diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100755--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,4613 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.2002136426064398,
+  "eval_steps": 500,
+  "global_step": 656,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0003052037234854265,
+      "grad_norm": 19.476922880741295,
+      "learning_rate": 1.0101010101010103e-07,
+      "loss": 1.1728,
+      "step": 1
+    },
+    {
+      "epoch": 0.000610407446970853,
+      "grad_norm": 29.879020388476594,
+      "learning_rate": 2.0202020202020205e-07,
+      "loss": 1.0955,
+      "step": 2
+    },
+    {
+      "epoch": 0.0009156111704562796,
+      "grad_norm": 24.931945947136526,
+      "learning_rate": 3.0303030303030305e-07,
+      "loss": 0.9541,
+      "step": 3
+    },
+    {
+      "epoch": 0.001220814893941706,
+      "grad_norm": 27.83214939667906,
+      "learning_rate": 4.040404040404041e-07,
+      "loss": 1.0735,
+      "step": 4
+    },
+    {
+      "epoch": 0.0015260186174271325,
+      "grad_norm": 21.219233961021736,
+      "learning_rate": 5.05050505050505e-07,
+      "loss": 1.0455,
+      "step": 5
+    },
+    {
+      "epoch": 0.0018312223409125592,
+      "grad_norm": 20.022707446211225,
+      "learning_rate": 6.060606060606061e-07,
+      "loss": 0.9675,
+      "step": 6
+    },
+    {
+      "epoch": 0.0021364260643979855,
+      "grad_norm": 26.532427830157193,
+      "learning_rate": 7.070707070707071e-07,
+      "loss": 1.1393,
+      "step": 7
+    },
+    {
+      "epoch": 0.002441629787883412,
+      "grad_norm": 27.89728780710031,
+      "learning_rate": 8.080808080808082e-07,
+      "loss": 1.0952,
+      "step": 8
+    },
+    {
+      "epoch": 0.0027468335113688385,
+      "grad_norm": 20.346264005570532,
+      "learning_rate": 9.090909090909091e-07,
+      "loss": 0.9626,
+      "step": 9
+    },
+    {
+      "epoch": 0.003052037234854265,
+      "grad_norm": 18.804489508720884,
+      "learning_rate": 1.01010101010101e-06,
+      "loss": 1.0255,
+      "step": 10
+    },
+    {
+      "epoch": 0.003357240958339692,
+      "grad_norm": 19.776534785573535,
+      "learning_rate": 1.111111111111111e-06,
+      "loss": 0.7399,
+      "step": 11
+    },
+    {
+      "epoch": 0.0036624446818251184,
+      "grad_norm": 21.16130386460154,
+      "learning_rate": 1.2121212121212122e-06,
+      "loss": 0.5413,
+      "step": 12
+    },
+    {
+      "epoch": 0.0039676484053105445,
+      "grad_norm": 16.482713371526263,
+      "learning_rate": 1.3131313131313134e-06,
+      "loss": 0.5773,
+      "step": 13
+    },
+    {
+      "epoch": 0.004272852128795971,
+      "grad_norm": 10.780528168770594,
+      "learning_rate": 1.4141414141414143e-06,
+      "loss": 0.6782,
+      "step": 14
+    },
+    {
+      "epoch": 0.0045780558522813975,
+      "grad_norm": 7.0900135030469915,
+      "learning_rate": 1.5151515151515152e-06,
+      "loss": 0.9153,
+      "step": 15
+    },
+    {
+      "epoch": 0.004883259575766824,
+      "grad_norm": 8.490445320662754,
+      "learning_rate": 1.6161616161616164e-06,
+      "loss": 0.4798,
+      "step": 16
+    },
+    {
+      "epoch": 0.0051884632992522505,
+      "grad_norm": 6.677142812986669,
+      "learning_rate": 1.7171717171717173e-06,
+      "loss": 0.4782,
+      "step": 17
+    },
+    {
+      "epoch": 0.005493667022737677,
+      "grad_norm": 5.9204247946017485,
+      "learning_rate": 1.8181818181818183e-06,
+      "loss": 0.3191,
+      "step": 18
+    },
+    {
+      "epoch": 0.0057988707462231035,
+      "grad_norm": 5.012462343754674,
+      "learning_rate": 1.9191919191919192e-06,
+      "loss": 0.4115,
+      "step": 19
+    },
+    {
+      "epoch": 0.00610407446970853,
+      "grad_norm": 3.9095937836899113,
+      "learning_rate": 2.02020202020202e-06,
+      "loss": 0.6158,
+      "step": 20
+    },
+    {
+      "epoch": 0.006409278193193957,
+      "grad_norm": 4.438163815129716,
+      "learning_rate": 2.1212121212121216e-06,
+      "loss": 0.7388,
+      "step": 21
+    },
+    {
+      "epoch": 0.006714481916679384,
+      "grad_norm": 3.62875198348435,
+      "learning_rate": 2.222222222222222e-06,
+      "loss": 0.2875,
+      "step": 22
+    },
+    {
+      "epoch": 0.00701968564016481,
+      "grad_norm": 4.963543929599541,
+      "learning_rate": 2.3232323232323234e-06,
+      "loss": 0.4662,
+      "step": 23
+    },
+    {
+      "epoch": 0.007324889363650237,
+      "grad_norm": 4.274904100558248,
+      "learning_rate": 2.4242424242424244e-06,
+      "loss": 0.5171,
+      "step": 24
+    },
+    {
+      "epoch": 0.007630093087135663,
+      "grad_norm": 2.670885047669819,
+      "learning_rate": 2.5252525252525258e-06,
+      "loss": 0.4488,
+      "step": 25
+    },
+    {
+      "epoch": 0.007935296810621089,
+      "grad_norm": 2.6864388610994014,
+      "learning_rate": 2.6262626262626267e-06,
+      "loss": 0.372,
+      "step": 26
+    },
+    {
+      "epoch": 0.008240500534106516,
+      "grad_norm": 3.804357369452407,
+      "learning_rate": 2.7272727272727272e-06,
+      "loss": 0.2646,
+      "step": 27
+    },
+    {
+      "epoch": 0.008545704257591942,
+      "grad_norm": 4.059008227452532,
+      "learning_rate": 2.8282828282828286e-06,
+      "loss": 0.5907,
+      "step": 28
+    },
+    {
+      "epoch": 0.008850907981077369,
+      "grad_norm": 4.9062443629918855,
+      "learning_rate": 2.9292929292929295e-06,
+      "loss": 0.2972,
+      "step": 29
+    },
+    {
+      "epoch": 0.009156111704562795,
+      "grad_norm": 3.5391495380267064,
+      "learning_rate": 3.0303030303030305e-06,
+      "loss": 0.3821,
+      "step": 30
+    },
+    {
+      "epoch": 0.009461315428048222,
+      "grad_norm": 2.5896920322264854,
+      "learning_rate": 3.131313131313132e-06,
+      "loss": 0.4164,
+      "step": 31
+    },
+    {
+      "epoch": 0.009766519151533648,
+      "grad_norm": 3.0230775761822937,
+      "learning_rate": 3.232323232323233e-06,
+      "loss": 0.4237,
+      "step": 32
+    },
+    {
+      "epoch": 0.010071722875019075,
+      "grad_norm": 2.8417717057519423,
+      "learning_rate": 3.3333333333333333e-06,
+      "loss": 0.3353,
+      "step": 33
+    },
+    {
+      "epoch": 0.010376926598504501,
+      "grad_norm": 2.5789157463945878,
+      "learning_rate": 3.4343434343434347e-06,
+      "loss": 0.3769,
+      "step": 34
+    },
+    {
+      "epoch": 0.010682130321989928,
+      "grad_norm": 2.5222241581850096,
+      "learning_rate": 3.5353535353535356e-06,
+      "loss": 0.519,
+      "step": 35
+    },
+    {
+      "epoch": 0.010987334045475354,
+      "grad_norm": 2.8704682168269127,
+      "learning_rate": 3.6363636363636366e-06,
+      "loss": 0.2829,
+      "step": 36
+    },
+    {
+      "epoch": 0.01129253776896078,
+      "grad_norm": 3.24684532820184,
+      "learning_rate": 3.737373737373738e-06,
+      "loss": 0.3586,
+      "step": 37
+    },
+    {
+      "epoch": 0.011597741492446207,
+      "grad_norm": 5.24792475783676,
+      "learning_rate": 3.8383838383838385e-06,
+      "loss": 0.402,
+      "step": 38
+    },
+    {
+      "epoch": 0.011902945215931634,
+      "grad_norm": 3.111184671834165,
+      "learning_rate": 3.93939393939394e-06,
+      "loss": 0.466,
+      "step": 39
+    },
+    {
+      "epoch": 0.01220814893941706,
+      "grad_norm": 3.165565566985893,
+      "learning_rate": 4.04040404040404e-06,
+      "loss": 0.2678,
+      "step": 40
+    },
+    {
+      "epoch": 0.012513352662902488,
+      "grad_norm": 2.5486933296193257,
+      "learning_rate": 4.141414141414142e-06,
+      "loss": 0.5457,
+      "step": 41
+    },
+    {
+      "epoch": 0.012818556386387915,
+      "grad_norm": 3.4373721012250438,
+      "learning_rate": 4.242424242424243e-06,
+      "loss": 0.3862,
+      "step": 42
+    },
+    {
+      "epoch": 0.013123760109873341,
+      "grad_norm": 2.863317221380458,
+      "learning_rate": 4.343434343434344e-06,
+      "loss": 0.3601,
+      "step": 43
+    },
+    {
+      "epoch": 0.013428963833358768,
+      "grad_norm": 2.1041128573446035,
+      "learning_rate": 4.444444444444444e-06,
+      "loss": 0.3693,
+      "step": 44
+    },
+    {
+      "epoch": 0.013734167556844194,
+      "grad_norm": 2.286990324679626,
+      "learning_rate": 4.5454545454545455e-06,
+      "loss": 0.2513,
+      "step": 45
+    },
+    {
+      "epoch": 0.01403937128032962,
+      "grad_norm": 8.793466778432636,
+      "learning_rate": 4.646464646464647e-06,
+      "loss": 0.4343,
+      "step": 46
+    },
+    {
+      "epoch": 0.014344575003815047,
+      "grad_norm": 1.8648737533834159,
+      "learning_rate": 4.747474747474748e-06,
+      "loss": 0.2631,
+      "step": 47
+    },
+    {
+      "epoch": 0.014649778727300474,
+      "grad_norm": 2.3081781364995324,
+      "learning_rate": 4.848484848484849e-06,
+      "loss": 0.2755,
+      "step": 48
+    },
+    {
+      "epoch": 0.0149549824507859,
+      "grad_norm": 2.284005369243557,
+      "learning_rate": 4.94949494949495e-06,
+      "loss": 0.4186,
+      "step": 49
+    },
+    {
+      "epoch": 0.015260186174271327,
+      "grad_norm": 2.6759709423238096,
+      "learning_rate": 5.0505050505050515e-06,
+      "loss": 0.6459,
+      "step": 50
+    },
+    {
+      "epoch": 0.015565389897756753,
+      "grad_norm": 2.8773749120652523,
+      "learning_rate": 5.151515151515152e-06,
+      "loss": 0.3324,
+      "step": 51
+    },
+    {
+      "epoch": 0.015870593621242178,
+      "grad_norm": 2.8060164424498786,
+      "learning_rate": 5.252525252525253e-06,
+      "loss": 0.3608,
+      "step": 52
+    },
+    {
+      "epoch": 0.016175797344727606,
+      "grad_norm": 2.3060494229726793,
+      "learning_rate": 5.353535353535354e-06,
+      "loss": 0.3818,
+      "step": 53
+    },
+    {
+      "epoch": 0.01648100106821303,
+      "grad_norm": 2.073464811557714,
+      "learning_rate": 5.4545454545454545e-06,
+      "loss": 0.2667,
+      "step": 54
+    },
+    {
+      "epoch": 0.01678620479169846,
+      "grad_norm": 2.3474749655399245,
+      "learning_rate": 5.555555555555557e-06,
+      "loss": 0.35,
+      "step": 55
+    },
+    {
+      "epoch": 0.017091408515183884,
+      "grad_norm": 3.6988890036672086,
+      "learning_rate": 5.656565656565657e-06,
+      "loss": 0.284,
+      "step": 56
+    },
+    {
+      "epoch": 0.017396612238669312,
+      "grad_norm": 2.313501192849839,
+      "learning_rate": 5.7575757575757586e-06,
+      "loss": 0.3308,
+      "step": 57
+    },
+    {
+      "epoch": 0.017701815962154737,
+      "grad_norm": 2.411936098122121,
+      "learning_rate": 5.858585858585859e-06,
+      "loss": 0.3982,
+      "step": 58
+    },
+    {
+      "epoch": 0.018007019685640165,
+      "grad_norm": 2.724660127775508,
+      "learning_rate": 5.95959595959596e-06,
+      "loss": 0.3587,
+      "step": 59
+    },
+    {
+      "epoch": 0.01831222340912559,
+      "grad_norm": 3.130895013540925,
+      "learning_rate": 6.060606060606061e-06,
+      "loss": 0.3427,
+      "step": 60
+    },
+    {
+      "epoch": 0.01861742713261102,
+      "grad_norm": 3.4261489723004614,
+      "learning_rate": 6.1616161616161615e-06,
+      "loss": 0.4578,
+      "step": 61
+    },
+    {
+      "epoch": 0.018922630856096443,
+      "grad_norm": 2.413871881063889,
+      "learning_rate": 6.262626262626264e-06,
+      "loss": 0.2067,
+      "step": 62
+    },
+    {
+      "epoch": 0.01922783457958187,
+      "grad_norm": 2.0941348505038366,
+      "learning_rate": 6.363636363636364e-06,
+      "loss": 0.27,
+      "step": 63
+    },
+    {
+      "epoch": 0.019533038303067296,
+      "grad_norm": 2.2153240133926153,
+      "learning_rate": 6.464646464646466e-06,
+      "loss": 0.3298,
+      "step": 64
+    },
+    {
+      "epoch": 0.019838242026552724,
+      "grad_norm": 2.422022070572305,
+      "learning_rate": 6.565656565656566e-06,
+      "loss": 0.4894,
+      "step": 65
+    },
+    {
+      "epoch": 0.02014344575003815,
+      "grad_norm": 2.45442660843552,
+      "learning_rate": 6.666666666666667e-06,
+      "loss": 0.3684,
+      "step": 66
+    },
+    {
+      "epoch": 0.020448649473523577,
+      "grad_norm": 3.5398238081108304,
+      "learning_rate": 6.767676767676769e-06,
+      "loss": 0.4233,
+      "step": 67
+    },
+    {
+      "epoch": 0.020753853197009002,
+      "grad_norm": 2.530397719080883,
+      "learning_rate": 6.868686868686869e-06,
+      "loss": 0.2676,
+      "step": 68
+    },
+    {
+      "epoch": 0.02105905692049443,
+      "grad_norm": 2.259346305696615,
+      "learning_rate": 6.969696969696971e-06,
+      "loss": 0.4409,
+      "step": 69
+    },
+    {
+      "epoch": 0.021364260643979855,
+      "grad_norm": 2.3339543424453764,
+      "learning_rate": 7.070707070707071e-06,
+      "loss": 0.3882,
+      "step": 70
+    },
+    {
+      "epoch": 0.021669464367465283,
+      "grad_norm": 2.348843038116063,
+      "learning_rate": 7.171717171717172e-06,
+      "loss": 0.3904,
+      "step": 71
+    },
+    {
+      "epoch": 0.021974668090950708,
+      "grad_norm": 2.7011363922899965,
+      "learning_rate": 7.272727272727273e-06,
+      "loss": 0.3586,
+      "step": 72
+    },
+    {
+      "epoch": 0.022279871814436136,
+      "grad_norm": 2.6923381814173486,
+      "learning_rate": 7.373737373737374e-06,
+      "loss": 0.4331,
+      "step": 73
+    },
+    {
+      "epoch": 0.02258507553792156,
+      "grad_norm": 2.0435337430530924,
+      "learning_rate": 7.474747474747476e-06,
+      "loss": 0.2739,
+      "step": 74
+    },
+    {
+      "epoch": 0.02289027926140699,
+      "grad_norm": 2.257183264462076,
+      "learning_rate": 7.5757575757575764e-06,
+      "loss": 0.4554,
+      "step": 75
+    },
+    {
+      "epoch": 0.023195482984892414,
+      "grad_norm": 2.5384248372961626,
+      "learning_rate": 7.676767676767677e-06,
+      "loss": 0.4934,
+      "step": 76
+    },
+    {
+      "epoch": 0.023500686708377842,
+      "grad_norm": 2.1578730127908488,
+      "learning_rate": 7.77777777777778e-06,
+      "loss": 0.3519,
+      "step": 77
+    },
+    {
+      "epoch": 0.023805890431863267,
+      "grad_norm": 2.1316764516757476,
+      "learning_rate": 7.87878787878788e-06,
+      "loss": 0.3268,
+      "step": 78
+    },
+    {
+      "epoch": 0.024111094155348695,
+      "grad_norm": 2.095996278024237,
+      "learning_rate": 7.97979797979798e-06,
+      "loss": 0.3318,
+      "step": 79
+    },
+    {
+      "epoch": 0.02441629787883412,
+      "grad_norm": 1.9985574049541877,
+      "learning_rate": 8.08080808080808e-06,
+      "loss": 0.1852,
+      "step": 80
+    },
+    {
+      "epoch": 0.02472150160231955,
+      "grad_norm": 1.7092921737326583,
+      "learning_rate": 8.181818181818183e-06,
+      "loss": 0.2412,
+      "step": 81
+    },
+    {
+      "epoch": 0.025026705325804977,
+      "grad_norm": 1.9609482601524066,
+      "learning_rate": 8.282828282828283e-06,
+      "loss": 0.3349,
+      "step": 82
+    },
+    {
+      "epoch": 0.0253319090492904,
+      "grad_norm": 2.5619254980161412,
+      "learning_rate": 8.383838383838384e-06,
+      "loss": 0.3327,
+      "step": 83
+    },
+    {
+      "epoch": 0.02563711277277583,
+      "grad_norm": 2.1734116421771827,
+      "learning_rate": 8.484848484848486e-06,
+      "loss": 0.5005,
+      "step": 84
+    },
+    {
+      "epoch": 0.025942316496261254,
+      "grad_norm": 2.4612836321871785,
+      "learning_rate": 8.585858585858587e-06,
+      "loss": 0.5919,
+      "step": 85
+    },
+    {
+      "epoch": 0.026247520219746683,
+      "grad_norm": 2.050264187978962,
+      "learning_rate": 8.686868686868687e-06,
+      "loss": 0.2654,
+      "step": 86
+    },
+    {
+      "epoch": 0.026552723943232107,
+      "grad_norm": 1.7466792206761999,
+      "learning_rate": 8.787878787878788e-06,
+      "loss": 0.2875,
+      "step": 87
+    },
+    {
+      "epoch": 0.026857927666717536,
+      "grad_norm": 1.9114055019911376,
+      "learning_rate": 8.888888888888888e-06,
+      "loss": 0.3317,
+      "step": 88
+    },
+    {
+      "epoch": 0.02716313139020296,
+      "grad_norm": 2.136028617695754,
+      "learning_rate": 8.98989898989899e-06,
+      "loss": 0.4322,
+      "step": 89
+    },
+    {
+      "epoch": 0.02746833511368839,
+      "grad_norm": 2.0559196693817303,
+      "learning_rate": 9.090909090909091e-06,
+      "loss": 0.3372,
+      "step": 90
+    },
+    {
+      "epoch": 0.027773538837173813,
+      "grad_norm": 1.6053810559753854,
+      "learning_rate": 9.191919191919193e-06,
+      "loss": 0.2833,
+      "step": 91
+    },
+    {
+      "epoch": 0.02807874256065924,
+      "grad_norm": 1.9190338968500587,
+      "learning_rate": 9.292929292929294e-06,
+      "loss": 0.2358,
+      "step": 92
+    },
+    {
+      "epoch": 0.028383946284144666,
+      "grad_norm": 1.7424429804531956,
+      "learning_rate": 9.393939393939396e-06,
+      "loss": 0.2805,
+      "step": 93
+    },
+    {
+      "epoch": 0.028689150007630095,
+      "grad_norm": 1.5616301594921251,
+      "learning_rate": 9.494949494949497e-06,
+      "loss": 0.326,
+      "step": 94
+    },
+    {
+      "epoch": 0.02899435373111552,
+      "grad_norm": 2.6517363851490297,
+      "learning_rate": 9.595959595959597e-06,
+      "loss": 0.5839,
+      "step": 95
+    },
+    {
+      "epoch": 0.029299557454600948,
+      "grad_norm": 1.9068377479857994,
+      "learning_rate": 9.696969696969698e-06,
+      "loss": 0.4213,
+      "step": 96
+    },
+    {
+      "epoch": 0.029604761178086372,
+      "grad_norm": 2.147263972819766,
+      "learning_rate": 9.797979797979798e-06,
+      "loss": 0.3776,
+      "step": 97
+    },
+    {
+      "epoch": 0.0299099649015718,
+      "grad_norm": 2.3466004395170685,
+      "learning_rate": 9.8989898989899e-06,
+      "loss": 0.4828,
+      "step": 98
+    },
+    {
+      "epoch": 0.030215168625057225,
+      "grad_norm": 1.9328188798162316,
+      "learning_rate": 1e-05,
+      "loss": 0.3816,
+      "step": 99
+    },
+    {
+      "epoch": 0.030520372348542654,
+      "grad_norm": 2.120656679761712,
+      "learning_rate": 9.999997555414177e-06,
+      "loss": 0.287,
+      "step": 100
+    },
+    {
+      "epoch": 0.03082557607202808,
+      "grad_norm": 1.8272767014289886,
+      "learning_rate": 9.999990221659095e-06,
+      "loss": 0.2529,
+      "step": 101
+    },
+    {
+      "epoch": 0.031130779795513507,
+      "grad_norm": 2.108876035097533,
+      "learning_rate": 9.999977998741925e-06,
+      "loss": 0.4,
+      "step": 102
+    },
+    {
+      "epoch": 0.031435983518998935,
+      "grad_norm": 2.611227326027621,
+      "learning_rate": 9.999960886674623e-06,
+      "loss": 0.5577,
+      "step": 103
+    },
+    {
+      "epoch": 0.031741187242484356,
+      "grad_norm": 2.012760226088087,
+      "learning_rate": 9.999938885473916e-06,
+      "loss": 0.2397,
+      "step": 104
+    },
+    {
+      "epoch": 0.032046390965969784,
+      "grad_norm": 3.4069313977643088,
+      "learning_rate": 9.999911995161323e-06,
+      "loss": 0.3074,
+      "step": 105
+    },
+    {
+      "epoch": 0.03235159468945521,
+      "grad_norm": 1.5281487804348939,
+      "learning_rate": 9.999880215763133e-06,
+      "loss": 0.306,
+      "step": 106
+    },
+    {
+      "epoch": 0.03265679841294064,
+      "grad_norm": 1.5733903167529437,
+      "learning_rate": 9.999843547310427e-06,
+      "loss": 0.3123,
+      "step": 107
+    },
+    {
+      "epoch": 0.03296200213642606,
+      "grad_norm": 2.2084260837102776,
+      "learning_rate": 9.999801989839055e-06,
+      "loss": 0.2686,
+      "step": 108
+    },
+    {
+      "epoch": 0.03326720585991149,
+      "grad_norm": 2.0235527329790477,
+      "learning_rate": 9.999755543389658e-06,
+      "loss": 0.362,
+      "step": 109
+    },
+    {
+      "epoch": 0.03357240958339692,
+      "grad_norm": 1.4126246608311444,
+      "learning_rate": 9.999704208007647e-06,
+      "loss": 0.1868,
+      "step": 110
+    },
+    {
+      "epoch": 0.03387761330688235,
+      "grad_norm": 1.9363750145032863,
+      "learning_rate": 9.999647983743227e-06,
+      "loss": 0.4674,
+      "step": 111
+    },
+    {
+      "epoch": 0.03418281703036777,
+      "grad_norm": 2.306492812857686,
+      "learning_rate": 9.999586870651372e-06,
+      "loss": 0.7454,
+      "step": 112
+    },
+    {
+      "epoch": 0.034488020753853196,
+      "grad_norm": 1.9927578577114744,
+      "learning_rate": 9.999520868791839e-06,
+      "loss": 0.2964,
+      "step": 113
+    },
+    {
+      "epoch": 0.034793224477338625,
+      "grad_norm": 2.897230200199283,
+      "learning_rate": 9.99944997822917e-06,
+      "loss": 0.3507,
+      "step": 114
+    },
+    {
+      "epoch": 0.03509842820082405,
+      "grad_norm": 1.7040567211820554,
+      "learning_rate": 9.999374199032682e-06,
+      "loss": 0.358,
+      "step": 115
+    },
+    {
+      "epoch": 0.035403631924309474,
+      "grad_norm": 1.7684725864001616,
+      "learning_rate": 9.999293531276475e-06,
+      "loss": 0.469,
+      "step": 116
+    },
+    {
+      "epoch": 0.0357088356477949,
+      "grad_norm": 2.151331613378997,
+      "learning_rate": 9.999207975039429e-06,
+      "loss": 0.4007,
+      "step": 117
+    },
+    {
+      "epoch": 0.03601403937128033,
+      "grad_norm": 2.1827006415812678,
+      "learning_rate": 9.999117530405205e-06,
+      "loss": 0.373,
+      "step": 118
+    },
+    {
+      "epoch": 0.03631924309476576,
+      "grad_norm": 2.0424756244526283,
+      "learning_rate": 9.99902219746224e-06,
+      "loss": 0.4664,
+      "step": 119
+    },
+    {
+      "epoch": 0.03662444681825118,
+      "grad_norm": 2.4438750213097014,
+      "learning_rate": 9.998921976303757e-06,
+      "loss": 0.5884,
+      "step": 120
+    },
+    {
+      "epoch": 0.03692965054173661,
+      "grad_norm": 1.6168805259489245,
+      "learning_rate": 9.998816867027753e-06,
+      "loss": 0.3874,
+      "step": 121
+    },
+    {
+      "epoch": 0.03723485426522204,
+      "grad_norm": 2.4836564854380914,
+      "learning_rate": 9.99870686973701e-06,
+      "loss": 0.3865,
+      "step": 122
+    },
+    {
+      "epoch": 0.037540057988707465,
+      "grad_norm": 2.187549263535683,
+      "learning_rate": 9.998591984539085e-06,
+      "loss": 0.4419,
+      "step": 123
+    },
+    {
+      "epoch": 0.037845261712192886,
+      "grad_norm": 2.3145724108896366,
+      "learning_rate": 9.998472211546317e-06,
+      "loss": 0.5048,
+      "step": 124
+    },
+    {
+      "epoch": 0.038150465435678314,
+      "grad_norm": 2.6043824271784377,
+      "learning_rate": 9.998347550875825e-06,
+      "loss": 0.4323,
+      "step": 125
+    },
+    {
+      "epoch": 0.03845566915916374,
+      "grad_norm": 1.7266964407358079,
+      "learning_rate": 9.998218002649507e-06,
+      "loss": 0.3093,
+      "step": 126
+    },
+    {
+      "epoch": 0.03876087288264917,
+      "grad_norm": 2.3091863655820397,
+      "learning_rate": 9.99808356699404e-06,
+      "loss": 0.5394,
+      "step": 127
+    },
+    {
+      "epoch": 0.03906607660613459,
+      "grad_norm": 2.178584103245907,
+      "learning_rate": 9.997944244040877e-06,
+      "loss": 0.562,
+      "step": 128
+    },
+    {
+      "epoch": 0.03937128032962002,
+      "grad_norm": 1.4762803065381216,
+      "learning_rate": 9.997800033926252e-06,
+      "loss": 0.3012,
+      "step": 129
+    },
+    {
+      "epoch": 0.03967648405310545,
+      "grad_norm": 1.6768704233807339,
+      "learning_rate": 9.997650936791183e-06,
+      "loss": 0.3314,
+      "step": 130
+    },
+    {
+      "epoch": 0.03998168777659088,
+      "grad_norm": 1.8423584681568375,
+      "learning_rate": 9.997496952781461e-06,
+      "loss": 0.5373,
+      "step": 131
+    },
+    {
+      "epoch": 0.0402868915000763,
+      "grad_norm": 1.4926628434179245,
+      "learning_rate": 9.997338082047656e-06,
+      "loss": 0.1992,
+      "step": 132
+    },
+    {
+      "epoch": 0.040592095223561726,
+      "grad_norm": 1.6323074947028773,
+      "learning_rate": 9.997174324745117e-06,
+      "loss": 0.4872,
+      "step": 133
+    },
+    {
+      "epoch": 0.040897298947047155,
+      "grad_norm": 2.159688005520465,
+      "learning_rate": 9.997005681033973e-06,
+      "loss": 0.5076,
+      "step": 134
+    },
+    {
+      "epoch": 0.04120250267053258,
+      "grad_norm": 2.207163038792008,
+      "learning_rate": 9.996832151079127e-06,
+      "loss": 0.2677,
+      "step": 135
+    },
+    {
+      "epoch": 0.041507706394018004,
+      "grad_norm": 1.3990677420334965,
+      "learning_rate": 9.996653735050265e-06,
+      "loss": 0.2526,
+      "step": 136
+    },
+    {
+      "epoch": 0.04181291011750343,
+      "grad_norm": 1.7368886105229604,
+      "learning_rate": 9.996470433121847e-06,
+      "loss": 0.2874,
+      "step": 137
+    },
+    {
+      "epoch": 0.04211811384098886,
+      "grad_norm": 1.8138446424045762,
+      "learning_rate": 9.996282245473113e-06,
+      "loss": 0.2986,
+      "step": 138
+    },
+    {
+      "epoch": 0.04242331756447429,
+      "grad_norm": 1.8564789601928355,
+      "learning_rate": 9.996089172288078e-06,
+      "loss": 0.3954,
+      "step": 139
+    },
+    {
+      "epoch": 0.04272852128795971,
+      "grad_norm": 1.9085920361180522,
+      "learning_rate": 9.995891213755536e-06,
+      "loss": 0.2739,
+      "step": 140
+    },
+    {
+      "epoch": 0.04303372501144514,
+      "grad_norm": 1.8924678931794556,
+      "learning_rate": 9.99568837006906e-06,
+      "loss": 0.2766,
+      "step": 141
+    },
+    {
+      "epoch": 0.04333892873493057,
+      "grad_norm": 1.8418836037208652,
+      "learning_rate": 9.995480641426992e-06,
+      "loss": 0.488,
+      "step": 142
+    },
+    {
+      "epoch": 0.043644132458415995,
+      "grad_norm": 1.6305125707231247,
+      "learning_rate": 9.99526802803246e-06,
+      "loss": 0.3045,
+      "step": 143
+    },
+    {
+      "epoch": 0.043949336181901416,
+      "grad_norm": 2.143051665423358,
+      "learning_rate": 9.995050530093366e-06,
+      "loss": 0.3567,
+      "step": 144
+    },
+    {
+      "epoch": 0.044254539905386844,
+      "grad_norm": 1.994194545633334,
+      "learning_rate": 9.994828147822387e-06,
+      "loss": 0.3655,
+      "step": 145
+    },
+    {
+      "epoch": 0.04455974362887227,
+      "grad_norm": 1.8553346605537173,
+      "learning_rate": 9.994600881436972e-06,
+      "loss": 0.3249,
+      "step": 146
+    },
+    {
+      "epoch": 0.0448649473523577,
+      "grad_norm": 2.1613773805709857,
+      "learning_rate": 9.994368731159351e-06,
+      "loss": 0.4863,
+      "step": 147
+    },
+    {
+      "epoch": 0.04517015107584312,
+      "grad_norm": 2.199571706523493,
+      "learning_rate": 9.99413169721653e-06,
+      "loss": 0.465,
+      "step": 148
+    },
+    {
+      "epoch": 0.04547535479932855,
+      "grad_norm": 1.681707967900651,
+      "learning_rate": 9.99388977984029e-06,
+      "loss": 0.3472,
+      "step": 149
+    },
+    {
+      "epoch": 0.04578055852281398,
+      "grad_norm": 1.6586587053140593,
+      "learning_rate": 9.993642979267184e-06,
+      "loss": 0.3626,
+      "step": 150
+    },
+    {
+      "epoch": 0.04608576224629941,
+      "grad_norm": 2.12592721793332,
+      "learning_rate": 9.993391295738542e-06,
+      "loss": 0.3218,
+      "step": 151
+    },
+    {
+      "epoch": 0.04639096596978483,
+      "grad_norm": 1.6765944279655143,
+      "learning_rate": 9.99313472950047e-06,
+      "loss": 0.3402,
+      "step": 152
+    },
+    {
+      "epoch": 0.046696169693270256,
+      "grad_norm": 1.6019038139070678,
+      "learning_rate": 9.992873280803848e-06,
+      "loss": 0.4554,
+      "step": 153
+    },
+    {
+      "epoch": 0.047001373416755685,
+      "grad_norm": 1.6429860881882794,
+      "learning_rate": 9.99260694990433e-06,
+      "loss": 0.4086,
+      "step": 154
+    },
+    {
+      "epoch": 0.04730657714024111,
+      "grad_norm": 1.98592334325083,
+      "learning_rate": 9.992335737062338e-06,
+      "loss": 0.5733,
+      "step": 155
+    },
+    {
+      "epoch": 0.047611780863726534,
+      "grad_norm": 1.5624846648417388,
+      "learning_rate": 9.992059642543076e-06,
+      "loss": 0.2524,
+      "step": 156
+    },
+    {
+      "epoch": 0.04791698458721196,
+      "grad_norm": 1.4438198320418865,
+      "learning_rate": 9.991778666616523e-06,
+      "loss": 0.1756,
+      "step": 157
+    },
+    {
+      "epoch": 0.04822218831069739,
+      "grad_norm": 1.6284817295660008,
+      "learning_rate": 9.991492809557424e-06,
+      "loss": 0.4144,
+      "step": 158
+    },
+    {
+      "epoch": 0.04852739203418282,
+      "grad_norm": 1.2236340789910145,
+      "learning_rate": 9.991202071645298e-06,
+      "loss": 0.1664,
+      "step": 159
+    },
+    {
+      "epoch": 0.04883259575766824,
+      "grad_norm": 1.4874398163232816,
+      "learning_rate": 9.99090645316444e-06,
+      "loss": 0.3323,
+      "step": 160
+    },
+    {
+      "epoch": 0.04913779948115367,
+      "grad_norm": 2.5394515927833403,
+      "learning_rate": 9.990605954403917e-06,
+      "loss": 0.27,
+      "step": 161
+    },
+    {
+      "epoch": 0.0494430032046391,
+      "grad_norm": 1.7966332314422868,
+      "learning_rate": 9.990300575657565e-06,
+      "loss": 0.4453,
+      "step": 162
+    },
+    {
+      "epoch": 0.049748206928124525,
+      "grad_norm": 1.825976682624809,
+      "learning_rate": 9.989990317223995e-06,
+      "loss": 0.2646,
+      "step": 163
+    },
+    {
+      "epoch": 0.05005341065160995,
+      "grad_norm": 1.6554541925183588,
+      "learning_rate": 9.989675179406588e-06,
+      "loss": 0.445,
+      "step": 164
+    },
+    {
+      "epoch": 0.050358614375095374,
+      "grad_norm": 1.6711133844293076,
+      "learning_rate": 9.989355162513496e-06,
+      "loss": 0.3685,
+      "step": 165
+    },
+    {
+      "epoch": 0.0506638180985808,
+      "grad_norm": 1.8033315345252203,
+      "learning_rate": 9.989030266857644e-06,
+      "loss": 0.2566,
+      "step": 166
+    },
+    {
+      "epoch": 0.05096902182206623,
+      "grad_norm": 1.6879852444966537,
+      "learning_rate": 9.988700492756726e-06,
+      "loss": 0.4086,
+      "step": 167
+    },
+    {
+      "epoch": 0.05127422554555166,
+      "grad_norm": 1.6855038740169574,
+      "learning_rate": 9.988365840533204e-06,
+      "loss": 0.3081,
+      "step": 168
+    },
+    {
+      "epoch": 0.05157942926903708,
+      "grad_norm": 2.245121010490438,
+      "learning_rate": 9.988026310514316e-06,
+      "loss": 0.5646,
+      "step": 169
+    },
+    {
+      "epoch": 0.05188463299252251,
+      "grad_norm": 1.531117336209479,
+      "learning_rate": 9.987681903032065e-06,
+      "loss": 0.3598,
+      "step": 170
+    },
+    {
+      "epoch": 0.05218983671600794,
+      "grad_norm": 1.4368727600956301,
+      "learning_rate": 9.987332618423221e-06,
+      "loss": 0.3864,
+      "step": 171
+    },
+    {
+      "epoch": 0.052495040439493365,
+      "grad_norm": 2.039026486601271,
+      "learning_rate": 9.98697845702933e-06,
+      "loss": 0.2728,
+      "step": 172
+    },
+    {
+      "epoch": 0.052800244162978786,
+      "grad_norm": 1.5481974795842472,
+      "learning_rate": 9.986619419196704e-06,
+      "loss": 0.2376,
+      "step": 173
+    },
+    {
+      "epoch": 0.053105447886464215,
+      "grad_norm": 1.583025735121783,
+      "learning_rate": 9.986255505276418e-06,
+      "loss": 0.3941,
+      "step": 174
+    },
+    {
+      "epoch": 0.05341065160994964,
+      "grad_norm": 2.025610033619695,
+      "learning_rate": 9.985886715624326e-06,
+      "loss": 0.432,
+      "step": 175
+    },
+    {
+      "epoch": 0.05371585533343507,
+      "grad_norm": 1.9370365819159912,
+      "learning_rate": 9.985513050601037e-06,
+      "loss": 0.3311,
+      "step": 176
+    },
+    {
+      "epoch": 0.05402105905692049,
+      "grad_norm": 1.534591376747653,
+      "learning_rate": 9.985134510571936e-06,
+      "loss": 0.3804,
+      "step": 177
+    },
+    {
+      "epoch": 0.05432626278040592,
+      "grad_norm": 1.5627980520171343,
+      "learning_rate": 9.984751095907175e-06,
+      "loss": 0.3991,
+      "step": 178
+    },
+    {
+      "epoch": 0.05463146650389135,
+      "grad_norm": 1.858760828475349,
+      "learning_rate": 9.984362806981665e-06,
+      "loss": 0.4124,
+      "step": 179
+    },
+    {
+      "epoch": 0.05493667022737678,
+      "grad_norm": 1.4922057145689682,
+      "learning_rate": 9.983969644175092e-06,
+      "loss": 0.2571,
+      "step": 180
+    },
+    {
+      "epoch": 0.0552418739508622,
+      "grad_norm": 1.4358215484460224,
+      "learning_rate": 9.983571607871903e-06,
+      "loss": 0.3351,
+      "step": 181
+    },
+    {
+      "epoch": 0.05554707767434763,
+      "grad_norm": 1.7105120125454414,
+      "learning_rate": 9.983168698461312e-06,
+      "loss": 0.4374,
+      "step": 182
+    },
+    {
+      "epoch": 0.055852281397833055,
+      "grad_norm": 1.4100459259074987,
+      "learning_rate": 9.982760916337296e-06,
+      "loss": 0.3958,
+      "step": 183
+    },
+    {
+      "epoch": 0.05615748512131848,
+      "grad_norm": 1.667173817085955,
+      "learning_rate": 9.982348261898598e-06,
+      "loss": 0.2867,
+      "step": 184
+    },
+    {
+      "epoch": 0.056462688844803904,
+      "grad_norm": 1.8278737995984025,
+      "learning_rate": 9.981930735548731e-06,
+      "loss": 0.3738,
+      "step": 185
+    },
+    {
+      "epoch": 0.05676789256828933,
+      "grad_norm": 1.806852289121097,
+      "learning_rate": 9.98150833769596e-06,
+      "loss": 0.5608,
+      "step": 186
+    },
+    {
+      "epoch": 0.05707309629177476,
+      "grad_norm": 1.6986308867720055,
+      "learning_rate": 9.981081068753324e-06,
+      "loss": 0.4253,
+      "step": 187
+    },
+    {
+      "epoch": 0.05737830001526019,
+      "grad_norm": 1.6392088091109513,
+      "learning_rate": 9.98064892913862e-06,
+      "loss": 0.2444,
+      "step": 188
+    },
+    {
+      "epoch": 0.05768350373874561,
+      "grad_norm": 1.7762995408711126,
+      "learning_rate": 9.980211919274407e-06,
+      "loss": 0.3866,
+      "step": 189
+    },
+    {
+      "epoch": 0.05798870746223104,
+      "grad_norm": 1.7144647062044762,
+      "learning_rate": 9.979770039588013e-06,
+      "loss": 0.4504,
+      "step": 190
+    },
+    {
+      "epoch": 0.05829391118571647,
+      "grad_norm": 1.9069269572943617,
+      "learning_rate": 9.979323290511517e-06,
+      "loss": 0.4972,
+      "step": 191
+    },
+    {
+      "epoch": 0.058599114909201895,
+      "grad_norm": 1.831943664409223,
+      "learning_rate": 9.978871672481774e-06,
+      "loss": 0.3884,
+      "step": 192
+    },
+    {
+      "epoch": 0.058904318632687316,
+      "grad_norm": 1.60483584957947,
+      "learning_rate": 9.978415185940383e-06,
+      "loss": 0.3366,
+      "step": 193
+    },
+    {
+      "epoch": 0.059209522356172745,
+      "grad_norm": 2.041633475935638,
+      "learning_rate": 9.977953831333718e-06,
+      "loss": 0.4928,
+      "step": 194
+    },
+    {
+      "epoch": 0.05951472607965817,
+      "grad_norm": 2.1574861604284243,
+      "learning_rate": 9.977487609112904e-06,
+      "loss": 0.7092,
+      "step": 195
+    },
+    {
+      "epoch": 0.0598199298031436,
+      "grad_norm": 1.5382345073334531,
+      "learning_rate": 9.97701651973383e-06,
+      "loss": 0.2236,
+      "step": 196
+    },
+    {
+      "epoch": 0.06012513352662902,
+      "grad_norm": 2.1479787995768014,
+      "learning_rate": 9.976540563657143e-06,
+      "loss": 0.5182,
+      "step": 197
+    },
+    {
+      "epoch": 0.06043033725011445,
+      "grad_norm": 1.8579437774142544,
+      "learning_rate": 9.976059741348252e-06,
+      "loss": 0.3093,
+      "step": 198
+    },
+    {
+      "epoch": 0.06073554097359988,
+      "grad_norm": 1.5409701380525285,
+      "learning_rate": 9.975574053277317e-06,
+      "loss": 0.2877,
+      "step": 199
+    },
+    {
+      "epoch": 0.06104074469708531,
+      "grad_norm": 1.5474598097011698,
+      "learning_rate": 9.975083499919264e-06,
+      "loss": 0.2981,
+      "step": 200
+    },
+    {
+      "epoch": 0.06134594842057073,
+      "grad_norm": 1.9202152932180157,
+      "learning_rate": 9.974588081753773e-06,
+      "loss": 0.5369,
+      "step": 201
+    },
+    {
+      "epoch": 0.06165115214405616,
+      "grad_norm": 1.4598442515817716,
+      "learning_rate": 9.974087799265279e-06,
+      "loss": 0.3696,
+      "step": 202
+    },
+    {
+      "epoch": 0.061956355867541585,
+      "grad_norm": 1.48078814360119,
+      "learning_rate": 9.973582652942975e-06,
+      "loss": 0.284,
+      "step": 203
+    },
+    {
+      "epoch": 0.06226155959102701,
+      "grad_norm": 2.100326004155181,
+      "learning_rate": 9.973072643280813e-06,
+      "loss": 0.5681,
+      "step": 204
+    },
+    {
+      "epoch": 0.06256676331451244,
+      "grad_norm": 1.976128330719915,
+      "learning_rate": 9.972557770777496e-06,
+      "loss": 0.3655,
+      "step": 205
+    },
+    {
+      "epoch": 0.06287196703799787,
+      "grad_norm": 1.2103730393566896,
+      "learning_rate": 9.972038035936483e-06,
+      "loss": 0.2471,
+      "step": 206
+    },
+    {
+      "epoch": 0.06317717076148328,
+      "grad_norm": 1.670449906238349,
+      "learning_rate": 9.971513439265992e-06,
+      "loss": 0.2184,
+      "step": 207
+    },
+    {
+      "epoch": 0.06348237448496871,
+      "grad_norm": 1.5020544764497652,
+      "learning_rate": 9.970983981278989e-06,
+      "loss": 0.3196,
+      "step": 208
+    },
+    {
+      "epoch": 0.06378757820845414,
+      "grad_norm": 1.7833251911345853,
+      "learning_rate": 9.970449662493195e-06,
+      "loss": 0.4122,
+      "step": 209
+    },
+    {
+      "epoch": 0.06409278193193957,
+      "grad_norm": 1.4149595334362772,
+      "learning_rate": 9.96991048343109e-06,
+      "loss": 0.2947,
+      "step": 210
+    },
+    {
+      "epoch": 0.064397985655425,
+      "grad_norm": 1.5991867680932033,
+      "learning_rate": 9.969366444619898e-06,
+      "loss": 0.1902,
+      "step": 211
+    },
+    {
+      "epoch": 0.06470318937891043,
+      "grad_norm": 1.4132064841734169,
+      "learning_rate": 9.968817546591601e-06,
+      "loss": 0.3389,
+      "step": 212
+    },
+    {
+      "epoch": 0.06500839310239585,
+      "grad_norm": 1.7671902900221814,
+      "learning_rate": 9.968263789882926e-06,
+      "loss": 0.4294,
+      "step": 213
+    },
+    {
+      "epoch": 0.06531359682588128,
+      "grad_norm": 1.5709821497329826,
+      "learning_rate": 9.96770517503536e-06,
+      "loss": 0.2765,
+      "step": 214
+    },
+    {
+      "epoch": 0.0656188005493667,
+      "grad_norm": 1.5211731343844295,
+      "learning_rate": 9.967141702595134e-06,
+      "loss": 0.387,
+      "step": 215
+    },
+    {
+      "epoch": 0.06592400427285212,
+      "grad_norm": 1.5499265222668686,
+      "learning_rate": 9.96657337311323e-06,
+      "loss": 0.4535,
+      "step": 216
+    },
+    {
+      "epoch": 0.06622920799633755,
+      "grad_norm": 1.4736546539447488,
+      "learning_rate": 9.966000187145383e-06,
+      "loss": 0.3834,
+      "step": 217
+    },
+    {
+      "epoch": 0.06653441171982298,
+      "grad_norm": 1.3306288958233108,
+      "learning_rate": 9.965422145252072e-06,
+      "loss": 0.3172,
+      "step": 218
+    },
+    {
+      "epoch": 0.06683961544330841,
+      "grad_norm": 1.5745937005003143,
+      "learning_rate": 9.964839247998524e-06,
+      "loss": 0.2725,
+      "step": 219
+    },
+    {
+      "epoch": 0.06714481916679384,
+      "grad_norm": 1.7546511557153388,
+      "learning_rate": 9.96425149595472e-06,
+      "loss": 0.3577,
+      "step": 220
+    },
+    {
+      "epoch": 0.06745002289027927,
+      "grad_norm": 2.0422588449754286,
+      "learning_rate": 9.96365888969538e-06,
+      "loss": 0.4976,
+      "step": 221
+    },
+    {
+      "epoch": 0.0677552266137647,
+      "grad_norm": 1.4661824124133862,
+      "learning_rate": 9.963061429799979e-06,
+      "loss": 0.3672,
+      "step": 222
+    },
+    {
+      "epoch": 0.06806043033725011,
+      "grad_norm": 2.0959067552369666,
+      "learning_rate": 9.96245911685273e-06,
+      "loss": 0.5381,
+      "step": 223
+    },
+    {
+      "epoch": 0.06836563406073554,
+      "grad_norm": 1.3296813372997014,
+      "learning_rate": 9.961851951442599e-06,
+      "loss": 0.2799,
+      "step": 224
+    },
+    {
+      "epoch": 0.06867083778422096,
+      "grad_norm": 1.7385807765114274,
+      "learning_rate": 9.96123993416329e-06,
+      "loss": 0.5183,
+      "step": 225
+    },
+    {
+      "epoch": 0.06897604150770639,
+      "grad_norm": 1.5190119701865645,
+      "learning_rate": 9.960623065613254e-06,
+      "loss": 0.4608,
+      "step": 226
+    },
+    {
+      "epoch": 0.06928124523119182,
+      "grad_norm": 1.4393894383331207,
+      "learning_rate": 9.96000134639569e-06,
+      "loss": 0.3455,
+      "step": 227
+    },
+    {
+      "epoch": 0.06958644895467725,
+      "grad_norm": 1.7132863682619555,
+      "learning_rate": 9.959374777118533e-06,
+      "loss": 0.316,
+      "step": 228
+    },
+    {
+      "epoch": 0.06989165267816268,
+      "grad_norm": 1.3227120889592454,
+      "learning_rate": 9.958743358394464e-06,
+      "loss": 0.2467,
+      "step": 229
+    },
+    {
+      "epoch": 0.0701968564016481,
+      "grad_norm": 1.5331153407144422,
+      "learning_rate": 9.95810709084091e-06,
+      "loss": 0.3138,
+      "step": 230
+    },
+    {
+      "epoch": 0.07050206012513352,
+      "grad_norm": 1.7990748995190806,
+      "learning_rate": 9.957465975080031e-06,
+      "loss": 0.4747,
+      "step": 231
+    },
+    {
+      "epoch": 0.07080726384861895,
+      "grad_norm": 1.1638981235859056,
+      "learning_rate": 9.956820011738736e-06,
+      "loss": 0.2265,
+      "step": 232
+    },
+    {
+      "epoch": 0.07111246757210438,
+      "grad_norm": 1.5739388418179414,
+      "learning_rate": 9.956169201448665e-06,
+      "loss": 0.5066,
+      "step": 233
+    },
+    {
+      "epoch": 0.0714176712955898,
+      "grad_norm": 1.6803933013620869,
+      "learning_rate": 9.955513544846205e-06,
+      "loss": 0.4415,
+      "step": 234
+    },
+    {
+      "epoch": 0.07172287501907523,
+      "grad_norm": 1.4014872110785643,
+      "learning_rate": 9.954853042572479e-06,
+      "loss": 0.3271,
+      "step": 235
+    },
+    {
+      "epoch": 0.07202807874256066,
+      "grad_norm": 1.5310222689941932,
+      "learning_rate": 9.954187695273352e-06,
+      "loss": 0.3289,
+      "step": 236
+    },
+    {
+      "epoch": 0.07233328246604609,
+      "grad_norm": 2.166268226472017,
+      "learning_rate": 9.953517503599419e-06,
+      "loss": 0.622,
+      "step": 237
+    },
+    {
+      "epoch": 0.07263848618953152,
+      "grad_norm": 2.258081862277545,
+      "learning_rate": 9.952842468206019e-06,
+      "loss": 0.5071,
+      "step": 238
+    },
+    {
+      "epoch": 0.07294368991301693,
+      "grad_norm": 1.7322119894263104,
+      "learning_rate": 9.952162589753224e-06,
+      "loss": 0.5097,
+      "step": 239
+    },
+    {
+      "epoch": 0.07324889363650236,
+      "grad_norm": 1.9966284228033864,
+      "learning_rate": 9.951477868905843e-06,
+      "loss": 0.2263,
+      "step": 240
+    },
+    {
+      "epoch": 0.07355409735998779,
+      "grad_norm": 1.6793267860774614,
+      "learning_rate": 9.95078830633342e-06,
+      "loss": 0.2065,
+      "step": 241
+    },
+    {
+      "epoch": 0.07385930108347322,
+      "grad_norm": 2.122564153881175,
+      "learning_rate": 9.95009390271023e-06,
+      "loss": 0.2665,
+      "step": 242
+    },
+    {
+      "epoch": 0.07416450480695864,
+      "grad_norm": 1.5852282963187305,
+      "learning_rate": 9.949394658715289e-06,
+      "loss": 0.4453,
+      "step": 243
+    },
+    {
+      "epoch": 0.07446970853044407,
+      "grad_norm": 1.7534712016120517,
+      "learning_rate": 9.948690575032338e-06,
+      "loss": 0.3628,
+      "step": 244
+    },
+    {
+      "epoch": 0.0747749122539295,
+      "grad_norm": 1.351810586905304,
+      "learning_rate": 9.947981652349854e-06,
+      "loss": 0.3984,
+      "step": 245
+    },
+    {
+      "epoch": 0.07508011597741493,
+      "grad_norm": 1.8377506474408298,
+      "learning_rate": 9.947267891361051e-06,
+      "loss": 0.3677,
+      "step": 246
+    },
+    {
+      "epoch": 0.07538531970090036,
+      "grad_norm": 1.4655632998364951,
+      "learning_rate": 9.946549292763865e-06,
+      "loss": 0.3516,
+      "step": 247
+    },
+    {
+      "epoch": 0.07569052342438577,
+      "grad_norm": 3.240838121636416,
+      "learning_rate": 9.945825857260967e-06,
+      "loss": 0.2627,
+      "step": 248
+    },
+    {
+      "epoch": 0.0759957271478712,
+      "grad_norm": 1.4085823215183912,
+      "learning_rate": 9.945097585559757e-06,
+      "loss": 0.2716,
+      "step": 249
+    },
+    {
+      "epoch": 0.07630093087135663,
+      "grad_norm": 1.6361471921651585,
+      "learning_rate": 9.944364478372364e-06,
+      "loss": 0.3595,
+      "step": 250
+    },
+    {
+      "epoch": 0.07660613459484206,
+      "grad_norm": 1.0912978886499554,
+      "learning_rate": 9.943626536415647e-06,
+      "loss": 0.1968,
+      "step": 251
+    },
+    {
+      "epoch": 0.07691133831832749,
+      "grad_norm": 1.9515717700893849,
+      "learning_rate": 9.942883760411188e-06,
+      "loss": 0.374,
+      "step": 252
+    },
+    {
+      "epoch": 0.07721654204181291,
+      "grad_norm": 1.5560755068838334,
+      "learning_rate": 9.942136151085302e-06,
+      "loss": 0.44,
+      "step": 253
+    },
+    {
+      "epoch": 0.07752174576529834,
+      "grad_norm": 1.4843235207715992,
+      "learning_rate": 9.941383709169024e-06,
+      "loss": 0.3175,
+      "step": 254
+    },
+    {
+      "epoch": 0.07782694948878377,
+      "grad_norm": 1.5210960196158274,
+      "learning_rate": 9.94062643539812e-06,
+      "loss": 0.3722,
+      "step": 255
+    },
+    {
+      "epoch": 0.07813215321226918,
+      "grad_norm": 1.6656094376801425,
+      "learning_rate": 9.939864330513079e-06,
+      "loss": 0.3511,
+      "step": 256
+    },
+    {
+      "epoch": 0.07843735693575461,
+      "grad_norm": 1.2732857455769802,
+      "learning_rate": 9.939097395259108e-06,
+      "loss": 0.2619,
+      "step": 257
+    },
+    {
+      "epoch": 0.07874256065924004,
+      "grad_norm": 1.8947301386622588,
+      "learning_rate": 9.938325630386149e-06,
+      "loss": 0.3933,
+      "step": 258
+    },
+    {
+      "epoch": 0.07904776438272547,
+      "grad_norm": 1.5625416559388712,
+      "learning_rate": 9.937549036648857e-06,
+      "loss": 0.4491,
+      "step": 259
+    },
+    {
+      "epoch": 0.0793529681062109,
+      "grad_norm": 1.5125179888703784,
+      "learning_rate": 9.936767614806612e-06,
+      "loss": 0.3674,
+      "step": 260
+    },
+    {
+      "epoch": 0.07965817182969633,
+      "grad_norm": 1.5026525250547669,
+      "learning_rate": 9.935981365623516e-06,
+      "loss": 0.4103,
+      "step": 261
+    },
+    {
+      "epoch": 0.07996337555318175,
+      "grad_norm": 2.3948536293362115,
+      "learning_rate": 9.93519028986839e-06,
+      "loss": 0.4009,
+      "step": 262
+    },
+    {
+      "epoch": 0.08026857927666718,
+      "grad_norm": 2.416554371647352,
+      "learning_rate": 9.934394388314775e-06,
+      "loss": 0.4265,
+      "step": 263
+    },
+    {
+      "epoch": 0.0805737830001526,
+      "grad_norm": 1.560923734953618,
+      "learning_rate": 9.933593661740933e-06,
+      "loss": 0.303,
+      "step": 264
+    },
+    {
+      "epoch": 0.08087898672363802,
+      "grad_norm": 1.6053945705234087,
+      "learning_rate": 9.932788110929837e-06,
+      "loss": 0.3295,
+      "step": 265
+    },
+    {
+      "epoch": 0.08118419044712345,
+      "grad_norm": 1.7775437462596928,
+      "learning_rate": 9.931977736669185e-06,
+      "loss": 0.2197,
+      "step": 266
+    },
+    {
+      "epoch": 0.08148939417060888,
+      "grad_norm": 1.701318325041301,
+      "learning_rate": 9.931162539751392e-06,
+      "loss": 0.3581,
+      "step": 267
+    },
+    {
+      "epoch": 0.08179459789409431,
+      "grad_norm": 1.5974548511363529,
+      "learning_rate": 9.93034252097358e-06,
+      "loss": 0.3432,
+      "step": 268
+    },
+    {
+      "epoch": 0.08209980161757974,
+      "grad_norm": 1.8669593065073864,
+      "learning_rate": 9.929517681137594e-06,
+      "loss": 0.4133,
+      "step": 269
+    },
+    {
+      "epoch": 0.08240500534106517,
+      "grad_norm": 1.4895827642408586,
+      "learning_rate": 9.928688021049991e-06,
+      "loss": 0.3111,
+      "step": 270
+    },
+    {
+      "epoch": 0.0827102090645506,
+      "grad_norm": 1.4317804244871846,
+      "learning_rate": 9.927853541522041e-06,
+      "loss": 0.2915,
+      "step": 271
+    },
+    {
+      "epoch": 0.08301541278803601,
+      "grad_norm": 1.252478145781798,
+      "learning_rate": 9.927014243369727e-06,
+      "loss": 0.2794,
+      "step": 272
+    },
+    {
+      "epoch": 0.08332061651152144,
+      "grad_norm": 1.6973954865497314,
+      "learning_rate": 9.926170127413743e-06,
+      "loss": 0.6183,
+      "step": 273
+    },
+    {
+      "epoch": 0.08362582023500686,
+      "grad_norm": 1.4723277244112698,
+      "learning_rate": 9.925321194479494e-06,
+      "loss": 0.2815,
+      "step": 274
+    },
+    {
+      "epoch": 0.08393102395849229,
+      "grad_norm": 1.7075555550514414,
+      "learning_rate": 9.924467445397097e-06,
+      "loss": 0.4178,
+      "step": 275
+    },
+    {
+      "epoch": 0.08423622768197772,
+      "grad_norm": 1.5354808046910606,
+      "learning_rate": 9.923608881001377e-06,
+      "loss": 0.2355,
+      "step": 276
+    },
+    {
+      "epoch": 0.08454143140546315,
+      "grad_norm": 1.1795750747565834,
+      "learning_rate": 9.922745502131865e-06,
+      "loss": 0.3404,
+      "step": 277
+    },
+    {
+      "epoch": 0.08484663512894858,
+      "grad_norm": 1.427067758888222,
+      "learning_rate": 9.921877309632805e-06,
+      "loss": 0.3141,
+      "step": 278
+    },
+    {
+      "epoch": 0.085151838852434,
+      "grad_norm": 1.3691564278772157,
+      "learning_rate": 9.921004304353147e-06,
+      "loss": 0.287,
+      "step": 279
+    },
+    {
+      "epoch": 0.08545704257591942,
+      "grad_norm": 1.9220775714586407,
+      "learning_rate": 9.920126487146544e-06,
+      "loss": 0.6617,
+      "step": 280
+    },
+    {
+      "epoch": 0.08576224629940485,
+      "grad_norm": 1.6761030408371134,
+      "learning_rate": 9.919243858871355e-06,
+      "loss": 0.466,
+      "step": 281
+    },
+    {
+      "epoch": 0.08606745002289028,
+      "grad_norm": 1.6120747264173168,
+      "learning_rate": 9.918356420390645e-06,
+      "loss": 0.5351,
+      "step": 282
+    },
+    {
+      "epoch": 0.0863726537463757,
+      "grad_norm": 1.5236961732014556,
+      "learning_rate": 9.91746417257218e-06,
+      "loss": 0.33,
+      "step": 283
+    },
+    {
+      "epoch": 0.08667785746986113,
+      "grad_norm": 1.6328635321860312,
+      "learning_rate": 9.916567116288434e-06,
+      "loss": 0.4301,
+      "step": 284
+    },
+    {
+      "epoch": 0.08698306119334656,
+      "grad_norm": 1.4120804188821041,
+      "learning_rate": 9.915665252416577e-06,
+      "loss": 0.3025,
+      "step": 285
+    },
+    {
+      "epoch": 0.08728826491683199,
+      "grad_norm": 1.8410843798908767,
+      "learning_rate": 9.914758581838482e-06,
+      "loss": 0.5415,
+      "step": 286
+    },
+    {
+      "epoch": 0.08759346864031742,
+      "grad_norm": 1.1807475096034001,
+      "learning_rate": 9.913847105440725e-06,
+      "loss": 0.3184,
+      "step": 287
+    },
+    {
+      "epoch": 0.08789867236380283,
+      "grad_norm": 1.52681276111022,
+      "learning_rate": 9.912930824114577e-06,
+      "loss": 0.4266,
+      "step": 288
+    },
+    {
+      "epoch": 0.08820387608728826,
+      "grad_norm": 1.4904538614169496,
+      "learning_rate": 9.91200973875601e-06,
+      "loss": 0.3404,
+      "step": 289
+    },
+    {
+      "epoch": 0.08850907981077369,
+      "grad_norm": 1.7385111110311349,
+      "learning_rate": 9.911083850265692e-06,
+      "loss": 0.3371,
+      "step": 290
+    },
+    {
+      "epoch": 0.08881428353425912,
+      "grad_norm": 1.6013762575114376,
+      "learning_rate": 9.91015315954899e-06,
+      "loss": 0.4475,
+      "step": 291
+    },
+    {
+      "epoch": 0.08911948725774455,
+      "grad_norm": 1.5474202900018152,
+      "learning_rate": 9.909217667515964e-06,
+      "loss": 0.4162,
+      "step": 292
+    },
+    {
+      "epoch": 0.08942469098122997,
+      "grad_norm": 1.875769203080621,
+      "learning_rate": 9.908277375081371e-06,
+      "loss": 0.4446,
+      "step": 293
+    },
+    {
+      "epoch": 0.0897298947047154,
+      "grad_norm": 1.4914731218024286,
+      "learning_rate": 9.907332283164663e-06,
+      "loss": 0.4274,
+      "step": 294
+    },
+    {
+      "epoch": 0.09003509842820083,
+      "grad_norm": 1.6551811079983538,
+      "learning_rate": 9.90638239268998e-06,
+      "loss": 0.4883,
+      "step": 295
+    },
+    {
+      "epoch": 0.09034030215168624,
+      "grad_norm": 1.645510927644492,
+      "learning_rate": 9.905427704586158e-06,
+      "loss": 0.4885,
+      "step": 296
+    },
+    {
+      "epoch": 0.09064550587517167,
+      "grad_norm": 1.6759165462483547,
+      "learning_rate": 9.904468219786727e-06,
+      "loss": 0.3878,
+      "step": 297
+    },
+    {
+      "epoch": 0.0909507095986571,
+      "grad_norm": 1.596800484010474,
+      "learning_rate": 9.903503939229901e-06,
+      "loss": 0.2725,
+      "step": 298
+    },
+    {
+      "epoch": 0.09125591332214253,
+      "grad_norm": 1.4035704196730787,
+      "learning_rate": 9.902534863858588e-06,
+      "loss": 0.2147,
+      "step": 299
+    },
+    {
+      "epoch": 0.09156111704562796,
+      "grad_norm": 1.7460761357385464,
+      "learning_rate": 9.90156099462038e-06,
+      "loss": 0.3495,
+      "step": 300
+    },
+    {
+      "epoch": 0.09186632076911339,
+      "grad_norm": 1.3373562156184522,
+      "learning_rate": 9.900582332467566e-06,
+      "loss": 0.342,
+      "step": 301
+    },
+    {
+      "epoch": 0.09217152449259881,
+      "grad_norm": 1.1466755748188362,
+      "learning_rate": 9.89959887835711e-06,
+      "loss": 0.1737,
+      "step": 302
+    },
+    {
+      "epoch": 0.09247672821608424,
+      "grad_norm": 1.8078659273922337,
+      "learning_rate": 9.898610633250669e-06,
+      "loss": 0.3111,
+      "step": 303
+    },
+    {
+      "epoch": 0.09278193193956966,
+      "grad_norm": 1.5400638324339648,
+      "learning_rate": 9.897617598114584e-06,
+      "loss": 0.4746,
+      "step": 304
+    },
+    {
+      "epoch": 0.09308713566305508,
+      "grad_norm": 1.558728128630052,
+      "learning_rate": 9.896619773919878e-06,
+      "loss": 0.3085,
+      "step": 305
+    },
+    {
+      "epoch": 0.09339233938654051,
+      "grad_norm": 4.094736926672729,
+      "learning_rate": 9.895617161642257e-06,
+      "loss": 0.4664,
+      "step": 306
+    },
+    {
+      "epoch": 0.09369754311002594,
+      "grad_norm": 1.63116898024897,
+      "learning_rate": 9.89460976226211e-06,
+      "loss": 0.3878,
+      "step": 307
+    },
+    {
+      "epoch": 0.09400274683351137,
+      "grad_norm": 1.7238364123731507,
+      "learning_rate": 9.893597576764508e-06,
+      "loss": 0.2989,
+      "step": 308
+    },
+    {
+      "epoch": 0.0943079505569968,
+      "grad_norm": 1.2496662648050174,
+      "learning_rate": 9.8925806061392e-06,
+      "loss": 0.3054,
+      "step": 309
+    },
+    {
+      "epoch": 0.09461315428048223,
+      "grad_norm": 0.8807197003313585,
+      "learning_rate": 9.891558851380614e-06,
+      "loss": 0.1904,
+      "step": 310
+    },
+    {
+      "epoch": 0.09491835800396765,
+      "grad_norm": 1.5076918479598347,
+      "learning_rate": 9.890532313487858e-06,
+      "loss": 0.2679,
+      "step": 311
+    },
+    {
+      "epoch": 0.09522356172745307,
+      "grad_norm": 1.8465691043660122,
+      "learning_rate": 9.889500993464716e-06,
+      "loss": 0.5002,
+      "step": 312
+    },
+    {
+      "epoch": 0.0955287654509385,
+      "grad_norm": 1.9183643810942494,
+      "learning_rate": 9.888464892319647e-06,
+      "loss": 0.4869,
+      "step": 313
+    },
+    {
+      "epoch": 0.09583396917442392,
+      "grad_norm": 1.6515373264151805,
+      "learning_rate": 9.887424011065788e-06,
+      "loss": 0.4507,
+      "step": 314
+    },
+    {
+      "epoch": 0.09613917289790935,
+      "grad_norm": 1.6223391241834122,
+      "learning_rate": 9.886378350720945e-06,
+      "loss": 0.3445,
+      "step": 315
+    },
+    {
+      "epoch": 0.09644437662139478,
+      "grad_norm": 1.4416645097808285,
+      "learning_rate": 9.885327912307604e-06,
+      "loss": 0.2808,
+      "step": 316
+    },
+    {
+      "epoch": 0.09674958034488021,
+      "grad_norm": 1.4777192121308136,
+      "learning_rate": 9.88427269685292e-06,
+      "loss": 0.4335,
+      "step": 317
+    },
+    {
+      "epoch": 0.09705478406836564,
+      "grad_norm": 1.6934694740555867,
+      "learning_rate": 9.883212705388715e-06,
+      "loss": 0.4299,
+      "step": 318
+    },
+    {
+      "epoch": 0.09735998779185107,
+      "grad_norm": 1.9031284601590377,
+      "learning_rate": 9.882147938951489e-06,
+      "loss": 0.5364,
+      "step": 319
+    },
+    {
+      "epoch": 0.09766519151533648,
+      "grad_norm": 1.990035566558448,
+      "learning_rate": 9.881078398582406e-06,
+      "loss": 0.6476,
+      "step": 320
+    },
+    {
+      "epoch": 0.09797039523882191,
+      "grad_norm": 1.4458600630840748,
+      "learning_rate": 9.8800040853273e-06,
+      "loss": 0.268,
+      "step": 321
+    },
+    {
+      "epoch": 0.09827559896230734,
+      "grad_norm": 1.473557254783057,
+      "learning_rate": 9.878925000236667e-06,
+      "loss": 0.3889,
+      "step": 322
+    },
+    {
+      "epoch": 0.09858080268579276,
+      "grad_norm": 1.429462352597184,
+      "learning_rate": 9.877841144365681e-06,
+      "loss": 0.3348,
+      "step": 323
+    },
+    {
+      "epoch": 0.0988860064092782,
+      "grad_norm": 1.9126483909533352,
+      "learning_rate": 9.876752518774167e-06,
+      "loss": 0.5004,
+      "step": 324
+    },
+    {
+      "epoch": 0.09919121013276362,
+      "grad_norm": 1.528278815830415,
+      "learning_rate": 9.875659124526622e-06,
+      "loss": 0.1931,
+      "step": 325
+    },
+    {
+      "epoch": 0.09949641385624905,
+      "grad_norm": 1.6064809314060318,
+      "learning_rate": 9.874560962692207e-06,
+      "loss": 0.2627,
+      "step": 326
+    },
+    {
+      "epoch": 0.09980161757973448,
+      "grad_norm": 1.8583002911468363,
+      "learning_rate": 9.873458034344741e-06,
+      "loss": 0.4795,
+      "step": 327
+    },
+    {
+      "epoch": 0.1001068213032199,
+      "grad_norm": 2.180040993961252,
+      "learning_rate": 9.872350340562704e-06,
+      "loss": 0.3502,
+      "step": 328
+    },
+    {
+      "epoch": 0.10041202502670532,
+      "grad_norm": 2.2760944374886334,
+      "learning_rate": 9.871237882429237e-06,
+      "loss": 0.5504,
+      "step": 329
+    },
+    {
+      "epoch": 0.10071722875019075,
+      "grad_norm": 1.599604903553732,
+      "learning_rate": 9.87012066103214e-06,
+      "loss": 0.4642,
+      "step": 330
+    },
+    {
+      "epoch": 0.10102243247367618,
+      "grad_norm": 1.165592902920284,
+      "learning_rate": 9.868998677463874e-06,
+      "loss": 0.2118,
+      "step": 331
+    },
+    {
+      "epoch": 0.1013276361971616,
+      "grad_norm": 1.562687671834839,
+      "learning_rate": 9.867871932821549e-06,
+      "loss": 0.3389,
+      "step": 332
+    },
+    {
+      "epoch": 0.10163283992064703,
+      "grad_norm": 1.5690938291630006,
+      "learning_rate": 9.866740428206935e-06,
+      "loss": 0.3084,
+      "step": 333
+    },
+    {
+      "epoch": 0.10193804364413246,
+      "grad_norm": 1.6392469885959746,
+      "learning_rate": 9.865604164726456e-06,
+      "loss": 0.3935,
+      "step": 334
+    },
+    {
+      "epoch": 0.10224324736761789,
+      "grad_norm": 1.4807871775535164,
+      "learning_rate": 9.864463143491192e-06,
+      "loss": 0.4445,
+      "step": 335
+    },
+    {
+      "epoch": 0.10254845109110332,
+      "grad_norm": 1.1926217304533164,
+      "learning_rate": 9.86331736561687e-06,
+      "loss": 0.2623,
+      "step": 336
+    },
+    {
+      "epoch": 0.10285365481458873,
+      "grad_norm": 1.5461262133304665,
+      "learning_rate": 9.862166832223871e-06,
+      "loss": 0.4161,
+      "step": 337
+    },
+    {
+      "epoch": 0.10315885853807416,
+      "grad_norm": 1.5135662608063911,
+      "learning_rate": 9.861011544437226e-06,
+      "loss": 0.2864,
+      "step": 338
+    },
+    {
+      "epoch": 0.10346406226155959,
+      "grad_norm": 1.2723458882899108,
+      "learning_rate": 9.85985150338662e-06,
+      "loss": 0.3208,
+      "step": 339
+    },
+    {
+      "epoch": 0.10376926598504502,
+      "grad_norm": 1.5555768560283307,
+      "learning_rate": 9.858686710206373e-06,
+      "loss": 0.3341,
+      "step": 340
+    },
+    {
+      "epoch": 0.10407446970853045,
+      "grad_norm": 1.7562240544367693,
+      "learning_rate": 9.857517166035466e-06,
+      "loss": 0.4637,
+      "step": 341
+    },
+    {
+      "epoch": 0.10437967343201587,
+      "grad_norm": 3.9759170817857044,
+      "learning_rate": 9.856342872017515e-06,
+      "loss": 0.6559,
+      "step": 342
+    },
+    {
+      "epoch": 0.1046848771555013,
+      "grad_norm": 1.6864581247305628,
+      "learning_rate": 9.855163829300789e-06,
+      "loss": 0.4889,
+      "step": 343
+    },
+    {
+      "epoch": 0.10499008087898673,
+      "grad_norm": 1.5660173118774432,
+      "learning_rate": 9.853980039038193e-06,
+      "loss": 0.2512,
+      "step": 344
+    },
+    {
+      "epoch": 0.10529528460247214,
+      "grad_norm": 1.747919617181712,
+      "learning_rate": 9.85279150238728e-06,
+      "loss": 0.4796,
+      "step": 345
+    },
+    {
+      "epoch": 0.10560048832595757,
+      "grad_norm": 2.0064830609007496,
+      "learning_rate": 9.85159822051024e-06,
+      "loss": 0.6876,
+      "step": 346
+    },
+    {
+      "epoch": 0.105905692049443,
+      "grad_norm": 3.232705832740987,
+      "learning_rate": 9.850400194573908e-06,
+      "loss": 0.5043,
+      "step": 347
+    },
+    {
+      "epoch": 0.10621089577292843,
+      "grad_norm": 1.6750365694523466,
+      "learning_rate": 9.849197425749753e-06,
+      "loss": 0.4426,
+      "step": 348
+    },
+    {
+      "epoch": 0.10651609949641386,
+      "grad_norm": 1.2407353388150502,
+      "learning_rate": 9.847989915213883e-06,
+      "loss": 0.2728,
+      "step": 349
+    },
+    {
+      "epoch": 0.10682130321989929,
+      "grad_norm": 1.145770915094197,
+      "learning_rate": 9.846777664147046e-06,
+      "loss": 0.1828,
+      "step": 350
+    },
+    {
+      "epoch": 0.10712650694338471,
+      "grad_norm": 1.0259526919339048,
+      "learning_rate": 9.845560673734617e-06,
+      "loss": 0.1815,
+      "step": 351
+    },
+    {
+      "epoch": 0.10743171066687014,
+      "grad_norm": 1.5927868344920564,
+      "learning_rate": 9.844338945166619e-06,
+      "loss": 0.4563,
+      "step": 352
+    },
+    {
+      "epoch": 0.10773691439035556,
+      "grad_norm": 1.862511817843432,
+      "learning_rate": 9.843112479637692e-06,
+      "loss": 0.4873,
+      "step": 353
+    },
+    {
+      "epoch": 0.10804211811384098,
+      "grad_norm": 1.3346380863318066,
+      "learning_rate": 9.841881278347122e-06,
+      "loss": 0.344,
+      "step": 354
+    },
+    {
+      "epoch": 0.10834732183732641,
+      "grad_norm": 1.2558059799685006,
+      "learning_rate": 9.840645342498817e-06,
+      "loss": 0.3489,
+      "step": 355
+    },
+    {
+      "epoch": 0.10865252556081184,
+      "grad_norm": 1.734496310003891,
+      "learning_rate": 9.839404673301318e-06,
+      "loss": 0.4161,
+      "step": 356
+    },
+    {
+      "epoch": 0.10895772928429727,
+      "grad_norm": 1.5538063811233147,
+      "learning_rate": 9.838159271967795e-06,
+      "loss": 0.4971,
+      "step": 357
+    },
+    {
+      "epoch": 0.1092629330077827,
+      "grad_norm": 1.51967394631939,
+      "learning_rate": 9.836909139716044e-06,
+      "loss": 0.3422,
+      "step": 358
+    },
+    {
+      "epoch": 0.10956813673126813,
+      "grad_norm": 1.8153898943745974,
+      "learning_rate": 9.835654277768487e-06,
+      "loss": 0.5826,
+      "step": 359
+    },
+    {
+      "epoch": 0.10987334045475355,
+      "grad_norm": 1.3956832615282058,
+      "learning_rate": 9.834394687352168e-06,
+      "loss": 0.3444,
+      "step": 360
+    },
+    {
+      "epoch": 0.11017854417823897,
+      "grad_norm": 1.7028097158357791,
+      "learning_rate": 9.833130369698764e-06,
+      "loss": 0.4653,
+      "step": 361
+    },
+    {
+      "epoch": 0.1104837479017244,
+      "grad_norm": 1.984086744142622,
+      "learning_rate": 9.831861326044564e-06,
+      "loss": 0.7268,
+      "step": 362
+    },
+    {
+      "epoch": 0.11078895162520982,
+      "grad_norm": 1.6010962575202456,
+      "learning_rate": 9.830587557630481e-06,
+      "loss": 0.4979,
+      "step": 363
+    },
+    {
+      "epoch": 0.11109415534869525,
+      "grad_norm": 1.163967768763629,
+      "learning_rate": 9.829309065702054e-06,
+      "loss": 0.2721,
+      "step": 364
+    },
+    {
+      "epoch": 0.11139935907218068,
+      "grad_norm": 1.5116526665647354,
+      "learning_rate": 9.828025851509433e-06,
+      "loss": 0.4504,
+      "step": 365
+    },
+    {
+      "epoch": 0.11170456279566611,
+      "grad_norm": 1.3126613426949747,
+      "learning_rate": 9.82673791630739e-06,
+      "loss": 0.2915,
+      "step": 366
+    },
+    {
+      "epoch": 0.11200976651915154,
+      "grad_norm": 1.2919896962667308,
+      "learning_rate": 9.825445261355313e-06,
+      "loss": 0.2862,
+      "step": 367
+    },
+    {
+      "epoch": 0.11231497024263697,
+      "grad_norm": 1.7465438774619029,
+      "learning_rate": 9.824147887917201e-06,
+      "loss": 0.5347,
+      "step": 368
+    },
+    {
+      "epoch": 0.11262017396612238,
+      "grad_norm": 1.5226131458552274,
+      "learning_rate": 9.822845797261676e-06,
+      "loss": 0.4764,
+      "step": 369
+    },
+    {
+      "epoch": 0.11292537768960781,
+      "grad_norm": 1.157991129368488,
+      "learning_rate": 9.82153899066196e-06,
+      "loss": 0.2002,
+      "step": 370
+    },
+    {
+      "epoch": 0.11323058141309324,
+      "grad_norm": 0.9301400886155634,
+      "learning_rate": 9.8202274693959e-06,
+      "loss": 0.1684,
+      "step": 371
+    },
+    {
+      "epoch": 0.11353578513657867,
+      "grad_norm": 1.502487900950231,
+      "learning_rate": 9.818911234745942e-06,
+      "loss": 0.3093,
+      "step": 372
+    },
+    {
+      "epoch": 0.1138409888600641,
+      "grad_norm": 1.9784385010881411,
+      "learning_rate": 9.817590287999149e-06,
+      "loss": 0.4803,
+      "step": 373
+    },
+    {
+      "epoch": 0.11414619258354952,
+      "grad_norm": 0.9195545540370847,
+      "learning_rate": 9.816264630447186e-06,
+      "loss": 0.2159,
+      "step": 374
+    },
+    {
+      "epoch": 0.11445139630703495,
+      "grad_norm": 1.9332444000616873,
+      "learning_rate": 9.814934263386324e-06,
+      "loss": 0.704,
+      "step": 375
+    },
+    {
+      "epoch": 0.11475660003052038,
+      "grad_norm": 1.2982271338494382,
+      "learning_rate": 9.813599188117447e-06,
+      "loss": 0.2579,
+      "step": 376
+    },
+    {
+      "epoch": 0.11506180375400579,
+      "grad_norm": 1.4400207145843447,
+      "learning_rate": 9.812259405946033e-06,
+      "loss": 0.2581,
+      "step": 377
+    },
+    {
+      "epoch": 0.11536700747749122,
+      "grad_norm": 1.6214016909138647,
+      "learning_rate": 9.810914918182168e-06,
+      "loss": 0.3348,
+      "step": 378
+    },
+    {
+      "epoch": 0.11567221120097665,
+      "grad_norm": 1.3037284021243603,
+      "learning_rate": 9.80956572614054e-06,
+      "loss": 0.2009,
+      "step": 379
+    },
+    {
+      "epoch": 0.11597741492446208,
+      "grad_norm": 1.4597999101205241,
+      "learning_rate": 9.808211831140434e-06,
+      "loss": 0.4417,
+      "step": 380
+    },
+    {
+      "epoch": 0.1162826186479475,
+      "grad_norm": 1.3133108268461724,
+      "learning_rate": 9.806853234505736e-06,
+      "loss": 0.3305,
+      "step": 381
+    },
+    {
+      "epoch": 0.11658782237143293,
+      "grad_norm": 1.455145466929522,
+      "learning_rate": 9.805489937564926e-06,
+      "loss": 0.4611,
+      "step": 382
+    },
+    {
+      "epoch": 0.11689302609491836,
+      "grad_norm": 1.3560358010182432,
+      "learning_rate": 9.804121941651085e-06,
+      "loss": 0.2624,
+      "step": 383
+    },
+    {
+      "epoch": 0.11719822981840379,
+      "grad_norm": 1.5730489544680661,
+      "learning_rate": 9.802749248101885e-06,
+      "loss": 0.5959,
+      "step": 384
+    },
+    {
+      "epoch": 0.1175034335418892,
+      "grad_norm": 1.6080011966995038,
+      "learning_rate": 9.801371858259594e-06,
+      "loss": 0.3077,
+      "step": 385
+    },
+    {
+      "epoch": 0.11780863726537463,
+      "grad_norm": 1.3452073998773173,
+      "learning_rate": 9.799989773471071e-06,
+      "loss": 0.3877,
+      "step": 386
+    },
+    {
+      "epoch": 0.11811384098886006,
+      "grad_norm": 1.521531541360139,
+      "learning_rate": 9.798602995087764e-06,
+      "loss": 0.2978,
+      "step": 387
+    },
+    {
+      "epoch": 0.11841904471234549,
+      "grad_norm": 1.705542362457564,
+      "learning_rate": 9.797211524465715e-06,
+      "loss": 0.4298,
+      "step": 388
+    },
+    {
+      "epoch": 0.11872424843583092,
+      "grad_norm": 2.110762528312395,
+      "learning_rate": 9.79581536296555e-06,
+      "loss": 0.4766,
+      "step": 389
+    },
+    {
+      "epoch": 0.11902945215931635,
+      "grad_norm": 1.8442245131411212,
+      "learning_rate": 9.794414511952484e-06,
+      "loss": 0.2988,
+      "step": 390
+    },
+    {
+      "epoch": 0.11933465588280177,
+      "grad_norm": 1.2950053698748747,
+      "learning_rate": 9.793008972796318e-06,
+      "loss": 0.2907,
+      "step": 391
+    },
+    {
+      "epoch": 0.1196398596062872,
+      "grad_norm": 1.1692504994324242,
+      "learning_rate": 9.791598746871438e-06,
+      "loss": 0.2584,
+      "step": 392
+    },
+    {
+      "epoch": 0.11994506332977262,
+      "grad_norm": 1.7847769770487698,
+      "learning_rate": 9.790183835556806e-06,
+      "loss": 0.4874,
+      "step": 393
+    },
+    {
+      "epoch": 0.12025026705325804,
+      "grad_norm": 1.2562325547558533,
+      "learning_rate": 9.788764240235976e-06,
+      "loss": 0.3739,
+      "step": 394
+    },
+    {
+      "epoch": 0.12055547077674347,
+      "grad_norm": 1.340465345857484,
+      "learning_rate": 9.787339962297076e-06,
+      "loss": 0.1911,
+      "step": 395
+    },
+    {
+      "epoch": 0.1208606745002289,
+      "grad_norm": 1.5667496682899116,
+      "learning_rate": 9.785911003132811e-06,
+      "loss": 0.4243,
+      "step": 396
+    },
+    {
+      "epoch": 0.12116587822371433,
+      "grad_norm": 1.81118055527733,
+      "learning_rate": 9.78447736414047e-06,
+      "loss": 0.5246,
+      "step": 397
+    },
+    {
+      "epoch": 0.12147108194719976,
+      "grad_norm": 1.4053388398772566,
+      "learning_rate": 9.783039046721912e-06,
+      "loss": 0.2964,
+      "step": 398
+    },
+    {
+      "epoch": 0.12177628567068519,
+      "grad_norm": 1.5253960954907508,
+      "learning_rate": 9.781596052283573e-06,
+      "loss": 0.3939,
+      "step": 399
+    },
+    {
+      "epoch": 0.12208148939417061,
+      "grad_norm": 1.4250712259795966,
+      "learning_rate": 9.780148382236465e-06,
+      "loss": 0.2338,
+      "step": 400
+    },
+    {
+      "epoch": 0.12238669311765603,
+      "grad_norm": 1.6445234212439845,
+      "learning_rate": 9.778696037996167e-06,
+      "loss": 0.6218,
+      "step": 401
+    },
+    {
+      "epoch": 0.12269189684114146,
+      "grad_norm": 1.625655408880631,
+      "learning_rate": 9.777239020982834e-06,
+      "loss": 0.4565,
+      "step": 402
+    },
+    {
+      "epoch": 0.12299710056462689,
+      "grad_norm": 1.2734470484088918,
+      "learning_rate": 9.775777332621184e-06,
+      "loss": 0.3673,
+      "step": 403
+    },
+    {
+      "epoch": 0.12330230428811231,
+      "grad_norm": 1.5775928166525761,
+      "learning_rate": 9.774310974340506e-06,
+      "loss": 0.3673,
+      "step": 404
+    },
+    {
+      "epoch": 0.12360750801159774,
+      "grad_norm": 1.2252627733176171,
+      "learning_rate": 9.772839947574658e-06,
+      "loss": 0.4064,
+      "step": 405
+    },
+    {
+      "epoch": 0.12391271173508317,
+      "grad_norm": 1.7189364245488037,
+      "learning_rate": 9.77136425376206e-06,
+      "loss": 0.4633,
+      "step": 406
+    },
+    {
+      "epoch": 0.1242179154585686,
+      "grad_norm": 1.7558994988767551,
+      "learning_rate": 9.769883894345693e-06,
+      "loss": 0.5655,
+      "step": 407
+    },
+    {
+      "epoch": 0.12452311918205403,
+      "grad_norm": 1.4802133948901528,
+      "learning_rate": 9.768398870773108e-06,
+      "loss": 0.4175,
+      "step": 408
+    },
+    {
+      "epoch": 0.12482832290553945,
+      "grad_norm": 1.3302751078782091,
+      "learning_rate": 9.766909184496408e-06,
+      "loss": 0.3468,
+      "step": 409
+    },
+    {
+      "epoch": 0.12513352662902488,
+      "grad_norm": 1.704721594870751,
+      "learning_rate": 9.765414836972262e-06,
+      "loss": 0.4898,
+      "step": 410
+    },
+    {
+      "epoch": 0.1254387303525103,
+      "grad_norm": 1.2131375907547226,
+      "learning_rate": 9.763915829661891e-06,
+      "loss": 0.3886,
+      "step": 411
+    },
+    {
+      "epoch": 0.12574393407599574,
+      "grad_norm": 1.3401723742285034,
+      "learning_rate": 9.76241216403108e-06,
+      "loss": 0.4102,
+      "step": 412
+    },
+    {
+      "epoch": 0.12604913779948115,
+      "grad_norm": 1.6513004120286505,
+      "learning_rate": 9.760903841550164e-06,
+      "loss": 0.433,
+      "step": 413
+    },
+    {
+      "epoch": 0.12635434152296657,
+      "grad_norm": 2.2390646637016327,
+      "learning_rate": 9.75939086369403e-06,
+      "loss": 0.4213,
+      "step": 414
+    },
+    {
+      "epoch": 0.126659545246452,
+      "grad_norm": 1.585281461881457,
+      "learning_rate": 9.757873231942122e-06,
+      "loss": 0.2145,
+      "step": 415
+    },
+    {
+      "epoch": 0.12696474896993742,
+      "grad_norm": 1.616278056338578,
+      "learning_rate": 9.756350947778431e-06,
+      "loss": 0.3786,
+      "step": 416
+    },
+    {
+      "epoch": 0.12726995269342287,
+      "grad_norm": 1.3896314522086528,
+      "learning_rate": 9.754824012691499e-06,
+      "loss": 0.3886,
+      "step": 417
+    },
+    {
+      "epoch": 0.12757515641690828,
+      "grad_norm": 1.5388381717065547,
+      "learning_rate": 9.753292428174416e-06,
+      "loss": 0.4901,
+      "step": 418
+    },
+    {
+      "epoch": 0.12788036014039372,
+      "grad_norm": 1.2099556172325527,
+      "learning_rate": 9.75175619572482e-06,
+      "loss": 0.2379,
+      "step": 419
+    },
+    {
+      "epoch": 0.12818556386387914,
+      "grad_norm": 2.578673148403812,
+      "learning_rate": 9.750215316844886e-06,
+      "loss": 0.3093,
+      "step": 420
+    },
+    {
+      "epoch": 0.12849076758736458,
+      "grad_norm": 1.3557950817043143,
+      "learning_rate": 9.748669793041345e-06,
+      "loss": 0.2901,
+      "step": 421
+    },
+    {
+      "epoch": 0.12879597131085,
+      "grad_norm": 1.8058846206435177,
+      "learning_rate": 9.747119625825459e-06,
+      "loss": 0.5222,
+      "step": 422
+    },
+    {
+      "epoch": 0.1291011750343354,
+      "grad_norm": 1.449090093366816,
+      "learning_rate": 9.745564816713034e-06,
+      "loss": 0.3738,
+      "step": 423
+    },
+    {
+      "epoch": 0.12940637875782085,
+      "grad_norm": 1.3181509799847857,
+      "learning_rate": 9.74400536722442e-06,
+      "loss": 0.2624,
+      "step": 424
+    },
+    {
+      "epoch": 0.12971158248130626,
+      "grad_norm": 2.1599699357377675,
+      "learning_rate": 9.742441278884496e-06,
+      "loss": 0.4838,
+      "step": 425
+    },
+    {
+      "epoch": 0.1300167862047917,
+      "grad_norm": 1.6258096878519581,
+      "learning_rate": 9.740872553222685e-06,
+      "loss": 0.4999,
+      "step": 426
+    },
+    {
+      "epoch": 0.13032198992827712,
+      "grad_norm": 1.5046387312958875,
+      "learning_rate": 9.739299191772937e-06,
+      "loss": 0.3095,
+      "step": 427
+    },
+    {
+      "epoch": 0.13062719365176256,
+      "grad_norm": 3.2565177704876653,
+      "learning_rate": 9.737721196073742e-06,
+      "loss": 0.4886,
+      "step": 428
+    },
+    {
+      "epoch": 0.13093239737524798,
+      "grad_norm": 5.072257540244327,
+      "learning_rate": 9.736138567668117e-06,
+      "loss": 0.2433,
+      "step": 429
+    },
+    {
+      "epoch": 0.1312376010987334,
+      "grad_norm": 1.5984597272623309,
+      "learning_rate": 9.734551308103607e-06,
+      "loss": 0.3274,
+      "step": 430
+    },
+    {
+      "epoch": 0.13154280482221883,
+      "grad_norm": 1.9766310728941525,
+      "learning_rate": 9.732959418932297e-06,
+      "loss": 0.5601,
+      "step": 431
+    },
+    {
+      "epoch": 0.13184800854570425,
+      "grad_norm": 1.8500535183975242,
+      "learning_rate": 9.731362901710783e-06,
+      "loss": 0.8055,
+      "step": 432
+    },
+    {
+      "epoch": 0.1321532122691897,
+      "grad_norm": 1.5082490121163843,
+      "learning_rate": 9.7297617580002e-06,
+      "loss": 0.3715,
+      "step": 433
+    },
+    {
+      "epoch": 0.1324584159926751,
+      "grad_norm": 1.3484742708817405,
+      "learning_rate": 9.728155989366198e-06,
+      "loss": 0.2316,
+      "step": 434
+    },
+    {
+      "epoch": 0.13276361971616055,
+      "grad_norm": 1.4064281774599339,
+      "learning_rate": 9.726545597378953e-06,
+      "loss": 0.2529,
+      "step": 435
+    },
+    {
+      "epoch": 0.13306882343964596,
+      "grad_norm": 1.3619428913058238,
+      "learning_rate": 9.724930583613164e-06,
+      "loss": 0.2675,
+      "step": 436
+    },
+    {
+      "epoch": 0.1333740271631314,
+      "grad_norm": 1.8072201350967787,
+      "learning_rate": 9.723310949648044e-06,
+      "loss": 0.4832,
+      "step": 437
+    },
+    {
+      "epoch": 0.13367923088661682,
+      "grad_norm": 1.376837369774298,
+      "learning_rate": 9.721686697067328e-06,
+      "loss": 0.3855,
+      "step": 438
+    },
+    {
+      "epoch": 0.13398443461010223,
+      "grad_norm": 1.577084491012061,
+      "learning_rate": 9.720057827459264e-06,
+      "loss": 0.3307,
+      "step": 439
+    },
+    {
+      "epoch": 0.13428963833358767,
+      "grad_norm": 1.7335930513549882,
+      "learning_rate": 9.718424342416619e-06,
+      "loss": 0.3794,
+      "step": 440
+    },
+    {
+      "epoch": 0.1345948420570731,
+      "grad_norm": 1.3979759520207362,
+      "learning_rate": 9.716786243536672e-06,
+      "loss": 0.17,
+      "step": 441
+    },
+    {
+      "epoch": 0.13490004578055853,
+      "grad_norm": 1.5166020341173607,
+      "learning_rate": 9.715143532421208e-06,
+      "loss": 0.2782,
+      "step": 442
+    },
+    {
+      "epoch": 0.13520524950404395,
+      "grad_norm": 1.5547284451560954,
+      "learning_rate": 9.71349621067653e-06,
+      "loss": 0.2398,
+      "step": 443
+    },
+    {
+      "epoch": 0.1355104532275294,
+      "grad_norm": 1.4706416022210265,
+      "learning_rate": 9.711844279913443e-06,
+      "loss": 0.3486,
+      "step": 444
+    },
+    {
+      "epoch": 0.1358156569510148,
+      "grad_norm": 1.7726813901729983,
+      "learning_rate": 9.710187741747264e-06,
+      "loss": 0.4938,
+      "step": 445
+    },
+    {
+      "epoch": 0.13612086067450022,
+      "grad_norm": 1.3682846898358798,
+      "learning_rate": 9.708526597797812e-06,
+      "loss": 0.3399,
+      "step": 446
+    },
+    {
+      "epoch": 0.13642606439798566,
+      "grad_norm": 1.173206742476641,
+      "learning_rate": 9.70686084968941e-06,
+      "loss": 0.209,
+      "step": 447
+    },
+    {
+      "epoch": 0.13673126812147107,
+      "grad_norm": 1.5006667880365117,
+      "learning_rate": 9.705190499050885e-06,
+      "loss": 0.3918,
+      "step": 448
+    },
+    {
+      "epoch": 0.13703647184495651,
+      "grad_norm": 1.4081281623516984,
+      "learning_rate": 9.70351554751556e-06,
+      "loss": 0.3478,
+      "step": 449
+    },
+    {
+      "epoch": 0.13734167556844193,
+      "grad_norm": 1.3521857616183288,
+      "learning_rate": 9.701835996721267e-06,
+      "loss": 0.2527,
+      "step": 450
+    },
+    {
+      "epoch": 0.13764687929192737,
+      "grad_norm": 1.609701105059192,
+      "learning_rate": 9.70015184831032e-06,
+      "loss": 0.4019,
+      "step": 451
+    },
+    {
+      "epoch": 0.13795208301541279,
+      "grad_norm": 1.5247656094886828,
+      "learning_rate": 9.698463103929542e-06,
+      "loss": 0.3393,
+      "step": 452
+    },
+    {
+      "epoch": 0.13825728673889823,
+      "grad_norm": 1.7521263555747508,
+      "learning_rate": 9.696769765230244e-06,
+      "loss": 0.4711,
+      "step": 453
+    },
+    {
+      "epoch": 0.13856249046238364,
+      "grad_norm": 1.1750952661731386,
+      "learning_rate": 9.695071833868233e-06,
+      "loss": 0.3209,
+      "step": 454
+    },
+    {
+      "epoch": 0.13886769418586906,
+      "grad_norm": 1.1097718724387464,
+      "learning_rate": 9.693369311503801e-06,
+      "loss": 0.2077,
+      "step": 455
+    },
+    {
+      "epoch": 0.1391728979093545,
+      "grad_norm": 1.3477077308441543,
+      "learning_rate": 9.691662199801735e-06,
+      "loss": 0.3387,
+      "step": 456
+    },
+    {
+      "epoch": 0.1394781016328399,
+      "grad_norm": 1.5926509134118427,
+      "learning_rate": 9.689950500431306e-06,
+      "loss": 0.4737,
+      "step": 457
+    },
+    {
+      "epoch": 0.13978330535632535,
+      "grad_norm": 1.238821976588628,
+      "learning_rate": 9.688234215066274e-06,
+      "loss": 0.2941,
+      "step": 458
+    },
+    {
+      "epoch": 0.14008850907981077,
+      "grad_norm": 1.283103486116252,
+      "learning_rate": 9.68651334538488e-06,
+      "loss": 0.2298,
+      "step": 459
+    },
+    {
+      "epoch": 0.1403937128032962,
+      "grad_norm": 1.3964480095243228,
+      "learning_rate": 9.684787893069852e-06,
+      "loss": 0.2755,
+      "step": 460
+    },
+    {
+      "epoch": 0.14069891652678163,
+      "grad_norm": 4.5768637107742185,
+      "learning_rate": 9.683057859808394e-06,
+      "loss": 0.3969,
+      "step": 461
+    },
+    {
+      "epoch": 0.14100412025026704,
+      "grad_norm": 1.238974584705195,
+      "learning_rate": 9.681323247292193e-06,
+      "loss": 0.302,
+      "step": 462
+    },
+    {
+      "epoch": 0.14130932397375248,
+      "grad_norm": 1.484367852444757,
+      "learning_rate": 9.679584057217412e-06,
+      "loss": 0.3771,
+      "step": 463
+    },
+    {
+      "epoch": 0.1416145276972379,
+      "grad_norm": 1.8892302609803566,
+      "learning_rate": 9.677840291284693e-06,
+      "loss": 0.4296,
+      "step": 464
+    },
+    {
+      "epoch": 0.14191973142072334,
+      "grad_norm": 1.9264735588835349,
+      "learning_rate": 9.676091951199147e-06,
+      "loss": 0.1789,
+      "step": 465
+    },
+    {
+      "epoch": 0.14222493514420875,
+      "grad_norm": 1.6779527799580054,
+      "learning_rate": 9.674339038670362e-06,
+      "loss": 0.3394,
+      "step": 466
+    },
+    {
+      "epoch": 0.1425301388676942,
+      "grad_norm": 1.4269803792009585,
+      "learning_rate": 9.672581555412396e-06,
+      "loss": 0.3436,
+      "step": 467
+    },
+    {
+      "epoch": 0.1428353425911796,
+      "grad_norm": 1.6997561953129157,
+      "learning_rate": 9.67081950314378e-06,
+      "loss": 0.4244,
+      "step": 468
+    },
+    {
+      "epoch": 0.14314054631466505,
+      "grad_norm": 1.5847658235861504,
+      "learning_rate": 9.669052883587503e-06,
+      "loss": 0.4061,
+      "step": 469
+    },
+    {
+      "epoch": 0.14344575003815047,
+      "grad_norm": 1.5000132719088555,
+      "learning_rate": 9.667281698471032e-06,
+      "loss": 0.2349,
+      "step": 470
+    },
+    {
+      "epoch": 0.14375095376163588,
+      "grad_norm": 1.4700398564287758,
+      "learning_rate": 9.665505949526288e-06,
+      "loss": 0.4265,
+      "step": 471
+    },
+    {
+      "epoch": 0.14405615748512132,
+      "grad_norm": 1.1929108671224367,
+      "learning_rate": 9.663725638489662e-06,
+      "loss": 0.3107,
+      "step": 472
+    },
+    {
+      "epoch": 0.14436136120860674,
+      "grad_norm": 1.1577689014622543,
+      "learning_rate": 9.661940767102001e-06,
+      "loss": 0.1722,
+      "step": 473
+    },
+    {
+      "epoch": 0.14466656493209218,
+      "grad_norm": 2.093772204836488,
+      "learning_rate": 9.660151337108617e-06,
+      "loss": 0.4278,
+      "step": 474
+    },
+    {
+      "epoch": 0.1449717686555776,
+      "grad_norm": 1.6654665758895533,
+      "learning_rate": 9.658357350259274e-06,
+      "loss": 0.3394,
+      "step": 475
+    },
+    {
+      "epoch": 0.14527697237906304,
+      "grad_norm": 1.2202451289258887,
+      "learning_rate": 9.656558808308193e-06,
+      "loss": 0.3433,
+      "step": 476
+    },
+    {
+      "epoch": 0.14558217610254845,
+      "grad_norm": 1.5418473816537024,
+      "learning_rate": 9.654755713014052e-06,
+      "loss": 0.4099,
+      "step": 477
+    },
+    {
+      "epoch": 0.14588737982603386,
+      "grad_norm": 1.867172943441559,
+      "learning_rate": 9.652948066139978e-06,
+      "loss": 0.4445,
+      "step": 478
+    },
+    {
+      "epoch": 0.1461925835495193,
+      "grad_norm": 1.2274320399473075,
+      "learning_rate": 9.651135869453552e-06,
+      "loss": 0.3687,
+      "step": 479
+    },
+    {
+      "epoch": 0.14649778727300472,
+      "grad_norm": 1.3632570280967384,
+      "learning_rate": 9.649319124726799e-06,
+      "loss": 0.3554,
+      "step": 480
+    },
+    {
+      "epoch": 0.14680299099649016,
+      "grad_norm": 1.9368328668689925,
+      "learning_rate": 9.647497833736197e-06,
+      "loss": 0.433,
+      "step": 481
+    },
+    {
+      "epoch": 0.14710819471997558,
+      "grad_norm": 1.7750440538339176,
+      "learning_rate": 9.645671998262668e-06,
+      "loss": 0.4275,
+      "step": 482
+    },
+    {
+      "epoch": 0.14741339844346102,
+      "grad_norm": 1.7620981827052555,
+      "learning_rate": 9.643841620091572e-06,
+      "loss": 0.4327,
+      "step": 483
+    },
+    {
+      "epoch": 0.14771860216694643,
+      "grad_norm": 1.705808013569081,
+      "learning_rate": 9.642006701012719e-06,
+      "loss": 0.2914,
+      "step": 484
+    },
+    {
+      "epoch": 0.14802380589043188,
+      "grad_norm": 1.4490666476731855,
+      "learning_rate": 9.640167242820356e-06,
+      "loss": 0.4171,
+      "step": 485
+    },
+    {
+      "epoch": 0.1483290096139173,
+      "grad_norm": 1.4017898188106575,
+      "learning_rate": 9.638323247313167e-06,
+      "loss": 0.3609,
+      "step": 486
+    },
+    {
+      "epoch": 0.1486342133374027,
+      "grad_norm": 1.273585773097182,
+      "learning_rate": 9.636474716294275e-06,
+      "loss": 0.2781,
+      "step": 487
+    },
+    {
+      "epoch": 0.14893941706088815,
+      "grad_norm": 1.4426438722313946,
+      "learning_rate": 9.634621651571235e-06,
+      "loss": 0.3446,
+      "step": 488
+    },
+    {
+      "epoch": 0.14924462078437356,
+      "grad_norm": 1.3396568501630033,
+      "learning_rate": 9.632764054956042e-06,
+      "loss": 0.3209,
+      "step": 489
+    },
+    {
+      "epoch": 0.149549824507859,
+      "grad_norm": 1.869987463373602,
+      "learning_rate": 9.630901928265113e-06,
+      "loss": 0.6676,
+      "step": 490
+    },
+    {
+      "epoch": 0.14985502823134442,
+      "grad_norm": 1.3995020559881168,
+      "learning_rate": 9.6290352733193e-06,
+      "loss": 0.443,
+      "step": 491
+    },
+    {
+      "epoch": 0.15016023195482986,
+      "grad_norm": 1.5646801044524754,
+      "learning_rate": 9.627164091943886e-06,
+      "loss": 0.4808,
+      "step": 492
+    },
+    {
+      "epoch": 0.15046543567831527,
+      "grad_norm": 1.4064811516565765,
+      "learning_rate": 9.625288385968572e-06,
+      "loss": 0.3281,
+      "step": 493
+    },
+    {
+      "epoch": 0.15077063940180072,
+      "grad_norm": 1.3603824959747954,
+      "learning_rate": 9.623408157227493e-06,
+      "loss": 0.4203,
+      "step": 494
+    },
+    {
+      "epoch": 0.15107584312528613,
+      "grad_norm": 1.4648922518012057,
+      "learning_rate": 9.621523407559193e-06,
+      "loss": 0.4691,
+      "step": 495
+    },
+    {
+      "epoch": 0.15138104684877154,
+      "grad_norm": 1.2898880598011702,
+      "learning_rate": 9.619634138806653e-06,
+      "loss": 0.22,
+      "step": 496
+    },
+    {
+      "epoch": 0.151686250572257,
+      "grad_norm": 1.629312149082348,
+      "learning_rate": 9.61774035281726e-06,
+      "loss": 0.3852,
+      "step": 497
+    },
+    {
+      "epoch": 0.1519914542957424,
+      "grad_norm": 1.4209940770003642,
+      "learning_rate": 9.615842051442825e-06,
+      "loss": 0.3434,
+      "step": 498
+    },
+    {
+      "epoch": 0.15229665801922784,
+      "grad_norm": 1.5981471419786573,
+      "learning_rate": 9.613939236539571e-06,
+      "loss": 0.4405,
+      "step": 499
+    },
+    {
+      "epoch": 0.15260186174271326,
+      "grad_norm": 1.6345273007096384,
+      "learning_rate": 9.612031909968138e-06,
+      "loss": 0.4621,
+      "step": 500
+    },
+    {
+      "epoch": 0.1529070654661987,
+      "grad_norm": 1.7946527422515466,
+      "learning_rate": 9.610120073593574e-06,
+      "loss": 0.4215,
+      "step": 501
+    },
+    {
+      "epoch": 0.1532122691896841,
+      "grad_norm": 1.9696007964079152,
+      "learning_rate": 9.608203729285337e-06,
+      "loss": 0.2416,
+      "step": 502
+    },
+    {
+      "epoch": 0.15351747291316953,
+      "grad_norm": 1.186483550480542,
+      "learning_rate": 9.606282878917296e-06,
+      "loss": 0.1656,
+      "step": 503
+    },
+    {
+      "epoch": 0.15382267663665497,
+      "grad_norm": 1.3709781521921298,
+      "learning_rate": 9.604357524367723e-06,
+      "loss": 0.3374,
+      "step": 504
+    },
+    {
+      "epoch": 0.15412788036014038,
+      "grad_norm": 1.4744363645402312,
+      "learning_rate": 9.602427667519297e-06,
+      "loss": 0.3472,
+      "step": 505
+    },
+    {
+      "epoch": 0.15443308408362583,
+      "grad_norm": 1.7032963377834875,
+      "learning_rate": 9.600493310259098e-06,
+      "loss": 0.4352,
+      "step": 506
+    },
+    {
+      "epoch": 0.15473828780711124,
+      "grad_norm": 1.487020684601837,
+      "learning_rate": 9.598554454478608e-06,
+      "loss": 0.2661,
+      "step": 507
+    },
+    {
+      "epoch": 0.15504349153059668,
+      "grad_norm": 1.2499312381905126,
+      "learning_rate": 9.596611102073703e-06,
+      "loss": 0.2785,
+      "step": 508
+    },
+    {
+      "epoch": 0.1553486952540821,
+      "grad_norm": 1.529878897767237,
+      "learning_rate": 9.594663254944664e-06,
+      "loss": 0.3768,
+      "step": 509
+    },
+    {
+      "epoch": 0.15565389897756754,
+      "grad_norm": 1.5214931502474798,
+      "learning_rate": 9.592710914996162e-06,
+      "loss": 0.5126,
+      "step": 510
+    },
+    {
+      "epoch": 0.15595910270105295,
+      "grad_norm": 1.3836939529329817,
+      "learning_rate": 9.590754084137259e-06,
+      "loss": 0.3011,
+      "step": 511
+    },
+    {
+      "epoch": 0.15626430642453837,
+      "grad_norm": 1.4833094737816435,
+      "learning_rate": 9.588792764281413e-06,
+      "loss": 0.4761,
+      "step": 512
+    },
+    {
+      "epoch": 0.1565695101480238,
+      "grad_norm": 1.2346664760598378,
+      "learning_rate": 9.586826957346473e-06,
+      "loss": 0.2454,
+      "step": 513
+    },
+    {
+      "epoch": 0.15687471387150922,
+      "grad_norm": 1.6476636760719772,
+      "learning_rate": 9.584856665254667e-06,
+      "loss": 0.2867,
+      "step": 514
+    },
+    {
+      "epoch": 0.15717991759499467,
+      "grad_norm": 1.5637210652382973,
+      "learning_rate": 9.58288188993262e-06,
+      "loss": 0.2899,
+      "step": 515
+    },
+    {
+      "epoch": 0.15748512131848008,
+      "grad_norm": 1.3205463270086828,
+      "learning_rate": 9.580902633311331e-06,
+      "loss": 0.3756,
+      "step": 516
+    },
+    {
+      "epoch": 0.15779032504196552,
+      "grad_norm": 1.3975127161911243,
+      "learning_rate": 9.578918897326186e-06,
+      "loss": 0.5111,
+      "step": 517
+    },
+    {
+      "epoch": 0.15809552876545094,
+      "grad_norm": 1.4459383872321914,
+      "learning_rate": 9.57693068391695e-06,
+      "loss": 0.4283,
+      "step": 518
+    },
+    {
+      "epoch": 0.15840073248893635,
+      "grad_norm": 1.7061785001760192,
+      "learning_rate": 9.574937995027767e-06,
+      "loss": 0.3702,
+      "step": 519
+    },
+    {
+      "epoch": 0.1587059362124218,
+      "grad_norm": 1.822247438656905,
+      "learning_rate": 9.572940832607157e-06,
+      "loss": 0.3406,
+      "step": 520
+    },
+    {
+      "epoch": 0.1590111399359072,
+      "grad_norm": 2.44932543751886,
+      "learning_rate": 9.570939198608013e-06,
+      "loss": 0.3118,
+      "step": 521
+    },
+    {
+      "epoch": 0.15931634365939265,
+      "grad_norm": 1.6119202421775476,
+      "learning_rate": 9.5689330949876e-06,
+      "loss": 0.3903,
+      "step": 522
+    },
+    {
+      "epoch": 0.15962154738287807,
+      "grad_norm": 1.4848330945324375,
+      "learning_rate": 9.56692252370756e-06,
+      "loss": 0.2336,
+      "step": 523
+    },
+    {
+      "epoch": 0.1599267511063635,
+      "grad_norm": 1.410632362194396,
+      "learning_rate": 9.564907486733893e-06,
+      "loss": 0.2749,
+      "step": 524
+    },
+    {
+      "epoch": 0.16023195482984892,
+      "grad_norm": 1.596052638125191,
+      "learning_rate": 9.562887986036975e-06,
+      "loss": 0.4752,
+      "step": 525
+    },
+    {
+      "epoch": 0.16053715855333436,
+      "grad_norm": 1.7668740909494465,
+      "learning_rate": 9.560864023591541e-06,
+      "loss": 0.4457,
+      "step": 526
+    },
+    {
+      "epoch": 0.16084236227681978,
+      "grad_norm": 1.4009268145182425,
+      "learning_rate": 9.558835601376692e-06,
+      "loss": 0.2615,
+      "step": 527
+    },
+    {
+      "epoch": 0.1611475660003052,
+      "grad_norm": 1.7299333351168085,
+      "learning_rate": 9.55680272137589e-06,
+      "loss": 0.5216,
+      "step": 528
+    },
+    {
+      "epoch": 0.16145276972379063,
+      "grad_norm": 1.398003196407042,
+      "learning_rate": 9.554765385576951e-06,
+      "loss": 0.2917,
+      "step": 529
+    },
+    {
+      "epoch": 0.16175797344727605,
+      "grad_norm": 1.4037115710357768,
+      "learning_rate": 9.552723595972055e-06,
+      "loss": 0.2794,
+      "step": 530
+    },
+    {
+      "epoch": 0.1620631771707615,
+      "grad_norm": 1.4104804936912443,
+      "learning_rate": 9.550677354557734e-06,
+      "loss": 0.3294,
+      "step": 531
+    },
+    {
+      "epoch": 0.1623683808942469,
+      "grad_norm": 1.3043707731550427,
+      "learning_rate": 9.548626663334872e-06,
+      "loss": 0.3542,
+      "step": 532
+    },
+    {
+      "epoch": 0.16267358461773235,
+      "grad_norm": 1.4523817232860987,
+      "learning_rate": 9.546571524308707e-06,
+      "loss": 0.4394,
+      "step": 533
+    },
+    {
+      "epoch": 0.16297878834121776,
+      "grad_norm": 1.2378417959119585,
+      "learning_rate": 9.544511939488823e-06,
+      "loss": 0.2859,
+      "step": 534
+    },
+    {
+      "epoch": 0.16328399206470318,
+      "grad_norm": 1.1623856416463947,
+      "learning_rate": 9.542447910889152e-06,
+      "loss": 0.2682,
+      "step": 535
+    },
+    {
+      "epoch": 0.16358919578818862,
+      "grad_norm": 1.503292443987416,
+      "learning_rate": 9.540379440527974e-06,
+      "loss": 0.4513,
+      "step": 536
+    },
+    {
+      "epoch": 0.16389439951167403,
+      "grad_norm": 1.278183220840744,
+      "learning_rate": 9.538306530427908e-06,
+      "loss": 0.2486,
+      "step": 537
+    },
+    {
+      "epoch": 0.16419960323515947,
+      "grad_norm": 1.477438530587252,
+      "learning_rate": 9.536229182615919e-06,
+      "loss": 0.4748,
+      "step": 538
+    },
+    {
+      "epoch": 0.1645048069586449,
+      "grad_norm": 1.161000468008389,
+      "learning_rate": 9.534147399123308e-06,
+      "loss": 0.3166,
+      "step": 539
+    },
+    {
+      "epoch": 0.16481001068213033,
+      "grad_norm": 1.3151690275104762,
+      "learning_rate": 9.532061181985713e-06,
+      "loss": 0.3547,
+      "step": 540
+    },
+    {
+      "epoch": 0.16511521440561575,
+      "grad_norm": 1.750297725419665,
+      "learning_rate": 9.529970533243112e-06,
+      "loss": 0.4156,
+      "step": 541
+    },
+    {
+      "epoch": 0.1654204181291012,
+      "grad_norm": 1.3465531852012238,
+      "learning_rate": 9.52787545493981e-06,
+      "loss": 0.3366,
+      "step": 542
+    },
+    {
+      "epoch": 0.1657256218525866,
+      "grad_norm": 1.5440141830188223,
+      "learning_rate": 9.525775949124447e-06,
+      "loss": 0.3376,
+      "step": 543
+    },
+    {
+      "epoch": 0.16603082557607202,
+      "grad_norm": 1.5415130315298482,
+      "learning_rate": 9.523672017849995e-06,
+      "loss": 0.541,
+      "step": 544
+    },
+    {
+      "epoch": 0.16633602929955746,
+      "grad_norm": 1.7579856956776627,
+      "learning_rate": 9.521563663173746e-06,
+      "loss": 0.4806,
+      "step": 545
+    },
+    {
+      "epoch": 0.16664123302304287,
+      "grad_norm": 1.7818866729621587,
+      "learning_rate": 9.519450887157324e-06,
+      "loss": 0.5464,
+      "step": 546
+    },
+    {
+      "epoch": 0.16694643674652831,
+      "grad_norm": 1.6064088023016758,
+      "learning_rate": 9.517333691866672e-06,
+      "loss": 0.43,
+      "step": 547
+    },
+    {
+      "epoch": 0.16725164047001373,
+      "grad_norm": 1.4778043177300115,
+      "learning_rate": 9.515212079372059e-06,
+      "loss": 0.4399,
+      "step": 548
+    },
+    {
+      "epoch": 0.16755684419349917,
+      "grad_norm": 2.027608100840915,
+      "learning_rate": 9.513086051748069e-06,
+      "loss": 0.4069,
+      "step": 549
+    },
+    {
+      "epoch": 0.16786204791698459,
+      "grad_norm": 1.5850802204517391,
+      "learning_rate": 9.510955611073605e-06,
+      "loss": 0.3827,
+      "step": 550
+    },
+    {
+      "epoch": 0.16816725164047,
+      "grad_norm": 1.293889481401633,
+      "learning_rate": 9.508820759431883e-06,
+      "loss": 0.2572,
+      "step": 551
+    },
+    {
+      "epoch": 0.16847245536395544,
+      "grad_norm": 1.842230801320139,
+      "learning_rate": 9.506681498910437e-06,
+      "loss": 0.5275,
+      "step": 552
+    },
+    {
+      "epoch": 0.16877765908744086,
+      "grad_norm": 1.1873629641518748,
+      "learning_rate": 9.50453783160111e-06,
+      "loss": 0.3282,
+      "step": 553
+    },
+    {
+      "epoch": 0.1690828628109263,
+      "grad_norm": 1.6093453710106354,
+      "learning_rate": 9.50238975960005e-06,
+      "loss": 0.5784,
+      "step": 554
+    },
+    {
+      "epoch": 0.1693880665344117,
+      "grad_norm": 1.8900657646403543,
+      "learning_rate": 9.500237285007719e-06,
+      "loss": 0.5224,
+      "step": 555
+    },
+    {
+      "epoch": 0.16969327025789716,
+      "grad_norm": 1.4803576264157936,
+      "learning_rate": 9.498080409928878e-06,
+      "loss": 0.3726,
+      "step": 556
+    },
+    {
+      "epoch": 0.16999847398138257,
+      "grad_norm": 1.7675886459458987,
+      "learning_rate": 9.495919136472595e-06,
+      "loss": 0.4656,
+      "step": 557
+    },
+    {
+      "epoch": 0.170303677704868,
+      "grad_norm": 1.7910944298366285,
+      "learning_rate": 9.493753466752236e-06,
+      "loss": 0.8076,
+      "step": 558
+    },
+    {
+      "epoch": 0.17060888142835343,
+      "grad_norm": 1.4678825901279975,
+      "learning_rate": 9.49158340288547e-06,
+      "loss": 0.3575,
+      "step": 559
+    },
+    {
+      "epoch": 0.17091408515183884,
+      "grad_norm": 1.411596350121475,
+      "learning_rate": 9.489408946994256e-06,
+      "loss": 0.3316,
+      "step": 560
+    },
+    {
+      "epoch": 0.17121928887532428,
+      "grad_norm": 1.2961064094904746,
+      "learning_rate": 9.487230101204855e-06,
+      "loss": 0.3634,
+      "step": 561
+    },
+    {
+      "epoch": 0.1715244925988097,
+      "grad_norm": 1.2525457208629842,
+      "learning_rate": 9.485046867647816e-06,
+      "loss": 0.368,
+      "step": 562
+    },
+    {
+      "epoch": 0.17182969632229514,
+      "grad_norm": 1.5857071732762902,
+      "learning_rate": 9.48285924845798e-06,
+      "loss": 0.4546,
+      "step": 563
+    },
+    {
+      "epoch": 0.17213490004578055,
+      "grad_norm": 1.4510716523054648,
+      "learning_rate": 9.480667245774474e-06,
+      "loss": 0.2739,
+      "step": 564
+    },
+    {
+      "epoch": 0.172440103769266,
+      "grad_norm": 1.7890268513821783,
+      "learning_rate": 9.478470861740716e-06,
+      "loss": 0.4085,
+      "step": 565
+    },
+    {
+      "epoch": 0.1727453074927514,
+      "grad_norm": 1.2143129127325427,
+      "learning_rate": 9.476270098504405e-06,
+      "loss": 0.2669,
+      "step": 566
+    },
+    {
+      "epoch": 0.17305051121623685,
+      "grad_norm": 1.4725539275626134,
+      "learning_rate": 9.474064958217524e-06,
+      "loss": 0.3474,
+      "step": 567
+    },
+    {
+      "epoch": 0.17335571493972227,
+      "grad_norm": 1.6648347880329453,
+      "learning_rate": 9.471855443036333e-06,
+      "loss": 0.3059,
+      "step": 568
+    },
+    {
+      "epoch": 0.17366091866320768,
+      "grad_norm": 1.4826208089202084,
+      "learning_rate": 9.469641555121372e-06,
+      "loss": 0.3309,
+      "step": 569
+    },
+    {
+      "epoch": 0.17396612238669312,
+      "grad_norm": 1.2087588985869038,
+      "learning_rate": 9.467423296637458e-06,
+      "loss": 0.2765,
+      "step": 570
+    },
+    {
+      "epoch": 0.17427132611017854,
+      "grad_norm": 1.5271503296745377,
+      "learning_rate": 9.465200669753678e-06,
+      "loss": 0.4388,
+      "step": 571
+    },
+    {
+      "epoch": 0.17457652983366398,
+      "grad_norm": 1.720167996940521,
+      "learning_rate": 9.462973676643395e-06,
+      "loss": 0.4693,
+      "step": 572
+    },
+    {
+      "epoch": 0.1748817335571494,
+      "grad_norm": 1.4666930059033054,
+      "learning_rate": 9.46074231948424e-06,
+      "loss": 0.354,
+      "step": 573
+    },
+    {
+      "epoch": 0.17518693728063484,
+      "grad_norm": 1.8597430337784902,
+      "learning_rate": 9.458506600458106e-06,
+      "loss": 0.3892,
+      "step": 574
+    },
+    {
+      "epoch": 0.17549214100412025,
+      "grad_norm": 0.9696512337091734,
+      "learning_rate": 9.456266521751162e-06,
+      "loss": 0.2294,
+      "step": 575
+    },
+    {
+      "epoch": 0.17579734472760566,
+      "grad_norm": 1.781498807963985,
+      "learning_rate": 9.454022085553829e-06,
+      "loss": 0.4873,
+      "step": 576
+    },
+    {
+      "epoch": 0.1761025484510911,
+      "grad_norm": 1.3337716733106453,
+      "learning_rate": 9.451773294060797e-06,
+      "loss": 0.3031,
+      "step": 577
+    },
+    {
+      "epoch": 0.17640775217457652,
+      "grad_norm": 1.7055986970891146,
+      "learning_rate": 9.449520149471008e-06,
+      "loss": 0.6405,
+      "step": 578
+    },
+    {
+      "epoch": 0.17671295589806196,
+      "grad_norm": 1.5188407211523098,
+      "learning_rate": 9.447262653987668e-06,
+      "loss": 0.3739,
+      "step": 579
+    },
+    {
+      "epoch": 0.17701815962154738,
+      "grad_norm": 0.9308569789252137,
+      "learning_rate": 9.445000809818231e-06,
+      "loss": 0.2505,
+      "step": 580
+    },
+    {
+      "epoch": 0.17732336334503282,
+      "grad_norm": 1.5000700139554115,
+      "learning_rate": 9.442734619174408e-06,
+      "loss": 0.4158,
+      "step": 581
+    },
+    {
+      "epoch": 0.17762856706851823,
+      "grad_norm": 1.3548714588751447,
+      "learning_rate": 9.440464084272157e-06,
+      "loss": 0.3911,
+      "step": 582
+    },
+    {
+      "epoch": 0.17793377079200368,
+      "grad_norm": 1.3241166157002833,
+      "learning_rate": 9.438189207331684e-06,
+      "loss": 0.371,
+      "step": 583
+    },
+    {
+      "epoch": 0.1782389745154891,
+      "grad_norm": 1.251287417238196,
+      "learning_rate": 9.435909990577442e-06,
+      "loss": 0.4543,
+      "step": 584
+    },
+    {
+      "epoch": 0.1785441782389745,
+      "grad_norm": 1.6243898273186124,
+      "learning_rate": 9.433626436238128e-06,
+      "loss": 0.3955,
+      "step": 585
+    },
+    {
+      "epoch": 0.17884938196245995,
+      "grad_norm": 1.5933520250274826,
+      "learning_rate": 9.43133854654668e-06,
+      "loss": 0.4232,
+      "step": 586
+    },
+    {
+      "epoch": 0.17915458568594536,
+      "grad_norm": 1.271604048902552,
+      "learning_rate": 9.429046323740275e-06,
+      "loss": 0.186,
+      "step": 587
+    },
+    {
+      "epoch": 0.1794597894094308,
+      "grad_norm": 1.7480757897399084,
+      "learning_rate": 9.426749770060325e-06,
+      "loss": 0.3198,
+      "step": 588
+    },
+    {
+      "epoch": 0.17976499313291622,
+      "grad_norm": 1.6582908292655634,
+      "learning_rate": 9.424448887752485e-06,
+      "loss": 0.4497,
+      "step": 589
+    },
+    {
+      "epoch": 0.18007019685640166,
+      "grad_norm": 1.514808198729056,
+      "learning_rate": 9.42214367906663e-06,
+      "loss": 0.3135,
+      "step": 590
+    },
+    {
+      "epoch": 0.18037540057988707,
+      "grad_norm": 1.33735933544563,
+      "learning_rate": 9.419834146256875e-06,
+      "loss": 0.1512,
+      "step": 591
+    },
+    {
+      "epoch": 0.1806806043033725,
+      "grad_norm": 1.7983955915325747,
+      "learning_rate": 9.417520291581562e-06,
+      "loss": 0.397,
+      "step": 592
+    },
+    {
+      "epoch": 0.18098580802685793,
+      "grad_norm": 1.8357965942853254,
+      "learning_rate": 9.415202117303253e-06,
+      "loss": 0.3479,
+      "step": 593
+    },
+    {
+      "epoch": 0.18129101175034334,
+      "grad_norm": 2.1098502294891084,
+      "learning_rate": 9.412879625688742e-06,
+      "loss": 0.6081,
+      "step": 594
+    },
+    {
+      "epoch": 0.1815962154738288,
+      "grad_norm": 1.7002717361934219,
+      "learning_rate": 9.410552819009041e-06,
+      "loss": 0.2335,
+      "step": 595
+    },
+    {
+      "epoch": 0.1819014191973142,
+      "grad_norm": 1.6858392243118179,
+      "learning_rate": 9.408221699539381e-06,
+      "loss": 0.3502,
+      "step": 596
+    },
+    {
+      "epoch": 0.18220662292079964,
+      "grad_norm": 1.9779389304442994,
+      "learning_rate": 9.40588626955921e-06,
+      "loss": 0.5023,
+      "step": 597
+    },
+    {
+      "epoch": 0.18251182664428506,
+      "grad_norm": 1.984831269441273,
+      "learning_rate": 9.403546531352192e-06,
+      "loss": 0.1808,
+      "step": 598
+    },
+    {
+      "epoch": 0.1828170303677705,
+      "grad_norm": 1.7825668553552305,
+      "learning_rate": 9.401202487206205e-06,
+      "loss": 0.2451,
+      "step": 599
+    },
+    {
+      "epoch": 0.18312223409125591,
+      "grad_norm": 1.7395479880147604,
+      "learning_rate": 9.398854139413332e-06,
+      "loss": 0.4586,
+      "step": 600
+    },
+    {
+      "epoch": 0.18342743781474133,
+      "grad_norm": 1.7910877075998561,
+      "learning_rate": 9.396501490269871e-06,
+      "loss": 0.4334,
+      "step": 601
+    },
+    {
+      "epoch": 0.18373264153822677,
+      "grad_norm": 1.671153260357796,
+      "learning_rate": 9.394144542076321e-06,
+      "loss": 0.3457,
+      "step": 602
+    },
+    {
+      "epoch": 0.18403784526171219,
+      "grad_norm": 2.2927735747628057,
+      "learning_rate": 9.391783297137392e-06,
+      "loss": 0.5006,
+      "step": 603
+    },
+    {
+      "epoch": 0.18434304898519763,
+      "grad_norm": 1.4375155904173251,
+      "learning_rate": 9.389417757761983e-06,
+      "loss": 0.3412,
+      "step": 604
+    },
+    {
+      "epoch": 0.18464825270868304,
+      "grad_norm": 1.0876679937459988,
+      "learning_rate": 9.387047926263205e-06,
+      "loss": 0.2323,
+      "step": 605
+    },
+    {
+      "epoch": 0.18495345643216848,
+      "grad_norm": 1.5691166969698962,
+      "learning_rate": 9.384673804958357e-06,
+      "loss": 0.3929,
+      "step": 606
+    },
+    {
+      "epoch": 0.1852586601556539,
+      "grad_norm": 1.2508041656129554,
+      "learning_rate": 9.38229539616894e-06,
+      "loss": 0.2123,
+      "step": 607
+    },
+    {
+      "epoch": 0.1855638638791393,
+      "grad_norm": 1.6014009719897135,
+      "learning_rate": 9.379912702220641e-06,
+      "loss": 0.234,
+      "step": 608
+    },
+    {
+      "epoch": 0.18586906760262475,
+      "grad_norm": 1.4804611004553776,
+      "learning_rate": 9.377525725443341e-06,
+      "loss": 0.3951,
+      "step": 609
+    },
+    {
+      "epoch": 0.18617427132611017,
+      "grad_norm": 1.4589508747304376,
+      "learning_rate": 9.375134468171108e-06,
+      "loss": 0.2887,
+      "step": 610
+    },
+    {
+      "epoch": 0.1864794750495956,
+      "grad_norm": 1.9199770728371568,
+      "learning_rate": 9.372738932742193e-06,
+      "loss": 0.5627,
+      "step": 611
+    },
+    {
+      "epoch": 0.18678467877308103,
+      "grad_norm": 1.5903295576095668,
+      "learning_rate": 9.370339121499039e-06,
+      "loss": 0.4379,
+      "step": 612
+    },
+    {
+      "epoch": 0.18708988249656647,
+      "grad_norm": 1.6986262549100166,
+      "learning_rate": 9.367935036788257e-06,
+      "loss": 0.4873,
+      "step": 613
+    },
+    {
+      "epoch": 0.18739508622005188,
+      "grad_norm": 1.2194659257752518,
+      "learning_rate": 9.365526680960645e-06,
+      "loss": 0.3571,
+      "step": 614
+    },
+    {
+      "epoch": 0.18770028994353732,
+      "grad_norm": 1.3049138038493902,
+      "learning_rate": 9.363114056371178e-06,
+      "loss": 0.3114,
+      "step": 615
+    },
+    {
+      "epoch": 0.18800549366702274,
+      "grad_norm": 1.468685879220778,
+      "learning_rate": 9.360697165379004e-06,
+      "loss": 0.5043,
+      "step": 616
+    },
+    {
+      "epoch": 0.18831069739050815,
+      "grad_norm": 2.131454426646245,
+      "learning_rate": 9.35827601034744e-06,
+      "loss": 0.5165,
+      "step": 617
+    },
+    {
+      "epoch": 0.1886159011139936,
+      "grad_norm": 1.5563065422847613,
+      "learning_rate": 9.355850593643974e-06,
+      "loss": 0.4707,
+      "step": 618
+    },
+    {
+      "epoch": 0.188921104837479,
+      "grad_norm": 1.3992788671446874,
+      "learning_rate": 9.353420917640264e-06,
+      "loss": 0.3905,
+      "step": 619
+    },
+    {
+      "epoch": 0.18922630856096445,
+      "grad_norm": 1.2635210704880713,
+      "learning_rate": 9.350986984712128e-06,
+      "loss": 0.2399,
+      "step": 620
+    },
+    {
+      "epoch": 0.18953151228444987,
+      "grad_norm": 1.4071894580574145,
+      "learning_rate": 9.348548797239551e-06,
+      "loss": 0.3689,
+      "step": 621
+    },
+    {
+      "epoch": 0.1898367160079353,
+      "grad_norm": 1.8460324916935194,
+      "learning_rate": 9.346106357606675e-06,
+      "loss": 0.3337,
+      "step": 622
+    },
+    {
+      "epoch": 0.19014191973142072,
+      "grad_norm": 1.2199756132877848,
+      "learning_rate": 9.343659668201803e-06,
+      "loss": 0.3707,
+      "step": 623
+    },
+    {
+      "epoch": 0.19044712345490614,
+      "grad_norm": 1.3352934416971625,
+      "learning_rate": 9.34120873141739e-06,
+      "loss": 0.3332,
+      "step": 624
+    },
+    {
+      "epoch": 0.19075232717839158,
+      "grad_norm": 1.5847999740161538,
+      "learning_rate": 9.33875354965005e-06,
+      "loss": 0.4658,
+      "step": 625
+    },
+    {
+      "epoch": 0.191057530901877,
+      "grad_norm": 1.59267718540602,
+      "learning_rate": 9.336294125300538e-06,
+      "loss": 0.5138,
+      "step": 626
+    },
+    {
+      "epoch": 0.19136273462536244,
+      "grad_norm": 1.0558131110089597,
+      "learning_rate": 9.333830460773767e-06,
+      "loss": 0.2512,
+      "step": 627
+    },
+    {
+      "epoch": 0.19166793834884785,
+      "grad_norm": 1.6912848096424418,
+      "learning_rate": 9.331362558478793e-06,
+      "loss": 0.3129,
+      "step": 628
+    },
+    {
+      "epoch": 0.1919731420723333,
+      "grad_norm": 1.7565199783626735,
+      "learning_rate": 9.328890420828817e-06,
+      "loss": 0.2625,
+      "step": 629
+    },
+    {
+      "epoch": 0.1922783457958187,
+      "grad_norm": 1.6287483120468187,
+      "learning_rate": 9.326414050241176e-06,
+      "loss": 0.4631,
+      "step": 630
+    },
+    {
+      "epoch": 0.19258354951930415,
+      "grad_norm": 1.5343127080699748,
+      "learning_rate": 9.323933449137353e-06,
+      "loss": 0.35,
+      "step": 631
+    },
+    {
+      "epoch": 0.19288875324278956,
+      "grad_norm": 1.2901421851525343,
+      "learning_rate": 9.321448619942963e-06,
+      "loss": 0.3191,
+      "step": 632
+    },
+    {
+      "epoch": 0.19319395696627498,
+      "grad_norm": 1.3651871352420322,
+      "learning_rate": 9.318959565087761e-06,
+      "loss": 0.3063,
+      "step": 633
+    },
+    {
+      "epoch": 0.19349916068976042,
+      "grad_norm": 1.4900191765389657,
+      "learning_rate": 9.316466287005625e-06,
+      "loss": 0.3621,
+      "step": 634
+    },
+    {
+      "epoch": 0.19380436441324583,
+      "grad_norm": 1.836926327897149,
+      "learning_rate": 9.313968788134572e-06,
+      "loss": 0.6273,
+      "step": 635
+    },
+    {
+      "epoch": 0.19410956813673128,
+      "grad_norm": 1.6283757285797815,
+      "learning_rate": 9.311467070916743e-06,
+      "loss": 0.3191,
+      "step": 636
+    },
+    {
+      "epoch": 0.1944147718602167,
+      "grad_norm": 1.7047955780313857,
+      "learning_rate": 9.308961137798398e-06,
+      "loss": 0.5581,
+      "step": 637
+    },
+    {
+      "epoch": 0.19471997558370213,
+      "grad_norm": 1.4826549000090183,
+      "learning_rate": 9.306450991229927e-06,
+      "loss": 0.3157,
+      "step": 638
+    },
+    {
+      "epoch": 0.19502517930718755,
+      "grad_norm": 1.435361017145943,
+      "learning_rate": 9.30393663366584e-06,
+      "loss": 0.3084,
+      "step": 639
+    },
+    {
+      "epoch": 0.19533038303067296,
+      "grad_norm": 1.402358583674702,
+      "learning_rate": 9.301418067564758e-06,
+      "loss": 0.3351,
+      "step": 640
+    },
+    {
+      "epoch": 0.1956355867541584,
+      "grad_norm": 1.2933654176691274,
+      "learning_rate": 9.298895295389423e-06,
+      "loss": 0.2585,
+      "step": 641
+    },
+    {
+      "epoch": 0.19594079047764382,
+      "grad_norm": 1.5346090103364156,
+      "learning_rate": 9.29636831960669e-06,
+      "loss": 0.33,
+      "step": 642
+    },
+    {
+      "epoch": 0.19624599420112926,
+      "grad_norm": 1.3251725534327445,
+      "learning_rate": 9.293837142687522e-06,
+      "loss": 0.2104,
+      "step": 643
+    },
+    {
+      "epoch": 0.19655119792461467,
+      "grad_norm": 1.583285740923444,
+      "learning_rate": 9.291301767106986e-06,
+      "loss": 0.4326,
+      "step": 644
+    },
+    {
+      "epoch": 0.19685640164810012,
+      "grad_norm": 1.0905371753559963,
+      "learning_rate": 9.288762195344266e-06,
+      "loss": 0.274,
+      "step": 645
+    },
+    {
+      "epoch": 0.19716160537158553,
+      "grad_norm": 1.7263838041187525,
+      "learning_rate": 9.28621842988264e-06,
+      "loss": 0.5011,
+      "step": 646
+    },
+    {
+      "epoch": 0.19746680909507097,
+      "grad_norm": 1.4838510492072716,
+      "learning_rate": 9.283670473209488e-06,
+      "loss": 0.1956,
+      "step": 647
+    },
+    {
+      "epoch": 0.1977720128185564,
+      "grad_norm": 1.2036114489558822,
+      "learning_rate": 9.28111832781629e-06,
+      "loss": 0.2346,
+      "step": 648
+    },
+    {
+      "epoch": 0.1980772165420418,
+      "grad_norm": 1.3853687861302435,
+      "learning_rate": 9.278561996198622e-06,
+      "loss": 0.2313,
+      "step": 649
+    },
+    {
+      "epoch": 0.19838242026552724,
+      "grad_norm": 1.2708592907281826,
+      "learning_rate": 9.276001480856152e-06,
+      "loss": 0.3717,
+      "step": 650
+    },
+    {
+      "epoch": 0.19868762398901266,
+      "grad_norm": 1.6407943638167721,
+      "learning_rate": 9.273436784292641e-06,
+      "loss": 0.5533,
+      "step": 651
+    },
+    {
+      "epoch": 0.1989928277124981,
+      "grad_norm": 1.9614104046455092,
+      "learning_rate": 9.270867909015936e-06,
+      "loss": 0.4552,
+      "step": 652
+    },
+    {
+      "epoch": 0.1992980314359835,
+      "grad_norm": 1.6038724290911757,
+      "learning_rate": 9.268294857537973e-06,
+      "loss": 0.477,
+      "step": 653
+    },
+    {
+      "epoch": 0.19960323515946896,
+      "grad_norm": 1.9838192379267598,
+      "learning_rate": 9.26571763237477e-06,
+      "loss": 0.7118,
+      "step": 654
+    },
+    {
+      "epoch": 0.19990843888295437,
+      "grad_norm": 1.4066170955047037,
+      "learning_rate": 9.263136236046422e-06,
+      "loss": 0.4222,
+      "step": 655
+    },
+    {
+      "epoch": 0.2002136426064398,
+      "grad_norm": 1.800111592330467,
+      "learning_rate": 9.260550671077113e-06,
+      "loss": 0.4969,
+      "step": 656
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 3276,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 328,
+  "total_flos": 81687288602624.0,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}