": 5.949842005883428e-06,
+      "loss": 0.0566,
+      "step": 2177
+    },
+    {
+      "epoch": 0.45650806958708867,
+      "grad_norm": 0.0924360379576683,
+      "learning_rate": 5.9465087583537594e-06,
+      "loss": 0.0597,
+      "step": 2178
+    },
+    {
+      "epoch": 0.4567176692517292,
+      "grad_norm": 0.11675729602575302,
+      "learning_rate": 5.943175074483749e-06,
+      "loss": 0.0566,
+      "step": 2179
+    },
+    {
+      "epoch": 0.45692726891636976,
+      "grad_norm": 0.11411453783512115,
+      "learning_rate": 5.939840955810223e-06,
+      "loss": 0.0598,
+      "step": 2180
+    },
+    {
+      "epoch": 0.45713686858101027,
+      "grad_norm": 0.11106470972299576,
+      "learning_rate": 5.936506403870215e-06,
+      "loss": 0.0569,
+      "step": 2181
+    },
+    {
+      "epoch": 0.4573464682456508,
+      "grad_norm": 0.1371942162513733,
+      "learning_rate": 5.933171420200946e-06,
+      "loss": 0.0571,
+      "step": 2182
+    },
+    {
+      "epoch": 0.45755606791029135,
+      "grad_norm": 0.10696960240602493,
+      "learning_rate": 5.929836006339848e-06,
+      "loss": 0.0614,
+      "step": 2183
+    },
+    {
+      "epoch": 0.45776566757493187,
+      "grad_norm": 0.1150413230061531,
+      "learning_rate": 5.926500163824546e-06,
+      "loss": 0.0592,
+      "step": 2184
+    },
+    {
+      "epoch": 0.45797526723957244,
+      "grad_norm": 0.10095971077680588,
+      "learning_rate": 5.923163894192857e-06,
+      "loss": 0.0564,
+      "step": 2185
+    },
+    {
+      "epoch": 0.45818486690421295,
+      "grad_norm": 0.10621986538171768,
+      "learning_rate": 5.9198271989828075e-06,
+      "loss": 0.0544,
+      "step": 2186
+    },
+    {
+      "epoch": 0.45839446656885346,
+      "grad_norm": 0.12330467998981476,
+      "learning_rate": 5.91649007973261e-06,
+      "loss": 0.0565,
+      "step": 2187
+    },
+    {
+      "epoch": 0.45860406623349403,
+      "grad_norm": 0.10255663841962814,
+      "learning_rate": 5.913152537980674e-06,
+      "loss": 0.057,
+      "step": 2188
+    },
+    {
+      "epoch": 0.45881366589813455,
+      "grad_norm": 0.13345320522785187,
+      "learning_rate": 5.909814575265609e-06,
+      "loss": 0.0597,
+      "step": 2189
+    },
+    {
+      "epoch": 0.4590232655627751,
+      "grad_norm": 0.112071193754673,
+      "learning_rate": 5.9064761931262135e-06,
+      "loss": 0.0556,
+      "step": 2190
+    },
+    {
+      "epoch": 0.45923286522741563,
+      "grad_norm": 0.12068294733762741,
+      "learning_rate": 5.903137393101482e-06,
+      "loss": 0.0554,
+      "step": 2191
+    },
+    {
+      "epoch": 0.45944246489205615,
+      "grad_norm": 0.11650517582893372,
+      "learning_rate": 5.8997981767306e-06,
+      "loss": 0.0557,
+      "step": 2192
+    },
+    {
+      "epoch": 0.4596520645566967,
+      "grad_norm": 0.1526482254266739,
+      "learning_rate": 5.896458545552946e-06,
+      "loss": 0.0581,
+      "step": 2193
+    },
+    {
+      "epoch": 0.45986166422133723,
+      "grad_norm": 0.15084873139858246,
+      "learning_rate": 5.893118501108087e-06,
+      "loss": 0.0581,
+      "step": 2194
+    },
+    {
+      "epoch": 0.4600712638859778,
+      "grad_norm": 0.12740178406238556,
+      "learning_rate": 5.889778044935785e-06,
+      "loss": 0.0579,
+      "step": 2195
+    },
+    {
+      "epoch": 0.4602808635506183,
+      "grad_norm": 0.1175624206662178,
+      "learning_rate": 5.88643717857599e-06,
+      "loss": 0.0556,
+      "step": 2196
+    },
+    {
+      "epoch": 0.4604904632152589,
+      "grad_norm": 0.12416189163923264,
+      "learning_rate": 5.883095903568838e-06,
+      "loss": 0.0587,
+      "step": 2197
+    },
+    {
+      "epoch": 0.4607000628798994,
+      "grad_norm": 0.13820405304431915,
+      "learning_rate": 5.87975422145466e-06,
+      "loss": 0.0573,
+      "step": 2198
+    },
+    {
+      "epoch": 0.4609096625445399,
+      "grad_norm": 0.11369778960943222,
+      "learning_rate": 5.876412133773968e-06,
+      "loss": 0.0573,
+      "step": 2199
+    },
+    {
+      "epoch": 0.4611192622091805,
+      "grad_norm": 0.1153719425201416,
+      "learning_rate": 5.873069642067464e-06,
+      "loss": 0.0553,
+      "step": 2200
+    },
+    {
+      "epoch": 0.461328861873821,
+      "grad_norm": 0.10116839408874512,
+      "learning_rate": 5.869726747876036e-06,
+      "loss": 0.0588,
+      "step": 2201
+    },
+    {
+      "epoch": 0.46153846153846156,
+      "grad_norm": 0.1170465499162674,
+      "learning_rate": 5.866383452740758e-06,
+      "loss": 0.0555,
+      "step": 2202
+    },
+    {
+      "epoch": 0.4617480612031021,
+      "grad_norm": 0.14628863334655762,
+      "learning_rate": 5.863039758202889e-06,
+      "loss": 0.0561,
+      "step": 2203
+    },
+    {
+      "epoch": 0.4619576608677426,
+      "grad_norm": 0.15336976945400238,
+      "learning_rate": 5.85969566580387e-06,
+      "loss": 0.0593,
+      "step": 2204
+    },
+    {
+      "epoch": 0.46216726053238316,
+      "grad_norm": 0.14001086354255676,
+      "learning_rate": 5.856351177085327e-06,
+      "loss": 0.0572,
+      "step": 2205
+    },
+    {
+      "epoch": 0.4623768601970237,
+      "grad_norm": 0.09859530627727509,
+      "learning_rate": 5.853006293589071e-06,
+      "loss": 0.057,
+      "step": 2206
+    },
+    {
+      "epoch": 0.46258645986166425,
+      "grad_norm": 0.10227430611848831,
+      "learning_rate": 5.84966101685709e-06,
+      "loss": 0.0542,
+      "step": 2207
+    },
+    {
+      "epoch": 0.46279605952630476,
+      "grad_norm": 0.12590292096138,
+      "learning_rate": 5.846315348431555e-06,
+      "loss": 0.0573,
+      "step": 2208
+    },
+    {
+      "epoch": 0.4630056591909453,
+      "grad_norm": 0.11103103309869766,
+      "learning_rate": 5.842969289854823e-06,
+      "loss": 0.0576,
+      "step": 2209
+    },
+    {
+      "epoch": 0.46321525885558584,
+      "grad_norm": 0.12213583290576935,
+      "learning_rate": 5.839622842669423e-06,
+      "loss": 0.0586,
+      "step": 2210
+    },
+    {
+      "epoch": 0.46342485852022636,
+      "grad_norm": 0.10279441624879837,
+      "learning_rate": 5.836276008418065e-06,
+      "loss": 0.0569,
+      "step": 2211
+    },
+    {
+      "epoch": 0.4636344581848669,
+      "grad_norm": 0.09687735140323639,
+      "learning_rate": 5.832928788643644e-06,
+      "loss": 0.0582,
+      "step": 2212
+    },
+    {
+      "epoch": 0.46384405784950744,
+      "grad_norm": 0.09342148154973984,
+      "learning_rate": 5.8295811848892215e-06,
+      "loss": 0.0585,
+      "step": 2213
+    },
+    {
+      "epoch": 0.46405365751414795,
+      "grad_norm": 0.09567330032587051,
+      "learning_rate": 5.826233198698047e-06,
+      "loss": 0.0571,
+      "step": 2214
+    },
+    {
+      "epoch": 0.4642632571787885,
+      "grad_norm": 0.10038771480321884,
+      "learning_rate": 5.822884831613538e-06,
+      "loss": 0.0558,
+      "step": 2215
+    },
+    {
+      "epoch": 0.46447285684342904,
+      "grad_norm": 0.10112529247999191,
+      "learning_rate": 5.819536085179293e-06,
+      "loss": 0.0575,
+      "step": 2216
+    },
+    {
+      "epoch": 0.4646824565080696,
+      "grad_norm": 0.11409956961870193,
+      "learning_rate": 5.816186960939084e-06,
+      "loss": 0.0551,
+      "step": 2217
+    },
+    {
+      "epoch": 0.4648920561727101,
+      "grad_norm": 0.11404009908437729,
+      "learning_rate": 5.8128374604368534e-06,
+      "loss": 0.0572,
+      "step": 2218
+    },
+    {
+      "epoch": 0.46510165583735064,
+      "grad_norm": 0.1052129939198494,
+      "learning_rate": 5.809487585216725e-06,
+      "loss": 0.0569,
+      "step": 2219
+    },
+    {
+      "epoch": 0.4653112555019912,
+      "grad_norm": 0.1068800687789917,
+      "learning_rate": 5.806137336822987e-06,
+      "loss": 0.0568,
+      "step": 2220
+    },
+    {
+      "epoch": 0.4655208551666317,
+      "grad_norm": 0.1187024712562561,
+      "learning_rate": 5.802786716800102e-06,
+      "loss": 0.0567,
+      "step": 2221
+    },
+    {
+      "epoch": 0.4657304548312723,
+      "grad_norm": 0.11063768714666367,
+      "learning_rate": 5.79943572669271e-06,
+      "loss": 0.0565,
+      "step": 2222
+    },
+    {
+      "epoch": 0.4659400544959128,
+      "grad_norm": 0.11005666106939316,
+      "learning_rate": 5.796084368045612e-06,
+      "loss": 0.0578,
+      "step": 2223
+    },
+    {
+      "epoch": 0.4661496541605533,
+      "grad_norm": 0.09957437217235565,
+      "learning_rate": 5.7927326424037875e-06,
+      "loss": 0.0566,
+      "step": 2224
+    },
+    {
+      "epoch": 0.4663592538251939,
+      "grad_norm": 0.10687166452407837,
+      "learning_rate": 5.789380551312379e-06,
+      "loss": 0.0565,
+      "step": 2225
+    },
+    {
+      "epoch": 0.4665688534898344,
+      "grad_norm": 0.12073647975921631,
+      "learning_rate": 5.7860280963167e-06,
+      "loss": 0.0587,
+      "step": 2226
+    },
+    {
+      "epoch": 0.46677845315447497,
+      "grad_norm": 0.11012930423021317,
+      "learning_rate": 5.782675278962232e-06,
+      "loss": 0.056,
+      "step": 2227
+    },
+    {
+      "epoch": 0.4669880528191155,
+      "grad_norm": 0.10260272771120071,
+      "learning_rate": 5.7793221007946245e-06,
+      "loss": 0.0595,
+      "step": 2228
+    },
+    {
+      "epoch": 0.467197652483756,
+      "grad_norm": 0.07964638620615005,
+      "learning_rate": 5.775968563359688e-06,
+      "loss": 0.0563,
+      "step": 2229
+    },
+    {
+      "epoch": 0.46740725214839657,
+      "grad_norm": 0.09134583175182343,
+      "learning_rate": 5.7726146682034055e-06,
+      "loss": 0.0548,
+      "step": 2230
+    },
+    {
+      "epoch": 0.4676168518130371,
+      "grad_norm": 0.10940870642662048,
+      "learning_rate": 5.7692604168719225e-06,
+      "loss": 0.0535,
+      "step": 2231
+    },
+    {
+      "epoch": 0.46782645147767765,
+      "grad_norm": 0.1325157731771469,
+      "learning_rate": 5.765905810911546e-06,
+      "loss": 0.0571,
+      "step": 2232
+    },
+    {
+      "epoch": 0.46803605114231817,
+      "grad_norm": 0.13585898280143738,
+      "learning_rate": 5.762550851868751e-06,
+      "loss": 0.057,
+      "step": 2233
+    },
+    {
+      "epoch": 0.46824565080695874,
+      "grad_norm": 0.11632729321718216,
+      "learning_rate": 5.759195541290171e-06,
+      "loss": 0.0581,
+      "step": 2234
+    },
+    {
+      "epoch": 0.46845525047159925,
+      "grad_norm": 0.08722345530986786,
+      "learning_rate": 5.7558398807226045e-06,
+      "loss": 0.0561,
+      "step": 2235
+    },
+    {
+      "epoch": 0.46866485013623976,
+      "grad_norm": 0.1064450740814209,
+      "learning_rate": 5.7524838717130095e-06,
+      "loss": 0.0549,
+      "step": 2236
+    },
+    {
+      "epoch": 0.46887444980088033,
+      "grad_norm": 0.12046857923269272,
+      "learning_rate": 5.749127515808506e-06,
+      "loss": 0.0576,
+      "step": 2237
+    },
+    {
+      "epoch": 0.46908404946552085,
+      "grad_norm": 0.11908888071775436,
+      "learning_rate": 5.745770814556373e-06,
+      "loss": 0.0575,
+      "step": 2238
+    },
+    {
+      "epoch": 0.4692936491301614,
+      "grad_norm": 0.10716499388217926,
+      "learning_rate": 5.7424137695040495e-06,
+      "loss": 0.0562,
+      "step": 2239
+    },
+    {
+      "epoch": 0.46950324879480193,
+      "grad_norm": 0.09762990474700928,
+      "learning_rate": 5.7390563821991326e-06,
+      "loss": 0.0582,
+      "step": 2240
+    },
+    {
+      "epoch": 0.46971284845944244,
+      "grad_norm": 0.11566614359617233,
+      "learning_rate": 5.735698654189377e-06,
+      "loss": 0.0592,
+      "step": 2241
+    },
+    {
+      "epoch": 0.469922448124083,
+      "grad_norm": 0.09800199419260025,
+      "learning_rate": 5.7323405870226955e-06,
+      "loss": 0.0567,
+      "step": 2242
+    },
+    {
+      "epoch": 0.47013204778872353,
+      "grad_norm": 0.09443154186010361,
+      "learning_rate": 5.7289821822471545e-06,
+      "loss": 0.058,
+      "step": 2243
+    },
+    {
+      "epoch": 0.4703416474533641,
+      "grad_norm": 0.07052087038755417,
+      "learning_rate": 5.725623441410979e-06,
+      "loss": 0.0536,
+      "step": 2244
+    },
+    {
+      "epoch": 0.4705512471180046,
+      "grad_norm": 0.10066118091344833,
+      "learning_rate": 5.722264366062549e-06,
+      "loss": 0.0568,
+      "step": 2245
+    },
+    {
+      "epoch": 0.4707608467826451,
+      "grad_norm": 0.10691467672586441,
+      "learning_rate": 5.718904957750394e-06,
+      "loss": 0.0564,
+      "step": 2246
+    },
+    {
+      "epoch": 0.4709704464472857,
+      "grad_norm": 0.10959997028112411,
+      "learning_rate": 5.715545218023205e-06,
+      "loss": 0.0544,
+      "step": 2247
+    },
+    {
+      "epoch": 0.4711800461119262,
+      "grad_norm": 0.09174855053424835,
+      "learning_rate": 5.7121851484298184e-06,
+      "loss": 0.0534,
+      "step": 2248
+    },
+    {
+      "epoch": 0.4713896457765668,
+      "grad_norm": 0.10001726448535919,
+      "learning_rate": 5.708824750519225e-06,
+      "loss": 0.0574,
+      "step": 2249
+    },
+    {
+      "epoch": 0.4715992454412073,
+      "grad_norm": 0.10282893478870392,
+      "learning_rate": 5.705464025840571e-06,
+      "loss": 0.0535,
+      "step": 2250
+    },
+    {
+      "epoch": 0.4718088451058478,
+      "grad_norm": 0.0975334644317627,
+      "learning_rate": 5.702102975943147e-06,
+      "loss": 0.0609,
+      "step": 2251
+    },
+    {
+      "epoch": 0.4720184447704884,
+      "grad_norm": 0.08588805049657822,
+      "learning_rate": 5.698741602376395e-06,
+      "loss": 0.0538,
+      "step": 2252
+    },
+    {
+      "epoch": 0.4722280444351289,
+      "grad_norm": 0.08347409963607788,
+      "learning_rate": 5.695379906689912e-06,
+      "loss": 0.0594,
+      "step": 2253
+    },
+    {
+      "epoch": 0.47243764409976946,
+      "grad_norm": 0.0910545289516449,
+      "learning_rate": 5.6920178904334346e-06,
+      "loss": 0.0592,
+      "step": 2254
+    },
+    {
+      "epoch": 0.47264724376441,
+      "grad_norm": 0.08542585372924805,
+      "learning_rate": 5.688655555156854e-06,
+      "loss": 0.058,
+      "step": 2255
+    },
+    {
+      "epoch": 0.4728568434290505,
+      "grad_norm": 0.09314323216676712,
+      "learning_rate": 5.6852929024102065e-06,
+      "loss": 0.0546,
+      "step": 2256
+    },
+    {
+      "epoch": 0.47306644309369106,
+      "grad_norm": 0.09594455361366272,
+      "learning_rate": 5.681929933743672e-06,
+      "loss": 0.0546,
+      "step": 2257
+    },
+    {
+      "epoch": 0.47327604275833157,
+      "grad_norm": 0.09240654855966568,
+      "learning_rate": 5.67856665070758e-06,
+      "loss": 0.0568,
+      "step": 2258
+    },
+    {
+      "epoch": 0.47348564242297214,
+      "grad_norm": 0.08861672133207321,
+      "learning_rate": 5.675203054852403e-06,
+      "loss": 0.0584,
+      "step": 2259
+    },
+    {
+      "epoch": 0.47369524208761266,
+      "grad_norm": 0.08311017602682114,
+      "learning_rate": 5.671839147728758e-06,
+      "loss": 0.0548,
+      "step": 2260
+    },
+    {
+      "epoch": 0.47390484175225317,
+      "grad_norm": 0.08416654914617538,
+      "learning_rate": 5.668474930887406e-06,
+      "loss": 0.0569,
+      "step": 2261
+    },
+    {
+      "epoch": 0.47411444141689374,
+      "grad_norm": 0.07767292857170105,
+      "learning_rate": 5.6651104058792496e-06,
+      "loss": 0.0558,
+      "step": 2262
+    },
+    {
+      "epoch": 0.47432404108153425,
+      "grad_norm": 0.09354811161756516,
+      "learning_rate": 5.661745574255334e-06,
+      "loss": 0.0562,
+      "step": 2263
+    },
+    {
+      "epoch": 0.4745336407461748,
+      "grad_norm": 0.1013355404138565,
+      "learning_rate": 5.658380437566846e-06,
+      "loss": 0.0563,
+      "step": 2264
+    },
+    {
+      "epoch": 0.47474324041081534,
+      "grad_norm": 0.08834841102361679,
+      "learning_rate": 5.655014997365114e-06,
+      "loss": 0.0581,
+      "step": 2265
+    },
+    {
+      "epoch": 0.4749528400754559,
+      "grad_norm": 0.07844306528568268,
+      "learning_rate": 5.651649255201603e-06,
+      "loss": 0.0559,
+      "step": 2266
+    },
+    {
+      "epoch": 0.4751624397400964,
+      "grad_norm": 0.08669186383485794,
+      "learning_rate": 5.648283212627921e-06,
+      "loss": 0.0571,
+      "step": 2267
+    },
+    {
+      "epoch": 0.47537203940473693,
+      "grad_norm": 0.08236299455165863,
+      "learning_rate": 5.6449168711958135e-06,
+      "loss": 0.057,
+      "step": 2268
+    },
+    {
+      "epoch": 0.4755816390693775,
+      "grad_norm": 0.0980667769908905,
+      "learning_rate": 5.641550232457162e-06,
+      "loss": 0.0546,
+      "step": 2269
+    },
+    {
+      "epoch": 0.475791238734018,
+      "grad_norm": 0.11645791679620743,
+      "learning_rate": 5.638183297963986e-06,
+      "loss": 0.0549,
+      "step": 2270
+    },
+    {
+      "epoch": 0.4760008383986586,
+      "grad_norm": 0.12081831693649292,
+      "learning_rate": 5.634816069268442e-06,
+      "loss": 0.0582,
+      "step": 2271
+    },
+    {
+      "epoch": 0.4762104380632991,
+      "grad_norm": 0.1103171780705452,
+      "learning_rate": 5.631448547922822e-06,
+      "loss": 0.0567,
+      "step": 2272
+    },
+    {
+      "epoch": 0.4764200377279396,
+      "grad_norm": 0.1022624745965004,
+      "learning_rate": 5.628080735479553e-06,
+      "loss": 0.0577,
+      "step": 2273
+    },
+    {
+      "epoch": 0.4766296373925802,
+      "grad_norm": 0.0808848962187767,
+      "learning_rate": 5.624712633491196e-06,
+      "loss": 0.0593,
+      "step": 2274
+    },
+    {
+      "epoch": 0.4768392370572207,
+      "grad_norm": 0.07358387112617493,
+      "learning_rate": 5.621344243510444e-06,
+      "loss": 0.0568,
+      "step": 2275
+    },
+    {
+      "epoch": 0.47704883672186127,
+      "grad_norm": 0.07857120782136917,
+      "learning_rate": 5.6179755670901245e-06,
+      "loss": 0.0569,
+      "step": 2276
+    },
+    {
+      "epoch": 0.4772584363865018,
+      "grad_norm": 0.06889036297798157,
+      "learning_rate": 5.614606605783197e-06,
+      "loss": 0.0586,
+      "step": 2277
+    },
+    {
+      "epoch": 0.4774680360511423,
+      "grad_norm": 0.07760413736104965,
+      "learning_rate": 5.611237361142753e-06,
+      "loss": 0.0551,
+      "step": 2278
+    },
+    {
+      "epoch": 0.47767763571578287,
+      "grad_norm": 0.08692996203899384,
+      "learning_rate": 5.607867834722012e-06,
+      "loss": 0.0545,
+      "step": 2279
+    },
+    {
+      "epoch": 0.4778872353804234,
+      "grad_norm": 0.08127982169389725,
+      "learning_rate": 5.604498028074323e-06,
+      "loss": 0.0571,
+      "step": 2280
+    },
+    {
+      "epoch": 0.47809683504506395,
+      "grad_norm": 0.07126957178115845,
+      "learning_rate": 5.601127942753173e-06,
+      "loss": 0.0577,
+      "step": 2281
+    },
+    {
+      "epoch": 0.47830643470970446,
+      "grad_norm": 0.08008185029029846,
+      "learning_rate": 5.597757580312163e-06,
+      "loss": 0.0542,
+      "step": 2282
+    },
+    {
+      "epoch": 0.478516034374345,
+      "grad_norm": 0.07614471018314362,
+      "learning_rate": 5.594386942305035e-06,
+      "loss": 0.0554,
+      "step": 2283
+    },
+    {
+      "epoch": 0.47872563403898555,
+      "grad_norm": 0.06831208616495132,
+      "learning_rate": 5.5910160302856486e-06,
+      "loss": 0.0555,
+      "step": 2284
+    },
+    {
+      "epoch": 0.47893523370362606,
+      "grad_norm": 0.07799834758043289,
+      "learning_rate": 5.587644845807994e-06,
+      "loss": 0.0569,
+      "step": 2285
+    },
+    {
+      "epoch": 0.47914483336826663,
+      "grad_norm": 0.08827179670333862,
+      "learning_rate": 5.584273390426189e-06,
+      "loss": 0.0555,
+      "step": 2286
+    },
+    {
+      "epoch": 0.47935443303290715,
+      "grad_norm": 0.07891764491796494,
+      "learning_rate": 5.580901665694471e-06,
+      "loss": 0.0582,
+      "step": 2287
+    },
+    {
+      "epoch": 0.47956403269754766,
+      "grad_norm": 0.06794466078281403,
+      "learning_rate": 5.577529673167208e-06,
+      "loss": 0.0563,
+      "step": 2288
+    },
+    {
+      "epoch": 0.47977363236218823,
+      "grad_norm": 0.06872694939374924,
+      "learning_rate": 5.574157414398885e-06,
+      "loss": 0.0551,
+      "step": 2289
+    },
+    {
+      "epoch": 0.47998323202682874,
+      "grad_norm": 0.06112726777791977,
+      "learning_rate": 5.570784890944112e-06,
+      "loss": 0.0571,
+      "step": 2290
+    },
+    {
+      "epoch": 0.4801928316914693,
+      "grad_norm": 0.07286538183689117,
+      "learning_rate": 5.567412104357623e-06,
+      "loss": 0.057,
+      "step": 2291
+    },
+    {
+      "epoch": 0.4804024313561098,
+      "grad_norm": 0.06900748610496521,
+      "learning_rate": 5.564039056194274e-06,
+      "loss": 0.0567,
+      "step": 2292
+    },
+    {
+      "epoch": 0.48061203102075034,
+      "grad_norm": 0.07319528609514236,
+      "learning_rate": 5.560665748009034e-06,
+      "loss": 0.0562,
+      "step": 2293
+    },
+    {
+      "epoch": 0.4808216306853909,
+      "grad_norm": 0.07866884768009186,
+      "learning_rate": 5.557292181357003e-06,
+      "loss": 0.0561,
+      "step": 2294
+    },
+    {
+      "epoch": 0.4810312303500314,
+      "grad_norm": 0.07217524945735931,
+      "learning_rate": 5.553918357793391e-06,
+      "loss": 0.0565,
+      "step": 2295
+    },
+    {
+      "epoch": 0.481240830014672,
+      "grad_norm": 0.0580948144197464,
+      "learning_rate": 5.550544278873531e-06,
+      "loss": 0.0583,
+      "step": 2296
+    },
+    {
+      "epoch": 0.4814504296793125,
+      "grad_norm": 0.0548657588660717,
+      "learning_rate": 5.547169946152874e-06,
+      "loss": 0.0574,
+      "step": 2297
+    },
+    {
+      "epoch": 0.481660029343953,
+      "grad_norm": 0.06751734763383865,
+      "learning_rate": 5.543795361186984e-06,
+      "loss": 0.0567,
+      "step": 2298
+    },
+    {
+      "epoch": 0.4818696290085936,
+      "grad_norm": 0.0683041512966156,
+      "learning_rate": 5.540420525531547e-06,
+      "loss": 0.0561,
+      "step": 2299
+    },
+    {
+      "epoch": 0.4820792286732341,
+      "grad_norm": 0.0687454417347908,
+      "learning_rate": 5.537045440742359e-06,
+      "loss": 0.0569,
+      "step": 2300
+    },
+    {
+      "epoch": 0.4822888283378747,
+      "grad_norm": 0.060074321925640106,
+      "learning_rate": 5.533670108375334e-06,
+      "loss": 0.0553,
+      "step": 2301
+    },
+    {
+      "epoch": 0.4824984280025152,
+      "grad_norm": 0.05447131395339966,
+      "learning_rate": 5.5302945299865005e-06,
+      "loss": 0.057,
+      "step": 2302
+    },
+    {
+      "epoch": 0.48270802766715576,
+      "grad_norm": 0.051248107105493546,
+      "learning_rate": 5.526918707132e-06,
+      "loss": 0.0561,
+      "step": 2303
+    },
+    {
+      "epoch": 0.4829176273317963,
+      "grad_norm": 0.0613834448158741,
+      "learning_rate": 5.523542641368083e-06,
+      "loss": 0.0582,
+      "step": 2304
+    },
+    {
+      "epoch": 0.4831272269964368,
+      "grad_norm": 0.054669518023729324,
+      "learning_rate": 5.520166334251118e-06,
+      "loss": 0.056,
+      "step": 2305
+    },
+    {
+      "epoch": 0.48333682666107736,
+      "grad_norm": 0.058429256081581116,
+      "learning_rate": 5.5167897873375805e-06,
+      "loss": 0.0573,
+      "step": 2306
+    },
+    {
+      "epoch": 0.48354642632571787,
+      "grad_norm": 0.06477133929729462,
+      "learning_rate": 5.513413002184059e-06,
+      "loss": 0.0557,
+      "step": 2307
+    },
+    {
+      "epoch": 0.48375602599035844,
+      "grad_norm": 0.06107893958687782,
+      "learning_rate": 5.510035980347249e-06,
+      "loss": 0.0576,
+      "step": 2308
+    },
+    {
+      "epoch": 0.48396562565499895,
+      "grad_norm": 0.06686937808990479,
+      "learning_rate": 5.50665872338396e-06,
+      "loss": 0.0586,
+      "step": 2309
+    },
+    {
+      "epoch": 0.48417522531963947,
+      "grad_norm": 0.0777740553021431,
+      "learning_rate": 5.503281232851102e-06,
+      "loss": 0.0551,
+      "step": 2310
+    },
+    {
+      "epoch": 0.48438482498428004,
+      "grad_norm": 0.08693230897188187,
+      "learning_rate": 5.499903510305703e-06,
+      "loss": 0.0538,
+      "step": 2311
+    },
+    {
+      "epoch": 0.48459442464892055,
+      "grad_norm": 0.09828737378120422,
+      "learning_rate": 5.496525557304888e-06,
+      "loss": 0.0546,
+      "step": 2312
+    },
+    {
+      "epoch": 0.4848040243135611,
+      "grad_norm": 0.11315979063510895,
+      "learning_rate": 5.4931473754058935e-06,
+      "loss": 0.0586,
+      "step": 2313
+    },
+    {
+      "epoch": 0.48501362397820164,
+      "grad_norm": 0.1083892285823822,
+      "learning_rate": 5.489768966166064e-06,
+      "loss": 0.0606,
+      "step": 2314
+    },
+    {
+      "epoch": 0.48522322364284215,
+      "grad_norm": 0.08665548264980316,
+      "learning_rate": 5.486390331142841e-06,
+      "loss": 0.0561,
+      "step": 2315
+    },
+    {
+      "epoch": 0.4854328233074827,
+      "grad_norm": 0.06847358494997025,
+      "learning_rate": 5.483011471893775e-06,
+      "loss": 0.0552,
+      "step": 2316
+    },
+    {
+      "epoch": 0.48564242297212323,
+      "grad_norm": 0.08172042667865753,
+      "learning_rate": 5.479632389976524e-06,
+      "loss": 0.0575,
+      "step": 2317
+    },
+    {
+      "epoch": 0.4858520226367638,
+      "grad_norm": 0.120709627866745,
+      "learning_rate": 5.4762530869488385e-06,
+      "loss": 0.0572,
+      "step": 2318
+    },
+    {
+      "epoch": 0.4860616223014043,
+      "grad_norm": 0.1355757862329483,
+      "learning_rate": 5.4728735643685804e-06,
+      "loss": 0.0587,
+      "step": 2319
+    },
+    {
+      "epoch": 0.48627122196604483,
+      "grad_norm": 0.12048780918121338,
+      "learning_rate": 5.469493823793706e-06,
+      "loss": 0.0591,
+      "step": 2320
+    },
+    {
+      "epoch": 0.4864808216306854,
+      "grad_norm": 0.07901732623577118,
+      "learning_rate": 5.466113866782277e-06,
+      "loss": 0.0552,
+      "step": 2321
+    },
+    {
+      "epoch": 0.4866904212953259,
+      "grad_norm": 0.0863695964217186,
+      "learning_rate": 5.462733694892452e-06,
+      "loss": 0.0571,
+      "step": 2322
+    },
+    {
+      "epoch": 0.4869000209599665,
+      "grad_norm": 0.13559545576572418,
+      "learning_rate": 5.4593533096824906e-06,
+      "loss": 0.0566,
+      "step": 2323
+    },
+    {
+      "epoch": 0.487109620624607,
+      "grad_norm": 0.1430426388978958,
+      "learning_rate": 5.455972712710748e-06,
+      "loss": 0.058,
+      "step": 2324
+    },
+    {
+      "epoch": 0.4873192202892475,
+      "grad_norm": 0.09979075938463211,
+      "learning_rate": 5.4525919055356804e-06,
+      "loss": 0.0557,
+      "step": 2325
+    },
+    {
+      "epoch": 0.4875288199538881,
+      "grad_norm": 0.07715443521738052,
+      "learning_rate": 5.449210889715837e-06,
+      "loss": 0.0599,
+      "step": 2326
+    },
+    {
+      "epoch": 0.4877384196185286,
+      "grad_norm": 0.09540057182312012,
+      "learning_rate": 5.445829666809866e-06,
+      "loss": 0.0527,
+      "step": 2327
+    },
+    {
+      "epoch": 0.48794801928316917,
+      "grad_norm": 0.10082437098026276,
+      "learning_rate": 5.442448238376513e-06,
+      "loss": 0.058,
+      "step": 2328
+    },
+    {
+      "epoch": 0.4881576189478097,
+      "grad_norm": 0.09437216818332672,
+      "learning_rate": 5.439066605974615e-06,
+      "loss": 0.0533,
+      "step": 2329
+    },
+    {
+      "epoch": 0.4883672186124502,
+      "grad_norm": 0.10088086873292923,
+      "learning_rate": 5.4356847711631015e-06,
+      "loss": 0.0566,
+      "step": 2330
+    },
+    {
+      "epoch": 0.48857681827709076,
+      "grad_norm": 0.09661588817834854,
+      "learning_rate": 5.432302735500998e-06,
+      "loss": 0.0525,
+      "step": 2331
+    },
+    {
+      "epoch": 0.4887864179417313,
+      "grad_norm": 0.08586642891168594,
+      "learning_rate": 5.428920500547425e-06,
+      "loss": 0.0568,
+      "step": 2332
+    },
+    {
+      "epoch": 0.48899601760637185,
+      "grad_norm": 0.09433707594871521,
+      "learning_rate": 5.4255380678615885e-06,
+      "loss": 0.0546,
+      "step": 2333
+    },
+    {
+      "epoch": 0.48920561727101236,
+      "grad_norm": 0.1241917684674263,
+      "learning_rate": 5.422155439002793e-06,
+      "loss": 0.0577,
+      "step": 2334
+    },
+    {
+      "epoch": 0.48941521693565293,
+      "grad_norm": 0.11522339284420013,
+      "learning_rate": 5.418772615530426e-06,
+      "loss": 0.0575,
+      "step": 2335
+    },
+    {
+      "epoch": 0.48962481660029344,
+      "grad_norm": 0.09254017472267151,
+      "learning_rate": 5.415389599003972e-06,
+      "loss": 0.0549,
+      "step": 2336
+    },
+    {
+      "epoch": 0.48983441626493396,
+      "grad_norm": 0.1002906784415245,
+      "learning_rate": 5.412006390982999e-06,
+      "loss": 0.0554,
+      "step": 2337
+    },
+    {
+      "epoch": 0.49004401592957453,
+      "grad_norm": 0.09268494695425034,
+      "learning_rate": 5.4086229930271636e-06,
+      "loss": 0.059,
+      "step": 2338
+    },
+    {
+      "epoch": 0.49025361559421504,
+      "grad_norm": 0.0953909158706665,
+      "learning_rate": 5.405239406696216e-06,
+      "loss": 0.054,
+      "step": 2339
+    },
+    {
+      "epoch": 0.4904632152588556,
+      "grad_norm": 0.10069115459918976,
+      "learning_rate": 5.401855633549984e-06,
+      "loss": 0.0563,
+      "step": 2340
+    },
+    {
+      "epoch": 0.4906728149234961,
+      "grad_norm": 0.09953933954238892,
+      "learning_rate": 5.398471675148389e-06,
+      "loss": 0.0573,
+      "step": 2341
+    },
+    {
+      "epoch": 0.49088241458813664,
+      "grad_norm": 0.11302805691957474,
+      "learning_rate": 5.395087533051434e-06,
+      "loss": 0.0545,
+      "step": 2342
+    },
+    {
+      "epoch": 0.4910920142527772,
+      "grad_norm": 0.10684003680944443,
+      "learning_rate": 5.391703208819209e-06,
+      "loss": 0.0559,
+      "step": 2343
+    },
+    {
+      "epoch": 0.4913016139174177,
+      "grad_norm": 0.09628332406282425,
+      "learning_rate": 5.388318704011885e-06,
+      "loss": 0.0545,
+      "step": 2344
+    },
+    {
+      "epoch": 0.4915112135820583,
+      "grad_norm": 0.10096313059329987,
+      "learning_rate": 5.38493402018972e-06,
+      "loss": 0.059,
+      "step": 2345
+    },
+    {
+      "epoch": 0.4917208132466988,
+      "grad_norm": 0.08674436807632446,
+      "learning_rate": 5.3815491589130506e-06,
+      "loss": 0.0603,
+      "step": 2346
+    },
+    {
+      "epoch": 0.4919304129113393,
+      "grad_norm": 0.0984383150935173,
+      "learning_rate": 5.378164121742301e-06,
+      "loss": 0.0552,
+      "step": 2347
+    },
+    {
+      "epoch": 0.4921400125759799,
+      "grad_norm": 0.10171261429786682,
+      "learning_rate": 5.374778910237968e-06,
+      "loss": 0.0556,
+      "step": 2348
+    },
+    {
+      "epoch": 0.4923496122406204,
+      "grad_norm": 0.09578763693571091,
+      "learning_rate": 5.3713935259606345e-06,
+      "loss": 0.0573,
+      "step": 2349
+    },
+    {
+      "epoch": 0.492559211905261,
+      "grad_norm": 0.09438461065292358,
+      "learning_rate": 5.368007970470964e-06,
+      "loss": 0.0582,
+      "step": 2350
+    },
+    {
+      "epoch": 0.4927688115699015,
+      "grad_norm": 0.08926350623369217,
+      "learning_rate": 5.3646222453296936e-06,
+      "loss": 0.0568,
+      "step": 2351
+    },
+    {
+      "epoch": 0.492978411234542,
+      "grad_norm": 0.09703266620635986,
+      "learning_rate": 5.361236352097646e-06,
+      "loss": 0.057,
+      "step": 2352
+    },
+    {
+      "epoch": 0.49318801089918257,
+      "grad_norm": 0.0997847244143486,
+      "learning_rate": 5.357850292335715e-06,
+      "loss": 0.0538,
+      "step": 2353
+    },
+    {
+      "epoch": 0.4933976105638231,
+      "grad_norm": 0.09428234398365021,
+      "learning_rate": 5.354464067604872e-06,
+      "loss": 0.0584,
+      "step": 2354
+    },
+    {
+      "epoch": 0.49360721022846366,
+      "grad_norm": 0.10148053616285324,
+      "learning_rate": 5.35107767946617e-06,
+      "loss": 0.0593,
+      "step": 2355
+    },
+    {
+      "epoch": 0.49381680989310417,
+      "grad_norm": 0.11165421456098557,
+      "learning_rate": 5.3476911294807284e-06,
+      "loss": 0.057,
+      "step": 2356
+    },
+    {
+      "epoch": 0.4940264095577447,
+      "grad_norm": 0.0888749286532402,
+      "learning_rate": 5.344304419209748e-06,
+      "loss": 0.0558,
+      "step": 2357
+    },
+    {
+      "epoch": 0.49423600922238525,
+      "grad_norm": 0.10030533373355865,
+      "learning_rate": 5.340917550214504e-06,
+      "loss": 0.0587,
+      "step": 2358
+    },
+    {
+      "epoch": 0.49444560888702577,
+      "grad_norm": 0.10392221808433533,
+      "learning_rate": 5.337530524056338e-06,
+      "loss": 0.0557,
+      "step": 2359
+    },
+    {
+      "epoch": 0.49465520855166634,
+      "grad_norm": 0.09198975563049316,
+      "learning_rate": 5.334143342296672e-06,
+      "loss": 0.0559,
+      "step": 2360
+    },
+    {
+      "epoch": 0.49486480821630685,
+      "grad_norm": 0.0988771989941597,
+      "learning_rate": 5.3307560064969955e-06,
+      "loss": 0.0561,
+      "step": 2361
+    },
+    {
+      "epoch": 0.49507440788094736,
+      "grad_norm": 0.09191461652517319,
+      "learning_rate": 5.327368518218866e-06,
+      "loss": 0.0575,
+      "step": 2362
+    },
+    {
+      "epoch": 0.49528400754558793,
+      "grad_norm": 0.08047354221343994,
+      "learning_rate": 5.323980879023919e-06,
+      "loss": 0.0582,
+      "step": 2363
+    },
+    {
+      "epoch": 0.49549360721022845,
+      "grad_norm": 0.097519151866436,
+      "learning_rate": 5.3205930904738544e-06,
+      "loss": 0.0563,
+      "step": 2364
+    },
+    {
+      "epoch": 0.495703206874869,
+      "grad_norm": 0.09642237424850464,
+      "learning_rate": 5.317205154130442e-06,
+      "loss": 0.0562,
+      "step": 2365
+    },
+    {
+      "epoch": 0.49591280653950953,
+      "grad_norm": 0.09264837205410004,
+      "learning_rate": 5.3138170715555186e-06,
+      "loss": 0.0574,
+      "step": 2366
+    },
+    {
+      "epoch": 0.49612240620415005,
+      "grad_norm": 0.10126832872629166,
+      "learning_rate": 5.31042884431099e-06,
+      "loss": 0.0553,
+      "step": 2367
+    },
+    {
+      "epoch": 0.4963320058687906,
+      "grad_norm": 0.09630407392978668,
+      "learning_rate": 5.3070404739588285e-06,
+      "loss": 0.0527,
+      "step": 2368
+    },
+    {
+      "epoch": 0.49654160553343113,
+      "grad_norm": 0.0988563820719719,
+      "learning_rate": 5.303651962061074e-06,
+      "loss": 0.0599,
+      "step": 2369
+    },
+    {
+      "epoch": 0.4967512051980717,
+      "grad_norm": 0.09341637045145035,
+      "learning_rate": 5.300263310179826e-06,
+      "loss": 0.0563,
+      "step": 2370
+    },
+    {
+      "epoch": 0.4969608048627122,
+      "grad_norm": 0.08345013111829758,
+      "learning_rate": 5.296874519877256e-06,
+      "loss": 0.0572,
+      "step": 2371
+    },
+    {
+      "epoch": 0.4971704045273528,
+      "grad_norm": 0.08412396162748337,
+      "learning_rate": 5.293485592715593e-06,
+      "loss": 0.0528,
+      "step": 2372
+    },
+    {
+      "epoch": 0.4973800041919933,
+      "grad_norm": 0.09522940963506699,
+      "learning_rate": 5.290096530257134e-06,
+      "loss": 0.0576,
+      "step": 2373
+    },
+    {
+      "epoch": 0.4975896038566338,
+      "grad_norm": 0.0857052206993103,
+      "learning_rate": 5.286707334064234e-06,
+      "loss": 0.0547,
+      "step": 2374
+    },
+    {
+      "epoch": 0.4977992035212744,
+      "grad_norm": 0.08157432824373245,
+      "learning_rate": 5.283318005699313e-06,
+      "loss": 0.0595,
+      "step": 2375
+    },
+    {
+      "epoch": 0.4980088031859149,
+      "grad_norm": 0.08909313380718231,
+      "learning_rate": 5.279928546724849e-06,
+      "loss": 0.0555,
+      "step": 2376
+    },
+    {
+      "epoch": 0.49821840285055546,
+      "grad_norm": 0.0913192629814148,
+      "learning_rate": 5.276538958703384e-06,
+      "loss": 0.0553,
+      "step": 2377
+    },
+    {
+      "epoch": 0.498428002515196,
+      "grad_norm": 0.09321916103363037,
+      "learning_rate": 5.273149243197517e-06,
+      "loss": 0.0555,
+      "step": 2378
+    },
+    {
+      "epoch": 0.4986376021798365,
+      "grad_norm": 0.09494626522064209,
+      "learning_rate": 5.269759401769904e-06,
+      "loss": 0.0588,
+      "step": 2379
+    },
+    {
+      "epoch": 0.49884720184447706,
+      "grad_norm": 0.08014019578695297,
+      "learning_rate": 5.266369435983264e-06,
+      "loss": 0.0564,
+      "step": 2380
+    },
+    {
+      "epoch": 0.4990568015091176,
+      "grad_norm": 0.08060725778341293,
+      "learning_rate": 5.26297934740037e-06,
+      "loss": 0.0594,
+      "step": 2381
+    },
+    {
+      "epoch": 0.49926640117375815,
+      "grad_norm": 0.09616056829690933,
+      "learning_rate": 5.259589137584049e-06,
+      "loss": 0.0564,
+      "step": 2382
+    },
+    {
+      "epoch": 0.49947600083839866,
+      "grad_norm": 0.10162606090307236,
+      "learning_rate": 5.256198808097192e-06,
+      "loss": 0.0553,
+      "step": 2383
+    },
+    {
+      "epoch": 0.4996856005030392,
+      "grad_norm": 0.09933992475271225,
+      "learning_rate": 5.252808360502737e-06,
+      "loss": 0.0562,
+      "step": 2384
+    },
+    {
+      "epoch": 0.49989520016767974,
+      "grad_norm": 0.106983482837677,
+      "learning_rate": 5.2494177963636785e-06,
+      "loss": 0.0579,
+      "step": 2385
+    },
+    {
+      "epoch": 0.5001047998323203,
+      "grad_norm": 0.10930163413286209,
+      "learning_rate": 5.246027117243071e-06,
+      "loss": 0.0578,
+      "step": 2386
+    },
+    {
+      "epoch": 0.5003143994969608,
+      "grad_norm": 0.10855498164892197,
+      "learning_rate": 5.24263632470401e-06,
+      "loss": 0.0547,
+      "step": 2387
+    },
+    {
+      "epoch": 0.5005239991616014,
+      "grad_norm": 0.09061628580093384,
+      "learning_rate": 5.239245420309658e-06,
+      "loss": 0.0546,
+      "step": 2388
+    },
+    {
+      "epoch": 0.5007335988262419,
+      "grad_norm": 0.07383181154727936,
+      "learning_rate": 5.2358544056232145e-06,
+      "loss": 0.0568,
+      "step": 2389
+    },
+    {
+      "epoch": 0.5009431984908824,
+      "grad_norm": 0.07915870100259781,
+      "learning_rate": 5.232463282207937e-06,
+      "loss": 0.058,
+      "step": 2390
+    },
+    {
+      "epoch": 0.5011527981555229,
+      "grad_norm": 0.07858150452375412,
+      "learning_rate": 5.229072051627137e-06,
+      "loss": 0.0581,
+      "step": 2391
+    },
+    {
+      "epoch": 0.5013623978201635,
+      "grad_norm": 0.07529259473085403,
+      "learning_rate": 5.225680715444168e-06,
+      "loss": 0.0572,
+      "step": 2392
+    },
+    {
+      "epoch": 0.5015719974848041,
+      "grad_norm": 0.07334654778242111,
+      "learning_rate": 5.222289275222436e-06,
+      "loss": 0.0565,
+      "step": 2393
+    },
+    {
+      "epoch": 0.5017815971494446,
+      "grad_norm": 0.07713853567838669,
+      "learning_rate": 5.218897732525397e-06,
+      "loss": 0.0555,
+      "step": 2394
+    },
+    {
+      "epoch": 0.5019911968140851,
+      "grad_norm": 0.0780978575348854,
+      "learning_rate": 5.215506088916545e-06,
+      "loss": 0.0557,
+      "step": 2395
+    },
+    {
+      "epoch": 0.5022007964787256,
+      "grad_norm": 0.06955696642398834,
+      "learning_rate": 5.212114345959433e-06,
+      "loss": 0.0569,
+      "step": 2396
+    },
+    {
+      "epoch": 0.5024103961433661,
+      "grad_norm": 0.07240567356348038,
+      "learning_rate": 5.2087225052176515e-06,
+      "loss": 0.0579,
+      "step": 2397
+    },
+    {
+      "epoch": 0.5026199958080068,
+      "grad_norm": 0.08363047242164612,
+      "learning_rate": 5.205330568254837e-06,
+      "loss": 0.0534,
+      "step": 2398
+    },
+    {
+      "epoch": 0.5028295954726473,
+      "grad_norm": 0.08383505046367645,
+      "learning_rate": 5.201938536634674e-06,
+      "loss": 0.0569,
+      "step": 2399
+    },
+    {
+      "epoch": 0.5030391951372878,
+      "grad_norm": 0.07523591816425323,
+      "learning_rate": 5.198546411920888e-06,
+      "loss": 0.0564,
+      "step": 2400
+    },
+    {
+      "epoch": 0.5032487948019283,
+      "grad_norm": 0.08412330597639084,
+      "learning_rate": 5.195154195677247e-06,
+      "loss": 0.0541,
+      "step": 2401
+    },
+    {
+      "epoch": 0.5034583944665688,
+      "grad_norm": 0.08403821289539337,
+      "learning_rate": 5.1917618894675615e-06,
+      "loss": 0.0546,
+      "step": 2402
+    },
+    {
+      "epoch": 0.5036679941312094,
+      "grad_norm": 0.07715484499931335,
+      "learning_rate": 5.188369494855686e-06,
+      "loss": 0.0564,
+      "step": 2403
+    },
+    {
+      "epoch": 0.50387759379585,
+      "grad_norm": 0.07596497237682343,
+      "learning_rate": 5.184977013405512e-06,
+      "loss": 0.0592,
+      "step": 2404
+    },
+    {
+      "epoch": 0.5040871934604905,
+      "grad_norm": 0.0813627615571022,
+      "learning_rate": 5.181584446680974e-06,
+      "loss": 0.0554,
+      "step": 2405
+    },
+    {
+      "epoch": 0.504296793125131,
+      "grad_norm": 0.10759492963552475,
+      "learning_rate": 5.178191796246043e-06,
+      "loss": 0.0575,
+      "step": 2406
+    },
+    {
+      "epoch": 0.5045063927897715,
+      "grad_norm": 0.12914326786994934,
+      "learning_rate": 5.174799063664731e-06,
+      "loss": 0.0546,
+      "step": 2407
+    },
+    {
+      "epoch": 0.5047159924544121,
+      "grad_norm": 0.12072386592626572,
+      "learning_rate": 5.171406250501087e-06,
+      "loss": 0.0569,
+      "step": 2408
+    },
+    {
+      "epoch": 0.5049255921190526,
+      "grad_norm": 0.10323359817266464,
+      "learning_rate": 5.1680133583191975e-06,
+      "loss": 0.0564,
+      "step": 2409
+    },
+    {
+      "epoch": 0.5051351917836931,
+      "grad_norm": 0.10388664901256561,
+      "learning_rate": 5.164620388683183e-06,
+      "loss": 0.0547,
+      "step": 2410
+    },
+    {
+      "epoch": 0.5053447914483337,
+      "grad_norm": 0.09523274004459381,
+      "learning_rate": 5.1612273431572055e-06,
+      "loss": 0.0558,
+      "step": 2411
+    },
+    {
+      "epoch": 0.5055543911129742,
+      "grad_norm": 0.08987044543027878,
+      "learning_rate": 5.1578342233054555e-06,
+      "loss": 0.0577,
+      "step": 2412
+    },
+    {
+      "epoch": 0.5057639907776148,
+      "grad_norm": 0.08983159810304642,
+      "learning_rate": 5.154441030692162e-06,
+      "loss": 0.0571,
+      "step": 2413
+    },
+    {
+      "epoch": 0.5059735904422553,
+      "grad_norm": 0.07881300896406174,
+      "learning_rate": 5.151047766881585e-06,
+      "loss": 0.0565,
+      "step": 2414
+    },
+    {
+      "epoch": 0.5061831901068958,
+      "grad_norm": 0.07373305410146713,
+      "learning_rate": 5.147654433438018e-06,
+      "loss": 0.0545,
+      "step": 2415
+    },
+    {
+      "epoch": 0.5063927897715363,
+      "grad_norm": 0.07738931477069855,
+      "learning_rate": 5.144261031925789e-06,
+      "loss": 0.0551,
+      "step": 2416
+    },
+    {
+      "epoch": 0.5066023894361769,
+      "grad_norm": 0.07858271896839142,
+      "learning_rate": 5.140867563909254e-06,
+      "loss": 0.0541,
+      "step": 2417
+    },
+    {
+      "epoch": 0.5068119891008175,
+      "grad_norm": 0.08694379776716232,
+      "learning_rate": 5.137474030952801e-06,
+      "loss": 0.0558,
+      "step": 2418
+    },
+    {
+      "epoch": 0.507021588765458,
+      "grad_norm": 0.09215452522039413,
+      "learning_rate": 5.134080434620849e-06,
+      "loss": 0.0572,
+      "step": 2419
+    },
+    {
+      "epoch": 0.5072311884300985,
+      "grad_norm": 0.08612719178199768,
+      "learning_rate": 5.1306867764778445e-06,
+      "loss": 0.054,
+      "step": 2420
+    },
+    {
+      "epoch": 0.507440788094739,
+      "grad_norm": 0.0733933225274086,
+      "learning_rate": 5.127293058088261e-06,
+      "loss": 0.0533,
+      "step": 2421
+    },
+    {
+      "epoch": 0.5076503877593795,
+      "grad_norm": 0.06744138896465302,
+      "learning_rate": 5.1238992810166065e-06,
+      "loss": 0.0542,
+      "step": 2422
+    },
+    {
+      "epoch": 0.5078599874240202,
+      "grad_norm": 0.062108736485242844,
+      "learning_rate": 5.120505446827409e-06,
+      "loss": 0.0584,
+      "step": 2423
+    },
+    {
+      "epoch": 0.5080695870886607,
+      "grad_norm": 0.06326153874397278,
+      "learning_rate": 5.117111557085225e-06,
+      "loss": 0.0542,
+      "step": 2424
+    },
+    {
+      "epoch": 0.5082791867533012,
+      "grad_norm": 0.07352180778980255,
+      "learning_rate": 5.1137176133546376e-06,
+      "loss": 0.0584,
+      "step": 2425
+    },
+    {
+      "epoch": 0.5084887864179417,
+      "grad_norm": 0.07469355314970016,
+      "learning_rate": 5.1103236172002534e-06,
+      "loss": 0.0574,
+      "step": 2426
+    },
+    {
+      "epoch": 0.5086983860825822,
+      "grad_norm": 0.08976007997989655,
+      "learning_rate": 5.106929570186706e-06,
+      "loss": 0.057,
+      "step": 2427
+    },
+    {
+      "epoch": 0.5089079857472228,
+      "grad_norm": 0.09722107648849487,
+      "learning_rate": 5.1035354738786465e-06,
+      "loss": 0.0571,
+      "step": 2428
+    },
+    {
+      "epoch": 0.5091175854118634,
+      "grad_norm": 0.07933610677719116,
+      "learning_rate": 5.100141329840757e-06,
+      "loss": 0.0566,
+      "step": 2429
+    },
+    {
+      "epoch": 0.5093271850765039,
+      "grad_norm": 0.07435188442468643,
+      "learning_rate": 5.096747139637737e-06,
+      "loss": 0.0562,
+      "step": 2430
+    },
+    {
+      "epoch": 0.5095367847411444,
+      "grad_norm": 0.06793338805437088,
+      "learning_rate": 5.0933529048343025e-06,
+      "loss": 0.0573,
+      "step": 2431
+    },
+    {
+      "epoch": 0.5097463844057849,
+      "grad_norm": 0.07182233035564423,
+      "learning_rate": 5.089958626995199e-06,
+      "loss": 0.055,
+      "step": 2432
+    },
+    {
+      "epoch": 0.5099559840704255,
+      "grad_norm": 0.09133189171552658,
+      "learning_rate": 5.086564307685188e-06,
+      "loss": 0.0541,
+      "step": 2433
+    },
+    {
+      "epoch": 0.510165583735066,
+      "grad_norm": 0.08642099797725677,
+      "learning_rate": 5.083169948469049e-06,
+      "loss": 0.0556,
+      "step": 2434
+    },
+    {
+      "epoch": 0.5103751833997066,
+      "grad_norm": 0.08702914416790009,
+      "learning_rate": 5.07977555091158e-06,
+      "loss": 0.0571,
+      "step": 2435
+    },
+    {
+      "epoch": 0.5105847830643471,
+      "grad_norm": 0.08467281609773636,
+      "learning_rate": 5.0763811165776e-06,
+      "loss": 0.0562,
+      "step": 2436
+    },
+    {
+      "epoch": 0.5107943827289876,
+      "grad_norm": 0.06674730032682419,
+      "learning_rate": 5.07298664703194e-06,
+      "loss": 0.0522,
+      "step": 2437
+    },
+    {
+      "epoch": 0.5110039823936282,
+      "grad_norm": 0.07278907299041748,
+      "learning_rate": 5.069592143839452e-06,
+      "loss": 0.0542,
+      "step": 2438
+    },
+    {
+      "epoch": 0.5112135820582687,
+      "grad_norm": 0.0811818316578865,
+      "learning_rate": 5.066197608564999e-06,
+      "loss": 0.0573,
+      "step": 2439
+    },
+    {
+      "epoch": 0.5114231817229092,
+      "grad_norm": 0.07044421881437302,
+      "learning_rate": 5.062803042773463e-06,
+      "loss": 0.0575,
+      "step": 2440
+    },
+    {
+      "epoch": 0.5116327813875498,
+      "grad_norm": 0.06003498286008835,
+      "learning_rate": 5.059408448029737e-06,
+      "loss": 0.0577,
+      "step": 2441
+    },
+    {
+      "epoch": 0.5118423810521903,
+      "grad_norm": 0.06972794234752655,
+      "learning_rate": 5.056013825898729e-06,
+      "loss": 0.0543,
+      "step": 2442
+    },
+    {
+      "epoch": 0.5120519807168309,
+      "grad_norm": 0.07647814601659775,
+      "learning_rate": 5.052619177945358e-06,
+      "loss": 0.0548,
+      "step": 2443
+    },
+    {
+      "epoch": 0.5122615803814714,
+      "grad_norm": 0.06816070526838303,
+      "learning_rate": 5.049224505734558e-06,
+      "loss": 0.0582,
+      "step": 2444
+    },
+    {
+      "epoch": 0.5124711800461119,
+      "grad_norm": 0.0557573027908802,
+      "learning_rate": 5.045829810831271e-06,
+      "loss": 0.0547,
+      "step": 2445
+    },
+    {
+      "epoch": 0.5126807797107524,
+      "grad_norm": 0.06597109884023666,
+      "learning_rate": 5.042435094800451e-06,
+      "loss": 0.056,
+      "step": 2446
+    },
+    {
+      "epoch": 0.512890379375393,
+      "grad_norm": 0.0730803981423378,
+      "learning_rate": 5.0390403592070605e-06,
+      "loss": 0.0547,
+      "step": 2447
+    },
+    {
+      "epoch": 0.5130999790400336,
+      "grad_norm": 0.07265052199363708,
+      "learning_rate": 5.0356456056160715e-06,
+      "loss": 0.0567,
+      "step": 2448
+    },
+    {
+      "epoch": 0.5133095787046741,
+      "grad_norm": 0.07946612685918808,
+      "learning_rate": 5.032250835592467e-06,
+      "loss": 0.0576,
+      "step": 2449
+    },
+    {
+      "epoch": 0.5135191783693146,
+      "grad_norm": 0.09229972958564758,
+      "learning_rate": 5.028856050701234e-06,
+      "loss": 0.0536,
+      "step": 2450
+    },
+    {
+      "epoch": 0.5137287780339551,
+      "grad_norm": 0.1069934144616127,
+      "learning_rate": 5.025461252507367e-06,
+      "loss": 0.0562,
+      "step": 2451
+    },
+    {
+      "epoch": 0.5139383776985956,
+      "grad_norm": 0.12349528819322586,
+      "learning_rate": 5.0220664425758695e-06,
+      "loss": 0.0573,
+      "step": 2452
+    },
+    {
+      "epoch": 0.5141479773632363,
+      "grad_norm": 0.1274581402540207,
+      "learning_rate": 5.0186716224717445e-06,
+      "loss": 0.0576,
+      "step": 2453
+    },
+    {
+      "epoch": 0.5143575770278768,
+      "grad_norm": 0.11807071417570114,
+      "learning_rate": 5.0152767937600055e-06,
+      "loss": 0.0576,
+      "step": 2454
+    },
+    {
+      "epoch": 0.5145671766925173,
+      "grad_norm": 0.10544098168611526,
+      "learning_rate": 5.0118819580056686e-06,
+      "loss": 0.0589,
+      "step": 2455
+    },
+    {
+      "epoch": 0.5147767763571578,
+      "grad_norm": 0.091020368039608,
+      "learning_rate": 5.008487116773752e-06,
+      "loss": 0.0574,
+      "step": 2456
+    },
+    {
+      "epoch": 0.5149863760217984,
+      "grad_norm": 0.07477325946092606,
+      "learning_rate": 5.0050922716292745e-06,
+      "loss": 0.0584,
+      "step": 2457
+    },
+    {
+      "epoch": 0.5151959756864389,
+      "grad_norm": 0.06757844984531403,
+      "learning_rate": 5.001697424137264e-06,
+      "loss": 0.0552,
+      "step": 2458
+    },
+    {
+      "epoch": 0.5154055753510794,
+      "grad_norm": 0.06685573607683182,
+      "learning_rate": 4.9983025758627376e-06,
+      "loss": 0.0533,
+      "step": 2459
+    },
+    {
+      "epoch": 0.51561517501572,
+      "grad_norm": 0.0714949518442154,
+      "learning_rate": 4.9949077283707255e-06,
+      "loss": 0.0563,
+      "step": 2460
+    },
+    {
+      "epoch": 0.5158247746803605,
+      "grad_norm": 0.06642146408557892,
+      "learning_rate": 4.991512883226251e-06,
+      "loss": 0.0599,
+      "step": 2461
+    },
+    {
+      "epoch": 0.5160343743450011,
+      "grad_norm": 0.06495679169893265,
+      "learning_rate": 4.988118041994332e-06,
+      "loss": 0.0593,
+      "step": 2462
+    },
+    {
+      "epoch": 0.5162439740096416,
+      "grad_norm": 0.06817002594470978,
+      "learning_rate": 4.984723206239995e-06,
+      "loss": 0.0564,
+      "step": 2463
+    },
+    {
+      "epoch": 0.5164535736742821,
+      "grad_norm": 0.0714736357331276,
+      "learning_rate": 4.981328377528258e-06,
+      "loss": 0.0588,
+      "step": 2464
+    },
+    {
+      "epoch": 0.5166631733389226,
+      "grad_norm": 0.07612305879592896,
+      "learning_rate": 4.977933557424133e-06,
+      "loss": 0.0566,
+      "step": 2465
+    },
+    {
+      "epoch": 0.5168727730035632,
+      "grad_norm": 0.0783240795135498,
+      "learning_rate": 4.974538747492634e-06,
+      "loss": 0.056,
+      "step": 2466
+    },
+    {
+      "epoch": 0.5170823726682038,
+      "grad_norm": 0.07756256312131882,
+      "learning_rate": 4.971143949298769e-06,
+      "loss": 0.0546,
+      "step": 2467
+    },
+    {
+      "epoch": 0.5172919723328443,
+      "grad_norm": 0.07845515012741089,
+      "learning_rate": 4.967749164407535e-06,
+      "loss": 0.0533,
+      "step": 2468
+    },
+    {
+      "epoch": 0.5175015719974848,
+      "grad_norm": 0.08055856078863144,
+      "learning_rate": 4.964354394383929e-06,
+      "loss": 0.0586,
+      "step": 2469
+    },
+    {
+      "epoch": 0.5177111716621253,
+      "grad_norm": 0.07212040573358536,
+      "learning_rate": 4.960959640792943e-06,
+      "loss": 0.0549,
+      "step": 2470
+    },
+    {
+      "epoch": 0.5179207713267658,
+      "grad_norm": 0.07273006439208984,
+      "learning_rate": 4.9575649051995515e-06,
+      "loss": 0.0545,
+      "step": 2471
+    },
+    {
+      "epoch": 0.5181303709914065,
+      "grad_norm": 0.08731981366872787,
+      "learning_rate": 4.954170189168731e-06,
+      "loss": 0.0581,
+      "step": 2472
+    },
+    {
+      "epoch": 0.518339970656047,
+      "grad_norm": 0.07897140830755234,
+      "learning_rate": 4.950775494265443e-06,
+      "loss": 0.0562,
+      "step": 2473
+    },
+    {
+      "epoch": 0.5185495703206875,
+      "grad_norm": 0.055604368448257446,
+      "learning_rate": 4.947380822054643e-06,
+      "loss": 0.0545,
+      "step": 2474
+    },
+    {
+      "epoch": 0.518759169985328,
+      "grad_norm": 0.05771341547369957,
+      "learning_rate": 4.9439861741012726e-06,
+      "loss": 0.0541,
+      "step": 2475
+    },
+    {
+      "epoch": 0.5189687696499685,
+      "grad_norm": 0.08119264245033264,
+      "learning_rate": 4.940591551970264e-06,
+      "loss": 0.0558,
+      "step": 2476
+    },
+    {
+      "epoch": 0.5191783693146091,
+      "grad_norm": 0.08719661831855774,
+      "learning_rate": 4.93719695722654e-06,
+      "loss": 0.0553,
+      "step": 2477
+    },
+    {
+      "epoch": 0.5193879689792497,
+      "grad_norm": 0.08087395131587982,
+      "learning_rate": 4.933802391435002e-06,
+      "loss": 0.0536,
+      "step": 2478
+    },
+    {
+      "epoch": 0.5195975686438902,
+      "grad_norm": 0.07597414404153824,
+      "learning_rate": 4.93040785616055e-06,
+      "loss": 0.0563,
+      "step": 2479
+    },
+    {
+      "epoch": 0.5198071683085307,
+      "grad_norm": 0.07065032422542572,
+      "learning_rate": 4.927013352968063e-06,
+      "loss": 0.0543,
+      "step": 2480
+    },
+    {
+      "epoch": 0.5200167679731712,
+      "grad_norm": 0.06194977089762688,
+      "learning_rate": 4.9236188834224015e-06,
+      "loss": 0.0575,
+      "step": 2481
+    },
+    {
+      "epoch": 0.5202263676378118,
+      "grad_norm": 0.059151582419872284,
+      "learning_rate": 4.920224449088421e-06,
+      "loss": 0.0577,
+      "step": 2482
+    },
+    {
+      "epoch": 0.5204359673024523,
+      "grad_norm": 0.07860930263996124,
+      "learning_rate": 4.9168300515309515e-06,
+      "loss": 0.0566,
+      "step": 2483
+    },
+    {
+      "epoch": 0.5206455669670929,
+      "grad_norm": 0.11033756285905838,
+      "learning_rate": 4.913435692314814e-06,
+      "loss": 0.056,
+      "step": 2484
+    },
+    {
+      "epoch": 0.5208551666317334,
+      "grad_norm": 0.13464130461215973,
+      "learning_rate": 4.910041373004802e-06,
+      "loss": 0.0548,
+      "step": 2485
+    },
+    {
+      "epoch": 0.5210647662963739,
+      "grad_norm": 0.1375139057636261,
+      "learning_rate": 4.906647095165698e-06,
+      "loss": 0.0568,
+      "step": 2486
+    },
+    {
+      "epoch": 0.5212743659610145,
+      "grad_norm": 0.10741981118917465,
+      "learning_rate": 4.903252860362266e-06,
+      "loss": 0.0593,
+      "step": 2487
+    },
+    {
+      "epoch": 0.521483965625655,
+      "grad_norm": 0.08266045153141022,
+      "learning_rate": 4.8998586701592436e-06,
+      "loss": 0.0558,
+      "step": 2488
+    },
+    {
+      "epoch": 0.5216935652902955,
+      "grad_norm": 0.0874459519982338,
+      "learning_rate": 4.8964645261213535e-06,
+      "loss": 0.0557,
+      "step": 2489
+    },
+    {
+      "epoch": 0.521903164954936,
+      "grad_norm": 0.10738598555326462,
+      "learning_rate": 4.8930704298132965e-06,
+      "loss": 0.0556,
+      "step": 2490
+    },
+    {
+      "epoch": 0.5221127646195766,
+      "grad_norm": 0.11535639315843582,
+      "learning_rate": 4.889676382799748e-06,
+      "loss": 0.0566,
+      "step": 2491
+    },
+    {
+      "epoch": 0.5223223642842172,
+      "grad_norm": 0.10083887726068497,
+      "learning_rate": 4.886282386645363e-06,
+      "loss": 0.0567,
+      "step": 2492
+    },
+    {
+      "epoch": 0.5225319639488577,
+      "grad_norm": 0.07577608525753021,
+      "learning_rate": 4.8828884429147775e-06,
+      "loss": 0.0539,
+      "step": 2493
+    },
+    {
+      "epoch": 0.5227415636134982,
+      "grad_norm": 0.08031262457370758,
+      "learning_rate": 4.879494553172594e-06,
+      "loss": 0.0541,
+      "step": 2494
+    },
+    {
+      "epoch": 0.5229511632781387,
+      "grad_norm": 0.10877335071563721,
+      "learning_rate": 4.876100718983394e-06,
+      "loss": 0.0566,
+      "step": 2495
+    },
+    {
+      "epoch": 0.5231607629427792,
+      "grad_norm": 0.10455876588821411,
+      "learning_rate": 4.872706941911739e-06,
+      "loss": 0.0559,
+      "step": 2496
+    },
+    {
+      "epoch": 0.5233703626074199,
+      "grad_norm": 0.07597562670707703,
+      "learning_rate": 4.869313223522159e-06,
+      "loss": 0.0565,
+      "step": 2497
+    },
+    {
+      "epoch": 0.5235799622720604,
+      "grad_norm": 0.08931368589401245,
+      "learning_rate": 4.865919565379152e-06,
+      "loss": 0.0583,
+      "step": 2498
+    },
+    {
+      "epoch": 0.5237895619367009,
+      "grad_norm": 0.09384021908044815,
+      "learning_rate": 4.8625259690472e-06,
+      "loss": 0.0574,
+      "step": 2499
+    },
+    {
+      "epoch": 0.5239991616013414,
+      "grad_norm": 0.0911545678973198,
+      "learning_rate": 4.859132436090748e-06,
+      "loss": 0.0561,
+      "step": 2500
+    },
+    {
+      "epoch": 0.5242087612659819,
+      "grad_norm": 0.10391653329133987,
+      "learning_rate": 4.855738968074212e-06,
+      "loss": 0.0552,
+      "step": 2501
+    },
+    {
+      "epoch": 0.5244183609306226,
+      "grad_norm": 0.09222046285867691,
+      "learning_rate": 4.852345566561983e-06,
+      "loss": 0.059,
+      "step": 2502
+    },
+    {
+      "epoch": 0.5246279605952631,
+      "grad_norm": 0.07000227272510529,
+      "learning_rate": 4.848952233118417e-06,
+      "loss": 0.0584,
+      "step": 2503
+    },
+    {
+      "epoch": 0.5248375602599036,
+      "grad_norm": 0.08648461848497391,
+      "learning_rate": 4.845558969307839e-06,
+      "loss": 0.0541,
+      "step": 2504
+    },
+    {
+      "epoch": 0.5250471599245441,
+      "grad_norm": 0.09578924626111984,
+      "learning_rate": 4.842165776694545e-06,
+      "loss": 0.0551,
+      "step": 2505
+    },
+    {
+      "epoch": 0.5252567595891846,
+      "grad_norm": 0.091171033680439,
+      "learning_rate": 4.8387726568427945e-06,
+      "loss": 0.055,
+      "step": 2506
+    },
+    {
+      "epoch": 0.5254663592538252,
+      "grad_norm": 0.09794994443655014,
+      "learning_rate": 4.835379611316818e-06,
+      "loss": 0.0602,
+      "step": 2507
+    },
+    {
+      "epoch": 0.5256759589184657,
+      "grad_norm": 0.08770140260457993,
+      "learning_rate": 4.831986641680804e-06,
+      "loss": 0.0555,
+      "step": 2508
+    },
+    {
+      "epoch": 0.5258855585831063,
+      "grad_norm": 0.07761924713850021,
+      "learning_rate": 4.828593749498913e-06,
+      "loss": 0.0558,
+      "step": 2509
+    },
+    {
+      "epoch": 0.5260951582477468,
+      "grad_norm": 0.09006881713867188,
+      "learning_rate": 4.825200936335272e-06,
+      "loss": 0.0575,
+      "step": 2510
+    },
+    {
+      "epoch": 0.5263047579123873,
+      "grad_norm": 0.08472940325737,
+      "learning_rate": 4.821808203753959e-06,
+      "loss": 0.0555,
+      "step": 2511
+    },
+    {
+      "epoch": 0.5265143575770279,
+      "grad_norm": 0.06748061627149582,
+      "learning_rate": 4.818415553319027e-06,
+      "loss": 0.0555,
+      "step": 2512
+    },
+    {
+      "epoch": 0.5267239572416684,
+      "grad_norm": 0.07318470627069473,
+      "learning_rate": 4.815022986594491e-06,
+      "loss": 0.0564,
+      "step": 2513
+    },
+    {
+      "epoch": 0.5269335569063089,
+      "grad_norm": 0.07728718966245651,
+      "learning_rate": 4.811630505144316e-06,
+      "loss": 0.055,
+      "step": 2514
+    },
+    {
+      "epoch": 0.5271431565709495,
+      "grad_norm": 0.06775672733783722,
+      "learning_rate": 4.808238110532439e-06,
+      "loss": 0.0543,
+      "step": 2515
+    },
+    {
+      "epoch": 0.52735275623559,
+      "grad_norm": 0.08202599734067917,
+      "learning_rate": 4.804845804322756e-06,
+      "loss": 0.0562,
+      "step": 2516
+    },
+    {
+      "epoch": 0.5275623559002306,
+      "grad_norm": 0.09403207898139954,
+      "learning_rate": 4.801453588079113e-06,
+      "loss": 0.0548,
+      "step": 2517
+    },
+    {
+      "epoch": 0.5277719555648711,
+      "grad_norm": 0.08758095651865005,
+      "learning_rate": 4.798061463365327e-06,
+      "loss": 0.0533,
+      "step": 2518
+    },
+    {
+      "epoch": 0.5279815552295116,
+      "grad_norm": 0.07951238006353378,
+      "learning_rate": 4.7946694317451635e-06,
+      "loss": 0.0558,
+      "step": 2519
+    },
+    {
+      "epoch": 0.5281911548941521,
+      "grad_norm": 0.06479839980602264,
+      "learning_rate": 4.791277494782351e-06,
+      "loss": 0.056,
+      "step": 2520
+    },
+    {
+      "epoch": 0.5284007545587927,
+      "grad_norm": 0.060277167707681656,
+      "learning_rate": 4.787885654040569e-06,
+      "loss": 0.0572,
+      "step": 2521
+    },
+    {
+      "epoch": 0.5286103542234333,
+      "grad_norm": 0.06489501148462296,
+      "learning_rate": 4.784493911083455e-06,
+      "loss": 0.0559,
+      "step": 2522
+    },
+    {
+      "epoch": 0.5288199538880738,
+      "grad_norm": 0.07337628304958344,
+      "learning_rate": 4.781102267474606e-06,
+      "loss": 0.0576,
+      "step": 2523
+    },
+    {
+      "epoch": 0.5290295535527143,
+      "grad_norm": 0.08746179193258286,
+      "learning_rate": 4.777710724777565e-06,
+      "loss": 0.0556,
+      "step": 2524
+    },
+    {
+      "epoch": 0.5292391532173548,
+      "grad_norm": 0.09314849972724915,
+      "learning_rate": 4.774319284555833e-06,
+      "loss": 0.0566,
+      "step": 2525
+    },
+    {
+      "epoch": 0.5294487528819954,
+      "grad_norm": 0.08934000134468079,
+      "learning_rate": 4.770927948372865e-06,
+      "loss": 0.0522,
+      "step": 2526
+    },
+    {
+      "epoch": 0.529658352546636,
+      "grad_norm": 0.08320695906877518,
+      "learning_rate": 4.7675367177920645e-06,
+      "loss": 0.0547,
+      "step": 2527
+    },
+    {
+      "epoch": 0.5298679522112765,
+      "grad_norm": 0.07654823362827301,
+      "learning_rate": 4.764145594376788e-06,
+      "loss": 0.0531,
+      "step": 2528
+    },
+    {
+      "epoch": 0.530077551875917,
+      "grad_norm": 0.07065373659133911,
+      "learning_rate": 4.7607545796903444e-06,
+      "loss": 0.0584,
+      "step": 2529
+    },
+    {
+      "epoch": 0.5302871515405575,
+      "grad_norm": 0.06831829994916916,
+      "learning_rate": 4.757363675295991e-06,
+      "loss": 0.0557,
+      "step": 2530
+    },
+    {
+      "epoch": 0.5304967512051981,
+      "grad_norm": 0.06134999543428421,
+      "learning_rate": 4.753972882756931e-06,
+      "loss": 0.0548,
+      "step": 2531
+    },
+    {
+      "epoch": 0.5307063508698386,
+      "grad_norm": 0.05438331514596939,
+      "learning_rate": 4.7505822036363214e-06,
+      "loss": 0.0568,
+      "step": 2532
+    },
+    {
+      "epoch": 0.5309159505344792,
+      "grad_norm": 0.047348763793706894,
+      "learning_rate": 4.747191639497266e-06,
+      "loss": 0.0568,
+      "step": 2533
+    },
+    {
+      "epoch": 0.5311255501991197,
+      "grad_norm": 0.048679519444704056,
+      "learning_rate": 4.743801191902809e-06,
+      "loss": 0.0563,
+      "step": 2534
+    },
+    {
+      "epoch": 0.5313351498637602,
+      "grad_norm": 0.045131586492061615,
+      "learning_rate": 4.740410862415952e-06,
+      "loss": 0.0555,
+      "step": 2535
+    },
+    {
+      "epoch": 0.5315447495284008,
+      "grad_norm": 0.05211302638053894,
+      "learning_rate": 4.737020652599633e-06,
+      "loss": 0.0551,
+      "step": 2536
+    },
+    {
+      "epoch": 0.5317543491930413,
+      "grad_norm": 0.05385665223002434,
+      "learning_rate": 4.733630564016738e-06,
+      "loss": 0.0541,
+      "step": 2537
+    },
+    {
+      "epoch": 0.5319639488576818,
+      "grad_norm": 0.046267565339803696,
+      "learning_rate": 4.730240598230097e-06,
+      "loss": 0.0533,
+      "step": 2538
+    },
+    {
+      "epoch": 0.5321735485223223,
+      "grad_norm": 0.04819132760167122,
+      "learning_rate": 4.726850756802486e-06,
+      "loss": 0.0564,
+      "step": 2539
+    },
+    {
+      "epoch": 0.5323831481869629,
+      "grad_norm": 0.048858221620321274,
+      "learning_rate": 4.723461041296618e-06,
+      "loss": 0.0547,
+      "step": 2540
+    },
+    {
+      "epoch": 0.5325927478516035,
+      "grad_norm": 0.042201049625873566,
+      "learning_rate": 4.720071453275152e-06,
+      "loss": 0.0573,
+      "step": 2541
+    },
+    {
+      "epoch": 0.532802347516244,
+      "grad_norm": 0.04239390045404434,
+      "learning_rate": 4.716681994300688e-06,
+      "loss": 0.0573,
+      "step": 2542
+    },
+    {
+      "epoch": 0.5330119471808845,
+      "grad_norm": 0.05800594016909599,
+      "learning_rate": 4.7132926659357675e-06,
+      "loss": 0.0546,
+      "step": 2543
+    },
+    {
+      "epoch": 0.533221546845525,
+      "grad_norm": 0.05549288168549538,
+      "learning_rate": 4.7099034697428676e-06,
+      "loss": 0.0565,
+      "step": 2544
+    },
+    {
+      "epoch": 0.5334311465101655,
+      "grad_norm": 0.05524330586194992,
+      "learning_rate": 4.706514407284407e-06,
+      "loss": 0.0544,
+      "step": 2545
+    },
+    {
+      "epoch": 0.5336407461748062,
+      "grad_norm": 0.07277749478816986,
+      "learning_rate": 4.703125480122747e-06,
+      "loss": 0.0542,
+      "step": 2546
+    },
+    {
+      "epoch": 0.5338503458394467,
+      "grad_norm": 0.06170268729329109,
+      "learning_rate": 4.699736689820175e-06,
+      "loss": 0.0565,
+      "step": 2547
+    },
+    {
+      "epoch": 0.5340599455040872,
+      "grad_norm": 0.05076323822140694,
+      "learning_rate": 4.696348037938927e-06,
+      "loss": 0.0565,
+      "step": 2548
+    },
+    {
+      "epoch": 0.5342695451687277,
+      "grad_norm": 0.06921599805355072,
+      "learning_rate": 4.692959526041174e-06,
+      "loss": 0.0565,
+      "step": 2549
+    },
+    {
+      "epoch": 0.5344791448333682,
+      "grad_norm": 0.057804301381111145,
+      "learning_rate": 4.689571155689012e-06,
+      "loss": 0.053,
+      "step": 2550
+    },
+    {
+      "epoch": 0.5346887444980088,
+      "grad_norm": 0.04741832986474037,
+      "learning_rate": 4.686182928444484e-06,
+      "loss": 0.0552,
+      "step": 2551
+    },
+    {
+      "epoch": 0.5348983441626494,
+      "grad_norm": 0.07584595680236816,
+      "learning_rate": 4.682794845869559e-06,
+      "loss": 0.0525,
+      "step": 2552
+    },
+    {
+      "epoch": 0.5351079438272899,
+      "grad_norm": 0.07168301194906235,
+      "learning_rate": 4.679406909526147e-06,
+      "loss": 0.0579,
+      "step": 2553
+    },
+    {
+      "epoch": 0.5353175434919304,
+      "grad_norm": 0.04884837195277214,
+      "learning_rate": 4.676019120976082e-06,
+      "loss": 0.0574,
+      "step": 2554
+    },
+    {
+      "epoch": 0.5355271431565709,
+      "grad_norm": 0.06782133877277374,
+      "learning_rate": 4.672631481781134e-06,
+      "loss": 0.0565,
+      "step": 2555
+    },
+    {
+      "epoch": 0.5357367428212115,
+      "grad_norm": 0.07214924693107605,
+      "learning_rate": 4.669243993503008e-06,
+      "loss": 0.0584,
+      "step": 2556
+    },
+    {
+      "epoch": 0.535946342485852,
+      "grad_norm": 0.05515038222074509,
+      "learning_rate": 4.665856657703329e-06,
+      "loss": 0.0545,
+      "step": 2557
+    },
+    {
+      "epoch": 0.5361559421504926,
+      "grad_norm": 0.06522560119628906,
+      "learning_rate": 4.662469475943662e-06,
+      "loss": 0.0558,
+      "step": 2558
+    },
+    {
+      "epoch": 0.5363655418151331,
+      "grad_norm": 0.06113879382610321,
+      "learning_rate": 4.659082449785498e-06,
+      "loss": 0.0558,
+      "step": 2559
+    },
+    {
+      "epoch": 0.5365751414797736,
+      "grad_norm": 0.06243380904197693,
+      "learning_rate": 4.655695580790254e-06,
+      "loss": 0.0578,
+      "step": 2560
+    },
+    {
+      "epoch": 0.5367847411444142,
+      "grad_norm": 0.07164395600557327,
+      "learning_rate": 4.652308870519272e-06,
+      "loss": 0.0574,
+      "step": 2561
+    },
+    {
+      "epoch": 0.5369943408090547,
+      "grad_norm": 0.06726083904504776,
+      "learning_rate": 4.648922320533833e-06,
+      "loss": 0.0546,
+      "step": 2562
+    },
+    {
+      "epoch": 0.5372039404736952,
+      "grad_norm": 0.054155874997377396,
+      "learning_rate": 4.645535932395129e-06,
+      "loss": 0.0536,
+      "step": 2563
+    },
+    {
+      "epoch": 0.5374135401383358,
+      "grad_norm": 0.0629991814494133,
+      "learning_rate": 4.6421497076642864e-06,
+      "loss": 0.0576,
+      "step": 2564
+    },
+    {
+      "epoch": 0.5376231398029763,
+      "grad_norm": 0.06752461940050125,
+      "learning_rate": 4.638763647902355e-06,
+      "loss": 0.0567,
+      "step": 2565
+    },
+    {
+      "epoch": 0.5378327394676169,
+      "grad_norm": 0.0500505268573761,
+      "learning_rate": 4.635377754670307e-06,
+      "loss": 0.0546,
+      "step": 2566
+    },
+    {
+      "epoch": 0.5380423391322574,
+      "grad_norm": 0.043680962175130844,
+      "learning_rate": 4.631992029529037e-06,
+      "loss": 0.0556,
+      "step": 2567
+    },
+    {
+      "epoch": 0.5382519387968979,
+      "grad_norm": 0.06089196354150772,
+      "learning_rate": 4.628606474039366e-06,
+      "loss": 0.0545,
+      "step": 2568
+    },
+    {
+      "epoch": 0.5384615384615384,
+      "grad_norm": 0.07623324543237686,
+      "learning_rate": 4.625221089762034e-06,
+      "loss": 0.0577,
+      "step": 2569
+    },
+    {
+      "epoch": 0.538671138126179,
+      "grad_norm": 0.09632382541894913,
+      "learning_rate": 4.621835878257701e-06,
+      "loss": 0.0567,
+      "step": 2570
+    },
+    {
+      "epoch": 0.5388807377908196,
+      "grad_norm": 0.12334061414003372,
+      "learning_rate": 4.6184508410869486e-06,
+      "loss": 0.0533,
+      "step": 2571
+    },
+    {
+      "epoch": 0.5390903374554601,
+      "grad_norm": 0.1471899002790451,
+      "learning_rate": 4.615065979810282e-06,
+      "loss": 0.0571,
+      "step": 2572
+    },
+    {
+      "epoch": 0.5392999371201006,
+      "grad_norm": 0.16688735783100128,
+      "learning_rate": 4.6116812959881154e-06,
+      "loss": 0.0543,
+      "step": 2573
+    },
+    {
+      "epoch": 0.5395095367847411,
+      "grad_norm": 0.17000025510787964,
+      "learning_rate": 4.608296791180793e-06,
+      "loss": 0.0555,
+      "step": 2574
+    },
+    {
+      "epoch": 0.5397191364493816,
+      "grad_norm": 0.13026584684848785,
+      "learning_rate": 4.604912466948568e-06,
+      "loss": 0.058,
+      "step": 2575
+    },
+    {
+      "epoch": 0.5399287361140223,
+      "grad_norm": 0.06306587904691696,
+      "learning_rate": 4.601528324851613e-06,
+      "loss": 0.0546,
+      "step": 2576
+    },
+    {
+      "epoch": 0.5401383357786628,
+      "grad_norm": 0.06573072820901871,
+      "learning_rate": 4.598144366450018e-06,
+      "loss": 0.057,
+      "step": 2577
+    },
+    {
+      "epoch": 0.5403479354433033,
+      "grad_norm": 0.1011551097035408,
+      "learning_rate": 4.594760593303785e-06,
+      "loss": 0.0563,
+      "step": 2578
+    },
+    {
+      "epoch": 0.5405575351079438,
+      "grad_norm": 0.13201986253261566,
+      "learning_rate": 4.591377006972837e-06,
+      "loss": 0.0559,
+      "step": 2579
+    },
+    {
+      "epoch": 0.5407671347725843,
+      "grad_norm": 0.13169518113136292,
+      "learning_rate": 4.587993609017003e-06,
+      "loss": 0.0583,
+      "step": 2580
+    },
+    {
+      "epoch": 0.5409767344372249,
+      "grad_norm": 0.07675211131572723,
+      "learning_rate": 4.584610400996028e-06,
+      "loss": 0.0547,
+      "step": 2581
+    },
+    {
+      "epoch": 0.5411863341018655,
+      "grad_norm": 0.07180249691009521,
+      "learning_rate": 4.581227384469575e-06,
+      "loss": 0.056,
+      "step": 2582
+    },
+    {
+      "epoch": 0.541395933766506,
+      "grad_norm": 0.11214875429868698,
+      "learning_rate": 4.577844560997208e-06,
+      "loss": 0.0561,
+      "step": 2583
+    },
+    {
+      "epoch": 0.5416055334311465,
+      "grad_norm": 0.1112339124083519,
+      "learning_rate": 4.574461932138412e-06,
+      "loss": 0.0535,
+      "step": 2584
+    },
+    {
+      "epoch": 0.541815133095787,
+      "grad_norm": 0.08749670535326004,
+      "learning_rate": 4.571079499452578e-06,
+      "loss": 0.057,
+      "step": 2585
+    },
+    {
+      "epoch": 0.5420247327604276,
+      "grad_norm": 0.08394355326890945,
+      "learning_rate": 4.567697264499003e-06,
+      "loss": 0.0547,
+      "step": 2586
+    },
+    {
+      "epoch": 0.5422343324250681,
+      "grad_norm": 0.12321101129055023,
+      "learning_rate": 4.564315228836901e-06,
+      "loss": 0.0577,
+      "step": 2587
+    },
+    {
+      "epoch": 0.5424439320897086,
+      "grad_norm": 0.13158167898654938,
+      "learning_rate": 4.560933394025386e-06,
+      "loss": 0.0583,
+      "step": 2588
+    },
+    {
+      "epoch": 0.5426535317543492,
+      "grad_norm": 0.08250062167644501,
+      "learning_rate": 4.5575517616234874e-06,
+      "loss": 0.0547,
+      "step": 2589
+    },
+    {
+      "epoch": 0.5428631314189897,
+      "grad_norm": 0.07321292906999588,
+      "learning_rate": 4.5541703331901346e-06,
+      "loss": 0.0575,
+      "step": 2590
+    },
+    {
+      "epoch": 0.5430727310836303,
+      "grad_norm": 0.10240411758422852,
+      "learning_rate": 4.550789110284164e-06,
+      "loss": 0.0548,
+      "step": 2591
+    },
+    {
+      "epoch": 0.5432823307482708,
+      "grad_norm": 0.09402070939540863,
+      "learning_rate": 4.547408094464322e-06,
+      "loss": 0.0568,
+      "step": 2592
+    },
+    {
+      "epoch": 0.5434919304129113,
+      "grad_norm": 0.06143482029438019,
+      "learning_rate": 4.5440272872892546e-06,
+      "loss": 0.0552,
+      "step": 2593
+    },
+    {
+      "epoch": 0.5437015300775518,
+      "grad_norm": 0.07297394424676895,
+      "learning_rate": 4.54064669031751e-06,
+      "loss": 0.0561,
+      "step": 2594
+    },
+    {
+      "epoch": 0.5439111297421925,
+      "grad_norm": 0.08411424607038498,
+      "learning_rate": 4.537266305107549e-06,
+      "loss": 0.0554,
+      "step": 2595
+    },
+    {
+      "epoch": 0.544120729406833,
+      "grad_norm": 0.06542636454105377,
+      "learning_rate": 4.533886133217725e-06,
+      "loss": 0.0563,
+      "step": 2596
+    },
+    {
+      "epoch": 0.5443303290714735,
+      "grad_norm": 0.06550241261720657,
+      "learning_rate": 4.5305061762062945e-06,
+      "loss": 0.0535,
+      "step": 2597
+    },
+    {
+      "epoch": 0.544539928736114,
+      "grad_norm": 0.08619559556245804,
+      "learning_rate": 4.527126435631422e-06,
+      "loss": 0.0566,
+      "step": 2598
+    },
+    {
+      "epoch": 0.5447495284007545,
+      "grad_norm": 0.07745786011219025,
+      "learning_rate": 4.523746913051163e-06,
+      "loss": 0.0571,
+      "step": 2599
+    },
+    {
+      "epoch": 0.5449591280653951,
+      "grad_norm": 0.0609513558447361,
+      "learning_rate": 4.520367610023477e-06,
+      "loss": 0.0552,
+      "step": 2600
+    },
+    {
+      "epoch": 0.5451687277300357,
+      "grad_norm": 0.0679200291633606,
+      "learning_rate": 4.516988528106225e-06,
+      "loss": 0.0538,
+      "step": 2601
+    },
+    {
+      "epoch": 0.5453783273946762,
+      "grad_norm": 0.067595936357975,
+      "learning_rate": 4.513609668857162e-06,
+      "loss": 0.0541,
+      "step": 2602
+    },
+    {
+      "epoch": 0.5455879270593167,
+      "grad_norm": 0.06276869773864746,
+      "learning_rate": 4.510231033833938e-06,
+      "loss": 0.0549,
+      "step": 2603
+    },
+    {
+      "epoch": 0.5457975267239572,
+      "grad_norm": 0.058919332921504974,
+      "learning_rate": 4.506852624594107e-06,
+      "loss": 0.0553,
+      "step": 2604
+    },
+    {
+      "epoch": 0.5460071263885978,
+      "grad_norm": 0.06877582520246506,
+      "learning_rate": 4.503474442695115e-06,
+      "loss": 0.0558,
+      "step": 2605
+    },
+    {
+      "epoch": 0.5462167260532383,
+      "grad_norm": 0.07054632157087326,
+      "learning_rate": 4.500096489694299e-06,
+      "loss": 0.0591,
+      "step": 2606
+    },
+    {
+      "epoch": 0.5464263257178789,
+      "grad_norm": 0.05597640573978424,
+      "learning_rate": 4.496718767148898e-06,
+      "loss": 0.0552,
+      "step": 2607
+    },
+    {
+      "epoch": 0.5466359253825194,
+      "grad_norm": 0.052397601306438446,
+      "learning_rate": 4.493341276616044e-06,
+      "loss": 0.0543,
+      "step": 2608
+    },
+    {
+      "epoch": 0.5468455250471599,
+      "grad_norm": 0.06026868522167206,
+      "learning_rate": 4.489964019652752e-06,
+      "loss": 0.0578,
+      "step": 2609
+    },
+    {
+      "epoch": 0.5470551247118005,
+      "grad_norm": 0.05290864408016205,
+      "learning_rate": 4.486586997815942e-06,
+      "loss": 0.0533,
+      "step": 2610
+    },
+    {
+      "epoch": 0.547264724376441,
+      "grad_norm": 0.05662361532449722,
+      "learning_rate": 4.48321021266242e-06,
+      "loss": 0.0552,
+      "step": 2611
+    },
+    {
+      "epoch": 0.5474743240410815,
+      "grad_norm": 0.057523008435964584,
+      "learning_rate": 4.479833665748884e-06,
+      "loss": 0.0595,
+      "step": 2612
+    },
+    {
+      "epoch": 0.547683923705722,
+      "grad_norm": 0.054744523018598557,
+      "learning_rate": 4.476457358631918e-06,
+      "loss": 0.0589,
+      "step": 2613
+    },
+    {
+      "epoch": 0.5478935233703626,
+      "grad_norm": 0.06096423789858818,
+      "learning_rate": 4.473081292868001e-06,
+      "loss": 0.054,
+      "step": 2614
+    },
+    {
+      "epoch": 0.5481031230350032,
+      "grad_norm": 0.06122094392776489,
+      "learning_rate": 4.4697054700135e-06,
+      "loss": 0.0524,
+      "step": 2615
+    },
+    {
+      "epoch": 0.5483127226996437,
+      "grad_norm": 0.06579925119876862,
+      "learning_rate": 4.4663298916246665e-06,
+      "loss": 0.0554,
+      "step": 2616
+    },
+    {
+      "epoch": 0.5485223223642842,
+      "grad_norm": 0.0574062280356884,
+      "learning_rate": 4.4629545592576415e-06,
+      "loss": 0.0563,
+      "step": 2617
+    },
+    {
+      "epoch": 0.5487319220289247,
+      "grad_norm": 0.05878579616546631,
+      "learning_rate": 4.459579474468455e-06,
+      "loss": 0.0541,
+      "step": 2618
+    },
+    {
+      "epoch": 0.5489415216935652,
+      "grad_norm": 0.06138122081756592,
+      "learning_rate": 4.456204638813017e-06,
+      "loss": 0.0564,
+      "step": 2619
+    },
+    {
+      "epoch": 0.5491511213582059,
+      "grad_norm": 0.04858166351914406,
+      "learning_rate": 4.452830053847127e-06,
+      "loss": 0.0547,
+      "step": 2620
+    },
+    {
+      "epoch": 0.5493607210228464,
+      "grad_norm": 0.046725232154130936,
+      "learning_rate": 4.4494557211264715e-06,
+      "loss": 0.0551,
+      "step": 2621
+    },
+    {
+      "epoch": 0.5495703206874869,
+      "grad_norm": 0.05486287176609039,
+      "learning_rate": 4.446081642206611e-06,
+      "loss": 0.0574,
+      "step": 2622
+    },
+    {
+      "epoch": 0.5497799203521274,
+      "grad_norm": 0.052558574825525284,
+      "learning_rate": 4.442707818642999e-06,
+      "loss": 0.0535,
+      "step": 2623
+    },
+    {
+      "epoch": 0.5499895200167679,
+      "grad_norm": 0.04250088334083557,
+      "learning_rate": 4.439334251990966e-06,
+      "loss": 0.0561,
+      "step": 2624
+    },
+    {
+      "epoch": 0.5501991196814086,
+      "grad_norm": 0.047486599534749985,
+      "learning_rate": 4.435960943805729e-06,
+      "loss": 0.0572,
+      "step": 2625
+    },
+    {
+      "epoch": 0.5504087193460491,
+      "grad_norm": 0.04519723728299141,
+      "learning_rate": 4.432587895642378e-06,
+      "loss": 0.0536,
+      "step": 2626
+    },
+    {
+      "epoch": 0.5506183190106896,
+      "grad_norm": 0.042785726487636566,
+      "learning_rate": 4.4292151090558884e-06,
+      "loss": 0.0544,
+      "step": 2627
+    },
+    {
+      "epoch": 0.5508279186753301,
+      "grad_norm": 0.04462413862347603,
+      "learning_rate": 4.425842585601117e-06,
+      "loss": 0.0569,
+      "step": 2628
+    },
+    {
+      "epoch": 0.5510375183399706,
+      "grad_norm": 0.041346605867147446,
+      "learning_rate": 4.422470326832794e-06,
+      "loss": 0.0567,
+      "step": 2629
+    },
+    {
+      "epoch": 0.5512471180046112,
+      "grad_norm": 0.04267747327685356,
+      "learning_rate": 4.419098334305529e-06,
+      "loss": 0.0542,
+      "step": 2630
+    },
+    {
+      "epoch": 0.5514567176692518,
+      "grad_norm": 0.05388062447309494,
+      "learning_rate": 4.4157266095738125e-06,
+      "loss": 0.0559,
+      "step": 2631
+    },
+    {
+      "epoch": 0.5516663173338923,
+      "grad_norm": 0.064181849360466,
+      "learning_rate": 4.412355154192007e-06,
+      "loss": 0.0576,
+      "step": 2632
+    },
+    {
+      "epoch": 0.5518759169985328,
+      "grad_norm": 0.0693066269159317,
+      "learning_rate": 4.408983969714353e-06,
+      "loss": 0.056,
+      "step": 2633
+    },
+    {
+      "epoch": 0.5520855166631733,
+      "grad_norm": 0.07318708300590515,
+      "learning_rate": 4.405613057694967e-06,
+      "loss": 0.054,
+      "step": 2634
+    },
+    {
+      "epoch": 0.5522951163278139,
+      "grad_norm": 0.07104630768299103,
+      "learning_rate": 4.402242419687839e-06,
+      "loss": 0.0536,
+      "step": 2635
+    },
+    {
+      "epoch": 0.5525047159924544,
+      "grad_norm": 0.057221170514822006,
+      "learning_rate": 4.398872057246829e-06,
+      "loss": 0.0522,
+      "step": 2636
+    },
+    {
+      "epoch": 0.552714315657095,
+      "grad_norm": 0.05032278224825859,
+      "learning_rate": 4.395501971925677e-06,
+      "loss": 0.0552,
+      "step": 2637
+    },
+    {
+      "epoch": 0.5529239153217355,
+      "grad_norm": 0.050015486776828766,
+      "learning_rate": 4.392132165277991e-06,
+      "loss": 0.0571,
+      "step": 2638
+    },
+    {
+      "epoch": 0.553133514986376,
+      "grad_norm": 0.05708552524447441,
+      "learning_rate": 4.388762638857249e-06,
+      "loss": 0.054,
+      "step": 2639
+    },
+    {
+      "epoch": 0.5533431146510166,
+      "grad_norm": 0.06874597072601318,
+      "learning_rate": 4.385393394216804e-06,
+      "loss": 0.0553,
+      "step": 2640
+    },
+    {
+      "epoch": 0.5535527143156571,
+      "grad_norm": 0.059858787804841995,
+      "learning_rate": 4.382024432909878e-06,
+      "loss": 0.0597,
+      "step": 2641
+    },
+    {
+      "epoch": 0.5537623139802976,
+      "grad_norm": 0.04494311660528183,
+      "learning_rate": 4.378655756489558e-06,
+      "loss": 0.0579,
+      "step": 2642
+    },
+    {
+      "epoch": 0.5539719136449381,
+      "grad_norm": 0.062470775097608566,
+      "learning_rate": 4.3752873665088055e-06,
+      "loss": 0.0557,
+      "step": 2643
+    },
+    {
+      "epoch": 0.5541815133095787,
+      "grad_norm": 0.07735218852758408,
+      "learning_rate": 4.371919264520449e-06,
+      "loss": 0.0562,
+      "step": 2644
+    },
+    {
+      "epoch": 0.5543911129742193,
+      "grad_norm": 0.07211725413799286,
+      "learning_rate": 4.368551452077179e-06,
+      "loss": 0.056,
+      "step": 2645
+    },
+    {
+      "epoch": 0.5546007126388598,
+      "grad_norm": 0.0716887041926384,
+      "learning_rate": 4.365183930731559e-06,
+      "loss": 0.0586,
+      "step": 2646
+    },
+    {
+      "epoch": 0.5548103123035003,
+      "grad_norm": 0.06757423281669617,
+      "learning_rate": 4.361816702036015e-06,
+      "loss": 0.0548,
+      "step": 2647
+    },
+    {
+      "epoch": 0.5550199119681408,
+      "grad_norm": 0.05915999040007591,
+      "learning_rate": 4.35844976754284e-06,
+      "loss": 0.0561,
+      "step": 2648
+    },
+    {
+      "epoch": 0.5552295116327813,
+      "grad_norm": 0.06384500861167908,
+      "learning_rate": 4.355083128804188e-06,
+      "loss": 0.0545,
+      "step": 2649
+    },
+    {
+      "epoch": 0.555439111297422,
+      "grad_norm": 0.08520271629095078,
+      "learning_rate": 4.351716787372079e-06,
+      "loss": 0.0544,
+      "step": 2650
+    },
+    {
+      "epoch": 0.5556487109620625,
+      "grad_norm": 0.09145956486463547,
+      "learning_rate": 4.348350744798399e-06,
+      "loss": 0.0539,
+      "step": 2651
+    },
+    {
+      "epoch": 0.555858310626703,
+      "grad_norm": 0.08064839243888855,
+      "learning_rate": 4.344985002634888e-06,
+      "loss": 0.0575,
+      "step": 2652
+    },
+    {
+      "epoch": 0.5560679102913435,
+      "grad_norm": 0.055196382105350494,
+      "learning_rate": 4.341619562433154e-06,
+      "loss": 0.0543,
+      "step": 2653
+    },
+    {
+      "epoch": 0.556277509955984,
+      "grad_norm": 0.05988939851522446,
+      "learning_rate": 4.338254425744669e-06,
+      "loss": 0.0568,
+      "step": 2654
+    },
+    {
+      "epoch": 0.5564871096206246,
+      "grad_norm": 0.06533896923065186,
+      "learning_rate": 4.334889594120751e-06,
+      "loss": 0.0567,
+      "step": 2655
+    },
+    {
+      "epoch": 0.5566967092852652,
+      "grad_norm": 0.0704990103840828,
+      "learning_rate": 4.331525069112595e-06,
+      "loss": 0.0541,
+      "step": 2656
+    },
+    {
+      "epoch": 0.5569063089499057,
+      "grad_norm": 0.07413534820079803,
+      "learning_rate": 4.328160852271241e-06,
+      "loss": 0.0561,
+      "step": 2657
+    },
+    {
+      "epoch": 0.5571159086145462,
+      "grad_norm": 0.06026620417833328,
+      "learning_rate": 4.324796945147598e-06,
+      "loss": 0.0552,
+      "step": 2658
+    },
+    {
+      "epoch": 0.5573255082791867,
+      "grad_norm": 0.05583348870277405,
+      "learning_rate": 4.321433349292422e-06,
+      "loss": 0.0569,
+      "step": 2659
+    },
+    {
+      "epoch": 0.5575351079438273,
+      "grad_norm": 0.06051694229245186,
+      "learning_rate": 4.318070066256328e-06,
+      "loss": 0.061,
+      "step": 2660
+    },
+    {
+      "epoch": 0.5577447076084678,
+      "grad_norm": 0.06352775543928146,
+      "learning_rate": 4.314707097589796e-06,
+      "loss": 0.0561,
+      "step": 2661
+    },
+    {
+      "epoch": 0.5579543072731084,
+      "grad_norm": 0.06398675590753555,
+      "learning_rate": 4.311344444843147e-06,
+      "loss": 0.0535,
+      "step": 2662
+    },
+    {
+      "epoch": 0.5581639069377489,
+      "grad_norm": 0.06229983642697334,
+      "learning_rate": 4.307982109566566e-06,
+      "loss": 0.0519,
+      "step": 2663
+    },
+    {
+      "epoch": 0.5583735066023895,
+      "grad_norm": 0.05653541907668114,
+      "learning_rate": 4.3046200933100905e-06,
+      "loss": 0.0557,
+      "step": 2664
+    },
+    {
+      "epoch": 0.55858310626703,
+      "grad_norm": 0.047429151833057404,
+      "learning_rate": 4.301258397623606e-06,
+      "loss": 0.0534,
+      "step": 2665
+    },
+    {
+      "epoch": 0.5587927059316705,
+      "grad_norm": 0.05258271098136902,
+      "learning_rate": 4.2978970240568556e-06,
+      "loss": 0.0542,
+      "step": 2666
+    },
+    {
+      "epoch": 0.559002305596311,
+      "grad_norm": 0.05701644718647003,
+      "learning_rate": 4.2945359741594315e-06,
+      "loss": 0.0556,
+      "step": 2667
+    },
+    {
+      "epoch": 0.5592119052609515,
+      "grad_norm": 0.06717483699321747,
+      "learning_rate": 4.291175249480777e-06,
+      "loss": 0.054,
+      "step": 2668
+    },
+    {
+      "epoch": 0.5594215049255922,
+      "grad_norm": 0.06235239654779434,
+      "learning_rate": 4.287814851570183e-06,
+      "loss": 0.0587,
+      "step": 2669
+    },
+    {
+      "epoch": 0.5596311045902327,
+      "grad_norm": 0.06835221499204636,
+      "learning_rate": 4.284454781976796e-06,
+      "loss": 0.0562,
+      "step": 2670
+    },
+    {
+      "epoch": 0.5598407042548732,
+      "grad_norm": 0.07527367770671844,
+      "learning_rate": 4.281095042249608e-06,
+      "loss": 0.0565,
+      "step": 2671
+    },
+    {
+      "epoch": 0.5600503039195137,
+      "grad_norm": 0.062950000166893,
+      "learning_rate": 4.2777356339374526e-06,
+      "loss": 0.0572,
+      "step": 2672
+    },
+    {
+      "epoch": 0.5602599035841542,
+      "grad_norm": 0.06541959196329117,
+      "learning_rate": 4.274376558589022e-06,
+      "loss": 0.0521,
+      "step": 2673
+    },
+    {
+      "epoch": 0.5604695032487949,
+      "grad_norm": 0.08955197781324387,
+      "learning_rate": 4.271017817752847e-06,
+      "loss": 0.053,
+      "step": 2674
+    },
+    {
+      "epoch": 0.5606791029134354,
+      "grad_norm": 0.07426553964614868,
+      "learning_rate": 4.267659412977306e-06,
+      "loss": 0.0531,
+      "step": 2675
+    },
+    {
+      "epoch": 0.5608887025780759,
+      "grad_norm": 0.061013299971818924,
+      "learning_rate": 4.264301345810623e-06,
+      "loss": 0.0548,
+      "step": 2676
+    },
+    {
+      "epoch": 0.5610983022427164,
+      "grad_norm": 0.06869885325431824,
+      "learning_rate": 4.260943617800869e-06,
+      "loss": 0.0564,
+      "step": 2677
+    },
+    {
+      "epoch": 0.5613079019073569,
+      "grad_norm": 0.060993682593107224,
+      "learning_rate": 4.257586230495951e-06,
+      "loss": 0.0556,
+      "step": 2678
+    },
+    {
+      "epoch": 0.5615175015719975,
+      "grad_norm": 0.08102905750274658,
+      "learning_rate": 4.254229185443628e-06,
+      "loss": 0.0537,
+      "step": 2679
+    },
+    {
+      "epoch": 0.561727101236638,
+      "grad_norm": 0.0864529013633728,
+      "learning_rate": 4.250872484191495e-06,
+      "loss": 0.0589,
+      "step": 2680
+    },
+    {
+      "epoch": 0.5619367009012786,
+      "grad_norm": 0.07401063293218613,
+      "learning_rate": 4.247516128286992e-06,
+      "loss": 0.055,
+      "step": 2681
+    },
+    {
+      "epoch": 0.5621463005659191,
+      "grad_norm": 0.06574457883834839,
+      "learning_rate": 4.244160119277397e-06,
+      "loss": 0.059,
+      "step": 2682
+    },
+    {
+      "epoch": 0.5623559002305596,
+      "grad_norm": 0.05600839853286743,
+      "learning_rate": 4.24080445870983e-06,
+      "loss": 0.0577,
+      "step": 2683
+    },
+    {
+      "epoch": 0.5625654998952002,
+      "grad_norm": 0.058064643293619156,
+      "learning_rate": 4.2374491481312506e-06,
+      "loss": 0.0574,
+      "step": 2684
+    },
+    {
+      "epoch": 0.5627750995598407,
+      "grad_norm": 0.05819283798336983,
+      "learning_rate": 4.234094189088455e-06,
+      "loss": 0.0563,
+      "step": 2685
+    },
+    {
+      "epoch": 0.5629846992244812,
+      "grad_norm": 0.0627334788441658,
+      "learning_rate": 4.230739583128078e-06,
+      "loss": 0.0543,
+      "step": 2686
+    },
+    {
+      "epoch": 0.5631942988891218,
+      "grad_norm": 0.07298675179481506,
+      "learning_rate": 4.227385331796596e-06,
+      "loss": 0.0586,
+      "step": 2687
+    },
+    {
+      "epoch": 0.5634038985537623,
+      "grad_norm": 0.07162720710039139,
+      "learning_rate": 4.2240314366403135e-06,
+      "loss": 0.0575,
+      "step": 2688
+    },
+    {
+      "epoch": 0.5636134982184029,
+      "grad_norm": 0.06274513155221939,
+      "learning_rate": 4.220677899205376e-06,
+      "loss": 0.0544,
+      "step": 2689
+    },
+    {
+      "epoch": 0.5638230978830434,
+      "grad_norm": 0.049171797931194305,
+      "learning_rate": 4.21732472103777e-06,
+      "loss": 0.0564,
+      "step": 2690
+    },
+    {
+      "epoch": 0.5640326975476839,
+      "grad_norm": 0.04283327981829643,
+      "learning_rate": 4.213971903683301e-06,
+      "loss": 0.0541,
+      "step": 2691
+    },
+    {
+      "epoch": 0.5642422972123244,
+      "grad_norm": 0.05370490252971649,
+      "learning_rate": 4.210619448687622e-06,
+      "loss": 0.0539,
+      "step": 2692
+    },
+    {
+      "epoch": 0.564451896876965,
+      "grad_norm": 0.05934003368020058,
+      "learning_rate": 4.2072673575962125e-06,
+      "loss": 0.0545,
+      "step": 2693
+    },
+    {
+      "epoch": 0.5646614965416056,
+      "grad_norm": 0.06390709429979324,
+      "learning_rate": 4.203915631954389e-06,
+      "loss": 0.0548,
+      "step": 2694
+    },
+    {
+      "epoch": 0.5648710962062461,
+      "grad_norm": 0.0646679550409317,
+      "learning_rate": 4.200564273307292e-06,
+      "loss": 0.0554,
+      "step": 2695
+    },
+    {
+      "epoch": 0.5650806958708866,
+      "grad_norm": 0.05779599770903587,
+      "learning_rate": 4.197213283199898e-06,
+      "loss": 0.0591,
+      "step": 2696
+    },
+    {
+      "epoch": 0.5652902955355271,
+      "grad_norm": 0.06428996473550797,
+      "learning_rate": 4.193862663177016e-06,
+      "loss": 0.0544,
+      "step": 2697
+    },
+    {
+      "epoch": 0.5654998952001676,
+      "grad_norm": 0.07036624848842621,
+      "learning_rate": 4.190512414783278e-06,
+      "loss": 0.0551,
+      "step": 2698
+    },
+    {
+      "epoch": 0.5657094948648083,
+      "grad_norm": 0.07967984676361084,
+      "learning_rate": 4.187162539563147e-06,
+      "loss": 0.0584,
+      "step": 2699
+    },
+    {
+      "epoch": 0.5659190945294488,
+      "grad_norm": 0.09749884158372879,
+      "learning_rate": 4.183813039060919e-06,
+      "loss": 0.0543,
+      "step": 2700
+    },
+    {
+      "epoch": 0.5661286941940893,
+      "grad_norm": 0.09949739277362823,
+      "learning_rate": 4.180463914820709e-06,
+      "loss": 0.0562,
+      "step": 2701
+    },
+    {
+      "epoch": 0.5663382938587298,
+      "grad_norm": 0.09305305033922195,
+      "learning_rate": 4.177115168386463e-06,
+      "loss": 0.055,
+      "step": 2702
+    },
+    {
+      "epoch": 0.5665478935233703,
+      "grad_norm": 0.07777238637208939,
+      "learning_rate": 4.173766801301956e-06,
+      "loss": 0.0574,
+      "step": 2703
+    },
+    {
+      "epoch": 0.5667574931880109,
+      "grad_norm": 0.06988883018493652,
+      "learning_rate": 4.17041881511078e-06,
+      "loss": 0.0547,
+      "step": 2704
+    },
+    {
+      "epoch": 0.5669670928526515,
+      "grad_norm": 0.07308480888605118,
+      "learning_rate": 4.167071211356358e-06,
+      "loss": 0.0539,
+      "step": 2705
+    },
+    {
+      "epoch": 0.567176692517292,
+      "grad_norm": 0.053045183420181274,
+      "learning_rate": 4.163723991581935e-06,
+      "loss": 0.057,
+      "step": 2706
+    },
+    {
+      "epoch": 0.5673862921819325,
+      "grad_norm": 0.05672500282526016,
+      "learning_rate": 4.160377157330579e-06,
+      "loss": 0.0545,
+      "step": 2707
+    },
+    {
+      "epoch": 0.567595891846573,
+      "grad_norm": 0.06733475625514984,
+      "learning_rate": 4.157030710145178e-06,
+      "loss": 0.0554,
+      "step": 2708
+    },
+    {
+      "epoch": 0.5678054915112136,
+      "grad_norm": 0.07970572263002396,
+      "learning_rate": 4.153684651568445e-06,
+      "loss": 0.0561,
+      "step": 2709
+    },
+    {
+      "epoch": 0.5680150911758541,
+      "grad_norm": 0.1028270497918129,
+      "learning_rate": 4.150338983142913e-06,
+      "loss": 0.0553,
+      "step": 2710
+    },
+    {
+      "epoch": 0.5682246908404947,
+      "grad_norm": 0.10650909692049026,
+      "learning_rate": 4.1469937064109305e-06,
+      "loss": 0.0545,
+      "step": 2711
+    },
+    {
+      "epoch": 0.5684342905051352,
+      "grad_norm": 0.09105932712554932,
+      "learning_rate": 4.1436488229146735e-06,
+      "loss": 0.0579,
+      "step": 2712
+    },
+    {
+      "epoch": 0.5686438901697757,
+      "grad_norm": 0.0559769943356514,
+      "learning_rate": 4.140304334196133e-06,
+      "loss": 0.0559,
+      "step": 2713
+    },
+    {
+      "epoch": 0.5688534898344163,
+      "grad_norm": 0.056230757385492325,
+      "learning_rate": 4.136960241797113e-06,
+      "loss": 0.0553,
+      "step": 2714
+    },
+    {
+      "epoch": 0.5690630894990568,
+      "grad_norm": 0.07475091516971588,
+      "learning_rate": 4.1336165472592434e-06,
+      "loss": 0.0565,
+      "step": 2715
+    },
+    {
+      "epoch": 0.5692726891636973,
+      "grad_norm": 0.07072797417640686,
+      "learning_rate": 4.130273252123965e-06,
+      "loss": 0.0563,
+      "step": 2716
+    },
+    {
+      "epoch": 0.5694822888283378,
+      "grad_norm": 0.05199355632066727,
+      "learning_rate": 4.1269303579325385e-06,
+      "loss": 0.0579,
+      "step": 2717
+    },
+    {
+      "epoch": 0.5696918884929784,
+      "grad_norm": 0.047952428460121155,
+      "learning_rate": 4.123587866226035e-06,
+      "loss": 0.0544,
+      "step": 2718
+    },
+    {
+      "epoch": 0.569901488157619,
+      "grad_norm": 0.059139080345630646,
+      "learning_rate": 4.120245778545341e-06,
+      "loss": 0.0557,
+      "step": 2719
+    },
+    {
+      "epoch": 0.5701110878222595,
+      "grad_norm": 0.08245202153921127,
+      "learning_rate": 4.116904096431163e-06,
+      "loss": 0.0564,
+      "step": 2720
+    },
+    {
+      "epoch": 0.5703206874869,
+      "grad_norm": 0.09880422055721283,
+      "learning_rate": 4.113562821424012e-06,
+      "loss": 0.0558,
+      "step": 2721
+    },
+    {
+      "epoch": 0.5705302871515405,
+      "grad_norm": 0.09177903831005096,
+      "learning_rate": 4.1102219550642154e-06,
+      "loss": 0.0563,
+      "step": 2722
+    },
+    {
+      "epoch": 0.570739886816181,
+      "grad_norm": 0.06406284123659134,
+      "learning_rate": 4.1068814988919156e-06,
+      "loss": 0.0579,
+      "step": 2723
+    },
+    {
+      "epoch": 0.5709494864808217,
+      "grad_norm": 0.055943313986063004,
+      "learning_rate": 4.103541454447057e-06,
+      "loss": 0.0544,
+      "step": 2724
+    },
+    {
+      "epoch": 0.5711590861454622,
+      "grad_norm": 0.07301504909992218,
+      "learning_rate": 4.100201823269401e-06,
+      "loss": 0.054,
+      "step": 2725
+    },
+    {
+      "epoch": 0.5713686858101027,
+      "grad_norm": 0.08340940624475479,
+      "learning_rate": 4.0968626068985205e-06,
+      "loss": 0.0556,
+      "step": 2726
+    },
+    {
+      "epoch": 0.5715782854747432,
+      "grad_norm": 0.08021462708711624,
+      "learning_rate": 4.093523806873787e-06,
+      "loss": 0.0534,
+      "step": 2727
+    },
+    {
+      "epoch": 0.5717878851393837,
+      "grad_norm": 0.07112640142440796,
+      "learning_rate": 4.090185424734392e-06,
+      "loss": 0.0529,
+      "step": 2728
+    },
+    {
+      "epoch": 0.5719974848040243,
+      "grad_norm": 0.06481733918190002,
+      "learning_rate": 4.086847462019326e-06,
+      "loss": 0.0575,
+      "step": 2729
+    },
+    {
+      "epoch": 0.5722070844686649,
+      "grad_norm": 0.06660385429859161,
+      "learning_rate": 4.0835099202673926e-06,
+      "loss": 0.0535,
+      "step": 2730
+    },
+    {
+      "epoch": 0.5724166841333054,
+      "grad_norm": 0.07477893680334091,
+      "learning_rate": 4.080172801017195e-06,
+      "loss": 0.0567,
+      "step": 2731
+    },
+    {
+      "epoch": 0.5726262837979459,
+      "grad_norm": 0.07714486122131348,
+      "learning_rate": 4.076836105807143e-06,
+      "loss": 0.0576,
+      "step": 2732
+    },
+    {
+      "epoch": 0.5728358834625865,
+      "grad_norm": 0.05599776282906532,
+      "learning_rate": 4.073499836175457e-06,
+      "loss": 0.0559,
+      "step": 2733
+    },
+    {
+      "epoch": 0.573045483127227,
+      "grad_norm": 0.03909073770046234,
+      "learning_rate": 4.0701639936601535e-06,
+      "loss": 0.0551,
+      "step": 2734
+    },
+    {
+      "epoch": 0.5732550827918675,
+      "grad_norm": 0.06613650918006897,
+      "learning_rate": 4.066828579799054e-06,
+      "loss": 0.0566,
+      "step": 2735
+    },
+    {
+      "epoch": 0.5734646824565081,
+      "grad_norm": 0.0671084001660347,
+      "learning_rate": 4.063493596129788e-06,
+      "loss": 0.0562,
+      "step": 2736
+    },
+    {
+      "epoch": 0.5736742821211486,
+      "grad_norm": 0.058822277933359146,
+      "learning_rate": 4.060159044189778e-06,
+      "loss": 0.054,
+      "step": 2737
+    },
+    {
+      "epoch": 0.5738838817857892,
+      "grad_norm": 0.05256880447268486,
+      "learning_rate": 4.0568249255162526e-06,
+      "loss": 0.0573,
+      "step": 2738
+    },
+    {
+      "epoch": 0.5740934814504297,
+      "grad_norm": 0.05657995492219925,
+      "learning_rate": 4.053491241646242e-06,
+      "loss": 0.0553,
+      "step": 2739
+    },
+    {
+      "epoch": 0.5743030811150702,
+      "grad_norm": 0.05349545180797577,
+      "learning_rate": 4.050157994116573e-06,
+      "loss": 0.0562,
+      "step": 2740
+    },
+    {
+      "epoch": 0.5745126807797107,
+      "grad_norm": 0.04297221451997757,
+      "learning_rate": 4.04682518446387e-06,
+      "loss": 0.0556,
+      "step": 2741
+    },
+    {
+      "epoch": 0.5747222804443513,
+      "grad_norm": 0.04681278392672539,
+      "learning_rate": 4.043492814224559e-06,
+      "loss": 0.0543,
+      "step": 2742
+    },
+    {
+      "epoch": 0.5749318801089919,
+      "grad_norm": 0.06194974109530449,
+      "learning_rate": 4.040160884934864e-06,
+      "loss": 0.0546,
+      "step": 2743
+    },
+    {
+      "epoch": 0.5751414797736324,
+      "grad_norm": 0.08011812716722488,
+      "learning_rate": 4.036829398130799e-06,
+      "loss": 0.0552,
+      "step": 2744
+    },
+    {
+      "epoch": 0.5753510794382729,
+      "grad_norm": 0.08187726140022278,
+      "learning_rate": 4.033498355348183e-06,
+      "loss": 0.0533,
+      "step": 2745
+    },
+    {
+      "epoch": 0.5755606791029134,
+      "grad_norm": 0.06547623127698898,
+      "learning_rate": 4.030167758122625e-06,
+      "loss": 0.0533,
+      "step": 2746
+    },
+    {
+      "epoch": 0.5757702787675539,
+      "grad_norm": 0.0580391101539135,
+      "learning_rate": 4.026837607989527e-06,
+      "loss": 0.055,
+      "step": 2747
+    },
+    {
+      "epoch": 0.5759798784321946,
+      "grad_norm": 0.06372876465320587,
+      "learning_rate": 4.0235079064840905e-06,
+      "loss": 0.0547,
+      "step": 2748
+    },
+    {
+      "epoch": 0.5761894780968351,
+      "grad_norm": 0.0708019807934761,
+      "learning_rate": 4.020178655141307e-06,
+      "loss": 0.0543,
+      "step": 2749
+    },
+    {
+      "epoch": 0.5763990777614756,
+      "grad_norm": 0.06585292518138885,
+      "learning_rate": 4.016849855495959e-06,
+      "loss": 0.0568,
+      "step": 2750
+    },
+    {
+      "epoch": 0.5766086774261161,
+      "grad_norm": 0.05741509795188904,
+      "learning_rate": 4.013521509082624e-06,
+      "loss": 0.0556,
+      "step": 2751
+    },
+    {
+      "epoch": 0.5768182770907566,
+      "grad_norm": 0.05286962538957596,
+      "learning_rate": 4.0101936174356665e-06,
+      "loss": 0.0564,
+      "step": 2752
+    },
+    {
+      "epoch": 0.5770278767553972,
+      "grad_norm": 0.04501480236649513,
+      "learning_rate": 4.0068661820892485e-06,
+      "loss": 0.0571,
+      "step": 2753
+    },
+    {
+      "epoch": 0.5772374764200378,
+      "grad_norm": 0.045843351632356644,
+      "learning_rate": 4.003539204577313e-06,
+      "loss": 0.0557,
+      "step": 2754
+    },
+    {
+      "epoch": 0.5774470760846783,
+      "grad_norm": 0.062015220522880554,
+      "learning_rate": 4.000212686433597e-06,
+      "loss": 0.053,
+      "step": 2755
+    },
+    {
+      "epoch": 0.5776566757493188,
+      "grad_norm": 0.06968379020690918,
+      "learning_rate": 3.9968866291916254e-06,
+      "loss": 0.0549,
+      "step": 2756
+    },
+    {
+      "epoch": 0.5778662754139593,
+      "grad_norm": 0.06381241232156754,
+      "learning_rate": 3.99356103438471e-06,
+      "loss": 0.0555,
+      "step": 2757
+    },
+    {
+      "epoch": 0.5780758750785999,
+      "grad_norm": 0.051279522478580475,
+      "learning_rate": 3.990235903545947e-06,
+      "loss": 0.0565,
+      "step": 2758
+    },
+    {
+      "epoch": 0.5782854747432404,
+      "grad_norm": 0.04932436719536781,
+      "learning_rate": 3.9869112382082255e-06,
+      "loss": 0.0562,
+      "step": 2759
+    },
+    {
+      "epoch": 0.578495074407881,
+      "grad_norm": 0.05086817964911461,
+      "learning_rate": 3.98358703990421e-06,
+      "loss": 0.0573,
+      "step": 2760
+    },
+    {
+      "epoch": 0.5787046740725215,
+      "grad_norm": 0.03960997983813286,
+      "learning_rate": 3.980263310166359e-06,
+      "loss": 0.0557,
+      "step": 2761
+    },
+    {
+      "epoch": 0.578914273737162,
+      "grad_norm": 0.04365481063723564,
+      "learning_rate": 3.976940050526909e-06,
+      "loss": 0.0548,
+      "step": 2762
+    },
+    {
+      "epoch": 0.5791238734018026,
+      "grad_norm": 0.0648135393857956,
+      "learning_rate": 3.973617262517886e-06,
+      "loss": 0.0554,
+      "step": 2763
+    },
+    {
+      "epoch": 0.5793334730664431,
+      "grad_norm": 0.0678715854883194,
+      "learning_rate": 3.970294947671089e-06,
+      "loss": 0.0566,
+      "step": 2764
+    },
+    {
+      "epoch": 0.5795430727310836,
+      "grad_norm": 0.06538737565279007,
+      "learning_rate": 3.9669731075181074e-06,
+      "loss": 0.0556,
+      "step": 2765
+    },
+    {
+      "epoch": 0.5797526723957241,
+      "grad_norm": 0.058645885437726974,
+      "learning_rate": 3.963651743590311e-06,
+      "loss": 0.0545,
+      "step": 2766
+    },
+    {
+      "epoch": 0.5799622720603647,
+      "grad_norm": 0.04971649497747421,
+      "learning_rate": 3.960330857418844e-06,
+      "loss": 0.0548,
+      "step": 2767
+    },
+    {
+      "epoch": 0.5801718717250053,
+      "grad_norm": 0.05847443640232086,
+      "learning_rate": 3.9570104505346345e-06,
+      "loss": 0.0528,
+      "step": 2768
+    },
+    {
+      "epoch": 0.5803814713896458,
+      "grad_norm": 0.0770886018872261,
+      "learning_rate": 3.953690524468393e-06,
+      "loss": 0.058,
+      "step": 2769
+    },
+    {
+      "epoch": 0.5805910710542863,
+      "grad_norm": 0.08351907879114151,
+      "learning_rate": 3.950371080750602e-06,
+      "loss": 0.0587,
+      "step": 2770
+    },
+    {
+      "epoch": 0.5808006707189268,
+      "grad_norm": 0.06995301693677902,
+      "learning_rate": 3.947052120911523e-06,
+      "loss": 0.0554,
+      "step": 2771
+    },
+    {
+      "epoch": 0.5810102703835673,
+      "grad_norm": 0.05648371949791908,
+      "learning_rate": 3.9437336464812e-06,
+      "loss": 0.0523,
+      "step": 2772
+    },
+    {
+      "epoch": 0.581219870048208,
+      "grad_norm": 0.06814776360988617,
+      "learning_rate": 3.940415658989445e-06,
+      "loss": 0.0542,
+      "step": 2773
+    },
+    {
+      "epoch": 0.5814294697128485,
+      "grad_norm": 0.07342953979969025,
+      "learning_rate": 3.93709815996585e-06,
+      "loss": 0.0562,
+      "step": 2774
+    },
+    {
+      "epoch": 0.581639069377489,
+      "grad_norm": 0.06355999410152435,
+      "learning_rate": 3.933781150939784e-06,
+      "loss": 0.0555,
+      "step": 2775
+    },
+    {
+      "epoch": 0.5818486690421295,
+      "grad_norm": 0.05689748376607895,
+      "learning_rate": 3.9304646334403875e-06,
+      "loss": 0.054,
+      "step": 2776
+    },
+    {
+      "epoch": 0.58205826870677,
+      "grad_norm": 0.06319185346364975,
+      "learning_rate": 3.927148608996569e-06,
+      "loss": 0.0562,
+      "step": 2777
+    },
+    {
+      "epoch": 0.5822678683714106,
+      "grad_norm": 0.05822164565324783,
+      "learning_rate": 3.923833079137022e-06,
+      "loss": 0.0561,
+      "step": 2778
+    },
+    {
+      "epoch": 0.5824774680360512,
+      "grad_norm": 0.058942124247550964,
+      "learning_rate": 3.920518045390201e-06,
+      "loss": 0.0585,
+      "step": 2779
+    },
+    {
+      "epoch": 0.5826870677006917,
+      "grad_norm": 0.0649891048669815,
+      "learning_rate": 3.9172035092843365e-06,
+      "loss": 0.0539,
+      "step": 2780
+    },
+    {
+      "epoch": 0.5828966673653322,
+      "grad_norm": 0.06785421073436737,
+      "learning_rate": 3.91388947234743e-06,
+      "loss": 0.0573,
+      "step": 2781
+    },
+    {
+      "epoch": 0.5831062670299727,
+      "grad_norm": 0.07181856036186218,
+      "learning_rate": 3.9105759361072516e-06,
+      "loss": 0.0548,
+      "step": 2782
+    },
+    {
+      "epoch": 0.5833158666946133,
+      "grad_norm": 0.07179451733827591,
+      "learning_rate": 3.907262902091338e-06,
+      "loss": 0.0551,
+      "step": 2783
+    },
+    {
+      "epoch": 0.5835254663592538,
+      "grad_norm": 0.0645546168088913,
+      "learning_rate": 3.903950371827001e-06,
+      "loss": 0.0542,
+      "step": 2784
+    },
+    {
+      "epoch": 0.5837350660238944,
+      "grad_norm": 0.05469752475619316,
+      "learning_rate": 3.900638346841314e-06,
+      "loss": 0.0533,
+      "step": 2785
+    },
+    {
+      "epoch": 0.5839446656885349,
+      "grad_norm": 0.04990299046039581,
+      "learning_rate": 3.897326828661123e-06,
+      "loss": 0.0538,
+      "step": 2786
+    },
+    {
+      "epoch": 0.5841542653531754,
+      "grad_norm": 0.05425114557147026,
+      "learning_rate": 3.894015818813034e-06,
+      "loss": 0.0566,
+      "step": 2787
+    },
+    {
+      "epoch": 0.584363865017816,
+      "grad_norm": 0.07104803621768951,
+      "learning_rate": 3.890705318823421e-06,
+      "loss": 0.0536,
+      "step": 2788
+    },
+    {
+      "epoch": 0.5845734646824565,
+      "grad_norm": 0.0836774930357933,
+      "learning_rate": 3.887395330218429e-06,
+      "loss": 0.055,
+      "step": 2789
+    },
+    {
+      "epoch": 0.584783064347097,
+      "grad_norm": 0.08817489445209503,
+      "learning_rate": 3.884085854523956e-06,
+      "loss": 0.0544,
+      "step": 2790
+    },
+    {
+      "epoch": 0.5849926640117376,
+      "grad_norm": 0.08014960587024689,
+      "learning_rate": 3.880776893265673e-06,
+      "loss": 0.0536,
+      "step": 2791
+    },
+    {
+      "epoch": 0.5852022636763781,
+      "grad_norm": 0.07269839197397232,
+      "learning_rate": 3.877468447969011e-06,
+      "loss": 0.0537,
+      "step": 2792
+    },
+    {
+      "epoch": 0.5854118633410187,
+      "grad_norm": 0.06357665359973907,
+      "learning_rate": 3.874160520159159e-06,
+      "loss": 0.0549,
+      "step": 2793
+    },
+    {
+      "epoch": 0.5856214630056592,
+      "grad_norm": 0.051662154495716095,
+      "learning_rate": 3.8708531113610735e-06,
+      "loss": 0.0525,
+      "step": 2794
+    },
+    {
+      "epoch": 0.5858310626702997,
+      "grad_norm": 0.040864262729883194,
+      "learning_rate": 3.8675462230994725e-06,
+      "loss": 0.0547,
+      "step": 2795
+    },
+    {
+      "epoch": 0.5860406623349402,
+      "grad_norm": 0.04362620413303375,
+      "learning_rate": 3.864239856898824e-06,
+      "loss": 0.0537,
+      "step": 2796
+    },
+    {
+      "epoch": 0.5862502619995807,
+      "grad_norm": 0.047262124717235565,
+      "learning_rate": 3.860934014283366e-06,
+      "loss": 0.054,
+      "step": 2797
+    },
+    {
+      "epoch": 0.5864598616642214,
+      "grad_norm": 0.043658629059791565,
+      "learning_rate": 3.85762869677709e-06,
+      "loss": 0.0525,
+      "step": 2798
+    },
+    {
+      "epoch": 0.5866694613288619,
+      "grad_norm": 0.040076591074466705,
+      "learning_rate": 3.854323905903751e-06,
+      "loss": 0.0557,
+      "step": 2799
+    },
+    {
+      "epoch": 0.5868790609935024,
+      "grad_norm": 0.03843994066119194,
+      "learning_rate": 3.851019643186851e-06,
+      "loss": 0.0569,
+      "step": 2800
+    },
+    {
+      "epoch": 0.5870886606581429,
+      "grad_norm": 0.04533432796597481,
+      "learning_rate": 3.847715910149656e-06,
+      "loss": 0.0561,
+      "step": 2801
+    },
+    {
+      "epoch": 0.5872982603227835,
+      "grad_norm": 0.042377371340990067,
+      "learning_rate": 3.8444127083151885e-06,
+      "loss": 0.0557,
+      "step": 2802
+    },
+    {
+      "epoch": 0.587507859987424,
+      "grad_norm": 0.03570021688938141,
+      "learning_rate": 3.841110039206222e-06,
+      "loss": 0.0541,
+      "step": 2803
+    },
+    {
+      "epoch": 0.5877174596520646,
+      "grad_norm": 0.03966277092695236,
+      "learning_rate": 3.837807904345286e-06,
+      "loss": 0.0542,
+      "step": 2804
+    },
+    {
+      "epoch": 0.5879270593167051,
+      "grad_norm": 0.041379570960998535,
+      "learning_rate": 3.834506305254667e-06,
+      "loss": 0.0551,
+      "step": 2805
+    },
+    {
+      "epoch": 0.5881366589813456,
+      "grad_norm": 0.03879513218998909,
+      "learning_rate": 3.831205243456397e-06,
+      "loss": 0.056,
+      "step": 2806
+    },
+    {
+      "epoch": 0.5883462586459862,
+      "grad_norm": 0.041930824518203735,
+      "learning_rate": 3.827904720472267e-06,
+      "loss": 0.0544,
+      "step": 2807
+    },
+    {
+      "epoch": 0.5885558583106267,
+      "grad_norm": 0.04247865080833435,
+      "learning_rate": 3.824604737823819e-06,
+      "loss": 0.0552,
+      "step": 2808
+    },
+    {
+      "epoch": 0.5887654579752672,
+      "grad_norm": 0.03654211759567261,
+      "learning_rate": 3.821305297032342e-06,
+      "loss": 0.0553,
+      "step": 2809
+    },
+    {
+      "epoch": 0.5889750576399078,
+      "grad_norm": 0.04415356367826462,
+      "learning_rate": 3.818006399618877e-06,
+      "loss": 0.0576,
+      "step": 2810
+    },
+    {
+      "epoch": 0.5891846573045483,
+      "grad_norm": 0.058614566922187805,
+      "learning_rate": 3.8147080471042166e-06,
+      "loss": 0.057,
+      "step": 2811
+    },
+    {
+      "epoch": 0.5893942569691889,
+      "grad_norm": 0.05693338066339493,
+      "learning_rate": 3.811410241008902e-06,
+      "loss": 0.053,
+      "step": 2812
+    },
+    {
+      "epoch": 0.5896038566338294,
+      "grad_norm": 0.04473234340548515,
+      "learning_rate": 3.808112982853217e-06,
+      "loss": 0.0539,
+      "step": 2813
+    },
+    {
+      "epoch": 0.5898134562984699,
+      "grad_norm": 0.04146963357925415,
+      "learning_rate": 3.8048162741572008e-06,
+      "loss": 0.056,
+      "step": 2814
+    },
+    {
+      "epoch": 0.5900230559631104,
+      "grad_norm": 0.03515983745455742,
+      "learning_rate": 3.801520116440635e-06,
+      "loss": 0.0558,
+      "step": 2815
+    },
+    {
+      "epoch": 0.590232655627751,
+      "grad_norm": 0.040382951498031616,
+      "learning_rate": 3.798224511223044e-06,
+      "loss": 0.0587,
+      "step": 2816
+    },
+    {
+      "epoch": 0.5904422552923916,
+      "grad_norm": 0.042147841304540634,
+      "learning_rate": 3.794929460023705e-06,
+      "loss": 0.0541,
+      "step": 2817
+    },
+    {
+      "epoch": 0.5906518549570321,
+      "grad_norm": 0.0410742312669754,
+      "learning_rate": 3.7916349643616357e-06,
+      "loss": 0.0568,
+      "step": 2818
+    },
+    {
+      "epoch": 0.5908614546216726,
+      "grad_norm": 0.03527839481830597,
+      "learning_rate": 3.788341025755595e-06,
+      "loss": 0.0572,
+      "step": 2819
+    },
+    {
+      "epoch": 0.5910710542863131,
+      "grad_norm": 0.04356502741575241,
+      "learning_rate": 3.7850476457240905e-06,
+      "loss": 0.0555,
+      "step": 2820
+    },
+    {
+      "epoch": 0.5912806539509536,
+      "grad_norm": 0.053171779960393906,
+      "learning_rate": 3.781754825785368e-06,
+      "loss": 0.0572,
+      "step": 2821
+    },
+    {
+      "epoch": 0.5914902536155943,
+      "grad_norm": 0.06760048121213913,
+      "learning_rate": 3.77846256745742e-06,
+      "loss": 0.0539,
+      "step": 2822
+    },
+    {
+      "epoch": 0.5916998532802348,
+      "grad_norm": 0.06500531733036041,
+      "learning_rate": 3.7751708722579733e-06,
+      "loss": 0.0529,
+      "step": 2823
+    },
+    {
+      "epoch": 0.5919094529448753,
+      "grad_norm": 0.0501357838511467,
+      "learning_rate": 3.771879741704499e-06,
+      "loss": 0.053,
+      "step": 2824
+    },
+    {
+      "epoch": 0.5921190526095158,
+      "grad_norm": 0.05668150261044502,
+      "learning_rate": 3.768589177314211e-06,
+      "loss": 0.0561,
+      "step": 2825
+    },
+    {
+      "epoch": 0.5923286522741563,
+      "grad_norm": 0.05575096979737282,
+      "learning_rate": 3.765299180604055e-06,
+      "loss": 0.0568,
+      "step": 2826
+    },
+    {
+      "epoch": 0.592538251938797,
+      "grad_norm": 0.05972772836685181,
+      "learning_rate": 3.7620097530907196e-06,
+      "loss": 0.0563,
+      "step": 2827
+    },
+    {
+      "epoch": 0.5927478516034375,
+      "grad_norm": 0.05793767794966698,
+      "learning_rate": 3.758720896290634e-06,
+      "loss": 0.0539,
+      "step": 2828
+    },
+    {
+      "epoch": 0.592957451268078,
+      "grad_norm": 0.05584365129470825,
+      "learning_rate": 3.755432611719954e-06,
+      "loss": 0.0543,
+      "step": 2829
+    },
+    {
+      "epoch": 0.5931670509327185,
+      "grad_norm": 0.06453042477369308,
+      "learning_rate": 3.752144900894582e-06,
+      "loss": 0.054,
+      "step": 2830
+    },
+    {
+      "epoch": 0.593376650597359,
+      "grad_norm": 0.05443593114614487,
+      "learning_rate": 3.7488577653301538e-06,
+      "loss": 0.0543,
+      "step": 2831
+    },
+    {
+      "epoch": 0.5935862502619996,
+      "grad_norm": 0.05260578915476799,
+      "learning_rate": 3.7455712065420335e-06,
+      "loss": 0.0575,
+      "step": 2832
+    },
+    {
+      "epoch": 0.5937958499266401,
+      "grad_norm": 0.057504259049892426,
+      "learning_rate": 3.7422852260453274e-06,
+      "loss": 0.0539,
+      "step": 2833
+    },
+    {
+      "epoch": 0.5940054495912807,
+      "grad_norm": 0.05131387338042259,
+      "learning_rate": 3.7389998253548698e-06,
+      "loss": 0.055,
+      "step": 2834
+    },
+    {
+      "epoch": 0.5942150492559212,
+      "grad_norm": 0.04155094549059868,
+      "learning_rate": 3.7357150059852325e-06,
+      "loss": 0.0557,
+      "step": 2835
+    },
+    {
+      "epoch": 0.5944246489205617,
+      "grad_norm": 0.0489707812666893,
+      "learning_rate": 3.732430769450714e-06,
+      "loss": 0.0533,
+      "step": 2836
+    },
+    {
+      "epoch": 0.5946342485852023,
+      "grad_norm": 0.06095966696739197,
+      "learning_rate": 3.7291471172653463e-06,
+      "loss": 0.0531,
+      "step": 2837
+    },
+    {
+      "epoch": 0.5948438482498428,
+      "grad_norm": 0.05945379659533501,
+      "learning_rate": 3.7258640509428955e-06,
+      "loss": 0.0565,
+      "step": 2838
+    },
+    {
+      "epoch": 0.5950534479144833,
+      "grad_norm": 0.05267561972141266,
+      "learning_rate": 3.7225815719968522e-06,
+      "loss": 0.0533,
+      "step": 2839
+    },
+    {
+      "epoch": 0.5952630475791238,
+      "grad_norm": 0.04585307091474533,
+      "learning_rate": 3.719299681940437e-06,
+      "loss": 0.0541,
+      "step": 2840
+    },
+    {
+      "epoch": 0.5954726472437644,
+      "grad_norm": 0.057470791041851044,
+      "learning_rate": 3.716018382286605e-06,
+      "loss": 0.0551,
+      "step": 2841
+    },
+    {
+      "epoch": 0.595682246908405,
+      "grad_norm": 0.07041089981794357,
+      "learning_rate": 3.7127376745480313e-06,
+      "loss": 0.0542,
+      "step": 2842
+    },
+    {
+      "epoch": 0.5958918465730455,
+      "grad_norm": 0.07431340962648392,
+      "learning_rate": 3.709457560237121e-06,
+      "loss": 0.0535,
+      "step": 2843
+    },
+    {
+      "epoch": 0.596101446237686,
+      "grad_norm": 0.06186486408114433,
+      "learning_rate": 3.7061780408660075e-06,
+      "loss": 0.054,
+      "step": 2844
+    },
+    {
+      "epoch": 0.5963110459023265,
+      "grad_norm": 0.047516003251075745,
+      "learning_rate": 3.7028991179465502e-06,
+      "loss": 0.0556,
+      "step": 2845
+    },
+    {
+      "epoch": 0.596520645566967,
+      "grad_norm": 0.04875975474715233,
+      "learning_rate": 3.699620792990328e-06,
+      "loss": 0.0547,
+      "step": 2846
+    },
+    {
+      "epoch": 0.5967302452316077,
+      "grad_norm": 0.05895498767495155,
+      "learning_rate": 3.696343067508651e-06,
+      "loss": 0.0538,
+      "step": 2847
+    },
+    {
+      "epoch": 0.5969398448962482,
+      "grad_norm": 0.05717795342206955,
+      "learning_rate": 3.6930659430125506e-06,
+      "loss": 0.0563,
+      "step": 2848
+    },
+    {
+      "epoch": 0.5971494445608887,
+      "grad_norm": 0.06923867017030716,
+      "learning_rate": 3.6897894210127765e-06,
+      "loss": 0.056,
+      "step": 2849
+    },
+    {
+      "epoch": 0.5973590442255292,
+      "grad_norm": 0.07923050224781036,
+      "learning_rate": 3.6865135030198084e-06,
+      "loss": 0.0544,
+      "step": 2850
+    },
+    {
+      "epoch": 0.5975686438901697,
+      "grad_norm": 0.07306721806526184,
+      "learning_rate": 3.683238190543843e-06,
+      "loss": 0.055,
+      "step": 2851
+    },
+    {
+      "epoch": 0.5977782435548104,
+      "grad_norm": 0.05905516818165779,
+      "learning_rate": 3.679963485094797e-06,
+      "loss": 0.0547,
+      "step": 2852
+    },
+    {
+      "epoch": 0.5979878432194509,
+      "grad_norm": 0.05046588554978371,
+      "learning_rate": 3.6766893881823106e-06,
+      "loss": 0.0556,
+      "step": 2853
+    },
+    {
+      "epoch": 0.5981974428840914,
+      "grad_norm": 0.06019177287817001,
+      "learning_rate": 3.673415901315743e-06,
+      "loss": 0.0556,
+      "step": 2854
+    },
+    {
+      "epoch": 0.5984070425487319,
+      "grad_norm": 0.06949783116579056,
+      "learning_rate": 3.6701430260041672e-06,
+      "loss": 0.0553,
+      "step": 2855
+    },
+    {
+      "epoch": 0.5986166422133724,
+      "grad_norm": 0.06795257329940796,
+      "learning_rate": 3.6668707637563804e-06,
+      "loss": 0.0538,
+      "step": 2856
+    },
+    {
+      "epoch": 0.598826241878013,
+      "grad_norm": 0.05731310322880745,
+      "learning_rate": 3.6635991160808943e-06,
+      "loss": 0.0557,
+      "step": 2857
+    },
+    {
+      "epoch": 0.5990358415426535,
+      "grad_norm": 0.04602685570716858,
+      "learning_rate": 3.66032808448594e-06,
+      "loss": 0.0542,
+      "step": 2858
+    },
+    {
+      "epoch": 0.5992454412072941,
+      "grad_norm": 0.055196985602378845,
+      "learning_rate": 3.6570576704794593e-06,
+      "loss": 0.0562,
+      "step": 2859
+    },
+    {
+      "epoch": 0.5994550408719346,
+      "grad_norm": 0.06921806186437607,
+      "learning_rate": 3.6537878755691124e-06,
+      "loss": 0.0544,
+      "step": 2860
+    },
+    {
+      "epoch": 0.5996646405365751,
+      "grad_norm": 0.0728953555226326,
+      "learning_rate": 3.650518701262278e-06,
+      "loss": 0.0534,
+      "step": 2861
+    },
+    {
+      "epoch": 0.5998742402012157,
+      "grad_norm": 0.07024196535348892,
+      "learning_rate": 3.6472501490660407e-06,
+      "loss": 0.0554,
+      "step": 2862
+    },
+    {
+      "epoch": 0.6000838398658562,
+      "grad_norm": 0.06623408198356628,
+      "learning_rate": 3.643982220487202e-06,
+      "loss": 0.0524,
+      "step": 2863
+    },
+    {
+      "epoch": 0.6002934395304967,
+      "grad_norm": 0.0592159740626812,
+      "learning_rate": 3.640714917032281e-06,
+      "loss": 0.054,
+      "step": 2864
+    },
+    {
+      "epoch": 0.6005030391951373,
+      "grad_norm": 0.04894890636205673,
+      "learning_rate": 3.637448240207499e-06,
+      "loss": 0.0521,
+      "step": 2865
+    },
+    {
+      "epoch": 0.6007126388597778,
+      "grad_norm": 0.04575192928314209,
+      "learning_rate": 3.634182191518796e-06,
+      "loss": 0.059,
+      "step": 2866
+    },
+    {
+      "epoch": 0.6009222385244184,
+      "grad_norm": 0.05822861194610596,
+      "learning_rate": 3.630916772471817e-06,
+      "loss": 0.0532,
+      "step": 2867
+    },
+    {
+      "epoch": 0.6011318381890589,
+      "grad_norm": 0.08098480850458145,
+      "learning_rate": 3.6276519845719237e-06,
+      "loss": 0.0572,
+      "step": 2868
+    },
+    {
+      "epoch": 0.6013414378536994,
+      "grad_norm": 0.10072686523199081,
+      "learning_rate": 3.624387829324181e-06,
+      "loss": 0.0538,
+      "step": 2869
+    },
+    {
+      "epoch": 0.6015510375183399,
+      "grad_norm": 0.09352651983499527,
+      "learning_rate": 3.62112430823336e-06,
+      "loss": 0.0559,
+      "step": 2870
+    },
+    {
+      "epoch": 0.6017606371829806,
+      "grad_norm": 0.06254759430885315,
+      "learning_rate": 3.61786142280395e-06,
+      "loss": 0.0529,
+      "step": 2871
+    },
+    {
+      "epoch": 0.6019702368476211,
+      "grad_norm": 0.05310570448637009,
+      "learning_rate": 3.6145991745401354e-06,
+      "loss": 0.0568,
+      "step": 2872
+    },
+    {
+      "epoch": 0.6021798365122616,
+      "grad_norm": 0.051381923258304596,
+      "learning_rate": 3.6113375649458126e-06,
+      "loss": 0.0545,
+      "step": 2873
+    },
+    {
+      "epoch": 0.6023894361769021,
+      "grad_norm": 0.04709269851446152,
+      "learning_rate": 3.6080765955245867e-06,
+      "loss": 0.0551,
+      "step": 2874
+    },
+    {
+      "epoch": 0.6025990358415426,
+      "grad_norm": 0.05449269711971283,
+      "learning_rate": 3.6048162677797595e-06,
+      "loss": 0.053,
+      "step": 2875
+    },
+    {
+      "epoch": 0.6028086355061832,
+      "grad_norm": 0.0537823885679245,
+      "learning_rate": 3.601556583214342e-06,
+      "loss": 0.0567,
+      "step": 2876
+    },
+    {
+      "epoch": 0.6030182351708238,
+      "grad_norm": 0.041623324155807495,
+      "learning_rate": 3.5982975433310506e-06,
+      "loss": 0.0548,
+      "step": 2877
+    },
+    {
+      "epoch": 0.6032278348354643,
+      "grad_norm": 0.04960298910737038,
+      "learning_rate": 3.5950391496323007e-06,
+      "loss": 0.0561,
+      "step": 2878
+    },
+    {
+      "epoch": 0.6034374345001048,
+      "grad_norm": 0.06087161973118782,
+      "learning_rate": 3.591781403620209e-06,
+      "loss": 0.0552,
+      "step": 2879
+    },
+    {
+      "epoch": 0.6036470341647453,
+      "grad_norm": 0.050851162523031235,
+      "learning_rate": 3.5885243067965992e-06,
+      "loss": 0.0517,
+      "step": 2880
+    },
+    {
+      "epoch": 0.6038566338293859,
+      "grad_norm": 0.049509234726428986,
+      "learning_rate": 3.585267860662992e-06,
+      "loss": 0.0563,
+      "step": 2881
+    },
+    {
+      "epoch": 0.6040662334940264,
+      "grad_norm": 0.04799444228410721,
+      "learning_rate": 3.582012066720605e-06,
+      "loss": 0.0555,
+      "step": 2882
+    },
+    {
+      "epoch": 0.604275833158667,
+      "grad_norm": 0.05108596384525299,
+      "learning_rate": 3.5787569264703614e-06,
+      "loss": 0.0555,
+      "step": 2883
+    },
+    {
+      "epoch": 0.6044854328233075,
+      "grad_norm": 0.05567923188209534,
+      "learning_rate": 3.575502441412881e-06,
+      "loss": 0.0538,
+      "step": 2884
+    },
+    {
+      "epoch": 0.604695032487948,
+      "grad_norm": 0.052811216562986374,
+      "learning_rate": 3.572248613048477e-06,
+      "loss": 0.0541,
+      "step": 2885
+    },
+    {
+      "epoch": 0.6049046321525886,
+      "grad_norm": 0.05271197855472565,
+      "learning_rate": 3.568995442877167e-06,
+      "loss": 0.0536,
+      "step": 2886
+    },
+    {
+      "epoch": 0.6051142318172291,
+      "grad_norm": 0.056101392954587936,
+      "learning_rate": 3.565742932398661e-06,
+      "loss": 0.0548,
+      "step": 2887
+    },
+    {
+      "epoch": 0.6053238314818696,
+      "grad_norm": 0.06310348957777023,
+      "learning_rate": 3.5624910831123633e-06,
+      "loss": 0.0554,
+      "step": 2888
+    },
+    {
+      "epoch": 0.6055334311465101,
+      "grad_norm": 0.05690898001194,
+      "learning_rate": 3.559239896517379e-06,
+      "loss": 0.056,
+      "step": 2889
+    },
+    {
+      "epoch": 0.6057430308111507,
+      "grad_norm": 0.037709228694438934,
+      "learning_rate": 3.5559893741125018e-06,
+      "loss": 0.0556,
+      "step": 2890
+    },
+    {
+      "epoch": 0.6059526304757913,
+      "grad_norm": 0.043450452387332916,
+      "learning_rate": 3.5527395173962255e-06,
+      "loss": 0.053,
+      "step": 2891
+    },
+    {
+      "epoch": 0.6061622301404318,
+      "grad_norm": 0.051693037152290344,
+      "learning_rate": 3.5494903278667305e-06,
+      "loss": 0.0546,
+      "step": 2892
+    },
+    {
+      "epoch": 0.6063718298050723,
+      "grad_norm": 0.04153525084257126,
+      "learning_rate": 3.5462418070218913e-06,
+      "loss": 0.0546,
+      "step": 2893
+    },
+    {
+      "epoch": 0.6065814294697128,
+      "grad_norm": 0.043478380888700485,
+      "learning_rate": 3.5429939563592795e-06,
+      "loss": 0.0527,
+      "step": 2894
+    },
+    {
+      "epoch": 0.6067910291343533,
+      "grad_norm": 0.053386226296424866,
+      "learning_rate": 3.5397467773761495e-06,
+      "loss": 0.0544,
+      "step": 2895
+    },
+    {
+      "epoch": 0.607000628798994,
+      "grad_norm": 0.037870582193136215,
+      "learning_rate": 3.536500271569452e-06,
+      "loss": 0.0549,
+      "step": 2896
+    },
+    {
+      "epoch": 0.6072102284636345,
+      "grad_norm": 0.043722059577703476,
+      "learning_rate": 3.533254440435826e-06,
+      "loss": 0.0551,
+      "step": 2897
+    },
+    {
+      "epoch": 0.607419828128275,
+      "grad_norm": 0.05376339703798294,
+      "learning_rate": 3.5300092854715985e-06,
+      "loss": 0.0537,
+      "step": 2898
+    },
+    {
+      "epoch": 0.6076294277929155,
+      "grad_norm": 0.04392022639513016,
+      "learning_rate": 3.5267648081727834e-06,
+      "loss": 0.057,
+      "step": 2899
+    },
+    {
+      "epoch": 0.607839027457556,
+      "grad_norm": 0.04517117887735367,
+      "learning_rate": 3.523521010035089e-06,
+      "loss": 0.0543,
+      "step": 2900
+    },
+    {
+      "epoch": 0.6080486271221966,
+      "grad_norm": 0.05903865396976471,
+      "learning_rate": 3.520277892553899e-06,
+      "loss": 0.0538,
+      "step": 2901
+    },
+    {
+      "epoch": 0.6082582267868372,
+      "grad_norm": 0.049636900424957275,
+      "learning_rate": 3.5170354572242936e-06,
+      "loss": 0.0557,
+      "step": 2902
+    },
+    {
+      "epoch": 0.6084678264514777,
+      "grad_norm": 0.037783827632665634,
+      "learning_rate": 3.5137937055410343e-06,
+      "loss": 0.0566,
+      "step": 2903
+    },
+    {
+      "epoch": 0.6086774261161182,
+      "grad_norm": 0.04359547793865204,
+      "learning_rate": 3.51055263899857e-06,
+      "loss": 0.0542,
+      "step": 2904
+    },
+    {
+      "epoch": 0.6088870257807587,
+      "grad_norm": 0.042558085173368454,
+      "learning_rate": 3.5073122590910285e-06,
+      "loss": 0.057,
+      "step": 2905
+    },
+    {
+      "epoch": 0.6090966254453993,
+      "grad_norm": 0.04583901911973953,
+      "learning_rate": 3.5040725673122246e-06,
+      "loss": 0.0591,
+      "step": 2906
+    },
+    {
+      "epoch": 0.6093062251100398,
+      "grad_norm": 0.06440989673137665,
+      "learning_rate": 3.500833565155658e-06,
+      "loss": 0.057,
+      "step": 2907
+    },
+    {
+      "epoch": 0.6095158247746804,
+      "grad_norm": 0.057344041764736176,
+      "learning_rate": 3.4975952541145063e-06,
+      "loss": 0.0544,
+      "step": 2908
+    },
+    {
+      "epoch": 0.6097254244393209,
+      "grad_norm": 0.04767520725727081,
+      "learning_rate": 3.4943576356816287e-06,
+      "loss": 0.0573,
+      "step": 2909
+    },
+    {
+      "epoch": 0.6099350241039614,
+      "grad_norm": 0.04085380956530571,
+      "learning_rate": 3.4911207113495703e-06,
+      "loss": 0.0563,
+      "step": 2910
+    },
+    {
+      "epoch": 0.610144623768602,
+      "grad_norm": 0.03516289219260216,
+      "learning_rate": 3.4878844826105497e-06,
+      "loss": 0.0583,
+      "step": 2911
+    },
+    {
+      "epoch": 0.6103542234332425,
+      "grad_norm": 0.048390522599220276,
+      "learning_rate": 3.4846489509564674e-06,
+      "loss": 0.0521,
+      "step": 2912
+    },
+    {
+      "epoch": 0.610563823097883,
+      "grad_norm": 0.05823476240038872,
+      "learning_rate": 3.481414117878906e-06,
+      "loss": 0.0571,
+      "step": 2913
+    },
+    {
+      "epoch": 0.6107734227625236,
+      "grad_norm": 0.05212031677365303,
+      "learning_rate": 3.47817998486912e-06,
+      "loss": 0.0531,
+      "step": 2914
+    },
+    {
+      "epoch": 0.6109830224271641,
+      "grad_norm": 0.054088983684778214,
+      "learning_rate": 3.474946553418044e-06,
+      "loss": 0.0541,
+      "step": 2915
+    },
+    {
+      "epoch": 0.6111926220918047,
+      "grad_norm": 0.05518088862299919,
+      "learning_rate": 3.4717138250162908e-06,
+      "loss": 0.0524,
+      "step": 2916
+    },
+    {
+      "epoch": 0.6114022217564452,
+      "grad_norm": 0.04526456817984581,
+      "learning_rate": 3.4684818011541484e-06,
+      "loss": 0.0545,
+      "step": 2917
+    },
+    {
+      "epoch": 0.6116118214210857,
+      "grad_norm": 0.036848414689302444,
+      "learning_rate": 3.465250483321575e-06,
+      "loss": 0.053,
+      "step": 2918
+    },
+    {
+      "epoch": 0.6118214210857262,
+      "grad_norm": 0.03322865068912506,
+      "learning_rate": 3.462019873008211e-06,
+      "loss": 0.0532,
+      "step": 2919
+    },
+    {
+      "epoch": 0.6120310207503667,
+      "grad_norm": 0.035156525671482086,
+      "learning_rate": 3.458789971703367e-06,
+      "loss": 0.0583,
+      "step": 2920
+    },
+    {
+      "epoch": 0.6122406204150074,
+      "grad_norm": 0.03802697733044624,
+      "learning_rate": 3.4555607808960232e-06,
+      "loss": 0.0545,
+      "step": 2921
+    },
+    {
+      "epoch": 0.6124502200796479,
+      "grad_norm": 0.0401364266872406,
+      "learning_rate": 3.4523323020748413e-06,
+      "loss": 0.0569,
+      "step": 2922
+    },
+    {
+      "epoch": 0.6126598197442884,
+      "grad_norm": 0.04033052548766136,
+      "learning_rate": 3.449104536728146e-06,
+      "loss": 0.054,
+      "step": 2923
+    },
+    {
+      "epoch": 0.6128694194089289,
+      "grad_norm": 0.0390617661178112,
+      "learning_rate": 3.4458774863439366e-06,
+      "loss": 0.0552,
+      "step": 2924
+    },
+    {
+      "epoch": 0.6130790190735694,
+      "grad_norm": 0.042450230568647385,
+      "learning_rate": 3.4426511524098834e-06,
+      "loss": 0.0538,
+      "step": 2925
+    },
+    {
+      "epoch": 0.6132886187382101,
+      "grad_norm": 0.038516972213983536,
+      "learning_rate": 3.4394255364133245e-06,
+      "loss": 0.0565,
+      "step": 2926
+    },
+    {
+      "epoch": 0.6134982184028506,
+      "grad_norm": 0.029110699892044067,
+      "learning_rate": 3.436200639841271e-06,
+      "loss": 0.0553,
+      "step": 2927
+    },
+    {
+      "epoch": 0.6137078180674911,
+      "grad_norm": 0.0337090939283371,
+      "learning_rate": 3.432976464180397e-06,
+      "loss": 0.0579,
+      "step": 2928
+    },
+    {
+      "epoch": 0.6139174177321316,
+      "grad_norm": 0.04489186406135559,
+      "learning_rate": 3.4297530109170463e-06,
+      "loss": 0.0531,
+      "step": 2929
+    },
+    {
+      "epoch": 0.6141270173967721,
+      "grad_norm": 0.04716401919722557,
+      "learning_rate": 3.426530281537234e-06,
+      "loss": 0.0538,
+      "step": 2930
+    },
+    {
+      "epoch": 0.6143366170614127,
+      "grad_norm": 0.053189124912023544,
+      "learning_rate": 3.423308277526633e-06,
+      "loss": 0.0568,
+      "step": 2931
+    },
+    {
+      "epoch": 0.6145462167260533,
+      "grad_norm": 0.05599994957447052,
+      "learning_rate": 3.4200870003705883e-06,
+      "loss": 0.056,
+      "step": 2932
+    },
+    {
+      "epoch": 0.6147558163906938,
+      "grad_norm": 0.04985819011926651,
+      "learning_rate": 3.41686645155411e-06,
+      "loss": 0.0561,
+      "step": 2933
+    },
+    {
+      "epoch": 0.6149654160553343,
+      "grad_norm": 0.04989127814769745,
+      "learning_rate": 3.413646632561868e-06,
+      "loss": 0.0537,
+      "step": 2934
+    },
+    {
+      "epoch": 0.6151750157199748,
+      "grad_norm": 0.04989814758300781,
+      "learning_rate": 3.410427544878198e-06,
+      "loss": 0.0575,
+      "step": 2935
+    },
+    {
+      "epoch": 0.6153846153846154,
+      "grad_norm": 0.03400817885994911,
+      "learning_rate": 3.4072091899871016e-06,
+      "loss": 0.0551,
+      "step": 2936
+    },
+    {
+      "epoch": 0.6155942150492559,
+      "grad_norm": 0.034855592995882034,
+      "learning_rate": 3.403991569372235e-06,
+      "loss": 0.0567,
+      "step": 2937
+    },
+    {
+      "epoch": 0.6158038147138964,
+      "grad_norm": 0.042559489607810974,
+      "learning_rate": 3.4007746845169253e-06,
+      "loss": 0.0539,
+      "step": 2938
+    },
+    {
+      "epoch": 0.616013414378537,
+      "grad_norm": 0.042176634073257446,
+      "learning_rate": 3.397558536904152e-06,
+      "loss": 0.0521,
+      "step": 2939
+    },
+    {
+      "epoch": 0.6162230140431776,
+      "grad_norm": 0.04478650167584419,
+      "learning_rate": 3.394343128016563e-06,
+      "loss": 0.0573,
+      "step": 2940
+    },
+    {
+      "epoch": 0.6164326137078181,
+      "grad_norm": 0.04596136510372162,
+      "learning_rate": 3.3911284593364568e-06,
+      "loss": 0.056,
+      "step": 2941
+    },
+    {
+      "epoch": 0.6166422133724586,
+      "grad_norm": 0.044569674879312515,
+      "learning_rate": 3.387914532345796e-06,
+      "loss": 0.0557,
+      "step": 2942
+    },
+    {
+      "epoch": 0.6168518130370991,
+      "grad_norm": 0.03586115688085556,
+      "learning_rate": 3.3847013485262037e-06,
+      "loss": 0.0556,
+      "step": 2943
+    },
+    {
+      "epoch": 0.6170614127017396,
+      "grad_norm": 0.05500089004635811,
+      "learning_rate": 3.381488909358952e-06,
+      "loss": 0.0532,
+      "step": 2944
+    },
+    {
+      "epoch": 0.6172710123663803,
+      "grad_norm": 0.06483340263366699,
+      "learning_rate": 3.3782772163249767e-06,
+      "loss": 0.0562,
+      "step": 2945
+    },
+    {
+      "epoch": 0.6174806120310208,
+      "grad_norm": 0.057424839586019516,
+      "learning_rate": 3.375066270904869e-06,
+      "loss": 0.0568,
+      "step": 2946
+    },
+    {
+      "epoch": 0.6176902116956613,
+      "grad_norm": 0.0566512756049633,
+      "learning_rate": 3.3718560745788724e-06,
+      "loss": 0.0588,
+      "step": 2947
+    },
+    {
+      "epoch": 0.6178998113603018,
+      "grad_norm": 0.05279812961816788,
+      "learning_rate": 3.368646628826886e-06,
+      "loss": 0.0549,
+      "step": 2948
+    },
+    {
+      "epoch": 0.6181094110249423,
+      "grad_norm": 0.03966805338859558,
+      "learning_rate": 3.365437935128466e-06,
+      "loss": 0.0557,
+      "step": 2949
+    },
+    {
+      "epoch": 0.618319010689583,
+      "grad_norm": 0.03704335168004036,
+      "learning_rate": 3.3622299949628197e-06,
+      "loss": 0.0535,
+      "step": 2950
+    },
+    {
+      "epoch": 0.6185286103542235,
+      "grad_norm": 0.03752034902572632,
+      "learning_rate": 3.359022809808803e-06,
+      "loss": 0.055,
+      "step": 2951
+    },
+    {
+      "epoch": 0.618738210018864,
+      "grad_norm": 0.04068942740559578,
+      "learning_rate": 3.3558163811449317e-06,
+      "loss": 0.0546,
+      "step": 2952
+    },
+    {
+      "epoch": 0.6189478096835045,
+      "grad_norm": 0.03766583278775215,
+      "learning_rate": 3.352610710449368e-06,
+      "loss": 0.0553,
+      "step": 2953
+    },
+    {
+      "epoch": 0.619157409348145,
+      "grad_norm": 0.044414669275283813,
+      "learning_rate": 3.349405799199922e-06,
+      "loss": 0.0573,
+      "step": 2954
+    },
+    {
+      "epoch": 0.6193670090127856,
+      "grad_norm": 0.06159931421279907,
+      "learning_rate": 3.3462016488740612e-06,
+      "loss": 0.0544,
+      "step": 2955
+    },
+    {
+      "epoch": 0.6195766086774261,
+      "grad_norm": 0.05573199316859245,
+      "learning_rate": 3.3429982609488976e-06,
+      "loss": 0.0549,
+      "step": 2956
+    },
+    {
+      "epoch": 0.6197862083420667,
+      "grad_norm": 0.03836704418063164,
+      "learning_rate": 3.3397956369011897e-06,
+      "loss": 0.0552,
+      "step": 2957
+    },
+    {
+      "epoch": 0.6199958080067072,
+      "grad_norm": 0.05629691854119301,
+      "learning_rate": 3.3365937782073486e-06,
+      "loss": 0.0566,
+      "step": 2958
+    },
+    {
+      "epoch": 0.6202054076713477,
+      "grad_norm": 0.07051456719636917,
+      "learning_rate": 3.3333926863434317e-06,
+      "loss": 0.0524,
+      "step": 2959
+    },
+    {
+      "epoch": 0.6204150073359883,
+      "grad_norm": 0.0706133171916008,
+      "learning_rate": 3.3301923627851385e-06,
+      "loss": 0.0567,
+      "step": 2960
+    },
+    {
+      "epoch": 0.6206246070006288,
+      "grad_norm": 0.058882150799036026,
+      "learning_rate": 3.3269928090078204e-06,
+      "loss": 0.054,
+      "step": 2961
+    },
+    {
+      "epoch": 0.6208342066652693,
+      "grad_norm": 0.04647863656282425,
+      "learning_rate": 3.3237940264864684e-06,
+      "loss": 0.0558,
+      "step": 2962
+    },
+    {
+      "epoch": 0.6210438063299099,
+      "grad_norm": 0.046870891004800797,
+      "learning_rate": 3.320596016695724e-06,
+      "loss": 0.0534,
+      "step": 2963
+    },
+    {
+      "epoch": 0.6212534059945504,
+      "grad_norm": 0.043968793004751205,
+      "learning_rate": 3.3173987811098664e-06,
+      "loss": 0.0568,
+      "step": 2964
+    },
+    {
+      "epoch": 0.621463005659191,
+      "grad_norm": 0.05325428396463394,
+      "learning_rate": 3.3142023212028197e-06,
+      "loss": 0.0536,
+      "step": 2965
+    },
+    {
+      "epoch": 0.6216726053238315,
+      "grad_norm": 0.049214769154787064,
+      "learning_rate": 3.311006638448155e-06,
+      "loss": 0.0545,
+      "step": 2966
+    },
+    {
+      "epoch": 0.621882204988472,
+      "grad_norm": 0.04188595712184906,
+      "learning_rate": 3.307811734319078e-06,
+      "loss": 0.0546,
+      "step": 2967
+    },
+    {
+      "epoch": 0.6220918046531125,
+      "grad_norm": 0.040931567549705505,
+      "learning_rate": 3.304617610288439e-06,
+      "loss": 0.0557,
+      "step": 2968
+    },
+    {
+      "epoch": 0.622301404317753,
+      "grad_norm": 0.03652331605553627,
+      "learning_rate": 3.3014242678287327e-06,
+      "loss": 0.054,
+      "step": 2969
+    },
+    {
+      "epoch": 0.6225110039823937,
+      "grad_norm": 0.0368216410279274,
+      "learning_rate": 3.298231708412083e-06,
+      "loss": 0.0558,
+      "step": 2970
+    },
+    {
+      "epoch": 0.6227206036470342,
+      "grad_norm": 0.03968511521816254,
+      "learning_rate": 3.295039933510262e-06,
+      "loss": 0.0557,
+      "step": 2971
+    },
+    {
+      "epoch": 0.6229302033116747,
+      "grad_norm": 0.04443355277180672,
+      "learning_rate": 3.291848944594679e-06,
+      "loss": 0.0524,
+      "step": 2972
+    },
+    {
+      "epoch": 0.6231398029763152,
+      "grad_norm": 0.05209925025701523,
+      "learning_rate": 3.288658743136378e-06,
+      "loss": 0.0553,
+      "step": 2973
+    },
+    {
+      "epoch": 0.6233494026409557,
+      "grad_norm": 0.05424797162413597,
+      "learning_rate": 3.2854693306060407e-06,
+      "loss": 0.0547,
+      "step": 2974
+    },
+    {
+      "epoch": 0.6235590023055964,
+      "grad_norm": 0.04404807463288307,
+      "learning_rate": 3.282280708473985e-06,
+      "loss": 0.0548,
+      "step": 2975
+    },
+    {
+      "epoch": 0.6237686019702369,
+      "grad_norm": 0.03802521154284477,
+      "learning_rate": 3.2790928782101674e-06,
+      "loss": 0.0596,
+      "step": 2976
+    },
+    {
+      "epoch": 0.6239782016348774,
+      "grad_norm": 0.045362699776887894,
+      "learning_rate": 3.2759058412841742e-06,
+      "loss": 0.0551,
+      "step": 2977
+    },
+    {
+      "epoch": 0.6241878012995179,
+      "grad_norm": 0.05245572328567505,
+      "learning_rate": 3.2727195991652295e-06,
+      "loss": 0.0543,
+      "step": 2978
+    },
+    {
+      "epoch": 0.6243974009641584,
+      "grad_norm": 0.056894753128290176,
+      "learning_rate": 3.2695341533221926e-06,
+      "loss": 0.0573,
+      "step": 2979
+    },
+    {
+      "epoch": 0.624607000628799,
+      "grad_norm": 0.05126248672604561,
+      "learning_rate": 3.2663495052235505e-06,
+      "loss": 0.0536,
+      "step": 2980
+    },
+    {
+      "epoch": 0.6248166002934396,
+      "grad_norm": 0.05021638423204422,
+      "learning_rate": 3.263165656337426e-06,
+      "loss": 0.0562,
+      "step": 2981
+    },
+    {
+      "epoch": 0.6250261999580801,
+      "grad_norm": 0.04641691595315933,
+      "learning_rate": 3.2599826081315744e-06,
+      "loss": 0.0555,
+      "step": 2982
+    },
+    {
+      "epoch": 0.6252357996227206,
+      "grad_norm": 0.04853666201233864,
+      "learning_rate": 3.2568003620733778e-06,
+      "loss": 0.0559,
+      "step": 2983
+    },
+    {
+      "epoch": 0.6254453992873611,
+      "grad_norm": 0.05443096533417702,
+      "learning_rate": 3.2536189196298518e-06,
+      "loss": 0.0531,
+      "step": 2984
+    },
+    {
+      "epoch": 0.6256549989520017,
+      "grad_norm": 0.048564642667770386,
+      "learning_rate": 3.250438282267642e-06,
+      "loss": 0.0532,
+      "step": 2985
+    },
+    {
+      "epoch": 0.6258645986166422,
+      "grad_norm": 0.04952483996748924,
+      "learning_rate": 3.247258451453022e-06,
+      "loss": 0.0535,
+      "step": 2986
+    },
+    {
+      "epoch": 0.6260741982812827,
+      "grad_norm": 0.038559552282094955,
+      "learning_rate": 3.2440794286518896e-06,
+      "loss": 0.0568,
+      "step": 2987
+    },
+    {
+      "epoch": 0.6262837979459233,
+      "grad_norm": 0.0397375226020813,
+      "learning_rate": 3.2409012153297762e-06,
+      "loss": 0.0548,
+      "step": 2988
+    },
+    {
+      "epoch": 0.6264933976105638,
+      "grad_norm": 0.050432708114385605,
+      "learning_rate": 3.2377238129518392e-06,
+      "loss": 0.0539,
+      "step": 2989
+    },
+    {
+      "epoch": 0.6267029972752044,
+      "grad_norm": 0.04017667844891548,
+      "learning_rate": 3.2345472229828556e-06,
+      "loss": 0.0548,
+      "step": 2990
+    },
+    {
+      "epoch": 0.6269125969398449,
+      "grad_norm": 0.035998113453388214,
+      "learning_rate": 3.231371446887237e-06,
+      "loss": 0.0556,
+      "step": 2991
+    },
+    {
+      "epoch": 0.6271221966044854,
+      "grad_norm": 0.044560160487890244,
+      "learning_rate": 3.2281964861290146e-06,
+      "loss": 0.0553,
+      "step": 2992
+    },
+    {
+      "epoch": 0.6273317962691259,
+      "grad_norm": 0.04030834883451462,
+      "learning_rate": 3.225022342171842e-06,
+      "loss": 0.0535,
+      "step": 2993
+    },
+    {
+      "epoch": 0.6275413959337665,
+      "grad_norm": 0.03907003253698349,
+      "learning_rate": 3.2218490164790015e-06,
+      "loss": 0.0563,
+      "step": 2994
+    },
+    {
+      "epoch": 0.6277509955984071,
+      "grad_norm": 0.037859588861465454,
+      "learning_rate": 3.2186765105133955e-06,
+      "loss": 0.0553,
+      "step": 2995
+    },
+    {
+      "epoch": 0.6279605952630476,
+      "grad_norm": 0.04589460417628288,
+      "learning_rate": 3.215504825737549e-06,
+      "loss": 0.0531,
+      "step": 2996
+    },
+    {
+      "epoch": 0.6281701949276881,
+      "grad_norm": 0.05734090879559517,
+      "learning_rate": 3.2123339636136065e-06,
+      "loss": 0.0535,
+      "step": 2997
+    },
+    {
+      "epoch": 0.6283797945923286,
+      "grad_norm": 0.04286714643239975,
+      "learning_rate": 3.209163925603335e-06,
+      "loss": 0.0539,
+      "step": 2998
+    },
+    {
+      "epoch": 0.6285893942569691,
+      "grad_norm": 0.04124724492430687,
+      "learning_rate": 3.2059947131681226e-06,
+      "loss": 0.0566,
+      "step": 2999
+    },
+    {
+      "epoch": 0.6287989939216098,
+      "grad_norm": 0.06533816456794739,
+      "learning_rate": 3.202826327768974e-06,
+      "loss": 0.0547,
+      "step": 3000
+    },
+    {
+      "epoch": 0.6290085935862503,
+      "grad_norm": 0.08003267645835876,
+      "learning_rate": 3.199658770866515e-06,
+      "loss": 0.0532,
+      "step": 3001
+    },
+    {
+      "epoch": 0.6292181932508908,
+      "grad_norm": 0.05155742168426514,
+      "learning_rate": 3.19649204392099e-06,
+      "loss": 0.0543,
+      "step": 3002
+    },
+    {
+      "epoch": 0.6294277929155313,
+      "grad_norm": 0.0488094687461853,
+      "learning_rate": 3.193326148392257e-06,
+      "loss": 0.0571,
+      "step": 3003
+    },
+    {
+      "epoch": 0.6296373925801718,
+      "grad_norm": 0.07249227166175842,
+      "learning_rate": 3.1901610857397936e-06,
+      "loss": 0.0532,
+      "step": 3004
+    },
+    {
+      "epoch": 0.6298469922448124,
+      "grad_norm": 0.09708064049482346,
+      "learning_rate": 3.1869968574226966e-06,
+      "loss": 0.0545,
+      "step": 3005
+    },
+    {
+      "epoch": 0.630056591909453,
+      "grad_norm": 0.11073527485132217,
+      "learning_rate": 3.183833464899669e-06,
+      "loss": 0.055,
+      "step": 3006
+    },
+    {
+      "epoch": 0.6302661915740935,
+      "grad_norm": 0.10113838315010071,
+      "learning_rate": 3.1806709096290377e-06,
+      "loss": 0.0557,
+      "step": 3007
+    },
+    {
+      "epoch": 0.630475791238734,
+      "grad_norm": 0.06188422441482544,
+      "learning_rate": 3.1775091930687374e-06,
+      "loss": 0.0557,
+      "step": 3008
+    },
+    {
+      "epoch": 0.6306853909033746,
+      "grad_norm": 0.04998209327459335,
+      "learning_rate": 3.1743483166763234e-06,
+      "loss": 0.0554,
+      "step": 3009
+    },
+    {
+      "epoch": 0.6308949905680151,
+      "grad_norm": 0.11552390456199646,
+      "learning_rate": 3.1711882819089553e-06,
+      "loss": 0.0551,
+      "step": 3010
+    },
+    {
+      "epoch": 0.6311045902326556,
+      "grad_norm": 0.1522252857685089,
+      "learning_rate": 3.1680290902234078e-06,
+      "loss": 0.0541,
+      "step": 3011
+    },
+    {
+      "epoch": 0.6313141898972962,
+      "grad_norm": 0.12954400479793549,
+      "learning_rate": 3.164870743076072e-06,
+      "loss": 0.0543,
+      "step": 3012
+    },
+    {
+      "epoch": 0.6315237895619367,
+      "grad_norm": 0.06520622968673706,
+      "learning_rate": 3.16171324192294e-06,
+      "loss": 0.0547,
+      "step": 3013
+    },
+    {
+      "epoch": 0.6317333892265773,
+      "grad_norm": 0.06795618683099747,
+      "learning_rate": 3.1585565882196223e-06,
+      "loss": 0.052,
+      "step": 3014
+    },
+    {
+      "epoch": 0.6319429888912178,
+      "grad_norm": 0.12307886779308319,
+      "learning_rate": 3.1554007834213357e-06,
+      "loss": 0.0545,
+      "step": 3015
+    },
+    {
+      "epoch": 0.6321525885558583,
+      "grad_norm": 0.10787169635295868,
+      "learning_rate": 3.1522458289829045e-06,
+      "loss": 0.0564,
+      "step": 3016
+    },
+    {
+      "epoch": 0.6323621882204988,
+      "grad_norm": 0.04827128350734711,
+      "learning_rate": 3.1490917263587607e-06,
+      "loss": 0.0542,
+      "step": 3017
+    },
+    {
+      "epoch": 0.6325717878851393,
+      "grad_norm": 0.10941686481237411,
+      "learning_rate": 3.1459384770029476e-06,
+      "loss": 0.0537,
+      "step": 3018
+    },
+    {
+      "epoch": 0.63278138754978,
+      "grad_norm": 0.13342756032943726,
+      "learning_rate": 3.1427860823691136e-06,
+      "loss": 0.054,
+      "step": 3019
+    },
+    {
+      "epoch": 0.6329909872144205,
+      "grad_norm": 0.0916086882352829,
+      "learning_rate": 3.139634543910507e-06,
+      "loss": 0.0536,
+      "step": 3020
+    },
+    {
+      "epoch": 0.633200586879061,
+      "grad_norm": 0.06962735950946808,
+      "learning_rate": 3.13648386307999e-06,
+      "loss": 0.0578,
+      "step": 3021
+    },
+    {
+      "epoch": 0.6334101865437015,
+      "grad_norm": 0.11727678030729294,
+      "learning_rate": 3.1333340413300263e-06,
+      "loss": 0.054,
+      "step": 3022
+    },
+    {
+      "epoch": 0.633619786208342,
+      "grad_norm": 0.11866160482168198,
+      "learning_rate": 3.1301850801126797e-06,
+      "loss": 0.0551,
+      "step": 3023
+    },
+    {
+      "epoch": 0.6338293858729827,
+      "grad_norm": 0.06939590722322464,
+      "learning_rate": 3.127036980879624e-06,
+      "loss": 0.0589,
+      "step": 3024
+    },
+    {
+      "epoch": 0.6340389855376232,
+      "grad_norm": 0.07698585093021393,
+      "learning_rate": 3.123889745082132e-06,
+      "loss": 0.0568,
+      "step": 3025
+    },
+    {
+      "epoch": 0.6342485852022637,
+      "grad_norm": 0.10211010277271271,
+      "learning_rate": 3.1207433741710757e-06,
+      "loss": 0.0565,
+      "step": 3026
+    },
+    {
+      "epoch": 0.6344581848669042,
+      "grad_norm": 0.08808134496212006,
+      "learning_rate": 3.1175978695969344e-06,
+      "loss": 0.0556,
+      "step": 3027
+    },
+    {
+      "epoch": 0.6346677845315447,
+      "grad_norm": 0.05355629697442055,
+      "learning_rate": 3.1144532328097853e-06,
+      "loss": 0.0566,
+      "step": 3028
+    },
+    {
+      "epoch": 0.6348773841961853,
+      "grad_norm": 0.06826507300138474,
+      "learning_rate": 3.1113094652593023e-06,
+      "loss": 0.0551,
+      "step": 3029
+    },
+    {
+      "epoch": 0.6350869838608258,
+      "grad_norm": 0.07204960286617279,
+      "learning_rate": 3.108166568394765e-06,
+      "loss": 0.0561,
+      "step": 3030
+    },
+    {
+      "epoch": 0.6352965835254664,
+      "grad_norm": 0.06383077800273895,
+      "learning_rate": 3.105024543665045e-06,
+      "loss": 0.0559,
+      "step": 3031
+    },
+    {
+      "epoch": 0.6355061831901069,
+      "grad_norm": 0.050440918654203415,
+      "learning_rate": 3.1018833925186194e-06,
+      "loss": 0.0551,
+      "step": 3032
+    },
+    {
+      "epoch": 0.6357157828547474,
+      "grad_norm": 0.05051702260971069,
+      "learning_rate": 3.0987431164035542e-06,
+      "loss": 0.0548,
+      "step": 3033
+    },
+    {
+      "epoch": 0.635925382519388,
+      "grad_norm": 0.05574216693639755,
+      "learning_rate": 3.0956037167675164e-06,
+      "loss": 0.0549,
+      "step": 3034
+    },
+    {
+      "epoch": 0.6361349821840285,
+      "grad_norm": 0.047421135008335114,
+      "learning_rate": 3.092465195057771e-06,
+      "loss": 0.0559,
+      "step": 3035
+    },
+    {
+      "epoch": 0.636344581848669,
+      "grad_norm": 0.0535261407494545,
+      "learning_rate": 3.0893275527211742e-06,
+      "loss": 0.055,
+      "step": 3036
+    },
+    {
+      "epoch": 0.6365541815133096,
+      "grad_norm": 0.05517066270112991,
+      "learning_rate": 3.0861907912041776e-06,
+      "loss": 0.0564,
+      "step": 3037
+    },
+    {
+      "epoch": 0.6367637811779501,
+      "grad_norm": 0.04538208991289139,
+      "learning_rate": 3.083054911952831e-06,
+      "loss": 0.0563,
+      "step": 3038
+    },
+    {
+      "epoch": 0.6369733808425907,
+      "grad_norm": 0.054025497287511826,
+      "learning_rate": 3.0799199164127704e-06,
+      "loss": 0.0538,
+      "step": 3039
+    },
+    {
+      "epoch": 0.6371829805072312,
+      "grad_norm": 0.049201078712940216,
+      "learning_rate": 3.0767858060292285e-06,
+      "loss": 0.057,
+      "step": 3040
+    },
+    {
+      "epoch": 0.6373925801718717,
+      "grad_norm": 0.042893946170806885,
+      "learning_rate": 3.073652582247033e-06,
+      "loss": 0.0548,
+      "step": 3041
+    },
+    {
+      "epoch": 0.6376021798365122,
+      "grad_norm": 0.04716379567980766,
+      "learning_rate": 3.070520246510595e-06,
+      "loss": 0.0529,
+      "step": 3042
+    },
+    {
+      "epoch": 0.6378117795011528,
+      "grad_norm": 0.05528158321976662,
+      "learning_rate": 3.067388800263923e-06,
+      "loss": 0.055,
+      "step": 3043
+    },
+    {
+      "epoch": 0.6380213791657934,
+      "grad_norm": 0.052032019942998886,
+      "learning_rate": 3.064258244950612e-06,
+      "loss": 0.0524,
+      "step": 3044
+    },
+    {
+      "epoch": 0.6382309788304339,
+      "grad_norm": 0.041680578142404556,
+      "learning_rate": 3.0611285820138493e-06,
+      "loss": 0.0556,
+      "step": 3045
+    },
+    {
+      "epoch": 0.6384405784950744,
+      "grad_norm": 0.050715796649456024,
+      "learning_rate": 3.0579998128964073e-06,
+      "loss": 0.0546,
+      "step": 3046
+    },
+    {
+      "epoch": 0.6386501781597149,
+      "grad_norm": 0.057833991944789886,
+      "learning_rate": 3.0548719390406467e-06,
+      "loss": 0.0522,
+      "step": 3047
+    },
+    {
+      "epoch": 0.6388597778243554,
+      "grad_norm": 0.05053876340389252,
+      "learning_rate": 3.051744961888521e-06,
+      "loss": 0.0559,
+      "step": 3048
+    },
+    {
+      "epoch": 0.6390693774889961,
+      "grad_norm": 0.047344524413347244,
+      "learning_rate": 3.048618882881561e-06,
+      "loss": 0.0547,
+      "step": 3049
+    },
+    {
+      "epoch": 0.6392789771536366,
+      "grad_norm": 0.05755054950714111,
+      "learning_rate": 3.04549370346089e-06,
+      "loss": 0.0537,
+      "step": 3050
+    },
+    {
+      "epoch": 0.6394885768182771,
+      "grad_norm": 0.062311019748449326,
+      "learning_rate": 3.0423694250672182e-06,
+      "loss": 0.0551,
+      "step": 3051
+    },
+    {
+      "epoch": 0.6396981764829176,
+      "grad_norm": 0.05935823917388916,
+      "learning_rate": 3.0392460491408333e-06,
+      "loss": 0.0545,
+      "step": 3052
+    },
+    {
+      "epoch": 0.6399077761475581,
+      "grad_norm": 0.05066816136240959,
+      "learning_rate": 3.0361235771216114e-06,
+      "loss": 0.0509,
+      "step": 3053
+    },
+    {
+      "epoch": 0.6401173758121987,
+      "grad_norm": 0.06608299165964127,
+      "learning_rate": 3.033002010449014e-06,
+      "loss": 0.0565,
+      "step": 3054
+    },
+    {
+      "epoch": 0.6403269754768393,
+      "grad_norm": 0.05467569828033447,
+      "learning_rate": 3.0298813505620816e-06,
+      "loss": 0.0547,
+      "step": 3055
+    },
+    {
+      "epoch": 0.6405365751414798,
+      "grad_norm": 0.048877786844968796,
+      "learning_rate": 3.0267615988994357e-06,
+      "loss": 0.0563,
+      "step": 3056
+    },
+    {
+      "epoch": 0.6407461748061203,
+      "grad_norm": 0.053627368062734604,
+      "learning_rate": 3.0236427568992845e-06,
+      "loss": 0.0565,
+      "step": 3057
+    },
+    {
+      "epoch": 0.6409557744707608,
+      "grad_norm": 0.05301428213715553,
+      "learning_rate": 3.020524825999412e-06,
+      "loss": 0.0546,
+      "step": 3058
+    },
+    {
+      "epoch": 0.6411653741354014,
+      "grad_norm": 0.0528411827981472,
+      "learning_rate": 3.0174078076371815e-06,
+      "loss": 0.053,
+      "step": 3059
+    },
+    {
+      "epoch": 0.6413749738000419,
+      "grad_norm": 0.055189281702041626,
+      "learning_rate": 3.014291703249541e-06,
+      "loss": 0.0531,
+      "step": 3060
+    },
+    {
+      "epoch": 0.6415845734646825,
+      "grad_norm": 0.0514843687415123,
+      "learning_rate": 3.011176514273014e-06,
+      "loss": 0.0567,
+      "step": 3061
+    },
+    {
+      "epoch": 0.641794173129323,
+      "grad_norm": 0.04860663786530495,
+      "learning_rate": 3.008062242143699e-06,
+      "loss": 0.0563,
+      "step": 3062
+    },
+    {
+      "epoch": 0.6420037727939635,
+      "grad_norm": 0.05139222368597984,
+      "learning_rate": 3.0049488882972773e-06,
+      "loss": 0.0565,
+      "step": 3063
+    },
+    {
+      "epoch": 0.6422133724586041,
+      "grad_norm": 0.04456450790166855,
+      "learning_rate": 3.0018364541690048e-06,
+      "loss": 0.0534,
+      "step": 3064
+    },
+    {
+      "epoch": 0.6424229721232446,
+      "grad_norm": 0.05100685730576515,
+      "learning_rate": 2.9987249411937096e-06,
+      "loss": 0.0531,
+      "step": 3065
+    },
+    {
+      "epoch": 0.6426325717878851,
+      "grad_norm": 0.04359569400548935,
+      "learning_rate": 2.9956143508058023e-06,
+      "loss": 0.0551,
+      "step": 3066
+    },
+    {
+      "epoch": 0.6428421714525256,
+      "grad_norm": 0.049825966358184814,
+      "learning_rate": 2.9925046844392613e-06,
+      "loss": 0.0554,
+      "step": 3067
+    },
+    {
+      "epoch": 0.6430517711171662,
+      "grad_norm": 0.05026031658053398,
+      "learning_rate": 2.9893959435276464e-06,
+      "loss": 0.0543,
+      "step": 3068
+    },
+    {
+      "epoch": 0.6432613707818068,
+      "grad_norm": 0.04848875477910042,
+      "learning_rate": 2.9862881295040826e-06,
+      "loss": 0.055,
+      "step": 3069
+    },
+    {
+      "epoch": 0.6434709704464473,
+      "grad_norm": 0.0475265197455883,
+      "learning_rate": 2.983181243801272e-06,
+      "loss": 0.0545,
+      "step": 3070
+    },
+    {
+      "epoch": 0.6436805701110878,
+      "grad_norm": 0.05732462555170059,
+      "learning_rate": 2.9800752878514903e-06,
+      "loss": 0.0564,
+      "step": 3071
+    },
+    {
+      "epoch": 0.6438901697757283,
+      "grad_norm": 0.049768634140491486,
+      "learning_rate": 2.9769702630865814e-06,
+      "loss": 0.0524,
+      "step": 3072
+    },
+    {
+      "epoch": 0.6440997694403688,
+      "grad_norm": 0.0468343161046505,
+      "learning_rate": 2.973866170937959e-06,
+      "loss": 0.0559,
+      "step": 3073
+    },
+    {
+      "epoch": 0.6443093691050095,
+      "grad_norm": 0.05665425956249237,
+      "learning_rate": 2.9707630128366137e-06,
+      "loss": 0.0539,
+      "step": 3074
+    },
+    {
+      "epoch": 0.64451896876965,
+      "grad_norm": 0.04687637463212013,
+      "learning_rate": 2.9676607902130972e-06,
+      "loss": 0.0534,
+      "step": 3075
+    },
+    {
+      "epoch": 0.6447285684342905,
+      "grad_norm": 0.04909089207649231,
+      "learning_rate": 2.9645595044975328e-06,
+      "loss": 0.0555,
+      "step": 3076
+    },
+    {
+      "epoch": 0.644938168098931,
+      "grad_norm": 0.046408187597990036,
+      "learning_rate": 2.961459157119615e-06,
+      "loss": 0.0535,
+      "step": 3077
+    },
+    {
+      "epoch": 0.6451477677635716,
+      "grad_norm": 0.04547589272260666,
+      "learning_rate": 2.958359749508603e-06,
+      "loss": 0.0563,
+      "step": 3078
+    },
+    {
+      "epoch": 0.6453573674282121,
+      "grad_norm": 0.04302079603075981,
+      "learning_rate": 2.9552612830933216e-06,
+      "loss": 0.0552,
+      "step": 3079
+    },
+    {
+      "epoch": 0.6455669670928527,
+      "grad_norm": 0.0457628108561039,
+      "learning_rate": 2.9521637593021636e-06,
+      "loss": 0.0536,
+      "step": 3080
+    },
+    {
+      "epoch": 0.6457765667574932,
+      "grad_norm": 0.04492702707648277,
+      "learning_rate": 2.9490671795630884e-06,
+      "loss": 0.0543,
+      "step": 3081
+    },
+    {
+      "epoch": 0.6459861664221337,
+      "grad_norm": 0.04878842458128929,
+      "learning_rate": 2.9459715453036163e-06,
+      "loss": 0.0547,
+      "step": 3082
+    },
+    {
+      "epoch": 0.6461957660867743,
+      "grad_norm": 0.052640918642282486,
+      "learning_rate": 2.9428768579508347e-06,
+      "loss": 0.0546,
+      "step": 3083
+    },
+    {
+      "epoch": 0.6464053657514148,
+      "grad_norm": 0.05040021613240242,
+      "learning_rate": 2.9397831189313963e-06,
+      "loss": 0.0541,
+      "step": 3084
+    },
+    {
+      "epoch": 0.6466149654160553,
+      "grad_norm": 0.04280049726366997,
+      "learning_rate": 2.936690329671511e-06,
+      "loss": 0.0556,
+      "step": 3085
+    },
+    {
+      "epoch": 0.6468245650806959,
+      "grad_norm": 0.03978101909160614,
+      "learning_rate": 2.933598491596954e-06,
+      "loss": 0.0547,
+      "step": 3086
+    },
+    {
+      "epoch": 0.6470341647453364,
+      "grad_norm": 0.03757026419043541,
+      "learning_rate": 2.930507606133065e-06,
+      "loss": 0.0551,
+      "step": 3087
+    },
+    {
+      "epoch": 0.647243764409977,
+      "grad_norm": 0.039234839379787445,
+      "learning_rate": 2.927417674704739e-06,
+      "loss": 0.0582,
+      "step": 3088
+    },
+    {
+      "epoch": 0.6474533640746175,
+      "grad_norm": 0.03372887521982193,
+      "learning_rate": 2.924328698736434e-06,
+      "loss": 0.0578,
+      "step": 3089
+    },
+    {
+      "epoch": 0.647662963739258,
+      "grad_norm": 0.04041403532028198,
+      "learning_rate": 2.9212406796521675e-06,
+      "loss": 0.0554,
+      "step": 3090
+    },
+    {
+      "epoch": 0.6478725634038985,
+      "grad_norm": 0.04249253869056702,
+      "learning_rate": 2.9181536188755178e-06,
+      "loss": 0.0546,
+      "step": 3091
+    },
+    {
+      "epoch": 0.648082163068539,
+      "grad_norm": 0.04245641827583313,
+      "learning_rate": 2.915067517829615e-06,
+      "loss": 0.0551,
+      "step": 3092
+    },
+    {
+      "epoch": 0.6482917627331797,
+      "grad_norm": 0.044679149985313416,
+      "learning_rate": 2.9119823779371536e-06,
+      "loss": 0.0552,
+      "step": 3093
+    },
+    {
+      "epoch": 0.6485013623978202,
+      "grad_norm": 0.03879356011748314,
+      "learning_rate": 2.9088982006203835e-06,
+      "loss": 0.0526,
+      "step": 3094
+    },
+    {
+      "epoch": 0.6487109620624607,
+      "grad_norm": 0.04242287203669548,
+      "learning_rate": 2.90581498730111e-06,
+      "loss": 0.0546,
+      "step": 3095
+    },
+    {
+      "epoch": 0.6489205617271012,
+      "grad_norm": 0.05261940881609917,
+      "learning_rate": 2.902732739400689e-06,
+      "loss": 0.0519,
+      "step": 3096
+    },
+    {
+      "epoch": 0.6491301613917417,
+      "grad_norm": 0.054223619401454926,
+      "learning_rate": 2.8996514583400425e-06,
+      "loss": 0.0544,
+      "step": 3097
+    },
+    {
+      "epoch": 0.6493397610563824,
+      "grad_norm": 0.04815622419118881,
+      "learning_rate": 2.8965711455396343e-06,
+      "loss": 0.0534,
+      "step": 3098
+    },
+    {
+      "epoch": 0.6495493607210229,
+      "grad_norm": 0.0753355398774147,
+      "learning_rate": 2.893491802419492e-06,
+      "loss": 0.0551,
+      "step": 3099
+    },
+    {
+      "epoch": 0.6497589603856634,
+      "grad_norm": 0.07607095688581467,
+      "learning_rate": 2.8904134303991928e-06,
+      "loss": 0.0535,
+      "step": 3100
+    },
+    {
+      "epoch": 0.6499685600503039,
+      "grad_norm": 0.05531083047389984,
+      "learning_rate": 2.887336030897864e-06,
+      "loss": 0.0524,
+      "step": 3101
+    },
+    {
+      "epoch": 0.6501781597149444,
+      "grad_norm": 0.05754781886935234,
+      "learning_rate": 2.884259605334184e-06,
+      "loss": 0.0536,
+      "step": 3102
+    },
+    {
+      "epoch": 0.650387759379585,
+      "grad_norm": 0.05980090796947479,
+      "learning_rate": 2.8811841551263873e-06,
+      "loss": 0.0591,
+      "step": 3103
+    },
+    {
+      "epoch": 0.6505973590442256,
+      "grad_norm": 0.05938847362995148,
+      "learning_rate": 2.878109681692256e-06,
+      "loss": 0.0561,
+      "step": 3104
+    },
+    {
+      "epoch": 0.6508069587088661,
+      "grad_norm": 0.06025892496109009,
+      "learning_rate": 2.8750361864491195e-06,
+      "loss": 0.0583,
+      "step": 3105
+    },
+    {
+      "epoch": 0.6510165583735066,
+      "grad_norm": 0.0442025326192379,
+      "learning_rate": 2.871963670813861e-06,
+      "loss": 0.057,
+      "step": 3106
+    },
+    {
+      "epoch": 0.6512261580381471,
+      "grad_norm": 0.05523305386304855,
+      "learning_rate": 2.868892136202909e-06,
+      "loss": 0.0543,
+      "step": 3107
+    },
+    {
+      "epoch": 0.6514357577027877,
+      "grad_norm": 0.05948861315846443,
+      "learning_rate": 2.865821584032237e-06,
+      "loss": 0.0551,
+      "step": 3108
+    },
+    {
+      "epoch": 0.6516453573674282,
+      "grad_norm": 0.05912363529205322,
+      "learning_rate": 2.862752015717371e-06,
+      "loss": 0.0561,
+      "step": 3109
+    },
+    {
+      "epoch": 0.6518549570320687,
+      "grad_norm": 0.05048060417175293,
+      "learning_rate": 2.859683432673385e-06,
+      "loss": 0.056,
+      "step": 3110
+    },
+    {
+      "epoch": 0.6520645566967093,
+      "grad_norm": 0.07560472190380096,
+      "learning_rate": 2.856615836314889e-06,
+      "loss": 0.0548,
+      "step": 3111
+    },
+    {
+      "epoch": 0.6522741563613498,
+      "grad_norm": 0.07296677678823471,
+      "learning_rate": 2.8535492280560487e-06,
+      "loss": 0.055,
+      "step": 3112
+    },
+    {
+      "epoch": 0.6524837560259904,
+      "grad_norm": 0.0519208163022995,
+      "learning_rate": 2.850483609310567e-06,
+      "loss": 0.0532,
+      "step": 3113
+    },
+    {
+      "epoch": 0.6526933556906309,
+      "grad_norm": 0.05058110132813454,
+      "learning_rate": 2.8474189814916973e-06,
+      "loss": 0.0544,
+      "step": 3114
+    },
+    {
+      "epoch": 0.6529029553552714,
+      "grad_norm": 0.06586900353431702,
+      "learning_rate": 2.844355346012228e-06,
+      "loss": 0.0592,
+      "step": 3115
+    },
+    {
+      "epoch": 0.653112555019912,
+      "grad_norm": 0.056768354028463364,
+      "learning_rate": 2.8412927042844985e-06,
+      "loss": 0.0548,
+      "step": 3116
+    },
+    {
+      "epoch": 0.6533221546845525,
+      "grad_norm": 0.04887418821454048,
+      "learning_rate": 2.838231057720383e-06,
+      "loss": 0.053,
+      "step": 3117
+    },
+    {
+      "epoch": 0.6535317543491931,
+      "grad_norm": 0.06352341920137405,
+      "learning_rate": 2.8351704077312998e-06,
+      "loss": 0.0511,
+      "step": 3118
+    },
+    {
+      "epoch": 0.6537413540138336,
+      "grad_norm": 0.0803912952542305,
+      "learning_rate": 2.8321107557282083e-06,
+      "loss": 0.054,
+      "step": 3119
+    },
+    {
+      "epoch": 0.6539509536784741,
+      "grad_norm": 0.052244700491428375,
+      "learning_rate": 2.829052103121611e-06,
+      "loss": 0.0523,
+      "step": 3120
+    },
+    {
+      "epoch": 0.6541605533431146,
+      "grad_norm": 0.04686099290847778,
+      "learning_rate": 2.8259944513215405e-06,
+      "loss": 0.0567,
+      "step": 3121
+    },
+    {
+      "epoch": 0.6543701530077551,
+      "grad_norm": 0.07411143183708191,
+      "learning_rate": 2.8229378017375784e-06,
+      "loss": 0.055,
+      "step": 3122
+    },
+    {
+      "epoch": 0.6545797526723958,
+      "grad_norm": 0.06782218813896179,
+      "learning_rate": 2.819882155778836e-06,
+      "loss": 0.0541,
+      "step": 3123
+    },
+    {
+      "epoch": 0.6547893523370363,
+      "grad_norm": 0.04807867109775543,
+      "learning_rate": 2.816827514853968e-06,
+      "loss": 0.0563,
+      "step": 3124
+    },
+    {
+      "epoch": 0.6549989520016768,
+      "grad_norm": 0.047602880746126175,
+      "learning_rate": 2.8137738803711607e-06,
+      "loss": 0.0538,
+      "step": 3125
+    },
+    {
+      "epoch": 0.6552085516663173,
+      "grad_norm": 0.05647674947977066,
+      "learning_rate": 2.81072125373814e-06,
+      "loss": 0.0552,
+      "step": 3126
+    },
+    {
+      "epoch": 0.6554181513309578,
+      "grad_norm": 0.04610821232199669,
+      "learning_rate": 2.807669636362169e-06,
+      "loss": 0.0572,
+      "step": 3127
+    },
+    {
+      "epoch": 0.6556277509955984,
+      "grad_norm": 0.04513276740908623,
+      "learning_rate": 2.8046190296500407e-06,
+      "loss": 0.0562,
+      "step": 3128
+    },
+    {
+      "epoch": 0.655837350660239,
+      "grad_norm": 0.06186923012137413,
+      "learning_rate": 2.8015694350080813e-06,
+      "loss": 0.0527,
+      "step": 3129
+    },
+    {
+      "epoch": 0.6560469503248795,
+      "grad_norm": 0.04966472089290619,
+      "learning_rate": 2.7985208538421576e-06,
+      "loss": 0.0564,
+      "step": 3130
+    },
+    {
+      "epoch": 0.65625654998952,
+      "grad_norm": 0.04229920357465744,
+      "learning_rate": 2.7954732875576606e-06,
+      "loss": 0.0569,
+      "step": 3131
+    },
+    {
+      "epoch": 0.6564661496541605,
+      "grad_norm": 0.0393625944852829,
+      "learning_rate": 2.792426737559521e-06,
+      "loss": 0.0555,
+      "step": 3132
+    },
+    {
+      "epoch": 0.6566757493188011,
+      "grad_norm": 0.04377404600381851,
+      "learning_rate": 2.7893812052521984e-06,
+      "loss": 0.0534,
+      "step": 3133
+    },
+    {
+      "epoch": 0.6568853489834416,
+      "grad_norm": 0.04195486381649971,
+      "learning_rate": 2.7863366920396805e-06,
+      "loss": 0.0553,
+      "step": 3134
+    },
+    {
+      "epoch": 0.6570949486480822,
+      "grad_norm": 0.03995470330119133,
+      "learning_rate": 2.7832931993254865e-06,
+      "loss": 0.0516,
+      "step": 3135
+    },
+    {
+      "epoch": 0.6573045483127227,
+      "grad_norm": 0.041193921118974686,
+      "learning_rate": 2.7802507285126666e-06,
+      "loss": 0.0564,
+      "step": 3136
+    },
+    {
+      "epoch": 0.6575141479773632,
+      "grad_norm": 0.04172206670045853,
+      "learning_rate": 2.7772092810038027e-06,
+      "loss": 0.052,
+      "step": 3137
+    },
+    {
+      "epoch": 0.6577237476420038,
+      "grad_norm": 0.040663424879312515,
+      "learning_rate": 2.774168858200996e-06,
+      "loss": 0.0556,
+      "step": 3138
+    },
+    {
+      "epoch": 0.6579333473066443,
+      "grad_norm": 0.03617934137582779,
+      "learning_rate": 2.771129461505885e-06,
+      "loss": 0.0541,
+      "step": 3139
+    },
+    {
+      "epoch": 0.6581429469712848,
+      "grad_norm": 0.04274788871407509,
+      "learning_rate": 2.7680910923196293e-06,
+      "loss": 0.061,
+      "step": 3140
+    },
+    {
+      "epoch": 0.6583525466359254,
+      "grad_norm": 0.040159981697797775,
+      "learning_rate": 2.765053752042915e-06,
+      "loss": 0.0571,
+      "step": 3141
+    },
+    {
+      "epoch": 0.6585621463005659,
+      "grad_norm": 0.04021866247057915,
+      "learning_rate": 2.762017442075956e-06,
+      "loss": 0.0527,
+      "step": 3142
+    },
+    {
+      "epoch": 0.6587717459652065,
+      "grad_norm": 0.036748629063367844,
+      "learning_rate": 2.7589821638184942e-06,
+      "loss": 0.0545,
+      "step": 3143
+    },
+    {
+      "epoch": 0.658981345629847,
+      "grad_norm": 0.047550056129693985,
+      "learning_rate": 2.7559479186697868e-06,
+      "loss": 0.0544,
+      "step": 3144
+    },
+    {
+      "epoch": 0.6591909452944875,
+      "grad_norm": 0.05852500721812248,
+      "learning_rate": 2.7529147080286246e-06,
+      "loss": 0.0522,
+      "step": 3145
+    },
+    {
+      "epoch": 0.659400544959128,
+      "grad_norm": 0.05024390295147896,
+      "learning_rate": 2.7498825332933154e-06,
+      "loss": 0.0573,
+      "step": 3146
+    },
+    {
+      "epoch": 0.6596101446237687,
+      "grad_norm": 0.03166329488158226,
+      "learning_rate": 2.7468513958616895e-06,
+      "loss": 0.0534,
+      "step": 3147
+    },
+    {
+      "epoch": 0.6598197442884092,
+      "grad_norm": 0.03988707438111305,
+      "learning_rate": 2.7438212971311016e-06,
+      "loss": 0.0526,
+      "step": 3148
+    },
+    {
+      "epoch": 0.6600293439530497,
+      "grad_norm": 0.0417884886264801,
+      "learning_rate": 2.740792238498427e-06,
+      "loss": 0.0536,
+      "step": 3149
+    },
+    {
+      "epoch": 0.6602389436176902,
+      "grad_norm": 0.04264437034726143,
+      "learning_rate": 2.7377642213600652e-06,
+      "loss": 0.0557,
+      "step": 3150
+    },
+    {
+      "epoch": 0.6604485432823307,
+      "grad_norm": 0.039088521152734756,
+      "learning_rate": 2.734737247111924e-06,
+      "loss": 0.0562,
+      "step": 3151
+    },
+    {
+      "epoch": 0.6606581429469713,
+      "grad_norm": 0.0382147878408432,
+      "learning_rate": 2.7317113171494407e-06,
+      "loss": 0.054,
+      "step": 3152
+    },
+    {
+      "epoch": 0.6608677426116119,
+      "grad_norm": 0.05300328880548477,
+      "learning_rate": 2.728686432867571e-06,
+      "loss": 0.0522,
+      "step": 3153
+    },
+    {
+      "epoch": 0.6610773422762524,
+      "grad_norm": 0.043220195919275284,
+      "learning_rate": 2.7256625956607822e-06,
+      "loss": 0.0544,
+      "step": 3154
+    },
+    {
+      "epoch": 0.6612869419408929,
+      "grad_norm": 0.034808751195669174,
+      "learning_rate": 2.722639806923066e-06,
+      "loss": 0.0513,
+      "step": 3155
+    },
+    {
+      "epoch": 0.6614965416055334,
+      "grad_norm": 0.045243218541145325,
+      "learning_rate": 2.719618068047927e-06,
+      "loss": 0.0548,
+      "step": 3156
+    },
+    {
+      "epoch": 0.661706141270174,
+      "grad_norm": 0.0599578395485878,
+      "learning_rate": 2.716597380428382e-06,
+      "loss": 0.052,
+      "step": 3157
+    },
+    {
+      "epoch": 0.6619157409348145,
+      "grad_norm": 0.0630265325307846,
+      "learning_rate": 2.7135777454569703e-06,
+      "loss": 0.0534,
+      "step": 3158
+    },
+    {
+      "epoch": 0.662125340599455,
+      "grad_norm": 0.04426709935069084,
+      "learning_rate": 2.7105591645257432e-06,
+      "loss": 0.0565,
+      "step": 3159
+    },
+    {
+      "epoch": 0.6623349402640956,
+      "grad_norm": 0.030477387830615044,
+      "learning_rate": 2.7075416390262676e-06,
+      "loss": 0.0565,
+      "step": 3160
+    },
+    {
+      "epoch": 0.6625445399287361,
+      "grad_norm": 0.04804522171616554,
+      "learning_rate": 2.70452517034962e-06,
+      "loss": 0.0542,
+      "step": 3161
+    },
+    {
+      "epoch": 0.6627541395933767,
+      "grad_norm": 0.06287235021591187,
+      "learning_rate": 2.7015097598863906e-06,
+      "loss": 0.052,
+      "step": 3162
+    },
+    {
+      "epoch": 0.6629637392580172,
+      "grad_norm": 0.04516349360346794,
+      "learning_rate": 2.6984954090266856e-06,
+      "loss": 0.0546,
+      "step": 3163
+    },
+    {
+      "epoch": 0.6631733389226577,
+      "grad_norm": 0.03801724314689636,
+      "learning_rate": 2.6954821191601175e-06,
+      "loss": 0.0541,
+      "step": 3164
+    },
+    {
+      "epoch": 0.6633829385872982,
+      "grad_norm": 0.07248124480247498,
+      "learning_rate": 2.692469891675814e-06,
+      "loss": 0.0553,
+      "step": 3165
+    },
+    {
+      "epoch": 0.6635925382519388,
+      "grad_norm": 0.06941161304712296,
+      "learning_rate": 2.689458727962413e-06,
+      "loss": 0.0552,
+      "step": 3166
+    },
+    {
+      "epoch": 0.6638021379165794,
+      "grad_norm": 0.042739253491163254,
+      "learning_rate": 2.6864486294080585e-06,
+      "loss": 0.0544,
+      "step": 3167
+    },
+    {
+      "epoch": 0.6640117375812199,
+      "grad_norm": 0.05017360299825668,
+      "learning_rate": 2.683439597400403e-06,
+      "loss": 0.0566,
+      "step": 3168
+    },
+    {
+      "epoch": 0.6642213372458604,
+      "grad_norm": 0.07375095039606094,
+      "learning_rate": 2.680431633326614e-06,
+      "loss": 0.0569,
+      "step": 3169
+    },
+    {
+      "epoch": 0.6644309369105009,
+      "grad_norm": 0.0569414347410202,
+      "learning_rate": 2.677424738573359e-06,
+      "loss": 0.0524,
+      "step": 3170
+    },
+    {
+      "epoch": 0.6646405365751414,
+      "grad_norm": 0.041695497930049896,
+      "learning_rate": 2.6744189145268155e-06,
+      "loss": 0.0548,
+      "step": 3171
+    },
+    {
+      "epoch": 0.6648501362397821,
+      "grad_norm": 0.06951986998319626,
+      "learning_rate": 2.6714141625726725e-06,
+      "loss": 0.0538,
+      "step": 3172
+    },
+    {
+      "epoch": 0.6650597359044226,
+      "grad_norm": 0.06547250598669052,
+      "learning_rate": 2.6684104840961167e-06,
+      "loss": 0.0563,
+      "step": 3173
+    },
+    {
+      "epoch": 0.6652693355690631,
+      "grad_norm": 0.041869353502988815,
+      "learning_rate": 2.665407880481841e-06,
+      "loss": 0.0529,
+      "step": 3174
+    },
+    {
+      "epoch": 0.6654789352337036,
+      "grad_norm": 0.05186137557029724,
+      "learning_rate": 2.6624063531140477e-06,
+      "loss": 0.0539,
+      "step": 3175
+    },
+    {
+      "epoch": 0.6656885348983441,
+      "grad_norm": 0.06323316693305969,
+      "learning_rate": 2.659405903376442e-06,
+      "loss": 0.0561,
+      "step": 3176
+    },
+    {
+      "epoch": 0.6658981345629847,
+      "grad_norm": 0.05578729510307312,
+      "learning_rate": 2.656406532652227e-06,
+      "loss": 0.0543,
+      "step": 3177
+    },
+    {
+      "epoch": 0.6661077342276253,
+      "grad_norm": 0.04175778478384018,
+      "learning_rate": 2.6534082423241154e-06,
+      "loss": 0.0538,
+      "step": 3178
+    },
+    {
+      "epoch": 0.6663173338922658,
+      "grad_norm": 0.039774637669324875,
+      "learning_rate": 2.6504110337743166e-06,
+      "loss": 0.0528,
+      "step": 3179
+    },
+    {
+      "epoch": 0.6665269335569063,
+      "grad_norm": 0.04265711456537247,
+      "learning_rate": 2.6474149083845412e-06,
+      "loss": 0.0546,
+      "step": 3180
+    },
+    {
+      "epoch": 0.6667365332215468,
+      "grad_norm": 0.04271739721298218,
+      "learning_rate": 2.6444198675360044e-06,
+      "loss": 0.055,
+      "step": 3181
+    },
+    {
+      "epoch": 0.6669461328861874,
+      "grad_norm": 0.037498895078897476,
+      "learning_rate": 2.6414259126094192e-06,
+      "loss": 0.0567,
+      "step": 3182
+    },
+    {
+      "epoch": 0.6671557325508279,
+      "grad_norm": 0.040308877825737,
+      "learning_rate": 2.6384330449850028e-06,
+      "loss": 0.0524,
+      "step": 3183
+    },
+    {
+      "epoch": 0.6673653322154685,
+      "grad_norm": 0.04669109731912613,
+      "learning_rate": 2.6354412660424577e-06,
+      "loss": 0.0537,
+      "step": 3184
+    },
+    {
+      "epoch": 0.667574931880109,
+      "grad_norm": 0.03690009191632271,
+      "learning_rate": 2.6324505771609976e-06,
+      "loss": 0.0583,
+      "step": 3185
+    },
+    {
+      "epoch": 0.6677845315447495,
+      "grad_norm": 0.04083159938454628,
+      "learning_rate": 2.6294609797193326e-06,
+      "loss": 0.0584,
+      "step": 3186
+    },
+    {
+      "epoch": 0.6679941312093901,
+      "grad_norm": 0.041967883706092834,
+      "learning_rate": 2.6264724750956605e-06,
+      "loss": 0.055,
+      "step": 3187
+    },
+    {
+      "epoch": 0.6682037308740306,
+      "grad_norm": 0.034479230642318726,
+      "learning_rate": 2.623485064667687e-06,
+      "loss": 0.0565,
+      "step": 3188
+    },
+    {
+      "epoch": 0.6684133305386711,
+      "grad_norm": 0.04071040078997612,
+      "learning_rate": 2.6204987498126046e-06,
+      "loss": 0.0576,
+      "step": 3189
+    },
+    {
+      "epoch": 0.6686229302033116,
+      "grad_norm": 0.04605560004711151,
+      "learning_rate": 2.617513531907103e-06,
+      "loss": 0.0516,
+      "step": 3190
+    },
+    {
+      "epoch": 0.6688325298679522,
+      "grad_norm": 0.042507972568273544,
+      "learning_rate": 2.6145294123273677e-06,
+      "loss": 0.0521,
+      "step": 3191
+    },
+    {
+      "epoch": 0.6690421295325928,
+      "grad_norm": 0.03638184443116188,
+      "learning_rate": 2.6115463924490796e-06,
+      "loss": 0.0575,
+      "step": 3192
+    },
+    {
+      "epoch": 0.6692517291972333,
+      "grad_norm": 0.035932447761297226,
+      "learning_rate": 2.608564473647407e-06,
+      "loss": 0.0562,
+      "step": 3193
+    },
+    {
+      "epoch": 0.6694613288618738,
+      "grad_norm": 0.0350482314825058,
+      "learning_rate": 2.605583657297017e-06,
+      "loss": 0.0577,
+      "step": 3194
+    },
+    {
+      "epoch": 0.6696709285265143,
+      "grad_norm": 0.03746052831411362,
+      "learning_rate": 2.602603944772062e-06,
+      "loss": 0.0553,
+      "step": 3195
+    },
+    {
+      "epoch": 0.6698805281911548,
+      "grad_norm": 0.031886544078588486,
+      "learning_rate": 2.5996253374461924e-06,
+      "loss": 0.0555,
+      "step": 3196
+    },
+    {
+      "epoch": 0.6700901278557955,
+      "grad_norm": 0.03672702610492706,
+      "learning_rate": 2.5966478366925406e-06,
+      "loss": 0.055,
+      "step": 3197
+    },
+    {
+      "epoch": 0.670299727520436,
+      "grad_norm": 0.03667857125401497,
+      "learning_rate": 2.593671443883738e-06,
+      "loss": 0.0557,
+      "step": 3198
+    },
+    {
+      "epoch": 0.6705093271850765,
+      "grad_norm": 0.03002052754163742,
+      "learning_rate": 2.590696160391901e-06,
+      "loss": 0.0557,
+      "step": 3199
+    },
+    {
+      "epoch": 0.670718926849717,
+      "grad_norm": 0.034151580184698105,
+      "learning_rate": 2.587721987588635e-06,
+      "loss": 0.0556,
+      "step": 3200
+    },
+    {
+      "epoch": 0.6709285265143575,
+      "grad_norm": 0.037925586104393005,
+      "learning_rate": 2.5847489268450287e-06,
+      "loss": 0.0564,
+      "step": 3201
+    },
+    {
+      "epoch": 0.6711381261789982,
+      "grad_norm": 0.0316159762442112,
+      "learning_rate": 2.5817769795316674e-06,
+      "loss": 0.0528,
+      "step": 3202
+    },
+    {
+      "epoch": 0.6713477258436387,
+      "grad_norm": 0.03222265467047691,
+      "learning_rate": 2.578806147018614e-06,
+      "loss": 0.0547,
+      "step": 3203
+    },
+    {
+      "epoch": 0.6715573255082792,
+      "grad_norm": 0.0310065895318985,
+      "learning_rate": 2.5758364306754247e-06,
+      "loss": 0.0515,
+      "step": 3204
+    },
+    {
+      "epoch": 0.6717669251729197,
+      "grad_norm": 0.03238251060247421,
+      "learning_rate": 2.5728678318711385e-06,
+      "loss": 0.058,
+      "step": 3205
+    },
+    {
+      "epoch": 0.6719765248375602,
+      "grad_norm": 0.026389988139271736,
+      "learning_rate": 2.5699003519742783e-06,
+      "loss": 0.0564,
+      "step": 3206
+    },
+    {
+      "epoch": 0.6721861245022008,
+      "grad_norm": 0.028875682502985,
+      "learning_rate": 2.566933992352849e-06,
+      "loss": 0.0546,
+      "step": 3207
+    },
+    {
+      "epoch": 0.6723957241668413,
+      "grad_norm": 0.03320210054516792,
+      "learning_rate": 2.563968754374344e-06,
+      "loss": 0.0556,
+      "step": 3208
+    },
+    {
+      "epoch": 0.6726053238314819,
+      "grad_norm": 0.02969476394355297,
+      "learning_rate": 2.5610046394057386e-06,
+      "loss": 0.0555,
+      "step": 3209
+    },
+    {
+      "epoch": 0.6728149234961224,
+      "grad_norm": 0.029616905376315117,
+      "learning_rate": 2.5580416488134864e-06,
+      "loss": 0.056,
+      "step": 3210
+    },
+    {
+      "epoch": 0.6730245231607629,
+      "grad_norm": 0.023642316460609436,
+      "learning_rate": 2.5550797839635283e-06,
+      "loss": 0.0539,
+      "step": 3211
+    },
+    {
+      "epoch": 0.6732341228254035,
+      "grad_norm": 0.028511328622698784,
+      "learning_rate": 2.552119046221282e-06,
+      "loss": 0.0526,
+      "step": 3212
+    },
+    {
+      "epoch": 0.673443722490044,
+      "grad_norm": 0.030537253245711327,
+      "learning_rate": 2.5491594369516452e-06,
+      "loss": 0.0572,
+      "step": 3213
+    },
+    {
+      "epoch": 0.6736533221546845,
+      "grad_norm": 0.029667936265468597,
+      "learning_rate": 2.546200957518999e-06,
+      "loss": 0.0561,
+      "step": 3214
+    },
+    {
+      "epoch": 0.6738629218193251,
+      "grad_norm": 0.025500137358903885,
+      "learning_rate": 2.5432436092872036e-06,
+      "loss": 0.0569,
+      "step": 3215
+    },
+    {
+      "epoch": 0.6740725214839657,
+      "grad_norm": 0.024678369984030724,
+      "learning_rate": 2.5402873936195914e-06,
+      "loss": 0.0559,
+      "step": 3216
+    },
+    {
+      "epoch": 0.6742821211486062,
+      "grad_norm": 0.031020864844322205,
+      "learning_rate": 2.537332311878983e-06,
+      "loss": 0.0552,
+      "step": 3217
+    },
+    {
+      "epoch": 0.6744917208132467,
+      "grad_norm": 0.034975748509168625,
+      "learning_rate": 2.5343783654276644e-06,
+      "loss": 0.0541,
+      "step": 3218
+    },
+    {
+      "epoch": 0.6747013204778872,
+      "grad_norm": 0.03894275054335594,
+      "learning_rate": 2.5314255556274092e-06,
+      "loss": 0.0545,
+      "step": 3219
+    },
+    {
+      "epoch": 0.6749109201425277,
+      "grad_norm": 0.029810387641191483,
+      "learning_rate": 2.5284738838394586e-06,
+      "loss": 0.0543,
+      "step": 3220
+    },
+    {
+      "epoch": 0.6751205198071684,
+      "grad_norm": 0.02972550317645073,
+      "learning_rate": 2.5255233514245358e-06,
+      "loss": 0.0532,
+      "step": 3221
+    },
+    {
+      "epoch": 0.6753301194718089,
+      "grad_norm": 0.02595905028283596,
+      "learning_rate": 2.5225739597428334e-06,
+      "loss": 0.0548,
+      "step": 3222
+    },
+    {
+      "epoch": 0.6755397191364494,
+      "grad_norm": 0.025289386510849,
+      "learning_rate": 2.51962571015402e-06,
+      "loss": 0.0532,
+      "step": 3223
+    },
+    {
+      "epoch": 0.6757493188010899,
+      "grad_norm": 0.028770189732313156,
+      "learning_rate": 2.5166786040172387e-06,
+      "loss": 0.0516,
+      "step": 3224
+    },
+    {
+      "epoch": 0.6759589184657304,
+      "grad_norm": 0.024205055087804794,
+      "learning_rate": 2.5137326426911067e-06,
+      "loss": 0.052,
+      "step": 3225
+    },
+    {
+      "epoch": 0.676168518130371,
+      "grad_norm": 0.02634621225297451,
+      "learning_rate": 2.5107878275337084e-06,
+      "loss": 0.0547,
+      "step": 3226
+    },
+    {
+      "epoch": 0.6763781177950116,
+      "grad_norm": 0.031177837401628494,
+      "learning_rate": 2.507844159902606e-06,
+      "loss": 0.0581,
+      "step": 3227
+    },
+    {
+      "epoch": 0.6765877174596521,
+      "grad_norm": 0.032176099717617035,
+      "learning_rate": 2.5049016411548273e-06,
+      "loss": 0.0547,
+      "step": 3228
+    },
+    {
+      "epoch": 0.6767973171242926,
+      "grad_norm": 0.03147950395941734,
+      "learning_rate": 2.501960272646875e-06,
+      "loss": 0.0561,
+      "step": 3229
+    },
+    {
+      "epoch": 0.6770069167889331,
+      "grad_norm": 0.02897576615214348,
+      "learning_rate": 2.499020055734716e-06,
+      "loss": 0.0564,
+      "step": 3230
+    },
+    {
+      "epoch": 0.6772165164535737,
+      "grad_norm": 0.03353496268391609,
+      "learning_rate": 2.496080991773792e-06,
+      "loss": 0.0531,
+      "step": 3231
+    },
+    {
+      "epoch": 0.6774261161182142,
+      "grad_norm": 0.03561830520629883,
+      "learning_rate": 2.493143082119013e-06,
+      "loss": 0.0565,
+      "step": 3232
+    },
+    {
+      "epoch": 0.6776357157828548,
+      "grad_norm": 0.027927443385124207,
+      "learning_rate": 2.490206328124752e-06,
+      "loss": 0.0541,
+      "step": 3233
+    },
+    {
+      "epoch": 0.6778453154474953,
+      "grad_norm": 0.037114545702934265,
+      "learning_rate": 2.4872707311448504e-06,
+      "loss": 0.0534,
+      "step": 3234
+    },
+    {
+      "epoch": 0.6780549151121358,
+      "grad_norm": 0.03275038301944733,
+      "learning_rate": 2.484336292532622e-06,
+      "loss": 0.0519,
+      "step": 3235
+    },
+    {
+      "epoch": 0.6782645147767764,
+      "grad_norm": 0.02878388576209545,
+      "learning_rate": 2.481403013640838e-06,
+      "loss": 0.0572,
+      "step": 3236
+    },
+    {
+      "epoch": 0.6784741144414169,
+      "grad_norm": 0.02609802968800068,
+      "learning_rate": 2.478470895821742e-06,
+      "loss": 0.052,
+      "step": 3237
+    },
+    {
+      "epoch": 0.6786837141060574,
+      "grad_norm": 0.024750111624598503,
+      "learning_rate": 2.475539940427041e-06,
+      "loss": 0.0553,
+      "step": 3238
+    },
+    {
+      "epoch": 0.678893313770698,
+      "grad_norm": 0.02313164807856083,
+      "learning_rate": 2.472610148807903e-06,
+      "loss": 0.0561,
+      "step": 3239
+    },
+    {
+      "epoch": 0.6791029134353385,
+      "grad_norm": 0.02588365226984024,
+      "learning_rate": 2.469681522314959e-06,
+      "loss": 0.0566,
+      "step": 3240
+    },
+    {
+      "epoch": 0.6793125130999791,
+      "grad_norm": 0.02386978268623352,
+      "learning_rate": 2.4667540622983083e-06,
+      "loss": 0.053,
+      "step": 3241
+    },
+    {
+      "epoch": 0.6795221127646196,
+      "grad_norm": 0.025168968364596367,
+      "learning_rate": 2.4638277701075103e-06,
+      "loss": 0.0536,
+      "step": 3242
+    },
+    {
+      "epoch": 0.6797317124292601,
+      "grad_norm": 0.02500263787806034,
+      "learning_rate": 2.460902647091582e-06,
+      "loss": 0.0524,
+      "step": 3243
+    },
+    {
+      "epoch": 0.6799413120939006,
+      "grad_norm": 0.02609972469508648,
+      "learning_rate": 2.4579786945990073e-06,
+      "loss": 0.0566,
+      "step": 3244
+    },
+    {
+      "epoch": 0.6801509117585411,
+      "grad_norm": 0.03376591205596924,
+      "learning_rate": 2.4550559139777264e-06,
+      "loss": 0.0542,
+      "step": 3245
+    },
+    {
+      "epoch": 0.6803605114231818,
+      "grad_norm": 0.03760766610503197,
+      "learning_rate": 2.452134306575139e-06,
+      "loss": 0.0535,
+      "step": 3246
+    },
+    {
+      "epoch": 0.6805701110878223,
+      "grad_norm": 0.029737234115600586,
+      "learning_rate": 2.4492138737381066e-06,
+      "loss": 0.0522,
+      "step": 3247
+    },
+    {
+      "epoch": 0.6807797107524628,
+      "grad_norm": 0.030677396804094315,
+      "learning_rate": 2.446294616812951e-06,
+      "loss": 0.0523,
+      "step": 3248
+    },
+    {
+      "epoch": 0.6809893104171033,
+      "grad_norm": 0.05189737677574158,
+      "learning_rate": 2.443376537145444e-06,
+      "loss": 0.0541,
+      "step": 3249
+    },
+    {
+      "epoch": 0.6811989100817438,
+      "grad_norm": 0.0667051374912262,
+      "learning_rate": 2.4404596360808255e-06,
+      "loss": 0.056,
+      "step": 3250
+    },
+    {
+      "epoch": 0.6814085097463844,
+      "grad_norm": 0.05018071457743645,
+      "learning_rate": 2.437543914963782e-06,
+      "loss": 0.0542,
+      "step": 3251
+    },
+    {
+      "epoch": 0.681618109411025,
+      "grad_norm": 0.02675667405128479,
+      "learning_rate": 2.4346293751384597e-06,
+      "loss": 0.0551,
+      "step": 3252
+    },
+    {
+      "epoch": 0.6818277090756655,
+      "grad_norm": 0.04382751137018204,
+      "learning_rate": 2.431716017948462e-06,
+      "loss": 0.0544,
+      "step": 3253
+    },
+    {
+      "epoch": 0.682037308740306,
+      "grad_norm": 0.051815591752529144,
+      "learning_rate": 2.428803844736848e-06,
+      "loss": 0.0546,
+      "step": 3254
+    },
+    {
+      "epoch": 0.6822469084049465,
+      "grad_norm": 0.034245654940605164,
+      "learning_rate": 2.4258928568461303e-06,
+      "loss": 0.0518,
+      "step": 3255
+    },
+    {
+      "epoch": 0.6824565080695871,
+      "grad_norm": 0.029311811551451683,
+      "learning_rate": 2.422983055618267e-06,
+      "loss": 0.0555,
+      "step": 3256
+    },
+    {
+      "epoch": 0.6826661077342276,
+      "grad_norm": 0.03641844168305397,
+      "learning_rate": 2.4200744423946803e-06,
+      "loss": 0.0532,
+      "step": 3257
+    },
+    {
+      "epoch": 0.6828757073988682,
+      "grad_norm": 0.03218969702720642,
+      "learning_rate": 2.4171670185162415e-06,
+      "loss": 0.0543,
+      "step": 3258
+    },
+    {
+      "epoch": 0.6830853070635087,
+      "grad_norm": 0.024815967306494713,
+      "learning_rate": 2.4142607853232687e-06,
+      "loss": 0.0584,
+      "step": 3259
+    },
+    {
+      "epoch": 0.6832949067281492,
+      "grad_norm": 0.025883564725518227,
+      "learning_rate": 2.4113557441555384e-06,
+      "loss": 0.0544,
+      "step": 3260
+    },
+    {
+      "epoch": 0.6835045063927898,
+      "grad_norm": 0.03328178450465202,
+      "learning_rate": 2.4084518963522724e-06,
+      "loss": 0.0541,
+      "step": 3261
+    },
+    {
+      "epoch": 0.6837141060574303,
+      "grad_norm": 0.028745753690600395,
+      "learning_rate": 2.4055492432521428e-06,
+      "loss": 0.0541,
+      "step": 3262
+    },
+    {
+      "epoch": 0.6839237057220708,
+      "grad_norm": 0.023591186851263046,
+      "learning_rate": 2.402647786193272e-06,
+      "loss": 0.0531,
+      "step": 3263
+    },
+    {
+      "epoch": 0.6841333053867114,
+      "grad_norm": 0.026261072605848312,
+      "learning_rate": 2.3997475265132333e-06,
+      "loss": 0.058,
+      "step": 3264
+    },
+    {
+      "epoch": 0.6843429050513519,
+      "grad_norm": 0.023627085611224174,
+      "learning_rate": 2.3968484655490455e-06,
+      "loss": 0.0551,
+      "step": 3265
+    },
+    {
+      "epoch": 0.6845525047159925,
+      "grad_norm": 0.024966632947325706,
+      "learning_rate": 2.3939506046371753e-06,
+      "loss": 0.0562,
+      "step": 3266
+    },
+    {
+      "epoch": 0.684762104380633,
+      "grad_norm": 0.02931971289217472,
+      "learning_rate": 2.391053945113533e-06,
+      "loss": 0.0526,
+      "step": 3267
+    },
+    {
+      "epoch": 0.6849717040452735,
+      "grad_norm": 0.035215772688388824,
+      "learning_rate": 2.388158488313481e-06,
+      "loss": 0.0566,
+      "step": 3268
+    },
+    {
+      "epoch": 0.685181303709914,
+      "grad_norm": 0.02961471863090992,
+      "learning_rate": 2.3852642355718224e-06,
+      "loss": 0.0554,
+      "step": 3269
+    },
+    {
+      "epoch": 0.6853909033745545,
+      "grad_norm": 0.028823746368288994,
+      "learning_rate": 2.3823711882228077e-06,
+      "loss": 0.053,
+      "step": 3270
+    },
+    {
+      "epoch": 0.6856005030391952,
+      "grad_norm": 0.04585007578134537,
+      "learning_rate": 2.379479347600134e-06,
+      "loss": 0.0535,
+      "step": 3271
+    },
+    {
+      "epoch": 0.6858101027038357,
+      "grad_norm": 0.03799903392791748,
+      "learning_rate": 2.3765887150369366e-06,
+      "loss": 0.0529,
+      "step": 3272
+    },
+    {
+      "epoch": 0.6860197023684762,
+      "grad_norm": 0.02885265089571476,
+      "learning_rate": 2.373699291865794e-06,
+      "loss": 0.0559,
+      "step": 3273
+    },
+    {
+      "epoch": 0.6862293020331167,
+      "grad_norm": 0.03915941342711449,
+      "learning_rate": 2.370811079418735e-06,
+      "loss": 0.0528,
+      "step": 3274
+    },
+    {
+      "epoch": 0.6864389016977572,
+      "grad_norm": 0.03368467465043068,
+      "learning_rate": 2.3679240790272203e-06,
+      "loss": 0.0532,
+      "step": 3275
+    },
+    {
+      "epoch": 0.6866485013623979,
+      "grad_norm": 0.03050885535776615,
+      "learning_rate": 2.3650382920221578e-06,
+      "loss": 0.0564,
+      "step": 3276
+    },
+    {
+      "epoch": 0.6868581010270384,
+      "grad_norm": 0.03771447017788887,
+      "learning_rate": 2.3621537197338977e-06,
+      "loss": 0.0533,
+      "step": 3277
+    },
+    {
+      "epoch": 0.6870677006916789,
+      "grad_norm": 0.040800098329782486,
+      "learning_rate": 2.359270363492225e-06,
+      "loss": 0.056,
+      "step": 3278
+    },
+    {
+      "epoch": 0.6872773003563194,
+      "grad_norm": 0.03476854786276817,
+      "learning_rate": 2.356388224626364e-06,
+      "loss": 0.0529,
+      "step": 3279
+    },
+    {
+      "epoch": 0.6874869000209599,
+      "grad_norm": 0.026112914085388184,
+      "learning_rate": 2.3535073044649824e-06,
+      "loss": 0.055,
+      "step": 3280
+    },
+    {
+      "epoch": 0.6876964996856005,
+      "grad_norm": 0.03609404340386391,
+      "learning_rate": 2.350627604336186e-06,
+      "loss": 0.0527,
+      "step": 3281
+    },
+    {
+      "epoch": 0.687906099350241,
+      "grad_norm": 0.039491958916187286,
+      "learning_rate": 2.347749125567511e-06,
+      "loss": 0.054,
+      "step": 3282
+    },
+    {
+      "epoch": 0.6881156990148816,
+      "grad_norm": 0.02687055617570877,
+      "learning_rate": 2.344871869485941e-06,
+      "loss": 0.0556,
+      "step": 3283
+    },
+    {
+      "epoch": 0.6883252986795221,
+      "grad_norm": 0.03269565477967262,
+      "learning_rate": 2.341995837417887e-06,
+      "loss": 0.0552,
+      "step": 3284
+    },
+    {
+      "epoch": 0.6885348983441627,
+      "grad_norm": 0.037071727216243744,
+      "learning_rate": 2.3391210306891977e-06,
+      "loss": 0.0513,
+      "step": 3285
+    },
+    {
+      "epoch": 0.6887444980088032,
+      "grad_norm": 0.03246400132775307,
+      "learning_rate": 2.336247450625161e-06,
+      "loss": 0.0536,
+      "step": 3286
+    },
+    {
+      "epoch": 0.6889540976734437,
+      "grad_norm": 0.02836901880800724,
+      "learning_rate": 2.333375098550496e-06,
+      "loss": 0.0533,
+      "step": 3287
+    },
+    {
+      "epoch": 0.6891636973380842,
+      "grad_norm": 0.04530966654419899,
+      "learning_rate": 2.330503975789361e-06,
+      "loss": 0.0553,
+      "step": 3288
+    },
+    {
+      "epoch": 0.6893732970027248,
+      "grad_norm": 0.05676478147506714,
+      "learning_rate": 2.3276340836653342e-06,
+      "loss": 0.0544,
+      "step": 3289
+    },
+    {
+      "epoch": 0.6895828966673654,
+      "grad_norm": 0.03974361717700958,
+      "learning_rate": 2.3247654235014403e-06,
+      "loss": 0.056,
+      "step": 3290
+    },
+    {
+      "epoch": 0.6897924963320059,
+      "grad_norm": 0.022683048620820045,
+      "learning_rate": 2.321897996620132e-06,
+      "loss": 0.0571,
+      "step": 3291
+    },
+    {
+      "epoch": 0.6900020959966464,
+      "grad_norm": 0.03632507100701332,
+      "learning_rate": 2.3190318043432892e-06,
+      "loss": 0.053,
+      "step": 3292
+    },
+    {
+      "epoch": 0.6902116956612869,
+      "grad_norm": 0.029882922768592834,
+      "learning_rate": 2.3161668479922293e-06,
+      "loss": 0.0551,
+      "step": 3293
+    },
+    {
+      "epoch": 0.6904212953259274,
+      "grad_norm": 0.0291314534842968,
+      "learning_rate": 2.3133031288876955e-06,
+      "loss": 0.0527,
+      "step": 3294
+    },
+    {
+      "epoch": 0.6906308949905681,
+      "grad_norm": 0.03613777831196785,
+      "learning_rate": 2.3104406483498593e-06,
+      "loss": 0.0545,
+      "step": 3295
+    },
+    {
+      "epoch": 0.6908404946552086,
+      "grad_norm": 0.04006670415401459,
+      "learning_rate": 2.3075794076983254e-06,
+      "loss": 0.0551,
+      "step": 3296
+    },
+    {
+      "epoch": 0.6910500943198491,
+      "grad_norm": 0.031675100326538086,
+      "learning_rate": 2.304719408252127e-06,
+      "loss": 0.0566,
+      "step": 3297
+    },
+    {
+      "epoch": 0.6912596939844896,
+      "grad_norm": 0.03170279040932655,
+      "learning_rate": 2.30186065132972e-06,
+      "loss": 0.0545,
+      "step": 3298
+    },
+    {
+      "epoch": 0.6914692936491301,
+      "grad_norm": 0.04302024096250534,
+      "learning_rate": 2.2990031382489935e-06,
+      "loss": 0.055,
+      "step": 3299
+    },
+    {
+      "epoch": 0.6916788933137707,
+      "grad_norm": 0.03898413106799126,
+      "learning_rate": 2.296146870327258e-06,
+      "loss": 0.0566,
+      "step": 3300
+    },
+    {
+      "epoch": 0.6918884929784113,
+      "grad_norm": 0.021424876525998116,
+      "learning_rate": 2.293291848881255e-06,
+      "loss": 0.0572,
+      "step": 3301
+    },
+    {
+      "epoch": 0.6920980926430518,
+      "grad_norm": 0.03237781301140785,
+      "learning_rate": 2.290438075227146e-06,
+      "loss": 0.0543,
+      "step": 3302
+    },
+    {
+      "epoch": 0.6923076923076923,
+      "grad_norm": 0.03710688278079033,
+      "learning_rate": 2.2875855506805217e-06,
+      "loss": 0.0557,
+      "step": 3303
+    },
+    {
+      "epoch": 0.6925172919723328,
+      "grad_norm": 0.030332237482070923,
+      "learning_rate": 2.284734276556396e-06,
+      "loss": 0.056,
+      "step": 3304
+    },
+    {
+      "epoch": 0.6927268916369734,
+      "grad_norm": 0.02461421862244606,
+      "learning_rate": 2.281884254169206e-06,
+      "loss": 0.0517,
+      "step": 3305
+    },
+    {
+      "epoch": 0.6929364913016139,
+      "grad_norm": 0.02470548450946808,
+      "learning_rate": 2.279035484832808e-06,
+      "loss": 0.0545,
+      "step": 3306
+    },
+    {
+      "epoch": 0.6931460909662545,
+      "grad_norm": 0.025578899309039116,
+      "learning_rate": 2.276187969860488e-06,
+      "loss": 0.0556,
+      "step": 3307
+    },
+    {
+      "epoch": 0.693355690630895,
+      "grad_norm": 0.024182526394724846,
+      "learning_rate": 2.2733417105649458e-06,
+      "loss": 0.0549,
+      "step": 3308
+    },
+    {
+      "epoch": 0.6935652902955355,
+      "grad_norm": 0.02719123847782612,
+      "learning_rate": 2.270496708258309e-06,
+      "loss": 0.0541,
+      "step": 3309
+    },
+    {
+      "epoch": 0.6937748899601761,
+      "grad_norm": 0.03566061705350876,
+      "learning_rate": 2.2676529642521244e-06,
+      "loss": 0.0548,
+      "step": 3310
+    },
+    {
+      "epoch": 0.6939844896248166,
+      "grad_norm": 0.036818746477365494,
+      "learning_rate": 2.264810479857356e-06,
+      "loss": 0.0577,
+      "step": 3311
+    },
+    {
+      "epoch": 0.6941940892894571,
+      "grad_norm": 0.03650686517357826,
+      "learning_rate": 2.2619692563843865e-06,
+      "loss": 0.0557,
+      "step": 3312
+    },
+    {
+      "epoch": 0.6944036889540977,
+      "grad_norm": 0.02722245827317238,
+      "learning_rate": 2.259129295143021e-06,
+      "loss": 0.0572,
+      "step": 3313
+    },
+    {
+      "epoch": 0.6946132886187382,
+      "grad_norm": 0.023747239261865616,
+      "learning_rate": 2.2562905974424824e-06,
+      "loss": 0.0539,
+      "step": 3314
+    },
+    {
+      "epoch": 0.6948228882833788,
+      "grad_norm": 0.027470340952277184,
+      "learning_rate": 2.2534531645914078e-06,
+      "loss": 0.0524,
+      "step": 3315
+    },
+    {
+      "epoch": 0.6950324879480193,
+      "grad_norm": 0.02598888985812664,
+      "learning_rate": 2.2506169978978543e-06,
+      "loss": 0.0545,
+      "step": 3316
+    },
+    {
+      "epoch": 0.6952420876126598,
+      "grad_norm": 0.024863192811608315,
+      "learning_rate": 2.2477820986692945e-06,
+      "loss": 0.0519,
+      "step": 3317
+    },
+    {
+      "epoch": 0.6954516872773003,
+      "grad_norm": 0.0245378315448761,
+      "learning_rate": 2.2449484682126133e-06,
+      "loss": 0.0544,
+      "step": 3318
+    },
+    {
+      "epoch": 0.6956612869419408,
+      "grad_norm": 0.02471884898841381,
+      "learning_rate": 2.2421161078341163e-06,
+      "loss": 0.0544,
+      "step": 3319
+    },
+    {
+      "epoch": 0.6958708866065815,
+      "grad_norm": 0.02821512520313263,
+      "learning_rate": 2.2392850188395227e-06,
+      "loss": 0.0583,
+      "step": 3320
+    },
+    {
+      "epoch": 0.696080486271222,
+      "grad_norm": 0.024967767298221588,
+      "learning_rate": 2.23645520253396e-06,
+      "loss": 0.0552,
+      "step": 3321
+    },
+    {
+      "epoch": 0.6962900859358625,
+      "grad_norm": 0.022636888548731804,
+      "learning_rate": 2.2336266602219762e-06,
+      "loss": 0.0556,
+      "step": 3322
+    },
+    {
+      "epoch": 0.696499685600503,
+      "grad_norm": 0.02549467422068119,
+      "learning_rate": 2.230799393207526e-06,
+      "loss": 0.0551,
+      "step": 3323
+    },
+    {
+      "epoch": 0.6967092852651435,
+      "grad_norm": 0.022535137832164764,
+      "learning_rate": 2.227973402793982e-06,
+      "loss": 0.0537,
+      "step": 3324
+    },
+    {
+      "epoch": 0.6969188849297842,
+      "grad_norm": 0.028567228466272354,
+      "learning_rate": 2.225148690284122e-06,
+      "loss": 0.0531,
+      "step": 3325
+    },
+    {
+      "epoch": 0.6971284845944247,
+      "grad_norm": 0.025192799046635628,
+      "learning_rate": 2.2223252569801412e-06,
+      "loss": 0.0575,
+      "step": 3326
+    },
+    {
+      "epoch": 0.6973380842590652,
+      "grad_norm": 0.02626647800207138,
+      "learning_rate": 2.2195031041836396e-06,
+      "loss": 0.0538,
+      "step": 3327
+    },
+    {
+      "epoch": 0.6975476839237057,
+      "grad_norm": 0.026845086365938187,
+      "learning_rate": 2.216682233195628e-06,
+      "loss": 0.0532,
+      "step": 3328
+    },
+    {
+      "epoch": 0.6977572835883462,
+      "grad_norm": 0.027518998831510544,
+      "learning_rate": 2.2138626453165286e-06,
+      "loss": 0.0531,
+      "step": 3329
+    },
+    {
+      "epoch": 0.6979668832529868,
+      "grad_norm": 0.0274403914809227,
+      "learning_rate": 2.2110443418461723e-06,
+      "loss": 0.0507,
+      "step": 3330
+    },
+    {
+      "epoch": 0.6981764829176274,
+      "grad_norm": 0.03426744416356087,
+      "learning_rate": 2.208227324083794e-06,
+      "loss": 0.0554,
+      "step": 3331
+    },
+    {
+      "epoch": 0.6983860825822679,
+      "grad_norm": 0.029531171545386314,
+      "learning_rate": 2.2054115933280407e-06,
+      "loss": 0.0547,
+      "step": 3332
+    },
+    {
+      "epoch": 0.6985956822469084,
+      "grad_norm": 0.02566445991396904,
+      "learning_rate": 2.2025971508769607e-06,
+      "loss": 0.0586,
+      "step": 3333
+    },
+    {
+      "epoch": 0.6988052819115489,
+      "grad_norm": 0.04020659625530243,
+      "learning_rate": 2.199783998028015e-06,
+      "loss": 0.0545,
+      "step": 3334
+    },
+    {
+      "epoch": 0.6990148815761895,
+      "grad_norm": 0.03343269228935242,
+      "learning_rate": 2.1969721360780626e-06,
+      "loss": 0.0539,
+      "step": 3335
+    },
+    {
+      "epoch": 0.69922448124083,
+      "grad_norm": 0.02812224067747593,
+      "learning_rate": 2.1941615663233733e-06,
+      "loss": 0.0545,
+      "step": 3336
+    },
+    {
+      "epoch": 0.6994340809054705,
+      "grad_norm": 0.04151785001158714,
+      "learning_rate": 2.191352290059621e-06,
+      "loss": 0.055,
+      "step": 3337
+    },
+    {
+      "epoch": 0.6996436805701111,
+      "grad_norm": 0.028555648401379585,
+      "learning_rate": 2.1885443085818795e-06,
+      "loss": 0.0527,
+      "step": 3338
+    },
+    {
+      "epoch": 0.6998532802347516,
+      "grad_norm": 0.034111034125089645,
+      "learning_rate": 2.185737623184626e-06,
+      "loss": 0.0539,
+      "step": 3339
+    },
+    {
+      "epoch": 0.7000628798993922,
+      "grad_norm": 0.035869497805833817,
+      "learning_rate": 2.1829322351617456e-06,
+      "loss": 0.0548,
+      "step": 3340
+    },
+    {
+      "epoch": 0.7002724795640327,
+      "grad_norm": 0.029365766793489456,
+      "learning_rate": 2.1801281458065177e-06,
+      "loss": 0.0542,
+      "step": 3341
+    },
+    {
+      "epoch": 0.7004820792286732,
+      "grad_norm": 0.03768596425652504,
+      "learning_rate": 2.1773253564116288e-06,
+      "loss": 0.0529,
+      "step": 3342
+    },
+    {
+      "epoch": 0.7006916788933137,
+      "grad_norm": 0.03441116586327553,
+      "learning_rate": 2.174523868269166e-06,
+      "loss": 0.056,
+      "step": 3343
+    },
+    {
+      "epoch": 0.7009012785579543,
+      "grad_norm": 0.03129161149263382,
+      "learning_rate": 2.171723682670613e-06,
+      "loss": 0.054,
+      "step": 3344
+    },
+    {
+      "epoch": 0.7011108782225949,
+      "grad_norm": 0.03608183562755585,
+      "learning_rate": 2.1689248009068532e-06,
+      "loss": 0.0549,
+      "step": 3345
+    },
+    {
+      "epoch": 0.7013204778872354,
+      "grad_norm": 0.026896623894572258,
+      "learning_rate": 2.166127224268172e-06,
+      "loss": 0.055,
+      "step": 3346
+    },
+    {
+      "epoch": 0.7015300775518759,
+      "grad_norm": 0.03324010595679283,
+      "learning_rate": 2.163330954044253e-06,
+      "loss": 0.0538,
+      "step": 3347
+    },
+    {
+      "epoch": 0.7017396772165164,
+      "grad_norm": 0.02656789869070053,
+      "learning_rate": 2.160535991524174e-06,
+      "loss": 0.0549,
+      "step": 3348
+    },
+    {
+      "epoch": 0.7019492768811569,
+      "grad_norm": 0.029164310544729233,
+      "learning_rate": 2.1577423379964147e-06,
+      "loss": 0.0556,
+      "step": 3349
+    },
+    {
+      "epoch": 0.7021588765457976,
+      "grad_norm": 0.02474060468375683,
+      "learning_rate": 2.154949994748847e-06,
+      "loss": 0.0542,
+      "step": 3350
+    },
+    {
+      "epoch": 0.7023684762104381,
+      "grad_norm": 0.031023385003209114,
+      "learning_rate": 2.152158963068739e-06,
+      "loss": 0.0539,
+      "step": 3351
+    },
+    {
+      "epoch": 0.7025780758750786,
+      "grad_norm": 0.02580900862812996,
+      "learning_rate": 2.149369244242758e-06,
+      "loss": 0.0536,
+      "step": 3352
+    },
+    {
+      "epoch": 0.7027876755397191,
+      "grad_norm": 0.027426833286881447,
+      "learning_rate": 2.1465808395569644e-06,
+      "loss": 0.0532,
+      "step": 3353
+    },
+    {
+      "epoch": 0.7029972752043597,
+      "grad_norm": 0.031186439096927643,
+      "learning_rate": 2.1437937502968093e-06,
+      "loss": 0.0555,
+      "step": 3354
+    },
+    {
+      "epoch": 0.7032068748690002,
+      "grad_norm": 0.031483497470617294,
+      "learning_rate": 2.1410079777471444e-06,
+      "loss": 0.0527,
+      "step": 3355
+    },
+    {
+      "epoch": 0.7034164745336408,
+      "grad_norm": 0.03852154687047005,
+      "learning_rate": 2.1382235231922053e-06,
+      "loss": 0.0592,
+      "step": 3356
+    },
+    {
+      "epoch": 0.7036260741982813,
+      "grad_norm": 0.03263459354639053,
+      "learning_rate": 2.13544038791563e-06,
+      "loss": 0.0546,
+      "step": 3357
+    },
+    {
+      "epoch": 0.7038356738629218,
+      "grad_norm": 0.028121495619416237,
+      "learning_rate": 2.1326585732004384e-06,
+      "loss": 0.0508,
+      "step": 3358
+    },
+    {
+      "epoch": 0.7040452735275624,
+      "grad_norm": 0.036633770912885666,
+      "learning_rate": 2.1298780803290497e-06,
+      "loss": 0.0558,
+      "step": 3359
+    },
+    {
+      "epoch": 0.7042548731922029,
+      "grad_norm": 0.03133592754602432,
+      "learning_rate": 2.127098910583273e-06,
+      "loss": 0.0547,
+      "step": 3360
+    },
+    {
+      "epoch": 0.7044644728568434,
+      "grad_norm": 0.026572570204734802,
+      "learning_rate": 2.124321065244298e-06,
+      "loss": 0.0551,
+      "step": 3361
+    },
+    {
+      "epoch": 0.704674072521484,
+      "grad_norm": 0.023017263039946556,
+      "learning_rate": 2.121544545592715e-06,
+      "loss": 0.0539,
+      "step": 3362
+    },
+    {
+      "epoch": 0.7048836721861245,
+      "grad_norm": 0.021083930507302284,
+      "learning_rate": 2.1187693529085e-06,
+      "loss": 0.0557,
+      "step": 3363
+    },
+    {
+      "epoch": 0.7050932718507651,
+      "grad_norm": 0.02307903952896595,
+      "learning_rate": 2.1159954884710133e-06,
+      "loss": 0.0571,
+      "step": 3364
+    },
+    {
+      "epoch": 0.7053028715154056,
+      "grad_norm": 0.02702009305357933,
+      "learning_rate": 2.1132229535590092e-06,
+      "loss": 0.054,
+      "step": 3365
+    },
+    {
+      "epoch": 0.7055124711800461,
+      "grad_norm": 0.029108747839927673,
+      "learning_rate": 2.110451749450624e-06,
+      "loss": 0.0525,
+      "step": 3366
+    },
+    {
+      "epoch": 0.7057220708446866,
+      "grad_norm": 0.024553043767809868,
+      "learning_rate": 2.1076818774233796e-06,
+      "loss": 0.0525,
+      "step": 3367
+    },
+    {
+      "epoch": 0.7059316705093271,
+      "grad_norm": 0.03205152973532677,
+      "learning_rate": 2.104913338754189e-06,
+      "loss": 0.0532,
+      "step": 3368
+    },
+    {
+      "epoch": 0.7061412701739678,
+      "grad_norm": 0.04320382699370384,
+      "learning_rate": 2.102146134719349e-06,
+      "loss": 0.0527,
+      "step": 3369
+    },
+    {
+      "epoch": 0.7063508698386083,
+      "grad_norm": 0.04006104916334152,
+      "learning_rate": 2.0993802665945396e-06,
+      "loss": 0.0532,
+      "step": 3370
+    },
+    {
+      "epoch": 0.7065604695032488,
+      "grad_norm": 0.022940274327993393,
+      "learning_rate": 2.0966157356548255e-06,
+      "loss": 0.0541,
+      "step": 3371
+    },
+    {
+      "epoch": 0.7067700691678893,
+      "grad_norm": 0.03261176869273186,
+      "learning_rate": 2.093852543174652e-06,
+      "loss": 0.057,
+      "step": 3372
+    },
+    {
+      "epoch": 0.7069796688325298,
+      "grad_norm": 0.033332716673612595,
+      "learning_rate": 2.0910906904278542e-06,
+      "loss": 0.0528,
+      "step": 3373
+    },
+    {
+      "epoch": 0.7071892684971705,
+      "grad_norm": 0.02474953606724739,
+      "learning_rate": 2.088330178687642e-06,
+      "loss": 0.0564,
+      "step": 3374
+    },
+    {
+      "epoch": 0.707398868161811,
+      "grad_norm": 0.028840798884630203,
+      "learning_rate": 2.085571009226613e-06,
+      "loss": 0.0528,
+      "step": 3375
+    },
+    {
+      "epoch": 0.7076084678264515,
+      "grad_norm": 0.03333436697721481,
+      "learning_rate": 2.082813183316745e-06,
+      "loss": 0.0579,
+      "step": 3376
+    },
+    {
+      "epoch": 0.707818067491092,
+      "grad_norm": 0.03125330060720444,
+      "learning_rate": 2.080056702229393e-06,
+      "loss": 0.0536,
+      "step": 3377
+    },
+    {
+      "epoch": 0.7080276671557325,
+      "grad_norm": 0.03265562653541565,
+      "learning_rate": 2.0773015672352938e-06,
+      "loss": 0.0561,
+      "step": 3378
+    },
+    {
+      "epoch": 0.7082372668203731,
+      "grad_norm": 0.030047666281461716,
+      "learning_rate": 2.0745477796045664e-06,
+      "loss": 0.0535,
+      "step": 3379
+    },
+    {
+      "epoch": 0.7084468664850136,
+      "grad_norm": 0.03281790018081665,
+      "learning_rate": 2.0717953406067033e-06,
+      "loss": 0.0539,
+      "step": 3380
+    },
+    {
+      "epoch": 0.7086564661496542,
+      "grad_norm": 0.031788330525159836,
+      "learning_rate": 2.0690442515105797e-06,
+      "loss": 0.0548,
+      "step": 3381
+    },
+    {
+      "epoch": 0.7088660658142947,
+      "grad_norm": 0.03671342507004738,
+      "learning_rate": 2.0662945135844493e-06,
+      "loss": 0.0547,
+      "step": 3382
+    },
+    {
+      "epoch": 0.7090756654789352,
+      "grad_norm": 0.05137075111269951,
+      "learning_rate": 2.0635461280959386e-06,
+      "loss": 0.054,
+      "step": 3383
+    },
+    {
+      "epoch": 0.7092852651435758,
+      "grad_norm": 0.025410549715161324,
+      "learning_rate": 2.060799096312051e-06,
+      "loss": 0.0519,
+      "step": 3384
+    },
+    {
+      "epoch": 0.7094948648082163,
+      "grad_norm": 0.041060563176870346,
+      "learning_rate": 2.0580534194991696e-06,
+      "loss": 0.0533,
+      "step": 3385
+    },
+    {
+      "epoch": 0.7097044644728568,
+      "grad_norm": 0.04743944853544235,
+      "learning_rate": 2.0553090989230527e-06,
+      "loss": 0.0556,
+      "step": 3386
+    },
+    {
+      "epoch": 0.7099140641374974,
+      "grad_norm": 0.02690894342958927,
+      "learning_rate": 2.052566135848828e-06,
+      "loss": 0.0542,
+      "step": 3387
+    },
+    {
+      "epoch": 0.7101236638021379,
+      "grad_norm": 0.028761066496372223,
+      "learning_rate": 2.0498245315410037e-06,
+      "loss": 0.0551,
+      "step": 3388
+    },
+    {
+      "epoch": 0.7103332634667785,
+      "grad_norm": 0.027141369879245758,
+      "learning_rate": 2.047084287263458e-06,
+      "loss": 0.0542,
+      "step": 3389
+    },
+    {
+      "epoch": 0.710542863131419,
+      "grad_norm": 0.02281133271753788,
+      "learning_rate": 2.0443454042794404e-06,
+      "loss": 0.0546,
+      "step": 3390
+    },
+    {
+      "epoch": 0.7107524627960595,
+      "grad_norm": 0.025255044922232628,
+      "learning_rate": 2.0416078838515773e-06,
+      "loss": 0.0542,
+      "step": 3391
+    },
+    {
+      "epoch": 0.7109620624607,
+      "grad_norm": 0.02403366006910801,
+      "learning_rate": 2.0388717272418653e-06,
+      "loss": 0.0549,
+      "step": 3392
+    },
+    {
+      "epoch": 0.7111716621253406,
+      "grad_norm": 0.022493207827210426,
+      "learning_rate": 2.036136935711674e-06,
+      "loss": 0.054,
+      "step": 3393
+    },
+    {
+      "epoch": 0.7113812617899812,
+      "grad_norm": 0.022795764729380608,
+      "learning_rate": 2.0334035105217394e-06,
+      "loss": 0.0582,
+      "step": 3394
+    },
+    {
+      "epoch": 0.7115908614546217,
+      "grad_norm": 0.024398164823651314,
+      "learning_rate": 2.0306714529321687e-06,
+      "loss": 0.0541,
+      "step": 3395
+    },
+    {
+      "epoch": 0.7118004611192622,
+      "grad_norm": 0.022561099380254745,
+      "learning_rate": 2.0279407642024427e-06,
+      "loss": 0.0557,
+      "step": 3396
+    },
+    {
+      "epoch": 0.7120100607839027,
+      "grad_norm": 0.02206028625369072,
+      "learning_rate": 2.0252114455914056e-06,
+      "loss": 0.0549,
+      "step": 3397
+    },
+    {
+      "epoch": 0.7122196604485432,
+      "grad_norm": 0.021681906655430794,
+      "learning_rate": 2.0224834983572766e-06,
+      "loss": 0.0546,
+      "step": 3398
+    },
+    {
+      "epoch": 0.7124292601131839,
+      "grad_norm": 0.02355903573334217,
+      "learning_rate": 2.0197569237576352e-06,
+      "loss": 0.0529,
+      "step": 3399
+    },
+    {
+      "epoch": 0.7126388597778244,
+      "grad_norm": 0.02560109831392765,
+      "learning_rate": 2.017031723049432e-06,
+      "loss": 0.0537,
+      "step": 3400
+    },
+    {
+      "epoch": 0.7128484594424649,
+      "grad_norm": 0.021792568266391754,
+      "learning_rate": 2.0143078974889846e-06,
+      "loss": 0.0549,
+      "step": 3401
+    },
+    {
+      "epoch": 0.7130580591071054,
+      "grad_norm": 0.022112544625997543,
+      "learning_rate": 2.011585448331978e-06,
+      "loss": 0.0541,
+      "step": 3402
+    },
+    {
+      "epoch": 0.7132676587717459,
+      "grad_norm": 0.02488597482442856,
+      "learning_rate": 2.008864376833457e-06,
+      "loss": 0.0531,
+      "step": 3403
+    },
+    {
+      "epoch": 0.7134772584363865,
+      "grad_norm": 0.022757617756724358,
+      "learning_rate": 2.0061446842478393e-06,
+      "loss": 0.0544,
+      "step": 3404
+    },
+    {
+      "epoch": 0.713686858101027,
+      "grad_norm": 0.028682250529527664,
+      "learning_rate": 2.003426371828898e-06,
+      "loss": 0.0521,
+      "step": 3405
+    },
+    {
+      "epoch": 0.7138964577656676,
+      "grad_norm": 0.026006808504462242,
+      "learning_rate": 2.000709440829779e-06,
+      "loss": 0.055,
+      "step": 3406
+    },
+    {
+      "epoch": 0.7141060574303081,
+      "grad_norm": 0.020952891558408737,
+      "learning_rate": 1.997993892502983e-06,
+      "loss": 0.0541,
+      "step": 3407
+    },
+    {
+      "epoch": 0.7143156570949486,
+      "grad_norm": 0.030959462746977806,
+      "learning_rate": 1.9952797281003784e-06,
+      "loss": 0.0533,
+      "step": 3408
+    },
+    {
+      "epoch": 0.7145252567595892,
+      "grad_norm": 0.022278886288404465,
+      "learning_rate": 1.9925669488731975e-06,
+      "loss": 0.0538,
+      "step": 3409
+    },
+    {
+      "epoch": 0.7147348564242297,
+      "grad_norm": 0.02709982544183731,
+      "learning_rate": 1.989855556072028e-06,
+      "loss": 0.0511,
+      "step": 3410
+    },
+    {
+      "epoch": 0.7149444560888703,
+      "grad_norm": 0.02348734438419342,
+      "learning_rate": 1.98714555094682e-06,
+      "loss": 0.0545,
+      "step": 3411
+    },
+    {
+      "epoch": 0.7151540557535108,
+      "grad_norm": 0.02444126456975937,
+      "learning_rate": 1.9844369347468895e-06,
+      "loss": 0.0546,
+      "step": 3412
+    },
+    {
+      "epoch": 0.7153636554181513,
+      "grad_norm": 0.020223500207066536,
+      "learning_rate": 1.981729708720903e-06,
+      "loss": 0.0556,
+      "step": 3413
+    },
+    {
+      "epoch": 0.7155732550827919,
+      "grad_norm": 0.028548697009682655,
+      "learning_rate": 1.979023874116895e-06,
+      "loss": 0.0556,
+      "step": 3414
+    },
+    {
+      "epoch": 0.7157828547474324,
+      "grad_norm": 0.021214155480265617,
+      "learning_rate": 1.976319432182254e-06,
+      "loss": 0.0555,
+      "step": 3415
+    },
+    {
+      "epoch": 0.7159924544120729,
+      "grad_norm": 0.022887524217367172,
+      "learning_rate": 1.973616384163728e-06,
+      "loss": 0.0523,
+      "step": 3416
+    },
+    {
+      "epoch": 0.7162020540767134,
+      "grad_norm": 0.024027662351727486,
+      "learning_rate": 1.9709147313074177e-06,
+      "loss": 0.0544,
+      "step": 3417
+    },
+    {
+      "epoch": 0.716411653741354,
+      "grad_norm": 0.028380177915096283,
+      "learning_rate": 1.968214474858787e-06,
+      "loss": 0.053,
+      "step": 3418
+    },
+    {
+      "epoch": 0.7166212534059946,
+      "grad_norm": 0.025715503841638565,
+      "learning_rate": 1.965515616062656e-06,
+      "loss": 0.0545,
+      "step": 3419
+    },
+    {
+      "epoch": 0.7168308530706351,
+      "grad_norm": 0.03151887282729149,
+      "learning_rate": 1.962818156163194e-06,
+      "loss": 0.0551,
+      "step": 3420
+    },
+    {
+      "epoch": 0.7170404527352756,
+      "grad_norm": 0.04665770381689072,
+      "learning_rate": 1.9601220964039324e-06,
+      "loss": 0.056,
+      "step": 3421
+    },
+    {
+      "epoch": 0.7172500523999161,
+      "grad_norm": 0.02261737547814846,
+      "learning_rate": 1.957427438027753e-06,
+      "loss": 0.0548,
+      "step": 3422
+    },
+    {
+      "epoch": 0.7174596520645568,
+      "grad_norm": 0.03728951886296272,
+      "learning_rate": 1.9547341822768906e-06,
+      "loss": 0.0542,
+      "step": 3423
+    },
+    {
+      "epoch": 0.7176692517291973,
+      "grad_norm": 0.032367952167987823,
+      "learning_rate": 1.9520423303929383e-06,
+      "loss": 0.0532,
+      "step": 3424
+    },
+    {
+      "epoch": 0.7178788513938378,
+      "grad_norm": 0.028157230466604233,
+      "learning_rate": 1.9493518836168403e-06,
+      "loss": 0.0551,
+      "step": 3425
+    },
+    {
+      "epoch": 0.7180884510584783,
+      "grad_norm": 0.029264342039823532,
+      "learning_rate": 1.946662843188888e-06,
+      "loss": 0.0559,
+      "step": 3426
+    },
+    {
+      "epoch": 0.7182980507231188,
+      "grad_norm": 0.02802872471511364,
+      "learning_rate": 1.9439752103487324e-06,
+      "loss": 0.0525,
+      "step": 3427
+    },
+    {
+      "epoch": 0.7185076503877594,
+      "grad_norm": 0.02839348092675209,
+      "learning_rate": 1.9412889863353683e-06,
+      "loss": 0.0532,
+      "step": 3428
+    },
+    {
+      "epoch": 0.7187172500524,
+      "grad_norm": 0.026173194870352745,
+      "learning_rate": 1.9386041723871485e-06,
+      "loss": 0.0528,
+      "step": 3429
+    },
+    {
+      "epoch": 0.7189268497170405,
+      "grad_norm": 0.028481315821409225,
+      "learning_rate": 1.9359207697417677e-06,
+      "loss": 0.0537,
+      "step": 3430
+    },
+    {
+      "epoch": 0.719136449381681,
+      "grad_norm": 0.028069039806723595,
+      "learning_rate": 1.9332387796362744e-06,
+      "loss": 0.0566,
+      "step": 3431
+    },
+    {
+      "epoch": 0.7193460490463215,
+      "grad_norm": 0.031697165220975876,
+      "learning_rate": 1.9305582033070714e-06,
+      "loss": 0.0542,
+      "step": 3432
+    },
+    {
+      "epoch": 0.7195556487109621,
+      "grad_norm": 0.029638856649398804,
+      "learning_rate": 1.927879041989895e-06,
+      "loss": 0.054,
+      "step": 3433
+    },
+    {
+      "epoch": 0.7197652483756026,
+      "grad_norm": 0.022818326950073242,
+      "learning_rate": 1.925201296919842e-06,
+      "loss": 0.0546,
+      "step": 3434
+    },
+    {
+      "epoch": 0.7199748480402431,
+      "grad_norm": 0.03616581857204437,
+      "learning_rate": 1.9225249693313547e-06,
+      "loss": 0.0542,
+      "step": 3435
+    },
+    {
+      "epoch": 0.7201844477048837,
+      "grad_norm": 0.028139004483819008,
+      "learning_rate": 1.919850060458215e-06,
+      "loss": 0.053,
+      "step": 3436
+    },
+    {
+      "epoch": 0.7203940473695242,
+      "grad_norm": 0.028990762308239937,
+      "learning_rate": 1.91717657153356e-06,
+      "loss": 0.0549,
+      "step": 3437
+    },
+    {
+      "epoch": 0.7206036470341648,
+      "grad_norm": 0.022995000705122948,
+      "learning_rate": 1.914504503789863e-06,
+      "loss": 0.0554,
+      "step": 3438
+    },
+    {
+      "epoch": 0.7208132466988053,
+      "grad_norm": 0.026573555544018745,
+      "learning_rate": 1.9118338584589503e-06,
+      "loss": 0.0556,
+      "step": 3439
+    },
+    {
+      "epoch": 0.7210228463634458,
+      "grad_norm": 0.024811234325170517,
+      "learning_rate": 1.909164636771986e-06,
+      "loss": 0.0537,
+      "step": 3440
+    },
+    {
+      "epoch": 0.7212324460280863,
+      "grad_norm": 0.02453632839024067,
+      "learning_rate": 1.9064968399594818e-06,
+      "loss": 0.0517,
+      "step": 3441
+    },
+    {
+      "epoch": 0.7214420456927269,
+      "grad_norm": 0.024129144847393036,
+      "learning_rate": 1.9038304692512943e-06,
+      "loss": 0.0534,
+      "step": 3442
+    },
+    {
+      "epoch": 0.7216516453573675,
+      "grad_norm": 0.024486854672431946,
+      "learning_rate": 1.9011655258766165e-06,
+      "loss": 0.0531,
+      "step": 3443
+    },
+    {
+      "epoch": 0.721861245022008,
+      "grad_norm": 0.022883350029587746,
+      "learning_rate": 1.8985020110639862e-06,
+      "loss": 0.0555,
+      "step": 3444
+    },
+    {
+      "epoch": 0.7220708446866485,
+      "grad_norm": 0.02430296316742897,
+      "learning_rate": 1.8958399260412864e-06,
+      "loss": 0.0543,
+      "step": 3445
+    },
+    {
+      "epoch": 0.722280444351289,
+      "grad_norm": 0.023016586899757385,
+      "learning_rate": 1.893179272035734e-06,
+      "loss": 0.0543,
+      "step": 3446
+    },
+    {
+      "epoch": 0.7224900440159295,
+      "grad_norm": 0.022667940706014633,
+      "learning_rate": 1.890520050273892e-06,
+      "loss": 0.0557,
+      "step": 3447
+    },
+    {
+      "epoch": 0.7226996436805702,
+      "grad_norm": 0.025942491367459297,
+      "learning_rate": 1.8878622619816629e-06,
+      "loss": 0.0539,
+      "step": 3448
+    },
+    {
+      "epoch": 0.7229092433452107,
+      "grad_norm": 0.027111142873764038,
+      "learning_rate": 1.8852059083842838e-06,
+      "loss": 0.0553,
+      "step": 3449
+    },
+    {
+      "epoch": 0.7231188430098512,
+      "grad_norm": 0.024661056697368622,
+      "learning_rate": 1.8825509907063328e-06,
+      "loss": 0.055,
+      "step": 3450
+    },
+    {
+      "epoch": 0.7233284426744917,
+      "grad_norm": 0.028169486671686172,
+      "learning_rate": 1.8798975101717275e-06,
+      "loss": 0.0546,
+      "step": 3451
+    },
+    {
+      "epoch": 0.7235380423391322,
+      "grad_norm": 0.02489306405186653,
+      "learning_rate": 1.8772454680037233e-06,
+      "loss": 0.0551,
+      "step": 3452
+    },
+    {
+      "epoch": 0.7237476420037728,
+      "grad_norm": 0.024785684421658516,
+      "learning_rate": 1.8745948654249085e-06,
+      "loss": 0.0548,
+      "step": 3453
+    },
+    {
+      "epoch": 0.7239572416684134,
+      "grad_norm": 0.021179838106036186,
+      "learning_rate": 1.871945703657213e-06,
+      "loss": 0.0556,
+      "step": 3454
+    },
+    {
+      "epoch": 0.7241668413330539,
+      "grad_norm": 0.02177123725414276,
+      "learning_rate": 1.8692979839218984e-06,
+      "loss": 0.0575,
+      "step": 3455
+    },
+    {
+      "epoch": 0.7243764409976944,
+      "grad_norm": 0.025510409846901894,
+      "learning_rate": 1.8666517074395607e-06,
+      "loss": 0.0527,
+      "step": 3456
+    },
+    {
+      "epoch": 0.7245860406623349,
+      "grad_norm": 0.01962488703429699,
+      "learning_rate": 1.8640068754301355e-06,
+      "loss": 0.0554,
+      "step": 3457
+    },
+    {
+      "epoch": 0.7247956403269755,
+      "grad_norm": 0.023709991946816444,
+      "learning_rate": 1.8613634891128912e-06,
+      "loss": 0.0541,
+      "step": 3458
+    },
+    {
+      "epoch": 0.725005239991616,
+      "grad_norm": 0.02632421813905239,
+      "learning_rate": 1.8587215497064242e-06,
+      "loss": 0.0557,
+      "step": 3459
+    },
+    {
+      "epoch": 0.7252148396562565,
+      "grad_norm": 0.0217702928930521,
+      "learning_rate": 1.8560810584286726e-06,
+      "loss": 0.0557,
+      "step": 3460
+    },
+    {
+      "epoch": 0.7254244393208971,
+      "grad_norm": 0.02565830387175083,
+      "learning_rate": 1.853442016496898e-06,
+      "loss": 0.0528,
+      "step": 3461
+    },
+    {
+      "epoch": 0.7256340389855376,
+      "grad_norm": 0.034673016518354416,
+      "learning_rate": 1.8508044251277019e-06,
+      "loss": 0.0535,
+      "step": 3462
+    },
+    {
+      "epoch": 0.7258436386501782,
+      "grad_norm": 0.02693593129515648,
+      "learning_rate": 1.8481682855370098e-06,
+      "loss": 0.0535,
+      "step": 3463
+    },
+    {
+      "epoch": 0.7260532383148187,
+      "grad_norm": 0.018855255097150803,
+      "learning_rate": 1.8455335989400836e-06,
+      "loss": 0.0527,
+      "step": 3464
+    },
+    {
+      "epoch": 0.7262628379794592,
+      "grad_norm": 0.039858147501945496,
+      "learning_rate": 1.8429003665515165e-06,
+      "loss": 0.0535,
+      "step": 3465
+    },
+    {
+      "epoch": 0.7264724376440997,
+      "grad_norm": 0.03316653147339821,
+      "learning_rate": 1.8402685895852213e-06,
+      "loss": 0.0561,
+      "step": 3466
+    },
+    {
+      "epoch": 0.7266820373087403,
+      "grad_norm": 0.018629280850291252,
+      "learning_rate": 1.8376382692544498e-06,
+      "loss": 0.0525,
+      "step": 3467
+    },
+    {
+      "epoch": 0.7268916369733809,
+      "grad_norm": 0.024704501032829285,
+      "learning_rate": 1.835009406771781e-06,
+      "loss": 0.0541,
+      "step": 3468
+    },
+    {
+      "epoch": 0.7271012366380214,
+      "grad_norm": 0.020816821604967117,
+      "learning_rate": 1.8323820033491163e-06,
+      "loss": 0.0552,
+      "step": 3469
+    },
+    {
+      "epoch": 0.7273108363026619,
+      "grad_norm": 0.021062025800347328,
+      "learning_rate": 1.829756060197692e-06,
+      "loss": 0.0548,
+      "step": 3470
+    },
+    {
+      "epoch": 0.7275204359673024,
+      "grad_norm": 0.026151493191719055,
+      "learning_rate": 1.8271315785280658e-06,
+      "loss": 0.0545,
+      "step": 3471
+    },
+    {
+      "epoch": 0.7277300356319429,
+      "grad_norm": 0.019930191338062286,
+      "learning_rate": 1.8245085595501205e-06,
+      "loss": 0.0541,
+      "step": 3472
+    },
+    {
+      "epoch": 0.7279396352965836,
+      "grad_norm": 0.028385937213897705,
+      "learning_rate": 1.8218870044730702e-06,
+      "loss": 0.0535,
+      "step": 3473
+    },
+    {
+      "epoch": 0.7281492349612241,
+      "grad_norm": 0.021223368123173714,
+      "learning_rate": 1.8192669145054503e-06,
+      "loss": 0.0507,
+      "step": 3474
+    },
+    {
+      "epoch": 0.7283588346258646,
+      "grad_norm": 0.02485741302371025,
+      "learning_rate": 1.8166482908551242e-06,
+      "loss": 0.0555,
+      "step": 3475
+    },
+    {
+      "epoch": 0.7285684342905051,
+      "grad_norm": 0.022934934124350548,
+      "learning_rate": 1.8140311347292744e-06,
+      "loss": 0.0534,
+      "step": 3476
+    },
+    {
+      "epoch": 0.7287780339551456,
+      "grad_norm": 0.020589333027601242,
+      "learning_rate": 1.8114154473344081e-06,
+      "loss": 0.0537,
+      "step": 3477
+    },
+    {
+      "epoch": 0.7289876336197862,
+      "grad_norm": 0.021330052986741066,
+      "learning_rate": 1.8088012298763603e-06,
+      "loss": 0.0529,
+      "step": 3478
+    },
+    {
+      "epoch": 0.7291972332844268,
+      "grad_norm": 0.025633566081523895,
+      "learning_rate": 1.8061884835602805e-06,
+      "loss": 0.0564,
+      "step": 3479
+    },
+    {
+      "epoch": 0.7294068329490673,
+      "grad_norm": 0.021011043339967728,
+      "learning_rate": 1.8035772095906462e-06,
+      "loss": 0.0529,
+      "step": 3480
+    },
+    {
+      "epoch": 0.7296164326137078,
+      "grad_norm": 0.023229733109474182,
+      "learning_rate": 1.8009674091712565e-06,
+      "loss": 0.0541,
+      "step": 3481
+    },
+    {
+      "epoch": 0.7298260322783483,
+      "grad_norm": 0.02509014680981636,
+      "learning_rate": 1.7983590835052267e-06,
+      "loss": 0.054,
+      "step": 3482
+    },
+    {
+      "epoch": 0.7300356319429889,
+      "grad_norm": 0.021912094205617905,
+      "learning_rate": 1.7957522337949924e-06,
+      "loss": 0.0527,
+      "step": 3483
+    },
+    {
+      "epoch": 0.7302452316076294,
+      "grad_norm": 0.024635696783661842,
+      "learning_rate": 1.7931468612423142e-06,
+      "loss": 0.0531,
+      "step": 3484
+    },
+    {
+      "epoch": 0.73045483127227,
+      "grad_norm": 0.018132206052541733,
+      "learning_rate": 1.7905429670482654e-06,
+      "loss": 0.0552,
+      "step": 3485
+    },
+    {
+      "epoch": 0.7306644309369105,
+      "grad_norm": 0.023826826363801956,
+      "learning_rate": 1.7879405524132426e-06,
+      "loss": 0.0526,
+      "step": 3486
+    },
+    {
+      "epoch": 0.730874030601551,
+      "grad_norm": 0.020080413669347763,
+      "learning_rate": 1.7853396185369592e-06,
+      "loss": 0.0542,
+      "step": 3487
+    },
+    {
+      "epoch": 0.7310836302661916,
+      "grad_norm": 0.02045084349811077,
+      "learning_rate": 1.7827401666184434e-06,
+      "loss": 0.0543,
+      "step": 3488
+    },
+    {
+      "epoch": 0.7312932299308321,
+      "grad_norm": 0.023532617837190628,
+      "learning_rate": 1.7801421978560418e-06,
+      "loss": 0.0532,
+      "step": 3489
+    },
+    {
+      "epoch": 0.7315028295954726,
+      "grad_norm": 0.02120790258049965,
+      "learning_rate": 1.7775457134474177e-06,
+      "loss": 0.0526,
+      "step": 3490
+    },
+    {
+      "epoch": 0.7317124292601132,
+      "grad_norm": 0.024196181446313858,
+      "learning_rate": 1.7749507145895518e-06,
+      "loss": 0.0563,
+      "step": 3491
+    },
+    {
+      "epoch": 0.7319220289247538,
+      "grad_norm": 0.023171545937657356,
+      "learning_rate": 1.772357202478735e-06,
+      "loss": 0.0559,
+      "step": 3492
+    },
+    {
+      "epoch": 0.7321316285893943,
+      "grad_norm": 0.02514570765197277,
+      "learning_rate": 1.7697651783105795e-06,
+      "loss": 0.054,
+      "step": 3493
+    },
+    {
+      "epoch": 0.7323412282540348,
+      "grad_norm": 0.02238095924258232,
+      "learning_rate": 1.7671746432800058e-06,
+      "loss": 0.0535,
+      "step": 3494
+    },
+    {
+      "epoch": 0.7325508279186753,
+      "grad_norm": 0.030948419123888016,
+      "learning_rate": 1.7645855985812476e-06,
+      "loss": 0.0521,
+      "step": 3495
+    },
+    {
+      "epoch": 0.7327604275833158,
+      "grad_norm": 0.02502829022705555,
+      "learning_rate": 1.7619980454078572e-06,
+      "loss": 0.0525,
+      "step": 3496
+    },
+    {
+      "epoch": 0.7329700272479565,
+      "grad_norm": 0.02190042845904827,
+      "learning_rate": 1.7594119849526947e-06,
+      "loss": 0.0547,
+      "step": 3497
+    },
+    {
+      "epoch": 0.733179626912597,
+      "grad_norm": 0.031756095588207245,
+      "learning_rate": 1.756827418407936e-06,
+      "loss": 0.0517,
+      "step": 3498
+    },
+    {
+      "epoch": 0.7333892265772375,
+      "grad_norm": 0.03542139008641243,
+      "learning_rate": 1.7542443469650638e-06,
+      "loss": 0.0533,
+      "step": 3499
+    },
+    {
+      "epoch": 0.733598826241878,
+      "grad_norm": 0.024449272081255913,
+      "learning_rate": 1.7516627718148716e-06,
+      "loss": 0.0501,
+      "step": 3500
+    },
+    {
+      "epoch": 0.7338084259065185,
+      "grad_norm": 0.0264756977558136,
+      "learning_rate": 1.7490826941474687e-06,
+      "loss": 0.0556,
+      "step": 3501
+    },
+    {
+      "epoch": 0.7340180255711591,
+      "grad_norm": 0.026716774329543114,
+      "learning_rate": 1.7465041151522666e-06,
+      "loss": 0.0549,
+      "step": 3502
+    },
+    {
+      "epoch": 0.7342276252357997,
+      "grad_norm": 0.03247566148638725,
+      "learning_rate": 1.7439270360179933e-06,
+      "loss": 0.0522,
+      "step": 3503
+    },
+    {
+      "epoch": 0.7344372249004402,
+      "grad_norm": 0.04316284507513046,
+      "learning_rate": 1.7413514579326806e-06,
+      "loss": 0.0556,
+      "step": 3504
+    },
+    {
+      "epoch": 0.7346468245650807,
+      "grad_norm": 0.021153295412659645,
+      "learning_rate": 1.7387773820836668e-06,
+      "loss": 0.055,
+      "step": 3505
+    },
+    {
+      "epoch": 0.7348564242297212,
+      "grad_norm": 0.045009203255176544,
+      "learning_rate": 1.7362048096576023e-06,
+      "loss": 0.0544,
+      "step": 3506
+    },
+    {
+      "epoch": 0.7350660238943618,
+      "grad_norm": 0.040110033005476,
+      "learning_rate": 1.7336337418404442e-06,
+      "loss": 0.0585,
+      "step": 3507
+    },
+    {
+      "epoch": 0.7352756235590023,
+      "grad_norm": 0.024850115180015564,
+      "learning_rate": 1.7310641798174516e-06,
+      "loss": 0.0576,
+      "step": 3508
+    },
+    {
+      "epoch": 0.7354852232236428,
+      "grad_norm": 0.03829395771026611,
+      "learning_rate": 1.728496124773194e-06,
+      "loss": 0.0543,
+      "step": 3509
+    },
+    {
+      "epoch": 0.7356948228882834,
+      "grad_norm": 0.02593119814991951,
+      "learning_rate": 1.7259295778915419e-06,
+      "loss": 0.0555,
+      "step": 3510
+    },
+    {
+      "epoch": 0.7359044225529239,
+      "grad_norm": 0.027632324025034904,
+      "learning_rate": 1.7233645403556765e-06,
+      "loss": 0.0559,
+      "step": 3511
+    },
+    {
+      "epoch": 0.7361140222175645,
+      "grad_norm": 0.03341587260365486,
+      "learning_rate": 1.7208010133480751e-06,
+      "loss": 0.0545,
+      "step": 3512
+    },
+    {
+      "epoch": 0.736323621882205,
+      "grad_norm": 0.023744579404592514,
+      "learning_rate": 1.7182389980505254e-06,
+      "loss": 0.0524,
+      "step": 3513
+    },
+    {
+      "epoch": 0.7365332215468455,
+      "grad_norm": 0.02730092778801918,
+      "learning_rate": 1.7156784956441181e-06,
+      "loss": 0.058,
+      "step": 3514
+    },
+    {
+      "epoch": 0.736742821211486,
+      "grad_norm": 0.04322980344295502,
+      "learning_rate": 1.7131195073092417e-06,
+      "loss": 0.0545,
+      "step": 3515
+    },
+    {
+      "epoch": 0.7369524208761266,
+      "grad_norm": 0.02667373977601528,
+      "learning_rate": 1.710562034225588e-06,
+      "loss": 0.0557,
+      "step": 3516
+    },
+    {
+      "epoch": 0.7371620205407672,
+      "grad_norm": 0.03887426108121872,
+      "learning_rate": 1.7080060775721546e-06,
+      "loss": 0.0514,
+      "step": 3517
+    },
+    {
+      "epoch": 0.7373716202054077,
+      "grad_norm": 0.06282377243041992,
+      "learning_rate": 1.7054516385272345e-06,
+      "loss": 0.0567,
+      "step": 3518
+    },
+    {
+      "epoch": 0.7375812198700482,
+      "grad_norm": 0.041183438152074814,
+      "learning_rate": 1.7028987182684248e-06,
+      "loss": 0.0551,
+      "step": 3519
+    },
+    {
+      "epoch": 0.7377908195346887,
+      "grad_norm": 0.026868589222431183,
+      "learning_rate": 1.7003473179726226e-06,
+      "loss": 0.0558,
+      "step": 3520
+    },
+    {
+      "epoch": 0.7380004191993292,
+      "grad_norm": 0.061398621648550034,
+      "learning_rate": 1.6977974388160213e-06,
+      "loss": 0.0527,
+      "step": 3521
+    },
+    {
+      "epoch": 0.7382100188639699,
+      "grad_norm": 0.05005694553256035,
+      "learning_rate": 1.695249081974113e-06,
+      "loss": 0.0526,
+      "step": 3522
+    },
+    {
+      "epoch": 0.7384196185286104,
+      "grad_norm": 0.01858423836529255,
+      "learning_rate": 1.6927022486216915e-06,
+      "loss": 0.0522,
+      "step": 3523
+    },
+    {
+      "epoch": 0.7386292181932509,
+      "grad_norm": 0.045225538313388824,
+      "learning_rate": 1.6901569399328483e-06,
+      "loss": 0.0529,
+      "step": 3524
+    },
+    {
+      "epoch": 0.7388388178578914,
+      "grad_norm": 0.03772973269224167,
+      "learning_rate": 1.6876131570809667e-06,
+      "loss": 0.0558,
+      "step": 3525
+    },
+    {
+      "epoch": 0.7390484175225319,
+      "grad_norm": 0.029959680512547493,
+      "learning_rate": 1.6850709012387328e-06,
+      "loss": 0.0574,
+      "step": 3526
+    },
+    {
+      "epoch": 0.7392580171871725,
+      "grad_norm": 0.06689772009849548,
+      "learning_rate": 1.6825301735781252e-06,
+      "loss": 0.0507,
+      "step": 3527
+    },
+    {
+      "epoch": 0.7394676168518131,
+      "grad_norm": 0.04782269150018692,
+      "learning_rate": 1.679990975270417e-06,
+      "loss": 0.0543,
+      "step": 3528
+    },
+    {
+      "epoch": 0.7396772165164536,
+      "grad_norm": 0.035698726773262024,
+      "learning_rate": 1.6774533074861793e-06,
+      "loss": 0.053,
+      "step": 3529
+    },
+    {
+      "epoch": 0.7398868161810941,
+      "grad_norm": 0.07200644910335541,
+      "learning_rate": 1.6749171713952783e-06,
+      "loss": 0.0537,
+      "step": 3530
+    },
+    {
+      "epoch": 0.7400964158457346,
+      "grad_norm": 0.05012660101056099,
+      "learning_rate": 1.6723825681668692e-06,
+      "loss": 0.0529,
+      "step": 3531
+    },
+    {
+      "epoch": 0.7403060155103752,
+      "grad_norm": 0.031019916757941246,
+      "learning_rate": 1.6698494989694064e-06,
+      "loss": 0.0542,
+      "step": 3532
+    },
+    {
+      "epoch": 0.7405156151750157,
+      "grad_norm": 0.07774803787469864,
+      "learning_rate": 1.6673179649706312e-06,
+      "loss": 0.0546,
+      "step": 3533
+    },
+    {
+      "epoch": 0.7407252148396563,
+      "grad_norm": 0.06452113389968872,
+      "learning_rate": 1.664787967337584e-06,
+      "loss": 0.053,
+      "step": 3534
+    },
+    {
+      "epoch": 0.7409348145042968,
+      "grad_norm": 0.024198826402425766,
+      "learning_rate": 1.6622595072365887e-06,
+      "loss": 0.0523,
+      "step": 3535
+    },
+    {
+      "epoch": 0.7411444141689373,
+      "grad_norm": 0.07494359463453293,
+      "learning_rate": 1.6597325858332675e-06,
+      "loss": 0.0525,
+      "step": 3536
+    },
+    {
+      "epoch": 0.7413540138335779,
+      "grad_norm": 0.060714419931173325,
+      "learning_rate": 1.6572072042925335e-06,
+      "loss": 0.0551,
+      "step": 3537
+    },
+    {
+      "epoch": 0.7415636134982184,
+      "grad_norm": 0.028920724987983704,
+      "learning_rate": 1.6546833637785814e-06,
+      "loss": 0.055,
+      "step": 3538
+    },
+    {
+      "epoch": 0.7417732131628589,
+      "grad_norm": 0.07191255688667297,
+      "learning_rate": 1.652161065454903e-06,
+      "loss": 0.0569,
+      "step": 3539
+    },
+    {
+      "epoch": 0.7419828128274994,
+      "grad_norm": 0.045189227908849716,
+      "learning_rate": 1.64964031048428e-06,
+      "loss": 0.0538,
+      "step": 3540
+    },
+    {
+      "epoch": 0.74219241249214,
+      "grad_norm": 0.034395113587379456,
+      "learning_rate": 1.6471211000287762e-06,
+      "loss": 0.0559,
+      "step": 3541
+    },
+    {
+      "epoch": 0.7424020121567806,
+      "grad_norm": 0.062001634389162064,
+      "learning_rate": 1.6446034352497504e-06,
+      "loss": 0.0567,
+      "step": 3542
+    },
+    {
+      "epoch": 0.7426116118214211,
+      "grad_norm": 0.03360583260655403,
+      "learning_rate": 1.6420873173078422e-06,
+      "loss": 0.0552,
+      "step": 3543
+    },
+    {
+      "epoch": 0.7428212114860616,
+      "grad_norm": 0.0365772545337677,
+      "learning_rate": 1.6395727473629852e-06,
+      "loss": 0.051,
+      "step": 3544
+    },
+    {
+      "epoch": 0.7430308111507021,
+      "grad_norm": 0.046734243631362915,
+      "learning_rate": 1.637059726574392e-06,
+      "loss": 0.0532,
+      "step": 3545
+    },
+    {
+      "epoch": 0.7432404108153426,
+      "grad_norm": 0.021529680117964745,
+      "learning_rate": 1.634548256100566e-06,
+      "loss": 0.053,
+      "step": 3546
+    },
+    {
+      "epoch": 0.7434500104799833,
+      "grad_norm": 0.04237625375390053,
+      "learning_rate": 1.632038337099297e-06,
+      "loss": 0.0539,
+      "step": 3547
+    },
+    {
+      "epoch": 0.7436596101446238,
+      "grad_norm": 0.029763460159301758,
+      "learning_rate": 1.6295299707276546e-06,
+      "loss": 0.0511,
+      "step": 3548
+    },
+    {
+      "epoch": 0.7438692098092643,
+      "grad_norm": 0.031175674870610237,
+      "learning_rate": 1.6270231581419943e-06,
+      "loss": 0.0551,
+      "step": 3549
+    },
+    {
+      "epoch": 0.7440788094739048,
+      "grad_norm": 0.04852377995848656,
+      "learning_rate": 1.6245179004979588e-06,
+      "loss": 0.054,
+      "step": 3550
+    },
+    {
+      "epoch": 0.7442884091385453,
+      "grad_norm": 0.027154749259352684,
+      "learning_rate": 1.6220141989504683e-06,
+      "loss": 0.0528,
+      "step": 3551
+    },
+    {
+      "epoch": 0.744498008803186,
+      "grad_norm": 0.026955176144838333,
+      "learning_rate": 1.6195120546537307e-06,
+      "loss": 0.0542,
+      "step": 3552
+    },
+    {
+      "epoch": 0.7447076084678265,
+      "grad_norm": 0.029728444293141365,
+      "learning_rate": 1.6170114687612349e-06,
+      "loss": 0.0566,
+      "step": 3553
+    },
+    {
+      "epoch": 0.744917208132467,
+      "grad_norm": 0.018385691568255424,
+      "learning_rate": 1.6145124424257497e-06,
+      "loss": 0.0548,
+      "step": 3554
+    },
+    {
+      "epoch": 0.7451268077971075,
+      "grad_norm": 0.02659742347896099,
+      "learning_rate": 1.6120149767993237e-06,
+      "loss": 0.0572,
+      "step": 3555
+    },
+    {
+      "epoch": 0.745336407461748,
+      "grad_norm": 0.030509591102600098,
+      "learning_rate": 1.6095190730332893e-06,
+      "loss": 0.0547,
+      "step": 3556
+    },
+    {
+      "epoch": 0.7455460071263886,
+      "grad_norm": 0.02252962999045849,
+      "learning_rate": 1.60702473227826e-06,
+      "loss": 0.0554,
+      "step": 3557
+    },
+    {
+      "epoch": 0.7457556067910291,
+      "grad_norm": 0.029321059584617615,
+      "learning_rate": 1.6045319556841227e-06,
+      "loss": 0.0525,
+      "step": 3558
+    },
+    {
+      "epoch": 0.7459652064556697,
+      "grad_norm": 0.026201212778687477,
+      "learning_rate": 1.6020407444000497e-06,
+      "loss": 0.0564,
+      "step": 3559
+    },
+    {
+      "epoch": 0.7461748061203102,
+      "grad_norm": 0.028670430183410645,
+      "learning_rate": 1.5995510995744879e-06,
+      "loss": 0.0548,
+      "step": 3560
+    },
+    {
+      "epoch": 0.7463844057849508,
+      "grad_norm": 0.028581839054822922,
+      "learning_rate": 1.5970630223551614e-06,
+      "loss": 0.0519,
+      "step": 3561
+    },
+    {
+      "epoch": 0.7465940054495913,
+      "grad_norm": 0.025404004380106926,
+      "learning_rate": 1.5945765138890746e-06,
+      "loss": 0.0549,
+      "step": 3562
+    },
+    {
+      "epoch": 0.7468036051142318,
+      "grad_norm": 0.04105531796813011,
+      "learning_rate": 1.5920915753225097e-06,
+      "loss": 0.0537,
+      "step": 3563
+    },
+    {
+      "epoch": 0.7470132047788723,
+      "grad_norm": 0.02948727272450924,
+      "learning_rate": 1.5896082078010183e-06,
+      "loss": 0.0519,
+      "step": 3564
+    },
+    {
+      "epoch": 0.7472228044435129,
+      "grad_norm": 0.02374359965324402,
+      "learning_rate": 1.5871264124694368e-06,
+      "loss": 0.0529,
+      "step": 3565
+    },
+    {
+      "epoch": 0.7474324041081535,
+      "grad_norm": 0.02481861598789692,
+      "learning_rate": 1.5846461904718686e-06,
+      "loss": 0.0539,
+      "step": 3566
+    },
+    {
+      "epoch": 0.747642003772794,
+      "grad_norm": 0.019914066419005394,
+      "learning_rate": 1.5821675429516981e-06,
+      "loss": 0.0535,
+      "step": 3567
+    },
+    {
+      "epoch": 0.7478516034374345,
+      "grad_norm": 0.027355968952178955,
+      "learning_rate": 1.5796904710515792e-06,
+      "loss": 0.054,
+      "step": 3568
+    },
+    {
+      "epoch": 0.748061203102075,
+      "grad_norm": 0.02385944128036499,
+      "learning_rate": 1.577214975913443e-06,
+      "loss": 0.0551,
+      "step": 3569
+    },
+    {
+      "epoch": 0.7482708027667155,
+      "grad_norm": 0.022167477756738663,
+      "learning_rate": 1.574741058678495e-06,
+      "loss": 0.0533,
+      "step": 3570
+    },
+    {
+      "epoch": 0.7484804024313562,
+      "grad_norm": 0.024447597563266754,
+      "learning_rate": 1.5722687204872038e-06,
+      "loss": 0.0569,
+      "step": 3571
+    },
+    {
+      "epoch": 0.7486900020959967,
+      "grad_norm": 0.019285105168819427,
+      "learning_rate": 1.569797962479321e-06,
+      "loss": 0.0547,
+      "step": 3572
+    },
+    {
+      "epoch": 0.7488996017606372,
+      "grad_norm": 0.026579929515719414,
+      "learning_rate": 1.5673287857938663e-06,
+      "loss": 0.0544,
+      "step": 3573
+    },
+    {
+      "epoch": 0.7491092014252777,
+      "grad_norm": 0.021767770871520042,
+      "learning_rate": 1.564861191569127e-06,
+      "loss": 0.0549,
+      "step": 3574
+    },
+    {
+      "epoch": 0.7493188010899182,
+      "grad_norm": 0.022724144160747528,
+      "learning_rate": 1.5623951809426663e-06,
+      "loss": 0.0553,
+      "step": 3575
+    },
+    {
+      "epoch": 0.7495284007545588,
+      "grad_norm": 0.02660524845123291,
+      "learning_rate": 1.5599307550513132e-06,
+      "loss": 0.0526,
+      "step": 3576
+    },
+    {
+      "epoch": 0.7497380004191994,
+      "grad_norm": 0.02129141427576542,
+      "learning_rate": 1.5574679150311656e-06,
+      "loss": 0.0547,
+      "step": 3577
+    },
+    {
+      "epoch": 0.7499476000838399,
+      "grad_norm": 0.02047175168991089,
+      "learning_rate": 1.555006662017594e-06,
+      "loss": 0.0529,
+      "step": 3578
+    },
+    {
+      "epoch": 0.7501571997484804,
+      "grad_norm": 0.02189241163432598,
+      "learning_rate": 1.5525469971452362e-06,
+      "loss": 0.0533,
+      "step": 3579
+    },
+    {
+      "epoch": 0.7503667994131209,
+      "grad_norm": 0.019183902069926262,
+      "learning_rate": 1.5500889215479974e-06,
+      "loss": 0.0535,
+      "step": 3580
+    },
+    {
+      "epoch": 0.7505763990777615,
+      "grad_norm": 0.022731618955731392,
+      "learning_rate": 1.54763243635905e-06,
+      "loss": 0.054,
+      "step": 3581
+    },
+    {
+      "epoch": 0.750785998742402,
+      "grad_norm": 0.02083863690495491,
+      "learning_rate": 1.5451775427108302e-06,
+      "loss": 0.0566,
+      "step": 3582
+    },
+    {
+      "epoch": 0.7509955984070426,
+      "grad_norm": 0.021793080493807793,
+      "learning_rate": 1.5427242417350474e-06,
+      "loss": 0.0539,
+      "step": 3583
+    },
+    {
+      "epoch": 0.7512051980716831,
+      "grad_norm": 0.024363208562135696,
+      "learning_rate": 1.540272534562669e-06,
+      "loss": 0.0542,
+      "step": 3584
+    },
+    {
+      "epoch": 0.7514147977363236,
+      "grad_norm": 0.020569128915667534,
+      "learning_rate": 1.5378224223239341e-06,
+      "loss": 0.0554,
+      "step": 3585
+    },
+    {
+      "epoch": 0.7516243974009642,
+      "grad_norm": 0.02453795075416565,
+      "learning_rate": 1.5353739061483446e-06,
+      "loss": 0.0513,
+      "step": 3586
+    },
+    {
+      "epoch": 0.7518339970656047,
+      "grad_norm": 0.022484583780169487,
+      "learning_rate": 1.5329269871646646e-06,
+      "loss": 0.0539,
+      "step": 3587
+    },
+    {
+      "epoch": 0.7520435967302452,
+      "grad_norm": 0.021373063325881958,
+      "learning_rate": 1.530481666500922e-06,
+      "loss": 0.0554,
+      "step": 3588
+    },
+    {
+      "epoch": 0.7522531963948857,
+      "grad_norm": 0.026554672047495842,
+      "learning_rate": 1.5280379452844124e-06,
+      "loss": 0.0527,
+      "step": 3589
+    },
+    {
+      "epoch": 0.7524627960595263,
+      "grad_norm": 0.023381482809782028,
+      "learning_rate": 1.525595824641687e-06,
+      "loss": 0.0568,
+      "step": 3590
+    },
+    {
+      "epoch": 0.7526723957241669,
+      "grad_norm": 0.027381494641304016,
+      "learning_rate": 1.5231553056985642e-06,
+      "loss": 0.0522,
+      "step": 3591
+    },
+    {
+      "epoch": 0.7528819953888074,
+      "grad_norm": 0.03211834654211998,
+      "learning_rate": 1.5207163895801252e-06,
+      "loss": 0.0554,
+      "step": 3592
+    },
+    {
+      "epoch": 0.7530915950534479,
+      "grad_norm": 0.021453700959682465,
+      "learning_rate": 1.5182790774107082e-06,
+      "loss": 0.0549,
+      "step": 3593
+    },
+    {
+      "epoch": 0.7533011947180884,
+      "grad_norm": 0.037080686539411545,
+      "learning_rate": 1.5158433703139114e-06,
+      "loss": 0.0532,
+      "step": 3594
+    },
+    {
+      "epoch": 0.7535107943827289,
+      "grad_norm": 0.02473445050418377,
+      "learning_rate": 1.5134092694125968e-06,
+      "loss": 0.0549,
+      "step": 3595
+    },
+    {
+      "epoch": 0.7537203940473696,
+      "grad_norm": 0.02276427671313286,
+      "learning_rate": 1.510976775828887e-06,
+      "loss": 0.0568,
+      "step": 3596
+    },
+    {
+      "epoch": 0.7539299937120101,
+      "grad_norm": 0.020384134724736214,
+      "learning_rate": 1.508545890684157e-06,
+      "loss": 0.054,
+      "step": 3597
+    },
+    {
+      "epoch": 0.7541395933766506,
+      "grad_norm": 0.025425709784030914,
+      "learning_rate": 1.5061166150990475e-06,
+      "loss": 0.0536,
+      "step": 3598
+    },
+    {
+      "epoch": 0.7543491930412911,
+      "grad_norm": 0.022402847185730934,
+      "learning_rate": 1.5036889501934533e-06,
+      "loss": 0.0543,
+      "step": 3599
+    },
+    {
+      "epoch": 0.7545587927059316,
+      "grad_norm": 0.02631525509059429,
+      "learning_rate": 1.5012628970865245e-06,
+      "loss": 0.0548,
+      "step": 3600
+    },
+    {
+      "epoch": 0.7547683923705722,
+      "grad_norm": 0.019039105623960495,
+      "learning_rate": 1.498838456896674e-06,
+      "loss": 0.0555,
+      "step": 3601
+    },
+    {
+      "epoch": 0.7549779920352128,
+      "grad_norm": 0.028720203787088394,
+      "learning_rate": 1.4964156307415673e-06,
+      "loss": 0.0541,
+      "step": 3602
+    },
+    {
+      "epoch": 0.7551875916998533,
+      "grad_norm": 0.02376984804868698,
+      "learning_rate": 1.493994419738129e-06,
+      "loss": 0.053,
+      "step": 3603
+    },
+    {
+      "epoch": 0.7553971913644938,
+      "grad_norm": 0.028366727754473686,
+      "learning_rate": 1.4915748250025346e-06,
+      "loss": 0.0518,
+      "step": 3604
+    },
+    {
+      "epoch": 0.7556067910291343,
+      "grad_norm": 0.024544060230255127,
+      "learning_rate": 1.4891568476502154e-06,
+      "loss": 0.0542,
+      "step": 3605
+    },
+    {
+      "epoch": 0.7558163906937749,
+      "grad_norm": 0.029393775388598442,
+      "learning_rate": 1.486740488795862e-06,
+      "loss": 0.0566,
+      "step": 3606
+    },
+    {
+      "epoch": 0.7560259903584154,
+      "grad_norm": 0.030081739649176598,
+      "learning_rate": 1.484325749553412e-06,
+      "loss": 0.054,
+      "step": 3607
+    },
+    {
+      "epoch": 0.756235590023056,
+      "grad_norm": 0.03067159280180931,
+      "learning_rate": 1.4819126310360626e-06,
+      "loss": 0.0546,
+      "step": 3608
+    },
+    {
+      "epoch": 0.7564451896876965,
+      "grad_norm": 0.02665751427412033,
+      "learning_rate": 1.4795011343562594e-06,
+      "loss": 0.0545,
+      "step": 3609
+    },
+    {
+      "epoch": 0.756654789352337,
+      "grad_norm": 0.025258727371692657,
+      "learning_rate": 1.4770912606257003e-06,
+      "loss": 0.0552,
+      "step": 3610
+    },
+    {
+      "epoch": 0.7568643890169776,
+      "grad_norm": 0.028663890436291695,
+      "learning_rate": 1.4746830109553388e-06,
+      "loss": 0.0521,
+      "step": 3611
+    },
+    {
+      "epoch": 0.7570739886816181,
+      "grad_norm": 0.02796069160103798,
+      "learning_rate": 1.472276386455378e-06,
+      "loss": 0.0549,
+      "step": 3612
+    },
+    {
+      "epoch": 0.7572835883462586,
+      "grad_norm": 0.028017984703183174,
+      "learning_rate": 1.4698713882352694e-06,
+      "loss": 0.0551,
+      "step": 3613
+    },
+    {
+      "epoch": 0.7574931880108992,
+      "grad_norm": 0.036044858396053314,
+      "learning_rate": 1.4674680174037186e-06,
+      "loss": 0.0549,
+      "step": 3614
+    },
+    {
+      "epoch": 0.7577027876755397,
+      "grad_norm": 0.024651629850268364,
+      "learning_rate": 1.465066275068676e-06,
+      "loss": 0.0535,
+      "step": 3615
+    },
+    {
+      "epoch": 0.7579123873401803,
+      "grad_norm": 0.02148495241999626,
+      "learning_rate": 1.462666162337349e-06,
+      "loss": 0.0529,
+      "step": 3616
+    },
+    {
+      "epoch": 0.7581219870048208,
+      "grad_norm": 0.023804927244782448,
+      "learning_rate": 1.4602676803161842e-06,
+      "loss": 0.0557,
+      "step": 3617
+    },
+    {
+      "epoch": 0.7583315866694613,
+      "grad_norm": 0.017874160781502724,
+      "learning_rate": 1.4578708301108835e-06,
+      "loss": 0.0548,
+      "step": 3618
+    },
+    {
+      "epoch": 0.7585411863341018,
+      "grad_norm": 0.021595729514956474,
+      "learning_rate": 1.4554756128263958e-06,
+      "loss": 0.0567,
+      "step": 3619
+    },
+    {
+      "epoch": 0.7587507859987423,
+      "grad_norm": 0.022326629608869553,
+      "learning_rate": 1.4530820295669145e-06,
+      "loss": 0.0534,
+      "step": 3620
+    },
+    {
+      "epoch": 0.758960385663383,
+      "grad_norm": 0.021938448771834373,
+      "learning_rate": 1.4506900814358794e-06,
+      "loss": 0.0532,
+      "step": 3621
+    },
+    {
+      "epoch": 0.7591699853280235,
+      "grad_norm": 0.02145432122051716,
+      "learning_rate": 1.4482997695359807e-06,
+      "loss": 0.0548,
+      "step": 3622
+    },
+    {
+      "epoch": 0.759379584992664,
+      "grad_norm": 0.019530145451426506,
+      "learning_rate": 1.445911094969149e-06,
+      "loss": 0.0541,
+      "step": 3623
+    },
+    {
+      "epoch": 0.7595891846573045,
+      "grad_norm": 0.01897241175174713,
+      "learning_rate": 1.4435240588365645e-06,
+      "loss": 0.0528,
+      "step": 3624
+    },
+    {
+      "epoch": 0.759798784321945,
+      "grad_norm": 0.02352256327867508,
+      "learning_rate": 1.4411386622386519e-06,
+      "loss": 0.0556,
+      "step": 3625
+    },
+    {
+      "epoch": 0.7600083839865857,
+      "grad_norm": 0.016778651624917984,
+      "learning_rate": 1.4387549062750767e-06,
+      "loss": 0.055,
+      "step": 3626
+    },
+    {
+      "epoch": 0.7602179836512262,
+      "grad_norm": 0.02301890030503273,
+      "learning_rate": 1.4363727920447478e-06,
+      "loss": 0.053,
+      "step": 3627
+    },
+    {
+      "epoch": 0.7604275833158667,
+      "grad_norm": 0.020533403381705284,
+      "learning_rate": 1.4339923206458222e-06,
+      "loss": 0.0549,
+      "step": 3628
+    },
+    {
+      "epoch": 0.7606371829805072,
+      "grad_norm": 0.02557838149368763,
+      "learning_rate": 1.431613493175697e-06,
+      "loss": 0.0544,
+      "step": 3629
+    },
+    {
+      "epoch": 0.7608467826451477,
+      "grad_norm": 0.026523873209953308,
+      "learning_rate": 1.4292363107310091e-06,
+      "loss": 0.0518,
+      "step": 3630
+    },
+    {
+      "epoch": 0.7610563823097883,
+      "grad_norm": 0.02507561445236206,
+      "learning_rate": 1.4268607744076419e-06,
+      "loss": 0.0542,
+      "step": 3631
+    },
+    {
+      "epoch": 0.7612659819744289,
+      "grad_norm": 0.034792933613061905,
+      "learning_rate": 1.424486885300715e-06,
+      "loss": 0.0514,
+      "step": 3632
+    },
+    {
+      "epoch": 0.7614755816390694,
+      "grad_norm": 0.025153324007987976,
+      "learning_rate": 1.4221146445045903e-06,
+      "loss": 0.0534,
+      "step": 3633
+    },
+    {
+      "epoch": 0.7616851813037099,
+      "grad_norm": 0.0239273589104414,
+      "learning_rate": 1.419744053112871e-06,
+      "loss": 0.0546,
+      "step": 3634
+    },
+    {
+      "epoch": 0.7618947809683505,
+      "grad_norm": 0.027409644797444344,
+      "learning_rate": 1.417375112218401e-06,
+      "loss": 0.053,
+      "step": 3635
+    },
+    {
+      "epoch": 0.762104380632991,
+      "grad_norm": 0.021910734474658966,
+      "learning_rate": 1.4150078229132586e-06,
+      "loss": 0.0506,
+      "step": 3636
+    },
+    {
+      "epoch": 0.7623139802976315,
+      "grad_norm": 0.028264936059713364,
+      "learning_rate": 1.4126421862887668e-06,
+      "loss": 0.0526,
+      "step": 3637
+    },
+    {
+      "epoch": 0.762523579962272,
+      "grad_norm": 0.017890270799398422,
+      "learning_rate": 1.41027820343548e-06,
+      "loss": 0.0564,
+      "step": 3638
+    },
+    {
+      "epoch": 0.7627331796269126,
+      "grad_norm": 0.02689182385802269,
+      "learning_rate": 1.4079158754431981e-06,
+      "loss": 0.0544,
+      "step": 3639
+    },
+    {
+      "epoch": 0.7629427792915532,
+      "grad_norm": 0.024087607860565186,
+      "learning_rate": 1.4055552034009496e-06,
+      "loss": 0.0558,
+      "step": 3640
+    },
+    {
+      "epoch": 0.7631523789561937,
+      "grad_norm": 0.019843678921461105,
+      "learning_rate": 1.4031961883970053e-06,
+      "loss": 0.0523,
+      "step": 3641
+    },
+    {
+      "epoch": 0.7633619786208342,
+      "grad_norm": 0.029412388801574707,
+      "learning_rate": 1.4008388315188743e-06,
+      "loss": 0.0544,
+      "step": 3642
+    },
+    {
+      "epoch": 0.7635715782854747,
+      "grad_norm": 0.022888310253620148,
+      "learning_rate": 1.3984831338532916e-06,
+      "loss": 0.0546,
+      "step": 3643
+    },
+    {
+      "epoch": 0.7637811779501152,
+      "grad_norm": 0.02272786572575569,
+      "learning_rate": 1.3961290964862356e-06,
+      "loss": 0.053,
+      "step": 3644
+    },
+    {
+      "epoch": 0.7639907776147559,
+      "grad_norm": 0.028499163687229156,
+      "learning_rate": 1.3937767205029196e-06,
+      "loss": 0.0532,
+      "step": 3645
+    },
+    {
+      "epoch": 0.7642003772793964,
+      "grad_norm": 0.02279391512274742,
+      "learning_rate": 1.3914260069877844e-06,
+      "loss": 0.0518,
+      "step": 3646
+    },
+    {
+      "epoch": 0.7644099769440369,
+      "grad_norm": 0.03257404640316963,
+      "learning_rate": 1.3890769570245122e-06,
+      "loss": 0.056,
+      "step": 3647
+    },
+    {
+      "epoch": 0.7646195766086774,
+      "grad_norm": 0.030943913385272026,
+      "learning_rate": 1.386729571696011e-06,
+      "loss": 0.0548,
+      "step": 3648
+    },
+    {
+      "epoch": 0.7648291762733179,
+      "grad_norm": 0.017855705693364143,
+      "learning_rate": 1.3843838520844288e-06,
+      "loss": 0.0545,
+      "step": 3649
+    },
+    {
+      "epoch": 0.7650387759379585,
+      "grad_norm": 0.029183411970734596,
+      "learning_rate": 1.3820397992711377e-06,
+      "loss": 0.0562,
+      "step": 3650
+    },
+    {
+      "epoch": 0.7652483756025991,
+      "grad_norm": 0.02148706465959549,
+      "learning_rate": 1.3796974143367475e-06,
+      "loss": 0.0561,
+      "step": 3651
+    },
+    {
+      "epoch": 0.7654579752672396,
+      "grad_norm": 0.018655741587281227,
+      "learning_rate": 1.3773566983610992e-06,
+      "loss": 0.0533,
+      "step": 3652
+    },
+    {
+      "epoch": 0.7656675749318801,
+      "grad_norm": 0.022678757086396217,
+      "learning_rate": 1.3750176524232605e-06,
+      "loss": 0.0535,
+      "step": 3653
+    },
+    {
+      "epoch": 0.7658771745965206,
+      "grad_norm": 0.02211586944758892,
+      "learning_rate": 1.372680277601529e-06,
+      "loss": 0.0545,
+      "step": 3654
+    },
+    {
+      "epoch": 0.7660867742611612,
+      "grad_norm": 0.019549356773495674,
+      "learning_rate": 1.3703445749734384e-06,
+      "loss": 0.0548,
+      "step": 3655
+    },
+    {
+      "epoch": 0.7662963739258017,
+      "grad_norm": 0.02576516941189766,
+      "learning_rate": 1.3680105456157427e-06,
+      "loss": 0.0552,
+      "step": 3656
+    },
+    {
+      "epoch": 0.7665059735904423,
+      "grad_norm": 0.024852564558386803,
+      "learning_rate": 1.3656781906044315e-06,
+      "loss": 0.0538,
+      "step": 3657
+    },
+    {
+      "epoch": 0.7667155732550828,
+      "grad_norm": 0.022067122161388397,
+      "learning_rate": 1.3633475110147204e-06,
+      "loss": 0.0551,
+      "step": 3658
+    },
+    {
+      "epoch": 0.7669251729197233,
+      "grad_norm": 0.029038861393928528,
+      "learning_rate": 1.3610185079210514e-06,
+      "loss": 0.0531,
+      "step": 3659
+    },
+    {
+      "epoch": 0.7671347725843639,
+      "grad_norm": 0.019402772188186646,
+      "learning_rate": 1.3586911823970933e-06,
+      "loss": 0.0537,
+      "step": 3660
+    },
+    {
+      "epoch": 0.7673443722490044,
+      "grad_norm": 0.024672850966453552,
+      "learning_rate": 1.3563655355157434e-06,
+      "loss": 0.0522,
+      "step": 3661
+    },
+    {
+      "epoch": 0.7675539719136449,
+      "grad_norm": 0.022120898589491844,
+      "learning_rate": 1.3540415683491265e-06,
+      "loss": 0.0542,
+      "step": 3662
+    },
+    {
+      "epoch": 0.7677635715782855,
+      "grad_norm": 0.021284013986587524,
+      "learning_rate": 1.3517192819685875e-06,
+      "loss": 0.0516,
+      "step": 3663
+    },
+    {
+      "epoch": 0.767973171242926,
+      "grad_norm": 0.024125682190060616,
+      "learning_rate": 1.3493986774447032e-06,
+      "loss": 0.0575,
+      "step": 3664
+    },
+    {
+      "epoch": 0.7681827709075666,
+      "grad_norm": 0.022346889600157738,
+      "learning_rate": 1.3470797558472709e-06,
+      "loss": 0.0539,
+      "step": 3665
+    },
+    {
+      "epoch": 0.7683923705722071,
+      "grad_norm": 0.022748976945877075,
+      "learning_rate": 1.34476251824531e-06,
+      "loss": 0.0563,
+      "step": 3666
+    },
+    {
+      "epoch": 0.7686019702368476,
+      "grad_norm": 0.02310675010085106,
+      "learning_rate": 1.3424469657070693e-06,
+      "loss": 0.0534,
+      "step": 3667
+    },
+    {
+      "epoch": 0.7688115699014881,
+      "grad_norm": 0.016234230250120163,
+      "learning_rate": 1.3401330993000195e-06,
+      "loss": 0.0568,
+      "step": 3668
+    },
+    {
+      "epoch": 0.7690211695661286,
+      "grad_norm": 0.02276422083377838,
+      "learning_rate": 1.3378209200908487e-06,
+      "loss": 0.0533,
+      "step": 3669
+    },
+    {
+      "epoch": 0.7692307692307693,
+      "grad_norm": 0.020701710134744644,
+      "learning_rate": 1.3355104291454751e-06,
+      "loss": 0.0522,
+      "step": 3670
+    },
+    {
+      "epoch": 0.7694403688954098,
+      "grad_norm": 0.017248544842004776,
+      "learning_rate": 1.3332016275290304e-06,
+      "loss": 0.0541,
+      "step": 3671
+    },
+    {
+      "epoch": 0.7696499685600503,
+      "grad_norm": 0.01811116933822632,
+      "learning_rate": 1.3308945163058757e-06,
+      "loss": 0.0532,
+      "step": 3672
+    },
+    {
+      "epoch": 0.7698595682246908,
+      "grad_norm": 0.017278380692005157,
+      "learning_rate": 1.3285890965395853e-06,
+      "loss": 0.0548,
+      "step": 3673
+    },
+    {
+      "epoch": 0.7700691678893313,
+      "grad_norm": 0.017872022464871407,
+      "learning_rate": 1.3262853692929583e-06,
+      "loss": 0.0554,
+      "step": 3674
+    },
+    {
+      "epoch": 0.770278767553972,
+      "grad_norm": 0.019185813143849373,
+      "learning_rate": 1.3239833356280152e-06,
+      "loss": 0.0519,
+      "step": 3675
+    },
+    {
+      "epoch": 0.7704883672186125,
+      "grad_norm": 0.0181709136813879,
+      "learning_rate": 1.3216829966059902e-06,
+      "loss": 0.0555,
+      "step": 3676
+    },
+    {
+      "epoch": 0.770697966883253,
+      "grad_norm": 0.020711593329906464,
+      "learning_rate": 1.3193843532873385e-06,
+      "loss": 0.0529,
+      "step": 3677
+    },
+    {
+      "epoch": 0.7709075665478935,
+      "grad_norm": 0.015872521325945854,
+      "learning_rate": 1.3170874067317362e-06,
+      "loss": 0.0539,
+      "step": 3678
+    },
+    {
+      "epoch": 0.771117166212534,
+      "grad_norm": 0.02129794843494892,
+      "learning_rate": 1.3147921579980739e-06,
+      "loss": 0.0543,
+      "step": 3679
+    },
+    {
+      "epoch": 0.7713267658771746,
+      "grad_norm": 0.015224761329591274,
+      "learning_rate": 1.3124986081444625e-06,
+      "loss": 0.0548,
+      "step": 3680
+    },
+    {
+      "epoch": 0.7715363655418152,
+      "grad_norm": 0.016702933236956596,
+      "learning_rate": 1.3102067582282264e-06,
+      "loss": 0.0549,
+      "step": 3681
+    },
+    {
+      "epoch": 0.7717459652064557,
+      "grad_norm": 0.01623140461742878,
+      "learning_rate": 1.307916609305907e-06,
+      "loss": 0.0532,
+      "step": 3682
+    },
+    {
+      "epoch": 0.7719555648710962,
+      "grad_norm": 0.014785263687372208,
+      "learning_rate": 1.305628162433264e-06,
+      "loss": 0.055,
+      "step": 3683
+    },
+    {
+      "epoch": 0.7721651645357367,
+      "grad_norm": 0.014699485152959824,
+      "learning_rate": 1.3033414186652705e-06,
+      "loss": 0.0561,
+      "step": 3684
+    },
+    {
+      "epoch": 0.7723747642003773,
+      "grad_norm": 0.017208058387041092,
+      "learning_rate": 1.3010563790561165e-06,
+      "loss": 0.0548,
+      "step": 3685
+    },
+    {
+      "epoch": 0.7725843638650178,
+      "grad_norm": 0.019409308210015297,
+      "learning_rate": 1.298773044659204e-06,
+      "loss": 0.0524,
+      "step": 3686
+    },
+    {
+      "epoch": 0.7727939635296583,
+      "grad_norm": 0.015810973942279816,
+      "learning_rate": 1.296491416527147e-06,
+      "loss": 0.0546,
+      "step": 3687
+    },
+    {
+      "epoch": 0.7730035631942989,
+      "grad_norm": 0.020034223794937134,
+      "learning_rate": 1.2942114957117797e-06,
+      "loss": 0.0545,
+      "step": 3688
+    },
+    {
+      "epoch": 0.7732131628589394,
+      "grad_norm": 0.021472373977303505,
+      "learning_rate": 1.2919332832641413e-06,
+      "loss": 0.0534,
+      "step": 3689
+    },
+    {
+      "epoch": 0.77342276252358,
+      "grad_norm": 0.012456726282835007,
+      "learning_rate": 1.2896567802344888e-06,
+      "loss": 0.0529,
+      "step": 3690
+    },
+    {
+      "epoch": 0.7736323621882205,
+      "grad_norm": 0.018201405182480812,
+      "learning_rate": 1.287381987672292e-06,
+      "loss": 0.0526,
+      "step": 3691
+    },
+    {
+      "epoch": 0.773841961852861,
+      "grad_norm": 0.012917966581881046,
+      "learning_rate": 1.2851089066262272e-06,
+      "loss": 0.0538,
+      "step": 3692
+    },
+    {
+      "epoch": 0.7740515615175015,
+      "grad_norm": 0.019098002463579178,
+      "learning_rate": 1.2828375381441837e-06,
+      "loss": 0.055,
+      "step": 3693
+    },
+    {
+      "epoch": 0.774261161182142,
+      "grad_norm": 0.015969131141901016,
+      "learning_rate": 1.2805678832732627e-06,
+      "loss": 0.0544,
+      "step": 3694
+    },
+    {
+      "epoch": 0.7744707608467827,
+      "grad_norm": 0.016314983367919922,
+      "learning_rate": 1.2782999430597764e-06,
+      "loss": 0.0513,
+      "step": 3695
+    },
+    {
+      "epoch": 0.7746803605114232,
+      "grad_norm": 0.021401531994342804,
+      "learning_rate": 1.2760337185492423e-06,
+      "loss": 0.0557,
+      "step": 3696
+    },
+    {
+      "epoch": 0.7748899601760637,
+      "grad_norm": 0.015217507258057594,
+      "learning_rate": 1.2737692107863914e-06,
+      "loss": 0.0566,
+      "step": 3697
+    },
+    {
+      "epoch": 0.7750995598407042,
+      "grad_norm": 0.013445749878883362,
+      "learning_rate": 1.2715064208151606e-06,
+      "loss": 0.0534,
+      "step": 3698
+    },
+    {
+      "epoch": 0.7753091595053447,
+      "grad_norm": 0.01527884230017662,
+      "learning_rate": 1.2692453496786933e-06,
+      "loss": 0.0554,
+      "step": 3699
+    },
+    {
+      "epoch": 0.7755187591699854,
+      "grad_norm": 0.013215204700827599,
+      "learning_rate": 1.2669859984193456e-06,
+      "loss": 0.05,
+      "step": 3700
+    },
+    {
+      "epoch": 0.7757283588346259,
+      "grad_norm": 0.013395383022725582,
+      "learning_rate": 1.264728368078678e-06,
+      "loss": 0.0579,
+      "step": 3701
+    },
+    {
+      "epoch": 0.7759379584992664,
+      "grad_norm": 0.013562247157096863,
+      "learning_rate": 1.2624724596974557e-06,
+      "loss": 0.0546,
+      "step": 3702
+    },
+    {
+      "epoch": 0.7761475581639069,
+      "grad_norm": 0.0144569780677557,
+      "learning_rate": 1.260218274315655e-06,
+      "loss": 0.0544,
+      "step": 3703
+    },
+    {
+      "epoch": 0.7763571578285475,
+      "grad_norm": 0.014783012680709362,
+      "learning_rate": 1.2579658129724526e-06,
+      "loss": 0.0539,
+      "step": 3704
+    },
+    {
+      "epoch": 0.776566757493188,
+      "grad_norm": 0.014723209664225578,
+      "learning_rate": 1.2557150767062315e-06,
+      "loss": 0.0554,
+      "step": 3705
+    },
+    {
+      "epoch": 0.7767763571578286,
+      "grad_norm": 0.016170332208275795,
+      "learning_rate": 1.2534660665545822e-06,
+      "loss": 0.0526,
+      "step": 3706
+    },
+    {
+      "epoch": 0.7769859568224691,
+      "grad_norm": 0.019932882860302925,
+      "learning_rate": 1.2512187835542982e-06,
+      "loss": 0.0526,
+      "step": 3707
+    },
+    {
+      "epoch": 0.7771955564871096,
+      "grad_norm": 0.016744259744882584,
+      "learning_rate": 1.248973228741378e-06,
+      "loss": 0.0535,
+      "step": 3708
+    },
+    {
+      "epoch": 0.7774051561517502,
+      "grad_norm": 0.016560077667236328,
+      "learning_rate": 1.2467294031510202e-06,
+      "loss": 0.0526,
+      "step": 3709
+    },
+    {
+      "epoch": 0.7776147558163907,
+      "grad_norm": 0.016798479482531548,
+      "learning_rate": 1.2444873078176262e-06,
+      "loss": 0.0535,
+      "step": 3710
+    },
+    {
+      "epoch": 0.7778243554810312,
+      "grad_norm": 0.01939672976732254,
+      "learning_rate": 1.2422469437748046e-06,
+      "loss": 0.0531,
+      "step": 3711
+    },
+    {
+      "epoch": 0.7780339551456718,
+      "grad_norm": 0.019812364131212234,
+      "learning_rate": 1.2400083120553602e-06,
+      "loss": 0.0537,
+      "step": 3712
+    },
+    {
+      "epoch": 0.7782435548103123,
+      "grad_norm": 0.0169843677431345,
+      "learning_rate": 1.2377714136913028e-06,
+      "loss": 0.0537,
+      "step": 3713
+    },
+    {
+      "epoch": 0.7784531544749529,
+      "grad_norm": 0.019338861107826233,
+      "learning_rate": 1.235536249713845e-06,
+      "loss": 0.0538,
+      "step": 3714
+    },
+    {
+      "epoch": 0.7786627541395934,
+      "grad_norm": 0.018493298441171646,
+      "learning_rate": 1.2333028211533916e-06,
+      "loss": 0.0527,
+      "step": 3715
+    },
+    {
+      "epoch": 0.7788723538042339,
+      "grad_norm": 0.0183496605604887,
+      "learning_rate": 1.2310711290395545e-06,
+      "loss": 0.0549,
+      "step": 3716
+    },
+    {
+      "epoch": 0.7790819534688744,
+      "grad_norm": 0.019363224506378174,
+      "learning_rate": 1.2288411744011464e-06,
+      "loss": 0.0513,
+      "step": 3717
+    },
+    {
+      "epoch": 0.779291553133515,
+      "grad_norm": 0.015527610667049885,
+      "learning_rate": 1.2266129582661712e-06,
+      "loss": 0.0545,
+      "step": 3718
+    },
+    {
+      "epoch": 0.7795011527981556,
+      "grad_norm": 0.018898479640483856,
+      "learning_rate": 1.2243864816618407e-06,
+      "loss": 0.0539,
+      "step": 3719
+    },
+    {
+      "epoch": 0.7797107524627961,
+      "grad_norm": 0.014554295688867569,
+      "learning_rate": 1.2221617456145556e-06,
+      "loss": 0.0532,
+      "step": 3720
+    },
+    {
+      "epoch": 0.7799203521274366,
+      "grad_norm": 0.01795900985598564,
+      "learning_rate": 1.2199387511499234e-06,
+      "loss": 0.0542,
+      "step": 3721
+    },
+    {
+      "epoch": 0.7801299517920771,
+      "grad_norm": 0.018716957420110703,
+      "learning_rate": 1.21771749929274e-06,
+      "loss": 0.0534,
+      "step": 3722
+    },
+    {
+      "epoch": 0.7803395514567176,
+      "grad_norm": 0.015783434733748436,
+      "learning_rate": 1.2154979910670033e-06,
+      "loss": 0.055,
+      "step": 3723
+    },
+    {
+      "epoch": 0.7805491511213583,
+      "grad_norm": 0.016636742278933525,
+      "learning_rate": 1.2132802274959082e-06,
+      "loss": 0.0527,
+      "step": 3724
+    },
+    {
+      "epoch": 0.7807587507859988,
+      "grad_norm": 0.017774462699890137,
+      "learning_rate": 1.2110642096018421e-06,
+      "loss": 0.0556,
+      "step": 3725
+    },
+    {
+      "epoch": 0.7809683504506393,
+      "grad_norm": 0.01745484583079815,
+      "learning_rate": 1.2088499384063868e-06,
+      "loss": 0.0521,
+      "step": 3726
+    },
+    {
+      "epoch": 0.7811779501152798,
+      "grad_norm": 0.018969282507896423,
+      "learning_rate": 1.2066374149303234e-06,
+      "loss": 0.0524,
+      "step": 3727
+    },
+    {
+      "epoch": 0.7813875497799203,
+      "grad_norm": 0.0174677986651659,
+      "learning_rate": 1.2044266401936228e-06,
+      "loss": 0.0541,
+      "step": 3728
+    },
+    {
+      "epoch": 0.7815971494445609,
+      "grad_norm": 0.01889014057815075,
+      "learning_rate": 1.2022176152154525e-06,
+      "loss": 0.0532,
+      "step": 3729
+    },
+    {
+      "epoch": 0.7818067491092014,
+      "grad_norm": 0.020782019942998886,
+      "learning_rate": 1.200010341014174e-06,
+      "loss": 0.0517,
+      "step": 3730
+    },
+    {
+      "epoch": 0.782016348773842,
+      "grad_norm": 0.017878524959087372,
+      "learning_rate": 1.1978048186073388e-06,
+      "loss": 0.0548,
+      "step": 3731
+    },
+    {
+      "epoch": 0.7822259484384825,
+      "grad_norm": 0.01691311225295067,
+      "learning_rate": 1.1956010490116915e-06,
+      "loss": 0.0516,
+      "step": 3732
+    },
+    {
+      "epoch": 0.782435548103123,
+      "grad_norm": 0.014464441686868668,
+      "learning_rate": 1.1933990332431699e-06,
+      "loss": 0.055,
+      "step": 3733
+    },
+    {
+      "epoch": 0.7826451477677636,
+      "grad_norm": 0.02182583324611187,
+      "learning_rate": 1.191198772316905e-06,
+      "loss": 0.0515,
+      "step": 3734
+    },
+    {
+      "epoch": 0.7828547474324041,
+      "grad_norm": 0.01545246597379446,
+      "learning_rate": 1.1890002672472133e-06,
+      "loss": 0.0551,
+      "step": 3735
+    },
+    {
+      "epoch": 0.7830643470970446,
+      "grad_norm": 0.022399727255105972,
+      "learning_rate": 1.1868035190476085e-06,
+      "loss": 0.0572,
+      "step": 3736
+    },
+    {
+      "epoch": 0.7832739467616852,
+      "grad_norm": 0.01581631414592266,
+      "learning_rate": 1.184608528730789e-06,
+      "loss": 0.0539,
+      "step": 3737
+    },
+    {
+      "epoch": 0.7834835464263257,
+      "grad_norm": 0.022165510803461075,
+      "learning_rate": 1.1824152973086444e-06,
+      "loss": 0.0579,
+      "step": 3738
+    },
+    {
+      "epoch": 0.7836931460909663,
+      "grad_norm": 0.018284514546394348,
+      "learning_rate": 1.1802238257922543e-06,
+      "loss": 0.056,
+      "step": 3739
+    },
+    {
+      "epoch": 0.7839027457556068,
+      "grad_norm": 0.017083924263715744,
+      "learning_rate": 1.1780341151918883e-06,
+      "loss": 0.0517,
+      "step": 3740
+    },
+    {
+      "epoch": 0.7841123454202473,
+      "grad_norm": 0.01769246533513069,
+      "learning_rate": 1.1758461665170001e-06,
+      "loss": 0.0558,
+      "step": 3741
+    },
+    {
+      "epoch": 0.7843219450848878,
+      "grad_norm": 0.018215294927358627,
+      "learning_rate": 1.1736599807762366e-06,
+      "loss": 0.0544,
+      "step": 3742
+    },
+    {
+      "epoch": 0.7845315447495284,
+      "grad_norm": 0.018238717690110207,
+      "learning_rate": 1.1714755589774252e-06,
+      "loss": 0.0523,
+      "step": 3743
+    },
+    {
+      "epoch": 0.784741144414169,
+      "grad_norm": 0.01696399226784706,
+      "learning_rate": 1.1692929021275874e-06,
+      "loss": 0.0542,
+      "step": 3744
+    },
+    {
+      "epoch": 0.7849507440788095,
+      "grad_norm": 0.020423993468284607,
+      "learning_rate": 1.1671120112329248e-06,
+      "loss": 0.0547,
+      "step": 3745
+    },
+    {
+      "epoch": 0.78516034374345,
+      "grad_norm": 0.016111532226204872,
+      "learning_rate": 1.1649328872988286e-06,
+      "loss": 0.0546,
+      "step": 3746
+    },
+    {
+      "epoch": 0.7853699434080905,
+      "grad_norm": 0.023867281153798103,
+      "learning_rate": 1.1627555313298777e-06,
+      "loss": 0.0544,
+      "step": 3747
+    },
+    {
+      "epoch": 0.785579543072731,
+      "grad_norm": 0.017127446830272675,
+      "learning_rate": 1.160579944329827e-06,
+      "loss": 0.0551,
+      "step": 3748
+    },
+    {
+      "epoch": 0.7857891427373717,
+      "grad_norm": 0.0323978066444397,
+      "learning_rate": 1.1584061273016245e-06,
+      "loss": 0.0584,
+      "step": 3749
+    },
+    {
+      "epoch": 0.7859987424020122,
+      "grad_norm": 0.015841031447052956,
+      "learning_rate": 1.1562340812474004e-06,
+      "loss": 0.0511,
+      "step": 3750
+    },
+    {
+      "epoch": 0.7862083420666527,
+      "grad_norm": 0.03065073862671852,
+      "learning_rate": 1.154063807168465e-06,
+      "loss": 0.0526,
+      "step": 3751
+    },
+    {
+      "epoch": 0.7864179417312932,
+      "grad_norm": 0.02133885584771633,
+      "learning_rate": 1.1518953060653177e-06,
+      "loss": 0.054,
+      "step": 3752
+    },
+    {
+      "epoch": 0.7866275413959337,
+      "grad_norm": 0.02793465554714203,
+      "learning_rate": 1.1497285789376327e-06,
+      "loss": 0.0542,
+      "step": 3753
+    },
+    {
+      "epoch": 0.7868371410605743,
+      "grad_norm": 0.024401402100920677,
+      "learning_rate": 1.1475636267842754e-06,
+      "loss": 0.0541,
+      "step": 3754
+    },
+    {
+      "epoch": 0.7870467407252149,
+      "grad_norm": 0.02825610339641571,
+      "learning_rate": 1.145400450603284e-06,
+      "loss": 0.053,
+      "step": 3755
+    },
+    {
+      "epoch": 0.7872563403898554,
+      "grad_norm": 0.031988829374313354,
+      "learning_rate": 1.143239051391884e-06,
+      "loss": 0.0519,
+      "step": 3756
+    },
+    {
+      "epoch": 0.7874659400544959,
+      "grad_norm": 0.01913175918161869,
+      "learning_rate": 1.1410794301464817e-06,
+      "loss": 0.0559,
+      "step": 3757
+    },
+    {
+      "epoch": 0.7876755397191364,
+      "grad_norm": 0.02814309112727642,
+      "learning_rate": 1.1389215878626608e-06,
+      "loss": 0.0569,
+      "step": 3758
+    },
+    {
+      "epoch": 0.787885139383777,
+      "grad_norm": 0.019799359142780304,
+      "learning_rate": 1.1367655255351845e-06,
+      "loss": 0.0542,
+      "step": 3759
+    },
+    {
+      "epoch": 0.7880947390484175,
+      "grad_norm": 0.023299338296055794,
+      "learning_rate": 1.1346112441579998e-06,
+      "loss": 0.0548,
+      "step": 3760
+    },
+    {
+      "epoch": 0.788304338713058,
+      "grad_norm": 0.022608846426010132,
+      "learning_rate": 1.132458744724227e-06,
+      "loss": 0.0527,
+      "step": 3761
+    },
+    {
+      "epoch": 0.7885139383776986,
+      "grad_norm": 0.03028971515595913,
+      "learning_rate": 1.1303080282261698e-06,
+      "loss": 0.0561,
+      "step": 3762
+    },
+    {
+      "epoch": 0.7887235380423391,
+      "grad_norm": 0.019988486543297768,
+      "learning_rate": 1.128159095655309e-06,
+      "loss": 0.0526,
+      "step": 3763
+    },
+    {
+      "epoch": 0.7889331377069797,
+      "grad_norm": 0.036664996296167374,
+      "learning_rate": 1.1260119480023008e-06,
+      "loss": 0.0525,
+      "step": 3764
+    },
+    {
+      "epoch": 0.7891427373716202,
+      "grad_norm": 0.02409091591835022,
+      "learning_rate": 1.1238665862569786e-06,
+      "loss": 0.0527,
+      "step": 3765
+    },
+    {
+      "epoch": 0.7893523370362607,
+      "grad_norm": 0.02567540481686592,
+      "learning_rate": 1.121723011408355e-06,
+      "loss": 0.054,
+      "step": 3766
+    },
+    {
+      "epoch": 0.7895619367009012,
+      "grad_norm": 0.028370145708322525,
+      "learning_rate": 1.1195812244446185e-06,
+      "loss": 0.0561,
+      "step": 3767
+    },
+    {
+      "epoch": 0.7897715363655418,
+      "grad_norm": 0.015024359337985516,
+      "learning_rate": 1.117441226353131e-06,
+      "loss": 0.0532,
+      "step": 3768
+    },
+    {
+      "epoch": 0.7899811360301824,
+      "grad_norm": 0.02175975777208805,
+      "learning_rate": 1.115303018120432e-06,
+      "loss": 0.0534,
+      "step": 3769
+    },
+    {
+      "epoch": 0.7901907356948229,
+      "grad_norm": 0.01616404764354229,
+      "learning_rate": 1.1131666007322356e-06,
+      "loss": 0.0545,
+      "step": 3770
+    },
+    {
+      "epoch": 0.7904003353594634,
+      "grad_norm": 0.01787029393017292,
+      "learning_rate": 1.1110319751734271e-06,
+      "loss": 0.0541,
+      "step": 3771
+    },
+    {
+      "epoch": 0.7906099350241039,
+      "grad_norm": 0.015742896124720573,
+      "learning_rate": 1.1088991424280705e-06,
+      "loss": 0.0542,
+      "step": 3772
+    },
+    {
+      "epoch": 0.7908195346887446,
+      "grad_norm": 0.018519075587391853,
+      "learning_rate": 1.106768103479402e-06,
+      "loss": 0.0527,
+      "step": 3773
+    },
+    {
+      "epoch": 0.7910291343533851,
+      "grad_norm": 0.012830471619963646,
+      "learning_rate": 1.1046388593098284e-06,
+      "loss": 0.0523,
+      "step": 3774
+    },
+    {
+      "epoch": 0.7912387340180256,
+      "grad_norm": 0.018937064334750175,
+      "learning_rate": 1.1025114109009321e-06,
+      "loss": 0.0537,
+      "step": 3775
+    },
+    {
+      "epoch": 0.7914483336826661,
+      "grad_norm": 0.016385626047849655,
+      "learning_rate": 1.100385759233465e-06,
+      "loss": 0.0537,
+      "step": 3776
+    },
+    {
+      "epoch": 0.7916579333473066,
+      "grad_norm": 0.01458564680069685,
+      "learning_rate": 1.098261905287354e-06,
+      "loss": 0.0535,
+      "step": 3777
+    },
+    {
+      "epoch": 0.7918675330119472,
+      "grad_norm": 0.01804785430431366,
+      "learning_rate": 1.0961398500416926e-06,
+      "loss": 0.0504,
+      "step": 3778
+    },
+    {
+      "epoch": 0.7920771326765877,
+      "grad_norm": 0.015284632332623005,
+      "learning_rate": 1.0940195944747494e-06,
+      "loss": 0.0537,
+      "step": 3779
+    },
+    {
+      "epoch": 0.7922867323412283,
+      "grad_norm": 0.013636418618261814,
+      "learning_rate": 1.091901139563964e-06,
+      "loss": 0.0531,
+      "step": 3780
+    },
+    {
+      "epoch": 0.7924963320058688,
+      "grad_norm": 0.013568844646215439,
+      "learning_rate": 1.0897844862859407e-06,
+      "loss": 0.0519,
+      "step": 3781
+    },
+    {
+      "epoch": 0.7927059316705093,
+      "grad_norm": 0.015443854033946991,
+      "learning_rate": 1.0876696356164556e-06,
+      "loss": 0.0566,
+      "step": 3782
+    },
+    {
+      "epoch": 0.7929155313351499,
+      "grad_norm": 0.013935340568423271,
+      "learning_rate": 1.0855565885304575e-06,
+      "loss": 0.0503,
+      "step": 3783
+    },
+    {
+      "epoch": 0.7931251309997904,
+      "grad_norm": 0.01774810254573822,
+      "learning_rate": 1.0834453460020577e-06,
+      "loss": 0.0548,
+      "step": 3784
+    },
+    {
+      "epoch": 0.7933347306644309,
+      "grad_norm": 0.01960287056863308,
+      "learning_rate": 1.0813359090045412e-06,
+      "loss": 0.0535,
+      "step": 3785
+    },
+    {
+      "epoch": 0.7935443303290715,
+      "grad_norm": 0.014412653632462025,
+      "learning_rate": 1.0792282785103565e-06,
+      "loss": 0.0538,
+      "step": 3786
+    },
+    {
+      "epoch": 0.793753929993712,
+      "grad_norm": 0.01678672805428505,
+      "learning_rate": 1.0771224554911197e-06,
+      "loss": 0.0539,
+      "step": 3787
+    },
+    {
+      "epoch": 0.7939635296583526,
+      "grad_norm": 0.017591232433915138,
+      "learning_rate": 1.0750184409176156e-06,
+      "loss": 0.0577,
+      "step": 3788
+    },
+    {
+      "epoch": 0.7941731293229931,
+      "grad_norm": 0.025450995191931725,
+      "learning_rate": 1.0729162357597956e-06,
+      "loss": 0.0556,
+      "step": 3789
+    },
+    {
+      "epoch": 0.7943827289876336,
+      "grad_norm": 0.01504075713455677,
+      "learning_rate": 1.0708158409867763e-06,
+      "loss": 0.0564,
+      "step": 3790
+    },
+    {
+      "epoch": 0.7945923286522741,
+      "grad_norm": 0.029961617663502693,
+      "learning_rate": 1.0687172575668381e-06,
+      "loss": 0.0532,
+      "step": 3791
+    },
+    {
+      "epoch": 0.7948019283169147,
+      "grad_norm": 0.01688455045223236,
+      "learning_rate": 1.0666204864674263e-06,
+      "loss": 0.0561,
+      "step": 3792
+    },
+    {
+      "epoch": 0.7950115279815553,
+      "grad_norm": 0.022862810641527176,
+      "learning_rate": 1.0645255286551548e-06,
+      "loss": 0.0555,
+      "step": 3793
+    },
+    {
+      "epoch": 0.7952211276461958,
+      "grad_norm": 0.016880135983228683,
+      "learning_rate": 1.0624323850957952e-06,
+      "loss": 0.0528,
+      "step": 3794
+    },
+    {
+      "epoch": 0.7954307273108363,
+      "grad_norm": 0.02419593743979931,
+      "learning_rate": 1.0603410567542882e-06,
+      "loss": 0.054,
+      "step": 3795
+    },
+    {
+      "epoch": 0.7956403269754768,
+      "grad_norm": 0.02624059095978737,
+      "learning_rate": 1.0582515445947377e-06,
+      "loss": 0.0534,
+      "step": 3796
+    },
+    {
+      "epoch": 0.7958499266401173,
+      "grad_norm": 0.013864974491298199,
+      "learning_rate": 1.056163849580406e-06,
+      "loss": 0.0516,
+      "step": 3797
+    },
+    {
+      "epoch": 0.796059526304758,
+      "grad_norm": 0.018372228369116783,
+      "learning_rate": 1.0540779726737187e-06,
+      "loss": 0.0544,
+      "step": 3798
+    },
+    {
+      "epoch": 0.7962691259693985,
+      "grad_norm": 0.017514994367957115,
+      "learning_rate": 1.0519939148362667e-06,
+      "loss": 0.0556,
+      "step": 3799
+    },
+    {
+      "epoch": 0.796478725634039,
+      "grad_norm": 0.026039427146315575,
+      "learning_rate": 1.0499116770288015e-06,
+      "loss": 0.0531,
+      "step": 3800
+    },
+    {
+      "epoch": 0.7966883252986795,
+      "grad_norm": 0.019942620769143105,
+      "learning_rate": 1.0478312602112312e-06,
+      "loss": 0.0527,
+      "step": 3801
+    },
+    {
+      "epoch": 0.79689792496332,
+      "grad_norm": 0.03383920341730118,
+      "learning_rate": 1.0457526653426303e-06,
+      "loss": 0.0536,
+      "step": 3802
+    },
+    {
+      "epoch": 0.7971075246279606,
+      "grad_norm": 0.018814336508512497,
+      "learning_rate": 1.0436758933812292e-06,
+      "loss": 0.0547,
+      "step": 3803
+    },
+    {
+      "epoch": 0.7973171242926012,
+      "grad_norm": 0.035695578902959824,
+      "learning_rate": 1.0416009452844178e-06,
+      "loss": 0.0525,
+      "step": 3804
+    },
+    {
+      "epoch": 0.7975267239572417,
+      "grad_norm": 0.019596830010414124,
+      "learning_rate": 1.039527822008749e-06,
+      "loss": 0.0549,
+      "step": 3805
+    },
+    {
+      "epoch": 0.7977363236218822,
+      "grad_norm": 0.030918627977371216,
+      "learning_rate": 1.0374565245099328e-06,
+      "loss": 0.0572,
+      "step": 3806
+    },
+    {
+      "epoch": 0.7979459232865227,
+      "grad_norm": 0.020429056137800217,
+      "learning_rate": 1.035387053742834e-06,
+      "loss": 0.0537,
+      "step": 3807
+    },
+    {
+      "epoch": 0.7981555229511633,
+      "grad_norm": 0.026732752099633217,
+      "learning_rate": 1.0333194106614813e-06,
+      "loss": 0.0556,
+      "step": 3808
+    },
+    {
+      "epoch": 0.7983651226158038,
+      "grad_norm": 0.01719031110405922,
+      "learning_rate": 1.0312535962190567e-06,
+      "loss": 0.0563,
+      "step": 3809
+    },
+    {
+      "epoch": 0.7985747222804443,
+      "grad_norm": 0.03157135099172592,
+      "learning_rate": 1.0291896113678983e-06,
+      "loss": 0.0525,
+      "step": 3810
+    },
+    {
+      "epoch": 0.7987843219450849,
+      "grad_norm": 0.02437172830104828,
+      "learning_rate": 1.0271274570595041e-06,
+      "loss": 0.054,
+      "step": 3811
+    },
+    {
+      "epoch": 0.7989939216097254,
+      "grad_norm": 0.017890898510813713,
+      "learning_rate": 1.0250671342445273e-06,
+      "loss": 0.0557,
+      "step": 3812
+    },
+    {
+      "epoch": 0.799203521274366,
+      "grad_norm": 0.022037314251065254,
+      "learning_rate": 1.0230086438727771e-06,
+      "loss": 0.0532,
+      "step": 3813
+    },
+    {
+      "epoch": 0.7994131209390065,
+      "grad_norm": 0.014611390419304371,
+      "learning_rate": 1.020951986893216e-06,
+      "loss": 0.056,
+      "step": 3814
+    },
+    {
+      "epoch": 0.799622720603647,
+      "grad_norm": 0.02562752366065979,
+      "learning_rate": 1.0188971642539614e-06,
+      "loss": 0.0536,
+      "step": 3815
+    },
+    {
+      "epoch": 0.7998323202682875,
+      "grad_norm": 0.015589050948619843,
+      "learning_rate": 1.016844176902288e-06,
+      "loss": 0.0522,
+      "step": 3816
+    },
+    {
+      "epoch": 0.8000419199329281,
+      "grad_norm": 0.030634846538305283,
+      "learning_rate": 1.0147930257846206e-06,
+      "loss": 0.0525,
+      "step": 3817
+    },
+    {
+      "epoch": 0.8002515195975687,
+      "grad_norm": 0.018007168546319008,
+      "learning_rate": 1.0127437118465405e-06,
+      "loss": 0.0567,
+      "step": 3818
+    },
+    {
+      "epoch": 0.8004611192622092,
+      "grad_norm": 0.02537359483540058,
+      "learning_rate": 1.0106962360327832e-06,
+      "loss": 0.0564,
+      "step": 3819
+    },
+    {
+      "epoch": 0.8006707189268497,
+      "grad_norm": 0.015520203858613968,
+      "learning_rate": 1.0086505992872304e-06,
+      "loss": 0.0566,
+      "step": 3820
+    },
+    {
+      "epoch": 0.8008803185914902,
+      "grad_norm": 0.02692110277712345,
+      "learning_rate": 1.0066068025529219e-06,
+      "loss": 0.0548,
+      "step": 3821
+    },
+    {
+      "epoch": 0.8010899182561307,
+      "grad_norm": 0.021093720570206642,
+      "learning_rate": 1.0045648467720492e-06,
+      "loss": 0.0524,
+      "step": 3822
+    },
+    {
+      "epoch": 0.8012995179207714,
+      "grad_norm": 0.01891363225877285,
+      "learning_rate": 1.002524732885951e-06,
+      "loss": 0.053,
+      "step": 3823
+    },
+    {
+      "epoch": 0.8015091175854119,
+      "grad_norm": 0.016529306769371033,
+      "learning_rate": 1.0004864618351223e-06,
+      "loss": 0.054,
+      "step": 3824
+    },
+    {
+      "epoch": 0.8017187172500524,
+      "grad_norm": 0.01724756881594658,
+      "learning_rate": 9.984500345592023e-07,
+      "loss": 0.0582,
+      "step": 3825
+    },
+    {
+      "epoch": 0.8019283169146929,
+      "grad_norm": 0.014622312039136887,
+      "learning_rate": 9.964154519969865e-07,
+      "loss": 0.0617,
+      "step": 3826
+    },
+    {
+      "epoch": 0.8021379165793334,
+      "grad_norm": 0.016424983739852905,
+      "learning_rate": 9.943827150864143e-07,
+      "loss": 0.0527,
+      "step": 3827
+    },
+    {
+      "epoch": 0.802347516243974,
+      "grad_norm": 0.016418717801570892,
+      "learning_rate": 9.923518247645785e-07,
+      "loss": 0.0525,
+      "step": 3828
+    },
+    {
+      "epoch": 0.8025571159086146,
+      "grad_norm": 0.01681547239422798,
+      "learning_rate": 9.903227819677203e-07,
+      "loss": 0.0546,
+      "step": 3829
+    },
+    {
+      "epoch": 0.8027667155732551,
+      "grad_norm": 0.016537809744477272,
+      "learning_rate": 9.882955876312266e-07,
+      "loss": 0.052,
+      "step": 3830
+    },
+    {
+      "epoch": 0.8029763152378956,
+      "grad_norm": 0.018042949959635735,
+      "learning_rate": 9.862702426896327e-07,
+      "loss": 0.0549,
+      "step": 3831
+    },
+    {
+      "epoch": 0.8031859149025361,
+      "grad_norm": 0.019126849249005318,
+      "learning_rate": 9.842467480766243e-07,
+      "loss": 0.0537,
+      "step": 3832
+    },
+    {
+      "epoch": 0.8033955145671767,
+      "grad_norm": 0.01958855427801609,
+      "learning_rate": 9.822251047250298e-07,
+      "loss": 0.0544,
+      "step": 3833
+    },
+    {
+      "epoch": 0.8036051142318172,
+      "grad_norm": 0.017055636271834373,
+      "learning_rate": 9.80205313566827e-07,
+      "loss": 0.0512,
+      "step": 3834
+    },
+    {
+      "epoch": 0.8038147138964578,
+      "grad_norm": 0.0194768775254488,
+      "learning_rate": 9.781873755331412e-07,
+      "loss": 0.0541,
+      "step": 3835
+    },
+    {
+      "epoch": 0.8040243135610983,
+      "grad_norm": 0.015344620682299137,
+      "learning_rate": 9.7617129155424e-07,
+      "loss": 0.0538,
+      "step": 3836
+    },
+    {
+      "epoch": 0.8042339132257388,
+      "grad_norm": 0.016386136412620544,
+      "learning_rate": 9.74157062559536e-07,
+      "loss": 0.0537,
+      "step": 3837
+    },
+    {
+      "epoch": 0.8044435128903794,
+      "grad_norm": 0.016445597633719444,
+      "learning_rate": 9.7214468947759e-07,
+      "loss": 0.0538,
+      "step": 3838
+    },
+    {
+      "epoch": 0.8046531125550199,
+      "grad_norm": 0.017396828159689903,
+      "learning_rate": 9.701341732361068e-07,
+      "loss": 0.0533,
+      "step": 3839
+    },
+    {
+      "epoch": 0.8048627122196604,
+      "grad_norm": 0.018001766875386238,
+      "learning_rate": 9.681255147619317e-07,
+      "loss": 0.0519,
+      "step": 3840
+    },
+    {
+      "epoch": 0.805072311884301,
+      "grad_norm": 0.014202550053596497,
+      "learning_rate": 9.66118714981058e-07,
+      "loss": 0.0547,
+      "step": 3841
+    },
+    {
+      "epoch": 0.8052819115489416,
+      "grad_norm": 0.01720808818936348,
+      "learning_rate": 9.641137748186186e-07,
+      "loss": 0.0543,
+      "step": 3842
+    },
+    {
+      "epoch": 0.8054915112135821,
+      "grad_norm": 0.012487749569118023,
+      "learning_rate": 9.62110695198889e-07,
+      "loss": 0.0572,
+      "step": 3843
+    },
+    {
+      "epoch": 0.8057011108782226,
+      "grad_norm": 0.014534911140799522,
+      "learning_rate": 9.601094770452907e-07,
+      "loss": 0.0555,
+      "step": 3844
+    },
+    {
+      "epoch": 0.8059107105428631,
+      "grad_norm": 0.012713300064206123,
+      "learning_rate": 9.581101212803857e-07,
+      "loss": 0.0567,
+      "step": 3845
+    },
+    {
+      "epoch": 0.8061203102075036,
+      "grad_norm": 0.01954994909465313,
+      "learning_rate": 9.561126288258738e-07,
+      "loss": 0.0562,
+      "step": 3846
+    },
+    {
+      "epoch": 0.8063299098721443,
+      "grad_norm": 0.01227479986846447,
+      "learning_rate": 9.541170006026012e-07,
+      "loss": 0.0526,
+      "step": 3847
+    },
+    {
+      "epoch": 0.8065395095367848,
+      "grad_norm": 0.013360395096242428,
+      "learning_rate": 9.521232375305494e-07,
+      "loss": 0.0531,
+      "step": 3848
+    },
+    {
+      "epoch": 0.8067491092014253,
+      "grad_norm": 0.012572353705763817,
+      "learning_rate": 9.50131340528846e-07,
+      "loss": 0.0548,
+      "step": 3849
+    },
+    {
+      "epoch": 0.8069587088660658,
+      "grad_norm": 0.012260310351848602,
+      "learning_rate": 9.481413105157517e-07,
+      "loss": 0.0508,
+      "step": 3850
+    },
+    {
+      "epoch": 0.8071683085307063,
+      "grad_norm": 0.012266767211258411,
+      "learning_rate": 9.461531484086722e-07,
+      "loss": 0.0564,
+      "step": 3851
+    },
+    {
+      "epoch": 0.8073779081953469,
+      "grad_norm": 0.013467447832226753,
+      "learning_rate": 9.441668551241511e-07,
+      "loss": 0.0564,
+      "step": 3852
+    },
+    {
+      "epoch": 0.8075875078599875,
+      "grad_norm": 0.013181211426854134,
+      "learning_rate": 9.421824315778649e-07,
+      "loss": 0.0548,
+      "step": 3853
+    },
+    {
+      "epoch": 0.807797107524628,
+      "grad_norm": 0.014856358990073204,
+      "learning_rate": 9.401998786846356e-07,
+      "loss": 0.0543,
+      "step": 3854
+    },
+    {
+      "epoch": 0.8080067071892685,
+      "grad_norm": 0.013289416208863258,
+      "learning_rate": 9.382191973584193e-07,
+      "loss": 0.0524,
+      "step": 3855
+    },
+    {
+      "epoch": 0.808216306853909,
+      "grad_norm": 0.017398502677679062,
+      "learning_rate": 9.362403885123084e-07,
+      "loss": 0.0515,
+      "step": 3856
+    },
+    {
+      "epoch": 0.8084259065185496,
+      "grad_norm": 0.012360899709165096,
+      "learning_rate": 9.342634530585354e-07,
+      "loss": 0.0531,
+      "step": 3857
+    },
+    {
+      "epoch": 0.8086355061831901,
+      "grad_norm": 0.014594539068639278,
+      "learning_rate": 9.322883919084652e-07,
+      "loss": 0.0537,
+      "step": 3858
+    },
+    {
+      "epoch": 0.8088451058478306,
+      "grad_norm": 0.014027457684278488,
+      "learning_rate": 9.303152059726023e-07,
+      "loss": 0.0545,
+      "step": 3859
+    },
+    {
+      "epoch": 0.8090547055124712,
+      "grad_norm": 0.015334793366491795,
+      "learning_rate": 9.283438961605829e-07,
+      "loss": 0.0515,
+      "step": 3860
+    },
+    {
+      "epoch": 0.8092643051771117,
+      "grad_norm": 0.018460331484675407,
+      "learning_rate": 9.263744633811816e-07,
+      "loss": 0.0543,
+      "step": 3861
+    },
+    {
+      "epoch": 0.8094739048417523,
+      "grad_norm": 0.017743943259119987,
+      "learning_rate": 9.244069085423074e-07,
+      "loss": 0.0559,
+      "step": 3862
+    },
+    {
+      "epoch": 0.8096835045063928,
+      "grad_norm": 0.014662131667137146,
+      "learning_rate": 9.224412325510024e-07,
+      "loss": 0.055,
+      "step": 3863
+    },
+    {
+      "epoch": 0.8098931041710333,
+      "grad_norm": 0.016809482127428055,
+      "learning_rate": 9.204774363134405e-07,
+      "loss": 0.0564,
+      "step": 3864
+    },
+    {
+      "epoch": 0.8101027038356738,
+      "grad_norm": 0.022974979132413864,
+      "learning_rate": 9.185155207349344e-07,
+      "loss": 0.0545,
+      "step": 3865
+    },
+    {
+      "epoch": 0.8103123035003144,
+      "grad_norm": 0.018833018839359283,
+      "learning_rate": 9.165554867199245e-07,
+      "loss": 0.0561,
+      "step": 3866
+    },
+    {
+      "epoch": 0.810521903164955,
+      "grad_norm": 0.021687567234039307,
+      "learning_rate": 9.145973351719867e-07,
+      "loss": 0.0569,
+      "step": 3867
+    },
+    {
+      "epoch": 0.8107315028295955,
+      "grad_norm": 0.016966206952929497,
+      "learning_rate": 9.126410669938302e-07,
+      "loss": 0.0518,
+      "step": 3868
+    },
+    {
+      "epoch": 0.810941102494236,
+      "grad_norm": 0.023966066539287567,
+      "learning_rate": 9.106866830872929e-07,
+      "loss": 0.0569,
+      "step": 3869
+    },
+    {
+      "epoch": 0.8111507021588765,
+      "grad_norm": 0.016374409198760986,
+      "learning_rate": 9.087341843533437e-07,
+      "loss": 0.0552,
+      "step": 3870
+    },
+    {
+      "epoch": 0.811360301823517,
+      "grad_norm": 0.025269117206335068,
+      "learning_rate": 9.067835716920859e-07,
+      "loss": 0.0551,
+      "step": 3871
+    },
+    {
+      "epoch": 0.8115699014881577,
+      "grad_norm": 0.01591191627085209,
+      "learning_rate": 9.048348460027528e-07,
+      "loss": 0.0532,
+      "step": 3872
+    },
+    {
+      "epoch": 0.8117795011527982,
+      "grad_norm": 0.020468199625611305,
+      "learning_rate": 9.028880081837032e-07,
+      "loss": 0.0539,
+      "step": 3873
+    },
+    {
+      "epoch": 0.8119891008174387,
+      "grad_norm": 0.0173234511166811,
+      "learning_rate": 9.009430591324325e-07,
+      "loss": 0.0537,
+      "step": 3874
+    },
+    {
+      "epoch": 0.8121987004820792,
+      "grad_norm": 0.015647042542696,
+      "learning_rate": 8.989999997455601e-07,
+      "loss": 0.0563,
+      "step": 3875
+    },
+    {
+      "epoch": 0.8124083001467197,
+      "grad_norm": 0.022868547588586807,
+      "learning_rate": 8.970588309188343e-07,
+      "loss": 0.0527,
+      "step": 3876
+    },
+    {
+      "epoch": 0.8126178998113603,
+      "grad_norm": 0.018496761098504066,
+      "learning_rate": 8.951195535471357e-07,
+      "loss": 0.0546,
+      "step": 3877
+    },
+    {
+      "epoch": 0.8128274994760009,
+      "grad_norm": 0.023296529427170753,
+      "learning_rate": 8.931821685244712e-07,
+      "loss": 0.0546,
+      "step": 3878
+    },
+    {
+      "epoch": 0.8130370991406414,
+      "grad_norm": 0.019333289936184883,
+      "learning_rate": 8.912466767439726e-07,
+      "loss": 0.0532,
+      "step": 3879
+    },
+    {
+      "epoch": 0.8132466988052819,
+      "grad_norm": 0.021376557648181915,
+      "learning_rate": 8.893130790979038e-07,
+      "loss": 0.0532,
+      "step": 3880
+    },
+    {
+      "epoch": 0.8134562984699224,
+      "grad_norm": 0.015562393702566624,
+      "learning_rate": 8.873813764776506e-07,
+      "loss": 0.0536,
+      "step": 3881
+    },
+    {
+      "epoch": 0.813665898134563,
+      "grad_norm": 0.026157313957810402,
+      "learning_rate": 8.854515697737298e-07,
+      "loss": 0.0531,
+      "step": 3882
+    },
+    {
+      "epoch": 0.8138754977992035,
+      "grad_norm": 0.017569879069924355,
+      "learning_rate": 8.835236598757796e-07,
+      "loss": 0.0528,
+      "step": 3883
+    },
+    {
+      "epoch": 0.814085097463844,
+      "grad_norm": 0.021584592759609222,
+      "learning_rate": 8.815976476725668e-07,
+      "loss": 0.053,
+      "step": 3884
+    },
+    {
+      "epoch": 0.8142946971284846,
+      "grad_norm": 0.01839163526892662,
+      "learning_rate": 8.796735340519847e-07,
+      "loss": 0.0539,
+      "step": 3885
+    },
+    {
+      "epoch": 0.8145042967931251,
+      "grad_norm": 0.01819518581032753,
+      "learning_rate": 8.777513199010468e-07,
+      "loss": 0.0558,
+      "step": 3886
+    },
+    {
+      "epoch": 0.8147138964577657,
+      "grad_norm": 0.01771095208823681,
+      "learning_rate": 8.758310061058934e-07,
+      "loss": 0.0527,
+      "step": 3887
+    },
+    {
+      "epoch": 0.8149234961224062,
+      "grad_norm": 0.01995546743273735,
+      "learning_rate": 8.739125935517906e-07,
+      "loss": 0.0531,
+      "step": 3888
+    },
+    {
+      "epoch": 0.8151330957870467,
+      "grad_norm": 0.016256198287010193,
+      "learning_rate": 8.719960831231239e-07,
+      "loss": 0.0523,
+      "step": 3889
+    },
+    {
+      "epoch": 0.8153426954516872,
+      "grad_norm": 0.020824657753109932,
+      "learning_rate": 8.70081475703406e-07,
+      "loss": 0.055,
+      "step": 3890
+    },
+    {
+      "epoch": 0.8155522951163278,
+      "grad_norm": 0.0172143392264843,
+      "learning_rate": 8.681687721752719e-07,
+      "loss": 0.0564,
+      "step": 3891
+    },
+    {
+      "epoch": 0.8157618947809684,
+      "grad_norm": 0.022133000195026398,
+      "learning_rate": 8.66257973420473e-07,
+      "loss": 0.053,
+      "step": 3892
+    },
+    {
+      "epoch": 0.8159714944456089,
+      "grad_norm": 0.01741599105298519,
+      "learning_rate": 8.643490803198895e-07,
+      "loss": 0.0574,
+      "step": 3893
+    },
+    {
+      "epoch": 0.8161810941102494,
+      "grad_norm": 0.01774427853524685,
+      "learning_rate": 8.62442093753521e-07,
+      "loss": 0.0556,
+      "step": 3894
+    },
+    {
+      "epoch": 0.8163906937748899,
+      "grad_norm": 0.019820379093289375,
+      "learning_rate": 8.605370146004894e-07,
+      "loss": 0.0536,
+      "step": 3895
+    },
+    {
+      "epoch": 0.8166002934395304,
+      "grad_norm": 0.017903191968798637,
+      "learning_rate": 8.58633843739034e-07,
+      "loss": 0.056,
+      "step": 3896
+    },
+    {
+      "epoch": 0.8168098931041711,
+      "grad_norm": 0.02086867205798626,
+      "learning_rate": 8.567325820465156e-07,
+      "loss": 0.053,
+      "step": 3897
+    },
+    {
+      "epoch": 0.8170194927688116,
+      "grad_norm": 0.02020520158112049,
+      "learning_rate": 8.548332303994167e-07,
+      "loss": 0.055,
+      "step": 3898
+    },
+    {
+      "epoch": 0.8172290924334521,
+      "grad_norm": 0.015635685995221138,
+      "learning_rate": 8.52935789673337e-07,
+      "loss": 0.0539,
+      "step": 3899
+    },
+    {
+      "epoch": 0.8174386920980926,
+      "grad_norm": 0.02270682342350483,
+      "learning_rate": 8.510402607429963e-07,
+      "loss": 0.056,
+      "step": 3900
+    },
+    {
+      "epoch": 0.8176482917627331,
+      "grad_norm": 0.014987417496740818,
+      "learning_rate": 8.491466444822355e-07,
+      "loss": 0.0546,
+      "step": 3901
+    },
+    {
+      "epoch": 0.8178578914273738,
+      "grad_norm": 0.0187685564160347,
+      "learning_rate": 8.472549417640092e-07,
+      "loss": 0.0521,
+      "step": 3902
+    },
+    {
+      "epoch": 0.8180674910920143,
+      "grad_norm": 0.01810646429657936,
+      "learning_rate": 8.453651534603901e-07,
+      "loss": 0.0542,
+      "step": 3903
+    },
+    {
+      "epoch": 0.8182770907566548,
+      "grad_norm": 0.01981244422495365,
+      "learning_rate": 8.434772804425734e-07,
+      "loss": 0.0551,
+      "step": 3904
+    },
+    {
+      "epoch": 0.8184866904212953,
+      "grad_norm": 0.017939042299985886,
+      "learning_rate": 8.415913235808675e-07,
+      "loss": 0.0553,
+      "step": 3905
+    },
+    {
+      "epoch": 0.8186962900859358,
+      "grad_norm": 0.02355090342462063,
+      "learning_rate": 8.397072837446968e-07,
+      "loss": 0.052,
+      "step": 3906
+    },
+    {
+      "epoch": 0.8189058897505764,
+      "grad_norm": 0.01684572361409664,
+      "learning_rate": 8.378251618026051e-07,
+      "loss": 0.0561,
+      "step": 3907
+    },
+    {
+      "epoch": 0.819115489415217,
+      "grad_norm": 0.024392616003751755,
+      "learning_rate": 8.35944958622249e-07,
+      "loss": 0.0536,
+      "step": 3908
+    },
+    {
+      "epoch": 0.8193250890798575,
+      "grad_norm": 0.01632305234670639,
+      "learning_rate": 8.340666750704013e-07,
+      "loss": 0.0538,
+      "step": 3909
+    },
+    {
+      "epoch": 0.819534688744498,
+      "grad_norm": 0.023946644738316536,
+      "learning_rate": 8.32190312012951e-07,
+      "loss": 0.0542,
+      "step": 3910
+    },
+    {
+      "epoch": 0.8197442884091386,
+      "grad_norm": 0.016722572967410088,
+      "learning_rate": 8.303158703149023e-07,
+      "loss": 0.0546,
+      "step": 3911
+    },
+    {
+      "epoch": 0.8199538880737791,
+      "grad_norm": 0.02079438418149948,
+      "learning_rate": 8.2844335084037e-07,
+      "loss": 0.0552,
+      "step": 3912
+    },
+    {
+      "epoch": 0.8201634877384196,
+      "grad_norm": 0.016205811873078346,
+      "learning_rate": 8.265727544525876e-07,
+      "loss": 0.0574,
+      "step": 3913
+    },
+    {
+      "epoch": 0.8203730874030601,
+      "grad_norm": 0.020107150077819824,
+      "learning_rate": 8.247040820138985e-07,
+      "loss": 0.0529,
+      "step": 3914
+    },
+    {
+      "epoch": 0.8205826870677007,
+      "grad_norm": 0.02017829939723015,
+      "learning_rate": 8.228373343857593e-07,
+      "loss": 0.0559,
+      "step": 3915
+    },
+    {
+      "epoch": 0.8207922867323413,
+      "grad_norm": 0.015335801057517529,
+      "learning_rate": 8.20972512428741e-07,
+      "loss": 0.0539,
+      "step": 3916
+    },
+    {
+      "epoch": 0.8210018863969818,
+      "grad_norm": 0.01876046508550644,
+      "learning_rate": 8.191096170025265e-07,
+      "loss": 0.0531,
+      "step": 3917
+    },
+    {
+      "epoch": 0.8212114860616223,
+      "grad_norm": 0.022027568891644478,
+      "learning_rate": 8.172486489659115e-07,
+      "loss": 0.0528,
+      "step": 3918
+    },
+    {
+      "epoch": 0.8214210857262628,
+      "grad_norm": 0.013593902811408043,
+      "learning_rate": 8.153896091768004e-07,
+      "loss": 0.056,
+      "step": 3919
+    },
+    {
+      "epoch": 0.8216306853909033,
+      "grad_norm": 0.02417590655386448,
+      "learning_rate": 8.135324984922088e-07,
+      "loss": 0.0555,
+      "step": 3920
+    },
+    {
+      "epoch": 0.821840285055544,
+      "grad_norm": 0.016407081857323647,
+      "learning_rate": 8.116773177682674e-07,
+      "loss": 0.0532,
+      "step": 3921
+    },
+    {
+      "epoch": 0.8220498847201845,
+      "grad_norm": 0.021131981164216995,
+      "learning_rate": 8.09824067860211e-07,
+      "loss": 0.0534,
+      "step": 3922
+    },
+    {
+      "epoch": 0.822259484384825,
+      "grad_norm": 0.01895805634558201,
+      "learning_rate": 8.079727496223894e-07,
+      "loss": 0.0551,
+      "step": 3923
+    },
+    {
+      "epoch": 0.8224690840494655,
+      "grad_norm": 0.027057627215981483,
+      "learning_rate": 8.061233639082616e-07,
+      "loss": 0.0553,
+      "step": 3924
+    },
+    {
+      "epoch": 0.822678683714106,
+      "grad_norm": 0.016223613172769547,
+      "learning_rate": 8.042759115703891e-07,
+      "loss": 0.0534,
+      "step": 3925
+    },
+    {
+      "epoch": 0.8228882833787466,
+      "grad_norm": 0.030798260122537613,
+      "learning_rate": 8.024303934604505e-07,
+      "loss": 0.0537,
+      "step": 3926
+    },
+    {
+      "epoch": 0.8230978830433872,
+      "grad_norm": 0.01324604731053114,
+      "learning_rate": 8.005868104292291e-07,
+      "loss": 0.0519,
+      "step": 3927
+    },
+    {
+      "epoch": 0.8233074827080277,
+      "grad_norm": 0.023022161796689034,
+      "learning_rate": 7.987451633266153e-07,
+      "loss": 0.0524,
+      "step": 3928
+    },
+    {
+      "epoch": 0.8235170823726682,
+      "grad_norm": 0.015440089628100395,
+      "learning_rate": 7.969054530016091e-07,
+      "loss": 0.0538,
+      "step": 3929
+    },
+    {
+      "epoch": 0.8237266820373087,
+      "grad_norm": 0.023025307804346085,
+      "learning_rate": 7.950676803023149e-07,
+      "loss": 0.0521,
+      "step": 3930
+    },
+    {
+      "epoch": 0.8239362817019493,
+      "grad_norm": 0.01681152544915676,
+      "learning_rate": 7.93231846075948e-07,
+      "loss": 0.0525,
+      "step": 3931
+    },
+    {
+      "epoch": 0.8241458813665898,
+      "grad_norm": 0.016274144873023033,
+      "learning_rate": 7.913979511688252e-07,
+      "loss": 0.0558,
+      "step": 3932
+    },
+    {
+      "epoch": 0.8243554810312304,
+      "grad_norm": 0.016948441043496132,
+      "learning_rate": 7.895659964263725e-07,
+      "loss": 0.0526,
+      "step": 3933
+    },
+    {
+      "epoch": 0.8245650806958709,
+      "grad_norm": 0.01407658401876688,
+      "learning_rate": 7.877359826931225e-07,
+      "loss": 0.055,
+      "step": 3934
+    },
+    {
+      "epoch": 0.8247746803605114,
+      "grad_norm": 0.011790010146796703,
+      "learning_rate": 7.859079108127088e-07,
+      "loss": 0.0548,
+      "step": 3935
+    },
+    {
+      "epoch": 0.824984280025152,
+      "grad_norm": 0.014163525775074959,
+      "learning_rate": 7.840817816278723e-07,
+      "loss": 0.055,
+      "step": 3936
+    },
+    {
+      "epoch": 0.8251938796897925,
+      "grad_norm": 0.011464595794677734,
+      "learning_rate": 7.822575959804596e-07,
+      "loss": 0.0534,
+      "step": 3937
+    },
+    {
+      "epoch": 0.825403479354433,
+      "grad_norm": 0.012643976137042046,
+      "learning_rate": 7.80435354711418e-07,
+      "loss": 0.0527,
+      "step": 3938
+    },
+    {
+      "epoch": 0.8256130790190735,
+      "grad_norm": 0.012431315146386623,
+      "learning_rate": 7.78615058660801e-07,
+      "loss": 0.0529,
+      "step": 3939
+    },
+    {
+      "epoch": 0.8258226786837141,
+      "grad_norm": 0.011443213559687138,
+      "learning_rate": 7.767967086677669e-07,
+      "loss": 0.0546,
+      "step": 3940
+    },
+    {
+      "epoch": 0.8260322783483547,
+      "grad_norm": 0.012839680537581444,
+      "learning_rate": 7.749803055705723e-07,
+      "loss": 0.054,
+      "step": 3941
+    },
+    {
+      "epoch": 0.8262418780129952,
+      "grad_norm": 0.010854445397853851,
+      "learning_rate": 7.731658502065786e-07,
+      "loss": 0.0539,
+      "step": 3942
+    },
+    {
+      "epoch": 0.8264514776776357,
+      "grad_norm": 0.01064964011311531,
+      "learning_rate": 7.713533434122494e-07,
+      "loss": 0.0548,
+      "step": 3943
+    },
+    {
+      "epoch": 0.8266610773422762,
+      "grad_norm": 0.012149804271757603,
+      "learning_rate": 7.695427860231519e-07,
+      "loss": 0.0553,
+      "step": 3944
+    },
+    {
+      "epoch": 0.8268706770069167,
+      "grad_norm": 0.01205404382199049,
+      "learning_rate": 7.677341788739507e-07,
+      "loss": 0.0537,
+      "step": 3945
+    },
+    {
+      "epoch": 0.8270802766715574,
+      "grad_norm": 0.014352019876241684,
+      "learning_rate": 7.659275227984142e-07,
+      "loss": 0.0514,
+      "step": 3946
+    },
+    {
+      "epoch": 0.8272898763361979,
+      "grad_norm": 0.01178305596113205,
+      "learning_rate": 7.641228186294108e-07,
+      "loss": 0.054,
+      "step": 3947
+    },
+    {
+      "epoch": 0.8274994760008384,
+      "grad_norm": 0.020204555243253708,
+      "learning_rate": 7.623200671989067e-07,
+      "loss": 0.0541,
+      "step": 3948
+    },
+    {
+      "epoch": 0.8277090756654789,
+      "grad_norm": 0.011678210459649563,
+      "learning_rate": 7.605192693379715e-07,
+      "loss": 0.0558,
+      "step": 3949
+    },
+    {
+      "epoch": 0.8279186753301194,
+      "grad_norm": 0.014891520142555237,
+      "learning_rate": 7.587204258767733e-07,
+      "loss": 0.0533,
+      "step": 3950
+    },
+    {
+      "epoch": 0.82812827499476,
+      "grad_norm": 0.012660454027354717,
+      "learning_rate": 7.569235376445772e-07,
+      "loss": 0.0535,
+      "step": 3951
+    },
+    {
+      "epoch": 0.8283378746594006,
+      "grad_norm": 0.01889774762094021,
+      "learning_rate": 7.551286054697498e-07,
+      "loss": 0.0552,
+      "step": 3952
+    },
+    {
+      "epoch": 0.8285474743240411,
+      "grad_norm": 0.011206352151930332,
+      "learning_rate": 7.533356301797523e-07,
+      "loss": 0.0553,
+      "step": 3953
+    },
+    {
+      "epoch": 0.8287570739886816,
+      "grad_norm": 0.015002798289060593,
+      "learning_rate": 7.515446126011484e-07,
+      "loss": 0.0552,
+      "step": 3954
+    },
+    {
+      "epoch": 0.8289666736533221,
+      "grad_norm": 0.012600685469806194,
+      "learning_rate": 7.497555535595946e-07,
+      "loss": 0.0534,
+      "step": 3955
+    },
+    {
+      "epoch": 0.8291762733179627,
+      "grad_norm": 0.01465271133929491,
+      "learning_rate": 7.479684538798476e-07,
+      "loss": 0.0585,
+      "step": 3956
+    },
+    {
+      "epoch": 0.8293858729826032,
+      "grad_norm": 0.015729481354355812,
+      "learning_rate": 7.461833143857611e-07,
+      "loss": 0.0515,
+      "step": 3957
+    },
+    {
+      "epoch": 0.8295954726472438,
+      "grad_norm": 0.012866144999861717,
+      "learning_rate": 7.444001359002833e-07,
+      "loss": 0.0558,
+      "step": 3958
+    },
+    {
+      "epoch": 0.8298050723118843,
+      "grad_norm": 0.019521057605743408,
+      "learning_rate": 7.426189192454575e-07,
+      "loss": 0.0552,
+      "step": 3959
+    },
+    {
+      "epoch": 0.8300146719765248,
+      "grad_norm": 0.018146377056837082,
+      "learning_rate": 7.408396652424271e-07,
+      "loss": 0.0508,
+      "step": 3960
+    },
+    {
+      "epoch": 0.8302242716411654,
+      "grad_norm": 0.01978515274822712,
+      "learning_rate": 7.39062374711425e-07,
+      "loss": 0.0542,
+      "step": 3961
+    },
+    {
+      "epoch": 0.8304338713058059,
+      "grad_norm": 0.015556586906313896,
+      "learning_rate": 7.372870484717843e-07,
+      "loss": 0.0571,
+      "step": 3962
+    },
+    {
+      "epoch": 0.8306434709704464,
+      "grad_norm": 0.02219201996922493,
+      "learning_rate": 7.355136873419277e-07,
+      "loss": 0.0563,
+      "step": 3963
+    },
+    {
+      "epoch": 0.830853070635087,
+      "grad_norm": 0.01369245070964098,
+      "learning_rate": 7.337422921393767e-07,
+      "loss": 0.0554,
+      "step": 3964
+    },
+    {
+      "epoch": 0.8310626702997275,
+      "grad_norm": 0.021102454513311386,
+      "learning_rate": 7.319728636807411e-07,
+      "loss": 0.0518,
+      "step": 3965
+    },
+    {
+      "epoch": 0.8312722699643681,
+      "grad_norm": 0.013064729049801826,
+      "learning_rate": 7.302054027817291e-07,
+      "loss": 0.0556,
+      "step": 3966
+    },
+    {
+      "epoch": 0.8314818696290086,
+      "grad_norm": 0.015176662243902683,
+      "learning_rate": 7.28439910257141e-07,
+      "loss": 0.057,
+      "step": 3967
+    },
+    {
+      "epoch": 0.8316914692936491,
+      "grad_norm": 0.014463878236711025,
+      "learning_rate": 7.266763869208665e-07,
+      "loss": 0.0547,
+      "step": 3968
+    },
+    {
+      "epoch": 0.8319010689582896,
+      "grad_norm": 0.012370740063488483,
+      "learning_rate": 7.249148335858891e-07,
+      "loss": 0.0541,
+      "step": 3969
+    },
+    {
+      "epoch": 0.8321106686229301,
+      "grad_norm": 0.020610513165593147,
+      "learning_rate": 7.231552510642864e-07,
+      "loss": 0.0552,
+      "step": 3970
+    },
+    {
+      "epoch": 0.8323202682875708,
+      "grad_norm": 0.010926956310868263,
+      "learning_rate": 7.213976401672235e-07,
+      "loss": 0.0535,
+      "step": 3971
+    },
+    {
+      "epoch": 0.8325298679522113,
+      "grad_norm": 0.021732453256845474,
+      "learning_rate": 7.196420017049599e-07,
+      "loss": 0.0549,
+      "step": 3972
+    },
+    {
+      "epoch": 0.8327394676168518,
+      "grad_norm": 0.01048396248370409,
+      "learning_rate": 7.178883364868455e-07,
+      "loss": 0.0557,
+      "step": 3973
+    },
+    {
+      "epoch": 0.8329490672814923,
+      "grad_norm": 0.02181028202176094,
+      "learning_rate": 7.161366453213181e-07,
+      "loss": 0.0557,
+      "step": 3974
+    },
+    {
+      "epoch": 0.8331586669461328,
+      "grad_norm": 0.012369881384074688,
+      "learning_rate": 7.143869290159067e-07,
+      "loss": 0.0522,
+      "step": 3975
+    },
+    {
+      "epoch": 0.8333682666107735,
+      "grad_norm": 0.02149048261344433,
+      "learning_rate": 7.12639188377231e-07,
+      "loss": 0.059,
+      "step": 3976
+    },
+    {
+      "epoch": 0.833577866275414,
+      "grad_norm": 0.013264495879411697,
+      "learning_rate": 7.108934242109994e-07,
+      "loss": 0.0581,
+      "step": 3977
+    },
+    {
+      "epoch": 0.8337874659400545,
+      "grad_norm": 0.019909605383872986,
+      "learning_rate": 7.091496373220075e-07,
+      "loss": 0.0549,
+      "step": 3978
+    },
+    {
+      "epoch": 0.833997065604695,
+      "grad_norm": 0.02005838230252266,
+      "learning_rate": 7.074078285141428e-07,
+      "loss": 0.0534,
+      "step": 3979
+    },
+    {
+      "epoch": 0.8342066652693356,
+      "grad_norm": 0.0182628370821476,
+      "learning_rate": 7.056679985903774e-07,
+      "loss": 0.0565,
+      "step": 3980
+    },
+    {
+      "epoch": 0.8344162649339761,
+      "grad_norm": 0.01993626542389393,
+      "learning_rate": 7.03930148352771e-07,
+      "loss": 0.0518,
+      "step": 3981
+    },
+    {
+      "epoch": 0.8346258645986167,
+      "grad_norm": 0.017620259895920753,
+      "learning_rate": 7.021942786024743e-07,
+      "loss": 0.0566,
+      "step": 3982
+    },
+    {
+      "epoch": 0.8348354642632572,
+      "grad_norm": 0.012461712583899498,
+      "learning_rate": 7.004603901397239e-07,
+      "loss": 0.0549,
+      "step": 3983
+    },
+    {
+      "epoch": 0.8350450639278977,
+      "grad_norm": 0.009824151173233986,
+      "learning_rate": 6.987284837638391e-07,
+      "loss": 0.0525,
+      "step": 3984
+    },
+    {
+      "epoch": 0.8352546635925383,
+      "grad_norm": 0.018029648810625076,
+      "learning_rate": 6.969985602732309e-07,
+      "loss": 0.0559,
+      "step": 3985
+    },
+    {
+      "epoch": 0.8354642632571788,
+      "grad_norm": 0.01331211719661951,
+      "learning_rate": 6.952706204653914e-07,
+      "loss": 0.0536,
+      "step": 3986
+    },
+    {
+      "epoch": 0.8356738629218193,
+      "grad_norm": 0.0163306575268507,
+      "learning_rate": 6.935446651369027e-07,
+      "loss": 0.0557,
+      "step": 3987
+    },
+    {
+      "epoch": 0.8358834625864598,
+      "grad_norm": 0.012105715461075306,
+      "learning_rate": 6.918206950834283e-07,
+      "loss": 0.054,
+      "step": 3988
+    },
+    {
+      "epoch": 0.8360930622511004,
+      "grad_norm": 0.013904665596783161,
+      "learning_rate": 6.900987110997182e-07,
+      "loss": 0.0538,
+      "step": 3989
+    },
+    {
+      "epoch": 0.836302661915741,
+      "grad_norm": 0.012371689081192017,
+      "learning_rate": 6.883787139796078e-07,
+      "loss": 0.0539,
+      "step": 3990
+    },
+    {
+      "epoch": 0.8365122615803815,
+      "grad_norm": 0.012469745241105556,
+      "learning_rate": 6.866607045160151e-07,
+      "loss": 0.0568,
+      "step": 3991
+    },
+    {
+      "epoch": 0.836721861245022,
+      "grad_norm": 0.012731354683637619,
+      "learning_rate": 6.849446835009405e-07,
+      "loss": 0.0524,
+      "step": 3992
+    },
+    {
+      "epoch": 0.8369314609096625,
+      "grad_norm": 0.012761647813022137,
+      "learning_rate": 6.832306517254716e-07,
+      "loss": 0.0523,
+      "step": 3993
+    },
+    {
+      "epoch": 0.837141060574303,
+      "grad_norm": 0.013718758709728718,
+      "learning_rate": 6.815186099797744e-07,
+      "loss": 0.0522,
+      "step": 3994
+    },
+    {
+      "epoch": 0.8373506602389437,
+      "grad_norm": 0.015886375680565834,
+      "learning_rate": 6.798085590531012e-07,
+      "loss": 0.0548,
+      "step": 3995
+    },
+    {
+      "epoch": 0.8375602599035842,
+      "grad_norm": 0.01095049548894167,
+      "learning_rate": 6.78100499733787e-07,
+      "loss": 0.0551,
+      "step": 3996
+    },
+    {
+      "epoch": 0.8377698595682247,
+      "grad_norm": 0.015172593295574188,
+      "learning_rate": 6.76394432809242e-07,
+      "loss": 0.0546,
+      "step": 3997
+    },
+    {
+      "epoch": 0.8379794592328652,
+      "grad_norm": 0.015550515614449978,
+      "learning_rate": 6.746903590659659e-07,
+      "loss": 0.0529,
+      "step": 3998
+    },
+    {
+      "epoch": 0.8381890588975057,
+      "grad_norm": 0.014822781085968018,
+      "learning_rate": 6.729882792895359e-07,
+      "loss": 0.0542,
+      "step": 3999
+    },
+    {
+      "epoch": 0.8383986585621463,
+      "grad_norm": 0.01649445667862892,
+      "learning_rate": 6.71288194264611e-07,
+      "loss": 0.0533,
+      "step": 4000
+    },
+    {
+      "epoch": 0.8386082582267869,
+      "grad_norm": 0.015881575644016266,
+      "learning_rate": 6.695901047749298e-07,
+      "loss": 0.0535,
+      "step": 4001
+    },
+    {
+      "epoch": 0.8388178578914274,
+      "grad_norm": 0.020809467881917953,
+      "learning_rate": 6.678940116033095e-07,
+      "loss": 0.0549,
+      "step": 4002
+    },
+    {
+      "epoch": 0.8390274575560679,
+      "grad_norm": 0.013579810969531536,
+      "learning_rate": 6.661999155316512e-07,
+      "loss": 0.0518,
+      "step": 4003
+    },
+    {
+      "epoch": 0.8392370572207084,
+      "grad_norm": 0.028101500123739243,
+      "learning_rate": 6.645078173409303e-07,
+      "loss": 0.052,
+      "step": 4004
+    },
+    {
+      "epoch": 0.839446656885349,
+      "grad_norm": 0.013599547557532787,
+      "learning_rate": 6.628177178112055e-07,
+      "loss": 0.0546,
+      "step": 4005
+    },
+    {
+      "epoch": 0.8396562565499895,
+      "grad_norm": 0.028915736824274063,
+      "learning_rate": 6.611296177216125e-07,
+      "loss": 0.0564,
+      "step": 4006
+    },
+    {
+      "epoch": 0.8398658562146301,
+      "grad_norm": 0.016699662432074547,
+      "learning_rate": 6.594435178503644e-07,
+      "loss": 0.0523,
+      "step": 4007
+    },
+    {
+      "epoch": 0.8400754558792706,
+      "grad_norm": 0.02233414351940155,
+      "learning_rate": 6.577594189747521e-07,
+      "loss": 0.0537,
+      "step": 4008
+    },
+    {
+      "epoch": 0.8402850555439111,
+      "grad_norm": 0.021178172901272774,
+      "learning_rate": 6.560773218711458e-07,
+      "loss": 0.0539,
+      "step": 4009
+    },
+    {
+      "epoch": 0.8404946552085517,
+      "grad_norm": 0.022526554763317108,
+      "learning_rate": 6.543972273149928e-07,
+      "loss": 0.0546,
+      "step": 4010
+    },
+    {
+      "epoch": 0.8407042548731922,
+      "grad_norm": 0.02022298239171505,
+      "learning_rate": 6.527191360808144e-07,
+      "loss": 0.0568,
+      "step": 4011
+    },
+    {
+      "epoch": 0.8409138545378327,
+      "grad_norm": 0.020070351660251617,
+      "learning_rate": 6.51043048942212e-07,
+      "loss": 0.0559,
+      "step": 4012
+    },
+    {
+      "epoch": 0.8411234542024733,
+      "grad_norm": 0.019938215613365173,
+      "learning_rate": 6.493689666718611e-07,
+      "loss": 0.0538,
+      "step": 4013
+    },
+    {
+      "epoch": 0.8413330538671138,
+      "grad_norm": 0.021108483895659447,
+      "learning_rate": 6.476968900415115e-07,
+      "loss": 0.053,
+      "step": 4014
+    },
+    {
+      "epoch": 0.8415426535317544,
+      "grad_norm": 0.02614649012684822,
+      "learning_rate": 6.460268198219916e-07,
+      "loss": 0.0533,
+      "step": 4015
+    },
+    {
+      "epoch": 0.8417522531963949,
+      "grad_norm": 0.021795378997921944,
+      "learning_rate": 6.443587567832044e-07,
+      "loss": 0.0542,
+      "step": 4016
+    },
+    {
+      "epoch": 0.8419618528610354,
+      "grad_norm": 0.030953634530305862,
+      "learning_rate": 6.426927016941248e-07,
+      "loss": 0.0536,
+      "step": 4017
+    },
+    {
+      "epoch": 0.8421714525256759,
+      "grad_norm": 0.0239299014210701,
+      "learning_rate": 6.410286553228052e-07,
+      "loss": 0.0575,
+      "step": 4018
+    },
+    {
+      "epoch": 0.8423810521903164,
+      "grad_norm": 0.035018209367990494,
+      "learning_rate": 6.393666184363701e-07,
+      "loss": 0.0552,
+      "step": 4019
+    },
+    {
+      "epoch": 0.8425906518549571,
+      "grad_norm": 0.01738305762410164,
+      "learning_rate": 6.377065918010173e-07,
+      "loss": 0.0514,
+      "step": 4020
+    },
+    {
+      "epoch": 0.8428002515195976,
+      "grad_norm": 0.027201887220144272,
+      "learning_rate": 6.360485761820195e-07,
+      "loss": 0.0553,
+      "step": 4021
+    },
+    {
+      "epoch": 0.8430098511842381,
+      "grad_norm": 0.01937497965991497,
+      "learning_rate": 6.343925723437217e-07,
+      "loss": 0.0526,
+      "step": 4022
+    },
+    {
+      "epoch": 0.8432194508488786,
+      "grad_norm": 0.018232090398669243,
+      "learning_rate": 6.327385810495423e-07,
+      "loss": 0.0567,
+      "step": 4023
+    },
+    {
+      "epoch": 0.8434290505135191,
+      "grad_norm": 0.01994934491813183,
+      "learning_rate": 6.310866030619694e-07,
+      "loss": 0.0548,
+      "step": 4024
+    },
+    {
+      "epoch": 0.8436386501781598,
+      "grad_norm": 0.016245190054178238,
+      "learning_rate": 6.294366391425643e-07,
+      "loss": 0.0534,
+      "step": 4025
+    },
+    {
+      "epoch": 0.8438482498428003,
+      "grad_norm": 0.018503889441490173,
+      "learning_rate": 6.27788690051962e-07,
+      "loss": 0.0547,
+      "step": 4026
+    },
+    {
+      "epoch": 0.8440578495074408,
+      "grad_norm": 0.015852371230721474,
+      "learning_rate": 6.26142756549864e-07,
+      "loss": 0.0561,
+      "step": 4027
+    },
+    {
+      "epoch": 0.8442674491720813,
+      "grad_norm": 0.012544246390461922,
+      "learning_rate": 6.244988393950469e-07,
+      "loss": 0.0558,
+      "step": 4028
+    },
+    {
+      "epoch": 0.8444770488367218,
+      "grad_norm": 0.01918352209031582,
+      "learning_rate": 6.228569393453582e-07,
+      "loss": 0.052,
+      "step": 4029
+    },
+    {
+      "epoch": 0.8446866485013624,
+      "grad_norm": 0.014716164208948612,
+      "learning_rate": 6.212170571577087e-07,
+      "loss": 0.0517,
+      "step": 4030
+    },
+    {
+      "epoch": 0.844896248166003,
+      "grad_norm": 0.015544608235359192,
+      "learning_rate": 6.195791935880868e-07,
+      "loss": 0.0523,
+      "step": 4031
+    },
+    {
+      "epoch": 0.8451058478306435,
+      "grad_norm": 0.016333509236574173,
+      "learning_rate": 6.17943349391546e-07,
+      "loss": 0.0548,
+      "step": 4032
+    },
+    {
+      "epoch": 0.845315447495284,
+      "grad_norm": 0.018222004175186157,
+      "learning_rate": 6.163095253222129e-07,
+      "loss": 0.052,
+      "step": 4033
+    },
+    {
+      "epoch": 0.8455250471599245,
+      "grad_norm": 0.022488001734018326,
+      "learning_rate": 6.146777221332772e-07,
+      "loss": 0.0538,
+      "step": 4034
+    },
+    {
+      "epoch": 0.8457346468245651,
+      "grad_norm": 0.019987892359495163,
+      "learning_rate": 6.130479405770002e-07,
+      "loss": 0.0545,
+      "step": 4035
+    },
+    {
+      "epoch": 0.8459442464892056,
+      "grad_norm": 0.01848592422902584,
+      "learning_rate": 6.114201814047122e-07,
+      "loss": 0.053,
+      "step": 4036
+    },
+    {
+      "epoch": 0.8461538461538461,
+      "grad_norm": 0.011879503726959229,
+      "learning_rate": 6.097944453668081e-07,
+      "loss": 0.0537,
+      "step": 4037
+    },
+    {
+      "epoch": 0.8463634458184867,
+      "grad_norm": 0.019580353051424026,
+      "learning_rate": 6.081707332127523e-07,
+      "loss": 0.0525,
+      "step": 4038
+    },
+    {
+      "epoch": 0.8465730454831272,
+      "grad_norm": 0.015657026320695877,
+      "learning_rate": 6.065490456910777e-07,
+      "loss": 0.0554,
+      "step": 4039
+    },
+    {
+      "epoch": 0.8467826451477678,
+      "grad_norm": 0.021646184846758842,
+      "learning_rate": 6.049293835493803e-07,
+      "loss": 0.0539,
+      "step": 4040
+    },
+    {
+      "epoch": 0.8469922448124083,
+      "grad_norm": 0.019581638276576996,
+      "learning_rate": 6.03311747534323e-07,
+      "loss": 0.0546,
+      "step": 4041
+    },
+    {
+      "epoch": 0.8472018444770488,
+      "grad_norm": 0.019119396805763245,
+      "learning_rate": 6.016961383916381e-07,
+      "loss": 0.0526,
+      "step": 4042
+    },
+    {
+      "epoch": 0.8474114441416893,
+      "grad_norm": 0.018327955156564713,
+      "learning_rate": 6.000825568661184e-07,
+      "loss": 0.0533,
+      "step": 4043
+    },
+    {
+      "epoch": 0.8476210438063299,
+      "grad_norm": 0.012688934803009033,
+      "learning_rate": 5.984710037016267e-07,
+      "loss": 0.0557,
+      "step": 4044
+    },
+    {
+      "epoch": 0.8478306434709705,
+      "grad_norm": 0.01570408046245575,
+      "learning_rate": 5.968614796410882e-07,
+      "loss": 0.0533,
+      "step": 4045
+    },
+    {
+      "epoch": 0.848040243135611,
+      "grad_norm": 0.013904701918363571,
+      "learning_rate": 5.952539854264938e-07,
+      "loss": 0.0549,
+      "step": 4046
+    },
+    {
+      "epoch": 0.8482498428002515,
+      "grad_norm": 0.0192690659314394,
+      "learning_rate": 5.936485217988958e-07,
+      "loss": 0.0524,
+      "step": 4047
+    },
+    {
+      "epoch": 0.848459442464892,
+      "grad_norm": 0.013410338200628757,
+      "learning_rate": 5.920450894984137e-07,
+      "loss": 0.0537,
+      "step": 4048
+    },
+    {
+      "epoch": 0.8486690421295326,
+      "grad_norm": 0.014187265187501907,
+      "learning_rate": 5.904436892642306e-07,
+      "loss": 0.0548,
+      "step": 4049
+    },
+    {
+      "epoch": 0.8488786417941732,
+      "grad_norm": 0.012049532495439053,
+      "learning_rate": 5.888443218345907e-07,
+      "loss": 0.0562,
+      "step": 4050
+    },
+    {
+      "epoch": 0.8490882414588137,
+      "grad_norm": 0.012612142600119114,
+      "learning_rate": 5.872469879468024e-07,
+      "loss": 0.057,
+      "step": 4051
+    },
+    {
+      "epoch": 0.8492978411234542,
+      "grad_norm": 0.016537649556994438,
+      "learning_rate": 5.856516883372365e-07,
+      "loss": 0.0538,
+      "step": 4052
+    },
+    {
+      "epoch": 0.8495074407880947,
+      "grad_norm": 0.012568135745823383,
+      "learning_rate": 5.840584237413239e-07,
+      "loss": 0.0545,
+      "step": 4053
+    },
+    {
+      "epoch": 0.8497170404527353,
+      "grad_norm": 0.013640418648719788,
+      "learning_rate": 5.824671948935606e-07,
+      "loss": 0.0567,
+      "step": 4054
+    },
+    {
+      "epoch": 0.8499266401173758,
+      "grad_norm": 0.016170648857951164,
+      "learning_rate": 5.808780025275045e-07,
+      "loss": 0.054,
+      "step": 4055
+    },
+    {
+      "epoch": 0.8501362397820164,
+      "grad_norm": 0.014185959473252296,
+      "learning_rate": 5.792908473757697e-07,
+      "loss": 0.0535,
+      "step": 4056
+    },
+    {
+      "epoch": 0.8503458394466569,
+      "grad_norm": 0.01882059872150421,
+      "learning_rate": 5.777057301700372e-07,
+      "loss": 0.0532,
+      "step": 4057
+    },
+    {
+      "epoch": 0.8505554391112974,
+      "grad_norm": 0.01410512626171112,
+      "learning_rate": 5.761226516410434e-07,
+      "loss": 0.0541,
+      "step": 4058
+    },
+    {
+      "epoch": 0.850765038775938,
+      "grad_norm": 0.01758696883916855,
+      "learning_rate": 5.745416125185898e-07,
+      "loss": 0.0543,
+      "step": 4059
+    },
+    {
+      "epoch": 0.8509746384405785,
+      "grad_norm": 0.015843121334910393,
+      "learning_rate": 5.729626135315319e-07,
+      "loss": 0.0518,
+      "step": 4060
+    },
+    {
+      "epoch": 0.851184238105219,
+      "grad_norm": 0.015141872689127922,
+      "learning_rate": 5.713856554077901e-07,
+      "loss": 0.056,
+      "step": 4061
+    },
+    {
+      "epoch": 0.8513938377698596,
+      "grad_norm": 0.016413336619734764,
+      "learning_rate": 5.698107388743418e-07,
+      "loss": 0.0559,
+      "step": 4062
+    },
+    {
+      "epoch": 0.8516034374345001,
+      "grad_norm": 0.016290897503495216,
+      "learning_rate": 5.682378646572229e-07,
+      "loss": 0.0511,
+      "step": 4063
+    },
+    {
+      "epoch": 0.8518130370991407,
+      "grad_norm": 0.01656182110309601,
+      "learning_rate": 5.666670334815267e-07,
+      "loss": 0.0546,
+      "step": 4064
+    },
+    {
+      "epoch": 0.8520226367637812,
+      "grad_norm": 0.01612510345876217,
+      "learning_rate": 5.650982460714083e-07,
+      "loss": 0.0572,
+      "step": 4065
+    },
+    {
+      "epoch": 0.8522322364284217,
+      "grad_norm": 0.01949678361415863,
+      "learning_rate": 5.635315031500766e-07,
+      "loss": 0.0533,
+      "step": 4066
+    },
+    {
+      "epoch": 0.8524418360930622,
+      "grad_norm": 0.017564570531249046,
+      "learning_rate": 5.619668054398008e-07,
+      "loss": 0.0558,
+      "step": 4067
+    },
+    {
+      "epoch": 0.8526514357577027,
+      "grad_norm": 0.024784216657280922,
+      "learning_rate": 5.604041536619048e-07,
+      "loss": 0.0532,
+      "step": 4068
+    },
+    {
+      "epoch": 0.8528610354223434,
+      "grad_norm": 0.020440300926566124,
+      "learning_rate": 5.588435485367733e-07,
+      "loss": 0.0522,
+      "step": 4069
+    },
+    {
+      "epoch": 0.8530706350869839,
+      "grad_norm": 0.024173369631171227,
+      "learning_rate": 5.572849907838423e-07,
+      "loss": 0.0533,
+      "step": 4070
+    },
+    {
+      "epoch": 0.8532802347516244,
+      "grad_norm": 0.016502492129802704,
+      "learning_rate": 5.557284811216074e-07,
+      "loss": 0.0526,
+      "step": 4071
+    },
+    {
+      "epoch": 0.8534898344162649,
+      "grad_norm": 0.020477591082453728,
+      "learning_rate": 5.54174020267621e-07,
+      "loss": 0.051,
+      "step": 4072
+    },
+    {
+      "epoch": 0.8536994340809054,
+      "grad_norm": 0.01553434506058693,
+      "learning_rate": 5.526216089384872e-07,
+      "loss": 0.0532,
+      "step": 4073
+    },
+    {
+      "epoch": 0.853909033745546,
+      "grad_norm": 0.016066189855337143,
+      "learning_rate": 5.510712478498675e-07,
+      "loss": 0.0515,
+      "step": 4074
+    },
+    {
+      "epoch": 0.8541186334101866,
+      "grad_norm": 0.017956750467419624,
+      "learning_rate": 5.49522937716479e-07,
+      "loss": 0.054,
+      "step": 4075
+    },
+    {
+      "epoch": 0.8543282330748271,
+      "grad_norm": 0.016211628913879395,
+      "learning_rate": 5.479766792520908e-07,
+      "loss": 0.0554,
+      "step": 4076
+    },
+    {
+      "epoch": 0.8545378327394676,
+      "grad_norm": 0.02058112435042858,
+      "learning_rate": 5.464324731695286e-07,
+      "loss": 0.0564,
+      "step": 4077
+    },
+    {
+      "epoch": 0.8547474324041081,
+      "grad_norm": 0.020162228494882584,
+      "learning_rate": 5.448903201806727e-07,
+      "loss": 0.0581,
+      "step": 4078
+    },
+    {
+      "epoch": 0.8549570320687487,
+      "grad_norm": 0.017797335982322693,
+      "learning_rate": 5.433502209964531e-07,
+      "loss": 0.0549,
+      "step": 4079
+    },
+    {
+      "epoch": 0.8551666317333892,
+      "grad_norm": 0.017947765067219734,
+      "learning_rate": 5.418121763268553e-07,
+      "loss": 0.0527,
+      "step": 4080
+    },
+    {
+      "epoch": 0.8553762313980298,
+      "grad_norm": 0.02111784741282463,
+      "learning_rate": 5.402761868809181e-07,
+      "loss": 0.0541,
+      "step": 4081
+    },
+    {
+      "epoch": 0.8555858310626703,
+      "grad_norm": 0.023570656776428223,
+      "learning_rate": 5.387422533667336e-07,
+      "loss": 0.0538,
+      "step": 4082
+    },
+    {
+      "epoch": 0.8557954307273108,
+      "grad_norm": 0.023569107055664062,
+      "learning_rate": 5.372103764914421e-07,
+      "loss": 0.0557,
+      "step": 4083
+    },
+    {
+      "epoch": 0.8560050303919514,
+      "grad_norm": 0.02265586517751217,
+      "learning_rate": 5.356805569612417e-07,
+      "loss": 0.0549,
+      "step": 4084
+    },
+    {
+      "epoch": 0.8562146300565919,
+      "grad_norm": 0.03135437145829201,
+      "learning_rate": 5.341527954813763e-07,
+      "loss": 0.0536,
+      "step": 4085
+    },
+    {
+      "epoch": 0.8564242297212324,
+      "grad_norm": 0.015480936504900455,
+      "learning_rate": 5.326270927561444e-07,
+      "loss": 0.0559,
+      "step": 4086
+    },
+    {
+      "epoch": 0.856633829385873,
+      "grad_norm": 0.025915948674082756,
+      "learning_rate": 5.311034494888945e-07,
+      "loss": 0.0521,
+      "step": 4087
+    },
+    {
+      "epoch": 0.8568434290505135,
+      "grad_norm": 0.025078527629375458,
+      "learning_rate": 5.295818663820268e-07,
+      "loss": 0.0523,
+      "step": 4088
+    },
+    {
+      "epoch": 0.8570530287151541,
+      "grad_norm": 0.015537966974079609,
+      "learning_rate": 5.280623441369897e-07,
+      "loss": 0.0529,
+      "step": 4089
+    },
+    {
+      "epoch": 0.8572626283797946,
+      "grad_norm": 0.028883641585707664,
+      "learning_rate": 5.265448834542836e-07,
+      "loss": 0.0587,
+      "step": 4090
+    },
+    {
+      "epoch": 0.8574722280444351,
+      "grad_norm": 0.018950091674923897,
+      "learning_rate": 5.250294850334564e-07,
+      "loss": 0.0531,
+      "step": 4091
+    },
+    {
+      "epoch": 0.8576818277090756,
+      "grad_norm": 0.019102994352579117,
+      "learning_rate": 5.235161495731079e-07,
+      "loss": 0.0513,
+      "step": 4092
+    },
+    {
+      "epoch": 0.8578914273737162,
+      "grad_norm": 0.02294645458459854,
+      "learning_rate": 5.22004877770883e-07,
+      "loss": 0.0553,
+      "step": 4093
+    },
+    {
+      "epoch": 0.8581010270383568,
+      "grad_norm": 0.018397770822048187,
+      "learning_rate": 5.204956703234804e-07,
+      "loss": 0.0538,
+      "step": 4094
+    },
+    {
+      "epoch": 0.8583106267029973,
+      "grad_norm": 0.019318964332342148,
+      "learning_rate": 5.189885279266433e-07,
+      "loss": 0.0546,
+      "step": 4095
+    },
+    {
+      "epoch": 0.8585202263676378,
+      "grad_norm": 0.018967201933264732,
+      "learning_rate": 5.174834512751647e-07,
+      "loss": 0.055,
+      "step": 4096
+    },
+    {
+      "epoch": 0.8587298260322783,
+      "grad_norm": 0.016506996005773544,
+      "learning_rate": 5.159804410628827e-07,
+      "loss": 0.0524,
+      "step": 4097
+    },
+    {
+      "epoch": 0.8589394256969188,
+      "grad_norm": 0.015648337081074715,
+      "learning_rate": 5.144794979826867e-07,
+      "loss": 0.0533,
+      "step": 4098
+    },
+    {
+      "epoch": 0.8591490253615595,
+      "grad_norm": 0.016446251422166824,
+      "learning_rate": 5.129806227265099e-07,
+      "loss": 0.0533,
+      "step": 4099
+    },
+    {
+      "epoch": 0.8593586250262,
+      "grad_norm": 0.020193303003907204,
+      "learning_rate": 5.114838159853336e-07,
+      "loss": 0.0549,
+      "step": 4100
+    },
+    {
+      "epoch": 0.8595682246908405,
+      "grad_norm": 0.015607200562953949,
+      "learning_rate": 5.099890784491879e-07,
+      "loss": 0.0559,
+      "step": 4101
+    },
+    {
+      "epoch": 0.859777824355481,
+      "grad_norm": 0.01757110469043255,
+      "learning_rate": 5.084964108071428e-07,
+      "loss": 0.0546,
+      "step": 4102
+    },
+    {
+      "epoch": 0.8599874240201215,
+      "grad_norm": 0.021767864003777504,
+      "learning_rate": 5.070058137473188e-07,
+      "loss": 0.0536,
+      "step": 4103
+    },
+    {
+      "epoch": 0.8601970236847621,
+      "grad_norm": 0.01381634920835495,
+      "learning_rate": 5.055172879568815e-07,
+      "loss": 0.0543,
+      "step": 4104
+    },
+    {
+      "epoch": 0.8604066233494027,
+      "grad_norm": 0.016660235822200775,
+      "learning_rate": 5.040308341220418e-07,
+      "loss": 0.0551,
+      "step": 4105
+    },
+    {
+      "epoch": 0.8606162230140432,
+      "grad_norm": 0.0208145659416914,
+      "learning_rate": 5.025464529280538e-07,
+      "loss": 0.0551,
+      "step": 4106
+    },
+    {
+      "epoch": 0.8608258226786837,
+      "grad_norm": 0.013750380836427212,
+      "learning_rate": 5.010641450592158e-07,
+      "loss": 0.0538,
+      "step": 4107
+    },
+    {
+      "epoch": 0.8610354223433242,
+      "grad_norm": 0.01817614957690239,
+      "learning_rate": 4.995839111988737e-07,
+      "loss": 0.0553,
+      "step": 4108
+    },
+    {
+      "epoch": 0.8612450220079648,
+      "grad_norm": 0.019916629418730736,
+      "learning_rate": 4.981057520294124e-07,
+      "loss": 0.0522,
+      "step": 4109
+    },
+    {
+      "epoch": 0.8614546216726053,
+      "grad_norm": 0.012952596880495548,
+      "learning_rate": 4.966296682322641e-07,
+      "loss": 0.0554,
+      "step": 4110
+    },
+    {
+      "epoch": 0.8616642213372459,
+      "grad_norm": 0.01568775624036789,
+      "learning_rate": 4.951556604879049e-07,
+      "loss": 0.0523,
+      "step": 4111
+    },
+    {
+      "epoch": 0.8618738210018864,
+      "grad_norm": 0.020983314141631126,
+      "learning_rate": 4.936837294758501e-07,
+      "loss": 0.0545,
+      "step": 4112
+    },
+    {
+      "epoch": 0.8620834206665269,
+      "grad_norm": 0.017767738550901413,
+      "learning_rate": 4.922138758746587e-07,
+      "loss": 0.0548,
+      "step": 4113
+    },
+    {
+      "epoch": 0.8622930203311675,
+      "grad_norm": 0.016531750559806824,
+      "learning_rate": 4.907461003619346e-07,
+      "loss": 0.0573,
+      "step": 4114
+    },
+    {
+      "epoch": 0.862502619995808,
+      "grad_norm": 0.019049130380153656,
+      "learning_rate": 4.892804036143223e-07,
+      "loss": 0.0538,
+      "step": 4115
+    },
+    {
+      "epoch": 0.8627122196604485,
+      "grad_norm": 0.01858980767428875,
+      "learning_rate": 4.878167863075061e-07,
+      "loss": 0.054,
+      "step": 4116
+    },
+    {
+      "epoch": 0.862921819325089,
+      "grad_norm": 0.013377662748098373,
+      "learning_rate": 4.863552491162149e-07,
+      "loss": 0.0547,
+      "step": 4117
+    },
+    {
+      "epoch": 0.8631314189897297,
+      "grad_norm": 0.022448167204856873,
+      "learning_rate": 4.848957927142167e-07,
+      "loss": 0.0553,
+      "step": 4118
+    },
+    {
+      "epoch": 0.8633410186543702,
+      "grad_norm": 0.019559383392333984,
+      "learning_rate": 4.83438417774319e-07,
+      "loss": 0.0549,
+      "step": 4119
+    },
+    {
+      "epoch": 0.8635506183190107,
+      "grad_norm": 0.017378434538841248,
+      "learning_rate": 4.819831249683726e-07,
+      "loss": 0.0533,
+      "step": 4120
+    },
+    {
+      "epoch": 0.8637602179836512,
+      "grad_norm": 0.022112280130386353,
+      "learning_rate": 4.805299149672682e-07,
+      "loss": 0.0558,
+      "step": 4121
+    },
+    {
+      "epoch": 0.8639698176482917,
+      "grad_norm": 0.02191154658794403,
+      "learning_rate": 4.790787884409332e-07,
+      "loss": 0.0577,
+      "step": 4122
+    },
+    {
+      "epoch": 0.8641794173129324,
+      "grad_norm": 0.015262911096215248,
+      "learning_rate": 4.776297460583384e-07,
+      "loss": 0.057,
+      "step": 4123
+    },
+    {
+      "epoch": 0.8643890169775729,
+      "grad_norm": 0.01982884109020233,
+      "learning_rate": 4.7618278848749146e-07,
+      "loss": 0.0536,
+      "step": 4124
+    },
+    {
+      "epoch": 0.8645986166422134,
+      "grad_norm": 0.024044396355748177,
+      "learning_rate": 4.7473791639543853e-07,
+      "loss": 0.0529,
+      "step": 4125
+    },
+    {
+      "epoch": 0.8648082163068539,
+      "grad_norm": 0.015614238567650318,
+      "learning_rate": 4.732951304482658e-07,
+      "loss": 0.0541,
+      "step": 4126
+    },
+    {
+      "epoch": 0.8650178159714944,
+      "grad_norm": 0.02073797397315502,
+      "learning_rate": 4.7185443131109785e-07,
+      "loss": 0.0526,
+      "step": 4127
+    },
+    {
+      "epoch": 0.865227415636135,
+      "grad_norm": 0.026310456916689873,
+      "learning_rate": 4.7041581964809733e-07,
+      "loss": 0.0539,
+      "step": 4128
+    },
+    {
+      "epoch": 0.8654370153007755,
+      "grad_norm": 0.02114870585501194,
+      "learning_rate": 4.6897929612246317e-07,
+      "loss": 0.0532,
+      "step": 4129
+    },
+    {
+      "epoch": 0.8656466149654161,
+      "grad_norm": 0.020745746791362762,
+      "learning_rate": 4.675448613964317e-07,
+      "loss": 0.0538,
+      "step": 4130
+    },
+    {
+      "epoch": 0.8658562146300566,
+      "grad_norm": 0.02710759826004505,
+      "learning_rate": 4.6611251613127793e-07,
+      "loss": 0.0548,
+      "step": 4131
+    },
+    {
+      "epoch": 0.8660658142946971,
+      "grad_norm": 0.020626511424779892,
+      "learning_rate": 4.64682260987312e-07,
+      "loss": 0.0534,
+      "step": 4132
+    },
+    {
+      "epoch": 0.8662754139593377,
+      "grad_norm": 0.020137401297688484,
+      "learning_rate": 4.6325409662388133e-07,
+      "loss": 0.0547,
+      "step": 4133
+    },
+    {
+      "epoch": 0.8664850136239782,
+      "grad_norm": 0.0274630356580019,
+      "learning_rate": 4.618280236993711e-07,
+      "loss": 0.0579,
+      "step": 4134
+    },
+    {
+      "epoch": 0.8666946132886187,
+      "grad_norm": 0.02381911128759384,
+      "learning_rate": 4.6040404287119924e-07,
+      "loss": 0.0563,
+      "step": 4135
+    },
+    {
+      "epoch": 0.8669042129532593,
+      "grad_norm": 0.014909719116985798,
+      "learning_rate": 4.589821547958195e-07,
+      "loss": 0.0531,
+      "step": 4136
+    },
+    {
+      "epoch": 0.8671138126178998,
+      "grad_norm": 0.02371630258858204,
+      "learning_rate": 4.5756236012872324e-07,
+      "loss": 0.053,
+      "step": 4137
+    },
+    {
+      "epoch": 0.8673234122825404,
+      "grad_norm": 0.028402652591466904,
+      "learning_rate": 4.561446595244362e-07,
+      "loss": 0.0536,
+      "step": 4138
+    },
+    {
+      "epoch": 0.8675330119471809,
+      "grad_norm": 0.01664336584508419,
+      "learning_rate": 4.547290536365173e-07,
+      "loss": 0.0545,
+      "step": 4139
+    },
+    {
+      "epoch": 0.8677426116118214,
+      "grad_norm": 0.022892151027917862,
+      "learning_rate": 4.5331554311755956e-07,
+      "loss": 0.0527,
+      "step": 4140
+    },
+    {
+      "epoch": 0.8679522112764619,
+      "grad_norm": 0.027045048773288727,
+      "learning_rate": 4.519041286191933e-07,
+      "loss": 0.053,
+      "step": 4141
+    },
+    {
+      "epoch": 0.8681618109411025,
+      "grad_norm": 0.021929344162344933,
+      "learning_rate": 4.504948107920781e-07,
+      "loss": 0.0518,
+      "step": 4142
+    },
+    {
+      "epoch": 0.8683714106057431,
+      "grad_norm": 0.016940593719482422,
+      "learning_rate": 4.490875902859099e-07,
+      "loss": 0.0526,
+      "step": 4143
+    },
+    {
+      "epoch": 0.8685810102703836,
+      "grad_norm": 0.03185591474175453,
+      "learning_rate": 4.476824677494179e-07,
+      "loss": 0.0529,
+      "step": 4144
+    },
+    {
+      "epoch": 0.8687906099350241,
+      "grad_norm": 0.02771037444472313,
+      "learning_rate": 4.46279443830363e-07,
+      "loss": 0.0572,
+      "step": 4145
+    },
+    {
+      "epoch": 0.8690002095996646,
+      "grad_norm": 0.017459120601415634,
+      "learning_rate": 4.448785191755378e-07,
+      "loss": 0.0521,
+      "step": 4146
+    },
+    {
+      "epoch": 0.8692098092643051,
+      "grad_norm": 0.03185920789837837,
+      "learning_rate": 4.4347969443076956e-07,
+      "loss": 0.0549,
+      "step": 4147
+    },
+    {
+      "epoch": 0.8694194089289458,
+      "grad_norm": 0.03139664977788925,
+      "learning_rate": 4.420829702409152e-07,
+      "loss": 0.053,
+      "step": 4148
+    },
+    {
+      "epoch": 0.8696290085935863,
+      "grad_norm": 0.023265665397047997,
+      "learning_rate": 4.4068834724986466e-07,
+      "loss": 0.054,
+      "step": 4149
+    },
+    {
+      "epoch": 0.8698386082582268,
+      "grad_norm": 0.026613622903823853,
+      "learning_rate": 4.3929582610053976e-07,
+      "loss": 0.0566,
+      "step": 4150
+    },
+    {
+      "epoch": 0.8700482079228673,
+      "grad_norm": 0.03510505333542824,
+      "learning_rate": 4.3790540743489207e-07,
+      "loss": 0.0543,
+      "step": 4151
+    },
+    {
+      "epoch": 0.8702578075875078,
+      "grad_norm": 0.03017215058207512,
+      "learning_rate": 4.365170918939027e-07,
+      "loss": 0.0546,
+      "step": 4152
+    },
+    {
+      "epoch": 0.8704674072521484,
+      "grad_norm": 0.020732155069708824,
+      "learning_rate": 4.351308801175863e-07,
+      "loss": 0.0524,
+      "step": 4153
+    },
+    {
+      "epoch": 0.870677006916789,
+      "grad_norm": 0.032381054013967514,
+      "learning_rate": 4.337467727449862e-07,
+      "loss": 0.0544,
+      "step": 4154
+    },
+    {
+      "epoch": 0.8708866065814295,
+      "grad_norm": 0.03272484615445137,
+      "learning_rate": 4.323647704141754e-07,
+      "loss": 0.0546,
+      "step": 4155
+    },
+    {
+      "epoch": 0.87109620624607,
+      "grad_norm": 0.019726725295186043,
+      "learning_rate": 4.309848737622568e-07,
+      "loss": 0.0547,
+      "step": 4156
+    },
+    {
+      "epoch": 0.8713058059107105,
+      "grad_norm": 0.0256227757781744,
+      "learning_rate": 4.2960708342536295e-07,
+      "loss": 0.0544,
+      "step": 4157
+    },
+    {
+      "epoch": 0.8715154055753511,
+      "grad_norm": 0.03493373841047287,
+      "learning_rate": 4.2823140003865283e-07,
+      "loss": 0.0541,
+      "step": 4158
+    },
+    {
+      "epoch": 0.8717250052399916,
+      "grad_norm": 0.022653086110949516,
+      "learning_rate": 4.2685782423631806e-07,
+      "loss": 0.0534,
+      "step": 4159
+    },
+    {
+      "epoch": 0.8719346049046321,
+      "grad_norm": 0.021094655618071556,
+      "learning_rate": 4.2548635665157713e-07,
+      "loss": 0.0501,
+      "step": 4160
+    },
+    {
+      "epoch": 0.8721442045692727,
+      "grad_norm": 0.03143506497144699,
+      "learning_rate": 4.241169979166748e-07,
+      "loss": 0.0533,
+      "step": 4161
+    },
+    {
+      "epoch": 0.8723538042339132,
+      "grad_norm": 0.029018806293606758,
+      "learning_rate": 4.2274974866288675e-07,
+      "loss": 0.0539,
+      "step": 4162
+    },
+    {
+      "epoch": 0.8725634038985538,
+      "grad_norm": 0.01749360002577305,
+      "learning_rate": 4.213846095205126e-07,
+      "loss": 0.0518,
+      "step": 4163
+    },
+    {
+      "epoch": 0.8727730035631943,
+      "grad_norm": 0.02830035425722599,
+      "learning_rate": 4.2002158111888345e-07,
+      "loss": 0.0551,
+      "step": 4164
+    },
+    {
+      "epoch": 0.8729826032278348,
+      "grad_norm": 0.03221756964921951,
+      "learning_rate": 4.186606640863533e-07,
+      "loss": 0.054,
+      "step": 4165
+    },
+    {
+      "epoch": 0.8731922028924753,
+      "grad_norm": 0.021834442391991615,
+      "learning_rate": 4.1730185905030527e-07,
+      "loss": 0.0535,
+      "step": 4166
+    },
+    {
+      "epoch": 0.8734018025571159,
+      "grad_norm": 0.025496196001768112,
+      "learning_rate": 4.1594516663714946e-07,
+      "loss": 0.0545,
+      "step": 4167
+    },
+    {
+      "epoch": 0.8736114022217565,
+      "grad_norm": 0.031498175114393234,
+      "learning_rate": 4.145905874723194e-07,
+      "loss": 0.0569,
+      "step": 4168
+    },
+    {
+      "epoch": 0.873821001886397,
+      "grad_norm": 0.023885555565357208,
+      "learning_rate": 4.1323812218027506e-07,
+      "loss": 0.0548,
+      "step": 4169
+    },
+    {
+      "epoch": 0.8740306015510375,
+      "grad_norm": 0.019355174154043198,
+      "learning_rate": 4.1188777138450487e-07,
+      "loss": 0.0546,
+      "step": 4170
+    },
+    {
+      "epoch": 0.874240201215678,
+      "grad_norm": 0.029070964083075523,
+      "learning_rate": 4.1053953570751813e-07,
+      "loss": 0.055,
+      "step": 4171
+    },
+    {
+      "epoch": 0.8744498008803185,
+      "grad_norm": 0.030413752421736717,
+      "learning_rate": 4.0919341577085157e-07,
+      "loss": 0.0537,
+      "step": 4172
+    },
+    {
+      "epoch": 0.8746594005449592,
+      "grad_norm": 0.0210228580981493,
+      "learning_rate": 4.078494121950682e-07,
+      "loss": 0.0545,
+      "step": 4173
+    },
+    {
+      "epoch": 0.8748690002095997,
+      "grad_norm": 0.026022691279649734,
+      "learning_rate": 4.065075255997514e-07,
+      "loss": 0.0524,
+      "step": 4174
+    },
+    {
+      "epoch": 0.8750785998742402,
+      "grad_norm": 0.03338058665394783,
+      "learning_rate": 4.051677566035106e-07,
+      "loss": 0.0534,
+      "step": 4175
+    },
+    {
+      "epoch": 0.8752881995388807,
+      "grad_norm": 0.024644313380122185,
+      "learning_rate": 4.038301058239796e-07,
+      "loss": 0.0516,
+      "step": 4176
+    },
+    {
+      "epoch": 0.8754977992035212,
+      "grad_norm": 0.019061360508203506,
+      "learning_rate": 4.024945738778163e-07,
+      "loss": 0.0525,
+      "step": 4177
+    },
+    {
+      "epoch": 0.8757073988681618,
+      "grad_norm": 0.035856083035469055,
+      "learning_rate": 4.011611613806987e-07,
+      "loss": 0.0563,
+      "step": 4178
+    },
+    {
+      "epoch": 0.8759169985328024,
+      "grad_norm": 0.03140418231487274,
+      "learning_rate": 3.998298689473301e-07,
+      "loss": 0.0531,
+      "step": 4179
+    },
+    {
+      "epoch": 0.8761265981974429,
+      "grad_norm": 0.01924203895032406,
+      "learning_rate": 3.9850069719143735e-07,
+      "loss": 0.0564,
+      "step": 4180
+    },
+    {
+      "epoch": 0.8763361978620834,
+      "grad_norm": 0.028375735506415367,
+      "learning_rate": 3.971736467257659e-07,
+      "loss": 0.0528,
+      "step": 4181
+    },
+    {
+      "epoch": 0.8765457975267239,
+      "grad_norm": 0.03424501046538353,
+      "learning_rate": 3.958487181620879e-07,
+      "loss": 0.0542,
+      "step": 4182
+    },
+    {
+      "epoch": 0.8767553971913645,
+      "grad_norm": 0.02565137669444084,
+      "learning_rate": 3.9452591211119496e-07,
+      "loss": 0.0556,
+      "step": 4183
+    },
+    {
+      "epoch": 0.876964996856005,
+      "grad_norm": 0.021563977003097534,
+      "learning_rate": 3.9320522918289973e-07,
+      "loss": 0.0527,
+      "step": 4184
+    },
+    {
+      "epoch": 0.8771745965206456,
+      "grad_norm": 0.027671657502651215,
+      "learning_rate": 3.918866699860363e-07,
+      "loss": 0.0522,
+      "step": 4185
+    },
+    {
+      "epoch": 0.8773841961852861,
+      "grad_norm": 0.032790180295705795,
+      "learning_rate": 3.9057023512846127e-07,
+      "loss": 0.056,
+      "step": 4186
+    },
+    {
+      "epoch": 0.8775937958499267,
+      "grad_norm": 0.023974338546395302,
+      "learning_rate": 3.892559252170508e-07,
+      "loss": 0.0528,
+      "step": 4187
+    },
+    {
+      "epoch": 0.8778033955145672,
+      "grad_norm": 0.02199658937752247,
+      "learning_rate": 3.879437408576997e-07,
+      "loss": 0.0536,
+      "step": 4188
+    },
+    {
+      "epoch": 0.8780129951792077,
+      "grad_norm": 0.03704606741666794,
+      "learning_rate": 3.866336826553274e-07,
+      "loss": 0.0531,
+      "step": 4189
+    },
+    {
+      "epoch": 0.8782225948438482,
+      "grad_norm": 0.027375850826501846,
+      "learning_rate": 3.8532575121386916e-07,
+      "loss": 0.0519,
+      "step": 4190
+    },
+    {
+      "epoch": 0.8784321945084888,
+      "grad_norm": 0.01718190312385559,
+      "learning_rate": 3.8401994713628044e-07,
+      "loss": 0.0551,
+      "step": 4191
+    },
+    {
+      "epoch": 0.8786417941731294,
+      "grad_norm": 0.031402431428432465,
+      "learning_rate": 3.827162710245369e-07,
+      "loss": 0.0557,
+      "step": 4192
+    },
+    {
+      "epoch": 0.8788513938377699,
+      "grad_norm": 0.033082105219364166,
+      "learning_rate": 3.814147234796345e-07,
+      "loss": 0.0532,
+      "step": 4193
+    },
+    {
+      "epoch": 0.8790609935024104,
+      "grad_norm": 0.02236991748213768,
+      "learning_rate": 3.8011530510158377e-07,
+      "loss": 0.0544,
+      "step": 4194
+    },
+    {
+      "epoch": 0.8792705931670509,
+      "grad_norm": 0.02740001678466797,
+      "learning_rate": 3.78818016489419e-07,
+      "loss": 0.0549,
+      "step": 4195
+    },
+    {
+      "epoch": 0.8794801928316914,
+      "grad_norm": 0.03492108732461929,
+      "learning_rate": 3.775228582411877e-07,
+      "loss": 0.0536,
+      "step": 4196
+    },
+    {
+      "epoch": 0.8796897924963321,
+      "grad_norm": 0.029871255159378052,
+      "learning_rate": 3.7622983095395973e-07,
+      "loss": 0.0522,
+      "step": 4197
+    },
+    {
+      "epoch": 0.8798993921609726,
+      "grad_norm": 0.02398385852575302,
+      "learning_rate": 3.7493893522381866e-07,
+      "loss": 0.0529,
+      "step": 4198
+    },
+    {
+      "epoch": 0.8801089918256131,
+      "grad_norm": 0.028454309329390526,
+      "learning_rate": 3.736501716458668e-07,
+      "loss": 0.0572,
+      "step": 4199
+    },
+    {
+      "epoch": 0.8803185914902536,
+      "grad_norm": 0.038288019597530365,
+      "learning_rate": 3.723635408142262e-07,
+      "loss": 0.0538,
+      "step": 4200
+    },
+    {
+      "epoch": 0.8805281911548941,
+      "grad_norm": 0.030530139803886414,
+      "learning_rate": 3.710790433220324e-07,
+      "loss": 0.0522,
+      "step": 4201
+    },
+    {
+      "epoch": 0.8807377908195347,
+      "grad_norm": 0.021036000922322273,
+      "learning_rate": 3.6979667976143663e-07,
+      "loss": 0.0532,
+      "step": 4202
+    },
+    {
+      "epoch": 0.8809473904841753,
+      "grad_norm": 0.035266853868961334,
+      "learning_rate": 3.685164507236111e-07,
+      "loss": 0.0538,
+      "step": 4203
+    },
+    {
+      "epoch": 0.8811569901488158,
+      "grad_norm": 0.03357855603098869,
+      "learning_rate": 3.6723835679873864e-07,
+      "loss": 0.0546,
+      "step": 4204
+    },
+    {
+      "epoch": 0.8813665898134563,
+      "grad_norm": 0.025351261720061302,
+      "learning_rate": 3.6596239857602136e-07,
+      "loss": 0.0522,
+      "step": 4205
+    },
+    {
+      "epoch": 0.8815761894780968,
+      "grad_norm": 0.027122756466269493,
+      "learning_rate": 3.6468857664367753e-07,
+      "loss": 0.0541,
+      "step": 4206
+    },
+    {
+      "epoch": 0.8817857891427374,
+      "grad_norm": 0.03549753874540329,
+      "learning_rate": 3.6341689158893524e-07,
+      "loss": 0.0535,
+      "step": 4207
+    },
+    {
+      "epoch": 0.8819953888073779,
+      "grad_norm": 0.03277155011892319,
+      "learning_rate": 3.6214734399804277e-07,
+      "loss": 0.0532,
+      "step": 4208
+    },
+    {
+      "epoch": 0.8822049884720184,
+      "grad_norm": 0.019641457125544548,
+      "learning_rate": 3.6087993445626135e-07,
+      "loss": 0.0543,
+      "step": 4209
+    },
+    {
+      "epoch": 0.882414588136659,
+      "grad_norm": 0.027548063546419144,
+      "learning_rate": 3.596146635478676e-07,
+      "loss": 0.0563,
+      "step": 4210
+    },
+    {
+      "epoch": 0.8826241878012995,
+      "grad_norm": 0.03249534219503403,
+      "learning_rate": 3.583515318561498e-07,
+      "loss": 0.0534,
+      "step": 4211
+    },
+    {
+      "epoch": 0.8828337874659401,
+      "grad_norm": 0.030830688774585724,
+      "learning_rate": 3.570905399634111e-07,
+      "loss": 0.0559,
+      "step": 4212
+    },
+    {
+      "epoch": 0.8830433871305806,
+      "grad_norm": 0.019546935334801674,
+      "learning_rate": 3.558316884509694e-07,
+      "loss": 0.054,
+      "step": 4213
+    },
+    {
+      "epoch": 0.8832529867952211,
+      "grad_norm": 0.025954263284802437,
+      "learning_rate": 3.545749778991542e-07,
+      "loss": 0.0538,
+      "step": 4214
+    },
+    {
+      "epoch": 0.8834625864598616,
+      "grad_norm": 0.03318324312567711,
+      "learning_rate": 3.5332040888730935e-07,
+      "loss": 0.052,
+      "step": 4215
+    },
+    {
+      "epoch": 0.8836721861245022,
+      "grad_norm": 0.0273283664137125,
+      "learning_rate": 3.5206798199379166e-07,
+      "loss": 0.0548,
+      "step": 4216
+    },
+    {
+      "epoch": 0.8838817857891428,
+      "grad_norm": 0.02014850080013275,
+      "learning_rate": 3.508176977959682e-07,
+      "loss": 0.0536,
+      "step": 4217
+    },
+    {
+      "epoch": 0.8840913854537833,
+      "grad_norm": 0.02724394015967846,
+      "learning_rate": 3.495695568702201e-07,
+      "loss": 0.0559,
+      "step": 4218
+    },
+    {
+      "epoch": 0.8843009851184238,
+      "grad_norm": 0.028185635805130005,
+      "learning_rate": 3.483235597919404e-07,
+      "loss": 0.0514,
+      "step": 4219
+    },
+    {
+      "epoch": 0.8845105847830643,
+      "grad_norm": 0.02001328580081463,
+      "learning_rate": 3.4707970713553364e-07,
+      "loss": 0.0539,
+      "step": 4220
+    },
+    {
+      "epoch": 0.8847201844477048,
+      "grad_norm": 0.01998419128358364,
+      "learning_rate": 3.4583799947441487e-07,
+      "loss": 0.0523,
+      "step": 4221
+    },
+    {
+      "epoch": 0.8849297841123455,
+      "grad_norm": 0.030080687254667282,
+      "learning_rate": 3.445984373810124e-07,
+      "loss": 0.0544,
+      "step": 4222
+    },
+    {
+      "epoch": 0.885139383776986,
+      "grad_norm": 0.026855189353227615,
+      "learning_rate": 3.4336102142676354e-07,
+      "loss": 0.0522,
+      "step": 4223
+    },
+    {
+      "epoch": 0.8853489834416265,
+      "grad_norm": 0.01764325052499771,
+      "learning_rate": 3.421257521821153e-07,
+      "loss": 0.0546,
+      "step": 4224
+    },
+    {
+      "epoch": 0.885558583106267,
+      "grad_norm": 0.02428247220814228,
+      "learning_rate": 3.408926302165283e-07,
+      "loss": 0.0541,
+      "step": 4225
+    },
+    {
+      "epoch": 0.8857681827709075,
+      "grad_norm": 0.025871314108371735,
+      "learning_rate": 3.396616560984711e-07,
+      "loss": 0.0529,
+      "step": 4226
+    },
+    {
+      "epoch": 0.8859777824355481,
+      "grad_norm": 0.02489456534385681,
+      "learning_rate": 3.384328303954221e-07,
+      "loss": 0.0562,
+      "step": 4227
+    },
+    {
+      "epoch": 0.8861873821001887,
+      "grad_norm": 0.017987683415412903,
+      "learning_rate": 3.372061536738708e-07,
+      "loss": 0.0529,
+      "step": 4228
+    },
+    {
+      "epoch": 0.8863969817648292,
+      "grad_norm": 0.025899479165673256,
+      "learning_rate": 3.3598162649931374e-07,
+      "loss": 0.0577,
+      "step": 4229
+    },
+    {
+      "epoch": 0.8866065814294697,
+      "grad_norm": 0.030789077281951904,
+      "learning_rate": 3.347592494362578e-07,
+      "loss": 0.0536,
+      "step": 4230
+    },
+    {
+      "epoch": 0.8868161810941102,
+      "grad_norm": 0.024812959134578705,
+      "learning_rate": 3.3353902304821826e-07,
+      "loss": 0.0562,
+      "step": 4231
+    },
+    {
+      "epoch": 0.8870257807587508,
+      "grad_norm": 0.01769872196018696,
+      "learning_rate": 3.323209478977202e-07,
+      "loss": 0.0534,
+      "step": 4232
+    },
+    {
+      "epoch": 0.8872353804233913,
+      "grad_norm": 0.026812558993697166,
+      "learning_rate": 3.31105024546296e-07,
+      "loss": 0.054,
+      "step": 4233
+    },
+    {
+      "epoch": 0.8874449800880319,
+      "grad_norm": 0.0294424407184124,
+      "learning_rate": 3.2989125355448623e-07,
+      "loss": 0.0524,
+      "step": 4234
+    },
+    {
+      "epoch": 0.8876545797526724,
+      "grad_norm": 0.02184954471886158,
+      "learning_rate": 3.2867963548183723e-07,
+      "loss": 0.052,
+      "step": 4235
+    },
+    {
+      "epoch": 0.8878641794173129,
+      "grad_norm": 0.01684270054101944,
+      "learning_rate": 3.274701708869066e-07,
+      "loss": 0.0565,
+      "step": 4236
+    },
+    {
+      "epoch": 0.8880737790819535,
+      "grad_norm": 0.028860921040177345,
+      "learning_rate": 3.2626286032725664e-07,
+      "loss": 0.0527,
+      "step": 4237
+    },
+    {
+      "epoch": 0.888283378746594,
+      "grad_norm": 0.02757331356406212,
+      "learning_rate": 3.250577043594566e-07,
+      "loss": 0.0569,
+      "step": 4238
+    },
+    {
+      "epoch": 0.8884929784112345,
+      "grad_norm": 0.020717767998576164,
+      "learning_rate": 3.238547035390843e-07,
+      "loss": 0.0569,
+      "step": 4239
+    },
+    {
+      "epoch": 0.888702578075875,
+      "grad_norm": 0.019048750400543213,
+      "learning_rate": 3.226538584207228e-07,
+      "loss": 0.0549,
+      "step": 4240
+    },
+    {
+      "epoch": 0.8889121777405156,
+      "grad_norm": 0.030015893280506134,
+      "learning_rate": 3.214551695579604e-07,
+      "loss": 0.0528,
+      "step": 4241
+    },
+    {
+      "epoch": 0.8891217774051562,
+      "grad_norm": 0.02680326998233795,
+      "learning_rate": 3.202586375033928e-07,
+      "loss": 0.0538,
+      "step": 4242
+    },
+    {
+      "epoch": 0.8893313770697967,
+      "grad_norm": 0.020373962819576263,
+      "learning_rate": 3.190642628086221e-07,
+      "loss": 0.0539,
+      "step": 4243
+    },
+    {
+      "epoch": 0.8895409767344372,
+      "grad_norm": 0.020322667434811592,
+      "learning_rate": 3.178720460242535e-07,
+      "loss": 0.0539,
+      "step": 4244
+    },
+    {
+      "epoch": 0.8897505763990777,
+      "grad_norm": 0.025189222767949104,
+      "learning_rate": 3.166819876998989e-07,
+      "loss": 0.0532,
+      "step": 4245
+    },
+    {
+      "epoch": 0.8899601760637182,
+      "grad_norm": 0.027642201632261276,
+      "learning_rate": 3.154940883841756e-07,
+      "loss": 0.0551,
+      "step": 4246
+    },
+    {
+      "epoch": 0.8901697757283589,
+      "grad_norm": 0.023163793608546257,
+      "learning_rate": 3.1430834862470395e-07,
+      "loss": 0.0545,
+      "step": 4247
+    },
+    {
+      "epoch": 0.8903793753929994,
+      "grad_norm": 0.01851850189268589,
+      "learning_rate": 3.131247689681099e-07,
+      "loss": 0.0553,
+      "step": 4248
+    },
+    {
+      "epoch": 0.8905889750576399,
+      "grad_norm": 0.028064940124750137,
+      "learning_rate": 3.1194334996002497e-07,
+      "loss": 0.0544,
+      "step": 4249
+    },
+    {
+      "epoch": 0.8907985747222804,
+      "grad_norm": 0.02554197795689106,
+      "learning_rate": 3.1076409214508164e-07,
+      "loss": 0.0533,
+      "step": 4250
+    },
+    {
+      "epoch": 0.8910081743869209,
+      "grad_norm": 0.018633106723427773,
+      "learning_rate": 3.095869960669173e-07,
+      "loss": 0.0544,
+      "step": 4251
+    },
+    {
+      "epoch": 0.8912177740515616,
+      "grad_norm": 0.019019601866602898,
+      "learning_rate": 3.084120622681741e-07,
+      "loss": 0.0535,
+      "step": 4252
+    },
+    {
+      "epoch": 0.8914273737162021,
+      "grad_norm": 0.026201697066426277,
+      "learning_rate": 3.072392912904948e-07,
+      "loss": 0.0555,
+      "step": 4253
+    },
+    {
+      "epoch": 0.8916369733808426,
+      "grad_norm": 0.027933111414313316,
+      "learning_rate": 3.0606868367452746e-07,
+      "loss": 0.0539,
+      "step": 4254
+    },
+    {
+      "epoch": 0.8918465730454831,
+      "grad_norm": 0.018472185358405113,
+      "learning_rate": 3.049002399599232e-07,
+      "loss": 0.0544,
+      "step": 4255
+    },
+    {
+      "epoch": 0.8920561727101237,
+      "grad_norm": 0.017080184072256088,
+      "learning_rate": 3.0373396068533234e-07,
+      "loss": 0.054,
+      "step": 4256
+    },
+    {
+      "epoch": 0.8922657723747642,
+      "grad_norm": 0.02512308955192566,
+      "learning_rate": 3.0256984638840967e-07,
+      "loss": 0.0548,
+      "step": 4257
+    },
+    {
+      "epoch": 0.8924753720394047,
+      "grad_norm": 0.02607639506459236,
+      "learning_rate": 3.0140789760581214e-07,
+      "loss": 0.0531,
+      "step": 4258
+    },
+    {
+      "epoch": 0.8926849717040453,
+      "grad_norm": 0.01739831455051899,
+      "learning_rate": 3.0024811487319837e-07,
+      "loss": 0.0551,
+      "step": 4259
+    },
+    {
+      "epoch": 0.8928945713686858,
+      "grad_norm": 0.017008259892463684,
+      "learning_rate": 2.990904987252269e-07,
+      "loss": 0.0544,
+      "step": 4260
+    },
+    {
+      "epoch": 0.8931041710333264,
+      "grad_norm": 0.024107355624437332,
+      "learning_rate": 2.9793504969555965e-07,
+      "loss": 0.053,
+      "step": 4261
+    },
+    {
+      "epoch": 0.8933137706979669,
+      "grad_norm": 0.026618480682373047,
+      "learning_rate": 2.9678176831685776e-07,
+      "loss": 0.0535,
+      "step": 4262
+    },
+    {
+      "epoch": 0.8935233703626074,
+      "grad_norm": 0.019204149022698402,
+      "learning_rate": 2.9563065512078325e-07,
+      "loss": 0.053,
+      "step": 4263
+    },
+    {
+      "epoch": 0.8937329700272479,
+      "grad_norm": 0.018961802124977112,
+      "learning_rate": 2.9448171063799933e-07,
+      "loss": 0.0525,
+      "step": 4264
+    },
+    {
+      "epoch": 0.8939425696918885,
+      "grad_norm": 0.022955385968089104,
+      "learning_rate": 2.933349353981701e-07,
+      "loss": 0.0565,
+      "step": 4265
+    },
+    {
+      "epoch": 0.8941521693565291,
+      "grad_norm": 0.023665906861424446,
+      "learning_rate": 2.921903299299572e-07,
+      "loss": 0.0532,
+      "step": 4266
+    },
+    {
+      "epoch": 0.8943617690211696,
+      "grad_norm": 0.02021893300116062,
+      "learning_rate": 2.9104789476102515e-07,
+      "loss": 0.0554,
+      "step": 4267
+    },
+    {
+      "epoch": 0.8945713686858101,
+      "grad_norm": 0.02034766785800457,
+      "learning_rate": 2.899076304180348e-07,
+      "loss": 0.0534,
+      "step": 4268
+    },
+    {
+      "epoch": 0.8947809683504506,
+      "grad_norm": 0.019094116985797882,
+      "learning_rate": 2.8876953742664914e-07,
+      "loss": 0.0563,
+      "step": 4269
+    },
+    {
+      "epoch": 0.8949905680150911,
+      "grad_norm": 0.022905079647898674,
+      "learning_rate": 2.876336163115273e-07,
+      "loss": 0.0533,
+      "step": 4270
+    },
+    {
+      "epoch": 0.8952001676797318,
+      "grad_norm": 0.022755665704607964,
+      "learning_rate": 2.8649986759632985e-07,
+      "loss": 0.0598,
+      "step": 4271
+    },
+    {
+      "epoch": 0.8954097673443723,
+      "grad_norm": 0.01669159345328808,
+      "learning_rate": 2.8536829180371485e-07,
+      "loss": 0.0517,
+      "step": 4272
+    },
+    {
+      "epoch": 0.8956193670090128,
+      "grad_norm": 0.021552741527557373,
+      "learning_rate": 2.842388894553377e-07,
+      "loss": 0.0517,
+      "step": 4273
+    },
+    {
+      "epoch": 0.8958289666736533,
+      "grad_norm": 0.0235520601272583,
+      "learning_rate": 2.8311166107185263e-07,
+      "loss": 0.0553,
+      "step": 4274
+    },
+    {
+      "epoch": 0.8960385663382938,
+      "grad_norm": 0.019750304520130157,
+      "learning_rate": 2.819866071729127e-07,
+      "loss": 0.0529,
+      "step": 4275
+    },
+    {
+      "epoch": 0.8962481660029344,
+      "grad_norm": 0.01832771860063076,
+      "learning_rate": 2.808637282771659e-07,
+      "loss": 0.0518,
+      "step": 4276
+    },
+    {
+      "epoch": 0.896457765667575,
+      "grad_norm": 0.019599363207817078,
+      "learning_rate": 2.7974302490226034e-07,
+      "loss": 0.05,
+      "step": 4277
+    },
+    {
+      "epoch": 0.8966673653322155,
+      "grad_norm": 0.02147594839334488,
+      "learning_rate": 2.786244975648406e-07,
+      "loss": 0.0559,
+      "step": 4278
+    },
+    {
+      "epoch": 0.896876964996856,
+      "grad_norm": 0.020394155755639076,
+      "learning_rate": 2.7750814678054626e-07,
+      "loss": 0.0533,
+      "step": 4279
+    },
+    {
+      "epoch": 0.8970865646614965,
+      "grad_norm": 0.014958519488573074,
+      "learning_rate": 2.7639397306401527e-07,
+      "loss": 0.052,
+      "step": 4280
+    },
+    {
+      "epoch": 0.8972961643261371,
+      "grad_norm": 0.017371637746691704,
+      "learning_rate": 2.7528197692888114e-07,
+      "loss": 0.0491,
+      "step": 4281
+    },
+    {
+      "epoch": 0.8975057639907776,
+      "grad_norm": 0.021931590512394905,
+      "learning_rate": 2.7417215888777493e-07,
+      "loss": 0.0539,
+      "step": 4282
+    },
+    {
+      "epoch": 0.8977153636554182,
+      "grad_norm": 0.022836145013570786,
+      "learning_rate": 2.730645194523218e-07,
+      "loss": 0.0569,
+      "step": 4283
+    },
+    {
+      "epoch": 0.8979249633200587,
+      "grad_norm": 0.01579204574227333,
+      "learning_rate": 2.719590591331428e-07,
+      "loss": 0.0532,
+      "step": 4284
+    },
+    {
+      "epoch": 0.8981345629846992,
+      "grad_norm": 0.018433870747685432,
+      "learning_rate": 2.7085577843985634e-07,
+      "loss": 0.0533,
+      "step": 4285
+    },
+    {
+      "epoch": 0.8983441626493398,
+      "grad_norm": 0.025535134598612785,
+      "learning_rate": 2.697546778810728e-07,
+      "loss": 0.0572,
+      "step": 4286
+    },
+    {
+      "epoch": 0.8985537623139803,
+      "grad_norm": 0.022626137360930443,
+      "learning_rate": 2.6865575796440004e-07,
+      "loss": 0.0534,
+      "step": 4287
+    },
+    {
+      "epoch": 0.8987633619786208,
+      "grad_norm": 0.01860016956925392,
+      "learning_rate": 2.675590191964406e-07,
+      "loss": 0.0553,
+      "step": 4288
+    },
+    {
+      "epoch": 0.8989729616432613,
+      "grad_norm": 0.020209958776831627,
+      "learning_rate": 2.6646446208279054e-07,
+      "loss": 0.0535,
+      "step": 4289
+    },
+    {
+      "epoch": 0.8991825613079019,
+      "grad_norm": 0.023047301918268204,
+      "learning_rate": 2.653720871280396e-07,
+      "loss": 0.0545,
+      "step": 4290
+    },
+    {
+      "epoch": 0.8993921609725425,
+      "grad_norm": 0.022766290232539177,
+      "learning_rate": 2.6428189483577283e-07,
+      "loss": 0.0541,
+      "step": 4291
+    },
+    {
+      "epoch": 0.899601760637183,
+      "grad_norm": 0.021814832463860512,
+      "learning_rate": 2.631938857085697e-07,
+      "loss": 0.0514,
+      "step": 4292
+    },
+    {
+      "epoch": 0.8998113603018235,
+      "grad_norm": 0.017104798927903175,
+      "learning_rate": 2.6210806024800083e-07,
+      "loss": 0.055,
+      "step": 4293
+    },
+    {
+      "epoch": 0.900020959966464,
+      "grad_norm": 0.019604802131652832,
+      "learning_rate": 2.6102441895463237e-07,
+      "loss": 0.0555,
+      "step": 4294
+    },
+    {
+      "epoch": 0.9002305596311045,
+      "grad_norm": 0.022560935467481613,
+      "learning_rate": 2.5994296232802254e-07,
+      "loss": 0.0527,
+      "step": 4295
+    },
+    {
+      "epoch": 0.9004401592957452,
+      "grad_norm": 0.017508791759610176,
+      "learning_rate": 2.5886369086672193e-07,
+      "loss": 0.0542,
+      "step": 4296
+    },
+    {
+      "epoch": 0.9006497589603857,
+      "grad_norm": 0.010546859353780746,
+      "learning_rate": 2.577866050682748e-07,
+      "loss": 0.0537,
+      "step": 4297
+    },
+    {
+      "epoch": 0.9008593586250262,
+      "grad_norm": 0.017833339050412178,
+      "learning_rate": 2.567117054292184e-07,
+      "loss": 0.0552,
+      "step": 4298
+    },
+    {
+      "epoch": 0.9010689582896667,
+      "grad_norm": 0.02061723917722702,
+      "learning_rate": 2.5563899244507974e-07,
+      "loss": 0.0548,
+      "step": 4299
+    },
+    {
+      "epoch": 0.9012785579543072,
+      "grad_norm": 0.016052190214395523,
+      "learning_rate": 2.545684666103809e-07,
+      "loss": 0.0593,
+      "step": 4300
+    },
+    {
+      "epoch": 0.9014881576189478,
+      "grad_norm": 0.013846572488546371,
+      "learning_rate": 2.5350012841863283e-07,
+      "loss": 0.0579,
+      "step": 4301
+    },
+    {
+      "epoch": 0.9016977572835884,
+      "grad_norm": 0.014581656083464622,
+      "learning_rate": 2.5243397836233975e-07,
+      "loss": 0.0558,
+      "step": 4302
+    },
+    {
+      "epoch": 0.9019073569482289,
+      "grad_norm": 0.021665887907147408,
+      "learning_rate": 2.513700169329963e-07,
+      "loss": 0.0534,
+      "step": 4303
+    },
+    {
+      "epoch": 0.9021169566128694,
+      "grad_norm": 0.020303577184677124,
+      "learning_rate": 2.503082446210886e-07,
+      "loss": 0.0511,
+      "step": 4304
+    },
+    {
+      "epoch": 0.9023265562775099,
+      "grad_norm": 0.012806370854377747,
+      "learning_rate": 2.4924866191609387e-07,
+      "loss": 0.0514,
+      "step": 4305
+    },
+    {
+      "epoch": 0.9025361559421505,
+      "grad_norm": 0.013476556167006493,
+      "learning_rate": 2.4819126930647976e-07,
+      "loss": 0.0545,
+      "step": 4306
+    },
+    {
+      "epoch": 0.902745755606791,
+      "grad_norm": 0.019602788612246513,
+      "learning_rate": 2.471360672797019e-07,
+      "loss": 0.0532,
+      "step": 4307
+    },
+    {
+      "epoch": 0.9029553552714316,
+      "grad_norm": 0.021197538822889328,
+      "learning_rate": 2.460830563222111e-07,
+      "loss": 0.0515,
+      "step": 4308
+    },
+    {
+      "epoch": 0.9031649549360721,
+      "grad_norm": 0.014859777875244617,
+      "learning_rate": 2.45032236919443e-07,
+      "loss": 0.0555,
+      "step": 4309
+    },
+    {
+      "epoch": 0.9033745546007126,
+      "grad_norm": 0.010106687434017658,
+      "learning_rate": 2.439836095558262e-07,
+      "loss": 0.053,
+      "step": 4310
+    },
+    {
+      "epoch": 0.9035841542653532,
+      "grad_norm": 0.016903823241591454,
+      "learning_rate": 2.429371747147785e-07,
+      "loss": 0.0536,
+      "step": 4311
+    },
+    {
+      "epoch": 0.9037937539299937,
+      "grad_norm": 0.018577901646494865,
+      "learning_rate": 2.41892932878704e-07,
+      "loss": 0.0544,
+      "step": 4312
+    },
+    {
+      "epoch": 0.9040033535946342,
+      "grad_norm": 0.01648200862109661,
+      "learning_rate": 2.408508845289992e-07,
+      "loss": 0.054,
+      "step": 4313
+    },
+    {
+      "epoch": 0.9042129532592748,
+      "grad_norm": 0.012114167213439941,
+      "learning_rate": 2.3981103014604765e-07,
+      "loss": 0.0536,
+      "step": 4314
+    },
+    {
+      "epoch": 0.9044225529239153,
+      "grad_norm": 0.012820238247513771,
+      "learning_rate": 2.387733702092232e-07,
+      "loss": 0.0555,
+      "step": 4315
+    },
+    {
+      "epoch": 0.9046321525885559,
+      "grad_norm": 0.018219808116555214,
+      "learning_rate": 2.3773790519688644e-07,
+      "loss": 0.0574,
+      "step": 4316
+    },
+    {
+      "epoch": 0.9048417522531964,
+      "grad_norm": 0.022166509181261063,
+      "learning_rate": 2.3670463558638556e-07,
+      "loss": 0.0546,
+      "step": 4317
+    },
+    {
+      "epoch": 0.9050513519178369,
+      "grad_norm": 0.013082697987556458,
+      "learning_rate": 2.3567356185405844e-07,
+      "loss": 0.0553,
+      "step": 4318
+    },
+    {
+      "epoch": 0.9052609515824774,
+      "grad_norm": 0.0153275141492486,
+      "learning_rate": 2.3464468447522925e-07,
+      "loss": 0.0527,
+      "step": 4319
+    },
+    {
+      "epoch": 0.905470551247118,
+      "grad_norm": 0.019890129566192627,
+      "learning_rate": 2.3361800392421086e-07,
+      "loss": 0.0537,
+      "step": 4320
+    },
+    {
+      "epoch": 0.9056801509117586,
+      "grad_norm": 0.020730622112751007,
+      "learning_rate": 2.3259352067430298e-07,
+      "loss": 0.0548,
+      "step": 4321
+    },
+    {
+      "epoch": 0.9058897505763991,
+      "grad_norm": 0.019399119541049004,
+      "learning_rate": 2.3157123519779168e-07,
+      "loss": 0.0541,
+      "step": 4322
+    },
+    {
+      "epoch": 0.9060993502410396,
+      "grad_norm": 0.02034193091094494,
+      "learning_rate": 2.3055114796595e-07,
+      "loss": 0.0571,
+      "step": 4323
+    },
+    {
+      "epoch": 0.9063089499056801,
+      "grad_norm": 0.02012833207845688,
+      "learning_rate": 2.2953325944903848e-07,
+      "loss": 0.0539,
+      "step": 4324
+    },
+    {
+      "epoch": 0.9065185495703207,
+      "grad_norm": 0.020468810573220253,
+      "learning_rate": 2.2851757011630393e-07,
+      "loss": 0.057,
+      "step": 4325
+    },
+    {
+      "epoch": 0.9067281492349613,
+      "grad_norm": 0.0244952030479908,
+      "learning_rate": 2.2750408043597794e-07,
+      "loss": 0.0518,
+      "step": 4326
+    },
+    {
+      "epoch": 0.9069377488996018,
+      "grad_norm": 0.023987367749214172,
+      "learning_rate": 2.264927908752801e-07,
+      "loss": 0.0568,
+      "step": 4327
+    },
+    {
+      "epoch": 0.9071473485642423,
+      "grad_norm": 0.017585035413503647,
+      "learning_rate": 2.2548370190041414e-07,
+      "loss": 0.0534,
+      "step": 4328
+    },
+    {
+      "epoch": 0.9073569482288828,
+      "grad_norm": 0.017734866589307785,
+      "learning_rate": 2.2447681397656908e-07,
+      "loss": 0.0527,
+      "step": 4329
+    },
+    {
+      "epoch": 0.9075665478935234,
+      "grad_norm": 0.023448040708899498,
+      "learning_rate": 2.234721275679208e-07,
+      "loss": 0.0557,
+      "step": 4330
+    },
+    {
+      "epoch": 0.9077761475581639,
+      "grad_norm": 0.02103973738849163,
+      "learning_rate": 2.2246964313763053e-07,
+      "loss": 0.055,
+      "step": 4331
+    },
+    {
+      "epoch": 0.9079857472228045,
+      "grad_norm": 0.019696585834026337,
+      "learning_rate": 2.2146936114784134e-07,
+      "loss": 0.0544,
+      "step": 4332
+    },
+    {
+      "epoch": 0.908195346887445,
+      "grad_norm": 0.016549983993172646,
+      "learning_rate": 2.2047128205968494e-07,
+      "loss": 0.0539,
+      "step": 4333
+    },
+    {
+      "epoch": 0.9084049465520855,
+      "grad_norm": 0.02317041903734207,
+      "learning_rate": 2.1947540633327437e-07,
+      "loss": 0.0528,
+      "step": 4334
+    },
+    {
+      "epoch": 0.9086145462167261,
+      "grad_norm": 0.020171795040369034,
+      "learning_rate": 2.184817344277085e-07,
+      "loss": 0.0543,
+      "step": 4335
+    },
+    {
+      "epoch": 0.9088241458813666,
+      "grad_norm": 0.019286734983325005,
+      "learning_rate": 2.1749026680106922e-07,
+      "loss": 0.0522,
+      "step": 4336
+    },
+    {
+      "epoch": 0.9090337455460071,
+      "grad_norm": 0.011374424211680889,
+      "learning_rate": 2.1650100391042373e-07,
+      "loss": 0.0576,
+      "step": 4337
+    },
+    {
+      "epoch": 0.9092433452106476,
+      "grad_norm": 0.022341901436448097,
+      "learning_rate": 2.1551394621182277e-07,
+      "loss": 0.0552,
+      "step": 4338
+    },
+    {
+      "epoch": 0.9094529448752882,
+      "grad_norm": 0.026063280180096626,
+      "learning_rate": 2.1452909416029844e-07,
+      "loss": 0.0539,
+      "step": 4339
+    },
+    {
+      "epoch": 0.9096625445399288,
+      "grad_norm": 0.01974085345864296,
+      "learning_rate": 2.1354644820986703e-07,
+      "loss": 0.0536,
+      "step": 4340
+    },
+    {
+      "epoch": 0.9098721442045693,
+      "grad_norm": 0.011266632936894894,
+      "learning_rate": 2.1256600881352951e-07,
+      "loss": 0.0542,
+      "step": 4341
+    },
+    {
+      "epoch": 0.9100817438692098,
+      "grad_norm": 0.02346237748861313,
+      "learning_rate": 2.1158777642326656e-07,
+      "loss": 0.0558,
+      "step": 4342
+    },
+    {
+      "epoch": 0.9102913435338503,
+      "grad_norm": 0.020998260006308556,
+      "learning_rate": 2.1061175149004464e-07,
+      "loss": 0.0536,
+      "step": 4343
+    },
+    {
+      "epoch": 0.9105009431984908,
+      "grad_norm": 0.01797637529671192,
+      "learning_rate": 2.0963793446381053e-07,
+      "loss": 0.0532,
+      "step": 4344
+    },
+    {
+      "epoch": 0.9107105428631315,
+      "grad_norm": 0.016491252928972244,
+      "learning_rate": 2.08666325793494e-07,
+      "loss": 0.0522,
+      "step": 4345
+    },
+    {
+      "epoch": 0.910920142527772,
+      "grad_norm": 0.011900043115019798,
+      "learning_rate": 2.076969259270051e-07,
+      "loss": 0.0581,
+      "step": 4346
+    },
+    {
+      "epoch": 0.9111297421924125,
+      "grad_norm": 0.014170379377901554,
+      "learning_rate": 2.0672973531123796e-07,
+      "loss": 0.0534,
+      "step": 4347
+    },
+    {
+      "epoch": 0.911339341857053,
+      "grad_norm": 0.01760711334645748,
+      "learning_rate": 2.0576475439206767e-07,
+      "loss": 0.0555,
+      "step": 4348
+    },
+    {
+      "epoch": 0.9115489415216935,
+      "grad_norm": 0.017065905034542084,
+      "learning_rate": 2.0480198361435e-07,
+      "loss": 0.055,
+      "step": 4349
+    },
+    {
+      "epoch": 0.9117585411863341,
+      "grad_norm": 0.012907272204756737,
+      "learning_rate": 2.038414234219216e-07,
+      "loss": 0.0532,
+      "step": 4350
+    },
+    {
+      "epoch": 0.9119681408509747,
+      "grad_norm": 0.010160834528505802,
+      "learning_rate": 2.0288307425760046e-07,
+      "loss": 0.0543,
+      "step": 4351
+    },
+    {
+      "epoch": 0.9121777405156152,
+      "grad_norm": 0.013258119113743305,
+      "learning_rate": 2.0192693656318597e-07,
+      "loss": 0.0518,
+      "step": 4352
+    },
+    {
+      "epoch": 0.9123873401802557,
+      "grad_norm": 0.0151562774553895,
+      "learning_rate": 2.0097301077945607e-07,
+      "loss": 0.0527,
+      "step": 4353
+    },
+    {
+      "epoch": 0.9125969398448962,
+      "grad_norm": 0.016773242503404617,
+      "learning_rate": 2.0002129734617292e-07,
+      "loss": 0.0534,
+      "step": 4354
+    },
+    {
+      "epoch": 0.9128065395095368,
+      "grad_norm": 0.012901091016829014,
+      "learning_rate": 1.9907179670207387e-07,
+      "loss": 0.0518,
+      "step": 4355
+    },
+    {
+      "epoch": 0.9130161391741773,
+      "grad_norm": 0.010573726147413254,
+      "learning_rate": 1.9812450928487936e-07,
+      "loss": 0.0518,
+      "step": 4356
+    },
+    {
+      "epoch": 0.9132257388388179,
+      "grad_norm": 0.012841584160923958,
+      "learning_rate": 1.9717943553128893e-07,
+      "loss": 0.0556,
+      "step": 4357
+    },
+    {
+      "epoch": 0.9134353385034584,
+      "grad_norm": 0.014390083961188793,
+      "learning_rate": 1.962365758769802e-07,
+      "loss": 0.0526,
+      "step": 4358
+    },
+    {
+      "epoch": 0.9136449381680989,
+      "grad_norm": 0.014265509322285652,
+      "learning_rate": 1.9529593075661267e-07,
+      "loss": 0.0531,
+      "step": 4359
+    },
+    {
+      "epoch": 0.9138545378327395,
+      "grad_norm": 0.013005812652409077,
+      "learning_rate": 1.943575006038234e-07,
+      "loss": 0.0535,
+      "step": 4360
+    },
+    {
+      "epoch": 0.91406413749738,
+      "grad_norm": 0.009814724326133728,
+      "learning_rate": 1.9342128585122798e-07,
+      "loss": 0.0574,
+      "step": 4361
+    },
+    {
+      "epoch": 0.9142737371620205,
+      "grad_norm": 0.011820555664598942,
+      "learning_rate": 1.9248728693042117e-07,
+      "loss": 0.0519,
+      "step": 4362
+    },
+    {
+      "epoch": 0.914483336826661,
+      "grad_norm": 0.014412821270525455,
+      "learning_rate": 1.9155550427197577e-07,
+      "loss": 0.0579,
+      "step": 4363
+    },
+    {
+      "epoch": 0.9146929364913016,
+      "grad_norm": 0.013178630731999874,
+      "learning_rate": 1.906259383054454e-07,
+      "loss": 0.0536,
+      "step": 4364
+    },
+    {
+      "epoch": 0.9149025361559422,
+      "grad_norm": 0.01012821588665247,
+      "learning_rate": 1.8969858945935783e-07,
+      "loss": 0.0566,
+      "step": 4365
+    },
+    {
+      "epoch": 0.9151121358205827,
+      "grad_norm": 0.009884363040328026,
+      "learning_rate": 1.8877345816122162e-07,
+      "loss": 0.0551,
+      "step": 4366
+    },
+    {
+      "epoch": 0.9153217354852232,
+      "grad_norm": 0.011177174746990204,
+      "learning_rate": 1.8785054483752174e-07,
+      "loss": 0.0518,
+      "step": 4367
+    },
+    {
+      "epoch": 0.9155313351498637,
+      "grad_norm": 0.012874050997197628,
+      "learning_rate": 1.8692984991372065e-07,
+      "loss": 0.0554,
+      "step": 4368
+    },
+    {
+      "epoch": 0.9157409348145042,
+      "grad_norm": 0.012516334652900696,
+      "learning_rate": 1.8601137381425938e-07,
+      "loss": 0.0551,
+      "step": 4369
+    },
+    {
+      "epoch": 0.9159505344791449,
+      "grad_norm": 0.010518589057028294,
+      "learning_rate": 1.8509511696255421e-07,
+      "loss": 0.0543,
+      "step": 4370
+    },
+    {
+      "epoch": 0.9161601341437854,
+      "grad_norm": 0.00973634421825409,
+      "learning_rate": 1.841810797810012e-07,
+      "loss": 0.0542,
+      "step": 4371
+    },
+    {
+      "epoch": 0.9163697338084259,
+      "grad_norm": 0.01056087389588356,
+      "learning_rate": 1.8326926269096935e-07,
+      "loss": 0.0525,
+      "step": 4372
+    },
+    {
+      "epoch": 0.9165793334730664,
+      "grad_norm": 0.012748325243592262,
+      "learning_rate": 1.8235966611280687e-07,
+      "loss": 0.0534,
+      "step": 4373
+    },
+    {
+      "epoch": 0.9167889331377069,
+      "grad_norm": 0.010445375926792622,
+      "learning_rate": 1.8145229046583778e-07,
+      "loss": 0.0553,
+      "step": 4374
+    },
+    {
+      "epoch": 0.9169985328023476,
+      "grad_norm": 0.00820910930633545,
+      "learning_rate": 1.805471361683614e-07,
+      "loss": 0.0551,
+      "step": 4375
+    },
+    {
+      "epoch": 0.9172081324669881,
+      "grad_norm": 0.008532201871275902,
+      "learning_rate": 1.7964420363765444e-07,
+      "loss": 0.0562,
+      "step": 4376
+    },
+    {
+      "epoch": 0.9174177321316286,
+      "grad_norm": 0.01010076142847538,
+      "learning_rate": 1.78743493289969e-07,
+      "loss": 0.0537,
+      "step": 4377
+    },
+    {
+      "epoch": 0.9176273317962691,
+      "grad_norm": 0.011571811512112617,
+      "learning_rate": 1.778450055405312e-07,
+      "loss": 0.0552,
+      "step": 4378
+    },
+    {
+      "epoch": 0.9178369314609096,
+      "grad_norm": 0.01056272629648447,
+      "learning_rate": 1.7694874080354362e-07,
+      "loss": 0.056,
+      "step": 4379
+    },
+    {
+      "epoch": 0.9180465311255502,
+      "grad_norm": 0.008779754862189293,
+      "learning_rate": 1.760546994921858e-07,
+      "loss": 0.0541,
+      "step": 4380
+    },
+    {
+      "epoch": 0.9182561307901907,
+      "grad_norm": 0.008502528071403503,
+      "learning_rate": 1.7516288201860853e-07,
+      "loss": 0.0514,
+      "step": 4381
+    },
+    {
+      "epoch": 0.9184657304548313,
+      "grad_norm": 0.009822116233408451,
+      "learning_rate": 1.742732887939408e-07,
+      "loss": 0.0528,
+      "step": 4382
+    },
+    {
+      "epoch": 0.9186753301194718,
+      "grad_norm": 0.012791884131729603,
+      "learning_rate": 1.733859202282845e-07,
+      "loss": 0.0535,
+      "step": 4383
+    },
+    {
+      "epoch": 0.9188849297841123,
+      "grad_norm": 0.010147576220333576,
+      "learning_rate": 1.7250077673071685e-07,
+      "loss": 0.0537,
+      "step": 4384
+    },
+    {
+      "epoch": 0.9190945294487529,
+      "grad_norm": 0.009435161016881466,
+      "learning_rate": 1.716178587092876e-07,
+      "loss": 0.0526,
+      "step": 4385
+    },
+    {
+      "epoch": 0.9193041291133934,
+      "grad_norm": 0.007458524778485298,
+      "learning_rate": 1.7073716657102278e-07,
+      "loss": 0.0523,
+      "step": 4386
+    },
+    {
+      "epoch": 0.919513728778034,
+      "grad_norm": 0.008520993404090405,
+      "learning_rate": 1.6985870072192156e-07,
+      "loss": 0.0538,
+      "step": 4387
+    },
+    {
+      "epoch": 0.9197233284426745,
+      "grad_norm": 0.01144067570567131,
+      "learning_rate": 1.6898246156695552e-07,
+      "loss": 0.0559,
+      "step": 4388
+    },
+    {
+      "epoch": 0.919932928107315,
+      "grad_norm": 0.01067439466714859,
+      "learning_rate": 1.6810844951007099e-07,
+      "loss": 0.0563,
+      "step": 4389
+    },
+    {
+      "epoch": 0.9201425277719556,
+      "grad_norm": 0.012738638557493687,
+      "learning_rate": 1.6723666495418844e-07,
+      "loss": 0.0553,
+      "step": 4390
+    },
+    {
+      "epoch": 0.9203521274365961,
+      "grad_norm": 0.009737935848534107,
+      "learning_rate": 1.6636710830119863e-07,
+      "loss": 0.054,
+      "step": 4391
+    },
+    {
+      "epoch": 0.9205617271012366,
+      "grad_norm": 0.0069859870709478855,
+      "learning_rate": 1.6549977995196809e-07,
+      "loss": 0.0557,
+      "step": 4392
+    },
+    {
+      "epoch": 0.9207713267658771,
+      "grad_norm": 0.011632826179265976,
+      "learning_rate": 1.6463468030633478e-07,
+      "loss": 0.0526,
+      "step": 4393
+    },
+    {
+      "epoch": 0.9209809264305178,
+      "grad_norm": 0.011854654178023338,
+      "learning_rate": 1.6377180976310968e-07,
+      "loss": 0.0561,
+      "step": 4394
+    },
+    {
+      "epoch": 0.9211905260951583,
+      "grad_norm": 0.010901962406933308,
+      "learning_rate": 1.6291116872007573e-07,
+      "loss": 0.0511,
+      "step": 4395
+    },
+    {
+      "epoch": 0.9214001257597988,
+      "grad_norm": 0.009396993555128574,
+      "learning_rate": 1.6205275757398774e-07,
+      "loss": 0.0547,
+      "step": 4396
+    },
+    {
+      "epoch": 0.9216097254244393,
+      "grad_norm": 0.00782528892159462,
+      "learning_rate": 1.611965767205742e-07,
+      "loss": 0.0529,
+      "step": 4397
+    },
+    {
+      "epoch": 0.9218193250890798,
+      "grad_norm": 0.00804727990180254,
+      "learning_rate": 1.6034262655453269e-07,
+      "loss": 0.0538,
+      "step": 4398
+    },
+    {
+      "epoch": 0.9220289247537204,
+      "grad_norm": 0.011399338953197002,
+      "learning_rate": 1.594909074695361e-07,
+      "loss": 0.0523,
+      "step": 4399
+    },
+    {
+      "epoch": 0.922238524418361,
+      "grad_norm": 0.013382563367486,
+      "learning_rate": 1.586414198582259e-07,
+      "loss": 0.0557,
+      "step": 4400
+    },
+    {
+      "epoch": 0.9224481240830015,
+      "grad_norm": 0.010052556172013283,
+      "learning_rate": 1.5779416411221437e-07,
+      "loss": 0.0554,
+      "step": 4401
+    },
+    {
+      "epoch": 0.922657723747642,
+      "grad_norm": 0.010469266213476658,
+      "learning_rate": 1.5694914062208799e-07,
+      "loss": 0.0544,
+      "step": 4402
+    },
+    {
+      "epoch": 0.9228673234122825,
+      "grad_norm": 0.011628585867583752,
+      "learning_rate": 1.561063497774018e-07,
+      "loss": 0.0535,
+      "step": 4403
+    },
+    {
+      "epoch": 0.9230769230769231,
+      "grad_norm": 0.016579635441303253,
+      "learning_rate": 1.552657919666817e-07,
+      "loss": 0.0547,
+      "step": 4404
+    },
+    {
+      "epoch": 0.9232865227415636,
+      "grad_norm": 0.01323653757572174,
+      "learning_rate": 1.54427467577426e-07,
+      "loss": 0.0525,
+      "step": 4405
+    },
+    {
+      "epoch": 0.9234961224062042,
+      "grad_norm": 0.01618046499788761,
+      "learning_rate": 1.535913769961006e-07,
+      "loss": 0.0535,
+      "step": 4406
+    },
+    {
+      "epoch": 0.9237057220708447,
+      "grad_norm": 0.012092587538063526,
+      "learning_rate": 1.527575206081444e-07,
+      "loss": 0.0524,
+      "step": 4407
+    },
+    {
+      "epoch": 0.9239153217354852,
+      "grad_norm": 0.014828270301222801,
+      "learning_rate": 1.5192589879796383e-07,
+      "loss": 0.0533,
+      "step": 4408
+    },
+    {
+      "epoch": 0.9241249214001258,
+      "grad_norm": 0.008504010736942291,
+      "learning_rate": 1.510965119489366e-07,
+      "loss": 0.0541,
+      "step": 4409
+    },
+    {
+      "epoch": 0.9243345210647663,
+      "grad_norm": 0.010867651551961899,
+      "learning_rate": 1.5026936044341078e-07,
+      "loss": 0.0546,
+      "step": 4410
+    },
+    {
+      "epoch": 0.9245441207294068,
+      "grad_norm": 0.015342487022280693,
+      "learning_rate": 1.4944444466270248e-07,
+      "loss": 0.0543,
+      "step": 4411
+    },
+    {
+      "epoch": 0.9247537203940474,
+      "grad_norm": 0.01415248867124319,
+      "learning_rate": 1.486217649870969e-07,
+      "loss": 0.0539,
+      "step": 4412
+    },
+    {
+      "epoch": 0.9249633200586879,
+      "grad_norm": 0.009957171976566315,
+      "learning_rate": 1.478013217958507e-07,
+      "loss": 0.0545,
+      "step": 4413
+    },
+    {
+      "epoch": 0.9251729197233285,
+      "grad_norm": 0.00818433053791523,
+      "learning_rate": 1.4698311546718635e-07,
+      "loss": 0.0556,
+      "step": 4414
+    },
+    {
+      "epoch": 0.925382519387969,
+      "grad_norm": 0.010151896625757217,
+      "learning_rate": 1.4616714637829822e-07,
+      "loss": 0.0533,
+      "step": 4415
+    },
+    {
+      "epoch": 0.9255921190526095,
+      "grad_norm": 0.012620055116713047,
+      "learning_rate": 1.453534149053476e-07,
+      "loss": 0.0556,
+      "step": 4416
+    },
+    {
+      "epoch": 0.92580171871725,
+      "grad_norm": 0.011959872208535671,
+      "learning_rate": 1.4454192142346446e-07,
+      "loss": 0.0549,
+      "step": 4417
+    },
+    {
+      "epoch": 0.9260113183818905,
+      "grad_norm": 0.010501948185265064,
+      "learning_rate": 1.437326663067462e-07,
+      "loss": 0.0526,
+      "step": 4418
+    },
+    {
+      "epoch": 0.9262209180465312,
+      "grad_norm": 0.009827865287661552,
+      "learning_rate": 1.429256499282605e-07,
+      "loss": 0.0529,
+      "step": 4419
+    },
+    {
+      "epoch": 0.9264305177111717,
+      "grad_norm": 0.009340137243270874,
+      "learning_rate": 1.421208726600415e-07,
+      "loss": 0.0524,
+      "step": 4420
+    },
+    {
+      "epoch": 0.9266401173758122,
+      "grad_norm": 0.010111922398209572,
+      "learning_rate": 1.4131833487309122e-07,
+      "loss": 0.0547,
+      "step": 4421
+    },
+    {
+      "epoch": 0.9268497170404527,
+      "grad_norm": 0.01139815803617239,
+      "learning_rate": 1.4051803693737876e-07,
+      "loss": 0.0537,
+      "step": 4422
+    },
+    {
+      "epoch": 0.9270593167050932,
+      "grad_norm": 0.009999927133321762,
+      "learning_rate": 1.3971997922184234e-07,
+      "loss": 0.0532,
+      "step": 4423
+    },
+    {
+      "epoch": 0.9272689163697339,
+      "grad_norm": 0.010446673259139061,
+      "learning_rate": 1.3892416209438542e-07,
+      "loss": 0.0524,
+      "step": 4424
+    },
+    {
+      "epoch": 0.9274785160343744,
+      "grad_norm": 0.012159881182014942,
+      "learning_rate": 1.381305859218801e-07,
+      "loss": 0.051,
+      "step": 4425
+    },
+    {
+      "epoch": 0.9276881156990149,
+      "grad_norm": 0.013391540385782719,
+      "learning_rate": 1.3733925107016542e-07,
+      "loss": 0.0519,
+      "step": 4426
+    },
+    {
+      "epoch": 0.9278977153636554,
+      "grad_norm": 0.006831051781773567,
+      "learning_rate": 1.3655015790404568e-07,
+      "loss": 0.0536,
+      "step": 4427
+    },
+    {
+      "epoch": 0.9281073150282959,
+      "grad_norm": 0.007939697243273258,
+      "learning_rate": 1.3576330678729266e-07,
+      "loss": 0.0535,
+      "step": 4428
+    },
+    {
+      "epoch": 0.9283169146929365,
+      "grad_norm": 0.015715105459094048,
+      "learning_rate": 1.3497869808264453e-07,
+      "loss": 0.054,
+      "step": 4429
+    },
+    {
+      "epoch": 0.928526514357577,
+      "grad_norm": 0.008145746774971485,
+      "learning_rate": 1.341963321518064e-07,
+      "loss": 0.057,
+      "step": 4430
+    },
+    {
+      "epoch": 0.9287361140222176,
+      "grad_norm": 0.0074055916629731655,
+      "learning_rate": 1.3341620935544864e-07,
+      "loss": 0.0559,
+      "step": 4431
+    },
+    {
+      "epoch": 0.9289457136868581,
+      "grad_norm": 0.00775395892560482,
+      "learning_rate": 1.3263833005320747e-07,
+      "loss": 0.0521,
+      "step": 4432
+    },
+    {
+      "epoch": 0.9291553133514986,
+      "grad_norm": 0.008558698929846287,
+      "learning_rate": 1.3186269460368496e-07,
+      "loss": 0.0523,
+      "step": 4433
+    },
+    {
+      "epoch": 0.9293649130161392,
+      "grad_norm": 0.009375239722430706,
+      "learning_rate": 1.3108930336444893e-07,
+      "loss": 0.0522,
+      "step": 4434
+    },
+    {
+      "epoch": 0.9295745126807797,
+      "grad_norm": 0.007077595219016075,
+      "learning_rate": 1.303181566920325e-07,
+      "loss": 0.0538,
+      "step": 4435
+    },
+    {
+      "epoch": 0.9297841123454202,
+      "grad_norm": 0.0054700132459402084,
+      "learning_rate": 1.2954925494193472e-07,
+      "loss": 0.0545,
+      "step": 4436
+    },
+    {
+      "epoch": 0.9299937120100608,
+      "grad_norm": 0.008495796471834183,
+      "learning_rate": 1.2878259846861862e-07,
+      "loss": 0.0587,
+      "step": 4437
+    },
+    {
+      "epoch": 0.9302033116747013,
+      "grad_norm": 0.008338769897818565,
+      "learning_rate": 1.2801818762551265e-07,
+      "loss": 0.051,
+      "step": 4438
+    },
+    {
+      "epoch": 0.9304129113393419,
+      "grad_norm": 0.009145451709628105,
+      "learning_rate": 1.2725602276501047e-07,
+      "loss": 0.0544,
+      "step": 4439
+    },
+    {
+      "epoch": 0.9306225110039824,
+      "grad_norm": 0.00860004685819149,
+      "learning_rate": 1.2649610423846937e-07,
+      "loss": 0.0549,
+      "step": 4440
+    },
+    {
+      "epoch": 0.9308321106686229,
+      "grad_norm": 0.006629549898207188,
+      "learning_rate": 1.2573843239621185e-07,
+      "loss": 0.0548,
+      "step": 4441
+    },
+    {
+      "epoch": 0.9310417103332634,
+      "grad_norm": 0.007576174568384886,
+      "learning_rate": 1.2498300758752413e-07,
+      "loss": 0.0517,
+      "step": 4442
+    },
+    {
+      "epoch": 0.931251309997904,
+      "grad_norm": 0.007479529827833176,
+      "learning_rate": 1.2422983016065816e-07,
+      "loss": 0.0575,
+      "step": 4443
+    },
+    {
+      "epoch": 0.9314609096625446,
+      "grad_norm": 0.006873736623674631,
+      "learning_rate": 1.2347890046282785e-07,
+      "loss": 0.0543,
+      "step": 4444
+    },
+    {
+      "epoch": 0.9316705093271851,
+      "grad_norm": 0.008335386402904987,
+      "learning_rate": 1.2273021884021074e-07,
+      "loss": 0.0526,
+      "step": 4445
+    },
+    {
+      "epoch": 0.9318801089918256,
+      "grad_norm": 0.007241726852953434,
+      "learning_rate": 1.2198378563795023e-07,
+      "loss": 0.0534,
+      "step": 4446
+    },
+    {
+      "epoch": 0.9320897086564661,
+      "grad_norm": 0.008679443039000034,
+      "learning_rate": 1.2123960120015045e-07,
+      "loss": 0.0542,
+      "step": 4447
+    },
+    {
+      "epoch": 0.9322993083211066,
+      "grad_norm": 0.0065467506647109985,
+      "learning_rate": 1.2049766586988142e-07,
+      "loss": 0.0506,
+      "step": 4448
+    },
+    {
+      "epoch": 0.9325089079857473,
+      "grad_norm": 0.0077818529680371284,
+      "learning_rate": 1.1975797998917514e-07,
+      "loss": 0.0526,
+      "step": 4449
+    },
+    {
+      "epoch": 0.9327185076503878,
+      "grad_norm": 0.008923080749809742,
+      "learning_rate": 1.1902054389902662e-07,
+      "loss": 0.0511,
+      "step": 4450
+    },
+    {
+      "epoch": 0.9329281073150283,
+      "grad_norm": 0.009406541474163532,
+      "learning_rate": 1.1828535793939277e-07,
+      "loss": 0.0507,
+      "step": 4451
+    },
+    {
+      "epoch": 0.9331377069796688,
+      "grad_norm": 0.008926275186240673,
+      "learning_rate": 1.1755242244919528e-07,
+      "loss": 0.0553,
+      "step": 4452
+    },
+    {
+      "epoch": 0.9333473066443093,
+      "grad_norm": 0.007040916942059994,
+      "learning_rate": 1.1682173776631722e-07,
+      "loss": 0.054,
+      "step": 4453
+    },
+    {
+      "epoch": 0.9335569063089499,
+      "grad_norm": 0.009015001356601715,
+      "learning_rate": 1.16093304227603e-07,
+      "loss": 0.0555,
+      "step": 4454
+    },
+    {
+      "epoch": 0.9337665059735905,
+      "grad_norm": 0.010234599001705647,
+      "learning_rate": 1.1536712216886181e-07,
+      "loss": 0.0542,
+      "step": 4455
+    },
+    {
+      "epoch": 0.933976105638231,
+      "grad_norm": 0.008125402964651585,
+      "learning_rate": 1.1464319192486251e-07,
+      "loss": 0.0515,
+      "step": 4456
+    },
+    {
+      "epoch": 0.9341857053028715,
+      "grad_norm": 0.009704657830297947,
+      "learning_rate": 1.1392151382933647e-07,
+      "loss": 0.0555,
+      "step": 4457
+    },
+    {
+      "epoch": 0.934395304967512,
+      "grad_norm": 0.009614041075110435,
+      "learning_rate": 1.1320208821497758e-07,
+      "loss": 0.0529,
+      "step": 4458
+    },
+    {
+      "epoch": 0.9346049046321526,
+      "grad_norm": 0.007132851053029299,
+      "learning_rate": 1.1248491541344164e-07,
+      "loss": 0.0539,
+      "step": 4459
+    },
+    {
+      "epoch": 0.9348145042967931,
+      "grad_norm": 0.008845999836921692,
+      "learning_rate": 1.117699957553442e-07,
+      "loss": 0.0558,
+      "step": 4460
+    },
+    {
+      "epoch": 0.9350241039614337,
+      "grad_norm": 0.009133417159318924,
+      "learning_rate": 1.1105732957026272e-07,
+      "loss": 0.0528,
+      "step": 4461
+    },
+    {
+      "epoch": 0.9352337036260742,
+      "grad_norm": 0.008758967742323875,
+      "learning_rate": 1.1034691718673774e-07,
+      "loss": 0.0551,
+      "step": 4462
+    },
+    {
+      "epoch": 0.9354433032907148,
+      "grad_norm": 0.009887007996439934,
+      "learning_rate": 1.0963875893226728e-07,
+      "loss": 0.0545,
+      "step": 4463
+    },
+    {
+      "epoch": 0.9356529029553553,
+      "grad_norm": 0.00951511599123478,
+      "learning_rate": 1.0893285513331353e-07,
+      "loss": 0.0566,
+      "step": 4464
+    },
+    {
+      "epoch": 0.9358625026199958,
+      "grad_norm": 0.0072890762239694595,
+      "learning_rate": 1.0822920611529786e-07,
+      "loss": 0.0538,
+      "step": 4465
+    },
+    {
+      "epoch": 0.9360721022846363,
+      "grad_norm": 0.008058588020503521,
+      "learning_rate": 1.0752781220260245e-07,
+      "loss": 0.0518,
+      "step": 4466
+    },
+    {
+      "epoch": 0.9362817019492768,
+      "grad_norm": 0.008425997570157051,
+      "learning_rate": 1.0682867371856864e-07,
+      "loss": 0.0534,
+      "step": 4467
+    },
+    {
+      "epoch": 0.9364913016139175,
+      "grad_norm": 0.008281690068542957,
+      "learning_rate": 1.0613179098549975e-07,
+      "loss": 0.055,
+      "step": 4468
+    },
+    {
+      "epoch": 0.936700901278558,
+      "grad_norm": 0.009938735514879227,
+      "learning_rate": 1.0543716432465933e-07,
+      "loss": 0.0524,
+      "step": 4469
+    },
+    {
+      "epoch": 0.9369105009431985,
+      "grad_norm": 0.009986302815377712,
+      "learning_rate": 1.0474479405626847e-07,
+      "loss": 0.0527,
+      "step": 4470
+    },
+    {
+      "epoch": 0.937120100607839,
+      "grad_norm": 0.007245698943734169,
+      "learning_rate": 1.0405468049951184e-07,
+      "loss": 0.0539,
+      "step": 4471
+    },
+    {
+      "epoch": 0.9373297002724795,
+      "grad_norm": 0.009867161512374878,
+      "learning_rate": 1.0336682397252995e-07,
+      "loss": 0.0551,
+      "step": 4472
+    },
+    {
+      "epoch": 0.9375392999371202,
+      "grad_norm": 0.008164730854332447,
+      "learning_rate": 1.0268122479242526e-07,
+      "loss": 0.0547,
+      "step": 4473
+    },
+    {
+      "epoch": 0.9377488996017607,
+      "grad_norm": 0.0104014091193676,
+      "learning_rate": 1.0199788327525828e-07,
+      "loss": 0.0508,
+      "step": 4474
+    },
+    {
+      "epoch": 0.9379584992664012,
+      "grad_norm": 0.009538762271404266,
+      "learning_rate": 1.013167997360498e-07,
+      "loss": 0.0505,
+      "step": 4475
+    },
+    {
+      "epoch": 0.9381680989310417,
+      "grad_norm": 0.008720876649022102,
+      "learning_rate": 1.0063797448877977e-07,
+      "loss": 0.0534,
+      "step": 4476
+    },
+    {
+      "epoch": 0.9383776985956822,
+      "grad_norm": 0.010985384695231915,
+      "learning_rate": 9.996140784638564e-08,
+      "loss": 0.0572,
+      "step": 4477
+    },
+    {
+      "epoch": 0.9385872982603228,
+      "grad_norm": 0.008751952089369297,
+      "learning_rate": 9.928710012076404e-08,
+      "loss": 0.0542,
+      "step": 4478
+    },
+    {
+      "epoch": 0.9387968979249633,
+      "grad_norm": 0.007031100802123547,
+      "learning_rate": 9.861505162277186e-08,
+      "loss": 0.0552,
+      "step": 4479
+    },
+    {
+      "epoch": 0.9390064975896039,
+      "grad_norm": 0.008988458663225174,
+      "learning_rate": 9.794526266222237e-08,
+      "loss": 0.0531,
+      "step": 4480
+    },
+    {
+      "epoch": 0.9392160972542444,
+      "grad_norm": 0.007348395884037018,
+      "learning_rate": 9.727773354788861e-08,
+      "loss": 0.0547,
+      "step": 4481
+    },
+    {
+      "epoch": 0.9394256969188849,
+      "grad_norm": 0.009616075083613396,
+      "learning_rate": 9.661246458750162e-08,
+      "loss": 0.053,
+      "step": 4482
+    },
+    {
+      "epoch": 0.9396352965835255,
+      "grad_norm": 0.00715535506606102,
+      "learning_rate": 9.594945608774997e-08,
+      "loss": 0.0563,
+      "step": 4483
+    },
+    {
+      "epoch": 0.939844896248166,
+      "grad_norm": 0.008326810784637928,
+      "learning_rate": 9.528870835427972e-08,
+      "loss": 0.0532,
+      "step": 4484
+    },
+    {
+      "epoch": 0.9400544959128065,
+      "grad_norm": 0.009740813635289669,
+      "learning_rate": 9.463022169169666e-08,
+      "loss": 0.0514,
+      "step": 4485
+    },
+    {
+      "epoch": 0.9402640955774471,
+      "grad_norm": 0.0066096787340939045,
+      "learning_rate": 9.397399640356242e-08,
+      "loss": 0.0543,
+      "step": 4486
+    },
+    {
+      "epoch": 0.9404736952420876,
+      "grad_norm": 0.009233192540705204,
+      "learning_rate": 9.332003279239665e-08,
+      "loss": 0.0582,
+      "step": 4487
+    },
+    {
+      "epoch": 0.9406832949067282,
+      "grad_norm": 0.01031689066439867,
+      "learning_rate": 9.266833115967655e-08,
+      "loss": 0.0556,
+      "step": 4488
+    },
+    {
+      "epoch": 0.9408928945713687,
+      "grad_norm": 0.010228820145130157,
+      "learning_rate": 9.201889180583679e-08,
+      "loss": 0.0538,
+      "step": 4489
+    },
+    {
+      "epoch": 0.9411024942360092,
+      "grad_norm": 0.0081089548766613,
+      "learning_rate": 9.137171503026787e-08,
+      "loss": 0.0541,
+      "step": 4490
+    },
+    {
+      "epoch": 0.9413120939006497,
+      "grad_norm": 0.007212504744529724,
+      "learning_rate": 9.072680113131893e-08,
+      "loss": 0.0517,
+      "step": 4491
+    },
+    {
+      "epoch": 0.9415216935652903,
+      "grad_norm": 0.006384396459907293,
+      "learning_rate": 9.008415040629548e-08,
+      "loss": 0.0528,
+      "step": 4492
+    },
+    {
+      "epoch": 0.9417312932299309,
+      "grad_norm": 0.007916389964520931,
+      "learning_rate": 8.944376315145831e-08,
+      "loss": 0.0533,
+      "step": 4493
+    },
+    {
+      "epoch": 0.9419408928945714,
+      "grad_norm": 0.009434894658625126,
+      "learning_rate": 8.880563966202627e-08,
+      "loss": 0.0538,
+      "step": 4494
+    },
+    {
+      "epoch": 0.9421504925592119,
+      "grad_norm": 0.009904226288199425,
+      "learning_rate": 8.816978023217404e-08,
+      "loss": 0.0542,
+      "step": 4495
+    },
+    {
+      "epoch": 0.9423600922238524,
+      "grad_norm": 0.008942014537751675,
+      "learning_rate": 8.753618515503325e-08,
+      "loss": 0.0543,
+      "step": 4496
+    },
+    {
+      "epoch": 0.9425696918884929,
+      "grad_norm": 0.008763092570006847,
+      "learning_rate": 8.690485472269028e-08,
+      "loss": 0.0546,
+      "step": 4497
+    },
+    {
+      "epoch": 0.9427792915531336,
+      "grad_norm": 0.006246887147426605,
+      "learning_rate": 8.627578922618895e-08,
+      "loss": 0.0525,
+      "step": 4498
+    },
+    {
+      "epoch": 0.9429888912177741,
+      "grad_norm": 0.00860271230340004,
+      "learning_rate": 8.564898895552843e-08,
+      "loss": 0.0542,
+      "step": 4499
+    },
+    {
+      "epoch": 0.9431984908824146,
+      "grad_norm": 0.005993400234729052,
+      "learning_rate": 8.502445419966254e-08,
+      "loss": 0.0541,
+      "step": 4500
+    },
+    {
+      "epoch": 0.9434080905470551,
+      "grad_norm": 0.009251521900296211,
+      "learning_rate": 8.440218524650268e-08,
+      "loss": 0.0515,
+      "step": 4501
+    },
+    {
+      "epoch": 0.9436176902116956,
+      "grad_norm": 0.01032980252057314,
+      "learning_rate": 8.378218238291492e-08,
+      "loss": 0.0538,
+      "step": 4502
+    },
+    {
+      "epoch": 0.9438272898763362,
+      "grad_norm": 0.007855959236621857,
+      "learning_rate": 8.316444589471895e-08,
+      "loss": 0.0531,
+      "step": 4503
+    },
+    {
+      "epoch": 0.9440368895409768,
+      "grad_norm": 0.008385205641388893,
+      "learning_rate": 8.254897606669254e-08,
+      "loss": 0.0547,
+      "step": 4504
+    },
+    {
+      "epoch": 0.9442464892056173,
+      "grad_norm": 0.0065010362304747105,
+      "learning_rate": 8.193577318256707e-08,
+      "loss": 0.056,
+      "step": 4505
+    },
+    {
+      "epoch": 0.9444560888702578,
+      "grad_norm": 0.007750457618385553,
+      "learning_rate": 8.132483752502806e-08,
+      "loss": 0.0555,
+      "step": 4506
+    },
+    {
+      "epoch": 0.9446656885348983,
+      "grad_norm": 0.007757239043712616,
+      "learning_rate": 8.071616937571692e-08,
+      "loss": 0.0538,
+      "step": 4507
+    },
+    {
+      "epoch": 0.9448752881995389,
+      "grad_norm": 0.006676631513983011,
+      "learning_rate": 8.010976901523082e-08,
+      "loss": 0.0536,
+      "step": 4508
+    },
+    {
+      "epoch": 0.9450848878641794,
+      "grad_norm": 0.007698145695030689,
+      "learning_rate": 7.95056367231184e-08,
+      "loss": 0.0557,
+      "step": 4509
+    },
+    {
+      "epoch": 0.94529448752882,
+      "grad_norm": 0.009001096710562706,
+      "learning_rate": 7.89037727778863e-08,
+      "loss": 0.0539,
+      "step": 4510
+    },
+    {
+      "epoch": 0.9455040871934605,
+      "grad_norm": 0.008882983587682247,
+      "learning_rate": 7.830417745699204e-08,
+      "loss": 0.0516,
+      "step": 4511
+    },
+    {
+      "epoch": 0.945713686858101,
+      "grad_norm": 0.006819653324782848,
+      "learning_rate": 7.770685103685005e-08,
+      "loss": 0.0526,
+      "step": 4512
+    },
+    {
+      "epoch": 0.9459232865227416,
+      "grad_norm": 0.006686016917228699,
+      "learning_rate": 7.711179379282674e-08,
+      "loss": 0.0529,
+      "step": 4513
+    },
+    {
+      "epoch": 0.9461328861873821,
+      "grad_norm": 0.0065452903509140015,
+      "learning_rate": 7.651900599924378e-08,
+      "loss": 0.0526,
+      "step": 4514
+    },
+    {
+      "epoch": 0.9463424858520226,
+      "grad_norm": 0.006739321630448103,
+      "learning_rate": 7.592848792937701e-08,
+      "loss": 0.0535,
+      "step": 4515
+    },
+    {
+      "epoch": 0.9465520855166631,
+      "grad_norm": 0.009856310673058033,
+      "learning_rate": 7.53402398554548e-08,
+      "loss": 0.0527,
+      "step": 4516
+    },
+    {
+      "epoch": 0.9467616851813037,
+      "grad_norm": 0.00838090106844902,
+      "learning_rate": 7.475426204865854e-08,
+      "loss": 0.0539,
+      "step": 4517
+    },
+    {
+      "epoch": 0.9469712848459443,
+      "grad_norm": 0.0068512773141264915,
+      "learning_rate": 7.417055477912438e-08,
+      "loss": 0.0536,
+      "step": 4518
+    },
+    {
+      "epoch": 0.9471808845105848,
+      "grad_norm": 0.006120042875409126,
+      "learning_rate": 7.358911831594095e-08,
+      "loss": 0.0537,
+      "step": 4519
+    },
+    {
+      "epoch": 0.9473904841752253,
+      "grad_norm": 0.008014840073883533,
+      "learning_rate": 7.300995292715107e-08,
+      "loss": 0.0535,
+      "step": 4520
+    },
+    {
+      "epoch": 0.9476000838398658,
+      "grad_norm": 0.006846890784800053,
+      "learning_rate": 7.24330588797495e-08,
+      "loss": 0.0555,
+      "step": 4521
+    },
+    {
+      "epoch": 0.9478096835045063,
+      "grad_norm": 0.006406300701200962,
+      "learning_rate": 7.18584364396846e-08,
+      "loss": 0.0551,
+      "step": 4522
+    },
+    {
+      "epoch": 0.948019283169147,
+      "grad_norm": 0.006220379378646612,
+      "learning_rate": 7.128608587185615e-08,
+      "loss": 0.0531,
+      "step": 4523
+    },
+    {
+      "epoch": 0.9482288828337875,
+      "grad_norm": 0.006792591419070959,
+      "learning_rate": 7.071600744011865e-08,
+      "loss": 0.0516,
+      "step": 4524
+    },
+    {
+      "epoch": 0.948438482498428,
+      "grad_norm": 0.0063816942274570465,
+      "learning_rate": 7.014820140727797e-08,
+      "loss": 0.0542,
+      "step": 4525
+    },
+    {
+      "epoch": 0.9486480821630685,
+      "grad_norm": 0.006532501429319382,
+      "learning_rate": 6.958266803509195e-08,
+      "loss": 0.0515,
+      "step": 4526
+    },
+    {
+      "epoch": 0.948857681827709,
+      "grad_norm": 0.005698981694877148,
+      "learning_rate": 6.901940758427206e-08,
+      "loss": 0.0547,
+      "step": 4527
+    },
+    {
+      "epoch": 0.9490672814923496,
+      "grad_norm": 0.007572156842797995,
+      "learning_rate": 6.845842031448113e-08,
+      "loss": 0.0508,
+      "step": 4528
+    },
+    {
+      "epoch": 0.9492768811569902,
+      "grad_norm": 0.005259588360786438,
+      "learning_rate": 6.789970648433397e-08,
+      "loss": 0.0532,
+      "step": 4529
+    },
+    {
+      "epoch": 0.9494864808216307,
+      "grad_norm": 0.00802881084382534,
+      "learning_rate": 6.734326635139732e-08,
+      "loss": 0.0535,
+      "step": 4530
+    },
+    {
+      "epoch": 0.9496960804862712,
+      "grad_norm": 0.008471885696053505,
+      "learning_rate": 6.678910017219098e-08,
+      "loss": 0.0531,
+      "step": 4531
+    },
+    {
+      "epoch": 0.9499056801509118,
+      "grad_norm": 0.006917075254023075,
+      "learning_rate": 6.623720820218449e-08,
+      "loss": 0.0513,
+      "step": 4532
+    },
+    {
+      "epoch": 0.9501152798155523,
+      "grad_norm": 0.006756001152098179,
+      "learning_rate": 6.568759069579988e-08,
+      "loss": 0.0524,
+      "step": 4533
+    },
+    {
+      "epoch": 0.9503248794801928,
+      "grad_norm": 0.006958952639251947,
+      "learning_rate": 6.514024790641116e-08,
+      "loss": 0.054,
+      "step": 4534
+    },
+    {
+      "epoch": 0.9505344791448334,
+      "grad_norm": 0.006947326939553022,
+      "learning_rate": 6.459518008634313e-08,
+      "loss": 0.0548,
+      "step": 4535
+    },
+    {
+      "epoch": 0.9507440788094739,
+      "grad_norm": 0.006160185672342777,
+      "learning_rate": 6.405238748687203e-08,
+      "loss": 0.0538,
+      "step": 4536
+    },
+    {
+      "epoch": 0.9509536784741145,
+      "grad_norm": 0.007486341055482626,
+      "learning_rate": 6.351187035822492e-08,
+      "loss": 0.054,
+      "step": 4537
+    },
+    {
+      "epoch": 0.951163278138755,
+      "grad_norm": 0.006339498329907656,
+      "learning_rate": 6.297362894958025e-08,
+      "loss": 0.0563,
+      "step": 4538
+    },
+    {
+      "epoch": 0.9513728778033955,
+      "grad_norm": 0.006274237297475338,
+      "learning_rate": 6.243766350906733e-08,
+      "loss": 0.0538,
+      "step": 4539
+    },
+    {
+      "epoch": 0.951582477468036,
+      "grad_norm": 0.005623075179755688,
+      "learning_rate": 6.190397428376515e-08,
+      "loss": 0.0534,
+      "step": 4540
+    },
+    {
+      "epoch": 0.9517920771326766,
+      "grad_norm": 0.0076430076733231544,
+      "learning_rate": 6.137256151970583e-08,
+      "loss": 0.0539,
+      "step": 4541
+    },
+    {
+      "epoch": 0.9520016767973172,
+      "grad_norm": 0.006511925719678402,
+      "learning_rate": 6.08434254618695e-08,
+      "loss": 0.0525,
+      "step": 4542
+    },
+    {
+      "epoch": 0.9522112764619577,
+      "grad_norm": 0.0067582023330032825,
+      "learning_rate": 6.031656635418825e-08,
+      "loss": 0.0536,
+      "step": 4543
+    },
+    {
+      "epoch": 0.9524208761265982,
+      "grad_norm": 0.006065226625651121,
+      "learning_rate": 5.979198443954393e-08,
+      "loss": 0.0538,
+      "step": 4544
+    },
+    {
+      "epoch": 0.9526304757912387,
+      "grad_norm": 0.005706054624170065,
+      "learning_rate": 5.926967995976807e-08,
+      "loss": 0.0517,
+      "step": 4545
+    },
+    {
+      "epoch": 0.9528400754558792,
+      "grad_norm": 0.005716789048165083,
+      "learning_rate": 5.8749653155643626e-08,
+      "loss": 0.0539,
+      "step": 4546
+    },
+    {
+      "epoch": 0.9530496751205199,
+      "grad_norm": 0.0065559083595871925,
+      "learning_rate": 5.8231904266902726e-08,
+      "loss": 0.0522,
+      "step": 4547
+    },
+    {
+      "epoch": 0.9532592747851604,
+      "grad_norm": 0.005127377342432737,
+      "learning_rate": 5.771643353222778e-08,
+      "loss": 0.0544,
+      "step": 4548
+    },
+    {
+      "epoch": 0.9534688744498009,
+      "grad_norm": 0.005965971853584051,
+      "learning_rate": 5.720324118925036e-08,
+      "loss": 0.0529,
+      "step": 4549
+    },
+    {
+      "epoch": 0.9536784741144414,
+      "grad_norm": 0.0051533933728933334,
+      "learning_rate": 5.669232747455178e-08,
+      "loss": 0.0544,
+      "step": 4550
+    },
+    {
+      "epoch": 0.9538880737790819,
+      "grad_norm": 0.007161029148846865,
+      "learning_rate": 5.618369262366363e-08,
+      "loss": 0.0548,
+      "step": 4551
+    },
+    {
+      "epoch": 0.9540976734437225,
+      "grad_norm": 0.006360564846545458,
+      "learning_rate": 5.567733687106558e-08,
+      "loss": 0.0558,
+      "step": 4552
+    },
+    {
+      "epoch": 0.954307273108363,
+      "grad_norm": 0.006422116421163082,
+      "learning_rate": 5.517326045018867e-08,
+      "loss": 0.0535,
+      "step": 4553
+    },
+    {
+      "epoch": 0.9545168727730036,
+      "grad_norm": 0.00639104750007391,
+      "learning_rate": 5.4671463593412025e-08,
+      "loss": 0.0536,
+      "step": 4554
+    },
+    {
+      "epoch": 0.9547264724376441,
+      "grad_norm": 0.005469374358654022,
+      "learning_rate": 5.417194653206337e-08,
+      "loss": 0.0555,
+      "step": 4555
+    },
+    {
+      "epoch": 0.9549360721022846,
+      "grad_norm": 0.006682547274976969,
+      "learning_rate": 5.367470949641906e-08,
+      "loss": 0.0536,
+      "step": 4556
+    },
+    {
+      "epoch": 0.9551456717669252,
+      "grad_norm": 0.008950907737016678,
+      "learning_rate": 5.317975271570686e-08,
+      "loss": 0.0548,
+      "step": 4557
+    },
+    {
+      "epoch": 0.9553552714315657,
+      "grad_norm": 0.00793489534407854,
+      "learning_rate": 5.268707641810144e-08,
+      "loss": 0.0519,
+      "step": 4558
+    },
+    {
+      "epoch": 0.9555648710962062,
+      "grad_norm": 0.0063106887973845005,
+      "learning_rate": 5.2196680830725596e-08,
+      "loss": 0.0545,
+      "step": 4559
+    },
+    {
+      "epoch": 0.9557744707608468,
+      "grad_norm": 0.006787710823118687,
+      "learning_rate": 5.1708566179652363e-08,
+      "loss": 0.0508,
+      "step": 4560
+    },
+    {
+      "epoch": 0.9559840704254873,
+      "grad_norm": 0.008157514967024326,
+      "learning_rate": 5.122273268990285e-08,
+      "loss": 0.0546,
+      "step": 4561
+    },
+    {
+      "epoch": 0.9561936700901279,
+      "grad_norm": 0.00778998015448451,
+      "learning_rate": 5.073918058544458e-08,
+      "loss": 0.0536,
+      "step": 4562
+    },
+    {
+      "epoch": 0.9564032697547684,
+      "grad_norm": 0.009132279083132744,
+      "learning_rate": 5.025791008919645e-08,
+      "loss": 0.0527,
+      "step": 4563
+    },
+    {
+      "epoch": 0.9566128694194089,
+      "grad_norm": 0.006497305817902088,
+      "learning_rate": 4.977892142302376e-08,
+      "loss": 0.0539,
+      "step": 4564
+    },
+    {
+      "epoch": 0.9568224690840494,
+      "grad_norm": 0.005645697936415672,
+      "learning_rate": 4.930221480773989e-08,
+      "loss": 0.0531,
+      "step": 4565
+    },
+    {
+      "epoch": 0.95703206874869,
+      "grad_norm": 0.007060209289193153,
+      "learning_rate": 4.882779046310682e-08,
+      "loss": 0.053,
+      "step": 4566
+    },
+    {
+      "epoch": 0.9572416684133306,
+      "grad_norm": 0.005475896876305342,
+      "learning_rate": 4.835564860783404e-08,
+      "loss": 0.054,
+      "step": 4567
+    },
+    {
+      "epoch": 0.9574512680779711,
+      "grad_norm": 0.007286901585757732,
+      "learning_rate": 4.7885789459578e-08,
+      "loss": 0.0506,
+      "step": 4568
+    },
+    {
+      "epoch": 0.9576608677426116,
+      "grad_norm": 0.007155861239880323,
+      "learning_rate": 4.741821323494489e-08,
+      "loss": 0.0533,
+      "step": 4569
+    },
+    {
+      "epoch": 0.9578704674072521,
+      "grad_norm": 0.005204516928642988,
+      "learning_rate": 4.6952920149486715e-08,
+      "loss": 0.054,
+      "step": 4570
+    },
+    {
+      "epoch": 0.9580800670718926,
+      "grad_norm": 0.006855521816760302,
+      "learning_rate": 4.6489910417703564e-08,
+      "loss": 0.054,
+      "step": 4571
+    },
+    {
+      "epoch": 0.9582896667365333,
+      "grad_norm": 0.005472875200212002,
+      "learning_rate": 4.602918425304248e-08,
+      "loss": 0.0533,
+      "step": 4572
+    },
+    {
+      "epoch": 0.9584992664011738,
+      "grad_norm": 0.005690296180546284,
+      "learning_rate": 4.5570741867898563e-08,
+      "loss": 0.0535,
+      "step": 4573
+    },
+    {
+      "epoch": 0.9587088660658143,
+      "grad_norm": 0.0066895815543830395,
+      "learning_rate": 4.511458347361386e-08,
+      "loss": 0.0539,
+      "step": 4574
+    },
+    {
+      "epoch": 0.9589184657304548,
+      "grad_norm": 0.008229286409914494,
+      "learning_rate": 4.4660709280476275e-08,
+      "loss": 0.0512,
+      "step": 4575
+    },
+    {
+      "epoch": 0.9591280653950953,
+      "grad_norm": 0.0072631631046533585,
+      "learning_rate": 4.4209119497722883e-08,
+      "loss": 0.0546,
+      "step": 4576
+    },
+    {
+      "epoch": 0.959337665059736,
+      "grad_norm": 0.006190068554133177,
+      "learning_rate": 4.375981433353604e-08,
+      "loss": 0.0516,
+      "step": 4577
+    },
+    {
+      "epoch": 0.9595472647243765,
+      "grad_norm": 0.0051482743583619595,
+      "learning_rate": 4.331279399504507e-08,
+      "loss": 0.0529,
+      "step": 4578
+    },
+    {
+      "epoch": 0.959756864389017,
+      "grad_norm": 0.005935709923505783,
+      "learning_rate": 4.286805868832622e-08,
+      "loss": 0.0524,
+      "step": 4579
+    },
+    {
+      "epoch": 0.9599664640536575,
+      "grad_norm": 0.005591293331235647,
+      "learning_rate": 4.242560861840273e-08,
+      "loss": 0.0534,
+      "step": 4580
+    },
+    {
+      "epoch": 0.960176063718298,
+      "grad_norm": 0.006417642813175917,
+      "learning_rate": 4.19854439892442e-08,
+      "loss": 0.0546,
+      "step": 4581
+    },
+    {
+      "epoch": 0.9603856633829386,
+      "grad_norm": 0.006241221912205219,
+      "learning_rate": 4.154756500376611e-08,
+      "loss": 0.0545,
+      "step": 4582
+    },
+    {
+      "epoch": 0.9605952630475791,
+      "grad_norm": 0.004910702351480722,
+      "learning_rate": 4.1111971863830866e-08,
+      "loss": 0.054,
+      "step": 4583
+    },
+    {
+      "epoch": 0.9608048627122197,
+      "grad_norm": 0.005474326200783253,
+      "learning_rate": 4.0678664770246177e-08,
+      "loss": 0.0539,
+      "step": 4584
+    },
+    {
+      "epoch": 0.9610144623768602,
+      "grad_norm": 0.006352483760565519,
+      "learning_rate": 4.02476439227667e-08,
+      "loss": 0.0556,
+      "step": 4585
+    },
+    {
+      "epoch": 0.9612240620415007,
+      "grad_norm": 0.006457141134887934,
+      "learning_rate": 3.981890952009293e-08,
+      "loss": 0.0547,
+      "step": 4586
+    },
+    {
+      "epoch": 0.9614336617061413,
+      "grad_norm": 0.00615812698379159,
+      "learning_rate": 3.939246175987232e-08,
+      "loss": 0.0556,
+      "step": 4587
+    },
+    {
+      "epoch": 0.9616432613707818,
+      "grad_norm": 0.006070506758987904,
+      "learning_rate": 3.896830083869596e-08,
+      "loss": 0.0511,
+      "step": 4588
+    },
+    {
+      "epoch": 0.9618528610354223,
+      "grad_norm": 0.00562552735209465,
+      "learning_rate": 3.8546426952102425e-08,
+      "loss": 0.0568,
+      "step": 4589
+    },
+    {
+      "epoch": 0.9620624607000628,
+      "grad_norm": 0.006280606612563133,
+      "learning_rate": 3.812684029457614e-08,
+      "loss": 0.0551,
+      "step": 4590
+    },
+    {
+      "epoch": 0.9622720603647034,
+      "grad_norm": 0.006904164794832468,
+      "learning_rate": 3.770954105954461e-08,
+      "loss": 0.0524,
+      "step": 4591
+    },
+    {
+      "epoch": 0.962481660029344,
+      "grad_norm": 0.007641474716365337,
+      "learning_rate": 3.7294529439384494e-08,
+      "loss": 0.0559,
+      "step": 4592
+    },
+    {
+      "epoch": 0.9626912596939845,
+      "grad_norm": 0.005529316142201424,
+      "learning_rate": 3.6881805625415544e-08,
+      "loss": 0.0517,
+      "step": 4593
+    },
+    {
+      "epoch": 0.962900859358625,
+      "grad_norm": 0.004727725870907307,
+      "learning_rate": 3.647136980790333e-08,
+      "loss": 0.0551,
+      "step": 4594
+    },
+    {
+      "epoch": 0.9631104590232655,
+      "grad_norm": 0.00712125189602375,
+      "learning_rate": 3.606322217605873e-08,
+      "loss": 0.0546,
+      "step": 4595
+    },
+    {
+      "epoch": 0.963320058687906,
+      "grad_norm": 0.006122817751020193,
+      "learning_rate": 3.565736291803734e-08,
+      "loss": 0.0553,
+      "step": 4596
+    },
+    {
+      "epoch": 0.9635296583525467,
+      "grad_norm": 0.004824989475309849,
+      "learning_rate": 3.525379222094061e-08,
+      "loss": 0.0539,
+      "step": 4597
+    },
+    {
+      "epoch": 0.9637392580171872,
+      "grad_norm": 0.007551755756139755,
+      "learning_rate": 3.485251027081415e-08,
+      "loss": 0.0536,
+      "step": 4598
+    },
+    {
+      "epoch": 0.9639488576818277,
+      "grad_norm": 0.005360460840165615,
+      "learning_rate": 3.445351725264945e-08,
+      "loss": 0.0547,
+      "step": 4599
+    },
+    {
+      "epoch": 0.9641584573464682,
+      "grad_norm": 0.0053915646858513355,
+      "learning_rate": 3.405681335038158e-08,
+      "loss": 0.0536,
+      "step": 4600
+    },
+    {
+      "epoch": 0.9643680570111088,
+      "grad_norm": 0.0051850867457687855,
+      "learning_rate": 3.3662398746890924e-08,
+      "loss": 0.0531,
+      "step": 4601
+    },
+    {
+      "epoch": 0.9645776566757494,
+      "grad_norm": 0.0051136258989572525,
+      "learning_rate": 3.327027362400315e-08,
+      "loss": 0.0537,
+      "step": 4602
+    },
+    {
+      "epoch": 0.9647872563403899,
+      "grad_norm": 0.008604098111391068,
+      "learning_rate": 3.288043816248809e-08,
+      "loss": 0.0521,
+      "step": 4603
+    },
+    {
+      "epoch": 0.9649968560050304,
+      "grad_norm": 0.00728649040684104,
+      "learning_rate": 3.249289254205867e-08,
+      "loss": 0.0521,
+      "step": 4604
+    },
+    {
+      "epoch": 0.9652064556696709,
+      "grad_norm": 0.005971864331513643,
+      "learning_rate": 3.21076369413742e-08,
+      "loss": 0.0518,
+      "step": 4605
+    },
+    {
+      "epoch": 0.9654160553343115,
+      "grad_norm": 0.004847438540309668,
+      "learning_rate": 3.172467153803704e-08,
+      "loss": 0.0536,
+      "step": 4606
+    },
+    {
+      "epoch": 0.965625654998952,
+      "grad_norm": 0.004828069359064102,
+      "learning_rate": 3.134399650859432e-08,
+      "loss": 0.0529,
+      "step": 4607
+    },
+    {
+      "epoch": 0.9658352546635925,
+      "grad_norm": 0.004983065649867058,
+      "learning_rate": 3.096561202853676e-08,
+      "loss": 0.0563,
+      "step": 4608
+    },
+    {
+      "epoch": 0.9660448543282331,
+      "grad_norm": 0.005727593321353197,
+      "learning_rate": 3.0589518272300946e-08,
+      "loss": 0.052,
+      "step": 4609
+    },
+    {
+      "epoch": 0.9662544539928736,
+      "grad_norm": 0.005493259988725185,
+      "learning_rate": 3.0215715413264294e-08,
+      "loss": 0.0526,
+      "step": 4610
+    },
+    {
+      "epoch": 0.9664640536575142,
+      "grad_norm": 0.005342698656022549,
+      "learning_rate": 2.984420362375007e-08,
+      "loss": 0.0517,
+      "step": 4611
+    },
+    {
+      "epoch": 0.9666736533221547,
+      "grad_norm": 0.0063081239350140095,
+      "learning_rate": 2.9474983075026276e-08,
+      "loss": 0.0536,
+      "step": 4612
+    },
+    {
+      "epoch": 0.9668832529867952,
+      "grad_norm": 0.005974064581096172,
+      "learning_rate": 2.9108053937302316e-08,
+      "loss": 0.0562,
+      "step": 4613
+    },
+    {
+      "epoch": 0.9670928526514357,
+      "grad_norm": 0.005807976704090834,
+      "learning_rate": 2.8743416379733435e-08,
+      "loss": 0.0499,
+      "step": 4614
+    },
+    {
+      "epoch": 0.9673024523160763,
+      "grad_norm": 0.005039132200181484,
+      "learning_rate": 2.8381070570416835e-08,
+      "loss": 0.0521,
+      "step": 4615
+    },
+    {
+      "epoch": 0.9675120519807169,
+      "grad_norm": 0.007256725803017616,
+      "learning_rate": 2.8021016676393897e-08,
+      "loss": 0.0539,
+      "step": 4616
+    },
+    {
+      "epoch": 0.9677216516453574,
+      "grad_norm": 0.006315998733043671,
+      "learning_rate": 2.7663254863649625e-08,
+      "loss": 0.0531,
+      "step": 4617
+    },
+    {
+      "epoch": 0.9679312513099979,
+      "grad_norm": 0.006833828054368496,
+      "learning_rate": 2.7307785297111533e-08,
+      "loss": 0.0571,
+      "step": 4618
+    },
+    {
+      "epoch": 0.9681408509746384,
+      "grad_norm": 0.007867738604545593,
+      "learning_rate": 2.6954608140651872e-08,
+      "loss": 0.0566,
+      "step": 4619
+    },
+    {
+      "epoch": 0.9683504506392789,
+      "grad_norm": 0.005606554448604584,
+      "learning_rate": 2.6603723557085402e-08,
+      "loss": 0.0547,
+      "step": 4620
+    },
+    {
+      "epoch": 0.9685600503039196,
+      "grad_norm": 0.00556945102289319,
+      "learning_rate": 2.6255131708168845e-08,
+      "loss": 0.0523,
+      "step": 4621
+    },
+    {
+      "epoch": 0.9687696499685601,
+      "grad_norm": 0.006726409774273634,
+      "learning_rate": 2.5908832754603097e-08,
+      "loss": 0.0541,
+      "step": 4622
+    },
+    {
+      "epoch": 0.9689792496332006,
+      "grad_norm": 0.0066106924787163734,
+      "learning_rate": 2.5564826856032677e-08,
+      "loss": 0.054,
+      "step": 4623
+    },
+    {
+      "epoch": 0.9691888492978411,
+      "grad_norm": 0.004752601031213999,
+      "learning_rate": 2.5223114171043507e-08,
+      "loss": 0.0539,
+      "step": 4624
+    },
+    {
+      "epoch": 0.9693984489624816,
+      "grad_norm": 0.006882491055876017,
+      "learning_rate": 2.488369485716513e-08,
+      "loss": 0.0553,
+      "step": 4625
+    },
+    {
+      "epoch": 0.9696080486271222,
+      "grad_norm": 0.008670981973409653,
+      "learning_rate": 2.4546569070870717e-08,
+      "loss": 0.0556,
+      "step": 4626
+    },
+    {
+      "epoch": 0.9698176482917628,
+      "grad_norm": 0.006038348656147718,
+      "learning_rate": 2.4211736967574283e-08,
+      "loss": 0.0529,
+      "step": 4627
+    },
+    {
+      "epoch": 0.9700272479564033,
+      "grad_norm": 0.00571649894118309,
+      "learning_rate": 2.387919870163291e-08,
+      "loss": 0.054,
+      "step": 4628
+    },
+    {
+      "epoch": 0.9702368476210438,
+      "grad_norm": 0.004810945596545935,
+      "learning_rate": 2.354895442634786e-08,
+      "loss": 0.054,
+      "step": 4629
+    },
+    {
+      "epoch": 0.9704464472856843,
+      "grad_norm": 0.00811366643756628,
+      "learning_rate": 2.3221004293961237e-08,
+      "loss": 0.0542,
+      "step": 4630
+    },
+    {
+      "epoch": 0.9706560469503249,
+      "grad_norm": 0.005908562336117029,
+      "learning_rate": 2.289534845565766e-08,
+      "loss": 0.0528,
+      "step": 4631
+    },
+    {
+      "epoch": 0.9708656466149654,
+      "grad_norm": 0.00523727061226964,
+      "learning_rate": 2.2571987061564827e-08,
+      "loss": 0.0542,
+      "step": 4632
+    },
+    {
+      "epoch": 0.971075246279606,
+      "grad_norm": 0.006445094011723995,
+      "learning_rate": 2.225092026075182e-08,
+      "loss": 0.0511,
+      "step": 4633
+    },
+    {
+      "epoch": 0.9712848459442465,
+      "grad_norm": 0.006017652805894613,
+      "learning_rate": 2.193214820123024e-08,
+      "loss": 0.0585,
+      "step": 4634
+    },
+    {
+      "epoch": 0.971494445608887,
+      "grad_norm": 0.00453847274184227,
+      "learning_rate": 2.1615671029954765e-08,
+      "loss": 0.0548,
+      "step": 4635
+    },
+    {
+      "epoch": 0.9717040452735276,
+      "grad_norm": 0.0046917712315917015,
+      "learning_rate": 2.1301488892820908e-08,
+      "loss": 0.0533,
+      "step": 4636
+    },
+    {
+      "epoch": 0.9719136449381681,
+      "grad_norm": 0.004841329529881477,
+      "learning_rate": 2.098960193466615e-08,
+      "loss": 0.0536,
+      "step": 4637
+    },
+    {
+      "epoch": 0.9721232446028086,
+      "grad_norm": 0.0073344954289495945,
+      "learning_rate": 2.0680010299271024e-08,
+      "loss": 0.0557,
+      "step": 4638
+    },
+    {
+      "epoch": 0.9723328442674491,
+      "grad_norm": 0.006531344726681709,
+      "learning_rate": 2.0372714129356375e-08,
+      "loss": 0.0546,
+      "step": 4639
+    },
+    {
+      "epoch": 0.9725424439320897,
+      "grad_norm": 0.00542111974209547,
+      "learning_rate": 2.0067713566586654e-08,
+      "loss": 0.0539,
+      "step": 4640
+    },
+    {
+      "epoch": 0.9727520435967303,
+      "grad_norm": 0.004822442773729563,
+      "learning_rate": 1.9765008751566618e-08,
+      "loss": 0.0562,
+      "step": 4641
+    },
+    {
+      "epoch": 0.9729616432613708,
+      "grad_norm": 0.004666461609303951,
+      "learning_rate": 1.9464599823842966e-08,
+      "loss": 0.0558,
+      "step": 4642
+    },
+    {
+      "epoch": 0.9731712429260113,
+      "grad_norm": 0.005237584933638573,
+      "learning_rate": 1.9166486921903814e-08,
+      "loss": 0.0542,
+      "step": 4643
+    },
+    {
+      "epoch": 0.9733808425906518,
+      "grad_norm": 0.005628027021884918,
+      "learning_rate": 1.8870670183179783e-08,
+      "loss": 0.052,
+      "step": 4644
+    },
+    {
+      "epoch": 0.9735904422552923,
+      "grad_norm": 0.005040889140218496,
+      "learning_rate": 1.8577149744042343e-08,
+      "loss": 0.0522,
+      "step": 4645
+    },
+    {
+      "epoch": 0.973800041919933,
+      "grad_norm": 0.004789900500327349,
+      "learning_rate": 1.8285925739803812e-08,
+      "loss": 0.0521,
+      "step": 4646
+    },
+    {
+      "epoch": 0.9740096415845735,
+      "grad_norm": 0.00599799444898963,
+      "learning_rate": 1.7996998304719016e-08,
+      "loss": 0.0553,
+      "step": 4647
+    },
+    {
+      "epoch": 0.974219241249214,
+      "grad_norm": 0.004553031176328659,
+      "learning_rate": 1.7710367571983077e-08,
+      "loss": 0.0505,
+      "step": 4648
+    },
+    {
+      "epoch": 0.9744288409138545,
+      "grad_norm": 0.00728783430531621,
+      "learning_rate": 1.7426033673733077e-08,
+      "loss": 0.0537,
+      "step": 4649
+    },
+    {
+      "epoch": 0.974638440578495,
+      "grad_norm": 0.004845589864999056,
+      "learning_rate": 1.7143996741045832e-08,
+      "loss": 0.0554,
+      "step": 4650
+    },
+    {
+      "epoch": 0.9748480402431356,
+      "grad_norm": 0.005394340958446264,
+      "learning_rate": 1.686425690394178e-08,
+      "loss": 0.0516,
+      "step": 4651
+    },
+    {
+      "epoch": 0.9750576399077762,
+      "grad_norm": 0.006395083852112293,
+      "learning_rate": 1.6586814291379428e-08,
+      "loss": 0.0553,
+      "step": 4652
+    },
+    {
+      "epoch": 0.9752672395724167,
+      "grad_norm": 0.005952565465122461,
+      "learning_rate": 1.631166903126147e-08,
+      "loss": 0.0554,
+      "step": 4653
+    },
+    {
+      "epoch": 0.9754768392370572,
+      "grad_norm": 0.004673383664339781,
+      "learning_rate": 1.603882125042866e-08,
+      "loss": 0.0555,
+      "step": 4654
+    },
+    {
+      "epoch": 0.9756864389016977,
+      "grad_norm": 0.005786755122244358,
+      "learning_rate": 1.576827107466372e-08,
+      "loss": 0.0529,
+      "step": 4655
+    },
+    {
+      "epoch": 0.9758960385663383,
+      "grad_norm": 0.006783796474337578,
+      "learning_rate": 1.5500018628690216e-08,
+      "loss": 0.0526,
+      "step": 4656
+    },
+    {
+      "epoch": 0.9761056382309788,
+      "grad_norm": 0.006686339620500803,
+      "learning_rate": 1.5234064036173114e-08,
+      "loss": 0.0541,
+      "step": 4657
+    },
+    {
+      "epoch": 0.9763152378956194,
+      "grad_norm": 0.006544138304889202,
+      "learning_rate": 1.4970407419717116e-08,
+      "loss": 0.0517,
+      "step": 4658
+    },
+    {
+      "epoch": 0.9765248375602599,
+      "grad_norm": 0.004879649728536606,
+      "learning_rate": 1.4709048900867772e-08,
+      "loss": 0.0521,
+      "step": 4659
+    },
+    {
+      "epoch": 0.9767344372249004,
+      "grad_norm": 0.006231387611478567,
+      "learning_rate": 1.4449988600111486e-08,
+      "loss": 0.0551,
+      "step": 4660
+    },
+    {
+      "epoch": 0.976944036889541,
+      "grad_norm": 0.005122179165482521,
+      "learning_rate": 1.4193226636874391e-08,
+      "loss": 0.0516,
+      "step": 4661
+    },
+    {
+      "epoch": 0.9771536365541815,
+      "grad_norm": 0.004648258443921804,
+      "learning_rate": 1.393876312952458e-08,
+      "loss": 0.0523,
+      "step": 4662
+    },
+    {
+      "epoch": 0.977363236218822,
+      "grad_norm": 0.005532658658921719,
+      "learning_rate": 1.3686598195369327e-08,
+      "loss": 0.0589,
+      "step": 4663
+    },
+    {
+      "epoch": 0.9775728358834626,
+      "grad_norm": 0.00559549406170845,
+      "learning_rate": 1.34367319506562e-08,
+      "loss": 0.0499,
+      "step": 4664
+    },
+    {
+      "epoch": 0.9777824355481031,
+      "grad_norm": 0.004705341067165136,
+      "learning_rate": 1.318916451057417e-08,
+      "loss": 0.0535,
+      "step": 4665
+    },
+    {
+      "epoch": 0.9779920352127437,
+      "grad_norm": 0.005659153684973717,
+      "learning_rate": 1.2943895989251387e-08,
+      "loss": 0.0543,
+      "step": 4666
+    },
+    {
+      "epoch": 0.9782016348773842,
+      "grad_norm": 0.004608216229826212,
+      "learning_rate": 1.2700926499756295e-08,
+      "loss": 0.058,
+      "step": 4667
+    },
+    {
+      "epoch": 0.9784112345420247,
+      "grad_norm": 0.005427349358797073,
+      "learning_rate": 1.2460256154098738e-08,
+      "loss": 0.0566,
+      "step": 4668
+    },
+    {
+      "epoch": 0.9786208342066652,
+      "grad_norm": 0.0053130644373595715,
+      "learning_rate": 1.2221885063226635e-08,
+      "loss": 0.0516,
+      "step": 4669
+    },
+    {
+      "epoch": 0.9788304338713059,
+      "grad_norm": 0.004453903995454311,
+      "learning_rate": 1.198581333702986e-08,
+      "loss": 0.053,
+      "step": 4670
+    },
+    {
+      "epoch": 0.9790400335359464,
+      "grad_norm": 0.0048965164460241795,
+      "learning_rate": 1.1752041084336364e-08,
+      "loss": 0.0536,
+      "step": 4671
+    },
+    {
+      "epoch": 0.9792496332005869,
+      "grad_norm": 0.008032665587961674,
+      "learning_rate": 1.15205684129166e-08,
+      "loss": 0.0548,
+      "step": 4672
+    },
+    {
+      "epoch": 0.9794592328652274,
+      "grad_norm": 0.004445253871381283,
+      "learning_rate": 1.1291395429477991e-08,
+      "loss": 0.052,
+      "step": 4673
+    },
+    {
+      "epoch": 0.9796688325298679,
+      "grad_norm": 0.004714785609394312,
+      "learning_rate": 1.1064522239669916e-08,
+      "loss": 0.0553,
+      "step": 4674
+    },
+    {
+      "epoch": 0.9798784321945085,
+      "grad_norm": 0.004439793061465025,
+      "learning_rate": 1.0839948948080937e-08,
+      "loss": 0.0509,
+      "step": 4675
+    },
+    {
+      "epoch": 0.9800880318591491,
+      "grad_norm": 0.004678604193031788,
+      "learning_rate": 1.0617675658239345e-08,
+      "loss": 0.0568,
+      "step": 4676
+    },
+    {
+      "epoch": 0.9802976315237896,
+      "grad_norm": 0.007188865914940834,
+      "learning_rate": 1.0397702472612625e-08,
+      "loss": 0.053,
+      "step": 4677
+    },
+    {
+      "epoch": 0.9805072311884301,
+      "grad_norm": 0.006724335718899965,
+      "learning_rate": 1.0180029492608546e-08,
+      "loss": 0.0556,
+      "step": 4678
+    },
+    {
+      "epoch": 0.9807168308530706,
+      "grad_norm": 0.006979641038924456,
+      "learning_rate": 9.964656818574614e-09,
+      "loss": 0.0528,
+      "step": 4679
+    },
+    {
+      "epoch": 0.9809264305177112,
+      "grad_norm": 0.0065865106880664825,
+      "learning_rate": 9.751584549796966e-09,
+      "loss": 0.0538,
+      "step": 4680
+    },
+    {
+      "epoch": 0.9811360301823517,
+      "grad_norm": 0.005535759497433901,
+      "learning_rate": 9.54081278450314e-09,
+      "loss": 0.0547,
+      "step": 4681
+    },
+    {
+      "epoch": 0.9813456298469923,
+      "grad_norm": 0.006264574825763702,
+      "learning_rate": 9.332341619857078e-09,
+      "loss": 0.0541,
+      "step": 4682
+    },
+    {
+      "epoch": 0.9815552295116328,
+      "grad_norm": 0.006483092904090881,
+      "learning_rate": 9.126171151965235e-09,
+      "loss": 0.0536,
+      "step": 4683
+    },
+    {
+      "epoch": 0.9817648291762733,
+      "grad_norm": 0.005037569906562567,
+      "learning_rate": 8.922301475872141e-09,
+      "loss": 0.0548,
+      "step": 4684
+    },
+    {
+      "epoch": 0.9819744288409139,
+      "grad_norm": 0.005114252213388681,
+      "learning_rate": 8.72073268556095e-09,
+      "loss": 0.055,
+      "step": 4685
+    },
+    {
+      "epoch": 0.9821840285055544,
+      "grad_norm": 0.006643661763519049,
+      "learning_rate": 8.52146487395511e-09,
+      "loss": 0.0564,
+      "step": 4686
+    },
+    {
+      "epoch": 0.9823936281701949,
+      "grad_norm": 0.007684056647121906,
+      "learning_rate": 8.324498132917248e-09,
+      "loss": 0.055,
+      "step": 4687
+    },
+    {
+      "epoch": 0.9826032278348354,
+      "grad_norm": 0.005348839331418276,
+      "learning_rate": 8.129832553249173e-09,
+      "loss": 0.0527,
+      "step": 4688
+    },
+    {
+      "epoch": 0.982812827499476,
+      "grad_norm": 0.006276331376284361,
+      "learning_rate": 7.937468224691325e-09,
+      "loss": 0.0553,
+      "step": 4689
+    },
+    {
+      "epoch": 0.9830224271641166,
+      "grad_norm": 0.005827800370752811,
+      "learning_rate": 7.747405235923322e-09,
+      "loss": 0.0534,
+      "step": 4690
+    },
+    {
+      "epoch": 0.9832320268287571,
+      "grad_norm": 0.005027239676564932,
+      "learning_rate": 7.55964367456452e-09,
+      "loss": 0.0505,
+      "step": 4691
+    },
+    {
+      "epoch": 0.9834416264933976,
+      "grad_norm": 0.004755694884806871,
+      "learning_rate": 7.374183627173459e-09,
+      "loss": 0.0544,
+      "step": 4692
+    },
+    {
+      "epoch": 0.9836512261580381,
+      "grad_norm": 0.004315654281526804,
+      "learning_rate": 7.191025179246192e-09,
+      "loss": 0.0505,
+      "step": 4693
+    },
+    {
+      "epoch": 0.9838608258226786,
+      "grad_norm": 0.006336344871670008,
+      "learning_rate": 7.010168415219621e-09,
+      "loss": 0.0546,
+      "step": 4694
+    },
+    {
+      "epoch": 0.9840704254873193,
+      "grad_norm": 0.004240935202687979,
+      "learning_rate": 6.831613418468163e-09,
+      "loss": 0.0529,
+      "step": 4695
+    },
+    {
+      "epoch": 0.9842800251519598,
+      "grad_norm": 0.00471118651330471,
+      "learning_rate": 6.655360271305422e-09,
+      "loss": 0.0562,
+      "step": 4696
+    },
+    {
+      "epoch": 0.9844896248166003,
+      "grad_norm": 0.005362183786928654,
+      "learning_rate": 6.4814090549847334e-09,
+      "loss": 0.0535,
+      "step": 4697
+    },
+    {
+      "epoch": 0.9846992244812408,
+      "grad_norm": 0.003966295626014471,
+      "learning_rate": 6.30975984969695e-09,
+      "loss": 0.0536,
+      "step": 4698
+    },
+    {
+      "epoch": 0.9849088241458813,
+      "grad_norm": 0.005599298048764467,
+      "learning_rate": 6.140412734572665e-09,
+      "loss": 0.0529,
+      "step": 4699
+    },
+    {
+      "epoch": 0.985118423810522,
+      "grad_norm": 0.00638514244928956,
+      "learning_rate": 5.973367787681095e-09,
+      "loss": 0.0514,
+      "step": 4700
+    },
+    {
+      "epoch": 0.9853280234751625,
+      "grad_norm": 0.004646534100174904,
+      "learning_rate": 5.808625086029529e-09,
+      "loss": 0.0538,
+      "step": 4701
+    },
+    {
+      "epoch": 0.985537623139803,
+      "grad_norm": 0.0049830214120447636,
+      "learning_rate": 5.646184705563884e-09,
+      "loss": 0.0535,
+      "step": 4702
+    },
+    {
+      "epoch": 0.9857472228044435,
+      "grad_norm": 0.005619076080620289,
+      "learning_rate": 5.486046721170368e-09,
+      "loss": 0.0554,
+      "step": 4703
+    },
+    {
+      "epoch": 0.985956822469084,
+      "grad_norm": 0.004761868622153997,
+      "learning_rate": 5.328211206671596e-09,
+      "loss": 0.0546,
+      "step": 4704
+    },
+    {
+      "epoch": 0.9861664221337246,
+      "grad_norm": 0.004782952833920717,
+      "learning_rate": 5.17267823482992e-09,
+      "loss": 0.0549,
+      "step": 4705
+    },
+    {
+      "epoch": 0.9863760217983651,
+      "grad_norm": 0.005984222050756216,
+      "learning_rate": 5.019447877346317e-09,
+      "loss": 0.055,
+      "step": 4706
+    },
+    {
+      "epoch": 0.9865856214630057,
+      "grad_norm": 0.004718398675322533,
+      "learning_rate": 4.868520204859284e-09,
+      "loss": 0.0522,
+      "step": 4707
+    },
+    {
+      "epoch": 0.9867952211276462,
+      "grad_norm": 0.005698632914572954,
+      "learning_rate": 4.719895286947052e-09,
+      "loss": 0.0547,
+      "step": 4708
+    },
+    {
+      "epoch": 0.9870048207922867,
+      "grad_norm": 0.005634862929582596,
+      "learning_rate": 4.573573192125369e-09,
+      "loss": 0.0523,
+      "step": 4709
+    },
+    {
+      "epoch": 0.9872144204569273,
+      "grad_norm": 0.0050827860832214355,
+      "learning_rate": 4.429553987849167e-09,
+      "loss": 0.0547,
+      "step": 4710
+    },
+    {
+      "epoch": 0.9874240201215678,
+      "grad_norm": 0.005555478390306234,
+      "learning_rate": 4.287837740510336e-09,
+      "loss": 0.0524,
+      "step": 4711
+    },
+    {
+      "epoch": 0.9876336197862083,
+      "grad_norm": 0.0045521133579313755,
+      "learning_rate": 4.148424515441063e-09,
+      "loss": 0.0535,
+      "step": 4712
+    },
+    {
+      "epoch": 0.9878432194508489,
+      "grad_norm": 0.005545974709093571,
+      "learning_rate": 4.011314376909936e-09,
+      "loss": 0.0513,
+      "step": 4713
+    },
+    {
+      "epoch": 0.9880528191154894,
+      "grad_norm": 0.005865436978638172,
+      "learning_rate": 3.876507388125839e-09,
+      "loss": 0.0522,
+      "step": 4714
+    },
+    {
+      "epoch": 0.98826241878013,
+      "grad_norm": 0.004759861622005701,
+      "learning_rate": 3.744003611233505e-09,
+      "loss": 0.0529,
+      "step": 4715
+    },
+    {
+      "epoch": 0.9884720184447705,
+      "grad_norm": 0.004680339712649584,
+      "learning_rate": 3.613803107317959e-09,
+      "loss": 0.0536,
+      "step": 4716
+    },
+    {
+      "epoch": 0.988681618109411,
+      "grad_norm": 0.005101062823086977,
+      "learning_rate": 3.4859059364006354e-09,
+      "loss": 0.0506,
+      "step": 4717
+    },
+    {
+      "epoch": 0.9888912177740515,
+      "grad_norm": 0.005664270371198654,
+      "learning_rate": 3.3603121574438126e-09,
+      "loss": 0.0565,
+      "step": 4718
+    },
+    {
+      "epoch": 0.989100817438692,
+      "grad_norm": 0.005366276018321514,
+      "learning_rate": 3.237021828344511e-09,
+      "loss": 0.0542,
+      "step": 4719
+    },
+    {
+      "epoch": 0.9893104171033327,
+      "grad_norm": 0.004395569209009409,
+      "learning_rate": 3.1160350059405986e-09,
+      "loss": 0.0537,
+      "step": 4720
+    },
+    {
+      "epoch": 0.9895200167679732,
+      "grad_norm": 0.00447695842012763,
+      "learning_rate": 2.9973517460063496e-09,
+      "loss": 0.0532,
+      "step": 4721
+    },
+    {
+      "epoch": 0.9897296164326137,
+      "grad_norm": 0.00491610262542963,
+      "learning_rate": 2.8809721032552197e-09,
+      "loss": 0.054,
+      "step": 4722
+    },
+    {
+      "epoch": 0.9899392160972542,
+      "grad_norm": 0.0043014525435864925,
+      "learning_rate": 2.7668961313376263e-09,
+      "loss": 0.0561,
+      "step": 4723
+    },
+    {
+      "epoch": 0.9901488157618947,
+      "grad_norm": 0.007249589078128338,
+      "learning_rate": 2.6551238828431692e-09,
+      "loss": 0.0538,
+      "step": 4724
+    },
+    {
+      "epoch": 0.9903584154265354,
+      "grad_norm": 0.005701399873942137,
+      "learning_rate": 2.5456554092984087e-09,
+      "loss": 0.0568,
+      "step": 4725
+    },
+    {
+      "epoch": 0.9905680150911759,
+      "grad_norm": 0.00624644011259079,
+      "learning_rate": 2.438490761168533e-09,
+      "loss": 0.0554,
+      "step": 4726
+    },
+    {
+      "epoch": 0.9907776147558164,
+      "grad_norm": 0.005006265826523304,
+      "learning_rate": 2.3336299878562453e-09,
+      "loss": 0.0531,
+      "step": 4727
+    },
+    {
+      "epoch": 0.9909872144204569,
+      "grad_norm": 0.005804962012916803,
+      "learning_rate": 2.231073137702877e-09,
+      "loss": 0.0542,
+      "step": 4728
+    },
+    {
+      "epoch": 0.9911968140850974,
+      "grad_norm": 0.005389668978750706,
+      "learning_rate": 2.1308202579861657e-09,
+      "loss": 0.0529,
+      "step": 4729
+    },
+    {
+      "epoch": 0.991406413749738,
+      "grad_norm": 0.005474920384585857,
+      "learning_rate": 2.0328713949230304e-09,
+      "loss": 0.0545,
+      "step": 4730
+    },
+    {
+      "epoch": 0.9916160134143785,
+      "grad_norm": 0.0048176608979702,
+      "learning_rate": 1.937226593668462e-09,
+      "loss": 0.0534,
+      "step": 4731
+    },
+    {
+      "epoch": 0.9918256130790191,
+      "grad_norm": 0.004726291634142399,
+      "learning_rate": 1.8438858983138575e-09,
+      "loss": 0.0535,
+      "step": 4732
+    },
+    {
+      "epoch": 0.9920352127436596,
+      "grad_norm": 0.0048382277600467205,
+      "learning_rate": 1.752849351889796e-09,
+      "loss": 0.0548,
+      "step": 4733
+    },
+    {
+      "epoch": 0.9922448124083001,
+      "grad_norm": 0.006312186364084482,
+      "learning_rate": 1.6641169963638182e-09,
+      "loss": 0.0554,
+      "step": 4734
+    },
+    {
+      "epoch": 0.9924544120729407,
+      "grad_norm": 0.005486776586622,
+      "learning_rate": 1.5776888726420913e-09,
+      "loss": 0.0535,
+      "step": 4735
+    },
+    {
+      "epoch": 0.9926640117375812,
+      "grad_norm": 0.005428542383015156,
+      "learning_rate": 1.4935650205671893e-09,
+      "loss": 0.0557,
+      "step": 4736
+    },
+    {
+      "epoch": 0.9928736114022217,
+      "grad_norm": 0.0053994497284293175,
+      "learning_rate": 1.4117454789208673e-09,
+      "loss": 0.0514,
+      "step": 4737
+    },
+    {
+      "epoch": 0.9930832110668623,
+      "grad_norm": 0.005145637784153223,
+      "learning_rate": 1.3322302854212876e-09,
+      "loss": 0.0544,
+      "step": 4738
+    },
+    {
+      "epoch": 0.9932928107315029,
+      "grad_norm": 0.005606017541140318,
+      "learning_rate": 1.2550194767252387e-09,
+      "loss": 0.054,
+      "step": 4739
+    },
+    {
+      "epoch": 0.9935024103961434,
+      "grad_norm": 0.004900652449578047,
+      "learning_rate": 1.1801130884270262e-09,
+      "loss": 0.0527,
+      "step": 4740
+    },
+    {
+      "epoch": 0.9937120100607839,
+      "grad_norm": 0.0050704628229141235,
+      "learning_rate": 1.1075111550579166e-09,
+      "loss": 0.0548,
+      "step": 4741
+    },
+    {
+      "epoch": 0.9939216097254244,
+      "grad_norm": 0.004520857241004705,
+      "learning_rate": 1.0372137100883584e-09,
+      "loss": 0.0558,
+      "step": 4742
+    },
+    {
+      "epoch": 0.9941312093900649,
+      "grad_norm": 0.004727307241410017,
+      "learning_rate": 9.692207859246516e-10,
+      "loss": 0.0543,
+      "step": 4743
+    },
+    {
+      "epoch": 0.9943408090547056,
+      "grad_norm": 0.0052885450422763824,
+      "learning_rate": 9.03532413911723e-10,
+      "loss": 0.0537,
+      "step": 4744
+    },
+    {
+      "epoch": 0.9945504087193461,
+      "grad_norm": 0.00510826800018549,
+      "learning_rate": 8.401486243320156e-10,
+      "loss": 0.0523,
+      "step": 4745
+    },
+    {
+      "epoch": 0.9947600083839866,
+      "grad_norm": 0.005960457026958466,
+      "learning_rate": 7.790694464054893e-10,
+      "loss": 0.0528,
+      "step": 4746
+    },
+    {
+      "epoch": 0.9949696080486271,
+      "grad_norm": 0.006908085662871599,
+      "learning_rate": 7.202949082890654e-10,
+      "loss": 0.054,
+      "step": 4747
+    },
+    {
+      "epoch": 0.9951792077132676,
+      "grad_norm": 0.004360835067927837,
+      "learning_rate": 6.638250370788468e-10,
+      "loss": 0.0512,
+      "step": 4748
+    },
+    {
+      "epoch": 0.9953888073779082,
+      "grad_norm": 0.00422016391530633,
+      "learning_rate": 6.09659858806233e-10,
+      "loss": 0.0539,
+      "step": 4749
+    },
+    {
+      "epoch": 0.9955984070425488,
+      "grad_norm": 0.0072461930103600025,
+      "learning_rate": 5.577993984423602e-10,
+      "loss": 0.0521,
+      "step": 4750
+    },
+    {
+      "epoch": 0.9958080067071893,
+      "grad_norm": 0.0055481684394180775,
+      "learning_rate": 5.08243679894771e-10,
+      "loss": 0.054,
+      "step": 4751
+    },
+    {
+      "epoch": 0.9960176063718298,
+      "grad_norm": 0.004346661269664764,
+      "learning_rate": 4.609927260079694e-10,
+      "loss": 0.0544,
+      "step": 4752
+    },
+    {
+      "epoch": 0.9962272060364703,
+      "grad_norm": 0.004471504595130682,
+      "learning_rate": 4.1604655856508633e-10,
+      "loss": 0.0528,
+      "step": 4753
+    },
+    {
+      "epoch": 0.9964368057011109,
+      "grad_norm": 0.004563014954328537,
+      "learning_rate": 3.7340519828621415e-10,
+      "loss": 0.0531,
+      "step": 4754
+    },
+    {
+      "epoch": 0.9966464053657514,
+      "grad_norm": 0.005326093640178442,
+      "learning_rate": 3.330686648289616e-10,
+      "loss": 0.0527,
+      "step": 4755
+    },
+    {
+      "epoch": 0.996856005030392,
+      "grad_norm": 0.005793221294879913,
+      "learning_rate": 2.950369767884542e-10,
+      "loss": 0.0533,
+      "step": 4756
+    },
+    {
+      "epoch": 0.9970656046950325,
+      "grad_norm": 0.005273030139505863,
+      "learning_rate": 2.59310151697334e-10,
+      "loss": 0.0521,
+      "step": 4757
+    },
+    {
+      "epoch": 0.997275204359673,
+      "grad_norm": 0.004422128200531006,
+      "learning_rate": 2.2588820602631457e-10,
+      "loss": 0.0517,
+      "step": 4758
+    },
+    {
+      "epoch": 0.9974848040243136,
+      "grad_norm": 0.006338967010378838,
+      "learning_rate": 1.9477115518140577e-10,
+      "loss": 0.0555,
+      "step": 4759
+    },
+    {
+      "epoch": 0.9976944036889541,
+      "grad_norm": 0.004579038359224796,
+      "learning_rate": 1.6595901350890954e-10,
+      "loss": 0.0539,
+      "step": 4760
+    },
+    {
+      "epoch": 0.9979040033535946,
+      "grad_norm": 0.005021790973842144,
+      "learning_rate": 1.39451794290979e-10,
+      "loss": 0.0528,
+      "step": 4761
+    },
+    {
+      "epoch": 0.9981136030182352,
+      "grad_norm": 0.004821773152798414,
+      "learning_rate": 1.1524950974672878e-10,
+      "loss": 0.0525,
+      "step": 4762
+    },
+    {
+      "epoch": 0.9983232026828757,
+      "grad_norm": 0.004908620845526457,
+      "learning_rate": 9.335217103445538e-11,
+      "loss": 0.0524,
+      "step": 4763
+    },
+    {
+      "epoch": 0.9985328023475163,
+      "grad_norm": 0.007371780462563038,
+      "learning_rate": 7.375978824775143e-11,
+      "loss": 0.0526,
+      "step": 4764
+    },
+    {
+      "epoch": 0.9987424020121568,
+      "grad_norm": 0.005421594250947237,
+      "learning_rate": 5.6472370419391464e-11,
+      "loss": 0.0543,
+      "step": 4765
+    },
+    {
+      "epoch": 0.9989520016767973,
+      "grad_norm": 0.0057752360589802265,
+      "learning_rate": 4.148992551855635e-11,
+      "loss": 0.053,
+      "step": 4766
+    },
+    {
+      "epoch": 0.9991616013414378,
+      "grad_norm": 0.007910934276878834,
+      "learning_rate": 2.8812460452498637e-11,
+      "loss": 0.0517,
+      "step": 4767
+    },
+    {
+      "epoch": 0.9993712010060783,
+      "grad_norm": 0.005225565284490585,
+      "learning_rate": 1.843998106543232e-11,
+      "loss": 0.0547,
+      "step": 4768
+    },
+    {
+      "epoch": 0.999580800670719,
+      "grad_norm": 0.005672250874340534,
+      "learning_rate": 1.0372492138532864e-11,
+      "loss": 0.0564,
+      "step": 4769
+    },
+    {
+      "epoch": 0.9997904003353595,
+      "grad_norm": 0.005982316564768553,
+      "learning_rate": 4.609997391602505e-12,
+      "loss": 0.0538,
+      "step": 4770
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.005909937433898449,
+      "learning_rate": 1.1524994808498335e-12,
+      "loss": 0.0547,
+      "step": 4771
+    },
+    {
+      "epoch": 1.0,
+      "step": 4771,
+      "total_flos": 0.0,
+      "train_loss": 0.06302105758305591,
+      "train_runtime": 301007.0841,
+      "train_samples_per_second": 3.043,
+      "train_steps_per_second": 0.016
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 4771,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 0.0,
+  "train_batch_size": 6,
+  "trial_name": null,
+  "trial_params": null
+}