diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,5011 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 3.0,
+  "eval_steps": 500,
+  "global_step": 711,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.004232804232804233,
+      "grad_norm": 54.30406078717587,
+      "learning_rate": 0.0,
+      "loss": 1.2873,
+      "step": 1
+    },
+    {
+      "epoch": 0.008465608465608466,
+      "grad_norm": 56.84874016239752,
+      "learning_rate": 2.0833333333333333e-07,
+      "loss": 1.278,
+      "step": 2
+    },
+    {
+      "epoch": 0.012698412698412698,
+      "grad_norm": 54.878082362253245,
+      "learning_rate": 4.1666666666666667e-07,
+      "loss": 1.2261,
+      "step": 3
+    },
+    {
+      "epoch": 0.016931216931216932,
+      "grad_norm": 54.66242016872322,
+      "learning_rate": 6.25e-07,
+      "loss": 1.1957,
+      "step": 4
+    },
+    {
+      "epoch": 0.021164021164021163,
+      "grad_norm": 54.63578365153812,
+      "learning_rate": 8.333333333333333e-07,
+      "loss": 1.3671,
+      "step": 5
+    },
+    {
+      "epoch": 0.025396825396825397,
+      "grad_norm": 52.69479916022209,
+      "learning_rate": 1.0416666666666667e-06,
+      "loss": 1.0852,
+      "step": 6
+    },
+    {
+      "epoch": 0.02962962962962963,
+      "grad_norm": 51.80015106662938,
+      "learning_rate": 1.25e-06,
+      "loss": 1.1174,
+      "step": 7
+    },
+    {
+      "epoch": 0.033862433862433865,
+      "grad_norm": 46.94049102872614,
+      "learning_rate": 1.4583333333333335e-06,
+      "loss": 1.0722,
+      "step": 8
+    },
+    {
+      "epoch": 0.0380952380952381,
+      "grad_norm": 42.75063943291744,
+      "learning_rate": 1.6666666666666667e-06,
+      "loss": 0.998,
+      "step": 9
+    },
+    {
+      "epoch": 0.042328042328042326,
+      "grad_norm": 43.214149555747724,
+      "learning_rate": 1.8750000000000003e-06,
+      "loss": 0.942,
+      "step": 10
+    },
+    {
+      "epoch": 0.04656084656084656,
+      "grad_norm": 33.48353944926147,
+      "learning_rate": 2.0833333333333334e-06,
+      "loss": 0.6964,
+      "step": 11
+    },
+    {
+      "epoch": 0.050793650793650794,
+      "grad_norm": 33.50023841667552,
+      "learning_rate": 2.2916666666666666e-06,
+      "loss": 0.8639,
+      "step": 12
+    },
+    {
+      "epoch": 0.05502645502645503,
+      "grad_norm": 28.838714912036302,
+      "learning_rate": 2.5e-06,
+      "loss": 0.8166,
+      "step": 13
+    },
+    {
+      "epoch": 0.05925925925925926,
+      "grad_norm": 13.200524499309642,
+      "learning_rate": 2.7083333333333334e-06,
+      "loss": 0.6199,
+      "step": 14
+    },
+    {
+      "epoch": 0.06349206349206349,
+      "grad_norm": 12.22641881931881,
+      "learning_rate": 2.916666666666667e-06,
+      "loss": 0.5748,
+      "step": 15
+    },
+    {
+      "epoch": 0.06772486772486773,
+      "grad_norm": 11.38831601039375,
+      "learning_rate": 3.125e-06,
+      "loss": 0.6128,
+      "step": 16
+    },
+    {
+      "epoch": 0.07195767195767196,
+      "grad_norm": 9.044088753747445,
+      "learning_rate": 3.3333333333333333e-06,
+      "loss": 0.5587,
+      "step": 17
+    },
+    {
+      "epoch": 0.0761904761904762,
+      "grad_norm": 7.844366385332237,
+      "learning_rate": 3.5416666666666673e-06,
+      "loss": 0.5319,
+      "step": 18
+    },
+    {
+      "epoch": 0.08042328042328042,
+      "grad_norm": 5.731160381356699,
+      "learning_rate": 3.7500000000000005e-06,
+      "loss": 0.4898,
+      "step": 19
+    },
+    {
+      "epoch": 0.08465608465608465,
+      "grad_norm": 4.726984164569137,
+      "learning_rate": 3.958333333333333e-06,
+      "loss": 0.5398,
+      "step": 20
+    },
+    {
+      "epoch": 0.08888888888888889,
+      "grad_norm": 4.5781542459814535,
+      "learning_rate": 4.166666666666667e-06,
+      "loss": 0.3968,
+      "step": 21
+    },
+    {
+      "epoch": 0.09312169312169312,
+      "grad_norm": 3.7775510520748368,
+      "learning_rate": 4.3750000000000005e-06,
+      "loss": 0.3964,
+      "step": 22
+    },
+    {
+      "epoch": 0.09735449735449736,
+      "grad_norm": 3.05477842541282,
+      "learning_rate": 4.583333333333333e-06,
+      "loss": 0.4362,
+      "step": 23
+    },
+    {
+      "epoch": 0.10158730158730159,
+      "grad_norm": 3.438531926670594,
+      "learning_rate": 4.791666666666668e-06,
+      "loss": 0.458,
+      "step": 24
+    },
+    {
+      "epoch": 0.10582010582010581,
+      "grad_norm": 3.8238853915379662,
+      "learning_rate": 5e-06,
+      "loss": 0.4659,
+      "step": 25
+    },
+    {
+      "epoch": 0.11005291005291006,
+      "grad_norm": 2.8947366693177288,
+      "learning_rate": 5.208333333333334e-06,
+      "loss": 0.3804,
+      "step": 26
+    },
+    {
+      "epoch": 0.11428571428571428,
+      "grad_norm": 3.6390451997485584,
+      "learning_rate": 5.416666666666667e-06,
+      "loss": 0.4531,
+      "step": 27
+    },
+    {
+      "epoch": 0.11851851851851852,
+      "grad_norm": 2.9482232685782703,
+      "learning_rate": 5.625e-06,
+      "loss": 0.3826,
+      "step": 28
+    },
+    {
+      "epoch": 0.12275132275132275,
+      "grad_norm": 3.674602280740873,
+      "learning_rate": 5.833333333333334e-06,
+      "loss": 0.4131,
+      "step": 29
+    },
+    {
+      "epoch": 0.12698412698412698,
+      "grad_norm": 3.123169330914684,
+      "learning_rate": 6.041666666666667e-06,
+      "loss": 0.3337,
+      "step": 30
+    },
+    {
+      "epoch": 0.1312169312169312,
+      "grad_norm": 2.8563692549958866,
+      "learning_rate": 6.25e-06,
+      "loss": 0.3496,
+      "step": 31
+    },
+    {
+      "epoch": 0.13544973544973546,
+      "grad_norm": 2.5685868294187646,
+      "learning_rate": 6.458333333333334e-06,
+      "loss": 0.3464,
+      "step": 32
+    },
+    {
+      "epoch": 0.13968253968253969,
+      "grad_norm": 2.7561118762037933,
+      "learning_rate": 6.666666666666667e-06,
+      "loss": 0.3553,
+      "step": 33
+    },
+    {
+      "epoch": 0.1439153439153439,
+      "grad_norm": 2.799437451043737,
+      "learning_rate": 6.875e-06,
+      "loss": 0.3412,
+      "step": 34
+    },
+    {
+      "epoch": 0.14814814814814814,
+      "grad_norm": 3.292845577989745,
+      "learning_rate": 7.083333333333335e-06,
+      "loss": 0.3276,
+      "step": 35
+    },
+    {
+      "epoch": 0.1523809523809524,
+      "grad_norm": 2.3306789119643936,
+      "learning_rate": 7.291666666666667e-06,
+      "loss": 0.2384,
+      "step": 36
+    },
+    {
+      "epoch": 0.15661375661375662,
+      "grad_norm": 2.803053991451241,
+      "learning_rate": 7.500000000000001e-06,
+      "loss": 0.3443,
+      "step": 37
+    },
+    {
+      "epoch": 0.16084656084656085,
+      "grad_norm": 2.8701427259214065,
+      "learning_rate": 7.708333333333334e-06,
+      "loss": 0.3013,
+      "step": 38
+    },
+    {
+      "epoch": 0.16507936507936508,
+      "grad_norm": 2.537652900045535,
+      "learning_rate": 7.916666666666667e-06,
+      "loss": 0.3267,
+      "step": 39
+    },
+    {
+      "epoch": 0.1693121693121693,
+      "grad_norm": 2.6768501900523223,
+      "learning_rate": 8.125000000000001e-06,
+      "loss": 0.2639,
+      "step": 40
+    },
+    {
+      "epoch": 0.17354497354497356,
+      "grad_norm": 2.6577601401008333,
+      "learning_rate": 8.333333333333334e-06,
+      "loss": 0.3324,
+      "step": 41
+    },
+    {
+      "epoch": 0.17777777777777778,
+      "grad_norm": 2.5512677479663632,
+      "learning_rate": 8.541666666666666e-06,
+      "loss": 0.2672,
+      "step": 42
+    },
+    {
+      "epoch": 0.182010582010582,
+      "grad_norm": 2.8990280872171805,
+      "learning_rate": 8.750000000000001e-06,
+      "loss": 0.2439,
+      "step": 43
+    },
+    {
+      "epoch": 0.18624338624338624,
+      "grad_norm": 2.889097227933124,
+      "learning_rate": 8.958333333333334e-06,
+      "loss": 0.3163,
+      "step": 44
+    },
+    {
+      "epoch": 0.19047619047619047,
+      "grad_norm": 2.9104235083784102,
+      "learning_rate": 9.166666666666666e-06,
+      "loss": 0.3205,
+      "step": 45
+    },
+    {
+      "epoch": 0.19470899470899472,
+      "grad_norm": 3.083107801252212,
+      "learning_rate": 9.375000000000001e-06,
+      "loss": 0.3172,
+      "step": 46
+    },
+    {
+      "epoch": 0.19894179894179895,
+      "grad_norm": 2.3884460518664374,
+      "learning_rate": 9.583333333333335e-06,
+      "loss": 0.1934,
+      "step": 47
+    },
+    {
+      "epoch": 0.20317460317460317,
+      "grad_norm": 3.226585957430524,
+      "learning_rate": 9.791666666666666e-06,
+      "loss": 0.2701,
+      "step": 48
+    },
+    {
+      "epoch": 0.2074074074074074,
+      "grad_norm": 2.7423891340389615,
+      "learning_rate": 1e-05,
+      "loss": 0.2527,
+      "step": 49
+    },
+    {
+      "epoch": 0.21164021164021163,
+      "grad_norm": 2.820228218301591,
+      "learning_rate": 9.999969538288953e-06,
+      "loss": 0.2192,
+      "step": 50
+    },
+    {
+      "epoch": 0.21587301587301588,
+      "grad_norm": 2.467968682909313,
+      "learning_rate": 9.999878153526974e-06,
+      "loss": 0.2493,
+      "step": 51
+    },
+    {
+      "epoch": 0.2201058201058201,
+      "grad_norm": 2.66662023011886,
+      "learning_rate": 9.999725846827562e-06,
+      "loss": 0.2806,
+      "step": 52
+    },
+    {
+      "epoch": 0.22433862433862434,
+      "grad_norm": 3.129707362043057,
+      "learning_rate": 9.999512620046523e-06,
+      "loss": 0.309,
+      "step": 53
+    },
+    {
+      "epoch": 0.22857142857142856,
+      "grad_norm": 2.56704704796148,
+      "learning_rate": 9.999238475781957e-06,
+      "loss": 0.2664,
+      "step": 54
+    },
+    {
+      "epoch": 0.2328042328042328,
+      "grad_norm": 2.351657358687772,
+      "learning_rate": 9.998903417374228e-06,
+      "loss": 0.2318,
+      "step": 55
+    },
+    {
+      "epoch": 0.23703703703703705,
+      "grad_norm": 2.616314308180943,
+      "learning_rate": 9.998507448905917e-06,
+      "loss": 0.2548,
+      "step": 56
+    },
+    {
+      "epoch": 0.24126984126984127,
+      "grad_norm": 2.740317144109295,
+      "learning_rate": 9.998050575201772e-06,
+      "loss": 0.2506,
+      "step": 57
+    },
+    {
+      "epoch": 0.2455026455026455,
+      "grad_norm": 2.1871378697009396,
+      "learning_rate": 9.997532801828659e-06,
+      "loss": 0.2187,
+      "step": 58
+    },
+    {
+      "epoch": 0.24973544973544973,
+      "grad_norm": 3.3209760924838423,
+      "learning_rate": 9.99695413509548e-06,
+      "loss": 0.2841,
+      "step": 59
+    },
+    {
+      "epoch": 0.25396825396825395,
+      "grad_norm": 2.447101187980101,
+      "learning_rate": 9.996314582053106e-06,
+      "loss": 0.2218,
+      "step": 60
+    },
+    {
+      "epoch": 0.2582010582010582,
+      "grad_norm": 2.833727508307957,
+      "learning_rate": 9.995614150494293e-06,
+      "loss": 0.2551,
+      "step": 61
+    },
+    {
+      "epoch": 0.2624338624338624,
+      "grad_norm": 2.544105354348358,
+      "learning_rate": 9.994852848953574e-06,
+      "loss": 0.2213,
+      "step": 62
+    },
+    {
+      "epoch": 0.26666666666666666,
+      "grad_norm": 2.4267488774897927,
+      "learning_rate": 9.994030686707171e-06,
+      "loss": 0.1665,
+      "step": 63
+    },
+    {
+      "epoch": 0.2708994708994709,
+      "grad_norm": 1.8496876818825865,
+      "learning_rate": 9.993147673772869e-06,
+      "loss": 0.176,
+      "step": 64
+    },
+    {
+      "epoch": 0.2751322751322751,
+      "grad_norm": 2.2079265341473375,
+      "learning_rate": 9.992203820909906e-06,
+      "loss": 0.2061,
+      "step": 65
+    },
+    {
+      "epoch": 0.27936507936507937,
+      "grad_norm": 2.7008947539879693,
+      "learning_rate": 9.991199139618828e-06,
+      "loss": 0.2402,
+      "step": 66
+    },
+    {
+      "epoch": 0.28359788359788357,
+      "grad_norm": 1.9380719993127324,
+      "learning_rate": 9.990133642141359e-06,
+      "loss": 0.2095,
+      "step": 67
+    },
+    {
+      "epoch": 0.2878306878306878,
+      "grad_norm": 2.3742267443315432,
+      "learning_rate": 9.989007341460251e-06,
+      "loss": 0.2176,
+      "step": 68
+    },
+    {
+      "epoch": 0.2920634920634921,
+      "grad_norm": 2.1948771063063632,
+      "learning_rate": 9.987820251299121e-06,
+      "loss": 0.2145,
+      "step": 69
+    },
+    {
+      "epoch": 0.2962962962962963,
+      "grad_norm": 1.9006443729448308,
+      "learning_rate": 9.98657238612229e-06,
+      "loss": 0.2187,
+      "step": 70
+    },
+    {
+      "epoch": 0.30052910052910053,
+      "grad_norm": 2.0165710016106995,
+      "learning_rate": 9.985263761134602e-06,
+      "loss": 0.1908,
+      "step": 71
+    },
+    {
+      "epoch": 0.3047619047619048,
+      "grad_norm": 2.05912929127493,
+      "learning_rate": 9.983894392281237e-06,
+      "loss": 0.2137,
+      "step": 72
+    },
+    {
+      "epoch": 0.308994708994709,
+      "grad_norm": 2.810426086613887,
+      "learning_rate": 9.982464296247523e-06,
+      "loss": 0.2565,
+      "step": 73
+    },
+    {
+      "epoch": 0.31322751322751324,
+      "grad_norm": 2.3766628874747937,
+      "learning_rate": 9.980973490458728e-06,
+      "loss": 0.1442,
+      "step": 74
+    },
+    {
+      "epoch": 0.31746031746031744,
+      "grad_norm": 2.4709186657411233,
+      "learning_rate": 9.979421993079853e-06,
+      "loss": 0.2542,
+      "step": 75
+    },
+    {
+      "epoch": 0.3216931216931217,
+      "grad_norm": 1.9082809595138144,
+      "learning_rate": 9.9778098230154e-06,
+      "loss": 0.19,
+      "step": 76
+    },
+    {
+      "epoch": 0.32592592592592595,
+      "grad_norm": 2.071460019416873,
+      "learning_rate": 9.976136999909156e-06,
+      "loss": 0.1663,
+      "step": 77
+    },
+    {
+      "epoch": 0.33015873015873015,
+      "grad_norm": 2.516711677394165,
+      "learning_rate": 9.974403544143942e-06,
+      "loss": 0.1794,
+      "step": 78
+    },
+    {
+      "epoch": 0.3343915343915344,
+      "grad_norm": 1.7925945452723815,
+      "learning_rate": 9.972609476841368e-06,
+      "loss": 0.1671,
+      "step": 79
+    },
+    {
+      "epoch": 0.3386243386243386,
+      "grad_norm": 1.9441478155371381,
+      "learning_rate": 9.970754819861577e-06,
+      "loss": 0.1743,
+      "step": 80
+    },
+    {
+      "epoch": 0.34285714285714286,
+      "grad_norm": 2.6254833507520234,
+      "learning_rate": 9.968839595802982e-06,
+      "loss": 0.2414,
+      "step": 81
+    },
+    {
+      "epoch": 0.3470899470899471,
+      "grad_norm": 2.316238733652428,
+      "learning_rate": 9.966863828001982e-06,
+      "loss": 0.2089,
+      "step": 82
+    },
+    {
+      "epoch": 0.3513227513227513,
+      "grad_norm": 2.0519535959849593,
+      "learning_rate": 9.964827540532685e-06,
+      "loss": 0.2097,
+      "step": 83
+    },
+    {
+      "epoch": 0.35555555555555557,
+      "grad_norm": 2.057989705603544,
+      "learning_rate": 9.962730758206612e-06,
+      "loss": 0.1938,
+      "step": 84
+    },
+    {
+      "epoch": 0.35978835978835977,
+      "grad_norm": 2.672807955843952,
+      "learning_rate": 9.960573506572391e-06,
+      "loss": 0.258,
+      "step": 85
+    },
+    {
+      "epoch": 0.364021164021164,
+      "grad_norm": 2.2091594991845716,
+      "learning_rate": 9.958355811915452e-06,
+      "loss": 0.1949,
+      "step": 86
+    },
+    {
+      "epoch": 0.3682539682539683,
+      "grad_norm": 2.179853329030986,
+      "learning_rate": 9.95607770125771e-06,
+      "loss": 0.2111,
+      "step": 87
+    },
+    {
+      "epoch": 0.3724867724867725,
+      "grad_norm": 1.925480718448465,
+      "learning_rate": 9.953739202357219e-06,
+      "loss": 0.1621,
+      "step": 88
+    },
+    {
+      "epoch": 0.37671957671957673,
+      "grad_norm": 2.5097628215455994,
+      "learning_rate": 9.951340343707852e-06,
+      "loss": 0.1916,
+      "step": 89
+    },
+    {
+      "epoch": 0.38095238095238093,
+      "grad_norm": 2.2291743713915815,
+      "learning_rate": 9.948881154538946e-06,
+      "loss": 0.1901,
+      "step": 90
+    },
+    {
+      "epoch": 0.3851851851851852,
+      "grad_norm": 2.2427956711484147,
+      "learning_rate": 9.946361664814942e-06,
+      "loss": 0.1926,
+      "step": 91
+    },
+    {
+      "epoch": 0.38941798941798944,
+      "grad_norm": 2.1674595305785966,
+      "learning_rate": 9.94378190523503e-06,
+      "loss": 0.15,
+      "step": 92
+    },
+    {
+      "epoch": 0.39365079365079364,
+      "grad_norm": 2.065344580545721,
+      "learning_rate": 9.941141907232766e-06,
+      "loss": 0.151,
+      "step": 93
+    },
+    {
+      "epoch": 0.3978835978835979,
+      "grad_norm": 2.4498733708415332,
+      "learning_rate": 9.938441702975689e-06,
+      "loss": 0.1991,
+      "step": 94
+    },
+    {
+      "epoch": 0.4021164021164021,
+      "grad_norm": 2.196562226640055,
+      "learning_rate": 9.93568132536494e-06,
+      "loss": 0.1565,
+      "step": 95
+    },
+    {
+      "epoch": 0.40634920634920635,
+      "grad_norm": 2.45021933901518,
+      "learning_rate": 9.932860808034847e-06,
+      "loss": 0.1954,
+      "step": 96
+    },
+    {
+      "epoch": 0.4105820105820106,
+      "grad_norm": 1.6870920962679627,
+      "learning_rate": 9.929980185352525e-06,
+      "loss": 0.1415,
+      "step": 97
+    },
+    {
+      "epoch": 0.4148148148148148,
+      "grad_norm": 2.0433028628482277,
+      "learning_rate": 9.927039492417452e-06,
+      "loss": 0.1752,
+      "step": 98
+    },
+    {
+      "epoch": 0.41904761904761906,
+      "grad_norm": 2.3924574752175576,
+      "learning_rate": 9.924038765061042e-06,
+      "loss": 0.169,
+      "step": 99
+    },
+    {
+      "epoch": 0.42328042328042326,
+      "grad_norm": 2.1514028524429563,
+      "learning_rate": 9.92097803984621e-06,
+      "loss": 0.2042,
+      "step": 100
+    },
+    {
+      "epoch": 0.4275132275132275,
+      "grad_norm": 1.9233810049291866,
+      "learning_rate": 9.91785735406693e-06,
+      "loss": 0.1625,
+      "step": 101
+    },
+    {
+      "epoch": 0.43174603174603177,
+      "grad_norm": 2.0269343004100433,
+      "learning_rate": 9.914676745747772e-06,
+      "loss": 0.1783,
+      "step": 102
+    },
+    {
+      "epoch": 0.43597883597883597,
+      "grad_norm": 2.9018403399357657,
+      "learning_rate": 9.911436253643445e-06,
+      "loss": 0.2005,
+      "step": 103
+    },
+    {
+      "epoch": 0.4402116402116402,
+      "grad_norm": 2.3864889162085694,
+      "learning_rate": 9.908135917238321e-06,
+      "loss": 0.1899,
+      "step": 104
+    },
+    {
+      "epoch": 0.4444444444444444,
+      "grad_norm": 2.029884804297179,
+      "learning_rate": 9.904775776745959e-06,
+      "loss": 0.1646,
+      "step": 105
+    },
+    {
+      "epoch": 0.4486772486772487,
+      "grad_norm": 2.217474285635462,
+      "learning_rate": 9.901355873108611e-06,
+      "loss": 0.1975,
+      "step": 106
+    },
+    {
+      "epoch": 0.45291005291005293,
+      "grad_norm": 3.316060847140544,
+      "learning_rate": 9.89787624799672e-06,
+      "loss": 0.1757,
+      "step": 107
+    },
+    {
+      "epoch": 0.45714285714285713,
+      "grad_norm": 2.256786436988083,
+      "learning_rate": 9.894336943808426e-06,
+      "loss": 0.1808,
+      "step": 108
+    },
+    {
+      "epoch": 0.4613756613756614,
+      "grad_norm": 2.3520584460075855,
+      "learning_rate": 9.890738003669029e-06,
+      "loss": 0.1832,
+      "step": 109
+    },
+    {
+      "epoch": 0.4656084656084656,
+      "grad_norm": 2.819589162595303,
+      "learning_rate": 9.887079471430481e-06,
+      "loss": 0.1937,
+      "step": 110
+    },
+    {
+      "epoch": 0.46984126984126984,
+      "grad_norm": 2.0799214485519713,
+      "learning_rate": 9.883361391670841e-06,
+      "loss": 0.1204,
+      "step": 111
+    },
+    {
+      "epoch": 0.4740740740740741,
+      "grad_norm": 2.4098902856637148,
+      "learning_rate": 9.879583809693737e-06,
+      "loss": 0.1774,
+      "step": 112
+    },
+    {
+      "epoch": 0.4783068783068783,
+      "grad_norm": 2.4639460942435334,
+      "learning_rate": 9.875746771527817e-06,
+      "loss": 0.1954,
+      "step": 113
+    },
+    {
+      "epoch": 0.48253968253968255,
+      "grad_norm": 1.8744056955663004,
+      "learning_rate": 9.871850323926178e-06,
+      "loss": 0.186,
+      "step": 114
+    },
+    {
+      "epoch": 0.48677248677248675,
+      "grad_norm": 2.1213169931981226,
+      "learning_rate": 9.867894514365802e-06,
+      "loss": 0.1425,
+      "step": 115
+    },
+    {
+      "epoch": 0.491005291005291,
+      "grad_norm": 2.1632110720054962,
+      "learning_rate": 9.863879391046985e-06,
+      "loss": 0.1552,
+      "step": 116
+    },
+    {
+      "epoch": 0.49523809523809526,
+      "grad_norm": 2.7544098112396282,
+      "learning_rate": 9.859805002892733e-06,
+      "loss": 0.1432,
+      "step": 117
+    },
+    {
+      "epoch": 0.49947089947089945,
+      "grad_norm": 2.100186269105726,
+      "learning_rate": 9.85567139954818e-06,
+      "loss": 0.1648,
+      "step": 118
+    },
+    {
+      "epoch": 0.5037037037037037,
+      "grad_norm": 1.9132320908075728,
+      "learning_rate": 9.851478631379982e-06,
+      "loss": 0.1645,
+      "step": 119
+    },
+    {
+      "epoch": 0.5079365079365079,
+      "grad_norm": 2.080325184083409,
+      "learning_rate": 9.847226749475696e-06,
+      "loss": 0.1421,
+      "step": 120
+    },
+    {
+      "epoch": 0.5121693121693122,
+      "grad_norm": 2.1166691740240946,
+      "learning_rate": 9.842915805643156e-06,
+      "loss": 0.1739,
+      "step": 121
+    },
+    {
+      "epoch": 0.5164021164021164,
+      "grad_norm": 2.7777924615211402,
+      "learning_rate": 9.838545852409857e-06,
+      "loss": 0.2158,
+      "step": 122
+    },
+    {
+      "epoch": 0.5206349206349207,
+      "grad_norm": 2.011825936444935,
+      "learning_rate": 9.834116943022299e-06,
+      "loss": 0.1283,
+      "step": 123
+    },
+    {
+      "epoch": 0.5248677248677248,
+      "grad_norm": 2.3088090393013707,
+      "learning_rate": 9.829629131445342e-06,
+      "loss": 0.1594,
+      "step": 124
+    },
+    {
+      "epoch": 0.5291005291005291,
+      "grad_norm": 2.5599162277007506,
+      "learning_rate": 9.825082472361558e-06,
+      "loss": 0.1602,
+      "step": 125
+    },
+    {
+      "epoch": 0.5333333333333333,
+      "grad_norm": 2.8804298377274624,
+      "learning_rate": 9.82047702117055e-06,
+      "loss": 0.21,
+      "step": 126
+    },
+    {
+      "epoch": 0.5375661375661376,
+      "grad_norm": 2.322574871028054,
+      "learning_rate": 9.815812833988292e-06,
+      "loss": 0.1894,
+      "step": 127
+    },
+    {
+      "epoch": 0.5417989417989418,
+      "grad_norm": 2.3497044645708662,
+      "learning_rate": 9.811089967646427e-06,
+      "loss": 0.2065,
+      "step": 128
+    },
+    {
+      "epoch": 0.546031746031746,
+      "grad_norm": 2.1840543614377785,
+      "learning_rate": 9.806308479691595e-06,
+      "loss": 0.1752,
+      "step": 129
+    },
+    {
+      "epoch": 0.5502645502645502,
+      "grad_norm": 1.9149801234923844,
+      "learning_rate": 9.801468428384716e-06,
+      "loss": 0.1449,
+      "step": 130
+    },
+    {
+      "epoch": 0.5544973544973545,
+      "grad_norm": 1.9431409264703756,
+      "learning_rate": 9.796569872700287e-06,
+      "loss": 0.1562,
+      "step": 131
+    },
+    {
+      "epoch": 0.5587301587301587,
+      "grad_norm": 1.9277634151443268,
+      "learning_rate": 9.791612872325667e-06,
+      "loss": 0.1665,
+      "step": 132
+    },
+    {
+      "epoch": 0.562962962962963,
+      "grad_norm": 2.6741227436047703,
+      "learning_rate": 9.786597487660336e-06,
+      "loss": 0.195,
+      "step": 133
+    },
+    {
+      "epoch": 0.5671957671957671,
+      "grad_norm": 1.9587886510325367,
+      "learning_rate": 9.781523779815178e-06,
+      "loss": 0.1496,
+      "step": 134
+    },
+    {
+      "epoch": 0.5714285714285714,
+      "grad_norm": 2.167004574977241,
+      "learning_rate": 9.776391810611719e-06,
+      "loss": 0.1873,
+      "step": 135
+    },
+    {
+      "epoch": 0.5756613756613757,
+      "grad_norm": 2.3098964403422717,
+      "learning_rate": 9.771201642581384e-06,
+      "loss": 0.1448,
+      "step": 136
+    },
+    {
+      "epoch": 0.5798941798941799,
+      "grad_norm": 2.2442486200909886,
+      "learning_rate": 9.765953338964736e-06,
+      "loss": 0.1787,
+      "step": 137
+    },
+    {
+      "epoch": 0.5841269841269842,
+      "grad_norm": 2.213445354502761,
+      "learning_rate": 9.760646963710694e-06,
+      "loss": 0.2045,
+      "step": 138
+    },
+    {
+      "epoch": 0.5883597883597883,
+      "grad_norm": 1.7015924599869439,
+      "learning_rate": 9.755282581475769e-06,
+      "loss": 0.1105,
+      "step": 139
+    },
+    {
+      "epoch": 0.5925925925925926,
+      "grad_norm": 2.135699337583285,
+      "learning_rate": 9.749860257623262e-06,
+      "loss": 0.1696,
+      "step": 140
+    },
+    {
+      "epoch": 0.5968253968253968,
+      "grad_norm": 2.022004867298511,
+      "learning_rate": 9.744380058222483e-06,
+      "loss": 0.1436,
+      "step": 141
+    },
+    {
+      "epoch": 0.6010582010582011,
+      "grad_norm": 1.938130175850262,
+      "learning_rate": 9.73884205004793e-06,
+      "loss": 0.152,
+      "step": 142
+    },
+    {
+      "epoch": 0.6052910052910053,
+      "grad_norm": 2.2793353633377644,
+      "learning_rate": 9.733246300578482e-06,
+      "loss": 0.1801,
+      "step": 143
+    },
+    {
+      "epoch": 0.6095238095238096,
+      "grad_norm": 2.3515552282000023,
+      "learning_rate": 9.727592877996585e-06,
+      "loss": 0.1709,
+      "step": 144
+    },
+    {
+      "epoch": 0.6137566137566137,
+      "grad_norm": 2.4174135230244693,
+      "learning_rate": 9.721881851187406e-06,
+      "loss": 0.1888,
+      "step": 145
+    },
+    {
+      "epoch": 0.617989417989418,
+      "grad_norm": 2.02003900874957,
+      "learning_rate": 9.716113289738005e-06,
+      "loss": 0.1549,
+      "step": 146
+    },
+    {
+      "epoch": 0.6222222222222222,
+      "grad_norm": 2.1212067581354637,
+      "learning_rate": 9.710287263936485e-06,
+      "loss": 0.1463,
+      "step": 147
+    },
+    {
+      "epoch": 0.6264550264550265,
+      "grad_norm": 2.10847848049695,
+      "learning_rate": 9.704403844771128e-06,
+      "loss": 0.1548,
+      "step": 148
+    },
+    {
+      "epoch": 0.6306878306878307,
+      "grad_norm": 2.1977701291513783,
+      "learning_rate": 9.698463103929542e-06,
+      "loss": 0.1842,
+      "step": 149
+    },
+    {
+      "epoch": 0.6349206349206349,
+      "grad_norm": 2.0234900314907303,
+      "learning_rate": 9.69246511379778e-06,
+      "loss": 0.1564,
+      "step": 150
+    },
+    {
+      "epoch": 0.6391534391534391,
+      "grad_norm": 2.221863968702518,
+      "learning_rate": 9.68640994745946e-06,
+      "loss": 0.1492,
+      "step": 151
+    },
+    {
+      "epoch": 0.6433862433862434,
+      "grad_norm": 2.2957077394485865,
+      "learning_rate": 9.680297678694867e-06,
+      "loss": 0.1681,
+      "step": 152
+    },
+    {
+      "epoch": 0.6476190476190476,
+      "grad_norm": 2.319080436137813,
+      "learning_rate": 9.674128381980073e-06,
+      "loss": 0.1584,
+      "step": 153
+    },
+    {
+      "epoch": 0.6518518518518519,
+      "grad_norm": 2.081726848575624,
+      "learning_rate": 9.667902132486009e-06,
+      "loss": 0.1605,
+      "step": 154
+    },
+    {
+      "epoch": 0.656084656084656,
+      "grad_norm": 1.696442003069325,
+      "learning_rate": 9.661619006077562e-06,
+      "loss": 0.1175,
+      "step": 155
+    },
+    {
+      "epoch": 0.6603174603174603,
+      "grad_norm": 1.9060631377388697,
+      "learning_rate": 9.655279079312643e-06,
+      "loss": 0.1327,
+      "step": 156
+    },
+    {
+      "epoch": 0.6645502645502646,
+      "grad_norm": 2.437272536385666,
+      "learning_rate": 9.648882429441258e-06,
+      "loss": 0.1673,
+      "step": 157
+    },
+    {
+      "epoch": 0.6687830687830688,
+      "grad_norm": 2.5407710665248406,
+      "learning_rate": 9.642429134404568e-06,
+      "loss": 0.1795,
+      "step": 158
+    },
+    {
+      "epoch": 0.6730158730158731,
+      "grad_norm": 1.7280628828064413,
+      "learning_rate": 9.635919272833938e-06,
+      "loss": 0.1192,
+      "step": 159
+    },
+    {
+      "epoch": 0.6772486772486772,
+      "grad_norm": 2.0466311163540554,
+      "learning_rate": 9.629352924049975e-06,
+      "loss": 0.1641,
+      "step": 160
+    },
+    {
+      "epoch": 0.6814814814814815,
+      "grad_norm": 2.430559199002775,
+      "learning_rate": 9.622730168061568e-06,
+      "loss": 0.164,
+      "step": 161
+    },
+    {
+      "epoch": 0.6857142857142857,
+      "grad_norm": 1.8034268409827128,
+      "learning_rate": 9.616051085564905e-06,
+      "loss": 0.1151,
+      "step": 162
+    },
+    {
+      "epoch": 0.68994708994709,
+      "grad_norm": 2.076098544214192,
+      "learning_rate": 9.609315757942504e-06,
+      "loss": 0.1356,
+      "step": 163
+    },
+    {
+      "epoch": 0.6941798941798942,
+      "grad_norm": 1.8175196422774194,
+      "learning_rate": 9.602524267262202e-06,
+      "loss": 0.1286,
+      "step": 164
+    },
+    {
+      "epoch": 0.6984126984126984,
+      "grad_norm": 1.9844294374596974,
+      "learning_rate": 9.595676696276173e-06,
+      "loss": 0.1447,
+      "step": 165
+    },
+    {
+      "epoch": 0.7026455026455026,
+      "grad_norm": 2.0850373939499103,
+      "learning_rate": 9.588773128419907e-06,
+      "loss": 0.1641,
+      "step": 166
+    },
+    {
+      "epoch": 0.7068783068783069,
+      "grad_norm": 2.1830728529262213,
+      "learning_rate": 9.581813647811199e-06,
+      "loss": 0.146,
+      "step": 167
+    },
+    {
+      "epoch": 0.7111111111111111,
+      "grad_norm": 1.9107339137949033,
+      "learning_rate": 9.574798339249124e-06,
+      "loss": 0.1424,
+      "step": 168
+    },
+    {
+      "epoch": 0.7153439153439154,
+      "grad_norm": 1.9073629765575915,
+      "learning_rate": 9.567727288213005e-06,
+      "loss": 0.1321,
+      "step": 169
+    },
+    {
+      "epoch": 0.7195767195767195,
+      "grad_norm": 1.7910676565872312,
+      "learning_rate": 9.560600580861366e-06,
+      "loss": 0.124,
+      "step": 170
+    },
+    {
+      "epoch": 0.7238095238095238,
+      "grad_norm": 2.3559878393889444,
+      "learning_rate": 9.553418304030886e-06,
+      "loss": 0.1542,
+      "step": 171
+    },
+    {
+      "epoch": 0.728042328042328,
+      "grad_norm": 2.257035851112734,
+      "learning_rate": 9.546180545235344e-06,
+      "loss": 0.1811,
+      "step": 172
+    },
+    {
+      "epoch": 0.7322751322751323,
+      "grad_norm": 2.372408574995787,
+      "learning_rate": 9.538887392664544e-06,
+      "loss": 0.1415,
+      "step": 173
+    },
+    {
+      "epoch": 0.7365079365079366,
+      "grad_norm": 2.1859794226240643,
+      "learning_rate": 9.531538935183252e-06,
+      "loss": 0.164,
+      "step": 174
+    },
+    {
+      "epoch": 0.7407407407407407,
+      "grad_norm": 1.9315910825592963,
+      "learning_rate": 9.524135262330098e-06,
+      "loss": 0.1242,
+      "step": 175
+    },
+    {
+      "epoch": 0.744973544973545,
+      "grad_norm": 1.932293530910452,
+      "learning_rate": 9.516676464316505e-06,
+      "loss": 0.1223,
+      "step": 176
+    },
+    {
+      "epoch": 0.7492063492063492,
+      "grad_norm": 2.5245512957918037,
+      "learning_rate": 9.50916263202557e-06,
+      "loss": 0.1758,
+      "step": 177
+    },
+    {
+      "epoch": 0.7534391534391535,
+      "grad_norm": 2.2836095033018307,
+      "learning_rate": 9.501593857010968e-06,
+      "loss": 0.1472,
+      "step": 178
+    },
+    {
+      "epoch": 0.7576719576719577,
+      "grad_norm": 2.0376880346254227,
+      "learning_rate": 9.493970231495836e-06,
+      "loss": 0.1489,
+      "step": 179
+    },
+    {
+      "epoch": 0.7619047619047619,
+      "grad_norm": 1.9142818784287132,
+      "learning_rate": 9.486291848371642e-06,
+      "loss": 0.1242,
+      "step": 180
+    },
+    {
+      "epoch": 0.7661375661375661,
+      "grad_norm": 2.0433594909925823,
+      "learning_rate": 9.478558801197065e-06,
+      "loss": 0.1246,
+      "step": 181
+    },
+    {
+      "epoch": 0.7703703703703704,
+      "grad_norm": 1.814914680097962,
+      "learning_rate": 9.470771184196842e-06,
+      "loss": 0.1178,
+      "step": 182
+    },
+    {
+      "epoch": 0.7746031746031746,
+      "grad_norm": 2.038031930273932,
+      "learning_rate": 9.46292909226063e-06,
+      "loss": 0.1741,
+      "step": 183
+    },
+    {
+      "epoch": 0.7788359788359789,
+      "grad_norm": 2.2491615885983025,
+      "learning_rate": 9.45503262094184e-06,
+      "loss": 0.1428,
+      "step": 184
+    },
+    {
+      "epoch": 0.783068783068783,
+      "grad_norm": 1.9769912092154236,
+      "learning_rate": 9.44708186645649e-06,
+      "loss": 0.1431,
+      "step": 185
+    },
+    {
+      "epoch": 0.7873015873015873,
+      "grad_norm": 1.908015924501842,
+      "learning_rate": 9.439076925682006e-06,
+      "loss": 0.1249,
+      "step": 186
+    },
+    {
+      "epoch": 0.7915343915343915,
+      "grad_norm": 1.8504394720752184,
+      "learning_rate": 9.431017896156074e-06,
+      "loss": 0.1205,
+      "step": 187
+    },
+    {
+      "epoch": 0.7957671957671958,
+      "grad_norm": 2.0652458552127384,
+      "learning_rate": 9.42290487607542e-06,
+      "loss": 0.1333,
+      "step": 188
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.065167228642111,
+      "learning_rate": 9.414737964294636e-06,
+      "loss": 0.1407,
+      "step": 189
+    },
+    {
+      "epoch": 0.8042328042328042,
+      "grad_norm": 2.1818559331219234,
+      "learning_rate": 9.406517260324962e-06,
+      "loss": 0.127,
+      "step": 190
+    },
+    {
+      "epoch": 0.8084656084656084,
+      "grad_norm": 1.9594250023861293,
+      "learning_rate": 9.398242864333084e-06,
+      "loss": 0.1196,
+      "step": 191
+    },
+    {
+      "epoch": 0.8126984126984127,
+      "grad_norm": 2.143484340303363,
+      "learning_rate": 9.389914877139903e-06,
+      "loss": 0.1361,
+      "step": 192
+    },
+    {
+      "epoch": 0.816931216931217,
+      "grad_norm": 1.925094740498635,
+      "learning_rate": 9.381533400219319e-06,
+      "loss": 0.1089,
+      "step": 193
+    },
+    {
+      "epoch": 0.8211640211640212,
+      "grad_norm": 2.951061076480981,
+      "learning_rate": 9.37309853569698e-06,
+      "loss": 0.1371,
+      "step": 194
+    },
+    {
+      "epoch": 0.8253968253968254,
+      "grad_norm": 1.9339219951657884,
+      "learning_rate": 9.364610386349048e-06,
+      "loss": 0.126,
+      "step": 195
+    },
+    {
+      "epoch": 0.8296296296296296,
+      "grad_norm": 2.0008984944402566,
+      "learning_rate": 9.356069055600949e-06,
+      "loss": 0.1194,
+      "step": 196
+    },
+    {
+      "epoch": 0.8338624338624339,
+      "grad_norm": 2.2860380840573713,
+      "learning_rate": 9.347474647526095e-06,
+      "loss": 0.1311,
+      "step": 197
+    },
+    {
+      "epoch": 0.8380952380952381,
+      "grad_norm": 1.8033278898675713,
+      "learning_rate": 9.338827266844643e-06,
+      "loss": 0.1184,
+      "step": 198
+    },
+    {
+      "epoch": 0.8423280423280424,
+      "grad_norm": 1.9247101184163469,
+      "learning_rate": 9.330127018922195e-06,
+      "loss": 0.1195,
+      "step": 199
+    },
+    {
+      "epoch": 0.8465608465608465,
+      "grad_norm": 2.1205250196763927,
+      "learning_rate": 9.321374009768525e-06,
+      "loss": 0.1193,
+      "step": 200
+    },
+    {
+      "epoch": 0.8507936507936508,
+      "grad_norm": 1.92871242138401,
+      "learning_rate": 9.312568346036288e-06,
+      "loss": 0.1127,
+      "step": 201
+    },
+    {
+      "epoch": 0.855026455026455,
+      "grad_norm": 1.936940428750681,
+      "learning_rate": 9.30371013501972e-06,
+      "loss": 0.1159,
+      "step": 202
+    },
+    {
+      "epoch": 0.8592592592592593,
+      "grad_norm": 1.885933491210588,
+      "learning_rate": 9.294799484653323e-06,
+      "loss": 0.1196,
+      "step": 203
+    },
+    {
+      "epoch": 0.8634920634920635,
+      "grad_norm": 2.082748868345629,
+      "learning_rate": 9.285836503510562e-06,
+      "loss": 0.1136,
+      "step": 204
+    },
+    {
+      "epoch": 0.8677248677248677,
+      "grad_norm": 1.7285347264228283,
+      "learning_rate": 9.276821300802535e-06,
+      "loss": 0.1076,
+      "step": 205
+    },
+    {
+      "epoch": 0.8719576719576719,
+      "grad_norm": 2.1678906514126584,
+      "learning_rate": 9.267753986376638e-06,
+      "loss": 0.1309,
+      "step": 206
+    },
+    {
+      "epoch": 0.8761904761904762,
+      "grad_norm": 2.2303167600680838,
+      "learning_rate": 9.25863467071524e-06,
+      "loss": 0.1188,
+      "step": 207
+    },
+    {
+      "epoch": 0.8804232804232804,
+      "grad_norm": 1.8524486771122908,
+      "learning_rate": 9.24946346493432e-06,
+      "loss": 0.1037,
+      "step": 208
+    },
+    {
+      "epoch": 0.8846560846560847,
+      "grad_norm": 1.9545872165465215,
+      "learning_rate": 9.24024048078213e-06,
+      "loss": 0.1239,
+      "step": 209
+    },
+    {
+      "epoch": 0.8888888888888888,
+      "grad_norm": 2.2585920119228486,
+      "learning_rate": 9.230965830637821e-06,
+      "loss": 0.1426,
+      "step": 210
+    },
+    {
+      "epoch": 0.8931216931216931,
+      "grad_norm": 2.604112442901209,
+      "learning_rate": 9.221639627510076e-06,
+      "loss": 0.1691,
+      "step": 211
+    },
+    {
+      "epoch": 0.8973544973544973,
+      "grad_norm": 2.148863724504725,
+      "learning_rate": 9.21226198503574e-06,
+      "loss": 0.1239,
+      "step": 212
+    },
+    {
+      "epoch": 0.9015873015873016,
+      "grad_norm": 2.1443004360413283,
+      "learning_rate": 9.202833017478421e-06,
+      "loss": 0.1111,
+      "step": 213
+    },
+    {
+      "epoch": 0.9058201058201059,
+      "grad_norm": 2.0065790334640754,
+      "learning_rate": 9.193352839727122e-06,
+      "loss": 0.0986,
+      "step": 214
+    },
+    {
+      "epoch": 0.91005291005291,
+      "grad_norm": 1.685651884041455,
+      "learning_rate": 9.18382156729481e-06,
+      "loss": 0.0962,
+      "step": 215
+    },
+    {
+      "epoch": 0.9142857142857143,
+      "grad_norm": 1.9759626665833376,
+      "learning_rate": 9.174239316317034e-06,
+      "loss": 0.1228,
+      "step": 216
+    },
+    {
+      "epoch": 0.9185185185185185,
+      "grad_norm": 2.0642455639952093,
+      "learning_rate": 9.164606203550498e-06,
+      "loss": 0.1408,
+      "step": 217
+    },
+    {
+      "epoch": 0.9227513227513228,
+      "grad_norm": 2.032852061108819,
+      "learning_rate": 9.154922346371641e-06,
+      "loss": 0.1152,
+      "step": 218
+    },
+    {
+      "epoch": 0.926984126984127,
+      "grad_norm": 2.354712801581503,
+      "learning_rate": 9.145187862775208e-06,
+      "loss": 0.1428,
+      "step": 219
+    },
+    {
+      "epoch": 0.9312169312169312,
+      "grad_norm": 2.5192247867841546,
+      "learning_rate": 9.13540287137281e-06,
+      "loss": 0.1405,
+      "step": 220
+    },
+    {
+      "epoch": 0.9354497354497354,
+      "grad_norm": 1.72459255330544,
+      "learning_rate": 9.125567491391476e-06,
+      "loss": 0.1003,
+      "step": 221
+    },
+    {
+      "epoch": 0.9396825396825397,
+      "grad_norm": 1.8550339147298913,
+      "learning_rate": 9.115681842672211e-06,
+      "loss": 0.1154,
+      "step": 222
+    },
+    {
+      "epoch": 0.9439153439153439,
+      "grad_norm": 1.9003240348247867,
+      "learning_rate": 9.10574604566852e-06,
+      "loss": 0.1181,
+      "step": 223
+    },
+    {
+      "epoch": 0.9481481481481482,
+      "grad_norm": 2.156481954204094,
+      "learning_rate": 9.09576022144496e-06,
+      "loss": 0.1264,
+      "step": 224
+    },
+    {
+      "epoch": 0.9523809523809523,
+      "grad_norm": 1.998485739236388,
+      "learning_rate": 9.085724491675642e-06,
+      "loss": 0.1341,
+      "step": 225
+    },
+    {
+      "epoch": 0.9566137566137566,
+      "grad_norm": 1.822826676452421,
+      "learning_rate": 9.07563897864277e-06,
+      "loss": 0.1399,
+      "step": 226
+    },
+    {
+      "epoch": 0.9608465608465608,
+      "grad_norm": 2.2296945866049365,
+      "learning_rate": 9.065503805235139e-06,
+      "loss": 0.113,
+      "step": 227
+    },
+    {
+      "epoch": 0.9650793650793651,
+      "grad_norm": 2.533173971355714,
+      "learning_rate": 9.055319094946633e-06,
+      "loss": 0.1544,
+      "step": 228
+    },
+    {
+      "epoch": 0.9693121693121693,
+      "grad_norm": 2.189492349413132,
+      "learning_rate": 9.045084971874738e-06,
+      "loss": 0.138,
+      "step": 229
+    },
+    {
+      "epoch": 0.9735449735449735,
+      "grad_norm": 2.0227496012742434,
+      "learning_rate": 9.03480156071901e-06,
+      "loss": 0.1201,
+      "step": 230
+    },
+    {
+      "epoch": 0.9777777777777777,
+      "grad_norm": 2.19493710384594,
+      "learning_rate": 9.02446898677957e-06,
+      "loss": 0.1501,
+      "step": 231
+    },
+    {
+      "epoch": 0.982010582010582,
+      "grad_norm": 1.6357273412380897,
+      "learning_rate": 9.014087375955574e-06,
+      "loss": 0.1024,
+      "step": 232
+    },
+    {
+      "epoch": 0.9862433862433863,
+      "grad_norm": 1.9821532554149048,
+      "learning_rate": 9.003656854743667e-06,
+      "loss": 0.1149,
+      "step": 233
+    },
+    {
+      "epoch": 0.9904761904761905,
+      "grad_norm": 1.739947981961382,
+      "learning_rate": 8.993177550236464e-06,
+      "loss": 0.1128,
+      "step": 234
+    },
+    {
+      "epoch": 0.9947089947089947,
+      "grad_norm": 1.8038704679693724,
+      "learning_rate": 8.982649590120982e-06,
+      "loss": 0.0901,
+      "step": 235
+    },
+    {
+      "epoch": 0.9989417989417989,
+      "grad_norm": 1.755306767490001,
+      "learning_rate": 8.972073102677091e-06,
+      "loss": 0.0941,
+      "step": 236
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.1593265626408567,
+      "learning_rate": 8.961448216775955e-06,
+      "loss": 0.0896,
+      "step": 237
+    },
+    {
+      "epoch": 1.0042328042328041,
+      "grad_norm": 1.6382971000061148,
+      "learning_rate": 8.950775061878453e-06,
+      "loss": 0.0839,
+      "step": 238
+    },
+    {
+      "epoch": 1.0084656084656085,
+      "grad_norm": 1.733233185872628,
+      "learning_rate": 8.94005376803361e-06,
+      "loss": 0.0942,
+      "step": 239
+    },
+    {
+      "epoch": 1.0126984126984127,
+      "grad_norm": 1.6215967926850725,
+      "learning_rate": 8.92928446587701e-06,
+      "loss": 0.0823,
+      "step": 240
+    },
+    {
+      "epoch": 1.016931216931217,
+      "grad_norm": 1.700851779943189,
+      "learning_rate": 8.9184672866292e-06,
+      "loss": 0.0707,
+      "step": 241
+    },
+    {
+      "epoch": 1.0211640211640212,
+      "grad_norm": 1.5763379344310904,
+      "learning_rate": 8.907602362094094e-06,
+      "loss": 0.0709,
+      "step": 242
+    },
+    {
+      "epoch": 1.0253968253968253,
+      "grad_norm": 2.1897050252472425,
+      "learning_rate": 8.896689824657371e-06,
+      "loss": 0.0944,
+      "step": 243
+    },
+    {
+      "epoch": 1.0296296296296297,
+      "grad_norm": 1.7815709759982583,
+      "learning_rate": 8.885729807284855e-06,
+      "loss": 0.0782,
+      "step": 244
+    },
+    {
+      "epoch": 1.0338624338624338,
+      "grad_norm": 1.8077323351970311,
+      "learning_rate": 8.874722443520898e-06,
+      "loss": 0.0652,
+      "step": 245
+    },
+    {
+      "epoch": 1.0380952380952382,
+      "grad_norm": 2.1870219531658295,
+      "learning_rate": 8.863667867486756e-06,
+      "loss": 0.0924,
+      "step": 246
+    },
+    {
+      "epoch": 1.0423280423280423,
+      "grad_norm": 2.023053098789974,
+      "learning_rate": 8.852566213878947e-06,
+      "loss": 0.0879,
+      "step": 247
+    },
+    {
+      "epoch": 1.0465608465608465,
+      "grad_norm": 2.0022724327057855,
+      "learning_rate": 8.841417617967618e-06,
+      "loss": 0.081,
+      "step": 248
+    },
+    {
+      "epoch": 1.0507936507936508,
+      "grad_norm": 1.8561313251630824,
+      "learning_rate": 8.83022221559489e-06,
+      "loss": 0.064,
+      "step": 249
+    },
+    {
+      "epoch": 1.055026455026455,
+      "grad_norm": 2.0472910334735452,
+      "learning_rate": 8.818980143173212e-06,
+      "loss": 0.0654,
+      "step": 250
+    },
+    {
+      "epoch": 1.0592592592592593,
+      "grad_norm": 1.847056303737043,
+      "learning_rate": 8.807691537683685e-06,
+      "loss": 0.0645,
+      "step": 251
+    },
+    {
+      "epoch": 1.0634920634920635,
+      "grad_norm": 2.3192957709806685,
+      "learning_rate": 8.796356536674404e-06,
+      "loss": 0.0595,
+      "step": 252
+    },
+    {
+      "epoch": 1.0677248677248676,
+      "grad_norm": 1.8495895474699158,
+      "learning_rate": 8.784975278258783e-06,
+      "loss": 0.0547,
+      "step": 253
+    },
+    {
+      "epoch": 1.071957671957672,
+      "grad_norm": 2.3472935666710306,
+      "learning_rate": 8.773547901113862e-06,
+      "loss": 0.0955,
+      "step": 254
+    },
+    {
+      "epoch": 1.0761904761904761,
+      "grad_norm": 1.8466358550678363,
+      "learning_rate": 8.762074544478622e-06,
+      "loss": 0.0838,
+      "step": 255
+    },
+    {
+      "epoch": 1.0804232804232805,
+      "grad_norm": 1.8807145283212745,
+      "learning_rate": 8.750555348152299e-06,
+      "loss": 0.084,
+      "step": 256
+    },
+    {
+      "epoch": 1.0846560846560847,
+      "grad_norm": 2.0632365769348913,
+      "learning_rate": 8.73899045249266e-06,
+      "loss": 0.1131,
+      "step": 257
+    },
+    {
+      "epoch": 1.0888888888888888,
+      "grad_norm": 1.7220071689576308,
+      "learning_rate": 8.727379998414311e-06,
+      "loss": 0.0651,
+      "step": 258
+    },
+    {
+      "epoch": 1.0931216931216932,
+      "grad_norm": 1.8875007957585828,
+      "learning_rate": 8.715724127386971e-06,
+      "loss": 0.0912,
+      "step": 259
+    },
+    {
+      "epoch": 1.0973544973544973,
+      "grad_norm": 1.7491126079962518,
+      "learning_rate": 8.70402298143375e-06,
+      "loss": 0.077,
+      "step": 260
+    },
+    {
+      "epoch": 1.1015873015873017,
+      "grad_norm": 1.9113293742259723,
+      "learning_rate": 8.692276703129421e-06,
+      "loss": 0.065,
+      "step": 261
+    },
+    {
+      "epoch": 1.1058201058201058,
+      "grad_norm": 1.972668011890266,
+      "learning_rate": 8.680485435598674e-06,
+      "loss": 0.073,
+      "step": 262
+    },
+    {
+      "epoch": 1.11005291005291,
+      "grad_norm": 2.0401162544608704,
+      "learning_rate": 8.668649322514382e-06,
+      "loss": 0.0813,
+      "step": 263
+    },
+    {
+      "epoch": 1.1142857142857143,
+      "grad_norm": 1.705144782299304,
+      "learning_rate": 8.656768508095853e-06,
+      "loss": 0.0687,
+      "step": 264
+    },
+    {
+      "epoch": 1.1185185185185185,
+      "grad_norm": 1.8673386144289976,
+      "learning_rate": 8.644843137107058e-06,
+      "loss": 0.0736,
+      "step": 265
+    },
+    {
+      "epoch": 1.1227513227513228,
+      "grad_norm": 2.182984594576122,
+      "learning_rate": 8.632873354854881e-06,
+      "loss": 0.1059,
+      "step": 266
+    },
+    {
+      "epoch": 1.126984126984127,
+      "grad_norm": 2.195083172027126,
+      "learning_rate": 8.620859307187339e-06,
+      "loss": 0.0941,
+      "step": 267
+    },
+    {
+      "epoch": 1.1312169312169311,
+      "grad_norm": 1.8505253617934838,
+      "learning_rate": 8.608801140491811e-06,
+      "loss": 0.0835,
+      "step": 268
+    },
+    {
+      "epoch": 1.1354497354497355,
+      "grad_norm": 1.9135866347855481,
+      "learning_rate": 8.596699001693257e-06,
+      "loss": 0.0774,
+      "step": 269
+    },
+    {
+      "epoch": 1.1396825396825396,
+      "grad_norm": 1.954938336450862,
+      "learning_rate": 8.584553038252415e-06,
+      "loss": 0.0855,
+      "step": 270
+    },
+    {
+      "epoch": 1.143915343915344,
+      "grad_norm": 1.6669954564422538,
+      "learning_rate": 8.572363398164017e-06,
+      "loss": 0.0758,
+      "step": 271
+    },
+    {
+      "epoch": 1.1481481481481481,
+      "grad_norm": 1.6351480365128717,
+      "learning_rate": 8.560130229954985e-06,
+      "loss": 0.0683,
+      "step": 272
+    },
+    {
+      "epoch": 1.1523809523809523,
+      "grad_norm": 1.7579143840699374,
+      "learning_rate": 8.547853682682605e-06,
+      "loss": 0.0747,
+      "step": 273
+    },
+    {
+      "epoch": 1.1566137566137566,
+      "grad_norm": 1.5569433244482247,
+      "learning_rate": 8.535533905932739e-06,
+      "loss": 0.0671,
+      "step": 274
+    },
+    {
+      "epoch": 1.1608465608465608,
+      "grad_norm": 1.8671513527781665,
+      "learning_rate": 8.523171049817974e-06,
+      "loss": 0.0737,
+      "step": 275
+    },
+    {
+      "epoch": 1.1650793650793652,
+      "grad_norm": 1.871079630531322,
+      "learning_rate": 8.510765264975813e-06,
+      "loss": 0.067,
+      "step": 276
+    },
+    {
+      "epoch": 1.1693121693121693,
+      "grad_norm": 1.7124717161637384,
+      "learning_rate": 8.498316702566828e-06,
+      "loss": 0.0661,
+      "step": 277
+    },
+    {
+      "epoch": 1.1735449735449737,
+      "grad_norm": 1.5069679809087728,
+      "learning_rate": 8.485825514272824e-06,
+      "loss": 0.0626,
+      "step": 278
+    },
+    {
+      "epoch": 1.1777777777777778,
+      "grad_norm": 1.637366530387125,
+      "learning_rate": 8.473291852294986e-06,
+      "loss": 0.0676,
+      "step": 279
+    },
+    {
+      "epoch": 1.182010582010582,
+      "grad_norm": 1.870549503958005,
+      "learning_rate": 8.460715869352035e-06,
+      "loss": 0.0673,
+      "step": 280
+    },
+    {
+      "epoch": 1.1862433862433863,
+      "grad_norm": 2.6075150985041775,
+      "learning_rate": 8.44809771867835e-06,
+      "loss": 0.1037,
+      "step": 281
+    },
+    {
+      "epoch": 1.1904761904761905,
+      "grad_norm": 1.8129166707802178,
+      "learning_rate": 8.435437554022116e-06,
+      "loss": 0.0658,
+      "step": 282
+    },
+    {
+      "epoch": 1.1947089947089946,
+      "grad_norm": 1.9273840349175055,
+      "learning_rate": 8.422735529643445e-06,
+      "loss": 0.0692,
+      "step": 283
+    },
+    {
+      "epoch": 1.198941798941799,
+      "grad_norm": 1.946526317802552,
+      "learning_rate": 8.409991800312493e-06,
+      "loss": 0.0872,
+      "step": 284
+    },
+    {
+      "epoch": 1.2031746031746031,
+      "grad_norm": 2.093538504659771,
+      "learning_rate": 8.397206521307584e-06,
+      "loss": 0.0575,
+      "step": 285
+    },
+    {
+      "epoch": 1.2074074074074075,
+      "grad_norm": 1.8111794559902639,
+      "learning_rate": 8.384379848413304e-06,
+      "loss": 0.0536,
+      "step": 286
+    },
+    {
+      "epoch": 1.2116402116402116,
+      "grad_norm": 1.4485261428861709,
+      "learning_rate": 8.371511937918616e-06,
+      "loss": 0.0567,
+      "step": 287
+    },
+    {
+      "epoch": 1.215873015873016,
+      "grad_norm": 2.2184787541807047,
+      "learning_rate": 8.358602946614952e-06,
+      "loss": 0.0777,
+      "step": 288
+    },
+    {
+      "epoch": 1.2201058201058201,
+      "grad_norm": 2.019737455684998,
+      "learning_rate": 8.345653031794292e-06,
+      "loss": 0.0679,
+      "step": 289
+    },
+    {
+      "epoch": 1.2243386243386243,
+      "grad_norm": 1.6322133861673878,
+      "learning_rate": 8.332662351247262e-06,
+      "loss": 0.0473,
+      "step": 290
+    },
+    {
+      "epoch": 1.2285714285714286,
+      "grad_norm": 2.2122138272956797,
+      "learning_rate": 8.319631063261209e-06,
+      "loss": 0.0902,
+      "step": 291
+    },
+    {
+      "epoch": 1.2328042328042328,
+      "grad_norm": 1.7651396895281377,
+      "learning_rate": 8.30655932661826e-06,
+      "loss": 0.0732,
+      "step": 292
+    },
+    {
+      "epoch": 1.237037037037037,
+      "grad_norm": 1.8538276922333994,
+      "learning_rate": 8.293447300593402e-06,
+      "loss": 0.0688,
+      "step": 293
+    },
+    {
+      "epoch": 1.2412698412698413,
+      "grad_norm": 2.15194163484418,
+      "learning_rate": 8.280295144952537e-06,
+      "loss": 0.0764,
+      "step": 294
+    },
+    {
+      "epoch": 1.2455026455026454,
+      "grad_norm": 1.7234143791825733,
+      "learning_rate": 8.267103019950529e-06,
+      "loss": 0.0538,
+      "step": 295
+    },
+    {
+      "epoch": 1.2497354497354498,
+      "grad_norm": 2.293390918667569,
+      "learning_rate": 8.253871086329255e-06,
+      "loss": 0.0769,
+      "step": 296
+    },
+    {
+      "epoch": 1.253968253968254,
+      "grad_norm": 1.8852414195646272,
+      "learning_rate": 8.240599505315656e-06,
+      "loss": 0.0584,
+      "step": 297
+    },
+    {
+      "epoch": 1.2582010582010583,
+      "grad_norm": 2.1521764452864973,
+      "learning_rate": 8.227288438619754e-06,
+      "loss": 0.0718,
+      "step": 298
+    },
+    {
+      "epoch": 1.2624338624338625,
+      "grad_norm": 1.3679302426411497,
+      "learning_rate": 8.213938048432697e-06,
+      "loss": 0.0476,
+      "step": 299
+    },
+    {
+      "epoch": 1.2666666666666666,
+      "grad_norm": 1.8759901982931633,
+      "learning_rate": 8.200548497424779e-06,
+      "loss": 0.0588,
+      "step": 300
+    },
+    {
+      "epoch": 1.270899470899471,
+      "grad_norm": 2.1115534695308757,
+      "learning_rate": 8.18711994874345e-06,
+      "loss": 0.0474,
+      "step": 301
+    },
+    {
+      "epoch": 1.2751322751322751,
+      "grad_norm": 1.42054603929946,
+      "learning_rate": 8.173652566011339e-06,
+      "loss": 0.0491,
+      "step": 302
+    },
+    {
+      "epoch": 1.2793650793650793,
+      "grad_norm": 2.136699369797229,
+      "learning_rate": 8.160146513324256e-06,
+      "loss": 0.087,
+      "step": 303
+    },
+    {
+      "epoch": 1.2835978835978836,
+      "grad_norm": 1.7069088981866332,
+      "learning_rate": 8.146601955249187e-06,
+      "loss": 0.0535,
+      "step": 304
+    },
+    {
+      "epoch": 1.2878306878306878,
+      "grad_norm": 2.678199387264105,
+      "learning_rate": 8.133019056822303e-06,
+      "loss": 0.0858,
+      "step": 305
+    },
+    {
+      "epoch": 1.2920634920634921,
+      "grad_norm": 2.400768922479514,
+      "learning_rate": 8.119397983546932e-06,
+      "loss": 0.0838,
+      "step": 306
+    },
+    {
+      "epoch": 1.2962962962962963,
+      "grad_norm": 2.0384750677283283,
+      "learning_rate": 8.105738901391553e-06,
+      "loss": 0.0714,
+      "step": 307
+    },
+    {
+      "epoch": 1.3005291005291006,
+      "grad_norm": 1.8955000216446414,
+      "learning_rate": 8.092041976787772e-06,
+      "loss": 0.0653,
+      "step": 308
+    },
+    {
+      "epoch": 1.3047619047619048,
+      "grad_norm": 1.8920868109479223,
+      "learning_rate": 8.078307376628292e-06,
+      "loss": 0.0693,
+      "step": 309
+    },
+    {
+      "epoch": 1.308994708994709,
+      "grad_norm": 2.0890907162900434,
+      "learning_rate": 8.064535268264883e-06,
+      "loss": 0.0814,
+      "step": 310
+    },
+    {
+      "epoch": 1.3132275132275133,
+      "grad_norm": 1.8374660007465378,
+      "learning_rate": 8.05072581950634e-06,
+      "loss": 0.0697,
+      "step": 311
+    },
+    {
+      "epoch": 1.3174603174603174,
+      "grad_norm": 1.7534716195041071,
+      "learning_rate": 8.036879198616434e-06,
+      "loss": 0.0643,
+      "step": 312
+    },
+    {
+      "epoch": 1.3216931216931216,
+      "grad_norm": 1.9643071053476988,
+      "learning_rate": 8.022995574311876e-06,
+      "loss": 0.0808,
+      "step": 313
+    },
+    {
+      "epoch": 1.325925925925926,
+      "grad_norm": 1.7283546374071002,
+      "learning_rate": 8.009075115760243e-06,
+      "loss": 0.0772,
+      "step": 314
+    },
+    {
+      "epoch": 1.33015873015873,
+      "grad_norm": 1.5108710880067853,
+      "learning_rate": 7.99511799257793e-06,
+      "loss": 0.059,
+      "step": 315
+    },
+    {
+      "epoch": 1.3343915343915345,
+      "grad_norm": 1.4537965690657546,
+      "learning_rate": 7.981124374828079e-06,
+      "loss": 0.0594,
+      "step": 316
+    },
+    {
+      "epoch": 1.3386243386243386,
+      "grad_norm": 2.237037563797428,
+      "learning_rate": 7.967094433018508e-06,
+      "loss": 0.0768,
+      "step": 317
+    },
+    {
+      "epoch": 1.342857142857143,
+      "grad_norm": 1.6057750143434733,
+      "learning_rate": 7.953028338099628e-06,
+      "loss": 0.0696,
+      "step": 318
+    },
+    {
+      "epoch": 1.3470899470899471,
+      "grad_norm": 1.5702160775846645,
+      "learning_rate": 7.938926261462366e-06,
+      "loss": 0.065,
+      "step": 319
+    },
+    {
+      "epoch": 1.3513227513227513,
+      "grad_norm": 1.8059545714263667,
+      "learning_rate": 7.92478837493608e-06,
+      "loss": 0.0751,
+      "step": 320
+    },
+    {
+      "epoch": 1.3555555555555556,
+      "grad_norm": 1.8812566576792862,
+      "learning_rate": 7.910614850786448e-06,
+      "loss": 0.066,
+      "step": 321
+    },
+    {
+      "epoch": 1.3597883597883598,
+      "grad_norm": 1.64104762399391,
+      "learning_rate": 7.896405861713393e-06,
+      "loss": 0.0612,
+      "step": 322
+    },
+    {
+      "epoch": 1.364021164021164,
+      "grad_norm": 1.938404168025234,
+      "learning_rate": 7.882161580848966e-06,
+      "loss": 0.0716,
+      "step": 323
+    },
+    {
+      "epoch": 1.3682539682539683,
+      "grad_norm": 1.9354768854944182,
+      "learning_rate": 7.86788218175523e-06,
+      "loss": 0.0625,
+      "step": 324
+    },
+    {
+      "epoch": 1.3724867724867724,
+      "grad_norm": 1.7263366442747021,
+      "learning_rate": 7.85356783842216e-06,
+      "loss": 0.0641,
+      "step": 325
+    },
+    {
+      "epoch": 1.3767195767195768,
+      "grad_norm": 1.5396345057515881,
+      "learning_rate": 7.839218725265507e-06,
+      "loss": 0.0553,
+      "step": 326
+    },
+    {
+      "epoch": 1.380952380952381,
+      "grad_norm": 1.8525012205903337,
+      "learning_rate": 7.82483501712469e-06,
+      "loss": 0.0766,
+      "step": 327
+    },
+    {
+      "epoch": 1.3851851851851853,
+      "grad_norm": 1.6334096140515466,
+      "learning_rate": 7.810416889260653e-06,
+      "loss": 0.0572,
+      "step": 328
+    },
+    {
+      "epoch": 1.3894179894179894,
+      "grad_norm": 1.5922133675439893,
+      "learning_rate": 7.795964517353734e-06,
+      "loss": 0.0687,
+      "step": 329
+    },
+    {
+      "epoch": 1.3936507936507936,
+      "grad_norm": 1.2091317327665092,
+      "learning_rate": 7.781478077501526e-06,
+      "loss": 0.0424,
+      "step": 330
+    },
+    {
+      "epoch": 1.397883597883598,
+      "grad_norm": 2.132743017629856,
+      "learning_rate": 7.76695774621672e-06,
+      "loss": 0.0617,
+      "step": 331
+    },
+    {
+      "epoch": 1.402116402116402,
+      "grad_norm": 2.178343442230559,
+      "learning_rate": 7.752403700424978e-06,
+      "loss": 0.0767,
+      "step": 332
+    },
+    {
+      "epoch": 1.4063492063492062,
+      "grad_norm": 2.0956999898184905,
+      "learning_rate": 7.737816117462752e-06,
+      "loss": 0.0856,
+      "step": 333
+    },
+    {
+      "epoch": 1.4105820105820106,
+      "grad_norm": 2.328993538503913,
+      "learning_rate": 7.723195175075136e-06,
+      "loss": 0.0697,
+      "step": 334
+    },
+    {
+      "epoch": 1.4148148148148147,
+      "grad_norm": 2.004814063553197,
+      "learning_rate": 7.7085410514137e-06,
+      "loss": 0.0719,
+      "step": 335
+    },
+    {
+      "epoch": 1.4190476190476191,
+      "grad_norm": 1.560361912982766,
+      "learning_rate": 7.693853925034316e-06,
+      "loss": 0.0478,
+      "step": 336
+    },
+    {
+      "epoch": 1.4232804232804233,
+      "grad_norm": 1.6252770277096498,
+      "learning_rate": 7.679133974894984e-06,
+      "loss": 0.0593,
+      "step": 337
+    },
+    {
+      "epoch": 1.4275132275132276,
+      "grad_norm": 1.8015191094916037,
+      "learning_rate": 7.66438138035365e-06,
+      "loss": 0.0651,
+      "step": 338
+    },
+    {
+      "epoch": 1.4317460317460318,
+      "grad_norm": 2.3510718772860475,
+      "learning_rate": 7.649596321166024e-06,
+      "loss": 0.0774,
+      "step": 339
+    },
+    {
+      "epoch": 1.435978835978836,
+      "grad_norm": 2.7202765882488382,
+      "learning_rate": 7.634778977483389e-06,
+      "loss": 0.0861,
+      "step": 340
+    },
+    {
+      "epoch": 1.4402116402116403,
+      "grad_norm": 2.336843036814147,
+      "learning_rate": 7.619929529850397e-06,
+      "loss": 0.0755,
+      "step": 341
+    },
+    {
+      "epoch": 1.4444444444444444,
+      "grad_norm": 1.9373601062542976,
+      "learning_rate": 7.605048159202884e-06,
+      "loss": 0.0633,
+      "step": 342
+    },
+    {
+      "epoch": 1.4486772486772486,
+      "grad_norm": 1.6817059556405092,
+      "learning_rate": 7.590135046865652e-06,
+      "loss": 0.063,
+      "step": 343
+    },
+    {
+      "epoch": 1.452910052910053,
+      "grad_norm": 2.1941319252714404,
+      "learning_rate": 7.575190374550272e-06,
+      "loss": 0.0682,
+      "step": 344
+    },
+    {
+      "epoch": 1.457142857142857,
+      "grad_norm": 1.6374443562647756,
+      "learning_rate": 7.560214324352858e-06,
+      "loss": 0.0566,
+      "step": 345
+    },
+    {
+      "epoch": 1.4613756613756614,
+      "grad_norm": 2.007828522647094,
+      "learning_rate": 7.545207078751858e-06,
+      "loss": 0.0714,
+      "step": 346
+    },
+    {
+      "epoch": 1.4656084656084656,
+      "grad_norm": 1.4907319319460866,
+      "learning_rate": 7.530168820605819e-06,
+      "loss": 0.0528,
+      "step": 347
+    },
+    {
+      "epoch": 1.46984126984127,
+      "grad_norm": 1.7939591139887787,
+      "learning_rate": 7.515099733151177e-06,
+      "loss": 0.0539,
+      "step": 348
+    },
+    {
+      "epoch": 1.474074074074074,
+      "grad_norm": 1.4906209466069649,
+      "learning_rate": 7.500000000000001e-06,
+      "loss": 0.0605,
+      "step": 349
+    },
+    {
+      "epoch": 1.4783068783068782,
+      "grad_norm": 1.8613522048230187,
+      "learning_rate": 7.484869805137778e-06,
+      "loss": 0.0664,
+      "step": 350
+    },
+    {
+      "epoch": 1.4825396825396826,
+      "grad_norm": 1.8229049611749533,
+      "learning_rate": 7.469709332921155e-06,
+      "loss": 0.0568,
+      "step": 351
+    },
+    {
+      "epoch": 1.4867724867724867,
+      "grad_norm": 1.4355304492933953,
+      "learning_rate": 7.454518768075705e-06,
+      "loss": 0.0512,
+      "step": 352
+    },
+    {
+      "epoch": 1.491005291005291,
+      "grad_norm": 2.1617485048622775,
+      "learning_rate": 7.4392982956936644e-06,
+      "loss": 0.0866,
+      "step": 353
+    },
+    {
+      "epoch": 1.4952380952380953,
+      "grad_norm": 1.6436091016901637,
+      "learning_rate": 7.424048101231687e-06,
+      "loss": 0.0569,
+      "step": 354
+    },
+    {
+      "epoch": 1.4994708994708994,
+      "grad_norm": 2.117413284430222,
+      "learning_rate": 7.408768370508577e-06,
+      "loss": 0.073,
+      "step": 355
+    },
+    {
+      "epoch": 1.5037037037037035,
+      "grad_norm": 2.0982337999329355,
+      "learning_rate": 7.393459289703035e-06,
+      "loss": 0.0548,
+      "step": 356
+    },
+    {
+      "epoch": 1.507936507936508,
+      "grad_norm": 1.9819236384785228,
+      "learning_rate": 7.378121045351378e-06,
+      "loss": 0.065,
+      "step": 357
+    },
+    {
+      "epoch": 1.5121693121693123,
+      "grad_norm": 1.9155338730187241,
+      "learning_rate": 7.362753824345271e-06,
+      "loss": 0.0619,
+      "step": 358
+    },
+    {
+      "epoch": 1.5164021164021164,
+      "grad_norm": 1.834898522988967,
+      "learning_rate": 7.347357813929455e-06,
+      "loss": 0.0652,
+      "step": 359
+    },
+    {
+      "epoch": 1.5206349206349206,
+      "grad_norm": 2.2906521765290715,
+      "learning_rate": 7.3319332016994575e-06,
+      "loss": 0.0754,
+      "step": 360
+    },
+    {
+      "epoch": 1.524867724867725,
+      "grad_norm": 2.1765523809119127,
+      "learning_rate": 7.31648017559931e-06,
+      "loss": 0.0692,
+      "step": 361
+    },
+    {
+      "epoch": 1.529100529100529,
+      "grad_norm": 2.000515811428855,
+      "learning_rate": 7.300998923919259e-06,
+      "loss": 0.0611,
+      "step": 362
+    },
+    {
+      "epoch": 1.5333333333333332,
+      "grad_norm": 1.8581129150182212,
+      "learning_rate": 7.285489635293472e-06,
+      "loss": 0.0719,
+      "step": 363
+    },
+    {
+      "epoch": 1.5375661375661376,
+      "grad_norm": 1.9376344291681409,
+      "learning_rate": 7.269952498697734e-06,
+      "loss": 0.062,
+      "step": 364
+    },
+    {
+      "epoch": 1.541798941798942,
+      "grad_norm": 1.5469896978040778,
+      "learning_rate": 7.254387703447154e-06,
+      "loss": 0.0463,
+      "step": 365
+    },
+    {
+      "epoch": 1.5460317460317459,
+      "grad_norm": 1.7399436482112474,
+      "learning_rate": 7.238795439193849e-06,
+      "loss": 0.0636,
+      "step": 366
+    },
+    {
+      "epoch": 1.5502645502645502,
+      "grad_norm": 2.0308279574862516,
+      "learning_rate": 7.223175895924638e-06,
+      "loss": 0.0814,
+      "step": 367
+    },
+    {
+      "epoch": 1.5544973544973546,
+      "grad_norm": 1.771006156983727,
+      "learning_rate": 7.207529263958727e-06,
+      "loss": 0.0707,
+      "step": 368
+    },
+    {
+      "epoch": 1.5587301587301587,
+      "grad_norm": 1.692251815342732,
+      "learning_rate": 7.191855733945388e-06,
+      "loss": 0.0502,
+      "step": 369
+    },
+    {
+      "epoch": 1.5629629629629629,
+      "grad_norm": 1.9394355705804434,
+      "learning_rate": 7.176155496861639e-06,
+      "loss": 0.0665,
+      "step": 370
+    },
+    {
+      "epoch": 1.5671957671957673,
+      "grad_norm": 1.6884842235455637,
+      "learning_rate": 7.160428744009913e-06,
+      "loss": 0.0569,
+      "step": 371
+    },
+    {
+      "epoch": 1.5714285714285714,
+      "grad_norm": 1.7508558255498743,
+      "learning_rate": 7.1446756670157306e-06,
+      "loss": 0.0629,
+      "step": 372
+    },
+    {
+      "epoch": 1.5756613756613755,
+      "grad_norm": 1.3658017585158069,
+      "learning_rate": 7.128896457825364e-06,
+      "loss": 0.0419,
+      "step": 373
+    },
+    {
+      "epoch": 1.57989417989418,
+      "grad_norm": 1.7555663442176712,
+      "learning_rate": 7.113091308703498e-06,
+      "loss": 0.0592,
+      "step": 374
+    },
+    {
+      "epoch": 1.5841269841269843,
+      "grad_norm": 1.631900320996386,
+      "learning_rate": 7.0972604122308865e-06,
+      "loss": 0.0635,
+      "step": 375
+    },
+    {
+      "epoch": 1.5883597883597882,
+      "grad_norm": 1.7741916871006203,
+      "learning_rate": 7.081403961302007e-06,
+      "loss": 0.0633,
+      "step": 376
+    },
+    {
+      "epoch": 1.5925925925925926,
+      "grad_norm": 1.8151890739104601,
+      "learning_rate": 7.06552214912271e-06,
+      "loss": 0.0671,
+      "step": 377
+    },
+    {
+      "epoch": 1.596825396825397,
+      "grad_norm": 1.797004370573756,
+      "learning_rate": 7.049615169207864e-06,
+      "loss": 0.0588,
+      "step": 378
+    },
+    {
+      "epoch": 1.601058201058201,
+      "grad_norm": 2.3631789146103084,
+      "learning_rate": 7.033683215379002e-06,
+      "loss": 0.0752,
+      "step": 379
+    },
+    {
+      "epoch": 1.6052910052910052,
+      "grad_norm": 1.763180877149612,
+      "learning_rate": 7.0177264817619514e-06,
+      "loss": 0.0557,
+      "step": 380
+    },
+    {
+      "epoch": 1.6095238095238096,
+      "grad_norm": 1.8995850867394273,
+      "learning_rate": 7.0017451627844765e-06,
+      "loss": 0.0717,
+      "step": 381
+    },
+    {
+      "epoch": 1.6137566137566137,
+      "grad_norm": 1.801085177997209,
+      "learning_rate": 6.985739453173903e-06,
+      "loss": 0.0592,
+      "step": 382
+    },
+    {
+      "epoch": 1.6179894179894179,
+      "grad_norm": 1.7879752968994786,
+      "learning_rate": 6.9697095479547564e-06,
+      "loss": 0.0585,
+      "step": 383
+    },
+    {
+      "epoch": 1.6222222222222222,
+      "grad_norm": 1.6567664297191056,
+      "learning_rate": 6.953655642446368e-06,
+      "loss": 0.0457,
+      "step": 384
+    },
+    {
+      "epoch": 1.6264550264550266,
+      "grad_norm": 1.6155271306262087,
+      "learning_rate": 6.9375779322605154e-06,
+      "loss": 0.059,
+      "step": 385
+    },
+    {
+      "epoch": 1.6306878306878307,
+      "grad_norm": 1.68695010471739,
+      "learning_rate": 6.921476613299018e-06,
+      "loss": 0.0542,
+      "step": 386
+    },
+    {
+      "epoch": 1.6349206349206349,
+      "grad_norm": 1.9421206046284432,
+      "learning_rate": 6.905351881751372e-06,
+      "loss": 0.0717,
+      "step": 387
+    },
+    {
+      "epoch": 1.6391534391534393,
+      "grad_norm": 2.448162982815167,
+      "learning_rate": 6.889203934092337e-06,
+      "loss": 0.0742,
+      "step": 388
+    },
+    {
+      "epoch": 1.6433862433862434,
+      "grad_norm": 1.9166726296413361,
+      "learning_rate": 6.873032967079562e-06,
+      "loss": 0.0581,
+      "step": 389
+    },
+    {
+      "epoch": 1.6476190476190475,
+      "grad_norm": 1.8239195781910584,
+      "learning_rate": 6.856839177751175e-06,
+      "loss": 0.0656,
+      "step": 390
+    },
+    {
+      "epoch": 1.651851851851852,
+      "grad_norm": 1.3729431544806852,
+      "learning_rate": 6.840622763423391e-06,
+      "loss": 0.0402,
+      "step": 391
+    },
+    {
+      "epoch": 1.656084656084656,
+      "grad_norm": 1.7457900027219597,
+      "learning_rate": 6.824383921688098e-06,
+      "loss": 0.0581,
+      "step": 392
+    },
+    {
+      "epoch": 1.6603174603174602,
+      "grad_norm": 1.4041185285984514,
+      "learning_rate": 6.808122850410461e-06,
+      "loss": 0.0445,
+      "step": 393
+    },
+    {
+      "epoch": 1.6645502645502646,
+      "grad_norm": 2.05451457672099,
+      "learning_rate": 6.7918397477265e-06,
+      "loss": 0.071,
+      "step": 394
+    },
+    {
+      "epoch": 1.668783068783069,
+      "grad_norm": 1.6512575262339615,
+      "learning_rate": 6.775534812040686e-06,
+      "loss": 0.0553,
+      "step": 395
+    },
+    {
+      "epoch": 1.673015873015873,
+      "grad_norm": 1.8115387096252427,
+      "learning_rate": 6.759208242023509e-06,
+      "loss": 0.0576,
+      "step": 396
+    },
+    {
+      "epoch": 1.6772486772486772,
+      "grad_norm": 1.2702269626974747,
+      "learning_rate": 6.7428602366090764e-06,
+      "loss": 0.0381,
+      "step": 397
+    },
+    {
+      "epoch": 1.6814814814814816,
+      "grad_norm": 1.5144029396774048,
+      "learning_rate": 6.7264909949926735e-06,
+      "loss": 0.0501,
+      "step": 398
+    },
+    {
+      "epoch": 1.6857142857142857,
+      "grad_norm": 1.4102482360717825,
+      "learning_rate": 6.710100716628345e-06,
+      "loss": 0.0409,
+      "step": 399
+    },
+    {
+      "epoch": 1.6899470899470899,
+      "grad_norm": 2.1685643870197486,
+      "learning_rate": 6.693689601226458e-06,
+      "loss": 0.0737,
+      "step": 400
+    },
+    {
+      "epoch": 1.6941798941798942,
+      "grad_norm": 1.900913941941258,
+      "learning_rate": 6.677257848751276e-06,
+      "loss": 0.0688,
+      "step": 401
+    },
+    {
+      "epoch": 1.6984126984126984,
+      "grad_norm": 1.6471464997248573,
+      "learning_rate": 6.6608056594185166e-06,
+      "loss": 0.0491,
+      "step": 402
+    },
+    {
+      "epoch": 1.7026455026455025,
+      "grad_norm": 1.7754346406800103,
+      "learning_rate": 6.644333233692917e-06,
+      "loss": 0.0559,
+      "step": 403
+    },
+    {
+      "epoch": 1.7068783068783069,
+      "grad_norm": 2.201736043430736,
+      "learning_rate": 6.627840772285784e-06,
+      "loss": 0.0768,
+      "step": 404
+    },
+    {
+      "epoch": 1.7111111111111112,
+      "grad_norm": 1.728174746623607,
+      "learning_rate": 6.611328476152557e-06,
+      "loss": 0.0544,
+      "step": 405
+    },
+    {
+      "epoch": 1.7153439153439154,
+      "grad_norm": 1.653359901954905,
+      "learning_rate": 6.594796546490351e-06,
+      "loss": 0.0516,
+      "step": 406
+    },
+    {
+      "epoch": 1.7195767195767195,
+      "grad_norm": 2.0915084346199113,
+      "learning_rate": 6.578245184735513e-06,
+      "loss": 0.0677,
+      "step": 407
+    },
+    {
+      "epoch": 1.723809523809524,
+      "grad_norm": 1.3338655011947615,
+      "learning_rate": 6.561674592561164e-06,
+      "loss": 0.0452,
+      "step": 408
+    },
+    {
+      "epoch": 1.728042328042328,
+      "grad_norm": 1.916548372288022,
+      "learning_rate": 6.545084971874738e-06,
+      "loss": 0.0511,
+      "step": 409
+    },
+    {
+      "epoch": 1.7322751322751322,
+      "grad_norm": 1.4723543528268832,
+      "learning_rate": 6.5284765248155295e-06,
+      "loss": 0.0427,
+      "step": 410
+    },
+    {
+      "epoch": 1.7365079365079366,
+      "grad_norm": 1.7767150430559346,
+      "learning_rate": 6.5118494537522235e-06,
+      "loss": 0.0549,
+      "step": 411
+    },
+    {
+      "epoch": 1.7407407407407407,
+      "grad_norm": 2.073520060997652,
+      "learning_rate": 6.495203961280434e-06,
+      "loss": 0.0705,
+      "step": 412
+    },
+    {
+      "epoch": 1.7449735449735448,
+      "grad_norm": 1.9164477972239198,
+      "learning_rate": 6.4785402502202345e-06,
+      "loss": 0.0467,
+      "step": 413
+    },
+    {
+      "epoch": 1.7492063492063492,
+      "grad_norm": 1.4479847528255574,
+      "learning_rate": 6.461858523613684e-06,
+      "loss": 0.0411,
+      "step": 414
+    },
+    {
+      "epoch": 1.7534391534391536,
+      "grad_norm": 1.6478763296120762,
+      "learning_rate": 6.445158984722358e-06,
+      "loss": 0.0506,
+      "step": 415
+    },
+    {
+      "epoch": 1.7576719576719577,
+      "grad_norm": 1.7693910067270657,
+      "learning_rate": 6.428441837024868e-06,
+      "loss": 0.0453,
+      "step": 416
+    },
+    {
+      "epoch": 1.7619047619047619,
+      "grad_norm": 1.3800910764927643,
+      "learning_rate": 6.411707284214384e-06,
+      "loss": 0.0351,
+      "step": 417
+    },
+    {
+      "epoch": 1.7661375661375662,
+      "grad_norm": 1.7577530047799417,
+      "learning_rate": 6.3949555301961474e-06,
+      "loss": 0.0661,
+      "step": 418
+    },
+    {
+      "epoch": 1.7703703703703704,
+      "grad_norm": 1.8222505148681252,
+      "learning_rate": 6.378186779084996e-06,
+      "loss": 0.0628,
+      "step": 419
+    },
+    {
+      "epoch": 1.7746031746031745,
+      "grad_norm": 2.0358038447360047,
+      "learning_rate": 6.361401235202872e-06,
+      "loss": 0.0507,
+      "step": 420
+    },
+    {
+      "epoch": 1.7788359788359789,
+      "grad_norm": 2.1606830531676495,
+      "learning_rate": 6.344599103076329e-06,
+      "loss": 0.071,
+      "step": 421
+    },
+    {
+      "epoch": 1.783068783068783,
+      "grad_norm": 1.579905383421796,
+      "learning_rate": 6.327780587434045e-06,
+      "loss": 0.0464,
+      "step": 422
+    },
+    {
+      "epoch": 1.7873015873015872,
+      "grad_norm": 2.361618973258393,
+      "learning_rate": 6.310945893204324e-06,
+      "loss": 0.0559,
+      "step": 423
+    },
+    {
+      "epoch": 1.7915343915343915,
+      "grad_norm": 2.4256436411132274,
+      "learning_rate": 6.294095225512604e-06,
+      "loss": 0.0703,
+      "step": 424
+    },
+    {
+      "epoch": 1.795767195767196,
+      "grad_norm": 2.0483972204305405,
+      "learning_rate": 6.277228789678953e-06,
+      "loss": 0.076,
+      "step": 425
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 1.6486424050280866,
+      "learning_rate": 6.26034679121557e-06,
+      "loss": 0.0354,
+      "step": 426
+    },
+    {
+      "epoch": 1.8042328042328042,
+      "grad_norm": 1.7057849786251837,
+      "learning_rate": 6.243449435824276e-06,
+      "loss": 0.0459,
+      "step": 427
+    },
+    {
+      "epoch": 1.8084656084656086,
+      "grad_norm": 1.4964739470323853,
+      "learning_rate": 6.2265369293940135e-06,
+      "loss": 0.0468,
+      "step": 428
+    },
+    {
+      "epoch": 1.8126984126984127,
+      "grad_norm": 1.5712916952431084,
+      "learning_rate": 6.209609477998339e-06,
+      "loss": 0.0511,
+      "step": 429
+    },
+    {
+      "epoch": 1.8169312169312168,
+      "grad_norm": 1.6075446649270715,
+      "learning_rate": 6.192667287892905e-06,
+      "loss": 0.0488,
+      "step": 430
+    },
+    {
+      "epoch": 1.8211640211640212,
+      "grad_norm": 1.766038600491867,
+      "learning_rate": 6.17571056551295e-06,
+      "loss": 0.0549,
+      "step": 431
+    },
+    {
+      "epoch": 1.8253968253968254,
+      "grad_norm": 1.438358120751698,
+      "learning_rate": 6.158739517470786e-06,
+      "loss": 0.042,
+      "step": 432
+    },
+    {
+      "epoch": 1.8296296296296295,
+      "grad_norm": 1.855618360148089,
+      "learning_rate": 6.141754350553279e-06,
+      "loss": 0.0636,
+      "step": 433
+    },
+    {
+      "epoch": 1.8338624338624339,
+      "grad_norm": 1.6267674194949144,
+      "learning_rate": 6.124755271719326e-06,
+      "loss": 0.0565,
+      "step": 434
+    },
+    {
+      "epoch": 1.8380952380952382,
+      "grad_norm": 1.957873438669131,
+      "learning_rate": 6.107742488097338e-06,
+      "loss": 0.0589,
+      "step": 435
+    },
+    {
+      "epoch": 1.8423280423280424,
+      "grad_norm": 1.8169340288404146,
+      "learning_rate": 6.090716206982714e-06,
+      "loss": 0.0619,
+      "step": 436
+    },
+    {
+      "epoch": 1.8465608465608465,
+      "grad_norm": 1.9473267362173434,
+      "learning_rate": 6.073676635835317e-06,
+      "loss": 0.054,
+      "step": 437
+    },
+    {
+      "epoch": 1.8507936507936509,
+      "grad_norm": 1.6984158293128067,
+      "learning_rate": 6.056623982276945e-06,
+      "loss": 0.043,
+      "step": 438
+    },
+    {
+      "epoch": 1.855026455026455,
+      "grad_norm": 1.7978129108415464,
+      "learning_rate": 6.039558454088796e-06,
+      "loss": 0.0536,
+      "step": 439
+    },
+    {
+      "epoch": 1.8592592592592592,
+      "grad_norm": 1.4388535107743592,
+      "learning_rate": 6.022480259208951e-06,
+      "loss": 0.0461,
+      "step": 440
+    },
+    {
+      "epoch": 1.8634920634920635,
+      "grad_norm": 1.9385002527687403,
+      "learning_rate": 6.005389605729824e-06,
+      "loss": 0.0606,
+      "step": 441
+    },
+    {
+      "epoch": 1.8677248677248677,
+      "grad_norm": 1.715710356385189,
+      "learning_rate": 5.988286701895631e-06,
+      "loss": 0.0505,
+      "step": 442
+    },
+    {
+      "epoch": 1.8719576719576718,
+      "grad_norm": 1.9893593627181096,
+      "learning_rate": 5.97117175609986e-06,
+      "loss": 0.0491,
+      "step": 443
+    },
+    {
+      "epoch": 1.8761904761904762,
+      "grad_norm": 1.669051511290126,
+      "learning_rate": 5.954044976882725e-06,
+      "loss": 0.0443,
+      "step": 444
+    },
+    {
+      "epoch": 1.8804232804232806,
+      "grad_norm": 1.8334555873159546,
+      "learning_rate": 5.936906572928625e-06,
+      "loss": 0.0595,
+      "step": 445
+    },
+    {
+      "epoch": 1.8846560846560847,
+      "grad_norm": 1.482799215009145,
+      "learning_rate": 5.919756753063601e-06,
+      "loss": 0.0419,
+      "step": 446
+    },
+    {
+      "epoch": 1.8888888888888888,
+      "grad_norm": 2.158817135873301,
+      "learning_rate": 5.902595726252801e-06,
+      "loss": 0.0523,
+      "step": 447
+    },
+    {
+      "epoch": 1.8931216931216932,
+      "grad_norm": 1.4512278026518133,
+      "learning_rate": 5.885423701597918e-06,
+      "loss": 0.0341,
+      "step": 448
+    },
+    {
+      "epoch": 1.8973544973544973,
+      "grad_norm": 2.055857253134802,
+      "learning_rate": 5.8682408883346535e-06,
+      "loss": 0.0678,
+      "step": 449
+    },
+    {
+      "epoch": 1.9015873015873015,
+      "grad_norm": 1.4630518070623408,
+      "learning_rate": 5.851047495830163e-06,
+      "loss": 0.0428,
+      "step": 450
+    },
+    {
+      "epoch": 1.9058201058201059,
+      "grad_norm": 1.5001384600677683,
+      "learning_rate": 5.8338437335805124e-06,
+      "loss": 0.0487,
+      "step": 451
+    },
+    {
+      "epoch": 1.91005291005291,
+      "grad_norm": 1.5957812640910574,
+      "learning_rate": 5.816629811208112e-06,
+      "loss": 0.0454,
+      "step": 452
+    },
+    {
+      "epoch": 1.9142857142857141,
+      "grad_norm": 1.7392417324751044,
+      "learning_rate": 5.799405938459175e-06,
+      "loss": 0.0545,
+      "step": 453
+    },
+    {
+      "epoch": 1.9185185185185185,
+      "grad_norm": 1.2059754368705733,
+      "learning_rate": 5.782172325201155e-06,
+      "loss": 0.0332,
+      "step": 454
+    },
+    {
+      "epoch": 1.9227513227513229,
+      "grad_norm": 2.1573344543845696,
+      "learning_rate": 5.764929181420191e-06,
+      "loss": 0.0504,
+      "step": 455
+    },
+    {
+      "epoch": 1.926984126984127,
+      "grad_norm": 1.7501207563945467,
+      "learning_rate": 5.747676717218549e-06,
+      "loss": 0.0523,
+      "step": 456
+    },
+    {
+      "epoch": 1.9312169312169312,
+      "grad_norm": 1.4778971571819108,
+      "learning_rate": 5.730415142812059e-06,
+      "loss": 0.0374,
+      "step": 457
+    },
+    {
+      "epoch": 1.9354497354497355,
+      "grad_norm": 1.9397580291264296,
+      "learning_rate": 5.7131446685275595e-06,
+      "loss": 0.0483,
+      "step": 458
+    },
+    {
+      "epoch": 1.9396825396825397,
+      "grad_norm": 1.5774150951682633,
+      "learning_rate": 5.695865504800328e-06,
+      "loss": 0.0427,
+      "step": 459
+    },
+    {
+      "epoch": 1.9439153439153438,
+      "grad_norm": 1.5205801074165397,
+      "learning_rate": 5.678577862171523e-06,
+      "loss": 0.0403,
+      "step": 460
+    },
+    {
+      "epoch": 1.9481481481481482,
+      "grad_norm": 1.9814309441969933,
+      "learning_rate": 5.661281951285613e-06,
+      "loss": 0.0546,
+      "step": 461
+    },
+    {
+      "epoch": 1.9523809523809523,
+      "grad_norm": 1.8199362216465413,
+      "learning_rate": 5.643977982887815e-06,
+      "loss": 0.0556,
+      "step": 462
+    },
+    {
+      "epoch": 1.9566137566137565,
+      "grad_norm": 2.403211392183965,
+      "learning_rate": 5.626666167821522e-06,
+      "loss": 0.063,
+      "step": 463
+    },
+    {
+      "epoch": 1.9608465608465608,
+      "grad_norm": 1.3278174443006239,
+      "learning_rate": 5.609346717025738e-06,
+      "loss": 0.0412,
+      "step": 464
+    },
+    {
+      "epoch": 1.9650793650793652,
+      "grad_norm": 1.3097142822228083,
+      "learning_rate": 5.592019841532507e-06,
+      "loss": 0.038,
+      "step": 465
+    },
+    {
+      "epoch": 1.9693121693121693,
+      "grad_norm": 1.3209428741356222,
+      "learning_rate": 5.5746857524643335e-06,
+      "loss": 0.0321,
+      "step": 466
+    },
+    {
+      "epoch": 1.9735449735449735,
+      "grad_norm": 1.4633993153900189,
+      "learning_rate": 5.557344661031628e-06,
+      "loss": 0.0439,
+      "step": 467
+    },
+    {
+      "epoch": 1.9777777777777779,
+      "grad_norm": 1.6791239992400115,
+      "learning_rate": 5.539996778530114e-06,
+      "loss": 0.0496,
+      "step": 468
+    },
+    {
+      "epoch": 1.982010582010582,
+      "grad_norm": 1.6259211954562258,
+      "learning_rate": 5.522642316338268e-06,
+      "loss": 0.0361,
+      "step": 469
+    },
+    {
+      "epoch": 1.9862433862433861,
+      "grad_norm": 1.6806163306838742,
+      "learning_rate": 5.505281485914732e-06,
+      "loss": 0.0419,
+      "step": 470
+    },
+    {
+      "epoch": 1.9904761904761905,
+      "grad_norm": 1.9595273597942116,
+      "learning_rate": 5.487914498795748e-06,
+      "loss": 0.0555,
+      "step": 471
+    },
+    {
+      "epoch": 1.9947089947089947,
+      "grad_norm": 1.8422091219424643,
+      "learning_rate": 5.470541566592573e-06,
+      "loss": 0.0466,
+      "step": 472
+    },
+    {
+      "epoch": 1.9989417989417988,
+      "grad_norm": 1.7500818137119094,
+      "learning_rate": 5.453162900988902e-06,
+      "loss": 0.0438,
+      "step": 473
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 2.761502971744513,
+      "learning_rate": 5.435778713738292e-06,
+      "loss": 0.0405,
+      "step": 474
+    },
+    {
+      "epoch": 2.0042328042328044,
+      "grad_norm": 1.425482793293008,
+      "learning_rate": 5.41838921666158e-06,
+      "loss": 0.0281,
+      "step": 475
+    },
+    {
+      "epoch": 2.0084656084656083,
+      "grad_norm": 2.068969474891088,
+      "learning_rate": 5.400994621644294e-06,
+      "loss": 0.0328,
+      "step": 476
+    },
+    {
+      "epoch": 2.0126984126984127,
+      "grad_norm": 1.3694895890619094,
+      "learning_rate": 5.383595140634093e-06,
+      "loss": 0.0267,
+      "step": 477
+    },
+    {
+      "epoch": 2.016931216931217,
+      "grad_norm": 0.9597151794646276,
+      "learning_rate": 5.366190985638159e-06,
+      "loss": 0.0239,
+      "step": 478
+    },
+    {
+      "epoch": 2.0211640211640214,
+      "grad_norm": 0.9612439901820802,
+      "learning_rate": 5.348782368720627e-06,
+      "loss": 0.0195,
+      "step": 479
+    },
+    {
+      "epoch": 2.0253968253968253,
+      "grad_norm": 1.5229336713881063,
+      "learning_rate": 5.3313695020000026e-06,
+      "loss": 0.0286,
+      "step": 480
+    },
+    {
+      "epoch": 2.0296296296296297,
+      "grad_norm": 1.6211741677941989,
+      "learning_rate": 5.3139525976465675e-06,
+      "loss": 0.0297,
+      "step": 481
+    },
+    {
+      "epoch": 2.033862433862434,
+      "grad_norm": 1.4938054925255841,
+      "learning_rate": 5.296531867879809e-06,
+      "loss": 0.0246,
+      "step": 482
+    },
+    {
+      "epoch": 2.038095238095238,
+      "grad_norm": 1.4029085405633699,
+      "learning_rate": 5.27910752496582e-06,
+      "loss": 0.0225,
+      "step": 483
+    },
+    {
+      "epoch": 2.0423280423280423,
+      "grad_norm": 1.3430049638024157,
+      "learning_rate": 5.2616797812147205e-06,
+      "loss": 0.0208,
+      "step": 484
+    },
+    {
+      "epoch": 2.0465608465608467,
+      "grad_norm": 1.2721185303171878,
+      "learning_rate": 5.244248848978067e-06,
+      "loss": 0.0232,
+      "step": 485
+    },
+    {
+      "epoch": 2.0507936507936506,
+      "grad_norm": 1.463578245334908,
+      "learning_rate": 5.226814940646268e-06,
+      "loss": 0.029,
+      "step": 486
+    },
+    {
+      "epoch": 2.055026455026455,
+      "grad_norm": 1.702815967014725,
+      "learning_rate": 5.209378268645998e-06,
+      "loss": 0.0323,
+      "step": 487
+    },
+    {
+      "epoch": 2.0592592592592593,
+      "grad_norm": 1.26455994835701,
+      "learning_rate": 5.1919390454376e-06,
+      "loss": 0.0223,
+      "step": 488
+    },
+    {
+      "epoch": 2.0634920634920633,
+      "grad_norm": 1.110889327620989,
+      "learning_rate": 5.174497483512506e-06,
+      "loss": 0.0265,
+      "step": 489
+    },
+    {
+      "epoch": 2.0677248677248676,
+      "grad_norm": 1.2207931374567949,
+      "learning_rate": 5.157053795390642e-06,
+      "loss": 0.0226,
+      "step": 490
+    },
+    {
+      "epoch": 2.071957671957672,
+      "grad_norm": 1.0427216738524305,
+      "learning_rate": 5.139608193617846e-06,
+      "loss": 0.0175,
+      "step": 491
+    },
+    {
+      "epoch": 2.0761904761904764,
+      "grad_norm": 1.277813572552022,
+      "learning_rate": 5.1221608907632665e-06,
+      "loss": 0.0207,
+      "step": 492
+    },
+    {
+      "epoch": 2.0804232804232803,
+      "grad_norm": 1.4478523585989624,
+      "learning_rate": 5.1047120994167855e-06,
+      "loss": 0.0205,
+      "step": 493
+    },
+    {
+      "epoch": 2.0846560846560847,
+      "grad_norm": 0.9255571718220782,
+      "learning_rate": 5.087262032186418e-06,
+      "loss": 0.0176,
+      "step": 494
+    },
+    {
+      "epoch": 2.088888888888889,
+      "grad_norm": 1.321686704356327,
+      "learning_rate": 5.069810901695727e-06,
+      "loss": 0.0214,
+      "step": 495
+    },
+    {
+      "epoch": 2.093121693121693,
+      "grad_norm": 1.7093390536661062,
+      "learning_rate": 5.05235892058123e-06,
+      "loss": 0.0251,
+      "step": 496
+    },
+    {
+      "epoch": 2.0973544973544973,
+      "grad_norm": 1.8903454646946267,
+      "learning_rate": 5.034906301489808e-06,
+      "loss": 0.0328,
+      "step": 497
+    },
+    {
+      "epoch": 2.1015873015873017,
+      "grad_norm": 1.622033237548866,
+      "learning_rate": 5.0174532570761194e-06,
+      "loss": 0.0222,
+      "step": 498
+    },
+    {
+      "epoch": 2.105820105820106,
+      "grad_norm": 1.3898704763132579,
+      "learning_rate": 5e-06,
+      "loss": 0.0213,
+      "step": 499
+    },
+    {
+      "epoch": 2.11005291005291,
+      "grad_norm": 1.7802781541406807,
+      "learning_rate": 4.982546742923883e-06,
+      "loss": 0.0351,
+      "step": 500
+    },
+    {
+      "epoch": 2.1142857142857143,
+      "grad_norm": 1.3745676765685821,
+      "learning_rate": 4.965093698510192e-06,
+      "loss": 0.0229,
+      "step": 501
+    },
+    {
+      "epoch": 2.1185185185185187,
+      "grad_norm": 1.2480925927636604,
+      "learning_rate": 4.9476410794187726e-06,
+      "loss": 0.0224,
+      "step": 502
+    },
+    {
+      "epoch": 2.1227513227513226,
+      "grad_norm": 1.328383776404079,
+      "learning_rate": 4.9301890983042744e-06,
+      "loss": 0.021,
+      "step": 503
+    },
+    {
+      "epoch": 2.126984126984127,
+      "grad_norm": 1.2744683969863004,
+      "learning_rate": 4.9127379678135825e-06,
+      "loss": 0.0195,
+      "step": 504
+    },
+    {
+      "epoch": 2.1312169312169313,
+      "grad_norm": 1.5077037111075677,
+      "learning_rate": 4.895287900583216e-06,
+      "loss": 0.0203,
+      "step": 505
+    },
+    {
+      "epoch": 2.1354497354497353,
+      "grad_norm": 1.505975774545219,
+      "learning_rate": 4.877839109236735e-06,
+      "loss": 0.0272,
+      "step": 506
+    },
+    {
+      "epoch": 2.1396825396825396,
+      "grad_norm": 1.923885296333145,
+      "learning_rate": 4.860391806382157e-06,
+      "loss": 0.0328,
+      "step": 507
+    },
+    {
+      "epoch": 2.143915343915344,
+      "grad_norm": 1.1003948235042609,
+      "learning_rate": 4.842946204609359e-06,
+      "loss": 0.0205,
+      "step": 508
+    },
+    {
+      "epoch": 2.148148148148148,
+      "grad_norm": 1.5711105962165373,
+      "learning_rate": 4.825502516487497e-06,
+      "loss": 0.0239,
+      "step": 509
+    },
+    {
+      "epoch": 2.1523809523809523,
+      "grad_norm": 2.723763362350314,
+      "learning_rate": 4.8080609545624004e-06,
+      "loss": 0.0277,
+      "step": 510
+    },
+    {
+      "epoch": 2.1566137566137566,
+      "grad_norm": 1.1294689243975755,
+      "learning_rate": 4.7906217313540035e-06,
+      "loss": 0.025,
+      "step": 511
+    },
+    {
+      "epoch": 2.160846560846561,
+      "grad_norm": 1.2353799701389105,
+      "learning_rate": 4.7731850593537316e-06,
+      "loss": 0.0225,
+      "step": 512
+    },
+    {
+      "epoch": 2.165079365079365,
+      "grad_norm": 0.9875825016071724,
+      "learning_rate": 4.755751151021934e-06,
+      "loss": 0.0155,
+      "step": 513
+    },
+    {
+      "epoch": 2.1693121693121693,
+      "grad_norm": 1.058383517339813,
+      "learning_rate": 4.738320218785281e-06,
+      "loss": 0.0227,
+      "step": 514
+    },
+    {
+      "epoch": 2.1735449735449737,
+      "grad_norm": 1.3182471200484542,
+      "learning_rate": 4.720892475034181e-06,
+      "loss": 0.0205,
+      "step": 515
+    },
+    {
+      "epoch": 2.1777777777777776,
+      "grad_norm": 1.091377381304472,
+      "learning_rate": 4.703468132120193e-06,
+      "loss": 0.0248,
+      "step": 516
+    },
+    {
+      "epoch": 2.182010582010582,
+      "grad_norm": 1.505329948005569,
+      "learning_rate": 4.686047402353433e-06,
+      "loss": 0.0291,
+      "step": 517
+    },
+    {
+      "epoch": 2.1862433862433863,
+      "grad_norm": 1.0420378868483025,
+      "learning_rate": 4.668630498000001e-06,
+      "loss": 0.0178,
+      "step": 518
+    },
+    {
+      "epoch": 2.1904761904761907,
+      "grad_norm": 1.0770607200654685,
+      "learning_rate": 4.651217631279374e-06,
+      "loss": 0.0183,
+      "step": 519
+    },
+    {
+      "epoch": 2.1947089947089946,
+      "grad_norm": 1.0753499999979097,
+      "learning_rate": 4.6338090143618435e-06,
+      "loss": 0.0237,
+      "step": 520
+    },
+    {
+      "epoch": 2.198941798941799,
+      "grad_norm": 1.2102511948406212,
+      "learning_rate": 4.6164048593659076e-06,
+      "loss": 0.0212,
+      "step": 521
+    },
+    {
+      "epoch": 2.2031746031746033,
+      "grad_norm": 1.6612931544402387,
+      "learning_rate": 4.5990053783557066e-06,
+      "loss": 0.0306,
+      "step": 522
+    },
+    {
+      "epoch": 2.2074074074074073,
+      "grad_norm": 1.434922916344193,
+      "learning_rate": 4.581610783338424e-06,
+      "loss": 0.0173,
+      "step": 523
+    },
+    {
+      "epoch": 2.2116402116402116,
+      "grad_norm": 1.318046293179565,
+      "learning_rate": 4.564221286261709e-06,
+      "loss": 0.0188,
+      "step": 524
+    },
+    {
+      "epoch": 2.215873015873016,
+      "grad_norm": 1.5523824099307213,
+      "learning_rate": 4.546837099011101e-06,
+      "loss": 0.0279,
+      "step": 525
+    },
+    {
+      "epoch": 2.22010582010582,
+      "grad_norm": 1.265493250306095,
+      "learning_rate": 4.529458433407429e-06,
+      "loss": 0.0264,
+      "step": 526
+    },
+    {
+      "epoch": 2.2243386243386243,
+      "grad_norm": 1.4757784411924582,
+      "learning_rate": 4.512085501204254e-06,
+      "loss": 0.0264,
+      "step": 527
+    },
+    {
+      "epoch": 2.2285714285714286,
+      "grad_norm": 1.1688272093442453,
+      "learning_rate": 4.494718514085269e-06,
+      "loss": 0.0189,
+      "step": 528
+    },
+    {
+      "epoch": 2.2328042328042326,
+      "grad_norm": 1.4981131427762684,
+      "learning_rate": 4.477357683661734e-06,
+      "loss": 0.0238,
+      "step": 529
+    },
+    {
+      "epoch": 2.237037037037037,
+      "grad_norm": 1.060141826721873,
+      "learning_rate": 4.460003221469886e-06,
+      "loss": 0.0202,
+      "step": 530
+    },
+    {
+      "epoch": 2.2412698412698413,
+      "grad_norm": 1.9904078872978332,
+      "learning_rate": 4.442655338968373e-06,
+      "loss": 0.0201,
+      "step": 531
+    },
+    {
+      "epoch": 2.2455026455026457,
+      "grad_norm": 1.1693403834565992,
+      "learning_rate": 4.425314247535668e-06,
+      "loss": 0.021,
+      "step": 532
+    },
+    {
+      "epoch": 2.2497354497354496,
+      "grad_norm": 5.074805136944312,
+      "learning_rate": 4.4079801584674955e-06,
+      "loss": 0.0222,
+      "step": 533
+    },
+    {
+      "epoch": 2.253968253968254,
+      "grad_norm": 1.079029935941859,
+      "learning_rate": 4.390653282974264e-06,
+      "loss": 0.0199,
+      "step": 534
+    },
+    {
+      "epoch": 2.2582010582010583,
+      "grad_norm": 2.38714350665424,
+      "learning_rate": 4.373333832178478e-06,
+      "loss": 0.0298,
+      "step": 535
+    },
+    {
+      "epoch": 2.2624338624338622,
+      "grad_norm": 1.2813798644078875,
+      "learning_rate": 4.356022017112187e-06,
+      "loss": 0.0201,
+      "step": 536
+    },
+    {
+      "epoch": 2.2666666666666666,
+      "grad_norm": 1.305628889106575,
+      "learning_rate": 4.3387180487143875e-06,
+      "loss": 0.0173,
+      "step": 537
+    },
+    {
+      "epoch": 2.270899470899471,
+      "grad_norm": 1.094988751447929,
+      "learning_rate": 4.321422137828479e-06,
+      "loss": 0.0155,
+      "step": 538
+    },
+    {
+      "epoch": 2.2751322751322753,
+      "grad_norm": 1.348309678538079,
+      "learning_rate": 4.304134495199675e-06,
+      "loss": 0.0209,
+      "step": 539
+    },
+    {
+      "epoch": 2.2793650793650793,
+      "grad_norm": 1.5868771818843859,
+      "learning_rate": 4.286855331472442e-06,
+      "loss": 0.0282,
+      "step": 540
+    },
+    {
+      "epoch": 2.2835978835978836,
+      "grad_norm": 1.2350590787759976,
+      "learning_rate": 4.269584857187942e-06,
+      "loss": 0.0212,
+      "step": 541
+    },
+    {
+      "epoch": 2.287830687830688,
+      "grad_norm": 1.4470288037087002,
+      "learning_rate": 4.2523232827814534e-06,
+      "loss": 0.0192,
+      "step": 542
+    },
+    {
+      "epoch": 2.292063492063492,
+      "grad_norm": 1.4099588785263357,
+      "learning_rate": 4.23507081857981e-06,
+      "loss": 0.0184,
+      "step": 543
+    },
+    {
+      "epoch": 2.2962962962962963,
+      "grad_norm": 1.4782113067056102,
+      "learning_rate": 4.217827674798845e-06,
+      "loss": 0.023,
+      "step": 544
+    },
+    {
+      "epoch": 2.3005291005291006,
+      "grad_norm": 1.9086034541901193,
+      "learning_rate": 4.200594061540827e-06,
+      "loss": 0.0275,
+      "step": 545
+    },
+    {
+      "epoch": 2.3047619047619046,
+      "grad_norm": 1.2032912201710146,
+      "learning_rate": 4.183370188791891e-06,
+      "loss": 0.0196,
+      "step": 546
+    },
+    {
+      "epoch": 2.308994708994709,
+      "grad_norm": 1.3308907007952828,
+      "learning_rate": 4.166156266419489e-06,
+      "loss": 0.022,
+      "step": 547
+    },
+    {
+      "epoch": 2.3132275132275133,
+      "grad_norm": 1.0984488540294757,
+      "learning_rate": 4.148952504169839e-06,
+      "loss": 0.0228,
+      "step": 548
+    },
+    {
+      "epoch": 2.317460317460317,
+      "grad_norm": 0.7471478820780099,
+      "learning_rate": 4.131759111665349e-06,
+      "loss": 0.0111,
+      "step": 549
+    },
+    {
+      "epoch": 2.3216931216931216,
+      "grad_norm": 1.0717935625690418,
+      "learning_rate": 4.114576298402085e-06,
+      "loss": 0.0161,
+      "step": 550
+    },
+    {
+      "epoch": 2.325925925925926,
+      "grad_norm": 1.218749961666418,
+      "learning_rate": 4.0974042737472005e-06,
+      "loss": 0.0186,
+      "step": 551
+    },
+    {
+      "epoch": 2.3301587301587303,
+      "grad_norm": 2.8887759534542723,
+      "learning_rate": 4.0802432469364e-06,
+      "loss": 0.0254,
+      "step": 552
+    },
+    {
+      "epoch": 2.3343915343915342,
+      "grad_norm": 1.930475086228816,
+      "learning_rate": 4.063093427071376e-06,
+      "loss": 0.023,
+      "step": 553
+    },
+    {
+      "epoch": 2.3386243386243386,
+      "grad_norm": 1.8908317964947536,
+      "learning_rate": 4.045955023117276e-06,
+      "loss": 0.0285,
+      "step": 554
+    },
+    {
+      "epoch": 2.342857142857143,
+      "grad_norm": 1.1934657197447822,
+      "learning_rate": 4.028828243900141e-06,
+      "loss": 0.0201,
+      "step": 555
+    },
+    {
+      "epoch": 2.3470899470899473,
+      "grad_norm": 1.4865568918577254,
+      "learning_rate": 4.0117132981043695e-06,
+      "loss": 0.0273,
+      "step": 556
+    },
+    {
+      "epoch": 2.3513227513227513,
+      "grad_norm": 0.9529787490185494,
+      "learning_rate": 3.994610394270178e-06,
+      "loss": 0.0137,
+      "step": 557
+    },
+    {
+      "epoch": 2.3555555555555556,
+      "grad_norm": 1.4123895372263633,
+      "learning_rate": 3.977519740791049e-06,
+      "loss": 0.0247,
+      "step": 558
+    },
+    {
+      "epoch": 2.35978835978836,
+      "grad_norm": 1.4967579484639353,
+      "learning_rate": 3.960441545911205e-06,
+      "loss": 0.0223,
+      "step": 559
+    },
+    {
+      "epoch": 2.364021164021164,
+      "grad_norm": 1.3782252518557059,
+      "learning_rate": 3.943376017723058e-06,
+      "loss": 0.0203,
+      "step": 560
+    },
+    {
+      "epoch": 2.3682539682539683,
+      "grad_norm": 1.011290054301172,
+      "learning_rate": 3.926323364164684e-06,
+      "loss": 0.0174,
+      "step": 561
+    },
+    {
+      "epoch": 2.3724867724867726,
+      "grad_norm": 0.9372736736401631,
+      "learning_rate": 3.909283793017289e-06,
+      "loss": 0.0127,
+      "step": 562
+    },
+    {
+      "epoch": 2.3767195767195766,
+      "grad_norm": 1.0720245224295357,
+      "learning_rate": 3.892257511902664e-06,
+      "loss": 0.0185,
+      "step": 563
+    },
+    {
+      "epoch": 2.380952380952381,
+      "grad_norm": 1.3820051395972568,
+      "learning_rate": 3.875244728280676e-06,
+      "loss": 0.0245,
+      "step": 564
+    },
+    {
+      "epoch": 2.3851851851851853,
+      "grad_norm": 1.54167399561864,
+      "learning_rate": 3.8582456494467214e-06,
+      "loss": 0.0229,
+      "step": 565
+    },
+    {
+      "epoch": 2.389417989417989,
+      "grad_norm": 1.3749494144666705,
+      "learning_rate": 3.841260482529215e-06,
+      "loss": 0.0209,
+      "step": 566
+    },
+    {
+      "epoch": 2.3936507936507936,
+      "grad_norm": 0.9837959983275872,
+      "learning_rate": 3.82428943448705e-06,
+      "loss": 0.0153,
+      "step": 567
+    },
+    {
+      "epoch": 2.397883597883598,
+      "grad_norm": 1.2997925691857108,
+      "learning_rate": 3.8073327121070968e-06,
+      "loss": 0.0168,
+      "step": 568
+    },
+    {
+      "epoch": 2.402116402116402,
+      "grad_norm": 0.99630339528548,
+      "learning_rate": 3.790390522001662e-06,
+      "loss": 0.0162,
+      "step": 569
+    },
+    {
+      "epoch": 2.4063492063492062,
+      "grad_norm": 1.2303194376503306,
+      "learning_rate": 3.7734630706059873e-06,
+      "loss": 0.0185,
+      "step": 570
+    },
+    {
+      "epoch": 2.4105820105820106,
+      "grad_norm": 1.7621485315307714,
+      "learning_rate": 3.756550564175727e-06,
+      "loss": 0.0255,
+      "step": 571
+    },
+    {
+      "epoch": 2.414814814814815,
+      "grad_norm": 1.0569009320866272,
+      "learning_rate": 3.7396532087844318e-06,
+      "loss": 0.0143,
+      "step": 572
+    },
+    {
+      "epoch": 2.419047619047619,
+      "grad_norm": 1.7821677956105721,
+      "learning_rate": 3.7227712103210485e-06,
+      "loss": 0.0249,
+      "step": 573
+    },
+    {
+      "epoch": 2.4232804232804233,
+      "grad_norm": 1.3573086881674947,
+      "learning_rate": 3.705904774487396e-06,
+      "loss": 0.0204,
+      "step": 574
+    },
+    {
+      "epoch": 2.4275132275132276,
+      "grad_norm": 1.1718662977450818,
+      "learning_rate": 3.6890541067956775e-06,
+      "loss": 0.0146,
+      "step": 575
+    },
+    {
+      "epoch": 2.431746031746032,
+      "grad_norm": 1.614044036491686,
+      "learning_rate": 3.672219412565956e-06,
+      "loss": 0.0192,
+      "step": 576
+    },
+    {
+      "epoch": 2.435978835978836,
+      "grad_norm": 1.5234312497660505,
+      "learning_rate": 3.655400896923672e-06,
+      "loss": 0.024,
+      "step": 577
+    },
+    {
+      "epoch": 2.4402116402116403,
+      "grad_norm": 1.2229255498253058,
+      "learning_rate": 3.6385987647971287e-06,
+      "loss": 0.0213,
+      "step": 578
+    },
+    {
+      "epoch": 2.4444444444444446,
+      "grad_norm": 1.188233949021249,
+      "learning_rate": 3.6218132209150047e-06,
+      "loss": 0.0158,
+      "step": 579
+    },
+    {
+      "epoch": 2.4486772486772486,
+      "grad_norm": 1.2163503629423913,
+      "learning_rate": 3.6050444698038547e-06,
+      "loss": 0.0233,
+      "step": 580
+    },
+    {
+      "epoch": 2.452910052910053,
+      "grad_norm": 1.181425300400814,
+      "learning_rate": 3.5882927157856175e-06,
+      "loss": 0.0196,
+      "step": 581
+    },
+    {
+      "epoch": 2.4571428571428573,
+      "grad_norm": 1.7922575768503977,
+      "learning_rate": 3.571558162975133e-06,
+      "loss": 0.0218,
+      "step": 582
+    },
+    {
+      "epoch": 2.461375661375661,
+      "grad_norm": 5.6028053486201514,
+      "learning_rate": 3.5548410152776414e-06,
+      "loss": 0.0175,
+      "step": 583
+    },
+    {
+      "epoch": 2.4656084656084656,
+      "grad_norm": 1.4059375192080907,
+      "learning_rate": 3.538141476386317e-06,
+      "loss": 0.0217,
+      "step": 584
+    },
+    {
+      "epoch": 2.46984126984127,
+      "grad_norm": 1.2473008607933311,
+      "learning_rate": 3.521459749779769e-06,
+      "loss": 0.0211,
+      "step": 585
+    },
+    {
+      "epoch": 2.474074074074074,
+      "grad_norm": 1.0232237746517738,
+      "learning_rate": 3.5047960387195673e-06,
+      "loss": 0.0164,
+      "step": 586
+    },
+    {
+      "epoch": 2.4783068783068782,
+      "grad_norm": 1.1551458380571011,
+      "learning_rate": 3.488150546247778e-06,
+      "loss": 0.0177,
+      "step": 587
+    },
+    {
+      "epoch": 2.4825396825396826,
+      "grad_norm": 1.241989730027279,
+      "learning_rate": 3.471523475184472e-06,
+      "loss": 0.0213,
+      "step": 588
+    },
+    {
+      "epoch": 2.4867724867724865,
+      "grad_norm": 1.3270342015490493,
+      "learning_rate": 3.4549150281252635e-06,
+      "loss": 0.0227,
+      "step": 589
+    },
+    {
+      "epoch": 2.491005291005291,
+      "grad_norm": 1.3678587583649702,
+      "learning_rate": 3.4383254074388373e-06,
+      "loss": 0.0228,
+      "step": 590
+    },
+    {
+      "epoch": 2.4952380952380953,
+      "grad_norm": 2.2689660584590783,
+      "learning_rate": 3.4217548152644887e-06,
+      "loss": 0.0175,
+      "step": 591
+    },
+    {
+      "epoch": 2.4994708994708996,
+      "grad_norm": 1.5596005660525438,
+      "learning_rate": 3.40520345350965e-06,
+      "loss": 0.0199,
+      "step": 592
+    },
+    {
+      "epoch": 2.5037037037037035,
+      "grad_norm": 1.2720201006007885,
+      "learning_rate": 3.3886715238474454e-06,
+      "loss": 0.0177,
+      "step": 593
+    },
+    {
+      "epoch": 2.507936507936508,
+      "grad_norm": 1.093460165926427,
+      "learning_rate": 3.372159227714218e-06,
+      "loss": 0.0139,
+      "step": 594
+    },
+    {
+      "epoch": 2.5121693121693123,
+      "grad_norm": 1.6179479955437774,
+      "learning_rate": 3.355666766307084e-06,
+      "loss": 0.0232,
+      "step": 595
+    },
+    {
+      "epoch": 2.5164021164021166,
+      "grad_norm": 1.2102869187922056,
+      "learning_rate": 3.339194340581485e-06,
+      "loss": 0.0187,
+      "step": 596
+    },
+    {
+      "epoch": 2.5206349206349206,
+      "grad_norm": 1.1103393510900712,
+      "learning_rate": 3.322742151248726e-06,
+      "loss": 0.0143,
+      "step": 597
+    },
+    {
+      "epoch": 2.524867724867725,
+      "grad_norm": 1.2647278804921822,
+      "learning_rate": 3.3063103987735433e-06,
+      "loss": 0.0186,
+      "step": 598
+    },
+    {
+      "epoch": 2.5291005291005293,
+      "grad_norm": 1.1276120725215881,
+      "learning_rate": 3.289899283371657e-06,
+      "loss": 0.0148,
+      "step": 599
+    },
+    {
+      "epoch": 2.533333333333333,
+      "grad_norm": 1.0196598923468276,
+      "learning_rate": 3.273509005007327e-06,
+      "loss": 0.0166,
+      "step": 600
+    },
+    {
+      "epoch": 2.5375661375661376,
+      "grad_norm": 0.9388923647646054,
+      "learning_rate": 3.2571397633909252e-06,
+      "loss": 0.0147,
+      "step": 601
+    },
+    {
+      "epoch": 2.541798941798942,
+      "grad_norm": 1.09462020411353,
+      "learning_rate": 3.2407917579764914e-06,
+      "loss": 0.0152,
+      "step": 602
+    },
+    {
+      "epoch": 2.546031746031746,
+      "grad_norm": 1.066286773297906,
+      "learning_rate": 3.224465187959316e-06,
+      "loss": 0.0169,
+      "step": 603
+    },
+    {
+      "epoch": 2.5502645502645502,
+      "grad_norm": 1.5390006694829865,
+      "learning_rate": 3.2081602522734987e-06,
+      "loss": 0.0198,
+      "step": 604
+    },
+    {
+      "epoch": 2.5544973544973546,
+      "grad_norm": 1.4007819028143056,
+      "learning_rate": 3.1918771495895395e-06,
+      "loss": 0.0203,
+      "step": 605
+    },
+    {
+      "epoch": 2.5587301587301585,
+      "grad_norm": 0.9872842079869668,
+      "learning_rate": 3.1756160783119015e-06,
+      "loss": 0.014,
+      "step": 606
+    },
+    {
+      "epoch": 2.562962962962963,
+      "grad_norm": 1.0921886747672007,
+      "learning_rate": 3.1593772365766107e-06,
+      "loss": 0.0162,
+      "step": 607
+    },
+    {
+      "epoch": 2.5671957671957673,
+      "grad_norm": 1.1585886648374661,
+      "learning_rate": 3.1431608222488276e-06,
+      "loss": 0.0163,
+      "step": 608
+    },
+    {
+      "epoch": 2.571428571428571,
+      "grad_norm": 1.248592252763677,
+      "learning_rate": 3.12696703292044e-06,
+      "loss": 0.0199,
+      "step": 609
+    },
+    {
+      "epoch": 2.5756613756613755,
+      "grad_norm": 1.4112394419564622,
+      "learning_rate": 3.110796065907665e-06,
+      "loss": 0.022,
+      "step": 610
+    },
+    {
+      "epoch": 2.57989417989418,
+      "grad_norm": 1.165336623248628,
+      "learning_rate": 3.09464811824863e-06,
+      "loss": 0.0168,
+      "step": 611
+    },
+    {
+      "epoch": 2.5841269841269843,
+      "grad_norm": 1.4448704250044067,
+      "learning_rate": 3.078523386700982e-06,
+      "loss": 0.0157,
+      "step": 612
+    },
+    {
+      "epoch": 2.588359788359788,
+      "grad_norm": 1.3676792063312764,
+      "learning_rate": 3.0624220677394854e-06,
+      "loss": 0.0199,
+      "step": 613
+    },
+    {
+      "epoch": 2.5925925925925926,
+      "grad_norm": 1.129628210066895,
+      "learning_rate": 3.0463443575536324e-06,
+      "loss": 0.0178,
+      "step": 614
+    },
+    {
+      "epoch": 2.596825396825397,
+      "grad_norm": 1.0316857304867937,
+      "learning_rate": 3.030290452045245e-06,
+      "loss": 0.0199,
+      "step": 615
+    },
+    {
+      "epoch": 2.6010582010582013,
+      "grad_norm": 1.3226258154560515,
+      "learning_rate": 3.0142605468260976e-06,
+      "loss": 0.0152,
+      "step": 616
+    },
+    {
+      "epoch": 2.605291005291005,
+      "grad_norm": 1.5852740515104826,
+      "learning_rate": 2.9982548372155264e-06,
+      "loss": 0.0271,
+      "step": 617
+    },
+    {
+      "epoch": 2.6095238095238096,
+      "grad_norm": 1.1196079859503962,
+      "learning_rate": 2.98227351823805e-06,
+      "loss": 0.0138,
+      "step": 618
+    },
+    {
+      "epoch": 2.613756613756614,
+      "grad_norm": 1.8626892669889739,
+      "learning_rate": 2.966316784621e-06,
+      "loss": 0.0264,
+      "step": 619
+    },
+    {
+      "epoch": 2.617989417989418,
+      "grad_norm": 1.383225574585185,
+      "learning_rate": 2.9503848307921363e-06,
+      "loss": 0.0205,
+      "step": 620
+    },
+    {
+      "epoch": 2.6222222222222222,
+      "grad_norm": 2.733985735902233,
+      "learning_rate": 2.934477850877292e-06,
+      "loss": 0.0246,
+      "step": 621
+    },
+    {
+      "epoch": 2.6264550264550266,
+      "grad_norm": 1.1381292666769829,
+      "learning_rate": 2.918596038697995e-06,
+      "loss": 0.0188,
+      "step": 622
+    },
+    {
+      "epoch": 2.6306878306878305,
+      "grad_norm": 1.0370355502501467,
+      "learning_rate": 2.9027395877691143e-06,
+      "loss": 0.0162,
+      "step": 623
+    },
+    {
+      "epoch": 2.634920634920635,
+      "grad_norm": 1.439361804374129,
+      "learning_rate": 2.886908691296504e-06,
+      "loss": 0.0197,
+      "step": 624
+    },
+    {
+      "epoch": 2.6391534391534393,
+      "grad_norm": 1.0102400740735453,
+      "learning_rate": 2.871103542174637e-06,
+      "loss": 0.0175,
+      "step": 625
+    },
+    {
+      "epoch": 2.643386243386243,
+      "grad_norm": 1.5198907140103581,
+      "learning_rate": 2.8553243329842715e-06,
+      "loss": 0.0196,
+      "step": 626
+    },
+    {
+      "epoch": 2.6476190476190475,
+      "grad_norm": 1.0531509330878106,
+      "learning_rate": 2.839571255990088e-06,
+      "loss": 0.0141,
+      "step": 627
+    },
+    {
+      "epoch": 2.651851851851852,
+      "grad_norm": 1.0357940819910703,
+      "learning_rate": 2.8238445031383634e-06,
+      "loss": 0.0173,
+      "step": 628
+    },
+    {
+      "epoch": 2.656084656084656,
+      "grad_norm": 1.1307445427957337,
+      "learning_rate": 2.8081442660546126e-06,
+      "loss": 0.0184,
+      "step": 629
+    },
+    {
+      "epoch": 2.66031746031746,
+      "grad_norm": 1.3306823534784538,
+      "learning_rate": 2.7924707360412743e-06,
+      "loss": 0.0147,
+      "step": 630
+    },
+    {
+      "epoch": 2.6645502645502646,
+      "grad_norm": 1.5271179132418502,
+      "learning_rate": 2.776824104075364e-06,
+      "loss": 0.021,
+      "step": 631
+    },
+    {
+      "epoch": 2.668783068783069,
+      "grad_norm": 1.497782212123214,
+      "learning_rate": 2.761204560806152e-06,
+      "loss": 0.0186,
+      "step": 632
+    },
+    {
+      "epoch": 2.6730158730158733,
+      "grad_norm": 1.3368279908950362,
+      "learning_rate": 2.7456122965528475e-06,
+      "loss": 0.0206,
+      "step": 633
+    },
+    {
+      "epoch": 2.677248677248677,
+      "grad_norm": 1.2484325713697815,
+      "learning_rate": 2.7300475013022666e-06,
+      "loss": 0.0173,
+      "step": 634
+    },
+    {
+      "epoch": 2.6814814814814816,
+      "grad_norm": 0.865644253076397,
+      "learning_rate": 2.714510364706531e-06,
+      "loss": 0.0128,
+      "step": 635
+    },
+    {
+      "epoch": 2.685714285714286,
+      "grad_norm": 1.0152662338026661,
+      "learning_rate": 2.699001076080742e-06,
+      "loss": 0.0152,
+      "step": 636
+    },
+    {
+      "epoch": 2.68994708994709,
+      "grad_norm": 0.9844152479186873,
+      "learning_rate": 2.683519824400693e-06,
+      "loss": 0.0156,
+      "step": 637
+    },
+    {
+      "epoch": 2.6941798941798942,
+      "grad_norm": 0.7334924826589737,
+      "learning_rate": 2.6680667983005446e-06,
+      "loss": 0.0125,
+      "step": 638
+    },
+    {
+      "epoch": 2.6984126984126986,
+      "grad_norm": 1.1114018703107458,
+      "learning_rate": 2.6526421860705474e-06,
+      "loss": 0.018,
+      "step": 639
+    },
+    {
+      "epoch": 2.7026455026455025,
+      "grad_norm": 1.7960430432669203,
+      "learning_rate": 2.637246175654731e-06,
+      "loss": 0.0177,
+      "step": 640
+    },
+    {
+      "epoch": 2.706878306878307,
+      "grad_norm": 0.8414473229281895,
+      "learning_rate": 2.6218789546486235e-06,
+      "loss": 0.0127,
+      "step": 641
+    },
+    {
+      "epoch": 2.7111111111111112,
+      "grad_norm": 1.4403979439268044,
+      "learning_rate": 2.6065407102969664e-06,
+      "loss": 0.021,
+      "step": 642
+    },
+    {
+      "epoch": 2.715343915343915,
+      "grad_norm": 1.4586265785977015,
+      "learning_rate": 2.5912316294914232e-06,
+      "loss": 0.0216,
+      "step": 643
+    },
+    {
+      "epoch": 2.7195767195767195,
+      "grad_norm": 1.6315741402908153,
+      "learning_rate": 2.5759518987683154e-06,
+      "loss": 0.0162,
+      "step": 644
+    },
+    {
+      "epoch": 2.723809523809524,
+      "grad_norm": 1.34535240869925,
+      "learning_rate": 2.560701704306336e-06,
+      "loss": 0.019,
+      "step": 645
+    },
+    {
+      "epoch": 2.728042328042328,
+      "grad_norm": 0.9316829843215475,
+      "learning_rate": 2.545481231924296e-06,
+      "loss": 0.0172,
+      "step": 646
+    },
+    {
+      "epoch": 2.732275132275132,
+      "grad_norm": 1.4011841310159103,
+      "learning_rate": 2.5302906670788463e-06,
+      "loss": 0.0168,
+      "step": 647
+    },
+    {
+      "epoch": 2.7365079365079366,
+      "grad_norm": 0.6747870852771698,
+      "learning_rate": 2.5151301948622235e-06,
+      "loss": 0.0136,
+      "step": 648
+    },
+    {
+      "epoch": 2.7407407407407405,
+      "grad_norm": 1.057610980636594,
+      "learning_rate": 2.5000000000000015e-06,
+      "loss": 0.0131,
+      "step": 649
+    },
+    {
+      "epoch": 2.744973544973545,
+      "grad_norm": 1.6746201989469378,
+      "learning_rate": 2.484900266848825e-06,
+      "loss": 0.0205,
+      "step": 650
+    },
+    {
+      "epoch": 2.749206349206349,
+      "grad_norm": 0.9494040703209882,
+      "learning_rate": 2.469831179394182e-06,
+      "loss": 0.0135,
+      "step": 651
+    },
+    {
+      "epoch": 2.7534391534391536,
+      "grad_norm": 1.5003754041430553,
+      "learning_rate": 2.4547929212481436e-06,
+      "loss": 0.0158,
+      "step": 652
+    },
+    {
+      "epoch": 2.757671957671958,
+      "grad_norm": 0.8934711380462033,
+      "learning_rate": 2.4397856756471435e-06,
+      "loss": 0.0154,
+      "step": 653
+    },
+    {
+      "epoch": 2.761904761904762,
+      "grad_norm": 1.6609542321924584,
+      "learning_rate": 2.424809625449729e-06,
+      "loss": 0.0181,
+      "step": 654
+    },
+    {
+      "epoch": 2.7661375661375662,
+      "grad_norm": 1.6824205917017476,
+      "learning_rate": 2.40986495313435e-06,
+      "loss": 0.0169,
+      "step": 655
+    },
+    {
+      "epoch": 2.7703703703703706,
+      "grad_norm": 0.9183880056629511,
+      "learning_rate": 2.39495184079712e-06,
+      "loss": 0.0134,
+      "step": 656
+    },
+    {
+      "epoch": 2.7746031746031745,
+      "grad_norm": 1.0245037009884315,
+      "learning_rate": 2.380070470149605e-06,
+      "loss": 0.0139,
+      "step": 657
+    },
+    {
+      "epoch": 2.778835978835979,
+      "grad_norm": 1.3421089341613759,
+      "learning_rate": 2.3652210225166122e-06,
+      "loss": 0.0142,
+      "step": 658
+    },
+    {
+      "epoch": 2.7830687830687832,
+      "grad_norm": 1.3809956989606829,
+      "learning_rate": 2.3504036788339763e-06,
+      "loss": 0.0178,
+      "step": 659
+    },
+    {
+      "epoch": 2.787301587301587,
+      "grad_norm": 1.1273731561378648,
+      "learning_rate": 2.3356186196463497e-06,
+      "loss": 0.0185,
+      "step": 660
+    },
+    {
+      "epoch": 2.7915343915343915,
+      "grad_norm": 1.0143190004115585,
+      "learning_rate": 2.320866025105016e-06,
+      "loss": 0.0124,
+      "step": 661
+    },
+    {
+      "epoch": 2.795767195767196,
+      "grad_norm": 1.3534973057449449,
+      "learning_rate": 2.3061460749656844e-06,
+      "loss": 0.0208,
+      "step": 662
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 1.0435939785814092,
+      "learning_rate": 2.2914589485863015e-06,
+      "loss": 0.0163,
+      "step": 663
+    },
+    {
+      "epoch": 2.804232804232804,
+      "grad_norm": 0.9203370702295048,
+      "learning_rate": 2.2768048249248648e-06,
+      "loss": 0.0146,
+      "step": 664
+    },
+    {
+      "epoch": 2.8084656084656086,
+      "grad_norm": 1.098704858652637,
+      "learning_rate": 2.2621838825372496e-06,
+      "loss": 0.0123,
+      "step": 665
+    },
+    {
+      "epoch": 2.8126984126984125,
+      "grad_norm": 1.191480174922082,
+      "learning_rate": 2.2475962995750224e-06,
+      "loss": 0.0144,
+      "step": 666
+    },
+    {
+      "epoch": 2.816931216931217,
+      "grad_norm": 1.276715625287459,
+      "learning_rate": 2.23304225378328e-06,
+      "loss": 0.0188,
+      "step": 667
+    },
+    {
+      "epoch": 2.821164021164021,
+      "grad_norm": 1.131649710749666,
+      "learning_rate": 2.218521922498476e-06,
+      "loss": 0.0167,
+      "step": 668
+    },
+    {
+      "epoch": 2.825396825396825,
+      "grad_norm": 0.8016802676955552,
+      "learning_rate": 2.204035482646267e-06,
+      "loss": 0.0131,
+      "step": 669
+    },
+    {
+      "epoch": 2.8296296296296295,
+      "grad_norm": 0.6893558290783893,
+      "learning_rate": 2.1895831107393485e-06,
+      "loss": 0.0108,
+      "step": 670
+    },
+    {
+      "epoch": 2.833862433862434,
+      "grad_norm": 1.234838405201532,
+      "learning_rate": 2.175164982875311e-06,
+      "loss": 0.017,
+      "step": 671
+    },
+    {
+      "epoch": 2.8380952380952382,
+      "grad_norm": 1.2011736692296922,
+      "learning_rate": 2.1607812747344955e-06,
+      "loss": 0.0157,
+      "step": 672
+    },
+    {
+      "epoch": 2.8423280423280426,
+      "grad_norm": 1.1466267503085876,
+      "learning_rate": 2.146432161577842e-06,
+      "loss": 0.0157,
+      "step": 673
+    },
+    {
+      "epoch": 2.8465608465608465,
+      "grad_norm": 0.969199506256315,
+      "learning_rate": 2.132117818244771e-06,
+      "loss": 0.0152,
+      "step": 674
+    },
+    {
+      "epoch": 2.850793650793651,
+      "grad_norm": 1.3534092293875413,
+      "learning_rate": 2.1178384191510344e-06,
+      "loss": 0.0162,
+      "step": 675
+    },
+    {
+      "epoch": 2.8550264550264552,
+      "grad_norm": 1.2829883064220085,
+      "learning_rate": 2.103594138286607e-06,
+      "loss": 0.0195,
+      "step": 676
+    },
+    {
+      "epoch": 2.859259259259259,
+      "grad_norm": 1.1271167538927263,
+      "learning_rate": 2.0893851492135536e-06,
+      "loss": 0.0172,
+      "step": 677
+    },
+    {
+      "epoch": 2.8634920634920635,
+      "grad_norm": 1.343559581567661,
+      "learning_rate": 2.075211625063923e-06,
+      "loss": 0.0163,
+      "step": 678
+    },
+    {
+      "epoch": 2.867724867724868,
+      "grad_norm": 1.1918468332382046,
+      "learning_rate": 2.061073738537635e-06,
+      "loss": 0.0164,
+      "step": 679
+    },
+    {
+      "epoch": 2.871957671957672,
+      "grad_norm": 1.4193106995910565,
+      "learning_rate": 2.046971661900373e-06,
+      "loss": 0.017,
+      "step": 680
+    },
+    {
+      "epoch": 2.876190476190476,
+      "grad_norm": 0.7357092999278971,
+      "learning_rate": 2.0329055669814936e-06,
+      "loss": 0.0123,
+      "step": 681
+    },
+    {
+      "epoch": 2.8804232804232806,
+      "grad_norm": 1.4745946537108765,
+      "learning_rate": 2.0188756251719204e-06,
+      "loss": 0.0153,
+      "step": 682
+    },
+    {
+      "epoch": 2.8846560846560845,
+      "grad_norm": 1.1857296628720868,
+      "learning_rate": 2.0048820074220716e-06,
+      "loss": 0.0188,
+      "step": 683
+    },
+    {
+      "epoch": 2.888888888888889,
+      "grad_norm": 1.4131026911127977,
+      "learning_rate": 1.990924884239758e-06,
+      "loss": 0.0173,
+      "step": 684
+    },
+    {
+      "epoch": 2.893121693121693,
+      "grad_norm": 1.0848820434397748,
+      "learning_rate": 1.977004425688126e-06,
+      "loss": 0.0177,
+      "step": 685
+    },
+    {
+      "epoch": 2.897354497354497,
+      "grad_norm": 1.0664286982794255,
+      "learning_rate": 1.9631208013835677e-06,
+      "loss": 0.0124,
+      "step": 686
+    },
+    {
+      "epoch": 2.9015873015873015,
+      "grad_norm": 1.0870334070827417,
+      "learning_rate": 1.9492741804936623e-06,
+      "loss": 0.0134,
+      "step": 687
+    },
+    {
+      "epoch": 2.905820105820106,
+      "grad_norm": 0.9405135966991898,
+      "learning_rate": 1.9354647317351187e-06,
+      "loss": 0.0138,
+      "step": 688
+    },
+    {
+      "epoch": 2.91005291005291,
+      "grad_norm": 1.1557524847446539,
+      "learning_rate": 1.9216926233717087e-06,
+      "loss": 0.0154,
+      "step": 689
+    },
+    {
+      "epoch": 2.914285714285714,
+      "grad_norm": 1.145639729440071,
+      "learning_rate": 1.90795802321223e-06,
+      "loss": 0.0163,
+      "step": 690
+    },
+    {
+      "epoch": 2.9185185185185185,
+      "grad_norm": 1.3118126671264705,
+      "learning_rate": 1.8942610986084487e-06,
+      "loss": 0.0148,
+      "step": 691
+    },
+    {
+      "epoch": 2.922751322751323,
+      "grad_norm": 1.0498615239365812,
+      "learning_rate": 1.8806020164530702e-06,
+      "loss": 0.0135,
+      "step": 692
+    },
+    {
+      "epoch": 2.9269841269841272,
+      "grad_norm": 1.0097762792486051,
+      "learning_rate": 1.8669809431776991e-06,
+      "loss": 0.0134,
+      "step": 693
+    },
+    {
+      "epoch": 2.931216931216931,
+      "grad_norm": 1.0798461928002918,
+      "learning_rate": 1.8533980447508138e-06,
+      "loss": 0.0138,
+      "step": 694
+    },
+    {
+      "epoch": 2.9354497354497355,
+      "grad_norm": 0.7591375304655225,
+      "learning_rate": 1.8398534866757455e-06,
+      "loss": 0.0085,
+      "step": 695
+    },
+    {
+      "epoch": 2.93968253968254,
+      "grad_norm": 1.1784009394326262,
+      "learning_rate": 1.8263474339886628e-06,
+      "loss": 0.0132,
+      "step": 696
+    },
+    {
+      "epoch": 2.943915343915344,
+      "grad_norm": 1.220548304552571,
+      "learning_rate": 1.8128800512565514e-06,
+      "loss": 0.0223,
+      "step": 697
+    },
+    {
+      "epoch": 2.948148148148148,
+      "grad_norm": 1.1132676461069289,
+      "learning_rate": 1.799451502575222e-06,
+      "loss": 0.0144,
+      "step": 698
+    },
+    {
+      "epoch": 2.9523809523809526,
+      "grad_norm": 0.7581893865812536,
+      "learning_rate": 1.7860619515673034e-06,
+      "loss": 0.011,
+      "step": 699
+    },
+    {
+      "epoch": 2.9566137566137565,
+      "grad_norm": 0.9716146953090274,
+      "learning_rate": 1.7727115613802465e-06,
+      "loss": 0.0096,
+      "step": 700
+    },
+    {
+      "epoch": 2.960846560846561,
+      "grad_norm": 1.674856558431357,
+      "learning_rate": 1.7594004946843458e-06,
+      "loss": 0.0128,
+      "step": 701
+    },
+    {
+      "epoch": 2.965079365079365,
+      "grad_norm": 1.1843497789446338,
+      "learning_rate": 1.746128913670746e-06,
+      "loss": 0.0144,
+      "step": 702
+    },
+    {
+      "epoch": 2.969312169312169,
+      "grad_norm": 1.7482050222848124,
+      "learning_rate": 1.7328969800494727e-06,
+      "loss": 0.0191,
+      "step": 703
+    },
+    {
+      "epoch": 2.9735449735449735,
+      "grad_norm": 1.3527434122410733,
+      "learning_rate": 1.7197048550474643e-06,
+      "loss": 0.0199,
+      "step": 704
+    },
+    {
+      "epoch": 2.977777777777778,
+      "grad_norm": 1.04160773464941,
+      "learning_rate": 1.7065526994065973e-06,
+      "loss": 0.0187,
+      "step": 705
+    },
+    {
+      "epoch": 2.982010582010582,
+      "grad_norm": 1.262543944063061,
+      "learning_rate": 1.6934406733817417e-06,
+      "loss": 0.019,
+      "step": 706
+    },
+    {
+      "epoch": 2.986243386243386,
+      "grad_norm": 1.1849446771124943,
+      "learning_rate": 1.680368936738792e-06,
+      "loss": 0.0162,
+      "step": 707
+    },
+    {
+      "epoch": 2.9904761904761905,
+      "grad_norm": 1.0664910484388517,
+      "learning_rate": 1.6673376487527382e-06,
+      "loss": 0.0161,
+      "step": 708
+    },
+    {
+      "epoch": 2.9947089947089944,
+      "grad_norm": 2.0774774510855742,
+      "learning_rate": 1.6543469682057105e-06,
+      "loss": 0.0194,
+      "step": 709
+    },
+    {
+      "epoch": 2.998941798941799,
+      "grad_norm": 1.0920006522906185,
+      "learning_rate": 1.6413970533850498e-06,
+      "loss": 0.0117,
+      "step": 710
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 1.4997123713326503,
+      "learning_rate": 1.6284880620813847e-06,
+      "loss": 0.0124,
+      "step": 711
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 948,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 4,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 116192413286400.0,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}