diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,5073 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.3157721616139466,
+  "eval_steps": 500,
+  "global_step": 720,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.00043857244668603693,
+      "grad_norm": 1.5650805234909058,
+      "learning_rate": 1.4492753623188406e-08,
+      "loss": 0.4933,
+      "step": 1
+    },
+    {
+      "epoch": 0.0008771448933720739,
+      "grad_norm": 1.2313175201416016,
+      "learning_rate": 2.898550724637681e-08,
+      "loss": 0.4744,
+      "step": 2
+    },
+    {
+      "epoch": 0.001315717340058111,
+      "grad_norm": 1.4933998584747314,
+      "learning_rate": 4.347826086956521e-08,
+      "loss": 0.4723,
+      "step": 3
+    },
+    {
+      "epoch": 0.0017542897867441477,
+      "grad_norm": 1.4024336338043213,
+      "learning_rate": 5.797101449275362e-08,
+      "loss": 0.4569,
+      "step": 4
+    },
+    {
+      "epoch": 0.0021928622334301848,
+      "grad_norm": 1.2763197422027588,
+      "learning_rate": 7.246376811594203e-08,
+      "loss": 0.4702,
+      "step": 5
+    },
+    {
+      "epoch": 0.002631434680116222,
+      "grad_norm": 1.3957324028015137,
+      "learning_rate": 8.695652173913042e-08,
+      "loss": 0.4788,
+      "step": 6
+    },
+    {
+      "epoch": 0.003070007126802259,
+      "grad_norm": 1.3984495401382446,
+      "learning_rate": 1.0144927536231885e-07,
+      "loss": 0.4864,
+      "step": 7
+    },
+    {
+      "epoch": 0.0035085795734882954,
+      "grad_norm": 1.321620225906372,
+      "learning_rate": 1.1594202898550725e-07,
+      "loss": 0.4611,
+      "step": 8
+    },
+    {
+      "epoch": 0.0039471520201743325,
+      "grad_norm": 1.2592936754226685,
+      "learning_rate": 1.3043478260869563e-07,
+      "loss": 0.4508,
+      "step": 9
+    },
+    {
+      "epoch": 0.0043857244668603695,
+      "grad_norm": 1.4119383096694946,
+      "learning_rate": 1.4492753623188405e-07,
+      "loss": 0.5231,
+      "step": 10
+    },
+    {
+      "epoch": 0.0048242969135464065,
+      "grad_norm": 1.343310832977295,
+      "learning_rate": 1.5942028985507245e-07,
+      "loss": 0.4483,
+      "step": 11
+    },
+    {
+      "epoch": 0.005262869360232444,
+      "grad_norm": 1.522481918334961,
+      "learning_rate": 1.7391304347826085e-07,
+      "loss": 0.4727,
+      "step": 12
+    },
+    {
+      "epoch": 0.005701441806918481,
+      "grad_norm": 1.2751290798187256,
+      "learning_rate": 1.8840579710144927e-07,
+      "loss": 0.4973,
+      "step": 13
+    },
+    {
+      "epoch": 0.006140014253604518,
+      "grad_norm": 1.29275643825531,
+      "learning_rate": 2.028985507246377e-07,
+      "loss": 0.4627,
+      "step": 14
+    },
+    {
+      "epoch": 0.006578586700290554,
+      "grad_norm": 1.2619212865829468,
+      "learning_rate": 2.1739130434782607e-07,
+      "loss": 0.4771,
+      "step": 15
+    },
+    {
+      "epoch": 0.007017159146976591,
+      "grad_norm": 1.4083727598190308,
+      "learning_rate": 2.318840579710145e-07,
+      "loss": 0.4892,
+      "step": 16
+    },
+    {
+      "epoch": 0.007455731593662628,
+      "grad_norm": 1.468392252922058,
+      "learning_rate": 2.463768115942029e-07,
+      "loss": 0.4481,
+      "step": 17
+    },
+    {
+      "epoch": 0.007894304040348665,
+      "grad_norm": 1.4284064769744873,
+      "learning_rate": 2.6086956521739126e-07,
+      "loss": 0.454,
+      "step": 18
+    },
+    {
+      "epoch": 0.008332876487034702,
+      "grad_norm": 1.2147692441940308,
+      "learning_rate": 2.753623188405797e-07,
+      "loss": 0.4357,
+      "step": 19
+    },
+    {
+      "epoch": 0.008771448933720739,
+      "grad_norm": 1.2484781742095947,
+      "learning_rate": 2.898550724637681e-07,
+      "loss": 0.47,
+      "step": 20
+    },
+    {
+      "epoch": 0.009210021380406776,
+      "grad_norm": 1.3065662384033203,
+      "learning_rate": 3.043478260869565e-07,
+      "loss": 0.515,
+      "step": 21
+    },
+    {
+      "epoch": 0.009648593827092813,
+      "grad_norm": 1.2731220722198486,
+      "learning_rate": 3.188405797101449e-07,
+      "loss": 0.4567,
+      "step": 22
+    },
+    {
+      "epoch": 0.01008716627377885,
+      "grad_norm": 1.2372796535491943,
+      "learning_rate": 3.333333333333333e-07,
+      "loss": 0.4886,
+      "step": 23
+    },
+    {
+      "epoch": 0.010525738720464887,
+      "grad_norm": 1.2341394424438477,
+      "learning_rate": 3.478260869565217e-07,
+      "loss": 0.4618,
+      "step": 24
+    },
+    {
+      "epoch": 0.010964311167150924,
+      "grad_norm": 1.2708252668380737,
+      "learning_rate": 3.6231884057971015e-07,
+      "loss": 0.4695,
+      "step": 25
+    },
+    {
+      "epoch": 0.011402883613836961,
+      "grad_norm": 1.326284646987915,
+      "learning_rate": 3.7681159420289855e-07,
+      "loss": 0.4807,
+      "step": 26
+    },
+    {
+      "epoch": 0.011841456060522998,
+      "grad_norm": 1.267345666885376,
+      "learning_rate": 3.9130434782608694e-07,
+      "loss": 0.4729,
+      "step": 27
+    },
+    {
+      "epoch": 0.012280028507209035,
+      "grad_norm": 1.302316665649414,
+      "learning_rate": 4.057971014492754e-07,
+      "loss": 0.4427,
+      "step": 28
+    },
+    {
+      "epoch": 0.012718600953895072,
+      "grad_norm": 1.2135436534881592,
+      "learning_rate": 4.2028985507246374e-07,
+      "loss": 0.4371,
+      "step": 29
+    },
+    {
+      "epoch": 0.013157173400581108,
+      "grad_norm": 1.162211298942566,
+      "learning_rate": 4.3478260869565214e-07,
+      "loss": 0.4611,
+      "step": 30
+    },
+    {
+      "epoch": 0.013595745847267145,
+      "grad_norm": 1.305415391921997,
+      "learning_rate": 4.4927536231884053e-07,
+      "loss": 0.4789,
+      "step": 31
+    },
+    {
+      "epoch": 0.014034318293953182,
+      "grad_norm": 1.545601725578308,
+      "learning_rate": 4.63768115942029e-07,
+      "loss": 0.5328,
+      "step": 32
+    },
+    {
+      "epoch": 0.014472890740639219,
+      "grad_norm": 1.4781523942947388,
+      "learning_rate": 4.782608695652174e-07,
+      "loss": 0.4827,
+      "step": 33
+    },
+    {
+      "epoch": 0.014911463187325256,
+      "grad_norm": 1.2530491352081299,
+      "learning_rate": 4.927536231884058e-07,
+      "loss": 0.4836,
+      "step": 34
+    },
+    {
+      "epoch": 0.015350035634011293,
+      "grad_norm": 1.2580320835113525,
+      "learning_rate": 5.072463768115942e-07,
+      "loss": 0.4536,
+      "step": 35
+    },
+    {
+      "epoch": 0.01578860808069733,
+      "grad_norm": 1.2908357381820679,
+      "learning_rate": 5.217391304347825e-07,
+      "loss": 0.4578,
+      "step": 36
+    },
+    {
+      "epoch": 0.016227180527383367,
+      "grad_norm": 1.2363778352737427,
+      "learning_rate": 5.36231884057971e-07,
+      "loss": 0.4768,
+      "step": 37
+    },
+    {
+      "epoch": 0.016665752974069404,
+      "grad_norm": 1.395221471786499,
+      "learning_rate": 5.507246376811594e-07,
+      "loss": 0.4261,
+      "step": 38
+    },
+    {
+      "epoch": 0.01710432542075544,
+      "grad_norm": 1.5333712100982666,
+      "learning_rate": 5.652173913043477e-07,
+      "loss": 0.5309,
+      "step": 39
+    },
+    {
+      "epoch": 0.017542897867441478,
+      "grad_norm": 1.4437744617462158,
+      "learning_rate": 5.797101449275362e-07,
+      "loss": 0.5116,
+      "step": 40
+    },
+    {
+      "epoch": 0.017981470314127515,
+      "grad_norm": 1.3201457262039185,
+      "learning_rate": 5.942028985507246e-07,
+      "loss": 0.4506,
+      "step": 41
+    },
+    {
+      "epoch": 0.018420042760813552,
+      "grad_norm": 1.2283574342727661,
+      "learning_rate": 6.08695652173913e-07,
+      "loss": 0.5068,
+      "step": 42
+    },
+    {
+      "epoch": 0.01885861520749959,
+      "grad_norm": 1.355634093284607,
+      "learning_rate": 6.231884057971014e-07,
+      "loss": 0.5099,
+      "step": 43
+    },
+    {
+      "epoch": 0.019297187654185626,
+      "grad_norm": 1.284976601600647,
+      "learning_rate": 6.376811594202898e-07,
+      "loss": 0.4596,
+      "step": 44
+    },
+    {
+      "epoch": 0.019735760100871663,
+      "grad_norm": 1.3104331493377686,
+      "learning_rate": 6.521739130434782e-07,
+      "loss": 0.4222,
+      "step": 45
+    },
+    {
+      "epoch": 0.0201743325475577,
+      "grad_norm": 1.2699097394943237,
+      "learning_rate": 6.666666666666666e-07,
+      "loss": 0.4565,
+      "step": 46
+    },
+    {
+      "epoch": 0.020612904994243737,
+      "grad_norm": 1.401367425918579,
+      "learning_rate": 6.811594202898551e-07,
+      "loss": 0.4207,
+      "step": 47
+    },
+    {
+      "epoch": 0.021051477440929774,
+      "grad_norm": 1.2467551231384277,
+      "learning_rate": 6.956521739130434e-07,
+      "loss": 0.448,
+      "step": 48
+    },
+    {
+      "epoch": 0.02149004988761581,
+      "grad_norm": 1.304906964302063,
+      "learning_rate": 7.101449275362319e-07,
+      "loss": 0.451,
+      "step": 49
+    },
+    {
+      "epoch": 0.02192862233430185,
+      "grad_norm": 1.21262526512146,
+      "learning_rate": 7.246376811594203e-07,
+      "loss": 0.4607,
+      "step": 50
+    },
+    {
+      "epoch": 0.022367194780987885,
+      "grad_norm": 1.1871509552001953,
+      "learning_rate": 7.391304347826086e-07,
+      "loss": 0.4355,
+      "step": 51
+    },
+    {
+      "epoch": 0.022805767227673922,
+      "grad_norm": 1.2784358263015747,
+      "learning_rate": 7.536231884057971e-07,
+      "loss": 0.4316,
+      "step": 52
+    },
+    {
+      "epoch": 0.02324433967435996,
+      "grad_norm": 1.2877992391586304,
+      "learning_rate": 7.681159420289855e-07,
+      "loss": 0.4737,
+      "step": 53
+    },
+    {
+      "epoch": 0.023682912121045997,
+      "grad_norm": 1.315425157546997,
+      "learning_rate": 7.826086956521739e-07,
+      "loss": 0.4885,
+      "step": 54
+    },
+    {
+      "epoch": 0.024121484567732034,
+      "grad_norm": 1.2742252349853516,
+      "learning_rate": 7.971014492753623e-07,
+      "loss": 0.4956,
+      "step": 55
+    },
+    {
+      "epoch": 0.02456005701441807,
+      "grad_norm": 1.2137497663497925,
+      "learning_rate": 8.115942028985508e-07,
+      "loss": 0.3931,
+      "step": 56
+    },
+    {
+      "epoch": 0.024998629461104108,
+      "grad_norm": 1.2331868410110474,
+      "learning_rate": 8.260869565217391e-07,
+      "loss": 0.463,
+      "step": 57
+    },
+    {
+      "epoch": 0.025437201907790145,
+      "grad_norm": 1.3358384370803833,
+      "learning_rate": 8.405797101449275e-07,
+      "loss": 0.4744,
+      "step": 58
+    },
+    {
+      "epoch": 0.025875774354476182,
+      "grad_norm": 1.3751511573791504,
+      "learning_rate": 8.550724637681159e-07,
+      "loss": 0.4179,
+      "step": 59
+    },
+    {
+      "epoch": 0.026314346801162215,
+      "grad_norm": 1.3184670209884644,
+      "learning_rate": 8.695652173913043e-07,
+      "loss": 0.4813,
+      "step": 60
+    },
+    {
+      "epoch": 0.026752919247848252,
+      "grad_norm": 1.2712794542312622,
+      "learning_rate": 8.840579710144928e-07,
+      "loss": 0.4931,
+      "step": 61
+    },
+    {
+      "epoch": 0.02719149169453429,
+      "grad_norm": 1.3398113250732422,
+      "learning_rate": 8.985507246376811e-07,
+      "loss": 0.54,
+      "step": 62
+    },
+    {
+      "epoch": 0.027630064141220326,
+      "grad_norm": 1.2505016326904297,
+      "learning_rate": 9.130434782608695e-07,
+      "loss": 0.4464,
+      "step": 63
+    },
+    {
+      "epoch": 0.028068636587906363,
+      "grad_norm": 1.1980383396148682,
+      "learning_rate": 9.27536231884058e-07,
+      "loss": 0.4118,
+      "step": 64
+    },
+    {
+      "epoch": 0.0285072090345924,
+      "grad_norm": 1.27497136592865,
+      "learning_rate": 9.420289855072463e-07,
+      "loss": 0.4491,
+      "step": 65
+    },
+    {
+      "epoch": 0.028945781481278438,
+      "grad_norm": 1.2081972360610962,
+      "learning_rate": 9.565217391304349e-07,
+      "loss": 0.3582,
+      "step": 66
+    },
+    {
+      "epoch": 0.029384353927964475,
+      "grad_norm": 1.2842012643814087,
+      "learning_rate": 9.710144927536232e-07,
+      "loss": 0.4584,
+      "step": 67
+    },
+    {
+      "epoch": 0.02982292637465051,
+      "grad_norm": 1.2653955221176147,
+      "learning_rate": 9.855072463768117e-07,
+      "loss": 0.463,
+      "step": 68
+    },
+    {
+      "epoch": 0.03026149882133655,
+      "grad_norm": 1.190528392791748,
+      "learning_rate": 1e-06,
+      "loss": 0.4628,
+      "step": 69
+    },
+    {
+      "epoch": 0.030700071268022586,
+      "grad_norm": 1.1633963584899902,
+      "learning_rate": 9.999994952664242e-07,
+      "loss": 0.4204,
+      "step": 70
+    },
+    {
+      "epoch": 0.031138643714708623,
+      "grad_norm": 1.4251805543899536,
+      "learning_rate": 9.999979810667154e-07,
+      "loss": 0.492,
+      "step": 71
+    },
+    {
+      "epoch": 0.03157721616139466,
+      "grad_norm": 1.4169254302978516,
+      "learning_rate": 9.99995457403931e-07,
+      "loss": 0.4918,
+      "step": 72
+    },
+    {
+      "epoch": 0.0320157886080807,
+      "grad_norm": 1.2996279001235962,
+      "learning_rate": 9.999919242831662e-07,
+      "loss": 0.4389,
+      "step": 73
+    },
+    {
+      "epoch": 0.032454361054766734,
+      "grad_norm": 1.1896483898162842,
+      "learning_rate": 9.999873817115539e-07,
+      "loss": 0.4169,
+      "step": 74
+    },
+    {
+      "epoch": 0.032892933501452774,
+      "grad_norm": 1.3035953044891357,
+      "learning_rate": 9.999818296982652e-07,
+      "loss": 0.4625,
+      "step": 75
+    },
+    {
+      "epoch": 0.03333150594813881,
+      "grad_norm": 1.2109572887420654,
+      "learning_rate": 9.999752682545095e-07,
+      "loss": 0.4225,
+      "step": 76
+    },
+    {
+      "epoch": 0.03377007839482485,
+      "grad_norm": 1.1753894090652466,
+      "learning_rate": 9.999676973935336e-07,
+      "loss": 0.4265,
+      "step": 77
+    },
+    {
+      "epoch": 0.03420865084151088,
+      "grad_norm": 1.2878810167312622,
+      "learning_rate": 9.99959117130623e-07,
+      "loss": 0.4736,
+      "step": 78
+    },
+    {
+      "epoch": 0.03464722328819692,
+      "grad_norm": 1.29270601272583,
+      "learning_rate": 9.999495274831003e-07,
+      "loss": 0.4767,
+      "step": 79
+    },
+    {
+      "epoch": 0.035085795734882956,
+      "grad_norm": 1.1873911619186401,
+      "learning_rate": 9.999389284703264e-07,
+      "loss": 0.4282,
+      "step": 80
+    },
+    {
+      "epoch": 0.03552436818156899,
+      "grad_norm": 1.4078459739685059,
+      "learning_rate": 9.999273201137004e-07,
+      "loss": 0.4702,
+      "step": 81
+    },
+    {
+      "epoch": 0.03596294062825503,
+      "grad_norm": 1.209952473640442,
+      "learning_rate": 9.999147024366583e-07,
+      "loss": 0.4174,
+      "step": 82
+    },
+    {
+      "epoch": 0.036401513074941064,
+      "grad_norm": 1.1953450441360474,
+      "learning_rate": 9.999010754646748e-07,
+      "loss": 0.4217,
+      "step": 83
+    },
+    {
+      "epoch": 0.036840085521627104,
+      "grad_norm": 1.3063303232192993,
+      "learning_rate": 9.998864392252614e-07,
+      "loss": 0.4839,
+      "step": 84
+    },
+    {
+      "epoch": 0.03727865796831314,
+      "grad_norm": 1.1942335367202759,
+      "learning_rate": 9.998707937479682e-07,
+      "loss": 0.4443,
+      "step": 85
+    },
+    {
+      "epoch": 0.03771723041499918,
+      "grad_norm": 1.6526542901992798,
+      "learning_rate": 9.99854139064382e-07,
+      "loss": 0.4641,
+      "step": 86
+    },
+    {
+      "epoch": 0.03815580286168521,
+      "grad_norm": 1.4341564178466797,
+      "learning_rate": 9.998364752081277e-07,
+      "loss": 0.4928,
+      "step": 87
+    },
+    {
+      "epoch": 0.03859437530837125,
+      "grad_norm": 1.1596306562423706,
+      "learning_rate": 9.998178022148676e-07,
+      "loss": 0.4209,
+      "step": 88
+    },
+    {
+      "epoch": 0.039032947755057286,
+      "grad_norm": 1.2552282810211182,
+      "learning_rate": 9.997981201223009e-07,
+      "loss": 0.5077,
+      "step": 89
+    },
+    {
+      "epoch": 0.039471520201743326,
+      "grad_norm": 1.4964027404785156,
+      "learning_rate": 9.997774289701647e-07,
+      "loss": 0.4983,
+      "step": 90
+    },
+    {
+      "epoch": 0.03991009264842936,
+      "grad_norm": 1.221358060836792,
+      "learning_rate": 9.99755728800233e-07,
+      "loss": 0.4285,
+      "step": 91
+    },
+    {
+      "epoch": 0.0403486650951154,
+      "grad_norm": 1.2171251773834229,
+      "learning_rate": 9.997330196563169e-07,
+      "loss": 0.4797,
+      "step": 92
+    },
+    {
+      "epoch": 0.040787237541801434,
+      "grad_norm": 1.2266311645507812,
+      "learning_rate": 9.99709301584265e-07,
+      "loss": 0.4371,
+      "step": 93
+    },
+    {
+      "epoch": 0.041225809988487475,
+      "grad_norm": 1.4037154912948608,
+      "learning_rate": 9.99684574631962e-07,
+      "loss": 0.5075,
+      "step": 94
+    },
+    {
+      "epoch": 0.04166438243517351,
+      "grad_norm": 1.2918537855148315,
+      "learning_rate": 9.996588388493306e-07,
+      "loss": 0.4561,
+      "step": 95
+    },
+    {
+      "epoch": 0.04210295488185955,
+      "grad_norm": 1.422583818435669,
+      "learning_rate": 9.996320942883295e-07,
+      "loss": 0.4975,
+      "step": 96
+    },
+    {
+      "epoch": 0.04254152732854558,
+      "grad_norm": 1.3405108451843262,
+      "learning_rate": 9.996043410029537e-07,
+      "loss": 0.4725,
+      "step": 97
+    },
+    {
+      "epoch": 0.04298009977523162,
+      "grad_norm": 1.2558434009552002,
+      "learning_rate": 9.995755790492359e-07,
+      "loss": 0.4548,
+      "step": 98
+    },
+    {
+      "epoch": 0.043418672221917656,
+      "grad_norm": 1.3062516450881958,
+      "learning_rate": 9.99545808485244e-07,
+      "loss": 0.4432,
+      "step": 99
+    },
+    {
+      "epoch": 0.0438572446686037,
+      "grad_norm": 1.277416467666626,
+      "learning_rate": 9.995150293710838e-07,
+      "loss": 0.4464,
+      "step": 100
+    },
+    {
+      "epoch": 0.04429581711528973,
+      "grad_norm": 1.2594455480575562,
+      "learning_rate": 9.99483241768895e-07,
+      "loss": 0.4271,
+      "step": 101
+    },
+    {
+      "epoch": 0.04473438956197577,
+      "grad_norm": 1.4411756992340088,
+      "learning_rate": 9.994504457428556e-07,
+      "loss": 0.4759,
+      "step": 102
+    },
+    {
+      "epoch": 0.045172962008661804,
+      "grad_norm": 1.3192076683044434,
+      "learning_rate": 9.994166413591784e-07,
+      "loss": 0.4448,
+      "step": 103
+    },
+    {
+      "epoch": 0.045611534455347845,
+      "grad_norm": 1.247989296913147,
+      "learning_rate": 9.993818286861122e-07,
+      "loss": 0.491,
+      "step": 104
+    },
+    {
+      "epoch": 0.04605010690203388,
+      "grad_norm": 1.2985870838165283,
+      "learning_rate": 9.993460077939414e-07,
+      "loss": 0.4688,
+      "step": 105
+    },
+    {
+      "epoch": 0.04648867934871992,
+      "grad_norm": 1.3416420221328735,
+      "learning_rate": 9.993091787549862e-07,
+      "loss": 0.4958,
+      "step": 106
+    },
+    {
+      "epoch": 0.04692725179540595,
+      "grad_norm": 1.4006417989730835,
+      "learning_rate": 9.99271341643602e-07,
+      "loss": 0.4829,
+      "step": 107
+    },
+    {
+      "epoch": 0.04736582424209199,
+      "grad_norm": 1.422197699546814,
+      "learning_rate": 9.99232496536179e-07,
+      "loss": 0.4628,
+      "step": 108
+    },
+    {
+      "epoch": 0.04780439668877803,
+      "grad_norm": 1.2422194480895996,
+      "learning_rate": 9.991926435111437e-07,
+      "loss": 0.4086,
+      "step": 109
+    },
+    {
+      "epoch": 0.04824296913546407,
+      "grad_norm": 1.185133695602417,
+      "learning_rate": 9.99151782648956e-07,
+      "loss": 0.4374,
+      "step": 110
+    },
+    {
+      "epoch": 0.0486815415821501,
+      "grad_norm": 1.2822550535202026,
+      "learning_rate": 9.99109914032112e-07,
+      "loss": 0.457,
+      "step": 111
+    },
+    {
+      "epoch": 0.04912011402883614,
+      "grad_norm": 1.3902647495269775,
+      "learning_rate": 9.99067037745141e-07,
+      "loss": 0.4281,
+      "step": 112
+    },
+    {
+      "epoch": 0.049558686475522175,
+      "grad_norm": 1.244568943977356,
+      "learning_rate": 9.990231538746078e-07,
+      "loss": 0.443,
+      "step": 113
+    },
+    {
+      "epoch": 0.049997258922208215,
+      "grad_norm": 1.1944634914398193,
+      "learning_rate": 9.989782625091113e-07,
+      "loss": 0.3956,
+      "step": 114
+    },
+    {
+      "epoch": 0.05043583136889425,
+      "grad_norm": 1.330433964729309,
+      "learning_rate": 9.989323637392834e-07,
+      "loss": 0.5167,
+      "step": 115
+    },
+    {
+      "epoch": 0.05087440381558029,
+      "grad_norm": 1.2958663702011108,
+      "learning_rate": 9.988854576577913e-07,
+      "loss": 0.5158,
+      "step": 116
+    },
+    {
+      "epoch": 0.05131297626226632,
+      "grad_norm": 1.2222250699996948,
+      "learning_rate": 9.988375443593354e-07,
+      "loss": 0.4502,
+      "step": 117
+    },
+    {
+      "epoch": 0.051751548708952363,
+      "grad_norm": 1.3030247688293457,
+      "learning_rate": 9.987886239406491e-07,
+      "loss": 0.387,
+      "step": 118
+    },
+    {
+      "epoch": 0.0521901211556384,
+      "grad_norm": 1.281866431236267,
+      "learning_rate": 9.987386965004997e-07,
+      "loss": 0.4687,
+      "step": 119
+    },
+    {
+      "epoch": 0.05262869360232443,
+      "grad_norm": 1.273772120475769,
+      "learning_rate": 9.986877621396877e-07,
+      "loss": 0.4064,
+      "step": 120
+    },
+    {
+      "epoch": 0.05306726604901047,
+      "grad_norm": 1.3751946687698364,
+      "learning_rate": 9.986358209610457e-07,
+      "loss": 0.4919,
+      "step": 121
+    },
+    {
+      "epoch": 0.053505838495696505,
+      "grad_norm": 1.3724075555801392,
+      "learning_rate": 9.985828730694396e-07,
+      "loss": 0.4633,
+      "step": 122
+    },
+    {
+      "epoch": 0.053944410942382545,
+      "grad_norm": 1.2193125486373901,
+      "learning_rate": 9.985289185717683e-07,
+      "loss": 0.4356,
+      "step": 123
+    },
+    {
+      "epoch": 0.05438298338906858,
+      "grad_norm": 1.424628734588623,
+      "learning_rate": 9.984739575769617e-07,
+      "loss": 0.4395,
+      "step": 124
+    },
+    {
+      "epoch": 0.05482155583575462,
+      "grad_norm": 1.2202560901641846,
+      "learning_rate": 9.984179901959828e-07,
+      "loss": 0.4765,
+      "step": 125
+    },
+    {
+      "epoch": 0.05526012828244065,
+      "grad_norm": 1.3114961385726929,
+      "learning_rate": 9.983610165418259e-07,
+      "loss": 0.4197,
+      "step": 126
+    },
+    {
+      "epoch": 0.05569870072912669,
+      "grad_norm": 1.2029987573623657,
+      "learning_rate": 9.983030367295173e-07,
+      "loss": 0.429,
+      "step": 127
+    },
+    {
+      "epoch": 0.05613727317581273,
+      "grad_norm": 1.4868053197860718,
+      "learning_rate": 9.982440508761143e-07,
+      "loss": 0.4736,
+      "step": 128
+    },
+    {
+      "epoch": 0.05657584562249877,
+      "grad_norm": 1.212609887123108,
+      "learning_rate": 9.981840591007051e-07,
+      "loss": 0.4235,
+      "step": 129
+    },
+    {
+      "epoch": 0.0570144180691848,
+      "grad_norm": 1.3265501260757446,
+      "learning_rate": 9.981230615244099e-07,
+      "loss": 0.459,
+      "step": 130
+    },
+    {
+      "epoch": 0.05745299051587084,
+      "grad_norm": 1.2474690675735474,
+      "learning_rate": 9.980610582703782e-07,
+      "loss": 0.4498,
+      "step": 131
+    },
+    {
+      "epoch": 0.057891562962556875,
+      "grad_norm": 1.1922861337661743,
+      "learning_rate": 9.979980494637908e-07,
+      "loss": 0.4227,
+      "step": 132
+    },
+    {
+      "epoch": 0.058330135409242916,
+      "grad_norm": 1.351237416267395,
+      "learning_rate": 9.979340352318582e-07,
+      "loss": 0.441,
+      "step": 133
+    },
+    {
+      "epoch": 0.05876870785592895,
+      "grad_norm": 1.1860822439193726,
+      "learning_rate": 9.978690157038208e-07,
+      "loss": 0.412,
+      "step": 134
+    },
+    {
+      "epoch": 0.05920728030261499,
+      "grad_norm": 1.3116803169250488,
+      "learning_rate": 9.97802991010949e-07,
+      "loss": 0.5344,
+      "step": 135
+    },
+    {
+      "epoch": 0.05964585274930102,
+      "grad_norm": 1.3794771432876587,
+      "learning_rate": 9.977359612865422e-07,
+      "loss": 0.4982,
+      "step": 136
+    },
+    {
+      "epoch": 0.060084425195987064,
+      "grad_norm": 1.197117567062378,
+      "learning_rate": 9.976679266659292e-07,
+      "loss": 0.4201,
+      "step": 137
+    },
+    {
+      "epoch": 0.0605229976426731,
+      "grad_norm": 1.2389888763427734,
+      "learning_rate": 9.97598887286467e-07,
+      "loss": 0.4692,
+      "step": 138
+    },
+    {
+      "epoch": 0.06096157008935914,
+      "grad_norm": 1.3116774559020996,
+      "learning_rate": 9.975288432875422e-07,
+      "loss": 0.4854,
+      "step": 139
+    },
+    {
+      "epoch": 0.06140014253604517,
+      "grad_norm": 1.2350678443908691,
+      "learning_rate": 9.974577948105684e-07,
+      "loss": 0.4724,
+      "step": 140
+    },
+    {
+      "epoch": 0.06183871498273121,
+      "grad_norm": 1.3790651559829712,
+      "learning_rate": 9.973857419989881e-07,
+      "loss": 0.5353,
+      "step": 141
+    },
+    {
+      "epoch": 0.062277287429417245,
+      "grad_norm": 1.3146638870239258,
+      "learning_rate": 9.973126849982713e-07,
+      "loss": 0.4545,
+      "step": 142
+    },
+    {
+      "epoch": 0.06271585987610329,
+      "grad_norm": 1.4369558095932007,
+      "learning_rate": 9.972386239559152e-07,
+      "loss": 0.4897,
+      "step": 143
+    },
+    {
+      "epoch": 0.06315443232278932,
+      "grad_norm": 1.3575363159179688,
+      "learning_rate": 9.97163559021444e-07,
+      "loss": 0.4034,
+      "step": 144
+    },
+    {
+      "epoch": 0.06359300476947535,
+      "grad_norm": 1.2711018323898315,
+      "learning_rate": 9.97087490346409e-07,
+      "loss": 0.4584,
+      "step": 145
+    },
+    {
+      "epoch": 0.0640315772161614,
+      "grad_norm": 1.2842679023742676,
+      "learning_rate": 9.970104180843878e-07,
+      "loss": 0.4689,
+      "step": 146
+    },
+    {
+      "epoch": 0.06447014966284743,
+      "grad_norm": 1.2036633491516113,
+      "learning_rate": 9.969323423909846e-07,
+      "loss": 0.417,
+      "step": 147
+    },
+    {
+      "epoch": 0.06490872210953347,
+      "grad_norm": 1.199987769126892,
+      "learning_rate": 9.968532634238287e-07,
+      "loss": 0.4346,
+      "step": 148
+    },
+    {
+      "epoch": 0.0653472945562195,
+      "grad_norm": 1.280448317527771,
+      "learning_rate": 9.967731813425752e-07,
+      "loss": 0.4319,
+      "step": 149
+    },
+    {
+      "epoch": 0.06578586700290555,
+      "grad_norm": 1.2478801012039185,
+      "learning_rate": 9.966920963089051e-07,
+      "loss": 0.4424,
+      "step": 150
+    },
+    {
+      "epoch": 0.06622443944959158,
+      "grad_norm": 1.2198948860168457,
+      "learning_rate": 9.966100084865232e-07,
+      "loss": 0.4236,
+      "step": 151
+    },
+    {
+      "epoch": 0.06666301189627762,
+      "grad_norm": 1.28062903881073,
+      "learning_rate": 9.965269180411598e-07,
+      "loss": 0.4262,
+      "step": 152
+    },
+    {
+      "epoch": 0.06710158434296365,
+      "grad_norm": 1.137345790863037,
+      "learning_rate": 9.96442825140569e-07,
+      "loss": 0.4692,
+      "step": 153
+    },
+    {
+      "epoch": 0.0675401567896497,
+      "grad_norm": 1.2888954877853394,
+      "learning_rate": 9.963577299545286e-07,
+      "loss": 0.5076,
+      "step": 154
+    },
+    {
+      "epoch": 0.06797872923633573,
+      "grad_norm": 1.2218375205993652,
+      "learning_rate": 9.962716326548404e-07,
+      "loss": 0.4651,
+      "step": 155
+    },
+    {
+      "epoch": 0.06841730168302176,
+      "grad_norm": 1.3181045055389404,
+      "learning_rate": 9.961845334153293e-07,
+      "loss": 0.4151,
+      "step": 156
+    },
+    {
+      "epoch": 0.0688558741297078,
+      "grad_norm": 1.2577183246612549,
+      "learning_rate": 9.960964324118425e-07,
+      "loss": 0.4713,
+      "step": 157
+    },
+    {
+      "epoch": 0.06929444657639384,
+      "grad_norm": 1.2748234272003174,
+      "learning_rate": 9.960073298222508e-07,
+      "loss": 0.4551,
+      "step": 158
+    },
+    {
+      "epoch": 0.06973301902307988,
+      "grad_norm": 1.2437751293182373,
+      "learning_rate": 9.959172258264458e-07,
+      "loss": 0.4388,
+      "step": 159
+    },
+    {
+      "epoch": 0.07017159146976591,
+      "grad_norm": 1.3476532697677612,
+      "learning_rate": 9.95826120606342e-07,
+      "loss": 0.5177,
+      "step": 160
+    },
+    {
+      "epoch": 0.07061016391645195,
+      "grad_norm": 1.3056862354278564,
+      "learning_rate": 9.957340143458747e-07,
+      "loss": 0.4781,
+      "step": 161
+    },
+    {
+      "epoch": 0.07104873636313798,
+      "grad_norm": 1.4270120859146118,
+      "learning_rate": 9.956409072310004e-07,
+      "loss": 0.4588,
+      "step": 162
+    },
+    {
+      "epoch": 0.07148730880982403,
+      "grad_norm": 1.3080512285232544,
+      "learning_rate": 9.95546799449696e-07,
+      "loss": 0.4265,
+      "step": 163
+    },
+    {
+      "epoch": 0.07192588125651006,
+      "grad_norm": 1.4175631999969482,
+      "learning_rate": 9.954516911919595e-07,
+      "loss": 0.4487,
+      "step": 164
+    },
+    {
+      "epoch": 0.0723644537031961,
+      "grad_norm": 1.2619612216949463,
+      "learning_rate": 9.953555826498077e-07,
+      "loss": 0.4375,
+      "step": 165
+    },
+    {
+      "epoch": 0.07280302614988213,
+      "grad_norm": 1.301414966583252,
+      "learning_rate": 9.952584740172777e-07,
+      "loss": 0.4425,
+      "step": 166
+    },
+    {
+      "epoch": 0.07324159859656817,
+      "grad_norm": 1.2964798212051392,
+      "learning_rate": 9.951603654904254e-07,
+      "loss": 0.4687,
+      "step": 167
+    },
+    {
+      "epoch": 0.07368017104325421,
+      "grad_norm": 1.2225522994995117,
+      "learning_rate": 9.950612572673255e-07,
+      "loss": 0.4119,
+      "step": 168
+    },
+    {
+      "epoch": 0.07411874348994024,
+      "grad_norm": 1.2875328063964844,
+      "learning_rate": 9.949611495480708e-07,
+      "loss": 0.4658,
+      "step": 169
+    },
+    {
+      "epoch": 0.07455731593662628,
+      "grad_norm": 1.168454885482788,
+      "learning_rate": 9.948600425347724e-07,
+      "loss": 0.409,
+      "step": 170
+    },
+    {
+      "epoch": 0.07499588838331232,
+      "grad_norm": 1.3310086727142334,
+      "learning_rate": 9.947579364315587e-07,
+      "loss": 0.4657,
+      "step": 171
+    },
+    {
+      "epoch": 0.07543446082999836,
+      "grad_norm": 1.254629135131836,
+      "learning_rate": 9.946548314445751e-07,
+      "loss": 0.4406,
+      "step": 172
+    },
+    {
+      "epoch": 0.07587303327668439,
+      "grad_norm": 1.2079718112945557,
+      "learning_rate": 9.94550727781984e-07,
+      "loss": 0.4325,
+      "step": 173
+    },
+    {
+      "epoch": 0.07631160572337042,
+      "grad_norm": 1.2881447076797485,
+      "learning_rate": 9.944456256539636e-07,
+      "loss": 0.4459,
+      "step": 174
+    },
+    {
+      "epoch": 0.07675017817005647,
+      "grad_norm": 1.4596824645996094,
+      "learning_rate": 9.943395252727085e-07,
+      "loss": 0.4727,
+      "step": 175
+    },
+    {
+      "epoch": 0.0771887506167425,
+      "grad_norm": 1.2925268411636353,
+      "learning_rate": 9.94232426852428e-07,
+      "loss": 0.482,
+      "step": 176
+    },
+    {
+      "epoch": 0.07762732306342854,
+      "grad_norm": 1.2629894018173218,
+      "learning_rate": 9.94124330609347e-07,
+      "loss": 0.45,
+      "step": 177
+    },
+    {
+      "epoch": 0.07806589551011457,
+      "grad_norm": 1.238906741142273,
+      "learning_rate": 9.940152367617049e-07,
+      "loss": 0.4358,
+      "step": 178
+    },
+    {
+      "epoch": 0.07850446795680062,
+      "grad_norm": 1.33396315574646,
+      "learning_rate": 9.939051455297548e-07,
+      "loss": 0.4683,
+      "step": 179
+    },
+    {
+      "epoch": 0.07894304040348665,
+      "grad_norm": 1.4251528978347778,
+      "learning_rate": 9.937940571357636e-07,
+      "loss": 0.4883,
+      "step": 180
+    },
+    {
+      "epoch": 0.07938161285017269,
+      "grad_norm": 1.2432212829589844,
+      "learning_rate": 9.936819718040116e-07,
+      "loss": 0.4295,
+      "step": 181
+    },
+    {
+      "epoch": 0.07982018529685872,
+      "grad_norm": 1.2691503763198853,
+      "learning_rate": 9.935688897607915e-07,
+      "loss": 0.49,
+      "step": 182
+    },
+    {
+      "epoch": 0.08025875774354477,
+      "grad_norm": 1.3803023099899292,
+      "learning_rate": 9.934548112344087e-07,
+      "loss": 0.4467,
+      "step": 183
+    },
+    {
+      "epoch": 0.0806973301902308,
+      "grad_norm": 1.3414316177368164,
+      "learning_rate": 9.933397364551805e-07,
+      "loss": 0.4608,
+      "step": 184
+    },
+    {
+      "epoch": 0.08113590263691683,
+      "grad_norm": 1.3661813735961914,
+      "learning_rate": 9.93223665655435e-07,
+      "loss": 0.4901,
+      "step": 185
+    },
+    {
+      "epoch": 0.08157447508360287,
+      "grad_norm": 1.2298403978347778,
+      "learning_rate": 9.931065990695113e-07,
+      "loss": 0.4408,
+      "step": 186
+    },
+    {
+      "epoch": 0.08201304753028892,
+      "grad_norm": 1.1648515462875366,
+      "learning_rate": 9.929885369337596e-07,
+      "loss": 0.4097,
+      "step": 187
+    },
+    {
+      "epoch": 0.08245161997697495,
+      "grad_norm": 1.1954864263534546,
+      "learning_rate": 9.928694794865395e-07,
+      "loss": 0.4162,
+      "step": 188
+    },
+    {
+      "epoch": 0.08289019242366098,
+      "grad_norm": 1.2637189626693726,
+      "learning_rate": 9.9274942696822e-07,
+      "loss": 0.4361,
+      "step": 189
+    },
+    {
+      "epoch": 0.08332876487034702,
+      "grad_norm": 1.2444093227386475,
+      "learning_rate": 9.926283796211794e-07,
+      "loss": 0.4692,
+      "step": 190
+    },
+    {
+      "epoch": 0.08376733731703306,
+      "grad_norm": 1.282979965209961,
+      "learning_rate": 9.925063376898044e-07,
+      "loss": 0.3999,
+      "step": 191
+    },
+    {
+      "epoch": 0.0842059097637191,
+      "grad_norm": 1.2734895944595337,
+      "learning_rate": 9.923833014204893e-07,
+      "loss": 0.4343,
+      "step": 192
+    },
+    {
+      "epoch": 0.08464448221040513,
+      "grad_norm": 1.208871603012085,
+      "learning_rate": 9.922592710616364e-07,
+      "loss": 0.4414,
+      "step": 193
+    },
+    {
+      "epoch": 0.08508305465709116,
+      "grad_norm": 1.2236772775650024,
+      "learning_rate": 9.92134246863655e-07,
+      "loss": 0.4345,
+      "step": 194
+    },
+    {
+      "epoch": 0.08552162710377721,
+      "grad_norm": 1.2935453653335571,
+      "learning_rate": 9.920082290789607e-07,
+      "loss": 0.4637,
+      "step": 195
+    },
+    {
+      "epoch": 0.08596019955046325,
+      "grad_norm": 1.307611107826233,
+      "learning_rate": 9.91881217961975e-07,
+      "loss": 0.4666,
+      "step": 196
+    },
+    {
+      "epoch": 0.08639877199714928,
+      "grad_norm": 1.2390419244766235,
+      "learning_rate": 9.917532137691252e-07,
+      "loss": 0.4072,
+      "step": 197
+    },
+    {
+      "epoch": 0.08683734444383531,
+      "grad_norm": 1.3468120098114014,
+      "learning_rate": 9.916242167588432e-07,
+      "loss": 0.4486,
+      "step": 198
+    },
+    {
+      "epoch": 0.08727591689052135,
+      "grad_norm": 1.270461916923523,
+      "learning_rate": 9.914942271915655e-07,
+      "loss": 0.4438,
+      "step": 199
+    },
+    {
+      "epoch": 0.0877144893372074,
+      "grad_norm": 1.2275724411010742,
+      "learning_rate": 9.913632453297325e-07,
+      "loss": 0.4503,
+      "step": 200
+    },
+    {
+      "epoch": 0.08815306178389343,
+      "grad_norm": 1.2674055099487305,
+      "learning_rate": 9.912312714377879e-07,
+      "loss": 0.425,
+      "step": 201
+    },
+    {
+      "epoch": 0.08859163423057946,
+      "grad_norm": 1.2971616983413696,
+      "learning_rate": 9.910983057821786e-07,
+      "loss": 0.4633,
+      "step": 202
+    },
+    {
+      "epoch": 0.0890302066772655,
+      "grad_norm": 1.3597564697265625,
+      "learning_rate": 9.909643486313533e-07,
+      "loss": 0.4709,
+      "step": 203
+    },
+    {
+      "epoch": 0.08946877912395154,
+      "grad_norm": 1.403801441192627,
+      "learning_rate": 9.908294002557627e-07,
+      "loss": 0.4508,
+      "step": 204
+    },
+    {
+      "epoch": 0.08990735157063758,
+      "grad_norm": 1.1720523834228516,
+      "learning_rate": 9.906934609278588e-07,
+      "loss": 0.4182,
+      "step": 205
+    },
+    {
+      "epoch": 0.09034592401732361,
+      "grad_norm": 1.4372029304504395,
+      "learning_rate": 9.90556530922094e-07,
+      "loss": 0.4495,
+      "step": 206
+    },
+    {
+      "epoch": 0.09078449646400964,
+      "grad_norm": 1.2117644548416138,
+      "learning_rate": 9.904186105149211e-07,
+      "loss": 0.4486,
+      "step": 207
+    },
+    {
+      "epoch": 0.09122306891069569,
+      "grad_norm": 1.253118872642517,
+      "learning_rate": 9.902796999847923e-07,
+      "loss": 0.4683,
+      "step": 208
+    },
+    {
+      "epoch": 0.09166164135738172,
+      "grad_norm": 1.1815754175186157,
+      "learning_rate": 9.901397996121587e-07,
+      "loss": 0.4208,
+      "step": 209
+    },
+    {
+      "epoch": 0.09210021380406776,
+      "grad_norm": 1.1744418144226074,
+      "learning_rate": 9.899989096794704e-07,
+      "loss": 0.4086,
+      "step": 210
+    },
+    {
+      "epoch": 0.09253878625075379,
+      "grad_norm": 1.194718599319458,
+      "learning_rate": 9.898570304711746e-07,
+      "loss": 0.4343,
+      "step": 211
+    },
+    {
+      "epoch": 0.09297735869743984,
+      "grad_norm": 1.3020581007003784,
+      "learning_rate": 9.897141622737159e-07,
+      "loss": 0.4369,
+      "step": 212
+    },
+    {
+      "epoch": 0.09341593114412587,
+      "grad_norm": 1.3153276443481445,
+      "learning_rate": 9.895703053755363e-07,
+      "loss": 0.4609,
+      "step": 213
+    },
+    {
+      "epoch": 0.0938545035908119,
+      "grad_norm": 1.3371548652648926,
+      "learning_rate": 9.89425460067073e-07,
+      "loss": 0.459,
+      "step": 214
+    },
+    {
+      "epoch": 0.09429307603749794,
+      "grad_norm": 1.2718877792358398,
+      "learning_rate": 9.892796266407595e-07,
+      "loss": 0.4669,
+      "step": 215
+    },
+    {
+      "epoch": 0.09473164848418399,
+      "grad_norm": 1.2996882200241089,
+      "learning_rate": 9.891328053910237e-07,
+      "loss": 0.4598,
+      "step": 216
+    },
+    {
+      "epoch": 0.09517022093087002,
+      "grad_norm": 1.3045529127120972,
+      "learning_rate": 9.88984996614288e-07,
+      "loss": 0.4628,
+      "step": 217
+    },
+    {
+      "epoch": 0.09560879337755605,
+      "grad_norm": 1.2552134990692139,
+      "learning_rate": 9.888362006089688e-07,
+      "loss": 0.423,
+      "step": 218
+    },
+    {
+      "epoch": 0.09604736582424209,
+      "grad_norm": 1.3412526845932007,
+      "learning_rate": 9.886864176754754e-07,
+      "loss": 0.4708,
+      "step": 219
+    },
+    {
+      "epoch": 0.09648593827092813,
+      "grad_norm": 1.2526757717132568,
+      "learning_rate": 9.885356481162096e-07,
+      "loss": 0.4416,
+      "step": 220
+    },
+    {
+      "epoch": 0.09692451071761417,
+      "grad_norm": 1.2230411767959595,
+      "learning_rate": 9.883838922355653e-07,
+      "loss": 0.4032,
+      "step": 221
+    },
+    {
+      "epoch": 0.0973630831643002,
+      "grad_norm": 1.3482338190078735,
+      "learning_rate": 9.882311503399277e-07,
+      "loss": 0.445,
+      "step": 222
+    },
+    {
+      "epoch": 0.09780165561098624,
+      "grad_norm": 1.3570636510849,
+      "learning_rate": 9.880774227376725e-07,
+      "loss": 0.519,
+      "step": 223
+    },
+    {
+      "epoch": 0.09824022805767228,
+      "grad_norm": 1.2754287719726562,
+      "learning_rate": 9.879227097391658e-07,
+      "loss": 0.4919,
+      "step": 224
+    },
+    {
+      "epoch": 0.09867880050435832,
+      "grad_norm": 1.3529372215270996,
+      "learning_rate": 9.87767011656763e-07,
+      "loss": 0.4193,
+      "step": 225
+    },
+    {
+      "epoch": 0.09911737295104435,
+      "grad_norm": 1.3907129764556885,
+      "learning_rate": 9.876103288048084e-07,
+      "loss": 0.4202,
+      "step": 226
+    },
+    {
+      "epoch": 0.09955594539773038,
+      "grad_norm": 1.2041593790054321,
+      "learning_rate": 9.87452661499634e-07,
+      "loss": 0.4326,
+      "step": 227
+    },
+    {
+      "epoch": 0.09999451784441643,
+      "grad_norm": 1.3893024921417236,
+      "learning_rate": 9.872940100595597e-07,
+      "loss": 0.4933,
+      "step": 228
+    },
+    {
+      "epoch": 0.10043309029110246,
+      "grad_norm": 1.1358956098556519,
+      "learning_rate": 9.871343748048929e-07,
+      "loss": 0.3958,
+      "step": 229
+    },
+    {
+      "epoch": 0.1008716627377885,
+      "grad_norm": 1.258034586906433,
+      "learning_rate": 9.869737560579262e-07,
+      "loss": 0.4579,
+      "step": 230
+    },
+    {
+      "epoch": 0.10131023518447453,
+      "grad_norm": 1.3062182664871216,
+      "learning_rate": 9.868121541429386e-07,
+      "loss": 0.4671,
+      "step": 231
+    },
+    {
+      "epoch": 0.10174880763116058,
+      "grad_norm": 1.2269024848937988,
+      "learning_rate": 9.866495693861934e-07,
+      "loss": 0.4541,
+      "step": 232
+    },
+    {
+      "epoch": 0.10218738007784661,
+      "grad_norm": 1.3577250242233276,
+      "learning_rate": 9.86486002115939e-07,
+      "loss": 0.5296,
+      "step": 233
+    },
+    {
+      "epoch": 0.10262595252453265,
+      "grad_norm": 1.330257534980774,
+      "learning_rate": 9.863214526624063e-07,
+      "loss": 0.4678,
+      "step": 234
+    },
+    {
+      "epoch": 0.10306452497121868,
+      "grad_norm": 1.1935516595840454,
+      "learning_rate": 9.861559213578107e-07,
+      "loss": 0.4826,
+      "step": 235
+    },
+    {
+      "epoch": 0.10350309741790473,
+      "grad_norm": 1.4533356428146362,
+      "learning_rate": 9.859894085363485e-07,
+      "loss": 0.5029,
+      "step": 236
+    },
+    {
+      "epoch": 0.10394166986459076,
+      "grad_norm": 1.274011254310608,
+      "learning_rate": 9.85821914534198e-07,
+      "loss": 0.4471,
+      "step": 237
+    },
+    {
+      "epoch": 0.1043802423112768,
+      "grad_norm": 1.3055455684661865,
+      "learning_rate": 9.856534396895193e-07,
+      "loss": 0.4521,
+      "step": 238
+    },
+    {
+      "epoch": 0.10481881475796283,
+      "grad_norm": 1.290225625038147,
+      "learning_rate": 9.854839843424512e-07,
+      "loss": 0.4444,
+      "step": 239
+    },
+    {
+      "epoch": 0.10525738720464886,
+      "grad_norm": 1.256276249885559,
+      "learning_rate": 9.853135488351132e-07,
+      "loss": 0.4397,
+      "step": 240
+    },
+    {
+      "epoch": 0.10569595965133491,
+      "grad_norm": 1.2009029388427734,
+      "learning_rate": 9.851421335116036e-07,
+      "loss": 0.464,
+      "step": 241
+    },
+    {
+      "epoch": 0.10613453209802094,
+      "grad_norm": 1.3538483381271362,
+      "learning_rate": 9.849697387179987e-07,
+      "loss": 0.4543,
+      "step": 242
+    },
+    {
+      "epoch": 0.10657310454470698,
+      "grad_norm": 1.314831018447876,
+      "learning_rate": 9.84796364802352e-07,
+      "loss": 0.4675,
+      "step": 243
+    },
+    {
+      "epoch": 0.10701167699139301,
+      "grad_norm": 1.334395408630371,
+      "learning_rate": 9.846220121146943e-07,
+      "loss": 0.4541,
+      "step": 244
+    },
+    {
+      "epoch": 0.10745024943807906,
+      "grad_norm": 1.2025914192199707,
+      "learning_rate": 9.844466810070317e-07,
+      "loss": 0.3998,
+      "step": 245
+    },
+    {
+      "epoch": 0.10788882188476509,
+      "grad_norm": 1.438745141029358,
+      "learning_rate": 9.842703718333468e-07,
+      "loss": 0.4303,
+      "step": 246
+    },
+    {
+      "epoch": 0.10832739433145112,
+      "grad_norm": 1.2495355606079102,
+      "learning_rate": 9.84093084949596e-07,
+      "loss": 0.4415,
+      "step": 247
+    },
+    {
+      "epoch": 0.10876596677813716,
+      "grad_norm": 1.3335902690887451,
+      "learning_rate": 9.8391482071371e-07,
+      "loss": 0.4923,
+      "step": 248
+    },
+    {
+      "epoch": 0.1092045392248232,
+      "grad_norm": 1.359126091003418,
+      "learning_rate": 9.837355794855923e-07,
+      "loss": 0.4598,
+      "step": 249
+    },
+    {
+      "epoch": 0.10964311167150924,
+      "grad_norm": 1.1839964389801025,
+      "learning_rate": 9.835553616271194e-07,
+      "loss": 0.4053,
+      "step": 250
+    },
+    {
+      "epoch": 0.11008168411819527,
+      "grad_norm": 1.2885066270828247,
+      "learning_rate": 9.83374167502139e-07,
+      "loss": 0.4521,
+      "step": 251
+    },
+    {
+      "epoch": 0.1105202565648813,
+      "grad_norm": 1.381494402885437,
+      "learning_rate": 9.83191997476471e-07,
+      "loss": 0.5172,
+      "step": 252
+    },
+    {
+      "epoch": 0.11095882901156735,
+      "grad_norm": 1.3523175716400146,
+      "learning_rate": 9.830088519179035e-07,
+      "loss": 0.4813,
+      "step": 253
+    },
+    {
+      "epoch": 0.11139740145825339,
+      "grad_norm": 1.1988216638565063,
+      "learning_rate": 9.82824731196196e-07,
+      "loss": 0.4377,
+      "step": 254
+    },
+    {
+      "epoch": 0.11183597390493942,
+      "grad_norm": 1.1344822645187378,
+      "learning_rate": 9.826396356830764e-07,
+      "loss": 0.4638,
+      "step": 255
+    },
+    {
+      "epoch": 0.11227454635162545,
+      "grad_norm": 1.2201296091079712,
+      "learning_rate": 9.824535657522397e-07,
+      "loss": 0.4378,
+      "step": 256
+    },
+    {
+      "epoch": 0.1127131187983115,
+      "grad_norm": 1.4678555727005005,
+      "learning_rate": 9.822665217793496e-07,
+      "loss": 0.4644,
+      "step": 257
+    },
+    {
+      "epoch": 0.11315169124499753,
+      "grad_norm": 1.251297950744629,
+      "learning_rate": 9.820785041420348e-07,
+      "loss": 0.4329,
+      "step": 258
+    },
+    {
+      "epoch": 0.11359026369168357,
+      "grad_norm": 1.2199487686157227,
+      "learning_rate": 9.818895132198913e-07,
+      "loss": 0.4474,
+      "step": 259
+    },
+    {
+      "epoch": 0.1140288361383696,
+      "grad_norm": 1.311081051826477,
+      "learning_rate": 9.81699549394479e-07,
+      "loss": 0.4531,
+      "step": 260
+    },
+    {
+      "epoch": 0.11446740858505565,
+      "grad_norm": 1.3128409385681152,
+      "learning_rate": 9.815086130493221e-07,
+      "loss": 0.4166,
+      "step": 261
+    },
+    {
+      "epoch": 0.11490598103174168,
+      "grad_norm": 1.300299882888794,
+      "learning_rate": 9.81316704569909e-07,
+      "loss": 0.4667,
+      "step": 262
+    },
+    {
+      "epoch": 0.11534455347842772,
+      "grad_norm": 1.1987988948822021,
+      "learning_rate": 9.811238243436904e-07,
+      "loss": 0.4499,
+      "step": 263
+    },
+    {
+      "epoch": 0.11578312592511375,
+      "grad_norm": 1.2745237350463867,
+      "learning_rate": 9.809299727600783e-07,
+      "loss": 0.4198,
+      "step": 264
+    },
+    {
+      "epoch": 0.1162216983717998,
+      "grad_norm": 1.3099387884140015,
+      "learning_rate": 9.807351502104468e-07,
+      "loss": 0.4695,
+      "step": 265
+    },
+    {
+      "epoch": 0.11666027081848583,
+      "grad_norm": 1.3721712827682495,
+      "learning_rate": 9.805393570881295e-07,
+      "loss": 0.4978,
+      "step": 266
+    },
+    {
+      "epoch": 0.11709884326517186,
+      "grad_norm": 1.3487995862960815,
+      "learning_rate": 9.8034259378842e-07,
+      "loss": 0.4425,
+      "step": 267
+    },
+    {
+      "epoch": 0.1175374157118579,
+      "grad_norm": 1.3157742023468018,
+      "learning_rate": 9.801448607085704e-07,
+      "loss": 0.4446,
+      "step": 268
+    },
+    {
+      "epoch": 0.11797598815854395,
+      "grad_norm": 1.368003487586975,
+      "learning_rate": 9.799461582477909e-07,
+      "loss": 0.462,
+      "step": 269
+    },
+    {
+      "epoch": 0.11841456060522998,
+      "grad_norm": 1.254192590713501,
+      "learning_rate": 9.797464868072486e-07,
+      "loss": 0.4711,
+      "step": 270
+    },
+    {
+      "epoch": 0.11885313305191601,
+      "grad_norm": 1.3297892808914185,
+      "learning_rate": 9.795458467900672e-07,
+      "loss": 0.4489,
+      "step": 271
+    },
+    {
+      "epoch": 0.11929170549860205,
+      "grad_norm": 1.3227829933166504,
+      "learning_rate": 9.793442386013255e-07,
+      "loss": 0.4494,
+      "step": 272
+    },
+    {
+      "epoch": 0.1197302779452881,
+      "grad_norm": 1.3050018548965454,
+      "learning_rate": 9.79141662648057e-07,
+      "loss": 0.4533,
+      "step": 273
+    },
+    {
+      "epoch": 0.12016885039197413,
+      "grad_norm": 1.2946937084197998,
+      "learning_rate": 9.789381193392498e-07,
+      "loss": 0.4566,
+      "step": 274
+    },
+    {
+      "epoch": 0.12060742283866016,
+      "grad_norm": 1.1722913980484009,
+      "learning_rate": 9.787336090858441e-07,
+      "loss": 0.4402,
+      "step": 275
+    },
+    {
+      "epoch": 0.1210459952853462,
+      "grad_norm": 1.3350423574447632,
+      "learning_rate": 9.78528132300733e-07,
+      "loss": 0.4497,
+      "step": 276
+    },
+    {
+      "epoch": 0.12148456773203223,
+      "grad_norm": 1.2697076797485352,
+      "learning_rate": 9.7832168939876e-07,
+      "loss": 0.4525,
+      "step": 277
+    },
+    {
+      "epoch": 0.12192314017871828,
+      "grad_norm": 1.3002448081970215,
+      "learning_rate": 9.781142807967204e-07,
+      "loss": 0.4955,
+      "step": 278
+    },
+    {
+      "epoch": 0.12236171262540431,
+      "grad_norm": 1.319963812828064,
+      "learning_rate": 9.779059069133582e-07,
+      "loss": 0.4588,
+      "step": 279
+    },
+    {
+      "epoch": 0.12280028507209034,
+      "grad_norm": 1.4620566368103027,
+      "learning_rate": 9.776965681693666e-07,
+      "loss": 0.4813,
+      "step": 280
+    },
+    {
+      "epoch": 0.12323885751877638,
+      "grad_norm": 1.4124107360839844,
+      "learning_rate": 9.774862649873868e-07,
+      "loss": 0.4535,
+      "step": 281
+    },
+    {
+      "epoch": 0.12367742996546242,
+      "grad_norm": 1.3538119792938232,
+      "learning_rate": 9.772749977920071e-07,
+      "loss": 0.4501,
+      "step": 282
+    },
+    {
+      "epoch": 0.12411600241214846,
+      "grad_norm": 1.2902806997299194,
+      "learning_rate": 9.770627670097623e-07,
+      "loss": 0.4242,
+      "step": 283
+    },
+    {
+      "epoch": 0.12455457485883449,
+      "grad_norm": 1.2163552045822144,
+      "learning_rate": 9.768495730691321e-07,
+      "loss": 0.4376,
+      "step": 284
+    },
+    {
+      "epoch": 0.12499314730552052,
+      "grad_norm": 1.2694215774536133,
+      "learning_rate": 9.766354164005414e-07,
+      "loss": 0.4893,
+      "step": 285
+    },
+    {
+      "epoch": 0.12543171975220657,
+      "grad_norm": 1.3819620609283447,
+      "learning_rate": 9.76420297436358e-07,
+      "loss": 0.4442,
+      "step": 286
+    },
+    {
+      "epoch": 0.1258702921988926,
+      "grad_norm": 1.2240216732025146,
+      "learning_rate": 9.762042166108932e-07,
+      "loss": 0.4411,
+      "step": 287
+    },
+    {
+      "epoch": 0.12630886464557864,
+      "grad_norm": 1.3175026178359985,
+      "learning_rate": 9.759871743604001e-07,
+      "loss": 0.3919,
+      "step": 288
+    },
+    {
+      "epoch": 0.1267474370922647,
+      "grad_norm": 1.219504714012146,
+      "learning_rate": 9.757691711230727e-07,
+      "loss": 0.4241,
+      "step": 289
+    },
+    {
+      "epoch": 0.1271860095389507,
+      "grad_norm": 1.389089584350586,
+      "learning_rate": 9.75550207339045e-07,
+      "loss": 0.4885,
+      "step": 290
+    },
+    {
+      "epoch": 0.12762458198563675,
+      "grad_norm": 1.3088550567626953,
+      "learning_rate": 9.753302834503908e-07,
+      "loss": 0.4106,
+      "step": 291
+    },
+    {
+      "epoch": 0.1280631544323228,
+      "grad_norm": 1.2590484619140625,
+      "learning_rate": 9.751093999011216e-07,
+      "loss": 0.472,
+      "step": 292
+    },
+    {
+      "epoch": 0.12850172687900882,
+      "grad_norm": 1.2284749746322632,
+      "learning_rate": 9.74887557137187e-07,
+      "loss": 0.4192,
+      "step": 293
+    },
+    {
+      "epoch": 0.12894029932569487,
+      "grad_norm": 1.2713497877120972,
+      "learning_rate": 9.746647556064732e-07,
+      "loss": 0.4552,
+      "step": 294
+    },
+    {
+      "epoch": 0.1293788717723809,
+      "grad_norm": 1.276537537574768,
+      "learning_rate": 9.744409957588014e-07,
+      "loss": 0.4545,
+      "step": 295
+    },
+    {
+      "epoch": 0.12981744421906694,
+      "grad_norm": 1.3731132745742798,
+      "learning_rate": 9.742162780459281e-07,
+      "loss": 0.4143,
+      "step": 296
+    },
+    {
+      "epoch": 0.13025601666575298,
+      "grad_norm": 1.246085524559021,
+      "learning_rate": 9.73990602921544e-07,
+      "loss": 0.4145,
+      "step": 297
+    },
+    {
+      "epoch": 0.130694589112439,
+      "grad_norm": 1.2259469032287598,
+      "learning_rate": 9.737639708412721e-07,
+      "loss": 0.4474,
+      "step": 298
+    },
+    {
+      "epoch": 0.13113316155912505,
+      "grad_norm": 1.3381538391113281,
+      "learning_rate": 9.735363822626676e-07,
+      "loss": 0.4735,
+      "step": 299
+    },
+    {
+      "epoch": 0.1315717340058111,
+      "grad_norm": 1.2572318315505981,
+      "learning_rate": 9.73307837645217e-07,
+      "loss": 0.406,
+      "step": 300
+    },
+    {
+      "epoch": 0.13201030645249712,
+      "grad_norm": 1.3827505111694336,
+      "learning_rate": 9.730783374503369e-07,
+      "loss": 0.405,
+      "step": 301
+    },
+    {
+      "epoch": 0.13244887889918316,
+      "grad_norm": 1.2806872129440308,
+      "learning_rate": 9.728478821413728e-07,
+      "loss": 0.439,
+      "step": 302
+    },
+    {
+      "epoch": 0.13288745134586918,
+      "grad_norm": 1.2340537309646606,
+      "learning_rate": 9.726164721835995e-07,
+      "loss": 0.4199,
+      "step": 303
+    },
+    {
+      "epoch": 0.13332602379255523,
+      "grad_norm": 1.3651092052459717,
+      "learning_rate": 9.723841080442176e-07,
+      "loss": 0.4593,
+      "step": 304
+    },
+    {
+      "epoch": 0.13376459623924128,
+      "grad_norm": 1.4623818397521973,
+      "learning_rate": 9.721507901923559e-07,
+      "loss": 0.4769,
+      "step": 305
+    },
+    {
+      "epoch": 0.1342031686859273,
+      "grad_norm": 1.2197152376174927,
+      "learning_rate": 9.719165190990673e-07,
+      "loss": 0.4333,
+      "step": 306
+    },
+    {
+      "epoch": 0.13464174113261335,
+      "grad_norm": 1.3367750644683838,
+      "learning_rate": 9.716812952373297e-07,
+      "loss": 0.4313,
+      "step": 307
+    },
+    {
+      "epoch": 0.1350803135792994,
+      "grad_norm": 1.3246334791183472,
+      "learning_rate": 9.714451190820449e-07,
+      "loss": 0.4228,
+      "step": 308
+    },
+    {
+      "epoch": 0.1355188860259854,
+      "grad_norm": 1.2900984287261963,
+      "learning_rate": 9.71207991110037e-07,
+      "loss": 0.414,
+      "step": 309
+    },
+    {
+      "epoch": 0.13595745847267146,
+      "grad_norm": 1.2274436950683594,
+      "learning_rate": 9.709699118000517e-07,
+      "loss": 0.4576,
+      "step": 310
+    },
+    {
+      "epoch": 0.13639603091935748,
+      "grad_norm": 1.288940191268921,
+      "learning_rate": 9.707308816327556e-07,
+      "loss": 0.4692,
+      "step": 311
+    },
+    {
+      "epoch": 0.13683460336604353,
+      "grad_norm": 1.3872264623641968,
+      "learning_rate": 9.704909010907348e-07,
+      "loss": 0.4913,
+      "step": 312
+    },
+    {
+      "epoch": 0.13727317581272958,
+      "grad_norm": 1.2140966653823853,
+      "learning_rate": 9.702499706584943e-07,
+      "loss": 0.4335,
+      "step": 313
+    },
+    {
+      "epoch": 0.1377117482594156,
+      "grad_norm": 1.399438738822937,
+      "learning_rate": 9.700080908224567e-07,
+      "loss": 0.4552,
+      "step": 314
+    },
+    {
+      "epoch": 0.13815032070610164,
+      "grad_norm": 1.334566354751587,
+      "learning_rate": 9.697652620709615e-07,
+      "loss": 0.4204,
+      "step": 315
+    },
+    {
+      "epoch": 0.1385888931527877,
+      "grad_norm": 1.1500535011291504,
+      "learning_rate": 9.695214848942641e-07,
+      "loss": 0.419,
+      "step": 316
+    },
+    {
+      "epoch": 0.1390274655994737,
+      "grad_norm": 1.257188081741333,
+      "learning_rate": 9.692767597845349e-07,
+      "loss": 0.4364,
+      "step": 317
+    },
+    {
+      "epoch": 0.13946603804615976,
+      "grad_norm": 1.3126678466796875,
+      "learning_rate": 9.690310872358571e-07,
+      "loss": 0.4729,
+      "step": 318
+    },
+    {
+      "epoch": 0.13990461049284578,
+      "grad_norm": 1.2673767805099487,
+      "learning_rate": 9.687844677442282e-07,
+      "loss": 0.4365,
+      "step": 319
+    },
+    {
+      "epoch": 0.14034318293953182,
+      "grad_norm": 1.3040989637374878,
+      "learning_rate": 9.685369018075562e-07,
+      "loss": 0.4715,
+      "step": 320
+    },
+    {
+      "epoch": 0.14078175538621787,
+      "grad_norm": 1.3866188526153564,
+      "learning_rate": 9.682883899256607e-07,
+      "loss": 0.438,
+      "step": 321
+    },
+    {
+      "epoch": 0.1412203278329039,
+      "grad_norm": 1.2850979566574097,
+      "learning_rate": 9.680389326002707e-07,
+      "loss": 0.4325,
+      "step": 322
+    },
+    {
+      "epoch": 0.14165890027958994,
+      "grad_norm": 1.244288682937622,
+      "learning_rate": 9.677885303350244e-07,
+      "loss": 0.447,
+      "step": 323
+    },
+    {
+      "epoch": 0.14209747272627596,
+      "grad_norm": 1.327620029449463,
+      "learning_rate": 9.675371836354673e-07,
+      "loss": 0.4371,
+      "step": 324
+    },
+    {
+      "epoch": 0.142536045172962,
+      "grad_norm": 1.311292290687561,
+      "learning_rate": 9.672848930090522e-07,
+      "loss": 0.4506,
+      "step": 325
+    },
+    {
+      "epoch": 0.14297461761964805,
+      "grad_norm": 1.2936956882476807,
+      "learning_rate": 9.670316589651367e-07,
+      "loss": 0.4778,
+      "step": 326
+    },
+    {
+      "epoch": 0.14341319006633407,
+      "grad_norm": 1.2305073738098145,
+      "learning_rate": 9.667774820149843e-07,
+      "loss": 0.4253,
+      "step": 327
+    },
+    {
+      "epoch": 0.14385176251302012,
+      "grad_norm": 1.366324782371521,
+      "learning_rate": 9.665223626717613e-07,
+      "loss": 0.463,
+      "step": 328
+    },
+    {
+      "epoch": 0.14429033495970617,
+      "grad_norm": 1.3423768281936646,
+      "learning_rate": 9.66266301450537e-07,
+      "loss": 0.4535,
+      "step": 329
+    },
+    {
+      "epoch": 0.1447289074063922,
+      "grad_norm": 1.348443627357483,
+      "learning_rate": 9.66009298868282e-07,
+      "loss": 0.4935,
+      "step": 330
+    },
+    {
+      "epoch": 0.14516747985307824,
+      "grad_norm": 1.1361678838729858,
+      "learning_rate": 9.657513554438677e-07,
+      "loss": 0.3826,
+      "step": 331
+    },
+    {
+      "epoch": 0.14560605229976425,
+      "grad_norm": 1.1119284629821777,
+      "learning_rate": 9.65492471698065e-07,
+      "loss": 0.4397,
+      "step": 332
+    },
+    {
+      "epoch": 0.1460446247464503,
+      "grad_norm": 1.3069874048233032,
+      "learning_rate": 9.652326481535433e-07,
+      "loss": 0.4402,
+      "step": 333
+    },
+    {
+      "epoch": 0.14648319719313635,
+      "grad_norm": 1.3529688119888306,
+      "learning_rate": 9.64971885334869e-07,
+      "loss": 0.4139,
+      "step": 334
+    },
+    {
+      "epoch": 0.14692176963982237,
+      "grad_norm": 1.2486492395401,
+      "learning_rate": 9.647101837685052e-07,
+      "loss": 0.3979,
+      "step": 335
+    },
+    {
+      "epoch": 0.14736034208650842,
+      "grad_norm": 1.3289217948913574,
+      "learning_rate": 9.644475439828102e-07,
+      "loss": 0.451,
+      "step": 336
+    },
+    {
+      "epoch": 0.14779891453319446,
+      "grad_norm": 1.3076781034469604,
+      "learning_rate": 9.641839665080363e-07,
+      "loss": 0.482,
+      "step": 337
+    },
+    {
+      "epoch": 0.14823748697988048,
+      "grad_norm": 1.285366415977478,
+      "learning_rate": 9.639194518763294e-07,
+      "loss": 0.4418,
+      "step": 338
+    },
+    {
+      "epoch": 0.14867605942656653,
+      "grad_norm": 1.2070653438568115,
+      "learning_rate": 9.636540006217268e-07,
+      "loss": 0.439,
+      "step": 339
+    },
+    {
+      "epoch": 0.14911463187325255,
+      "grad_norm": 1.2946048974990845,
+      "learning_rate": 9.633876132801577e-07,
+      "loss": 0.4141,
+      "step": 340
+    },
+    {
+      "epoch": 0.1495532043199386,
+      "grad_norm": 1.31601881980896,
+      "learning_rate": 9.6312029038944e-07,
+      "loss": 0.4488,
+      "step": 341
+    },
+    {
+      "epoch": 0.14999177676662465,
+      "grad_norm": 1.2885645627975464,
+      "learning_rate": 9.628520324892816e-07,
+      "loss": 0.4398,
+      "step": 342
+    },
+    {
+      "epoch": 0.15043034921331067,
+      "grad_norm": 1.262040138244629,
+      "learning_rate": 9.625828401212772e-07,
+      "loss": 0.4839,
+      "step": 343
+    },
+    {
+      "epoch": 0.1508689216599967,
+      "grad_norm": 1.273505687713623,
+      "learning_rate": 9.623127138289087e-07,
+      "loss": 0.419,
+      "step": 344
+    },
+    {
+      "epoch": 0.15130749410668276,
+      "grad_norm": 1.374241590499878,
+      "learning_rate": 9.620416541575432e-07,
+      "loss": 0.4567,
+      "step": 345
+    },
+    {
+      "epoch": 0.15174606655336878,
+      "grad_norm": 1.118963360786438,
+      "learning_rate": 9.617696616544325e-07,
+      "loss": 0.3974,
+      "step": 346
+    },
+    {
+      "epoch": 0.15218463900005483,
+      "grad_norm": 1.2339282035827637,
+      "learning_rate": 9.614967368687115e-07,
+      "loss": 0.4428,
+      "step": 347
+    },
+    {
+      "epoch": 0.15262321144674085,
+      "grad_norm": 1.2473440170288086,
+      "learning_rate": 9.612228803513975e-07,
+      "loss": 0.4359,
+      "step": 348
+    },
+    {
+      "epoch": 0.1530617838934269,
+      "grad_norm": 1.1782147884368896,
+      "learning_rate": 9.609480926553887e-07,
+      "loss": 0.4493,
+      "step": 349
+    },
+    {
+      "epoch": 0.15350035634011294,
+      "grad_norm": 1.233314037322998,
+      "learning_rate": 9.606723743354637e-07,
+      "loss": 0.4602,
+      "step": 350
+    },
+    {
+      "epoch": 0.15393892878679896,
+      "grad_norm": 1.267967700958252,
+      "learning_rate": 9.60395725948279e-07,
+      "loss": 0.4777,
+      "step": 351
+    },
+    {
+      "epoch": 0.154377501233485,
+      "grad_norm": 1.3795247077941895,
+      "learning_rate": 9.601181480523702e-07,
+      "loss": 0.425,
+      "step": 352
+    },
+    {
+      "epoch": 0.15481607368017106,
+      "grad_norm": 1.411054253578186,
+      "learning_rate": 9.598396412081488e-07,
+      "loss": 0.479,
+      "step": 353
+    },
+    {
+      "epoch": 0.15525464612685708,
+      "grad_norm": 1.2811428308486938,
+      "learning_rate": 9.595602059779015e-07,
+      "loss": 0.4635,
+      "step": 354
+    },
+    {
+      "epoch": 0.15569321857354312,
+      "grad_norm": 1.3159114122390747,
+      "learning_rate": 9.592798429257899e-07,
+      "loss": 0.444,
+      "step": 355
+    },
+    {
+      "epoch": 0.15613179102022914,
+      "grad_norm": 1.1596899032592773,
+      "learning_rate": 9.589985526178484e-07,
+      "loss": 0.3732,
+      "step": 356
+    },
+    {
+      "epoch": 0.1565703634669152,
+      "grad_norm": 1.2319432497024536,
+      "learning_rate": 9.587163356219836e-07,
+      "loss": 0.426,
+      "step": 357
+    },
+    {
+      "epoch": 0.15700893591360124,
+      "grad_norm": 1.502126932144165,
+      "learning_rate": 9.584331925079734e-07,
+      "loss": 0.4757,
+      "step": 358
+    },
+    {
+      "epoch": 0.15744750836028726,
+      "grad_norm": 1.2502065896987915,
+      "learning_rate": 9.58149123847465e-07,
+      "loss": 0.4382,
+      "step": 359
+    },
+    {
+      "epoch": 0.1578860808069733,
+      "grad_norm": 1.2441209554672241,
+      "learning_rate": 9.578641302139742e-07,
+      "loss": 0.431,
+      "step": 360
+    },
+    {
+      "epoch": 0.15832465325365933,
+      "grad_norm": 1.2779874801635742,
+      "learning_rate": 9.575782121828845e-07,
+      "loss": 0.4354,
+      "step": 361
+    },
+    {
+      "epoch": 0.15876322570034537,
+      "grad_norm": 1.3990627527236938,
+      "learning_rate": 9.572913703314454e-07,
+      "loss": 0.4955,
+      "step": 362
+    },
+    {
+      "epoch": 0.15920179814703142,
+      "grad_norm": 1.2426034212112427,
+      "learning_rate": 9.570036052387724e-07,
+      "loss": 0.4749,
+      "step": 363
+    },
+    {
+      "epoch": 0.15964037059371744,
+      "grad_norm": 1.1871353387832642,
+      "learning_rate": 9.567149174858438e-07,
+      "loss": 0.4638,
+      "step": 364
+    },
+    {
+      "epoch": 0.1600789430404035,
+      "grad_norm": 1.3922412395477295,
+      "learning_rate": 9.564253076555013e-07,
+      "loss": 0.4672,
+      "step": 365
+    },
+    {
+      "epoch": 0.16051751548708953,
+      "grad_norm": 1.1922954320907593,
+      "learning_rate": 9.561347763324483e-07,
+      "loss": 0.4251,
+      "step": 366
+    },
+    {
+      "epoch": 0.16095608793377555,
+      "grad_norm": 1.1256937980651855,
+      "learning_rate": 9.558433241032483e-07,
+      "loss": 0.4816,
+      "step": 367
+    },
+    {
+      "epoch": 0.1613946603804616,
+      "grad_norm": 1.2206835746765137,
+      "learning_rate": 9.55550951556324e-07,
+      "loss": 0.442,
+      "step": 368
+    },
+    {
+      "epoch": 0.16183323282714762,
+      "grad_norm": 1.230305790901184,
+      "learning_rate": 9.55257659281957e-07,
+      "loss": 0.4258,
+      "step": 369
+    },
+    {
+      "epoch": 0.16227180527383367,
+      "grad_norm": 1.1930097341537476,
+      "learning_rate": 9.549634478722843e-07,
+      "loss": 0.4303,
+      "step": 370
+    },
+    {
+      "epoch": 0.16271037772051972,
+      "grad_norm": 1.1612834930419922,
+      "learning_rate": 9.546683179213e-07,
+      "loss": 0.4636,
+      "step": 371
+    },
+    {
+      "epoch": 0.16314895016720574,
+      "grad_norm": 1.4160550832748413,
+      "learning_rate": 9.54372270024852e-07,
+      "loss": 0.4658,
+      "step": 372
+    },
+    {
+      "epoch": 0.16358752261389178,
+      "grad_norm": 1.2340151071548462,
+      "learning_rate": 9.540753047806413e-07,
+      "loss": 0.4572,
+      "step": 373
+    },
+    {
+      "epoch": 0.16402609506057783,
+      "grad_norm": 1.3071835041046143,
+      "learning_rate": 9.537774227882215e-07,
+      "loss": 0.4409,
+      "step": 374
+    },
+    {
+      "epoch": 0.16446466750726385,
+      "grad_norm": 1.381076455116272,
+      "learning_rate": 9.534786246489966e-07,
+      "loss": 0.466,
+      "step": 375
+    },
+    {
+      "epoch": 0.1649032399539499,
+      "grad_norm": 1.2368775606155396,
+      "learning_rate": 9.531789109662204e-07,
+      "loss": 0.4359,
+      "step": 376
+    },
+    {
+      "epoch": 0.16534181240063592,
+      "grad_norm": 1.3651288747787476,
+      "learning_rate": 9.528782823449953e-07,
+      "loss": 0.4409,
+      "step": 377
+    },
+    {
+      "epoch": 0.16578038484732197,
+      "grad_norm": 1.466973900794983,
+      "learning_rate": 9.525767393922706e-07,
+      "loss": 0.4804,
+      "step": 378
+    },
+    {
+      "epoch": 0.166218957294008,
+      "grad_norm": 1.3845632076263428,
+      "learning_rate": 9.522742827168416e-07,
+      "loss": 0.4723,
+      "step": 379
+    },
+    {
+      "epoch": 0.16665752974069403,
+      "grad_norm": 1.242058277130127,
+      "learning_rate": 9.519709129293488e-07,
+      "loss": 0.4361,
+      "step": 380
+    },
+    {
+      "epoch": 0.16709610218738008,
+      "grad_norm": 1.2678145170211792,
+      "learning_rate": 9.516666306422755e-07,
+      "loss": 0.4524,
+      "step": 381
+    },
+    {
+      "epoch": 0.16753467463406613,
+      "grad_norm": 1.2828023433685303,
+      "learning_rate": 9.51361436469948e-07,
+      "loss": 0.4527,
+      "step": 382
+    },
+    {
+      "epoch": 0.16797324708075215,
+      "grad_norm": 1.3856935501098633,
+      "learning_rate": 9.510553310285331e-07,
+      "loss": 0.4676,
+      "step": 383
+    },
+    {
+      "epoch": 0.1684118195274382,
+      "grad_norm": 1.3793553113937378,
+      "learning_rate": 9.507483149360375e-07,
+      "loss": 0.4231,
+      "step": 384
+    },
+    {
+      "epoch": 0.16885039197412421,
+      "grad_norm": 1.2514710426330566,
+      "learning_rate": 9.504403888123066e-07,
+      "loss": 0.4119,
+      "step": 385
+    },
+    {
+      "epoch": 0.16928896442081026,
+      "grad_norm": 1.2418358325958252,
+      "learning_rate": 9.50131553279023e-07,
+      "loss": 0.4563,
+      "step": 386
+    },
+    {
+      "epoch": 0.1697275368674963,
+      "grad_norm": 1.291680097579956,
+      "learning_rate": 9.498218089597054e-07,
+      "loss": 0.4411,
+      "step": 387
+    },
+    {
+      "epoch": 0.17016610931418233,
+      "grad_norm": 1.2920475006103516,
+      "learning_rate": 9.495111564797073e-07,
+      "loss": 0.4543,
+      "step": 388
+    },
+    {
+      "epoch": 0.17060468176086838,
+      "grad_norm": 1.341176152229309,
+      "learning_rate": 9.491995964662154e-07,
+      "loss": 0.4597,
+      "step": 389
+    },
+    {
+      "epoch": 0.17104325420755442,
+      "grad_norm": 1.2180360555648804,
+      "learning_rate": 9.488871295482491e-07,
+      "loss": 0.4578,
+      "step": 390
+    },
+    {
+      "epoch": 0.17148182665424044,
+      "grad_norm": 1.3022912740707397,
+      "learning_rate": 9.485737563566585e-07,
+      "loss": 0.4191,
+      "step": 391
+    },
+    {
+      "epoch": 0.1719203991009265,
+      "grad_norm": 1.261922001838684,
+      "learning_rate": 9.482594775241236e-07,
+      "loss": 0.4556,
+      "step": 392
+    },
+    {
+      "epoch": 0.1723589715476125,
+      "grad_norm": 1.392162799835205,
+      "learning_rate": 9.479442936851526e-07,
+      "loss": 0.456,
+      "step": 393
+    },
+    {
+      "epoch": 0.17279754399429856,
+      "grad_norm": 1.3650363683700562,
+      "learning_rate": 9.476282054760809e-07,
+      "loss": 0.4934,
+      "step": 394
+    },
+    {
+      "epoch": 0.1732361164409846,
+      "grad_norm": 1.3042758703231812,
+      "learning_rate": 9.4731121353507e-07,
+      "loss": 0.4519,
+      "step": 395
+    },
+    {
+      "epoch": 0.17367468888767063,
+      "grad_norm": 1.3327823877334595,
+      "learning_rate": 9.469933185021058e-07,
+      "loss": 0.4212,
+      "step": 396
+    },
+    {
+      "epoch": 0.17411326133435667,
+      "grad_norm": 1.2559716701507568,
+      "learning_rate": 9.466745210189972e-07,
+      "loss": 0.4643,
+      "step": 397
+    },
+    {
+      "epoch": 0.1745518337810427,
+      "grad_norm": 1.188941478729248,
+      "learning_rate": 9.463548217293759e-07,
+      "loss": 0.4303,
+      "step": 398
+    },
+    {
+      "epoch": 0.17499040622772874,
+      "grad_norm": 1.3942642211914062,
+      "learning_rate": 9.460342212786932e-07,
+      "loss": 0.4702,
+      "step": 399
+    },
+    {
+      "epoch": 0.1754289786744148,
+      "grad_norm": 1.2623578310012817,
+      "learning_rate": 9.457127203142206e-07,
+      "loss": 0.4399,
+      "step": 400
+    },
+    {
+      "epoch": 0.1758675511211008,
+      "grad_norm": 1.3886511325836182,
+      "learning_rate": 9.453903194850475e-07,
+      "loss": 0.4714,
+      "step": 401
+    },
+    {
+      "epoch": 0.17630612356778685,
+      "grad_norm": 1.1634944677352905,
+      "learning_rate": 9.4506701944208e-07,
+      "loss": 0.4144,
+      "step": 402
+    },
+    {
+      "epoch": 0.1767446960144729,
+      "grad_norm": 1.3607139587402344,
+      "learning_rate": 9.447428208380395e-07,
+      "loss": 0.4516,
+      "step": 403
+    },
+    {
+      "epoch": 0.17718326846115892,
+      "grad_norm": 1.351024866104126,
+      "learning_rate": 9.444177243274617e-07,
+      "loss": 0.4194,
+      "step": 404
+    },
+    {
+      "epoch": 0.17762184090784497,
+      "grad_norm": 1.2279891967773438,
+      "learning_rate": 9.440917305666951e-07,
+      "loss": 0.482,
+      "step": 405
+    },
+    {
+      "epoch": 0.178060413354531,
+      "grad_norm": 1.2650517225265503,
+      "learning_rate": 9.437648402138998e-07,
+      "loss": 0.414,
+      "step": 406
+    },
+    {
+      "epoch": 0.17849898580121704,
+      "grad_norm": 1.3450144529342651,
+      "learning_rate": 9.434370539290459e-07,
+      "loss": 0.4864,
+      "step": 407
+    },
+    {
+      "epoch": 0.17893755824790308,
+      "grad_norm": 1.2379168272018433,
+      "learning_rate": 9.431083723739124e-07,
+      "loss": 0.4029,
+      "step": 408
+    },
+    {
+      "epoch": 0.1793761306945891,
+      "grad_norm": 1.3272401094436646,
+      "learning_rate": 9.427787962120857e-07,
+      "loss": 0.4524,
+      "step": 409
+    },
+    {
+      "epoch": 0.17981470314127515,
+      "grad_norm": 1.3062082529067993,
+      "learning_rate": 9.424483261089583e-07,
+      "loss": 0.4385,
+      "step": 410
+    },
+    {
+      "epoch": 0.1802532755879612,
+      "grad_norm": 1.2583765983581543,
+      "learning_rate": 9.421169627317278e-07,
+      "loss": 0.4247,
+      "step": 411
+    },
+    {
+      "epoch": 0.18069184803464722,
+      "grad_norm": 1.2366396188735962,
+      "learning_rate": 9.417847067493952e-07,
+      "loss": 0.4399,
+      "step": 412
+    },
+    {
+      "epoch": 0.18113042048133327,
+      "grad_norm": 1.334314227104187,
+      "learning_rate": 9.414515588327631e-07,
+      "loss": 0.438,
+      "step": 413
+    },
+    {
+      "epoch": 0.18156899292801928,
+      "grad_norm": 1.3514381647109985,
+      "learning_rate": 9.411175196544358e-07,
+      "loss": 0.5151,
+      "step": 414
+    },
+    {
+      "epoch": 0.18200756537470533,
+      "grad_norm": 1.2838636636734009,
+      "learning_rate": 9.40782589888816e-07,
+      "loss": 0.4465,
+      "step": 415
+    },
+    {
+      "epoch": 0.18244613782139138,
+      "grad_norm": 1.25214684009552,
+      "learning_rate": 9.404467702121051e-07,
+      "loss": 0.3996,
+      "step": 416
+    },
+    {
+      "epoch": 0.1828847102680774,
+      "grad_norm": 1.323198676109314,
+      "learning_rate": 9.40110061302301e-07,
+      "loss": 0.4093,
+      "step": 417
+    },
+    {
+      "epoch": 0.18332328271476345,
+      "grad_norm": 1.3139373064041138,
+      "learning_rate": 9.397724638391967e-07,
+      "loss": 0.4789,
+      "step": 418
+    },
+    {
+      "epoch": 0.1837618551614495,
+      "grad_norm": 1.1997904777526855,
+      "learning_rate": 9.394339785043794e-07,
+      "loss": 0.4365,
+      "step": 419
+    },
+    {
+      "epoch": 0.18420042760813551,
+      "grad_norm": 1.297034502029419,
+      "learning_rate": 9.390946059812289e-07,
+      "loss": 0.4518,
+      "step": 420
+    },
+    {
+      "epoch": 0.18463900005482156,
+      "grad_norm": 1.1115834712982178,
+      "learning_rate": 9.387543469549155e-07,
+      "loss": 0.4254,
+      "step": 421
+    },
+    {
+      "epoch": 0.18507757250150758,
+      "grad_norm": 1.141788363456726,
+      "learning_rate": 9.384132021124004e-07,
+      "loss": 0.3842,
+      "step": 422
+    },
+    {
+      "epoch": 0.18551614494819363,
+      "grad_norm": 1.3769128322601318,
+      "learning_rate": 9.380711721424326e-07,
+      "loss": 0.4629,
+      "step": 423
+    },
+    {
+      "epoch": 0.18595471739487968,
+      "grad_norm": 1.3783560991287231,
+      "learning_rate": 9.377282577355478e-07,
+      "loss": 0.4906,
+      "step": 424
+    },
+    {
+      "epoch": 0.1863932898415657,
+      "grad_norm": 1.255490779876709,
+      "learning_rate": 9.373844595840678e-07,
+      "loss": 0.4537,
+      "step": 425
+    },
+    {
+      "epoch": 0.18683186228825174,
+      "grad_norm": 1.354235053062439,
+      "learning_rate": 9.370397783820984e-07,
+      "loss": 0.4497,
+      "step": 426
+    },
+    {
+      "epoch": 0.1872704347349378,
+      "grad_norm": 1.347273826599121,
+      "learning_rate": 9.366942148255285e-07,
+      "loss": 0.4267,
+      "step": 427
+    },
+    {
+      "epoch": 0.1877090071816238,
+      "grad_norm": 1.2848392724990845,
+      "learning_rate": 9.363477696120283e-07,
+      "loss": 0.4738,
+      "step": 428
+    },
+    {
+      "epoch": 0.18814757962830986,
+      "grad_norm": 1.2392444610595703,
+      "learning_rate": 9.360004434410476e-07,
+      "loss": 0.4834,
+      "step": 429
+    },
+    {
+      "epoch": 0.18858615207499588,
+      "grad_norm": 1.195491909980774,
+      "learning_rate": 9.356522370138154e-07,
+      "loss": 0.3701,
+      "step": 430
+    },
+    {
+      "epoch": 0.18902472452168192,
+      "grad_norm": 1.2303284406661987,
+      "learning_rate": 9.353031510333373e-07,
+      "loss": 0.4281,
+      "step": 431
+    },
+    {
+      "epoch": 0.18946329696836797,
+      "grad_norm": 1.2569408416748047,
+      "learning_rate": 9.349531862043951e-07,
+      "loss": 0.4603,
+      "step": 432
+    },
+    {
+      "epoch": 0.189901869415054,
+      "grad_norm": 1.3959214687347412,
+      "learning_rate": 9.346023432335449e-07,
+      "loss": 0.4663,
+      "step": 433
+    },
+    {
+      "epoch": 0.19034044186174004,
+      "grad_norm": 1.31195068359375,
+      "learning_rate": 9.342506228291156e-07,
+      "loss": 0.4729,
+      "step": 434
+    },
+    {
+      "epoch": 0.1907790143084261,
+      "grad_norm": 1.33975088596344,
+      "learning_rate": 9.338980257012074e-07,
+      "loss": 0.4647,
+      "step": 435
+    },
+    {
+      "epoch": 0.1912175867551121,
+      "grad_norm": 1.283033013343811,
+      "learning_rate": 9.335445525616909e-07,
+      "loss": 0.4047,
+      "step": 436
+    },
+    {
+      "epoch": 0.19165615920179815,
+      "grad_norm": 1.341052770614624,
+      "learning_rate": 9.331902041242053e-07,
+      "loss": 0.4434,
+      "step": 437
+    },
+    {
+      "epoch": 0.19209473164848417,
+      "grad_norm": 1.3923901319503784,
+      "learning_rate": 9.328349811041564e-07,
+      "loss": 0.4544,
+      "step": 438
+    },
+    {
+      "epoch": 0.19253330409517022,
+      "grad_norm": 1.2661969661712646,
+      "learning_rate": 9.324788842187163e-07,
+      "loss": 0.4461,
+      "step": 439
+    },
+    {
+      "epoch": 0.19297187654185627,
+      "grad_norm": 1.2110265493392944,
+      "learning_rate": 9.321219141868215e-07,
+      "loss": 0.4707,
+      "step": 440
+    },
+    {
+      "epoch": 0.1934104489885423,
+      "grad_norm": 1.4487905502319336,
+      "learning_rate": 9.317640717291708e-07,
+      "loss": 0.4779,
+      "step": 441
+    },
+    {
+      "epoch": 0.19384902143522834,
+      "grad_norm": 1.3091298341751099,
+      "learning_rate": 9.314053575682246e-07,
+      "loss": 0.4674,
+      "step": 442
+    },
+    {
+      "epoch": 0.19428759388191436,
+      "grad_norm": 1.2794967889785767,
+      "learning_rate": 9.310457724282033e-07,
+      "loss": 0.3946,
+      "step": 443
+    },
+    {
+      "epoch": 0.1947261663286004,
+      "grad_norm": 1.2558822631835938,
+      "learning_rate": 9.306853170350854e-07,
+      "loss": 0.4358,
+      "step": 444
+    },
+    {
+      "epoch": 0.19516473877528645,
+      "grad_norm": 1.1746054887771606,
+      "learning_rate": 9.303239921166071e-07,
+      "loss": 0.434,
+      "step": 445
+    },
+    {
+      "epoch": 0.19560331122197247,
+      "grad_norm": 1.2854063510894775,
+      "learning_rate": 9.299617984022597e-07,
+      "loss": 0.4739,
+      "step": 446
+    },
+    {
+      "epoch": 0.19604188366865852,
+      "grad_norm": 1.214192509651184,
+      "learning_rate": 9.29598736623288e-07,
+      "loss": 0.4272,
+      "step": 447
+    },
+    {
+      "epoch": 0.19648045611534457,
+      "grad_norm": 1.248357892036438,
+      "learning_rate": 9.292348075126901e-07,
+      "loss": 0.431,
+      "step": 448
+    },
+    {
+      "epoch": 0.19691902856203058,
+      "grad_norm": 1.2610982656478882,
+      "learning_rate": 9.288700118052151e-07,
+      "loss": 0.4582,
+      "step": 449
+    },
+    {
+      "epoch": 0.19735760100871663,
+      "grad_norm": 1.3101803064346313,
+      "learning_rate": 9.285043502373615e-07,
+      "loss": 0.4896,
+      "step": 450
+    },
+    {
+      "epoch": 0.19779617345540265,
+      "grad_norm": 1.4143675565719604,
+      "learning_rate": 9.281378235473761e-07,
+      "loss": 0.4681,
+      "step": 451
+    },
+    {
+      "epoch": 0.1982347459020887,
+      "grad_norm": 1.2853251695632935,
+      "learning_rate": 9.27770432475252e-07,
+      "loss": 0.4826,
+      "step": 452
+    },
+    {
+      "epoch": 0.19867331834877475,
+      "grad_norm": 1.3457460403442383,
+      "learning_rate": 9.274021777627276e-07,
+      "loss": 0.4796,
+      "step": 453
+    },
+    {
+      "epoch": 0.19911189079546077,
+      "grad_norm": 1.3269720077514648,
+      "learning_rate": 9.270330601532854e-07,
+      "loss": 0.4544,
+      "step": 454
+    },
+    {
+      "epoch": 0.1995504632421468,
+      "grad_norm": 1.2965871095657349,
+      "learning_rate": 9.266630803921491e-07,
+      "loss": 0.4134,
+      "step": 455
+    },
+    {
+      "epoch": 0.19998903568883286,
+      "grad_norm": 1.250986933708191,
+      "learning_rate": 9.262922392262837e-07,
+      "loss": 0.4538,
+      "step": 456
+    },
+    {
+      "epoch": 0.20042760813551888,
+      "grad_norm": 1.3280874490737915,
+      "learning_rate": 9.259205374043933e-07,
+      "loss": 0.457,
+      "step": 457
+    },
+    {
+      "epoch": 0.20086618058220493,
+      "grad_norm": 1.3324693441390991,
+      "learning_rate": 9.255479756769193e-07,
+      "loss": 0.4568,
+      "step": 458
+    },
+    {
+      "epoch": 0.20130475302889095,
+      "grad_norm": 1.4459820985794067,
+      "learning_rate": 9.251745547960393e-07,
+      "loss": 0.5189,
+      "step": 459
+    },
+    {
+      "epoch": 0.201743325475577,
+      "grad_norm": 1.2799947261810303,
+      "learning_rate": 9.248002755156659e-07,
+      "loss": 0.4259,
+      "step": 460
+    },
+    {
+      "epoch": 0.20218189792226304,
+      "grad_norm": 1.3187311887741089,
+      "learning_rate": 9.244251385914437e-07,
+      "loss": 0.4845,
+      "step": 461
+    },
+    {
+      "epoch": 0.20262047036894906,
+      "grad_norm": 1.3244918584823608,
+      "learning_rate": 9.240491447807501e-07,
+      "loss": 0.4272,
+      "step": 462
+    },
+    {
+      "epoch": 0.2030590428156351,
+      "grad_norm": 1.355553150177002,
+      "learning_rate": 9.236722948426918e-07,
+      "loss": 0.4544,
+      "step": 463
+    },
+    {
+      "epoch": 0.20349761526232116,
+      "grad_norm": 1.241315245628357,
+      "learning_rate": 9.232945895381039e-07,
+      "loss": 0.419,
+      "step": 464
+    },
+    {
+      "epoch": 0.20393618770900718,
+      "grad_norm": 1.3401538133621216,
+      "learning_rate": 9.229160296295487e-07,
+      "loss": 0.3983,
+      "step": 465
+    },
+    {
+      "epoch": 0.20437476015569322,
+      "grad_norm": 1.4097830057144165,
+      "learning_rate": 9.225366158813138e-07,
+      "loss": 0.4733,
+      "step": 466
+    },
+    {
+      "epoch": 0.20481333260237924,
+      "grad_norm": 1.2525343894958496,
+      "learning_rate": 9.221563490594103e-07,
+      "loss": 0.4384,
+      "step": 467
+    },
+    {
+      "epoch": 0.2052519050490653,
+      "grad_norm": 1.1235498189926147,
+      "learning_rate": 9.217752299315724e-07,
+      "loss": 0.4004,
+      "step": 468
+    },
+    {
+      "epoch": 0.20569047749575134,
+      "grad_norm": 1.2831324338912964,
+      "learning_rate": 9.213932592672544e-07,
+      "loss": 0.4974,
+      "step": 469
+    },
+    {
+      "epoch": 0.20612904994243736,
+      "grad_norm": 1.2940030097961426,
+      "learning_rate": 9.2101043783763e-07,
+      "loss": 0.4909,
+      "step": 470
+    },
+    {
+      "epoch": 0.2065676223891234,
+      "grad_norm": 1.3095605373382568,
+      "learning_rate": 9.206267664155906e-07,
+      "loss": 0.4604,
+      "step": 471
+    },
+    {
+      "epoch": 0.20700619483580945,
+      "grad_norm": 1.2282830476760864,
+      "learning_rate": 9.202422457757433e-07,
+      "loss": 0.4773,
+      "step": 472
+    },
+    {
+      "epoch": 0.20744476728249547,
+      "grad_norm": 1.3521174192428589,
+      "learning_rate": 9.198568766944102e-07,
+      "loss": 0.4535,
+      "step": 473
+    },
+    {
+      "epoch": 0.20788333972918152,
+      "grad_norm": 1.260043978691101,
+      "learning_rate": 9.194706599496262e-07,
+      "loss": 0.4328,
+      "step": 474
+    },
+    {
+      "epoch": 0.20832191217586754,
+      "grad_norm": 1.3410409688949585,
+      "learning_rate": 9.190835963211376e-07,
+      "loss": 0.4879,
+      "step": 475
+    },
+    {
+      "epoch": 0.2087604846225536,
+      "grad_norm": 1.3885812759399414,
+      "learning_rate": 9.186956865904003e-07,
+      "loss": 0.4449,
+      "step": 476
+    },
+    {
+      "epoch": 0.20919905706923964,
+      "grad_norm": 1.3543665409088135,
+      "learning_rate": 9.183069315405784e-07,
+      "loss": 0.5023,
+      "step": 477
+    },
+    {
+      "epoch": 0.20963762951592566,
+      "grad_norm": 1.3189713954925537,
+      "learning_rate": 9.179173319565432e-07,
+      "loss": 0.4523,
+      "step": 478
+    },
+    {
+      "epoch": 0.2100762019626117,
+      "grad_norm": 1.262413501739502,
+      "learning_rate": 9.175268886248704e-07,
+      "loss": 0.4632,
+      "step": 479
+    },
+    {
+      "epoch": 0.21051477440929772,
+      "grad_norm": 1.3755730390548706,
+      "learning_rate": 9.171356023338395e-07,
+      "loss": 0.4475,
+      "step": 480
+    },
+    {
+      "epoch": 0.21095334685598377,
+      "grad_norm": 1.2970229387283325,
+      "learning_rate": 9.167434738734318e-07,
+      "loss": 0.4604,
+      "step": 481
+    },
+    {
+      "epoch": 0.21139191930266982,
+      "grad_norm": 1.2604976892471313,
+      "learning_rate": 9.163505040353287e-07,
+      "loss": 0.5005,
+      "step": 482
+    },
+    {
+      "epoch": 0.21183049174935584,
+      "grad_norm": 1.4260404109954834,
+      "learning_rate": 9.159566936129111e-07,
+      "loss": 0.4655,
+      "step": 483
+    },
+    {
+      "epoch": 0.21226906419604188,
+      "grad_norm": 1.4161478281021118,
+      "learning_rate": 9.155620434012556e-07,
+      "loss": 0.4763,
+      "step": 484
+    },
+    {
+      "epoch": 0.21270763664272793,
+      "grad_norm": 1.3012006282806396,
+      "learning_rate": 9.151665541971356e-07,
+      "loss": 0.4265,
+      "step": 485
+    },
+    {
+      "epoch": 0.21314620908941395,
+      "grad_norm": 1.3242242336273193,
+      "learning_rate": 9.147702267990177e-07,
+      "loss": 0.4468,
+      "step": 486
+    },
+    {
+      "epoch": 0.2135847815361,
+      "grad_norm": 1.350150465965271,
+      "learning_rate": 9.143730620070608e-07,
+      "loss": 0.4252,
+      "step": 487
+    },
+    {
+      "epoch": 0.21402335398278602,
+      "grad_norm": 1.3396698236465454,
+      "learning_rate": 9.139750606231145e-07,
+      "loss": 0.4238,
+      "step": 488
+    },
+    {
+      "epoch": 0.21446192642947207,
+      "grad_norm": 1.2412779331207275,
+      "learning_rate": 9.135762234507174e-07,
+      "loss": 0.4363,
+      "step": 489
+    },
+    {
+      "epoch": 0.2149004988761581,
+      "grad_norm": 1.2845489978790283,
+      "learning_rate": 9.131765512950958e-07,
+      "loss": 0.447,
+      "step": 490
+    },
+    {
+      "epoch": 0.21533907132284413,
+      "grad_norm": 1.3290574550628662,
+      "learning_rate": 9.127760449631612e-07,
+      "loss": 0.5195,
+      "step": 491
+    },
+    {
+      "epoch": 0.21577764376953018,
+      "grad_norm": 1.325376033782959,
+      "learning_rate": 9.123747052635098e-07,
+      "loss": 0.4565,
+      "step": 492
+    },
+    {
+      "epoch": 0.21621621621621623,
+      "grad_norm": 1.39933180809021,
+      "learning_rate": 9.119725330064201e-07,
+      "loss": 0.4568,
+      "step": 493
+    },
+    {
+      "epoch": 0.21665478866290225,
+      "grad_norm": 1.3452569246292114,
+      "learning_rate": 9.115695290038513e-07,
+      "loss": 0.4797,
+      "step": 494
+    },
+    {
+      "epoch": 0.2170933611095883,
+      "grad_norm": 1.330816388130188,
+      "learning_rate": 9.111656940694421e-07,
+      "loss": 0.4302,
+      "step": 495
+    },
+    {
+      "epoch": 0.21753193355627432,
+      "grad_norm": 1.2303009033203125,
+      "learning_rate": 9.107610290185087e-07,
+      "loss": 0.4297,
+      "step": 496
+    },
+    {
+      "epoch": 0.21797050600296036,
+      "grad_norm": 1.2003905773162842,
+      "learning_rate": 9.103555346680433e-07,
+      "loss": 0.4407,
+      "step": 497
+    },
+    {
+      "epoch": 0.2184090784496464,
+      "grad_norm": 1.2907445430755615,
+      "learning_rate": 9.099492118367122e-07,
+      "loss": 0.4384,
+      "step": 498
+    },
+    {
+      "epoch": 0.21884765089633243,
+      "grad_norm": 1.2537270784378052,
+      "learning_rate": 9.095420613448548e-07,
+      "loss": 0.4237,
+      "step": 499
+    },
+    {
+      "epoch": 0.21928622334301848,
+      "grad_norm": 1.3271981477737427,
+      "learning_rate": 9.091340840144807e-07,
+      "loss": 0.4436,
+      "step": 500
+    },
+    {
+      "epoch": 0.21972479578970452,
+      "grad_norm": 1.2225565910339355,
+      "learning_rate": 9.087252806692699e-07,
+      "loss": 0.4514,
+      "step": 501
+    },
+    {
+      "epoch": 0.22016336823639054,
+      "grad_norm": 1.273189663887024,
+      "learning_rate": 9.083156521345692e-07,
+      "loss": 0.4481,
+      "step": 502
+    },
+    {
+      "epoch": 0.2206019406830766,
+      "grad_norm": 1.3194257020950317,
+      "learning_rate": 9.079051992373916e-07,
+      "loss": 0.4672,
+      "step": 503
+    },
+    {
+      "epoch": 0.2210405131297626,
+      "grad_norm": 1.24362051486969,
+      "learning_rate": 9.074939228064147e-07,
+      "loss": 0.4002,
+      "step": 504
+    },
+    {
+      "epoch": 0.22147908557644866,
+      "grad_norm": 1.4185460805892944,
+      "learning_rate": 9.070818236719785e-07,
+      "loss": 0.4555,
+      "step": 505
+    },
+    {
+      "epoch": 0.2219176580231347,
+      "grad_norm": 1.2558276653289795,
+      "learning_rate": 9.066689026660842e-07,
+      "loss": 0.4683,
+      "step": 506
+    },
+    {
+      "epoch": 0.22235623046982073,
+      "grad_norm": 1.3859413862228394,
+      "learning_rate": 9.062551606223921e-07,
+      "loss": 0.4426,
+      "step": 507
+    },
+    {
+      "epoch": 0.22279480291650677,
+      "grad_norm": 1.3655338287353516,
+      "learning_rate": 9.058405983762201e-07,
+      "loss": 0.405,
+      "step": 508
+    },
+    {
+      "epoch": 0.22323337536319282,
+      "grad_norm": 1.2751038074493408,
+      "learning_rate": 9.054252167645425e-07,
+      "loss": 0.4462,
+      "step": 509
+    },
+    {
+      "epoch": 0.22367194780987884,
+      "grad_norm": 1.3899219036102295,
+      "learning_rate": 9.050090166259872e-07,
+      "loss": 0.412,
+      "step": 510
+    },
+    {
+      "epoch": 0.2241105202565649,
+      "grad_norm": 1.3597509860992432,
+      "learning_rate": 9.045919988008348e-07,
+      "loss": 0.4391,
+      "step": 511
+    },
+    {
+      "epoch": 0.2245490927032509,
+      "grad_norm": 1.4710191488265991,
+      "learning_rate": 9.041741641310172e-07,
+      "loss": 0.44,
+      "step": 512
+    },
+    {
+      "epoch": 0.22498766514993696,
+      "grad_norm": 1.2938218116760254,
+      "learning_rate": 9.037555134601149e-07,
+      "loss": 0.433,
+      "step": 513
+    },
+    {
+      "epoch": 0.225426237596623,
+      "grad_norm": 1.3852367401123047,
+      "learning_rate": 9.033360476333565e-07,
+      "loss": 0.4891,
+      "step": 514
+    },
+    {
+      "epoch": 0.22586481004330902,
+      "grad_norm": 1.2105050086975098,
+      "learning_rate": 9.029157674976154e-07,
+      "loss": 0.4549,
+      "step": 515
+    },
+    {
+      "epoch": 0.22630338248999507,
+      "grad_norm": 1.290735125541687,
+      "learning_rate": 9.0249467390141e-07,
+      "loss": 0.436,
+      "step": 516
+    },
+    {
+      "epoch": 0.2267419549366811,
+      "grad_norm": 1.3402304649353027,
+      "learning_rate": 9.020727676949004e-07,
+      "loss": 0.4032,
+      "step": 517
+    },
+    {
+      "epoch": 0.22718052738336714,
+      "grad_norm": 1.257806420326233,
+      "learning_rate": 9.016500497298876e-07,
+      "loss": 0.4829,
+      "step": 518
+    },
+    {
+      "epoch": 0.22761909983005318,
+      "grad_norm": 1.203140139579773,
+      "learning_rate": 9.012265208598113e-07,
+      "loss": 0.4132,
+      "step": 519
+    },
+    {
+      "epoch": 0.2280576722767392,
+      "grad_norm": 1.2450907230377197,
+      "learning_rate": 9.008021819397486e-07,
+      "loss": 0.4349,
+      "step": 520
+    },
+    {
+      "epoch": 0.22849624472342525,
+      "grad_norm": 1.3185664415359497,
+      "learning_rate": 9.003770338264118e-07,
+      "loss": 0.4742,
+      "step": 521
+    },
+    {
+      "epoch": 0.2289348171701113,
+      "grad_norm": 1.2235329151153564,
+      "learning_rate": 8.999510773781471e-07,
+      "loss": 0.4181,
+      "step": 522
+    },
+    {
+      "epoch": 0.22937338961679732,
+      "grad_norm": 1.2654236555099487,
+      "learning_rate": 8.995243134549326e-07,
+      "loss": 0.4767,
+      "step": 523
+    },
+    {
+      "epoch": 0.22981196206348337,
+      "grad_norm": 1.2756022214889526,
+      "learning_rate": 8.990967429183765e-07,
+      "loss": 0.4335,
+      "step": 524
+    },
+    {
+      "epoch": 0.23025053451016939,
+      "grad_norm": 1.337327241897583,
+      "learning_rate": 8.986683666317157e-07,
+      "loss": 0.4182,
+      "step": 525
+    },
+    {
+      "epoch": 0.23068910695685543,
+      "grad_norm": 1.3000355958938599,
+      "learning_rate": 8.982391854598137e-07,
+      "loss": 0.4844,
+      "step": 526
+    },
+    {
+      "epoch": 0.23112767940354148,
+      "grad_norm": 1.393452525138855,
+      "learning_rate": 8.97809200269159e-07,
+      "loss": 0.4648,
+      "step": 527
+    },
+    {
+      "epoch": 0.2315662518502275,
+      "grad_norm": 1.2905426025390625,
+      "learning_rate": 8.973784119278639e-07,
+      "loss": 0.4342,
+      "step": 528
+    },
+    {
+      "epoch": 0.23200482429691355,
+      "grad_norm": 1.1917750835418701,
+      "learning_rate": 8.969468213056613e-07,
+      "loss": 0.4271,
+      "step": 529
+    },
+    {
+      "epoch": 0.2324433967435996,
+      "grad_norm": 1.3469759225845337,
+      "learning_rate": 8.965144292739046e-07,
+      "loss": 0.4533,
+      "step": 530
+    },
+    {
+      "epoch": 0.23288196919028561,
+      "grad_norm": 1.2097269296646118,
+      "learning_rate": 8.960812367055646e-07,
+      "loss": 0.4292,
+      "step": 531
+    },
+    {
+      "epoch": 0.23332054163697166,
+      "grad_norm": 1.1486494541168213,
+      "learning_rate": 8.95647244475229e-07,
+      "loss": 0.4925,
+      "step": 532
+    },
+    {
+      "epoch": 0.23375911408365768,
+      "grad_norm": 1.2630372047424316,
+      "learning_rate": 8.952124534590993e-07,
+      "loss": 0.4106,
+      "step": 533
+    },
+    {
+      "epoch": 0.23419768653034373,
+      "grad_norm": 1.3153948783874512,
+      "learning_rate": 8.947768645349903e-07,
+      "loss": 0.4591,
+      "step": 534
+    },
+    {
+      "epoch": 0.23463625897702978,
+      "grad_norm": 1.280995488166809,
+      "learning_rate": 8.943404785823269e-07,
+      "loss": 0.4071,
+      "step": 535
+    },
+    {
+      "epoch": 0.2350748314237158,
+      "grad_norm": 1.2669576406478882,
+      "learning_rate": 8.939032964821442e-07,
+      "loss": 0.4185,
+      "step": 536
+    },
+    {
+      "epoch": 0.23551340387040184,
+      "grad_norm": 1.197329044342041,
+      "learning_rate": 8.93465319117084e-07,
+      "loss": 0.4476,
+      "step": 537
+    },
+    {
+      "epoch": 0.2359519763170879,
+      "grad_norm": 1.1159907579421997,
+      "learning_rate": 8.930265473713937e-07,
+      "loss": 0.429,
+      "step": 538
+    },
+    {
+      "epoch": 0.2363905487637739,
+      "grad_norm": 1.3145620822906494,
+      "learning_rate": 8.925869821309247e-07,
+      "loss": 0.4711,
+      "step": 539
+    },
+    {
+      "epoch": 0.23682912121045996,
+      "grad_norm": 1.306536078453064,
+      "learning_rate": 8.921466242831303e-07,
+      "loss": 0.4491,
+      "step": 540
+    },
+    {
+      "epoch": 0.23726769365714598,
+      "grad_norm": 1.2363736629486084,
+      "learning_rate": 8.917054747170642e-07,
+      "loss": 0.44,
+      "step": 541
+    },
+    {
+      "epoch": 0.23770626610383203,
+      "grad_norm": 1.340304970741272,
+      "learning_rate": 8.912635343233783e-07,
+      "loss": 0.4919,
+      "step": 542
+    },
+    {
+      "epoch": 0.23814483855051807,
+      "grad_norm": 1.2518609762191772,
+      "learning_rate": 8.908208039943213e-07,
+      "loss": 0.4362,
+      "step": 543
+    },
+    {
+      "epoch": 0.2385834109972041,
+      "grad_norm": 1.2911936044692993,
+      "learning_rate": 8.903772846237364e-07,
+      "loss": 0.4416,
+      "step": 544
+    },
+    {
+      "epoch": 0.23902198344389014,
+      "grad_norm": 1.2846801280975342,
+      "learning_rate": 8.899329771070602e-07,
+      "loss": 0.4885,
+      "step": 545
+    },
+    {
+      "epoch": 0.2394605558905762,
+      "grad_norm": 1.3153619766235352,
+      "learning_rate": 8.894878823413207e-07,
+      "loss": 0.4922,
+      "step": 546
+    },
+    {
+      "epoch": 0.2398991283372622,
+      "grad_norm": 1.4053038358688354,
+      "learning_rate": 8.890420012251346e-07,
+      "loss": 0.4912,
+      "step": 547
+    },
+    {
+      "epoch": 0.24033770078394825,
+      "grad_norm": 1.302511215209961,
+      "learning_rate": 8.885953346587065e-07,
+      "loss": 0.4019,
+      "step": 548
+    },
+    {
+      "epoch": 0.24077627323063427,
+      "grad_norm": 1.3971174955368042,
+      "learning_rate": 8.881478835438272e-07,
+      "loss": 0.4438,
+      "step": 549
+    },
+    {
+      "epoch": 0.24121484567732032,
+      "grad_norm": 1.2062628269195557,
+      "learning_rate": 8.876996487838711e-07,
+      "loss": 0.383,
+      "step": 550
+    },
+    {
+      "epoch": 0.24165341812400637,
+      "grad_norm": 1.3044462203979492,
+      "learning_rate": 8.872506312837944e-07,
+      "loss": 0.451,
+      "step": 551
+    },
+    {
+      "epoch": 0.2420919905706924,
+      "grad_norm": 1.3488870859146118,
+      "learning_rate": 8.868008319501341e-07,
+      "loss": 0.4249,
+      "step": 552
+    },
+    {
+      "epoch": 0.24253056301737844,
+      "grad_norm": 1.1751151084899902,
+      "learning_rate": 8.863502516910057e-07,
+      "loss": 0.4581,
+      "step": 553
+    },
+    {
+      "epoch": 0.24296913546406446,
+      "grad_norm": 1.3605632781982422,
+      "learning_rate": 8.858988914161009e-07,
+      "loss": 0.4447,
+      "step": 554
+    },
+    {
+      "epoch": 0.2434077079107505,
+      "grad_norm": 1.2312195301055908,
+      "learning_rate": 8.854467520366864e-07,
+      "loss": 0.4164,
+      "step": 555
+    },
+    {
+      "epoch": 0.24384628035743655,
+      "grad_norm": 1.3179501295089722,
+      "learning_rate": 8.849938344656021e-07,
+      "loss": 0.4674,
+      "step": 556
+    },
+    {
+      "epoch": 0.24428485280412257,
+      "grad_norm": 1.2114101648330688,
+      "learning_rate": 8.845401396172588e-07,
+      "loss": 0.4529,
+      "step": 557
+    },
+    {
+      "epoch": 0.24472342525080862,
+      "grad_norm": 1.238842487335205,
+      "learning_rate": 8.840856684076365e-07,
+      "loss": 0.4685,
+      "step": 558
+    },
+    {
+      "epoch": 0.24516199769749467,
+      "grad_norm": 1.1789189577102661,
+      "learning_rate": 8.836304217542828e-07,
+      "loss": 0.4282,
+      "step": 559
+    },
+    {
+      "epoch": 0.24560057014418069,
+      "grad_norm": 1.304688811302185,
+      "learning_rate": 8.831744005763107e-07,
+      "loss": 0.4424,
+      "step": 560
+    },
+    {
+      "epoch": 0.24603914259086673,
+      "grad_norm": 1.3729431629180908,
+      "learning_rate": 8.827176057943969e-07,
+      "loss": 0.4683,
+      "step": 561
+    },
+    {
+      "epoch": 0.24647771503755275,
+      "grad_norm": 1.3452732563018799,
+      "learning_rate": 8.822600383307802e-07,
+      "loss": 0.4498,
+      "step": 562
+    },
+    {
+      "epoch": 0.2469162874842388,
+      "grad_norm": 1.27022385597229,
+      "learning_rate": 8.818016991092594e-07,
+      "loss": 0.4617,
+      "step": 563
+    },
+    {
+      "epoch": 0.24735485993092485,
+      "grad_norm": 1.2468392848968506,
+      "learning_rate": 8.813425890551909e-07,
+      "loss": 0.4356,
+      "step": 564
+    },
+    {
+      "epoch": 0.24779343237761087,
+      "grad_norm": 1.2570581436157227,
+      "learning_rate": 8.808827090954881e-07,
+      "loss": 0.3973,
+      "step": 565
+    },
+    {
+      "epoch": 0.24823200482429691,
+      "grad_norm": 1.2952882051467896,
+      "learning_rate": 8.804220601586183e-07,
+      "loss": 0.4624,
+      "step": 566
+    },
+    {
+      "epoch": 0.24867057727098296,
+      "grad_norm": 1.2446449995040894,
+      "learning_rate": 8.799606431746013e-07,
+      "loss": 0.3957,
+      "step": 567
+    },
+    {
+      "epoch": 0.24910914971766898,
+      "grad_norm": 1.4092421531677246,
+      "learning_rate": 8.794984590750079e-07,
+      "loss": 0.4968,
+      "step": 568
+    },
+    {
+      "epoch": 0.24954772216435503,
+      "grad_norm": 1.3203057050704956,
+      "learning_rate": 8.790355087929573e-07,
+      "loss": 0.4874,
+      "step": 569
+    },
+    {
+      "epoch": 0.24998629461104105,
+      "grad_norm": 1.468639850616455,
+      "learning_rate": 8.785717932631155e-07,
+      "loss": 0.4666,
+      "step": 570
+    },
+    {
+      "epoch": 0.2504248670577271,
+      "grad_norm": 1.3520934581756592,
+      "learning_rate": 8.781073134216943e-07,
+      "loss": 0.4469,
+      "step": 571
+    },
+    {
+      "epoch": 0.25086343950441314,
+      "grad_norm": 1.266510248184204,
+      "learning_rate": 8.776420702064473e-07,
+      "loss": 0.3974,
+      "step": 572
+    },
+    {
+      "epoch": 0.2513020119510992,
+      "grad_norm": 1.2830793857574463,
+      "learning_rate": 8.771760645566705e-07,
+      "loss": 0.4821,
+      "step": 573
+    },
+    {
+      "epoch": 0.2517405843977852,
+      "grad_norm": 1.3247817754745483,
+      "learning_rate": 8.767092974131984e-07,
+      "loss": 0.4426,
+      "step": 574
+    },
+    {
+      "epoch": 0.25217915684447123,
+      "grad_norm": 1.340187907218933,
+      "learning_rate": 8.762417697184032e-07,
+      "loss": 0.4512,
+      "step": 575
+    },
+    {
+      "epoch": 0.2526177292911573,
+      "grad_norm": 1.2662535905838013,
+      "learning_rate": 8.757734824161929e-07,
+      "loss": 0.4661,
+      "step": 576
+    },
+    {
+      "epoch": 0.2530563017378433,
+      "grad_norm": 1.2929002046585083,
+      "learning_rate": 8.753044364520083e-07,
+      "loss": 0.4576,
+      "step": 577
+    },
+    {
+      "epoch": 0.2534948741845294,
+      "grad_norm": 1.3643664121627808,
+      "learning_rate": 8.748346327728228e-07,
+      "loss": 0.4642,
+      "step": 578
+    },
+    {
+      "epoch": 0.2539334466312154,
+      "grad_norm": 1.4776318073272705,
+      "learning_rate": 8.74364072327139e-07,
+      "loss": 0.4554,
+      "step": 579
+    },
+    {
+      "epoch": 0.2543720190779014,
+      "grad_norm": 1.1963412761688232,
+      "learning_rate": 8.738927560649876e-07,
+      "loss": 0.4026,
+      "step": 580
+    },
+    {
+      "epoch": 0.25481059152458746,
+      "grad_norm": 1.3636819124221802,
+      "learning_rate": 8.734206849379253e-07,
+      "loss": 0.4415,
+      "step": 581
+    },
+    {
+      "epoch": 0.2552491639712735,
+      "grad_norm": 1.3011325597763062,
+      "learning_rate": 8.729478598990323e-07,
+      "loss": 0.4289,
+      "step": 582
+    },
+    {
+      "epoch": 0.25568773641795955,
+      "grad_norm": 1.2742562294006348,
+      "learning_rate": 8.724742819029116e-07,
+      "loss": 0.4317,
+      "step": 583
+    },
+    {
+      "epoch": 0.2561263088646456,
+      "grad_norm": 1.4198004007339478,
+      "learning_rate": 8.719999519056859e-07,
+      "loss": 0.4656,
+      "step": 584
+    },
+    {
+      "epoch": 0.2565648813113316,
+      "grad_norm": 1.376230001449585,
+      "learning_rate": 8.715248708649963e-07,
+      "loss": 0.459,
+      "step": 585
+    },
+    {
+      "epoch": 0.25700345375801764,
+      "grad_norm": 1.1574681997299194,
+      "learning_rate": 8.710490397400005e-07,
+      "loss": 0.4235,
+      "step": 586
+    },
+    {
+      "epoch": 0.2574420262047037,
+      "grad_norm": 1.2351993322372437,
+      "learning_rate": 8.7057245949137e-07,
+      "loss": 0.4214,
+      "step": 587
+    },
+    {
+      "epoch": 0.25788059865138974,
+      "grad_norm": 1.2393819093704224,
+      "learning_rate": 8.70095131081289e-07,
+      "loss": 0.415,
+      "step": 588
+    },
+    {
+      "epoch": 0.2583191710980758,
+      "grad_norm": 1.267637014389038,
+      "learning_rate": 8.696170554734523e-07,
+      "loss": 0.4274,
+      "step": 589
+    },
+    {
+      "epoch": 0.2587577435447618,
+      "grad_norm": 1.3031575679779053,
+      "learning_rate": 8.691382336330631e-07,
+      "loss": 0.4582,
+      "step": 590
+    },
+    {
+      "epoch": 0.2591963159914478,
+      "grad_norm": 1.3642998933792114,
+      "learning_rate": 8.686586665268313e-07,
+      "loss": 0.465,
+      "step": 591
+    },
+    {
+      "epoch": 0.25963488843813387,
+      "grad_norm": 1.3312978744506836,
+      "learning_rate": 8.681783551229713e-07,
+      "loss": 0.4679,
+      "step": 592
+    },
+    {
+      "epoch": 0.2600734608848199,
+      "grad_norm": 1.3136823177337646,
+      "learning_rate": 8.676973003912004e-07,
+      "loss": 0.485,
+      "step": 593
+    },
+    {
+      "epoch": 0.26051203333150597,
+      "grad_norm": 1.2545801401138306,
+      "learning_rate": 8.672155033027364e-07,
+      "loss": 0.4402,
+      "step": 594
+    },
+    {
+      "epoch": 0.26095060577819196,
+      "grad_norm": 1.227077603340149,
+      "learning_rate": 8.667329648302959e-07,
+      "loss": 0.4597,
+      "step": 595
+    },
+    {
+      "epoch": 0.261389178224878,
+      "grad_norm": 1.3195209503173828,
+      "learning_rate": 8.662496859480925e-07,
+      "loss": 0.4567,
+      "step": 596
+    },
+    {
+      "epoch": 0.26182775067156405,
+      "grad_norm": 1.235021710395813,
+      "learning_rate": 8.657656676318345e-07,
+      "loss": 0.4215,
+      "step": 597
+    },
+    {
+      "epoch": 0.2622663231182501,
+      "grad_norm": 1.2290730476379395,
+      "learning_rate": 8.652809108587231e-07,
+      "loss": 0.4139,
+      "step": 598
+    },
+    {
+      "epoch": 0.26270489556493615,
+      "grad_norm": 1.2141824960708618,
+      "learning_rate": 8.647954166074503e-07,
+      "loss": 0.4234,
+      "step": 599
+    },
+    {
+      "epoch": 0.2631434680116222,
+      "grad_norm": 1.3154791593551636,
+      "learning_rate": 8.64309185858197e-07,
+      "loss": 0.4791,
+      "step": 600
+    },
+    {
+      "epoch": 0.2635820404583082,
+      "grad_norm": 1.2648977041244507,
+      "learning_rate": 8.638222195926313e-07,
+      "loss": 0.4227,
+      "step": 601
+    },
+    {
+      "epoch": 0.26402061290499423,
+      "grad_norm": 1.231683373451233,
+      "learning_rate": 8.633345187939061e-07,
+      "loss": 0.4806,
+      "step": 602
+    },
+    {
+      "epoch": 0.2644591853516803,
+      "grad_norm": 1.3356053829193115,
+      "learning_rate": 8.628460844466572e-07,
+      "loss": 0.4733,
+      "step": 603
+    },
+    {
+      "epoch": 0.26489775779836633,
+      "grad_norm": 1.3214858770370483,
+      "learning_rate": 8.623569175370016e-07,
+      "loss": 0.4439,
+      "step": 604
+    },
+    {
+      "epoch": 0.2653363302450524,
+      "grad_norm": 1.3323317766189575,
+      "learning_rate": 8.61867019052535e-07,
+      "loss": 0.4439,
+      "step": 605
+    },
+    {
+      "epoch": 0.26577490269173837,
+      "grad_norm": 1.388413906097412,
+      "learning_rate": 8.613763899823303e-07,
+      "loss": 0.46,
+      "step": 606
+    },
+    {
+      "epoch": 0.2662134751384244,
+      "grad_norm": 1.3175569772720337,
+      "learning_rate": 8.608850313169355e-07,
+      "loss": 0.3937,
+      "step": 607
+    },
+    {
+      "epoch": 0.26665204758511046,
+      "grad_norm": 1.3802878856658936,
+      "learning_rate": 8.603929440483713e-07,
+      "loss": 0.4245,
+      "step": 608
+    },
+    {
+      "epoch": 0.2670906200317965,
+      "grad_norm": 1.294541835784912,
+      "learning_rate": 8.599001291701294e-07,
+      "loss": 0.4257,
+      "step": 609
+    },
+    {
+      "epoch": 0.26752919247848256,
+      "grad_norm": 1.3904340267181396,
+      "learning_rate": 8.59406587677171e-07,
+      "loss": 0.47,
+      "step": 610
+    },
+    {
+      "epoch": 0.26796776492516855,
+      "grad_norm": 1.310609221458435,
+      "learning_rate": 8.589123205659237e-07,
+      "loss": 0.4693,
+      "step": 611
+    },
+    {
+      "epoch": 0.2684063373718546,
+      "grad_norm": 1.508705496788025,
+      "learning_rate": 8.584173288342806e-07,
+      "loss": 0.4818,
+      "step": 612
+    },
+    {
+      "epoch": 0.26884490981854064,
+      "grad_norm": 1.258613109588623,
+      "learning_rate": 8.579216134815972e-07,
+      "loss": 0.4391,
+      "step": 613
+    },
+    {
+      "epoch": 0.2692834822652267,
+      "grad_norm": 1.309943437576294,
+      "learning_rate": 8.574251755086905e-07,
+      "loss": 0.4345,
+      "step": 614
+    },
+    {
+      "epoch": 0.26972205471191274,
+      "grad_norm": 1.2155766487121582,
+      "learning_rate": 8.569280159178358e-07,
+      "loss": 0.4451,
+      "step": 615
+    },
+    {
+      "epoch": 0.2701606271585988,
+      "grad_norm": 1.3451566696166992,
+      "learning_rate": 8.564301357127662e-07,
+      "loss": 0.446,
+      "step": 616
+    },
+    {
+      "epoch": 0.2705991996052848,
+      "grad_norm": 1.581193447113037,
+      "learning_rate": 8.559315358986684e-07,
+      "loss": 0.455,
+      "step": 617
+    },
+    {
+      "epoch": 0.2710377720519708,
+      "grad_norm": 1.3008910417556763,
+      "learning_rate": 8.554322174821833e-07,
+      "loss": 0.4209,
+      "step": 618
+    },
+    {
+      "epoch": 0.2714763444986569,
+      "grad_norm": 1.2218880653381348,
+      "learning_rate": 8.549321814714017e-07,
+      "loss": 0.4627,
+      "step": 619
+    },
+    {
+      "epoch": 0.2719149169453429,
+      "grad_norm": 1.2023916244506836,
+      "learning_rate": 8.544314288758634e-07,
+      "loss": 0.4502,
+      "step": 620
+    },
+    {
+      "epoch": 0.27235348939202897,
+      "grad_norm": 1.3902664184570312,
+      "learning_rate": 8.539299607065551e-07,
+      "loss": 0.4435,
+      "step": 621
+    },
+    {
+      "epoch": 0.27279206183871496,
+      "grad_norm": 1.35061776638031,
+      "learning_rate": 8.534277779759081e-07,
+      "loss": 0.4476,
+      "step": 622
+    },
+    {
+      "epoch": 0.273230634285401,
+      "grad_norm": 1.2908838987350464,
+      "learning_rate": 8.529248816977963e-07,
+      "loss": 0.4829,
+      "step": 623
+    },
+    {
+      "epoch": 0.27366920673208706,
+      "grad_norm": 1.2177608013153076,
+      "learning_rate": 8.524212728875342e-07,
+      "loss": 0.4346,
+      "step": 624
+    },
+    {
+      "epoch": 0.2741077791787731,
+      "grad_norm": 1.3742519617080688,
+      "learning_rate": 8.51916952561875e-07,
+      "loss": 0.463,
+      "step": 625
+    },
+    {
+      "epoch": 0.27454635162545915,
+      "grad_norm": 1.1771491765975952,
+      "learning_rate": 8.514119217390084e-07,
+      "loss": 0.4665,
+      "step": 626
+    },
+    {
+      "epoch": 0.27498492407214514,
+      "grad_norm": 1.4156662225723267,
+      "learning_rate": 8.509061814385581e-07,
+      "loss": 0.5086,
+      "step": 627
+    },
+    {
+      "epoch": 0.2754234965188312,
+      "grad_norm": 1.4472264051437378,
+      "learning_rate": 8.503997326815811e-07,
+      "loss": 0.4758,
+      "step": 628
+    },
+    {
+      "epoch": 0.27586206896551724,
+      "grad_norm": 1.3086676597595215,
+      "learning_rate": 8.498925764905635e-07,
+      "loss": 0.4232,
+      "step": 629
+    },
+    {
+      "epoch": 0.2763006414122033,
+      "grad_norm": 1.2598360776901245,
+      "learning_rate": 8.493847138894208e-07,
+      "loss": 0.4545,
+      "step": 630
+    },
+    {
+      "epoch": 0.27673921385888933,
+      "grad_norm": 1.3427696228027344,
+      "learning_rate": 8.488761459034941e-07,
+      "loss": 0.4266,
+      "step": 631
+    },
+    {
+      "epoch": 0.2771777863055754,
+      "grad_norm": 1.257368803024292,
+      "learning_rate": 8.483668735595486e-07,
+      "loss": 0.4527,
+      "step": 632
+    },
+    {
+      "epoch": 0.27761635875226137,
+      "grad_norm": 1.3200676441192627,
+      "learning_rate": 8.478568978857721e-07,
+      "loss": 0.5008,
+      "step": 633
+    },
+    {
+      "epoch": 0.2780549311989474,
+      "grad_norm": 1.3147386312484741,
+      "learning_rate": 8.473462199117715e-07,
+      "loss": 0.4622,
+      "step": 634
+    },
+    {
+      "epoch": 0.27849350364563347,
+      "grad_norm": 1.241818904876709,
+      "learning_rate": 8.468348406685724e-07,
+      "loss": 0.4707,
+      "step": 635
+    },
+    {
+      "epoch": 0.2789320760923195,
+      "grad_norm": 1.1343870162963867,
+      "learning_rate": 8.463227611886157e-07,
+      "loss": 0.4625,
+      "step": 636
+    },
+    {
+      "epoch": 0.27937064853900556,
+      "grad_norm": 1.3791996240615845,
+      "learning_rate": 8.458099825057565e-07,
+      "loss": 0.4212,
+      "step": 637
+    },
+    {
+      "epoch": 0.27980922098569155,
+      "grad_norm": 1.2124180793762207,
+      "learning_rate": 8.45296505655261e-07,
+      "loss": 0.4738,
+      "step": 638
+    },
+    {
+      "epoch": 0.2802477934323776,
+      "grad_norm": 1.332551121711731,
+      "learning_rate": 8.447823316738054e-07,
+      "loss": 0.445,
+      "step": 639
+    },
+    {
+      "epoch": 0.28068636587906365,
+      "grad_norm": 1.3134421110153198,
+      "learning_rate": 8.442674615994731e-07,
+      "loss": 0.4764,
+      "step": 640
+    },
+    {
+      "epoch": 0.2811249383257497,
+      "grad_norm": 1.2080031633377075,
+      "learning_rate": 8.43751896471753e-07,
+      "loss": 0.4158,
+      "step": 641
+    },
+    {
+      "epoch": 0.28156351077243574,
+      "grad_norm": 1.3157594203948975,
+      "learning_rate": 8.432356373315371e-07,
+      "loss": 0.4535,
+      "step": 642
+    },
+    {
+      "epoch": 0.28200208321912174,
+      "grad_norm": 1.3137787580490112,
+      "learning_rate": 8.427186852211188e-07,
+      "loss": 0.4377,
+      "step": 643
+    },
+    {
+      "epoch": 0.2824406556658078,
+      "grad_norm": 1.255462408065796,
+      "learning_rate": 8.422010411841905e-07,
+      "loss": 0.3872,
+      "step": 644
+    },
+    {
+      "epoch": 0.28287922811249383,
+      "grad_norm": 1.292777180671692,
+      "learning_rate": 8.416827062658415e-07,
+      "loss": 0.4944,
+      "step": 645
+    },
+    {
+      "epoch": 0.2833178005591799,
+      "grad_norm": 1.4436018466949463,
+      "learning_rate": 8.411636815125558e-07,
+      "loss": 0.4772,
+      "step": 646
+    },
+    {
+      "epoch": 0.2837563730058659,
+      "grad_norm": 1.2209681272506714,
+      "learning_rate": 8.406439679722104e-07,
+      "loss": 0.4437,
+      "step": 647
+    },
+    {
+      "epoch": 0.2841949454525519,
+      "grad_norm": 1.3344581127166748,
+      "learning_rate": 8.401235666940727e-07,
+      "loss": 0.4715,
+      "step": 648
+    },
+    {
+      "epoch": 0.28463351789923796,
+      "grad_norm": 1.1497730016708374,
+      "learning_rate": 8.396024787287988e-07,
+      "loss": 0.4298,
+      "step": 649
+    },
+    {
+      "epoch": 0.285072090345924,
+      "grad_norm": 1.341781735420227,
+      "learning_rate": 8.390807051284309e-07,
+      "loss": 0.4257,
+      "step": 650
+    },
+    {
+      "epoch": 0.28551066279261006,
+      "grad_norm": 1.280103087425232,
+      "learning_rate": 8.385582469463959e-07,
+      "loss": 0.4368,
+      "step": 651
+    },
+    {
+      "epoch": 0.2859492352392961,
+      "grad_norm": 1.2544479370117188,
+      "learning_rate": 8.380351052375023e-07,
+      "loss": 0.4723,
+      "step": 652
+    },
+    {
+      "epoch": 0.28638780768598215,
+      "grad_norm": 1.1628601551055908,
+      "learning_rate": 8.375112810579389e-07,
+      "loss": 0.4515,
+      "step": 653
+    },
+    {
+      "epoch": 0.28682638013266815,
+      "grad_norm": 1.3643077611923218,
+      "learning_rate": 8.369867754652724e-07,
+      "loss": 0.4235,
+      "step": 654
+    },
+    {
+      "epoch": 0.2872649525793542,
+      "grad_norm": 1.205430030822754,
+      "learning_rate": 8.36461589518445e-07,
+      "loss": 0.4375,
+      "step": 655
+    },
+    {
+      "epoch": 0.28770352502604024,
+      "grad_norm": 1.2593382596969604,
+      "learning_rate": 8.359357242777728e-07,
+      "loss": 0.4494,
+      "step": 656
+    },
+    {
+      "epoch": 0.2881420974727263,
+      "grad_norm": 1.226905107498169,
+      "learning_rate": 8.354091808049431e-07,
+      "loss": 0.4476,
+      "step": 657
+    },
+    {
+      "epoch": 0.28858066991941234,
+      "grad_norm": 1.2723194360733032,
+      "learning_rate": 8.348819601630124e-07,
+      "loss": 0.4706,
+      "step": 658
+    },
+    {
+      "epoch": 0.28901924236609833,
+      "grad_norm": 1.2584813833236694,
+      "learning_rate": 8.343540634164047e-07,
+      "loss": 0.4518,
+      "step": 659
+    },
+    {
+      "epoch": 0.2894578148127844,
+      "grad_norm": 1.19351327419281,
+      "learning_rate": 8.338254916309089e-07,
+      "loss": 0.3974,
+      "step": 660
+    },
+    {
+      "epoch": 0.2898963872594704,
+      "grad_norm": 1.3015458583831787,
+      "learning_rate": 8.332962458736765e-07,
+      "loss": 0.451,
+      "step": 661
+    },
+    {
+      "epoch": 0.29033495970615647,
+      "grad_norm": 1.206496000289917,
+      "learning_rate": 8.327663272132202e-07,
+      "loss": 0.4202,
+      "step": 662
+    },
+    {
+      "epoch": 0.2907735321528425,
+      "grad_norm": 1.1564853191375732,
+      "learning_rate": 8.322357367194108e-07,
+      "loss": 0.4161,
+      "step": 663
+    },
+    {
+      "epoch": 0.2912121045995285,
+      "grad_norm": 1.2696150541305542,
+      "learning_rate": 8.317044754634756e-07,
+      "loss": 0.4461,
+      "step": 664
+    },
+    {
+      "epoch": 0.29165067704621456,
+      "grad_norm": 1.227945327758789,
+      "learning_rate": 8.311725445179964e-07,
+      "loss": 0.4303,
+      "step": 665
+    },
+    {
+      "epoch": 0.2920892494929006,
+      "grad_norm": 1.2026177644729614,
+      "learning_rate": 8.306399449569066e-07,
+      "loss": 0.4329,
+      "step": 666
+    },
+    {
+      "epoch": 0.29252782193958665,
+      "grad_norm": 1.330443263053894,
+      "learning_rate": 8.301066778554897e-07,
+      "loss": 0.4263,
+      "step": 667
+    },
+    {
+      "epoch": 0.2929663943862727,
+      "grad_norm": 1.2969058752059937,
+      "learning_rate": 8.295727442903772e-07,
+      "loss": 0.4558,
+      "step": 668
+    },
+    {
+      "epoch": 0.29340496683295875,
+      "grad_norm": 1.3471171855926514,
+      "learning_rate": 8.290381453395457e-07,
+      "loss": 0.4873,
+      "step": 669
+    },
+    {
+      "epoch": 0.29384353927964474,
+      "grad_norm": 1.3648743629455566,
+      "learning_rate": 8.285028820823153e-07,
+      "loss": 0.4626,
+      "step": 670
+    },
+    {
+      "epoch": 0.2942821117263308,
+      "grad_norm": 1.481045126914978,
+      "learning_rate": 8.279669555993475e-07,
+      "loss": 0.4924,
+      "step": 671
+    },
+    {
+      "epoch": 0.29472068417301683,
+      "grad_norm": 1.1982755661010742,
+      "learning_rate": 8.274303669726426e-07,
+      "loss": 0.4365,
+      "step": 672
+    },
+    {
+      "epoch": 0.2951592566197029,
+      "grad_norm": 1.2949172258377075,
+      "learning_rate": 8.268931172855378e-07,
+      "loss": 0.4929,
+      "step": 673
+    },
+    {
+      "epoch": 0.29559782906638893,
+      "grad_norm": 1.2132493257522583,
+      "learning_rate": 8.263552076227047e-07,
+      "loss": 0.413,
+      "step": 674
+    },
+    {
+      "epoch": 0.2960364015130749,
+      "grad_norm": 1.3627359867095947,
+      "learning_rate": 8.258166390701481e-07,
+      "loss": 0.486,
+      "step": 675
+    },
+    {
+      "epoch": 0.29647497395976097,
+      "grad_norm": 1.3661818504333496,
+      "learning_rate": 8.25277412715202e-07,
+      "loss": 0.4704,
+      "step": 676
+    },
+    {
+      "epoch": 0.296913546406447,
+      "grad_norm": 1.2108001708984375,
+      "learning_rate": 8.247375296465293e-07,
+      "loss": 0.402,
+      "step": 677
+    },
+    {
+      "epoch": 0.29735211885313306,
+      "grad_norm": 1.4531514644622803,
+      "learning_rate": 8.241969909541183e-07,
+      "loss": 0.4896,
+      "step": 678
+    },
+    {
+      "epoch": 0.2977906912998191,
+      "grad_norm": 1.3670005798339844,
+      "learning_rate": 8.236557977292813e-07,
+      "loss": 0.4511,
+      "step": 679
+    },
+    {
+      "epoch": 0.2982292637465051,
+      "grad_norm": 1.4220718145370483,
+      "learning_rate": 8.231139510646515e-07,
+      "loss": 0.4765,
+      "step": 680
+    },
+    {
+      "epoch": 0.29866783619319115,
+      "grad_norm": 1.2339059114456177,
+      "learning_rate": 8.22571452054182e-07,
+      "loss": 0.4748,
+      "step": 681
+    },
+    {
+      "epoch": 0.2991064086398772,
+      "grad_norm": 1.2986462116241455,
+      "learning_rate": 8.220283017931427e-07,
+      "loss": 0.4738,
+      "step": 682
+    },
+    {
+      "epoch": 0.29954498108656324,
+      "grad_norm": 1.2862370014190674,
+      "learning_rate": 8.214845013781183e-07,
+      "loss": 0.4658,
+      "step": 683
+    },
+    {
+      "epoch": 0.2999835535332493,
+      "grad_norm": 1.3527040481567383,
+      "learning_rate": 8.209400519070057e-07,
+      "loss": 0.4905,
+      "step": 684
+    },
+    {
+      "epoch": 0.3004221259799353,
+      "grad_norm": 1.2541968822479248,
+      "learning_rate": 8.203949544790129e-07,
+      "loss": 0.4487,
+      "step": 685
+    },
+    {
+      "epoch": 0.30086069842662133,
+      "grad_norm": 1.3078125715255737,
+      "learning_rate": 8.198492101946562e-07,
+      "loss": 0.4253,
+      "step": 686
+    },
+    {
+      "epoch": 0.3012992708733074,
+      "grad_norm": 1.210807204246521,
+      "learning_rate": 8.193028201557567e-07,
+      "loss": 0.414,
+      "step": 687
+    },
+    {
+      "epoch": 0.3017378433199934,
+      "grad_norm": 1.3149670362472534,
+      "learning_rate": 8.187557854654406e-07,
+      "loss": 0.4341,
+      "step": 688
+    },
+    {
+      "epoch": 0.3021764157666795,
+      "grad_norm": 1.2983028888702393,
+      "learning_rate": 8.182081072281346e-07,
+      "loss": 0.471,
+      "step": 689
+    },
+    {
+      "epoch": 0.3026149882133655,
+      "grad_norm": 1.1206141710281372,
+      "learning_rate": 8.176597865495653e-07,
+      "loss": 0.4074,
+      "step": 690
+    },
+    {
+      "epoch": 0.3030535606600515,
+      "grad_norm": 1.276363492012024,
+      "learning_rate": 8.171108245367561e-07,
+      "loss": 0.4624,
+      "step": 691
+    },
+    {
+      "epoch": 0.30349213310673756,
+      "grad_norm": 1.3328988552093506,
+      "learning_rate": 8.165612222980251e-07,
+      "loss": 0.4394,
+      "step": 692
+    },
+    {
+      "epoch": 0.3039307055534236,
+      "grad_norm": 1.4199296236038208,
+      "learning_rate": 8.160109809429834e-07,
+      "loss": 0.4394,
+      "step": 693
+    },
+    {
+      "epoch": 0.30436927800010966,
+      "grad_norm": 1.363855004310608,
+      "learning_rate": 8.154601015825318e-07,
+      "loss": 0.4267,
+      "step": 694
+    },
+    {
+      "epoch": 0.3048078504467957,
+      "grad_norm": 1.2294096946716309,
+      "learning_rate": 8.149085853288597e-07,
+      "loss": 0.4361,
+      "step": 695
+    },
+    {
+      "epoch": 0.3052464228934817,
+      "grad_norm": 1.3705590963363647,
+      "learning_rate": 8.143564332954425e-07,
+      "loss": 0.465,
+      "step": 696
+    },
+    {
+      "epoch": 0.30568499534016774,
+      "grad_norm": 1.2903246879577637,
+      "learning_rate": 8.138036465970384e-07,
+      "loss": 0.4469,
+      "step": 697
+    },
+    {
+      "epoch": 0.3061235677868538,
+      "grad_norm": 1.2863341569900513,
+      "learning_rate": 8.132502263496875e-07,
+      "loss": 0.4226,
+      "step": 698
+    },
+    {
+      "epoch": 0.30656214023353984,
+      "grad_norm": 1.333102822303772,
+      "learning_rate": 8.126961736707091e-07,
+      "loss": 0.4326,
+      "step": 699
+    },
+    {
+      "epoch": 0.3070007126802259,
+      "grad_norm": 1.2394230365753174,
+      "learning_rate": 8.121414896786991e-07,
+      "loss": 0.4383,
+      "step": 700
+    },
+    {
+      "epoch": 0.3074392851269119,
+      "grad_norm": 1.4252898693084717,
+      "learning_rate": 8.115861754935279e-07,
+      "loss": 0.4935,
+      "step": 701
+    },
+    {
+      "epoch": 0.3078778575735979,
+      "grad_norm": 1.2032008171081543,
+      "learning_rate": 8.110302322363387e-07,
+      "loss": 0.4592,
+      "step": 702
+    },
+    {
+      "epoch": 0.30831643002028397,
+      "grad_norm": 1.212705135345459,
+      "learning_rate": 8.10473661029544e-07,
+      "loss": 0.4354,
+      "step": 703
+    },
+    {
+      "epoch": 0.30875500246697,
+      "grad_norm": 1.3495712280273438,
+      "learning_rate": 8.099164629968247e-07,
+      "loss": 0.4419,
+      "step": 704
+    },
+    {
+      "epoch": 0.30919357491365607,
+      "grad_norm": 1.325252652168274,
+      "learning_rate": 8.093586392631271e-07,
+      "loss": 0.4354,
+      "step": 705
+    },
+    {
+      "epoch": 0.3096321473603421,
+      "grad_norm": 1.2952181100845337,
+      "learning_rate": 8.088001909546606e-07,
+      "loss": 0.4495,
+      "step": 706
+    },
+    {
+      "epoch": 0.3100707198070281,
+      "grad_norm": 1.4086304903030396,
+      "learning_rate": 8.082411191988956e-07,
+      "loss": 0.4763,
+      "step": 707
+    },
+    {
+      "epoch": 0.31050929225371415,
+      "grad_norm": 1.5200958251953125,
+      "learning_rate": 8.076814251245612e-07,
+      "loss": 0.4554,
+      "step": 708
+    },
+    {
+      "epoch": 0.3109478647004002,
+      "grad_norm": 1.2769299745559692,
+      "learning_rate": 8.071211098616433e-07,
+      "loss": 0.4089,
+      "step": 709
+    },
+    {
+      "epoch": 0.31138643714708625,
+      "grad_norm": 1.294661283493042,
+      "learning_rate": 8.06560174541381e-07,
+      "loss": 0.5231,
+      "step": 710
+    },
+    {
+      "epoch": 0.3118250095937723,
+      "grad_norm": 1.3333851099014282,
+      "learning_rate": 8.059986202962666e-07,
+      "loss": 0.4307,
+      "step": 711
+    },
+    {
+      "epoch": 0.3122635820404583,
+      "grad_norm": 1.2841682434082031,
+      "learning_rate": 8.054364482600405e-07,
+      "loss": 0.4259,
+      "step": 712
+    },
+    {
+      "epoch": 0.31270215448714433,
+      "grad_norm": 1.257988452911377,
+      "learning_rate": 8.048736595676916e-07,
+      "loss": 0.4293,
+      "step": 713
+    },
+    {
+      "epoch": 0.3131407269338304,
+      "grad_norm": 1.2623865604400635,
+      "learning_rate": 8.043102553554531e-07,
+      "loss": 0.4311,
+      "step": 714
+    },
+    {
+      "epoch": 0.31357929938051643,
+      "grad_norm": 1.398133635520935,
+      "learning_rate": 8.037462367608012e-07,
+      "loss": 0.4415,
+      "step": 715
+    },
+    {
+      "epoch": 0.3140178718272025,
+      "grad_norm": 1.3084006309509277,
+      "learning_rate": 8.031816049224523e-07,
+      "loss": 0.4699,
+      "step": 716
+    },
+    {
+      "epoch": 0.31445644427388847,
+      "grad_norm": 1.1205732822418213,
+      "learning_rate": 8.026163609803611e-07,
+      "loss": 0.4297,
+      "step": 717
+    },
+    {
+      "epoch": 0.3148950167205745,
+      "grad_norm": 1.3080092668533325,
+      "learning_rate": 8.020505060757178e-07,
+      "loss": 0.3942,
+      "step": 718
+    },
+    {
+      "epoch": 0.31533358916726056,
+      "grad_norm": 1.2918123006820679,
+      "learning_rate": 8.014840413509464e-07,
+      "loss": 0.4946,
+      "step": 719
+    },
+    {
+      "epoch": 0.3157721616139466,
+      "grad_norm": 1.257453441619873,
+      "learning_rate": 8.009169679497019e-07,
+      "loss": 0.4411,
+      "step": 720
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 2280,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 40,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1.522795283589628e+18,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}