diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,22962 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.9998473981382573,
+  "eval_steps": 500,
+  "global_step": 3276,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0003052037234854265,
+      "grad_norm": 19.476922880741295,
+      "learning_rate": 1.0101010101010103e-07,
+      "loss": 1.1728,
+      "step": 1
+    },
+    {
+      "epoch": 0.000610407446970853,
+      "grad_norm": 29.879020388476594,
+      "learning_rate": 2.0202020202020205e-07,
+      "loss": 1.0955,
+      "step": 2
+    },
+    {
+      "epoch": 0.0009156111704562796,
+      "grad_norm": 24.931945947136526,
+      "learning_rate": 3.0303030303030305e-07,
+      "loss": 0.9541,
+      "step": 3
+    },
+    {
+      "epoch": 0.001220814893941706,
+      "grad_norm": 27.83214939667906,
+      "learning_rate": 4.040404040404041e-07,
+      "loss": 1.0735,
+      "step": 4
+    },
+    {
+      "epoch": 0.0015260186174271325,
+      "grad_norm": 21.219233961021736,
+      "learning_rate": 5.05050505050505e-07,
+      "loss": 1.0455,
+      "step": 5
+    },
+    {
+      "epoch": 0.0018312223409125592,
+      "grad_norm": 20.022707446211225,
+      "learning_rate": 6.060606060606061e-07,
+      "loss": 0.9675,
+      "step": 6
+    },
+    {
+      "epoch": 0.0021364260643979855,
+      "grad_norm": 26.532427830157193,
+      "learning_rate": 7.070707070707071e-07,
+      "loss": 1.1393,
+      "step": 7
+    },
+    {
+      "epoch": 0.002441629787883412,
+      "grad_norm": 27.89728780710031,
+      "learning_rate": 8.080808080808082e-07,
+      "loss": 1.0952,
+      "step": 8
+    },
+    {
+      "epoch": 0.0027468335113688385,
+      "grad_norm": 20.346264005570532,
+      "learning_rate": 9.090909090909091e-07,
+      "loss": 0.9626,
+      "step": 9
+    },
+    {
+      "epoch": 0.003052037234854265,
+      "grad_norm": 18.804489508720884,
+      "learning_rate": 1.01010101010101e-06,
+      "loss": 1.0255,
+      "step": 10
+    },
+    {
+      "epoch": 0.003357240958339692,
+      "grad_norm": 19.776534785573535,
+      "learning_rate": 1.111111111111111e-06,
+      "loss": 0.7399,
+      "step": 11
+    },
+    {
+      "epoch": 0.0036624446818251184,
+      "grad_norm": 21.16130386460154,
+      "learning_rate": 1.2121212121212122e-06,
+      "loss": 0.5413,
+      "step": 12
+    },
+    {
+      "epoch": 0.0039676484053105445,
+      "grad_norm": 16.482713371526263,
+      "learning_rate": 1.3131313131313134e-06,
+      "loss": 0.5773,
+      "step": 13
+    },
+    {
+      "epoch": 0.004272852128795971,
+      "grad_norm": 10.780528168770594,
+      "learning_rate": 1.4141414141414143e-06,
+      "loss": 0.6782,
+      "step": 14
+    },
+    {
+      "epoch": 0.0045780558522813975,
+      "grad_norm": 7.0900135030469915,
+      "learning_rate": 1.5151515151515152e-06,
+      "loss": 0.9153,
+      "step": 15
+    },
+    {
+      "epoch": 0.004883259575766824,
+      "grad_norm": 8.490445320662754,
+      "learning_rate": 1.6161616161616164e-06,
+      "loss": 0.4798,
+      "step": 16
+    },
+    {
+      "epoch": 0.0051884632992522505,
+      "grad_norm": 6.677142812986669,
+      "learning_rate": 1.7171717171717173e-06,
+      "loss": 0.4782,
+      "step": 17
+    },
+    {
+      "epoch": 0.005493667022737677,
+      "grad_norm": 5.9204247946017485,
+      "learning_rate": 1.8181818181818183e-06,
+      "loss": 0.3191,
+      "step": 18
+    },
+    {
+      "epoch": 0.0057988707462231035,
+      "grad_norm": 5.012462343754674,
+      "learning_rate": 1.9191919191919192e-06,
+      "loss": 0.4115,
+      "step": 19
+    },
+    {
+      "epoch": 0.00610407446970853,
+      "grad_norm": 3.9095937836899113,
+      "learning_rate": 2.02020202020202e-06,
+      "loss": 0.6158,
+      "step": 20
+    },
+    {
+      "epoch": 0.006409278193193957,
+      "grad_norm": 4.438163815129716,
+      "learning_rate": 2.1212121212121216e-06,
+      "loss": 0.7388,
+      "step": 21
+    },
+    {
+      "epoch": 0.006714481916679384,
+      "grad_norm": 3.62875198348435,
+      "learning_rate": 2.222222222222222e-06,
+      "loss": 0.2875,
+      "step": 22
+    },
+    {
+      "epoch": 0.00701968564016481,
+      "grad_norm": 4.963543929599541,
+      "learning_rate": 2.3232323232323234e-06,
+      "loss": 0.4662,
+      "step": 23
+    },
+    {
+      "epoch": 0.007324889363650237,
+      "grad_norm": 4.274904100558248,
+      "learning_rate": 2.4242424242424244e-06,
+      "loss": 0.5171,
+      "step": 24
+    },
+    {
+      "epoch": 0.007630093087135663,
+      "grad_norm": 2.670885047669819,
+      "learning_rate": 2.5252525252525258e-06,
+      "loss": 0.4488,
+      "step": 25
+    },
+    {
+      "epoch": 0.007935296810621089,
+      "grad_norm": 2.6864388610994014,
+      "learning_rate": 2.6262626262626267e-06,
+      "loss": 0.372,
+      "step": 26
+    },
+    {
+      "epoch": 0.008240500534106516,
+      "grad_norm": 3.804357369452407,
+      "learning_rate": 2.7272727272727272e-06,
+      "loss": 0.2646,
+      "step": 27
+    },
+    {
+      "epoch": 0.008545704257591942,
+      "grad_norm": 4.059008227452532,
+      "learning_rate": 2.8282828282828286e-06,
+      "loss": 0.5907,
+      "step": 28
+    },
+    {
+      "epoch": 0.008850907981077369,
+      "grad_norm": 4.9062443629918855,
+      "learning_rate": 2.9292929292929295e-06,
+      "loss": 0.2972,
+      "step": 29
+    },
+    {
+      "epoch": 0.009156111704562795,
+      "grad_norm": 3.5391495380267064,
+      "learning_rate": 3.0303030303030305e-06,
+      "loss": 0.3821,
+      "step": 30
+    },
+    {
+      "epoch": 0.009461315428048222,
+      "grad_norm": 2.5896920322264854,
+      "learning_rate": 3.131313131313132e-06,
+      "loss": 0.4164,
+      "step": 31
+    },
+    {
+      "epoch": 0.009766519151533648,
+      "grad_norm": 3.0230775761822937,
+      "learning_rate": 3.232323232323233e-06,
+      "loss": 0.4237,
+      "step": 32
+    },
+    {
+      "epoch": 0.010071722875019075,
+      "grad_norm": 2.8417717057519423,
+      "learning_rate": 3.3333333333333333e-06,
+      "loss": 0.3353,
+      "step": 33
+    },
+    {
+      "epoch": 0.010376926598504501,
+      "grad_norm": 2.5789157463945878,
+      "learning_rate": 3.4343434343434347e-06,
+      "loss": 0.3769,
+      "step": 34
+    },
+    {
+      "epoch": 0.010682130321989928,
+      "grad_norm": 2.5222241581850096,
+      "learning_rate": 3.5353535353535356e-06,
+      "loss": 0.519,
+      "step": 35
+    },
+    {
+      "epoch": 0.010987334045475354,
+      "grad_norm": 2.8704682168269127,
+      "learning_rate": 3.6363636363636366e-06,
+      "loss": 0.2829,
+      "step": 36
+    },
+    {
+      "epoch": 0.01129253776896078,
+      "grad_norm": 3.24684532820184,
+      "learning_rate": 3.737373737373738e-06,
+      "loss": 0.3586,
+      "step": 37
+    },
+    {
+      "epoch": 0.011597741492446207,
+      "grad_norm": 5.24792475783676,
+      "learning_rate": 3.8383838383838385e-06,
+      "loss": 0.402,
+      "step": 38
+    },
+    {
+      "epoch": 0.011902945215931634,
+      "grad_norm": 3.111184671834165,
+      "learning_rate": 3.93939393939394e-06,
+      "loss": 0.466,
+      "step": 39
+    },
+    {
+      "epoch": 0.01220814893941706,
+      "grad_norm": 3.165565566985893,
+      "learning_rate": 4.04040404040404e-06,
+      "loss": 0.2678,
+      "step": 40
+    },
+    {
+      "epoch": 0.012513352662902488,
+      "grad_norm": 2.5486933296193257,
+      "learning_rate": 4.141414141414142e-06,
+      "loss": 0.5457,
+      "step": 41
+    },
+    {
+      "epoch": 0.012818556386387915,
+      "grad_norm": 3.4373721012250438,
+      "learning_rate": 4.242424242424243e-06,
+      "loss": 0.3862,
+      "step": 42
+    },
+    {
+      "epoch": 0.013123760109873341,
+      "grad_norm": 2.863317221380458,
+      "learning_rate": 4.343434343434344e-06,
+      "loss": 0.3601,
+      "step": 43
+    },
+    {
+      "epoch": 0.013428963833358768,
+      "grad_norm": 2.1041128573446035,
+      "learning_rate": 4.444444444444444e-06,
+      "loss": 0.3693,
+      "step": 44
+    },
+    {
+      "epoch": 0.013734167556844194,
+      "grad_norm": 2.286990324679626,
+      "learning_rate": 4.5454545454545455e-06,
+      "loss": 0.2513,
+      "step": 45
+    },
+    {
+      "epoch": 0.01403937128032962,
+      "grad_norm": 8.793466778432636,
+      "learning_rate": 4.646464646464647e-06,
+      "loss": 0.4343,
+      "step": 46
+    },
+    {
+      "epoch": 0.014344575003815047,
+      "grad_norm": 1.8648737533834159,
+      "learning_rate": 4.747474747474748e-06,
+      "loss": 0.2631,
+      "step": 47
+    },
+    {
+      "epoch": 0.014649778727300474,
+      "grad_norm": 2.3081781364995324,
+      "learning_rate": 4.848484848484849e-06,
+      "loss": 0.2755,
+      "step": 48
+    },
+    {
+      "epoch": 0.0149549824507859,
+      "grad_norm": 2.284005369243557,
+      "learning_rate": 4.94949494949495e-06,
+      "loss": 0.4186,
+      "step": 49
+    },
+    {
+      "epoch": 0.015260186174271327,
+      "grad_norm": 2.6759709423238096,
+      "learning_rate": 5.0505050505050515e-06,
+      "loss": 0.6459,
+      "step": 50
+    },
+    {
+      "epoch": 0.015565389897756753,
+      "grad_norm": 2.8773749120652523,
+      "learning_rate": 5.151515151515152e-06,
+      "loss": 0.3324,
+      "step": 51
+    },
+    {
+      "epoch": 0.015870593621242178,
+      "grad_norm": 2.8060164424498786,
+      "learning_rate": 5.252525252525253e-06,
+      "loss": 0.3608,
+      "step": 52
+    },
+    {
+      "epoch": 0.016175797344727606,
+      "grad_norm": 2.3060494229726793,
+      "learning_rate": 5.353535353535354e-06,
+      "loss": 0.3818,
+      "step": 53
+    },
+    {
+      "epoch": 0.01648100106821303,
+      "grad_norm": 2.073464811557714,
+      "learning_rate": 5.4545454545454545e-06,
+      "loss": 0.2667,
+      "step": 54
+    },
+    {
+      "epoch": 0.01678620479169846,
+      "grad_norm": 2.3474749655399245,
+      "learning_rate": 5.555555555555557e-06,
+      "loss": 0.35,
+      "step": 55
+    },
+    {
+      "epoch": 0.017091408515183884,
+      "grad_norm": 3.6988890036672086,
+      "learning_rate": 5.656565656565657e-06,
+      "loss": 0.284,
+      "step": 56
+    },
+    {
+      "epoch": 0.017396612238669312,
+      "grad_norm": 2.313501192849839,
+      "learning_rate": 5.7575757575757586e-06,
+      "loss": 0.3308,
+      "step": 57
+    },
+    {
+      "epoch": 0.017701815962154737,
+      "grad_norm": 2.411936098122121,
+      "learning_rate": 5.858585858585859e-06,
+      "loss": 0.3982,
+      "step": 58
+    },
+    {
+      "epoch": 0.018007019685640165,
+      "grad_norm": 2.724660127775508,
+      "learning_rate": 5.95959595959596e-06,
+      "loss": 0.3587,
+      "step": 59
+    },
+    {
+      "epoch": 0.01831222340912559,
+      "grad_norm": 3.130895013540925,
+      "learning_rate": 6.060606060606061e-06,
+      "loss": 0.3427,
+      "step": 60
+    },
+    {
+      "epoch": 0.01861742713261102,
+      "grad_norm": 3.4261489723004614,
+      "learning_rate": 6.1616161616161615e-06,
+      "loss": 0.4578,
+      "step": 61
+    },
+    {
+      "epoch": 0.018922630856096443,
+      "grad_norm": 2.413871881063889,
+      "learning_rate": 6.262626262626264e-06,
+      "loss": 0.2067,
+      "step": 62
+    },
+    {
+      "epoch": 0.01922783457958187,
+      "grad_norm": 2.0941348505038366,
+      "learning_rate": 6.363636363636364e-06,
+      "loss": 0.27,
+      "step": 63
+    },
+    {
+      "epoch": 0.019533038303067296,
+      "grad_norm": 2.2153240133926153,
+      "learning_rate": 6.464646464646466e-06,
+      "loss": 0.3298,
+      "step": 64
+    },
+    {
+      "epoch": 0.019838242026552724,
+      "grad_norm": 2.422022070572305,
+      "learning_rate": 6.565656565656566e-06,
+      "loss": 0.4894,
+      "step": 65
+    },
+    {
+      "epoch": 0.02014344575003815,
+      "grad_norm": 2.45442660843552,
+      "learning_rate": 6.666666666666667e-06,
+      "loss": 0.3684,
+      "step": 66
+    },
+    {
+      "epoch": 0.020448649473523577,
+      "grad_norm": 3.5398238081108304,
+      "learning_rate": 6.767676767676769e-06,
+      "loss": 0.4233,
+      "step": 67
+    },
+    {
+      "epoch": 0.020753853197009002,
+      "grad_norm": 2.530397719080883,
+      "learning_rate": 6.868686868686869e-06,
+      "loss": 0.2676,
+      "step": 68
+    },
+    {
+      "epoch": 0.02105905692049443,
+      "grad_norm": 2.259346305696615,
+      "learning_rate": 6.969696969696971e-06,
+      "loss": 0.4409,
+      "step": 69
+    },
+    {
+      "epoch": 0.021364260643979855,
+      "grad_norm": 2.3339543424453764,
+      "learning_rate": 7.070707070707071e-06,
+      "loss": 0.3882,
+      "step": 70
+    },
+    {
+      "epoch": 0.021669464367465283,
+      "grad_norm": 2.348843038116063,
+      "learning_rate": 7.171717171717172e-06,
+      "loss": 0.3904,
+      "step": 71
+    },
+    {
+      "epoch": 0.021974668090950708,
+      "grad_norm": 2.7011363922899965,
+      "learning_rate": 7.272727272727273e-06,
+      "loss": 0.3586,
+      "step": 72
+    },
+    {
+      "epoch": 0.022279871814436136,
+      "grad_norm": 2.6923381814173486,
+      "learning_rate": 7.373737373737374e-06,
+      "loss": 0.4331,
+      "step": 73
+    },
+    {
+      "epoch": 0.02258507553792156,
+      "grad_norm": 2.0435337430530924,
+      "learning_rate": 7.474747474747476e-06,
+      "loss": 0.2739,
+      "step": 74
+    },
+    {
+      "epoch": 0.02289027926140699,
+      "grad_norm": 2.257183264462076,
+      "learning_rate": 7.5757575757575764e-06,
+      "loss": 0.4554,
+      "step": 75
+    },
+    {
+      "epoch": 0.023195482984892414,
+      "grad_norm": 2.5384248372961626,
+      "learning_rate": 7.676767676767677e-06,
+      "loss": 0.4934,
+      "step": 76
+    },
+    {
+      "epoch": 0.023500686708377842,
+      "grad_norm": 2.1578730127908488,
+      "learning_rate": 7.77777777777778e-06,
+      "loss": 0.3519,
+      "step": 77
+    },
+    {
+      "epoch": 0.023805890431863267,
+      "grad_norm": 2.1316764516757476,
+      "learning_rate": 7.87878787878788e-06,
+      "loss": 0.3268,
+      "step": 78
+    },
+    {
+      "epoch": 0.024111094155348695,
+      "grad_norm": 2.095996278024237,
+      "learning_rate": 7.97979797979798e-06,
+      "loss": 0.3318,
+      "step": 79
+    },
+    {
+      "epoch": 0.02441629787883412,
+      "grad_norm": 1.9985574049541877,
+      "learning_rate": 8.08080808080808e-06,
+      "loss": 0.1852,
+      "step": 80
+    },
+    {
+      "epoch": 0.02472150160231955,
+      "grad_norm": 1.7092921737326583,
+      "learning_rate": 8.181818181818183e-06,
+      "loss": 0.2412,
+      "step": 81
+    },
+    {
+      "epoch": 0.025026705325804977,
+      "grad_norm": 1.9609482601524066,
+      "learning_rate": 8.282828282828283e-06,
+      "loss": 0.3349,
+      "step": 82
+    },
+    {
+      "epoch": 0.0253319090492904,
+      "grad_norm": 2.5619254980161412,
+      "learning_rate": 8.383838383838384e-06,
+      "loss": 0.3327,
+      "step": 83
+    },
+    {
+      "epoch": 0.02563711277277583,
+      "grad_norm": 2.1734116421771827,
+      "learning_rate": 8.484848484848486e-06,
+      "loss": 0.5005,
+      "step": 84
+    },
+    {
+      "epoch": 0.025942316496261254,
+      "grad_norm": 2.4612836321871785,
+      "learning_rate": 8.585858585858587e-06,
+      "loss": 0.5919,
+      "step": 85
+    },
+    {
+      "epoch": 0.026247520219746683,
+      "grad_norm": 2.050264187978962,
+      "learning_rate": 8.686868686868687e-06,
+      "loss": 0.2654,
+      "step": 86
+    },
+    {
+      "epoch": 0.026552723943232107,
+      "grad_norm": 1.7466792206761999,
+      "learning_rate": 8.787878787878788e-06,
+      "loss": 0.2875,
+      "step": 87
+    },
+    {
+      "epoch": 0.026857927666717536,
+      "grad_norm": 1.9114055019911376,
+      "learning_rate": 8.888888888888888e-06,
+      "loss": 0.3317,
+      "step": 88
+    },
+    {
+      "epoch": 0.02716313139020296,
+      "grad_norm": 2.136028617695754,
+      "learning_rate": 8.98989898989899e-06,
+      "loss": 0.4322,
+      "step": 89
+    },
+    {
+      "epoch": 0.02746833511368839,
+      "grad_norm": 2.0559196693817303,
+      "learning_rate": 9.090909090909091e-06,
+      "loss": 0.3372,
+      "step": 90
+    },
+    {
+      "epoch": 0.027773538837173813,
+      "grad_norm": 1.6053810559753854,
+      "learning_rate": 9.191919191919193e-06,
+      "loss": 0.2833,
+      "step": 91
+    },
+    {
+      "epoch": 0.02807874256065924,
+      "grad_norm": 1.9190338968500587,
+      "learning_rate": 9.292929292929294e-06,
+      "loss": 0.2358,
+      "step": 92
+    },
+    {
+      "epoch": 0.028383946284144666,
+      "grad_norm": 1.7424429804531956,
+      "learning_rate": 9.393939393939396e-06,
+      "loss": 0.2805,
+      "step": 93
+    },
+    {
+      "epoch": 0.028689150007630095,
+      "grad_norm": 1.5616301594921251,
+      "learning_rate": 9.494949494949497e-06,
+      "loss": 0.326,
+      "step": 94
+    },
+    {
+      "epoch": 0.02899435373111552,
+      "grad_norm": 2.6517363851490297,
+      "learning_rate": 9.595959595959597e-06,
+      "loss": 0.5839,
+      "step": 95
+    },
+    {
+      "epoch": 0.029299557454600948,
+      "grad_norm": 1.9068377479857994,
+      "learning_rate": 9.696969696969698e-06,
+      "loss": 0.4213,
+      "step": 96
+    },
+    {
+      "epoch": 0.029604761178086372,
+      "grad_norm": 2.147263972819766,
+      "learning_rate": 9.797979797979798e-06,
+      "loss": 0.3776,
+      "step": 97
+    },
+    {
+      "epoch": 0.0299099649015718,
+      "grad_norm": 2.3466004395170685,
+      "learning_rate": 9.8989898989899e-06,
+      "loss": 0.4828,
+      "step": 98
+    },
+    {
+      "epoch": 0.030215168625057225,
+      "grad_norm": 1.9328188798162316,
+      "learning_rate": 1e-05,
+      "loss": 0.3816,
+      "step": 99
+    },
+    {
+      "epoch": 0.030520372348542654,
+      "grad_norm": 2.120656679761712,
+      "learning_rate": 9.999997555414177e-06,
+      "loss": 0.287,
+      "step": 100
+    },
+    {
+      "epoch": 0.03082557607202808,
+      "grad_norm": 1.8272767014289886,
+      "learning_rate": 9.999990221659095e-06,
+      "loss": 0.2529,
+      "step": 101
+    },
+    {
+      "epoch": 0.031130779795513507,
+      "grad_norm": 2.108876035097533,
+      "learning_rate": 9.999977998741925e-06,
+      "loss": 0.4,
+      "step": 102
+    },
+    {
+      "epoch": 0.031435983518998935,
+      "grad_norm": 2.611227326027621,
+      "learning_rate": 9.999960886674623e-06,
+      "loss": 0.5577,
+      "step": 103
+    },
+    {
+      "epoch": 0.031741187242484356,
+      "grad_norm": 2.012760226088087,
+      "learning_rate": 9.999938885473916e-06,
+      "loss": 0.2397,
+      "step": 104
+    },
+    {
+      "epoch": 0.032046390965969784,
+      "grad_norm": 3.4069313977643088,
+      "learning_rate": 9.999911995161323e-06,
+      "loss": 0.3074,
+      "step": 105
+    },
+    {
+      "epoch": 0.03235159468945521,
+      "grad_norm": 1.5281487804348939,
+      "learning_rate": 9.999880215763133e-06,
+      "loss": 0.306,
+      "step": 106
+    },
+    {
+      "epoch": 0.03265679841294064,
+      "grad_norm": 1.5733903167529437,
+      "learning_rate": 9.999843547310427e-06,
+      "loss": 0.3123,
+      "step": 107
+    },
+    {
+      "epoch": 0.03296200213642606,
+      "grad_norm": 2.2084260837102776,
+      "learning_rate": 9.999801989839055e-06,
+      "loss": 0.2686,
+      "step": 108
+    },
+    {
+      "epoch": 0.03326720585991149,
+      "grad_norm": 2.0235527329790477,
+      "learning_rate": 9.999755543389658e-06,
+      "loss": 0.362,
+      "step": 109
+    },
+    {
+      "epoch": 0.03357240958339692,
+      "grad_norm": 1.4126246608311444,
+      "learning_rate": 9.999704208007647e-06,
+      "loss": 0.1868,
+      "step": 110
+    },
+    {
+      "epoch": 0.03387761330688235,
+      "grad_norm": 1.9363750145032863,
+      "learning_rate": 9.999647983743227e-06,
+      "loss": 0.4674,
+      "step": 111
+    },
+    {
+      "epoch": 0.03418281703036777,
+      "grad_norm": 2.306492812857686,
+      "learning_rate": 9.999586870651372e-06,
+      "loss": 0.7454,
+      "step": 112
+    },
+    {
+      "epoch": 0.034488020753853196,
+      "grad_norm": 1.9927578577114744,
+      "learning_rate": 9.999520868791839e-06,
+      "loss": 0.2964,
+      "step": 113
+    },
+    {
+      "epoch": 0.034793224477338625,
+      "grad_norm": 2.897230200199283,
+      "learning_rate": 9.99944997822917e-06,
+      "loss": 0.3507,
+      "step": 114
+    },
+    {
+      "epoch": 0.03509842820082405,
+      "grad_norm": 1.7040567211820554,
+      "learning_rate": 9.999374199032682e-06,
+      "loss": 0.358,
+      "step": 115
+    },
+    {
+      "epoch": 0.035403631924309474,
+      "grad_norm": 1.7684725864001616,
+      "learning_rate": 9.999293531276475e-06,
+      "loss": 0.469,
+      "step": 116
+    },
+    {
+      "epoch": 0.0357088356477949,
+      "grad_norm": 2.151331613378997,
+      "learning_rate": 9.999207975039429e-06,
+      "loss": 0.4007,
+      "step": 117
+    },
+    {
+      "epoch": 0.03601403937128033,
+      "grad_norm": 2.1827006415812678,
+      "learning_rate": 9.999117530405205e-06,
+      "loss": 0.373,
+      "step": 118
+    },
+    {
+      "epoch": 0.03631924309476576,
+      "grad_norm": 2.0424756244526283,
+      "learning_rate": 9.99902219746224e-06,
+      "loss": 0.4664,
+      "step": 119
+    },
+    {
+      "epoch": 0.03662444681825118,
+      "grad_norm": 2.4438750213097014,
+      "learning_rate": 9.998921976303757e-06,
+      "loss": 0.5884,
+      "step": 120
+    },
+    {
+      "epoch": 0.03692965054173661,
+      "grad_norm": 1.6168805259489245,
+      "learning_rate": 9.998816867027753e-06,
+      "loss": 0.3874,
+      "step": 121
+    },
+    {
+      "epoch": 0.03723485426522204,
+      "grad_norm": 2.4836564854380914,
+      "learning_rate": 9.99870686973701e-06,
+      "loss": 0.3865,
+      "step": 122
+    },
+    {
+      "epoch": 0.037540057988707465,
+      "grad_norm": 2.187549263535683,
+      "learning_rate": 9.998591984539085e-06,
+      "loss": 0.4419,
+      "step": 123
+    },
+    {
+      "epoch": 0.037845261712192886,
+      "grad_norm": 2.3145724108896366,
+      "learning_rate": 9.998472211546317e-06,
+      "loss": 0.5048,
+      "step": 124
+    },
+    {
+      "epoch": 0.038150465435678314,
+      "grad_norm": 2.6043824271784377,
+      "learning_rate": 9.998347550875825e-06,
+      "loss": 0.4323,
+      "step": 125
+    },
+    {
+      "epoch": 0.03845566915916374,
+      "grad_norm": 1.7266964407358079,
+      "learning_rate": 9.998218002649507e-06,
+      "loss": 0.3093,
+      "step": 126
+    },
+    {
+      "epoch": 0.03876087288264917,
+      "grad_norm": 2.3091863655820397,
+      "learning_rate": 9.99808356699404e-06,
+      "loss": 0.5394,
+      "step": 127
+    },
+    {
+      "epoch": 0.03906607660613459,
+      "grad_norm": 2.178584103245907,
+      "learning_rate": 9.997944244040877e-06,
+      "loss": 0.562,
+      "step": 128
+    },
+    {
+      "epoch": 0.03937128032962002,
+      "grad_norm": 1.4762803065381216,
+      "learning_rate": 9.997800033926252e-06,
+      "loss": 0.3012,
+      "step": 129
+    },
+    {
+      "epoch": 0.03967648405310545,
+      "grad_norm": 1.6768704233807339,
+      "learning_rate": 9.997650936791183e-06,
+      "loss": 0.3314,
+      "step": 130
+    },
+    {
+      "epoch": 0.03998168777659088,
+      "grad_norm": 1.8423584681568375,
+      "learning_rate": 9.997496952781461e-06,
+      "loss": 0.5373,
+      "step": 131
+    },
+    {
+      "epoch": 0.0402868915000763,
+      "grad_norm": 1.4926628434179245,
+      "learning_rate": 9.997338082047656e-06,
+      "loss": 0.1992,
+      "step": 132
+    },
+    {
+      "epoch": 0.040592095223561726,
+      "grad_norm": 1.6323074947028773,
+      "learning_rate": 9.997174324745117e-06,
+      "loss": 0.4872,
+      "step": 133
+    },
+    {
+      "epoch": 0.040897298947047155,
+      "grad_norm": 2.159688005520465,
+      "learning_rate": 9.997005681033973e-06,
+      "loss": 0.5076,
+      "step": 134
+    },
+    {
+      "epoch": 0.04120250267053258,
+      "grad_norm": 2.207163038792008,
+      "learning_rate": 9.996832151079127e-06,
+      "loss": 0.2677,
+      "step": 135
+    },
+    {
+      "epoch": 0.041507706394018004,
+      "grad_norm": 1.3990677420334965,
+      "learning_rate": 9.996653735050265e-06,
+      "loss": 0.2526,
+      "step": 136
+    },
+    {
+      "epoch": 0.04181291011750343,
+      "grad_norm": 1.7368886105229604,
+      "learning_rate": 9.996470433121847e-06,
+      "loss": 0.2874,
+      "step": 137
+    },
+    {
+      "epoch": 0.04211811384098886,
+      "grad_norm": 1.8138446424045762,
+      "learning_rate": 9.996282245473113e-06,
+      "loss": 0.2986,
+      "step": 138
+    },
+    {
+      "epoch": 0.04242331756447429,
+      "grad_norm": 1.8564789601928355,
+      "learning_rate": 9.996089172288078e-06,
+      "loss": 0.3954,
+      "step": 139
+    },
+    {
+      "epoch": 0.04272852128795971,
+      "grad_norm": 1.9085920361180522,
+      "learning_rate": 9.995891213755536e-06,
+      "loss": 0.2739,
+      "step": 140
+    },
+    {
+      "epoch": 0.04303372501144514,
+      "grad_norm": 1.8924678931794556,
+      "learning_rate": 9.99568837006906e-06,
+      "loss": 0.2766,
+      "step": 141
+    },
+    {
+      "epoch": 0.04333892873493057,
+      "grad_norm": 1.8418836037208652,
+      "learning_rate": 9.995480641426992e-06,
+      "loss": 0.488,
+      "step": 142
+    },
+    {
+      "epoch": 0.043644132458415995,
+      "grad_norm": 1.6305125707231247,
+      "learning_rate": 9.99526802803246e-06,
+      "loss": 0.3045,
+      "step": 143
+    },
+    {
+      "epoch": 0.043949336181901416,
+      "grad_norm": 2.143051665423358,
+      "learning_rate": 9.995050530093366e-06,
+      "loss": 0.3567,
+      "step": 144
+    },
+    {
+      "epoch": 0.044254539905386844,
+      "grad_norm": 1.994194545633334,
+      "learning_rate": 9.994828147822387e-06,
+      "loss": 0.3655,
+      "step": 145
+    },
+    {
+      "epoch": 0.04455974362887227,
+      "grad_norm": 1.8553346605537173,
+      "learning_rate": 9.994600881436972e-06,
+      "loss": 0.3249,
+      "step": 146
+    },
+    {
+      "epoch": 0.0448649473523577,
+      "grad_norm": 2.1613773805709857,
+      "learning_rate": 9.994368731159351e-06,
+      "loss": 0.4863,
+      "step": 147
+    },
+    {
+      "epoch": 0.04517015107584312,
+      "grad_norm": 2.199571706523493,
+      "learning_rate": 9.99413169721653e-06,
+      "loss": 0.465,
+      "step": 148
+    },
+    {
+      "epoch": 0.04547535479932855,
+      "grad_norm": 1.681707967900651,
+      "learning_rate": 9.99388977984029e-06,
+      "loss": 0.3472,
+      "step": 149
+    },
+    {
+      "epoch": 0.04578055852281398,
+      "grad_norm": 1.6586587053140593,
+      "learning_rate": 9.993642979267184e-06,
+      "loss": 0.3626,
+      "step": 150
+    },
+    {
+      "epoch": 0.04608576224629941,
+      "grad_norm": 2.12592721793332,
+      "learning_rate": 9.993391295738542e-06,
+      "loss": 0.3218,
+      "step": 151
+    },
+    {
+      "epoch": 0.04639096596978483,
+      "grad_norm": 1.6765944279655143,
+      "learning_rate": 9.99313472950047e-06,
+      "loss": 0.3402,
+      "step": 152
+    },
+    {
+      "epoch": 0.046696169693270256,
+      "grad_norm": 1.6019038139070678,
+      "learning_rate": 9.992873280803848e-06,
+      "loss": 0.4554,
+      "step": 153
+    },
+    {
+      "epoch": 0.047001373416755685,
+      "grad_norm": 1.6429860881882794,
+      "learning_rate": 9.99260694990433e-06,
+      "loss": 0.4086,
+      "step": 154
+    },
+    {
+      "epoch": 0.04730657714024111,
+      "grad_norm": 1.98592334325083,
+      "learning_rate": 9.992335737062338e-06,
+      "loss": 0.5733,
+      "step": 155
+    },
+    {
+      "epoch": 0.047611780863726534,
+      "grad_norm": 1.5624846648417388,
+      "learning_rate": 9.992059642543076e-06,
+      "loss": 0.2524,
+      "step": 156
+    },
+    {
+      "epoch": 0.04791698458721196,
+      "grad_norm": 1.4438198320418865,
+      "learning_rate": 9.991778666616523e-06,
+      "loss": 0.1756,
+      "step": 157
+    },
+    {
+      "epoch": 0.04822218831069739,
+      "grad_norm": 1.6284817295660008,
+      "learning_rate": 9.991492809557424e-06,
+      "loss": 0.4144,
+      "step": 158
+    },
+    {
+      "epoch": 0.04852739203418282,
+      "grad_norm": 1.2236340789910145,
+      "learning_rate": 9.991202071645298e-06,
+      "loss": 0.1664,
+      "step": 159
+    },
+    {
+      "epoch": 0.04883259575766824,
+      "grad_norm": 1.4874398163232816,
+      "learning_rate": 9.99090645316444e-06,
+      "loss": 0.3323,
+      "step": 160
+    },
+    {
+      "epoch": 0.04913779948115367,
+      "grad_norm": 2.5394515927833403,
+      "learning_rate": 9.990605954403917e-06,
+      "loss": 0.27,
+      "step": 161
+    },
+    {
+      "epoch": 0.0494430032046391,
+      "grad_norm": 1.7966332314422868,
+      "learning_rate": 9.990300575657565e-06,
+      "loss": 0.4453,
+      "step": 162
+    },
+    {
+      "epoch": 0.049748206928124525,
+      "grad_norm": 1.825976682624809,
+      "learning_rate": 9.989990317223995e-06,
+      "loss": 0.2646,
+      "step": 163
+    },
+    {
+      "epoch": 0.05005341065160995,
+      "grad_norm": 1.6554541925183588,
+      "learning_rate": 9.989675179406588e-06,
+      "loss": 0.445,
+      "step": 164
+    },
+    {
+      "epoch": 0.050358614375095374,
+      "grad_norm": 1.6711133844293076,
+      "learning_rate": 9.989355162513496e-06,
+      "loss": 0.3685,
+      "step": 165
+    },
+    {
+      "epoch": 0.0506638180985808,
+      "grad_norm": 1.8033315345252203,
+      "learning_rate": 9.989030266857644e-06,
+      "loss": 0.2566,
+      "step": 166
+    },
+    {
+      "epoch": 0.05096902182206623,
+      "grad_norm": 1.6879852444966537,
+      "learning_rate": 9.988700492756726e-06,
+      "loss": 0.4086,
+      "step": 167
+    },
+    {
+      "epoch": 0.05127422554555166,
+      "grad_norm": 1.6855038740169574,
+      "learning_rate": 9.988365840533204e-06,
+      "loss": 0.3081,
+      "step": 168
+    },
+    {
+      "epoch": 0.05157942926903708,
+      "grad_norm": 2.245121010490438,
+      "learning_rate": 9.988026310514316e-06,
+      "loss": 0.5646,
+      "step": 169
+    },
+    {
+      "epoch": 0.05188463299252251,
+      "grad_norm": 1.531117336209479,
+      "learning_rate": 9.987681903032065e-06,
+      "loss": 0.3598,
+      "step": 170
+    },
+    {
+      "epoch": 0.05218983671600794,
+      "grad_norm": 1.4368727600956301,
+      "learning_rate": 9.987332618423221e-06,
+      "loss": 0.3864,
+      "step": 171
+    },
+    {
+      "epoch": 0.052495040439493365,
+      "grad_norm": 2.039026486601271,
+      "learning_rate": 9.98697845702933e-06,
+      "loss": 0.2728,
+      "step": 172
+    },
+    {
+      "epoch": 0.052800244162978786,
+      "grad_norm": 1.5481974795842472,
+      "learning_rate": 9.986619419196704e-06,
+      "loss": 0.2376,
+      "step": 173
+    },
+    {
+      "epoch": 0.053105447886464215,
+      "grad_norm": 1.583025735121783,
+      "learning_rate": 9.986255505276418e-06,
+      "loss": 0.3941,
+      "step": 174
+    },
+    {
+      "epoch": 0.05341065160994964,
+      "grad_norm": 2.025610033619695,
+      "learning_rate": 9.985886715624326e-06,
+      "loss": 0.432,
+      "step": 175
+    },
+    {
+      "epoch": 0.05371585533343507,
+      "grad_norm": 1.9370365819159912,
+      "learning_rate": 9.985513050601037e-06,
+      "loss": 0.3311,
+      "step": 176
+    },
+    {
+      "epoch": 0.05402105905692049,
+      "grad_norm": 1.534591376747653,
+      "learning_rate": 9.985134510571936e-06,
+      "loss": 0.3804,
+      "step": 177
+    },
+    {
+      "epoch": 0.05432626278040592,
+      "grad_norm": 1.5627980520171343,
+      "learning_rate": 9.984751095907175e-06,
+      "loss": 0.3991,
+      "step": 178
+    },
+    {
+      "epoch": 0.05463146650389135,
+      "grad_norm": 1.858760828475349,
+      "learning_rate": 9.984362806981665e-06,
+      "loss": 0.4124,
+      "step": 179
+    },
+    {
+      "epoch": 0.05493667022737678,
+      "grad_norm": 1.4922057145689682,
+      "learning_rate": 9.983969644175092e-06,
+      "loss": 0.2571,
+      "step": 180
+    },
+    {
+      "epoch": 0.0552418739508622,
+      "grad_norm": 1.4358215484460224,
+      "learning_rate": 9.983571607871903e-06,
+      "loss": 0.3351,
+      "step": 181
+    },
+    {
+      "epoch": 0.05554707767434763,
+      "grad_norm": 1.7105120125454414,
+      "learning_rate": 9.983168698461312e-06,
+      "loss": 0.4374,
+      "step": 182
+    },
+    {
+      "epoch": 0.055852281397833055,
+      "grad_norm": 1.4100459259074987,
+      "learning_rate": 9.982760916337296e-06,
+      "loss": 0.3958,
+      "step": 183
+    },
+    {
+      "epoch": 0.05615748512131848,
+      "grad_norm": 1.667173817085955,
+      "learning_rate": 9.982348261898598e-06,
+      "loss": 0.2867,
+      "step": 184
+    },
+    {
+      "epoch": 0.056462688844803904,
+      "grad_norm": 1.8278737995984025,
+      "learning_rate": 9.981930735548731e-06,
+      "loss": 0.3738,
+      "step": 185
+    },
+    {
+      "epoch": 0.05676789256828933,
+      "grad_norm": 1.806852289121097,
+      "learning_rate": 9.98150833769596e-06,
+      "loss": 0.5608,
+      "step": 186
+    },
+    {
+      "epoch": 0.05707309629177476,
+      "grad_norm": 1.6986308867720055,
+      "learning_rate": 9.981081068753324e-06,
+      "loss": 0.4253,
+      "step": 187
+    },
+    {
+      "epoch": 0.05737830001526019,
+      "grad_norm": 1.6392088091109513,
+      "learning_rate": 9.98064892913862e-06,
+      "loss": 0.2444,
+      "step": 188
+    },
+    {
+      "epoch": 0.05768350373874561,
+      "grad_norm": 1.7762995408711126,
+      "learning_rate": 9.980211919274407e-06,
+      "loss": 0.3866,
+      "step": 189
+    },
+    {
+      "epoch": 0.05798870746223104,
+      "grad_norm": 1.7144647062044762,
+      "learning_rate": 9.979770039588013e-06,
+      "loss": 0.4504,
+      "step": 190
+    },
+    {
+      "epoch": 0.05829391118571647,
+      "grad_norm": 1.9069269572943617,
+      "learning_rate": 9.979323290511517e-06,
+      "loss": 0.4972,
+      "step": 191
+    },
+    {
+      "epoch": 0.058599114909201895,
+      "grad_norm": 1.831943664409223,
+      "learning_rate": 9.978871672481774e-06,
+      "loss": 0.3884,
+      "step": 192
+    },
+    {
+      "epoch": 0.058904318632687316,
+      "grad_norm": 1.60483584957947,
+      "learning_rate": 9.978415185940383e-06,
+      "loss": 0.3366,
+      "step": 193
+    },
+    {
+      "epoch": 0.059209522356172745,
+      "grad_norm": 2.041633475935638,
+      "learning_rate": 9.977953831333718e-06,
+      "loss": 0.4928,
+      "step": 194
+    },
+    {
+      "epoch": 0.05951472607965817,
+      "grad_norm": 2.1574861604284243,
+      "learning_rate": 9.977487609112904e-06,
+      "loss": 0.7092,
+      "step": 195
+    },
+    {
+      "epoch": 0.0598199298031436,
+      "grad_norm": 1.5382345073334531,
+      "learning_rate": 9.97701651973383e-06,
+      "loss": 0.2236,
+      "step": 196
+    },
+    {
+      "epoch": 0.06012513352662902,
+      "grad_norm": 2.1479787995768014,
+      "learning_rate": 9.976540563657143e-06,
+      "loss": 0.5182,
+      "step": 197
+    },
+    {
+      "epoch": 0.06043033725011445,
+      "grad_norm": 1.8579437774142544,
+      "learning_rate": 9.976059741348252e-06,
+      "loss": 0.3093,
+      "step": 198
+    },
+    {
+      "epoch": 0.06073554097359988,
+      "grad_norm": 1.5409701380525285,
+      "learning_rate": 9.975574053277317e-06,
+      "loss": 0.2877,
+      "step": 199
+    },
+    {
+      "epoch": 0.06104074469708531,
+      "grad_norm": 1.5474598097011698,
+      "learning_rate": 9.975083499919264e-06,
+      "loss": 0.2981,
+      "step": 200
+    },
+    {
+      "epoch": 0.06134594842057073,
+      "grad_norm": 1.9202152932180157,
+      "learning_rate": 9.974588081753773e-06,
+      "loss": 0.5369,
+      "step": 201
+    },
+    {
+      "epoch": 0.06165115214405616,
+      "grad_norm": 1.4598442515817716,
+      "learning_rate": 9.974087799265279e-06,
+      "loss": 0.3696,
+      "step": 202
+    },
+    {
+      "epoch": 0.061956355867541585,
+      "grad_norm": 1.48078814360119,
+      "learning_rate": 9.973582652942975e-06,
+      "loss": 0.284,
+      "step": 203
+    },
+    {
+      "epoch": 0.06226155959102701,
+      "grad_norm": 2.100326004155181,
+      "learning_rate": 9.973072643280813e-06,
+      "loss": 0.5681,
+      "step": 204
+    },
+    {
+      "epoch": 0.06256676331451244,
+      "grad_norm": 1.976128330719915,
+      "learning_rate": 9.972557770777496e-06,
+      "loss": 0.3655,
+      "step": 205
+    },
+    {
+      "epoch": 0.06287196703799787,
+      "grad_norm": 1.2103730393566896,
+      "learning_rate": 9.972038035936483e-06,
+      "loss": 0.2471,
+      "step": 206
+    },
+    {
+      "epoch": 0.06317717076148328,
+      "grad_norm": 1.670449906238349,
+      "learning_rate": 9.971513439265992e-06,
+      "loss": 0.2184,
+      "step": 207
+    },
+    {
+      "epoch": 0.06348237448496871,
+      "grad_norm": 1.5020544764497652,
+      "learning_rate": 9.970983981278989e-06,
+      "loss": 0.3196,
+      "step": 208
+    },
+    {
+      "epoch": 0.06378757820845414,
+      "grad_norm": 1.7833251911345853,
+      "learning_rate": 9.970449662493195e-06,
+      "loss": 0.4122,
+      "step": 209
+    },
+    {
+      "epoch": 0.06409278193193957,
+      "grad_norm": 1.4149595334362772,
+      "learning_rate": 9.96991048343109e-06,
+      "loss": 0.2947,
+      "step": 210
+    },
+    {
+      "epoch": 0.064397985655425,
+      "grad_norm": 1.5991867680932033,
+      "learning_rate": 9.969366444619898e-06,
+      "loss": 0.1902,
+      "step": 211
+    },
+    {
+      "epoch": 0.06470318937891043,
+      "grad_norm": 1.4132064841734169,
+      "learning_rate": 9.968817546591601e-06,
+      "loss": 0.3389,
+      "step": 212
+    },
+    {
+      "epoch": 0.06500839310239585,
+      "grad_norm": 1.7671902900221814,
+      "learning_rate": 9.968263789882926e-06,
+      "loss": 0.4294,
+      "step": 213
+    },
+    {
+      "epoch": 0.06531359682588128,
+      "grad_norm": 1.5709821497329826,
+      "learning_rate": 9.96770517503536e-06,
+      "loss": 0.2765,
+      "step": 214
+    },
+    {
+      "epoch": 0.0656188005493667,
+      "grad_norm": 1.5211731343844295,
+      "learning_rate": 9.967141702595134e-06,
+      "loss": 0.387,
+      "step": 215
+    },
+    {
+      "epoch": 0.06592400427285212,
+      "grad_norm": 1.5499265222668686,
+      "learning_rate": 9.96657337311323e-06,
+      "loss": 0.4535,
+      "step": 216
+    },
+    {
+      "epoch": 0.06622920799633755,
+      "grad_norm": 1.4736546539447488,
+      "learning_rate": 9.966000187145383e-06,
+      "loss": 0.3834,
+      "step": 217
+    },
+    {
+      "epoch": 0.06653441171982298,
+      "grad_norm": 1.3306288958233108,
+      "learning_rate": 9.965422145252072e-06,
+      "loss": 0.3172,
+      "step": 218
+    },
+    {
+      "epoch": 0.06683961544330841,
+      "grad_norm": 1.5745937005003143,
+      "learning_rate": 9.964839247998524e-06,
+      "loss": 0.2725,
+      "step": 219
+    },
+    {
+      "epoch": 0.06714481916679384,
+      "grad_norm": 1.7546511557153388,
+      "learning_rate": 9.96425149595472e-06,
+      "loss": 0.3577,
+      "step": 220
+    },
+    {
+      "epoch": 0.06745002289027927,
+      "grad_norm": 2.0422588449754286,
+      "learning_rate": 9.96365888969538e-06,
+      "loss": 0.4976,
+      "step": 221
+    },
+    {
+      "epoch": 0.0677552266137647,
+      "grad_norm": 1.4661824124133862,
+      "learning_rate": 9.963061429799979e-06,
+      "loss": 0.3672,
+      "step": 222
+    },
+    {
+      "epoch": 0.06806043033725011,
+      "grad_norm": 2.0959067552369666,
+      "learning_rate": 9.96245911685273e-06,
+      "loss": 0.5381,
+      "step": 223
+    },
+    {
+      "epoch": 0.06836563406073554,
+      "grad_norm": 1.3296813372997014,
+      "learning_rate": 9.961851951442599e-06,
+      "loss": 0.2799,
+      "step": 224
+    },
+    {
+      "epoch": 0.06867083778422096,
+      "grad_norm": 1.7385807765114274,
+      "learning_rate": 9.96123993416329e-06,
+      "loss": 0.5183,
+      "step": 225
+    },
+    {
+      "epoch": 0.06897604150770639,
+      "grad_norm": 1.5190119701865645,
+      "learning_rate": 9.960623065613254e-06,
+      "loss": 0.4608,
+      "step": 226
+    },
+    {
+      "epoch": 0.06928124523119182,
+      "grad_norm": 1.4393894383331207,
+      "learning_rate": 9.96000134639569e-06,
+      "loss": 0.3455,
+      "step": 227
+    },
+    {
+      "epoch": 0.06958644895467725,
+      "grad_norm": 1.7132863682619555,
+      "learning_rate": 9.959374777118533e-06,
+      "loss": 0.316,
+      "step": 228
+    },
+    {
+      "epoch": 0.06989165267816268,
+      "grad_norm": 1.3227120889592454,
+      "learning_rate": 9.958743358394464e-06,
+      "loss": 0.2467,
+      "step": 229
+    },
+    {
+      "epoch": 0.0701968564016481,
+      "grad_norm": 1.5331153407144422,
+      "learning_rate": 9.95810709084091e-06,
+      "loss": 0.3138,
+      "step": 230
+    },
+    {
+      "epoch": 0.07050206012513352,
+      "grad_norm": 1.7990748995190806,
+      "learning_rate": 9.957465975080031e-06,
+      "loss": 0.4747,
+      "step": 231
+    },
+    {
+      "epoch": 0.07080726384861895,
+      "grad_norm": 1.1638981235859056,
+      "learning_rate": 9.956820011738736e-06,
+      "loss": 0.2265,
+      "step": 232
+    },
+    {
+      "epoch": 0.07111246757210438,
+      "grad_norm": 1.5739388418179414,
+      "learning_rate": 9.956169201448665e-06,
+      "loss": 0.5066,
+      "step": 233
+    },
+    {
+      "epoch": 0.0714176712955898,
+      "grad_norm": 1.6803933013620869,
+      "learning_rate": 9.955513544846205e-06,
+      "loss": 0.4415,
+      "step": 234
+    },
+    {
+      "epoch": 0.07172287501907523,
+      "grad_norm": 1.4014872110785643,
+      "learning_rate": 9.954853042572479e-06,
+      "loss": 0.3271,
+      "step": 235
+    },
+    {
+      "epoch": 0.07202807874256066,
+      "grad_norm": 1.5310222689941932,
+      "learning_rate": 9.954187695273352e-06,
+      "loss": 0.3289,
+      "step": 236
+    },
+    {
+      "epoch": 0.07233328246604609,
+      "grad_norm": 2.166268226472017,
+      "learning_rate": 9.953517503599419e-06,
+      "loss": 0.622,
+      "step": 237
+    },
+    {
+      "epoch": 0.07263848618953152,
+      "grad_norm": 2.258081862277545,
+      "learning_rate": 9.952842468206019e-06,
+      "loss": 0.5071,
+      "step": 238
+    },
+    {
+      "epoch": 0.07294368991301693,
+      "grad_norm": 1.7322119894263104,
+      "learning_rate": 9.952162589753224e-06,
+      "loss": 0.5097,
+      "step": 239
+    },
+    {
+      "epoch": 0.07324889363650236,
+      "grad_norm": 1.9966284228033864,
+      "learning_rate": 9.951477868905843e-06,
+      "loss": 0.2263,
+      "step": 240
+    },
+    {
+      "epoch": 0.07355409735998779,
+      "grad_norm": 1.6793267860774614,
+      "learning_rate": 9.95078830633342e-06,
+      "loss": 0.2065,
+      "step": 241
+    },
+    {
+      "epoch": 0.07385930108347322,
+      "grad_norm": 2.122564153881175,
+      "learning_rate": 9.95009390271023e-06,
+      "loss": 0.2665,
+      "step": 242
+    },
+    {
+      "epoch": 0.07416450480695864,
+      "grad_norm": 1.5852282963187305,
+      "learning_rate": 9.949394658715289e-06,
+      "loss": 0.4453,
+      "step": 243
+    },
+    {
+      "epoch": 0.07446970853044407,
+      "grad_norm": 1.7534712016120517,
+      "learning_rate": 9.948690575032338e-06,
+      "loss": 0.3628,
+      "step": 244
+    },
+    {
+      "epoch": 0.0747749122539295,
+      "grad_norm": 1.351810586905304,
+      "learning_rate": 9.947981652349854e-06,
+      "loss": 0.3984,
+      "step": 245
+    },
+    {
+      "epoch": 0.07508011597741493,
+      "grad_norm": 1.8377506474408298,
+      "learning_rate": 9.947267891361051e-06,
+      "loss": 0.3677,
+      "step": 246
+    },
+    {
+      "epoch": 0.07538531970090036,
+      "grad_norm": 1.4655632998364951,
+      "learning_rate": 9.946549292763865e-06,
+      "loss": 0.3516,
+      "step": 247
+    },
+    {
+      "epoch": 0.07569052342438577,
+      "grad_norm": 3.240838121636416,
+      "learning_rate": 9.945825857260967e-06,
+      "loss": 0.2627,
+      "step": 248
+    },
+    {
+      "epoch": 0.0759957271478712,
+      "grad_norm": 1.4085823215183912,
+      "learning_rate": 9.945097585559757e-06,
+      "loss": 0.2716,
+      "step": 249
+    },
+    {
+      "epoch": 0.07630093087135663,
+      "grad_norm": 1.6361471921651585,
+      "learning_rate": 9.944364478372364e-06,
+      "loss": 0.3595,
+      "step": 250
+    },
+    {
+      "epoch": 0.07660613459484206,
+      "grad_norm": 1.0912978886499554,
+      "learning_rate": 9.943626536415647e-06,
+      "loss": 0.1968,
+      "step": 251
+    },
+    {
+      "epoch": 0.07691133831832749,
+      "grad_norm": 1.9515717700893849,
+      "learning_rate": 9.942883760411188e-06,
+      "loss": 0.374,
+      "step": 252
+    },
+    {
+      "epoch": 0.07721654204181291,
+      "grad_norm": 1.5560755068838334,
+      "learning_rate": 9.942136151085302e-06,
+      "loss": 0.44,
+      "step": 253
+    },
+    {
+      "epoch": 0.07752174576529834,
+      "grad_norm": 1.4843235207715992,
+      "learning_rate": 9.941383709169024e-06,
+      "loss": 0.3175,
+      "step": 254
+    },
+    {
+      "epoch": 0.07782694948878377,
+      "grad_norm": 1.5210960196158274,
+      "learning_rate": 9.94062643539812e-06,
+      "loss": 0.3722,
+      "step": 255
+    },
+    {
+      "epoch": 0.07813215321226918,
+      "grad_norm": 1.6656094376801425,
+      "learning_rate": 9.939864330513079e-06,
+      "loss": 0.3511,
+      "step": 256
+    },
+    {
+      "epoch": 0.07843735693575461,
+      "grad_norm": 1.2732857455769802,
+      "learning_rate": 9.939097395259108e-06,
+      "loss": 0.2619,
+      "step": 257
+    },
+    {
+      "epoch": 0.07874256065924004,
+      "grad_norm": 1.8947301386622588,
+      "learning_rate": 9.938325630386149e-06,
+      "loss": 0.3933,
+      "step": 258
+    },
+    {
+      "epoch": 0.07904776438272547,
+      "grad_norm": 1.5625416559388712,
+      "learning_rate": 9.937549036648857e-06,
+      "loss": 0.4491,
+      "step": 259
+    },
+    {
+      "epoch": 0.0793529681062109,
+      "grad_norm": 1.5125179888703784,
+      "learning_rate": 9.936767614806612e-06,
+      "loss": 0.3674,
+      "step": 260
+    },
+    {
+      "epoch": 0.07965817182969633,
+      "grad_norm": 1.5026525250547669,
+      "learning_rate": 9.935981365623516e-06,
+      "loss": 0.4103,
+      "step": 261
+    },
+    {
+      "epoch": 0.07996337555318175,
+      "grad_norm": 2.3948536293362115,
+      "learning_rate": 9.93519028986839e-06,
+      "loss": 0.4009,
+      "step": 262
+    },
+    {
+      "epoch": 0.08026857927666718,
+      "grad_norm": 2.416554371647352,
+      "learning_rate": 9.934394388314775e-06,
+      "loss": 0.4265,
+      "step": 263
+    },
+    {
+      "epoch": 0.0805737830001526,
+      "grad_norm": 1.560923734953618,
+      "learning_rate": 9.933593661740933e-06,
+      "loss": 0.303,
+      "step": 264
+    },
+    {
+      "epoch": 0.08087898672363802,
+      "grad_norm": 1.6053945705234087,
+      "learning_rate": 9.932788110929837e-06,
+      "loss": 0.3295,
+      "step": 265
+    },
+    {
+      "epoch": 0.08118419044712345,
+      "grad_norm": 1.7775437462596928,
+      "learning_rate": 9.931977736669185e-06,
+      "loss": 0.2197,
+      "step": 266
+    },
+    {
+      "epoch": 0.08148939417060888,
+      "grad_norm": 1.701318325041301,
+      "learning_rate": 9.931162539751392e-06,
+      "loss": 0.3581,
+      "step": 267
+    },
+    {
+      "epoch": 0.08179459789409431,
+      "grad_norm": 1.5974548511363529,
+      "learning_rate": 9.93034252097358e-06,
+      "loss": 0.3432,
+      "step": 268
+    },
+    {
+      "epoch": 0.08209980161757974,
+      "grad_norm": 1.8669593065073864,
+      "learning_rate": 9.929517681137594e-06,
+      "loss": 0.4133,
+      "step": 269
+    },
+    {
+      "epoch": 0.08240500534106517,
+      "grad_norm": 1.4895827642408586,
+      "learning_rate": 9.928688021049991e-06,
+      "loss": 0.3111,
+      "step": 270
+    },
+    {
+      "epoch": 0.0827102090645506,
+      "grad_norm": 1.4317804244871846,
+      "learning_rate": 9.927853541522041e-06,
+      "loss": 0.2915,
+      "step": 271
+    },
+    {
+      "epoch": 0.08301541278803601,
+      "grad_norm": 1.252478145781798,
+      "learning_rate": 9.927014243369727e-06,
+      "loss": 0.2794,
+      "step": 272
+    },
+    {
+      "epoch": 0.08332061651152144,
+      "grad_norm": 1.6973954865497314,
+      "learning_rate": 9.926170127413743e-06,
+      "loss": 0.6183,
+      "step": 273
+    },
+    {
+      "epoch": 0.08362582023500686,
+      "grad_norm": 1.4723277244112698,
+      "learning_rate": 9.925321194479494e-06,
+      "loss": 0.2815,
+      "step": 274
+    },
+    {
+      "epoch": 0.08393102395849229,
+      "grad_norm": 1.7075555550514414,
+      "learning_rate": 9.924467445397097e-06,
+      "loss": 0.4178,
+      "step": 275
+    },
+    {
+      "epoch": 0.08423622768197772,
+      "grad_norm": 1.5354808046910606,
+      "learning_rate": 9.923608881001377e-06,
+      "loss": 0.2355,
+      "step": 276
+    },
+    {
+      "epoch": 0.08454143140546315,
+      "grad_norm": 1.1795750747565834,
+      "learning_rate": 9.922745502131865e-06,
+      "loss": 0.3404,
+      "step": 277
+    },
+    {
+      "epoch": 0.08484663512894858,
+      "grad_norm": 1.427067758888222,
+      "learning_rate": 9.921877309632805e-06,
+      "loss": 0.3141,
+      "step": 278
+    },
+    {
+      "epoch": 0.085151838852434,
+      "grad_norm": 1.3691564278772157,
+      "learning_rate": 9.921004304353147e-06,
+      "loss": 0.287,
+      "step": 279
+    },
+    {
+      "epoch": 0.08545704257591942,
+      "grad_norm": 1.9220775714586407,
+      "learning_rate": 9.920126487146544e-06,
+      "loss": 0.6617,
+      "step": 280
+    },
+    {
+      "epoch": 0.08576224629940485,
+      "grad_norm": 1.6761030408371134,
+      "learning_rate": 9.919243858871355e-06,
+      "loss": 0.466,
+      "step": 281
+    },
+    {
+      "epoch": 0.08606745002289028,
+      "grad_norm": 1.6120747264173168,
+      "learning_rate": 9.918356420390645e-06,
+      "loss": 0.5351,
+      "step": 282
+    },
+    {
+      "epoch": 0.0863726537463757,
+      "grad_norm": 1.5236961732014556,
+      "learning_rate": 9.91746417257218e-06,
+      "loss": 0.33,
+      "step": 283
+    },
+    {
+      "epoch": 0.08667785746986113,
+      "grad_norm": 1.6328635321860312,
+      "learning_rate": 9.916567116288434e-06,
+      "loss": 0.4301,
+      "step": 284
+    },
+    {
+      "epoch": 0.08698306119334656,
+      "grad_norm": 1.4120804188821041,
+      "learning_rate": 9.915665252416577e-06,
+      "loss": 0.3025,
+      "step": 285
+    },
+    {
+      "epoch": 0.08728826491683199,
+      "grad_norm": 1.8410843798908767,
+      "learning_rate": 9.914758581838482e-06,
+      "loss": 0.5415,
+      "step": 286
+    },
+    {
+      "epoch": 0.08759346864031742,
+      "grad_norm": 1.1807475096034001,
+      "learning_rate": 9.913847105440725e-06,
+      "loss": 0.3184,
+      "step": 287
+    },
+    {
+      "epoch": 0.08789867236380283,
+      "grad_norm": 1.52681276111022,
+      "learning_rate": 9.912930824114577e-06,
+      "loss": 0.4266,
+      "step": 288
+    },
+    {
+      "epoch": 0.08820387608728826,
+      "grad_norm": 1.4904538614169496,
+      "learning_rate": 9.91200973875601e-06,
+      "loss": 0.3404,
+      "step": 289
+    },
+    {
+      "epoch": 0.08850907981077369,
+      "grad_norm": 1.7385111110311349,
+      "learning_rate": 9.911083850265692e-06,
+      "loss": 0.3371,
+      "step": 290
+    },
+    {
+      "epoch": 0.08881428353425912,
+      "grad_norm": 1.6013762575114376,
+      "learning_rate": 9.91015315954899e-06,
+      "loss": 0.4475,
+      "step": 291
+    },
+    {
+      "epoch": 0.08911948725774455,
+      "grad_norm": 1.5474202900018152,
+      "learning_rate": 9.909217667515964e-06,
+      "loss": 0.4162,
+      "step": 292
+    },
+    {
+      "epoch": 0.08942469098122997,
+      "grad_norm": 1.875769203080621,
+      "learning_rate": 9.908277375081371e-06,
+      "loss": 0.4446,
+      "step": 293
+    },
+    {
+      "epoch": 0.0897298947047154,
+      "grad_norm": 1.4914731218024286,
+      "learning_rate": 9.907332283164663e-06,
+      "loss": 0.4274,
+      "step": 294
+    },
+    {
+      "epoch": 0.09003509842820083,
+      "grad_norm": 1.6551811079983538,
+      "learning_rate": 9.90638239268998e-06,
+      "loss": 0.4883,
+      "step": 295
+    },
+    {
+      "epoch": 0.09034030215168624,
+      "grad_norm": 1.645510927644492,
+      "learning_rate": 9.905427704586158e-06,
+      "loss": 0.4885,
+      "step": 296
+    },
+    {
+      "epoch": 0.09064550587517167,
+      "grad_norm": 1.6759165462483547,
+      "learning_rate": 9.904468219786727e-06,
+      "loss": 0.3878,
+      "step": 297
+    },
+    {
+      "epoch": 0.0909507095986571,
+      "grad_norm": 1.596800484010474,
+      "learning_rate": 9.903503939229901e-06,
+      "loss": 0.2725,
+      "step": 298
+    },
+    {
+      "epoch": 0.09125591332214253,
+      "grad_norm": 1.4035704196730787,
+      "learning_rate": 9.902534863858588e-06,
+      "loss": 0.2147,
+      "step": 299
+    },
+    {
+      "epoch": 0.09156111704562796,
+      "grad_norm": 1.7460761357385464,
+      "learning_rate": 9.90156099462038e-06,
+      "loss": 0.3495,
+      "step": 300
+    },
+    {
+      "epoch": 0.09186632076911339,
+      "grad_norm": 1.3373562156184522,
+      "learning_rate": 9.900582332467566e-06,
+      "loss": 0.342,
+      "step": 301
+    },
+    {
+      "epoch": 0.09217152449259881,
+      "grad_norm": 1.1466755748188362,
+      "learning_rate": 9.89959887835711e-06,
+      "loss": 0.1737,
+      "step": 302
+    },
+    {
+      "epoch": 0.09247672821608424,
+      "grad_norm": 1.8078659273922337,
+      "learning_rate": 9.898610633250669e-06,
+      "loss": 0.3111,
+      "step": 303
+    },
+    {
+      "epoch": 0.09278193193956966,
+      "grad_norm": 1.5400638324339648,
+      "learning_rate": 9.897617598114584e-06,
+      "loss": 0.4746,
+      "step": 304
+    },
+    {
+      "epoch": 0.09308713566305508,
+      "grad_norm": 1.558728128630052,
+      "learning_rate": 9.896619773919878e-06,
+      "loss": 0.3085,
+      "step": 305
+    },
+    {
+      "epoch": 0.09339233938654051,
+      "grad_norm": 4.094736926672729,
+      "learning_rate": 9.895617161642257e-06,
+      "loss": 0.4664,
+      "step": 306
+    },
+    {
+      "epoch": 0.09369754311002594,
+      "grad_norm": 1.63116898024897,
+      "learning_rate": 9.89460976226211e-06,
+      "loss": 0.3878,
+      "step": 307
+    },
+    {
+      "epoch": 0.09400274683351137,
+      "grad_norm": 1.7238364123731507,
+      "learning_rate": 9.893597576764508e-06,
+      "loss": 0.2989,
+      "step": 308
+    },
+    {
+      "epoch": 0.0943079505569968,
+      "grad_norm": 1.2496662648050174,
+      "learning_rate": 9.8925806061392e-06,
+      "loss": 0.3054,
+      "step": 309
+    },
+    {
+      "epoch": 0.09461315428048223,
+      "grad_norm": 0.8807197003313585,
+      "learning_rate": 9.891558851380614e-06,
+      "loss": 0.1904,
+      "step": 310
+    },
+    {
+      "epoch": 0.09491835800396765,
+      "grad_norm": 1.5076918479598347,
+      "learning_rate": 9.890532313487858e-06,
+      "loss": 0.2679,
+      "step": 311
+    },
+    {
+      "epoch": 0.09522356172745307,
+      "grad_norm": 1.8465691043660122,
+      "learning_rate": 9.889500993464716e-06,
+      "loss": 0.5002,
+      "step": 312
+    },
+    {
+      "epoch": 0.0955287654509385,
+      "grad_norm": 1.9183643810942494,
+      "learning_rate": 9.888464892319647e-06,
+      "loss": 0.4869,
+      "step": 313
+    },
+    {
+      "epoch": 0.09583396917442392,
+      "grad_norm": 1.6515373264151805,
+      "learning_rate": 9.887424011065788e-06,
+      "loss": 0.4507,
+      "step": 314
+    },
+    {
+      "epoch": 0.09613917289790935,
+      "grad_norm": 1.6223391241834122,
+      "learning_rate": 9.886378350720945e-06,
+      "loss": 0.3445,
+      "step": 315
+    },
+    {
+      "epoch": 0.09644437662139478,
+      "grad_norm": 1.4416645097808285,
+      "learning_rate": 9.885327912307604e-06,
+      "loss": 0.2808,
+      "step": 316
+    },
+    {
+      "epoch": 0.09674958034488021,
+      "grad_norm": 1.4777192121308136,
+      "learning_rate": 9.88427269685292e-06,
+      "loss": 0.4335,
+      "step": 317
+    },
+    {
+      "epoch": 0.09705478406836564,
+      "grad_norm": 1.6934694740555867,
+      "learning_rate": 9.883212705388715e-06,
+      "loss": 0.4299,
+      "step": 318
+    },
+    {
+      "epoch": 0.09735998779185107,
+      "grad_norm": 1.9031284601590377,
+      "learning_rate": 9.882147938951489e-06,
+      "loss": 0.5364,
+      "step": 319
+    },
+    {
+      "epoch": 0.09766519151533648,
+      "grad_norm": 1.990035566558448,
+      "learning_rate": 9.881078398582406e-06,
+      "loss": 0.6476,
+      "step": 320
+    },
+    {
+      "epoch": 0.09797039523882191,
+      "grad_norm": 1.4458600630840748,
+      "learning_rate": 9.8800040853273e-06,
+      "loss": 0.268,
+      "step": 321
+    },
+    {
+      "epoch": 0.09827559896230734,
+      "grad_norm": 1.473557254783057,
+      "learning_rate": 9.878925000236667e-06,
+      "loss": 0.3889,
+      "step": 322
+    },
+    {
+      "epoch": 0.09858080268579276,
+      "grad_norm": 1.429462352597184,
+      "learning_rate": 9.877841144365681e-06,
+      "loss": 0.3348,
+      "step": 323
+    },
+    {
+      "epoch": 0.0988860064092782,
+      "grad_norm": 1.9126483909533352,
+      "learning_rate": 9.876752518774167e-06,
+      "loss": 0.5004,
+      "step": 324
+    },
+    {
+      "epoch": 0.09919121013276362,
+      "grad_norm": 1.528278815830415,
+      "learning_rate": 9.875659124526622e-06,
+      "loss": 0.1931,
+      "step": 325
+    },
+    {
+      "epoch": 0.09949641385624905,
+      "grad_norm": 1.6064809314060318,
+      "learning_rate": 9.874560962692207e-06,
+      "loss": 0.2627,
+      "step": 326
+    },
+    {
+      "epoch": 0.09980161757973448,
+      "grad_norm": 1.8583002911468363,
+      "learning_rate": 9.873458034344741e-06,
+      "loss": 0.4795,
+      "step": 327
+    },
+    {
+      "epoch": 0.1001068213032199,
+      "grad_norm": 2.180040993961252,
+      "learning_rate": 9.872350340562704e-06,
+      "loss": 0.3502,
+      "step": 328
+    },
+    {
+      "epoch": 0.10041202502670532,
+      "grad_norm": 2.2760944374886334,
+      "learning_rate": 9.871237882429237e-06,
+      "loss": 0.5504,
+      "step": 329
+    },
+    {
+      "epoch": 0.10071722875019075,
+      "grad_norm": 1.599604903553732,
+      "learning_rate": 9.87012066103214e-06,
+      "loss": 0.4642,
+      "step": 330
+    },
+    {
+      "epoch": 0.10102243247367618,
+      "grad_norm": 1.165592902920284,
+      "learning_rate": 9.868998677463874e-06,
+      "loss": 0.2118,
+      "step": 331
+    },
+    {
+      "epoch": 0.1013276361971616,
+      "grad_norm": 1.562687671834839,
+      "learning_rate": 9.867871932821549e-06,
+      "loss": 0.3389,
+      "step": 332
+    },
+    {
+      "epoch": 0.10163283992064703,
+      "grad_norm": 1.5690938291630006,
+      "learning_rate": 9.866740428206935e-06,
+      "loss": 0.3084,
+      "step": 333
+    },
+    {
+      "epoch": 0.10193804364413246,
+      "grad_norm": 1.6392469885959746,
+      "learning_rate": 9.865604164726456e-06,
+      "loss": 0.3935,
+      "step": 334
+    },
+    {
+      "epoch": 0.10224324736761789,
+      "grad_norm": 1.4807871775535164,
+      "learning_rate": 9.864463143491192e-06,
+      "loss": 0.4445,
+      "step": 335
+    },
+    {
+      "epoch": 0.10254845109110332,
+      "grad_norm": 1.1926217304533164,
+      "learning_rate": 9.86331736561687e-06,
+      "loss": 0.2623,
+      "step": 336
+    },
+    {
+      "epoch": 0.10285365481458873,
+      "grad_norm": 1.5461262133304665,
+      "learning_rate": 9.862166832223871e-06,
+      "loss": 0.4161,
+      "step": 337
+    },
+    {
+      "epoch": 0.10315885853807416,
+      "grad_norm": 1.5135662608063911,
+      "learning_rate": 9.861011544437226e-06,
+      "loss": 0.2864,
+      "step": 338
+    },
+    {
+      "epoch": 0.10346406226155959,
+      "grad_norm": 1.2723458882899108,
+      "learning_rate": 9.85985150338662e-06,
+      "loss": 0.3208,
+      "step": 339
+    },
+    {
+      "epoch": 0.10376926598504502,
+      "grad_norm": 1.5555768560283307,
+      "learning_rate": 9.858686710206373e-06,
+      "loss": 0.3341,
+      "step": 340
+    },
+    {
+      "epoch": 0.10407446970853045,
+      "grad_norm": 1.7562240544367693,
+      "learning_rate": 9.857517166035466e-06,
+      "loss": 0.4637,
+      "step": 341
+    },
+    {
+      "epoch": 0.10437967343201587,
+      "grad_norm": 3.9759170817857044,
+      "learning_rate": 9.856342872017515e-06,
+      "loss": 0.6559,
+      "step": 342
+    },
+    {
+      "epoch": 0.1046848771555013,
+      "grad_norm": 1.6864581247305628,
+      "learning_rate": 9.855163829300789e-06,
+      "loss": 0.4889,
+      "step": 343
+    },
+    {
+      "epoch": 0.10499008087898673,
+      "grad_norm": 1.5660173118774432,
+      "learning_rate": 9.853980039038193e-06,
+      "loss": 0.2512,
+      "step": 344
+    },
+    {
+      "epoch": 0.10529528460247214,
+      "grad_norm": 1.747919617181712,
+      "learning_rate": 9.85279150238728e-06,
+      "loss": 0.4796,
+      "step": 345
+    },
+    {
+      "epoch": 0.10560048832595757,
+      "grad_norm": 2.0064830609007496,
+      "learning_rate": 9.85159822051024e-06,
+      "loss": 0.6876,
+      "step": 346
+    },
+    {
+      "epoch": 0.105905692049443,
+      "grad_norm": 3.232705832740987,
+      "learning_rate": 9.850400194573908e-06,
+      "loss": 0.5043,
+      "step": 347
+    },
+    {
+      "epoch": 0.10621089577292843,
+      "grad_norm": 1.6750365694523466,
+      "learning_rate": 9.849197425749753e-06,
+      "loss": 0.4426,
+      "step": 348
+    },
+    {
+      "epoch": 0.10651609949641386,
+      "grad_norm": 1.2407353388150502,
+      "learning_rate": 9.847989915213883e-06,
+      "loss": 0.2728,
+      "step": 349
+    },
+    {
+      "epoch": 0.10682130321989929,
+      "grad_norm": 1.145770915094197,
+      "learning_rate": 9.846777664147046e-06,
+      "loss": 0.1828,
+      "step": 350
+    },
+    {
+      "epoch": 0.10712650694338471,
+      "grad_norm": 1.0259526919339048,
+      "learning_rate": 9.845560673734617e-06,
+      "loss": 0.1815,
+      "step": 351
+    },
+    {
+      "epoch": 0.10743171066687014,
+      "grad_norm": 1.5927868344920564,
+      "learning_rate": 9.844338945166619e-06,
+      "loss": 0.4563,
+      "step": 352
+    },
+    {
+      "epoch": 0.10773691439035556,
+      "grad_norm": 1.862511817843432,
+      "learning_rate": 9.843112479637692e-06,
+      "loss": 0.4873,
+      "step": 353
+    },
+    {
+      "epoch": 0.10804211811384098,
+      "grad_norm": 1.3346380863318066,
+      "learning_rate": 9.841881278347122e-06,
+      "loss": 0.344,
+      "step": 354
+    },
+    {
+      "epoch": 0.10834732183732641,
+      "grad_norm": 1.2558059799685006,
+      "learning_rate": 9.840645342498817e-06,
+      "loss": 0.3489,
+      "step": 355
+    },
+    {
+      "epoch": 0.10865252556081184,
+      "grad_norm": 1.734496310003891,
+      "learning_rate": 9.839404673301318e-06,
+      "loss": 0.4161,
+      "step": 356
+    },
+    {
+      "epoch": 0.10895772928429727,
+      "grad_norm": 1.5538063811233147,
+      "learning_rate": 9.838159271967795e-06,
+      "loss": 0.4971,
+      "step": 357
+    },
+    {
+      "epoch": 0.1092629330077827,
+      "grad_norm": 1.51967394631939,
+      "learning_rate": 9.836909139716044e-06,
+      "loss": 0.3422,
+      "step": 358
+    },
+    {
+      "epoch": 0.10956813673126813,
+      "grad_norm": 1.8153898943745974,
+      "learning_rate": 9.835654277768487e-06,
+      "loss": 0.5826,
+      "step": 359
+    },
+    {
+      "epoch": 0.10987334045475355,
+      "grad_norm": 1.3956832615282058,
+      "learning_rate": 9.834394687352168e-06,
+      "loss": 0.3444,
+      "step": 360
+    },
+    {
+      "epoch": 0.11017854417823897,
+      "grad_norm": 1.7028097158357791,
+      "learning_rate": 9.833130369698764e-06,
+      "loss": 0.4653,
+      "step": 361
+    },
+    {
+      "epoch": 0.1104837479017244,
+      "grad_norm": 1.984086744142622,
+      "learning_rate": 9.831861326044564e-06,
+      "loss": 0.7268,
+      "step": 362
+    },
+    {
+      "epoch": 0.11078895162520982,
+      "grad_norm": 1.6010962575202456,
+      "learning_rate": 9.830587557630481e-06,
+      "loss": 0.4979,
+      "step": 363
+    },
+    {
+      "epoch": 0.11109415534869525,
+      "grad_norm": 1.163967768763629,
+      "learning_rate": 9.829309065702054e-06,
+      "loss": 0.2721,
+      "step": 364
+    },
+    {
+      "epoch": 0.11139935907218068,
+      "grad_norm": 1.5116526665647354,
+      "learning_rate": 9.828025851509433e-06,
+      "loss": 0.4504,
+      "step": 365
+    },
+    {
+      "epoch": 0.11170456279566611,
+      "grad_norm": 1.3126613426949747,
+      "learning_rate": 9.82673791630739e-06,
+      "loss": 0.2915,
+      "step": 366
+    },
+    {
+      "epoch": 0.11200976651915154,
+      "grad_norm": 1.2919896962667308,
+      "learning_rate": 9.825445261355313e-06,
+      "loss": 0.2862,
+      "step": 367
+    },
+    {
+      "epoch": 0.11231497024263697,
+      "grad_norm": 1.7465438774619029,
+      "learning_rate": 9.824147887917201e-06,
+      "loss": 0.5347,
+      "step": 368
+    },
+    {
+      "epoch": 0.11262017396612238,
+      "grad_norm": 1.5226131458552274,
+      "learning_rate": 9.822845797261676e-06,
+      "loss": 0.4764,
+      "step": 369
+    },
+    {
+      "epoch": 0.11292537768960781,
+      "grad_norm": 1.157991129368488,
+      "learning_rate": 9.82153899066196e-06,
+      "loss": 0.2002,
+      "step": 370
+    },
+    {
+      "epoch": 0.11323058141309324,
+      "grad_norm": 0.9301400886155634,
+      "learning_rate": 9.8202274693959e-06,
+      "loss": 0.1684,
+      "step": 371
+    },
+    {
+      "epoch": 0.11353578513657867,
+      "grad_norm": 1.502487900950231,
+      "learning_rate": 9.818911234745942e-06,
+      "loss": 0.3093,
+      "step": 372
+    },
+    {
+      "epoch": 0.1138409888600641,
+      "grad_norm": 1.9784385010881411,
+      "learning_rate": 9.817590287999149e-06,
+      "loss": 0.4803,
+      "step": 373
+    },
+    {
+      "epoch": 0.11414619258354952,
+      "grad_norm": 0.9195545540370847,
+      "learning_rate": 9.816264630447186e-06,
+      "loss": 0.2159,
+      "step": 374
+    },
+    {
+      "epoch": 0.11445139630703495,
+      "grad_norm": 1.9332444000616873,
+      "learning_rate": 9.814934263386324e-06,
+      "loss": 0.704,
+      "step": 375
+    },
+    {
+      "epoch": 0.11475660003052038,
+      "grad_norm": 1.2982271338494382,
+      "learning_rate": 9.813599188117447e-06,
+      "loss": 0.2579,
+      "step": 376
+    },
+    {
+      "epoch": 0.11506180375400579,
+      "grad_norm": 1.4400207145843447,
+      "learning_rate": 9.812259405946033e-06,
+      "loss": 0.2581,
+      "step": 377
+    },
+    {
+      "epoch": 0.11536700747749122,
+      "grad_norm": 1.6214016909138647,
+      "learning_rate": 9.810914918182168e-06,
+      "loss": 0.3348,
+      "step": 378
+    },
+    {
+      "epoch": 0.11567221120097665,
+      "grad_norm": 1.3037284021243603,
+      "learning_rate": 9.80956572614054e-06,
+      "loss": 0.2009,
+      "step": 379
+    },
+    {
+      "epoch": 0.11597741492446208,
+      "grad_norm": 1.4597999101205241,
+      "learning_rate": 9.808211831140434e-06,
+      "loss": 0.4417,
+      "step": 380
+    },
+    {
+      "epoch": 0.1162826186479475,
+      "grad_norm": 1.3133108268461724,
+      "learning_rate": 9.806853234505736e-06,
+      "loss": 0.3305,
+      "step": 381
+    },
+    {
+      "epoch": 0.11658782237143293,
+      "grad_norm": 1.455145466929522,
+      "learning_rate": 9.805489937564926e-06,
+      "loss": 0.4611,
+      "step": 382
+    },
+    {
+      "epoch": 0.11689302609491836,
+      "grad_norm": 1.3560358010182432,
+      "learning_rate": 9.804121941651085e-06,
+      "loss": 0.2624,
+      "step": 383
+    },
+    {
+      "epoch": 0.11719822981840379,
+      "grad_norm": 1.5730489544680661,
+      "learning_rate": 9.802749248101885e-06,
+      "loss": 0.5959,
+      "step": 384
+    },
+    {
+      "epoch": 0.1175034335418892,
+      "grad_norm": 1.6080011966995038,
+      "learning_rate": 9.801371858259594e-06,
+      "loss": 0.3077,
+      "step": 385
+    },
+    {
+      "epoch": 0.11780863726537463,
+      "grad_norm": 1.3452073998773173,
+      "learning_rate": 9.799989773471071e-06,
+      "loss": 0.3877,
+      "step": 386
+    },
+    {
+      "epoch": 0.11811384098886006,
+      "grad_norm": 1.521531541360139,
+      "learning_rate": 9.798602995087764e-06,
+      "loss": 0.2978,
+      "step": 387
+    },
+    {
+      "epoch": 0.11841904471234549,
+      "grad_norm": 1.705542362457564,
+      "learning_rate": 9.797211524465715e-06,
+      "loss": 0.4298,
+      "step": 388
+    },
+    {
+      "epoch": 0.11872424843583092,
+      "grad_norm": 2.110762528312395,
+      "learning_rate": 9.79581536296555e-06,
+      "loss": 0.4766,
+      "step": 389
+    },
+    {
+      "epoch": 0.11902945215931635,
+      "grad_norm": 1.8442245131411212,
+      "learning_rate": 9.794414511952484e-06,
+      "loss": 0.2988,
+      "step": 390
+    },
+    {
+      "epoch": 0.11933465588280177,
+      "grad_norm": 1.2950053698748747,
+      "learning_rate": 9.793008972796318e-06,
+      "loss": 0.2907,
+      "step": 391
+    },
+    {
+      "epoch": 0.1196398596062872,
+      "grad_norm": 1.1692504994324242,
+      "learning_rate": 9.791598746871438e-06,
+      "loss": 0.2584,
+      "step": 392
+    },
+    {
+      "epoch": 0.11994506332977262,
+      "grad_norm": 1.7847769770487698,
+      "learning_rate": 9.790183835556806e-06,
+      "loss": 0.4874,
+      "step": 393
+    },
+    {
+      "epoch": 0.12025026705325804,
+      "grad_norm": 1.2562325547558533,
+      "learning_rate": 9.788764240235976e-06,
+      "loss": 0.3739,
+      "step": 394
+    },
+    {
+      "epoch": 0.12055547077674347,
+      "grad_norm": 1.340465345857484,
+      "learning_rate": 9.787339962297076e-06,
+      "loss": 0.1911,
+      "step": 395
+    },
+    {
+      "epoch": 0.1208606745002289,
+      "grad_norm": 1.5667496682899116,
+      "learning_rate": 9.785911003132811e-06,
+      "loss": 0.4243,
+      "step": 396
+    },
+    {
+      "epoch": 0.12116587822371433,
+      "grad_norm": 1.81118055527733,
+      "learning_rate": 9.78447736414047e-06,
+      "loss": 0.5246,
+      "step": 397
+    },
+    {
+      "epoch": 0.12147108194719976,
+      "grad_norm": 1.4053388398772566,
+      "learning_rate": 9.783039046721912e-06,
+      "loss": 0.2964,
+      "step": 398
+    },
+    {
+      "epoch": 0.12177628567068519,
+      "grad_norm": 1.5253960954907508,
+      "learning_rate": 9.781596052283573e-06,
+      "loss": 0.3939,
+      "step": 399
+    },
+    {
+      "epoch": 0.12208148939417061,
+      "grad_norm": 1.4250712259795966,
+      "learning_rate": 9.780148382236465e-06,
+      "loss": 0.2338,
+      "step": 400
+    },
+    {
+      "epoch": 0.12238669311765603,
+      "grad_norm": 1.6445234212439845,
+      "learning_rate": 9.778696037996167e-06,
+      "loss": 0.6218,
+      "step": 401
+    },
+    {
+      "epoch": 0.12269189684114146,
+      "grad_norm": 1.625655408880631,
+      "learning_rate": 9.777239020982834e-06,
+      "loss": 0.4565,
+      "step": 402
+    },
+    {
+      "epoch": 0.12299710056462689,
+      "grad_norm": 1.2734470484088918,
+      "learning_rate": 9.775777332621184e-06,
+      "loss": 0.3673,
+      "step": 403
+    },
+    {
+      "epoch": 0.12330230428811231,
+      "grad_norm": 1.5775928166525761,
+      "learning_rate": 9.774310974340506e-06,
+      "loss": 0.3673,
+      "step": 404
+    },
+    {
+      "epoch": 0.12360750801159774,
+      "grad_norm": 1.2252627733176171,
+      "learning_rate": 9.772839947574658e-06,
+      "loss": 0.4064,
+      "step": 405
+    },
+    {
+      "epoch": 0.12391271173508317,
+      "grad_norm": 1.7189364245488037,
+      "learning_rate": 9.77136425376206e-06,
+      "loss": 0.4633,
+      "step": 406
+    },
+    {
+      "epoch": 0.1242179154585686,
+      "grad_norm": 1.7558994988767551,
+      "learning_rate": 9.769883894345693e-06,
+      "loss": 0.5655,
+      "step": 407
+    },
+    {
+      "epoch": 0.12452311918205403,
+      "grad_norm": 1.4802133948901528,
+      "learning_rate": 9.768398870773108e-06,
+      "loss": 0.4175,
+      "step": 408
+    },
+    {
+      "epoch": 0.12482832290553945,
+      "grad_norm": 1.3302751078782091,
+      "learning_rate": 9.766909184496408e-06,
+      "loss": 0.3468,
+      "step": 409
+    },
+    {
+      "epoch": 0.12513352662902488,
+      "grad_norm": 1.704721594870751,
+      "learning_rate": 9.765414836972262e-06,
+      "loss": 0.4898,
+      "step": 410
+    },
+    {
+      "epoch": 0.1254387303525103,
+      "grad_norm": 1.2131375907547226,
+      "learning_rate": 9.763915829661891e-06,
+      "loss": 0.3886,
+      "step": 411
+    },
+    {
+      "epoch": 0.12574393407599574,
+      "grad_norm": 1.3401723742285034,
+      "learning_rate": 9.76241216403108e-06,
+      "loss": 0.4102,
+      "step": 412
+    },
+    {
+      "epoch": 0.12604913779948115,
+      "grad_norm": 1.6513004120286505,
+      "learning_rate": 9.760903841550164e-06,
+      "loss": 0.433,
+      "step": 413
+    },
+    {
+      "epoch": 0.12635434152296657,
+      "grad_norm": 2.2390646637016327,
+      "learning_rate": 9.75939086369403e-06,
+      "loss": 0.4213,
+      "step": 414
+    },
+    {
+      "epoch": 0.126659545246452,
+      "grad_norm": 1.585281461881457,
+      "learning_rate": 9.757873231942122e-06,
+      "loss": 0.2145,
+      "step": 415
+    },
+    {
+      "epoch": 0.12696474896993742,
+      "grad_norm": 1.616278056338578,
+      "learning_rate": 9.756350947778431e-06,
+      "loss": 0.3786,
+      "step": 416
+    },
+    {
+      "epoch": 0.12726995269342287,
+      "grad_norm": 1.3896314522086528,
+      "learning_rate": 9.754824012691499e-06,
+      "loss": 0.3886,
+      "step": 417
+    },
+    {
+      "epoch": 0.12757515641690828,
+      "grad_norm": 1.5388381717065547,
+      "learning_rate": 9.753292428174416e-06,
+      "loss": 0.4901,
+      "step": 418
+    },
+    {
+      "epoch": 0.12788036014039372,
+      "grad_norm": 1.2099556172325527,
+      "learning_rate": 9.75175619572482e-06,
+      "loss": 0.2379,
+      "step": 419
+    },
+    {
+      "epoch": 0.12818556386387914,
+      "grad_norm": 2.578673148403812,
+      "learning_rate": 9.750215316844886e-06,
+      "loss": 0.3093,
+      "step": 420
+    },
+    {
+      "epoch": 0.12849076758736458,
+      "grad_norm": 1.3557950817043143,
+      "learning_rate": 9.748669793041345e-06,
+      "loss": 0.2901,
+      "step": 421
+    },
+    {
+      "epoch": 0.12879597131085,
+      "grad_norm": 1.8058846206435177,
+      "learning_rate": 9.747119625825459e-06,
+      "loss": 0.5222,
+      "step": 422
+    },
+    {
+      "epoch": 0.1291011750343354,
+      "grad_norm": 1.449090093366816,
+      "learning_rate": 9.745564816713034e-06,
+      "loss": 0.3738,
+      "step": 423
+    },
+    {
+      "epoch": 0.12940637875782085,
+      "grad_norm": 1.3181509799847857,
+      "learning_rate": 9.74400536722442e-06,
+      "loss": 0.2624,
+      "step": 424
+    },
+    {
+      "epoch": 0.12971158248130626,
+      "grad_norm": 2.1599699357377675,
+      "learning_rate": 9.742441278884496e-06,
+      "loss": 0.4838,
+      "step": 425
+    },
+    {
+      "epoch": 0.1300167862047917,
+      "grad_norm": 1.6258096878519581,
+      "learning_rate": 9.740872553222685e-06,
+      "loss": 0.4999,
+      "step": 426
+    },
+    {
+      "epoch": 0.13032198992827712,
+      "grad_norm": 1.5046387312958875,
+      "learning_rate": 9.739299191772937e-06,
+      "loss": 0.3095,
+      "step": 427
+    },
+    {
+      "epoch": 0.13062719365176256,
+      "grad_norm": 3.2565177704876653,
+      "learning_rate": 9.737721196073742e-06,
+      "loss": 0.4886,
+      "step": 428
+    },
+    {
+      "epoch": 0.13093239737524798,
+      "grad_norm": 5.072257540244327,
+      "learning_rate": 9.736138567668117e-06,
+      "loss": 0.2433,
+      "step": 429
+    },
+    {
+      "epoch": 0.1312376010987334,
+      "grad_norm": 1.5984597272623309,
+      "learning_rate": 9.734551308103607e-06,
+      "loss": 0.3274,
+      "step": 430
+    },
+    {
+      "epoch": 0.13154280482221883,
+      "grad_norm": 1.9766310728941525,
+      "learning_rate": 9.732959418932297e-06,
+      "loss": 0.5601,
+      "step": 431
+    },
+    {
+      "epoch": 0.13184800854570425,
+      "grad_norm": 1.8500535183975242,
+      "learning_rate": 9.731362901710783e-06,
+      "loss": 0.8055,
+      "step": 432
+    },
+    {
+      "epoch": 0.1321532122691897,
+      "grad_norm": 1.5082490121163843,
+      "learning_rate": 9.7297617580002e-06,
+      "loss": 0.3715,
+      "step": 433
+    },
+    {
+      "epoch": 0.1324584159926751,
+      "grad_norm": 1.3484742708817405,
+      "learning_rate": 9.728155989366198e-06,
+      "loss": 0.2316,
+      "step": 434
+    },
+    {
+      "epoch": 0.13276361971616055,
+      "grad_norm": 1.4064281774599339,
+      "learning_rate": 9.726545597378953e-06,
+      "loss": 0.2529,
+      "step": 435
+    },
+    {
+      "epoch": 0.13306882343964596,
+      "grad_norm": 1.3619428913058238,
+      "learning_rate": 9.724930583613164e-06,
+      "loss": 0.2675,
+      "step": 436
+    },
+    {
+      "epoch": 0.1333740271631314,
+      "grad_norm": 1.8072201350967787,
+      "learning_rate": 9.723310949648044e-06,
+      "loss": 0.4832,
+      "step": 437
+    },
+    {
+      "epoch": 0.13367923088661682,
+      "grad_norm": 1.376837369774298,
+      "learning_rate": 9.721686697067328e-06,
+      "loss": 0.3855,
+      "step": 438
+    },
+    {
+      "epoch": 0.13398443461010223,
+      "grad_norm": 1.577084491012061,
+      "learning_rate": 9.720057827459264e-06,
+      "loss": 0.3307,
+      "step": 439
+    },
+    {
+      "epoch": 0.13428963833358767,
+      "grad_norm": 1.7335930513549882,
+      "learning_rate": 9.718424342416619e-06,
+      "loss": 0.3794,
+      "step": 440
+    },
+    {
+      "epoch": 0.1345948420570731,
+      "grad_norm": 1.3979759520207362,
+      "learning_rate": 9.716786243536672e-06,
+      "loss": 0.17,
+      "step": 441
+    },
+    {
+      "epoch": 0.13490004578055853,
+      "grad_norm": 1.5166020341173607,
+      "learning_rate": 9.715143532421208e-06,
+      "loss": 0.2782,
+      "step": 442
+    },
+    {
+      "epoch": 0.13520524950404395,
+      "grad_norm": 1.5547284451560954,
+      "learning_rate": 9.71349621067653e-06,
+      "loss": 0.2398,
+      "step": 443
+    },
+    {
+      "epoch": 0.1355104532275294,
+      "grad_norm": 1.4706416022210265,
+      "learning_rate": 9.711844279913443e-06,
+      "loss": 0.3486,
+      "step": 444
+    },
+    {
+      "epoch": 0.1358156569510148,
+      "grad_norm": 1.7726813901729983,
+      "learning_rate": 9.710187741747264e-06,
+      "loss": 0.4938,
+      "step": 445
+    },
+    {
+      "epoch": 0.13612086067450022,
+      "grad_norm": 1.3682846898358798,
+      "learning_rate": 9.708526597797812e-06,
+      "loss": 0.3399,
+      "step": 446
+    },
+    {
+      "epoch": 0.13642606439798566,
+      "grad_norm": 1.173206742476641,
+      "learning_rate": 9.70686084968941e-06,
+      "loss": 0.209,
+      "step": 447
+    },
+    {
+      "epoch": 0.13673126812147107,
+      "grad_norm": 1.5006667880365117,
+      "learning_rate": 9.705190499050885e-06,
+      "loss": 0.3918,
+      "step": 448
+    },
+    {
+      "epoch": 0.13703647184495651,
+      "grad_norm": 1.4081281623516984,
+      "learning_rate": 9.70351554751556e-06,
+      "loss": 0.3478,
+      "step": 449
+    },
+    {
+      "epoch": 0.13734167556844193,
+      "grad_norm": 1.3521857616183288,
+      "learning_rate": 9.701835996721267e-06,
+      "loss": 0.2527,
+      "step": 450
+    },
+    {
+      "epoch": 0.13764687929192737,
+      "grad_norm": 1.609701105059192,
+      "learning_rate": 9.70015184831032e-06,
+      "loss": 0.4019,
+      "step": 451
+    },
+    {
+      "epoch": 0.13795208301541279,
+      "grad_norm": 1.5247656094886828,
+      "learning_rate": 9.698463103929542e-06,
+      "loss": 0.3393,
+      "step": 452
+    },
+    {
+      "epoch": 0.13825728673889823,
+      "grad_norm": 1.7521263555747508,
+      "learning_rate": 9.696769765230244e-06,
+      "loss": 0.4711,
+      "step": 453
+    },
+    {
+      "epoch": 0.13856249046238364,
+      "grad_norm": 1.1750952661731386,
+      "learning_rate": 9.695071833868233e-06,
+      "loss": 0.3209,
+      "step": 454
+    },
+    {
+      "epoch": 0.13886769418586906,
+      "grad_norm": 1.1097718724387464,
+      "learning_rate": 9.693369311503801e-06,
+      "loss": 0.2077,
+      "step": 455
+    },
+    {
+      "epoch": 0.1391728979093545,
+      "grad_norm": 1.3477077308441543,
+      "learning_rate": 9.691662199801735e-06,
+      "loss": 0.3387,
+      "step": 456
+    },
+    {
+      "epoch": 0.1394781016328399,
+      "grad_norm": 1.5926509134118427,
+      "learning_rate": 9.689950500431306e-06,
+      "loss": 0.4737,
+      "step": 457
+    },
+    {
+      "epoch": 0.13978330535632535,
+      "grad_norm": 1.238821976588628,
+      "learning_rate": 9.688234215066274e-06,
+      "loss": 0.2941,
+      "step": 458
+    },
+    {
+      "epoch": 0.14008850907981077,
+      "grad_norm": 1.283103486116252,
+      "learning_rate": 9.68651334538488e-06,
+      "loss": 0.2298,
+      "step": 459
+    },
+    {
+      "epoch": 0.1403937128032962,
+      "grad_norm": 1.3964480095243228,
+      "learning_rate": 9.684787893069852e-06,
+      "loss": 0.2755,
+      "step": 460
+    },
+    {
+      "epoch": 0.14069891652678163,
+      "grad_norm": 4.5768637107742185,
+      "learning_rate": 9.683057859808394e-06,
+      "loss": 0.3969,
+      "step": 461
+    },
+    {
+      "epoch": 0.14100412025026704,
+      "grad_norm": 1.238974584705195,
+      "learning_rate": 9.681323247292193e-06,
+      "loss": 0.302,
+      "step": 462
+    },
+    {
+      "epoch": 0.14130932397375248,
+      "grad_norm": 1.484367852444757,
+      "learning_rate": 9.679584057217412e-06,
+      "loss": 0.3771,
+      "step": 463
+    },
+    {
+      "epoch": 0.1416145276972379,
+      "grad_norm": 1.8892302609803566,
+      "learning_rate": 9.677840291284693e-06,
+      "loss": 0.4296,
+      "step": 464
+    },
+    {
+      "epoch": 0.14191973142072334,
+      "grad_norm": 1.9264735588835349,
+      "learning_rate": 9.676091951199147e-06,
+      "loss": 0.1789,
+      "step": 465
+    },
+    {
+      "epoch": 0.14222493514420875,
+      "grad_norm": 1.6779527799580054,
+      "learning_rate": 9.674339038670362e-06,
+      "loss": 0.3394,
+      "step": 466
+    },
+    {
+      "epoch": 0.1425301388676942,
+      "grad_norm": 1.4269803792009585,
+      "learning_rate": 9.672581555412396e-06,
+      "loss": 0.3436,
+      "step": 467
+    },
+    {
+      "epoch": 0.1428353425911796,
+      "grad_norm": 1.6997561953129157,
+      "learning_rate": 9.67081950314378e-06,
+      "loss": 0.4244,
+      "step": 468
+    },
+    {
+      "epoch": 0.14314054631466505,
+      "grad_norm": 1.5847658235861504,
+      "learning_rate": 9.669052883587503e-06,
+      "loss": 0.4061,
+      "step": 469
+    },
+    {
+      "epoch": 0.14344575003815047,
+      "grad_norm": 1.5000132719088555,
+      "learning_rate": 9.667281698471032e-06,
+      "loss": 0.2349,
+      "step": 470
+    },
+    {
+      "epoch": 0.14375095376163588,
+      "grad_norm": 1.4700398564287758,
+      "learning_rate": 9.665505949526288e-06,
+      "loss": 0.4265,
+      "step": 471
+    },
+    {
+      "epoch": 0.14405615748512132,
+      "grad_norm": 1.1929108671224367,
+      "learning_rate": 9.663725638489662e-06,
+      "loss": 0.3107,
+      "step": 472
+    },
+    {
+      "epoch": 0.14436136120860674,
+      "grad_norm": 1.1577689014622543,
+      "learning_rate": 9.661940767102001e-06,
+      "loss": 0.1722,
+      "step": 473
+    },
+    {
+      "epoch": 0.14466656493209218,
+      "grad_norm": 2.093772204836488,
+      "learning_rate": 9.660151337108617e-06,
+      "loss": 0.4278,
+      "step": 474
+    },
+    {
+      "epoch": 0.1449717686555776,
+      "grad_norm": 1.6654665758895533,
+      "learning_rate": 9.658357350259274e-06,
+      "loss": 0.3394,
+      "step": 475
+    },
+    {
+      "epoch": 0.14527697237906304,
+      "grad_norm": 1.2202451289258887,
+      "learning_rate": 9.656558808308193e-06,
+      "loss": 0.3433,
+      "step": 476
+    },
+    {
+      "epoch": 0.14558217610254845,
+      "grad_norm": 1.5418473816537024,
+      "learning_rate": 9.654755713014052e-06,
+      "loss": 0.4099,
+      "step": 477
+    },
+    {
+      "epoch": 0.14588737982603386,
+      "grad_norm": 1.867172943441559,
+      "learning_rate": 9.652948066139978e-06,
+      "loss": 0.4445,
+      "step": 478
+    },
+    {
+      "epoch": 0.1461925835495193,
+      "grad_norm": 1.2274320399473075,
+      "learning_rate": 9.651135869453552e-06,
+      "loss": 0.3687,
+      "step": 479
+    },
+    {
+      "epoch": 0.14649778727300472,
+      "grad_norm": 1.3632570280967384,
+      "learning_rate": 9.649319124726799e-06,
+      "loss": 0.3554,
+      "step": 480
+    },
+    {
+      "epoch": 0.14680299099649016,
+      "grad_norm": 1.9368328668689925,
+      "learning_rate": 9.647497833736197e-06,
+      "loss": 0.433,
+      "step": 481
+    },
+    {
+      "epoch": 0.14710819471997558,
+      "grad_norm": 1.7750440538339176,
+      "learning_rate": 9.645671998262668e-06,
+      "loss": 0.4275,
+      "step": 482
+    },
+    {
+      "epoch": 0.14741339844346102,
+      "grad_norm": 1.7620981827052555,
+      "learning_rate": 9.643841620091572e-06,
+      "loss": 0.4327,
+      "step": 483
+    },
+    {
+      "epoch": 0.14771860216694643,
+      "grad_norm": 1.705808013569081,
+      "learning_rate": 9.642006701012719e-06,
+      "loss": 0.2914,
+      "step": 484
+    },
+    {
+      "epoch": 0.14802380589043188,
+      "grad_norm": 1.4490666476731855,
+      "learning_rate": 9.640167242820356e-06,
+      "loss": 0.4171,
+      "step": 485
+    },
+    {
+      "epoch": 0.1483290096139173,
+      "grad_norm": 1.4017898188106575,
+      "learning_rate": 9.638323247313167e-06,
+      "loss": 0.3609,
+      "step": 486
+    },
+    {
+      "epoch": 0.1486342133374027,
+      "grad_norm": 1.273585773097182,
+      "learning_rate": 9.636474716294275e-06,
+      "loss": 0.2781,
+      "step": 487
+    },
+    {
+      "epoch": 0.14893941706088815,
+      "grad_norm": 1.4426438722313946,
+      "learning_rate": 9.634621651571235e-06,
+      "loss": 0.3446,
+      "step": 488
+    },
+    {
+      "epoch": 0.14924462078437356,
+      "grad_norm": 1.3396568501630033,
+      "learning_rate": 9.632764054956042e-06,
+      "loss": 0.3209,
+      "step": 489
+    },
+    {
+      "epoch": 0.149549824507859,
+      "grad_norm": 1.869987463373602,
+      "learning_rate": 9.630901928265113e-06,
+      "loss": 0.6676,
+      "step": 490
+    },
+    {
+      "epoch": 0.14985502823134442,
+      "grad_norm": 1.3995020559881168,
+      "learning_rate": 9.6290352733193e-06,
+      "loss": 0.443,
+      "step": 491
+    },
+    {
+      "epoch": 0.15016023195482986,
+      "grad_norm": 1.5646801044524754,
+      "learning_rate": 9.627164091943886e-06,
+      "loss": 0.4808,
+      "step": 492
+    },
+    {
+      "epoch": 0.15046543567831527,
+      "grad_norm": 1.4064811516565765,
+      "learning_rate": 9.625288385968572e-06,
+      "loss": 0.3281,
+      "step": 493
+    },
+    {
+      "epoch": 0.15077063940180072,
+      "grad_norm": 1.3603824959747954,
+      "learning_rate": 9.623408157227493e-06,
+      "loss": 0.4203,
+      "step": 494
+    },
+    {
+      "epoch": 0.15107584312528613,
+      "grad_norm": 1.4648922518012057,
+      "learning_rate": 9.621523407559193e-06,
+      "loss": 0.4691,
+      "step": 495
+    },
+    {
+      "epoch": 0.15138104684877154,
+      "grad_norm": 1.2898880598011702,
+      "learning_rate": 9.619634138806653e-06,
+      "loss": 0.22,
+      "step": 496
+    },
+    {
+      "epoch": 0.151686250572257,
+      "grad_norm": 1.629312149082348,
+      "learning_rate": 9.61774035281726e-06,
+      "loss": 0.3852,
+      "step": 497
+    },
+    {
+      "epoch": 0.1519914542957424,
+      "grad_norm": 1.4209940770003642,
+      "learning_rate": 9.615842051442825e-06,
+      "loss": 0.3434,
+      "step": 498
+    },
+    {
+      "epoch": 0.15229665801922784,
+      "grad_norm": 1.5981471419786573,
+      "learning_rate": 9.613939236539571e-06,
+      "loss": 0.4405,
+      "step": 499
+    },
+    {
+      "epoch": 0.15260186174271326,
+      "grad_norm": 1.6345273007096384,
+      "learning_rate": 9.612031909968138e-06,
+      "loss": 0.4621,
+      "step": 500
+    },
+    {
+      "epoch": 0.1529070654661987,
+      "grad_norm": 1.7946527422515466,
+      "learning_rate": 9.610120073593574e-06,
+      "loss": 0.4215,
+      "step": 501
+    },
+    {
+      "epoch": 0.1532122691896841,
+      "grad_norm": 1.9696007964079152,
+      "learning_rate": 9.608203729285337e-06,
+      "loss": 0.2416,
+      "step": 502
+    },
+    {
+      "epoch": 0.15351747291316953,
+      "grad_norm": 1.186483550480542,
+      "learning_rate": 9.606282878917296e-06,
+      "loss": 0.1656,
+      "step": 503
+    },
+    {
+      "epoch": 0.15382267663665497,
+      "grad_norm": 1.3709781521921298,
+      "learning_rate": 9.604357524367723e-06,
+      "loss": 0.3374,
+      "step": 504
+    },
+    {
+      "epoch": 0.15412788036014038,
+      "grad_norm": 1.4744363645402312,
+      "learning_rate": 9.602427667519297e-06,
+      "loss": 0.3472,
+      "step": 505
+    },
+    {
+      "epoch": 0.15443308408362583,
+      "grad_norm": 1.7032963377834875,
+      "learning_rate": 9.600493310259098e-06,
+      "loss": 0.4352,
+      "step": 506
+    },
+    {
+      "epoch": 0.15473828780711124,
+      "grad_norm": 1.487020684601837,
+      "learning_rate": 9.598554454478608e-06,
+      "loss": 0.2661,
+      "step": 507
+    },
+    {
+      "epoch": 0.15504349153059668,
+      "grad_norm": 1.2499312381905126,
+      "learning_rate": 9.596611102073703e-06,
+      "loss": 0.2785,
+      "step": 508
+    },
+    {
+      "epoch": 0.1553486952540821,
+      "grad_norm": 1.529878897767237,
+      "learning_rate": 9.594663254944664e-06,
+      "loss": 0.3768,
+      "step": 509
+    },
+    {
+      "epoch": 0.15565389897756754,
+      "grad_norm": 1.5214931502474798,
+      "learning_rate": 9.592710914996162e-06,
+      "loss": 0.5126,
+      "step": 510
+    },
+    {
+      "epoch": 0.15595910270105295,
+      "grad_norm": 1.3836939529329817,
+      "learning_rate": 9.590754084137259e-06,
+      "loss": 0.3011,
+      "step": 511
+    },
+    {
+      "epoch": 0.15626430642453837,
+      "grad_norm": 1.4833094737816435,
+      "learning_rate": 9.588792764281413e-06,
+      "loss": 0.4761,
+      "step": 512
+    },
+    {
+      "epoch": 0.1565695101480238,
+      "grad_norm": 1.2346664760598378,
+      "learning_rate": 9.586826957346473e-06,
+      "loss": 0.2454,
+      "step": 513
+    },
+    {
+      "epoch": 0.15687471387150922,
+      "grad_norm": 1.6476636760719772,
+      "learning_rate": 9.584856665254667e-06,
+      "loss": 0.2867,
+      "step": 514
+    },
+    {
+      "epoch": 0.15717991759499467,
+      "grad_norm": 1.5637210652382973,
+      "learning_rate": 9.58288188993262e-06,
+      "loss": 0.2899,
+      "step": 515
+    },
+    {
+      "epoch": 0.15748512131848008,
+      "grad_norm": 1.3205463270086828,
+      "learning_rate": 9.580902633311331e-06,
+      "loss": 0.3756,
+      "step": 516
+    },
+    {
+      "epoch": 0.15779032504196552,
+      "grad_norm": 1.3975127161911243,
+      "learning_rate": 9.578918897326186e-06,
+      "loss": 0.5111,
+      "step": 517
+    },
+    {
+      "epoch": 0.15809552876545094,
+      "grad_norm": 1.4459383872321914,
+      "learning_rate": 9.57693068391695e-06,
+      "loss": 0.4283,
+      "step": 518
+    },
+    {
+      "epoch": 0.15840073248893635,
+      "grad_norm": 1.7061785001760192,
+      "learning_rate": 9.574937995027767e-06,
+      "loss": 0.3702,
+      "step": 519
+    },
+    {
+      "epoch": 0.1587059362124218,
+      "grad_norm": 1.822247438656905,
+      "learning_rate": 9.572940832607157e-06,
+      "loss": 0.3406,
+      "step": 520
+    },
+    {
+      "epoch": 0.1590111399359072,
+      "grad_norm": 2.44932543751886,
+      "learning_rate": 9.570939198608013e-06,
+      "loss": 0.3118,
+      "step": 521
+    },
+    {
+      "epoch": 0.15931634365939265,
+      "grad_norm": 1.6119202421775476,
+      "learning_rate": 9.5689330949876e-06,
+      "loss": 0.3903,
+      "step": 522
+    },
+    {
+      "epoch": 0.15962154738287807,
+      "grad_norm": 1.4848330945324375,
+      "learning_rate": 9.56692252370756e-06,
+      "loss": 0.2336,
+      "step": 523
+    },
+    {
+      "epoch": 0.1599267511063635,
+      "grad_norm": 1.410632362194396,
+      "learning_rate": 9.564907486733893e-06,
+      "loss": 0.2749,
+      "step": 524
+    },
+    {
+      "epoch": 0.16023195482984892,
+      "grad_norm": 1.596052638125191,
+      "learning_rate": 9.562887986036975e-06,
+      "loss": 0.4752,
+      "step": 525
+    },
+    {
+      "epoch": 0.16053715855333436,
+      "grad_norm": 1.7668740909494465,
+      "learning_rate": 9.560864023591541e-06,
+      "loss": 0.4457,
+      "step": 526
+    },
+    {
+      "epoch": 0.16084236227681978,
+      "grad_norm": 1.4009268145182425,
+      "learning_rate": 9.558835601376692e-06,
+      "loss": 0.2615,
+      "step": 527
+    },
+    {
+      "epoch": 0.1611475660003052,
+      "grad_norm": 1.7299333351168085,
+      "learning_rate": 9.55680272137589e-06,
+      "loss": 0.5216,
+      "step": 528
+    },
+    {
+      "epoch": 0.16145276972379063,
+      "grad_norm": 1.398003196407042,
+      "learning_rate": 9.554765385576951e-06,
+      "loss": 0.2917,
+      "step": 529
+    },
+    {
+      "epoch": 0.16175797344727605,
+      "grad_norm": 1.4037115710357768,
+      "learning_rate": 9.552723595972055e-06,
+      "loss": 0.2794,
+      "step": 530
+    },
+    {
+      "epoch": 0.1620631771707615,
+      "grad_norm": 1.4104804936912443,
+      "learning_rate": 9.550677354557734e-06,
+      "loss": 0.3294,
+      "step": 531
+    },
+    {
+      "epoch": 0.1623683808942469,
+      "grad_norm": 1.3043707731550427,
+      "learning_rate": 9.548626663334872e-06,
+      "loss": 0.3542,
+      "step": 532
+    },
+    {
+      "epoch": 0.16267358461773235,
+      "grad_norm": 1.4523817232860987,
+      "learning_rate": 9.546571524308707e-06,
+      "loss": 0.4394,
+      "step": 533
+    },
+    {
+      "epoch": 0.16297878834121776,
+      "grad_norm": 1.2378417959119585,
+      "learning_rate": 9.544511939488823e-06,
+      "loss": 0.2859,
+      "step": 534
+    },
+    {
+      "epoch": 0.16328399206470318,
+      "grad_norm": 1.1623856416463947,
+      "learning_rate": 9.542447910889152e-06,
+      "loss": 0.2682,
+      "step": 535
+    },
+    {
+      "epoch": 0.16358919578818862,
+      "grad_norm": 1.503292443987416,
+      "learning_rate": 9.540379440527974e-06,
+      "loss": 0.4513,
+      "step": 536
+    },
+    {
+      "epoch": 0.16389439951167403,
+      "grad_norm": 1.278183220840744,
+      "learning_rate": 9.538306530427908e-06,
+      "loss": 0.2486,
+      "step": 537
+    },
+    {
+      "epoch": 0.16419960323515947,
+      "grad_norm": 1.477438530587252,
+      "learning_rate": 9.536229182615919e-06,
+      "loss": 0.4748,
+      "step": 538
+    },
+    {
+      "epoch": 0.1645048069586449,
+      "grad_norm": 1.161000468008389,
+      "learning_rate": 9.534147399123308e-06,
+      "loss": 0.3166,
+      "step": 539
+    },
+    {
+      "epoch": 0.16481001068213033,
+      "grad_norm": 1.3151690275104762,
+      "learning_rate": 9.532061181985713e-06,
+      "loss": 0.3547,
+      "step": 540
+    },
+    {
+      "epoch": 0.16511521440561575,
+      "grad_norm": 1.750297725419665,
+      "learning_rate": 9.529970533243112e-06,
+      "loss": 0.4156,
+      "step": 541
+    },
+    {
+      "epoch": 0.1654204181291012,
+      "grad_norm": 1.3465531852012238,
+      "learning_rate": 9.52787545493981e-06,
+      "loss": 0.3366,
+      "step": 542
+    },
+    {
+      "epoch": 0.1657256218525866,
+      "grad_norm": 1.5440141830188223,
+      "learning_rate": 9.525775949124447e-06,
+      "loss": 0.3376,
+      "step": 543
+    },
+    {
+      "epoch": 0.16603082557607202,
+      "grad_norm": 1.5415130315298482,
+      "learning_rate": 9.523672017849995e-06,
+      "loss": 0.541,
+      "step": 544
+    },
+    {
+      "epoch": 0.16633602929955746,
+      "grad_norm": 1.7579856956776627,
+      "learning_rate": 9.521563663173746e-06,
+      "loss": 0.4806,
+      "step": 545
+    },
+    {
+      "epoch": 0.16664123302304287,
+      "grad_norm": 1.7818866729621587,
+      "learning_rate": 9.519450887157324e-06,
+      "loss": 0.5464,
+      "step": 546
+    },
+    {
+      "epoch": 0.16694643674652831,
+      "grad_norm": 1.6064088023016758,
+      "learning_rate": 9.517333691866672e-06,
+      "loss": 0.43,
+      "step": 547
+    },
+    {
+      "epoch": 0.16725164047001373,
+      "grad_norm": 1.4778043177300115,
+      "learning_rate": 9.515212079372059e-06,
+      "loss": 0.4399,
+      "step": 548
+    },
+    {
+      "epoch": 0.16755684419349917,
+      "grad_norm": 2.027608100840915,
+      "learning_rate": 9.513086051748069e-06,
+      "loss": 0.4069,
+      "step": 549
+    },
+    {
+      "epoch": 0.16786204791698459,
+      "grad_norm": 1.5850802204517391,
+      "learning_rate": 9.510955611073605e-06,
+      "loss": 0.3827,
+      "step": 550
+    },
+    {
+      "epoch": 0.16816725164047,
+      "grad_norm": 1.293889481401633,
+      "learning_rate": 9.508820759431883e-06,
+      "loss": 0.2572,
+      "step": 551
+    },
+    {
+      "epoch": 0.16847245536395544,
+      "grad_norm": 1.842230801320139,
+      "learning_rate": 9.506681498910437e-06,
+      "loss": 0.5275,
+      "step": 552
+    },
+    {
+      "epoch": 0.16877765908744086,
+      "grad_norm": 1.1873629641518748,
+      "learning_rate": 9.50453783160111e-06,
+      "loss": 0.3282,
+      "step": 553
+    },
+    {
+      "epoch": 0.1690828628109263,
+      "grad_norm": 1.6093453710106354,
+      "learning_rate": 9.50238975960005e-06,
+      "loss": 0.5784,
+      "step": 554
+    },
+    {
+      "epoch": 0.1693880665344117,
+      "grad_norm": 1.8900657646403543,
+      "learning_rate": 9.500237285007719e-06,
+      "loss": 0.5224,
+      "step": 555
+    },
+    {
+      "epoch": 0.16969327025789716,
+      "grad_norm": 1.4803576264157936,
+      "learning_rate": 9.498080409928878e-06,
+      "loss": 0.3726,
+      "step": 556
+    },
+    {
+      "epoch": 0.16999847398138257,
+      "grad_norm": 1.7675886459458987,
+      "learning_rate": 9.495919136472595e-06,
+      "loss": 0.4656,
+      "step": 557
+    },
+    {
+      "epoch": 0.170303677704868,
+      "grad_norm": 1.7910944298366285,
+      "learning_rate": 9.493753466752236e-06,
+      "loss": 0.8076,
+      "step": 558
+    },
+    {
+      "epoch": 0.17060888142835343,
+      "grad_norm": 1.4678825901279975,
+      "learning_rate": 9.49158340288547e-06,
+      "loss": 0.3575,
+      "step": 559
+    },
+    {
+      "epoch": 0.17091408515183884,
+      "grad_norm": 1.411596350121475,
+      "learning_rate": 9.489408946994256e-06,
+      "loss": 0.3316,
+      "step": 560
+    },
+    {
+      "epoch": 0.17121928887532428,
+      "grad_norm": 1.2961064094904746,
+      "learning_rate": 9.487230101204855e-06,
+      "loss": 0.3634,
+      "step": 561
+    },
+    {
+      "epoch": 0.1715244925988097,
+      "grad_norm": 1.2525457208629842,
+      "learning_rate": 9.485046867647816e-06,
+      "loss": 0.368,
+      "step": 562
+    },
+    {
+      "epoch": 0.17182969632229514,
+      "grad_norm": 1.5857071732762902,
+      "learning_rate": 9.48285924845798e-06,
+      "loss": 0.4546,
+      "step": 563
+    },
+    {
+      "epoch": 0.17213490004578055,
+      "grad_norm": 1.4510716523054648,
+      "learning_rate": 9.480667245774474e-06,
+      "loss": 0.2739,
+      "step": 564
+    },
+    {
+      "epoch": 0.172440103769266,
+      "grad_norm": 1.7890268513821783,
+      "learning_rate": 9.478470861740716e-06,
+      "loss": 0.4085,
+      "step": 565
+    },
+    {
+      "epoch": 0.1727453074927514,
+      "grad_norm": 1.2143129127325427,
+      "learning_rate": 9.476270098504405e-06,
+      "loss": 0.2669,
+      "step": 566
+    },
+    {
+      "epoch": 0.17305051121623685,
+      "grad_norm": 1.4725539275626134,
+      "learning_rate": 9.474064958217524e-06,
+      "loss": 0.3474,
+      "step": 567
+    },
+    {
+      "epoch": 0.17335571493972227,
+      "grad_norm": 1.6648347880329453,
+      "learning_rate": 9.471855443036333e-06,
+      "loss": 0.3059,
+      "step": 568
+    },
+    {
+      "epoch": 0.17366091866320768,
+      "grad_norm": 1.4826208089202084,
+      "learning_rate": 9.469641555121372e-06,
+      "loss": 0.3309,
+      "step": 569
+    },
+    {
+      "epoch": 0.17396612238669312,
+      "grad_norm": 1.2087588985869038,
+      "learning_rate": 9.467423296637458e-06,
+      "loss": 0.2765,
+      "step": 570
+    },
+    {
+      "epoch": 0.17427132611017854,
+      "grad_norm": 1.5271503296745377,
+      "learning_rate": 9.465200669753678e-06,
+      "loss": 0.4388,
+      "step": 571
+    },
+    {
+      "epoch": 0.17457652983366398,
+      "grad_norm": 1.720167996940521,
+      "learning_rate": 9.462973676643395e-06,
+      "loss": 0.4693,
+      "step": 572
+    },
+    {
+      "epoch": 0.1748817335571494,
+      "grad_norm": 1.4666930059033054,
+      "learning_rate": 9.46074231948424e-06,
+      "loss": 0.354,
+      "step": 573
+    },
+    {
+      "epoch": 0.17518693728063484,
+      "grad_norm": 1.8597430337784902,
+      "learning_rate": 9.458506600458106e-06,
+      "loss": 0.3892,
+      "step": 574
+    },
+    {
+      "epoch": 0.17549214100412025,
+      "grad_norm": 0.9696512337091734,
+      "learning_rate": 9.456266521751162e-06,
+      "loss": 0.2294,
+      "step": 575
+    },
+    {
+      "epoch": 0.17579734472760566,
+      "grad_norm": 1.781498807963985,
+      "learning_rate": 9.454022085553829e-06,
+      "loss": 0.4873,
+      "step": 576
+    },
+    {
+      "epoch": 0.1761025484510911,
+      "grad_norm": 1.3337716733106453,
+      "learning_rate": 9.451773294060797e-06,
+      "loss": 0.3031,
+      "step": 577
+    },
+    {
+      "epoch": 0.17640775217457652,
+      "grad_norm": 1.7055986970891146,
+      "learning_rate": 9.449520149471008e-06,
+      "loss": 0.6405,
+      "step": 578
+    },
+    {
+      "epoch": 0.17671295589806196,
+      "grad_norm": 1.5188407211523098,
+      "learning_rate": 9.447262653987668e-06,
+      "loss": 0.3739,
+      "step": 579
+    },
+    {
+      "epoch": 0.17701815962154738,
+      "grad_norm": 0.9308569789252137,
+      "learning_rate": 9.445000809818231e-06,
+      "loss": 0.2505,
+      "step": 580
+    },
+    {
+      "epoch": 0.17732336334503282,
+      "grad_norm": 1.5000700139554115,
+      "learning_rate": 9.442734619174408e-06,
+      "loss": 0.4158,
+      "step": 581
+    },
+    {
+      "epoch": 0.17762856706851823,
+      "grad_norm": 1.3548714588751447,
+      "learning_rate": 9.440464084272157e-06,
+      "loss": 0.3911,
+      "step": 582
+    },
+    {
+      "epoch": 0.17793377079200368,
+      "grad_norm": 1.3241166157002833,
+      "learning_rate": 9.438189207331684e-06,
+      "loss": 0.371,
+      "step": 583
+    },
+    {
+      "epoch": 0.1782389745154891,
+      "grad_norm": 1.251287417238196,
+      "learning_rate": 9.435909990577442e-06,
+      "loss": 0.4543,
+      "step": 584
+    },
+    {
+      "epoch": 0.1785441782389745,
+      "grad_norm": 1.6243898273186124,
+      "learning_rate": 9.433626436238128e-06,
+      "loss": 0.3955,
+      "step": 585
+    },
+    {
+      "epoch": 0.17884938196245995,
+      "grad_norm": 1.5933520250274826,
+      "learning_rate": 9.43133854654668e-06,
+      "loss": 0.4232,
+      "step": 586
+    },
+    {
+      "epoch": 0.17915458568594536,
+      "grad_norm": 1.271604048902552,
+      "learning_rate": 9.429046323740275e-06,
+      "loss": 0.186,
+      "step": 587
+    },
+    {
+      "epoch": 0.1794597894094308,
+      "grad_norm": 1.7480757897399084,
+      "learning_rate": 9.426749770060325e-06,
+      "loss": 0.3198,
+      "step": 588
+    },
+    {
+      "epoch": 0.17976499313291622,
+      "grad_norm": 1.6582908292655634,
+      "learning_rate": 9.424448887752485e-06,
+      "loss": 0.4497,
+      "step": 589
+    },
+    {
+      "epoch": 0.18007019685640166,
+      "grad_norm": 1.514808198729056,
+      "learning_rate": 9.42214367906663e-06,
+      "loss": 0.3135,
+      "step": 590
+    },
+    {
+      "epoch": 0.18037540057988707,
+      "grad_norm": 1.33735933544563,
+      "learning_rate": 9.419834146256875e-06,
+      "loss": 0.1512,
+      "step": 591
+    },
+    {
+      "epoch": 0.1806806043033725,
+      "grad_norm": 1.7983955915325747,
+      "learning_rate": 9.417520291581562e-06,
+      "loss": 0.397,
+      "step": 592
+    },
+    {
+      "epoch": 0.18098580802685793,
+      "grad_norm": 1.8357965942853254,
+      "learning_rate": 9.415202117303253e-06,
+      "loss": 0.3479,
+      "step": 593
+    },
+    {
+      "epoch": 0.18129101175034334,
+      "grad_norm": 2.1098502294891084,
+      "learning_rate": 9.412879625688742e-06,
+      "loss": 0.6081,
+      "step": 594
+    },
+    {
+      "epoch": 0.1815962154738288,
+      "grad_norm": 1.7002717361934219,
+      "learning_rate": 9.410552819009041e-06,
+      "loss": 0.2335,
+      "step": 595
+    },
+    {
+      "epoch": 0.1819014191973142,
+      "grad_norm": 1.6858392243118179,
+      "learning_rate": 9.408221699539381e-06,
+      "loss": 0.3502,
+      "step": 596
+    },
+    {
+      "epoch": 0.18220662292079964,
+      "grad_norm": 1.9779389304442994,
+      "learning_rate": 9.40588626955921e-06,
+      "loss": 0.5023,
+      "step": 597
+    },
+    {
+      "epoch": 0.18251182664428506,
+      "grad_norm": 1.984831269441273,
+      "learning_rate": 9.403546531352192e-06,
+      "loss": 0.1808,
+      "step": 598
+    },
+    {
+      "epoch": 0.1828170303677705,
+      "grad_norm": 1.7825668553552305,
+      "learning_rate": 9.401202487206205e-06,
+      "loss": 0.2451,
+      "step": 599
+    },
+    {
+      "epoch": 0.18312223409125591,
+      "grad_norm": 1.7395479880147604,
+      "learning_rate": 9.398854139413332e-06,
+      "loss": 0.4586,
+      "step": 600
+    },
+    {
+      "epoch": 0.18342743781474133,
+      "grad_norm": 1.7910877075998561,
+      "learning_rate": 9.396501490269871e-06,
+      "loss": 0.4334,
+      "step": 601
+    },
+    {
+      "epoch": 0.18373264153822677,
+      "grad_norm": 1.671153260357796,
+      "learning_rate": 9.394144542076321e-06,
+      "loss": 0.3457,
+      "step": 602
+    },
+    {
+      "epoch": 0.18403784526171219,
+      "grad_norm": 2.2927735747628057,
+      "learning_rate": 9.391783297137392e-06,
+      "loss": 0.5006,
+      "step": 603
+    },
+    {
+      "epoch": 0.18434304898519763,
+      "grad_norm": 1.4375155904173251,
+      "learning_rate": 9.389417757761983e-06,
+      "loss": 0.3412,
+      "step": 604
+    },
+    {
+      "epoch": 0.18464825270868304,
+      "grad_norm": 1.0876679937459988,
+      "learning_rate": 9.387047926263205e-06,
+      "loss": 0.2323,
+      "step": 605
+    },
+    {
+      "epoch": 0.18495345643216848,
+      "grad_norm": 1.5691166969698962,
+      "learning_rate": 9.384673804958357e-06,
+      "loss": 0.3929,
+      "step": 606
+    },
+    {
+      "epoch": 0.1852586601556539,
+      "grad_norm": 1.2508041656129554,
+      "learning_rate": 9.38229539616894e-06,
+      "loss": 0.2123,
+      "step": 607
+    },
+    {
+      "epoch": 0.1855638638791393,
+      "grad_norm": 1.6014009719897135,
+      "learning_rate": 9.379912702220641e-06,
+      "loss": 0.234,
+      "step": 608
+    },
+    {
+      "epoch": 0.18586906760262475,
+      "grad_norm": 1.4804611004553776,
+      "learning_rate": 9.377525725443341e-06,
+      "loss": 0.3951,
+      "step": 609
+    },
+    {
+      "epoch": 0.18617427132611017,
+      "grad_norm": 1.4589508747304376,
+      "learning_rate": 9.375134468171108e-06,
+      "loss": 0.2887,
+      "step": 610
+    },
+    {
+      "epoch": 0.1864794750495956,
+      "grad_norm": 1.9199770728371568,
+      "learning_rate": 9.372738932742193e-06,
+      "loss": 0.5627,
+      "step": 611
+    },
+    {
+      "epoch": 0.18678467877308103,
+      "grad_norm": 1.5903295576095668,
+      "learning_rate": 9.370339121499039e-06,
+      "loss": 0.4379,
+      "step": 612
+    },
+    {
+      "epoch": 0.18708988249656647,
+      "grad_norm": 1.6986262549100166,
+      "learning_rate": 9.367935036788257e-06,
+      "loss": 0.4873,
+      "step": 613
+    },
+    {
+      "epoch": 0.18739508622005188,
+      "grad_norm": 1.2194659257752518,
+      "learning_rate": 9.365526680960645e-06,
+      "loss": 0.3571,
+      "step": 614
+    },
+    {
+      "epoch": 0.18770028994353732,
+      "grad_norm": 1.3049138038493902,
+      "learning_rate": 9.363114056371178e-06,
+      "loss": 0.3114,
+      "step": 615
+    },
+    {
+      "epoch": 0.18800549366702274,
+      "grad_norm": 1.468685879220778,
+      "learning_rate": 9.360697165379004e-06,
+      "loss": 0.5043,
+      "step": 616
+    },
+    {
+      "epoch": 0.18831069739050815,
+      "grad_norm": 2.131454426646245,
+      "learning_rate": 9.35827601034744e-06,
+      "loss": 0.5165,
+      "step": 617
+    },
+    {
+      "epoch": 0.1886159011139936,
+      "grad_norm": 1.5563065422847613,
+      "learning_rate": 9.355850593643974e-06,
+      "loss": 0.4707,
+      "step": 618
+    },
+    {
+      "epoch": 0.188921104837479,
+      "grad_norm": 1.3992788671446874,
+      "learning_rate": 9.353420917640264e-06,
+      "loss": 0.3905,
+      "step": 619
+    },
+    {
+      "epoch": 0.18922630856096445,
+      "grad_norm": 1.2635210704880713,
+      "learning_rate": 9.350986984712128e-06,
+      "loss": 0.2399,
+      "step": 620
+    },
+    {
+      "epoch": 0.18953151228444987,
+      "grad_norm": 1.4071894580574145,
+      "learning_rate": 9.348548797239551e-06,
+      "loss": 0.3689,
+      "step": 621
+    },
+    {
+      "epoch": 0.1898367160079353,
+      "grad_norm": 1.8460324916935194,
+      "learning_rate": 9.346106357606675e-06,
+      "loss": 0.3337,
+      "step": 622
+    },
+    {
+      "epoch": 0.19014191973142072,
+      "grad_norm": 1.2199756132877848,
+      "learning_rate": 9.343659668201803e-06,
+      "loss": 0.3707,
+      "step": 623
+    },
+    {
+      "epoch": 0.19044712345490614,
+      "grad_norm": 1.3352934416971625,
+      "learning_rate": 9.34120873141739e-06,
+      "loss": 0.3332,
+      "step": 624
+    },
+    {
+      "epoch": 0.19075232717839158,
+      "grad_norm": 1.5847999740161538,
+      "learning_rate": 9.33875354965005e-06,
+      "loss": 0.4658,
+      "step": 625
+    },
+    {
+      "epoch": 0.191057530901877,
+      "grad_norm": 1.59267718540602,
+      "learning_rate": 9.336294125300538e-06,
+      "loss": 0.5138,
+      "step": 626
+    },
+    {
+      "epoch": 0.19136273462536244,
+      "grad_norm": 1.0558131110089597,
+      "learning_rate": 9.333830460773767e-06,
+      "loss": 0.2512,
+      "step": 627
+    },
+    {
+      "epoch": 0.19166793834884785,
+      "grad_norm": 1.6912848096424418,
+      "learning_rate": 9.331362558478793e-06,
+      "loss": 0.3129,
+      "step": 628
+    },
+    {
+      "epoch": 0.1919731420723333,
+      "grad_norm": 1.7565199783626735,
+      "learning_rate": 9.328890420828817e-06,
+      "loss": 0.2625,
+      "step": 629
+    },
+    {
+      "epoch": 0.1922783457958187,
+      "grad_norm": 1.6287483120468187,
+      "learning_rate": 9.326414050241176e-06,
+      "loss": 0.4631,
+      "step": 630
+    },
+    {
+      "epoch": 0.19258354951930415,
+      "grad_norm": 1.5343127080699748,
+      "learning_rate": 9.323933449137353e-06,
+      "loss": 0.35,
+      "step": 631
+    },
+    {
+      "epoch": 0.19288875324278956,
+      "grad_norm": 1.2901421851525343,
+      "learning_rate": 9.321448619942963e-06,
+      "loss": 0.3191,
+      "step": 632
+    },
+    {
+      "epoch": 0.19319395696627498,
+      "grad_norm": 1.3651871352420322,
+      "learning_rate": 9.318959565087761e-06,
+      "loss": 0.3063,
+      "step": 633
+    },
+    {
+      "epoch": 0.19349916068976042,
+      "grad_norm": 1.4900191765389657,
+      "learning_rate": 9.316466287005625e-06,
+      "loss": 0.3621,
+      "step": 634
+    },
+    {
+      "epoch": 0.19380436441324583,
+      "grad_norm": 1.836926327897149,
+      "learning_rate": 9.313968788134572e-06,
+      "loss": 0.6273,
+      "step": 635
+    },
+    {
+      "epoch": 0.19410956813673128,
+      "grad_norm": 1.6283757285797815,
+      "learning_rate": 9.311467070916743e-06,
+      "loss": 0.3191,
+      "step": 636
+    },
+    {
+      "epoch": 0.1944147718602167,
+      "grad_norm": 1.7047955780313857,
+      "learning_rate": 9.308961137798398e-06,
+      "loss": 0.5581,
+      "step": 637
+    },
+    {
+      "epoch": 0.19471997558370213,
+      "grad_norm": 1.4826549000090183,
+      "learning_rate": 9.306450991229927e-06,
+      "loss": 0.3157,
+      "step": 638
+    },
+    {
+      "epoch": 0.19502517930718755,
+      "grad_norm": 1.435361017145943,
+      "learning_rate": 9.30393663366584e-06,
+      "loss": 0.3084,
+      "step": 639
+    },
+    {
+      "epoch": 0.19533038303067296,
+      "grad_norm": 1.402358583674702,
+      "learning_rate": 9.301418067564758e-06,
+      "loss": 0.3351,
+      "step": 640
+    },
+    {
+      "epoch": 0.1956355867541584,
+      "grad_norm": 1.2933654176691274,
+      "learning_rate": 9.298895295389423e-06,
+      "loss": 0.2585,
+      "step": 641
+    },
+    {
+      "epoch": 0.19594079047764382,
+      "grad_norm": 1.5346090103364156,
+      "learning_rate": 9.29636831960669e-06,
+      "loss": 0.33,
+      "step": 642
+    },
+    {
+      "epoch": 0.19624599420112926,
+      "grad_norm": 1.3251725534327445,
+      "learning_rate": 9.293837142687522e-06,
+      "loss": 0.2104,
+      "step": 643
+    },
+    {
+      "epoch": 0.19655119792461467,
+      "grad_norm": 1.583285740923444,
+      "learning_rate": 9.291301767106986e-06,
+      "loss": 0.4326,
+      "step": 644
+    },
+    {
+      "epoch": 0.19685640164810012,
+      "grad_norm": 1.0905371753559963,
+      "learning_rate": 9.288762195344266e-06,
+      "loss": 0.274,
+      "step": 645
+    },
+    {
+      "epoch": 0.19716160537158553,
+      "grad_norm": 1.7263838041187525,
+      "learning_rate": 9.28621842988264e-06,
+      "loss": 0.5011,
+      "step": 646
+    },
+    {
+      "epoch": 0.19746680909507097,
+      "grad_norm": 1.4838510492072716,
+      "learning_rate": 9.283670473209488e-06,
+      "loss": 0.1956,
+      "step": 647
+    },
+    {
+      "epoch": 0.1977720128185564,
+      "grad_norm": 1.2036114489558822,
+      "learning_rate": 9.28111832781629e-06,
+      "loss": 0.2346,
+      "step": 648
+    },
+    {
+      "epoch": 0.1980772165420418,
+      "grad_norm": 1.3853687861302435,
+      "learning_rate": 9.278561996198622e-06,
+      "loss": 0.2313,
+      "step": 649
+    },
+    {
+      "epoch": 0.19838242026552724,
+      "grad_norm": 1.2708592907281826,
+      "learning_rate": 9.276001480856152e-06,
+      "loss": 0.3717,
+      "step": 650
+    },
+    {
+      "epoch": 0.19868762398901266,
+      "grad_norm": 1.6407943638167721,
+      "learning_rate": 9.273436784292641e-06,
+      "loss": 0.5533,
+      "step": 651
+    },
+    {
+      "epoch": 0.1989928277124981,
+      "grad_norm": 1.9614104046455092,
+      "learning_rate": 9.270867909015936e-06,
+      "loss": 0.4552,
+      "step": 652
+    },
+    {
+      "epoch": 0.1992980314359835,
+      "grad_norm": 1.6038724290911757,
+      "learning_rate": 9.268294857537973e-06,
+      "loss": 0.477,
+      "step": 653
+    },
+    {
+      "epoch": 0.19960323515946896,
+      "grad_norm": 1.9838192379267598,
+      "learning_rate": 9.26571763237477e-06,
+      "loss": 0.7118,
+      "step": 654
+    },
+    {
+      "epoch": 0.19990843888295437,
+      "grad_norm": 1.4066170955047037,
+      "learning_rate": 9.263136236046422e-06,
+      "loss": 0.4222,
+      "step": 655
+    },
+    {
+      "epoch": 0.2002136426064398,
+      "grad_norm": 1.800111592330467,
+      "learning_rate": 9.260550671077113e-06,
+      "loss": 0.4969,
+      "step": 656
+    },
+    {
+      "epoch": 0.20051884632992523,
+      "grad_norm": 1.450700768131887,
+      "learning_rate": 9.257960939995093e-06,
+      "loss": 0.3938,
+      "step": 657
+    },
+    {
+      "epoch": 0.20082405005341064,
+      "grad_norm": 1.6208519445035616,
+      "learning_rate": 9.255367045332693e-06,
+      "loss": 0.5284,
+      "step": 658
+    },
+    {
+      "epoch": 0.20112925377689608,
+      "grad_norm": 1.4936718972189802,
+      "learning_rate": 9.25276898962631e-06,
+      "loss": 0.3314,
+      "step": 659
+    },
+    {
+      "epoch": 0.2014344575003815,
+      "grad_norm": 1.8465064182825872,
+      "learning_rate": 9.250166775416412e-06,
+      "loss": 0.5858,
+      "step": 660
+    },
+    {
+      "epoch": 0.20173966122386694,
+      "grad_norm": 1.7205660848788409,
+      "learning_rate": 9.247560405247535e-06,
+      "loss": 0.4412,
+      "step": 661
+    },
+    {
+      "epoch": 0.20204486494735235,
+      "grad_norm": 1.3534767838887214,
+      "learning_rate": 9.244949881668276e-06,
+      "loss": 0.2469,
+      "step": 662
+    },
+    {
+      "epoch": 0.2023500686708378,
+      "grad_norm": 1.283164106498675,
+      "learning_rate": 9.242335207231297e-06,
+      "loss": 0.2035,
+      "step": 663
+    },
+    {
+      "epoch": 0.2026552723943232,
+      "grad_norm": 1.6915673582245812,
+      "learning_rate": 9.239716384493313e-06,
+      "loss": 0.2989,
+      "step": 664
+    },
+    {
+      "epoch": 0.20296047611780862,
+      "grad_norm": 2.2410065734371956,
+      "learning_rate": 9.2370934160151e-06,
+      "loss": 0.4781,
+      "step": 665
+    },
+    {
+      "epoch": 0.20326567984129407,
+      "grad_norm": 1.3632387520798106,
+      "learning_rate": 9.234466304361487e-06,
+      "loss": 0.2167,
+      "step": 666
+    },
+    {
+      "epoch": 0.20357088356477948,
+      "grad_norm": 1.249315694419475,
+      "learning_rate": 9.231835052101352e-06,
+      "loss": 0.323,
+      "step": 667
+    },
+    {
+      "epoch": 0.20387608728826492,
+      "grad_norm": 1.2758684884196045,
+      "learning_rate": 9.229199661807628e-06,
+      "loss": 0.2939,
+      "step": 668
+    },
+    {
+      "epoch": 0.20418129101175034,
+      "grad_norm": 1.6065093517583355,
+      "learning_rate": 9.226560136057286e-06,
+      "loss": 0.5757,
+      "step": 669
+    },
+    {
+      "epoch": 0.20448649473523578,
+      "grad_norm": 1.4713850530122647,
+      "learning_rate": 9.223916477431348e-06,
+      "loss": 0.1891,
+      "step": 670
+    },
+    {
+      "epoch": 0.2047916984587212,
+      "grad_norm": 1.2115833160052654,
+      "learning_rate": 9.221268688514872e-06,
+      "loss": 0.2403,
+      "step": 671
+    },
+    {
+      "epoch": 0.20509690218220664,
+      "grad_norm": 1.2023485328623802,
+      "learning_rate": 9.218616771896958e-06,
+      "loss": 0.2996,
+      "step": 672
+    },
+    {
+      "epoch": 0.20540210590569205,
+      "grad_norm": 1.489677195048235,
+      "learning_rate": 9.21596073017074e-06,
+      "loss": 0.4833,
+      "step": 673
+    },
+    {
+      "epoch": 0.20570730962917746,
+      "grad_norm": 1.3322943821875144,
+      "learning_rate": 9.213300565933387e-06,
+      "loss": 0.4981,
+      "step": 674
+    },
+    {
+      "epoch": 0.2060125133526629,
+      "grad_norm": 1.657242243821286,
+      "learning_rate": 9.210636281786101e-06,
+      "loss": 0.4101,
+      "step": 675
+    },
+    {
+      "epoch": 0.20631771707614832,
+      "grad_norm": 1.5710052935767878,
+      "learning_rate": 9.207967880334107e-06,
+      "loss": 0.5269,
+      "step": 676
+    },
+    {
+      "epoch": 0.20662292079963376,
+      "grad_norm": 1.7349735636167414,
+      "learning_rate": 9.205295364186664e-06,
+      "loss": 0.4501,
+      "step": 677
+    },
+    {
+      "epoch": 0.20692812452311918,
+      "grad_norm": 1.2131120415162016,
+      "learning_rate": 9.202618735957044e-06,
+      "loss": 0.2405,
+      "step": 678
+    },
+    {
+      "epoch": 0.20723332824660462,
+      "grad_norm": 1.873473817856673,
+      "learning_rate": 9.199937998262553e-06,
+      "loss": 0.534,
+      "step": 679
+    },
+    {
+      "epoch": 0.20753853197009003,
+      "grad_norm": 1.5561433809212002,
+      "learning_rate": 9.197253153724504e-06,
+      "loss": 0.3787,
+      "step": 680
+    },
+    {
+      "epoch": 0.20784373569357545,
+      "grad_norm": 1.656918804520659,
+      "learning_rate": 9.194564204968231e-06,
+      "loss": 0.5373,
+      "step": 681
+    },
+    {
+      "epoch": 0.2081489394170609,
+      "grad_norm": 1.2717305075559726,
+      "learning_rate": 9.19187115462308e-06,
+      "loss": 0.3069,
+      "step": 682
+    },
+    {
+      "epoch": 0.2084541431405463,
+      "grad_norm": 1.6153063314109386,
+      "learning_rate": 9.189174005322408e-06,
+      "loss": 0.5305,
+      "step": 683
+    },
+    {
+      "epoch": 0.20875934686403175,
+      "grad_norm": 1.336223638576054,
+      "learning_rate": 9.18647275970358e-06,
+      "loss": 0.3889,
+      "step": 684
+    },
+    {
+      "epoch": 0.20906455058751716,
+      "grad_norm": 1.8056451147641948,
+      "learning_rate": 9.18376742040797e-06,
+      "loss": 0.5624,
+      "step": 685
+    },
+    {
+      "epoch": 0.2093697543110026,
+      "grad_norm": 1.4375442829774812,
+      "learning_rate": 9.181057990080945e-06,
+      "loss": 0.4177,
+      "step": 686
+    },
+    {
+      "epoch": 0.20967495803448802,
+      "grad_norm": 1.4711699333864836,
+      "learning_rate": 9.178344471371886e-06,
+      "loss": 0.3406,
+      "step": 687
+    },
+    {
+      "epoch": 0.20998016175797346,
+      "grad_norm": 1.3838222287012096,
+      "learning_rate": 9.17562686693416e-06,
+      "loss": 0.3066,
+      "step": 688
+    },
+    {
+      "epoch": 0.21028536548145887,
+      "grad_norm": 1.4022212067938096,
+      "learning_rate": 9.172905179425136e-06,
+      "loss": 0.2756,
+      "step": 689
+    },
+    {
+      "epoch": 0.2105905692049443,
+      "grad_norm": 1.3536281346270633,
+      "learning_rate": 9.170179411506175e-06,
+      "loss": 0.4636,
+      "step": 690
+    },
+    {
+      "epoch": 0.21089577292842973,
+      "grad_norm": 1.8927154333736904,
+      "learning_rate": 9.167449565842622e-06,
+      "loss": 0.4806,
+      "step": 691
+    },
+    {
+      "epoch": 0.21120097665191515,
+      "grad_norm": 1.3316444887041978,
+      "learning_rate": 9.164715645103818e-06,
+      "loss": 0.3236,
+      "step": 692
+    },
+    {
+      "epoch": 0.2115061803754006,
+      "grad_norm": 1.5414957862521723,
+      "learning_rate": 9.161977651963082e-06,
+      "loss": 0.3174,
+      "step": 693
+    },
+    {
+      "epoch": 0.211811384098886,
+      "grad_norm": 1.3526532837600678,
+      "learning_rate": 9.15923558909772e-06,
+      "loss": 0.3286,
+      "step": 694
+    },
+    {
+      "epoch": 0.21211658782237144,
+      "grad_norm": 1.6741032947314405,
+      "learning_rate": 9.156489459189011e-06,
+      "loss": 0.5021,
+      "step": 695
+    },
+    {
+      "epoch": 0.21242179154585686,
+      "grad_norm": 1.3755135010833581,
+      "learning_rate": 9.153739264922221e-06,
+      "loss": 0.3659,
+      "step": 696
+    },
+    {
+      "epoch": 0.21272699526934227,
+      "grad_norm": 1.1828934010664012,
+      "learning_rate": 9.150985008986579e-06,
+      "loss": 0.273,
+      "step": 697
+    },
+    {
+      "epoch": 0.21303219899282771,
+      "grad_norm": 1.6368043280832338,
+      "learning_rate": 9.148226694075295e-06,
+      "loss": 0.3457,
+      "step": 698
+    },
+    {
+      "epoch": 0.21333740271631313,
+      "grad_norm": 1.7649363297171154,
+      "learning_rate": 9.14546432288554e-06,
+      "loss": 0.504,
+      "step": 699
+    },
+    {
+      "epoch": 0.21364260643979857,
+      "grad_norm": 1.5869344950607664,
+      "learning_rate": 9.14269789811846e-06,
+      "loss": 0.2746,
+      "step": 700
+    },
+    {
+      "epoch": 0.21394781016328399,
+      "grad_norm": 2.1981528156810266,
+      "learning_rate": 9.139927422479157e-06,
+      "loss": 0.3494,
+      "step": 701
+    },
+    {
+      "epoch": 0.21425301388676943,
+      "grad_norm": 1.5364449020919386,
+      "learning_rate": 9.137152898676698e-06,
+      "loss": 0.259,
+      "step": 702
+    },
+    {
+      "epoch": 0.21455821761025484,
+      "grad_norm": 1.5976641591815994,
+      "learning_rate": 9.134374329424107e-06,
+      "loss": 0.3714,
+      "step": 703
+    },
+    {
+      "epoch": 0.21486342133374028,
+      "grad_norm": 1.2686502539601563,
+      "learning_rate": 9.131591717438366e-06,
+      "loss": 0.3223,
+      "step": 704
+    },
+    {
+      "epoch": 0.2151686250572257,
+      "grad_norm": 1.246057539745606,
+      "learning_rate": 9.128805065440405e-06,
+      "loss": 0.2376,
+      "step": 705
+    },
+    {
+      "epoch": 0.2154738287807111,
+      "grad_norm": 1.7361721316316239,
+      "learning_rate": 9.126014376155113e-06,
+      "loss": 0.4981,
+      "step": 706
+    },
+    {
+      "epoch": 0.21577903250419656,
+      "grad_norm": 1.5341841942267511,
+      "learning_rate": 9.12321965231132e-06,
+      "loss": 0.3948,
+      "step": 707
+    },
+    {
+      "epoch": 0.21608423622768197,
+      "grad_norm": 1.564805123380453,
+      "learning_rate": 9.1204208966418e-06,
+      "loss": 0.3948,
+      "step": 708
+    },
+    {
+      "epoch": 0.2163894399511674,
+      "grad_norm": 1.3738642249897421,
+      "learning_rate": 9.117618111883276e-06,
+      "loss": 0.3615,
+      "step": 709
+    },
+    {
+      "epoch": 0.21669464367465283,
+      "grad_norm": 1.2994164991334436,
+      "learning_rate": 9.114811300776405e-06,
+      "loss": 0.2186,
+      "step": 710
+    },
+    {
+      "epoch": 0.21699984739813827,
+      "grad_norm": 1.7972182236315297,
+      "learning_rate": 9.112000466065785e-06,
+      "loss": 0.5655,
+      "step": 711
+    },
+    {
+      "epoch": 0.21730505112162368,
+      "grad_norm": 1.278135561724537,
+      "learning_rate": 9.109185610499943e-06,
+      "loss": 0.3853,
+      "step": 712
+    },
+    {
+      "epoch": 0.2176102548451091,
+      "grad_norm": 1.213829377522313,
+      "learning_rate": 9.106366736831347e-06,
+      "loss": 0.3143,
+      "step": 713
+    },
+    {
+      "epoch": 0.21791545856859454,
+      "grad_norm": 1.5786296679622351,
+      "learning_rate": 9.103543847816384e-06,
+      "loss": 0.3748,
+      "step": 714
+    },
+    {
+      "epoch": 0.21822066229207995,
+      "grad_norm": 1.1980380231780912,
+      "learning_rate": 9.100716946215374e-06,
+      "loss": 0.3257,
+      "step": 715
+    },
+    {
+      "epoch": 0.2185258660155654,
+      "grad_norm": 1.5748488156857616,
+      "learning_rate": 9.097886034792557e-06,
+      "loss": 0.3859,
+      "step": 716
+    },
+    {
+      "epoch": 0.2188310697390508,
+      "grad_norm": 1.4962734980956909,
+      "learning_rate": 9.095051116316095e-06,
+      "loss": 0.4612,
+      "step": 717
+    },
+    {
+      "epoch": 0.21913627346253625,
+      "grad_norm": 1.4269067789650183,
+      "learning_rate": 9.092212193558072e-06,
+      "loss": 0.3177,
+      "step": 718
+    },
+    {
+      "epoch": 0.21944147718602167,
+      "grad_norm": 1.2842950689313655,
+      "learning_rate": 9.089369269294483e-06,
+      "loss": 0.238,
+      "step": 719
+    },
+    {
+      "epoch": 0.2197466809095071,
+      "grad_norm": 1.3869072584515276,
+      "learning_rate": 9.086522346305233e-06,
+      "loss": 0.3534,
+      "step": 720
+    },
+    {
+      "epoch": 0.22005188463299252,
+      "grad_norm": 1.5861416389361607,
+      "learning_rate": 9.083671427374144e-06,
+      "loss": 0.404,
+      "step": 721
+    },
+    {
+      "epoch": 0.22035708835647794,
+      "grad_norm": 1.3575612826767118,
+      "learning_rate": 9.080816515288944e-06,
+      "loss": 0.2564,
+      "step": 722
+    },
+    {
+      "epoch": 0.22066229207996338,
+      "grad_norm": 1.6224055486381046,
+      "learning_rate": 9.077957612841262e-06,
+      "loss": 0.5335,
+      "step": 723
+    },
+    {
+      "epoch": 0.2209674958034488,
+      "grad_norm": 1.1424184312646575,
+      "learning_rate": 9.07509472282663e-06,
+      "loss": 0.2248,
+      "step": 724
+    },
+    {
+      "epoch": 0.22127269952693424,
+      "grad_norm": 0.9653403668195691,
+      "learning_rate": 9.07222784804448e-06,
+      "loss": 0.1492,
+      "step": 725
+    },
+    {
+      "epoch": 0.22157790325041965,
+      "grad_norm": 1.7736448595259362,
+      "learning_rate": 9.069356991298145e-06,
+      "loss": 0.3365,
+      "step": 726
+    },
+    {
+      "epoch": 0.2218831069739051,
+      "grad_norm": 1.5382460544138177,
+      "learning_rate": 9.066482155394843e-06,
+      "loss": 0.3641,
+      "step": 727
+    },
+    {
+      "epoch": 0.2221883106973905,
+      "grad_norm": 1.5490656282179018,
+      "learning_rate": 9.063603343145689e-06,
+      "loss": 0.451,
+      "step": 728
+    },
+    {
+      "epoch": 0.22249351442087595,
+      "grad_norm": 1.6477503496903847,
+      "learning_rate": 9.060720557365683e-06,
+      "loss": 0.3152,
+      "step": 729
+    },
+    {
+      "epoch": 0.22279871814436136,
+      "grad_norm": 1.3334962278440752,
+      "learning_rate": 9.057833800873712e-06,
+      "loss": 0.3514,
+      "step": 730
+    },
+    {
+      "epoch": 0.22310392186784678,
+      "grad_norm": 1.3285303282716252,
+      "learning_rate": 9.054943076492548e-06,
+      "loss": 0.2835,
+      "step": 731
+    },
+    {
+      "epoch": 0.22340912559133222,
+      "grad_norm": 1.275406047266298,
+      "learning_rate": 9.05204838704884e-06,
+      "loss": 0.2324,
+      "step": 732
+    },
+    {
+      "epoch": 0.22371432931481763,
+      "grad_norm": 1.6512597679062968,
+      "learning_rate": 9.04914973537311e-06,
+      "loss": 0.6219,
+      "step": 733
+    },
+    {
+      "epoch": 0.22401953303830308,
+      "grad_norm": 1.2059290028925398,
+      "learning_rate": 9.046247124299766e-06,
+      "loss": 0.3296,
+      "step": 734
+    },
+    {
+      "epoch": 0.2243247367617885,
+      "grad_norm": 1.5740026131575533,
+      "learning_rate": 9.043340556667076e-06,
+      "loss": 0.4929,
+      "step": 735
+    },
+    {
+      "epoch": 0.22462994048527393,
+      "grad_norm": 1.525534179809715,
+      "learning_rate": 9.040430035317184e-06,
+      "loss": 0.3855,
+      "step": 736
+    },
+    {
+      "epoch": 0.22493514420875935,
+      "grad_norm": 1.5007721185282292,
+      "learning_rate": 9.037515563096096e-06,
+      "loss": 0.2867,
+      "step": 737
+    },
+    {
+      "epoch": 0.22524034793224476,
+      "grad_norm": 1.7439351574891226,
+      "learning_rate": 9.034597142853685e-06,
+      "loss": 0.5206,
+      "step": 738
+    },
+    {
+      "epoch": 0.2255455516557302,
+      "grad_norm": 1.3832916895025305,
+      "learning_rate": 9.031674777443679e-06,
+      "loss": 0.3525,
+      "step": 739
+    },
+    {
+      "epoch": 0.22585075537921562,
+      "grad_norm": 1.689345092847421,
+      "learning_rate": 9.028748469723671e-06,
+      "loss": 0.605,
+      "step": 740
+    },
+    {
+      "epoch": 0.22615595910270106,
+      "grad_norm": 1.5414939302466633,
+      "learning_rate": 9.025818222555105e-06,
+      "loss": 0.4616,
+      "step": 741
+    },
+    {
+      "epoch": 0.22646116282618647,
+      "grad_norm": 1.572475405138349,
+      "learning_rate": 9.022884038803275e-06,
+      "loss": 0.4125,
+      "step": 742
+    },
+    {
+      "epoch": 0.22676636654967192,
+      "grad_norm": 1.6244915386812042,
+      "learning_rate": 9.019945921337328e-06,
+      "loss": 0.3892,
+      "step": 743
+    },
+    {
+      "epoch": 0.22707157027315733,
+      "grad_norm": 1.109983250603302,
+      "learning_rate": 9.017003873030254e-06,
+      "loss": 0.2775,
+      "step": 744
+    },
+    {
+      "epoch": 0.22737677399664277,
+      "grad_norm": 1.2908690769048217,
+      "learning_rate": 9.014057896758892e-06,
+      "loss": 0.2799,
+      "step": 745
+    },
+    {
+      "epoch": 0.2276819777201282,
+      "grad_norm": 1.3869598608885343,
+      "learning_rate": 9.011107995403917e-06,
+      "loss": 0.3109,
+      "step": 746
+    },
+    {
+      "epoch": 0.2279871814436136,
+      "grad_norm": 1.9493518241147973,
+      "learning_rate": 9.008154171849843e-06,
+      "loss": 0.3222,
+      "step": 747
+    },
+    {
+      "epoch": 0.22829238516709904,
+      "grad_norm": 1.1260119760970688,
+      "learning_rate": 9.005196428985024e-06,
+      "loss": 0.2015,
+      "step": 748
+    },
+    {
+      "epoch": 0.22859758889058446,
+      "grad_norm": 1.3395279047737458,
+      "learning_rate": 9.002234769701637e-06,
+      "loss": 0.3174,
+      "step": 749
+    },
+    {
+      "epoch": 0.2289027926140699,
+      "grad_norm": 1.127299131698768,
+      "learning_rate": 8.999269196895698e-06,
+      "loss": 0.3164,
+      "step": 750
+    },
+    {
+      "epoch": 0.22920799633755531,
+      "grad_norm": 1.1374371018678644,
+      "learning_rate": 8.996299713467044e-06,
+      "loss": 0.3252,
+      "step": 751
+    },
+    {
+      "epoch": 0.22951320006104076,
+      "grad_norm": 1.4696910359597422,
+      "learning_rate": 8.993326322319339e-06,
+      "loss": 0.4006,
+      "step": 752
+    },
+    {
+      "epoch": 0.22981840378452617,
+      "grad_norm": 1.862503625238974,
+      "learning_rate": 8.990349026360065e-06,
+      "loss": 0.5648,
+      "step": 753
+    },
+    {
+      "epoch": 0.23012360750801159,
+      "grad_norm": 1.768109354521003,
+      "learning_rate": 8.987367828500528e-06,
+      "loss": 0.4002,
+      "step": 754
+    },
+    {
+      "epoch": 0.23042881123149703,
+      "grad_norm": 1.6108609348753753,
+      "learning_rate": 8.984382731655842e-06,
+      "loss": 0.3703,
+      "step": 755
+    },
+    {
+      "epoch": 0.23073401495498244,
+      "grad_norm": 2.0879815191569153,
+      "learning_rate": 8.98139373874494e-06,
+      "loss": 0.4966,
+      "step": 756
+    },
+    {
+      "epoch": 0.23103921867846788,
+      "grad_norm": 1.627248748628783,
+      "learning_rate": 8.978400852690557e-06,
+      "loss": 0.4972,
+      "step": 757
+    },
+    {
+      "epoch": 0.2313444224019533,
+      "grad_norm": 1.6245326323749618,
+      "learning_rate": 8.975404076419244e-06,
+      "loss": 0.2992,
+      "step": 758
+    },
+    {
+      "epoch": 0.23164962612543874,
+      "grad_norm": 1.295122180020727,
+      "learning_rate": 8.972403412861354e-06,
+      "loss": 0.1872,
+      "step": 759
+    },
+    {
+      "epoch": 0.23195482984892415,
+      "grad_norm": 1.5772423116895944,
+      "learning_rate": 8.969398864951034e-06,
+      "loss": 0.3683,
+      "step": 760
+    },
+    {
+      "epoch": 0.2322600335724096,
+      "grad_norm": 1.2952086072241094,
+      "learning_rate": 8.966390435626233e-06,
+      "loss": 0.3407,
+      "step": 761
+    },
+    {
+      "epoch": 0.232565237295895,
+      "grad_norm": 1.350730242849987,
+      "learning_rate": 8.963378127828703e-06,
+      "loss": 0.3012,
+      "step": 762
+    },
+    {
+      "epoch": 0.23287044101938043,
+      "grad_norm": 1.757344365832517,
+      "learning_rate": 8.960361944503976e-06,
+      "loss": 0.4483,
+      "step": 763
+    },
+    {
+      "epoch": 0.23317564474286587,
+      "grad_norm": 1.4971687299475023,
+      "learning_rate": 8.957341888601382e-06,
+      "loss": 0.3751,
+      "step": 764
+    },
+    {
+      "epoch": 0.23348084846635128,
+      "grad_norm": 0.9570785510771707,
+      "learning_rate": 8.954317963074035e-06,
+      "loss": 0.1987,
+      "step": 765
+    },
+    {
+      "epoch": 0.23378605218983672,
+      "grad_norm": 1.8380667817910563,
+      "learning_rate": 8.951290170878834e-06,
+      "loss": 0.4731,
+      "step": 766
+    },
+    {
+      "epoch": 0.23409125591332214,
+      "grad_norm": 1.112944434897937,
+      "learning_rate": 8.948258514976456e-06,
+      "loss": 0.2407,
+      "step": 767
+    },
+    {
+      "epoch": 0.23439645963680758,
+      "grad_norm": 1.309366755129546,
+      "learning_rate": 8.94522299833136e-06,
+      "loss": 0.319,
+      "step": 768
+    },
+    {
+      "epoch": 0.234701663360293,
+      "grad_norm": 1.1451808921483004,
+      "learning_rate": 8.94218362391178e-06,
+      "loss": 0.2171,
+      "step": 769
+    },
+    {
+      "epoch": 0.2350068670837784,
+      "grad_norm": 1.428457846554008,
+      "learning_rate": 8.939140394689716e-06,
+      "loss": 0.294,
+      "step": 770
+    },
+    {
+      "epoch": 0.23531207080726385,
+      "grad_norm": 1.6548235526629411,
+      "learning_rate": 8.936093313640947e-06,
+      "loss": 0.4041,
+      "step": 771
+    },
+    {
+      "epoch": 0.23561727453074927,
+      "grad_norm": 1.4221218858946143,
+      "learning_rate": 8.93304238374501e-06,
+      "loss": 0.2095,
+      "step": 772
+    },
+    {
+      "epoch": 0.2359224782542347,
+      "grad_norm": 1.5383523764762757,
+      "learning_rate": 8.929987607985212e-06,
+      "loss": 0.478,
+      "step": 773
+    },
+    {
+      "epoch": 0.23622768197772012,
+      "grad_norm": 1.5292649150056814,
+      "learning_rate": 8.926928989348612e-06,
+      "loss": 0.4167,
+      "step": 774
+    },
+    {
+      "epoch": 0.23653288570120556,
+      "grad_norm": 1.6476214228386405,
+      "learning_rate": 8.923866530826037e-06,
+      "loss": 0.4843,
+      "step": 775
+    },
+    {
+      "epoch": 0.23683808942469098,
+      "grad_norm": 1.5924467106022067,
+      "learning_rate": 8.920800235412067e-06,
+      "loss": 0.407,
+      "step": 776
+    },
+    {
+      "epoch": 0.23714329314817642,
+      "grad_norm": 1.5594856941518516,
+      "learning_rate": 8.917730106105022e-06,
+      "loss": 0.3124,
+      "step": 777
+    },
+    {
+      "epoch": 0.23744849687166183,
+      "grad_norm": 1.386797749592837,
+      "learning_rate": 8.914656145906988e-06,
+      "loss": 0.3016,
+      "step": 778
+    },
+    {
+      "epoch": 0.23775370059514725,
+      "grad_norm": 1.3692351348081182,
+      "learning_rate": 8.911578357823784e-06,
+      "loss": 0.3401,
+      "step": 779
+    },
+    {
+      "epoch": 0.2380589043186327,
+      "grad_norm": 1.389706254561915,
+      "learning_rate": 8.90849674486498e-06,
+      "loss": 0.3724,
+      "step": 780
+    },
+    {
+      "epoch": 0.2383641080421181,
+      "grad_norm": 1.05409970328404,
+      "learning_rate": 8.90541131004388e-06,
+      "loss": 0.2848,
+      "step": 781
+    },
+    {
+      "epoch": 0.23866931176560355,
+      "grad_norm": 1.5742105715707222,
+      "learning_rate": 8.90232205637753e-06,
+      "loss": 0.6582,
+      "step": 782
+    },
+    {
+      "epoch": 0.23897451548908896,
+      "grad_norm": 1.4335601553241863,
+      "learning_rate": 8.899228986886709e-06,
+      "loss": 0.4478,
+      "step": 783
+    },
+    {
+      "epoch": 0.2392797192125744,
+      "grad_norm": 1.692276024021936,
+      "learning_rate": 8.896132104595925e-06,
+      "loss": 0.2317,
+      "step": 784
+    },
+    {
+      "epoch": 0.23958492293605982,
+      "grad_norm": 1.4666941437892842,
+      "learning_rate": 8.893031412533415e-06,
+      "loss": 0.4121,
+      "step": 785
+    },
+    {
+      "epoch": 0.23989012665954523,
+      "grad_norm": 1.7867397905919786,
+      "learning_rate": 8.889926913731144e-06,
+      "loss": 0.3591,
+      "step": 786
+    },
+    {
+      "epoch": 0.24019533038303068,
+      "grad_norm": 1.5710332933779532,
+      "learning_rate": 8.886818611224798e-06,
+      "loss": 0.3421,
+      "step": 787
+    },
+    {
+      "epoch": 0.2405005341065161,
+      "grad_norm": 1.0993405381050079,
+      "learning_rate": 8.883706508053778e-06,
+      "loss": 0.1907,
+      "step": 788
+    },
+    {
+      "epoch": 0.24080573783000153,
+      "grad_norm": 1.5126727738496022,
+      "learning_rate": 8.88059060726121e-06,
+      "loss": 0.5117,
+      "step": 789
+    },
+    {
+      "epoch": 0.24111094155348695,
+      "grad_norm": 1.0050194650159943,
+      "learning_rate": 8.877470911893927e-06,
+      "loss": 0.2927,
+      "step": 790
+    },
+    {
+      "epoch": 0.2414161452769724,
+      "grad_norm": 1.549032075208785,
+      "learning_rate": 8.874347425002474e-06,
+      "loss": 0.4823,
+      "step": 791
+    },
+    {
+      "epoch": 0.2417213490004578,
+      "grad_norm": 1.4199913878918462,
+      "learning_rate": 8.871220149641101e-06,
+      "loss": 0.372,
+      "step": 792
+    },
+    {
+      "epoch": 0.24202655272394324,
+      "grad_norm": 1.2204976389548388,
+      "learning_rate": 8.868089088867772e-06,
+      "loss": 0.4162,
+      "step": 793
+    },
+    {
+      "epoch": 0.24233175644742866,
+      "grad_norm": 1.2773847194249226,
+      "learning_rate": 8.86495424574414e-06,
+      "loss": 0.3559,
+      "step": 794
+    },
+    {
+      "epoch": 0.24263696017091407,
+      "grad_norm": 1.5146172702910354,
+      "learning_rate": 8.861815623335564e-06,
+      "loss": 0.4853,
+      "step": 795
+    },
+    {
+      "epoch": 0.24294216389439952,
+      "grad_norm": 1.2421815500176878,
+      "learning_rate": 8.858673224711097e-06,
+      "loss": 0.2322,
+      "step": 796
+    },
+    {
+      "epoch": 0.24324736761788493,
+      "grad_norm": 1.573303107817239,
+      "learning_rate": 8.855527052943482e-06,
+      "loss": 0.4221,
+      "step": 797
+    },
+    {
+      "epoch": 0.24355257134137037,
+      "grad_norm": 1.3348685555629807,
+      "learning_rate": 8.852377111109158e-06,
+      "loss": 0.4016,
+      "step": 798
+    },
+    {
+      "epoch": 0.2438577750648558,
+      "grad_norm": 1.4724771810683823,
+      "learning_rate": 8.849223402288244e-06,
+      "loss": 0.3673,
+      "step": 799
+    },
+    {
+      "epoch": 0.24416297878834123,
+      "grad_norm": 1.4577243169053915,
+      "learning_rate": 8.846065929564542e-06,
+      "loss": 0.3661,
+      "step": 800
+    },
+    {
+      "epoch": 0.24446818251182664,
+      "grad_norm": 1.2504498625919398,
+      "learning_rate": 8.842904696025542e-06,
+      "loss": 0.3142,
+      "step": 801
+    },
+    {
+      "epoch": 0.24477338623531206,
+      "grad_norm": 1.4501852015903842,
+      "learning_rate": 8.839739704762404e-06,
+      "loss": 0.4218,
+      "step": 802
+    },
+    {
+      "epoch": 0.2450785899587975,
+      "grad_norm": 1.814308645288012,
+      "learning_rate": 8.836570958869966e-06,
+      "loss": 0.2846,
+      "step": 803
+    },
+    {
+      "epoch": 0.2453837936822829,
+      "grad_norm": 1.4130518552602014,
+      "learning_rate": 8.833398461446736e-06,
+      "loss": 0.3661,
+      "step": 804
+    },
+    {
+      "epoch": 0.24568899740576836,
+      "grad_norm": 1.207043545854772,
+      "learning_rate": 8.83022221559489e-06,
+      "loss": 0.3083,
+      "step": 805
+    },
+    {
+      "epoch": 0.24599420112925377,
+      "grad_norm": 1.3695879979407652,
+      "learning_rate": 8.827042224420273e-06,
+      "loss": 0.3913,
+      "step": 806
+    },
+    {
+      "epoch": 0.2462994048527392,
+      "grad_norm": 1.509429933800135,
+      "learning_rate": 8.823858491032388e-06,
+      "loss": 0.4189,
+      "step": 807
+    },
+    {
+      "epoch": 0.24660460857622463,
+      "grad_norm": 1.5440694625132374,
+      "learning_rate": 8.820671018544398e-06,
+      "loss": 0.4917,
+      "step": 808
+    },
+    {
+      "epoch": 0.24690981229971007,
+      "grad_norm": 1.2648979736165928,
+      "learning_rate": 8.817479810073124e-06,
+      "loss": 0.2272,
+      "step": 809
+    },
+    {
+      "epoch": 0.24721501602319548,
+      "grad_norm": 1.5676001662475934,
+      "learning_rate": 8.814284868739038e-06,
+      "loss": 0.3194,
+      "step": 810
+    },
+    {
+      "epoch": 0.2475202197466809,
+      "grad_norm": 1.5083433807769027,
+      "learning_rate": 8.811086197666266e-06,
+      "loss": 0.4213,
+      "step": 811
+    },
+    {
+      "epoch": 0.24782542347016634,
+      "grad_norm": 1.428992764464654,
+      "learning_rate": 8.807883799982574e-06,
+      "loss": 0.3399,
+      "step": 812
+    },
+    {
+      "epoch": 0.24813062719365175,
+      "grad_norm": 1.0166441279138219,
+      "learning_rate": 8.80467767881938e-06,
+      "loss": 0.2195,
+      "step": 813
+    },
+    {
+      "epoch": 0.2484358309171372,
+      "grad_norm": 1.2422483417317973,
+      "learning_rate": 8.801467837311738e-06,
+      "loss": 0.2906,
+      "step": 814
+    },
+    {
+      "epoch": 0.2487410346406226,
+      "grad_norm": 1.419560487453445,
+      "learning_rate": 8.798254278598345e-06,
+      "loss": 0.4495,
+      "step": 815
+    },
+    {
+      "epoch": 0.24904623836410805,
+      "grad_norm": 1.16665115800268,
+      "learning_rate": 8.795037005821521e-06,
+      "loss": 0.2503,
+      "step": 816
+    },
+    {
+      "epoch": 0.24935144208759347,
+      "grad_norm": 1.4481353411677762,
+      "learning_rate": 8.791816022127233e-06,
+      "loss": 0.3363,
+      "step": 817
+    },
+    {
+      "epoch": 0.2496566458110789,
+      "grad_norm": 1.338264713753052,
+      "learning_rate": 8.788591330665065e-06,
+      "loss": 0.2325,
+      "step": 818
+    },
+    {
+      "epoch": 0.24996184953456432,
+      "grad_norm": 1.3941866237825566,
+      "learning_rate": 8.785362934588234e-06,
+      "loss": 0.4524,
+      "step": 819
+    },
+    {
+      "epoch": 0.25026705325804977,
+      "grad_norm": 1.508098753010285,
+      "learning_rate": 8.782130837053575e-06,
+      "loss": 0.3796,
+      "step": 820
+    },
+    {
+      "epoch": 0.2505722569815352,
+      "grad_norm": 1.3573838043362958,
+      "learning_rate": 8.778895041221544e-06,
+      "loss": 0.2623,
+      "step": 821
+    },
+    {
+      "epoch": 0.2508774607050206,
+      "grad_norm": 1.3709683700574562,
+      "learning_rate": 8.775655550256214e-06,
+      "loss": 0.3998,
+      "step": 822
+    },
+    {
+      "epoch": 0.251182664428506,
+      "grad_norm": 1.5964366482889705,
+      "learning_rate": 8.772412367325269e-06,
+      "loss": 0.455,
+      "step": 823
+    },
+    {
+      "epoch": 0.2514878681519915,
+      "grad_norm": 1.6611709904306184,
+      "learning_rate": 8.769165495600007e-06,
+      "loss": 0.5094,
+      "step": 824
+    },
+    {
+      "epoch": 0.2517930718754769,
+      "grad_norm": 1.1400080018013656,
+      "learning_rate": 8.76591493825533e-06,
+      "loss": 0.3997,
+      "step": 825
+    },
+    {
+      "epoch": 0.2520982755989623,
+      "grad_norm": 1.862472646627605,
+      "learning_rate": 8.76266069846974e-06,
+      "loss": 0.3198,
+      "step": 826
+    },
+    {
+      "epoch": 0.2524034793224477,
+      "grad_norm": 1.6623689872345544,
+      "learning_rate": 8.75940277942535e-06,
+      "loss": 0.5344,
+      "step": 827
+    },
+    {
+      "epoch": 0.25270868304593314,
+      "grad_norm": 1.4234557534815584,
+      "learning_rate": 8.756141184307865e-06,
+      "loss": 0.3674,
+      "step": 828
+    },
+    {
+      "epoch": 0.2530138867694186,
+      "grad_norm": 1.7406127381363719,
+      "learning_rate": 8.75287591630658e-06,
+      "loss": 0.4348,
+      "step": 829
+    },
+    {
+      "epoch": 0.253319090492904,
+      "grad_norm": 1.4903918741773559,
+      "learning_rate": 8.749606978614391e-06,
+      "loss": 0.2545,
+      "step": 830
+    },
+    {
+      "epoch": 0.25362429421638943,
+      "grad_norm": 1.8076324214942547,
+      "learning_rate": 8.746334374427774e-06,
+      "loss": 0.4559,
+      "step": 831
+    },
+    {
+      "epoch": 0.25392949793987485,
+      "grad_norm": 1.3532948940123553,
+      "learning_rate": 8.743058106946798e-06,
+      "loss": 0.2642,
+      "step": 832
+    },
+    {
+      "epoch": 0.2542347016633603,
+      "grad_norm": 1.6339304623976283,
+      "learning_rate": 8.739778179375105e-06,
+      "loss": 0.3866,
+      "step": 833
+    },
+    {
+      "epoch": 0.25453990538684573,
+      "grad_norm": 1.2532848113555968,
+      "learning_rate": 8.736494594919922e-06,
+      "loss": 0.2563,
+      "step": 834
+    },
+    {
+      "epoch": 0.25484510911033115,
+      "grad_norm": 1.062909552234949,
+      "learning_rate": 8.733207356792053e-06,
+      "loss": 0.2942,
+      "step": 835
+    },
+    {
+      "epoch": 0.25515031283381656,
+      "grad_norm": 1.3812270046063335,
+      "learning_rate": 8.72991646820587e-06,
+      "loss": 0.4591,
+      "step": 836
+    },
+    {
+      "epoch": 0.255455516557302,
+      "grad_norm": 1.401063935434944,
+      "learning_rate": 8.726621932379319e-06,
+      "loss": 0.3184,
+      "step": 837
+    },
+    {
+      "epoch": 0.25576072028078745,
+      "grad_norm": 1.394771394969279,
+      "learning_rate": 8.723323752533907e-06,
+      "loss": 0.2373,
+      "step": 838
+    },
+    {
+      "epoch": 0.25606592400427286,
+      "grad_norm": 1.4771850726658662,
+      "learning_rate": 8.72002193189471e-06,
+      "loss": 0.384,
+      "step": 839
+    },
+    {
+      "epoch": 0.2563711277277583,
+      "grad_norm": 1.278173381384045,
+      "learning_rate": 8.716716473690361e-06,
+      "loss": 0.2333,
+      "step": 840
+    },
+    {
+      "epoch": 0.2566763314512437,
+      "grad_norm": 1.5231335458997086,
+      "learning_rate": 8.713407381153052e-06,
+      "loss": 0.4792,
+      "step": 841
+    },
+    {
+      "epoch": 0.25698153517472916,
+      "grad_norm": 1.2138032535770222,
+      "learning_rate": 8.710094657518524e-06,
+      "loss": 0.1745,
+      "step": 842
+    },
+    {
+      "epoch": 0.2572867388982146,
+      "grad_norm": 1.6203262902374684,
+      "learning_rate": 8.706778306026073e-06,
+      "loss": 0.392,
+      "step": 843
+    },
+    {
+      "epoch": 0.2575919426217,
+      "grad_norm": 1.5129887571452891,
+      "learning_rate": 8.703458329918541e-06,
+      "loss": 0.3458,
+      "step": 844
+    },
+    {
+      "epoch": 0.2578971463451854,
+      "grad_norm": 1.579903535191777,
+      "learning_rate": 8.700134732442318e-06,
+      "loss": 0.4183,
+      "step": 845
+    },
+    {
+      "epoch": 0.2582023500686708,
+      "grad_norm": 1.6503110245689208,
+      "learning_rate": 8.69680751684733e-06,
+      "loss": 0.5822,
+      "step": 846
+    },
+    {
+      "epoch": 0.2585075537921563,
+      "grad_norm": 1.4889049274535986,
+      "learning_rate": 8.693476686387038e-06,
+      "loss": 0.4368,
+      "step": 847
+    },
+    {
+      "epoch": 0.2588127575156417,
+      "grad_norm": 1.1464358161897683,
+      "learning_rate": 8.69014224431845e-06,
+      "loss": 0.2515,
+      "step": 848
+    },
+    {
+      "epoch": 0.2591179612391271,
+      "grad_norm": 1.759904896160663,
+      "learning_rate": 8.686804193902092e-06,
+      "loss": 0.3729,
+      "step": 849
+    },
+    {
+      "epoch": 0.25942316496261253,
+      "grad_norm": 1.6474376376075388,
+      "learning_rate": 8.683462538402026e-06,
+      "loss": 0.2528,
+      "step": 850
+    },
+    {
+      "epoch": 0.25972836868609794,
+      "grad_norm": 1.6153335633088148,
+      "learning_rate": 8.680117281085839e-06,
+      "loss": 0.4734,
+      "step": 851
+    },
+    {
+      "epoch": 0.2600335724095834,
+      "grad_norm": 1.7594579836313058,
+      "learning_rate": 8.676768425224636e-06,
+      "loss": 0.3666,
+      "step": 852
+    },
+    {
+      "epoch": 0.2603387761330688,
+      "grad_norm": 2.7725868317444973,
+      "learning_rate": 8.673415974093046e-06,
+      "loss": 0.3024,
+      "step": 853
+    },
+    {
+      "epoch": 0.26064397985655424,
+      "grad_norm": 1.3139609425748013,
+      "learning_rate": 8.67005993096921e-06,
+      "loss": 0.2145,
+      "step": 854
+    },
+    {
+      "epoch": 0.26094918358003966,
+      "grad_norm": 1.5222687762088085,
+      "learning_rate": 8.66670029913478e-06,
+      "loss": 0.306,
+      "step": 855
+    },
+    {
+      "epoch": 0.2612543873035251,
+      "grad_norm": 1.225174136338348,
+      "learning_rate": 8.663337081874921e-06,
+      "loss": 0.1945,
+      "step": 856
+    },
+    {
+      "epoch": 0.26155959102701054,
+      "grad_norm": 1.5844026100416482,
+      "learning_rate": 8.659970282478303e-06,
+      "loss": 0.5302,
+      "step": 857
+    },
+    {
+      "epoch": 0.26186479475049596,
+      "grad_norm": 1.4779095030916518,
+      "learning_rate": 8.656599904237097e-06,
+      "loss": 0.492,
+      "step": 858
+    },
+    {
+      "epoch": 0.26216999847398137,
+      "grad_norm": 1.6248924513219443,
+      "learning_rate": 8.653225950446973e-06,
+      "loss": 0.5415,
+      "step": 859
+    },
+    {
+      "epoch": 0.2624752021974668,
+      "grad_norm": 1.3916992153237053,
+      "learning_rate": 8.649848424407103e-06,
+      "loss": 0.2716,
+      "step": 860
+    },
+    {
+      "epoch": 0.26278040592095225,
+      "grad_norm": 1.5314790982505428,
+      "learning_rate": 8.646467329420145e-06,
+      "loss": 0.486,
+      "step": 861
+    },
+    {
+      "epoch": 0.26308560964443767,
+      "grad_norm": 1.474685202158443,
+      "learning_rate": 8.64308266879225e-06,
+      "loss": 0.4944,
+      "step": 862
+    },
+    {
+      "epoch": 0.2633908133679231,
+      "grad_norm": 2.089424447750544,
+      "learning_rate": 8.639694445833056e-06,
+      "loss": 0.5607,
+      "step": 863
+    },
+    {
+      "epoch": 0.2636960170914085,
+      "grad_norm": 1.6425002507132653,
+      "learning_rate": 8.636302663855682e-06,
+      "loss": 0.3179,
+      "step": 864
+    },
+    {
+      "epoch": 0.26400122081489397,
+      "grad_norm": 1.2523110959560677,
+      "learning_rate": 8.632907326176732e-06,
+      "loss": 0.3304,
+      "step": 865
+    },
+    {
+      "epoch": 0.2643064245383794,
+      "grad_norm": 1.9713067301422675,
+      "learning_rate": 8.629508436116281e-06,
+      "loss": 0.3482,
+      "step": 866
+    },
+    {
+      "epoch": 0.2646116282618648,
+      "grad_norm": 1.574438794377994,
+      "learning_rate": 8.626105996997881e-06,
+      "loss": 0.3681,
+      "step": 867
+    },
+    {
+      "epoch": 0.2649168319853502,
+      "grad_norm": 1.975886896641686,
+      "learning_rate": 8.622700012148554e-06,
+      "loss": 0.3096,
+      "step": 868
+    },
+    {
+      "epoch": 0.2652220357088356,
+      "grad_norm": 1.205037763961463,
+      "learning_rate": 8.619290484898791e-06,
+      "loss": 0.2542,
+      "step": 869
+    },
+    {
+      "epoch": 0.2655272394323211,
+      "grad_norm": 1.410569402556539,
+      "learning_rate": 8.615877418582541e-06,
+      "loss": 0.3157,
+      "step": 870
+    },
+    {
+      "epoch": 0.2658324431558065,
+      "grad_norm": 1.4086097839086094,
+      "learning_rate": 8.61246081653722e-06,
+      "loss": 0.2501,
+      "step": 871
+    },
+    {
+      "epoch": 0.2661376468792919,
+      "grad_norm": 1.8258092836473052,
+      "learning_rate": 8.609040682103697e-06,
+      "loss": 0.5621,
+      "step": 872
+    },
+    {
+      "epoch": 0.26644285060277734,
+      "grad_norm": 1.5606382340865774,
+      "learning_rate": 8.6056170186263e-06,
+      "loss": 0.4155,
+      "step": 873
+    },
+    {
+      "epoch": 0.2667480543262628,
+      "grad_norm": 1.4738547712233099,
+      "learning_rate": 8.6021898294528e-06,
+      "loss": 0.3736,
+      "step": 874
+    },
+    {
+      "epoch": 0.2670532580497482,
+      "grad_norm": 1.2521166999589783,
+      "learning_rate": 8.598759117934424e-06,
+      "loss": 0.2707,
+      "step": 875
+    },
+    {
+      "epoch": 0.26735846177323364,
+      "grad_norm": 1.854567909402273,
+      "learning_rate": 8.595324887425839e-06,
+      "loss": 0.6055,
+      "step": 876
+    },
+    {
+      "epoch": 0.26766366549671905,
+      "grad_norm": 1.3156991751956812,
+      "learning_rate": 8.59188714128515e-06,
+      "loss": 0.1861,
+      "step": 877
+    },
+    {
+      "epoch": 0.26796886922020446,
+      "grad_norm": 1.2377078296039872,
+      "learning_rate": 8.58844588287391e-06,
+      "loss": 0.3001,
+      "step": 878
+    },
+    {
+      "epoch": 0.26827407294368993,
+      "grad_norm": 1.2660582471980335,
+      "learning_rate": 8.585001115557093e-06,
+      "loss": 0.3137,
+      "step": 879
+    },
+    {
+      "epoch": 0.26857927666717535,
+      "grad_norm": 1.6544970466723443,
+      "learning_rate": 8.581552842703114e-06,
+      "loss": 0.518,
+      "step": 880
+    },
+    {
+      "epoch": 0.26888448039066076,
+      "grad_norm": 1.7821640881601462,
+      "learning_rate": 8.57810106768381e-06,
+      "loss": 0.3717,
+      "step": 881
+    },
+    {
+      "epoch": 0.2691896841141462,
+      "grad_norm": 1.4045088267172914,
+      "learning_rate": 8.57464579387445e-06,
+      "loss": 0.507,
+      "step": 882
+    },
+    {
+      "epoch": 0.2694948878376316,
+      "grad_norm": 1.2562243938208713,
+      "learning_rate": 8.571187024653715e-06,
+      "loss": 0.2897,
+      "step": 883
+    },
+    {
+      "epoch": 0.26980009156111706,
+      "grad_norm": 1.647258608185494,
+      "learning_rate": 8.567724763403709e-06,
+      "loss": 0.4629,
+      "step": 884
+    },
+    {
+      "epoch": 0.2701052952846025,
+      "grad_norm": 1.4132863649325282,
+      "learning_rate": 8.564259013509952e-06,
+      "loss": 0.2711,
+      "step": 885
+    },
+    {
+      "epoch": 0.2704104990080879,
+      "grad_norm": 1.0546579427462797,
+      "learning_rate": 8.560789778361372e-06,
+      "loss": 0.1637,
+      "step": 886
+    },
+    {
+      "epoch": 0.2707157027315733,
+      "grad_norm": 1.8546260807501882,
+      "learning_rate": 8.557317061350303e-06,
+      "loss": 0.3391,
+      "step": 887
+    },
+    {
+      "epoch": 0.2710209064550588,
+      "grad_norm": 1.23880229781815,
+      "learning_rate": 8.553840865872492e-06,
+      "loss": 0.1623,
+      "step": 888
+    },
+    {
+      "epoch": 0.2713261101785442,
+      "grad_norm": 1.5924230549187588,
+      "learning_rate": 8.550361195327079e-06,
+      "loss": 0.3142,
+      "step": 889
+    },
+    {
+      "epoch": 0.2716313139020296,
+      "grad_norm": 1.3037915837064973,
+      "learning_rate": 8.546878053116609e-06,
+      "loss": 0.2978,
+      "step": 890
+    },
+    {
+      "epoch": 0.271936517625515,
+      "grad_norm": 1.2055323902273756,
+      "learning_rate": 8.543391442647013e-06,
+      "loss": 0.2708,
+      "step": 891
+    },
+    {
+      "epoch": 0.27224172134900043,
+      "grad_norm": 1.1232141519442895,
+      "learning_rate": 8.539901367327622e-06,
+      "loss": 0.2381,
+      "step": 892
+    },
+    {
+      "epoch": 0.2725469250724859,
+      "grad_norm": 1.6705142749495014,
+      "learning_rate": 8.53640783057115e-06,
+      "loss": 0.3586,
+      "step": 893
+    },
+    {
+      "epoch": 0.2728521287959713,
+      "grad_norm": 1.1042663181369634,
+      "learning_rate": 8.532910835793697e-06,
+      "loss": 0.142,
+      "step": 894
+    },
+    {
+      "epoch": 0.27315733251945673,
+      "grad_norm": 1.4442013980293982,
+      "learning_rate": 8.529410386414747e-06,
+      "loss": 0.3902,
+      "step": 895
+    },
+    {
+      "epoch": 0.27346253624294214,
+      "grad_norm": 1.1510191464990058,
+      "learning_rate": 8.525906485857156e-06,
+      "loss": 0.2468,
+      "step": 896
+    },
+    {
+      "epoch": 0.2737677399664276,
+      "grad_norm": 1.5572115857341773,
+      "learning_rate": 8.522399137547162e-06,
+      "loss": 0.3039,
+      "step": 897
+    },
+    {
+      "epoch": 0.27407294368991303,
+      "grad_norm": 1.7209680290555225,
+      "learning_rate": 8.518888344914365e-06,
+      "loss": 0.5024,
+      "step": 898
+    },
+    {
+      "epoch": 0.27437814741339844,
+      "grad_norm": 1.3970675444552212,
+      "learning_rate": 8.515374111391745e-06,
+      "loss": 0.3327,
+      "step": 899
+    },
+    {
+      "epoch": 0.27468335113688386,
+      "grad_norm": 1.5862907429686486,
+      "learning_rate": 8.511856440415635e-06,
+      "loss": 0.4007,
+      "step": 900
+    },
+    {
+      "epoch": 0.27498855486036927,
+      "grad_norm": 1.5069995329007524,
+      "learning_rate": 8.508335335425739e-06,
+      "loss": 0.415,
+      "step": 901
+    },
+    {
+      "epoch": 0.27529375858385474,
+      "grad_norm": 1.3954829540134184,
+      "learning_rate": 8.50481079986511e-06,
+      "loss": 0.2788,
+      "step": 902
+    },
+    {
+      "epoch": 0.27559896230734016,
+      "grad_norm": 1.8025662725066627,
+      "learning_rate": 8.501282837180165e-06,
+      "loss": 0.3723,
+      "step": 903
+    },
+    {
+      "epoch": 0.27590416603082557,
+      "grad_norm": 1.520821384052033,
+      "learning_rate": 8.49775145082066e-06,
+      "loss": 0.2988,
+      "step": 904
+    },
+    {
+      "epoch": 0.276209369754311,
+      "grad_norm": 1.604241860679225,
+      "learning_rate": 8.494216644239712e-06,
+      "loss": 0.2951,
+      "step": 905
+    },
+    {
+      "epoch": 0.27651457347779645,
+      "grad_norm": 1.3350559025036193,
+      "learning_rate": 8.490678420893773e-06,
+      "loss": 0.2941,
+      "step": 906
+    },
+    {
+      "epoch": 0.27681977720128187,
+      "grad_norm": 1.3989463178848165,
+      "learning_rate": 8.487136784242641e-06,
+      "loss": 0.3683,
+      "step": 907
+    },
+    {
+      "epoch": 0.2771249809247673,
+      "grad_norm": 1.6081399715718512,
+      "learning_rate": 8.483591737749448e-06,
+      "loss": 0.4503,
+      "step": 908
+    },
+    {
+      "epoch": 0.2774301846482527,
+      "grad_norm": 1.6045658886374856,
+      "learning_rate": 8.480043284880666e-06,
+      "loss": 0.3672,
+      "step": 909
+    },
+    {
+      "epoch": 0.2777353883717381,
+      "grad_norm": 1.5365693477891538,
+      "learning_rate": 8.476491429106088e-06,
+      "loss": 0.5018,
+      "step": 910
+    },
+    {
+      "epoch": 0.2780405920952236,
+      "grad_norm": 1.306165577473847,
+      "learning_rate": 8.472936173898846e-06,
+      "loss": 0.3336,
+      "step": 911
+    },
+    {
+      "epoch": 0.278345795818709,
+      "grad_norm": 1.2410262335833262,
+      "learning_rate": 8.469377522735387e-06,
+      "loss": 0.3175,
+      "step": 912
+    },
+    {
+      "epoch": 0.2786509995421944,
+      "grad_norm": 0.9618771928411501,
+      "learning_rate": 8.465815479095485e-06,
+      "loss": 0.2373,
+      "step": 913
+    },
+    {
+      "epoch": 0.2789562032656798,
+      "grad_norm": 2.139593836663555,
+      "learning_rate": 8.462250046462226e-06,
+      "loss": 0.4121,
+      "step": 914
+    },
+    {
+      "epoch": 0.2792614069891653,
+      "grad_norm": 1.3283828036351077,
+      "learning_rate": 8.458681228322014e-06,
+      "loss": 0.38,
+      "step": 915
+    },
+    {
+      "epoch": 0.2795666107126507,
+      "grad_norm": 1.4749702062668928,
+      "learning_rate": 8.455109028164561e-06,
+      "loss": 0.4707,
+      "step": 916
+    },
+    {
+      "epoch": 0.2798718144361361,
+      "grad_norm": 1.2064530552545678,
+      "learning_rate": 8.45153344948289e-06,
+      "loss": 0.3217,
+      "step": 917
+    },
+    {
+      "epoch": 0.28017701815962154,
+      "grad_norm": 1.6256035637532993,
+      "learning_rate": 8.44795449577332e-06,
+      "loss": 0.4455,
+      "step": 918
+    },
+    {
+      "epoch": 0.28048222188310695,
+      "grad_norm": 1.3626337563191635,
+      "learning_rate": 8.444372170535478e-06,
+      "loss": 0.3391,
+      "step": 919
+    },
+    {
+      "epoch": 0.2807874256065924,
+      "grad_norm": 1.2635973004795618,
+      "learning_rate": 8.440786477272282e-06,
+      "loss": 0.2515,
+      "step": 920
+    },
+    {
+      "epoch": 0.28109262933007784,
+      "grad_norm": 1.1911478731743688,
+      "learning_rate": 8.437197419489948e-06,
+      "loss": 0.2891,
+      "step": 921
+    },
+    {
+      "epoch": 0.28139783305356325,
+      "grad_norm": 1.3587095012788262,
+      "learning_rate": 8.43360500069798e-06,
+      "loss": 0.416,
+      "step": 922
+    },
+    {
+      "epoch": 0.28170303677704867,
+      "grad_norm": 1.3537329737331374,
+      "learning_rate": 8.430009224409167e-06,
+      "loss": 0.2851,
+      "step": 923
+    },
+    {
+      "epoch": 0.2820082405005341,
+      "grad_norm": 1.0905607319343713,
+      "learning_rate": 8.426410094139582e-06,
+      "loss": 0.219,
+      "step": 924
+    },
+    {
+      "epoch": 0.28231344422401955,
+      "grad_norm": 1.2957621936139718,
+      "learning_rate": 8.422807613408583e-06,
+      "loss": 0.2757,
+      "step": 925
+    },
+    {
+      "epoch": 0.28261864794750496,
+      "grad_norm": 1.5464703579715704,
+      "learning_rate": 8.419201785738794e-06,
+      "loss": 0.488,
+      "step": 926
+    },
+    {
+      "epoch": 0.2829238516709904,
+      "grad_norm": 1.5027192422666622,
+      "learning_rate": 8.415592614656119e-06,
+      "loss": 0.298,
+      "step": 927
+    },
+    {
+      "epoch": 0.2832290553944758,
+      "grad_norm": 1.5041494197684437,
+      "learning_rate": 8.411980103689727e-06,
+      "loss": 0.2037,
+      "step": 928
+    },
+    {
+      "epoch": 0.28353425911796126,
+      "grad_norm": 1.1718395482104975,
+      "learning_rate": 8.40836425637206e-06,
+      "loss": 0.1936,
+      "step": 929
+    },
+    {
+      "epoch": 0.2838394628414467,
+      "grad_norm": 1.4034578790843537,
+      "learning_rate": 8.404745076238816e-06,
+      "loss": 0.3402,
+      "step": 930
+    },
+    {
+      "epoch": 0.2841446665649321,
+      "grad_norm": 1.471548215059981,
+      "learning_rate": 8.40112256682895e-06,
+      "loss": 0.3633,
+      "step": 931
+    },
+    {
+      "epoch": 0.2844498702884175,
+      "grad_norm": 2.534571034931892,
+      "learning_rate": 8.39749673168468e-06,
+      "loss": 0.3617,
+      "step": 932
+    },
+    {
+      "epoch": 0.2847550740119029,
+      "grad_norm": 1.5477637618039504,
+      "learning_rate": 8.39386757435147e-06,
+      "loss": 0.4332,
+      "step": 933
+    },
+    {
+      "epoch": 0.2850602777353884,
+      "grad_norm": 1.058523928884,
+      "learning_rate": 8.390235098378037e-06,
+      "loss": 0.2018,
+      "step": 934
+    },
+    {
+      "epoch": 0.2853654814588738,
+      "grad_norm": 1.051963849726592,
+      "learning_rate": 8.38659930731634e-06,
+      "loss": 0.2024,
+      "step": 935
+    },
+    {
+      "epoch": 0.2856706851823592,
+      "grad_norm": 1.4307030578090003,
+      "learning_rate": 8.382960204721575e-06,
+      "loss": 0.3186,
+      "step": 936
+    },
+    {
+      "epoch": 0.28597588890584463,
+      "grad_norm": 1.2191946734542334,
+      "learning_rate": 8.37931779415219e-06,
+      "loss": 0.2907,
+      "step": 937
+    },
+    {
+      "epoch": 0.2862810926293301,
+      "grad_norm": 1.4629392730406188,
+      "learning_rate": 8.375672079169854e-06,
+      "loss": 0.3628,
+      "step": 938
+    },
+    {
+      "epoch": 0.2865862963528155,
+      "grad_norm": 1.2671740435957073,
+      "learning_rate": 8.372023063339472e-06,
+      "loss": 0.3895,
+      "step": 939
+    },
+    {
+      "epoch": 0.28689150007630093,
+      "grad_norm": 1.2843029586747219,
+      "learning_rate": 8.36837075022918e-06,
+      "loss": 0.2448,
+      "step": 940
+    },
+    {
+      "epoch": 0.28719670379978635,
+      "grad_norm": 1.225493384282322,
+      "learning_rate": 8.36471514341033e-06,
+      "loss": 0.3751,
+      "step": 941
+    },
+    {
+      "epoch": 0.28750190752327176,
+      "grad_norm": 1.118820597376438,
+      "learning_rate": 8.361056246457509e-06,
+      "loss": 0.2959,
+      "step": 942
+    },
+    {
+      "epoch": 0.28780711124675723,
+      "grad_norm": 1.5826287542337327,
+      "learning_rate": 8.357394062948501e-06,
+      "loss": 0.1865,
+      "step": 943
+    },
+    {
+      "epoch": 0.28811231497024264,
+      "grad_norm": 1.7322615384777234,
+      "learning_rate": 8.353728596464324e-06,
+      "loss": 0.3319,
+      "step": 944
+    },
+    {
+      "epoch": 0.28841751869372806,
+      "grad_norm": 1.1819946263218575,
+      "learning_rate": 8.350059850589189e-06,
+      "loss": 0.258,
+      "step": 945
+    },
+    {
+      "epoch": 0.2887227224172135,
+      "grad_norm": 1.2768421179691603,
+      "learning_rate": 8.34638782891053e-06,
+      "loss": 0.3505,
+      "step": 946
+    },
+    {
+      "epoch": 0.28902792614069894,
+      "grad_norm": 1.6765785010321739,
+      "learning_rate": 8.342712535018968e-06,
+      "loss": 0.3988,
+      "step": 947
+    },
+    {
+      "epoch": 0.28933312986418436,
+      "grad_norm": 1.6460901695759294,
+      "learning_rate": 8.339033972508335e-06,
+      "loss": 0.5964,
+      "step": 948
+    },
+    {
+      "epoch": 0.28963833358766977,
+      "grad_norm": 1.3645220250147991,
+      "learning_rate": 8.335352144975657e-06,
+      "loss": 0.3997,
+      "step": 949
+    },
+    {
+      "epoch": 0.2899435373111552,
+      "grad_norm": 1.565124139219312,
+      "learning_rate": 8.331667056021151e-06,
+      "loss": 0.2581,
+      "step": 950
+    },
+    {
+      "epoch": 0.2902487410346406,
+      "grad_norm": 1.2163034972670224,
+      "learning_rate": 8.32797870924822e-06,
+      "loss": 0.3549,
+      "step": 951
+    },
+    {
+      "epoch": 0.29055394475812607,
+      "grad_norm": 1.9963649856615253,
+      "learning_rate": 8.324287108263462e-06,
+      "loss": 0.4499,
+      "step": 952
+    },
+    {
+      "epoch": 0.2908591484816115,
+      "grad_norm": 1.5738812787701257,
+      "learning_rate": 8.320592256676646e-06,
+      "loss": 0.4303,
+      "step": 953
+    },
+    {
+      "epoch": 0.2911643522050969,
+      "grad_norm": 1.51492812662889,
+      "learning_rate": 8.316894158100728e-06,
+      "loss": 0.3071,
+      "step": 954
+    },
+    {
+      "epoch": 0.2914695559285823,
+      "grad_norm": 1.5701047086118967,
+      "learning_rate": 8.313192816151832e-06,
+      "loss": 0.3527,
+      "step": 955
+    },
+    {
+      "epoch": 0.29177475965206773,
+      "grad_norm": 1.6660195604391899,
+      "learning_rate": 8.309488234449261e-06,
+      "loss": 0.4016,
+      "step": 956
+    },
+    {
+      "epoch": 0.2920799633755532,
+      "grad_norm": 1.5308116363222675,
+      "learning_rate": 8.30578041661548e-06,
+      "loss": 0.3991,
+      "step": 957
+    },
+    {
+      "epoch": 0.2923851670990386,
+      "grad_norm": 1.5118616317530167,
+      "learning_rate": 8.302069366276121e-06,
+      "loss": 0.4968,
+      "step": 958
+    },
+    {
+      "epoch": 0.292690370822524,
+      "grad_norm": 1.6934474407382258,
+      "learning_rate": 8.298355087059978e-06,
+      "loss": 0.5735,
+      "step": 959
+    },
+    {
+      "epoch": 0.29299557454600944,
+      "grad_norm": 1.4097627634984986,
+      "learning_rate": 8.294637582598998e-06,
+      "loss": 0.3816,
+      "step": 960
+    },
+    {
+      "epoch": 0.2933007782694949,
+      "grad_norm": 1.4214156530194402,
+      "learning_rate": 8.290916856528288e-06,
+      "loss": 0.3706,
+      "step": 961
+    },
+    {
+      "epoch": 0.2936059819929803,
+      "grad_norm": 1.3762680622131698,
+      "learning_rate": 8.287192912486098e-06,
+      "loss": 0.2482,
+      "step": 962
+    },
+    {
+      "epoch": 0.29391118571646574,
+      "grad_norm": 1.6238119477399238,
+      "learning_rate": 8.283465754113832e-06,
+      "loss": 0.483,
+      "step": 963
+    },
+    {
+      "epoch": 0.29421638943995115,
+      "grad_norm": 1.580806305043778,
+      "learning_rate": 8.279735385056028e-06,
+      "loss": 0.3277,
+      "step": 964
+    },
+    {
+      "epoch": 0.29452159316343657,
+      "grad_norm": 1.9306479469286622,
+      "learning_rate": 8.276001808960374e-06,
+      "loss": 0.448,
+      "step": 965
+    },
+    {
+      "epoch": 0.29482679688692204,
+      "grad_norm": 1.5497884575246978,
+      "learning_rate": 8.272265029477686e-06,
+      "loss": 0.3917,
+      "step": 966
+    },
+    {
+      "epoch": 0.29513200061040745,
+      "grad_norm": 1.5393309383304865,
+      "learning_rate": 8.26852505026192e-06,
+      "loss": 0.3033,
+      "step": 967
+    },
+    {
+      "epoch": 0.29543720433389287,
+      "grad_norm": 1.6980365016594954,
+      "learning_rate": 8.26478187497015e-06,
+      "loss": 0.2664,
+      "step": 968
+    },
+    {
+      "epoch": 0.2957424080573783,
+      "grad_norm": 1.363460408765191,
+      "learning_rate": 8.261035507262582e-06,
+      "loss": 0.254,
+      "step": 969
+    },
+    {
+      "epoch": 0.29604761178086375,
+      "grad_norm": 1.6405225449314822,
+      "learning_rate": 8.257285950802547e-06,
+      "loss": 0.4717,
+      "step": 970
+    },
+    {
+      "epoch": 0.29635281550434917,
+      "grad_norm": 1.4402490121062985,
+      "learning_rate": 8.253533209256486e-06,
+      "loss": 0.2763,
+      "step": 971
+    },
+    {
+      "epoch": 0.2966580192278346,
+      "grad_norm": 1.8581821237579783,
+      "learning_rate": 8.24977728629396e-06,
+      "loss": 0.5867,
+      "step": 972
+    },
+    {
+      "epoch": 0.29696322295132,
+      "grad_norm": 1.6644196462856506,
+      "learning_rate": 8.246018185587642e-06,
+      "loss": 0.6931,
+      "step": 973
+    },
+    {
+      "epoch": 0.2972684266748054,
+      "grad_norm": 1.7226101616124239,
+      "learning_rate": 8.242255910813308e-06,
+      "loss": 0.4707,
+      "step": 974
+    },
+    {
+      "epoch": 0.2975736303982909,
+      "grad_norm": 1.559017343729282,
+      "learning_rate": 8.238490465649837e-06,
+      "loss": 0.3913,
+      "step": 975
+    },
+    {
+      "epoch": 0.2978788341217763,
+      "grad_norm": 2.355733929284259,
+      "learning_rate": 8.234721853779212e-06,
+      "loss": 0.3725,
+      "step": 976
+    },
+    {
+      "epoch": 0.2981840378452617,
+      "grad_norm": 1.800766090445105,
+      "learning_rate": 8.230950078886512e-06,
+      "loss": 0.3071,
+      "step": 977
+    },
+    {
+      "epoch": 0.2984892415687471,
+      "grad_norm": 2.5023327434487097,
+      "learning_rate": 8.227175144659908e-06,
+      "loss": 0.5576,
+      "step": 978
+    },
+    {
+      "epoch": 0.2987944452922326,
+      "grad_norm": 1.657243682467184,
+      "learning_rate": 8.22339705479066e-06,
+      "loss": 0.3635,
+      "step": 979
+    },
+    {
+      "epoch": 0.299099649015718,
+      "grad_norm": 1.4497278352698486,
+      "learning_rate": 8.219615812973111e-06,
+      "loss": 0.3942,
+      "step": 980
+    },
+    {
+      "epoch": 0.2994048527392034,
+      "grad_norm": 1.4676463261815744,
+      "learning_rate": 8.215831422904694e-06,
+      "loss": 0.3701,
+      "step": 981
+    },
+    {
+      "epoch": 0.29971005646268883,
+      "grad_norm": 1.4714044973041538,
+      "learning_rate": 8.212043888285913e-06,
+      "loss": 0.3646,
+      "step": 982
+    },
+    {
+      "epoch": 0.30001526018617425,
+      "grad_norm": 1.8962188440528207,
+      "learning_rate": 8.208253212820349e-06,
+      "loss": 0.459,
+      "step": 983
+    },
+    {
+      "epoch": 0.3003204639096597,
+      "grad_norm": 1.5089138139920484,
+      "learning_rate": 8.204459400214657e-06,
+      "loss": 0.3636,
+      "step": 984
+    },
+    {
+      "epoch": 0.30062566763314513,
+      "grad_norm": 1.5444120585905727,
+      "learning_rate": 8.200662454178553e-06,
+      "loss": 0.3727,
+      "step": 985
+    },
+    {
+      "epoch": 0.30093087135663055,
+      "grad_norm": 1.3153314475890103,
+      "learning_rate": 8.196862378424826e-06,
+      "loss": 0.2808,
+      "step": 986
+    },
+    {
+      "epoch": 0.30123607508011596,
+      "grad_norm": 1.645327268934486,
+      "learning_rate": 8.193059176669317e-06,
+      "loss": 0.4536,
+      "step": 987
+    },
+    {
+      "epoch": 0.30154127880360143,
+      "grad_norm": 1.343818307959044,
+      "learning_rate": 8.189252852630927e-06,
+      "loss": 0.3137,
+      "step": 988
+    },
+    {
+      "epoch": 0.30184648252708685,
+      "grad_norm": 1.079353461920414,
+      "learning_rate": 8.185443410031613e-06,
+      "loss": 0.1587,
+      "step": 989
+    },
+    {
+      "epoch": 0.30215168625057226,
+      "grad_norm": 1.3923715489039352,
+      "learning_rate": 8.181630852596378e-06,
+      "loss": 0.5496,
+      "step": 990
+    },
+    {
+      "epoch": 0.3024568899740577,
+      "grad_norm": 1.3696798223323383,
+      "learning_rate": 8.17781518405327e-06,
+      "loss": 0.2037,
+      "step": 991
+    },
+    {
+      "epoch": 0.3027620936975431,
+      "grad_norm": 1.6749786375590874,
+      "learning_rate": 8.173996408133382e-06,
+      "loss": 0.523,
+      "step": 992
+    },
+    {
+      "epoch": 0.30306729742102856,
+      "grad_norm": 1.129331462466354,
+      "learning_rate": 8.170174528570845e-06,
+      "loss": 0.2696,
+      "step": 993
+    },
+    {
+      "epoch": 0.303372501144514,
+      "grad_norm": 1.4335553322615482,
+      "learning_rate": 8.16634954910282e-06,
+      "loss": 0.3102,
+      "step": 994
+    },
+    {
+      "epoch": 0.3036777048679994,
+      "grad_norm": 1.9054489954275935,
+      "learning_rate": 8.162521473469509e-06,
+      "loss": 0.4552,
+      "step": 995
+    },
+    {
+      "epoch": 0.3039829085914848,
+      "grad_norm": 1.8579421733648953,
+      "learning_rate": 8.158690305414132e-06,
+      "loss": 0.3536,
+      "step": 996
+    },
+    {
+      "epoch": 0.3042881123149702,
+      "grad_norm": 2.2709514225577223,
+      "learning_rate": 8.154856048682938e-06,
+      "loss": 0.4632,
+      "step": 997
+    },
+    {
+      "epoch": 0.3045933160384557,
+      "grad_norm": 1.5737905369537974,
+      "learning_rate": 8.151018707025194e-06,
+      "loss": 0.2503,
+      "step": 998
+    },
+    {
+      "epoch": 0.3048985197619411,
+      "grad_norm": 1.298904393760468,
+      "learning_rate": 8.147178284193185e-06,
+      "loss": 0.3482,
+      "step": 999
+    },
+    {
+      "epoch": 0.3052037234854265,
+      "grad_norm": 2.3561598742827132,
+      "learning_rate": 8.143334783942207e-06,
+      "loss": 0.3419,
+      "step": 1000
+    },
+    {
+      "epoch": 0.30550892720891193,
+      "grad_norm": 1.1397332664109803,
+      "learning_rate": 8.13948821003057e-06,
+      "loss": 0.264,
+      "step": 1001
+    },
+    {
+      "epoch": 0.3058141309323974,
+      "grad_norm": 1.664228189305695,
+      "learning_rate": 8.135638566219581e-06,
+      "loss": 0.6021,
+      "step": 1002
+    },
+    {
+      "epoch": 0.3061193346558828,
+      "grad_norm": 1.2828873662440865,
+      "learning_rate": 8.131785856273558e-06,
+      "loss": 0.3529,
+      "step": 1003
+    },
+    {
+      "epoch": 0.3064245383793682,
+      "grad_norm": 1.54799572968202,
+      "learning_rate": 8.127930083959813e-06,
+      "loss": 0.41,
+      "step": 1004
+    },
+    {
+      "epoch": 0.30672974210285364,
+      "grad_norm": 1.6015124057285708,
+      "learning_rate": 8.124071253048648e-06,
+      "loss": 0.5602,
+      "step": 1005
+    },
+    {
+      "epoch": 0.30703494582633906,
+      "grad_norm": 1.5019375684946967,
+      "learning_rate": 8.120209367313364e-06,
+      "loss": 0.5064,
+      "step": 1006
+    },
+    {
+      "epoch": 0.3073401495498245,
+      "grad_norm": 1.4841104372065577,
+      "learning_rate": 8.116344430530248e-06,
+      "loss": 0.2734,
+      "step": 1007
+    },
+    {
+      "epoch": 0.30764535327330994,
+      "grad_norm": 1.4979575079530576,
+      "learning_rate": 8.112476446478561e-06,
+      "loss": 0.4476,
+      "step": 1008
+    },
+    {
+      "epoch": 0.30795055699679535,
+      "grad_norm": 1.48012915266827,
+      "learning_rate": 8.108605418940555e-06,
+      "loss": 0.3478,
+      "step": 1009
+    },
+    {
+      "epoch": 0.30825576072028077,
+      "grad_norm": 1.7441608560206838,
+      "learning_rate": 8.104731351701456e-06,
+      "loss": 0.4198,
+      "step": 1010
+    },
+    {
+      "epoch": 0.30856096444376624,
+      "grad_norm": 1.6603083282826114,
+      "learning_rate": 8.100854248549453e-06,
+      "loss": 0.2977,
+      "step": 1011
+    },
+    {
+      "epoch": 0.30886616816725165,
+      "grad_norm": 1.4905130468269074,
+      "learning_rate": 8.096974113275716e-06,
+      "loss": 0.4249,
+      "step": 1012
+    },
+    {
+      "epoch": 0.30917137189073707,
+      "grad_norm": 1.3913148176340926,
+      "learning_rate": 8.093090949674373e-06,
+      "loss": 0.3232,
+      "step": 1013
+    },
+    {
+      "epoch": 0.3094765756142225,
+      "grad_norm": 1.896351174136801,
+      "learning_rate": 8.089204761542515e-06,
+      "loss": 0.6316,
+      "step": 1014
+    },
+    {
+      "epoch": 0.3097817793377079,
+      "grad_norm": 1.3098254883257765,
+      "learning_rate": 8.08531555268019e-06,
+      "loss": 0.3526,
+      "step": 1015
+    },
+    {
+      "epoch": 0.31008698306119337,
+      "grad_norm": 1.2414711860204384,
+      "learning_rate": 8.081423326890399e-06,
+      "loss": 0.2527,
+      "step": 1016
+    },
+    {
+      "epoch": 0.3103921867846788,
+      "grad_norm": 1.651694240025547,
+      "learning_rate": 8.077528087979095e-06,
+      "loss": 0.4595,
+      "step": 1017
+    },
+    {
+      "epoch": 0.3106973905081642,
+      "grad_norm": 1.2490722074076572,
+      "learning_rate": 8.073629839755177e-06,
+      "loss": 0.3143,
+      "step": 1018
+    },
+    {
+      "epoch": 0.3110025942316496,
+      "grad_norm": 1.4922248875592679,
+      "learning_rate": 8.069728586030483e-06,
+      "loss": 0.4735,
+      "step": 1019
+    },
+    {
+      "epoch": 0.3113077979551351,
+      "grad_norm": 1.3134160251209652,
+      "learning_rate": 8.065824330619796e-06,
+      "loss": 0.3399,
+      "step": 1020
+    },
+    {
+      "epoch": 0.3116130016786205,
+      "grad_norm": 1.3819445456159205,
+      "learning_rate": 8.061917077340828e-06,
+      "loss": 0.4378,
+      "step": 1021
+    },
+    {
+      "epoch": 0.3119182054021059,
+      "grad_norm": 1.3067268833489771,
+      "learning_rate": 8.05800683001423e-06,
+      "loss": 0.3642,
+      "step": 1022
+    },
+    {
+      "epoch": 0.3122234091255913,
+      "grad_norm": 1.2386206994963245,
+      "learning_rate": 8.054093592463572e-06,
+      "loss": 0.3116,
+      "step": 1023
+    },
+    {
+      "epoch": 0.31252861284907674,
+      "grad_norm": 1.5821932486019623,
+      "learning_rate": 8.050177368515353e-06,
+      "loss": 0.3499,
+      "step": 1024
+    },
+    {
+      "epoch": 0.3128338165725622,
+      "grad_norm": 1.2768808628569455,
+      "learning_rate": 8.046258161998993e-06,
+      "loss": 0.1659,
+      "step": 1025
+    },
+    {
+      "epoch": 0.3131390202960476,
+      "grad_norm": 1.2575220756033998,
+      "learning_rate": 8.042335976746822e-06,
+      "loss": 0.2405,
+      "step": 1026
+    },
+    {
+      "epoch": 0.31344422401953304,
+      "grad_norm": 1.8245135116139561,
+      "learning_rate": 8.038410816594093e-06,
+      "loss": 0.4366,
+      "step": 1027
+    },
+    {
+      "epoch": 0.31374942774301845,
+      "grad_norm": 1.4466165793811705,
+      "learning_rate": 8.034482685378959e-06,
+      "loss": 0.4082,
+      "step": 1028
+    },
+    {
+      "epoch": 0.31405463146650386,
+      "grad_norm": 1.5442738868035615,
+      "learning_rate": 8.030551586942484e-06,
+      "loss": 0.4539,
+      "step": 1029
+    },
+    {
+      "epoch": 0.31435983518998933,
+      "grad_norm": 1.2964103969559602,
+      "learning_rate": 8.026617525128628e-06,
+      "loss": 0.3801,
+      "step": 1030
+    },
+    {
+      "epoch": 0.31466503891347475,
+      "grad_norm": 1.872123164314918,
+      "learning_rate": 8.022680503784252e-06,
+      "loss": 0.3698,
+      "step": 1031
+    },
+    {
+      "epoch": 0.31497024263696016,
+      "grad_norm": 1.3752871993705764,
+      "learning_rate": 8.018740526759115e-06,
+      "loss": 0.3825,
+      "step": 1032
+    },
+    {
+      "epoch": 0.3152754463604456,
+      "grad_norm": 0.9331829492009497,
+      "learning_rate": 8.014797597905856e-06,
+      "loss": 0.2683,
+      "step": 1033
+    },
+    {
+      "epoch": 0.31558065008393105,
+      "grad_norm": 1.5771299945709583,
+      "learning_rate": 8.01085172108001e-06,
+      "loss": 0.5906,
+      "step": 1034
+    },
+    {
+      "epoch": 0.31588585380741646,
+      "grad_norm": 1.239300714325851,
+      "learning_rate": 8.006902900139989e-06,
+      "loss": 0.3648,
+      "step": 1035
+    },
+    {
+      "epoch": 0.3161910575309019,
+      "grad_norm": 1.058310326503882,
+      "learning_rate": 8.002951138947085e-06,
+      "loss": 0.228,
+      "step": 1036
+    },
+    {
+      "epoch": 0.3164962612543873,
+      "grad_norm": 1.1824475268280543,
+      "learning_rate": 7.99899644136547e-06,
+      "loss": 0.2105,
+      "step": 1037
+    },
+    {
+      "epoch": 0.3168014649778727,
+      "grad_norm": 1.5959459771356375,
+      "learning_rate": 7.995038811262176e-06,
+      "loss": 0.4822,
+      "step": 1038
+    },
+    {
+      "epoch": 0.3171066687013582,
+      "grad_norm": 1.0890532877681574,
+      "learning_rate": 7.991078252507117e-06,
+      "loss": 0.301,
+      "step": 1039
+    },
+    {
+      "epoch": 0.3174118724248436,
+      "grad_norm": 1.5858524092975068,
+      "learning_rate": 7.987114768973059e-06,
+      "loss": 0.3329,
+      "step": 1040
+    },
+    {
+      "epoch": 0.317717076148329,
+      "grad_norm": 1.2871444766825402,
+      "learning_rate": 7.983148364535633e-06,
+      "loss": 0.3842,
+      "step": 1041
+    },
+    {
+      "epoch": 0.3180222798718144,
+      "grad_norm": 1.6762774250879164,
+      "learning_rate": 7.979179043073327e-06,
+      "loss": 0.6428,
+      "step": 1042
+    },
+    {
+      "epoch": 0.3183274835952999,
+      "grad_norm": 1.1260557519041783,
+      "learning_rate": 7.975206808467477e-06,
+      "loss": 0.2912,
+      "step": 1043
+    },
+    {
+      "epoch": 0.3186326873187853,
+      "grad_norm": 1.4814440250076806,
+      "learning_rate": 7.971231664602273e-06,
+      "loss": 0.428,
+      "step": 1044
+    },
+    {
+      "epoch": 0.3189378910422707,
+      "grad_norm": 1.5470069289542376,
+      "learning_rate": 7.967253615364746e-06,
+      "loss": 0.326,
+      "step": 1045
+    },
+    {
+      "epoch": 0.31924309476575613,
+      "grad_norm": 1.6492330791767074,
+      "learning_rate": 7.963272664644765e-06,
+      "loss": 0.4851,
+      "step": 1046
+    },
+    {
+      "epoch": 0.31954829848924154,
+      "grad_norm": 1.1387453660755529,
+      "learning_rate": 7.95928881633505e-06,
+      "loss": 0.1819,
+      "step": 1047
+    },
+    {
+      "epoch": 0.319853502212727,
+      "grad_norm": 1.528342152399863,
+      "learning_rate": 7.955302074331136e-06,
+      "loss": 0.2465,
+      "step": 1048
+    },
+    {
+      "epoch": 0.32015870593621243,
+      "grad_norm": 1.3953796712292619,
+      "learning_rate": 7.9513124425314e-06,
+      "loss": 0.3372,
+      "step": 1049
+    },
+    {
+      "epoch": 0.32046390965969784,
+      "grad_norm": 1.5878443787469112,
+      "learning_rate": 7.94731992483704e-06,
+      "loss": 0.3593,
+      "step": 1050
+    },
+    {
+      "epoch": 0.32076911338318326,
+      "grad_norm": 1.5355544022885335,
+      "learning_rate": 7.943324525152077e-06,
+      "loss": 0.3395,
+      "step": 1051
+    },
+    {
+      "epoch": 0.3210743171066687,
+      "grad_norm": 1.2530592675223244,
+      "learning_rate": 7.93932624738335e-06,
+      "loss": 0.2164,
+      "step": 1052
+    },
+    {
+      "epoch": 0.32137952083015414,
+      "grad_norm": 4.526521517561823,
+      "learning_rate": 7.935325095440511e-06,
+      "loss": 0.2893,
+      "step": 1053
+    },
+    {
+      "epoch": 0.32168472455363956,
+      "grad_norm": 1.4025910565877562,
+      "learning_rate": 7.931321073236024e-06,
+      "loss": 0.3703,
+      "step": 1054
+    },
+    {
+      "epoch": 0.32198992827712497,
+      "grad_norm": 1.6381535397997637,
+      "learning_rate": 7.927314184685164e-06,
+      "loss": 0.3823,
+      "step": 1055
+    },
+    {
+      "epoch": 0.3222951320006104,
+      "grad_norm": 1.6423873881343087,
+      "learning_rate": 7.923304433705999e-06,
+      "loss": 0.3951,
+      "step": 1056
+    },
+    {
+      "epoch": 0.32260033572409585,
+      "grad_norm": 1.8735606709031154,
+      "learning_rate": 7.919291824219402e-06,
+      "loss": 0.2828,
+      "step": 1057
+    },
+    {
+      "epoch": 0.32290553944758127,
+      "grad_norm": 1.3931902680953308,
+      "learning_rate": 7.91527636014904e-06,
+      "loss": 0.2629,
+      "step": 1058
+    },
+    {
+      "epoch": 0.3232107431710667,
+      "grad_norm": 1.3616683731784345,
+      "learning_rate": 7.911258045421374e-06,
+      "loss": 0.3049,
+      "step": 1059
+    },
+    {
+      "epoch": 0.3235159468945521,
+      "grad_norm": 1.8477138002235685,
+      "learning_rate": 7.90723688396565e-06,
+      "loss": 0.5091,
+      "step": 1060
+    },
+    {
+      "epoch": 0.32382115061803757,
+      "grad_norm": 2.1715762804280048,
+      "learning_rate": 7.903212879713894e-06,
+      "loss": 0.3296,
+      "step": 1061
+    },
+    {
+      "epoch": 0.324126354341523,
+      "grad_norm": 1.2259722218688094,
+      "learning_rate": 7.89918603660092e-06,
+      "loss": 0.2084,
+      "step": 1062
+    },
+    {
+      "epoch": 0.3244315580650084,
+      "grad_norm": 1.5260897510578109,
+      "learning_rate": 7.895156358564308e-06,
+      "loss": 0.3144,
+      "step": 1063
+    },
+    {
+      "epoch": 0.3247367617884938,
+      "grad_norm": 1.2101539291645254,
+      "learning_rate": 7.891123849544421e-06,
+      "loss": 0.2139,
+      "step": 1064
+    },
+    {
+      "epoch": 0.3250419655119792,
+      "grad_norm": 1.4587532074802312,
+      "learning_rate": 7.887088513484383e-06,
+      "loss": 0.3954,
+      "step": 1065
+    },
+    {
+      "epoch": 0.3253471692354647,
+      "grad_norm": 1.6684881111981207,
+      "learning_rate": 7.883050354330085e-06,
+      "loss": 0.3655,
+      "step": 1066
+    },
+    {
+      "epoch": 0.3256523729589501,
+      "grad_norm": 1.307546988348603,
+      "learning_rate": 7.879009376030173e-06,
+      "loss": 0.2016,
+      "step": 1067
+    },
+    {
+      "epoch": 0.3259575766824355,
+      "grad_norm": 1.382846724092064,
+      "learning_rate": 7.87496558253606e-06,
+      "loss": 0.3516,
+      "step": 1068
+    },
+    {
+      "epoch": 0.32626278040592094,
+      "grad_norm": 1.420119155129035,
+      "learning_rate": 7.870918977801902e-06,
+      "loss": 0.4247,
+      "step": 1069
+    },
+    {
+      "epoch": 0.32656798412940635,
+      "grad_norm": 2.0170283200834227,
+      "learning_rate": 7.866869565784611e-06,
+      "loss": 0.4074,
+      "step": 1070
+    },
+    {
+      "epoch": 0.3268731878528918,
+      "grad_norm": 1.3132837770926757,
+      "learning_rate": 7.86281735044384e-06,
+      "loss": 0.3554,
+      "step": 1071
+    },
+    {
+      "epoch": 0.32717839157637724,
+      "grad_norm": 1.491650229990386,
+      "learning_rate": 7.858762335741984e-06,
+      "loss": 0.4311,
+      "step": 1072
+    },
+    {
+      "epoch": 0.32748359529986265,
+      "grad_norm": 1.3015830744734889,
+      "learning_rate": 7.854704525644175e-06,
+      "loss": 0.4368,
+      "step": 1073
+    },
+    {
+      "epoch": 0.32778879902334807,
+      "grad_norm": 1.240655590227942,
+      "learning_rate": 7.85064392411828e-06,
+      "loss": 0.3503,
+      "step": 1074
+    },
+    {
+      "epoch": 0.32809400274683354,
+      "grad_norm": 1.0370573041527442,
+      "learning_rate": 7.846580535134895e-06,
+      "loss": 0.2173,
+      "step": 1075
+    },
+    {
+      "epoch": 0.32839920647031895,
+      "grad_norm": 1.490198377421961,
+      "learning_rate": 7.842514362667341e-06,
+      "loss": 0.4797,
+      "step": 1076
+    },
+    {
+      "epoch": 0.32870441019380436,
+      "grad_norm": 1.2483676742868937,
+      "learning_rate": 7.83844541069166e-06,
+      "loss": 0.3117,
+      "step": 1077
+    },
+    {
+      "epoch": 0.3290096139172898,
+      "grad_norm": 1.1118596151916136,
+      "learning_rate": 7.834373683186614e-06,
+      "loss": 0.3425,
+      "step": 1078
+    },
+    {
+      "epoch": 0.3293148176407752,
+      "grad_norm": 1.1498236044067436,
+      "learning_rate": 7.830299184133676e-06,
+      "loss": 0.3116,
+      "step": 1079
+    },
+    {
+      "epoch": 0.32962002136426066,
+      "grad_norm": 1.330890961706903,
+      "learning_rate": 7.826221917517034e-06,
+      "loss": 0.326,
+      "step": 1080
+    },
+    {
+      "epoch": 0.3299252250877461,
+      "grad_norm": 1.1391107096366442,
+      "learning_rate": 7.822141887323577e-06,
+      "loss": 0.2253,
+      "step": 1081
+    },
+    {
+      "epoch": 0.3302304288112315,
+      "grad_norm": 1.0899804689249892,
+      "learning_rate": 7.8180590975429e-06,
+      "loss": 0.3972,
+      "step": 1082
+    },
+    {
+      "epoch": 0.3305356325347169,
+      "grad_norm": 1.1421995399769407,
+      "learning_rate": 7.813973552167293e-06,
+      "loss": 0.1951,
+      "step": 1083
+    },
+    {
+      "epoch": 0.3308408362582024,
+      "grad_norm": 1.6301828043180877,
+      "learning_rate": 7.809885255191745e-06,
+      "loss": 0.5203,
+      "step": 1084
+    },
+    {
+      "epoch": 0.3311460399816878,
+      "grad_norm": 1.1015959389663672,
+      "learning_rate": 7.805794210613934e-06,
+      "loss": 0.1868,
+      "step": 1085
+    },
+    {
+      "epoch": 0.3314512437051732,
+      "grad_norm": 1.0719373857232206,
+      "learning_rate": 7.801700422434218e-06,
+      "loss": 0.3546,
+      "step": 1086
+    },
+    {
+      "epoch": 0.3317564474286586,
+      "grad_norm": 1.7069619729935679,
+      "learning_rate": 7.79760389465565e-06,
+      "loss": 0.2883,
+      "step": 1087
+    },
+    {
+      "epoch": 0.33206165115214403,
+      "grad_norm": 1.1374914944508905,
+      "learning_rate": 7.793504631283952e-06,
+      "loss": 0.2634,
+      "step": 1088
+    },
+    {
+      "epoch": 0.3323668548756295,
+      "grad_norm": 1.407791140920363,
+      "learning_rate": 7.789402636327526e-06,
+      "loss": 0.4067,
+      "step": 1089
+    },
+    {
+      "epoch": 0.3326720585991149,
+      "grad_norm": 1.1160900354196726,
+      "learning_rate": 7.785297913797441e-06,
+      "loss": 0.2304,
+      "step": 1090
+    },
+    {
+      "epoch": 0.33297726232260033,
+      "grad_norm": 1.2290853807544064,
+      "learning_rate": 7.78119046770744e-06,
+      "loss": 0.3382,
+      "step": 1091
+    },
+    {
+      "epoch": 0.33328246604608575,
+      "grad_norm": 1.582064705954116,
+      "learning_rate": 7.777080302073922e-06,
+      "loss": 0.4928,
+      "step": 1092
+    },
+    {
+      "epoch": 0.3335876697695712,
+      "grad_norm": 1.1020461299123707,
+      "learning_rate": 7.772967420915948e-06,
+      "loss": 0.2539,
+      "step": 1093
+    },
+    {
+      "epoch": 0.33389287349305663,
+      "grad_norm": 1.2964060291640898,
+      "learning_rate": 7.768851828255233e-06,
+      "loss": 0.3398,
+      "step": 1094
+    },
+    {
+      "epoch": 0.33419807721654204,
+      "grad_norm": 1.2035385015753215,
+      "learning_rate": 7.764733528116148e-06,
+      "loss": 0.2841,
+      "step": 1095
+    },
+    {
+      "epoch": 0.33450328094002746,
+      "grad_norm": 1.4026090323213516,
+      "learning_rate": 7.760612524525708e-06,
+      "loss": 0.3719,
+      "step": 1096
+    },
+    {
+      "epoch": 0.3348084846635129,
+      "grad_norm": 1.44136903296084,
+      "learning_rate": 7.75648882151357e-06,
+      "loss": 0.3998,
+      "step": 1097
+    },
+    {
+      "epoch": 0.33511368838699834,
+      "grad_norm": 1.491339317646278,
+      "learning_rate": 7.752362423112032e-06,
+      "loss": 0.4328,
+      "step": 1098
+    },
+    {
+      "epoch": 0.33541889211048376,
+      "grad_norm": 1.0421144222506495,
+      "learning_rate": 7.74823333335603e-06,
+      "loss": 0.1908,
+      "step": 1099
+    },
+    {
+      "epoch": 0.33572409583396917,
+      "grad_norm": 1.2708409522823558,
+      "learning_rate": 7.744101556283129e-06,
+      "loss": 0.3129,
+      "step": 1100
+    },
+    {
+      "epoch": 0.3360292995574546,
+      "grad_norm": 1.6982843745883154,
+      "learning_rate": 7.73996709593352e-06,
+      "loss": 0.4516,
+      "step": 1101
+    },
+    {
+      "epoch": 0.33633450328094,
+      "grad_norm": 1.3745112417433676,
+      "learning_rate": 7.735829956350026e-06,
+      "loss": 0.3835,
+      "step": 1102
+    },
+    {
+      "epoch": 0.33663970700442547,
+      "grad_norm": 1.0289081222992171,
+      "learning_rate": 7.731690141578076e-06,
+      "loss": 0.23,
+      "step": 1103
+    },
+    {
+      "epoch": 0.3369449107279109,
+      "grad_norm": 1.3309142499739386,
+      "learning_rate": 7.72754765566573e-06,
+      "loss": 0.2879,
+      "step": 1104
+    },
+    {
+      "epoch": 0.3372501144513963,
+      "grad_norm": 1.7026299273281889,
+      "learning_rate": 7.72340250266365e-06,
+      "loss": 0.6118,
+      "step": 1105
+    },
+    {
+      "epoch": 0.3375553181748817,
+      "grad_norm": 1.3708586314610167,
+      "learning_rate": 7.71925468662511e-06,
+      "loss": 0.3695,
+      "step": 1106
+    },
+    {
+      "epoch": 0.3378605218983672,
+      "grad_norm": 1.138227058539724,
+      "learning_rate": 7.715104211605987e-06,
+      "loss": 0.3156,
+      "step": 1107
+    },
+    {
+      "epoch": 0.3381657256218526,
+      "grad_norm": 1.6312680765955443,
+      "learning_rate": 7.710951081664758e-06,
+      "loss": 0.3716,
+      "step": 1108
+    },
+    {
+      "epoch": 0.338470929345338,
+      "grad_norm": 1.43186060111883,
+      "learning_rate": 7.706795300862493e-06,
+      "loss": 0.4303,
+      "step": 1109
+    },
+    {
+      "epoch": 0.3387761330688234,
+      "grad_norm": 1.295302897418865,
+      "learning_rate": 7.702636873262859e-06,
+      "loss": 0.2334,
+      "step": 1110
+    },
+    {
+      "epoch": 0.33908133679230884,
+      "grad_norm": 1.2674378022051007,
+      "learning_rate": 7.69847580293211e-06,
+      "loss": 0.3056,
+      "step": 1111
+    },
+    {
+      "epoch": 0.3393865405157943,
+      "grad_norm": 1.326251694716828,
+      "learning_rate": 7.694312093939084e-06,
+      "loss": 0.2555,
+      "step": 1112
+    },
+    {
+      "epoch": 0.3396917442392797,
+      "grad_norm": 1.525503000572152,
+      "learning_rate": 7.690145750355198e-06,
+      "loss": 0.3448,
+      "step": 1113
+    },
+    {
+      "epoch": 0.33999694796276514,
+      "grad_norm": 1.3733356980578115,
+      "learning_rate": 7.685976776254446e-06,
+      "loss": 0.2901,
+      "step": 1114
+    },
+    {
+      "epoch": 0.34030215168625055,
+      "grad_norm": 1.23597237288206,
+      "learning_rate": 7.681805175713392e-06,
+      "loss": 0.2144,
+      "step": 1115
+    },
+    {
+      "epoch": 0.340607355409736,
+      "grad_norm": 1.5777935019830391,
+      "learning_rate": 7.677630952811172e-06,
+      "loss": 0.5106,
+      "step": 1116
+    },
+    {
+      "epoch": 0.34091255913322144,
+      "grad_norm": 1.171188713025467,
+      "learning_rate": 7.673454111629486e-06,
+      "loss": 0.1928,
+      "step": 1117
+    },
+    {
+      "epoch": 0.34121776285670685,
+      "grad_norm": 1.3533361626972884,
+      "learning_rate": 7.669274656252588e-06,
+      "loss": 0.2716,
+      "step": 1118
+    },
+    {
+      "epoch": 0.34152296658019227,
+      "grad_norm": 1.3706061645549534,
+      "learning_rate": 7.665092590767298e-06,
+      "loss": 0.2415,
+      "step": 1119
+    },
+    {
+      "epoch": 0.3418281703036777,
+      "grad_norm": 1.444083686347945,
+      "learning_rate": 7.66090791926298e-06,
+      "loss": 0.2384,
+      "step": 1120
+    },
+    {
+      "epoch": 0.34213337402716315,
+      "grad_norm": 1.7114127557234347,
+      "learning_rate": 7.65672064583155e-06,
+      "loss": 0.422,
+      "step": 1121
+    },
+    {
+      "epoch": 0.34243857775064857,
+      "grad_norm": 1.5259644507270844,
+      "learning_rate": 7.652530774567468e-06,
+      "loss": 0.5226,
+      "step": 1122
+    },
+    {
+      "epoch": 0.342743781474134,
+      "grad_norm": 0.9911746164065847,
+      "learning_rate": 7.648338309567735e-06,
+      "loss": 0.2267,
+      "step": 1123
+    },
+    {
+      "epoch": 0.3430489851976194,
+      "grad_norm": 1.0166151648645703,
+      "learning_rate": 7.644143254931887e-06,
+      "loss": 0.2104,
+      "step": 1124
+    },
+    {
+      "epoch": 0.34335418892110486,
+      "grad_norm": 1.6720832623689992,
+      "learning_rate": 7.63994561476199e-06,
+      "loss": 0.5396,
+      "step": 1125
+    },
+    {
+      "epoch": 0.3436593926445903,
+      "grad_norm": 1.5678839432039253,
+      "learning_rate": 7.635745393162643e-06,
+      "loss": 0.4492,
+      "step": 1126
+    },
+    {
+      "epoch": 0.3439645963680757,
+      "grad_norm": 1.8116112207083335,
+      "learning_rate": 7.631542594240968e-06,
+      "loss": 0.3977,
+      "step": 1127
+    },
+    {
+      "epoch": 0.3442698000915611,
+      "grad_norm": 1.473465027051407,
+      "learning_rate": 7.627337222106602e-06,
+      "loss": 0.3924,
+      "step": 1128
+    },
+    {
+      "epoch": 0.3445750038150465,
+      "grad_norm": 1.8018368091168058,
+      "learning_rate": 7.623129280871706e-06,
+      "loss": 0.6324,
+      "step": 1129
+    },
+    {
+      "epoch": 0.344880207538532,
+      "grad_norm": 1.2871463289874823,
+      "learning_rate": 7.618918774650947e-06,
+      "loss": 0.317,
+      "step": 1130
+    },
+    {
+      "epoch": 0.3451854112620174,
+      "grad_norm": 1.3104766508523527,
+      "learning_rate": 7.614705707561505e-06,
+      "loss": 0.1865,
+      "step": 1131
+    },
+    {
+      "epoch": 0.3454906149855028,
+      "grad_norm": 1.4958005772283107,
+      "learning_rate": 7.610490083723059e-06,
+      "loss": 0.3286,
+      "step": 1132
+    },
+    {
+      "epoch": 0.34579581870898823,
+      "grad_norm": 1.3334807622385083,
+      "learning_rate": 7.606271907257793e-06,
+      "loss": 0.1803,
+      "step": 1133
+    },
+    {
+      "epoch": 0.3461010224324737,
+      "grad_norm": 1.58655509762167,
+      "learning_rate": 7.602051182290382e-06,
+      "loss": 0.428,
+      "step": 1134
+    },
+    {
+      "epoch": 0.3464062261559591,
+      "grad_norm": 1.6285506850668747,
+      "learning_rate": 7.597827912947998e-06,
+      "loss": 0.3046,
+      "step": 1135
+    },
+    {
+      "epoch": 0.34671142987944453,
+      "grad_norm": 1.3766123681447917,
+      "learning_rate": 7.593602103360298e-06,
+      "loss": 0.2579,
+      "step": 1136
+    },
+    {
+      "epoch": 0.34701663360292995,
+      "grad_norm": 1.7542194180322948,
+      "learning_rate": 7.589373757659424e-06,
+      "loss": 0.2563,
+      "step": 1137
+    },
+    {
+      "epoch": 0.34732183732641536,
+      "grad_norm": 1.1682456208270866,
+      "learning_rate": 7.585142879979998e-06,
+      "loss": 0.3068,
+      "step": 1138
+    },
+    {
+      "epoch": 0.34762704104990083,
+      "grad_norm": 1.1214377908787607,
+      "learning_rate": 7.580909474459117e-06,
+      "loss": 0.1778,
+      "step": 1139
+    },
+    {
+      "epoch": 0.34793224477338625,
+      "grad_norm": 1.5446511669234666,
+      "learning_rate": 7.576673545236349e-06,
+      "loss": 0.2251,
+      "step": 1140
+    },
+    {
+      "epoch": 0.34823744849687166,
+      "grad_norm": 1.2705911290898333,
+      "learning_rate": 7.572435096453734e-06,
+      "loss": 0.2938,
+      "step": 1141
+    },
+    {
+      "epoch": 0.3485426522203571,
+      "grad_norm": 1.493297862812399,
+      "learning_rate": 7.5681941322557685e-06,
+      "loss": 0.3066,
+      "step": 1142
+    },
+    {
+      "epoch": 0.3488478559438425,
+      "grad_norm": 1.0677076510295394,
+      "learning_rate": 7.563950656789416e-06,
+      "loss": 0.2338,
+      "step": 1143
+    },
+    {
+      "epoch": 0.34915305966732796,
+      "grad_norm": 1.2516527217201623,
+      "learning_rate": 7.559704674204091e-06,
+      "loss": 0.3347,
+      "step": 1144
+    },
+    {
+      "epoch": 0.3494582633908134,
+      "grad_norm": 1.702080362504898,
+      "learning_rate": 7.555456188651665e-06,
+      "loss": 0.4124,
+      "step": 1145
+    },
+    {
+      "epoch": 0.3497634671142988,
+      "grad_norm": 1.539828192304715,
+      "learning_rate": 7.551205204286447e-06,
+      "loss": 0.462,
+      "step": 1146
+    },
+    {
+      "epoch": 0.3500686708377842,
+      "grad_norm": 1.3087519763541762,
+      "learning_rate": 7.546951725265198e-06,
+      "loss": 0.2483,
+      "step": 1147
+    },
+    {
+      "epoch": 0.35037387456126967,
+      "grad_norm": 1.43063223220777,
+      "learning_rate": 7.542695755747116e-06,
+      "loss": 0.3615,
+      "step": 1148
+    },
+    {
+      "epoch": 0.3506790782847551,
+      "grad_norm": 1.5523078380591047,
+      "learning_rate": 7.538437299893836e-06,
+      "loss": 0.5231,
+      "step": 1149
+    },
+    {
+      "epoch": 0.3509842820082405,
+      "grad_norm": 1.2101188107160756,
+      "learning_rate": 7.534176361869418e-06,
+      "loss": 0.3176,
+      "step": 1150
+    },
+    {
+      "epoch": 0.3512894857317259,
+      "grad_norm": 1.4239347860105078,
+      "learning_rate": 7.529912945840359e-06,
+      "loss": 0.3347,
+      "step": 1151
+    },
+    {
+      "epoch": 0.35159468945521133,
+      "grad_norm": 1.3610805029033388,
+      "learning_rate": 7.5256470559755665e-06,
+      "loss": 0.3266,
+      "step": 1152
+    },
+    {
+      "epoch": 0.3518998931786968,
+      "grad_norm": 1.6132233172860933,
+      "learning_rate": 7.521378696446381e-06,
+      "loss": 0.3327,
+      "step": 1153
+    },
+    {
+      "epoch": 0.3522050969021822,
+      "grad_norm": 1.2930810424354338,
+      "learning_rate": 7.517107871426548e-06,
+      "loss": 0.1709,
+      "step": 1154
+    },
+    {
+      "epoch": 0.3525103006256676,
+      "grad_norm": 1.6807461972866704,
+      "learning_rate": 7.512834585092226e-06,
+      "loss": 0.3168,
+      "step": 1155
+    },
+    {
+      "epoch": 0.35281550434915304,
+      "grad_norm": 1.7732192926139976,
+      "learning_rate": 7.508558841621981e-06,
+      "loss": 0.5005,
+      "step": 1156
+    },
+    {
+      "epoch": 0.3531207080726385,
+      "grad_norm": 1.4750885241824656,
+      "learning_rate": 7.504280645196784e-06,
+      "loss": 0.3684,
+      "step": 1157
+    },
+    {
+      "epoch": 0.3534259117961239,
+      "grad_norm": 1.5202750221957153,
+      "learning_rate": 7.500000000000001e-06,
+      "loss": 0.1899,
+      "step": 1158
+    },
+    {
+      "epoch": 0.35373111551960934,
+      "grad_norm": 1.1595814112433622,
+      "learning_rate": 7.495716910217393e-06,
+      "loss": 0.3105,
+      "step": 1159
+    },
+    {
+      "epoch": 0.35403631924309475,
+      "grad_norm": 1.4534366232509097,
+      "learning_rate": 7.491431380037113e-06,
+      "loss": 0.3465,
+      "step": 1160
+    },
+    {
+      "epoch": 0.35434152296658017,
+      "grad_norm": 1.156548538845248,
+      "learning_rate": 7.4871434136497e-06,
+      "loss": 0.2893,
+      "step": 1161
+    },
+    {
+      "epoch": 0.35464672669006564,
+      "grad_norm": 1.4538056615012076,
+      "learning_rate": 7.482853015248075e-06,
+      "loss": 0.3667,
+      "step": 1162
+    },
+    {
+      "epoch": 0.35495193041355105,
+      "grad_norm": 1.6556975684961233,
+      "learning_rate": 7.478560189027536e-06,
+      "loss": 0.2493,
+      "step": 1163
+    },
+    {
+      "epoch": 0.35525713413703647,
+      "grad_norm": 1.6844460438765754,
+      "learning_rate": 7.474264939185756e-06,
+      "loss": 0.1798,
+      "step": 1164
+    },
+    {
+      "epoch": 0.3555623378605219,
+      "grad_norm": 1.9475210131391785,
+      "learning_rate": 7.469967269922777e-06,
+      "loss": 0.5204,
+      "step": 1165
+    },
+    {
+      "epoch": 0.35586754158400735,
+      "grad_norm": 2.793715839902533,
+      "learning_rate": 7.465667185441009e-06,
+      "loss": 0.3077,
+      "step": 1166
+    },
+    {
+      "epoch": 0.35617274530749277,
+      "grad_norm": 1.4888770645456522,
+      "learning_rate": 7.4613646899452205e-06,
+      "loss": 0.4135,
+      "step": 1167
+    },
+    {
+      "epoch": 0.3564779490309782,
+      "grad_norm": 1.1987516506211549,
+      "learning_rate": 7.457059787642541e-06,
+      "loss": 0.2382,
+      "step": 1168
+    },
+    {
+      "epoch": 0.3567831527544636,
+      "grad_norm": 1.4813551853352487,
+      "learning_rate": 7.452752482742452e-06,
+      "loss": 0.217,
+      "step": 1169
+    },
+    {
+      "epoch": 0.357088356477949,
+      "grad_norm": 1.6930510736367361,
+      "learning_rate": 7.448442779456781e-06,
+      "loss": 0.5214,
+      "step": 1170
+    },
+    {
+      "epoch": 0.3573935602014345,
+      "grad_norm": 2.190573141270535,
+      "learning_rate": 7.444130681999708e-06,
+      "loss": 0.2428,
+      "step": 1171
+    },
+    {
+      "epoch": 0.3576987639249199,
+      "grad_norm": 1.3779387713699958,
+      "learning_rate": 7.439816194587748e-06,
+      "loss": 0.3164,
+      "step": 1172
+    },
+    {
+      "epoch": 0.3580039676484053,
+      "grad_norm": 1.092055070131237,
+      "learning_rate": 7.435499321439754e-06,
+      "loss": 0.2772,
+      "step": 1173
+    },
+    {
+      "epoch": 0.3583091713718907,
+      "grad_norm": 1.3792533977854593,
+      "learning_rate": 7.431180066776913e-06,
+      "loss": 0.3562,
+      "step": 1174
+    },
+    {
+      "epoch": 0.35861437509537614,
+      "grad_norm": 1.2656833023017504,
+      "learning_rate": 7.426858434822742e-06,
+      "loss": 0.2438,
+      "step": 1175
+    },
+    {
+      "epoch": 0.3589195788188616,
+      "grad_norm": 1.4217271832774543,
+      "learning_rate": 7.42253442980308e-06,
+      "loss": 0.4452,
+      "step": 1176
+    },
+    {
+      "epoch": 0.359224782542347,
+      "grad_norm": 1.2647369471621235,
+      "learning_rate": 7.418208055946088e-06,
+      "loss": 0.3769,
+      "step": 1177
+    },
+    {
+      "epoch": 0.35952998626583244,
+      "grad_norm": 1.5573904799259919,
+      "learning_rate": 7.413879317482242e-06,
+      "loss": 0.2497,
+      "step": 1178
+    },
+    {
+      "epoch": 0.35983518998931785,
+      "grad_norm": 1.1158649645978311,
+      "learning_rate": 7.409548218644332e-06,
+      "loss": 0.3706,
+      "step": 1179
+    },
+    {
+      "epoch": 0.3601403937128033,
+      "grad_norm": 1.325219870264579,
+      "learning_rate": 7.4052147636674545e-06,
+      "loss": 0.2988,
+      "step": 1180
+    },
+    {
+      "epoch": 0.36044559743628873,
+      "grad_norm": 1.339413943736962,
+      "learning_rate": 7.400878956789011e-06,
+      "loss": 0.3791,
+      "step": 1181
+    },
+    {
+      "epoch": 0.36075080115977415,
+      "grad_norm": 1.3236780103464618,
+      "learning_rate": 7.396540802248704e-06,
+      "loss": 0.3051,
+      "step": 1182
+    },
+    {
+      "epoch": 0.36105600488325956,
+      "grad_norm": 1.4621900605001774,
+      "learning_rate": 7.392200304288526e-06,
+      "loss": 0.4029,
+      "step": 1183
+    },
+    {
+      "epoch": 0.361361208606745,
+      "grad_norm": 1.3004193418247747,
+      "learning_rate": 7.387857467152767e-06,
+      "loss": 0.3573,
+      "step": 1184
+    },
+    {
+      "epoch": 0.36166641233023045,
+      "grad_norm": 1.3403584017356904,
+      "learning_rate": 7.383512295088002e-06,
+      "loss": 0.301,
+      "step": 1185
+    },
+    {
+      "epoch": 0.36197161605371586,
+      "grad_norm": 1.3326412828848118,
+      "learning_rate": 7.379164792343091e-06,
+      "loss": 0.2346,
+      "step": 1186
+    },
+    {
+      "epoch": 0.3622768197772013,
+      "grad_norm": 1.3231894845074195,
+      "learning_rate": 7.37481496316917e-06,
+      "loss": 0.3924,
+      "step": 1187
+    },
+    {
+      "epoch": 0.3625820235006867,
+      "grad_norm": 1.2230476508538641,
+      "learning_rate": 7.370462811819651e-06,
+      "loss": 0.3316,
+      "step": 1188
+    },
+    {
+      "epoch": 0.36288722722417216,
+      "grad_norm": 1.6324841474899627,
+      "learning_rate": 7.366108342550217e-06,
+      "loss": 0.2683,
+      "step": 1189
+    },
+    {
+      "epoch": 0.3631924309476576,
+      "grad_norm": 1.1650127142511193,
+      "learning_rate": 7.361751559618819e-06,
+      "loss": 0.3151,
+      "step": 1190
+    },
+    {
+      "epoch": 0.363497634671143,
+      "grad_norm": 1.7617352605413967,
+      "learning_rate": 7.357392467285667e-06,
+      "loss": 0.3985,
+      "step": 1191
+    },
+    {
+      "epoch": 0.3638028383946284,
+      "grad_norm": 1.6177177849776165,
+      "learning_rate": 7.353031069813231e-06,
+      "loss": 0.3853,
+      "step": 1192
+    },
+    {
+      "epoch": 0.3641080421181138,
+      "grad_norm": 1.4672814189844252,
+      "learning_rate": 7.348667371466238e-06,
+      "loss": 0.3228,
+      "step": 1193
+    },
+    {
+      "epoch": 0.3644132458415993,
+      "grad_norm": 1.429990475762935,
+      "learning_rate": 7.344301376511659e-06,
+      "loss": 0.3216,
+      "step": 1194
+    },
+    {
+      "epoch": 0.3647184495650847,
+      "grad_norm": 1.2518101936436028,
+      "learning_rate": 7.339933089218716e-06,
+      "loss": 0.3541,
+      "step": 1195
+    },
+    {
+      "epoch": 0.3650236532885701,
+      "grad_norm": 1.5100278244233902,
+      "learning_rate": 7.335562513858868e-06,
+      "loss": 0.3777,
+      "step": 1196
+    },
+    {
+      "epoch": 0.36532885701205553,
+      "grad_norm": 1.3864156902279452,
+      "learning_rate": 7.331189654705816e-06,
+      "loss": 0.3451,
+      "step": 1197
+    },
+    {
+      "epoch": 0.365634060735541,
+      "grad_norm": 1.3148051409795818,
+      "learning_rate": 7.326814516035491e-06,
+      "loss": 0.2568,
+      "step": 1198
+    },
+    {
+      "epoch": 0.3659392644590264,
+      "grad_norm": 1.1783966682135776,
+      "learning_rate": 7.322437102126052e-06,
+      "loss": 0.409,
+      "step": 1199
+    },
+    {
+      "epoch": 0.36624446818251183,
+      "grad_norm": 1.3482269405611969,
+      "learning_rate": 7.318057417257886e-06,
+      "loss": 0.4714,
+      "step": 1200
+    },
+    {
+      "epoch": 0.36654967190599724,
+      "grad_norm": 1.6310707549202617,
+      "learning_rate": 7.3136754657136e-06,
+      "loss": 0.2499,
+      "step": 1201
+    },
+    {
+      "epoch": 0.36685487562948266,
+      "grad_norm": 1.039768581410541,
+      "learning_rate": 7.309291251778016e-06,
+      "loss": 0.272,
+      "step": 1202
+    },
+    {
+      "epoch": 0.3671600793529681,
+      "grad_norm": 0.9895520991699935,
+      "learning_rate": 7.304904779738169e-06,
+      "loss": 0.2435,
+      "step": 1203
+    },
+    {
+      "epoch": 0.36746528307645354,
+      "grad_norm": 1.229493544913369,
+      "learning_rate": 7.300516053883301e-06,
+      "loss": 0.3142,
+      "step": 1204
+    },
+    {
+      "epoch": 0.36777048679993896,
+      "grad_norm": 1.4306220663631386,
+      "learning_rate": 7.29612507850486e-06,
+      "loss": 0.3279,
+      "step": 1205
+    },
+    {
+      "epoch": 0.36807569052342437,
+      "grad_norm": 1.6062925874875447,
+      "learning_rate": 7.291731857896493e-06,
+      "loss": 0.4255,
+      "step": 1206
+    },
+    {
+      "epoch": 0.3683808942469098,
+      "grad_norm": 1.268726927448817,
+      "learning_rate": 7.287336396354039e-06,
+      "loss": 0.3437,
+      "step": 1207
+    },
+    {
+      "epoch": 0.36868609797039525,
+      "grad_norm": 1.6882520342147596,
+      "learning_rate": 7.282938698175535e-06,
+      "loss": 0.5873,
+      "step": 1208
+    },
+    {
+      "epoch": 0.36899130169388067,
+      "grad_norm": 1.0404068703903802,
+      "learning_rate": 7.278538767661198e-06,
+      "loss": 0.2389,
+      "step": 1209
+    },
+    {
+      "epoch": 0.3692965054173661,
+      "grad_norm": 1.1653062972796002,
+      "learning_rate": 7.274136609113433e-06,
+      "loss": 0.2512,
+      "step": 1210
+    },
+    {
+      "epoch": 0.3696017091408515,
+      "grad_norm": 1.3873057667640636,
+      "learning_rate": 7.269732226836821e-06,
+      "loss": 0.3267,
+      "step": 1211
+    },
+    {
+      "epoch": 0.36990691286433697,
+      "grad_norm": 1.5224889687825975,
+      "learning_rate": 7.265325625138119e-06,
+      "loss": 0.3719,
+      "step": 1212
+    },
+    {
+      "epoch": 0.3702121165878224,
+      "grad_norm": 1.2204715600675486,
+      "learning_rate": 7.260916808326252e-06,
+      "loss": 0.2686,
+      "step": 1213
+    },
+    {
+      "epoch": 0.3705173203113078,
+      "grad_norm": 1.3842524558223457,
+      "learning_rate": 7.256505780712314e-06,
+      "loss": 0.2467,
+      "step": 1214
+    },
+    {
+      "epoch": 0.3708225240347932,
+      "grad_norm": 1.7657099340484863,
+      "learning_rate": 7.252092546609558e-06,
+      "loss": 0.2798,
+      "step": 1215
+    },
+    {
+      "epoch": 0.3711277277582786,
+      "grad_norm": 1.3329438693740925,
+      "learning_rate": 7.247677110333397e-06,
+      "loss": 0.2828,
+      "step": 1216
+    },
+    {
+      "epoch": 0.3714329314817641,
+      "grad_norm": 1.104320779277888,
+      "learning_rate": 7.2432594762013945e-06,
+      "loss": 0.3003,
+      "step": 1217
+    },
+    {
+      "epoch": 0.3717381352052495,
+      "grad_norm": 1.2445465814923657,
+      "learning_rate": 7.238839648533266e-06,
+      "loss": 0.2837,
+      "step": 1218
+    },
+    {
+      "epoch": 0.3720433389287349,
+      "grad_norm": 1.540336275418335,
+      "learning_rate": 7.234417631650872e-06,
+      "loss": 0.5497,
+      "step": 1219
+    },
+    {
+      "epoch": 0.37234854265222034,
+      "grad_norm": 1.2167853145142045,
+      "learning_rate": 7.2299934298782105e-06,
+      "loss": 0.2327,
+      "step": 1220
+    },
+    {
+      "epoch": 0.3726537463757058,
+      "grad_norm": 1.6897903723185226,
+      "learning_rate": 7.225567047541418e-06,
+      "loss": 0.4309,
+      "step": 1221
+    },
+    {
+      "epoch": 0.3729589500991912,
+      "grad_norm": 1.3675420682807005,
+      "learning_rate": 7.221138488968763e-06,
+      "loss": 0.3553,
+      "step": 1222
+    },
+    {
+      "epoch": 0.37326415382267664,
+      "grad_norm": 1.1874100500472131,
+      "learning_rate": 7.2167077584906416e-06,
+      "loss": 0.2332,
+      "step": 1223
+    },
+    {
+      "epoch": 0.37356935754616205,
+      "grad_norm": 1.7310995153583675,
+      "learning_rate": 7.2122748604395765e-06,
+      "loss": 0.3075,
+      "step": 1224
+    },
+    {
+      "epoch": 0.37387456126964747,
+      "grad_norm": 1.3940442942063533,
+      "learning_rate": 7.207839799150206e-06,
+      "loss": 0.43,
+      "step": 1225
+    },
+    {
+      "epoch": 0.37417976499313294,
+      "grad_norm": 1.6425564976559994,
+      "learning_rate": 7.203402578959285e-06,
+      "loss": 0.5069,
+      "step": 1226
+    },
+    {
+      "epoch": 0.37448496871661835,
+      "grad_norm": 0.9576849748308918,
+      "learning_rate": 7.1989632042056816e-06,
+      "loss": 0.2351,
+      "step": 1227
+    },
+    {
+      "epoch": 0.37479017244010376,
+      "grad_norm": 1.3701544258348275,
+      "learning_rate": 7.1945216792303666e-06,
+      "loss": 0.3381,
+      "step": 1228
+    },
+    {
+      "epoch": 0.3750953761635892,
+      "grad_norm": 1.140854485823984,
+      "learning_rate": 7.190078008376415e-06,
+      "loss": 0.2751,
+      "step": 1229
+    },
+    {
+      "epoch": 0.37540057988707465,
+      "grad_norm": 1.3323249381262718,
+      "learning_rate": 7.185632195989005e-06,
+      "loss": 0.3349,
+      "step": 1230
+    },
+    {
+      "epoch": 0.37570578361056006,
+      "grad_norm": 1.0885163567382525,
+      "learning_rate": 7.1811842464154e-06,
+      "loss": 0.2448,
+      "step": 1231
+    },
+    {
+      "epoch": 0.3760109873340455,
+      "grad_norm": 1.4080823193078549,
+      "learning_rate": 7.1767341640049595e-06,
+      "loss": 0.3132,
+      "step": 1232
+    },
+    {
+      "epoch": 0.3763161910575309,
+      "grad_norm": 1.353912867702153,
+      "learning_rate": 7.172281953109128e-06,
+      "loss": 0.3432,
+      "step": 1233
+    },
+    {
+      "epoch": 0.3766213947810163,
+      "grad_norm": 1.3812789171739044,
+      "learning_rate": 7.167827618081426e-06,
+      "loss": 0.2137,
+      "step": 1234
+    },
+    {
+      "epoch": 0.3769265985045018,
+      "grad_norm": 1.2478616544202277,
+      "learning_rate": 7.1633711632774605e-06,
+      "loss": 0.2565,
+      "step": 1235
+    },
+    {
+      "epoch": 0.3772318022279872,
+      "grad_norm": 1.4006239352207315,
+      "learning_rate": 7.158912593054904e-06,
+      "loss": 0.4235,
+      "step": 1236
+    },
+    {
+      "epoch": 0.3775370059514726,
+      "grad_norm": 1.2649826963093111,
+      "learning_rate": 7.154451911773499e-06,
+      "loss": 0.3147,
+      "step": 1237
+    },
+    {
+      "epoch": 0.377842209674958,
+      "grad_norm": 1.9419141139672966,
+      "learning_rate": 7.149989123795054e-06,
+      "loss": 0.2825,
+      "step": 1238
+    },
+    {
+      "epoch": 0.3781474133984435,
+      "grad_norm": 1.4118815933485622,
+      "learning_rate": 7.145524233483434e-06,
+      "loss": 0.252,
+      "step": 1239
+    },
+    {
+      "epoch": 0.3784526171219289,
+      "grad_norm": 1.55813966795455,
+      "learning_rate": 7.1410572452045635e-06,
+      "loss": 0.3432,
+      "step": 1240
+    },
+    {
+      "epoch": 0.3787578208454143,
+      "grad_norm": 0.9637101653261684,
+      "learning_rate": 7.136588163326417e-06,
+      "loss": 0.2123,
+      "step": 1241
+    },
+    {
+      "epoch": 0.37906302456889973,
+      "grad_norm": 1.4156183777374252,
+      "learning_rate": 7.1321169922190144e-06,
+      "loss": 0.4448,
+      "step": 1242
+    },
+    {
+      "epoch": 0.37936822829238515,
+      "grad_norm": 1.2011574845111053,
+      "learning_rate": 7.127643736254424e-06,
+      "loss": 0.1914,
+      "step": 1243
+    },
+    {
+      "epoch": 0.3796734320158706,
+      "grad_norm": 1.2719810910228972,
+      "learning_rate": 7.123168399806747e-06,
+      "loss": 0.2365,
+      "step": 1244
+    },
+    {
+      "epoch": 0.37997863573935603,
+      "grad_norm": 1.1627952282839493,
+      "learning_rate": 7.118690987252121e-06,
+      "loss": 0.2535,
+      "step": 1245
+    },
+    {
+      "epoch": 0.38028383946284144,
+      "grad_norm": 1.526033273441638,
+      "learning_rate": 7.114211502968712e-06,
+      "loss": 0.563,
+      "step": 1246
+    },
+    {
+      "epoch": 0.38058904318632686,
+      "grad_norm": 1.6215492435897827,
+      "learning_rate": 7.1097299513367166e-06,
+      "loss": 0.238,
+      "step": 1247
+    },
+    {
+      "epoch": 0.3808942469098123,
+      "grad_norm": 1.689691321719832,
+      "learning_rate": 7.105246336738348e-06,
+      "loss": 0.1984,
+      "step": 1248
+    },
+    {
+      "epoch": 0.38119945063329774,
+      "grad_norm": 3.3706702256948455,
+      "learning_rate": 7.100760663557841e-06,
+      "loss": 0.2772,
+      "step": 1249
+    },
+    {
+      "epoch": 0.38150465435678316,
+      "grad_norm": 1.5153837503559604,
+      "learning_rate": 7.0962729361814355e-06,
+      "loss": 0.2683,
+      "step": 1250
+    },
+    {
+      "epoch": 0.38180985808026857,
+      "grad_norm": 1.269653690009095,
+      "learning_rate": 7.0917831589973895e-06,
+      "loss": 0.2611,
+      "step": 1251
+    },
+    {
+      "epoch": 0.382115061803754,
+      "grad_norm": 1.409968651632736,
+      "learning_rate": 7.0872913363959614e-06,
+      "loss": 0.3647,
+      "step": 1252
+    },
+    {
+      "epoch": 0.38242026552723946,
+      "grad_norm": 1.5218215588345414,
+      "learning_rate": 7.082797472769408e-06,
+      "loss": 0.466,
+      "step": 1253
+    },
+    {
+      "epoch": 0.38272546925072487,
+      "grad_norm": 2.3890410517745835,
+      "learning_rate": 7.078301572511984e-06,
+      "loss": 0.3646,
+      "step": 1254
+    },
+    {
+      "epoch": 0.3830306729742103,
+      "grad_norm": 1.641992489517636,
+      "learning_rate": 7.073803640019936e-06,
+      "loss": 0.3105,
+      "step": 1255
+    },
+    {
+      "epoch": 0.3833358766976957,
+      "grad_norm": 1.299025900561255,
+      "learning_rate": 7.0693036796914945e-06,
+      "loss": 0.3399,
+      "step": 1256
+    },
+    {
+      "epoch": 0.3836410804211811,
+      "grad_norm": 1.5380435417024074,
+      "learning_rate": 7.064801695926877e-06,
+      "loss": 0.2588,
+      "step": 1257
+    },
+    {
+      "epoch": 0.3839462841446666,
+      "grad_norm": 1.2372269820552038,
+      "learning_rate": 7.060297693128277e-06,
+      "loss": 0.3599,
+      "step": 1258
+    },
+    {
+      "epoch": 0.384251487868152,
+      "grad_norm": 1.1614198321930482,
+      "learning_rate": 7.055791675699863e-06,
+      "loss": 0.1642,
+      "step": 1259
+    },
+    {
+      "epoch": 0.3845566915916374,
+      "grad_norm": 1.4934775483085228,
+      "learning_rate": 7.051283648047775e-06,
+      "loss": 0.4615,
+      "step": 1260
+    },
+    {
+      "epoch": 0.3848618953151228,
+      "grad_norm": 1.4582552843552379,
+      "learning_rate": 7.046773614580116e-06,
+      "loss": 0.3002,
+      "step": 1261
+    },
+    {
+      "epoch": 0.3851670990386083,
+      "grad_norm": 4.333102342366705,
+      "learning_rate": 7.042261579706951e-06,
+      "loss": 0.3606,
+      "step": 1262
+    },
+    {
+      "epoch": 0.3854723027620937,
+      "grad_norm": 1.5981293143466662,
+      "learning_rate": 7.037747547840303e-06,
+      "loss": 0.3758,
+      "step": 1263
+    },
+    {
+      "epoch": 0.3857775064855791,
+      "grad_norm": 1.1060333427728648,
+      "learning_rate": 7.033231523394147e-06,
+      "loss": 0.1347,
+      "step": 1264
+    },
+    {
+      "epoch": 0.38608271020906454,
+      "grad_norm": 1.7104158346504792,
+      "learning_rate": 7.028713510784409e-06,
+      "loss": 0.6244,
+      "step": 1265
+    },
+    {
+      "epoch": 0.38638791393254995,
+      "grad_norm": 1.2041816530098084,
+      "learning_rate": 7.024193514428953e-06,
+      "loss": 0.2692,
+      "step": 1266
+    },
+    {
+      "epoch": 0.3866931176560354,
+      "grad_norm": 1.2271375340838422,
+      "learning_rate": 7.0196715387475885e-06,
+      "loss": 0.2536,
+      "step": 1267
+    },
+    {
+      "epoch": 0.38699832137952084,
+      "grad_norm": 1.3208524965326127,
+      "learning_rate": 7.015147588162061e-06,
+      "loss": 0.2425,
+      "step": 1268
+    },
+    {
+      "epoch": 0.38730352510300625,
+      "grad_norm": 1.6679415198075516,
+      "learning_rate": 7.010621667096041e-06,
+      "loss": 0.3849,
+      "step": 1269
+    },
+    {
+      "epoch": 0.38760872882649167,
+      "grad_norm": 1.336956505558018,
+      "learning_rate": 7.0060937799751316e-06,
+      "loss": 0.2577,
+      "step": 1270
+    },
+    {
+      "epoch": 0.38791393254997714,
+      "grad_norm": 1.1363423042029048,
+      "learning_rate": 7.0015639312268555e-06,
+      "loss": 0.2241,
+      "step": 1271
+    },
+    {
+      "epoch": 0.38821913627346255,
+      "grad_norm": 1.283661967883495,
+      "learning_rate": 6.997032125280655e-06,
+      "loss": 0.2903,
+      "step": 1272
+    },
+    {
+      "epoch": 0.38852433999694796,
+      "grad_norm": 1.6001822129445207,
+      "learning_rate": 6.992498366567884e-06,
+      "loss": 0.2261,
+      "step": 1273
+    },
+    {
+      "epoch": 0.3888295437204334,
+      "grad_norm": 1.3026652853973728,
+      "learning_rate": 6.987962659521808e-06,
+      "loss": 0.3057,
+      "step": 1274
+    },
+    {
+      "epoch": 0.3891347474439188,
+      "grad_norm": 1.4690995307739332,
+      "learning_rate": 6.983425008577598e-06,
+      "loss": 0.5416,
+      "step": 1275
+    },
+    {
+      "epoch": 0.38943995116740426,
+      "grad_norm": 1.8375650537094952,
+      "learning_rate": 6.978885418172325e-06,
+      "loss": 0.3456,
+      "step": 1276
+    },
+    {
+      "epoch": 0.3897451548908897,
+      "grad_norm": 2.194246189266676,
+      "learning_rate": 6.974343892744954e-06,
+      "loss": 0.2907,
+      "step": 1277
+    },
+    {
+      "epoch": 0.3900503586143751,
+      "grad_norm": 1.6774755773457783,
+      "learning_rate": 6.969800436736347e-06,
+      "loss": 0.3058,
+      "step": 1278
+    },
+    {
+      "epoch": 0.3903555623378605,
+      "grad_norm": 1.4187678667867623,
+      "learning_rate": 6.965255054589252e-06,
+      "loss": 0.4065,
+      "step": 1279
+    },
+    {
+      "epoch": 0.3906607660613459,
+      "grad_norm": 1.2370997427013408,
+      "learning_rate": 6.960707750748296e-06,
+      "loss": 0.276,
+      "step": 1280
+    },
+    {
+      "epoch": 0.3909659697848314,
+      "grad_norm": 1.2143352462379313,
+      "learning_rate": 6.956158529659991e-06,
+      "loss": 0.2169,
+      "step": 1281
+    },
+    {
+      "epoch": 0.3912711735083168,
+      "grad_norm": 1.3152454363292938,
+      "learning_rate": 6.951607395772721e-06,
+      "loss": 0.3729,
+      "step": 1282
+    },
+    {
+      "epoch": 0.3915763772318022,
+      "grad_norm": 1.7624342452965842,
+      "learning_rate": 6.947054353536742e-06,
+      "loss": 0.3491,
+      "step": 1283
+    },
+    {
+      "epoch": 0.39188158095528763,
+      "grad_norm": 1.0107668961163137,
+      "learning_rate": 6.942499407404175e-06,
+      "loss": 0.2528,
+      "step": 1284
+    },
+    {
+      "epoch": 0.3921867846787731,
+      "grad_norm": 1.3640381643739714,
+      "learning_rate": 6.937942561829001e-06,
+      "loss": 0.2922,
+      "step": 1285
+    },
+    {
+      "epoch": 0.3924919884022585,
+      "grad_norm": 1.3678373263857668,
+      "learning_rate": 6.933383821267062e-06,
+      "loss": 0.4763,
+      "step": 1286
+    },
+    {
+      "epoch": 0.39279719212574393,
+      "grad_norm": 1.4067856510214773,
+      "learning_rate": 6.928823190176051e-06,
+      "loss": 0.441,
+      "step": 1287
+    },
+    {
+      "epoch": 0.39310239584922935,
+      "grad_norm": 1.1456457967166047,
+      "learning_rate": 6.924260673015507e-06,
+      "loss": 0.2726,
+      "step": 1288
+    },
+    {
+      "epoch": 0.39340759957271476,
+      "grad_norm": 1.2577366776447958,
+      "learning_rate": 6.919696274246818e-06,
+      "loss": 0.2356,
+      "step": 1289
+    },
+    {
+      "epoch": 0.39371280329620023,
+      "grad_norm": 1.5018478773369082,
+      "learning_rate": 6.91512999833321e-06,
+      "loss": 0.3858,
+      "step": 1290
+    },
+    {
+      "epoch": 0.39401800701968565,
+      "grad_norm": 1.4382055458480731,
+      "learning_rate": 6.910561849739743e-06,
+      "loss": 0.4623,
+      "step": 1291
+    },
+    {
+      "epoch": 0.39432321074317106,
+      "grad_norm": 1.4150013204507765,
+      "learning_rate": 6.905991832933312e-06,
+      "loss": 0.2598,
+      "step": 1292
+    },
+    {
+      "epoch": 0.3946284144666565,
+      "grad_norm": 1.313619408763559,
+      "learning_rate": 6.901419952382633e-06,
+      "loss": 0.1856,
+      "step": 1293
+    },
+    {
+      "epoch": 0.39493361819014194,
+      "grad_norm": 1.1626329284469816,
+      "learning_rate": 6.89684621255825e-06,
+      "loss": 0.2655,
+      "step": 1294
+    },
+    {
+      "epoch": 0.39523882191362736,
+      "grad_norm": 1.4286438514880226,
+      "learning_rate": 6.892270617932523e-06,
+      "loss": 0.3111,
+      "step": 1295
+    },
+    {
+      "epoch": 0.3955440256371128,
+      "grad_norm": 1.467025068592859,
+      "learning_rate": 6.887693172979624e-06,
+      "loss": 0.1943,
+      "step": 1296
+    },
+    {
+      "epoch": 0.3958492293605982,
+      "grad_norm": 1.5302799032399699,
+      "learning_rate": 6.883113882175536e-06,
+      "loss": 0.2896,
+      "step": 1297
+    },
+    {
+      "epoch": 0.3961544330840836,
+      "grad_norm": 1.4221040310691428,
+      "learning_rate": 6.878532749998049e-06,
+      "loss": 0.2751,
+      "step": 1298
+    },
+    {
+      "epoch": 0.39645963680756907,
+      "grad_norm": 1.377710099164584,
+      "learning_rate": 6.873949780926747e-06,
+      "loss": 0.2793,
+      "step": 1299
+    },
+    {
+      "epoch": 0.3967648405310545,
+      "grad_norm": 1.3417904782871544,
+      "learning_rate": 6.869364979443019e-06,
+      "loss": 0.2602,
+      "step": 1300
+    },
+    {
+      "epoch": 0.3970700442545399,
+      "grad_norm": 1.3467916175186017,
+      "learning_rate": 6.86477835003004e-06,
+      "loss": 0.2857,
+      "step": 1301
+    },
+    {
+      "epoch": 0.3973752479780253,
+      "grad_norm": 1.4529951304165758,
+      "learning_rate": 6.860189897172772e-06,
+      "loss": 0.5097,
+      "step": 1302
+    },
+    {
+      "epoch": 0.3976804517015108,
+      "grad_norm": 1.3839926990470812,
+      "learning_rate": 6.8555996253579645e-06,
+      "loss": 0.4489,
+      "step": 1303
+    },
+    {
+      "epoch": 0.3979856554249962,
+      "grad_norm": 1.4801954354304339,
+      "learning_rate": 6.85100753907414e-06,
+      "loss": 0.5139,
+      "step": 1304
+    },
+    {
+      "epoch": 0.3982908591484816,
+      "grad_norm": 1.0192216424990155,
+      "learning_rate": 6.846413642811598e-06,
+      "loss": 0.231,
+      "step": 1305
+    },
+    {
+      "epoch": 0.398596062871967,
+      "grad_norm": 1.4352380952459303,
+      "learning_rate": 6.841817941062412e-06,
+      "loss": 0.3393,
+      "step": 1306
+    },
+    {
+      "epoch": 0.39890126659545244,
+      "grad_norm": 1.5851110550635767,
+      "learning_rate": 6.837220438320411e-06,
+      "loss": 0.41,
+      "step": 1307
+    },
+    {
+      "epoch": 0.3992064703189379,
+      "grad_norm": 1.569272508271095,
+      "learning_rate": 6.832621139081196e-06,
+      "loss": 0.3163,
+      "step": 1308
+    },
+    {
+      "epoch": 0.3995116740424233,
+      "grad_norm": 1.3018066671809359,
+      "learning_rate": 6.8280200478421185e-06,
+      "loss": 0.2329,
+      "step": 1309
+    },
+    {
+      "epoch": 0.39981687776590874,
+      "grad_norm": 1.660594639958246,
+      "learning_rate": 6.823417169102282e-06,
+      "loss": 0.3668,
+      "step": 1310
+    },
+    {
+      "epoch": 0.40012208148939415,
+      "grad_norm": 1.522549727556404,
+      "learning_rate": 6.81881250736254e-06,
+      "loss": 0.3363,
+      "step": 1311
+    },
+    {
+      "epoch": 0.4004272852128796,
+      "grad_norm": 1.497940397892529,
+      "learning_rate": 6.8142060671254905e-06,
+      "loss": 0.393,
+      "step": 1312
+    },
+    {
+      "epoch": 0.40073248893636504,
+      "grad_norm": 1.4706728908070485,
+      "learning_rate": 6.8095978528954665e-06,
+      "loss": 0.4207,
+      "step": 1313
+    },
+    {
+      "epoch": 0.40103769265985045,
+      "grad_norm": 1.2561900413969935,
+      "learning_rate": 6.80498786917854e-06,
+      "loss": 0.2689,
+      "step": 1314
+    },
+    {
+      "epoch": 0.40134289638333587,
+      "grad_norm": 2.2206434310633494,
+      "learning_rate": 6.800376120482508e-06,
+      "loss": 0.3934,
+      "step": 1315
+    },
+    {
+      "epoch": 0.4016481001068213,
+      "grad_norm": 1.180675774729955,
+      "learning_rate": 6.795762611316901e-06,
+      "loss": 0.1996,
+      "step": 1316
+    },
+    {
+      "epoch": 0.40195330383030675,
+      "grad_norm": 1.2644690898256274,
+      "learning_rate": 6.791147346192965e-06,
+      "loss": 0.3932,
+      "step": 1317
+    },
+    {
+      "epoch": 0.40225850755379217,
+      "grad_norm": 1.9397161791331279,
+      "learning_rate": 6.786530329623663e-06,
+      "loss": 0.2068,
+      "step": 1318
+    },
+    {
+      "epoch": 0.4025637112772776,
+      "grad_norm": 1.2683770669890404,
+      "learning_rate": 6.7819115661236745e-06,
+      "loss": 0.3456,
+      "step": 1319
+    },
+    {
+      "epoch": 0.402868915000763,
+      "grad_norm": 1.2741738279318962,
+      "learning_rate": 6.7772910602093845e-06,
+      "loss": 0.226,
+      "step": 1320
+    },
+    {
+      "epoch": 0.4031741187242484,
+      "grad_norm": 1.5848234426074979,
+      "learning_rate": 6.772668816398882e-06,
+      "loss": 0.4086,
+      "step": 1321
+    },
+    {
+      "epoch": 0.4034793224477339,
+      "grad_norm": 1.1692837868067145,
+      "learning_rate": 6.768044839211955e-06,
+      "loss": 0.2752,
+      "step": 1322
+    },
+    {
+      "epoch": 0.4037845261712193,
+      "grad_norm": 1.3967415526137874,
+      "learning_rate": 6.763419133170089e-06,
+      "loss": 0.4226,
+      "step": 1323
+    },
+    {
+      "epoch": 0.4040897298947047,
+      "grad_norm": 1.4254602358278508,
+      "learning_rate": 6.758791702796454e-06,
+      "loss": 0.3899,
+      "step": 1324
+    },
+    {
+      "epoch": 0.4043949336181901,
+      "grad_norm": 1.5572617271717888,
+      "learning_rate": 6.754162552615916e-06,
+      "loss": 0.4574,
+      "step": 1325
+    },
+    {
+      "epoch": 0.4047001373416756,
+      "grad_norm": 1.7167631282679396,
+      "learning_rate": 6.749531687155014e-06,
+      "loss": 0.298,
+      "step": 1326
+    },
+    {
+      "epoch": 0.405005341065161,
+      "grad_norm": 1.2196148469218278,
+      "learning_rate": 6.744899110941967e-06,
+      "loss": 0.3353,
+      "step": 1327
+    },
+    {
+      "epoch": 0.4053105447886464,
+      "grad_norm": 1.425829742861132,
+      "learning_rate": 6.740264828506668e-06,
+      "loss": 0.3501,
+      "step": 1328
+    },
+    {
+      "epoch": 0.40561574851213184,
+      "grad_norm": 1.3938860001518678,
+      "learning_rate": 6.7356288443806765e-06,
+      "loss": 0.2435,
+      "step": 1329
+    },
+    {
+      "epoch": 0.40592095223561725,
+      "grad_norm": 1.2919006084530176,
+      "learning_rate": 6.730991163097217e-06,
+      "loss": 0.419,
+      "step": 1330
+    },
+    {
+      "epoch": 0.4062261559591027,
+      "grad_norm": 1.1955154346409425,
+      "learning_rate": 6.7263517891911745e-06,
+      "loss": 0.2766,
+      "step": 1331
+    },
+    {
+      "epoch": 0.40653135968258813,
+      "grad_norm": 1.272620752447219,
+      "learning_rate": 6.721710727199087e-06,
+      "loss": 0.3691,
+      "step": 1332
+    },
+    {
+      "epoch": 0.40683656340607355,
+      "grad_norm": 1.2911916084787372,
+      "learning_rate": 6.717067981659145e-06,
+      "loss": 0.2437,
+      "step": 1333
+    },
+    {
+      "epoch": 0.40714176712955896,
+      "grad_norm": 1.3485651903926066,
+      "learning_rate": 6.712423557111184e-06,
+      "loss": 0.3076,
+      "step": 1334
+    },
+    {
+      "epoch": 0.40744697085304443,
+      "grad_norm": 1.1329951698085787,
+      "learning_rate": 6.707777458096682e-06,
+      "loss": 0.2218,
+      "step": 1335
+    },
+    {
+      "epoch": 0.40775217457652985,
+      "grad_norm": 1.667913860338061,
+      "learning_rate": 6.703129689158755e-06,
+      "loss": 0.3711,
+      "step": 1336
+    },
+    {
+      "epoch": 0.40805737830001526,
+      "grad_norm": 1.145899868969619,
+      "learning_rate": 6.698480254842148e-06,
+      "loss": 0.2636,
+      "step": 1337
+    },
+    {
+      "epoch": 0.4083625820235007,
+      "grad_norm": 1.1099719738214417,
+      "learning_rate": 6.693829159693241e-06,
+      "loss": 0.2096,
+      "step": 1338
+    },
+    {
+      "epoch": 0.4086677857469861,
+      "grad_norm": 2.3227758208085665,
+      "learning_rate": 6.689176408260033e-06,
+      "loss": 0.4818,
+      "step": 1339
+    },
+    {
+      "epoch": 0.40897298947047156,
+      "grad_norm": 1.1176996724319626,
+      "learning_rate": 6.684522005092142e-06,
+      "loss": 0.3045,
+      "step": 1340
+    },
+    {
+      "epoch": 0.409278193193957,
+      "grad_norm": 1.2902343971744434,
+      "learning_rate": 6.679865954740808e-06,
+      "loss": 0.2739,
+      "step": 1341
+    },
+    {
+      "epoch": 0.4095833969174424,
+      "grad_norm": 1.24221657772939,
+      "learning_rate": 6.675208261758874e-06,
+      "loss": 0.248,
+      "step": 1342
+    },
+    {
+      "epoch": 0.4098886006409278,
+      "grad_norm": 1.466634078458504,
+      "learning_rate": 6.670548930700791e-06,
+      "loss": 0.2954,
+      "step": 1343
+    },
+    {
+      "epoch": 0.4101938043644133,
+      "grad_norm": 1.4954932379638273,
+      "learning_rate": 6.665887966122616e-06,
+      "loss": 0.4283,
+      "step": 1344
+    },
+    {
+      "epoch": 0.4104990080878987,
+      "grad_norm": 1.3273678584881572,
+      "learning_rate": 6.661225372581996e-06,
+      "loss": 0.2329,
+      "step": 1345
+    },
+    {
+      "epoch": 0.4108042118113841,
+      "grad_norm": 1.2699849887771746,
+      "learning_rate": 6.6565611546381794e-06,
+      "loss": 0.1797,
+      "step": 1346
+    },
+    {
+      "epoch": 0.4111094155348695,
+      "grad_norm": 1.2890458539408176,
+      "learning_rate": 6.651895316851996e-06,
+      "loss": 0.2587,
+      "step": 1347
+    },
+    {
+      "epoch": 0.41141461925835493,
+      "grad_norm": 1.216750093543553,
+      "learning_rate": 6.6472278637858615e-06,
+      "loss": 0.3274,
+      "step": 1348
+    },
+    {
+      "epoch": 0.4117198229818404,
+      "grad_norm": 1.3058663712935152,
+      "learning_rate": 6.642558800003775e-06,
+      "loss": 0.2467,
+      "step": 1349
+    },
+    {
+      "epoch": 0.4120250267053258,
+      "grad_norm": 1.4706561117776265,
+      "learning_rate": 6.637888130071304e-06,
+      "loss": 0.4008,
+      "step": 1350
+    },
+    {
+      "epoch": 0.41233023042881123,
+      "grad_norm": 1.1481689346703796,
+      "learning_rate": 6.633215858555592e-06,
+      "loss": 0.2391,
+      "step": 1351
+    },
+    {
+      "epoch": 0.41263543415229664,
+      "grad_norm": 1.2458895811994175,
+      "learning_rate": 6.628541990025346e-06,
+      "loss": 0.2592,
+      "step": 1352
+    },
+    {
+      "epoch": 0.41294063787578206,
+      "grad_norm": 2.065000463305553,
+      "learning_rate": 6.623866529050834e-06,
+      "loss": 0.5728,
+      "step": 1353
+    },
+    {
+      "epoch": 0.4132458415992675,
+      "grad_norm": 1.626428050000369,
+      "learning_rate": 6.6191894802038826e-06,
+      "loss": 0.495,
+      "step": 1354
+    },
+    {
+      "epoch": 0.41355104532275294,
+      "grad_norm": 1.2186318364692617,
+      "learning_rate": 6.614510848057874e-06,
+      "loss": 0.3359,
+      "step": 1355
+    },
+    {
+      "epoch": 0.41385624904623836,
+      "grad_norm": 1.3483193797759696,
+      "learning_rate": 6.609830637187729e-06,
+      "loss": 0.3744,
+      "step": 1356
+    },
+    {
+      "epoch": 0.41416145276972377,
+      "grad_norm": 1.2559337920214608,
+      "learning_rate": 6.6051488521699235e-06,
+      "loss": 0.3737,
+      "step": 1357
+    },
+    {
+      "epoch": 0.41446665649320924,
+      "grad_norm": 1.3874775704297853,
+      "learning_rate": 6.600465497582466e-06,
+      "loss": 0.1995,
+      "step": 1358
+    },
+    {
+      "epoch": 0.41477186021669465,
+      "grad_norm": 1.6050931393702568,
+      "learning_rate": 6.595780578004901e-06,
+      "loss": 0.5452,
+      "step": 1359
+    },
+    {
+      "epoch": 0.41507706394018007,
+      "grad_norm": 1.4120118673382418,
+      "learning_rate": 6.591094098018306e-06,
+      "loss": 0.3246,
+      "step": 1360
+    },
+    {
+      "epoch": 0.4153822676636655,
+      "grad_norm": 1.1696817855430393,
+      "learning_rate": 6.586406062205278e-06,
+      "loss": 0.1567,
+      "step": 1361
+    },
+    {
+      "epoch": 0.4156874713871509,
+      "grad_norm": 1.4850231161148664,
+      "learning_rate": 6.581716475149943e-06,
+      "loss": 0.3303,
+      "step": 1362
+    },
+    {
+      "epoch": 0.41599267511063637,
+      "grad_norm": 1.4803085844420174,
+      "learning_rate": 6.577025341437939e-06,
+      "loss": 0.3144,
+      "step": 1363
+    },
+    {
+      "epoch": 0.4162978788341218,
+      "grad_norm": 1.314596318555117,
+      "learning_rate": 6.572332665656417e-06,
+      "loss": 0.3133,
+      "step": 1364
+    },
+    {
+      "epoch": 0.4166030825576072,
+      "grad_norm": 1.2386593406906814,
+      "learning_rate": 6.567638452394036e-06,
+      "loss": 0.2295,
+      "step": 1365
+    },
+    {
+      "epoch": 0.4169082862810926,
+      "grad_norm": 1.5178670129535232,
+      "learning_rate": 6.56294270624096e-06,
+      "loss": 0.3895,
+      "step": 1366
+    },
+    {
+      "epoch": 0.4172134900045781,
+      "grad_norm": 1.1161753197408095,
+      "learning_rate": 6.558245431788851e-06,
+      "loss": 0.2753,
+      "step": 1367
+    },
+    {
+      "epoch": 0.4175186937280635,
+      "grad_norm": 1.6916589702312108,
+      "learning_rate": 6.553546633630865e-06,
+      "loss": 0.5491,
+      "step": 1368
+    },
+    {
+      "epoch": 0.4178238974515489,
+      "grad_norm": 1.2497529739433018,
+      "learning_rate": 6.548846316361648e-06,
+      "loss": 0.275,
+      "step": 1369
+    },
+    {
+      "epoch": 0.4181291011750343,
+      "grad_norm": 1.3872597941831253,
+      "learning_rate": 6.544144484577331e-06,
+      "loss": 0.4461,
+      "step": 1370
+    },
+    {
+      "epoch": 0.41843430489851974,
+      "grad_norm": 1.290246777845889,
+      "learning_rate": 6.539441142875528e-06,
+      "loss": 0.3839,
+      "step": 1371
+    },
+    {
+      "epoch": 0.4187395086220052,
+      "grad_norm": 1.9737467857948305,
+      "learning_rate": 6.534736295855325e-06,
+      "loss": 0.2692,
+      "step": 1372
+    },
+    {
+      "epoch": 0.4190447123454906,
+      "grad_norm": 1.4088078442454266,
+      "learning_rate": 6.530029948117284e-06,
+      "loss": 0.4093,
+      "step": 1373
+    },
+    {
+      "epoch": 0.41934991606897604,
+      "grad_norm": 1.2700736897033225,
+      "learning_rate": 6.525322104263439e-06,
+      "loss": 0.4136,
+      "step": 1374
+    },
+    {
+      "epoch": 0.41965511979246145,
+      "grad_norm": 1.7721019538973242,
+      "learning_rate": 6.520612768897272e-06,
+      "loss": 0.3371,
+      "step": 1375
+    },
+    {
+      "epoch": 0.4199603235159469,
+      "grad_norm": 1.2889345047723617,
+      "learning_rate": 6.51590194662374e-06,
+      "loss": 0.303,
+      "step": 1376
+    },
+    {
+      "epoch": 0.42026552723943233,
+      "grad_norm": 1.1153113343630081,
+      "learning_rate": 6.511189642049244e-06,
+      "loss": 0.2793,
+      "step": 1377
+    },
+    {
+      "epoch": 0.42057073096291775,
+      "grad_norm": 1.3638733722885776,
+      "learning_rate": 6.506475859781637e-06,
+      "loss": 0.374,
+      "step": 1378
+    },
+    {
+      "epoch": 0.42087593468640316,
+      "grad_norm": 1.4684349696171455,
+      "learning_rate": 6.501760604430218e-06,
+      "loss": 0.3168,
+      "step": 1379
+    },
+    {
+      "epoch": 0.4211811384098886,
+      "grad_norm": 1.1514319999384148,
+      "learning_rate": 6.497043880605725e-06,
+      "loss": 0.2929,
+      "step": 1380
+    },
+    {
+      "epoch": 0.42148634213337405,
+      "grad_norm": 1.169458160460905,
+      "learning_rate": 6.492325692920329e-06,
+      "loss": 0.2506,
+      "step": 1381
+    },
+    {
+      "epoch": 0.42179154585685946,
+      "grad_norm": 1.7740829570344059,
+      "learning_rate": 6.487606045987645e-06,
+      "loss": 0.647,
+      "step": 1382
+    },
+    {
+      "epoch": 0.4220967495803449,
+      "grad_norm": 1.5764040486432587,
+      "learning_rate": 6.482884944422696e-06,
+      "loss": 0.403,
+      "step": 1383
+    },
+    {
+      "epoch": 0.4224019533038303,
+      "grad_norm": 1.4963369146995207,
+      "learning_rate": 6.478162392841942e-06,
+      "loss": 0.4148,
+      "step": 1384
+    },
+    {
+      "epoch": 0.42270715702731576,
+      "grad_norm": 1.4435630313971244,
+      "learning_rate": 6.473438395863259e-06,
+      "loss": 0.4409,
+      "step": 1385
+    },
+    {
+      "epoch": 0.4230123607508012,
+      "grad_norm": 1.3799044100508142,
+      "learning_rate": 6.468712958105926e-06,
+      "loss": 0.2855,
+      "step": 1386
+    },
+    {
+      "epoch": 0.4233175644742866,
+      "grad_norm": 1.160815119293516,
+      "learning_rate": 6.463986084190644e-06,
+      "loss": 0.2924,
+      "step": 1387
+    },
+    {
+      "epoch": 0.423622768197772,
+      "grad_norm": 1.4073095780293283,
+      "learning_rate": 6.459257778739511e-06,
+      "loss": 0.4324,
+      "step": 1388
+    },
+    {
+      "epoch": 0.4239279719212574,
+      "grad_norm": 1.3577032223044758,
+      "learning_rate": 6.4545280463760226e-06,
+      "loss": 0.2674,
+      "step": 1389
+    },
+    {
+      "epoch": 0.4242331756447429,
+      "grad_norm": 1.2994968614411917,
+      "learning_rate": 6.449796891725081e-06,
+      "loss": 0.2802,
+      "step": 1390
+    },
+    {
+      "epoch": 0.4245383793682283,
+      "grad_norm": 1.343062113855798,
+      "learning_rate": 6.445064319412966e-06,
+      "loss": 0.218,
+      "step": 1391
+    },
+    {
+      "epoch": 0.4248435830917137,
+      "grad_norm": 1.907900174162606,
+      "learning_rate": 6.44033033406735e-06,
+      "loss": 0.3629,
+      "step": 1392
+    },
+    {
+      "epoch": 0.42514878681519913,
+      "grad_norm": 1.3076657321115293,
+      "learning_rate": 6.43559494031729e-06,
+      "loss": 0.4244,
+      "step": 1393
+    },
+    {
+      "epoch": 0.42545399053868455,
+      "grad_norm": 1.393965833570882,
+      "learning_rate": 6.430858142793213e-06,
+      "loss": 0.2388,
+      "step": 1394
+    },
+    {
+      "epoch": 0.42575919426217,
+      "grad_norm": 1.4458587439041957,
+      "learning_rate": 6.426119946126922e-06,
+      "loss": 0.5452,
+      "step": 1395
+    },
+    {
+      "epoch": 0.42606439798565543,
+      "grad_norm": 1.2918032092044962,
+      "learning_rate": 6.4213803549515896e-06,
+      "loss": 0.3516,
+      "step": 1396
+    },
+    {
+      "epoch": 0.42636960170914084,
+      "grad_norm": 1.4058848012801475,
+      "learning_rate": 6.41663937390175e-06,
+      "loss": 0.2097,
+      "step": 1397
+    },
+    {
+      "epoch": 0.42667480543262626,
+      "grad_norm": 1.6133906071062372,
+      "learning_rate": 6.4118970076132995e-06,
+      "loss": 0.5124,
+      "step": 1398
+    },
+    {
+      "epoch": 0.42698000915611173,
+      "grad_norm": 1.3377311221640538,
+      "learning_rate": 6.407153260723484e-06,
+      "loss": 0.315,
+      "step": 1399
+    },
+    {
+      "epoch": 0.42728521287959714,
+      "grad_norm": 1.5917236311274028,
+      "learning_rate": 6.402408137870903e-06,
+      "loss": 0.4754,
+      "step": 1400
+    },
+    {
+      "epoch": 0.42759041660308256,
+      "grad_norm": 1.7638143522011498,
+      "learning_rate": 6.397661643695502e-06,
+      "loss": 0.3344,
+      "step": 1401
+    },
+    {
+      "epoch": 0.42789562032656797,
+      "grad_norm": 1.3144896730327111,
+      "learning_rate": 6.392913782838563e-06,
+      "loss": 0.2085,
+      "step": 1402
+    },
+    {
+      "epoch": 0.4282008240500534,
+      "grad_norm": 1.1905378835451088,
+      "learning_rate": 6.388164559942709e-06,
+      "loss": 0.2553,
+      "step": 1403
+    },
+    {
+      "epoch": 0.42850602777353886,
+      "grad_norm": 2.0174215427728277,
+      "learning_rate": 6.383413979651893e-06,
+      "loss": 0.396,
+      "step": 1404
+    },
+    {
+      "epoch": 0.42881123149702427,
+      "grad_norm": 1.1441372658734617,
+      "learning_rate": 6.378662046611397e-06,
+      "loss": 0.2705,
+      "step": 1405
+    },
+    {
+      "epoch": 0.4291164352205097,
+      "grad_norm": 1.5452084281598708,
+      "learning_rate": 6.373908765467821e-06,
+      "loss": 0.3395,
+      "step": 1406
+    },
+    {
+      "epoch": 0.4294216389439951,
+      "grad_norm": 1.0332304693111856,
+      "learning_rate": 6.369154140869088e-06,
+      "loss": 0.2871,
+      "step": 1407
+    },
+    {
+      "epoch": 0.42972684266748057,
+      "grad_norm": 1.6720861141214347,
+      "learning_rate": 6.364398177464435e-06,
+      "loss": 0.3322,
+      "step": 1408
+    },
+    {
+      "epoch": 0.430032046390966,
+      "grad_norm": 1.3643272344010366,
+      "learning_rate": 6.359640879904405e-06,
+      "loss": 0.3161,
+      "step": 1409
+    },
+    {
+      "epoch": 0.4303372501144514,
+      "grad_norm": 1.1447056451932383,
+      "learning_rate": 6.354882252840845e-06,
+      "loss": 0.2654,
+      "step": 1410
+    },
+    {
+      "epoch": 0.4306424538379368,
+      "grad_norm": 1.1667911883159097,
+      "learning_rate": 6.350122300926906e-06,
+      "loss": 0.2105,
+      "step": 1411
+    },
+    {
+      "epoch": 0.4309476575614222,
+      "grad_norm": 1.3662993243668309,
+      "learning_rate": 6.3453610288170335e-06,
+      "loss": 0.3964,
+      "step": 1412
+    },
+    {
+      "epoch": 0.4312528612849077,
+      "grad_norm": 1.5696957264190432,
+      "learning_rate": 6.340598441166958e-06,
+      "loss": 0.3675,
+      "step": 1413
+    },
+    {
+      "epoch": 0.4315580650083931,
+      "grad_norm": 1.2281182014723664,
+      "learning_rate": 6.335834542633707e-06,
+      "loss": 0.2694,
+      "step": 1414
+    },
+    {
+      "epoch": 0.4318632687318785,
+      "grad_norm": 1.4302582982340493,
+      "learning_rate": 6.331069337875581e-06,
+      "loss": 0.2888,
+      "step": 1415
+    },
+    {
+      "epoch": 0.43216847245536394,
+      "grad_norm": 0.9906163729704169,
+      "learning_rate": 6.32630283155216e-06,
+      "loss": 0.22,
+      "step": 1416
+    },
+    {
+      "epoch": 0.4324736761788494,
+      "grad_norm": 1.4635044332996883,
+      "learning_rate": 6.321535028324302e-06,
+      "loss": 0.4671,
+      "step": 1417
+    },
+    {
+      "epoch": 0.4327788799023348,
+      "grad_norm": 1.4378547230792813,
+      "learning_rate": 6.316765932854122e-06,
+      "loss": 0.3855,
+      "step": 1418
+    },
+    {
+      "epoch": 0.43308408362582024,
+      "grad_norm": 1.5956913213398347,
+      "learning_rate": 6.31199554980501e-06,
+      "loss": 0.3164,
+      "step": 1419
+    },
+    {
+      "epoch": 0.43338928734930565,
+      "grad_norm": 1.3379057277745874,
+      "learning_rate": 6.307223883841609e-06,
+      "loss": 0.3991,
+      "step": 1420
+    },
+    {
+      "epoch": 0.43369449107279107,
+      "grad_norm": 1.0448334558458165,
+      "learning_rate": 6.302450939629818e-06,
+      "loss": 0.2132,
+      "step": 1421
+    },
+    {
+      "epoch": 0.43399969479627654,
+      "grad_norm": 1.6769341191161795,
+      "learning_rate": 6.297676721836785e-06,
+      "loss": 0.4639,
+      "step": 1422
+    },
+    {
+      "epoch": 0.43430489851976195,
+      "grad_norm": 1.508740075559931,
+      "learning_rate": 6.292901235130904e-06,
+      "loss": 0.435,
+      "step": 1423
+    },
+    {
+      "epoch": 0.43461010224324736,
+      "grad_norm": 1.6165557954399556,
+      "learning_rate": 6.288124484181811e-06,
+      "loss": 0.3716,
+      "step": 1424
+    },
+    {
+      "epoch": 0.4349153059667328,
+      "grad_norm": 1.4933007366732682,
+      "learning_rate": 6.283346473660377e-06,
+      "loss": 0.4222,
+      "step": 1425
+    },
+    {
+      "epoch": 0.4352205096902182,
+      "grad_norm": 1.31895828771684,
+      "learning_rate": 6.278567208238705e-06,
+      "loss": 0.3421,
+      "step": 1426
+    },
+    {
+      "epoch": 0.43552571341370366,
+      "grad_norm": 1.6584626299167318,
+      "learning_rate": 6.273786692590122e-06,
+      "loss": 0.5482,
+      "step": 1427
+    },
+    {
+      "epoch": 0.4358309171371891,
+      "grad_norm": 1.4286268291630235,
+      "learning_rate": 6.269004931389184e-06,
+      "loss": 0.2923,
+      "step": 1428
+    },
+    {
+      "epoch": 0.4361361208606745,
+      "grad_norm": 1.330259694437434,
+      "learning_rate": 6.264221929311659e-06,
+      "loss": 0.228,
+      "step": 1429
+    },
+    {
+      "epoch": 0.4364413245841599,
+      "grad_norm": 0.9297451354040828,
+      "learning_rate": 6.259437691034531e-06,
+      "loss": 0.1884,
+      "step": 1430
+    },
+    {
+      "epoch": 0.4367465283076454,
+      "grad_norm": 1.2652771083096295,
+      "learning_rate": 6.2546522212359925e-06,
+      "loss": 0.3444,
+      "step": 1431
+    },
+    {
+      "epoch": 0.4370517320311308,
+      "grad_norm": 1.6100392868745599,
+      "learning_rate": 6.2498655245954396e-06,
+      "loss": 0.5201,
+      "step": 1432
+    },
+    {
+      "epoch": 0.4373569357546162,
+      "grad_norm": 1.2278762390726963,
+      "learning_rate": 6.245077605793471e-06,
+      "loss": 0.2888,
+      "step": 1433
+    },
+    {
+      "epoch": 0.4376621394781016,
+      "grad_norm": 1.386480821426643,
+      "learning_rate": 6.2402884695118745e-06,
+      "loss": 0.3163,
+      "step": 1434
+    },
+    {
+      "epoch": 0.43796734320158703,
+      "grad_norm": 1.0406764982843713,
+      "learning_rate": 6.235498120433635e-06,
+      "loss": 0.2533,
+      "step": 1435
+    },
+    {
+      "epoch": 0.4382725469250725,
+      "grad_norm": 1.5763703213054279,
+      "learning_rate": 6.230706563242919e-06,
+      "loss": 0.4144,
+      "step": 1436
+    },
+    {
+      "epoch": 0.4385777506485579,
+      "grad_norm": 1.639255933375323,
+      "learning_rate": 6.225913802625076e-06,
+      "loss": 0.3643,
+      "step": 1437
+    },
+    {
+      "epoch": 0.43888295437204333,
+      "grad_norm": 1.7824581131688386,
+      "learning_rate": 6.22111984326663e-06,
+      "loss": 0.2619,
+      "step": 1438
+    },
+    {
+      "epoch": 0.43918815809552875,
+      "grad_norm": 1.6342850757025045,
+      "learning_rate": 6.216324689855283e-06,
+      "loss": 0.4316,
+      "step": 1439
+    },
+    {
+      "epoch": 0.4394933618190142,
+      "grad_norm": 1.3924162825479758,
+      "learning_rate": 6.211528347079896e-06,
+      "loss": 0.5224,
+      "step": 1440
+    },
+    {
+      "epoch": 0.43979856554249963,
+      "grad_norm": 1.4266758450691195,
+      "learning_rate": 6.206730819630502e-06,
+      "loss": 0.5211,
+      "step": 1441
+    },
+    {
+      "epoch": 0.44010376926598505,
+      "grad_norm": 1.3623827402595172,
+      "learning_rate": 6.201932112198285e-06,
+      "loss": 0.3893,
+      "step": 1442
+    },
+    {
+      "epoch": 0.44040897298947046,
+      "grad_norm": 1.026045410355969,
+      "learning_rate": 6.197132229475586e-06,
+      "loss": 0.2376,
+      "step": 1443
+    },
+    {
+      "epoch": 0.4407141767129559,
+      "grad_norm": 1.1848190562207745,
+      "learning_rate": 6.192331176155897e-06,
+      "loss": 0.3109,
+      "step": 1444
+    },
+    {
+      "epoch": 0.44101938043644134,
+      "grad_norm": 1.5310615112137054,
+      "learning_rate": 6.1875289569338505e-06,
+      "loss": 0.5635,
+      "step": 1445
+    },
+    {
+      "epoch": 0.44132458415992676,
+      "grad_norm": 1.414171583488937,
+      "learning_rate": 6.182725576505222e-06,
+      "loss": 0.5202,
+      "step": 1446
+    },
+    {
+      "epoch": 0.4416297878834122,
+      "grad_norm": 1.881243926890763,
+      "learning_rate": 6.177921039566925e-06,
+      "loss": 0.4076,
+      "step": 1447
+    },
+    {
+      "epoch": 0.4419349916068976,
+      "grad_norm": 1.3255828762960546,
+      "learning_rate": 6.173115350816995e-06,
+      "loss": 0.3769,
+      "step": 1448
+    },
+    {
+      "epoch": 0.44224019533038306,
+      "grad_norm": 1.5119486789167886,
+      "learning_rate": 6.168308514954603e-06,
+      "loss": 0.4031,
+      "step": 1449
+    },
+    {
+      "epoch": 0.44254539905386847,
+      "grad_norm": 2.0489266865517695,
+      "learning_rate": 6.1635005366800385e-06,
+      "loss": 0.3346,
+      "step": 1450
+    },
+    {
+      "epoch": 0.4428506027773539,
+      "grad_norm": 1.4973886329367188,
+      "learning_rate": 6.158691420694706e-06,
+      "loss": 0.4739,
+      "step": 1451
+    },
+    {
+      "epoch": 0.4431558065008393,
+      "grad_norm": 1.0605172000337302,
+      "learning_rate": 6.153881171701124e-06,
+      "loss": 0.2526,
+      "step": 1452
+    },
+    {
+      "epoch": 0.4434610102243247,
+      "grad_norm": 1.118949194708475,
+      "learning_rate": 6.149069794402922e-06,
+      "loss": 0.2583,
+      "step": 1453
+    },
+    {
+      "epoch": 0.4437662139478102,
+      "grad_norm": 1.5444477187990262,
+      "learning_rate": 6.144257293504826e-06,
+      "loss": 0.4095,
+      "step": 1454
+    },
+    {
+      "epoch": 0.4440714176712956,
+      "grad_norm": 1.2399450249944732,
+      "learning_rate": 6.139443673712666e-06,
+      "loss": 0.3351,
+      "step": 1455
+    },
+    {
+      "epoch": 0.444376621394781,
+      "grad_norm": 1.5587921398740923,
+      "learning_rate": 6.134628939733367e-06,
+      "loss": 0.4804,
+      "step": 1456
+    },
+    {
+      "epoch": 0.4446818251182664,
+      "grad_norm": 1.3042968347851616,
+      "learning_rate": 6.129813096274937e-06,
+      "loss": 0.37,
+      "step": 1457
+    },
+    {
+      "epoch": 0.4449870288417519,
+      "grad_norm": 1.3791281113704872,
+      "learning_rate": 6.124996148046478e-06,
+      "loss": 0.3613,
+      "step": 1458
+    },
+    {
+      "epoch": 0.4452922325652373,
+      "grad_norm": 1.2376992094205213,
+      "learning_rate": 6.120178099758163e-06,
+      "loss": 0.2888,
+      "step": 1459
+    },
+    {
+      "epoch": 0.4455974362887227,
+      "grad_norm": 1.5950209094264465,
+      "learning_rate": 6.1153589561212465e-06,
+      "loss": 0.4905,
+      "step": 1460
+    },
+    {
+      "epoch": 0.44590264001220814,
+      "grad_norm": 1.144267549829832,
+      "learning_rate": 6.110538721848055e-06,
+      "loss": 0.2428,
+      "step": 1461
+    },
+    {
+      "epoch": 0.44620784373569355,
+      "grad_norm": 1.8243651239455192,
+      "learning_rate": 6.1057174016519746e-06,
+      "loss": 0.4422,
+      "step": 1462
+    },
+    {
+      "epoch": 0.446513047459179,
+      "grad_norm": 1.4471328406739752,
+      "learning_rate": 6.100895000247461e-06,
+      "loss": 0.365,
+      "step": 1463
+    },
+    {
+      "epoch": 0.44681825118266444,
+      "grad_norm": 1.1681201032981883,
+      "learning_rate": 6.096071522350023e-06,
+      "loss": 0.3077,
+      "step": 1464
+    },
+    {
+      "epoch": 0.44712345490614985,
+      "grad_norm": 1.0017560798569047,
+      "learning_rate": 6.091246972676221e-06,
+      "loss": 0.2612,
+      "step": 1465
+    },
+    {
+      "epoch": 0.44742865862963527,
+      "grad_norm": 1.1198051570109302,
+      "learning_rate": 6.086421355943669e-06,
+      "loss": 0.2661,
+      "step": 1466
+    },
+    {
+      "epoch": 0.4477338623531207,
+      "grad_norm": 1.5027878603280684,
+      "learning_rate": 6.081594676871018e-06,
+      "loss": 0.3801,
+      "step": 1467
+    },
+    {
+      "epoch": 0.44803906607660615,
+      "grad_norm": 1.5399520549494288,
+      "learning_rate": 6.076766940177961e-06,
+      "loss": 0.472,
+      "step": 1468
+    },
+    {
+      "epoch": 0.44834426980009157,
+      "grad_norm": 1.0842776340009068,
+      "learning_rate": 6.071938150585225e-06,
+      "loss": 0.2661,
+      "step": 1469
+    },
+    {
+      "epoch": 0.448649473523577,
+      "grad_norm": 1.5030565908492635,
+      "learning_rate": 6.067108312814565e-06,
+      "loss": 0.4655,
+      "step": 1470
+    },
+    {
+      "epoch": 0.4489546772470624,
+      "grad_norm": 1.538417777942964,
+      "learning_rate": 6.062277431588764e-06,
+      "loss": 0.4456,
+      "step": 1471
+    },
+    {
+      "epoch": 0.44925988097054786,
+      "grad_norm": 1.4993872980628737,
+      "learning_rate": 6.0574455116316226e-06,
+      "loss": 0.4185,
+      "step": 1472
+    },
+    {
+      "epoch": 0.4495650846940333,
+      "grad_norm": 1.5907228930614425,
+      "learning_rate": 6.052612557667958e-06,
+      "loss": 0.3011,
+      "step": 1473
+    },
+    {
+      "epoch": 0.4498702884175187,
+      "grad_norm": 1.2701054140331611,
+      "learning_rate": 6.0477785744236e-06,
+      "loss": 0.3505,
+      "step": 1474
+    },
+    {
+      "epoch": 0.4501754921410041,
+      "grad_norm": 1.195165338370715,
+      "learning_rate": 6.042943566625381e-06,
+      "loss": 0.342,
+      "step": 1475
+    },
+    {
+      "epoch": 0.4504806958644895,
+      "grad_norm": 1.751726797192751,
+      "learning_rate": 6.038107539001139e-06,
+      "loss": 0.3494,
+      "step": 1476
+    },
+    {
+      "epoch": 0.450785899587975,
+      "grad_norm": 1.6552717810878512,
+      "learning_rate": 6.033270496279707e-06,
+      "loss": 0.2819,
+      "step": 1477
+    },
+    {
+      "epoch": 0.4510911033114604,
+      "grad_norm": 1.3423149963148238,
+      "learning_rate": 6.0284324431909126e-06,
+      "loss": 0.4002,
+      "step": 1478
+    },
+    {
+      "epoch": 0.4513963070349458,
+      "grad_norm": 1.5006613862735463,
+      "learning_rate": 6.023593384465569e-06,
+      "loss": 0.4365,
+      "step": 1479
+    },
+    {
+      "epoch": 0.45170151075843124,
+      "grad_norm": 1.4902782908677048,
+      "learning_rate": 6.018753324835474e-06,
+      "loss": 0.1634,
+      "step": 1480
+    },
+    {
+      "epoch": 0.4520067144819167,
+      "grad_norm": 1.1448537743161962,
+      "learning_rate": 6.013912269033405e-06,
+      "loss": 0.196,
+      "step": 1481
+    },
+    {
+      "epoch": 0.4523119182054021,
+      "grad_norm": 1.4541649327454527,
+      "learning_rate": 6.009070221793113e-06,
+      "loss": 0.413,
+      "step": 1482
+    },
+    {
+      "epoch": 0.45261712192888753,
+      "grad_norm": 1.1981770218692536,
+      "learning_rate": 6.004227187849315e-06,
+      "loss": 0.2623,
+      "step": 1483
+    },
+    {
+      "epoch": 0.45292232565237295,
+      "grad_norm": 1.0048017849150068,
+      "learning_rate": 5.999383171937699e-06,
+      "loss": 0.2374,
+      "step": 1484
+    },
+    {
+      "epoch": 0.45322752937585836,
+      "grad_norm": 1.3030578120266822,
+      "learning_rate": 5.99453817879491e-06,
+      "loss": 0.32,
+      "step": 1485
+    },
+    {
+      "epoch": 0.45353273309934383,
+      "grad_norm": 1.7326148096661058,
+      "learning_rate": 5.989692213158547e-06,
+      "loss": 0.4677,
+      "step": 1486
+    },
+    {
+      "epoch": 0.45383793682282925,
+      "grad_norm": 1.6077233885750193,
+      "learning_rate": 5.984845279767161e-06,
+      "loss": 0.5707,
+      "step": 1487
+    },
+    {
+      "epoch": 0.45414314054631466,
+      "grad_norm": 1.290704871562614,
+      "learning_rate": 5.979997383360252e-06,
+      "loss": 0.3204,
+      "step": 1488
+    },
+    {
+      "epoch": 0.4544483442698001,
+      "grad_norm": 1.4568229301374171,
+      "learning_rate": 5.975148528678259e-06,
+      "loss": 0.3551,
+      "step": 1489
+    },
+    {
+      "epoch": 0.45475354799328555,
+      "grad_norm": 1.3095427303268419,
+      "learning_rate": 5.970298720462557e-06,
+      "loss": 0.2769,
+      "step": 1490
+    },
+    {
+      "epoch": 0.45505875171677096,
+      "grad_norm": 1.1544985522380489,
+      "learning_rate": 5.965447963455457e-06,
+      "loss": 0.3078,
+      "step": 1491
+    },
+    {
+      "epoch": 0.4553639554402564,
+      "grad_norm": 1.5958447621121368,
+      "learning_rate": 5.960596262400195e-06,
+      "loss": 0.2531,
+      "step": 1492
+    },
+    {
+      "epoch": 0.4556691591637418,
+      "grad_norm": 1.5606312830397104,
+      "learning_rate": 5.95574362204093e-06,
+      "loss": 0.3338,
+      "step": 1493
+    },
+    {
+      "epoch": 0.4559743628872272,
+      "grad_norm": 1.2261609920664052,
+      "learning_rate": 5.950890047122742e-06,
+      "loss": 0.3511,
+      "step": 1494
+    },
+    {
+      "epoch": 0.4562795666107127,
+      "grad_norm": 1.434002311342896,
+      "learning_rate": 5.9460355423916205e-06,
+      "loss": 0.3288,
+      "step": 1495
+    },
+    {
+      "epoch": 0.4565847703341981,
+      "grad_norm": 0.9862685988187032,
+      "learning_rate": 5.941180112594469e-06,
+      "loss": 0.2005,
+      "step": 1496
+    },
+    {
+      "epoch": 0.4568899740576835,
+      "grad_norm": 1.464340245497521,
+      "learning_rate": 5.936323762479093e-06,
+      "loss": 0.2452,
+      "step": 1497
+    },
+    {
+      "epoch": 0.4571951777811689,
+      "grad_norm": 1.554706899243958,
+      "learning_rate": 5.9314664967942e-06,
+      "loss": 0.33,
+      "step": 1498
+    },
+    {
+      "epoch": 0.45750038150465433,
+      "grad_norm": 1.4569953320859805,
+      "learning_rate": 5.926608320289388e-06,
+      "loss": 0.3072,
+      "step": 1499
+    },
+    {
+      "epoch": 0.4578055852281398,
+      "grad_norm": 1.4440211945507897,
+      "learning_rate": 5.92174923771515e-06,
+      "loss": 0.4295,
+      "step": 1500
+    },
+    {
+      "epoch": 0.4581107889516252,
+      "grad_norm": 1.669630451316585,
+      "learning_rate": 5.916889253822865e-06,
+      "loss": 0.5226,
+      "step": 1501
+    },
+    {
+      "epoch": 0.45841599267511063,
+      "grad_norm": 1.4250182736212258,
+      "learning_rate": 5.912028373364791e-06,
+      "loss": 0.3333,
+      "step": 1502
+    },
+    {
+      "epoch": 0.45872119639859604,
+      "grad_norm": 1.0328058008851029,
+      "learning_rate": 5.907166601094063e-06,
+      "loss": 0.2765,
+      "step": 1503
+    },
+    {
+      "epoch": 0.4590264001220815,
+      "grad_norm": 1.6987471002662056,
+      "learning_rate": 5.9023039417646895e-06,
+      "loss": 0.3082,
+      "step": 1504
+    },
+    {
+      "epoch": 0.4593316038455669,
+      "grad_norm": 1.3004454216451125,
+      "learning_rate": 5.897440400131546e-06,
+      "loss": 0.3204,
+      "step": 1505
+    },
+    {
+      "epoch": 0.45963680756905234,
+      "grad_norm": 1.4760594394056248,
+      "learning_rate": 5.8925759809503725e-06,
+      "loss": 0.4925,
+      "step": 1506
+    },
+    {
+      "epoch": 0.45994201129253776,
+      "grad_norm": 1.0003988543450875,
+      "learning_rate": 5.887710688977762e-06,
+      "loss": 0.2494,
+      "step": 1507
+    },
+    {
+      "epoch": 0.46024721501602317,
+      "grad_norm": 1.4474390829441592,
+      "learning_rate": 5.882844528971164e-06,
+      "loss": 0.421,
+      "step": 1508
+    },
+    {
+      "epoch": 0.46055241873950864,
+      "grad_norm": 1.4506829363533515,
+      "learning_rate": 5.8779775056888795e-06,
+      "loss": 0.3237,
+      "step": 1509
+    },
+    {
+      "epoch": 0.46085762246299405,
+      "grad_norm": 1.102194097730089,
+      "learning_rate": 5.873109623890047e-06,
+      "loss": 0.3195,
+      "step": 1510
+    },
+    {
+      "epoch": 0.46116282618647947,
+      "grad_norm": 1.4653483831299812,
+      "learning_rate": 5.8682408883346535e-06,
+      "loss": 0.4283,
+      "step": 1511
+    },
+    {
+      "epoch": 0.4614680299099649,
+      "grad_norm": 1.5580902433592914,
+      "learning_rate": 5.863371303783511e-06,
+      "loss": 0.4983,
+      "step": 1512
+    },
+    {
+      "epoch": 0.46177323363345035,
+      "grad_norm": 1.859287163719373,
+      "learning_rate": 5.858500874998268e-06,
+      "loss": 0.4035,
+      "step": 1513
+    },
+    {
+      "epoch": 0.46207843735693577,
+      "grad_norm": 1.3899736057492478,
+      "learning_rate": 5.853629606741398e-06,
+      "loss": 0.3503,
+      "step": 1514
+    },
+    {
+      "epoch": 0.4623836410804212,
+      "grad_norm": 1.9654382371342167,
+      "learning_rate": 5.848757503776194e-06,
+      "loss": 0.2825,
+      "step": 1515
+    },
+    {
+      "epoch": 0.4626888448039066,
+      "grad_norm": 1.749083142563443,
+      "learning_rate": 5.843884570866764e-06,
+      "loss": 0.2898,
+      "step": 1516
+    },
+    {
+      "epoch": 0.462994048527392,
+      "grad_norm": 2.06586332345819,
+      "learning_rate": 5.839010812778033e-06,
+      "loss": 0.4441,
+      "step": 1517
+    },
+    {
+      "epoch": 0.4632992522508775,
+      "grad_norm": 1.416665899987107,
+      "learning_rate": 5.834136234275725e-06,
+      "loss": 0.2629,
+      "step": 1518
+    },
+    {
+      "epoch": 0.4636044559743629,
+      "grad_norm": 1.491807979233761,
+      "learning_rate": 5.829260840126371e-06,
+      "loss": 0.3957,
+      "step": 1519
+    },
+    {
+      "epoch": 0.4639096596978483,
+      "grad_norm": 1.823810409162459,
+      "learning_rate": 5.824384635097299e-06,
+      "loss": 0.3113,
+      "step": 1520
+    },
+    {
+      "epoch": 0.4642148634213337,
+      "grad_norm": 1.43257042224072,
+      "learning_rate": 5.819507623956631e-06,
+      "loss": 0.3604,
+      "step": 1521
+    },
+    {
+      "epoch": 0.4645200671448192,
+      "grad_norm": 1.6999961039554605,
+      "learning_rate": 5.814629811473274e-06,
+      "loss": 0.3102,
+      "step": 1522
+    },
+    {
+      "epoch": 0.4648252708683046,
+      "grad_norm": 1.35742920798098,
+      "learning_rate": 5.809751202416922e-06,
+      "loss": 0.2888,
+      "step": 1523
+    },
+    {
+      "epoch": 0.46513047459179,
+      "grad_norm": 1.7134434015762743,
+      "learning_rate": 5.804871801558045e-06,
+      "loss": 0.4024,
+      "step": 1524
+    },
+    {
+      "epoch": 0.46543567831527544,
+      "grad_norm": 1.1985209171155202,
+      "learning_rate": 5.79999161366789e-06,
+      "loss": 0.2442,
+      "step": 1525
+    },
+    {
+      "epoch": 0.46574088203876085,
+      "grad_norm": 1.3407548748143259,
+      "learning_rate": 5.795110643518473e-06,
+      "loss": 0.1733,
+      "step": 1526
+    },
+    {
+      "epoch": 0.4660460857622463,
+      "grad_norm": 1.5247299580605493,
+      "learning_rate": 5.790228895882571e-06,
+      "loss": 0.3787,
+      "step": 1527
+    },
+    {
+      "epoch": 0.46635128948573173,
+      "grad_norm": 1.535528006931427,
+      "learning_rate": 5.785346375533726e-06,
+      "loss": 0.4336,
+      "step": 1528
+    },
+    {
+      "epoch": 0.46665649320921715,
+      "grad_norm": 1.3568775430354145,
+      "learning_rate": 5.7804630872462354e-06,
+      "loss": 0.1561,
+      "step": 1529
+    },
+    {
+      "epoch": 0.46696169693270256,
+      "grad_norm": 1.5873546563844085,
+      "learning_rate": 5.775579035795145e-06,
+      "loss": 0.4776,
+      "step": 1530
+    },
+    {
+      "epoch": 0.467266900656188,
+      "grad_norm": 1.4509069270195842,
+      "learning_rate": 5.770694225956248e-06,
+      "loss": 0.3115,
+      "step": 1531
+    },
+    {
+      "epoch": 0.46757210437967345,
+      "grad_norm": 1.4283327448223122,
+      "learning_rate": 5.765808662506079e-06,
+      "loss": 0.3693,
+      "step": 1532
+    },
+    {
+      "epoch": 0.46787730810315886,
+      "grad_norm": 1.6978039694364109,
+      "learning_rate": 5.760922350221909e-06,
+      "loss": 0.2827,
+      "step": 1533
+    },
+    {
+      "epoch": 0.4681825118266443,
+      "grad_norm": 1.6065097227778868,
+      "learning_rate": 5.756035293881744e-06,
+      "loss": 0.2631,
+      "step": 1534
+    },
+    {
+      "epoch": 0.4684877155501297,
+      "grad_norm": 1.5416742616758474,
+      "learning_rate": 5.751147498264314e-06,
+      "loss": 0.5121,
+      "step": 1535
+    },
+    {
+      "epoch": 0.46879291927361516,
+      "grad_norm": 1.4642775598491673,
+      "learning_rate": 5.746258968149073e-06,
+      "loss": 0.411,
+      "step": 1536
+    },
+    {
+      "epoch": 0.4690981229971006,
+      "grad_norm": 1.8272083951571823,
+      "learning_rate": 5.7413697083161944e-06,
+      "loss": 0.3571,
+      "step": 1537
+    },
+    {
+      "epoch": 0.469403326720586,
+      "grad_norm": 1.4317064048986723,
+      "learning_rate": 5.736479723546564e-06,
+      "loss": 0.3125,
+      "step": 1538
+    },
+    {
+      "epoch": 0.4697085304440714,
+      "grad_norm": 1.314235538878401,
+      "learning_rate": 5.731589018621777e-06,
+      "loss": 0.1967,
+      "step": 1539
+    },
+    {
+      "epoch": 0.4700137341675568,
+      "grad_norm": 1.5035862966267712,
+      "learning_rate": 5.726697598324132e-06,
+      "loss": 0.3313,
+      "step": 1540
+    },
+    {
+      "epoch": 0.4703189378910423,
+      "grad_norm": 1.3404934923098695,
+      "learning_rate": 5.721805467436628e-06,
+      "loss": 0.3265,
+      "step": 1541
+    },
+    {
+      "epoch": 0.4706241416145277,
+      "grad_norm": 1.1548514284630382,
+      "learning_rate": 5.71691263074296e-06,
+      "loss": 0.1728,
+      "step": 1542
+    },
+    {
+      "epoch": 0.4709293453380131,
+      "grad_norm": 1.3183575665496856,
+      "learning_rate": 5.71201909302751e-06,
+      "loss": 0.431,
+      "step": 1543
+    },
+    {
+      "epoch": 0.47123454906149853,
+      "grad_norm": 1.4548282280692988,
+      "learning_rate": 5.707124859075346e-06,
+      "loss": 0.4382,
+      "step": 1544
+    },
+    {
+      "epoch": 0.471539752784984,
+      "grad_norm": 1.3806205661714575,
+      "learning_rate": 5.70222993367222e-06,
+      "loss": 0.1854,
+      "step": 1545
+    },
+    {
+      "epoch": 0.4718449565084694,
+      "grad_norm": 1.6537668579349547,
+      "learning_rate": 5.697334321604557e-06,
+      "loss": 0.4702,
+      "step": 1546
+    },
+    {
+      "epoch": 0.47215016023195483,
+      "grad_norm": 1.3352342505798407,
+      "learning_rate": 5.692438027659457e-06,
+      "loss": 0.4055,
+      "step": 1547
+    },
+    {
+      "epoch": 0.47245536395544024,
+      "grad_norm": 1.288274350677023,
+      "learning_rate": 5.68754105662468e-06,
+      "loss": 0.3408,
+      "step": 1548
+    },
+    {
+      "epoch": 0.47276056767892566,
+      "grad_norm": 1.4151500924241693,
+      "learning_rate": 5.682643413288656e-06,
+      "loss": 0.4746,
+      "step": 1549
+    },
+    {
+      "epoch": 0.47306577140241113,
+      "grad_norm": 1.2957954969534842,
+      "learning_rate": 5.677745102440468e-06,
+      "loss": 0.3128,
+      "step": 1550
+    },
+    {
+      "epoch": 0.47337097512589654,
+      "grad_norm": 1.4892185093430534,
+      "learning_rate": 5.672846128869853e-06,
+      "loss": 0.1999,
+      "step": 1551
+    },
+    {
+      "epoch": 0.47367617884938196,
+      "grad_norm": 1.159495669766031,
+      "learning_rate": 5.667946497367191e-06,
+      "loss": 0.2935,
+      "step": 1552
+    },
+    {
+      "epoch": 0.47398138257286737,
+      "grad_norm": 1.3909682803839567,
+      "learning_rate": 5.663046212723516e-06,
+      "loss": 0.4137,
+      "step": 1553
+    },
+    {
+      "epoch": 0.47428658629635284,
+      "grad_norm": 1.2459299583771315,
+      "learning_rate": 5.658145279730491e-06,
+      "loss": 0.255,
+      "step": 1554
+    },
+    {
+      "epoch": 0.47459179001983826,
+      "grad_norm": 1.3887686179326861,
+      "learning_rate": 5.65324370318042e-06,
+      "loss": 0.3186,
+      "step": 1555
+    },
+    {
+      "epoch": 0.47489699374332367,
+      "grad_norm": 1.6068867827569175,
+      "learning_rate": 5.648341487866228e-06,
+      "loss": 0.3965,
+      "step": 1556
+    },
+    {
+      "epoch": 0.4752021974668091,
+      "grad_norm": 1.5957802201770763,
+      "learning_rate": 5.643438638581472e-06,
+      "loss": 0.527,
+      "step": 1557
+    },
+    {
+      "epoch": 0.4755074011902945,
+      "grad_norm": 1.7927880289589633,
+      "learning_rate": 5.638535160120328e-06,
+      "loss": 0.5276,
+      "step": 1558
+    },
+    {
+      "epoch": 0.47581260491377997,
+      "grad_norm": 1.473208619804551,
+      "learning_rate": 5.633631057277582e-06,
+      "loss": 0.3348,
+      "step": 1559
+    },
+    {
+      "epoch": 0.4761178086372654,
+      "grad_norm": 1.1516190658610226,
+      "learning_rate": 5.6287263348486356e-06,
+      "loss": 0.2193,
+      "step": 1560
+    },
+    {
+      "epoch": 0.4764230123607508,
+      "grad_norm": 1.4141650926667002,
+      "learning_rate": 5.6238209976294955e-06,
+      "loss": 0.2721,
+      "step": 1561
+    },
+    {
+      "epoch": 0.4767282160842362,
+      "grad_norm": 1.342084606698652,
+      "learning_rate": 5.618915050416768e-06,
+      "loss": 0.3196,
+      "step": 1562
+    },
+    {
+      "epoch": 0.4770334198077217,
+      "grad_norm": 1.3503394795013224,
+      "learning_rate": 5.614008498007658e-06,
+      "loss": 0.3637,
+      "step": 1563
+    },
+    {
+      "epoch": 0.4773386235312071,
+      "grad_norm": 1.328424891104901,
+      "learning_rate": 5.60910134519996e-06,
+      "loss": 0.3218,
+      "step": 1564
+    },
+    {
+      "epoch": 0.4776438272546925,
+      "grad_norm": 1.3485172340980611,
+      "learning_rate": 5.604193596792054e-06,
+      "loss": 0.3297,
+      "step": 1565
+    },
+    {
+      "epoch": 0.4779490309781779,
+      "grad_norm": 1.2738033775398612,
+      "learning_rate": 5.599285257582911e-06,
+      "loss": 0.2263,
+      "step": 1566
+    },
+    {
+      "epoch": 0.47825423470166334,
+      "grad_norm": 1.3885241723681556,
+      "learning_rate": 5.594376332372068e-06,
+      "loss": 0.3373,
+      "step": 1567
+    },
+    {
+      "epoch": 0.4785594384251488,
+      "grad_norm": 1.1021865808381979,
+      "learning_rate": 5.589466825959642e-06,
+      "loss": 0.2387,
+      "step": 1568
+    },
+    {
+      "epoch": 0.4788646421486342,
+      "grad_norm": 1.428554398571414,
+      "learning_rate": 5.5845567431463175e-06,
+      "loss": 0.2956,
+      "step": 1569
+    },
+    {
+      "epoch": 0.47916984587211964,
+      "grad_norm": 1.3897546337563043,
+      "learning_rate": 5.579646088733343e-06,
+      "loss": 0.2354,
+      "step": 1570
+    },
+    {
+      "epoch": 0.47947504959560505,
+      "grad_norm": 1.4542362517109815,
+      "learning_rate": 5.574734867522523e-06,
+      "loss": 0.3921,
+      "step": 1571
+    },
+    {
+      "epoch": 0.47978025331909047,
+      "grad_norm": 1.7263399558968118,
+      "learning_rate": 5.569823084316219e-06,
+      "loss": 0.3029,
+      "step": 1572
+    },
+    {
+      "epoch": 0.48008545704257594,
+      "grad_norm": 1.621517558023657,
+      "learning_rate": 5.564910743917341e-06,
+      "loss": 0.3932,
+      "step": 1573
+    },
+    {
+      "epoch": 0.48039066076606135,
+      "grad_norm": 2.035620579678143,
+      "learning_rate": 5.559997851129346e-06,
+      "loss": 0.1695,
+      "step": 1574
+    },
+    {
+      "epoch": 0.48069586448954676,
+      "grad_norm": 1.3107818984028823,
+      "learning_rate": 5.555084410756226e-06,
+      "loss": 0.2542,
+      "step": 1575
+    },
+    {
+      "epoch": 0.4810010682130322,
+      "grad_norm": 1.6572997886831133,
+      "learning_rate": 5.550170427602513e-06,
+      "loss": 0.2124,
+      "step": 1576
+    },
+    {
+      "epoch": 0.48130627193651765,
+      "grad_norm": 2.748635560446825,
+      "learning_rate": 5.54525590647327e-06,
+      "loss": 0.3052,
+      "step": 1577
+    },
+    {
+      "epoch": 0.48161147566000306,
+      "grad_norm": 1.995315190898183,
+      "learning_rate": 5.540340852174083e-06,
+      "loss": 0.3363,
+      "step": 1578
+    },
+    {
+      "epoch": 0.4819166793834885,
+      "grad_norm": 1.5715192267042497,
+      "learning_rate": 5.53542526951106e-06,
+      "loss": 0.2548,
+      "step": 1579
+    },
+    {
+      "epoch": 0.4822218831069739,
+      "grad_norm": 1.4893926040435048,
+      "learning_rate": 5.530509163290829e-06,
+      "loss": 0.371,
+      "step": 1580
+    },
+    {
+      "epoch": 0.4825270868304593,
+      "grad_norm": 1.6286037539259315,
+      "learning_rate": 5.525592538320526e-06,
+      "loss": 0.376,
+      "step": 1581
+    },
+    {
+      "epoch": 0.4828322905539448,
+      "grad_norm": 1.7629836588949888,
+      "learning_rate": 5.520675399407795e-06,
+      "loss": 0.5786,
+      "step": 1582
+    },
+    {
+      "epoch": 0.4831374942774302,
+      "grad_norm": 1.4728271223346432,
+      "learning_rate": 5.515757751360784e-06,
+      "loss": 0.3431,
+      "step": 1583
+    },
+    {
+      "epoch": 0.4834426980009156,
+      "grad_norm": 1.1580318948342616,
+      "learning_rate": 5.510839598988137e-06,
+      "loss": 0.3826,
+      "step": 1584
+    },
+    {
+      "epoch": 0.483747901724401,
+      "grad_norm": 1.337925018091232,
+      "learning_rate": 5.505920947098995e-06,
+      "loss": 0.2725,
+      "step": 1585
+    },
+    {
+      "epoch": 0.4840531054478865,
+      "grad_norm": 1.6461638187220182,
+      "learning_rate": 5.50100180050298e-06,
+      "loss": 0.4827,
+      "step": 1586
+    },
+    {
+      "epoch": 0.4843583091713719,
+      "grad_norm": 1.4549773520578322,
+      "learning_rate": 5.496082164010207e-06,
+      "loss": 0.2952,
+      "step": 1587
+    },
+    {
+      "epoch": 0.4846635128948573,
+      "grad_norm": 1.4982046668394888,
+      "learning_rate": 5.491162042431264e-06,
+      "loss": 0.46,
+      "step": 1588
+    },
+    {
+      "epoch": 0.48496871661834273,
+      "grad_norm": 1.7000123023541676,
+      "learning_rate": 5.486241440577214e-06,
+      "loss": 0.2381,
+      "step": 1589
+    },
+    {
+      "epoch": 0.48527392034182815,
+      "grad_norm": 1.4018936497489025,
+      "learning_rate": 5.48132036325959e-06,
+      "loss": 0.3872,
+      "step": 1590
+    },
+    {
+      "epoch": 0.4855791240653136,
+      "grad_norm": 1.4151807546921091,
+      "learning_rate": 5.476398815290394e-06,
+      "loss": 0.381,
+      "step": 1591
+    },
+    {
+      "epoch": 0.48588432778879903,
+      "grad_norm": 0.9876963577129196,
+      "learning_rate": 5.471476801482079e-06,
+      "loss": 0.1868,
+      "step": 1592
+    },
+    {
+      "epoch": 0.48618953151228445,
+      "grad_norm": 1.6726232975328992,
+      "learning_rate": 5.466554326647564e-06,
+      "loss": 0.3994,
+      "step": 1593
+    },
+    {
+      "epoch": 0.48649473523576986,
+      "grad_norm": 1.2261133525745127,
+      "learning_rate": 5.461631395600208e-06,
+      "loss": 0.3083,
+      "step": 1594
+    },
+    {
+      "epoch": 0.48679993895925533,
+      "grad_norm": 2.0566916319054735,
+      "learning_rate": 5.456708013153829e-06,
+      "loss": 0.4756,
+      "step": 1595
+    },
+    {
+      "epoch": 0.48710514268274074,
+      "grad_norm": 1.4786102788703337,
+      "learning_rate": 5.451784184122676e-06,
+      "loss": 0.5202,
+      "step": 1596
+    },
+    {
+      "epoch": 0.48741034640622616,
+      "grad_norm": 1.3638063310410486,
+      "learning_rate": 5.446859913321437e-06,
+      "loss": 0.316,
+      "step": 1597
+    },
+    {
+      "epoch": 0.4877155501297116,
+      "grad_norm": 1.4931742815624802,
+      "learning_rate": 5.4419352055652345e-06,
+      "loss": 0.3344,
+      "step": 1598
+    },
+    {
+      "epoch": 0.488020753853197,
+      "grad_norm": 1.5569878051025439,
+      "learning_rate": 5.437010065669618e-06,
+      "loss": 0.3727,
+      "step": 1599
+    },
+    {
+      "epoch": 0.48832595757668246,
+      "grad_norm": 1.1973477657449398,
+      "learning_rate": 5.432084498450555e-06,
+      "loss": 0.253,
+      "step": 1600
+    },
+    {
+      "epoch": 0.48863116130016787,
+      "grad_norm": 1.4804547392250123,
+      "learning_rate": 5.427158508724436e-06,
+      "loss": 0.1928,
+      "step": 1601
+    },
+    {
+      "epoch": 0.4889363650236533,
+      "grad_norm": 1.5889388374815445,
+      "learning_rate": 5.422232101308064e-06,
+      "loss": 0.3242,
+      "step": 1602
+    },
+    {
+      "epoch": 0.4892415687471387,
+      "grad_norm": 1.4732208383918126,
+      "learning_rate": 5.417305281018648e-06,
+      "loss": 0.3179,
+      "step": 1603
+    },
+    {
+      "epoch": 0.4895467724706241,
+      "grad_norm": 1.5778071772757023,
+      "learning_rate": 5.4123780526738025e-06,
+      "loss": 0.3705,
+      "step": 1604
+    },
+    {
+      "epoch": 0.4898519761941096,
+      "grad_norm": 1.5952772414956238,
+      "learning_rate": 5.4074504210915405e-06,
+      "loss": 0.4752,
+      "step": 1605
+    },
+    {
+      "epoch": 0.490157179917595,
+      "grad_norm": 1.5327834411417531,
+      "learning_rate": 5.402522391090268e-06,
+      "loss": 0.2701,
+      "step": 1606
+    },
+    {
+      "epoch": 0.4904623836410804,
+      "grad_norm": 1.4174549863957768,
+      "learning_rate": 5.397593967488784e-06,
+      "loss": 0.305,
+      "step": 1607
+    },
+    {
+      "epoch": 0.4907675873645658,
+      "grad_norm": 1.4105364426993232,
+      "learning_rate": 5.3926651551062696e-06,
+      "loss": 0.2822,
+      "step": 1608
+    },
+    {
+      "epoch": 0.4910727910880513,
+      "grad_norm": 1.496829496898576,
+      "learning_rate": 5.387735958762284e-06,
+      "loss": 0.2455,
+      "step": 1609
+    },
+    {
+      "epoch": 0.4913779948115367,
+      "grad_norm": 1.3744912073153184,
+      "learning_rate": 5.38280638327677e-06,
+      "loss": 0.222,
+      "step": 1610
+    },
+    {
+      "epoch": 0.4916831985350221,
+      "grad_norm": 2.181978942265985,
+      "learning_rate": 5.377876433470031e-06,
+      "loss": 0.2868,
+      "step": 1611
+    },
+    {
+      "epoch": 0.49198840225850754,
+      "grad_norm": 1.6941287922726327,
+      "learning_rate": 5.372946114162745e-06,
+      "loss": 0.334,
+      "step": 1612
+    },
+    {
+      "epoch": 0.49229360598199295,
+      "grad_norm": 1.2726505398841743,
+      "learning_rate": 5.3680154301759435e-06,
+      "loss": 0.258,
+      "step": 1613
+    },
+    {
+      "epoch": 0.4925988097054784,
+      "grad_norm": 1.6061588484137834,
+      "learning_rate": 5.363084386331021e-06,
+      "loss": 0.4377,
+      "step": 1614
+    },
+    {
+      "epoch": 0.49290401342896384,
+      "grad_norm": 1.318008850970961,
+      "learning_rate": 5.358152987449722e-06,
+      "loss": 0.2085,
+      "step": 1615
+    },
+    {
+      "epoch": 0.49320921715244925,
+      "grad_norm": 1.2747270104863346,
+      "learning_rate": 5.353221238354136e-06,
+      "loss": 0.2869,
+      "step": 1616
+    },
+    {
+      "epoch": 0.49351442087593467,
+      "grad_norm": 1.6264464469792288,
+      "learning_rate": 5.348289143866696e-06,
+      "loss": 0.5463,
+      "step": 1617
+    },
+    {
+      "epoch": 0.49381962459942014,
+      "grad_norm": 1.455135472338877,
+      "learning_rate": 5.343356708810176e-06,
+      "loss": 0.4526,
+      "step": 1618
+    },
+    {
+      "epoch": 0.49412482832290555,
+      "grad_norm": 1.1443888604966979,
+      "learning_rate": 5.338423938007677e-06,
+      "loss": 0.2367,
+      "step": 1619
+    },
+    {
+      "epoch": 0.49443003204639097,
+      "grad_norm": 1.351694662875841,
+      "learning_rate": 5.3334908362826355e-06,
+      "loss": 0.1923,
+      "step": 1620
+    },
+    {
+      "epoch": 0.4947352357698764,
+      "grad_norm": 1.3722275613293962,
+      "learning_rate": 5.328557408458805e-06,
+      "loss": 0.2233,
+      "step": 1621
+    },
+    {
+      "epoch": 0.4950404394933618,
+      "grad_norm": 1.4038649360296689,
+      "learning_rate": 5.323623659360261e-06,
+      "loss": 0.2664,
+      "step": 1622
+    },
+    {
+      "epoch": 0.49534564321684726,
+      "grad_norm": 1.3069014353110344,
+      "learning_rate": 5.3186895938113934e-06,
+      "loss": 0.3407,
+      "step": 1623
+    },
+    {
+      "epoch": 0.4956508469403327,
+      "grad_norm": 1.2914138618332602,
+      "learning_rate": 5.3137552166369e-06,
+      "loss": 0.2242,
+      "step": 1624
+    },
+    {
+      "epoch": 0.4959560506638181,
+      "grad_norm": 1.4023515706388823,
+      "learning_rate": 5.3088205326617846e-06,
+      "loss": 0.432,
+      "step": 1625
+    },
+    {
+      "epoch": 0.4962612543873035,
+      "grad_norm": 4.034863885145975,
+      "learning_rate": 5.303885546711352e-06,
+      "loss": 0.3231,
+      "step": 1626
+    },
+    {
+      "epoch": 0.496566458110789,
+      "grad_norm": 1.3258072319474576,
+      "learning_rate": 5.298950263611197e-06,
+      "loss": 0.4098,
+      "step": 1627
+    },
+    {
+      "epoch": 0.4968716618342744,
+      "grad_norm": 1.4994044711169068,
+      "learning_rate": 5.294014688187213e-06,
+      "loss": 0.1783,
+      "step": 1628
+    },
+    {
+      "epoch": 0.4971768655577598,
+      "grad_norm": 1.0794348016638573,
+      "learning_rate": 5.289078825265573e-06,
+      "loss": 0.2446,
+      "step": 1629
+    },
+    {
+      "epoch": 0.4974820692812452,
+      "grad_norm": 1.2325800145854007,
+      "learning_rate": 5.284142679672733e-06,
+      "loss": 0.3066,
+      "step": 1630
+    },
+    {
+      "epoch": 0.49778727300473063,
+      "grad_norm": 1.5888945724293768,
+      "learning_rate": 5.279206256235429e-06,
+      "loss": 0.2676,
+      "step": 1631
+    },
+    {
+      "epoch": 0.4980924767282161,
+      "grad_norm": 1.5276496756436473,
+      "learning_rate": 5.27426955978066e-06,
+      "loss": 0.3483,
+      "step": 1632
+    },
+    {
+      "epoch": 0.4983976804517015,
+      "grad_norm": 1.621231110050268,
+      "learning_rate": 5.2693325951357e-06,
+      "loss": 0.4674,
+      "step": 1633
+    },
+    {
+      "epoch": 0.49870288417518693,
+      "grad_norm": 1.625742375918665,
+      "learning_rate": 5.264395367128085e-06,
+      "loss": 0.5141,
+      "step": 1634
+    },
+    {
+      "epoch": 0.49900808789867235,
+      "grad_norm": 1.3288540802436513,
+      "learning_rate": 5.259457880585599e-06,
+      "loss": 0.2809,
+      "step": 1635
+    },
+    {
+      "epoch": 0.4993132916221578,
+      "grad_norm": 1.4498227241274997,
+      "learning_rate": 5.254520140336293e-06,
+      "loss": 0.3109,
+      "step": 1636
+    },
+    {
+      "epoch": 0.49961849534564323,
+      "grad_norm": 1.411890754290964,
+      "learning_rate": 5.249582151208456e-06,
+      "loss": 0.3035,
+      "step": 1637
+    },
+    {
+      "epoch": 0.49992369906912865,
+      "grad_norm": 1.827351333060968,
+      "learning_rate": 5.244643918030623e-06,
+      "loss": 0.2952,
+      "step": 1638
+    },
+    {
+      "epoch": 0.5002289027926141,
+      "grad_norm": 1.4941225138046528,
+      "learning_rate": 5.23970544563157e-06,
+      "loss": 0.349,
+      "step": 1639
+    },
+    {
+      "epoch": 0.5005341065160995,
+      "grad_norm": 1.359880901338943,
+      "learning_rate": 5.234766738840301e-06,
+      "loss": 0.3015,
+      "step": 1640
+    },
+    {
+      "epoch": 0.500839310239585,
+      "grad_norm": 1.0754563848784184,
+      "learning_rate": 5.229827802486057e-06,
+      "loss": 0.2577,
+      "step": 1641
+    },
+    {
+      "epoch": 0.5011445139630704,
+      "grad_norm": 1.1574633649791777,
+      "learning_rate": 5.224888641398298e-06,
+      "loss": 0.2131,
+      "step": 1642
+    },
+    {
+      "epoch": 0.5014497176865558,
+      "grad_norm": 1.3888886144425863,
+      "learning_rate": 5.219949260406702e-06,
+      "loss": 0.2927,
+      "step": 1643
+    },
+    {
+      "epoch": 0.5017549214100412,
+      "grad_norm": 1.2080626787509892,
+      "learning_rate": 5.215009664341172e-06,
+      "loss": 0.1779,
+      "step": 1644
+    },
+    {
+      "epoch": 0.5020601251335266,
+      "grad_norm": 1.5390181946549573,
+      "learning_rate": 5.210069858031809e-06,
+      "loss": 0.4169,
+      "step": 1645
+    },
+    {
+      "epoch": 0.502365328857012,
+      "grad_norm": 1.6666794379062893,
+      "learning_rate": 5.205129846308927e-06,
+      "loss": 0.5059,
+      "step": 1646
+    },
+    {
+      "epoch": 0.5026705325804974,
+      "grad_norm": 1.1965539506640186,
+      "learning_rate": 5.200189634003041e-06,
+      "loss": 0.2838,
+      "step": 1647
+    },
+    {
+      "epoch": 0.502975736303983,
+      "grad_norm": 1.5427230446178664,
+      "learning_rate": 5.195249225944858e-06,
+      "loss": 0.5243,
+      "step": 1648
+    },
+    {
+      "epoch": 0.5032809400274684,
+      "grad_norm": 1.713220684130789,
+      "learning_rate": 5.190308626965278e-06,
+      "loss": 0.4358,
+      "step": 1649
+    },
+    {
+      "epoch": 0.5035861437509538,
+      "grad_norm": 1.972570557196598,
+      "learning_rate": 5.18536784189539e-06,
+      "loss": 0.1848,
+      "step": 1650
+    },
+    {
+      "epoch": 0.5038913474744392,
+      "grad_norm": 1.6571295939699278,
+      "learning_rate": 5.180426875566461e-06,
+      "loss": 0.6615,
+      "step": 1651
+    },
+    {
+      "epoch": 0.5041965511979246,
+      "grad_norm": 1.437005165181072,
+      "learning_rate": 5.175485732809941e-06,
+      "loss": 0.308,
+      "step": 1652
+    },
+    {
+      "epoch": 0.50450175492141,
+      "grad_norm": 1.0623456057400325,
+      "learning_rate": 5.170544418457447e-06,
+      "loss": 0.1621,
+      "step": 1653
+    },
+    {
+      "epoch": 0.5048069586448954,
+      "grad_norm": 1.7579201898436299,
+      "learning_rate": 5.165602937340765e-06,
+      "loss": 0.3202,
+      "step": 1654
+    },
+    {
+      "epoch": 0.5051121623683809,
+      "grad_norm": 1.5000292457272657,
+      "learning_rate": 5.160661294291845e-06,
+      "loss": 0.2291,
+      "step": 1655
+    },
+    {
+      "epoch": 0.5054173660918663,
+      "grad_norm": 0.9777395860180313,
+      "learning_rate": 5.155719494142799e-06,
+      "loss": 0.1873,
+      "step": 1656
+    },
+    {
+      "epoch": 0.5057225698153518,
+      "grad_norm": 1.503059921917863,
+      "learning_rate": 5.150777541725884e-06,
+      "loss": 0.2805,
+      "step": 1657
+    },
+    {
+      "epoch": 0.5060277735388372,
+      "grad_norm": 1.672907929102723,
+      "learning_rate": 5.145835441873514e-06,
+      "loss": 0.3044,
+      "step": 1658
+    },
+    {
+      "epoch": 0.5063329772623226,
+      "grad_norm": 2.0294129500910034,
+      "learning_rate": 5.140893199418238e-06,
+      "loss": 0.3157,
+      "step": 1659
+    },
+    {
+      "epoch": 0.506638180985808,
+      "grad_norm": 1.3093133570470568,
+      "learning_rate": 5.1359508191927585e-06,
+      "loss": 0.2408,
+      "step": 1660
+    },
+    {
+      "epoch": 0.5069433847092935,
+      "grad_norm": 1.3018779083966656,
+      "learning_rate": 5.131008306029901e-06,
+      "loss": 0.3438,
+      "step": 1661
+    },
+    {
+      "epoch": 0.5072485884327789,
+      "grad_norm": 1.5384066970844072,
+      "learning_rate": 5.126065664762623e-06,
+      "loss": 0.5485,
+      "step": 1662
+    },
+    {
+      "epoch": 0.5075537921562643,
+      "grad_norm": 1.249830902106062,
+      "learning_rate": 5.12112290022401e-06,
+      "loss": 0.2663,
+      "step": 1663
+    },
+    {
+      "epoch": 0.5078589958797497,
+      "grad_norm": 1.2530061812921935,
+      "learning_rate": 5.11618001724727e-06,
+      "loss": 0.2481,
+      "step": 1664
+    },
+    {
+      "epoch": 0.5081641996032351,
+      "grad_norm": 1.344296410824843,
+      "learning_rate": 5.111237020665718e-06,
+      "loss": 0.225,
+      "step": 1665
+    },
+    {
+      "epoch": 0.5084694033267206,
+      "grad_norm": 1.423868312190065,
+      "learning_rate": 5.106293915312788e-06,
+      "loss": 0.2623,
+      "step": 1666
+    },
+    {
+      "epoch": 0.508774607050206,
+      "grad_norm": 1.3149812496139412,
+      "learning_rate": 5.10135070602202e-06,
+      "loss": 0.2228,
+      "step": 1667
+    },
+    {
+      "epoch": 0.5090798107736915,
+      "grad_norm": 1.4409444425833837,
+      "learning_rate": 5.096407397627051e-06,
+      "loss": 0.5072,
+      "step": 1668
+    },
+    {
+      "epoch": 0.5093850144971769,
+      "grad_norm": 1.348503222582244,
+      "learning_rate": 5.09146399496162e-06,
+      "loss": 0.388,
+      "step": 1669
+    },
+    {
+      "epoch": 0.5096902182206623,
+      "grad_norm": 1.2483916903268428,
+      "learning_rate": 5.086520502859556e-06,
+      "loss": 0.2071,
+      "step": 1670
+    },
+    {
+      "epoch": 0.5099954219441477,
+      "grad_norm": 1.3998356654583803,
+      "learning_rate": 5.081576926154772e-06,
+      "loss": 0.2387,
+      "step": 1671
+    },
+    {
+      "epoch": 0.5103006256676331,
+      "grad_norm": 1.5289407561496389,
+      "learning_rate": 5.07663326968127e-06,
+      "loss": 0.5692,
+      "step": 1672
+    },
+    {
+      "epoch": 0.5106058293911185,
+      "grad_norm": 1.234600239737849,
+      "learning_rate": 5.071689538273126e-06,
+      "loss": 0.2844,
+      "step": 1673
+    },
+    {
+      "epoch": 0.510911033114604,
+      "grad_norm": 1.2635731960722638,
+      "learning_rate": 5.0667457367644904e-06,
+      "loss": 0.3551,
+      "step": 1674
+    },
+    {
+      "epoch": 0.5112162368380895,
+      "grad_norm": 1.3538142063487089,
+      "learning_rate": 5.061801869989582e-06,
+      "loss": 0.3053,
+      "step": 1675
+    },
+    {
+      "epoch": 0.5115214405615749,
+      "grad_norm": 1.2172009821134098,
+      "learning_rate": 5.056857942782682e-06,
+      "loss": 0.2138,
+      "step": 1676
+    },
+    {
+      "epoch": 0.5118266442850603,
+      "grad_norm": 1.5879432509969746,
+      "learning_rate": 5.051913959978136e-06,
+      "loss": 0.4392,
+      "step": 1677
+    },
+    {
+      "epoch": 0.5121318480085457,
+      "grad_norm": 1.9377433870094682,
+      "learning_rate": 5.046969926410336e-06,
+      "loss": 0.3714,
+      "step": 1678
+    },
+    {
+      "epoch": 0.5124370517320311,
+      "grad_norm": 1.2846490850622998,
+      "learning_rate": 5.042025846913731e-06,
+      "loss": 0.3023,
+      "step": 1679
+    },
+    {
+      "epoch": 0.5127422554555165,
+      "grad_norm": 1.584798018284521,
+      "learning_rate": 5.03708172632281e-06,
+      "loss": 0.5201,
+      "step": 1680
+    },
+    {
+      "epoch": 0.513047459179002,
+      "grad_norm": 1.3353130822121289,
+      "learning_rate": 5.032137569472103e-06,
+      "loss": 0.2252,
+      "step": 1681
+    },
+    {
+      "epoch": 0.5133526629024874,
+      "grad_norm": 1.2506857898127242,
+      "learning_rate": 5.027193381196178e-06,
+      "loss": 0.3043,
+      "step": 1682
+    },
+    {
+      "epoch": 0.5136578666259728,
+      "grad_norm": 1.7478079009136256,
+      "learning_rate": 5.022249166329633e-06,
+      "loss": 0.2608,
+      "step": 1683
+    },
+    {
+      "epoch": 0.5139630703494583,
+      "grad_norm": 1.4253991019869026,
+      "learning_rate": 5.017304929707085e-06,
+      "loss": 0.3544,
+      "step": 1684
+    },
+    {
+      "epoch": 0.5142682740729437,
+      "grad_norm": 1.0280421678748557,
+      "learning_rate": 5.012360676163186e-06,
+      "loss": 0.1793,
+      "step": 1685
+    },
+    {
+      "epoch": 0.5145734777964291,
+      "grad_norm": 1.1098929257342118,
+      "learning_rate": 5.007416410532593e-06,
+      "loss": 0.1695,
+      "step": 1686
+    },
+    {
+      "epoch": 0.5148786815199146,
+      "grad_norm": 1.192133647057363,
+      "learning_rate": 5.0024721376499775e-06,
+      "loss": 0.2837,
+      "step": 1687
+    },
+    {
+      "epoch": 0.5151838852434,
+      "grad_norm": 1.5272365054360038,
+      "learning_rate": 4.997527862350023e-06,
+      "loss": 0.4632,
+      "step": 1688
+    },
+    {
+      "epoch": 0.5154890889668854,
+      "grad_norm": 1.7459738602751647,
+      "learning_rate": 4.992583589467409e-06,
+      "loss": 0.3587,
+      "step": 1689
+    },
+    {
+      "epoch": 0.5157942926903708,
+      "grad_norm": 1.59178534193483,
+      "learning_rate": 4.987639323836815e-06,
+      "loss": 0.4888,
+      "step": 1690
+    },
+    {
+      "epoch": 0.5160994964138562,
+      "grad_norm": 1.452831197590841,
+      "learning_rate": 4.982695070292916e-06,
+      "loss": 0.3795,
+      "step": 1691
+    },
+    {
+      "epoch": 0.5164047001373416,
+      "grad_norm": 1.4771465781176318,
+      "learning_rate": 4.9777508336703705e-06,
+      "loss": 0.3817,
+      "step": 1692
+    },
+    {
+      "epoch": 0.516709903860827,
+      "grad_norm": 1.6254273366231557,
+      "learning_rate": 4.972806618803823e-06,
+      "loss": 0.4285,
+      "step": 1693
+    },
+    {
+      "epoch": 0.5170151075843126,
+      "grad_norm": 1.4842668995896693,
+      "learning_rate": 4.967862430527899e-06,
+      "loss": 0.2605,
+      "step": 1694
+    },
+    {
+      "epoch": 0.517320311307798,
+      "grad_norm": 1.4032108102229814,
+      "learning_rate": 4.962918273677192e-06,
+      "loss": 0.3153,
+      "step": 1695
+    },
+    {
+      "epoch": 0.5176255150312834,
+      "grad_norm": 1.5402384491648182,
+      "learning_rate": 4.957974153086269e-06,
+      "loss": 0.4933,
+      "step": 1696
+    },
+    {
+      "epoch": 0.5179307187547688,
+      "grad_norm": 1.361335436939541,
+      "learning_rate": 4.953030073589663e-06,
+      "loss": 0.3291,
+      "step": 1697
+    },
+    {
+      "epoch": 0.5182359224782542,
+      "grad_norm": 1.6322832481817982,
+      "learning_rate": 4.948086040021865e-06,
+      "loss": 0.4752,
+      "step": 1698
+    },
+    {
+      "epoch": 0.5185411262017396,
+      "grad_norm": 1.4430096400012284,
+      "learning_rate": 4.943142057217318e-06,
+      "loss": 0.2248,
+      "step": 1699
+    },
+    {
+      "epoch": 0.5188463299252251,
+      "grad_norm": 1.4209700838358612,
+      "learning_rate": 4.938198130010419e-06,
+      "loss": 0.2256,
+      "step": 1700
+    },
+    {
+      "epoch": 0.5191515336487105,
+      "grad_norm": 1.3308099870220456,
+      "learning_rate": 4.933254263235511e-06,
+      "loss": 0.2792,
+      "step": 1701
+    },
+    {
+      "epoch": 0.5194567373721959,
+      "grad_norm": 1.435096555910569,
+      "learning_rate": 4.928310461726876e-06,
+      "loss": 0.2201,
+      "step": 1702
+    },
+    {
+      "epoch": 0.5197619410956814,
+      "grad_norm": 1.2795476536614092,
+      "learning_rate": 4.923366730318731e-06,
+      "loss": 0.3959,
+      "step": 1703
+    },
+    {
+      "epoch": 0.5200671448191668,
+      "grad_norm": 1.445568000273355,
+      "learning_rate": 4.91842307384523e-06,
+      "loss": 0.3239,
+      "step": 1704
+    },
+    {
+      "epoch": 0.5203723485426522,
+      "grad_norm": 1.3882780522624698,
+      "learning_rate": 4.913479497140447e-06,
+      "loss": 0.3565,
+      "step": 1705
+    },
+    {
+      "epoch": 0.5206775522661377,
+      "grad_norm": 1.4812796996153128,
+      "learning_rate": 4.9085360050383805e-06,
+      "loss": 0.3752,
+      "step": 1706
+    },
+    {
+      "epoch": 0.5209827559896231,
+      "grad_norm": 1.257425047155564,
+      "learning_rate": 4.90359260237295e-06,
+      "loss": 0.1718,
+      "step": 1707
+    },
+    {
+      "epoch": 0.5212879597131085,
+      "grad_norm": 1.6987322933124722,
+      "learning_rate": 4.898649293977982e-06,
+      "loss": 0.349,
+      "step": 1708
+    },
+    {
+      "epoch": 0.5215931634365939,
+      "grad_norm": 1.1895206726638172,
+      "learning_rate": 4.893706084687213e-06,
+      "loss": 0.2822,
+      "step": 1709
+    },
+    {
+      "epoch": 0.5218983671600793,
+      "grad_norm": 1.3543095537752436,
+      "learning_rate": 4.888762979334285e-06,
+      "loss": 0.2793,
+      "step": 1710
+    },
+    {
+      "epoch": 0.5222035708835647,
+      "grad_norm": 1.6001037653415573,
+      "learning_rate": 4.883819982752733e-06,
+      "loss": 0.5458,
+      "step": 1711
+    },
+    {
+      "epoch": 0.5225087746070503,
+      "grad_norm": 1.5693779437099424,
+      "learning_rate": 4.878877099775991e-06,
+      "loss": 0.3162,
+      "step": 1712
+    },
+    {
+      "epoch": 0.5228139783305357,
+      "grad_norm": 1.4074639908298856,
+      "learning_rate": 4.8739343352373765e-06,
+      "loss": 0.3206,
+      "step": 1713
+    },
+    {
+      "epoch": 0.5231191820540211,
+      "grad_norm": 1.087432984227466,
+      "learning_rate": 4.868991693970099e-06,
+      "loss": 0.2167,
+      "step": 1714
+    },
+    {
+      "epoch": 0.5234243857775065,
+      "grad_norm": 1.4493333303392464,
+      "learning_rate": 4.864049180807242e-06,
+      "loss": 0.3923,
+      "step": 1715
+    },
+    {
+      "epoch": 0.5237295895009919,
+      "grad_norm": 1.6976032864998114,
+      "learning_rate": 4.859106800581762e-06,
+      "loss": 0.3282,
+      "step": 1716
+    },
+    {
+      "epoch": 0.5240347932244773,
+      "grad_norm": 1.1794888158561472,
+      "learning_rate": 4.854164558126489e-06,
+      "loss": 0.2228,
+      "step": 1717
+    },
+    {
+      "epoch": 0.5243399969479627,
+      "grad_norm": 1.2969136060001154,
+      "learning_rate": 4.849222458274118e-06,
+      "loss": 0.2679,
+      "step": 1718
+    },
+    {
+      "epoch": 0.5246452006714482,
+      "grad_norm": 1.7402174549586045,
+      "learning_rate": 4.844280505857203e-06,
+      "loss": 0.2213,
+      "step": 1719
+    },
+    {
+      "epoch": 0.5249504043949336,
+      "grad_norm": 1.604727543792755,
+      "learning_rate": 4.839338705708156e-06,
+      "loss": 0.4238,
+      "step": 1720
+    },
+    {
+      "epoch": 0.5252556081184191,
+      "grad_norm": 1.2421408590521474,
+      "learning_rate": 4.834397062659236e-06,
+      "loss": 0.3278,
+      "step": 1721
+    },
+    {
+      "epoch": 0.5255608118419045,
+      "grad_norm": 1.724100143577579,
+      "learning_rate": 4.829455581542555e-06,
+      "loss": 0.3619,
+      "step": 1722
+    },
+    {
+      "epoch": 0.5258660155653899,
+      "grad_norm": 1.3189581973354423,
+      "learning_rate": 4.824514267190061e-06,
+      "loss": 0.2873,
+      "step": 1723
+    },
+    {
+      "epoch": 0.5261712192888753,
+      "grad_norm": 1.2491657334100055,
+      "learning_rate": 4.81957312443354e-06,
+      "loss": 0.2806,
+      "step": 1724
+    },
+    {
+      "epoch": 0.5264764230123608,
+      "grad_norm": 1.2884089807374433,
+      "learning_rate": 4.814632158104612e-06,
+      "loss": 0.2944,
+      "step": 1725
+    },
+    {
+      "epoch": 0.5267816267358462,
+      "grad_norm": 1.5336454688828358,
+      "learning_rate": 4.809691373034725e-06,
+      "loss": 0.3756,
+      "step": 1726
+    },
+    {
+      "epoch": 0.5270868304593316,
+      "grad_norm": 1.1703041738460964,
+      "learning_rate": 4.804750774055145e-06,
+      "loss": 0.3417,
+      "step": 1727
+    },
+    {
+      "epoch": 0.527392034182817,
+      "grad_norm": 1.4976332589595074,
+      "learning_rate": 4.799810365996961e-06,
+      "loss": 0.4417,
+      "step": 1728
+    },
+    {
+      "epoch": 0.5276972379063024,
+      "grad_norm": 1.65105783191874,
+      "learning_rate": 4.794870153691072e-06,
+      "loss": 0.2322,
+      "step": 1729
+    },
+    {
+      "epoch": 0.5280024416297879,
+      "grad_norm": 1.122193650919566,
+      "learning_rate": 4.789930141968191e-06,
+      "loss": 0.2484,
+      "step": 1730
+    },
+    {
+      "epoch": 0.5283076453532733,
+      "grad_norm": 1.3721395563321508,
+      "learning_rate": 4.784990335658829e-06,
+      "loss": 0.2416,
+      "step": 1731
+    },
+    {
+      "epoch": 0.5286128490767588,
+      "grad_norm": 2.1794325156426564,
+      "learning_rate": 4.780050739593298e-06,
+      "loss": 0.3152,
+      "step": 1732
+    },
+    {
+      "epoch": 0.5289180528002442,
+      "grad_norm": 1.4591445937013092,
+      "learning_rate": 4.7751113586017036e-06,
+      "loss": 0.3311,
+      "step": 1733
+    },
+    {
+      "epoch": 0.5292232565237296,
+      "grad_norm": 1.3029702586412188,
+      "learning_rate": 4.770172197513944e-06,
+      "loss": 0.2442,
+      "step": 1734
+    },
+    {
+      "epoch": 0.529528460247215,
+      "grad_norm": 1.4553221632341957,
+      "learning_rate": 4.7652332611596996e-06,
+      "loss": 0.2228,
+      "step": 1735
+    },
+    {
+      "epoch": 0.5298336639707004,
+      "grad_norm": 1.7556590813614814,
+      "learning_rate": 4.760294554368432e-06,
+      "loss": 0.4196,
+      "step": 1736
+    },
+    {
+      "epoch": 0.5301388676941858,
+      "grad_norm": 1.518317279472581,
+      "learning_rate": 4.755356081969378e-06,
+      "loss": 0.3037,
+      "step": 1737
+    },
+    {
+      "epoch": 0.5304440714176712,
+      "grad_norm": 1.6122834673847743,
+      "learning_rate": 4.750417848791545e-06,
+      "loss": 0.3145,
+      "step": 1738
+    },
+    {
+      "epoch": 0.5307492751411568,
+      "grad_norm": 1.4647223256965953,
+      "learning_rate": 4.7454798596637075e-06,
+      "loss": 0.4074,
+      "step": 1739
+    },
+    {
+      "epoch": 0.5310544788646422,
+      "grad_norm": 1.6148928519923968,
+      "learning_rate": 4.740542119414402e-06,
+      "loss": 0.3707,
+      "step": 1740
+    },
+    {
+      "epoch": 0.5313596825881276,
+      "grad_norm": 1.7828794020487633,
+      "learning_rate": 4.735604632871919e-06,
+      "loss": 0.3663,
+      "step": 1741
+    },
+    {
+      "epoch": 0.531664886311613,
+      "grad_norm": 1.7313254412247303,
+      "learning_rate": 4.730667404864301e-06,
+      "loss": 0.3853,
+      "step": 1742
+    },
+    {
+      "epoch": 0.5319700900350984,
+      "grad_norm": 1.6644547407620942,
+      "learning_rate": 4.725730440219342e-06,
+      "loss": 0.3523,
+      "step": 1743
+    },
+    {
+      "epoch": 0.5322752937585838,
+      "grad_norm": 1.491477837500527,
+      "learning_rate": 4.720793743764574e-06,
+      "loss": 0.293,
+      "step": 1744
+    },
+    {
+      "epoch": 0.5325804974820693,
+      "grad_norm": 1.649985215091933,
+      "learning_rate": 4.715857320327267e-06,
+      "loss": 0.3803,
+      "step": 1745
+    },
+    {
+      "epoch": 0.5328857012055547,
+      "grad_norm": 1.331924120938343,
+      "learning_rate": 4.7109211747344276e-06,
+      "loss": 0.2943,
+      "step": 1746
+    },
+    {
+      "epoch": 0.5331909049290401,
+      "grad_norm": 1.651195514943811,
+      "learning_rate": 4.705985311812788e-06,
+      "loss": 0.3402,
+      "step": 1747
+    },
+    {
+      "epoch": 0.5334961086525256,
+      "grad_norm": 1.858088007872116,
+      "learning_rate": 4.701049736388804e-06,
+      "loss": 0.757,
+      "step": 1748
+    },
+    {
+      "epoch": 0.533801312376011,
+      "grad_norm": 1.3341596049865094,
+      "learning_rate": 4.69611445328865e-06,
+      "loss": 0.3517,
+      "step": 1749
+    },
+    {
+      "epoch": 0.5341065160994964,
+      "grad_norm": 1.349613438115718,
+      "learning_rate": 4.691179467338217e-06,
+      "loss": 0.239,
+      "step": 1750
+    },
+    {
+      "epoch": 0.5344117198229819,
+      "grad_norm": 1.3311127211303257,
+      "learning_rate": 4.6862447833631015e-06,
+      "loss": 0.1743,
+      "step": 1751
+    },
+    {
+      "epoch": 0.5347169235464673,
+      "grad_norm": 1.6728820619848093,
+      "learning_rate": 4.681310406188607e-06,
+      "loss": 0.3636,
+      "step": 1752
+    },
+    {
+      "epoch": 0.5350221272699527,
+      "grad_norm": 1.340753007659583,
+      "learning_rate": 4.67637634063974e-06,
+      "loss": 0.3137,
+      "step": 1753
+    },
+    {
+      "epoch": 0.5353273309934381,
+      "grad_norm": 1.6789840289703037,
+      "learning_rate": 4.671442591541196e-06,
+      "loss": 0.6145,
+      "step": 1754
+    },
+    {
+      "epoch": 0.5356325347169235,
+      "grad_norm": 1.3813820892281865,
+      "learning_rate": 4.666509163717365e-06,
+      "loss": 0.3854,
+      "step": 1755
+    },
+    {
+      "epoch": 0.5359377384404089,
+      "grad_norm": 1.519369945447102,
+      "learning_rate": 4.6615760619923235e-06,
+      "loss": 0.4617,
+      "step": 1756
+    },
+    {
+      "epoch": 0.5362429421638945,
+      "grad_norm": 1.2407672848532267,
+      "learning_rate": 4.656643291189826e-06,
+      "loss": 0.2891,
+      "step": 1757
+    },
+    {
+      "epoch": 0.5365481458873799,
+      "grad_norm": 1.6907484083561595,
+      "learning_rate": 4.6517108561333055e-06,
+      "loss": 0.439,
+      "step": 1758
+    },
+    {
+      "epoch": 0.5368533496108653,
+      "grad_norm": 1.307296793447627,
+      "learning_rate": 4.646778761645867e-06,
+      "loss": 0.3114,
+      "step": 1759
+    },
+    {
+      "epoch": 0.5371585533343507,
+      "grad_norm": 1.4060761238262538,
+      "learning_rate": 4.641847012550281e-06,
+      "loss": 0.3102,
+      "step": 1760
+    },
+    {
+      "epoch": 0.5374637570578361,
+      "grad_norm": 1.1585526091871678,
+      "learning_rate": 4.63691561366898e-06,
+      "loss": 0.344,
+      "step": 1761
+    },
+    {
+      "epoch": 0.5377689607813215,
+      "grad_norm": 1.2551860517157176,
+      "learning_rate": 4.6319845698240564e-06,
+      "loss": 0.3574,
+      "step": 1762
+    },
+    {
+      "epoch": 0.5380741645048069,
+      "grad_norm": 1.057066149533723,
+      "learning_rate": 4.627053885837257e-06,
+      "loss": 0.2527,
+      "step": 1763
+    },
+    {
+      "epoch": 0.5383793682282924,
+      "grad_norm": 1.3899142559966882,
+      "learning_rate": 4.622123566529969e-06,
+      "loss": 0.4238,
+      "step": 1764
+    },
+    {
+      "epoch": 0.5386845719517778,
+      "grad_norm": 1.5136220222385257,
+      "learning_rate": 4.617193616723231e-06,
+      "loss": 0.4436,
+      "step": 1765
+    },
+    {
+      "epoch": 0.5389897756752632,
+      "grad_norm": 1.439164731912172,
+      "learning_rate": 4.612264041237717e-06,
+      "loss": 0.3895,
+      "step": 1766
+    },
+    {
+      "epoch": 0.5392949793987487,
+      "grad_norm": 1.2109877052974534,
+      "learning_rate": 4.607334844893732e-06,
+      "loss": 0.2577,
+      "step": 1767
+    },
+    {
+      "epoch": 0.5396001831222341,
+      "grad_norm": 1.6738329679225448,
+      "learning_rate": 4.602406032511217e-06,
+      "loss": 0.5504,
+      "step": 1768
+    },
+    {
+      "epoch": 0.5399053868457195,
+      "grad_norm": 1.3467379772470351,
+      "learning_rate": 4.597477608909733e-06,
+      "loss": 0.2681,
+      "step": 1769
+    },
+    {
+      "epoch": 0.540210590569205,
+      "grad_norm": 1.3966677245240011,
+      "learning_rate": 4.592549578908462e-06,
+      "loss": 0.3906,
+      "step": 1770
+    },
+    {
+      "epoch": 0.5405157942926904,
+      "grad_norm": 1.305199574012114,
+      "learning_rate": 4.587621947326198e-06,
+      "loss": 0.2883,
+      "step": 1771
+    },
+    {
+      "epoch": 0.5408209980161758,
+      "grad_norm": 1.028638133088597,
+      "learning_rate": 4.582694718981354e-06,
+      "loss": 0.1888,
+      "step": 1772
+    },
+    {
+      "epoch": 0.5411262017396612,
+      "grad_norm": 1.511774973846459,
+      "learning_rate": 4.577767898691937e-06,
+      "loss": 0.3098,
+      "step": 1773
+    },
+    {
+      "epoch": 0.5414314054631466,
+      "grad_norm": 1.6115715074241228,
+      "learning_rate": 4.572841491275565e-06,
+      "loss": 0.4813,
+      "step": 1774
+    },
+    {
+      "epoch": 0.541736609186632,
+      "grad_norm": 1.1577809482656338,
+      "learning_rate": 4.5679155015494475e-06,
+      "loss": 0.1947,
+      "step": 1775
+    },
+    {
+      "epoch": 0.5420418129101175,
+      "grad_norm": 1.4664633787039183,
+      "learning_rate": 4.562989934330385e-06,
+      "loss": 0.2036,
+      "step": 1776
+    },
+    {
+      "epoch": 0.542347016633603,
+      "grad_norm": 1.2041716543620122,
+      "learning_rate": 4.558064794434766e-06,
+      "loss": 0.2272,
+      "step": 1777
+    },
+    {
+      "epoch": 0.5426522203570884,
+      "grad_norm": 1.3641306682040004,
+      "learning_rate": 4.553140086678563e-06,
+      "loss": 0.3755,
+      "step": 1778
+    },
+    {
+      "epoch": 0.5429574240805738,
+      "grad_norm": 1.3739568047542712,
+      "learning_rate": 4.548215815877324e-06,
+      "loss": 0.2787,
+      "step": 1779
+    },
+    {
+      "epoch": 0.5432626278040592,
+      "grad_norm": 2.6505197986898854,
+      "learning_rate": 4.543291986846172e-06,
+      "loss": 0.4521,
+      "step": 1780
+    },
+    {
+      "epoch": 0.5435678315275446,
+      "grad_norm": 1.4710712352129807,
+      "learning_rate": 4.5383686043997916e-06,
+      "loss": 0.3916,
+      "step": 1781
+    },
+    {
+      "epoch": 0.54387303525103,
+      "grad_norm": 1.1925169723230629,
+      "learning_rate": 4.533445673352438e-06,
+      "loss": 0.4002,
+      "step": 1782
+    },
+    {
+      "epoch": 0.5441782389745154,
+      "grad_norm": 1.251342815115979,
+      "learning_rate": 4.528523198517922e-06,
+      "loss": 0.2429,
+      "step": 1783
+    },
+    {
+      "epoch": 0.5444834426980009,
+      "grad_norm": 1.5975295485269585,
+      "learning_rate": 4.523601184709608e-06,
+      "loss": 0.2684,
+      "step": 1784
+    },
+    {
+      "epoch": 0.5447886464214864,
+      "grad_norm": 1.6643586947322488,
+      "learning_rate": 4.518679636740411e-06,
+      "loss": 0.3698,
+      "step": 1785
+    },
+    {
+      "epoch": 0.5450938501449718,
+      "grad_norm": 1.3920856048936912,
+      "learning_rate": 4.513758559422788e-06,
+      "loss": 0.2586,
+      "step": 1786
+    },
+    {
+      "epoch": 0.5453990538684572,
+      "grad_norm": 1.7161184105155805,
+      "learning_rate": 4.508837957568737e-06,
+      "loss": 0.4346,
+      "step": 1787
+    },
+    {
+      "epoch": 0.5457042575919426,
+      "grad_norm": 1.9392455451877773,
+      "learning_rate": 4.503917835989795e-06,
+      "loss": 0.3078,
+      "step": 1788
+    },
+    {
+      "epoch": 0.546009461315428,
+      "grad_norm": 1.5955708891333396,
+      "learning_rate": 4.4989981994970215e-06,
+      "loss": 0.3537,
+      "step": 1789
+    },
+    {
+      "epoch": 0.5463146650389135,
+      "grad_norm": 1.6369794760823198,
+      "learning_rate": 4.494079052901008e-06,
+      "loss": 0.3795,
+      "step": 1790
+    },
+    {
+      "epoch": 0.5466198687623989,
+      "grad_norm": 1.3368239111142433,
+      "learning_rate": 4.489160401011865e-06,
+      "loss": 0.3277,
+      "step": 1791
+    },
+    {
+      "epoch": 0.5469250724858843,
+      "grad_norm": 1.53161274252946,
+      "learning_rate": 4.484242248639219e-06,
+      "loss": 0.3846,
+      "step": 1792
+    },
+    {
+      "epoch": 0.5472302762093697,
+      "grad_norm": 1.847090459090323,
+      "learning_rate": 4.479324600592207e-06,
+      "loss": 0.5765,
+      "step": 1793
+    },
+    {
+      "epoch": 0.5475354799328552,
+      "grad_norm": 1.2508761196625122,
+      "learning_rate": 4.474407461679474e-06,
+      "loss": 0.262,
+      "step": 1794
+    },
+    {
+      "epoch": 0.5478406836563406,
+      "grad_norm": 1.4004402524276096,
+      "learning_rate": 4.46949083670917e-06,
+      "loss": 0.2882,
+      "step": 1795
+    },
+    {
+      "epoch": 0.5481458873798261,
+      "grad_norm": 1.5857922717718818,
+      "learning_rate": 4.46457473048894e-06,
+      "loss": 0.239,
+      "step": 1796
+    },
+    {
+      "epoch": 0.5484510911033115,
+      "grad_norm": 1.6062020439380993,
+      "learning_rate": 4.459659147825918e-06,
+      "loss": 0.3717,
+      "step": 1797
+    },
+    {
+      "epoch": 0.5487562948267969,
+      "grad_norm": 1.353016506717472,
+      "learning_rate": 4.45474409352673e-06,
+      "loss": 0.3397,
+      "step": 1798
+    },
+    {
+      "epoch": 0.5490614985502823,
+      "grad_norm": 1.8447201003228528,
+      "learning_rate": 4.4498295723974874e-06,
+      "loss": 0.4021,
+      "step": 1799
+    },
+    {
+      "epoch": 0.5493667022737677,
+      "grad_norm": 1.5434900103493834,
+      "learning_rate": 4.444915589243775e-06,
+      "loss": 0.4346,
+      "step": 1800
+    },
+    {
+      "epoch": 0.5496719059972531,
+      "grad_norm": 1.4692641636605677,
+      "learning_rate": 4.4400021488706554e-06,
+      "loss": 0.4139,
+      "step": 1801
+    },
+    {
+      "epoch": 0.5499771097207385,
+      "grad_norm": 1.7164322223877362,
+      "learning_rate": 4.43508925608266e-06,
+      "loss": 0.2815,
+      "step": 1802
+    },
+    {
+      "epoch": 0.5502823134442241,
+      "grad_norm": 1.5813457746488064,
+      "learning_rate": 4.430176915683782e-06,
+      "loss": 0.296,
+      "step": 1803
+    },
+    {
+      "epoch": 0.5505875171677095,
+      "grad_norm": 1.6394357639908674,
+      "learning_rate": 4.425265132477478e-06,
+      "loss": 0.5515,
+      "step": 1804
+    },
+    {
+      "epoch": 0.5508927208911949,
+      "grad_norm": 2.6314268002948626,
+      "learning_rate": 4.420353911266659e-06,
+      "loss": 0.3059,
+      "step": 1805
+    },
+    {
+      "epoch": 0.5511979246146803,
+      "grad_norm": 1.3058880975401832,
+      "learning_rate": 4.415443256853683e-06,
+      "loss": 0.3403,
+      "step": 1806
+    },
+    {
+      "epoch": 0.5515031283381657,
+      "grad_norm": 1.377286483777208,
+      "learning_rate": 4.410533174040359e-06,
+      "loss": 0.3375,
+      "step": 1807
+    },
+    {
+      "epoch": 0.5518083320616511,
+      "grad_norm": 0.8967185784549286,
+      "learning_rate": 4.4056236676279345e-06,
+      "loss": 0.1389,
+      "step": 1808
+    },
+    {
+      "epoch": 0.5521135357851366,
+      "grad_norm": 1.3009551794026946,
+      "learning_rate": 4.400714742417091e-06,
+      "loss": 0.3164,
+      "step": 1809
+    },
+    {
+      "epoch": 0.552418739508622,
+      "grad_norm": 1.0558516117873247,
+      "learning_rate": 4.395806403207947e-06,
+      "loss": 0.1837,
+      "step": 1810
+    },
+    {
+      "epoch": 0.5527239432321074,
+      "grad_norm": 1.5935819948692123,
+      "learning_rate": 4.3908986548000406e-06,
+      "loss": 0.4112,
+      "step": 1811
+    },
+    {
+      "epoch": 0.5530291469555929,
+      "grad_norm": 1.633330413595172,
+      "learning_rate": 4.3859915019923436e-06,
+      "loss": 0.3029,
+      "step": 1812
+    },
+    {
+      "epoch": 0.5533343506790783,
+      "grad_norm": 1.5346955440421504,
+      "learning_rate": 4.381084949583233e-06,
+      "loss": 0.3364,
+      "step": 1813
+    },
+    {
+      "epoch": 0.5536395544025637,
+      "grad_norm": 1.346352430252736,
+      "learning_rate": 4.376179002370505e-06,
+      "loss": 0.217,
+      "step": 1814
+    },
+    {
+      "epoch": 0.5539447581260492,
+      "grad_norm": 1.262614354535023,
+      "learning_rate": 4.371273665151366e-06,
+      "loss": 0.2082,
+      "step": 1815
+    },
+    {
+      "epoch": 0.5542499618495346,
+      "grad_norm": 1.5174607632889023,
+      "learning_rate": 4.36636894272242e-06,
+      "loss": 0.4906,
+      "step": 1816
+    },
+    {
+      "epoch": 0.55455516557302,
+      "grad_norm": 1.3534777967663216,
+      "learning_rate": 4.361464839879674e-06,
+      "loss": 0.2799,
+      "step": 1817
+    },
+    {
+      "epoch": 0.5548603692965054,
+      "grad_norm": 1.4373919404920255,
+      "learning_rate": 4.356561361418529e-06,
+      "loss": 0.2769,
+      "step": 1818
+    },
+    {
+      "epoch": 0.5551655730199908,
+      "grad_norm": 1.6256631818406408,
+      "learning_rate": 4.351658512133773e-06,
+      "loss": 0.3239,
+      "step": 1819
+    },
+    {
+      "epoch": 0.5554707767434762,
+      "grad_norm": 1.5206222736100585,
+      "learning_rate": 4.346756296819582e-06,
+      "loss": 0.3495,
+      "step": 1820
+    },
+    {
+      "epoch": 0.5557759804669617,
+      "grad_norm": 1.6079779922535562,
+      "learning_rate": 4.34185472026951e-06,
+      "loss": 0.404,
+      "step": 1821
+    },
+    {
+      "epoch": 0.5560811841904472,
+      "grad_norm": 1.3156486617835121,
+      "learning_rate": 4.3369537872764854e-06,
+      "loss": 0.2266,
+      "step": 1822
+    },
+    {
+      "epoch": 0.5563863879139326,
+      "grad_norm": 1.4051731649097803,
+      "learning_rate": 4.3320535026328095e-06,
+      "loss": 0.377,
+      "step": 1823
+    },
+    {
+      "epoch": 0.556691591637418,
+      "grad_norm": 1.5976076748279182,
+      "learning_rate": 4.327153871130151e-06,
+      "loss": 0.4271,
+      "step": 1824
+    },
+    {
+      "epoch": 0.5569967953609034,
+      "grad_norm": 1.5932764584008285,
+      "learning_rate": 4.322254897559534e-06,
+      "loss": 0.272,
+      "step": 1825
+    },
+    {
+      "epoch": 0.5573019990843888,
+      "grad_norm": 1.2476323593040315,
+      "learning_rate": 4.317356586711345e-06,
+      "loss": 0.3005,
+      "step": 1826
+    },
+    {
+      "epoch": 0.5576072028078742,
+      "grad_norm": 1.4480223948712203,
+      "learning_rate": 4.312458943375319e-06,
+      "loss": 0.5113,
+      "step": 1827
+    },
+    {
+      "epoch": 0.5579124065313597,
+      "grad_norm": 1.736339659164446,
+      "learning_rate": 4.307561972340545e-06,
+      "loss": 0.3381,
+      "step": 1828
+    },
+    {
+      "epoch": 0.5582176102548451,
+      "grad_norm": 1.5027439927263628,
+      "learning_rate": 4.302665678395444e-06,
+      "loss": 0.2734,
+      "step": 1829
+    },
+    {
+      "epoch": 0.5585228139783306,
+      "grad_norm": 1.7299477371965306,
+      "learning_rate": 4.297770066327781e-06,
+      "loss": 0.3715,
+      "step": 1830
+    },
+    {
+      "epoch": 0.558828017701816,
+      "grad_norm": 1.5375612944503692,
+      "learning_rate": 4.292875140924656e-06,
+      "loss": 0.2312,
+      "step": 1831
+    },
+    {
+      "epoch": 0.5591332214253014,
+      "grad_norm": 1.6590692529416964,
+      "learning_rate": 4.287980906972492e-06,
+      "loss": 0.5114,
+      "step": 1832
+    },
+    {
+      "epoch": 0.5594384251487868,
+      "grad_norm": 1.6149299823711942,
+      "learning_rate": 4.2830873692570406e-06,
+      "loss": 0.5849,
+      "step": 1833
+    },
+    {
+      "epoch": 0.5597436288722722,
+      "grad_norm": 1.2257140803122564,
+      "learning_rate": 4.278194532563373e-06,
+      "loss": 0.2985,
+      "step": 1834
+    },
+    {
+      "epoch": 0.5600488325957577,
+      "grad_norm": 1.4168339331371003,
+      "learning_rate": 4.273302401675869e-06,
+      "loss": 0.3212,
+      "step": 1835
+    },
+    {
+      "epoch": 0.5603540363192431,
+      "grad_norm": 1.24088703914615,
+      "learning_rate": 4.2684109813782235e-06,
+      "loss": 0.2954,
+      "step": 1836
+    },
+    {
+      "epoch": 0.5606592400427285,
+      "grad_norm": 1.2898878287553455,
+      "learning_rate": 4.263520276453438e-06,
+      "loss": 0.3189,
+      "step": 1837
+    },
+    {
+      "epoch": 0.5609644437662139,
+      "grad_norm": 1.2371405994749598,
+      "learning_rate": 4.258630291683807e-06,
+      "loss": 0.2873,
+      "step": 1838
+    },
+    {
+      "epoch": 0.5612696474896993,
+      "grad_norm": 1.6176357661582368,
+      "learning_rate": 4.253741031850928e-06,
+      "loss": 0.3103,
+      "step": 1839
+    },
+    {
+      "epoch": 0.5615748512131848,
+      "grad_norm": 1.7539077452927228,
+      "learning_rate": 4.248852501735688e-06,
+      "loss": 0.3436,
+      "step": 1840
+    },
+    {
+      "epoch": 0.5618800549366703,
+      "grad_norm": 1.464784829544187,
+      "learning_rate": 4.2439647061182585e-06,
+      "loss": 0.3883,
+      "step": 1841
+    },
+    {
+      "epoch": 0.5621852586601557,
+      "grad_norm": 1.0819455329738283,
+      "learning_rate": 4.2390776497780924e-06,
+      "loss": 0.2173,
+      "step": 1842
+    },
+    {
+      "epoch": 0.5624904623836411,
+      "grad_norm": 1.5239350900118207,
+      "learning_rate": 4.234191337493924e-06,
+      "loss": 0.2023,
+      "step": 1843
+    },
+    {
+      "epoch": 0.5627956661071265,
+      "grad_norm": 1.78640496384631,
+      "learning_rate": 4.229305774043753e-06,
+      "loss": 0.4053,
+      "step": 1844
+    },
+    {
+      "epoch": 0.5631008698306119,
+      "grad_norm": 1.7446265147225073,
+      "learning_rate": 4.224420964204856e-06,
+      "loss": 0.3202,
+      "step": 1845
+    },
+    {
+      "epoch": 0.5634060735540973,
+      "grad_norm": 1.409118270328708,
+      "learning_rate": 4.219536912753765e-06,
+      "loss": 0.3765,
+      "step": 1846
+    },
+    {
+      "epoch": 0.5637112772775827,
+      "grad_norm": 1.8518819412683283,
+      "learning_rate": 4.2146536244662755e-06,
+      "loss": 0.3897,
+      "step": 1847
+    },
+    {
+      "epoch": 0.5640164810010682,
+      "grad_norm": 1.418487033756496,
+      "learning_rate": 4.209771104117431e-06,
+      "loss": 0.2562,
+      "step": 1848
+    },
+    {
+      "epoch": 0.5643216847245537,
+      "grad_norm": 2.3571698323477404,
+      "learning_rate": 4.204889356481529e-06,
+      "loss": 0.3302,
+      "step": 1849
+    },
+    {
+      "epoch": 0.5646268884480391,
+      "grad_norm": 1.6473829322170688,
+      "learning_rate": 4.200008386332112e-06,
+      "loss": 0.4377,
+      "step": 1850
+    },
+    {
+      "epoch": 0.5649320921715245,
+      "grad_norm": 1.6536132408794952,
+      "learning_rate": 4.195128198441956e-06,
+      "loss": 0.3028,
+      "step": 1851
+    },
+    {
+      "epoch": 0.5652372958950099,
+      "grad_norm": 1.655984317415034,
+      "learning_rate": 4.19024879758308e-06,
+      "loss": 0.4111,
+      "step": 1852
+    },
+    {
+      "epoch": 0.5655424996184953,
+      "grad_norm": 1.3841085018145172,
+      "learning_rate": 4.185370188526729e-06,
+      "loss": 0.3595,
+      "step": 1853
+    },
+    {
+      "epoch": 0.5658477033419808,
+      "grad_norm": 1.5438987536712128,
+      "learning_rate": 4.180492376043371e-06,
+      "loss": 0.3239,
+      "step": 1854
+    },
+    {
+      "epoch": 0.5661529070654662,
+      "grad_norm": 1.4582024742633912,
+      "learning_rate": 4.175615364902702e-06,
+      "loss": 0.2405,
+      "step": 1855
+    },
+    {
+      "epoch": 0.5664581107889516,
+      "grad_norm": 1.563635680413248,
+      "learning_rate": 4.1707391598736315e-06,
+      "loss": 0.3507,
+      "step": 1856
+    },
+    {
+      "epoch": 0.566763314512437,
+      "grad_norm": 1.415615851433773,
+      "learning_rate": 4.1658637657242775e-06,
+      "loss": 0.3625,
+      "step": 1857
+    },
+    {
+      "epoch": 0.5670685182359225,
+      "grad_norm": 1.3449463065679088,
+      "learning_rate": 4.160989187221969e-06,
+      "loss": 0.3278,
+      "step": 1858
+    },
+    {
+      "epoch": 0.5673737219594079,
+      "grad_norm": 1.2270830349794986,
+      "learning_rate": 4.156115429133237e-06,
+      "loss": 0.2514,
+      "step": 1859
+    },
+    {
+      "epoch": 0.5676789256828934,
+      "grad_norm": 1.4667141379274422,
+      "learning_rate": 4.151242496223807e-06,
+      "loss": 0.318,
+      "step": 1860
+    },
+    {
+      "epoch": 0.5679841294063788,
+      "grad_norm": 1.6246601629699675,
+      "learning_rate": 4.146370393258603e-06,
+      "loss": 0.5131,
+      "step": 1861
+    },
+    {
+      "epoch": 0.5682893331298642,
+      "grad_norm": 1.614142901328195,
+      "learning_rate": 4.141499125001733e-06,
+      "loss": 0.5237,
+      "step": 1862
+    },
+    {
+      "epoch": 0.5685945368533496,
+      "grad_norm": 1.4567239964083918,
+      "learning_rate": 4.13662869621649e-06,
+      "loss": 0.2776,
+      "step": 1863
+    },
+    {
+      "epoch": 0.568899740576835,
+      "grad_norm": 1.2668036165155088,
+      "learning_rate": 4.131759111665349e-06,
+      "loss": 0.2393,
+      "step": 1864
+    },
+    {
+      "epoch": 0.5692049443003204,
+      "grad_norm": 1.8092268624549248,
+      "learning_rate": 4.126890376109953e-06,
+      "loss": 0.317,
+      "step": 1865
+    },
+    {
+      "epoch": 0.5695101480238058,
+      "grad_norm": 1.7242479647679296,
+      "learning_rate": 4.122022494311122e-06,
+      "loss": 0.3034,
+      "step": 1866
+    },
+    {
+      "epoch": 0.5698153517472914,
+      "grad_norm": 1.1713728273815491,
+      "learning_rate": 4.117155471028837e-06,
+      "loss": 0.1709,
+      "step": 1867
+    },
+    {
+      "epoch": 0.5701205554707768,
+      "grad_norm": 1.3046020205700666,
+      "learning_rate": 4.112289311022239e-06,
+      "loss": 0.2232,
+      "step": 1868
+    },
+    {
+      "epoch": 0.5704257591942622,
+      "grad_norm": 1.4203182543479589,
+      "learning_rate": 4.107424019049628e-06,
+      "loss": 0.3207,
+      "step": 1869
+    },
+    {
+      "epoch": 0.5707309629177476,
+      "grad_norm": 1.469352436289963,
+      "learning_rate": 4.102559599868454e-06,
+      "loss": 0.4105,
+      "step": 1870
+    },
+    {
+      "epoch": 0.571036166641233,
+      "grad_norm": 1.32659963638931,
+      "learning_rate": 4.097696058235311e-06,
+      "loss": 0.3715,
+      "step": 1871
+    },
+    {
+      "epoch": 0.5713413703647184,
+      "grad_norm": 1.351981831146365,
+      "learning_rate": 4.092833398905939e-06,
+      "loss": 0.413,
+      "step": 1872
+    },
+    {
+      "epoch": 0.5716465740882039,
+      "grad_norm": 1.1676271889652488,
+      "learning_rate": 4.087971626635212e-06,
+      "loss": 0.2227,
+      "step": 1873
+    },
+    {
+      "epoch": 0.5719517778116893,
+      "grad_norm": 1.395861378667927,
+      "learning_rate": 4.083110746177137e-06,
+      "loss": 0.3457,
+      "step": 1874
+    },
+    {
+      "epoch": 0.5722569815351747,
+      "grad_norm": 1.7052386406046318,
+      "learning_rate": 4.078250762284852e-06,
+      "loss": 0.3004,
+      "step": 1875
+    },
+    {
+      "epoch": 0.5725621852586602,
+      "grad_norm": 1.5593654471737393,
+      "learning_rate": 4.073391679710613e-06,
+      "loss": 0.1962,
+      "step": 1876
+    },
+    {
+      "epoch": 0.5728673889821456,
+      "grad_norm": 1.453389215546987,
+      "learning_rate": 4.068533503205801e-06,
+      "loss": 0.2628,
+      "step": 1877
+    },
+    {
+      "epoch": 0.573172592705631,
+      "grad_norm": 1.4620964634717861,
+      "learning_rate": 4.063676237520907e-06,
+      "loss": 0.3129,
+      "step": 1878
+    },
+    {
+      "epoch": 0.5734777964291164,
+      "grad_norm": 1.4098166270766606,
+      "learning_rate": 4.058819887405532e-06,
+      "loss": 0.2979,
+      "step": 1879
+    },
+    {
+      "epoch": 0.5737830001526019,
+      "grad_norm": 1.405420779416245,
+      "learning_rate": 4.053964457608381e-06,
+      "loss": 0.3358,
+      "step": 1880
+    },
+    {
+      "epoch": 0.5740882038760873,
+      "grad_norm": 1.3379669835930859,
+      "learning_rate": 4.049109952877261e-06,
+      "loss": 0.2897,
+      "step": 1881
+    },
+    {
+      "epoch": 0.5743934075995727,
+      "grad_norm": 1.527556109482659,
+      "learning_rate": 4.044256377959072e-06,
+      "loss": 0.4784,
+      "step": 1882
+    },
+    {
+      "epoch": 0.5746986113230581,
+      "grad_norm": 1.3768514825628073,
+      "learning_rate": 4.039403737599807e-06,
+      "loss": 0.3001,
+      "step": 1883
+    },
+    {
+      "epoch": 0.5750038150465435,
+      "grad_norm": 1.2508930830606682,
+      "learning_rate": 4.034552036544544e-06,
+      "loss": 0.3061,
+      "step": 1884
+    },
+    {
+      "epoch": 0.575309018770029,
+      "grad_norm": 1.1647278593095605,
+      "learning_rate": 4.029701279537444e-06,
+      "loss": 0.1376,
+      "step": 1885
+    },
+    {
+      "epoch": 0.5756142224935145,
+      "grad_norm": 1.2338571548694106,
+      "learning_rate": 4.024851471321744e-06,
+      "loss": 0.2517,
+      "step": 1886
+    },
+    {
+      "epoch": 0.5759194262169999,
+      "grad_norm": 1.5621581657327868,
+      "learning_rate": 4.02000261663975e-06,
+      "loss": 0.3779,
+      "step": 1887
+    },
+    {
+      "epoch": 0.5762246299404853,
+      "grad_norm": 1.496075264042124,
+      "learning_rate": 4.01515472023284e-06,
+      "loss": 0.22,
+      "step": 1888
+    },
+    {
+      "epoch": 0.5765298336639707,
+      "grad_norm": 1.9209284738850185,
+      "learning_rate": 4.010307786841456e-06,
+      "loss": 0.2617,
+      "step": 1889
+    },
+    {
+      "epoch": 0.5768350373874561,
+      "grad_norm": 1.4182439699870193,
+      "learning_rate": 4.005461821205092e-06,
+      "loss": 0.2029,
+      "step": 1890
+    },
+    {
+      "epoch": 0.5771402411109415,
+      "grad_norm": 1.632558848560275,
+      "learning_rate": 4.000616828062302e-06,
+      "loss": 0.339,
+      "step": 1891
+    },
+    {
+      "epoch": 0.577445444834427,
+      "grad_norm": 1.5939429671208698,
+      "learning_rate": 3.995772812150687e-06,
+      "loss": 0.343,
+      "step": 1892
+    },
+    {
+      "epoch": 0.5777506485579124,
+      "grad_norm": 1.86362055921937,
+      "learning_rate": 3.990929778206889e-06,
+      "loss": 0.3299,
+      "step": 1893
+    },
+    {
+      "epoch": 0.5780558522813979,
+      "grad_norm": 1.669041308861478,
+      "learning_rate": 3.986087730966596e-06,
+      "loss": 0.4438,
+      "step": 1894
+    },
+    {
+      "epoch": 0.5783610560048833,
+      "grad_norm": 1.9885482758668145,
+      "learning_rate": 3.981246675164526e-06,
+      "loss": 0.3986,
+      "step": 1895
+    },
+    {
+      "epoch": 0.5786662597283687,
+      "grad_norm": 1.4523473319942632,
+      "learning_rate": 3.976406615534433e-06,
+      "loss": 0.2753,
+      "step": 1896
+    },
+    {
+      "epoch": 0.5789714634518541,
+      "grad_norm": 1.7124036664389317,
+      "learning_rate": 3.971567556809089e-06,
+      "loss": 0.2557,
+      "step": 1897
+    },
+    {
+      "epoch": 0.5792766671753395,
+      "grad_norm": 1.3490546890467419,
+      "learning_rate": 3.966729503720294e-06,
+      "loss": 0.3153,
+      "step": 1898
+    },
+    {
+      "epoch": 0.579581870898825,
+      "grad_norm": 1.711569891289086,
+      "learning_rate": 3.961892460998863e-06,
+      "loss": 0.4145,
+      "step": 1899
+    },
+    {
+      "epoch": 0.5798870746223104,
+      "grad_norm": 1.6758973408215243,
+      "learning_rate": 3.95705643337462e-06,
+      "loss": 0.3659,
+      "step": 1900
+    },
+    {
+      "epoch": 0.5801922783457958,
+      "grad_norm": 1.5917092515664464,
+      "learning_rate": 3.952221425576401e-06,
+      "loss": 0.4488,
+      "step": 1901
+    },
+    {
+      "epoch": 0.5804974820692812,
+      "grad_norm": 1.4033121574658662,
+      "learning_rate": 3.947387442332043e-06,
+      "loss": 0.2452,
+      "step": 1902
+    },
+    {
+      "epoch": 0.5808026857927667,
+      "grad_norm": 1.473499329969696,
+      "learning_rate": 3.942554488368378e-06,
+      "loss": 0.3778,
+      "step": 1903
+    },
+    {
+      "epoch": 0.5811078895162521,
+      "grad_norm": 1.3946820345871305,
+      "learning_rate": 3.937722568411237e-06,
+      "loss": 0.238,
+      "step": 1904
+    },
+    {
+      "epoch": 0.5814130932397376,
+      "grad_norm": 1.735278332506307,
+      "learning_rate": 3.932891687185438e-06,
+      "loss": 0.3191,
+      "step": 1905
+    },
+    {
+      "epoch": 0.581718296963223,
+      "grad_norm": 1.728510245621926,
+      "learning_rate": 3.928061849414778e-06,
+      "loss": 0.4613,
+      "step": 1906
+    },
+    {
+      "epoch": 0.5820235006867084,
+      "grad_norm": 1.7322903038045538,
+      "learning_rate": 3.923233059822041e-06,
+      "loss": 0.4713,
+      "step": 1907
+    },
+    {
+      "epoch": 0.5823287044101938,
+      "grad_norm": 1.6221554140736476,
+      "learning_rate": 3.918405323128986e-06,
+      "loss": 0.3462,
+      "step": 1908
+    },
+    {
+      "epoch": 0.5826339081336792,
+      "grad_norm": 1.4204907228891313,
+      "learning_rate": 3.913578644056332e-06,
+      "loss": 0.2559,
+      "step": 1909
+    },
+    {
+      "epoch": 0.5829391118571646,
+      "grad_norm": 1.4410886337055444,
+      "learning_rate": 3.90875302732378e-06,
+      "loss": 0.3235,
+      "step": 1910
+    },
+    {
+      "epoch": 0.58324431558065,
+      "grad_norm": 1.3315830763234417,
+      "learning_rate": 3.903928477649978e-06,
+      "loss": 0.2323,
+      "step": 1911
+    },
+    {
+      "epoch": 0.5835495193041355,
+      "grad_norm": 1.4552983264589225,
+      "learning_rate": 3.899104999752541e-06,
+      "loss": 0.2523,
+      "step": 1912
+    },
+    {
+      "epoch": 0.583854723027621,
+      "grad_norm": 1.4988653341908547,
+      "learning_rate": 3.894282598348026e-06,
+      "loss": 0.3828,
+      "step": 1913
+    },
+    {
+      "epoch": 0.5841599267511064,
+      "grad_norm": 1.894276143625022,
+      "learning_rate": 3.889461278151946e-06,
+      "loss": 0.2661,
+      "step": 1914
+    },
+    {
+      "epoch": 0.5844651304745918,
+      "grad_norm": 1.2550040219915009,
+      "learning_rate": 3.884641043878754e-06,
+      "loss": 0.2631,
+      "step": 1915
+    },
+    {
+      "epoch": 0.5847703341980772,
+      "grad_norm": 1.5480246847190493,
+      "learning_rate": 3.879821900241838e-06,
+      "loss": 0.556,
+      "step": 1916
+    },
+    {
+      "epoch": 0.5850755379215626,
+      "grad_norm": 1.4156532402674349,
+      "learning_rate": 3.8750038519535235e-06,
+      "loss": 0.3244,
+      "step": 1917
+    },
+    {
+      "epoch": 0.585380741645048,
+      "grad_norm": 1.5455677404105437,
+      "learning_rate": 3.870186903725064e-06,
+      "loss": 0.5752,
+      "step": 1918
+    },
+    {
+      "epoch": 0.5856859453685335,
+      "grad_norm": 1.4636896495284506,
+      "learning_rate": 3.865371060266635e-06,
+      "loss": 0.3238,
+      "step": 1919
+    },
+    {
+      "epoch": 0.5859911490920189,
+      "grad_norm": 1.4740107047366455,
+      "learning_rate": 3.860556326287335e-06,
+      "loss": 0.3049,
+      "step": 1920
+    },
+    {
+      "epoch": 0.5862963528155043,
+      "grad_norm": 1.6439367384050791,
+      "learning_rate": 3.855742706495177e-06,
+      "loss": 0.4124,
+      "step": 1921
+    },
+    {
+      "epoch": 0.5866015565389898,
+      "grad_norm": 1.2368282132313706,
+      "learning_rate": 3.850930205597081e-06,
+      "loss": 0.3479,
+      "step": 1922
+    },
+    {
+      "epoch": 0.5869067602624752,
+      "grad_norm": 1.3429435261592815,
+      "learning_rate": 3.846118828298878e-06,
+      "loss": 0.2947,
+      "step": 1923
+    },
+    {
+      "epoch": 0.5872119639859606,
+      "grad_norm": 1.6180408087819254,
+      "learning_rate": 3.841308579305298e-06,
+      "loss": 0.4857,
+      "step": 1924
+    },
+    {
+      "epoch": 0.5875171677094461,
+      "grad_norm": 1.2893210411892835,
+      "learning_rate": 3.836499463319962e-06,
+      "loss": 0.4012,
+      "step": 1925
+    },
+    {
+      "epoch": 0.5878223714329315,
+      "grad_norm": 1.4427577353418675,
+      "learning_rate": 3.831691485045397e-06,
+      "loss": 0.5265,
+      "step": 1926
+    },
+    {
+      "epoch": 0.5881275751564169,
+      "grad_norm": 1.241609594881084,
+      "learning_rate": 3.826884649183006e-06,
+      "loss": 0.2704,
+      "step": 1927
+    },
+    {
+      "epoch": 0.5884327788799023,
+      "grad_norm": 1.424738756292409,
+      "learning_rate": 3.822078960433076e-06,
+      "loss": 0.3,
+      "step": 1928
+    },
+    {
+      "epoch": 0.5887379826033877,
+      "grad_norm": 1.7578843843808665,
+      "learning_rate": 3.817274423494778e-06,
+      "loss": 0.3955,
+      "step": 1929
+    },
+    {
+      "epoch": 0.5890431863268731,
+      "grad_norm": 1.5726966788282515,
+      "learning_rate": 3.8124710430661503e-06,
+      "loss": 0.2467,
+      "step": 1930
+    },
+    {
+      "epoch": 0.5893483900503587,
+      "grad_norm": 1.1631106382833136,
+      "learning_rate": 3.807668823844104e-06,
+      "loss": 0.3261,
+      "step": 1931
+    },
+    {
+      "epoch": 0.5896535937738441,
+      "grad_norm": 1.5045848553943075,
+      "learning_rate": 3.8028677705244153e-06,
+      "loss": 0.3358,
+      "step": 1932
+    },
+    {
+      "epoch": 0.5899587974973295,
+      "grad_norm": 1.8852065445237391,
+      "learning_rate": 3.798067887801717e-06,
+      "loss": 0.6849,
+      "step": 1933
+    },
+    {
+      "epoch": 0.5902640012208149,
+      "grad_norm": 1.3753020214954932,
+      "learning_rate": 3.7932691803694992e-06,
+      "loss": 0.4061,
+      "step": 1934
+    },
+    {
+      "epoch": 0.5905692049443003,
+      "grad_norm": 1.563977115522218,
+      "learning_rate": 3.7884716529201047e-06,
+      "loss": 0.2681,
+      "step": 1935
+    },
+    {
+      "epoch": 0.5908744086677857,
+      "grad_norm": 1.3038261905545871,
+      "learning_rate": 3.783675310144719e-06,
+      "loss": 0.3416,
+      "step": 1936
+    },
+    {
+      "epoch": 0.5911796123912711,
+      "grad_norm": 1.422085924538324,
+      "learning_rate": 3.7788801567333712e-06,
+      "loss": 0.3074,
+      "step": 1937
+    },
+    {
+      "epoch": 0.5914848161147566,
+      "grad_norm": 1.3398381447188235,
+      "learning_rate": 3.774086197374926e-06,
+      "loss": 0.3046,
+      "step": 1938
+    },
+    {
+      "epoch": 0.591790019838242,
+      "grad_norm": 1.215628514615742,
+      "learning_rate": 3.7692934367570823e-06,
+      "loss": 0.2865,
+      "step": 1939
+    },
+    {
+      "epoch": 0.5920952235617275,
+      "grad_norm": 1.0941402284130506,
+      "learning_rate": 3.7645018795663667e-06,
+      "loss": 0.2081,
+      "step": 1940
+    },
+    {
+      "epoch": 0.5924004272852129,
+      "grad_norm": 1.5331642486114183,
+      "learning_rate": 3.759711530488127e-06,
+      "loss": 0.1809,
+      "step": 1941
+    },
+    {
+      "epoch": 0.5927056310086983,
+      "grad_norm": 1.501084412551385,
+      "learning_rate": 3.7549223942065303e-06,
+      "loss": 0.3311,
+      "step": 1942
+    },
+    {
+      "epoch": 0.5930108347321837,
+      "grad_norm": 1.6530406735695713,
+      "learning_rate": 3.7501344754045604e-06,
+      "loss": 0.3101,
+      "step": 1943
+    },
+    {
+      "epoch": 0.5933160384556692,
+      "grad_norm": 1.6341490317636014,
+      "learning_rate": 3.745347778764008e-06,
+      "loss": 0.4414,
+      "step": 1944
+    },
+    {
+      "epoch": 0.5936212421791546,
+      "grad_norm": 1.45040770751033,
+      "learning_rate": 3.740562308965471e-06,
+      "loss": 0.2298,
+      "step": 1945
+    },
+    {
+      "epoch": 0.59392644590264,
+      "grad_norm": 1.67431355829628,
+      "learning_rate": 3.7357780706883423e-06,
+      "loss": 0.3863,
+      "step": 1946
+    },
+    {
+      "epoch": 0.5942316496261254,
+      "grad_norm": 1.2794014225529522,
+      "learning_rate": 3.730995068610817e-06,
+      "loss": 0.3324,
+      "step": 1947
+    },
+    {
+      "epoch": 0.5945368533496108,
+      "grad_norm": 1.325879701050942,
+      "learning_rate": 3.7262133074098793e-06,
+      "loss": 0.1979,
+      "step": 1948
+    },
+    {
+      "epoch": 0.5948420570730963,
+      "grad_norm": 1.251910133040228,
+      "learning_rate": 3.721432791761297e-06,
+      "loss": 0.2215,
+      "step": 1949
+    },
+    {
+      "epoch": 0.5951472607965818,
+      "grad_norm": 1.5840427750100448,
+      "learning_rate": 3.716653526339623e-06,
+      "loss": 0.3227,
+      "step": 1950
+    },
+    {
+      "epoch": 0.5954524645200672,
+      "grad_norm": 1.5473806922537334,
+      "learning_rate": 3.7118755158181896e-06,
+      "loss": 0.3352,
+      "step": 1951
+    },
+    {
+      "epoch": 0.5957576682435526,
+      "grad_norm": 1.521976650845937,
+      "learning_rate": 3.707098764869097e-06,
+      "loss": 0.4004,
+      "step": 1952
+    },
+    {
+      "epoch": 0.596062871967038,
+      "grad_norm": 1.5737464518124529,
+      "learning_rate": 3.702323278163217e-06,
+      "loss": 0.2717,
+      "step": 1953
+    },
+    {
+      "epoch": 0.5963680756905234,
+      "grad_norm": 1.680802228214579,
+      "learning_rate": 3.6975490603701846e-06,
+      "loss": 0.3889,
+      "step": 1954
+    },
+    {
+      "epoch": 0.5966732794140088,
+      "grad_norm": 1.3239251995767745,
+      "learning_rate": 3.692776116158393e-06,
+      "loss": 0.2214,
+      "step": 1955
+    },
+    {
+      "epoch": 0.5969784831374942,
+      "grad_norm": 1.571247334990142,
+      "learning_rate": 3.688004450194992e-06,
+      "loss": 0.1805,
+      "step": 1956
+    },
+    {
+      "epoch": 0.5972836868609797,
+      "grad_norm": 1.5040931486442533,
+      "learning_rate": 3.6832340671458804e-06,
+      "loss": 0.4679,
+      "step": 1957
+    },
+    {
+      "epoch": 0.5975888905844652,
+      "grad_norm": 1.1553720285455236,
+      "learning_rate": 3.6784649716757003e-06,
+      "loss": 0.1948,
+      "step": 1958
+    },
+    {
+      "epoch": 0.5978940943079506,
+      "grad_norm": 1.5530514183498014,
+      "learning_rate": 3.6736971684478403e-06,
+      "loss": 0.4196,
+      "step": 1959
+    },
+    {
+      "epoch": 0.598199298031436,
+      "grad_norm": 1.360045695098193,
+      "learning_rate": 3.6689306621244196e-06,
+      "loss": 0.3248,
+      "step": 1960
+    },
+    {
+      "epoch": 0.5985045017549214,
+      "grad_norm": 1.6265003907056328,
+      "learning_rate": 3.6641654573662943e-06,
+      "loss": 0.4469,
+      "step": 1961
+    },
+    {
+      "epoch": 0.5988097054784068,
+      "grad_norm": 1.3520633893925043,
+      "learning_rate": 3.659401558833042e-06,
+      "loss": 0.3628,
+      "step": 1962
+    },
+    {
+      "epoch": 0.5991149092018923,
+      "grad_norm": 1.3981669073598408,
+      "learning_rate": 3.654638971182969e-06,
+      "loss": 0.2122,
+      "step": 1963
+    },
+    {
+      "epoch": 0.5994201129253777,
+      "grad_norm": 1.204357061326749,
+      "learning_rate": 3.6498776990730956e-06,
+      "loss": 0.3035,
+      "step": 1964
+    },
+    {
+      "epoch": 0.5997253166488631,
+      "grad_norm": 1.5117911388009846,
+      "learning_rate": 3.6451177471591564e-06,
+      "loss": 0.3495,
+      "step": 1965
+    },
+    {
+      "epoch": 0.6000305203723485,
+      "grad_norm": 1.350551955098285,
+      "learning_rate": 3.640359120095597e-06,
+      "loss": 0.2792,
+      "step": 1966
+    },
+    {
+      "epoch": 0.600335724095834,
+      "grad_norm": 0.9937686486563697,
+      "learning_rate": 3.635601822535567e-06,
+      "loss": 0.1301,
+      "step": 1967
+    },
+    {
+      "epoch": 0.6006409278193194,
+      "grad_norm": 1.2543699173760188,
+      "learning_rate": 3.6308458591309124e-06,
+      "loss": 0.3023,
+      "step": 1968
+    },
+    {
+      "epoch": 0.6009461315428049,
+      "grad_norm": 1.5915859714577205,
+      "learning_rate": 3.6260912345321807e-06,
+      "loss": 0.4957,
+      "step": 1969
+    },
+    {
+      "epoch": 0.6012513352662903,
+      "grad_norm": 1.2913315194645163,
+      "learning_rate": 3.6213379533886063e-06,
+      "loss": 0.1803,
+      "step": 1970
+    },
+    {
+      "epoch": 0.6015565389897757,
+      "grad_norm": 1.4490209069122582,
+      "learning_rate": 3.6165860203481084e-06,
+      "loss": 0.4276,
+      "step": 1971
+    },
+    {
+      "epoch": 0.6018617427132611,
+      "grad_norm": 1.4139380637367027,
+      "learning_rate": 3.611835440057293e-06,
+      "loss": 0.386,
+      "step": 1972
+    },
+    {
+      "epoch": 0.6021669464367465,
+      "grad_norm": 1.4566161354402127,
+      "learning_rate": 3.6070862171614394e-06,
+      "loss": 0.5218,
+      "step": 1973
+    },
+    {
+      "epoch": 0.6024721501602319,
+      "grad_norm": 1.38118497255283,
+      "learning_rate": 3.602338356304501e-06,
+      "loss": 0.2619,
+      "step": 1974
+    },
+    {
+      "epoch": 0.6027773538837173,
+      "grad_norm": 1.365310688641884,
+      "learning_rate": 3.597591862129097e-06,
+      "loss": 0.3383,
+      "step": 1975
+    },
+    {
+      "epoch": 0.6030825576072029,
+      "grad_norm": 1.2206504871463535,
+      "learning_rate": 3.5928467392765164e-06,
+      "loss": 0.2515,
+      "step": 1976
+    },
+    {
+      "epoch": 0.6033877613306883,
+      "grad_norm": 1.0845150310236251,
+      "learning_rate": 3.5881029923867017e-06,
+      "loss": 0.2363,
+      "step": 1977
+    },
+    {
+      "epoch": 0.6036929650541737,
+      "grad_norm": 1.4588781519857335,
+      "learning_rate": 3.5833606260982503e-06,
+      "loss": 0.436,
+      "step": 1978
+    },
+    {
+      "epoch": 0.6039981687776591,
+      "grad_norm": 1.4518968868371211,
+      "learning_rate": 3.5786196450484112e-06,
+      "loss": 0.3958,
+      "step": 1979
+    },
+    {
+      "epoch": 0.6043033725011445,
+      "grad_norm": 1.3211666256268693,
+      "learning_rate": 3.5738800538730802e-06,
+      "loss": 0.2995,
+      "step": 1980
+    },
+    {
+      "epoch": 0.6046085762246299,
+      "grad_norm": 1.0657353229411386,
+      "learning_rate": 3.5691418572067895e-06,
+      "loss": 0.2288,
+      "step": 1981
+    },
+    {
+      "epoch": 0.6049137799481153,
+      "grad_norm": 1.3448858561527488,
+      "learning_rate": 3.5644050596827106e-06,
+      "loss": 0.3239,
+      "step": 1982
+    },
+    {
+      "epoch": 0.6052189836716008,
+      "grad_norm": 1.3294545418957675,
+      "learning_rate": 3.55966966593265e-06,
+      "loss": 0.3452,
+      "step": 1983
+    },
+    {
+      "epoch": 0.6055241873950862,
+      "grad_norm": 1.3759155693099883,
+      "learning_rate": 3.554935680587035e-06,
+      "loss": 0.2258,
+      "step": 1984
+    },
+    {
+      "epoch": 0.6058293911185716,
+      "grad_norm": 1.5070104491764489,
+      "learning_rate": 3.55020310827492e-06,
+      "loss": 0.2878,
+      "step": 1985
+    },
+    {
+      "epoch": 0.6061345948420571,
+      "grad_norm": 1.879756425844785,
+      "learning_rate": 3.545471953623978e-06,
+      "loss": 0.2489,
+      "step": 1986
+    },
+    {
+      "epoch": 0.6064397985655425,
+      "grad_norm": 1.4696428547590852,
+      "learning_rate": 3.540742221260492e-06,
+      "loss": 0.2002,
+      "step": 1987
+    },
+    {
+      "epoch": 0.606745002289028,
+      "grad_norm": 1.494970951117057,
+      "learning_rate": 3.5360139158093576e-06,
+      "loss": 0.4269,
+      "step": 1988
+    },
+    {
+      "epoch": 0.6070502060125134,
+      "grad_norm": 1.6030926349786003,
+      "learning_rate": 3.5312870418940758e-06,
+      "loss": 0.4339,
+      "step": 1989
+    },
+    {
+      "epoch": 0.6073554097359988,
+      "grad_norm": 1.5754902818308054,
+      "learning_rate": 3.5265616041367445e-06,
+      "loss": 0.3722,
+      "step": 1990
+    },
+    {
+      "epoch": 0.6076606134594842,
+      "grad_norm": 1.5153266377103214,
+      "learning_rate": 3.5218376071580572e-06,
+      "loss": 0.352,
+      "step": 1991
+    },
+    {
+      "epoch": 0.6079658171829696,
+      "grad_norm": 1.1711146621923718,
+      "learning_rate": 3.5171150555773036e-06,
+      "loss": 0.2412,
+      "step": 1992
+    },
+    {
+      "epoch": 0.608271020906455,
+      "grad_norm": 1.4680707456552144,
+      "learning_rate": 3.5123939540123564e-06,
+      "loss": 0.3318,
+      "step": 1993
+    },
+    {
+      "epoch": 0.6085762246299404,
+      "grad_norm": 1.4459308023423367,
+      "learning_rate": 3.50767430707967e-06,
+      "loss": 0.4519,
+      "step": 1994
+    },
+    {
+      "epoch": 0.608881428353426,
+      "grad_norm": 1.3206132625233349,
+      "learning_rate": 3.5029561193942772e-06,
+      "loss": 0.4462,
+      "step": 1995
+    },
+    {
+      "epoch": 0.6091866320769114,
+      "grad_norm": 2.04442630508687,
+      "learning_rate": 3.4982393955697835e-06,
+      "loss": 0.3718,
+      "step": 1996
+    },
+    {
+      "epoch": 0.6094918358003968,
+      "grad_norm": 1.5146144368766,
+      "learning_rate": 3.4935241402183647e-06,
+      "loss": 0.2436,
+      "step": 1997
+    },
+    {
+      "epoch": 0.6097970395238822,
+      "grad_norm": 1.5272208942456742,
+      "learning_rate": 3.488810357950757e-06,
+      "loss": 0.3587,
+      "step": 1998
+    },
+    {
+      "epoch": 0.6101022432473676,
+      "grad_norm": 1.3609917333157049,
+      "learning_rate": 3.484098053376261e-06,
+      "loss": 0.45,
+      "step": 1999
+    },
+    {
+      "epoch": 0.610407446970853,
+      "grad_norm": 1.5254999529433142,
+      "learning_rate": 3.4793872311027286e-06,
+      "loss": 0.4317,
+      "step": 2000
+    },
+    {
+      "epoch": 0.6107126506943384,
+      "grad_norm": 1.1504364263616889,
+      "learning_rate": 3.4746778957365633e-06,
+      "loss": 0.2089,
+      "step": 2001
+    },
+    {
+      "epoch": 0.6110178544178239,
+      "grad_norm": 1.5797256057625808,
+      "learning_rate": 3.469970051882716e-06,
+      "loss": 0.3793,
+      "step": 2002
+    },
+    {
+      "epoch": 0.6113230581413093,
+      "grad_norm": 1.5176225849644855,
+      "learning_rate": 3.4652637041446768e-06,
+      "loss": 0.4404,
+      "step": 2003
+    },
+    {
+      "epoch": 0.6116282618647948,
+      "grad_norm": 1.3316765366062373,
+      "learning_rate": 3.4605588571244754e-06,
+      "loss": 0.2728,
+      "step": 2004
+    },
+    {
+      "epoch": 0.6119334655882802,
+      "grad_norm": 1.4620157434526462,
+      "learning_rate": 3.455855515422672e-06,
+      "loss": 0.1825,
+      "step": 2005
+    },
+    {
+      "epoch": 0.6122386693117656,
+      "grad_norm": 1.3467097842675653,
+      "learning_rate": 3.451153683638355e-06,
+      "loss": 0.1818,
+      "step": 2006
+    },
+    {
+      "epoch": 0.612543873035251,
+      "grad_norm": 1.5570818229152807,
+      "learning_rate": 3.4464533663691357e-06,
+      "loss": 0.3795,
+      "step": 2007
+    },
+    {
+      "epoch": 0.6128490767587365,
+      "grad_norm": 1.1553422096400865,
+      "learning_rate": 3.44175456821115e-06,
+      "loss": 0.317,
+      "step": 2008
+    },
+    {
+      "epoch": 0.6131542804822219,
+      "grad_norm": 1.585321917649714,
+      "learning_rate": 3.4370572937590394e-06,
+      "loss": 0.4378,
+      "step": 2009
+    },
+    {
+      "epoch": 0.6134594842057073,
+      "grad_norm": 1.5798636952287084,
+      "learning_rate": 3.432361547605965e-06,
+      "loss": 0.5201,
+      "step": 2010
+    },
+    {
+      "epoch": 0.6137646879291927,
+      "grad_norm": 1.7181031830680185,
+      "learning_rate": 3.4276673343435846e-06,
+      "loss": 0.2838,
+      "step": 2011
+    },
+    {
+      "epoch": 0.6140698916526781,
+      "grad_norm": 1.2099924152704025,
+      "learning_rate": 3.422974658562062e-06,
+      "loss": 0.3456,
+      "step": 2012
+    },
+    {
+      "epoch": 0.6143750953761636,
+      "grad_norm": 1.416064620650194,
+      "learning_rate": 3.4182835248500584e-06,
+      "loss": 0.2049,
+      "step": 2013
+    },
+    {
+      "epoch": 0.614680299099649,
+      "grad_norm": 1.4845303102857086,
+      "learning_rate": 3.4135939377947225e-06,
+      "loss": 0.4289,
+      "step": 2014
+    },
+    {
+      "epoch": 0.6149855028231345,
+      "grad_norm": 1.6544476184824486,
+      "learning_rate": 3.4089059019816957e-06,
+      "loss": 0.4024,
+      "step": 2015
+    },
+    {
+      "epoch": 0.6152907065466199,
+      "grad_norm": 1.2671491134981838,
+      "learning_rate": 3.4042194219951002e-06,
+      "loss": 0.3065,
+      "step": 2016
+    },
+    {
+      "epoch": 0.6155959102701053,
+      "grad_norm": 1.4331331682533894,
+      "learning_rate": 3.3995345024175354e-06,
+      "loss": 0.3072,
+      "step": 2017
+    },
+    {
+      "epoch": 0.6159011139935907,
+      "grad_norm": 1.2838500089236313,
+      "learning_rate": 3.394851147830078e-06,
+      "loss": 0.2258,
+      "step": 2018
+    },
+    {
+      "epoch": 0.6162063177170761,
+      "grad_norm": 1.6058392990442283,
+      "learning_rate": 3.3901693628122735e-06,
+      "loss": 0.5359,
+      "step": 2019
+    },
+    {
+      "epoch": 0.6165115214405615,
+      "grad_norm": 1.433938299450454,
+      "learning_rate": 3.3854891519421295e-06,
+      "loss": 0.4852,
+      "step": 2020
+    },
+    {
+      "epoch": 0.616816725164047,
+      "grad_norm": 1.5971858071124367,
+      "learning_rate": 3.3808105197961183e-06,
+      "loss": 0.3521,
+      "step": 2021
+    },
+    {
+      "epoch": 0.6171219288875325,
+      "grad_norm": 1.5979296909615484,
+      "learning_rate": 3.376133470949169e-06,
+      "loss": 0.3326,
+      "step": 2022
+    },
+    {
+      "epoch": 0.6174271326110179,
+      "grad_norm": 1.6015247619648738,
+      "learning_rate": 3.3714580099746565e-06,
+      "loss": 0.2547,
+      "step": 2023
+    },
+    {
+      "epoch": 0.6177323363345033,
+      "grad_norm": 1.7584158984983258,
+      "learning_rate": 3.366784141444409e-06,
+      "loss": 0.2062,
+      "step": 2024
+    },
+    {
+      "epoch": 0.6180375400579887,
+      "grad_norm": 2.0900618679983007,
+      "learning_rate": 3.362111869928696e-06,
+      "loss": 0.1623,
+      "step": 2025
+    },
+    {
+      "epoch": 0.6183427437814741,
+      "grad_norm": 1.4144306862691312,
+      "learning_rate": 3.3574411999962262e-06,
+      "loss": 0.2689,
+      "step": 2026
+    },
+    {
+      "epoch": 0.6186479475049596,
+      "grad_norm": 1.083391952151385,
+      "learning_rate": 3.3527721362141384e-06,
+      "loss": 0.2124,
+      "step": 2027
+    },
+    {
+      "epoch": 0.618953151228445,
+      "grad_norm": 1.5982387385586947,
+      "learning_rate": 3.3481046831480045e-06,
+      "loss": 0.4185,
+      "step": 2028
+    },
+    {
+      "epoch": 0.6192583549519304,
+      "grad_norm": 1.5038134261380083,
+      "learning_rate": 3.343438845361822e-06,
+      "loss": 0.3102,
+      "step": 2029
+    },
+    {
+      "epoch": 0.6195635586754158,
+      "grad_norm": 1.6651988559686899,
+      "learning_rate": 3.3387746274180045e-06,
+      "loss": 0.4605,
+      "step": 2030
+    },
+    {
+      "epoch": 0.6198687623989013,
+      "grad_norm": 1.4493202523521755,
+      "learning_rate": 3.3341120338773854e-06,
+      "loss": 0.2189,
+      "step": 2031
+    },
+    {
+      "epoch": 0.6201739661223867,
+      "grad_norm": 1.3263552824004858,
+      "learning_rate": 3.32945106929921e-06,
+      "loss": 0.2189,
+      "step": 2032
+    },
+    {
+      "epoch": 0.6204791698458721,
+      "grad_norm": 1.103331639979082,
+      "learning_rate": 3.3247917382411276e-06,
+      "loss": 0.2551,
+      "step": 2033
+    },
+    {
+      "epoch": 0.6207843735693576,
+      "grad_norm": 1.7271412703578817,
+      "learning_rate": 3.3201340452591924e-06,
+      "loss": 0.3365,
+      "step": 2034
+    },
+    {
+      "epoch": 0.621089577292843,
+      "grad_norm": 1.4270357648434941,
+      "learning_rate": 3.3154779949078587e-06,
+      "loss": 0.2515,
+      "step": 2035
+    },
+    {
+      "epoch": 0.6213947810163284,
+      "grad_norm": 1.561423579417306,
+      "learning_rate": 3.310823591739969e-06,
+      "loss": 0.4633,
+      "step": 2036
+    },
+    {
+      "epoch": 0.6216999847398138,
+      "grad_norm": 1.4440339903172814,
+      "learning_rate": 3.3061708403067605e-06,
+      "loss": 0.2857,
+      "step": 2037
+    },
+    {
+      "epoch": 0.6220051884632992,
+      "grad_norm": 1.3703992339885327,
+      "learning_rate": 3.301519745157854e-06,
+      "loss": 0.3527,
+      "step": 2038
+    },
+    {
+      "epoch": 0.6223103921867846,
+      "grad_norm": 2.3498289492425277,
+      "learning_rate": 3.296870310841248e-06,
+      "loss": 0.3941,
+      "step": 2039
+    },
+    {
+      "epoch": 0.6226155959102702,
+      "grad_norm": 1.755215094880646,
+      "learning_rate": 3.2922225419033182e-06,
+      "loss": 0.3497,
+      "step": 2040
+    },
+    {
+      "epoch": 0.6229207996337556,
+      "grad_norm": 1.751629343382298,
+      "learning_rate": 3.287576442888816e-06,
+      "loss": 0.2313,
+      "step": 2041
+    },
+    {
+      "epoch": 0.623226003357241,
+      "grad_norm": 1.4654791916014522,
+      "learning_rate": 3.2829320183408565e-06,
+      "loss": 0.4198,
+      "step": 2042
+    },
+    {
+      "epoch": 0.6235312070807264,
+      "grad_norm": 1.445719068924257,
+      "learning_rate": 3.278289272800914e-06,
+      "loss": 0.2312,
+      "step": 2043
+    },
+    {
+      "epoch": 0.6238364108042118,
+      "grad_norm": 1.323658332280364,
+      "learning_rate": 3.2736482108088263e-06,
+      "loss": 0.1885,
+      "step": 2044
+    },
+    {
+      "epoch": 0.6241416145276972,
+      "grad_norm": 1.3388854400481305,
+      "learning_rate": 3.2690088369027845e-06,
+      "loss": 0.2285,
+      "step": 2045
+    },
+    {
+      "epoch": 0.6244468182511826,
+      "grad_norm": 1.6613302948823414,
+      "learning_rate": 3.264371155619325e-06,
+      "loss": 0.3752,
+      "step": 2046
+    },
+    {
+      "epoch": 0.6247520219746681,
+      "grad_norm": 1.319712304928442,
+      "learning_rate": 3.2597351714933333e-06,
+      "loss": 0.3389,
+      "step": 2047
+    },
+    {
+      "epoch": 0.6250572256981535,
+      "grad_norm": 1.7970433031956077,
+      "learning_rate": 3.255100889058035e-06,
+      "loss": 0.2808,
+      "step": 2048
+    },
+    {
+      "epoch": 0.625362429421639,
+      "grad_norm": 1.344036604013848,
+      "learning_rate": 3.2504683128449877e-06,
+      "loss": 0.2621,
+      "step": 2049
+    },
+    {
+      "epoch": 0.6256676331451244,
+      "grad_norm": 1.1089242368639294,
+      "learning_rate": 3.245837447384085e-06,
+      "loss": 0.2054,
+      "step": 2050
+    },
+    {
+      "epoch": 0.6259728368686098,
+      "grad_norm": 1.2462381022200077,
+      "learning_rate": 3.2412082972035474e-06,
+      "loss": 0.207,
+      "step": 2051
+    },
+    {
+      "epoch": 0.6262780405920952,
+      "grad_norm": 2.099058566922468,
+      "learning_rate": 3.236580866829914e-06,
+      "loss": 0.2684,
+      "step": 2052
+    },
+    {
+      "epoch": 0.6265832443155807,
+      "grad_norm": 1.5959101977981203,
+      "learning_rate": 3.2319551607880465e-06,
+      "loss": 0.1538,
+      "step": 2053
+    },
+    {
+      "epoch": 0.6268884480390661,
+      "grad_norm": 1.5808452163488298,
+      "learning_rate": 3.2273311836011212e-06,
+      "loss": 0.4676,
+      "step": 2054
+    },
+    {
+      "epoch": 0.6271936517625515,
+      "grad_norm": 1.5783441976772852,
+      "learning_rate": 3.2227089397906176e-06,
+      "loss": 0.2694,
+      "step": 2055
+    },
+    {
+      "epoch": 0.6274988554860369,
+      "grad_norm": 1.3233154276606143,
+      "learning_rate": 3.2180884338763263e-06,
+      "loss": 0.2815,
+      "step": 2056
+    },
+    {
+      "epoch": 0.6278040592095223,
+      "grad_norm": 1.3683772986237586,
+      "learning_rate": 3.213469670376337e-06,
+      "loss": 0.4691,
+      "step": 2057
+    },
+    {
+      "epoch": 0.6281092629330077,
+      "grad_norm": 1.3814349020532382,
+      "learning_rate": 3.208852653807035e-06,
+      "loss": 0.3432,
+      "step": 2058
+    },
+    {
+      "epoch": 0.6284144666564933,
+      "grad_norm": 1.7053781007379099,
+      "learning_rate": 3.2042373886830997e-06,
+      "loss": 0.1941,
+      "step": 2059
+    },
+    {
+      "epoch": 0.6287196703799787,
+      "grad_norm": 1.8130342419647294,
+      "learning_rate": 3.1996238795174915e-06,
+      "loss": 0.3539,
+      "step": 2060
+    },
+    {
+      "epoch": 0.6290248741034641,
+      "grad_norm": 1.5884229605259044,
+      "learning_rate": 3.195012130821461e-06,
+      "loss": 0.2894,
+      "step": 2061
+    },
+    {
+      "epoch": 0.6293300778269495,
+      "grad_norm": 1.5727476151253794,
+      "learning_rate": 3.1904021471045343e-06,
+      "loss": 0.2491,
+      "step": 2062
+    },
+    {
+      "epoch": 0.6296352815504349,
+      "grad_norm": 1.3468078596546862,
+      "learning_rate": 3.1857939328745103e-06,
+      "loss": 0.2732,
+      "step": 2063
+    },
+    {
+      "epoch": 0.6299404852739203,
+      "grad_norm": 1.6412286283753263,
+      "learning_rate": 3.181187492637461e-06,
+      "loss": 0.3381,
+      "step": 2064
+    },
+    {
+      "epoch": 0.6302456889974057,
+      "grad_norm": 1.2694648607941608,
+      "learning_rate": 3.1765828308977194e-06,
+      "loss": 0.1918,
+      "step": 2065
+    },
+    {
+      "epoch": 0.6305508927208912,
+      "grad_norm": 1.6855358419576623,
+      "learning_rate": 3.1719799521578827e-06,
+      "loss": 0.3533,
+      "step": 2066
+    },
+    {
+      "epoch": 0.6308560964443766,
+      "grad_norm": 1.5392790511318661,
+      "learning_rate": 3.167378860918805e-06,
+      "loss": 0.4304,
+      "step": 2067
+    },
+    {
+      "epoch": 0.6311613001678621,
+      "grad_norm": 1.1533721472892073,
+      "learning_rate": 3.1627795616795897e-06,
+      "loss": 0.2677,
+      "step": 2068
+    },
+    {
+      "epoch": 0.6314665038913475,
+      "grad_norm": 2.1180949501028907,
+      "learning_rate": 3.1581820589375905e-06,
+      "loss": 0.3112,
+      "step": 2069
+    },
+    {
+      "epoch": 0.6317717076148329,
+      "grad_norm": 1.5671800333683827,
+      "learning_rate": 3.153586357188404e-06,
+      "loss": 0.325,
+      "step": 2070
+    },
+    {
+      "epoch": 0.6320769113383183,
+      "grad_norm": 2.097007040082022,
+      "learning_rate": 3.1489924609258625e-06,
+      "loss": 0.4118,
+      "step": 2071
+    },
+    {
+      "epoch": 0.6323821150618038,
+      "grad_norm": 1.4393021440971188,
+      "learning_rate": 3.144400374642038e-06,
+      "loss": 0.1812,
+      "step": 2072
+    },
+    {
+      "epoch": 0.6326873187852892,
+      "grad_norm": 1.3994869790842874,
+      "learning_rate": 3.1398101028272276e-06,
+      "loss": 0.2627,
+      "step": 2073
+    },
+    {
+      "epoch": 0.6329925225087746,
+      "grad_norm": 1.5025259049351598,
+      "learning_rate": 3.13522164996996e-06,
+      "loss": 0.5017,
+      "step": 2074
+    },
+    {
+      "epoch": 0.63329772623226,
+      "grad_norm": 1.4470049047960356,
+      "learning_rate": 3.1306350205569815e-06,
+      "loss": 0.2555,
+      "step": 2075
+    },
+    {
+      "epoch": 0.6336029299557454,
+      "grad_norm": 1.2392322727033596,
+      "learning_rate": 3.1260502190732526e-06,
+      "loss": 0.2755,
+      "step": 2076
+    },
+    {
+      "epoch": 0.6339081336792309,
+      "grad_norm": 1.511805726538404,
+      "learning_rate": 3.121467250001952e-06,
+      "loss": 0.3459,
+      "step": 2077
+    },
+    {
+      "epoch": 0.6342133374027163,
+      "grad_norm": 1.3525250485792497,
+      "learning_rate": 3.1168861178244647e-06,
+      "loss": 0.3096,
+      "step": 2078
+    },
+    {
+      "epoch": 0.6345185411262018,
+      "grad_norm": 1.307945705908503,
+      "learning_rate": 3.112306827020377e-06,
+      "loss": 0.2578,
+      "step": 2079
+    },
+    {
+      "epoch": 0.6348237448496872,
+      "grad_norm": 1.5394497301963965,
+      "learning_rate": 3.107729382067478e-06,
+      "loss": 0.3444,
+      "step": 2080
+    },
+    {
+      "epoch": 0.6351289485731726,
+      "grad_norm": 1.598444737094519,
+      "learning_rate": 3.1031537874417518e-06,
+      "loss": 0.4928,
+      "step": 2081
+    },
+    {
+      "epoch": 0.635434152296658,
+      "grad_norm": 1.5971914795206672,
+      "learning_rate": 3.0985800476173684e-06,
+      "loss": 0.4915,
+      "step": 2082
+    },
+    {
+      "epoch": 0.6357393560201434,
+      "grad_norm": 1.6527534139352973,
+      "learning_rate": 3.09400816706669e-06,
+      "loss": 0.4085,
+      "step": 2083
+    },
+    {
+      "epoch": 0.6360445597436288,
+      "grad_norm": 1.6401130967569915,
+      "learning_rate": 3.089438150260259e-06,
+      "loss": 0.3713,
+      "step": 2084
+    },
+    {
+      "epoch": 0.6363497634671142,
+      "grad_norm": 1.541484340849266,
+      "learning_rate": 3.084870001666792e-06,
+      "loss": 0.4459,
+      "step": 2085
+    },
+    {
+      "epoch": 0.6366549671905998,
+      "grad_norm": 1.0952222452240987,
+      "learning_rate": 3.0803037257531832e-06,
+      "loss": 0.2727,
+      "step": 2086
+    },
+    {
+      "epoch": 0.6369601709140852,
+      "grad_norm": 1.5642281702733645,
+      "learning_rate": 3.075739326984496e-06,
+      "loss": 0.3154,
+      "step": 2087
+    },
+    {
+      "epoch": 0.6372653746375706,
+      "grad_norm": 1.7115958702034013,
+      "learning_rate": 3.071176809823952e-06,
+      "loss": 0.3098,
+      "step": 2088
+    },
+    {
+      "epoch": 0.637570578361056,
+      "grad_norm": 1.4996254771292976,
+      "learning_rate": 3.0666161787329385e-06,
+      "loss": 0.3766,
+      "step": 2089
+    },
+    {
+      "epoch": 0.6378757820845414,
+      "grad_norm": 1.6333493167069202,
+      "learning_rate": 3.062057438170999e-06,
+      "loss": 0.2892,
+      "step": 2090
+    },
+    {
+      "epoch": 0.6381809858080268,
+      "grad_norm": 1.5987303464634444,
+      "learning_rate": 3.0575005925958262e-06,
+      "loss": 0.3212,
+      "step": 2091
+    },
+    {
+      "epoch": 0.6384861895315123,
+      "grad_norm": 1.4072617176331819,
+      "learning_rate": 3.0529456464632582e-06,
+      "loss": 0.1471,
+      "step": 2092
+    },
+    {
+      "epoch": 0.6387913932549977,
+      "grad_norm": 1.571466126550538,
+      "learning_rate": 3.0483926042272794e-06,
+      "loss": 0.3683,
+      "step": 2093
+    },
+    {
+      "epoch": 0.6390965969784831,
+      "grad_norm": 1.2227746095740997,
+      "learning_rate": 3.043841470340011e-06,
+      "loss": 0.2164,
+      "step": 2094
+    },
+    {
+      "epoch": 0.6394018007019686,
+      "grad_norm": 1.7729233319896496,
+      "learning_rate": 3.0392922492517056e-06,
+      "loss": 0.6999,
+      "step": 2095
+    },
+    {
+      "epoch": 0.639707004425454,
+      "grad_norm": 1.5388646652175832,
+      "learning_rate": 3.0347449454107502e-06,
+      "loss": 0.3488,
+      "step": 2096
+    },
+    {
+      "epoch": 0.6400122081489394,
+      "grad_norm": 1.8500513275861752,
+      "learning_rate": 3.030199563263654e-06,
+      "loss": 0.245,
+      "step": 2097
+    },
+    {
+      "epoch": 0.6403174118724249,
+      "grad_norm": 1.4391914036577291,
+      "learning_rate": 3.0256561072550474e-06,
+      "loss": 0.3902,
+      "step": 2098
+    },
+    {
+      "epoch": 0.6406226155959103,
+      "grad_norm": 1.3225074896212763,
+      "learning_rate": 3.0211145818276766e-06,
+      "loss": 0.2795,
+      "step": 2099
+    },
+    {
+      "epoch": 0.6409278193193957,
+      "grad_norm": 1.4008863095885327,
+      "learning_rate": 3.016574991422404e-06,
+      "loss": 0.4512,
+      "step": 2100
+    },
+    {
+      "epoch": 0.6412330230428811,
+      "grad_norm": 1.3077431718505468,
+      "learning_rate": 3.0120373404781934e-06,
+      "loss": 0.345,
+      "step": 2101
+    },
+    {
+      "epoch": 0.6415382267663665,
+      "grad_norm": 1.410336892347925,
+      "learning_rate": 3.0075016334321176e-06,
+      "loss": 0.1603,
+      "step": 2102
+    },
+    {
+      "epoch": 0.6418434304898519,
+      "grad_norm": 1.296606564603937,
+      "learning_rate": 3.002967874719348e-06,
+      "loss": 0.2145,
+      "step": 2103
+    },
+    {
+      "epoch": 0.6421486342133375,
+      "grad_norm": 2.160768722154708,
+      "learning_rate": 2.998436068773146e-06,
+      "loss": 0.2316,
+      "step": 2104
+    },
+    {
+      "epoch": 0.6424538379368229,
+      "grad_norm": 1.6235477120050208,
+      "learning_rate": 2.9939062200248697e-06,
+      "loss": 0.4334,
+      "step": 2105
+    },
+    {
+      "epoch": 0.6427590416603083,
+      "grad_norm": 1.586342896010167,
+      "learning_rate": 2.989378332903959e-06,
+      "loss": 0.2349,
+      "step": 2106
+    },
+    {
+      "epoch": 0.6430642453837937,
+      "grad_norm": 1.4002403274390032,
+      "learning_rate": 2.9848524118379403e-06,
+      "loss": 0.3905,
+      "step": 2107
+    },
+    {
+      "epoch": 0.6433694491072791,
+      "grad_norm": 1.5982996011356052,
+      "learning_rate": 2.9803284612524115e-06,
+      "loss": 0.41,
+      "step": 2108
+    },
+    {
+      "epoch": 0.6436746528307645,
+      "grad_norm": 2.297067231937714,
+      "learning_rate": 2.9758064855710477e-06,
+      "loss": 0.3621,
+      "step": 2109
+    },
+    {
+      "epoch": 0.6439798565542499,
+      "grad_norm": 2.318012369052118,
+      "learning_rate": 2.9712864892155934e-06,
+      "loss": 0.3104,
+      "step": 2110
+    },
+    {
+      "epoch": 0.6442850602777354,
+      "grad_norm": 2.142990253492185,
+      "learning_rate": 2.966768476605854e-06,
+      "loss": 0.3539,
+      "step": 2111
+    },
+    {
+      "epoch": 0.6445902640012208,
+      "grad_norm": 1.4203788515210105,
+      "learning_rate": 2.9622524521596986e-06,
+      "loss": 0.4049,
+      "step": 2112
+    },
+    {
+      "epoch": 0.6448954677247063,
+      "grad_norm": 1.46424450633585,
+      "learning_rate": 2.9577384202930515e-06,
+      "loss": 0.3637,
+      "step": 2113
+    },
+    {
+      "epoch": 0.6452006714481917,
+      "grad_norm": 1.3445167128438134,
+      "learning_rate": 2.9532263854198864e-06,
+      "loss": 0.2368,
+      "step": 2114
+    },
+    {
+      "epoch": 0.6455058751716771,
+      "grad_norm": 1.4888215773741194,
+      "learning_rate": 2.9487163519522268e-06,
+      "loss": 0.3003,
+      "step": 2115
+    },
+    {
+      "epoch": 0.6458110788951625,
+      "grad_norm": 1.7424695252828115,
+      "learning_rate": 2.9442083243001383e-06,
+      "loss": 0.4063,
+      "step": 2116
+    },
+    {
+      "epoch": 0.646116282618648,
+      "grad_norm": 1.5240331023468454,
+      "learning_rate": 2.9397023068717245e-06,
+      "loss": 0.3378,
+      "step": 2117
+    },
+    {
+      "epoch": 0.6464214863421334,
+      "grad_norm": 1.449502099811024,
+      "learning_rate": 2.935198304073125e-06,
+      "loss": 0.4407,
+      "step": 2118
+    },
+    {
+      "epoch": 0.6467266900656188,
+      "grad_norm": 1.5464626494768008,
+      "learning_rate": 2.9306963203085076e-06,
+      "loss": 0.4029,
+      "step": 2119
+    },
+    {
+      "epoch": 0.6470318937891042,
+      "grad_norm": 1.4483921541854292,
+      "learning_rate": 2.9261963599800663e-06,
+      "loss": 0.5365,
+      "step": 2120
+    },
+    {
+      "epoch": 0.6473370975125896,
+      "grad_norm": 1.6113587532600442,
+      "learning_rate": 2.9216984274880174e-06,
+      "loss": 0.3575,
+      "step": 2121
+    },
+    {
+      "epoch": 0.6476423012360751,
+      "grad_norm": 1.3748376056834286,
+      "learning_rate": 2.917202527230592e-06,
+      "loss": 0.3129,
+      "step": 2122
+    },
+    {
+      "epoch": 0.6479475049595605,
+      "grad_norm": 1.446286260072825,
+      "learning_rate": 2.912708663604039e-06,
+      "loss": 0.3331,
+      "step": 2123
+    },
+    {
+      "epoch": 0.648252708683046,
+      "grad_norm": 1.3725686685307052,
+      "learning_rate": 2.90821684100261e-06,
+      "loss": 0.1887,
+      "step": 2124
+    },
+    {
+      "epoch": 0.6485579124065314,
+      "grad_norm": 1.2658837136411039,
+      "learning_rate": 2.903727063818565e-06,
+      "loss": 0.181,
+      "step": 2125
+    },
+    {
+      "epoch": 0.6488631161300168,
+      "grad_norm": 1.6770300846847512,
+      "learning_rate": 2.899239336442162e-06,
+      "loss": 0.3232,
+      "step": 2126
+    },
+    {
+      "epoch": 0.6491683198535022,
+      "grad_norm": 1.5126111454755953,
+      "learning_rate": 2.8947536632616514e-06,
+      "loss": 0.3019,
+      "step": 2127
+    },
+    {
+      "epoch": 0.6494735235769876,
+      "grad_norm": 1.9696856500616016,
+      "learning_rate": 2.890270048663284e-06,
+      "loss": 0.2439,
+      "step": 2128
+    },
+    {
+      "epoch": 0.649778727300473,
+      "grad_norm": 1.7741220640229292,
+      "learning_rate": 2.8857884970312895e-06,
+      "loss": 0.3048,
+      "step": 2129
+    },
+    {
+      "epoch": 0.6500839310239585,
+      "grad_norm": 1.3726947707371107,
+      "learning_rate": 2.8813090127478805e-06,
+      "loss": 0.2802,
+      "step": 2130
+    },
+    {
+      "epoch": 0.6503891347474439,
+      "grad_norm": 1.3405879764735342,
+      "learning_rate": 2.8768316001932544e-06,
+      "loss": 0.3458,
+      "step": 2131
+    },
+    {
+      "epoch": 0.6506943384709294,
+      "grad_norm": 1.4428183815455966,
+      "learning_rate": 2.8723562637455775e-06,
+      "loss": 0.3397,
+      "step": 2132
+    },
+    {
+      "epoch": 0.6509995421944148,
+      "grad_norm": 1.4755108029659387,
+      "learning_rate": 2.8678830077809855e-06,
+      "loss": 0.2699,
+      "step": 2133
+    },
+    {
+      "epoch": 0.6513047459179002,
+      "grad_norm": 5.025887231644773,
+      "learning_rate": 2.8634118366735853e-06,
+      "loss": 0.3069,
+      "step": 2134
+    },
+    {
+      "epoch": 0.6516099496413856,
+      "grad_norm": 1.7675406942039875,
+      "learning_rate": 2.8589427547954395e-06,
+      "loss": 0.3095,
+      "step": 2135
+    },
+    {
+      "epoch": 0.651915153364871,
+      "grad_norm": 1.6515770975584783,
+      "learning_rate": 2.854475766516568e-06,
+      "loss": 0.4581,
+      "step": 2136
+    },
+    {
+      "epoch": 0.6522203570883565,
+      "grad_norm": 1.5412767621738763,
+      "learning_rate": 2.850010876204949e-06,
+      "loss": 0.426,
+      "step": 2137
+    },
+    {
+      "epoch": 0.6525255608118419,
+      "grad_norm": 1.6086969521289778,
+      "learning_rate": 2.8455480882265013e-06,
+      "loss": 0.5286,
+      "step": 2138
+    },
+    {
+      "epoch": 0.6528307645353273,
+      "grad_norm": 1.2747819975932901,
+      "learning_rate": 2.8410874069450965e-06,
+      "loss": 0.1708,
+      "step": 2139
+    },
+    {
+      "epoch": 0.6531359682588127,
+      "grad_norm": 1.6086675329642104,
+      "learning_rate": 2.836628836722539e-06,
+      "loss": 0.3442,
+      "step": 2140
+    },
+    {
+      "epoch": 0.6534411719822982,
+      "grad_norm": 1.9070796724428367,
+      "learning_rate": 2.8321723819185745e-06,
+      "loss": 0.5537,
+      "step": 2141
+    },
+    {
+      "epoch": 0.6537463757057836,
+      "grad_norm": 1.3332865329317232,
+      "learning_rate": 2.827718046890875e-06,
+      "loss": 0.2816,
+      "step": 2142
+    },
+    {
+      "epoch": 0.6540515794292691,
+      "grad_norm": 1.730187662342969,
+      "learning_rate": 2.8232658359950413e-06,
+      "loss": 0.2651,
+      "step": 2143
+    },
+    {
+      "epoch": 0.6543567831527545,
+      "grad_norm": 1.9822012341616055,
+      "learning_rate": 2.8188157535846017e-06,
+      "loss": 0.3211,
+      "step": 2144
+    },
+    {
+      "epoch": 0.6546619868762399,
+      "grad_norm": 1.5474652021744844,
+      "learning_rate": 2.8143678040109977e-06,
+      "loss": 0.3248,
+      "step": 2145
+    },
+    {
+      "epoch": 0.6549671905997253,
+      "grad_norm": 1.5206826366158495,
+      "learning_rate": 2.8099219916235846e-06,
+      "loss": 0.2097,
+      "step": 2146
+    },
+    {
+      "epoch": 0.6552723943232107,
+      "grad_norm": 1.3269303727241766,
+      "learning_rate": 2.8054783207696355e-06,
+      "loss": 0.2827,
+      "step": 2147
+    },
+    {
+      "epoch": 0.6555775980466961,
+      "grad_norm": 1.6297771220740915,
+      "learning_rate": 2.801036795794321e-06,
+      "loss": 0.3147,
+      "step": 2148
+    },
+    {
+      "epoch": 0.6558828017701815,
+      "grad_norm": 1.3416452558163536,
+      "learning_rate": 2.7965974210407155e-06,
+      "loss": 0.4029,
+      "step": 2149
+    },
+    {
+      "epoch": 0.6561880054936671,
+      "grad_norm": 2.1646703424182663,
+      "learning_rate": 2.7921602008497957e-06,
+      "loss": 0.4962,
+      "step": 2150
+    },
+    {
+      "epoch": 0.6564932092171525,
+      "grad_norm": 1.4042343953416563,
+      "learning_rate": 2.7877251395604256e-06,
+      "loss": 0.3924,
+      "step": 2151
+    },
+    {
+      "epoch": 0.6567984129406379,
+      "grad_norm": 1.6602055082306306,
+      "learning_rate": 2.783292241509359e-06,
+      "loss": 0.2895,
+      "step": 2152
+    },
+    {
+      "epoch": 0.6571036166641233,
+      "grad_norm": 1.643207080368195,
+      "learning_rate": 2.77886151103124e-06,
+      "loss": 0.3346,
+      "step": 2153
+    },
+    {
+      "epoch": 0.6574088203876087,
+      "grad_norm": 1.9432752121379677,
+      "learning_rate": 2.7744329524585855e-06,
+      "loss": 0.3341,
+      "step": 2154
+    },
+    {
+      "epoch": 0.6577140241110941,
+      "grad_norm": 1.724986120527517,
+      "learning_rate": 2.7700065701217915e-06,
+      "loss": 0.2615,
+      "step": 2155
+    },
+    {
+      "epoch": 0.6580192278345796,
+      "grad_norm": 1.7374368395405981,
+      "learning_rate": 2.7655823683491286e-06,
+      "loss": 0.3564,
+      "step": 2156
+    },
+    {
+      "epoch": 0.658324431558065,
+      "grad_norm": 1.3937644239286564,
+      "learning_rate": 2.761160351466734e-06,
+      "loss": 0.3398,
+      "step": 2157
+    },
+    {
+      "epoch": 0.6586296352815504,
+      "grad_norm": 1.5512659963171873,
+      "learning_rate": 2.7567405237986067e-06,
+      "loss": 0.2583,
+      "step": 2158
+    },
+    {
+      "epoch": 0.6589348390050359,
+      "grad_norm": 1.7238301193948424,
+      "learning_rate": 2.7523228896666037e-06,
+      "loss": 0.4185,
+      "step": 2159
+    },
+    {
+      "epoch": 0.6592400427285213,
+      "grad_norm": 1.493719782184466,
+      "learning_rate": 2.747907453390443e-06,
+      "loss": 0.3503,
+      "step": 2160
+    },
+    {
+      "epoch": 0.6595452464520067,
+      "grad_norm": 1.6238057076028836,
+      "learning_rate": 2.743494219287688e-06,
+      "loss": 0.3759,
+      "step": 2161
+    },
+    {
+      "epoch": 0.6598504501754922,
+      "grad_norm": 1.7162390660926063,
+      "learning_rate": 2.7390831916737485e-06,
+      "loss": 0.3259,
+      "step": 2162
+    },
+    {
+      "epoch": 0.6601556538989776,
+      "grad_norm": 1.5950502066049759,
+      "learning_rate": 2.734674374861883e-06,
+      "loss": 0.5201,
+      "step": 2163
+    },
+    {
+      "epoch": 0.660460857622463,
+      "grad_norm": 1.306778973153302,
+      "learning_rate": 2.730267773163181e-06,
+      "loss": 0.2982,
+      "step": 2164
+    },
+    {
+      "epoch": 0.6607660613459484,
+      "grad_norm": 1.496021876588278,
+      "learning_rate": 2.725863390886568e-06,
+      "loss": 0.4278,
+      "step": 2165
+    },
+    {
+      "epoch": 0.6610712650694338,
+      "grad_norm": 1.227396687435002,
+      "learning_rate": 2.721461232338804e-06,
+      "loss": 0.2761,
+      "step": 2166
+    },
+    {
+      "epoch": 0.6613764687929192,
+      "grad_norm": 1.7441506038435841,
+      "learning_rate": 2.7170613018244683e-06,
+      "loss": 0.555,
+      "step": 2167
+    },
+    {
+      "epoch": 0.6616816725164048,
+      "grad_norm": 1.6919704844062524,
+      "learning_rate": 2.712663603645962e-06,
+      "loss": 0.371,
+      "step": 2168
+    },
+    {
+      "epoch": 0.6619868762398902,
+      "grad_norm": 1.4900654188298506,
+      "learning_rate": 2.7082681421035094e-06,
+      "loss": 0.3146,
+      "step": 2169
+    },
+    {
+      "epoch": 0.6622920799633756,
+      "grad_norm": 1.1272813130706185,
+      "learning_rate": 2.703874921495142e-06,
+      "loss": 0.2102,
+      "step": 2170
+    },
+    {
+      "epoch": 0.662597283686861,
+      "grad_norm": 1.4795927410136391,
+      "learning_rate": 2.6994839461167e-06,
+      "loss": 0.2223,
+      "step": 2171
+    },
+    {
+      "epoch": 0.6629024874103464,
+      "grad_norm": 1.4827372299722401,
+      "learning_rate": 2.6950952202618313e-06,
+      "loss": 0.4663,
+      "step": 2172
+    },
+    {
+      "epoch": 0.6632076911338318,
+      "grad_norm": 1.861175690980772,
+      "learning_rate": 2.6907087482219848e-06,
+      "loss": 0.4224,
+      "step": 2173
+    },
+    {
+      "epoch": 0.6635128948573172,
+      "grad_norm": 2.5564241718377256,
+      "learning_rate": 2.6863245342864008e-06,
+      "loss": 0.1718,
+      "step": 2174
+    },
+    {
+      "epoch": 0.6638180985808027,
+      "grad_norm": 1.352942583505333,
+      "learning_rate": 2.681942582742114e-06,
+      "loss": 0.3322,
+      "step": 2175
+    },
+    {
+      "epoch": 0.6641233023042881,
+      "grad_norm": 1.4334146247404056,
+      "learning_rate": 2.677562897873949e-06,
+      "loss": 0.3147,
+      "step": 2176
+    },
+    {
+      "epoch": 0.6644285060277736,
+      "grad_norm": 1.3552893264263521,
+      "learning_rate": 2.673185483964512e-06,
+      "loss": 0.3708,
+      "step": 2177
+    },
+    {
+      "epoch": 0.664733709751259,
+      "grad_norm": 1.8684004670444918,
+      "learning_rate": 2.668810345294185e-06,
+      "loss": 0.4266,
+      "step": 2178
+    },
+    {
+      "epoch": 0.6650389134747444,
+      "grad_norm": 1.4768731330410994,
+      "learning_rate": 2.6644374861411327e-06,
+      "loss": 0.3179,
+      "step": 2179
+    },
+    {
+      "epoch": 0.6653441171982298,
+      "grad_norm": 2.1939141106150575,
+      "learning_rate": 2.660066910781286e-06,
+      "loss": 0.3913,
+      "step": 2180
+    },
+    {
+      "epoch": 0.6656493209217152,
+      "grad_norm": 1.7390756913307748,
+      "learning_rate": 2.655698623488341e-06,
+      "loss": 0.332,
+      "step": 2181
+    },
+    {
+      "epoch": 0.6659545246452007,
+      "grad_norm": 1.6594161953455198,
+      "learning_rate": 2.651332628533764e-06,
+      "loss": 0.4666,
+      "step": 2182
+    },
+    {
+      "epoch": 0.6662597283686861,
+      "grad_norm": 1.6759364627548272,
+      "learning_rate": 2.6469689301867707e-06,
+      "loss": 0.2634,
+      "step": 2183
+    },
+    {
+      "epoch": 0.6665649320921715,
+      "grad_norm": 1.4169243783848449,
+      "learning_rate": 2.6426075327143348e-06,
+      "loss": 0.327,
+      "step": 2184
+    },
+    {
+      "epoch": 0.6668701358156569,
+      "grad_norm": 1.6951437307987798,
+      "learning_rate": 2.638248440381184e-06,
+      "loss": 0.3784,
+      "step": 2185
+    },
+    {
+      "epoch": 0.6671753395391424,
+      "grad_norm": 1.4855364622342466,
+      "learning_rate": 2.633891657449785e-06,
+      "loss": 0.4026,
+      "step": 2186
+    },
+    {
+      "epoch": 0.6674805432626278,
+      "grad_norm": 1.6989385264177839,
+      "learning_rate": 2.6295371881803505e-06,
+      "loss": 0.3153,
+      "step": 2187
+    },
+    {
+      "epoch": 0.6677857469861133,
+      "grad_norm": 1.6589449424000402,
+      "learning_rate": 2.6251850368308307e-06,
+      "loss": 0.3181,
+      "step": 2188
+    },
+    {
+      "epoch": 0.6680909507095987,
+      "grad_norm": 1.69333748104809,
+      "learning_rate": 2.62083520765691e-06,
+      "loss": 0.3024,
+      "step": 2189
+    },
+    {
+      "epoch": 0.6683961544330841,
+      "grad_norm": 1.596754047867073,
+      "learning_rate": 2.6164877049119984e-06,
+      "loss": 0.1607,
+      "step": 2190
+    },
+    {
+      "epoch": 0.6687013581565695,
+      "grad_norm": 1.5331805768146902,
+      "learning_rate": 2.6121425328472333e-06,
+      "loss": 0.3948,
+      "step": 2191
+    },
+    {
+      "epoch": 0.6690065618800549,
+      "grad_norm": 1.3316754623873768,
+      "learning_rate": 2.607799695711476e-06,
+      "loss": 0.2833,
+      "step": 2192
+    },
+    {
+      "epoch": 0.6693117656035403,
+      "grad_norm": 1.4953180996621576,
+      "learning_rate": 2.603459197751299e-06,
+      "loss": 0.3014,
+      "step": 2193
+    },
+    {
+      "epoch": 0.6696169693270257,
+      "grad_norm": 1.5357227784615837,
+      "learning_rate": 2.599121043210989e-06,
+      "loss": 0.3256,
+      "step": 2194
+    },
+    {
+      "epoch": 0.6699221730505113,
+      "grad_norm": 1.3627066726724995,
+      "learning_rate": 2.5947852363325463e-06,
+      "loss": 0.2163,
+      "step": 2195
+    },
+    {
+      "epoch": 0.6702273767739967,
+      "grad_norm": 1.6218927459627959,
+      "learning_rate": 2.5904517813556703e-06,
+      "loss": 0.257,
+      "step": 2196
+    },
+    {
+      "epoch": 0.6705325804974821,
+      "grad_norm": 1.537367996503828,
+      "learning_rate": 2.5861206825177586e-06,
+      "loss": 0.4094,
+      "step": 2197
+    },
+    {
+      "epoch": 0.6708377842209675,
+      "grad_norm": 2.2108715034043542,
+      "learning_rate": 2.581791944053914e-06,
+      "loss": 0.3225,
+      "step": 2198
+    },
+    {
+      "epoch": 0.6711429879444529,
+      "grad_norm": 1.31507339693338,
+      "learning_rate": 2.5774655701969226e-06,
+      "loss": 0.2566,
+      "step": 2199
+    },
+    {
+      "epoch": 0.6714481916679383,
+      "grad_norm": 1.1941951940145592,
+      "learning_rate": 2.5731415651772597e-06,
+      "loss": 0.2822,
+      "step": 2200
+    },
+    {
+      "epoch": 0.6717533953914238,
+      "grad_norm": 1.5930828119126677,
+      "learning_rate": 2.568819933223089e-06,
+      "loss": 0.3602,
+      "step": 2201
+    },
+    {
+      "epoch": 0.6720585991149092,
+      "grad_norm": 1.5396556969732447,
+      "learning_rate": 2.564500678560249e-06,
+      "loss": 0.4319,
+      "step": 2202
+    },
+    {
+      "epoch": 0.6723638028383946,
+      "grad_norm": 1.2802585394346762,
+      "learning_rate": 2.560183805412254e-06,
+      "loss": 0.2013,
+      "step": 2203
+    },
+    {
+      "epoch": 0.67266900656188,
+      "grad_norm": 1.4333977422362991,
+      "learning_rate": 2.5558693180002925e-06,
+      "loss": 0.3595,
+      "step": 2204
+    },
+    {
+      "epoch": 0.6729742102853655,
+      "grad_norm": 1.539821534405401,
+      "learning_rate": 2.5515572205432193e-06,
+      "loss": 0.3156,
+      "step": 2205
+    },
+    {
+      "epoch": 0.6732794140088509,
+      "grad_norm": 1.5023269882496997,
+      "learning_rate": 2.5472475172575508e-06,
+      "loss": 0.4035,
+      "step": 2206
+    },
+    {
+      "epoch": 0.6735846177323364,
+      "grad_norm": 1.4676758104903433,
+      "learning_rate": 2.5429402123574593e-06,
+      "loss": 0.2568,
+      "step": 2207
+    },
+    {
+      "epoch": 0.6738898214558218,
+      "grad_norm": 1.4353821121434278,
+      "learning_rate": 2.5386353100547807e-06,
+      "loss": 0.3064,
+      "step": 2208
+    },
+    {
+      "epoch": 0.6741950251793072,
+      "grad_norm": 2.7468215607358695,
+      "learning_rate": 2.5343328145589933e-06,
+      "loss": 0.3209,
+      "step": 2209
+    },
+    {
+      "epoch": 0.6745002289027926,
+      "grad_norm": 1.6208529741654805,
+      "learning_rate": 2.5300327300772243e-06,
+      "loss": 0.345,
+      "step": 2210
+    },
+    {
+      "epoch": 0.674805432626278,
+      "grad_norm": 1.7500265664400212,
+      "learning_rate": 2.525735060814246e-06,
+      "loss": 0.2432,
+      "step": 2211
+    },
+    {
+      "epoch": 0.6751106363497634,
+      "grad_norm": 1.5826410319323236,
+      "learning_rate": 2.5214398109724666e-06,
+      "loss": 0.3547,
+      "step": 2212
+    },
+    {
+      "epoch": 0.6754158400732488,
+      "grad_norm": 1.6468757354329329,
+      "learning_rate": 2.5171469847519257e-06,
+      "loss": 0.4582,
+      "step": 2213
+    },
+    {
+      "epoch": 0.6757210437967344,
+      "grad_norm": 1.524464891699435,
+      "learning_rate": 2.5128565863503006e-06,
+      "loss": 0.2804,
+      "step": 2214
+    },
+    {
+      "epoch": 0.6760262475202198,
+      "grad_norm": 1.757213236067236,
+      "learning_rate": 2.508568619962889e-06,
+      "loss": 0.3666,
+      "step": 2215
+    },
+    {
+      "epoch": 0.6763314512437052,
+      "grad_norm": 1.6424191791501594,
+      "learning_rate": 2.5042830897826083e-06,
+      "loss": 0.5182,
+      "step": 2216
+    },
+    {
+      "epoch": 0.6766366549671906,
+      "grad_norm": 1.5743705733094293,
+      "learning_rate": 2.5000000000000015e-06,
+      "loss": 0.3995,
+      "step": 2217
+    },
+    {
+      "epoch": 0.676941858690676,
+      "grad_norm": 1.9861037166016167,
+      "learning_rate": 2.4957193548032187e-06,
+      "loss": 0.4346,
+      "step": 2218
+    },
+    {
+      "epoch": 0.6772470624141614,
+      "grad_norm": 1.4295765151007014,
+      "learning_rate": 2.49144115837802e-06,
+      "loss": 0.2974,
+      "step": 2219
+    },
+    {
+      "epoch": 0.6775522661376469,
+      "grad_norm": 1.2637981841035508,
+      "learning_rate": 2.4871654149077747e-06,
+      "loss": 0.1472,
+      "step": 2220
+    },
+    {
+      "epoch": 0.6778574698611323,
+      "grad_norm": 1.55571410056159,
+      "learning_rate": 2.482892128573454e-06,
+      "loss": 0.3477,
+      "step": 2221
+    },
+    {
+      "epoch": 0.6781626735846177,
+      "grad_norm": 1.5650642715760328,
+      "learning_rate": 2.4786213035536205e-06,
+      "loss": 0.3266,
+      "step": 2222
+    },
+    {
+      "epoch": 0.6784678773081032,
+      "grad_norm": 1.4957203212256605,
+      "learning_rate": 2.474352944024433e-06,
+      "loss": 0.3351,
+      "step": 2223
+    },
+    {
+      "epoch": 0.6787730810315886,
+      "grad_norm": 1.693859550058235,
+      "learning_rate": 2.4700870541596434e-06,
+      "loss": 0.2397,
+      "step": 2224
+    },
+    {
+      "epoch": 0.679078284755074,
+      "grad_norm": 1.1375829813106715,
+      "learning_rate": 2.4658236381305834e-06,
+      "loss": 0.2042,
+      "step": 2225
+    },
+    {
+      "epoch": 0.6793834884785594,
+      "grad_norm": 1.5914490489420068,
+      "learning_rate": 2.4615627001061655e-06,
+      "loss": 0.3779,
+      "step": 2226
+    },
+    {
+      "epoch": 0.6796886922020449,
+      "grad_norm": 1.747681989893777,
+      "learning_rate": 2.457304244252885e-06,
+      "loss": 0.3251,
+      "step": 2227
+    },
+    {
+      "epoch": 0.6799938959255303,
+      "grad_norm": 1.236007287216926,
+      "learning_rate": 2.4530482747348045e-06,
+      "loss": 0.2769,
+      "step": 2228
+    },
+    {
+      "epoch": 0.6802990996490157,
+      "grad_norm": 1.765434187309303,
+      "learning_rate": 2.448794795713555e-06,
+      "loss": 0.4044,
+      "step": 2229
+    },
+    {
+      "epoch": 0.6806043033725011,
+      "grad_norm": 1.3105969483407403,
+      "learning_rate": 2.444543811348338e-06,
+      "loss": 0.3482,
+      "step": 2230
+    },
+    {
+      "epoch": 0.6809095070959865,
+      "grad_norm": 2.476701221130886,
+      "learning_rate": 2.4402953257959107e-06,
+      "loss": 0.2611,
+      "step": 2231
+    },
+    {
+      "epoch": 0.681214710819472,
+      "grad_norm": 1.3629428476314511,
+      "learning_rate": 2.436049343210585e-06,
+      "loss": 0.198,
+      "step": 2232
+    },
+    {
+      "epoch": 0.6815199145429575,
+      "grad_norm": 1.2075854760475797,
+      "learning_rate": 2.4318058677442336e-06,
+      "loss": 0.261,
+      "step": 2233
+    },
+    {
+      "epoch": 0.6818251182664429,
+      "grad_norm": 1.5747515437022757,
+      "learning_rate": 2.4275649035462695e-06,
+      "loss": 0.3806,
+      "step": 2234
+    },
+    {
+      "epoch": 0.6821303219899283,
+      "grad_norm": 1.7471483702517483,
+      "learning_rate": 2.423326454763652e-06,
+      "loss": 0.3672,
+      "step": 2235
+    },
+    {
+      "epoch": 0.6824355257134137,
+      "grad_norm": 1.3886323003052499,
+      "learning_rate": 2.4190905255408853e-06,
+      "loss": 0.3586,
+      "step": 2236
+    },
+    {
+      "epoch": 0.6827407294368991,
+      "grad_norm": 1.6604409366430954,
+      "learning_rate": 2.4148571200200025e-06,
+      "loss": 0.2567,
+      "step": 2237
+    },
+    {
+      "epoch": 0.6830459331603845,
+      "grad_norm": 1.0963401642522104,
+      "learning_rate": 2.4106262423405754e-06,
+      "loss": 0.2167,
+      "step": 2238
+    },
+    {
+      "epoch": 0.68335113688387,
+      "grad_norm": 1.5618556410622801,
+      "learning_rate": 2.4063978966397017e-06,
+      "loss": 0.3866,
+      "step": 2239
+    },
+    {
+      "epoch": 0.6836563406073554,
+      "grad_norm": 1.7270739043944339,
+      "learning_rate": 2.4021720870520033e-06,
+      "loss": 0.2361,
+      "step": 2240
+    },
+    {
+      "epoch": 0.6839615443308409,
+      "grad_norm": 1.773884452256481,
+      "learning_rate": 2.397948817709618e-06,
+      "loss": 0.4827,
+      "step": 2241
+    },
+    {
+      "epoch": 0.6842667480543263,
+      "grad_norm": 1.6337555712580818,
+      "learning_rate": 2.3937280927422087e-06,
+      "loss": 0.2982,
+      "step": 2242
+    },
+    {
+      "epoch": 0.6845719517778117,
+      "grad_norm": 1.6348042728499839,
+      "learning_rate": 2.3895099162769426e-06,
+      "loss": 0.536,
+      "step": 2243
+    },
+    {
+      "epoch": 0.6848771555012971,
+      "grad_norm": 1.9449403092705355,
+      "learning_rate": 2.3852942924384953e-06,
+      "loss": 0.4189,
+      "step": 2244
+    },
+    {
+      "epoch": 0.6851823592247825,
+      "grad_norm": 1.6119447210408322,
+      "learning_rate": 2.3810812253490533e-06,
+      "loss": 0.464,
+      "step": 2245
+    },
+    {
+      "epoch": 0.685487562948268,
+      "grad_norm": 1.6223325109804345,
+      "learning_rate": 2.3768707191282957e-06,
+      "loss": 0.2066,
+      "step": 2246
+    },
+    {
+      "epoch": 0.6857927666717534,
+      "grad_norm": 1.5158096672896404,
+      "learning_rate": 2.3726627778933997e-06,
+      "loss": 0.2889,
+      "step": 2247
+    },
+    {
+      "epoch": 0.6860979703952388,
+      "grad_norm": 1.5524089736175863,
+      "learning_rate": 2.3684574057590337e-06,
+      "loss": 0.2266,
+      "step": 2248
+    },
+    {
+      "epoch": 0.6864031741187242,
+      "grad_norm": 1.7073229926701183,
+      "learning_rate": 2.364254606837358e-06,
+      "loss": 0.3773,
+      "step": 2249
+    },
+    {
+      "epoch": 0.6867083778422097,
+      "grad_norm": 1.7264873781734367,
+      "learning_rate": 2.360054385238012e-06,
+      "loss": 0.2424,
+      "step": 2250
+    },
+    {
+      "epoch": 0.6870135815656951,
+      "grad_norm": 1.8255696491964866,
+      "learning_rate": 2.355856745068115e-06,
+      "loss": 0.4534,
+      "step": 2251
+    },
+    {
+      "epoch": 0.6873187852891806,
+      "grad_norm": 1.826256343091619,
+      "learning_rate": 2.351661690432267e-06,
+      "loss": 0.276,
+      "step": 2252
+    },
+    {
+      "epoch": 0.687623989012666,
+      "grad_norm": 1.6053853628217218,
+      "learning_rate": 2.3474692254325328e-06,
+      "loss": 0.282,
+      "step": 2253
+    },
+    {
+      "epoch": 0.6879291927361514,
+      "grad_norm": 1.5373106925225748,
+      "learning_rate": 2.3432793541684502e-06,
+      "loss": 0.3329,
+      "step": 2254
+    },
+    {
+      "epoch": 0.6882343964596368,
+      "grad_norm": 2.2342562343804526,
+      "learning_rate": 2.339092080737021e-06,
+      "loss": 0.2579,
+      "step": 2255
+    },
+    {
+      "epoch": 0.6885396001831222,
+      "grad_norm": 1.5812283963020564,
+      "learning_rate": 2.334907409232704e-06,
+      "loss": 0.3708,
+      "step": 2256
+    },
+    {
+      "epoch": 0.6888448039066076,
+      "grad_norm": 1.5494931345932486,
+      "learning_rate": 2.330725343747412e-06,
+      "loss": 0.2758,
+      "step": 2257
+    },
+    {
+      "epoch": 0.689150007630093,
+      "grad_norm": 1.6512441741586468,
+      "learning_rate": 2.326545888370516e-06,
+      "loss": 0.2963,
+      "step": 2258
+    },
+    {
+      "epoch": 0.6894552113535786,
+      "grad_norm": 1.396566065819781,
+      "learning_rate": 2.3223690471888287e-06,
+      "loss": 0.14,
+      "step": 2259
+    },
+    {
+      "epoch": 0.689760415077064,
+      "grad_norm": 1.6348208984364616,
+      "learning_rate": 2.318194824286608e-06,
+      "loss": 0.2872,
+      "step": 2260
+    },
+    {
+      "epoch": 0.6900656188005494,
+      "grad_norm": 1.3921842084817904,
+      "learning_rate": 2.3140232237455557e-06,
+      "loss": 0.2908,
+      "step": 2261
+    },
+    {
+      "epoch": 0.6903708225240348,
+      "grad_norm": 1.5516133802355774,
+      "learning_rate": 2.3098542496448035e-06,
+      "loss": 0.3063,
+      "step": 2262
+    },
+    {
+      "epoch": 0.6906760262475202,
+      "grad_norm": 1.1592235990721869,
+      "learning_rate": 2.3056879060609154e-06,
+      "loss": 0.253,
+      "step": 2263
+    },
+    {
+      "epoch": 0.6909812299710056,
+      "grad_norm": 1.4434332099581564,
+      "learning_rate": 2.3015241970678904e-06,
+      "loss": 0.4221,
+      "step": 2264
+    },
+    {
+      "epoch": 0.691286433694491,
+      "grad_norm": 1.4893591474232895,
+      "learning_rate": 2.297363126737143e-06,
+      "loss": 0.2535,
+      "step": 2265
+    },
+    {
+      "epoch": 0.6915916374179765,
+      "grad_norm": 1.583421487194619,
+      "learning_rate": 2.2932046991375083e-06,
+      "loss": 0.417,
+      "step": 2266
+    },
+    {
+      "epoch": 0.6918968411414619,
+      "grad_norm": 1.5192571946940916,
+      "learning_rate": 2.289048918335245e-06,
+      "loss": 0.3871,
+      "step": 2267
+    },
+    {
+      "epoch": 0.6922020448649474,
+      "grad_norm": 1.4427150996998601,
+      "learning_rate": 2.2848957883940155e-06,
+      "loss": 0.1737,
+      "step": 2268
+    },
+    {
+      "epoch": 0.6925072485884328,
+      "grad_norm": 1.5122006458988937,
+      "learning_rate": 2.2807453133748904e-06,
+      "loss": 0.4822,
+      "step": 2269
+    },
+    {
+      "epoch": 0.6928124523119182,
+      "grad_norm": 1.6984069989930315,
+      "learning_rate": 2.276597497336349e-06,
+      "loss": 0.3834,
+      "step": 2270
+    },
+    {
+      "epoch": 0.6931176560354037,
+      "grad_norm": 1.3572682245688865,
+      "learning_rate": 2.27245234433427e-06,
+      "loss": 0.2541,
+      "step": 2271
+    },
+    {
+      "epoch": 0.6934228597588891,
+      "grad_norm": 1.8251110565674988,
+      "learning_rate": 2.268309858421925e-06,
+      "loss": 0.3142,
+      "step": 2272
+    },
+    {
+      "epoch": 0.6937280634823745,
+      "grad_norm": 1.3607310404566657,
+      "learning_rate": 2.264170043649976e-06,
+      "loss": 0.2144,
+      "step": 2273
+    },
+    {
+      "epoch": 0.6940332672058599,
+      "grad_norm": 1.7300473770093052,
+      "learning_rate": 2.2600329040664805e-06,
+      "loss": 0.3721,
+      "step": 2274
+    },
+    {
+      "epoch": 0.6943384709293453,
+      "grad_norm": 1.7843543323716198,
+      "learning_rate": 2.2558984437168735e-06,
+      "loss": 0.4995,
+      "step": 2275
+    },
+    {
+      "epoch": 0.6946436746528307,
+      "grad_norm": 1.2786625111949597,
+      "learning_rate": 2.251766666643971e-06,
+      "loss": 0.2811,
+      "step": 2276
+    },
+    {
+      "epoch": 0.6949488783763161,
+      "grad_norm": 1.6238076163532298,
+      "learning_rate": 2.2476375768879686e-06,
+      "loss": 0.3075,
+      "step": 2277
+    },
+    {
+      "epoch": 0.6952540820998017,
+      "grad_norm": 1.5830049508861397,
+      "learning_rate": 2.2435111784864323e-06,
+      "loss": 0.3474,
+      "step": 2278
+    },
+    {
+      "epoch": 0.6955592858232871,
+      "grad_norm": 1.847689670639536,
+      "learning_rate": 2.239387475474293e-06,
+      "loss": 0.2911,
+      "step": 2279
+    },
+    {
+      "epoch": 0.6958644895467725,
+      "grad_norm": 1.5764459421342933,
+      "learning_rate": 2.2352664718838522e-06,
+      "loss": 0.3912,
+      "step": 2280
+    },
+    {
+      "epoch": 0.6961696932702579,
+      "grad_norm": 1.6883475682950468,
+      "learning_rate": 2.231148171744768e-06,
+      "loss": 0.2597,
+      "step": 2281
+    },
+    {
+      "epoch": 0.6964748969937433,
+      "grad_norm": 1.5012572106181077,
+      "learning_rate": 2.2270325790840538e-06,
+      "loss": 0.3866,
+      "step": 2282
+    },
+    {
+      "epoch": 0.6967801007172287,
+      "grad_norm": 1.8193213912210353,
+      "learning_rate": 2.2229196979260803e-06,
+      "loss": 0.4478,
+      "step": 2283
+    },
+    {
+      "epoch": 0.6970853044407141,
+      "grad_norm": 1.578617887670166,
+      "learning_rate": 2.2188095322925623e-06,
+      "loss": 0.1706,
+      "step": 2284
+    },
+    {
+      "epoch": 0.6973905081641996,
+      "grad_norm": 1.355551417575489,
+      "learning_rate": 2.2147020862025598e-06,
+      "loss": 0.3604,
+      "step": 2285
+    },
+    {
+      "epoch": 0.697695711887685,
+      "grad_norm": 1.638895267359216,
+      "learning_rate": 2.2105973636724746e-06,
+      "loss": 0.291,
+      "step": 2286
+    },
+    {
+      "epoch": 0.6980009156111705,
+      "grad_norm": 1.4815833895565786,
+      "learning_rate": 2.2064953687160496e-06,
+      "loss": 0.2214,
+      "step": 2287
+    },
+    {
+      "epoch": 0.6983061193346559,
+      "grad_norm": 1.660687242747742,
+      "learning_rate": 2.202396105344352e-06,
+      "loss": 0.3362,
+      "step": 2288
+    },
+    {
+      "epoch": 0.6986113230581413,
+      "grad_norm": 1.3490028619474306,
+      "learning_rate": 2.1982995775657823e-06,
+      "loss": 0.1904,
+      "step": 2289
+    },
+    {
+      "epoch": 0.6989165267816267,
+      "grad_norm": 1.3466313542246475,
+      "learning_rate": 2.194205789386069e-06,
+      "loss": 0.278,
+      "step": 2290
+    },
+    {
+      "epoch": 0.6992217305051122,
+      "grad_norm": 1.631021932392484,
+      "learning_rate": 2.1901147448082568e-06,
+      "loss": 0.2299,
+      "step": 2291
+    },
+    {
+      "epoch": 0.6995269342285976,
+      "grad_norm": 1.347854555439276,
+      "learning_rate": 2.186026447832707e-06,
+      "loss": 0.3104,
+      "step": 2292
+    },
+    {
+      "epoch": 0.699832137952083,
+      "grad_norm": 1.9764058047217288,
+      "learning_rate": 2.181940902457102e-06,
+      "loss": 0.3163,
+      "step": 2293
+    },
+    {
+      "epoch": 0.7001373416755684,
+      "grad_norm": 1.5671620816634946,
+      "learning_rate": 2.1778581126764253e-06,
+      "loss": 0.5232,
+      "step": 2294
+    },
+    {
+      "epoch": 0.7004425453990538,
+      "grad_norm": 1.4410796170129498,
+      "learning_rate": 2.1737780824829673e-06,
+      "loss": 0.2524,
+      "step": 2295
+    },
+    {
+      "epoch": 0.7007477491225393,
+      "grad_norm": 1.6871822199250446,
+      "learning_rate": 2.1697008158663252e-06,
+      "loss": 0.3339,
+      "step": 2296
+    },
+    {
+      "epoch": 0.7010529528460248,
+      "grad_norm": 1.6588896822957648,
+      "learning_rate": 2.165626316813389e-06,
+      "loss": 0.2508,
+      "step": 2297
+    },
+    {
+      "epoch": 0.7013581565695102,
+      "grad_norm": 1.6510372542874876,
+      "learning_rate": 2.1615545893083413e-06,
+      "loss": 0.4034,
+      "step": 2298
+    },
+    {
+      "epoch": 0.7016633602929956,
+      "grad_norm": 1.581491938842722,
+      "learning_rate": 2.1574856373326607e-06,
+      "loss": 0.4049,
+      "step": 2299
+    },
+    {
+      "epoch": 0.701968564016481,
+      "grad_norm": 1.453233694297758,
+      "learning_rate": 2.153419464865107e-06,
+      "loss": 0.3468,
+      "step": 2300
+    },
+    {
+      "epoch": 0.7022737677399664,
+      "grad_norm": 1.808627365162433,
+      "learning_rate": 2.14935607588172e-06,
+      "loss": 0.303,
+      "step": 2301
+    },
+    {
+      "epoch": 0.7025789714634518,
+      "grad_norm": 1.4415485756344446,
+      "learning_rate": 2.1452954743558245e-06,
+      "loss": 0.3154,
+      "step": 2302
+    },
+    {
+      "epoch": 0.7028841751869372,
+      "grad_norm": 1.592144175830078,
+      "learning_rate": 2.141237664258017e-06,
+      "loss": 0.3909,
+      "step": 2303
+    },
+    {
+      "epoch": 0.7031893789104227,
+      "grad_norm": 1.4403194475823704,
+      "learning_rate": 2.1371826495561615e-06,
+      "loss": 0.2841,
+      "step": 2304
+    },
+    {
+      "epoch": 0.7034945826339082,
+      "grad_norm": 1.6852591435866713,
+      "learning_rate": 2.133130434215389e-06,
+      "loss": 0.512,
+      "step": 2305
+    },
+    {
+      "epoch": 0.7037997863573936,
+      "grad_norm": 1.5938343698945718,
+      "learning_rate": 2.1290810221980986e-06,
+      "loss": 0.2011,
+      "step": 2306
+    },
+    {
+      "epoch": 0.704104990080879,
+      "grad_norm": 1.5326866105983223,
+      "learning_rate": 2.125034417463942e-06,
+      "loss": 0.345,
+      "step": 2307
+    },
+    {
+      "epoch": 0.7044101938043644,
+      "grad_norm": 2.587166577662469,
+      "learning_rate": 2.1209906239698274e-06,
+      "loss": 0.3219,
+      "step": 2308
+    },
+    {
+      "epoch": 0.7047153975278498,
+      "grad_norm": 1.4756136474431192,
+      "learning_rate": 2.1169496456699175e-06,
+      "loss": 0.2406,
+      "step": 2309
+    },
+    {
+      "epoch": 0.7050206012513353,
+      "grad_norm": 1.613624517761201,
+      "learning_rate": 2.1129114865156187e-06,
+      "loss": 0.2712,
+      "step": 2310
+    },
+    {
+      "epoch": 0.7053258049748207,
+      "grad_norm": 1.5627764647993212,
+      "learning_rate": 2.1088761504555787e-06,
+      "loss": 0.2583,
+      "step": 2311
+    },
+    {
+      "epoch": 0.7056310086983061,
+      "grad_norm": 1.4190231900460788,
+      "learning_rate": 2.104843641435692e-06,
+      "loss": 0.3631,
+      "step": 2312
+    },
+    {
+      "epoch": 0.7059362124217915,
+      "grad_norm": 1.120858835736163,
+      "learning_rate": 2.100813963399083e-06,
+      "loss": 0.2264,
+      "step": 2313
+    },
+    {
+      "epoch": 0.706241416145277,
+      "grad_norm": 5.592583902383602,
+      "learning_rate": 2.096787120286107e-06,
+      "loss": 0.3734,
+      "step": 2314
+    },
+    {
+      "epoch": 0.7065466198687624,
+      "grad_norm": 1.4254198426262648,
+      "learning_rate": 2.092763116034352e-06,
+      "loss": 0.2948,
+      "step": 2315
+    },
+    {
+      "epoch": 0.7068518235922479,
+      "grad_norm": 1.4066840455637433,
+      "learning_rate": 2.0887419545786276e-06,
+      "loss": 0.3381,
+      "step": 2316
+    },
+    {
+      "epoch": 0.7071570273157333,
+      "grad_norm": 1.58796329496228,
+      "learning_rate": 2.0847236398509602e-06,
+      "loss": 0.3611,
+      "step": 2317
+    },
+    {
+      "epoch": 0.7074622310392187,
+      "grad_norm": 1.6806456206316778,
+      "learning_rate": 2.080708175780598e-06,
+      "loss": 0.2421,
+      "step": 2318
+    },
+    {
+      "epoch": 0.7077674347627041,
+      "grad_norm": 1.2883188125380933,
+      "learning_rate": 2.0766955662940023e-06,
+      "loss": 0.2541,
+      "step": 2319
+    },
+    {
+      "epoch": 0.7080726384861895,
+      "grad_norm": 1.7411635601553608,
+      "learning_rate": 2.072685815314838e-06,
+      "loss": 0.3755,
+      "step": 2320
+    },
+    {
+      "epoch": 0.7083778422096749,
+      "grad_norm": 1.3239666634134253,
+      "learning_rate": 2.0686789267639744e-06,
+      "loss": 0.3451,
+      "step": 2321
+    },
+    {
+      "epoch": 0.7086830459331603,
+      "grad_norm": 1.983676997898493,
+      "learning_rate": 2.0646749045594906e-06,
+      "loss": 0.42,
+      "step": 2322
+    },
+    {
+      "epoch": 0.7089882496566459,
+      "grad_norm": 1.0878535320868636,
+      "learning_rate": 2.060673752616653e-06,
+      "loss": 0.2291,
+      "step": 2323
+    },
+    {
+      "epoch": 0.7092934533801313,
+      "grad_norm": 1.2600542551045928,
+      "learning_rate": 2.0566754748479244e-06,
+      "loss": 0.2867,
+      "step": 2324
+    },
+    {
+      "epoch": 0.7095986571036167,
+      "grad_norm": 1.570453238941518,
+      "learning_rate": 2.0526800751629616e-06,
+      "loss": 0.2137,
+      "step": 2325
+    },
+    {
+      "epoch": 0.7099038608271021,
+      "grad_norm": 1.3212249131288691,
+      "learning_rate": 2.048687557468602e-06,
+      "loss": 0.3396,
+      "step": 2326
+    },
+    {
+      "epoch": 0.7102090645505875,
+      "grad_norm": 1.6117839376605163,
+      "learning_rate": 2.0446979256688647e-06,
+      "loss": 0.4791,
+      "step": 2327
+    },
+    {
+      "epoch": 0.7105142682740729,
+      "grad_norm": 1.2685593861176456,
+      "learning_rate": 2.0407111836649524e-06,
+      "loss": 0.3699,
+      "step": 2328
+    },
+    {
+      "epoch": 0.7108194719975583,
+      "grad_norm": 1.2768400639904869,
+      "learning_rate": 2.036727335355235e-06,
+      "loss": 0.2184,
+      "step": 2329
+    },
+    {
+      "epoch": 0.7111246757210438,
+      "grad_norm": 1.2921283676105053,
+      "learning_rate": 2.0327463846352562e-06,
+      "loss": 0.2455,
+      "step": 2330
+    },
+    {
+      "epoch": 0.7114298794445292,
+      "grad_norm": 1.6655785187324519,
+      "learning_rate": 2.0287683353977294e-06,
+      "loss": 0.3405,
+      "step": 2331
+    },
+    {
+      "epoch": 0.7117350831680147,
+      "grad_norm": 1.9153355938351166,
+      "learning_rate": 2.0247931915325254e-06,
+      "loss": 0.3889,
+      "step": 2332
+    },
+    {
+      "epoch": 0.7120402868915001,
+      "grad_norm": 1.4741604759613922,
+      "learning_rate": 2.0208209569266744e-06,
+      "loss": 0.2703,
+      "step": 2333
+    },
+    {
+      "epoch": 0.7123454906149855,
+      "grad_norm": 1.9778531651588995,
+      "learning_rate": 2.016851635464368e-06,
+      "loss": 0.2542,
+      "step": 2334
+    },
+    {
+      "epoch": 0.712650694338471,
+      "grad_norm": 1.590275061585144,
+      "learning_rate": 2.0128852310269413e-06,
+      "loss": 0.403,
+      "step": 2335
+    },
+    {
+      "epoch": 0.7129558980619564,
+      "grad_norm": 1.70148536079918,
+      "learning_rate": 2.008921747492884e-06,
+      "loss": 0.2813,
+      "step": 2336
+    },
+    {
+      "epoch": 0.7132611017854418,
+      "grad_norm": 1.5707428755209287,
+      "learning_rate": 2.004961188737823e-06,
+      "loss": 0.3936,
+      "step": 2337
+    },
+    {
+      "epoch": 0.7135663055089272,
+      "grad_norm": 1.3604719624909132,
+      "learning_rate": 2.0010035586345323e-06,
+      "loss": 0.2977,
+      "step": 2338
+    },
+    {
+      "epoch": 0.7138715092324126,
+      "grad_norm": 1.311140764412078,
+      "learning_rate": 1.997048861052916e-06,
+      "loss": 0.2976,
+      "step": 2339
+    },
+    {
+      "epoch": 0.714176712955898,
+      "grad_norm": 1.5495168302002966,
+      "learning_rate": 1.9930970998600122e-06,
+      "loss": 0.3235,
+      "step": 2340
+    },
+    {
+      "epoch": 0.7144819166793834,
+      "grad_norm": 1.55097318369583,
+      "learning_rate": 1.989148278919992e-06,
+      "loss": 0.2715,
+      "step": 2341
+    },
+    {
+      "epoch": 0.714787120402869,
+      "grad_norm": 1.5986632365767137,
+      "learning_rate": 1.985202402094146e-06,
+      "loss": 0.3065,
+      "step": 2342
+    },
+    {
+      "epoch": 0.7150923241263544,
+      "grad_norm": 1.5353380246081771,
+      "learning_rate": 1.9812594732408867e-06,
+      "loss": 0.278,
+      "step": 2343
+    },
+    {
+      "epoch": 0.7153975278498398,
+      "grad_norm": 1.5029875884822175,
+      "learning_rate": 1.9773194962157483e-06,
+      "loss": 0.3471,
+      "step": 2344
+    },
+    {
+      "epoch": 0.7157027315733252,
+      "grad_norm": 1.5119065752079335,
+      "learning_rate": 1.9733824748713747e-06,
+      "loss": 0.1674,
+      "step": 2345
+    },
+    {
+      "epoch": 0.7160079352968106,
+      "grad_norm": 1.4828013058914185,
+      "learning_rate": 1.9694484130575177e-06,
+      "loss": 0.3072,
+      "step": 2346
+    },
+    {
+      "epoch": 0.716313139020296,
+      "grad_norm": 1.3822579861494915,
+      "learning_rate": 1.965517314621042e-06,
+      "loss": 0.4649,
+      "step": 2347
+    },
+    {
+      "epoch": 0.7166183427437814,
+      "grad_norm": 1.458170018815355,
+      "learning_rate": 1.961589183405909e-06,
+      "loss": 0.3916,
+      "step": 2348
+    },
+    {
+      "epoch": 0.7169235464672669,
+      "grad_norm": 1.3859819200831016,
+      "learning_rate": 1.9576640232531785e-06,
+      "loss": 0.1681,
+      "step": 2349
+    },
+    {
+      "epoch": 0.7172287501907523,
+      "grad_norm": 1.515528410360513,
+      "learning_rate": 1.9537418380010096e-06,
+      "loss": 0.27,
+      "step": 2350
+    },
+    {
+      "epoch": 0.7175339539142378,
+      "grad_norm": 1.4128114424484064,
+      "learning_rate": 1.9498226314846473e-06,
+      "loss": 0.3298,
+      "step": 2351
+    },
+    {
+      "epoch": 0.7178391576377232,
+      "grad_norm": 1.2879356313934298,
+      "learning_rate": 1.9459064075364297e-06,
+      "loss": 0.2601,
+      "step": 2352
+    },
+    {
+      "epoch": 0.7181443613612086,
+      "grad_norm": 1.2039121554636452,
+      "learning_rate": 1.94199316998577e-06,
+      "loss": 0.2421,
+      "step": 2353
+    },
+    {
+      "epoch": 0.718449565084694,
+      "grad_norm": 1.4974907709515553,
+      "learning_rate": 1.938082922659172e-06,
+      "loss": 0.2043,
+      "step": 2354
+    },
+    {
+      "epoch": 0.7187547688081795,
+      "grad_norm": 1.6285164272848978,
+      "learning_rate": 1.934175669380206e-06,
+      "loss": 0.4642,
+      "step": 2355
+    },
+    {
+      "epoch": 0.7190599725316649,
+      "grad_norm": 1.7399320500358815,
+      "learning_rate": 1.9302714139695176e-06,
+      "loss": 0.4048,
+      "step": 2356
+    },
+    {
+      "epoch": 0.7193651762551503,
+      "grad_norm": 1.4081572420684747,
+      "learning_rate": 1.926370160244825e-06,
+      "loss": 0.2177,
+      "step": 2357
+    },
+    {
+      "epoch": 0.7196703799786357,
+      "grad_norm": 1.4367667276773715,
+      "learning_rate": 1.922471912020907e-06,
+      "loss": 0.3391,
+      "step": 2358
+    },
+    {
+      "epoch": 0.7199755837021211,
+      "grad_norm": 1.5369055490455528,
+      "learning_rate": 1.9185766731096015e-06,
+      "loss": 0.4479,
+      "step": 2359
+    },
+    {
+      "epoch": 0.7202807874256066,
+      "grad_norm": 1.5895659932815216,
+      "learning_rate": 1.9146844473198118e-06,
+      "loss": 0.396,
+      "step": 2360
+    },
+    {
+      "epoch": 0.720585991149092,
+      "grad_norm": 1.5298209236412752,
+      "learning_rate": 1.910795238457487e-06,
+      "loss": 0.4029,
+      "step": 2361
+    },
+    {
+      "epoch": 0.7208911948725775,
+      "grad_norm": 1.4132651931753095,
+      "learning_rate": 1.9069090503256277e-06,
+      "loss": 0.5109,
+      "step": 2362
+    },
+    {
+      "epoch": 0.7211963985960629,
+      "grad_norm": 1.0504476592048764,
+      "learning_rate": 1.903025886724285e-06,
+      "loss": 0.2154,
+      "step": 2363
+    },
+    {
+      "epoch": 0.7215016023195483,
+      "grad_norm": 1.5375095025986196,
+      "learning_rate": 1.8991457514505491e-06,
+      "loss": 0.3494,
+      "step": 2364
+    },
+    {
+      "epoch": 0.7218068060430337,
+      "grad_norm": 1.4249222349397213,
+      "learning_rate": 1.8952686482985466e-06,
+      "loss": 0.298,
+      "step": 2365
+    },
+    {
+      "epoch": 0.7221120097665191,
+      "grad_norm": 1.3483580600151477,
+      "learning_rate": 1.8913945810594458e-06,
+      "loss": 0.3369,
+      "step": 2366
+    },
+    {
+      "epoch": 0.7224172134900045,
+      "grad_norm": 1.6537907894810377,
+      "learning_rate": 1.8875235535214415e-06,
+      "loss": 0.4102,
+      "step": 2367
+    },
+    {
+      "epoch": 0.72272241721349,
+      "grad_norm": 1.7329477847290056,
+      "learning_rate": 1.8836555694697528e-06,
+      "loss": 0.3852,
+      "step": 2368
+    },
+    {
+      "epoch": 0.7230276209369755,
+      "grad_norm": 0.9438474667978664,
+      "learning_rate": 1.8797906326866355e-06,
+      "loss": 0.2417,
+      "step": 2369
+    },
+    {
+      "epoch": 0.7233328246604609,
+      "grad_norm": 1.3714424841819393,
+      "learning_rate": 1.875928746951353e-06,
+      "loss": 0.3425,
+      "step": 2370
+    },
+    {
+      "epoch": 0.7236380283839463,
+      "grad_norm": 1.5856665769786433,
+      "learning_rate": 1.8720699160401883e-06,
+      "loss": 0.417,
+      "step": 2371
+    },
+    {
+      "epoch": 0.7239432321074317,
+      "grad_norm": 1.617833989608368,
+      "learning_rate": 1.8682141437264423e-06,
+      "loss": 0.2736,
+      "step": 2372
+    },
+    {
+      "epoch": 0.7242484358309171,
+      "grad_norm": 1.3707649727949278,
+      "learning_rate": 1.8643614337804194e-06,
+      "loss": 0.2649,
+      "step": 2373
+    },
+    {
+      "epoch": 0.7245536395544026,
+      "grad_norm": 1.7067003431325545,
+      "learning_rate": 1.8605117899694325e-06,
+      "loss": 0.357,
+      "step": 2374
+    },
+    {
+      "epoch": 0.724858843277888,
+      "grad_norm": 1.4575250111130236,
+      "learning_rate": 1.856665216057793e-06,
+      "loss": 0.2518,
+      "step": 2375
+    },
+    {
+      "epoch": 0.7251640470013734,
+      "grad_norm": 1.8392255842033116,
+      "learning_rate": 1.8528217158068168e-06,
+      "loss": 0.3795,
+      "step": 2376
+    },
+    {
+      "epoch": 0.7254692507248588,
+      "grad_norm": 1.2997742383501494,
+      "learning_rate": 1.848981292974808e-06,
+      "loss": 0.2735,
+      "step": 2377
+    },
+    {
+      "epoch": 0.7257744544483443,
+      "grad_norm": 1.513637379899284,
+      "learning_rate": 1.8451439513170633e-06,
+      "loss": 0.408,
+      "step": 2378
+    },
+    {
+      "epoch": 0.7260796581718297,
+      "grad_norm": 1.5385949058679458,
+      "learning_rate": 1.8413096945858695e-06,
+      "loss": 0.3891,
+      "step": 2379
+    },
+    {
+      "epoch": 0.7263848618953151,
+      "grad_norm": 2.232541699335092,
+      "learning_rate": 1.837478526530493e-06,
+      "loss": 0.4276,
+      "step": 2380
+    },
+    {
+      "epoch": 0.7266900656188006,
+      "grad_norm": 1.6072156919899125,
+      "learning_rate": 1.8336504508971798e-06,
+      "loss": 0.4644,
+      "step": 2381
+    },
+    {
+      "epoch": 0.726995269342286,
+      "grad_norm": 1.7098489721428152,
+      "learning_rate": 1.8298254714291575e-06,
+      "loss": 0.4162,
+      "step": 2382
+    },
+    {
+      "epoch": 0.7273004730657714,
+      "grad_norm": 1.5804989022260485,
+      "learning_rate": 1.8260035918666203e-06,
+      "loss": 0.4063,
+      "step": 2383
+    },
+    {
+      "epoch": 0.7276056767892568,
+      "grad_norm": 1.2030976032258889,
+      "learning_rate": 1.8221848159467293e-06,
+      "loss": 0.3434,
+      "step": 2384
+    },
+    {
+      "epoch": 0.7279108805127422,
+      "grad_norm": 1.7012053704276349,
+      "learning_rate": 1.8183691474036224e-06,
+      "loss": 0.3693,
+      "step": 2385
+    },
+    {
+      "epoch": 0.7282160842362276,
+      "grad_norm": 1.488585433512296,
+      "learning_rate": 1.8145565899683875e-06,
+      "loss": 0.3843,
+      "step": 2386
+    },
+    {
+      "epoch": 0.7285212879597132,
+      "grad_norm": 1.5594380704245687,
+      "learning_rate": 1.8107471473690725e-06,
+      "loss": 0.2597,
+      "step": 2387
+    },
+    {
+      "epoch": 0.7288264916831986,
+      "grad_norm": 1.4362419262319646,
+      "learning_rate": 1.8069408233306846e-06,
+      "loss": 0.2128,
+      "step": 2388
+    },
+    {
+      "epoch": 0.729131695406684,
+      "grad_norm": 1.453994982767172,
+      "learning_rate": 1.8031376215751762e-06,
+      "loss": 0.4371,
+      "step": 2389
+    },
+    {
+      "epoch": 0.7294368991301694,
+      "grad_norm": 1.545719061618233,
+      "learning_rate": 1.7993375458214468e-06,
+      "loss": 0.2536,
+      "step": 2390
+    },
+    {
+      "epoch": 0.7297421028536548,
+      "grad_norm": 1.398782740764591,
+      "learning_rate": 1.7955405997853448e-06,
+      "loss": 0.2635,
+      "step": 2391
+    },
+    {
+      "epoch": 0.7300473065771402,
+      "grad_norm": 1.281366482533797,
+      "learning_rate": 1.7917467871796518e-06,
+      "loss": 0.25,
+      "step": 2392
+    },
+    {
+      "epoch": 0.7303525103006256,
+      "grad_norm": 1.7206023205122598,
+      "learning_rate": 1.7879561117140875e-06,
+      "loss": 0.1821,
+      "step": 2393
+    },
+    {
+      "epoch": 0.7306577140241111,
+      "grad_norm": 1.2659729373353488,
+      "learning_rate": 1.7841685770953071e-06,
+      "loss": 0.2269,
+      "step": 2394
+    },
+    {
+      "epoch": 0.7309629177475965,
+      "grad_norm": 1.7040532233697345,
+      "learning_rate": 1.7803841870268902e-06,
+      "loss": 0.5313,
+      "step": 2395
+    },
+    {
+      "epoch": 0.731268121471082,
+      "grad_norm": 1.7637367616801418,
+      "learning_rate": 1.776602945209342e-06,
+      "loss": 0.407,
+      "step": 2396
+    },
+    {
+      "epoch": 0.7315733251945674,
+      "grad_norm": 1.6319978132607265,
+      "learning_rate": 1.772824855340094e-06,
+      "loss": 0.4076,
+      "step": 2397
+    },
+    {
+      "epoch": 0.7318785289180528,
+      "grad_norm": 1.2482438148484345,
+      "learning_rate": 1.7690499211134898e-06,
+      "loss": 0.2887,
+      "step": 2398
+    },
+    {
+      "epoch": 0.7321837326415382,
+      "grad_norm": 1.9050410455001443,
+      "learning_rate": 1.7652781462207908e-06,
+      "loss": 0.3476,
+      "step": 2399
+    },
+    {
+      "epoch": 0.7324889363650237,
+      "grad_norm": 1.5632417062831914,
+      "learning_rate": 1.7615095343501637e-06,
+      "loss": 0.3904,
+      "step": 2400
+    },
+    {
+      "epoch": 0.7327941400885091,
+      "grad_norm": 1.2516479120148944,
+      "learning_rate": 1.7577440891866938e-06,
+      "loss": 0.3661,
+      "step": 2401
+    },
+    {
+      "epoch": 0.7330993438119945,
+      "grad_norm": 1.4281327595906879,
+      "learning_rate": 1.7539818144123589e-06,
+      "loss": 0.2554,
+      "step": 2402
+    },
+    {
+      "epoch": 0.7334045475354799,
+      "grad_norm": 1.9938065476006872,
+      "learning_rate": 1.7502227137060385e-06,
+      "loss": 0.211,
+      "step": 2403
+    },
+    {
+      "epoch": 0.7337097512589653,
+      "grad_norm": 1.5656770069370214,
+      "learning_rate": 1.7464667907435145e-06,
+      "loss": 0.3249,
+      "step": 2404
+    },
+    {
+      "epoch": 0.7340149549824508,
+      "grad_norm": 1.4857685975873842,
+      "learning_rate": 1.7427140491974553e-06,
+      "loss": 0.3922,
+      "step": 2405
+    },
+    {
+      "epoch": 0.7343201587059363,
+      "grad_norm": 1.5962279261031407,
+      "learning_rate": 1.7389644927374188e-06,
+      "loss": 0.4693,
+      "step": 2406
+    },
+    {
+      "epoch": 0.7346253624294217,
+      "grad_norm": 1.3060063075487254,
+      "learning_rate": 1.7352181250298527e-06,
+      "loss": 0.3797,
+      "step": 2407
+    },
+    {
+      "epoch": 0.7349305661529071,
+      "grad_norm": 1.6413985109214866,
+      "learning_rate": 1.7314749497380835e-06,
+      "loss": 0.4944,
+      "step": 2408
+    },
+    {
+      "epoch": 0.7352357698763925,
+      "grad_norm": 1.5695092723289383,
+      "learning_rate": 1.7277349705223133e-06,
+      "loss": 0.3746,
+      "step": 2409
+    },
+    {
+      "epoch": 0.7355409735998779,
+      "grad_norm": 1.694154475715984,
+      "learning_rate": 1.7239981910396274e-06,
+      "loss": 0.4035,
+      "step": 2410
+    },
+    {
+      "epoch": 0.7358461773233633,
+      "grad_norm": 1.4321803471625723,
+      "learning_rate": 1.720264614943974e-06,
+      "loss": 0.3191,
+      "step": 2411
+    },
+    {
+      "epoch": 0.7361513810468487,
+      "grad_norm": 1.3771418016095573,
+      "learning_rate": 1.7165342458861706e-06,
+      "loss": 0.2529,
+      "step": 2412
+    },
+    {
+      "epoch": 0.7364565847703342,
+      "grad_norm": 1.6635343761280355,
+      "learning_rate": 1.712807087513903e-06,
+      "loss": 0.3062,
+      "step": 2413
+    },
+    {
+      "epoch": 0.7367617884938196,
+      "grad_norm": 0.9998690400201703,
+      "learning_rate": 1.7090831434717142e-06,
+      "loss": 0.1487,
+      "step": 2414
+    },
+    {
+      "epoch": 0.7370669922173051,
+      "grad_norm": 1.5637897508967649,
+      "learning_rate": 1.7053624174010019e-06,
+      "loss": 0.434,
+      "step": 2415
+    },
+    {
+      "epoch": 0.7373721959407905,
+      "grad_norm": 1.4102781459716118,
+      "learning_rate": 1.7016449129400232e-06,
+      "loss": 0.2888,
+      "step": 2416
+    },
+    {
+      "epoch": 0.7376773996642759,
+      "grad_norm": 1.1743694967533793,
+      "learning_rate": 1.6979306337238787e-06,
+      "loss": 0.2099,
+      "step": 2417
+    },
+    {
+      "epoch": 0.7379826033877613,
+      "grad_norm": 1.7745821447548338,
+      "learning_rate": 1.694219583384521e-06,
+      "loss": 0.3803,
+      "step": 2418
+    },
+    {
+      "epoch": 0.7382878071112468,
+      "grad_norm": 1.6354627501520815,
+      "learning_rate": 1.690511765550739e-06,
+      "loss": 0.4103,
+      "step": 2419
+    },
+    {
+      "epoch": 0.7385930108347322,
+      "grad_norm": 1.5297920916047307,
+      "learning_rate": 1.6868071838481686e-06,
+      "loss": 0.3585,
+      "step": 2420
+    },
+    {
+      "epoch": 0.7388982145582176,
+      "grad_norm": 1.6670000537581684,
+      "learning_rate": 1.6831058418992747e-06,
+      "loss": 0.3463,
+      "step": 2421
+    },
+    {
+      "epoch": 0.739203418281703,
+      "grad_norm": 1.377165649462315,
+      "learning_rate": 1.6794077433233542e-06,
+      "loss": 0.2459,
+      "step": 2422
+    },
+    {
+      "epoch": 0.7395086220051884,
+      "grad_norm": 1.254668387413534,
+      "learning_rate": 1.675712891736539e-06,
+      "loss": 0.3237,
+      "step": 2423
+    },
+    {
+      "epoch": 0.7398138257286739,
+      "grad_norm": 1.6620188597654555,
+      "learning_rate": 1.6720212907517803e-06,
+      "loss": 0.4299,
+      "step": 2424
+    },
+    {
+      "epoch": 0.7401190294521593,
+      "grad_norm": 1.6523645277948942,
+      "learning_rate": 1.6683329439788498e-06,
+      "loss": 0.3996,
+      "step": 2425
+    },
+    {
+      "epoch": 0.7404242331756448,
+      "grad_norm": 1.6187362257912699,
+      "learning_rate": 1.6646478550243433e-06,
+      "loss": 0.3153,
+      "step": 2426
+    },
+    {
+      "epoch": 0.7407294368991302,
+      "grad_norm": 1.243329081404492,
+      "learning_rate": 1.6609660274916662e-06,
+      "loss": 0.203,
+      "step": 2427
+    },
+    {
+      "epoch": 0.7410346406226156,
+      "grad_norm": 1.4617906822883615,
+      "learning_rate": 1.6572874649810334e-06,
+      "loss": 0.3921,
+      "step": 2428
+    },
+    {
+      "epoch": 0.741339844346101,
+      "grad_norm": 1.4772782785804854,
+      "learning_rate": 1.653612171089473e-06,
+      "loss": 0.2399,
+      "step": 2429
+    },
+    {
+      "epoch": 0.7416450480695864,
+      "grad_norm": 1.516653989731904,
+      "learning_rate": 1.6499401494108124e-06,
+      "loss": 0.325,
+      "step": 2430
+    },
+    {
+      "epoch": 0.7419502517930718,
+      "grad_norm": 1.726246317224904,
+      "learning_rate": 1.646271403535678e-06,
+      "loss": 0.4074,
+      "step": 2431
+    },
+    {
+      "epoch": 0.7422554555165572,
+      "grad_norm": 1.6689973112145233,
+      "learning_rate": 1.6426059370515002e-06,
+      "loss": 0.2204,
+      "step": 2432
+    },
+    {
+      "epoch": 0.7425606592400428,
+      "grad_norm": 1.13846425278691,
+      "learning_rate": 1.638943753542493e-06,
+      "loss": 0.1386,
+      "step": 2433
+    },
+    {
+      "epoch": 0.7428658629635282,
+      "grad_norm": 1.4215926843752273,
+      "learning_rate": 1.6352848565896696e-06,
+      "loss": 0.2048,
+      "step": 2434
+    },
+    {
+      "epoch": 0.7431710666870136,
+      "grad_norm": 1.5202445192418594,
+      "learning_rate": 1.6316292497708208e-06,
+      "loss": 0.421,
+      "step": 2435
+    },
+    {
+      "epoch": 0.743476270410499,
+      "grad_norm": 1.6762691756708405,
+      "learning_rate": 1.6279769366605287e-06,
+      "loss": 0.4006,
+      "step": 2436
+    },
+    {
+      "epoch": 0.7437814741339844,
+      "grad_norm": 1.7012411775717595,
+      "learning_rate": 1.6243279208301483e-06,
+      "loss": 0.2386,
+      "step": 2437
+    },
+    {
+      "epoch": 0.7440866778574698,
+      "grad_norm": 1.4287681750645878,
+      "learning_rate": 1.620682205847811e-06,
+      "loss": 0.2714,
+      "step": 2438
+    },
+    {
+      "epoch": 0.7443918815809553,
+      "grad_norm": 1.6610026277708503,
+      "learning_rate": 1.617039795278425e-06,
+      "loss": 0.3636,
+      "step": 2439
+    },
+    {
+      "epoch": 0.7446970853044407,
+      "grad_norm": 2.0639538120528167,
+      "learning_rate": 1.6134006926836636e-06,
+      "loss": 0.247,
+      "step": 2440
+    },
+    {
+      "epoch": 0.7450022890279261,
+      "grad_norm": 1.5850709699512209,
+      "learning_rate": 1.609764901621964e-06,
+      "loss": 0.4474,
+      "step": 2441
+    },
+    {
+      "epoch": 0.7453074927514116,
+      "grad_norm": 1.6495669143346545,
+      "learning_rate": 1.6061324256485305e-06,
+      "loss": 0.4433,
+      "step": 2442
+    },
+    {
+      "epoch": 0.745612696474897,
+      "grad_norm": 1.7739809522756622,
+      "learning_rate": 1.6025032683153219e-06,
+      "loss": 0.3559,
+      "step": 2443
+    },
+    {
+      "epoch": 0.7459179001983824,
+      "grad_norm": 1.2253523767428416,
+      "learning_rate": 1.5988774331710511e-06,
+      "loss": 0.2543,
+      "step": 2444
+    },
+    {
+      "epoch": 0.7462231039218679,
+      "grad_norm": 1.267811099891557,
+      "learning_rate": 1.5952549237611864e-06,
+      "loss": 0.211,
+      "step": 2445
+    },
+    {
+      "epoch": 0.7465283076453533,
+      "grad_norm": 1.8148702710661058,
+      "learning_rate": 1.5916357436279416e-06,
+      "loss": 0.4958,
+      "step": 2446
+    },
+    {
+      "epoch": 0.7468335113688387,
+      "grad_norm": 1.4019785966150915,
+      "learning_rate": 1.5880198963102727e-06,
+      "loss": 0.3548,
+      "step": 2447
+    },
+    {
+      "epoch": 0.7471387150923241,
+      "grad_norm": 1.514170863443681,
+      "learning_rate": 1.5844073853438835e-06,
+      "loss": 0.3233,
+      "step": 2448
+    },
+    {
+      "epoch": 0.7474439188158095,
+      "grad_norm": 2.458336027327101,
+      "learning_rate": 1.5807982142612072e-06,
+      "loss": 0.3463,
+      "step": 2449
+    },
+    {
+      "epoch": 0.7477491225392949,
+      "grad_norm": 1.8685681972292403,
+      "learning_rate": 1.5771923865914186e-06,
+      "loss": 0.4082,
+      "step": 2450
+    },
+    {
+      "epoch": 0.7480543262627805,
+      "grad_norm": 1.4607828639804248,
+      "learning_rate": 1.5735899058604165e-06,
+      "loss": 0.2455,
+      "step": 2451
+    },
+    {
+      "epoch": 0.7483595299862659,
+      "grad_norm": 1.5437532772384637,
+      "learning_rate": 1.569990775590834e-06,
+      "loss": 0.2209,
+      "step": 2452
+    },
+    {
+      "epoch": 0.7486647337097513,
+      "grad_norm": 1.5998717524867132,
+      "learning_rate": 1.5663949993020216e-06,
+      "loss": 0.3259,
+      "step": 2453
+    },
+    {
+      "epoch": 0.7489699374332367,
+      "grad_norm": 1.5833261723942027,
+      "learning_rate": 1.5628025805100527e-06,
+      "loss": 0.3406,
+      "step": 2454
+    },
+    {
+      "epoch": 0.7492751411567221,
+      "grad_norm": 1.5819079692514202,
+      "learning_rate": 1.5592135227277194e-06,
+      "loss": 0.4312,
+      "step": 2455
+    },
+    {
+      "epoch": 0.7495803448802075,
+      "grad_norm": 2.0089884479621603,
+      "learning_rate": 1.5556278294645243e-06,
+      "loss": 0.3901,
+      "step": 2456
+    },
+    {
+      "epoch": 0.7498855486036929,
+      "grad_norm": 1.381615503205179,
+      "learning_rate": 1.5520455042266809e-06,
+      "loss": 0.2985,
+      "step": 2457
+    },
+    {
+      "epoch": 0.7501907523271784,
+      "grad_norm": 1.3001022628763312,
+      "learning_rate": 1.548466550517112e-06,
+      "loss": 0.2003,
+      "step": 2458
+    },
+    {
+      "epoch": 0.7504959560506638,
+      "grad_norm": 1.4329564811556512,
+      "learning_rate": 1.5448909718354398e-06,
+      "loss": 0.2839,
+      "step": 2459
+    },
+    {
+      "epoch": 0.7508011597741493,
+      "grad_norm": 1.5403711013032622,
+      "learning_rate": 1.5413187716779865e-06,
+      "loss": 0.3578,
+      "step": 2460
+    },
+    {
+      "epoch": 0.7511063634976347,
+      "grad_norm": 1.4433461602913331,
+      "learning_rate": 1.5377499535377755e-06,
+      "loss": 0.3038,
+      "step": 2461
+    },
+    {
+      "epoch": 0.7514115672211201,
+      "grad_norm": 1.3903299940281486,
+      "learning_rate": 1.5341845209045175e-06,
+      "loss": 0.3495,
+      "step": 2462
+    },
+    {
+      "epoch": 0.7517167709446055,
+      "grad_norm": 1.777219998589263,
+      "learning_rate": 1.5306224772646138e-06,
+      "loss": 0.3205,
+      "step": 2463
+    },
+    {
+      "epoch": 0.752021974668091,
+      "grad_norm": 1.3511000266469209,
+      "learning_rate": 1.527063826101156e-06,
+      "loss": 0.2054,
+      "step": 2464
+    },
+    {
+      "epoch": 0.7523271783915764,
+      "grad_norm": 1.4184063533061904,
+      "learning_rate": 1.523508570893914e-06,
+      "loss": 0.2443,
+      "step": 2465
+    },
+    {
+      "epoch": 0.7526323821150618,
+      "grad_norm": 1.4611404196422997,
+      "learning_rate": 1.5199567151193362e-06,
+      "loss": 0.2695,
+      "step": 2466
+    },
+    {
+      "epoch": 0.7529375858385472,
+      "grad_norm": 1.3855112755769896,
+      "learning_rate": 1.516408262250551e-06,
+      "loss": 0.2605,
+      "step": 2467
+    },
+    {
+      "epoch": 0.7532427895620326,
+      "grad_norm": 1.3116126921299953,
+      "learning_rate": 1.5128632157573597e-06,
+      "loss": 0.2025,
+      "step": 2468
+    },
+    {
+      "epoch": 0.7535479932855181,
+      "grad_norm": 1.4796773360325253,
+      "learning_rate": 1.5093215791062282e-06,
+      "loss": 0.378,
+      "step": 2469
+    },
+    {
+      "epoch": 0.7538531970090036,
+      "grad_norm": 1.432234949087124,
+      "learning_rate": 1.5057833557602886e-06,
+      "loss": 0.3638,
+      "step": 2470
+    },
+    {
+      "epoch": 0.754158400732489,
+      "grad_norm": 1.4684559142132647,
+      "learning_rate": 1.5022485491793404e-06,
+      "loss": 0.3034,
+      "step": 2471
+    },
+    {
+      "epoch": 0.7544636044559744,
+      "grad_norm": 1.712348391093848,
+      "learning_rate": 1.4987171628198376e-06,
+      "loss": 0.6474,
+      "step": 2472
+    },
+    {
+      "epoch": 0.7547688081794598,
+      "grad_norm": 1.2683167267429447,
+      "learning_rate": 1.49518920013489e-06,
+      "loss": 0.2659,
+      "step": 2473
+    },
+    {
+      "epoch": 0.7550740119029452,
+      "grad_norm": 1.4067118416141682,
+      "learning_rate": 1.4916646645742623e-06,
+      "loss": 0.3122,
+      "step": 2474
+    },
+    {
+      "epoch": 0.7553792156264306,
+      "grad_norm": 1.3294614462999204,
+      "learning_rate": 1.4881435595843658e-06,
+      "loss": 0.2014,
+      "step": 2475
+    },
+    {
+      "epoch": 0.755684419349916,
+      "grad_norm": 1.5514381996604707,
+      "learning_rate": 1.4846258886082559e-06,
+      "loss": 0.4586,
+      "step": 2476
+    },
+    {
+      "epoch": 0.7559896230734015,
+      "grad_norm": 1.2414279270860267,
+      "learning_rate": 1.4811116550856358e-06,
+      "loss": 0.1905,
+      "step": 2477
+    },
+    {
+      "epoch": 0.756294826796887,
+      "grad_norm": 1.5658974903410336,
+      "learning_rate": 1.4776008624528416e-06,
+      "loss": 0.1939,
+      "step": 2478
+    },
+    {
+      "epoch": 0.7566000305203724,
+      "grad_norm": 1.7196002591058608,
+      "learning_rate": 1.4740935141428447e-06,
+      "loss": 0.3984,
+      "step": 2479
+    },
+    {
+      "epoch": 0.7569052342438578,
+      "grad_norm": 1.7639793896757774,
+      "learning_rate": 1.470589613585255e-06,
+      "loss": 0.2218,
+      "step": 2480
+    },
+    {
+      "epoch": 0.7572104379673432,
+      "grad_norm": 1.3225571551889375,
+      "learning_rate": 1.4670891642063046e-06,
+      "loss": 0.177,
+      "step": 2481
+    },
+    {
+      "epoch": 0.7575156416908286,
+      "grad_norm": 1.6469842371461352,
+      "learning_rate": 1.4635921694288512e-06,
+      "loss": 0.3551,
+      "step": 2482
+    },
+    {
+      "epoch": 0.757820845414314,
+      "grad_norm": 1.6689084552871396,
+      "learning_rate": 1.4600986326723783e-06,
+      "loss": 0.3513,
+      "step": 2483
+    },
+    {
+      "epoch": 0.7581260491377995,
+      "grad_norm": 1.6027332771701015,
+      "learning_rate": 1.4566085573529876e-06,
+      "loss": 0.4021,
+      "step": 2484
+    },
+    {
+      "epoch": 0.7584312528612849,
+      "grad_norm": 1.583972182973655,
+      "learning_rate": 1.453121946883393e-06,
+      "loss": 0.297,
+      "step": 2485
+    },
+    {
+      "epoch": 0.7587364565847703,
+      "grad_norm": 1.758010242783844,
+      "learning_rate": 1.4496388046729205e-06,
+      "loss": 0.4719,
+      "step": 2486
+    },
+    {
+      "epoch": 0.7590416603082557,
+      "grad_norm": 1.4587979076269033,
+      "learning_rate": 1.446159134127509e-06,
+      "loss": 0.2238,
+      "step": 2487
+    },
+    {
+      "epoch": 0.7593468640317412,
+      "grad_norm": 1.516910361878622,
+      "learning_rate": 1.4426829386496983e-06,
+      "loss": 0.2117,
+      "step": 2488
+    },
+    {
+      "epoch": 0.7596520677552266,
+      "grad_norm": 1.027697194150736,
+      "learning_rate": 1.43921022163863e-06,
+      "loss": 0.1316,
+      "step": 2489
+    },
+    {
+      "epoch": 0.7599572714787121,
+      "grad_norm": 1.7174808324552195,
+      "learning_rate": 1.4357409864900495e-06,
+      "loss": 0.309,
+      "step": 2490
+    },
+    {
+      "epoch": 0.7602624752021975,
+      "grad_norm": 1.54123979105666,
+      "learning_rate": 1.4322752365962921e-06,
+      "loss": 0.3638,
+      "step": 2491
+    },
+    {
+      "epoch": 0.7605676789256829,
+      "grad_norm": 1.3864146154300274,
+      "learning_rate": 1.4288129753462858e-06,
+      "loss": 0.2661,
+      "step": 2492
+    },
+    {
+      "epoch": 0.7608728826491683,
+      "grad_norm": 1.7307300241219294,
+      "learning_rate": 1.4253542061255515e-06,
+      "loss": 0.3089,
+      "step": 2493
+    },
+    {
+      "epoch": 0.7611780863726537,
+      "grad_norm": 1.332318406461796,
+      "learning_rate": 1.4218989323161909e-06,
+      "loss": 0.3856,
+      "step": 2494
+    },
+    {
+      "epoch": 0.7614832900961391,
+      "grad_norm": 1.4346339638172307,
+      "learning_rate": 1.4184471572968873e-06,
+      "loss": 0.3232,
+      "step": 2495
+    },
+    {
+      "epoch": 0.7617884938196245,
+      "grad_norm": 1.3945504408939717,
+      "learning_rate": 1.4149988844429086e-06,
+      "loss": 0.2484,
+      "step": 2496
+    },
+    {
+      "epoch": 0.7620936975431101,
+      "grad_norm": 1.3073203652361365,
+      "learning_rate": 1.4115541171260928e-06,
+      "loss": 0.1359,
+      "step": 2497
+    },
+    {
+      "epoch": 0.7623989012665955,
+      "grad_norm": 1.2954470574924586,
+      "learning_rate": 1.4081128587148496e-06,
+      "loss": 0.2196,
+      "step": 2498
+    },
+    {
+      "epoch": 0.7627041049900809,
+      "grad_norm": 1.7836561181237647,
+      "learning_rate": 1.4046751125741615e-06,
+      "loss": 0.2989,
+      "step": 2499
+    },
+    {
+      "epoch": 0.7630093087135663,
+      "grad_norm": 1.6338560429242976,
+      "learning_rate": 1.4012408820655765e-06,
+      "loss": 0.2952,
+      "step": 2500
+    },
+    {
+      "epoch": 0.7633145124370517,
+      "grad_norm": 1.8965014723096463,
+      "learning_rate": 1.3978101705471992e-06,
+      "loss": 0.3649,
+      "step": 2501
+    },
+    {
+      "epoch": 0.7636197161605371,
+      "grad_norm": 1.403675435260702,
+      "learning_rate": 1.394382981373701e-06,
+      "loss": 0.2878,
+      "step": 2502
+    },
+    {
+      "epoch": 0.7639249198840226,
+      "grad_norm": 1.9008802859293734,
+      "learning_rate": 1.3909593178963032e-06,
+      "loss": 0.559,
+      "step": 2503
+    },
+    {
+      "epoch": 0.764230123607508,
+      "grad_norm": 1.6882297915798834,
+      "learning_rate": 1.3875391834627815e-06,
+      "loss": 0.2705,
+      "step": 2504
+    },
+    {
+      "epoch": 0.7645353273309934,
+      "grad_norm": 1.3132465146910093,
+      "learning_rate": 1.3841225814174592e-06,
+      "loss": 0.2154,
+      "step": 2505
+    },
+    {
+      "epoch": 0.7648405310544789,
+      "grad_norm": 1.503727255734048,
+      "learning_rate": 1.3807095151012101e-06,
+      "loss": 0.3739,
+      "step": 2506
+    },
+    {
+      "epoch": 0.7651457347779643,
+      "grad_norm": 1.8532239967908612,
+      "learning_rate": 1.3772999878514464e-06,
+      "loss": 0.3879,
+      "step": 2507
+    },
+    {
+      "epoch": 0.7654509385014497,
+      "grad_norm": 1.5263100172234858,
+      "learning_rate": 1.3738940030021191e-06,
+      "loss": 0.2597,
+      "step": 2508
+    },
+    {
+      "epoch": 0.7657561422249352,
+      "grad_norm": 1.5510558070282685,
+      "learning_rate": 1.37049156388372e-06,
+      "loss": 0.1667,
+      "step": 2509
+    },
+    {
+      "epoch": 0.7660613459484206,
+      "grad_norm": 1.7420929339963271,
+      "learning_rate": 1.3670926738232699e-06,
+      "loss": 0.5169,
+      "step": 2510
+    },
+    {
+      "epoch": 0.766366549671906,
+      "grad_norm": 1.5027245573035968,
+      "learning_rate": 1.3636973361443184e-06,
+      "loss": 0.4286,
+      "step": 2511
+    },
+    {
+      "epoch": 0.7666717533953914,
+      "grad_norm": 1.3230944336327843,
+      "learning_rate": 1.3603055541669457e-06,
+      "loss": 0.2573,
+      "step": 2512
+    },
+    {
+      "epoch": 0.7669769571188768,
+      "grad_norm": 2.2833142282099708,
+      "learning_rate": 1.3569173312077522e-06,
+      "loss": 0.3053,
+      "step": 2513
+    },
+    {
+      "epoch": 0.7672821608423622,
+      "grad_norm": 1.4527561170731167,
+      "learning_rate": 1.3535326705798563e-06,
+      "loss": 0.2039,
+      "step": 2514
+    },
+    {
+      "epoch": 0.7675873645658478,
+      "grad_norm": 1.1661107010154332,
+      "learning_rate": 1.3501515755928968e-06,
+      "loss": 0.1814,
+      "step": 2515
+    },
+    {
+      "epoch": 0.7678925682893332,
+      "grad_norm": 1.745056731933207,
+      "learning_rate": 1.346774049553027e-06,
+      "loss": 0.2475,
+      "step": 2516
+    },
+    {
+      "epoch": 0.7681977720128186,
+      "grad_norm": 1.7971733177340519,
+      "learning_rate": 1.3434000957629035e-06,
+      "loss": 0.2691,
+      "step": 2517
+    },
+    {
+      "epoch": 0.768502975736304,
+      "grad_norm": 1.9578135862598618,
+      "learning_rate": 1.3400297175216982e-06,
+      "loss": 0.5148,
+      "step": 2518
+    },
+    {
+      "epoch": 0.7688081794597894,
+      "grad_norm": 1.843697886619509,
+      "learning_rate": 1.33666291812508e-06,
+      "loss": 0.3523,
+      "step": 2519
+    },
+    {
+      "epoch": 0.7691133831832748,
+      "grad_norm": 1.1772101454940205,
+      "learning_rate": 1.3332997008652204e-06,
+      "loss": 0.2361,
+      "step": 2520
+    },
+    {
+      "epoch": 0.7694185869067602,
+      "grad_norm": 1.7516590156472773,
+      "learning_rate": 1.3299400690307917e-06,
+      "loss": 0.2935,
+      "step": 2521
+    },
+    {
+      "epoch": 0.7697237906302457,
+      "grad_norm": 1.538795564226342,
+      "learning_rate": 1.3265840259069551e-06,
+      "loss": 0.4383,
+      "step": 2522
+    },
+    {
+      "epoch": 0.7700289943537311,
+      "grad_norm": 1.415002162918217,
+      "learning_rate": 1.3232315747753639e-06,
+      "loss": 0.258,
+      "step": 2523
+    },
+    {
+      "epoch": 0.7703341980772166,
+      "grad_norm": 1.6090919467930516,
+      "learning_rate": 1.3198827189141627e-06,
+      "loss": 0.4418,
+      "step": 2524
+    },
+    {
+      "epoch": 0.770639401800702,
+      "grad_norm": 1.825167684820497,
+      "learning_rate": 1.3165374615979759e-06,
+      "loss": 0.2328,
+      "step": 2525
+    },
+    {
+      "epoch": 0.7709446055241874,
+      "grad_norm": 1.7249947589297079,
+      "learning_rate": 1.313195806097911e-06,
+      "loss": 0.4641,
+      "step": 2526
+    },
+    {
+      "epoch": 0.7712498092476728,
+      "grad_norm": 1.4212044612463415,
+      "learning_rate": 1.309857755681553e-06,
+      "loss": 0.2251,
+      "step": 2527
+    },
+    {
+      "epoch": 0.7715550129711582,
+      "grad_norm": 2.061260197546575,
+      "learning_rate": 1.3065233136129635e-06,
+      "loss": 0.4621,
+      "step": 2528
+    },
+    {
+      "epoch": 0.7718602166946437,
+      "grad_norm": 1.6672609064589439,
+      "learning_rate": 1.3031924831526737e-06,
+      "loss": 0.3449,
+      "step": 2529
+    },
+    {
+      "epoch": 0.7721654204181291,
+      "grad_norm": 1.4680252722090177,
+      "learning_rate": 1.2998652675576835e-06,
+      "loss": 0.3795,
+      "step": 2530
+    },
+    {
+      "epoch": 0.7724706241416145,
+      "grad_norm": 1.7029476248392246,
+      "learning_rate": 1.296541670081458e-06,
+      "loss": 0.3317,
+      "step": 2531
+    },
+    {
+      "epoch": 0.7727758278650999,
+      "grad_norm": 1.4895697995440238,
+      "learning_rate": 1.2932216939739284e-06,
+      "loss": 0.4766,
+      "step": 2532
+    },
+    {
+      "epoch": 0.7730810315885854,
+      "grad_norm": 1.7170191285718017,
+      "learning_rate": 1.2899053424814766e-06,
+      "loss": 0.1474,
+      "step": 2533
+    },
+    {
+      "epoch": 0.7733862353120708,
+      "grad_norm": 1.7030782430687559,
+      "learning_rate": 1.2865926188469497e-06,
+      "loss": 0.3246,
+      "step": 2534
+    },
+    {
+      "epoch": 0.7736914390355563,
+      "grad_norm": 1.6892554018601282,
+      "learning_rate": 1.2832835263096393e-06,
+      "loss": 0.4806,
+      "step": 2535
+    },
+    {
+      "epoch": 0.7739966427590417,
+      "grad_norm": 1.958577107076153,
+      "learning_rate": 1.2799780681052892e-06,
+      "loss": 0.4276,
+      "step": 2536
+    },
+    {
+      "epoch": 0.7743018464825271,
+      "grad_norm": 1.4386387336719104,
+      "learning_rate": 1.2766762474660931e-06,
+      "loss": 0.3012,
+      "step": 2537
+    },
+    {
+      "epoch": 0.7746070502060125,
+      "grad_norm": 1.4934041919813408,
+      "learning_rate": 1.2733780676206826e-06,
+      "loss": 0.3041,
+      "step": 2538
+    },
+    {
+      "epoch": 0.7749122539294979,
+      "grad_norm": 1.4600687963136778,
+      "learning_rate": 1.2700835317941296e-06,
+      "loss": 0.281,
+      "step": 2539
+    },
+    {
+      "epoch": 0.7752174576529833,
+      "grad_norm": 1.888694218658229,
+      "learning_rate": 1.2667926432079474e-06,
+      "loss": 0.2526,
+      "step": 2540
+    },
+    {
+      "epoch": 0.7755226613764687,
+      "grad_norm": 1.2069766331633431,
+      "learning_rate": 1.2635054050800788e-06,
+      "loss": 0.189,
+      "step": 2541
+    },
+    {
+      "epoch": 0.7758278650999543,
+      "grad_norm": 1.7914796184516408,
+      "learning_rate": 1.2602218206248962e-06,
+      "loss": 0.3041,
+      "step": 2542
+    },
+    {
+      "epoch": 0.7761330688234397,
+      "grad_norm": 1.6355628982899726,
+      "learning_rate": 1.256941893053204e-06,
+      "loss": 0.3401,
+      "step": 2543
+    },
+    {
+      "epoch": 0.7764382725469251,
+      "grad_norm": 1.5394860473922316,
+      "learning_rate": 1.2536656255722268e-06,
+      "loss": 0.2876,
+      "step": 2544
+    },
+    {
+      "epoch": 0.7767434762704105,
+      "grad_norm": 1.6669326172673098,
+      "learning_rate": 1.2503930213856098e-06,
+      "loss": 0.3818,
+      "step": 2545
+    },
+    {
+      "epoch": 0.7770486799938959,
+      "grad_norm": 1.9309726404813077,
+      "learning_rate": 1.2471240836934207e-06,
+      "loss": 0.3303,
+      "step": 2546
+    },
+    {
+      "epoch": 0.7773538837173813,
+      "grad_norm": 1.789663488734216,
+      "learning_rate": 1.2438588156921378e-06,
+      "loss": 0.595,
+      "step": 2547
+    },
+    {
+      "epoch": 0.7776590874408668,
+      "grad_norm": 1.8420347867118425,
+      "learning_rate": 1.2405972205746503e-06,
+      "loss": 0.2228,
+      "step": 2548
+    },
+    {
+      "epoch": 0.7779642911643522,
+      "grad_norm": 1.9279536438509446,
+      "learning_rate": 1.23733930153026e-06,
+      "loss": 0.2951,
+      "step": 2549
+    },
+    {
+      "epoch": 0.7782694948878376,
+      "grad_norm": 1.7373150486929299,
+      "learning_rate": 1.2340850617446725e-06,
+      "loss": 0.3925,
+      "step": 2550
+    },
+    {
+      "epoch": 0.7785746986113231,
+      "grad_norm": 1.5091105195836365,
+      "learning_rate": 1.230834504399994e-06,
+      "loss": 0.3783,
+      "step": 2551
+    },
+    {
+      "epoch": 0.7788799023348085,
+      "grad_norm": 1.6718016991634157,
+      "learning_rate": 1.2275876326747305e-06,
+      "loss": 0.2561,
+      "step": 2552
+    },
+    {
+      "epoch": 0.7791851060582939,
+      "grad_norm": 1.5749814289754123,
+      "learning_rate": 1.2243444497437867e-06,
+      "loss": 0.3802,
+      "step": 2553
+    },
+    {
+      "epoch": 0.7794903097817794,
+      "grad_norm": 1.4463613467598617,
+      "learning_rate": 1.2211049587784574e-06,
+      "loss": 0.2736,
+      "step": 2554
+    },
+    {
+      "epoch": 0.7797955135052648,
+      "grad_norm": 1.6206257707657987,
+      "learning_rate": 1.2178691629464257e-06,
+      "loss": 0.4952,
+      "step": 2555
+    },
+    {
+      "epoch": 0.7801007172287502,
+      "grad_norm": 1.4752179033349433,
+      "learning_rate": 1.2146370654117674e-06,
+      "loss": 0.3609,
+      "step": 2556
+    },
+    {
+      "epoch": 0.7804059209522356,
+      "grad_norm": 1.8063204424892583,
+      "learning_rate": 1.2114086693349365e-06,
+      "loss": 0.4268,
+      "step": 2557
+    },
+    {
+      "epoch": 0.780711124675721,
+      "grad_norm": 1.5945172426542091,
+      "learning_rate": 1.2081839778727683e-06,
+      "loss": 0.455,
+      "step": 2558
+    },
+    {
+      "epoch": 0.7810163283992064,
+      "grad_norm": 2.4498786584817176,
+      "learning_rate": 1.2049629941784801e-06,
+      "loss": 0.4397,
+      "step": 2559
+    },
+    {
+      "epoch": 0.7813215321226918,
+      "grad_norm": 1.3785621278506959,
+      "learning_rate": 1.2017457214016582e-06,
+      "loss": 0.2884,
+      "step": 2560
+    },
+    {
+      "epoch": 0.7816267358461774,
+      "grad_norm": 1.2967205185400668,
+      "learning_rate": 1.1985321626882617e-06,
+      "loss": 0.3354,
+      "step": 2561
+    },
+    {
+      "epoch": 0.7819319395696628,
+      "grad_norm": 2.1178940160744526,
+      "learning_rate": 1.1953223211806209e-06,
+      "loss": 0.2787,
+      "step": 2562
+    },
+    {
+      "epoch": 0.7822371432931482,
+      "grad_norm": 1.650534430564501,
+      "learning_rate": 1.192116200017428e-06,
+      "loss": 0.3282,
+      "step": 2563
+    },
+    {
+      "epoch": 0.7825423470166336,
+      "grad_norm": 1.6846552288380163,
+      "learning_rate": 1.1889138023337365e-06,
+      "loss": 0.1773,
+      "step": 2564
+    },
+    {
+      "epoch": 0.782847550740119,
+      "grad_norm": 1.586187483906093,
+      "learning_rate": 1.1857151312609622e-06,
+      "loss": 0.2041,
+      "step": 2565
+    },
+    {
+      "epoch": 0.7831527544636044,
+      "grad_norm": 1.3146422024869324,
+      "learning_rate": 1.1825201899268774e-06,
+      "loss": 0.359,
+      "step": 2566
+    },
+    {
+      "epoch": 0.7834579581870899,
+      "grad_norm": 1.6338815064392982,
+      "learning_rate": 1.1793289814556036e-06,
+      "loss": 0.3058,
+      "step": 2567
+    },
+    {
+      "epoch": 0.7837631619105753,
+      "grad_norm": 1.7511584670919065,
+      "learning_rate": 1.1761415089676126e-06,
+      "loss": 0.2958,
+      "step": 2568
+    },
+    {
+      "epoch": 0.7840683656340607,
+      "grad_norm": 1.475500219191894,
+      "learning_rate": 1.1729577755797278e-06,
+      "loss": 0.3746,
+      "step": 2569
+    },
+    {
+      "epoch": 0.7843735693575462,
+      "grad_norm": 1.4671615777046416,
+      "learning_rate": 1.1697777844051105e-06,
+      "loss": 0.3707,
+      "step": 2570
+    },
+    {
+      "epoch": 0.7846787730810316,
+      "grad_norm": 1.4370461452351335,
+      "learning_rate": 1.1666015385532648e-06,
+      "loss": 0.322,
+      "step": 2571
+    },
+    {
+      "epoch": 0.784983976804517,
+      "grad_norm": 1.4175708727468457,
+      "learning_rate": 1.1634290411300354e-06,
+      "loss": 0.2171,
+      "step": 2572
+    },
+    {
+      "epoch": 0.7852891805280025,
+      "grad_norm": 1.5392806000287926,
+      "learning_rate": 1.160260295237598e-06,
+      "loss": 0.3984,
+      "step": 2573
+    },
+    {
+      "epoch": 0.7855943842514879,
+      "grad_norm": 1.4241266552501894,
+      "learning_rate": 1.1570953039744592e-06,
+      "loss": 0.2882,
+      "step": 2574
+    },
+    {
+      "epoch": 0.7858995879749733,
+      "grad_norm": 1.7183588449833307,
+      "learning_rate": 1.153934070435459e-06,
+      "loss": 0.2638,
+      "step": 2575
+    },
+    {
+      "epoch": 0.7862047916984587,
+      "grad_norm": 1.4766861798193391,
+      "learning_rate": 1.1507765977117596e-06,
+      "loss": 0.2883,
+      "step": 2576
+    },
+    {
+      "epoch": 0.7865099954219441,
+      "grad_norm": 1.4461283262829343,
+      "learning_rate": 1.1476228888908437e-06,
+      "loss": 0.3625,
+      "step": 2577
+    },
+    {
+      "epoch": 0.7868151991454295,
+      "grad_norm": 1.5577204272457295,
+      "learning_rate": 1.1444729470565191e-06,
+      "loss": 0.3019,
+      "step": 2578
+    },
+    {
+      "epoch": 0.787120402868915,
+      "grad_norm": 2.044754210906538,
+      "learning_rate": 1.1413267752889062e-06,
+      "loss": 0.2757,
+      "step": 2579
+    },
+    {
+      "epoch": 0.7874256065924005,
+      "grad_norm": 1.5613863218607207,
+      "learning_rate": 1.1381843766644379e-06,
+      "loss": 0.1602,
+      "step": 2580
+    },
+    {
+      "epoch": 0.7877308103158859,
+      "grad_norm": 1.3626170466832803,
+      "learning_rate": 1.1350457542558608e-06,
+      "loss": 0.2227,
+      "step": 2581
+    },
+    {
+      "epoch": 0.7880360140393713,
+      "grad_norm": 1.2994120955756945,
+      "learning_rate": 1.131910911132229e-06,
+      "loss": 0.2259,
+      "step": 2582
+    },
+    {
+      "epoch": 0.7883412177628567,
+      "grad_norm": 1.5530753666950647,
+      "learning_rate": 1.1287798503588994e-06,
+      "loss": 0.3569,
+      "step": 2583
+    },
+    {
+      "epoch": 0.7886464214863421,
+      "grad_norm": 1.5007414574736635,
+      "learning_rate": 1.1256525749975278e-06,
+      "loss": 0.2333,
+      "step": 2584
+    },
+    {
+      "epoch": 0.7889516252098275,
+      "grad_norm": 1.6337499528343469,
+      "learning_rate": 1.1225290881060747e-06,
+      "loss": 0.3932,
+      "step": 2585
+    },
+    {
+      "epoch": 0.789256828933313,
+      "grad_norm": 1.46669837021519,
+      "learning_rate": 1.1194093927387917e-06,
+      "loss": 0.3024,
+      "step": 2586
+    },
+    {
+      "epoch": 0.7895620326567984,
+      "grad_norm": 1.610861230889205,
+      "learning_rate": 1.1162934919462226e-06,
+      "loss": 0.3625,
+      "step": 2587
+    },
+    {
+      "epoch": 0.7898672363802839,
+      "grad_norm": 1.3004736550965181,
+      "learning_rate": 1.1131813887752046e-06,
+      "loss": 0.2931,
+      "step": 2588
+    },
+    {
+      "epoch": 0.7901724401037693,
+      "grad_norm": 1.5664590650677594,
+      "learning_rate": 1.110073086268858e-06,
+      "loss": 0.3256,
+      "step": 2589
+    },
+    {
+      "epoch": 0.7904776438272547,
+      "grad_norm": 1.3693046525986896,
+      "learning_rate": 1.1069685874665858e-06,
+      "loss": 0.3568,
+      "step": 2590
+    },
+    {
+      "epoch": 0.7907828475507401,
+      "grad_norm": 1.4960672162046136,
+      "learning_rate": 1.103867895404077e-06,
+      "loss": 0.2358,
+      "step": 2591
+    },
+    {
+      "epoch": 0.7910880512742255,
+      "grad_norm": 1.5834616893473803,
+      "learning_rate": 1.100771013113293e-06,
+      "loss": 0.3178,
+      "step": 2592
+    },
+    {
+      "epoch": 0.791393254997711,
+      "grad_norm": 1.2548552157562565,
+      "learning_rate": 1.0976779436224706e-06,
+      "loss": 0.2395,
+      "step": 2593
+    },
+    {
+      "epoch": 0.7916984587211964,
+      "grad_norm": 1.581735088742958,
+      "learning_rate": 1.0945886899561215e-06,
+      "loss": 0.4003,
+      "step": 2594
+    },
+    {
+      "epoch": 0.7920036624446818,
+      "grad_norm": 1.6164845583620946,
+      "learning_rate": 1.0915032551350225e-06,
+      "loss": 0.3452,
+      "step": 2595
+    },
+    {
+      "epoch": 0.7923088661681672,
+      "grad_norm": 1.665959311424618,
+      "learning_rate": 1.0884216421762167e-06,
+      "loss": 0.3503,
+      "step": 2596
+    },
+    {
+      "epoch": 0.7926140698916527,
+      "grad_norm": 1.3877622305018942,
+      "learning_rate": 1.0853438540930127e-06,
+      "loss": 0.2358,
+      "step": 2597
+    },
+    {
+      "epoch": 0.7929192736151381,
+      "grad_norm": 1.6745241514089109,
+      "learning_rate": 1.0822698938949783e-06,
+      "loss": 0.4222,
+      "step": 2598
+    },
+    {
+      "epoch": 0.7932244773386236,
+      "grad_norm": 1.5809693339955015,
+      "learning_rate": 1.079199764587936e-06,
+      "loss": 0.3156,
+      "step": 2599
+    },
+    {
+      "epoch": 0.793529681062109,
+      "grad_norm": 1.9384104527651729,
+      "learning_rate": 1.0761334691739623e-06,
+      "loss": 0.6036,
+      "step": 2600
+    },
+    {
+      "epoch": 0.7938348847855944,
+      "grad_norm": 1.3628061772973745,
+      "learning_rate": 1.0730710106513887e-06,
+      "loss": 0.2319,
+      "step": 2601
+    },
+    {
+      "epoch": 0.7941400885090798,
+      "grad_norm": 1.5729552847393045,
+      "learning_rate": 1.0700123920147915e-06,
+      "loss": 0.3517,
+      "step": 2602
+    },
+    {
+      "epoch": 0.7944452922325652,
+      "grad_norm": 1.5823186919378267,
+      "learning_rate": 1.0669576162549905e-06,
+      "loss": 0.3823,
+      "step": 2603
+    },
+    {
+      "epoch": 0.7947504959560506,
+      "grad_norm": 1.5621083340900117,
+      "learning_rate": 1.063906686359054e-06,
+      "loss": 0.2805,
+      "step": 2604
+    },
+    {
+      "epoch": 0.795055699679536,
+      "grad_norm": 1.5411503760694267,
+      "learning_rate": 1.0608596053102849e-06,
+      "loss": 0.1962,
+      "step": 2605
+    },
+    {
+      "epoch": 0.7953609034030216,
+      "grad_norm": 1.6821783005946354,
+      "learning_rate": 1.057816376088221e-06,
+      "loss": 0.4098,
+      "step": 2606
+    },
+    {
+      "epoch": 0.795666107126507,
+      "grad_norm": 1.7612649878790774,
+      "learning_rate": 1.0547770016686403e-06,
+      "loss": 0.3115,
+      "step": 2607
+    },
+    {
+      "epoch": 0.7959713108499924,
+      "grad_norm": 1.9945711125156906,
+      "learning_rate": 1.0517414850235452e-06,
+      "loss": 0.4042,
+      "step": 2608
+    },
+    {
+      "epoch": 0.7962765145734778,
+      "grad_norm": 1.6001281269747107,
+      "learning_rate": 1.0487098291211678e-06,
+      "loss": 0.3009,
+      "step": 2609
+    },
+    {
+      "epoch": 0.7965817182969632,
+      "grad_norm": 1.2878584353560845,
+      "learning_rate": 1.0456820369259667e-06,
+      "loss": 0.1765,
+      "step": 2610
+    },
+    {
+      "epoch": 0.7968869220204486,
+      "grad_norm": 1.6267063278824643,
+      "learning_rate": 1.04265811139862e-06,
+      "loss": 0.328,
+      "step": 2611
+    },
+    {
+      "epoch": 0.797192125743934,
+      "grad_norm": 1.5193173765477481,
+      "learning_rate": 1.0396380554960255e-06,
+      "loss": 0.2889,
+      "step": 2612
+    },
+    {
+      "epoch": 0.7974973294674195,
+      "grad_norm": 1.604876629493802,
+      "learning_rate": 1.0366218721712983e-06,
+      "loss": 0.3451,
+      "step": 2613
+    },
+    {
+      "epoch": 0.7978025331909049,
+      "grad_norm": 1.5038961831288478,
+      "learning_rate": 1.033609564373767e-06,
+      "loss": 0.2933,
+      "step": 2614
+    },
+    {
+      "epoch": 0.7981077369143904,
+      "grad_norm": 1.467630324807371,
+      "learning_rate": 1.0306011350489686e-06,
+      "loss": 0.2042,
+      "step": 2615
+    },
+    {
+      "epoch": 0.7984129406378758,
+      "grad_norm": 1.5211283079825,
+      "learning_rate": 1.0275965871386474e-06,
+      "loss": 0.2607,
+      "step": 2616
+    },
+    {
+      "epoch": 0.7987181443613612,
+      "grad_norm": 1.6333004163162896,
+      "learning_rate": 1.024595923580755e-06,
+      "loss": 0.29,
+      "step": 2617
+    },
+    {
+      "epoch": 0.7990233480848467,
+      "grad_norm": 1.6102677619000663,
+      "learning_rate": 1.0215991473094439e-06,
+      "loss": 0.2565,
+      "step": 2618
+    },
+    {
+      "epoch": 0.7993285518083321,
+      "grad_norm": 1.552044332470833,
+      "learning_rate": 1.0186062612550618e-06,
+      "loss": 0.2537,
+      "step": 2619
+    },
+    {
+      "epoch": 0.7996337555318175,
+      "grad_norm": 2.716022745905198,
+      "learning_rate": 1.015617268344159e-06,
+      "loss": 0.4498,
+      "step": 2620
+    },
+    {
+      "epoch": 0.7999389592553029,
+      "grad_norm": 1.854019016287683,
+      "learning_rate": 1.0126321714994741e-06,
+      "loss": 0.5988,
+      "step": 2621
+    },
+    {
+      "epoch": 0.8002441629787883,
+      "grad_norm": 1.546455711798656,
+      "learning_rate": 1.009650973639935e-06,
+      "loss": 0.3001,
+      "step": 2622
+    },
+    {
+      "epoch": 0.8005493667022737,
+      "grad_norm": 1.732303654057111,
+      "learning_rate": 1.006673677680663e-06,
+      "loss": 0.4049,
+      "step": 2623
+    },
+    {
+      "epoch": 0.8008545704257592,
+      "grad_norm": 1.6992859574981316,
+      "learning_rate": 1.0037002865329582e-06,
+      "loss": 0.499,
+      "step": 2624
+    },
+    {
+      "epoch": 0.8011597741492447,
+      "grad_norm": 1.5594763682716168,
+      "learning_rate": 1.0007308031043035e-06,
+      "loss": 0.4562,
+      "step": 2625
+    },
+    {
+      "epoch": 0.8014649778727301,
+      "grad_norm": 1.6350669081877012,
+      "learning_rate": 9.97765230298365e-07,
+      "loss": 0.4341,
+      "step": 2626
+    },
+    {
+      "epoch": 0.8017701815962155,
+      "grad_norm": 1.6414271255756192,
+      "learning_rate": 9.948035710149788e-07,
+      "loss": 0.3881,
+      "step": 2627
+    },
+    {
+      "epoch": 0.8020753853197009,
+      "grad_norm": 1.6181320159630346,
+      "learning_rate": 9.91845828150157e-07,
+      "loss": 0.4622,
+      "step": 2628
+    },
+    {
+      "epoch": 0.8023805890431863,
+      "grad_norm": 1.6530374283842706,
+      "learning_rate": 9.888920045960848e-07,
+      "loss": 0.4549,
+      "step": 2629
+    },
+    {
+      "epoch": 0.8026857927666717,
+      "grad_norm": 1.409891753607114,
+      "learning_rate": 9.859421032411087e-07,
+      "loss": 0.1985,
+      "step": 2630
+    },
+    {
+      "epoch": 0.8029909964901571,
+      "grad_norm": 3.1747841108318884,
+      "learning_rate": 9.82996126969747e-07,
+      "loss": 0.299,
+      "step": 2631
+    },
+    {
+      "epoch": 0.8032962002136426,
+      "grad_norm": 1.8120883605751632,
+      "learning_rate": 9.800540786626733e-07,
+      "loss": 0.5064,
+      "step": 2632
+    },
+    {
+      "epoch": 0.803601403937128,
+      "grad_norm": 1.5712675160696425,
+      "learning_rate": 9.771159611967263e-07,
+      "loss": 0.4155,
+      "step": 2633
+    },
+    {
+      "epoch": 0.8039066076606135,
+      "grad_norm": 1.4720110301920981,
+      "learning_rate": 9.741817774448964e-07,
+      "loss": 0.2089,
+      "step": 2634
+    },
+    {
+      "epoch": 0.8042118113840989,
+      "grad_norm": 1.4869879762975362,
+      "learning_rate": 9.712515302763287e-07,
+      "loss": 0.1924,
+      "step": 2635
+    },
+    {
+      "epoch": 0.8045170151075843,
+      "grad_norm": 1.7374620886309813,
+      "learning_rate": 9.683252225563212e-07,
+      "loss": 0.3555,
+      "step": 2636
+    },
+    {
+      "epoch": 0.8048222188310697,
+      "grad_norm": 1.4234749312771073,
+      "learning_rate": 9.654028571463169e-07,
+      "loss": 0.2412,
+      "step": 2637
+    },
+    {
+      "epoch": 0.8051274225545552,
+      "grad_norm": 1.4787099249497095,
+      "learning_rate": 9.624844369039048e-07,
+      "loss": 0.4092,
+      "step": 2638
+    },
+    {
+      "epoch": 0.8054326262780406,
+      "grad_norm": 1.699621814847962,
+      "learning_rate": 9.595699646828178e-07,
+      "loss": 0.2133,
+      "step": 2639
+    },
+    {
+      "epoch": 0.805737830001526,
+      "grad_norm": 1.5439514121430349,
+      "learning_rate": 9.566594433329257e-07,
+      "loss": 0.2268,
+      "step": 2640
+    },
+    {
+      "epoch": 0.8060430337250114,
+      "grad_norm": 1.7387800208324073,
+      "learning_rate": 9.537528757002357e-07,
+      "loss": 0.3704,
+      "step": 2641
+    },
+    {
+      "epoch": 0.8063482374484968,
+      "grad_norm": 1.8876369603468202,
+      "learning_rate": 9.508502646268908e-07,
+      "loss": 0.2616,
+      "step": 2642
+    },
+    {
+      "epoch": 0.8066534411719823,
+      "grad_norm": 1.3815542412463353,
+      "learning_rate": 9.479516129511635e-07,
+      "loss": 0.1867,
+      "step": 2643
+    },
+    {
+      "epoch": 0.8069586448954678,
+      "grad_norm": 1.4669823256287975,
+      "learning_rate": 9.450569235074531e-07,
+      "loss": 0.3232,
+      "step": 2644
+    },
+    {
+      "epoch": 0.8072638486189532,
+      "grad_norm": 1.6840137214554676,
+      "learning_rate": 9.421661991262893e-07,
+      "loss": 0.3368,
+      "step": 2645
+    },
+    {
+      "epoch": 0.8075690523424386,
+      "grad_norm": 1.7677239963764457,
+      "learning_rate": 9.392794426343189e-07,
+      "loss": 0.4677,
+      "step": 2646
+    },
+    {
+      "epoch": 0.807874256065924,
+      "grad_norm": 1.592833461133337,
+      "learning_rate": 9.36396656854312e-07,
+      "loss": 0.2972,
+      "step": 2647
+    },
+    {
+      "epoch": 0.8081794597894094,
+      "grad_norm": 1.3250285685356402,
+      "learning_rate": 9.335178446051579e-07,
+      "loss": 0.2016,
+      "step": 2648
+    },
+    {
+      "epoch": 0.8084846635128948,
+      "grad_norm": 1.8020969309505135,
+      "learning_rate": 9.306430087018565e-07,
+      "loss": 0.3618,
+      "step": 2649
+    },
+    {
+      "epoch": 0.8087898672363802,
+      "grad_norm": 1.6753558706039484,
+      "learning_rate": 9.277721519555194e-07,
+      "loss": 0.3168,
+      "step": 2650
+    },
+    {
+      "epoch": 0.8090950709598657,
+      "grad_norm": 1.9916679752084858,
+      "learning_rate": 9.249052771733713e-07,
+      "loss": 0.4444,
+      "step": 2651
+    },
+    {
+      "epoch": 0.8094002746833512,
+      "grad_norm": 1.558934683546651,
+      "learning_rate": 9.220423871587403e-07,
+      "loss": 0.373,
+      "step": 2652
+    },
+    {
+      "epoch": 0.8097054784068366,
+      "grad_norm": 1.9602276858604508,
+      "learning_rate": 9.19183484711057e-07,
+      "loss": 0.3209,
+      "step": 2653
+    },
+    {
+      "epoch": 0.810010682130322,
+      "grad_norm": 1.6310920960452386,
+      "learning_rate": 9.163285726258563e-07,
+      "loss": 0.2726,
+      "step": 2654
+    },
+    {
+      "epoch": 0.8103158858538074,
+      "grad_norm": 1.4738232266496092,
+      "learning_rate": 9.134776536947687e-07,
+      "loss": 0.3224,
+      "step": 2655
+    },
+    {
+      "epoch": 0.8106210895772928,
+      "grad_norm": 1.258303528657992,
+      "learning_rate": 9.106307307055201e-07,
+      "loss": 0.2699,
+      "step": 2656
+    },
+    {
+      "epoch": 0.8109262933007783,
+      "grad_norm": 2.01664885310773,
+      "learning_rate": 9.077878064419283e-07,
+      "loss": 0.3761,
+      "step": 2657
+    },
+    {
+      "epoch": 0.8112314970242637,
+      "grad_norm": 2.1735184852545903,
+      "learning_rate": 9.049488836839049e-07,
+      "loss": 0.2646,
+      "step": 2658
+    },
+    {
+      "epoch": 0.8115367007477491,
+      "grad_norm": 1.2358409048111307,
+      "learning_rate": 9.021139652074451e-07,
+      "loss": 0.1907,
+      "step": 2659
+    },
+    {
+      "epoch": 0.8118419044712345,
+      "grad_norm": 1.9648143849539013,
+      "learning_rate": 8.992830537846275e-07,
+      "loss": 0.4223,
+      "step": 2660
+    },
+    {
+      "epoch": 0.81214710819472,
+      "grad_norm": 1.99769292327075,
+      "learning_rate": 8.964561521836179e-07,
+      "loss": 0.304,
+      "step": 2661
+    },
+    {
+      "epoch": 0.8124523119182054,
+      "grad_norm": 1.4146455858753013,
+      "learning_rate": 8.936332631686545e-07,
+      "loss": 0.1802,
+      "step": 2662
+    },
+    {
+      "epoch": 0.8127575156416909,
+      "grad_norm": 1.703787018785887,
+      "learning_rate": 8.908143895000565e-07,
+      "loss": 0.3692,
+      "step": 2663
+    },
+    {
+      "epoch": 0.8130627193651763,
+      "grad_norm": 1.8186227708079596,
+      "learning_rate": 8.879995339342168e-07,
+      "loss": 0.1969,
+      "step": 2664
+    },
+    {
+      "epoch": 0.8133679230886617,
+      "grad_norm": 1.5891890243827214,
+      "learning_rate": 8.851886992235964e-07,
+      "loss": 0.3539,
+      "step": 2665
+    },
+    {
+      "epoch": 0.8136731268121471,
+      "grad_norm": 1.8341778052946631,
+      "learning_rate": 8.823818881167246e-07,
+      "loss": 0.2046,
+      "step": 2666
+    },
+    {
+      "epoch": 0.8139783305356325,
+      "grad_norm": 1.6571117534443234,
+      "learning_rate": 8.795791033582007e-07,
+      "loss": 0.1426,
+      "step": 2667
+    },
+    {
+      "epoch": 0.8142835342591179,
+      "grad_norm": 1.6438546499147626,
+      "learning_rate": 8.767803476886821e-07,
+      "loss": 0.2573,
+      "step": 2668
+    },
+    {
+      "epoch": 0.8145887379826033,
+      "grad_norm": 1.6007966978669126,
+      "learning_rate": 8.739856238448868e-07,
+      "loss": 0.3303,
+      "step": 2669
+    },
+    {
+      "epoch": 0.8148939417060889,
+      "grad_norm": 1.881959840806181,
+      "learning_rate": 8.711949345595949e-07,
+      "loss": 0.2803,
+      "step": 2670
+    },
+    {
+      "epoch": 0.8151991454295743,
+      "grad_norm": 1.6541498652581024,
+      "learning_rate": 8.684082825616363e-07,
+      "loss": 0.2841,
+      "step": 2671
+    },
+    {
+      "epoch": 0.8155043491530597,
+      "grad_norm": 1.9023082189355414,
+      "learning_rate": 8.656256705758936e-07,
+      "loss": 0.3925,
+      "step": 2672
+    },
+    {
+      "epoch": 0.8158095528765451,
+      "grad_norm": 1.3554873946169075,
+      "learning_rate": 8.628471013233036e-07,
+      "loss": 0.2266,
+      "step": 2673
+    },
+    {
+      "epoch": 0.8161147566000305,
+      "grad_norm": 1.4971356063509114,
+      "learning_rate": 8.600725775208446e-07,
+      "loss": 0.2315,
+      "step": 2674
+    },
+    {
+      "epoch": 0.8164199603235159,
+      "grad_norm": 1.628357280189394,
+      "learning_rate": 8.573021018815409e-07,
+      "loss": 0.3426,
+      "step": 2675
+    },
+    {
+      "epoch": 0.8167251640470014,
+      "grad_norm": 1.588611622088212,
+      "learning_rate": 8.545356771144603e-07,
+      "loss": 0.2657,
+      "step": 2676
+    },
+    {
+      "epoch": 0.8170303677704868,
+      "grad_norm": 1.5671981369981116,
+      "learning_rate": 8.517733059247074e-07,
+      "loss": 0.3451,
+      "step": 2677
+    },
+    {
+      "epoch": 0.8173355714939722,
+      "grad_norm": 1.4950183160482764,
+      "learning_rate": 8.490149910134233e-07,
+      "loss": 0.1883,
+      "step": 2678
+    },
+    {
+      "epoch": 0.8176407752174577,
+      "grad_norm": 1.733033632386829,
+      "learning_rate": 8.462607350777796e-07,
+      "loss": 0.2291,
+      "step": 2679
+    },
+    {
+      "epoch": 0.8179459789409431,
+      "grad_norm": 1.503043980308405,
+      "learning_rate": 8.435105408109889e-07,
+      "loss": 0.2581,
+      "step": 2680
+    },
+    {
+      "epoch": 0.8182511826644285,
+      "grad_norm": 1.518032325124027,
+      "learning_rate": 8.407644109022822e-07,
+      "loss": 0.2657,
+      "step": 2681
+    },
+    {
+      "epoch": 0.818556386387914,
+      "grad_norm": 1.5121513753077411,
+      "learning_rate": 8.380223480369182e-07,
+      "loss": 0.2234,
+      "step": 2682
+    },
+    {
+      "epoch": 0.8188615901113994,
+      "grad_norm": 1.6772203646134276,
+      "learning_rate": 8.352843548961831e-07,
+      "loss": 0.318,
+      "step": 2683
+    },
+    {
+      "epoch": 0.8191667938348848,
+      "grad_norm": 1.541296484896679,
+      "learning_rate": 8.325504341573792e-07,
+      "loss": 0.2169,
+      "step": 2684
+    },
+    {
+      "epoch": 0.8194719975583702,
+      "grad_norm": 1.6917734077991005,
+      "learning_rate": 8.298205884938265e-07,
+      "loss": 0.2167,
+      "step": 2685
+    },
+    {
+      "epoch": 0.8197772012818556,
+      "grad_norm": 1.5189316848902874,
+      "learning_rate": 8.270948205748642e-07,
+      "loss": 0.3513,
+      "step": 2686
+    },
+    {
+      "epoch": 0.820082405005341,
+      "grad_norm": 1.4605678151811097,
+      "learning_rate": 8.243731330658411e-07,
+      "loss": 0.2117,
+      "step": 2687
+    },
+    {
+      "epoch": 0.8203876087288265,
+      "grad_norm": 1.6154899348869585,
+      "learning_rate": 8.216555286281153e-07,
+      "loss": 0.221,
+      "step": 2688
+    },
+    {
+      "epoch": 0.820692812452312,
+      "grad_norm": 1.513179260818861,
+      "learning_rate": 8.189420099190559e-07,
+      "loss": 0.2856,
+      "step": 2689
+    },
+    {
+      "epoch": 0.8209980161757974,
+      "grad_norm": 1.922241858227103,
+      "learning_rate": 8.162325795920334e-07,
+      "loss": 0.4977,
+      "step": 2690
+    },
+    {
+      "epoch": 0.8213032198992828,
+      "grad_norm": 1.6229844431340152,
+      "learning_rate": 8.135272402964211e-07,
+      "loss": 0.2974,
+      "step": 2691
+    },
+    {
+      "epoch": 0.8216084236227682,
+      "grad_norm": 2.1257005546532133,
+      "learning_rate": 8.108259946775943e-07,
+      "loss": 0.4415,
+      "step": 2692
+    },
+    {
+      "epoch": 0.8219136273462536,
+      "grad_norm": 1.654955231905261,
+      "learning_rate": 8.081288453769226e-07,
+      "loss": 0.3712,
+      "step": 2693
+    },
+    {
+      "epoch": 0.822218831069739,
+      "grad_norm": 1.4861717994079104,
+      "learning_rate": 8.054357950317709e-07,
+      "loss": 0.2555,
+      "step": 2694
+    },
+    {
+      "epoch": 0.8225240347932244,
+      "grad_norm": 1.8518695174363622,
+      "learning_rate": 8.027468462754967e-07,
+      "loss": 0.3779,
+      "step": 2695
+    },
+    {
+      "epoch": 0.8228292385167099,
+      "grad_norm": 1.7312777933582089,
+      "learning_rate": 8.000620017374472e-07,
+      "loss": 0.3539,
+      "step": 2696
+    },
+    {
+      "epoch": 0.8231344422401954,
+      "grad_norm": 1.2929073445931385,
+      "learning_rate": 7.973812640429557e-07,
+      "loss": 0.1687,
+      "step": 2697
+    },
+    {
+      "epoch": 0.8234396459636808,
+      "grad_norm": 1.7167304919084254,
+      "learning_rate": 7.94704635813337e-07,
+      "loss": 0.3658,
+      "step": 2698
+    },
+    {
+      "epoch": 0.8237448496871662,
+      "grad_norm": 1.4959828308262608,
+      "learning_rate": 7.920321196658931e-07,
+      "loss": 0.3476,
+      "step": 2699
+    },
+    {
+      "epoch": 0.8240500534106516,
+      "grad_norm": 1.3697764271322341,
+      "learning_rate": 7.893637182139002e-07,
+      "loss": 0.2692,
+      "step": 2700
+    },
+    {
+      "epoch": 0.824355257134137,
+      "grad_norm": 1.5335971204552559,
+      "learning_rate": 7.866994340666129e-07,
+      "loss": 0.2137,
+      "step": 2701
+    },
+    {
+      "epoch": 0.8246604608576225,
+      "grad_norm": 1.657534119497938,
+      "learning_rate": 7.840392698292615e-07,
+      "loss": 0.371,
+      "step": 2702
+    },
+    {
+      "epoch": 0.8249656645811079,
+      "grad_norm": 1.8020708013844835,
+      "learning_rate": 7.813832281030442e-07,
+      "loss": 0.3245,
+      "step": 2703
+    },
+    {
+      "epoch": 0.8252708683045933,
+      "grad_norm": 1.6667883669607235,
+      "learning_rate": 7.787313114851292e-07,
+      "loss": 0.2716,
+      "step": 2704
+    },
+    {
+      "epoch": 0.8255760720280787,
+      "grad_norm": 1.2478884983014769,
+      "learning_rate": 7.760835225686536e-07,
+      "loss": 0.2216,
+      "step": 2705
+    },
+    {
+      "epoch": 0.8258812757515641,
+      "grad_norm": 1.5816860999845315,
+      "learning_rate": 7.734398639427154e-07,
+      "loss": 0.2371,
+      "step": 2706
+    },
+    {
+      "epoch": 0.8261864794750496,
+      "grad_norm": 1.5788263707479415,
+      "learning_rate": 7.708003381923734e-07,
+      "loss": 0.3195,
+      "step": 2707
+    },
+    {
+      "epoch": 0.826491683198535,
+      "grad_norm": 1.4651010516210756,
+      "learning_rate": 7.681649478986486e-07,
+      "loss": 0.3439,
+      "step": 2708
+    },
+    {
+      "epoch": 0.8267968869220205,
+      "grad_norm": 1.6232033113632907,
+      "learning_rate": 7.655336956385156e-07,
+      "loss": 0.2262,
+      "step": 2709
+    },
+    {
+      "epoch": 0.8271020906455059,
+      "grad_norm": 1.7164045108999924,
+      "learning_rate": 7.629065839849015e-07,
+      "loss": 0.2985,
+      "step": 2710
+    },
+    {
+      "epoch": 0.8274072943689913,
+      "grad_norm": 1.6014691581507452,
+      "learning_rate": 7.60283615506689e-07,
+      "loss": 0.4344,
+      "step": 2711
+    },
+    {
+      "epoch": 0.8277124980924767,
+      "grad_norm": 1.443248285088709,
+      "learning_rate": 7.576647927687041e-07,
+      "loss": 0.3966,
+      "step": 2712
+    },
+    {
+      "epoch": 0.8280177018159621,
+      "grad_norm": 1.4925582984260413,
+      "learning_rate": 7.550501183317238e-07,
+      "loss": 0.2209,
+      "step": 2713
+    },
+    {
+      "epoch": 0.8283229055394475,
+      "grad_norm": 1.6487842936271624,
+      "learning_rate": 7.524395947524648e-07,
+      "loss": 0.4352,
+      "step": 2714
+    },
+    {
+      "epoch": 0.828628109262933,
+      "grad_norm": 1.3713077041321657,
+      "learning_rate": 7.498332245835882e-07,
+      "loss": 0.3803,
+      "step": 2715
+    },
+    {
+      "epoch": 0.8289333129864185,
+      "grad_norm": 1.7482914076955032,
+      "learning_rate": 7.472310103736913e-07,
+      "loss": 0.4693,
+      "step": 2716
+    },
+    {
+      "epoch": 0.8292385167099039,
+      "grad_norm": 1.550940671168507,
+      "learning_rate": 7.446329546673075e-07,
+      "loss": 0.2152,
+      "step": 2717
+    },
+    {
+      "epoch": 0.8295437204333893,
+      "grad_norm": 1.4576270800196935,
+      "learning_rate": 7.420390600049077e-07,
+      "loss": 0.3354,
+      "step": 2718
+    },
+    {
+      "epoch": 0.8298489241568747,
+      "grad_norm": 1.8982575884682047,
+      "learning_rate": 7.394493289228887e-07,
+      "loss": 0.2404,
+      "step": 2719
+    },
+    {
+      "epoch": 0.8301541278803601,
+      "grad_norm": 1.691515066685838,
+      "learning_rate": 7.368637639535786e-07,
+      "loss": 0.4312,
+      "step": 2720
+    },
+    {
+      "epoch": 0.8304593316038456,
+      "grad_norm": 1.4694249651618714,
+      "learning_rate": 7.342823676252331e-07,
+      "loss": 0.2509,
+      "step": 2721
+    },
+    {
+      "epoch": 0.830764535327331,
+      "grad_norm": 1.546455249285673,
+      "learning_rate": 7.317051424620292e-07,
+      "loss": 0.2546,
+      "step": 2722
+    },
+    {
+      "epoch": 0.8310697390508164,
+      "grad_norm": 1.7556179334661928,
+      "learning_rate": 7.291320909840649e-07,
+      "loss": 0.371,
+      "step": 2723
+    },
+    {
+      "epoch": 0.8313749427743018,
+      "grad_norm": 1.8313804397736553,
+      "learning_rate": 7.265632157073604e-07,
+      "loss": 0.3004,
+      "step": 2724
+    },
+    {
+      "epoch": 0.8316801464977873,
+      "grad_norm": 1.5378078245585296,
+      "learning_rate": 7.239985191438498e-07,
+      "loss": 0.2243,
+      "step": 2725
+    },
+    {
+      "epoch": 0.8319853502212727,
+      "grad_norm": 1.5004354480661528,
+      "learning_rate": 7.214380038013797e-07,
+      "loss": 0.2521,
+      "step": 2726
+    },
+    {
+      "epoch": 0.8322905539447581,
+      "grad_norm": 1.7167295197520305,
+      "learning_rate": 7.188816721837116e-07,
+      "loss": 0.3701,
+      "step": 2727
+    },
+    {
+      "epoch": 0.8325957576682436,
+      "grad_norm": 1.4893673915884809,
+      "learning_rate": 7.163295267905135e-07,
+      "loss": 0.4652,
+      "step": 2728
+    },
+    {
+      "epoch": 0.832900961391729,
+      "grad_norm": 1.9899160562954143,
+      "learning_rate": 7.137815701173617e-07,
+      "loss": 0.3553,
+      "step": 2729
+    },
+    {
+      "epoch": 0.8332061651152144,
+      "grad_norm": 1.9035527893233544,
+      "learning_rate": 7.112378046557339e-07,
+      "loss": 0.2428,
+      "step": 2730
+    },
+    {
+      "epoch": 0.8335113688386998,
+      "grad_norm": 1.7219634100895571,
+      "learning_rate": 7.086982328930142e-07,
+      "loss": 0.4137,
+      "step": 2731
+    },
+    {
+      "epoch": 0.8338165725621852,
+      "grad_norm": 1.6987942570976622,
+      "learning_rate": 7.061628573124807e-07,
+      "loss": 0.3738,
+      "step": 2732
+    },
+    {
+      "epoch": 0.8341217762856706,
+      "grad_norm": 1.5903429002305511,
+      "learning_rate": 7.036316803933107e-07,
+      "loss": 0.4614,
+      "step": 2733
+    },
+    {
+      "epoch": 0.8344269800091562,
+      "grad_norm": 1.5979300639731906,
+      "learning_rate": 7.011047046105773e-07,
+      "loss": 0.3753,
+      "step": 2734
+    },
+    {
+      "epoch": 0.8347321837326416,
+      "grad_norm": 1.4488125049184104,
+      "learning_rate": 6.985819324352438e-07,
+      "loss": 0.1889,
+      "step": 2735
+    },
+    {
+      "epoch": 0.835037387456127,
+      "grad_norm": 1.6759672617434724,
+      "learning_rate": 6.960633663341615e-07,
+      "loss": 0.1517,
+      "step": 2736
+    },
+    {
+      "epoch": 0.8353425911796124,
+      "grad_norm": 1.625179574287605,
+      "learning_rate": 6.935490087700736e-07,
+      "loss": 0.3266,
+      "step": 2737
+    },
+    {
+      "epoch": 0.8356477949030978,
+      "grad_norm": 1.4067868373637953,
+      "learning_rate": 6.91038862201604e-07,
+      "loss": 0.1633,
+      "step": 2738
+    },
+    {
+      "epoch": 0.8359529986265832,
+      "grad_norm": 1.8386286065687836,
+      "learning_rate": 6.885329290832593e-07,
+      "loss": 0.2931,
+      "step": 2739
+    },
+    {
+      "epoch": 0.8362582023500686,
+      "grad_norm": 1.6406484693029308,
+      "learning_rate": 6.860312118654283e-07,
+      "loss": 0.4988,
+      "step": 2740
+    },
+    {
+      "epoch": 0.8365634060735541,
+      "grad_norm": 1.3974679324270254,
+      "learning_rate": 6.835337129943759e-07,
+      "loss": 0.2291,
+      "step": 2741
+    },
+    {
+      "epoch": 0.8368686097970395,
+      "grad_norm": 1.683103485593189,
+      "learning_rate": 6.810404349122407e-07,
+      "loss": 0.3066,
+      "step": 2742
+    },
+    {
+      "epoch": 0.837173813520525,
+      "grad_norm": 1.6876120353524549,
+      "learning_rate": 6.785513800570376e-07,
+      "loss": 0.4332,
+      "step": 2743
+    },
+    {
+      "epoch": 0.8374790172440104,
+      "grad_norm": 1.5257702305049623,
+      "learning_rate": 6.76066550862648e-07,
+      "loss": 0.1981,
+      "step": 2744
+    },
+    {
+      "epoch": 0.8377842209674958,
+      "grad_norm": 1.6135370453517366,
+      "learning_rate": 6.735859497588254e-07,
+      "loss": 0.1967,
+      "step": 2745
+    },
+    {
+      "epoch": 0.8380894246909812,
+      "grad_norm": 1.7321804011662434,
+      "learning_rate": 6.711095791711841e-07,
+      "loss": 0.2733,
+      "step": 2746
+    },
+    {
+      "epoch": 0.8383946284144667,
+      "grad_norm": 1.8118392792385045,
+      "learning_rate": 6.686374415212066e-07,
+      "loss": 0.2028,
+      "step": 2747
+    },
+    {
+      "epoch": 0.8386998321379521,
+      "grad_norm": 2.0445068871019902,
+      "learning_rate": 6.661695392262335e-07,
+      "loss": 0.4171,
+      "step": 2748
+    },
+    {
+      "epoch": 0.8390050358614375,
+      "grad_norm": 2.0394915717592106,
+      "learning_rate": 6.637058746994629e-07,
+      "loss": 0.3445,
+      "step": 2749
+    },
+    {
+      "epoch": 0.8393102395849229,
+      "grad_norm": 1.8957888489698793,
+      "learning_rate": 6.612464503499521e-07,
+      "loss": 0.4731,
+      "step": 2750
+    },
+    {
+      "epoch": 0.8396154433084083,
+      "grad_norm": 4.581169704342479,
+      "learning_rate": 6.587912685826104e-07,
+      "loss": 0.3414,
+      "step": 2751
+    },
+    {
+      "epoch": 0.8399206470318938,
+      "grad_norm": 1.8018825911654097,
+      "learning_rate": 6.563403317981975e-07,
+      "loss": 0.1865,
+      "step": 2752
+    },
+    {
+      "epoch": 0.8402258507553793,
+      "grad_norm": 1.8405863889002556,
+      "learning_rate": 6.538936423933257e-07,
+      "loss": 0.4019,
+      "step": 2753
+    },
+    {
+      "epoch": 0.8405310544788647,
+      "grad_norm": 1.3557187600114369,
+      "learning_rate": 6.514512027604508e-07,
+      "loss": 0.2392,
+      "step": 2754
+    },
+    {
+      "epoch": 0.8408362582023501,
+      "grad_norm": 1.6872412871795575,
+      "learning_rate": 6.490130152878727e-07,
+      "loss": 0.4288,
+      "step": 2755
+    },
+    {
+      "epoch": 0.8411414619258355,
+      "grad_norm": 1.643255033052677,
+      "learning_rate": 6.465790823597373e-07,
+      "loss": 0.4054,
+      "step": 2756
+    },
+    {
+      "epoch": 0.8414466656493209,
+      "grad_norm": 1.7998639002315395,
+      "learning_rate": 6.441494063560272e-07,
+      "loss": 0.4727,
+      "step": 2757
+    },
+    {
+      "epoch": 0.8417518693728063,
+      "grad_norm": 1.5893664944174015,
+      "learning_rate": 6.417239896525607e-07,
+      "loss": 0.2931,
+      "step": 2758
+    },
+    {
+      "epoch": 0.8420570730962917,
+      "grad_norm": 2.1674995533239834,
+      "learning_rate": 6.393028346209968e-07,
+      "loss": 0.376,
+      "step": 2759
+    },
+    {
+      "epoch": 0.8423622768197772,
+      "grad_norm": 3.081440490653117,
+      "learning_rate": 6.368859436288227e-07,
+      "loss": 0.4837,
+      "step": 2760
+    },
+    {
+      "epoch": 0.8426674805432627,
+      "grad_norm": 1.4416792282850845,
+      "learning_rate": 6.344733190393554e-07,
+      "loss": 0.2417,
+      "step": 2761
+    },
+    {
+      "epoch": 0.8429726842667481,
+      "grad_norm": 1.525500656242817,
+      "learning_rate": 6.320649632117443e-07,
+      "loss": 0.219,
+      "step": 2762
+    },
+    {
+      "epoch": 0.8432778879902335,
+      "grad_norm": 1.6337898651139318,
+      "learning_rate": 6.296608785009634e-07,
+      "loss": 0.437,
+      "step": 2763
+    },
+    {
+      "epoch": 0.8435830917137189,
+      "grad_norm": 1.6056735982404093,
+      "learning_rate": 6.272610672578072e-07,
+      "loss": 0.3207,
+      "step": 2764
+    },
+    {
+      "epoch": 0.8438882954372043,
+      "grad_norm": 1.7759983210340933,
+      "learning_rate": 6.248655318288932e-07,
+      "loss": 0.3323,
+      "step": 2765
+    },
+    {
+      "epoch": 0.8441934991606898,
+      "grad_norm": 1.433669667741416,
+      "learning_rate": 6.224742745566603e-07,
+      "loss": 0.2073,
+      "step": 2766
+    },
+    {
+      "epoch": 0.8444987028841752,
+      "grad_norm": 1.7109539431822582,
+      "learning_rate": 6.200872977793604e-07,
+      "loss": 0.4429,
+      "step": 2767
+    },
+    {
+      "epoch": 0.8448039066076606,
+      "grad_norm": 1.5934193586060046,
+      "learning_rate": 6.177046038310608e-07,
+      "loss": 0.3145,
+      "step": 2768
+    },
+    {
+      "epoch": 0.845109110331146,
+      "grad_norm": 1.6459467724042265,
+      "learning_rate": 6.153261950416433e-07,
+      "loss": 0.2244,
+      "step": 2769
+    },
+    {
+      "epoch": 0.8454143140546315,
+      "grad_norm": 2.6281456945377912,
+      "learning_rate": 6.129520737367972e-07,
+      "loss": 0.2321,
+      "step": 2770
+    },
+    {
+      "epoch": 0.8457195177781169,
+      "grad_norm": 1.1481705958782344,
+      "learning_rate": 6.105822422380181e-07,
+      "loss": 0.1813,
+      "step": 2771
+    },
+    {
+      "epoch": 0.8460247215016024,
+      "grad_norm": 2.063357319525523,
+      "learning_rate": 6.082167028626102e-07,
+      "loss": 0.2958,
+      "step": 2772
+    },
+    {
+      "epoch": 0.8463299252250878,
+      "grad_norm": 1.5858495528166288,
+      "learning_rate": 6.058554579236791e-07,
+      "loss": 0.228,
+      "step": 2773
+    },
+    {
+      "epoch": 0.8466351289485732,
+      "grad_norm": 1.5093068046249114,
+      "learning_rate": 6.034985097301299e-07,
+      "loss": 0.3244,
+      "step": 2774
+    },
+    {
+      "epoch": 0.8469403326720586,
+      "grad_norm": 1.69358407084056,
+      "learning_rate": 6.011458605866694e-07,
+      "loss": 0.4273,
+      "step": 2775
+    },
+    {
+      "epoch": 0.847245536395544,
+      "grad_norm": 1.5431909328757925,
+      "learning_rate": 5.987975127937973e-07,
+      "loss": 0.459,
+      "step": 2776
+    },
+    {
+      "epoch": 0.8475507401190294,
+      "grad_norm": 1.510231015426264,
+      "learning_rate": 5.964534686478074e-07,
+      "loss": 0.1519,
+      "step": 2777
+    },
+    {
+      "epoch": 0.8478559438425148,
+      "grad_norm": 1.6896869475566794,
+      "learning_rate": 5.941137304407901e-07,
+      "loss": 0.2564,
+      "step": 2778
+    },
+    {
+      "epoch": 0.8481611475660003,
+      "grad_norm": 1.7422526394422695,
+      "learning_rate": 5.917783004606192e-07,
+      "loss": 0.2066,
+      "step": 2779
+    },
+    {
+      "epoch": 0.8484663512894858,
+      "grad_norm": 1.9404947455240935,
+      "learning_rate": 5.894471809909585e-07,
+      "loss": 0.2178,
+      "step": 2780
+    },
+    {
+      "epoch": 0.8487715550129712,
+      "grad_norm": 1.551273604097384,
+      "learning_rate": 5.87120374311258e-07,
+      "loss": 0.3228,
+      "step": 2781
+    },
+    {
+      "epoch": 0.8490767587364566,
+      "grad_norm": 1.492001906432092,
+      "learning_rate": 5.847978826967482e-07,
+      "loss": 0.3276,
+      "step": 2782
+    },
+    {
+      "epoch": 0.849381962459942,
+      "grad_norm": 2.0452200918982286,
+      "learning_rate": 5.824797084184409e-07,
+      "loss": 0.127,
+      "step": 2783
+    },
+    {
+      "epoch": 0.8496871661834274,
+      "grad_norm": 1.866882235420747,
+      "learning_rate": 5.801658537431259e-07,
+      "loss": 0.392,
+      "step": 2784
+    },
+    {
+      "epoch": 0.8499923699069128,
+      "grad_norm": 1.5550305165427303,
+      "learning_rate": 5.778563209333715e-07,
+      "loss": 0.3541,
+      "step": 2785
+    },
+    {
+      "epoch": 0.8502975736303983,
+      "grad_norm": 1.5586397169022426,
+      "learning_rate": 5.755511122475172e-07,
+      "loss": 0.255,
+      "step": 2786
+    },
+    {
+      "epoch": 0.8506027773538837,
+      "grad_norm": 1.7528566478080552,
+      "learning_rate": 5.732502299396742e-07,
+      "loss": 0.2115,
+      "step": 2787
+    },
+    {
+      "epoch": 0.8509079810773691,
+      "grad_norm": 2.717382152707819,
+      "learning_rate": 5.709536762597262e-07,
+      "loss": 0.2892,
+      "step": 2788
+    },
+    {
+      "epoch": 0.8512131848008546,
+      "grad_norm": 1.9158547243459207,
+      "learning_rate": 5.686614534533213e-07,
+      "loss": 0.2301,
+      "step": 2789
+    },
+    {
+      "epoch": 0.85151838852434,
+      "grad_norm": 1.4491063817306546,
+      "learning_rate": 5.663735637618728e-07,
+      "loss": 0.1835,
+      "step": 2790
+    },
+    {
+      "epoch": 0.8518235922478254,
+      "grad_norm": 1.4648970530666563,
+      "learning_rate": 5.640900094225593e-07,
+      "loss": 0.357,
+      "step": 2791
+    },
+    {
+      "epoch": 0.8521287959713109,
+      "grad_norm": 1.7272508034615215,
+      "learning_rate": 5.61810792668318e-07,
+      "loss": 0.218,
+      "step": 2792
+    },
+    {
+      "epoch": 0.8524339996947963,
+      "grad_norm": 2.0758367574819987,
+      "learning_rate": 5.595359157278435e-07,
+      "loss": 0.4254,
+      "step": 2793
+    },
+    {
+      "epoch": 0.8527392034182817,
+      "grad_norm": 1.5052228440725954,
+      "learning_rate": 5.572653808255918e-07,
+      "loss": 0.3254,
+      "step": 2794
+    },
+    {
+      "epoch": 0.8530444071417671,
+      "grad_norm": 1.9500607065385915,
+      "learning_rate": 5.549991901817686e-07,
+      "loss": 0.2944,
+      "step": 2795
+    },
+    {
+      "epoch": 0.8533496108652525,
+      "grad_norm": 1.445294024375155,
+      "learning_rate": 5.527373460123314e-07,
+      "loss": 0.2834,
+      "step": 2796
+    },
+    {
+      "epoch": 0.8536548145887379,
+      "grad_norm": 1.546931256128644,
+      "learning_rate": 5.504798505289916e-07,
+      "loss": 0.3251,
+      "step": 2797
+    },
+    {
+      "epoch": 0.8539600183122235,
+      "grad_norm": 1.7930685443573522,
+      "learning_rate": 5.482267059392049e-07,
+      "loss": 0.2334,
+      "step": 2798
+    },
+    {
+      "epoch": 0.8542652220357089,
+      "grad_norm": 1.4881243285105927,
+      "learning_rate": 5.459779144461713e-07,
+      "loss": 0.3803,
+      "step": 2799
+    },
+    {
+      "epoch": 0.8545704257591943,
+      "grad_norm": 1.9628258841546715,
+      "learning_rate": 5.437334782488396e-07,
+      "loss": 0.3018,
+      "step": 2800
+    },
+    {
+      "epoch": 0.8548756294826797,
+      "grad_norm": 1.6535321374289942,
+      "learning_rate": 5.414933995418947e-07,
+      "loss": 0.324,
+      "step": 2801
+    },
+    {
+      "epoch": 0.8551808332061651,
+      "grad_norm": 2.0209536827929493,
+      "learning_rate": 5.392576805157617e-07,
+      "loss": 0.3422,
+      "step": 2802
+    },
+    {
+      "epoch": 0.8554860369296505,
+      "grad_norm": 1.6412923999431686,
+      "learning_rate": 5.370263233566053e-07,
+      "loss": 0.3786,
+      "step": 2803
+    },
+    {
+      "epoch": 0.8557912406531359,
+      "grad_norm": 1.9568745746447607,
+      "learning_rate": 5.347993302463228e-07,
+      "loss": 0.3732,
+      "step": 2804
+    },
+    {
+      "epoch": 0.8560964443766214,
+      "grad_norm": 1.6531349253942424,
+      "learning_rate": 5.325767033625429e-07,
+      "loss": 0.332,
+      "step": 2805
+    },
+    {
+      "epoch": 0.8564016481001068,
+      "grad_norm": 1.479312979039126,
+      "learning_rate": 5.303584448786286e-07,
+      "loss": 0.2376,
+      "step": 2806
+    },
+    {
+      "epoch": 0.8567068518235923,
+      "grad_norm": 1.6752264353727613,
+      "learning_rate": 5.281445569636684e-07,
+      "loss": 0.207,
+      "step": 2807
+    },
+    {
+      "epoch": 0.8570120555470777,
+      "grad_norm": 1.7475890854815774,
+      "learning_rate": 5.259350417824777e-07,
+      "loss": 0.3528,
+      "step": 2808
+    },
+    {
+      "epoch": 0.8573172592705631,
+      "grad_norm": 1.6438561727950982,
+      "learning_rate": 5.237299014955955e-07,
+      "loss": 0.164,
+      "step": 2809
+    },
+    {
+      "epoch": 0.8576224629940485,
+      "grad_norm": 1.9847062202427255,
+      "learning_rate": 5.215291382592841e-07,
+      "loss": 0.218,
+      "step": 2810
+    },
+    {
+      "epoch": 0.857927666717534,
+      "grad_norm": 2.064912252339862,
+      "learning_rate": 5.193327542255266e-07,
+      "loss": 0.3854,
+      "step": 2811
+    },
+    {
+      "epoch": 0.8582328704410194,
+      "grad_norm": 1.8378487236605714,
+      "learning_rate": 5.171407515420213e-07,
+      "loss": 0.3145,
+      "step": 2812
+    },
+    {
+      "epoch": 0.8585380741645048,
+      "grad_norm": 1.5437728911027,
+      "learning_rate": 5.14953132352185e-07,
+      "loss": 0.1919,
+      "step": 2813
+    },
+    {
+      "epoch": 0.8588432778879902,
+      "grad_norm": 1.9243326987424028,
+      "learning_rate": 5.127698987951457e-07,
+      "loss": 0.3098,
+      "step": 2814
+    },
+    {
+      "epoch": 0.8591484816114756,
+      "grad_norm": 1.6447618916260383,
+      "learning_rate": 5.105910530057439e-07,
+      "loss": 0.227,
+      "step": 2815
+    },
+    {
+      "epoch": 0.8594536853349611,
+      "grad_norm": 1.622402609560446,
+      "learning_rate": 5.084165971145316e-07,
+      "loss": 0.1718,
+      "step": 2816
+    },
+    {
+      "epoch": 0.8597588890584466,
+      "grad_norm": 1.7108956945877176,
+      "learning_rate": 5.06246533247765e-07,
+      "loss": 0.3257,
+      "step": 2817
+    },
+    {
+      "epoch": 0.860064092781932,
+      "grad_norm": 1.89074485966348,
+      "learning_rate": 5.040808635274063e-07,
+      "loss": 0.3174,
+      "step": 2818
+    },
+    {
+      "epoch": 0.8603692965054174,
+      "grad_norm": 2.2435927822106434,
+      "learning_rate": 5.019195900711233e-07,
+      "loss": 0.214,
+      "step": 2819
+    },
+    {
+      "epoch": 0.8606745002289028,
+      "grad_norm": 2.2042387183178556,
+      "learning_rate": 4.997627149922829e-07,
+      "loss": 0.3325,
+      "step": 2820
+    },
+    {
+      "epoch": 0.8609797039523882,
+      "grad_norm": 2.001996950739807,
+      "learning_rate": 4.976102403999511e-07,
+      "loss": 0.362,
+      "step": 2821
+    },
+    {
+      "epoch": 0.8612849076758736,
+      "grad_norm": 1.688043577454316,
+      "learning_rate": 4.954621683988919e-07,
+      "loss": 0.3231,
+      "step": 2822
+    },
+    {
+      "epoch": 0.861590111399359,
+      "grad_norm": 1.8101403233630515,
+      "learning_rate": 4.933185010895641e-07,
+      "loss": 0.3479,
+      "step": 2823
+    },
+    {
+      "epoch": 0.8618953151228445,
+      "grad_norm": 1.6240100779760942,
+      "learning_rate": 4.911792405681182e-07,
+      "loss": 0.2984,
+      "step": 2824
+    },
+    {
+      "epoch": 0.86220051884633,
+      "grad_norm": 1.349047354727378,
+      "learning_rate": 4.890443889263974e-07,
+      "loss": 0.2936,
+      "step": 2825
+    },
+    {
+      "epoch": 0.8625057225698154,
+      "grad_norm": 1.9930665713311155,
+      "learning_rate": 4.869139482519325e-07,
+      "loss": 0.2475,
+      "step": 2826
+    },
+    {
+      "epoch": 0.8628109262933008,
+      "grad_norm": 2.0426209483761584,
+      "learning_rate": 4.847879206279421e-07,
+      "loss": 0.3413,
+      "step": 2827
+    },
+    {
+      "epoch": 0.8631161300167862,
+      "grad_norm": 1.6504839447544135,
+      "learning_rate": 4.826663081333283e-07,
+      "loss": 0.291,
+      "step": 2828
+    },
+    {
+      "epoch": 0.8634213337402716,
+      "grad_norm": 1.6082023867071948,
+      "learning_rate": 4.805491128426776e-07,
+      "loss": 0.2673,
+      "step": 2829
+    },
+    {
+      "epoch": 0.863726537463757,
+      "grad_norm": 2.147514006066084,
+      "learning_rate": 4.784363368262557e-07,
+      "loss": 0.4531,
+      "step": 2830
+    },
+    {
+      "epoch": 0.8640317411872425,
+      "grad_norm": 1.4923935193747502,
+      "learning_rate": 4.763279821500061e-07,
+      "loss": 0.2804,
+      "step": 2831
+    },
+    {
+      "epoch": 0.8643369449107279,
+      "grad_norm": 1.5802741948758767,
+      "learning_rate": 4.74224050875553e-07,
+      "loss": 0.298,
+      "step": 2832
+    },
+    {
+      "epoch": 0.8646421486342133,
+      "grad_norm": 1.9097476826521518,
+      "learning_rate": 4.721245450601913e-07,
+      "loss": 0.4915,
+      "step": 2833
+    },
+    {
+      "epoch": 0.8649473523576988,
+      "grad_norm": 1.323127589594884,
+      "learning_rate": 4.7002946675688943e-07,
+      "loss": 0.2669,
+      "step": 2834
+    },
+    {
+      "epoch": 0.8652525560811842,
+      "grad_norm": 1.3578549798744834,
+      "learning_rate": 4.6793881801428796e-07,
+      "loss": 0.3498,
+      "step": 2835
+    },
+    {
+      "epoch": 0.8655577598046696,
+      "grad_norm": 2.281989513356405,
+      "learning_rate": 4.6585260087669394e-07,
+      "loss": 0.2315,
+      "step": 2836
+    },
+    {
+      "epoch": 0.8658629635281551,
+      "grad_norm": 1.5256072414416366,
+      "learning_rate": 4.6377081738408214e-07,
+      "loss": 0.3382,
+      "step": 2837
+    },
+    {
+      "epoch": 0.8661681672516405,
+      "grad_norm": 1.5990002488928738,
+      "learning_rate": 4.6169346957209313e-07,
+      "loss": 0.407,
+      "step": 2838
+    },
+    {
+      "epoch": 0.8664733709751259,
+      "grad_norm": 1.5824569317831327,
+      "learning_rate": 4.596205594720282e-07,
+      "loss": 0.1483,
+      "step": 2839
+    },
+    {
+      "epoch": 0.8667785746986113,
+      "grad_norm": 1.8002191357207018,
+      "learning_rate": 4.5755208911084927e-07,
+      "loss": 0.389,
+      "step": 2840
+    },
+    {
+      "epoch": 0.8670837784220967,
+      "grad_norm": 1.5460094766264687,
+      "learning_rate": 4.554880605111789e-07,
+      "loss": 0.3566,
+      "step": 2841
+    },
+    {
+      "epoch": 0.8673889821455821,
+      "grad_norm": 1.8934339353984617,
+      "learning_rate": 4.534284756912943e-07,
+      "loss": 0.2947,
+      "step": 2842
+    },
+    {
+      "epoch": 0.8676941858690677,
+      "grad_norm": 1.8798336032520344,
+      "learning_rate": 4.5137333666512794e-07,
+      "loss": 0.562,
+      "step": 2843
+    },
+    {
+      "epoch": 0.8679993895925531,
+      "grad_norm": 1.6991690091273968,
+      "learning_rate": 4.493226454422661e-07,
+      "loss": 0.3847,
+      "step": 2844
+    },
+    {
+      "epoch": 0.8683045933160385,
+      "grad_norm": 1.6244784765393927,
+      "learning_rate": 4.4727640402794536e-07,
+      "loss": 0.381,
+      "step": 2845
+    },
+    {
+      "epoch": 0.8686097970395239,
+      "grad_norm": 1.8197468545946518,
+      "learning_rate": 4.4523461442305016e-07,
+      "loss": 0.4407,
+      "step": 2846
+    },
+    {
+      "epoch": 0.8689150007630093,
+      "grad_norm": 1.7345323920789553,
+      "learning_rate": 4.431972786241118e-07,
+      "loss": 0.3216,
+      "step": 2847
+    },
+    {
+      "epoch": 0.8692202044864947,
+      "grad_norm": 1.4723264289459734,
+      "learning_rate": 4.4116439862330887e-07,
+      "loss": 0.3166,
+      "step": 2848
+    },
+    {
+      "epoch": 0.8695254082099801,
+      "grad_norm": 1.7425234346715797,
+      "learning_rate": 4.3913597640846016e-07,
+      "loss": 0.2847,
+      "step": 2849
+    },
+    {
+      "epoch": 0.8698306119334656,
+      "grad_norm": 1.7469750554837415,
+      "learning_rate": 4.371120139630258e-07,
+      "loss": 0.222,
+      "step": 2850
+    },
+    {
+      "epoch": 0.870135815656951,
+      "grad_norm": 1.5602645141141633,
+      "learning_rate": 4.350925132661077e-07,
+      "loss": 0.3852,
+      "step": 2851
+    },
+    {
+      "epoch": 0.8704410193804364,
+      "grad_norm": 2.6057014418616484,
+      "learning_rate": 4.330774762924417e-07,
+      "loss": 0.364,
+      "step": 2852
+    },
+    {
+      "epoch": 0.8707462231039219,
+      "grad_norm": 1.6420251593666035,
+      "learning_rate": 4.3106690501239945e-07,
+      "loss": 0.2929,
+      "step": 2853
+    },
+    {
+      "epoch": 0.8710514268274073,
+      "grad_norm": 1.6694773656836528,
+      "learning_rate": 4.290608013919889e-07,
+      "loss": 0.2865,
+      "step": 2854
+    },
+    {
+      "epoch": 0.8713566305508927,
+      "grad_norm": 1.5571008860954931,
+      "learning_rate": 4.270591673928448e-07,
+      "loss": 0.2187,
+      "step": 2855
+    },
+    {
+      "epoch": 0.8716618342743782,
+      "grad_norm": 1.8049190752481343,
+      "learning_rate": 4.250620049722337e-07,
+      "loss": 0.2377,
+      "step": 2856
+    },
+    {
+      "epoch": 0.8719670379978636,
+      "grad_norm": 1.2194201265448192,
+      "learning_rate": 4.2306931608305126e-07,
+      "loss": 0.196,
+      "step": 2857
+    },
+    {
+      "epoch": 0.872272241721349,
+      "grad_norm": 1.5577095602262914,
+      "learning_rate": 4.2108110267381596e-07,
+      "loss": 0.3266,
+      "step": 2858
+    },
+    {
+      "epoch": 0.8725774454448344,
+      "grad_norm": 1.8261355799367787,
+      "learning_rate": 4.190973666886705e-07,
+      "loss": 0.4057,
+      "step": 2859
+    },
+    {
+      "epoch": 0.8728826491683198,
+      "grad_norm": 1.7034846634774858,
+      "learning_rate": 4.17118110067381e-07,
+      "loss": 0.272,
+      "step": 2860
+    },
+    {
+      "epoch": 0.8731878528918052,
+      "grad_norm": 1.8111684446224605,
+      "learning_rate": 4.1514333474533263e-07,
+      "loss": 0.2453,
+      "step": 2861
+    },
+    {
+      "epoch": 0.8734930566152908,
+      "grad_norm": 1.5276556062497046,
+      "learning_rate": 4.1317304265352855e-07,
+      "loss": 0.3509,
+      "step": 2862
+    },
+    {
+      "epoch": 0.8737982603387762,
+      "grad_norm": 1.8832433057327023,
+      "learning_rate": 4.1120723571858645e-07,
+      "loss": 0.1538,
+      "step": 2863
+    },
+    {
+      "epoch": 0.8741034640622616,
+      "grad_norm": 1.9218201047919379,
+      "learning_rate": 4.092459158627421e-07,
+      "loss": 0.43,
+      "step": 2864
+    },
+    {
+      "epoch": 0.874408667785747,
+      "grad_norm": 1.895909765470433,
+      "learning_rate": 4.0728908500384024e-07,
+      "loss": 0.3998,
+      "step": 2865
+    },
+    {
+      "epoch": 0.8747138715092324,
+      "grad_norm": 1.494713208557475,
+      "learning_rate": 4.053367450553364e-07,
+      "loss": 0.2971,
+      "step": 2866
+    },
+    {
+      "epoch": 0.8750190752327178,
+      "grad_norm": 1.8677719031637467,
+      "learning_rate": 4.033888979262973e-07,
+      "loss": 0.3805,
+      "step": 2867
+    },
+    {
+      "epoch": 0.8753242789562032,
+      "grad_norm": 1.849333697261009,
+      "learning_rate": 4.014455455213939e-07,
+      "loss": 0.262,
+      "step": 2868
+    },
+    {
+      "epoch": 0.8756294826796887,
+      "grad_norm": 1.534919079812837,
+      "learning_rate": 3.995066897409022e-07,
+      "loss": 0.2033,
+      "step": 2869
+    },
+    {
+      "epoch": 0.8759346864031741,
+      "grad_norm": 1.834784872349212,
+      "learning_rate": 3.9757233248070403e-07,
+      "loss": 0.4,
+      "step": 2870
+    },
+    {
+      "epoch": 0.8762398901266596,
+      "grad_norm": 1.9071103015159971,
+      "learning_rate": 3.9564247563227854e-07,
+      "loss": 0.1831,
+      "step": 2871
+    },
+    {
+      "epoch": 0.876545093850145,
+      "grad_norm": 1.592166637699512,
+      "learning_rate": 3.9371712108270567e-07,
+      "loss": 0.2476,
+      "step": 2872
+    },
+    {
+      "epoch": 0.8768502975736304,
+      "grad_norm": 2.0964013009468907,
+      "learning_rate": 3.9179627071466443e-07,
+      "loss": 0.6438,
+      "step": 2873
+    },
+    {
+      "epoch": 0.8771555012971158,
+      "grad_norm": 1.5601353872677923,
+      "learning_rate": 3.898799264064279e-07,
+      "loss": 0.3256,
+      "step": 2874
+    },
+    {
+      "epoch": 0.8774607050206013,
+      "grad_norm": 2.2184170755256747,
+      "learning_rate": 3.8796809003186274e-07,
+      "loss": 0.3217,
+      "step": 2875
+    },
+    {
+      "epoch": 0.8777659087440867,
+      "grad_norm": 1.5778596862760628,
+      "learning_rate": 3.8606076346042855e-07,
+      "loss": 0.2969,
+      "step": 2876
+    },
+    {
+      "epoch": 0.8780711124675721,
+      "grad_norm": 1.5917046081388069,
+      "learning_rate": 3.8415794855717556e-07,
+      "loss": 0.4407,
+      "step": 2877
+    },
+    {
+      "epoch": 0.8783763161910575,
+      "grad_norm": 1.6609436322132212,
+      "learning_rate": 3.8225964718274114e-07,
+      "loss": 0.3823,
+      "step": 2878
+    },
+    {
+      "epoch": 0.8786815199145429,
+      "grad_norm": 1.6035716782319969,
+      "learning_rate": 3.8036586119334806e-07,
+      "loss": 0.2148,
+      "step": 2879
+    },
+    {
+      "epoch": 0.8789867236380284,
+      "grad_norm": 1.758931121934974,
+      "learning_rate": 3.784765924408074e-07,
+      "loss": 0.3072,
+      "step": 2880
+    },
+    {
+      "epoch": 0.8792919273615138,
+      "grad_norm": 1.7537087921934966,
+      "learning_rate": 3.7659184277251017e-07,
+      "loss": 0.4087,
+      "step": 2881
+    },
+    {
+      "epoch": 0.8795971310849993,
+      "grad_norm": 1.656573066152407,
+      "learning_rate": 3.747116140314283e-07,
+      "loss": 0.2921,
+      "step": 2882
+    },
+    {
+      "epoch": 0.8799023348084847,
+      "grad_norm": 1.6821172129718753,
+      "learning_rate": 3.728359080561156e-07,
+      "loss": 0.3312,
+      "step": 2883
+    },
+    {
+      "epoch": 0.8802075385319701,
+      "grad_norm": 2.029673624908971,
+      "learning_rate": 3.7096472668070095e-07,
+      "loss": 0.3149,
+      "step": 2884
+    },
+    {
+      "epoch": 0.8805127422554555,
+      "grad_norm": 1.4923235446767098,
+      "learning_rate": 3.690980717348885e-07,
+      "loss": 0.4473,
+      "step": 2885
+    },
+    {
+      "epoch": 0.8808179459789409,
+      "grad_norm": 1.884813820020141,
+      "learning_rate": 3.6723594504396054e-07,
+      "loss": 0.2947,
+      "step": 2886
+    },
+    {
+      "epoch": 0.8811231497024263,
+      "grad_norm": 2.0386964900644418,
+      "learning_rate": 3.6537834842876596e-07,
+      "loss": 0.2797,
+      "step": 2887
+    },
+    {
+      "epoch": 0.8814283534259117,
+      "grad_norm": 1.4766396800058812,
+      "learning_rate": 3.635252837057274e-07,
+      "loss": 0.2852,
+      "step": 2888
+    },
+    {
+      "epoch": 0.8817335571493973,
+      "grad_norm": 1.2583942844879477,
+      "learning_rate": 3.616767526868353e-07,
+      "loss": 0.1721,
+      "step": 2889
+    },
+    {
+      "epoch": 0.8820387608728827,
+      "grad_norm": 1.7116168341336353,
+      "learning_rate": 3.598327571796467e-07,
+      "loss": 0.3355,
+      "step": 2890
+    },
+    {
+      "epoch": 0.8823439645963681,
+      "grad_norm": 1.7783296973427543,
+      "learning_rate": 3.5799329898728254e-07,
+      "loss": 0.4913,
+      "step": 2891
+    },
+    {
+      "epoch": 0.8826491683198535,
+      "grad_norm": 1.660842789455405,
+      "learning_rate": 3.5615837990842894e-07,
+      "loss": 0.207,
+      "step": 2892
+    },
+    {
+      "epoch": 0.8829543720433389,
+      "grad_norm": 1.9116154189531032,
+      "learning_rate": 3.5432800173733464e-07,
+      "loss": 0.366,
+      "step": 2893
+    },
+    {
+      "epoch": 0.8832595757668243,
+      "grad_norm": 2.2885818758039442,
+      "learning_rate": 3.5250216626380395e-07,
+      "loss": 0.3826,
+      "step": 2894
+    },
+    {
+      "epoch": 0.8835647794903098,
+      "grad_norm": 3.0762827050967285,
+      "learning_rate": 3.506808752732016e-07,
+      "loss": 0.3125,
+      "step": 2895
+    },
+    {
+      "epoch": 0.8838699832137952,
+      "grad_norm": 1.5947558650859617,
+      "learning_rate": 3.488641305464496e-07,
+      "loss": 0.374,
+      "step": 2896
+    },
+    {
+      "epoch": 0.8841751869372806,
+      "grad_norm": 1.588661672926618,
+      "learning_rate": 3.470519338600231e-07,
+      "loss": 0.3316,
+      "step": 2897
+    },
+    {
+      "epoch": 0.8844803906607661,
+      "grad_norm": 1.91229278339448,
+      "learning_rate": 3.45244286985949e-07,
+      "loss": 0.3497,
+      "step": 2898
+    },
+    {
+      "epoch": 0.8847855943842515,
+      "grad_norm": 2.162437627697701,
+      "learning_rate": 3.434411916918079e-07,
+      "loss": 0.2772,
+      "step": 2899
+    },
+    {
+      "epoch": 0.8850907981077369,
+      "grad_norm": 2.1766614246548697,
+      "learning_rate": 3.4164264974072767e-07,
+      "loss": 0.3205,
+      "step": 2900
+    },
+    {
+      "epoch": 0.8853960018312224,
+      "grad_norm": 1.766540155139548,
+      "learning_rate": 3.3984866289138384e-07,
+      "loss": 0.3762,
+      "step": 2901
+    },
+    {
+      "epoch": 0.8857012055547078,
+      "grad_norm": 1.656932762057204,
+      "learning_rate": 3.3805923289799925e-07,
+      "loss": 0.2464,
+      "step": 2902
+    },
+    {
+      "epoch": 0.8860064092781932,
+      "grad_norm": 1.9518060121489198,
+      "learning_rate": 3.3627436151033976e-07,
+      "loss": 0.355,
+      "step": 2903
+    },
+    {
+      "epoch": 0.8863116130016786,
+      "grad_norm": 1.2265407293051998,
+      "learning_rate": 3.344940504737132e-07,
+      "loss": 0.2153,
+      "step": 2904
+    },
+    {
+      "epoch": 0.886616816725164,
+      "grad_norm": 2.1094249578140785,
+      "learning_rate": 3.3271830152897034e-07,
+      "loss": 0.3913,
+      "step": 2905
+    },
+    {
+      "epoch": 0.8869220204486494,
+      "grad_norm": 1.8163593901465602,
+      "learning_rate": 3.309471164124978e-07,
+      "loss": 0.1928,
+      "step": 2906
+    },
+    {
+      "epoch": 0.887227224172135,
+      "grad_norm": 1.611692741062144,
+      "learning_rate": 3.2918049685622176e-07,
+      "loss": 0.2319,
+      "step": 2907
+    },
+    {
+      "epoch": 0.8875324278956204,
+      "grad_norm": 1.7094278163526146,
+      "learning_rate": 3.274184445876033e-07,
+      "loss": 0.4093,
+      "step": 2908
+    },
+    {
+      "epoch": 0.8878376316191058,
+      "grad_norm": 2.231266019533923,
+      "learning_rate": 3.2566096132963855e-07,
+      "loss": 0.3184,
+      "step": 2909
+    },
+    {
+      "epoch": 0.8881428353425912,
+      "grad_norm": 1.7527878217750263,
+      "learning_rate": 3.2390804880085446e-07,
+      "loss": 0.2561,
+      "step": 2910
+    },
+    {
+      "epoch": 0.8884480390660766,
+      "grad_norm": 1.4361536936264583,
+      "learning_rate": 3.2215970871530833e-07,
+      "loss": 0.2711,
+      "step": 2911
+    },
+    {
+      "epoch": 0.888753242789562,
+      "grad_norm": 1.6304932691615903,
+      "learning_rate": 3.2041594278258814e-07,
+      "loss": 0.3421,
+      "step": 2912
+    },
+    {
+      "epoch": 0.8890584465130474,
+      "grad_norm": 1.6973082577670313,
+      "learning_rate": 3.1867675270780775e-07,
+      "loss": 0.4015,
+      "step": 2913
+    },
+    {
+      "epoch": 0.8893636502365329,
+      "grad_norm": 1.6377867352985322,
+      "learning_rate": 3.1694214019160617e-07,
+      "loss": 0.1819,
+      "step": 2914
+    },
+    {
+      "epoch": 0.8896688539600183,
+      "grad_norm": 2.070385280265544,
+      "learning_rate": 3.1521210693014895e-07,
+      "loss": 0.2974,
+      "step": 2915
+    },
+    {
+      "epoch": 0.8899740576835038,
+      "grad_norm": 1.3208479839358807,
+      "learning_rate": 3.1348665461511995e-07,
+      "loss": 0.2575,
+      "step": 2916
+    },
+    {
+      "epoch": 0.8902792614069892,
+      "grad_norm": 1.8611178524534187,
+      "learning_rate": 3.1176578493372634e-07,
+      "loss": 0.5982,
+      "step": 2917
+    },
+    {
+      "epoch": 0.8905844651304746,
+      "grad_norm": 1.8773806399667003,
+      "learning_rate": 3.1004949956869425e-07,
+      "loss": 0.3908,
+      "step": 2918
+    },
+    {
+      "epoch": 0.89088966885396,
+      "grad_norm": 1.9302383439732715,
+      "learning_rate": 3.083378001982662e-07,
+      "loss": 0.3582,
+      "step": 2919
+    },
+    {
+      "epoch": 0.8911948725774455,
+      "grad_norm": 1.632939018932989,
+      "learning_rate": 3.0663068849619947e-07,
+      "loss": 0.2021,
+      "step": 2920
+    },
+    {
+      "epoch": 0.8915000763009309,
+      "grad_norm": 1.7351259547529747,
+      "learning_rate": 3.0492816613176824e-07,
+      "loss": 0.2321,
+      "step": 2921
+    },
+    {
+      "epoch": 0.8918052800244163,
+      "grad_norm": 2.35469591611334,
+      "learning_rate": 3.0323023476975633e-07,
+      "loss": 0.3903,
+      "step": 2922
+    },
+    {
+      "epoch": 0.8921104837479017,
+      "grad_norm": 1.8756225823879709,
+      "learning_rate": 3.015368960704584e-07,
+      "loss": 0.3714,
+      "step": 2923
+    },
+    {
+      "epoch": 0.8924156874713871,
+      "grad_norm": 1.5504329107606214,
+      "learning_rate": 2.9984815168968005e-07,
+      "loss": 0.2628,
+      "step": 2924
+    },
+    {
+      "epoch": 0.8927208911948725,
+      "grad_norm": 1.5052450982942016,
+      "learning_rate": 2.9816400327873475e-07,
+      "loss": 0.311,
+      "step": 2925
+    },
+    {
+      "epoch": 0.893026094918358,
+      "grad_norm": 1.6022165591136466,
+      "learning_rate": 2.96484452484439e-07,
+      "loss": 0.4598,
+      "step": 2926
+    },
+    {
+      "epoch": 0.8933312986418435,
+      "grad_norm": 2.724851142247225,
+      "learning_rate": 2.948095009491164e-07,
+      "loss": 0.3631,
+      "step": 2927
+    },
+    {
+      "epoch": 0.8936365023653289,
+      "grad_norm": 1.5278134611567196,
+      "learning_rate": 2.9313915031059124e-07,
+      "loss": 0.1995,
+      "step": 2928
+    },
+    {
+      "epoch": 0.8939417060888143,
+      "grad_norm": 1.6693975328119404,
+      "learning_rate": 2.914734022021892e-07,
+      "loss": 0.3837,
+      "step": 2929
+    },
+    {
+      "epoch": 0.8942469098122997,
+      "grad_norm": 1.8591924625843064,
+      "learning_rate": 2.8981225825273685e-07,
+      "loss": 0.477,
+      "step": 2930
+    },
+    {
+      "epoch": 0.8945521135357851,
+      "grad_norm": 1.4329841835389625,
+      "learning_rate": 2.8815572008655813e-07,
+      "loss": 0.1959,
+      "step": 2931
+    },
+    {
+      "epoch": 0.8948573172592705,
+      "grad_norm": 1.4974132326746141,
+      "learning_rate": 2.865037893234718e-07,
+      "loss": 0.1497,
+      "step": 2932
+    },
+    {
+      "epoch": 0.895162520982756,
+      "grad_norm": 1.881911572696749,
+      "learning_rate": 2.8485646757879346e-07,
+      "loss": 0.3683,
+      "step": 2933
+    },
+    {
+      "epoch": 0.8954677247062414,
+      "grad_norm": 1.8397958024371963,
+      "learning_rate": 2.832137564633303e-07,
+      "loss": 0.3345,
+      "step": 2934
+    },
+    {
+      "epoch": 0.8957729284297269,
+      "grad_norm": 1.296334671153873,
+      "learning_rate": 2.815756575833822e-07,
+      "loss": 0.266,
+      "step": 2935
+    },
+    {
+      "epoch": 0.8960781321532123,
+      "grad_norm": 1.6612062252985271,
+      "learning_rate": 2.79942172540737e-07,
+      "loss": 0.4569,
+      "step": 2936
+    },
+    {
+      "epoch": 0.8963833358766977,
+      "grad_norm": 1.5904279005922972,
+      "learning_rate": 2.7831330293267456e-07,
+      "loss": 0.2959,
+      "step": 2937
+    },
+    {
+      "epoch": 0.8966885396001831,
+      "grad_norm": 1.961044430191955,
+      "learning_rate": 2.7668905035195814e-07,
+      "loss": 0.4133,
+      "step": 2938
+    },
+    {
+      "epoch": 0.8969937433236685,
+      "grad_norm": 1.6163743043142387,
+      "learning_rate": 2.750694163868373e-07,
+      "loss": 0.3433,
+      "step": 2939
+    },
+    {
+      "epoch": 0.897298947047154,
+      "grad_norm": 1.722356168945633,
+      "learning_rate": 2.734544026210473e-07,
+      "loss": 0.3398,
+      "step": 2940
+    },
+    {
+      "epoch": 0.8976041507706394,
+      "grad_norm": 1.5628068241229134,
+      "learning_rate": 2.7184401063380274e-07,
+      "loss": 0.1959,
+      "step": 2941
+    },
+    {
+      "epoch": 0.8979093544941248,
+      "grad_norm": 1.3166091717266857,
+      "learning_rate": 2.702382419998001e-07,
+      "loss": 0.1876,
+      "step": 2942
+    },
+    {
+      "epoch": 0.8982145582176102,
+      "grad_norm": 1.4971338546018425,
+      "learning_rate": 2.6863709828921613e-07,
+      "loss": 0.298,
+      "step": 2943
+    },
+    {
+      "epoch": 0.8985197619410957,
+      "grad_norm": 2.1887573988008957,
+      "learning_rate": 2.67040581067704e-07,
+      "loss": 0.2459,
+      "step": 2944
+    },
+    {
+      "epoch": 0.8988249656645811,
+      "grad_norm": 1.9489520851849282,
+      "learning_rate": 2.654486918963922e-07,
+      "loss": 0.3013,
+      "step": 2945
+    },
+    {
+      "epoch": 0.8991301693880666,
+      "grad_norm": 1.8230958541721705,
+      "learning_rate": 2.6386143233188555e-07,
+      "loss": 0.1776,
+      "step": 2946
+    },
+    {
+      "epoch": 0.899435373111552,
+      "grad_norm": 1.7492627907133111,
+      "learning_rate": 2.6227880392626027e-07,
+      "loss": 0.4107,
+      "step": 2947
+    },
+    {
+      "epoch": 0.8997405768350374,
+      "grad_norm": 1.5771168425144115,
+      "learning_rate": 2.60700808227064e-07,
+      "loss": 0.2886,
+      "step": 2948
+    },
+    {
+      "epoch": 0.9000457805585228,
+      "grad_norm": 1.432403735775991,
+      "learning_rate": 2.591274467773169e-07,
+      "loss": 0.3455,
+      "step": 2949
+    },
+    {
+      "epoch": 0.9003509842820082,
+      "grad_norm": 1.6271872105899479,
+      "learning_rate": 2.5755872111550507e-07,
+      "loss": 0.3225,
+      "step": 2950
+    },
+    {
+      "epoch": 0.9006561880054936,
+      "grad_norm": 1.551260924442995,
+      "learning_rate": 2.5599463277558135e-07,
+      "loss": 0.3506,
+      "step": 2951
+    },
+    {
+      "epoch": 0.900961391728979,
+      "grad_norm": 1.6455674338929827,
+      "learning_rate": 2.544351832869668e-07,
+      "loss": 0.3607,
+      "step": 2952
+    },
+    {
+      "epoch": 0.9012665954524646,
+      "grad_norm": 1.7086200783597483,
+      "learning_rate": 2.5288037417454337e-07,
+      "loss": 0.2798,
+      "step": 2953
+    },
+    {
+      "epoch": 0.90157179917595,
+      "grad_norm": 1.3498086299184577,
+      "learning_rate": 2.513302069586565e-07,
+      "loss": 0.1468,
+      "step": 2954
+    },
+    {
+      "epoch": 0.9018770028994354,
+      "grad_norm": 1.6086819091472078,
+      "learning_rate": 2.497846831551143e-07,
+      "loss": 0.2781,
+      "step": 2955
+    },
+    {
+      "epoch": 0.9021822066229208,
+      "grad_norm": 1.7869933710855876,
+      "learning_rate": 2.4824380427518225e-07,
+      "loss": 0.2661,
+      "step": 2956
+    },
+    {
+      "epoch": 0.9024874103464062,
+      "grad_norm": 1.8348454250746127,
+      "learning_rate": 2.46707571825584e-07,
+      "loss": 0.1668,
+      "step": 2957
+    },
+    {
+      "epoch": 0.9027926140698916,
+      "grad_norm": 1.7159862145001317,
+      "learning_rate": 2.4517598730850076e-07,
+      "loss": 0.3124,
+      "step": 2958
+    },
+    {
+      "epoch": 0.9030978177933771,
+      "grad_norm": 1.8220207483217812,
+      "learning_rate": 2.436490522215695e-07,
+      "loss": 0.2875,
+      "step": 2959
+    },
+    {
+      "epoch": 0.9034030215168625,
+      "grad_norm": 1.892282941255574,
+      "learning_rate": 2.421267680578787e-07,
+      "loss": 0.288,
+      "step": 2960
+    },
+    {
+      "epoch": 0.9037082252403479,
+      "grad_norm": 1.5044806634351904,
+      "learning_rate": 2.406091363059698e-07,
+      "loss": 0.2717,
+      "step": 2961
+    },
+    {
+      "epoch": 0.9040134289638334,
+      "grad_norm": 1.379994986497369,
+      "learning_rate": 2.3909615844983633e-07,
+      "loss": 0.281,
+      "step": 2962
+    },
+    {
+      "epoch": 0.9043186326873188,
+      "grad_norm": 1.6561392261240486,
+      "learning_rate": 2.3758783596891988e-07,
+      "loss": 0.3448,
+      "step": 2963
+    },
+    {
+      "epoch": 0.9046238364108042,
+      "grad_norm": 1.666811253316016,
+      "learning_rate": 2.360841703381084e-07,
+      "loss": 0.3559,
+      "step": 2964
+    },
+    {
+      "epoch": 0.9049290401342897,
+      "grad_norm": 1.77491503767383,
+      "learning_rate": 2.3458516302773905e-07,
+      "loss": 0.4251,
+      "step": 2965
+    },
+    {
+      "epoch": 0.9052342438577751,
+      "grad_norm": 2.1718214289144027,
+      "learning_rate": 2.3309081550359324e-07,
+      "loss": 0.288,
+      "step": 2966
+    },
+    {
+      "epoch": 0.9055394475812605,
+      "grad_norm": 2.085574000543535,
+      "learning_rate": 2.3160112922689326e-07,
+      "loss": 0.3233,
+      "step": 2967
+    },
+    {
+      "epoch": 0.9058446513047459,
+      "grad_norm": 2.1440465827621273,
+      "learning_rate": 2.301161056543072e-07,
+      "loss": 0.2738,
+      "step": 2968
+    },
+    {
+      "epoch": 0.9061498550282313,
+      "grad_norm": 1.9519208934810752,
+      "learning_rate": 2.2863574623794181e-07,
+      "loss": 0.4609,
+      "step": 2969
+    },
+    {
+      "epoch": 0.9064550587517167,
+      "grad_norm": 1.6276575512078413,
+      "learning_rate": 2.2716005242534255e-07,
+      "loss": 0.4467,
+      "step": 2970
+    },
+    {
+      "epoch": 0.9067602624752022,
+      "grad_norm": 1.7591843062973673,
+      "learning_rate": 2.2568902565949457e-07,
+      "loss": 0.2674,
+      "step": 2971
+    },
+    {
+      "epoch": 0.9070654661986877,
+      "grad_norm": 1.9227596820909427,
+      "learning_rate": 2.2422266737881838e-07,
+      "loss": 0.2263,
+      "step": 2972
+    },
+    {
+      "epoch": 0.9073706699221731,
+      "grad_norm": 1.827078756270843,
+      "learning_rate": 2.2276097901716753e-07,
+      "loss": 0.3184,
+      "step": 2973
+    },
+    {
+      "epoch": 0.9076758736456585,
+      "grad_norm": 1.7897558076617492,
+      "learning_rate": 2.2130396200383264e-07,
+      "loss": 0.2699,
+      "step": 2974
+    },
+    {
+      "epoch": 0.9079810773691439,
+      "grad_norm": 1.4481076816647764,
+      "learning_rate": 2.1985161776353513e-07,
+      "loss": 0.3225,
+      "step": 2975
+    },
+    {
+      "epoch": 0.9082862810926293,
+      "grad_norm": 1.751007198952909,
+      "learning_rate": 2.1840394771642737e-07,
+      "loss": 0.5209,
+      "step": 2976
+    },
+    {
+      "epoch": 0.9085914848161147,
+      "grad_norm": 1.678714346091653,
+      "learning_rate": 2.169609532780892e-07,
+      "loss": 0.3714,
+      "step": 2977
+    },
+    {
+      "epoch": 0.9088966885396002,
+      "grad_norm": 1.6436678325582152,
+      "learning_rate": 2.1552263585953136e-07,
+      "loss": 0.2791,
+      "step": 2978
+    },
+    {
+      "epoch": 0.9092018922630856,
+      "grad_norm": 1.752840733728362,
+      "learning_rate": 2.1408899686719e-07,
+      "loss": 0.2642,
+      "step": 2979
+    },
+    {
+      "epoch": 0.9095070959865711,
+      "grad_norm": 1.539056923173589,
+      "learning_rate": 2.1266003770292533e-07,
+      "loss": 0.1958,
+      "step": 2980
+    },
+    {
+      "epoch": 0.9098122997100565,
+      "grad_norm": 1.3348877111637576,
+      "learning_rate": 2.1123575976402467e-07,
+      "loss": 0.217,
+      "step": 2981
+    },
+    {
+      "epoch": 0.9101175034335419,
+      "grad_norm": 1.5200835256465404,
+      "learning_rate": 2.0981616444319452e-07,
+      "loss": 0.2368,
+      "step": 2982
+    },
+    {
+      "epoch": 0.9104227071570273,
+      "grad_norm": 1.5057324703588917,
+      "learning_rate": 2.0840125312856396e-07,
+      "loss": 0.1525,
+      "step": 2983
+    },
+    {
+      "epoch": 0.9107279108805127,
+      "grad_norm": 1.518402857411774,
+      "learning_rate": 2.0699102720368236e-07,
+      "loss": 0.4341,
+      "step": 2984
+    },
+    {
+      "epoch": 0.9110331146039982,
+      "grad_norm": 2.0636749388863187,
+      "learning_rate": 2.0558548804751666e-07,
+      "loss": 0.3891,
+      "step": 2985
+    },
+    {
+      "epoch": 0.9113383183274836,
+      "grad_norm": 1.8381990828593886,
+      "learning_rate": 2.041846370344508e-07,
+      "loss": 0.2212,
+      "step": 2986
+    },
+    {
+      "epoch": 0.911643522050969,
+      "grad_norm": 1.6369897440529062,
+      "learning_rate": 2.0278847553428682e-07,
+      "loss": 0.2761,
+      "step": 2987
+    },
+    {
+      "epoch": 0.9119487257744544,
+      "grad_norm": 1.956432319408319,
+      "learning_rate": 2.0139700491223768e-07,
+      "loss": 0.3502,
+      "step": 2988
+    },
+    {
+      "epoch": 0.9122539294979398,
+      "grad_norm": 1.4843700208078496,
+      "learning_rate": 2.000102265289311e-07,
+      "loss": 0.2639,
+      "step": 2989
+    },
+    {
+      "epoch": 0.9125591332214253,
+      "grad_norm": 1.8702280992730664,
+      "learning_rate": 1.9862814174040678e-07,
+      "loss": 0.2676,
+      "step": 2990
+    },
+    {
+      "epoch": 0.9128643369449108,
+      "grad_norm": 1.6415192846117646,
+      "learning_rate": 1.9725075189811594e-07,
+      "loss": 0.2689,
+      "step": 2991
+    },
+    {
+      "epoch": 0.9131695406683962,
+      "grad_norm": 1.7393193606196415,
+      "learning_rate": 1.9587805834891617e-07,
+      "loss": 0.3546,
+      "step": 2992
+    },
+    {
+      "epoch": 0.9134747443918816,
+      "grad_norm": 1.6258605732371942,
+      "learning_rate": 1.9451006243507487e-07,
+      "loss": 0.2548,
+      "step": 2993
+    },
+    {
+      "epoch": 0.913779948115367,
+      "grad_norm": 1.6662873631242507,
+      "learning_rate": 1.9314676549426593e-07,
+      "loss": 0.2709,
+      "step": 2994
+    },
+    {
+      "epoch": 0.9140851518388524,
+      "grad_norm": 3.438359239868652,
+      "learning_rate": 1.9178816885956742e-07,
+      "loss": 0.2072,
+      "step": 2995
+    },
+    {
+      "epoch": 0.9143903555623378,
+      "grad_norm": 1.552789881451876,
+      "learning_rate": 1.9043427385946057e-07,
+      "loss": 0.3087,
+      "step": 2996
+    },
+    {
+      "epoch": 0.9146955592858232,
+      "grad_norm": 1.6744143729085579,
+      "learning_rate": 1.8908508181783247e-07,
+      "loss": 0.282,
+      "step": 2997
+    },
+    {
+      "epoch": 0.9150007630093087,
+      "grad_norm": 1.862795592539928,
+      "learning_rate": 1.8774059405396894e-07,
+      "loss": 0.2934,
+      "step": 2998
+    },
+    {
+      "epoch": 0.9153059667327942,
+      "grad_norm": 1.5699772263351266,
+      "learning_rate": 1.8640081188255444e-07,
+      "loss": 0.2005,
+      "step": 2999
+    },
+    {
+      "epoch": 0.9156111704562796,
+      "grad_norm": 1.4547961889979046,
+      "learning_rate": 1.8506573661367655e-07,
+      "loss": 0.231,
+      "step": 3000
+    },
+    {
+      "epoch": 0.915916374179765,
+      "grad_norm": 1.9805592172038058,
+      "learning_rate": 1.83735369552816e-07,
+      "loss": 0.4442,
+      "step": 3001
+    },
+    {
+      "epoch": 0.9162215779032504,
+      "grad_norm": 1.6046233910046397,
+      "learning_rate": 1.8240971200085166e-07,
+      "loss": 0.3761,
+      "step": 3002
+    },
+    {
+      "epoch": 0.9165267816267358,
+      "grad_norm": 1.8319659841661153,
+      "learning_rate": 1.8108876525405772e-07,
+      "loss": 0.4478,
+      "step": 3003
+    },
+    {
+      "epoch": 0.9168319853502213,
+      "grad_norm": 1.522913681354371,
+      "learning_rate": 1.7977253060410093e-07,
+      "loss": 0.1907,
+      "step": 3004
+    },
+    {
+      "epoch": 0.9171371890737067,
+      "grad_norm": 1.6945324346838082,
+      "learning_rate": 1.7846100933803955e-07,
+      "loss": 0.27,
+      "step": 3005
+    },
+    {
+      "epoch": 0.9174423927971921,
+      "grad_norm": 1.6314053837810694,
+      "learning_rate": 1.771542027383255e-07,
+      "loss": 0.2314,
+      "step": 3006
+    },
+    {
+      "epoch": 0.9177475965206775,
+      "grad_norm": 1.6985445636319214,
+      "learning_rate": 1.7585211208279884e-07,
+      "loss": 0.3697,
+      "step": 3007
+    },
+    {
+      "epoch": 0.918052800244163,
+      "grad_norm": 1.5613154689723483,
+      "learning_rate": 1.7455473864468887e-07,
+      "loss": 0.2834,
+      "step": 3008
+    },
+    {
+      "epoch": 0.9183580039676484,
+      "grad_norm": 1.6859199756351442,
+      "learning_rate": 1.7326208369261023e-07,
+      "loss": 0.464,
+      "step": 3009
+    },
+    {
+      "epoch": 0.9186632076911339,
+      "grad_norm": 1.7305503808738993,
+      "learning_rate": 1.7197414849056793e-07,
+      "loss": 0.3251,
+      "step": 3010
+    },
+    {
+      "epoch": 0.9189684114146193,
+      "grad_norm": 1.7233544143749935,
+      "learning_rate": 1.706909342979468e-07,
+      "loss": 0.4452,
+      "step": 3011
+    },
+    {
+      "epoch": 0.9192736151381047,
+      "grad_norm": 1.6595896034336173,
+      "learning_rate": 1.694124423695187e-07,
+      "loss": 0.3588,
+      "step": 3012
+    },
+    {
+      "epoch": 0.9195788188615901,
+      "grad_norm": 1.6874345837382116,
+      "learning_rate": 1.681386739554375e-07,
+      "loss": 0.3651,
+      "step": 3013
+    },
+    {
+      "epoch": 0.9198840225850755,
+      "grad_norm": 1.8692354598745753,
+      "learning_rate": 1.6686963030123683e-07,
+      "loss": 0.4362,
+      "step": 3014
+    },
+    {
+      "epoch": 0.9201892263085609,
+      "grad_norm": 1.5592937379506653,
+      "learning_rate": 1.656053126478313e-07,
+      "loss": 0.2329,
+      "step": 3015
+    },
+    {
+      "epoch": 0.9204944300320463,
+      "grad_norm": 1.9022517563902894,
+      "learning_rate": 1.6434572223151423e-07,
+      "loss": 0.2243,
+      "step": 3016
+    },
+    {
+      "epoch": 0.9207996337555319,
+      "grad_norm": 1.464763181378052,
+      "learning_rate": 1.6309086028395648e-07,
+      "loss": 0.2692,
+      "step": 3017
+    },
+    {
+      "epoch": 0.9211048374790173,
+      "grad_norm": 1.8338574902859095,
+      "learning_rate": 1.6184072803220485e-07,
+      "loss": 0.332,
+      "step": 3018
+    },
+    {
+      "epoch": 0.9214100412025027,
+      "grad_norm": 2.0532311909937064,
+      "learning_rate": 1.6059532669868204e-07,
+      "loss": 0.3979,
+      "step": 3019
+    },
+    {
+      "epoch": 0.9217152449259881,
+      "grad_norm": 1.6680209220927975,
+      "learning_rate": 1.5935465750118396e-07,
+      "loss": 0.1827,
+      "step": 3020
+    },
+    {
+      "epoch": 0.9220204486494735,
+      "grad_norm": 1.593221314947013,
+      "learning_rate": 1.58118721652879e-07,
+      "loss": 0.2762,
+      "step": 3021
+    },
+    {
+      "epoch": 0.9223256523729589,
+      "grad_norm": 1.5754246033424293,
+      "learning_rate": 1.5688752036230881e-07,
+      "loss": 0.2798,
+      "step": 3022
+    },
+    {
+      "epoch": 0.9226308560964444,
+      "grad_norm": 1.6539477776029077,
+      "learning_rate": 1.556610548333831e-07,
+      "loss": 0.3146,
+      "step": 3023
+    },
+    {
+      "epoch": 0.9229360598199298,
+      "grad_norm": 1.7674358163103439,
+      "learning_rate": 1.5443932626538317e-07,
+      "loss": 0.3525,
+      "step": 3024
+    },
+    {
+      "epoch": 0.9232412635434152,
+      "grad_norm": 1.6245734315139786,
+      "learning_rate": 1.5322233585295608e-07,
+      "loss": 0.3127,
+      "step": 3025
+    },
+    {
+      "epoch": 0.9235464672669007,
+      "grad_norm": 1.5218617432817847,
+      "learning_rate": 1.5201008478611723e-07,
+      "loss": 0.3303,
+      "step": 3026
+    },
+    {
+      "epoch": 0.9238516709903861,
+      "grad_norm": 1.4627888418728086,
+      "learning_rate": 1.508025742502478e-07,
+      "loss": 0.3914,
+      "step": 3027
+    },
+    {
+      "epoch": 0.9241568747138715,
+      "grad_norm": 1.7738867371365525,
+      "learning_rate": 1.495998054260922e-07,
+      "loss": 0.3212,
+      "step": 3028
+    },
+    {
+      "epoch": 0.924462078437357,
+      "grad_norm": 1.8511057101111497,
+      "learning_rate": 1.4840177948975964e-07,
+      "loss": 0.3011,
+      "step": 3029
+    },
+    {
+      "epoch": 0.9247672821608424,
+      "grad_norm": 1.8829133572612307,
+      "learning_rate": 1.4720849761272138e-07,
+      "loss": 0.4673,
+      "step": 3030
+    },
+    {
+      "epoch": 0.9250724858843278,
+      "grad_norm": 1.8364459693162152,
+      "learning_rate": 1.4601996096180792e-07,
+      "loss": 0.2928,
+      "step": 3031
+    },
+    {
+      "epoch": 0.9253776896078132,
+      "grad_norm": 1.8570840008809846,
+      "learning_rate": 1.4483617069921296e-07,
+      "loss": 0.4844,
+      "step": 3032
+    },
+    {
+      "epoch": 0.9256828933312986,
+      "grad_norm": 1.6856272867234632,
+      "learning_rate": 1.436571279824861e-07,
+      "loss": 0.2067,
+      "step": 3033
+    },
+    {
+      "epoch": 0.925988097054784,
+      "grad_norm": 1.8831068260007167,
+      "learning_rate": 1.4248283396453566e-07,
+      "loss": 0.3718,
+      "step": 3034
+    },
+    {
+      "epoch": 0.9262933007782695,
+      "grad_norm": 1.5416696221950486,
+      "learning_rate": 1.4131328979362813e-07,
+      "loss": 0.2602,
+      "step": 3035
+    },
+    {
+      "epoch": 0.926598504501755,
+      "grad_norm": 1.3032519729192462,
+      "learning_rate": 1.4014849661338258e-07,
+      "loss": 0.1346,
+      "step": 3036
+    },
+    {
+      "epoch": 0.9269037082252404,
+      "grad_norm": 1.5945893115883587,
+      "learning_rate": 1.3898845556277352e-07,
+      "loss": 0.2103,
+      "step": 3037
+    },
+    {
+      "epoch": 0.9272089119487258,
+      "grad_norm": 2.267391384429593,
+      "learning_rate": 1.378331677761302e-07,
+      "loss": 0.3795,
+      "step": 3038
+    },
+    {
+      "epoch": 0.9275141156722112,
+      "grad_norm": 1.667049109767016,
+      "learning_rate": 1.3668263438313177e-07,
+      "loss": 0.3487,
+      "step": 3039
+    },
+    {
+      "epoch": 0.9278193193956966,
+      "grad_norm": 1.8954270594757712,
+      "learning_rate": 1.3553685650880998e-07,
+      "loss": 0.3996,
+      "step": 3040
+    },
+    {
+      "epoch": 0.928124523119182,
+      "grad_norm": 1.7837118570036214,
+      "learning_rate": 1.3439583527354417e-07,
+      "loss": 0.2491,
+      "step": 3041
+    },
+    {
+      "epoch": 0.9284297268426674,
+      "grad_norm": 1.8231666030295597,
+      "learning_rate": 1.3325957179306625e-07,
+      "loss": 0.1648,
+      "step": 3042
+    },
+    {
+      "epoch": 0.9287349305661529,
+      "grad_norm": 1.771683485119366,
+      "learning_rate": 1.321280671784525e-07,
+      "loss": 0.3665,
+      "step": 3043
+    },
+    {
+      "epoch": 0.9290401342896384,
+      "grad_norm": 1.8934447643684014,
+      "learning_rate": 1.3100132253612675e-07,
+      "loss": 0.1926,
+      "step": 3044
+    },
+    {
+      "epoch": 0.9293453380131238,
+      "grad_norm": 1.907266439635279,
+      "learning_rate": 1.2987933896785932e-07,
+      "loss": 0.3228,
+      "step": 3045
+    },
+    {
+      "epoch": 0.9296505417366092,
+      "grad_norm": 2.2551799064647726,
+      "learning_rate": 1.2876211757076373e-07,
+      "loss": 0.3462,
+      "step": 3046
+    },
+    {
+      "epoch": 0.9299557454600946,
+      "grad_norm": 1.7869325977714636,
+      "learning_rate": 1.2764965943729724e-07,
+      "loss": 0.2372,
+      "step": 3047
+    },
+    {
+      "epoch": 0.93026094918358,
+      "grad_norm": 1.7434704765049673,
+      "learning_rate": 1.2654196565526077e-07,
+      "loss": 0.3657,
+      "step": 3048
+    },
+    {
+      "epoch": 0.9305661529070655,
+      "grad_norm": 2.217071758149955,
+      "learning_rate": 1.2543903730779406e-07,
+      "loss": 0.305,
+      "step": 3049
+    },
+    {
+      "epoch": 0.9308713566305509,
+      "grad_norm": 1.9085185201264485,
+      "learning_rate": 1.2434087547337824e-07,
+      "loss": 0.1456,
+      "step": 3050
+    },
+    {
+      "epoch": 0.9311765603540363,
+      "grad_norm": 1.8140467916592178,
+      "learning_rate": 1.232474812258344e-07,
+      "loss": 0.2688,
+      "step": 3051
+    },
+    {
+      "epoch": 0.9314817640775217,
+      "grad_norm": 3.537591961315723,
+      "learning_rate": 1.2215885563432117e-07,
+      "loss": 0.2682,
+      "step": 3052
+    },
+    {
+      "epoch": 0.9317869678010072,
+      "grad_norm": 1.68609744525017,
+      "learning_rate": 1.2107499976333314e-07,
+      "loss": 0.3311,
+      "step": 3053
+    },
+    {
+      "epoch": 0.9320921715244926,
+      "grad_norm": 1.7453938218503213,
+      "learning_rate": 1.1999591467270255e-07,
+      "loss": 0.2957,
+      "step": 3054
+    },
+    {
+      "epoch": 0.932397375247978,
+      "grad_norm": 1.8908716466504132,
+      "learning_rate": 1.1892160141759479e-07,
+      "loss": 0.4493,
+      "step": 3055
+    },
+    {
+      "epoch": 0.9327025789714635,
+      "grad_norm": 1.4688529932150731,
+      "learning_rate": 1.1785206104851122e-07,
+      "loss": 0.1949,
+      "step": 3056
+    },
+    {
+      "epoch": 0.9330077826949489,
+      "grad_norm": 1.7005755740395718,
+      "learning_rate": 1.1678729461128524e-07,
+      "loss": 0.4439,
+      "step": 3057
+    },
+    {
+      "epoch": 0.9333129864184343,
+      "grad_norm": 1.62211918398401,
+      "learning_rate": 1.1572730314708181e-07,
+      "loss": 0.1903,
+      "step": 3058
+    },
+    {
+      "epoch": 0.9336181901419197,
+      "grad_norm": 1.4639416167390211,
+      "learning_rate": 1.1467208769239624e-07,
+      "loss": 0.1994,
+      "step": 3059
+    },
+    {
+      "epoch": 0.9339233938654051,
+      "grad_norm": 1.7931644773342565,
+      "learning_rate": 1.1362164927905595e-07,
+      "loss": 0.4354,
+      "step": 3060
+    },
+    {
+      "epoch": 0.9342285975888905,
+      "grad_norm": 2.07586259951712,
+      "learning_rate": 1.1257598893421429e-07,
+      "loss": 0.4668,
+      "step": 3061
+    },
+    {
+      "epoch": 0.934533801312376,
+      "grad_norm": 1.7073213169330927,
+      "learning_rate": 1.1153510768035447e-07,
+      "loss": 0.2002,
+      "step": 3062
+    },
+    {
+      "epoch": 0.9348390050358615,
+      "grad_norm": 1.294379579617933,
+      "learning_rate": 1.1049900653528512e-07,
+      "loss": 0.3173,
+      "step": 3063
+    },
+    {
+      "epoch": 0.9351442087593469,
+      "grad_norm": 1.7475252364141403,
+      "learning_rate": 1.0946768651214245e-07,
+      "loss": 0.1948,
+      "step": 3064
+    },
+    {
+      "epoch": 0.9354494124828323,
+      "grad_norm": 1.772644403383591,
+      "learning_rate": 1.0844114861938648e-07,
+      "loss": 0.4511,
+      "step": 3065
+    },
+    {
+      "epoch": 0.9357546162063177,
+      "grad_norm": 1.7161844699737987,
+      "learning_rate": 1.0741939386080091e-07,
+      "loss": 0.3736,
+      "step": 3066
+    },
+    {
+      "epoch": 0.9360598199298031,
+      "grad_norm": 1.6309186547749837,
+      "learning_rate": 1.0640242323549266e-07,
+      "loss": 0.2649,
+      "step": 3067
+    },
+    {
+      "epoch": 0.9363650236532886,
+      "grad_norm": 1.7457238814732614,
+      "learning_rate": 1.0539023773789069e-07,
+      "loss": 0.5385,
+      "step": 3068
+    },
+    {
+      "epoch": 0.936670227376774,
+      "grad_norm": 1.4956914694554349,
+      "learning_rate": 1.0438283835774387e-07,
+      "loss": 0.2701,
+      "step": 3069
+    },
+    {
+      "epoch": 0.9369754311002594,
+      "grad_norm": 1.4626932456651007,
+      "learning_rate": 1.0338022608012365e-07,
+      "loss": 0.2981,
+      "step": 3070
+    },
+    {
+      "epoch": 0.9372806348237448,
+      "grad_norm": 1.6248818868113604,
+      "learning_rate": 1.0238240188541748e-07,
+      "loss": 0.5032,
+      "step": 3071
+    },
+    {
+      "epoch": 0.9375858385472303,
+      "grad_norm": 1.458176886023303,
+      "learning_rate": 1.0138936674933098e-07,
+      "loss": 0.2469,
+      "step": 3072
+    },
+    {
+      "epoch": 0.9378910422707157,
+      "grad_norm": 1.741700724522758,
+      "learning_rate": 1.0040112164289073e-07,
+      "loss": 0.3523,
+      "step": 3073
+    },
+    {
+      "epoch": 0.9381962459942012,
+      "grad_norm": 2.2307663166219363,
+      "learning_rate": 9.941766753243543e-08,
+      "loss": 0.3739,
+      "step": 3074
+    },
+    {
+      "epoch": 0.9385014497176866,
+      "grad_norm": 1.6894420469064177,
+      "learning_rate": 9.84390053796197e-08,
+      "loss": 0.2126,
+      "step": 3075
+    },
+    {
+      "epoch": 0.938806653441172,
+      "grad_norm": 1.8517504890392837,
+      "learning_rate": 9.746513614141362e-08,
+      "loss": 0.1941,
+      "step": 3076
+    },
+    {
+      "epoch": 0.9391118571646574,
+      "grad_norm": 1.7329188927338595,
+      "learning_rate": 9.649606077009988e-08,
+      "loss": 0.5339,
+      "step": 3077
+    },
+    {
+      "epoch": 0.9394170608881428,
+      "grad_norm": 1.5848102039576724,
+      "learning_rate": 9.553178021327326e-08,
+      "loss": 0.1557,
+      "step": 3078
+    },
+    {
+      "epoch": 0.9397222646116282,
+      "grad_norm": 1.8113953743059201,
+      "learning_rate": 9.457229541384117e-08,
+      "loss": 0.1842,
+      "step": 3079
+    },
+    {
+      "epoch": 0.9400274683351136,
+      "grad_norm": 1.6202392532140553,
+      "learning_rate": 9.361760731002035e-08,
+      "loss": 0.336,
+      "step": 3080
+    },
+    {
+      "epoch": 0.9403326720585992,
+      "grad_norm": 1.6042187504755832,
+      "learning_rate": 9.266771683533738e-08,
+      "loss": 0.3475,
+      "step": 3081
+    },
+    {
+      "epoch": 0.9406378757820846,
+      "grad_norm": 1.3922444889550674,
+      "learning_rate": 9.172262491862816e-08,
+      "loss": 0.2794,
+      "step": 3082
+    },
+    {
+      "epoch": 0.94094307950557,
+      "grad_norm": 2.1203607518574707,
+      "learning_rate": 9.078233248403568e-08,
+      "loss": 0.1426,
+      "step": 3083
+    },
+    {
+      "epoch": 0.9412482832290554,
+      "grad_norm": 1.6426230480024866,
+      "learning_rate": 8.984684045101055e-08,
+      "loss": 0.2625,
+      "step": 3084
+    },
+    {
+      "epoch": 0.9415534869525408,
+      "grad_norm": 1.6979257158221603,
+      "learning_rate": 8.891614973430829e-08,
+      "loss": 0.3273,
+      "step": 3085
+    },
+    {
+      "epoch": 0.9418586906760262,
+      "grad_norm": 1.9710500693423525,
+      "learning_rate": 8.799026124399146e-08,
+      "loss": 0.274,
+      "step": 3086
+    },
+    {
+      "epoch": 0.9421638943995116,
+      "grad_norm": 1.870458825680719,
+      "learning_rate": 8.706917588542418e-08,
+      "loss": 0.174,
+      "step": 3087
+    },
+    {
+      "epoch": 0.9424690981229971,
+      "grad_norm": 1.8847617668151897,
+      "learning_rate": 8.6152894559276e-08,
+      "loss": 0.3027,
+      "step": 3088
+    },
+    {
+      "epoch": 0.9427743018464825,
+      "grad_norm": 1.881554084103904,
+      "learning_rate": 8.524141816151854e-08,
+      "loss": 0.1893,
+      "step": 3089
+    },
+    {
+      "epoch": 0.943079505569968,
+      "grad_norm": 1.739369461100097,
+      "learning_rate": 8.433474758342441e-08,
+      "loss": 0.2814,
+      "step": 3090
+    },
+    {
+      "epoch": 0.9433847092934534,
+      "grad_norm": 1.6545810567967778,
+      "learning_rate": 8.343288371156665e-08,
+      "loss": 0.2362,
+      "step": 3091
+    },
+    {
+      "epoch": 0.9436899130169388,
+      "grad_norm": 1.619856102286329,
+      "learning_rate": 8.253582742782041e-08,
+      "loss": 0.3785,
+      "step": 3092
+    },
+    {
+      "epoch": 0.9439951167404242,
+      "grad_norm": 1.651335785317375,
+      "learning_rate": 8.164357960935675e-08,
+      "loss": 0.4398,
+      "step": 3093
+    },
+    {
+      "epoch": 0.9443003204639097,
+      "grad_norm": 1.7539490012754981,
+      "learning_rate": 8.07561411286456e-08,
+      "loss": 0.3187,
+      "step": 3094
+    },
+    {
+      "epoch": 0.9446055241873951,
+      "grad_norm": 1.8747356228407501,
+      "learning_rate": 7.987351285345668e-08,
+      "loss": 0.3133,
+      "step": 3095
+    },
+    {
+      "epoch": 0.9449107279108805,
+      "grad_norm": 2.0530899857738967,
+      "learning_rate": 7.899569564685294e-08,
+      "loss": 0.1711,
+      "step": 3096
+    },
+    {
+      "epoch": 0.9452159316343659,
+      "grad_norm": 1.5156446632604146,
+      "learning_rate": 7.812269036719444e-08,
+      "loss": 0.3256,
+      "step": 3097
+    },
+    {
+      "epoch": 0.9455211353578513,
+      "grad_norm": 1.529266707901095,
+      "learning_rate": 7.725449786813554e-08,
+      "loss": 0.2665,
+      "step": 3098
+    },
+    {
+      "epoch": 0.9458263390813368,
+      "grad_norm": 1.7932842035311636,
+      "learning_rate": 7.63911189986255e-08,
+      "loss": 0.4121,
+      "step": 3099
+    },
+    {
+      "epoch": 0.9461315428048223,
+      "grad_norm": 1.753857720262275,
+      "learning_rate": 7.553255460290399e-08,
+      "loss": 0.2428,
+      "step": 3100
+    },
+    {
+      "epoch": 0.9464367465283077,
+      "grad_norm": 1.9349085797539933,
+      "learning_rate": 7.467880552050721e-08,
+      "loss": 0.4053,
+      "step": 3101
+    },
+    {
+      "epoch": 0.9467419502517931,
+      "grad_norm": 1.716596329121474,
+      "learning_rate": 7.382987258625851e-08,
+      "loss": 0.173,
+      "step": 3102
+    },
+    {
+      "epoch": 0.9470471539752785,
+      "grad_norm": 1.974453249229349,
+      "learning_rate": 7.298575663027385e-08,
+      "loss": 0.313,
+      "step": 3103
+    },
+    {
+      "epoch": 0.9473523576987639,
+      "grad_norm": 1.6405383859205218,
+      "learning_rate": 7.214645847795909e-08,
+      "loss": 0.4044,
+      "step": 3104
+    },
+    {
+      "epoch": 0.9476575614222493,
+      "grad_norm": 1.7514331943448285,
+      "learning_rate": 7.131197895000941e-08,
+      "loss": 0.2446,
+      "step": 3105
+    },
+    {
+      "epoch": 0.9479627651457347,
+      "grad_norm": 1.4290549958459764,
+      "learning_rate": 7.048231886240653e-08,
+      "loss": 0.1826,
+      "step": 3106
+    },
+    {
+      "epoch": 0.9482679688692202,
+      "grad_norm": 1.2450984220399632,
+      "learning_rate": 6.965747902642095e-08,
+      "loss": 0.2012,
+      "step": 3107
+    },
+    {
+      "epoch": 0.9485731725927057,
+      "grad_norm": 1.4945845598921765,
+      "learning_rate": 6.883746024860971e-08,
+      "loss": 0.2961,
+      "step": 3108
+    },
+    {
+      "epoch": 0.9488783763161911,
+      "grad_norm": 1.728767333588549,
+      "learning_rate": 6.802226333081474e-08,
+      "loss": 0.4283,
+      "step": 3109
+    },
+    {
+      "epoch": 0.9491835800396765,
+      "grad_norm": 1.7245224167187219,
+      "learning_rate": 6.721188907016396e-08,
+      "loss": 0.3148,
+      "step": 3110
+    },
+    {
+      "epoch": 0.9494887837631619,
+      "grad_norm": 1.5365620551148431,
+      "learning_rate": 6.640633825906906e-08,
+      "loss": 0.2251,
+      "step": 3111
+    },
+    {
+      "epoch": 0.9497939874866473,
+      "grad_norm": 1.5399204709213705,
+      "learning_rate": 6.56056116852255e-08,
+      "loss": 0.3458,
+      "step": 3112
+    },
+    {
+      "epoch": 0.9500991912101328,
+      "grad_norm": 1.4310328089740316,
+      "learning_rate": 6.48097101316103e-08,
+      "loss": 0.1783,
+      "step": 3113
+    },
+    {
+      "epoch": 0.9504043949336182,
+      "grad_norm": 1.6172332020074571,
+      "learning_rate": 6.40186343764848e-08,
+      "loss": 0.1566,
+      "step": 3114
+    },
+    {
+      "epoch": 0.9507095986571036,
+      "grad_norm": 1.8201053485143524,
+      "learning_rate": 6.32323851933886e-08,
+      "loss": 0.4466,
+      "step": 3115
+    },
+    {
+      "epoch": 0.951014802380589,
+      "grad_norm": 1.608563412958756,
+      "learning_rate": 6.245096335114387e-08,
+      "loss": 0.2211,
+      "step": 3116
+    },
+    {
+      "epoch": 0.9513200061040745,
+      "grad_norm": 1.6555680364455492,
+      "learning_rate": 6.16743696138522e-08,
+      "loss": 0.2064,
+      "step": 3117
+    },
+    {
+      "epoch": 0.9516252098275599,
+      "grad_norm": 2.2928571787631715,
+      "learning_rate": 6.090260474089227e-08,
+      "loss": 0.3908,
+      "step": 3118
+    },
+    {
+      "epoch": 0.9519304135510454,
+      "grad_norm": 1.796168710970393,
+      "learning_rate": 6.013566948692317e-08,
+      "loss": 0.4037,
+      "step": 3119
+    },
+    {
+      "epoch": 0.9522356172745308,
+      "grad_norm": 1.7894655798058297,
+      "learning_rate": 5.9373564601880595e-08,
+      "loss": 0.3374,
+      "step": 3120
+    },
+    {
+      "epoch": 0.9525408209980162,
+      "grad_norm": 1.1926506172885767,
+      "learning_rate": 5.8616290830976194e-08,
+      "loss": 0.1459,
+      "step": 3121
+    },
+    {
+      "epoch": 0.9528460247215016,
+      "grad_norm": 1.9998501483091524,
+      "learning_rate": 5.786384891469932e-08,
+      "loss": 0.3278,
+      "step": 3122
+    },
+    {
+      "epoch": 0.953151228444987,
+      "grad_norm": 1.780004366965777,
+      "learning_rate": 5.711623958881196e-08,
+      "loss": 0.2719,
+      "step": 3123
+    },
+    {
+      "epoch": 0.9534564321684724,
+      "grad_norm": 1.649335065432672,
+      "learning_rate": 5.6373463584353806e-08,
+      "loss": 0.203,
+      "step": 3124
+    },
+    {
+      "epoch": 0.9537616358919578,
+      "grad_norm": 1.96435364816284,
+      "learning_rate": 5.563552162763608e-08,
+      "loss": 0.4157,
+      "step": 3125
+    },
+    {
+      "epoch": 0.9540668396154434,
+      "grad_norm": 1.6124475278375099,
+      "learning_rate": 5.490241444024325e-08,
+      "loss": 0.2539,
+      "step": 3126
+    },
+    {
+      "epoch": 0.9543720433389288,
+      "grad_norm": 1.9688108525635462,
+      "learning_rate": 5.4174142739033545e-08,
+      "loss": 0.2505,
+      "step": 3127
+    },
+    {
+      "epoch": 0.9546772470624142,
+      "grad_norm": 1.9030120114515199,
+      "learning_rate": 5.3450707236135656e-08,
+      "loss": 0.393,
+      "step": 3128
+    },
+    {
+      "epoch": 0.9549824507858996,
+      "grad_norm": 2.1274750824846382,
+      "learning_rate": 5.2732108638949285e-08,
+      "loss": 0.3988,
+      "step": 3129
+    },
+    {
+      "epoch": 0.955287654509385,
+      "grad_norm": 1.9063736922359853,
+      "learning_rate": 5.201834765014568e-08,
+      "loss": 0.2716,
+      "step": 3130
+    },
+    {
+      "epoch": 0.9555928582328704,
+      "grad_norm": 1.746942983543036,
+      "learning_rate": 5.130942496766433e-08,
+      "loss": 0.3204,
+      "step": 3131
+    },
+    {
+      "epoch": 0.9558980619563558,
+      "grad_norm": 1.8650127179323246,
+      "learning_rate": 5.0605341284713504e-08,
+      "loss": 0.3323,
+      "step": 3132
+    },
+    {
+      "epoch": 0.9562032656798413,
+      "grad_norm": 1.8057055223573084,
+      "learning_rate": 4.990609728977136e-08,
+      "loss": 0.2977,
+      "step": 3133
+    },
+    {
+      "epoch": 0.9565084694033267,
+      "grad_norm": 1.8039711855851885,
+      "learning_rate": 4.9211693666582625e-08,
+      "loss": 0.2697,
+      "step": 3134
+    },
+    {
+      "epoch": 0.9568136731268121,
+      "grad_norm": 1.7222014711353566,
+      "learning_rate": 4.85221310941586e-08,
+      "loss": 0.2201,
+      "step": 3135
+    },
+    {
+      "epoch": 0.9571188768502976,
+      "grad_norm": 1.473317450672231,
+      "learning_rate": 4.7837410246777125e-08,
+      "loss": 0.2856,
+      "step": 3136
+    },
+    {
+      "epoch": 0.957424080573783,
+      "grad_norm": 1.6241791926305946,
+      "learning_rate": 4.715753179398208e-08,
+      "loss": 0.2776,
+      "step": 3137
+    },
+    {
+      "epoch": 0.9577292842972684,
+      "grad_norm": 1.5922351444442016,
+      "learning_rate": 4.648249640058222e-08,
+      "loss": 0.2124,
+      "step": 3138
+    },
+    {
+      "epoch": 0.9580344880207539,
+      "grad_norm": 1.724860409603394,
+      "learning_rate": 4.5812304726649546e-08,
+      "loss": 0.3272,
+      "step": 3139
+    },
+    {
+      "epoch": 0.9583396917442393,
+      "grad_norm": 1.6747404836910174,
+      "learning_rate": 4.514695742752151e-08,
+      "loss": 0.3928,
+      "step": 3140
+    },
+    {
+      "epoch": 0.9586448954677247,
+      "grad_norm": 1.6626282721345953,
+      "learning_rate": 4.448645515379657e-08,
+      "loss": 0.3282,
+      "step": 3141
+    },
+    {
+      "epoch": 0.9589500991912101,
+      "grad_norm": 1.4388094204216442,
+      "learning_rate": 4.383079855133699e-08,
+      "loss": 0.2552,
+      "step": 3142
+    },
+    {
+      "epoch": 0.9592553029146955,
+      "grad_norm": 1.6750453942823442,
+      "learning_rate": 4.317998826126601e-08,
+      "loss": 0.4144,
+      "step": 3143
+    },
+    {
+      "epoch": 0.9595605066381809,
+      "grad_norm": 1.691395819017345,
+      "learning_rate": 4.253402491996905e-08,
+      "loss": 0.3391,
+      "step": 3144
+    },
+    {
+      "epoch": 0.9598657103616665,
+      "grad_norm": 1.6872227758940446,
+      "learning_rate": 4.1892909159090275e-08,
+      "loss": 0.2,
+      "step": 3145
+    },
+    {
+      "epoch": 0.9601709140851519,
+      "grad_norm": 1.603624161312316,
+      "learning_rate": 4.125664160553544e-08,
+      "loss": 0.3261,
+      "step": 3146
+    },
+    {
+      "epoch": 0.9604761178086373,
+      "grad_norm": 1.6339332348223208,
+      "learning_rate": 4.062522288146853e-08,
+      "loss": 0.228,
+      "step": 3147
+    },
+    {
+      "epoch": 0.9607813215321227,
+      "grad_norm": 1.4318885744634444,
+      "learning_rate": 3.9998653604311763e-08,
+      "loss": 0.3181,
+      "step": 3148
+    },
+    {
+      "epoch": 0.9610865252556081,
+      "grad_norm": 1.5052302093852121,
+      "learning_rate": 3.937693438674728e-08,
+      "loss": 0.2756,
+      "step": 3149
+    },
+    {
+      "epoch": 0.9613917289790935,
+      "grad_norm": 1.8201246696782578,
+      "learning_rate": 3.876006583671266e-08,
+      "loss": 0.4108,
+      "step": 3150
+    },
+    {
+      "epoch": 0.9616969327025789,
+      "grad_norm": 1.5679880275025952,
+      "learning_rate": 3.8148048557403174e-08,
+      "loss": 0.3295,
+      "step": 3151
+    },
+    {
+      "epoch": 0.9620021364260644,
+      "grad_norm": 1.647552976089387,
+      "learning_rate": 3.754088314727067e-08,
+      "loss": 0.4898,
+      "step": 3152
+    },
+    {
+      "epoch": 0.9623073401495498,
+      "grad_norm": 1.9339834726015996,
+      "learning_rate": 3.693857020002245e-08,
+      "loss": 0.3003,
+      "step": 3153
+    },
+    {
+      "epoch": 0.9626125438730353,
+      "grad_norm": 1.5214906430390787,
+      "learning_rate": 3.634111030462073e-08,
+      "loss": 0.1723,
+      "step": 3154
+    },
+    {
+      "epoch": 0.9629177475965207,
+      "grad_norm": 1.947860519769557,
+      "learning_rate": 3.574850404528152e-08,
+      "loss": 0.2321,
+      "step": 3155
+    },
+    {
+      "epoch": 0.9632229513200061,
+      "grad_norm": 1.377300115923031,
+      "learning_rate": 3.5160752001476286e-08,
+      "loss": 0.2842,
+      "step": 3156
+    },
+    {
+      "epoch": 0.9635281550434915,
+      "grad_norm": 1.7866288336382636,
+      "learning_rate": 3.457785474792974e-08,
+      "loss": 0.3875,
+      "step": 3157
+    },
+    {
+      "epoch": 0.963833358766977,
+      "grad_norm": 2.1990459367465216,
+      "learning_rate": 3.399981285461706e-08,
+      "loss": 0.3652,
+      "step": 3158
+    },
+    {
+      "epoch": 0.9641385624904624,
+      "grad_norm": 1.47797661131238,
+      "learning_rate": 3.342662688676945e-08,
+      "loss": 0.417,
+      "step": 3159
+    },
+    {
+      "epoch": 0.9644437662139478,
+      "grad_norm": 1.8380294244712645,
+      "learning_rate": 3.2858297404866344e-08,
+      "loss": 0.3075,
+      "step": 3160
+    },
+    {
+      "epoch": 0.9647489699374332,
+      "grad_norm": 1.9320740555812037,
+      "learning_rate": 3.2294824964640424e-08,
+      "loss": 0.2917,
+      "step": 3161
+    },
+    {
+      "epoch": 0.9650541736609186,
+      "grad_norm": 1.5679615699136098,
+      "learning_rate": 3.173621011707484e-08,
+      "loss": 0.3805,
+      "step": 3162
+    },
+    {
+      "epoch": 0.9653593773844041,
+      "grad_norm": 1.8590137547501744,
+      "learning_rate": 3.118245340840154e-08,
+      "loss": 0.2933,
+      "step": 3163
+    },
+    {
+      "epoch": 0.9656645811078896,
+      "grad_norm": 1.5404897355750176,
+      "learning_rate": 3.0633555380102933e-08,
+      "loss": 0.1512,
+      "step": 3164
+    },
+    {
+      "epoch": 0.965969784831375,
+      "grad_norm": 1.943192149879615,
+      "learning_rate": 3.0089516568910795e-08,
+      "loss": 0.5274,
+      "step": 3165
+    },
+    {
+      "epoch": 0.9662749885548604,
+      "grad_norm": 2.6836705092734,
+      "learning_rate": 2.9550337506804583e-08,
+      "loss": 0.2219,
+      "step": 3166
+    },
+    {
+      "epoch": 0.9665801922783458,
+      "grad_norm": 1.5473289979487448,
+      "learning_rate": 2.9016018721012562e-08,
+      "loss": 0.2946,
+      "step": 3167
+    },
+    {
+      "epoch": 0.9668853960018312,
+      "grad_norm": 1.492487772343577,
+      "learning_rate": 2.8486560734009573e-08,
+      "loss": 0.2715,
+      "step": 3168
+    },
+    {
+      "epoch": 0.9671905997253166,
+      "grad_norm": 1.6104415766610036,
+      "learning_rate": 2.796196406351759e-08,
+      "loss": 0.4105,
+      "step": 3169
+    },
+    {
+      "epoch": 0.967495803448802,
+      "grad_norm": 1.5281427737335482,
+      "learning_rate": 2.7442229222505724e-08,
+      "loss": 0.3069,
+      "step": 3170
+    },
+    {
+      "epoch": 0.9678010071722875,
+      "grad_norm": 1.7751502953956564,
+      "learning_rate": 2.692735671918856e-08,
+      "loss": 0.3891,
+      "step": 3171
+    },
+    {
+      "epoch": 0.968106210895773,
+      "grad_norm": 1.7123520111962385,
+      "learning_rate": 2.641734705702559e-08,
+      "loss": 0.47,
+      "step": 3172
+    },
+    {
+      "epoch": 0.9684114146192584,
+      "grad_norm": 1.3369111645529905,
+      "learning_rate": 2.5912200734722336e-08,
+      "loss": 0.316,
+      "step": 3173
+    },
+    {
+      "epoch": 0.9687166183427438,
+      "grad_norm": 1.6674558757794788,
+      "learning_rate": 2.541191824622813e-08,
+      "loss": 0.1658,
+      "step": 3174
+    },
+    {
+      "epoch": 0.9690218220662292,
+      "grad_norm": 1.827558772111376,
+      "learning_rate": 2.4916500080736094e-08,
+      "loss": 0.2408,
+      "step": 3175
+    },
+    {
+      "epoch": 0.9693270257897146,
+      "grad_norm": 1.5216273582175108,
+      "learning_rate": 2.4425946722683725e-08,
+      "loss": 0.3275,
+      "step": 3176
+    },
+    {
+      "epoch": 0.9696322295132,
+      "grad_norm": 3.150742512944137,
+      "learning_rate": 2.394025865174954e-08,
+      "loss": 0.3721,
+      "step": 3177
+    },
+    {
+      "epoch": 0.9699374332366855,
+      "grad_norm": 1.7535139634941797,
+      "learning_rate": 2.345943634285752e-08,
+      "loss": 0.1866,
+      "step": 3178
+    },
+    {
+      "epoch": 0.9702426369601709,
+      "grad_norm": 1.4921787121898464,
+      "learning_rate": 2.2983480266171586e-08,
+      "loss": 0.3459,
+      "step": 3179
+    },
+    {
+      "epoch": 0.9705478406836563,
+      "grad_norm": 1.3992680475442059,
+      "learning_rate": 2.251239088709778e-08,
+      "loss": 0.246,
+      "step": 3180
+    },
+    {
+      "epoch": 0.9708530444071418,
+      "grad_norm": 1.5859029230567245,
+      "learning_rate": 2.2046168666284284e-08,
+      "loss": 0.3011,
+      "step": 3181
+    },
+    {
+      "epoch": 0.9711582481306272,
+      "grad_norm": 1.7028286177302412,
+      "learning_rate": 2.1584814059618098e-08,
+      "loss": 0.377,
+      "step": 3182
+    },
+    {
+      "epoch": 0.9714634518541126,
+      "grad_norm": 2.0593667940430684,
+      "learning_rate": 2.1128327518227797e-08,
+      "loss": 0.3509,
+      "step": 3183
+    },
+    {
+      "epoch": 0.9717686555775981,
+      "grad_norm": 1.6996737167043607,
+      "learning_rate": 2.067670948848244e-08,
+      "loss": 0.5195,
+      "step": 3184
+    },
+    {
+      "epoch": 0.9720738593010835,
+      "grad_norm": 1.247752027949812,
+      "learning_rate": 2.022996041198877e-08,
+      "loss": 0.1139,
+      "step": 3185
+    },
+    {
+      "epoch": 0.9723790630245689,
+      "grad_norm": 1.5635366434232196,
+      "learning_rate": 1.978808072559346e-08,
+      "loss": 0.2432,
+      "step": 3186
+    },
+    {
+      "epoch": 0.9726842667480543,
+      "grad_norm": 1.9556147461058806,
+      "learning_rate": 1.9351070861381995e-08,
+      "loss": 0.3601,
+      "step": 3187
+    },
+    {
+      "epoch": 0.9729894704715397,
+      "grad_norm": 1.507737627295227,
+      "learning_rate": 1.891893124667754e-08,
+      "loss": 0.3307,
+      "step": 3188
+    },
+    {
+      "epoch": 0.9732946741950251,
+      "grad_norm": 1.8446244574541686,
+      "learning_rate": 1.8491662304041536e-08,
+      "loss": 0.2839,
+      "step": 3189
+    },
+    {
+      "epoch": 0.9735998779185107,
+      "grad_norm": 1.459507369709913,
+      "learning_rate": 1.8069264451270884e-08,
+      "loss": 0.2402,
+      "step": 3190
+    },
+    {
+      "epoch": 0.9739050816419961,
+      "grad_norm": 1.7408320195966298,
+      "learning_rate": 1.7651738101401862e-08,
+      "loss": 0.2565,
+      "step": 3191
+    },
+    {
+      "epoch": 0.9742102853654815,
+      "grad_norm": 1.7614963839448081,
+      "learning_rate": 1.7239083662705657e-08,
+      "loss": 0.422,
+      "step": 3192
+    },
+    {
+      "epoch": 0.9745154890889669,
+      "grad_norm": 1.9687097938912876,
+      "learning_rate": 1.6831301538689503e-08,
+      "loss": 0.2555,
+      "step": 3193
+    },
+    {
+      "epoch": 0.9748206928124523,
+      "grad_norm": 1.791694936603218,
+      "learning_rate": 1.642839212809777e-08,
+      "loss": 0.2699,
+      "step": 3194
+    },
+    {
+      "epoch": 0.9751258965359377,
+      "grad_norm": 1.8067707406692137,
+      "learning_rate": 1.6030355824908638e-08,
+      "loss": 0.3262,
+      "step": 3195
+    },
+    {
+      "epoch": 0.9754311002594231,
+      "grad_norm": 2.2498554607165793,
+      "learning_rate": 1.5637193018335218e-08,
+      "loss": 0.2446,
+      "step": 3196
+    },
+    {
+      "epoch": 0.9757363039829086,
+      "grad_norm": 1.5439681667347085,
+      "learning_rate": 1.5248904092826088e-08,
+      "loss": 0.3062,
+      "step": 3197
+    },
+    {
+      "epoch": 0.976041507706394,
+      "grad_norm": 1.9424093866999719,
+      "learning_rate": 1.486548942806365e-08,
+      "loss": 0.3704,
+      "step": 3198
+    },
+    {
+      "epoch": 0.9763467114298795,
+      "grad_norm": 1.6517890738670324,
+      "learning_rate": 1.4486949398963557e-08,
+      "loss": 0.3108,
+      "step": 3199
+    },
+    {
+      "epoch": 0.9766519151533649,
+      "grad_norm": 1.5416845458085093,
+      "learning_rate": 1.4113284375675273e-08,
+      "loss": 0.2964,
+      "step": 3200
+    },
+    {
+      "epoch": 0.9769571188768503,
+      "grad_norm": 1.5593385374893778,
+      "learning_rate": 1.3744494723581526e-08,
+      "loss": 0.3386,
+      "step": 3201
+    },
+    {
+      "epoch": 0.9772623226003357,
+      "grad_norm": 1.6733622130653938,
+      "learning_rate": 1.3380580803297183e-08,
+      "loss": 0.3067,
+      "step": 3202
+    },
+    {
+      "epoch": 0.9775675263238212,
+      "grad_norm": 1.5936790151308489,
+      "learning_rate": 1.302154297066982e-08,
+      "loss": 0.1751,
+      "step": 3203
+    },
+    {
+      "epoch": 0.9778727300473066,
+      "grad_norm": 1.8938042887847097,
+      "learning_rate": 1.2667381576779714e-08,
+      "loss": 0.3145,
+      "step": 3204
+    },
+    {
+      "epoch": 0.978177933770792,
+      "grad_norm": 1.985342675905468,
+      "learning_rate": 1.231809696793651e-08,
+      "loss": 0.3656,
+      "step": 3205
+    },
+    {
+      "epoch": 0.9784831374942774,
+      "grad_norm": 1.7338526171051585,
+      "learning_rate": 1.1973689485684226e-08,
+      "loss": 0.2738,
+      "step": 3206
+    },
+    {
+      "epoch": 0.9787883412177628,
+      "grad_norm": 2.329815449261681,
+      "learning_rate": 1.1634159466795691e-08,
+      "loss": 0.4249,
+      "step": 3207
+    },
+    {
+      "epoch": 0.9790935449412482,
+      "grad_norm": 1.5805277897757888,
+      "learning_rate": 1.1299507243274222e-08,
+      "loss": 0.2954,
+      "step": 3208
+    },
+    {
+      "epoch": 0.9793987486647338,
+      "grad_norm": 1.850668453903233,
+      "learning_rate": 1.0969733142355832e-08,
+      "loss": 0.3293,
+      "step": 3209
+    },
+    {
+      "epoch": 0.9797039523882192,
+      "grad_norm": 1.7705096977547878,
+      "learning_rate": 1.064483748650369e-08,
+      "loss": 0.3052,
+      "step": 3210
+    },
+    {
+      "epoch": 0.9800091561117046,
+      "grad_norm": 2.1184759407252858,
+      "learning_rate": 1.0324820593412554e-08,
+      "loss": 0.2946,
+      "step": 3211
+    },
+    {
+      "epoch": 0.98031435983519,
+      "grad_norm": 1.6815254978375824,
+      "learning_rate": 1.0009682776005447e-08,
+      "loss": 0.4507,
+      "step": 3212
+    },
+    {
+      "epoch": 0.9806195635586754,
+      "grad_norm": 1.815273399253701,
+      "learning_rate": 9.69942434243587e-09,
+      "loss": 0.3258,
+      "step": 3213
+    },
+    {
+      "epoch": 0.9809247672821608,
+      "grad_norm": 1.7076222250047222,
+      "learning_rate": 9.394045596083923e-09,
+      "loss": 0.2681,
+      "step": 3214
+    },
+    {
+      "epoch": 0.9812299710056462,
+      "grad_norm": 1.7427164138755578,
+      "learning_rate": 9.093546835560185e-09,
+      "loss": 0.3873,
+      "step": 3215
+    },
+    {
+      "epoch": 0.9815351747291317,
+      "grad_norm": 1.5944511516812474,
+      "learning_rate": 8.79792835470239e-09,
+      "loss": 0.2452,
+      "step": 3216
+    },
+    {
+      "epoch": 0.9818403784526171,
+      "grad_norm": 1.9765614747056073,
+      "learning_rate": 8.507190442577085e-09,
+      "loss": 0.2779,
+      "step": 3217
+    },
+    {
+      "epoch": 0.9821455821761026,
+      "grad_norm": 1.9152803867141897,
+      "learning_rate": 8.221333383476859e-09,
+      "loss": 0.3015,
+      "step": 3218
+    },
+    {
+      "epoch": 0.982450785899588,
+      "grad_norm": 1.8976746584555748,
+      "learning_rate": 7.94035745692312e-09,
+      "loss": 0.3744,
+      "step": 3219
+    },
+    {
+      "epoch": 0.9827559896230734,
+      "grad_norm": 1.6438120086925618,
+      "learning_rate": 7.664262937663314e-09,
+      "loss": 0.223,
+      "step": 3220
+    },
+    {
+      "epoch": 0.9830611933465588,
+      "grad_norm": 2.0132378925455385,
+      "learning_rate": 7.393050095672594e-09,
+      "loss": 0.2209,
+      "step": 3221
+    },
+    {
+      "epoch": 0.9833663970700443,
+      "grad_norm": 1.4970602792900094,
+      "learning_rate": 7.126719196152154e-09,
+      "loss": 0.2667,
+      "step": 3222
+    },
+    {
+      "epoch": 0.9836716007935297,
+      "grad_norm": 1.9208437626266621,
+      "learning_rate": 6.865270499529786e-09,
+      "loss": 0.2817,
+      "step": 3223
+    },
+    {
+      "epoch": 0.9839768045170151,
+      "grad_norm": 1.9665921207890467,
+      "learning_rate": 6.608704261457655e-09,
+      "loss": 0.2463,
+      "step": 3224
+    },
+    {
+      "epoch": 0.9842820082405005,
+      "grad_norm": 1.7756663025534054,
+      "learning_rate": 6.357020732816743e-09,
+      "loss": 0.2913,
+      "step": 3225
+    },
+    {
+      "epoch": 0.9845872119639859,
+      "grad_norm": 1.478284205437885,
+      "learning_rate": 6.110220159710744e-09,
+      "loss": 0.2409,
+      "step": 3226
+    },
+    {
+      "epoch": 0.9848924156874714,
+      "grad_norm": 1.9575325634076046,
+      "learning_rate": 5.868302783469948e-09,
+      "loss": 0.45,
+      "step": 3227
+    },
+    {
+      "epoch": 0.9851976194109568,
+      "grad_norm": 1.889011356491274,
+      "learning_rate": 5.631268840650128e-09,
+      "loss": 0.2935,
+      "step": 3228
+    },
+    {
+      "epoch": 0.9855028231344423,
+      "grad_norm": 1.7479822242758298,
+      "learning_rate": 5.399118563030325e-09,
+      "loss": 0.3301,
+      "step": 3229
+    },
+    {
+      "epoch": 0.9858080268579277,
+      "grad_norm": 1.6485662184083025,
+      "learning_rate": 5.1718521776150665e-09,
+      "loss": 0.3427,
+      "step": 3230
+    },
+    {
+      "epoch": 0.9861132305814131,
+      "grad_norm": 1.6028818046415272,
+      "learning_rate": 4.9494699066338084e-09,
+      "loss": 0.196,
+      "step": 3231
+    },
+    {
+      "epoch": 0.9864184343048985,
+      "grad_norm": 2.015249646153553,
+      "learning_rate": 4.731971967539828e-09,
+      "loss": 0.439,
+      "step": 3232
+    },
+    {
+      "epoch": 0.9867236380283839,
+      "grad_norm": 1.834696313638363,
+      "learning_rate": 4.519358573009114e-09,
+      "loss": 0.2964,
+      "step": 3233
+    },
+    {
+      "epoch": 0.9870288417518693,
+      "grad_norm": 2.1060195796813526,
+      "learning_rate": 4.3116299309425845e-09,
+      "loss": 0.437,
+      "step": 3234
+    },
+    {
+      "epoch": 0.9873340454753547,
+      "grad_norm": 1.6522688611963112,
+      "learning_rate": 4.108786244464979e-09,
+      "loss": 0.3729,
+      "step": 3235
+    },
+    {
+      "epoch": 0.9876392491988403,
+      "grad_norm": 1.8505254489001164,
+      "learning_rate": 3.910827711923193e-09,
+      "loss": 0.4158,
+      "step": 3236
+    },
+    {
+      "epoch": 0.9879444529223257,
+      "grad_norm": 1.5994443703529546,
+      "learning_rate": 3.71775452688794e-09,
+      "loss": 0.2584,
+      "step": 3237
+    },
+    {
+      "epoch": 0.9882496566458111,
+      "grad_norm": 2.123646024299586,
+      "learning_rate": 3.529566878153756e-09,
+      "loss": 0.2636,
+      "step": 3238
+    },
+    {
+      "epoch": 0.9885548603692965,
+      "grad_norm": 1.6746975611466164,
+      "learning_rate": 3.346264949735667e-09,
+      "loss": 0.2883,
+      "step": 3239
+    },
+    {
+      "epoch": 0.9888600640927819,
+      "grad_norm": 1.7888428665675282,
+      "learning_rate": 3.1678489208736286e-09,
+      "loss": 0.3256,
+      "step": 3240
+    },
+    {
+      "epoch": 0.9891652678162673,
+      "grad_norm": 1.7498548992492244,
+      "learning_rate": 2.994318966028087e-09,
+      "loss": 0.2499,
+      "step": 3241
+    },
+    {
+      "epoch": 0.9894704715397528,
+      "grad_norm": 1.6301950894956327,
+      "learning_rate": 2.825675254883309e-09,
+      "loss": 0.4248,
+      "step": 3242
+    },
+    {
+      "epoch": 0.9897756752632382,
+      "grad_norm": 1.6474807638570632,
+      "learning_rate": 2.661917952344606e-09,
+      "loss": 0.4375,
+      "step": 3243
+    },
+    {
+      "epoch": 0.9900808789867236,
+      "grad_norm": 1.5130833816949905,
+      "learning_rate": 2.503047218539445e-09,
+      "loss": 0.2229,
+      "step": 3244
+    },
+    {
+      "epoch": 0.9903860827102091,
+      "grad_norm": 1.710721633740625,
+      "learning_rate": 2.349063208816893e-09,
+      "loss": 0.3004,
+      "step": 3245
+    },
+    {
+      "epoch": 0.9906912864336945,
+      "grad_norm": 1.4238009142793833,
+      "learning_rate": 2.199966073748172e-09,
+      "loss": 0.1267,
+      "step": 3246
+    },
+    {
+      "epoch": 0.9909964901571799,
+      "grad_norm": 2.179942563901439,
+      "learning_rate": 2.0557559591255495e-09,
+      "loss": 0.3234,
+      "step": 3247
+    },
+    {
+      "epoch": 0.9913016938806654,
+      "grad_norm": 2.6007007168155325,
+      "learning_rate": 1.916433005962337e-09,
+      "loss": 0.3738,
+      "step": 3248
+    },
+    {
+      "epoch": 0.9916068976041508,
+      "grad_norm": 2.0242736738285108,
+      "learning_rate": 1.7819973504940024e-09,
+      "loss": 0.1664,
+      "step": 3249
+    },
+    {
+      "epoch": 0.9919121013276362,
+      "grad_norm": 1.610472147760261,
+      "learning_rate": 1.6524491241753927e-09,
+      "loss": 0.3363,
+      "step": 3250
+    },
+    {
+      "epoch": 0.9922173050511216,
+      "grad_norm": 2.081484724675542,
+      "learning_rate": 1.5277884536835098e-09,
+      "loss": 0.2283,
+      "step": 3251
+    },
+    {
+      "epoch": 0.992522508774607,
+      "grad_norm": 1.7297711139376675,
+      "learning_rate": 1.408015460916401e-09,
+      "loss": 0.2762,
+      "step": 3252
+    },
+    {
+      "epoch": 0.9928277124980924,
+      "grad_norm": 1.5367615050441692,
+      "learning_rate": 1.2931302629914933e-09,
+      "loss": 0.2469,
+      "step": 3253
+    },
+    {
+      "epoch": 0.993132916221578,
+      "grad_norm": 1.7292033578434607,
+      "learning_rate": 1.1831329722478135e-09,
+      "loss": 0.3093,
+      "step": 3254
+    },
+    {
+      "epoch": 0.9934381199450634,
+      "grad_norm": 1.6099734628753108,
+      "learning_rate": 1.0780236962448787e-09,
+      "loss": 0.2235,
+      "step": 3255
+    },
+    {
+      "epoch": 0.9937433236685488,
+      "grad_norm": 1.4622875646213087,
+      "learning_rate": 9.778025377610301e-10,
+      "loss": 0.2812,
+      "step": 3256
+    },
+    {
+      "epoch": 0.9940485273920342,
+      "grad_norm": 1.5434701611866395,
+      "learning_rate": 8.824695947967643e-10,
+      "loss": 0.4045,
+      "step": 3257
+    },
+    {
+      "epoch": 0.9943537311155196,
+      "grad_norm": 1.786815381491549,
+      "learning_rate": 7.920249605719576e-10,
+      "loss": 0.3155,
+      "step": 3258
+    },
+    {
+      "epoch": 0.994658934839005,
+      "grad_norm": 1.9817553672905523,
+      "learning_rate": 7.064687235264211e-10,
+      "loss": 0.2771,
+      "step": 3259
+    },
+    {
+      "epoch": 0.9949641385624904,
+      "grad_norm": 1.6066809020592328,
+      "learning_rate": 6.258009673199006e-10,
+      "loss": 0.2929,
+      "step": 3260
+    },
+    {
+      "epoch": 0.9952693422859759,
+      "grad_norm": 2.435620194396852,
+      "learning_rate": 5.500217708320765e-10,
+      "loss": 0.3177,
+      "step": 3261
+    },
+    {
+      "epoch": 0.9955745460094613,
+      "grad_norm": 1.328916605483517,
+      "learning_rate": 4.791312081620092e-10,
+      "loss": 0.322,
+      "step": 3262
+    },
+    {
+      "epoch": 0.9958797497329468,
+      "grad_norm": 1.7408370185083828,
+      "learning_rate": 4.131293486298038e-10,
+      "loss": 0.1873,
+      "step": 3263
+    },
+    {
+      "epoch": 0.9961849534564322,
+      "grad_norm": 1.4655986010811306,
+      "learning_rate": 3.52016256773835e-10,
+      "loss": 0.3255,
+      "step": 3264
+    },
+    {
+      "epoch": 0.9964901571799176,
+      "grad_norm": 1.5263478185751251,
+      "learning_rate": 2.9579199235241216e-10,
+      "loss": 0.2787,
+      "step": 3265
+    },
+    {
+      "epoch": 0.996795360903403,
+      "grad_norm": 1.676123239615716,
+      "learning_rate": 2.4445661034377957e-10,
+      "loss": 0.3825,
+      "step": 3266
+    },
+    {
+      "epoch": 0.9971005646268885,
+      "grad_norm": 1.8073283108414324,
+      "learning_rate": 1.9801016094556092e-10,
+      "loss": 0.2098,
+      "step": 3267
+    },
+    {
+      "epoch": 0.9974057683503739,
+      "grad_norm": 1.7580545894492703,
+      "learning_rate": 1.5645268957420467e-10,
+      "loss": 0.4099,
+      "step": 3268
+    },
+    {
+      "epoch": 0.9977109720738593,
+      "grad_norm": 1.900901796255548,
+      "learning_rate": 1.1978423686664908e-10,
+      "loss": 0.4666,
+      "step": 3269
+    },
+    {
+      "epoch": 0.9980161757973447,
+      "grad_norm": 1.5234889095141848,
+      "learning_rate": 8.800483867865694e-11,
+      "loss": 0.3132,
+      "step": 3270
+    },
+    {
+      "epoch": 0.9983213795208301,
+      "grad_norm": 1.4627559992084171,
+      "learning_rate": 6.111452608426049e-11,
+      "loss": 0.2896,
+      "step": 3271
+    },
+    {
+      "epoch": 0.9986265832443156,
+      "grad_norm": 2.0341419467364257,
+      "learning_rate": 3.911332537853696e-11,
+      "loss": 0.3179,
+      "step": 3272
+    },
+    {
+      "epoch": 0.998931786967801,
+      "grad_norm": 1.884623562691759,
+      "learning_rate": 2.200125807538811e-11,
+      "loss": 0.2754,
+      "step": 3273
+    },
+    {
+      "epoch": 0.9992369906912865,
+      "grad_norm": 1.8192708059580842,
+      "learning_rate": 9.77834090643004e-12,
+      "loss": 0.3777,
+      "step": 3274
+    },
+    {
+      "epoch": 0.9995421944147719,
+      "grad_norm": 1.852195020067947,
+      "learning_rate": 2.4445858243238305e-12,
+      "loss": 0.4301,
+      "step": 3275
+    },
+    {
+      "epoch": 0.9998473981382573,
+      "grad_norm": 1.7682018549462064,
+      "learning_rate": 0.0,
+      "loss": 0.3041,
+      "step": 3276
+    },
+    {
+      "epoch": 0.9998473981382573,
+      "step": 3276,
+      "total_flos": 407180904620032.0,
+      "train_loss": 0.3420985525812822,
+      "train_runtime": 61899.4198,
+      "train_samples_per_second": 1.694,
+      "train_steps_per_second": 0.053
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 3276,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 328,
+  "total_flos": 407180904620032.0,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}