{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 3.91644908616188, "eval_steps": 500, "global_step": 3000, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.0013054830287206266, "grad_norm": 71.86194610595703, "learning_rate": 0.0, "loss": 7.193, "step": 1 }, { "epoch": 0.0026109660574412533, "grad_norm": 58.82057571411133, "learning_rate": 4.347826086956522e-08, "loss": 6.7844, "step": 2 }, { "epoch": 0.0039164490861618795, "grad_norm": 58.3909912109375, "learning_rate": 8.695652173913044e-08, "loss": 6.8077, "step": 3 }, { "epoch": 0.005221932114882507, "grad_norm": 59.3935432434082, "learning_rate": 1.3043478260869566e-07, "loss": 6.893, "step": 4 }, { "epoch": 0.006527415143603133, "grad_norm": 61.26504135131836, "learning_rate": 1.7391304347826088e-07, "loss": 7.1543, "step": 5 }, { "epoch": 0.007832898172323759, "grad_norm": 63.678253173828125, "learning_rate": 2.173913043478261e-07, "loss": 7.147, "step": 6 }, { "epoch": 0.009138381201044387, "grad_norm": 80.34788513183594, "learning_rate": 2.608695652173913e-07, "loss": 7.0721, "step": 7 }, { "epoch": 0.010443864229765013, "grad_norm": 61.98629379272461, "learning_rate": 3.0434782608695656e-07, "loss": 6.9481, "step": 8 }, { "epoch": 0.01174934725848564, "grad_norm": 57.9570426940918, "learning_rate": 3.4782608695652175e-07, "loss": 6.5879, "step": 9 }, { "epoch": 0.013054830287206266, "grad_norm": 59.829891204833984, "learning_rate": 3.91304347826087e-07, "loss": 6.8486, "step": 10 }, { "epoch": 0.014360313315926894, "grad_norm": 56.93695831298828, "learning_rate": 4.347826086956522e-07, "loss": 6.4561, "step": 11 }, { "epoch": 0.015665796344647518, "grad_norm": 56.91865158081055, "learning_rate": 4.782608695652174e-07, "loss": 6.578, "step": 12 }, { "epoch": 0.016971279373368148, "grad_norm": 61.51951217651367, "learning_rate": 5.217391304347826e-07, "loss": 7.0115, "step": 13 }, { "epoch": 0.018276762402088774, "grad_norm": 61.58373260498047, "learning_rate": 5.652173913043478e-07, "loss": 6.2652, "step": 14 }, { "epoch": 0.0195822454308094, "grad_norm": 54.762638092041016, "learning_rate": 6.086956521739131e-07, "loss": 6.1993, "step": 15 }, { "epoch": 0.020887728459530026, "grad_norm": 58.63661193847656, "learning_rate": 6.521739130434783e-07, "loss": 6.6445, "step": 16 }, { "epoch": 0.022193211488250653, "grad_norm": 61.23152542114258, "learning_rate": 6.956521739130435e-07, "loss": 6.4053, "step": 17 }, { "epoch": 0.02349869451697128, "grad_norm": 58.712669372558594, "learning_rate": 7.391304347826088e-07, "loss": 6.5313, "step": 18 }, { "epoch": 0.024804177545691905, "grad_norm": 56.684932708740234, "learning_rate": 7.82608695652174e-07, "loss": 6.2405, "step": 19 }, { "epoch": 0.02610966057441253, "grad_norm": 63.16392135620117, "learning_rate": 8.260869565217392e-07, "loss": 6.7795, "step": 20 }, { "epoch": 0.02741514360313316, "grad_norm": 58.998809814453125, "learning_rate": 8.695652173913044e-07, "loss": 6.2874, "step": 21 }, { "epoch": 0.028720626631853787, "grad_norm": 58.92256546020508, "learning_rate": 9.130434782608697e-07, "loss": 6.036, "step": 22 }, { "epoch": 0.030026109660574413, "grad_norm": 59.328006744384766, "learning_rate": 9.565217391304349e-07, "loss": 5.806, "step": 23 }, { "epoch": 0.031331592689295036, "grad_norm": 58.19785690307617, "learning_rate": 1.0000000000000002e-06, "loss": 6.027, "step": 24 }, { "epoch": 0.03263707571801567, "grad_norm": 50.844459533691406, "learning_rate": 1.0434782608695653e-06, "loss": 5.8045, "step": 25 }, { "epoch": 0.033942558746736295, "grad_norm": 51.77566146850586, "learning_rate": 1.0869565217391306e-06, "loss": 5.5728, "step": 26 }, { "epoch": 0.03524804177545692, "grad_norm": 50.79306411743164, "learning_rate": 1.1304347826086956e-06, "loss": 5.5591, "step": 27 }, { "epoch": 0.03655352480417755, "grad_norm": 49.766334533691406, "learning_rate": 1.173913043478261e-06, "loss": 5.4663, "step": 28 }, { "epoch": 0.037859007832898174, "grad_norm": 38.867008209228516, "learning_rate": 1.2173913043478262e-06, "loss": 4.867, "step": 29 }, { "epoch": 0.0391644908616188, "grad_norm": 34.57432556152344, "learning_rate": 1.2608695652173913e-06, "loss": 4.5507, "step": 30 }, { "epoch": 0.04046997389033943, "grad_norm": 30.742935180664062, "learning_rate": 1.3043478260869566e-06, "loss": 4.3219, "step": 31 }, { "epoch": 0.04177545691906005, "grad_norm": 31.543106079101562, "learning_rate": 1.347826086956522e-06, "loss": 4.2897, "step": 32 }, { "epoch": 0.04308093994778068, "grad_norm": 27.904544830322266, "learning_rate": 1.391304347826087e-06, "loss": 4.1886, "step": 33 }, { "epoch": 0.044386422976501305, "grad_norm": 27.338912963867188, "learning_rate": 1.4347826086956523e-06, "loss": 4.1003, "step": 34 }, { "epoch": 0.04569190600522193, "grad_norm": 28.656185150146484, "learning_rate": 1.4782608695652176e-06, "loss": 3.9418, "step": 35 }, { "epoch": 0.04699738903394256, "grad_norm": 26.85036849975586, "learning_rate": 1.521739130434783e-06, "loss": 3.747, "step": 36 }, { "epoch": 0.048302872062663184, "grad_norm": 29.934864044189453, "learning_rate": 1.565217391304348e-06, "loss": 3.5666, "step": 37 }, { "epoch": 0.04960835509138381, "grad_norm": 23.61662483215332, "learning_rate": 1.608695652173913e-06, "loss": 3.6716, "step": 38 }, { "epoch": 0.050913838120104436, "grad_norm": 24.50541114807129, "learning_rate": 1.6521739130434784e-06, "loss": 3.3028, "step": 39 }, { "epoch": 0.05221932114882506, "grad_norm": 34.06825637817383, "learning_rate": 1.6956521739130435e-06, "loss": 3.0356, "step": 40 }, { "epoch": 0.05352480417754569, "grad_norm": 21.728958129882812, "learning_rate": 1.7391304347826088e-06, "loss": 2.8856, "step": 41 }, { "epoch": 0.05483028720626632, "grad_norm": 16.015396118164062, "learning_rate": 1.782608695652174e-06, "loss": 2.9847, "step": 42 }, { "epoch": 0.05613577023498695, "grad_norm": 14.58934497833252, "learning_rate": 1.8260869565217394e-06, "loss": 2.8832, "step": 43 }, { "epoch": 0.057441253263707574, "grad_norm": 12.964859962463379, "learning_rate": 1.8695652173913044e-06, "loss": 2.9581, "step": 44 }, { "epoch": 0.0587467362924282, "grad_norm": 19.35062026977539, "learning_rate": 1.9130434782608697e-06, "loss": 2.5569, "step": 45 }, { "epoch": 0.06005221932114883, "grad_norm": 17.58714485168457, "learning_rate": 1.956521739130435e-06, "loss": 2.7137, "step": 46 }, { "epoch": 0.06135770234986945, "grad_norm": 18.694185256958008, "learning_rate": 2.0000000000000003e-06, "loss": 2.4987, "step": 47 }, { "epoch": 0.06266318537859007, "grad_norm": 16.908672332763672, "learning_rate": 2.0434782608695656e-06, "loss": 2.6439, "step": 48 }, { "epoch": 0.0639686684073107, "grad_norm": 11.977069854736328, "learning_rate": 2.0869565217391305e-06, "loss": 2.5277, "step": 49 }, { "epoch": 0.06527415143603134, "grad_norm": 11.961771965026855, "learning_rate": 2.130434782608696e-06, "loss": 2.2827, "step": 50 }, { "epoch": 0.06657963446475196, "grad_norm": 10.58080768585205, "learning_rate": 2.173913043478261e-06, "loss": 2.3296, "step": 51 }, { "epoch": 0.06788511749347259, "grad_norm": 9.911226272583008, "learning_rate": 2.2173913043478264e-06, "loss": 2.1657, "step": 52 }, { "epoch": 0.06919060052219321, "grad_norm": 10.240909576416016, "learning_rate": 2.2608695652173913e-06, "loss": 2.1944, "step": 53 }, { "epoch": 0.07049608355091384, "grad_norm": 11.288805961608887, "learning_rate": 2.3043478260869566e-06, "loss": 2.1338, "step": 54 }, { "epoch": 0.07180156657963446, "grad_norm": 8.695660591125488, "learning_rate": 2.347826086956522e-06, "loss": 1.806, "step": 55 }, { "epoch": 0.0731070496083551, "grad_norm": 8.733399391174316, "learning_rate": 2.391304347826087e-06, "loss": 2.0047, "step": 56 }, { "epoch": 0.07441253263707572, "grad_norm": 8.374787330627441, "learning_rate": 2.4347826086956525e-06, "loss": 1.9046, "step": 57 }, { "epoch": 0.07571801566579635, "grad_norm": 7.685578346252441, "learning_rate": 2.4782608695652178e-06, "loss": 1.6568, "step": 58 }, { "epoch": 0.07702349869451697, "grad_norm": 7.295040130615234, "learning_rate": 2.5217391304347826e-06, "loss": 1.6446, "step": 59 }, { "epoch": 0.0783289817232376, "grad_norm": 8.285629272460938, "learning_rate": 2.5652173913043484e-06, "loss": 1.6175, "step": 60 }, { "epoch": 0.07963446475195822, "grad_norm": 7.258594512939453, "learning_rate": 2.6086956521739132e-06, "loss": 1.7802, "step": 61 }, { "epoch": 0.08093994778067885, "grad_norm": 7.439572811126709, "learning_rate": 2.6521739130434785e-06, "loss": 2.0383, "step": 62 }, { "epoch": 0.08224543080939947, "grad_norm": 6.78455114364624, "learning_rate": 2.695652173913044e-06, "loss": 1.8115, "step": 63 }, { "epoch": 0.0835509138381201, "grad_norm": 6.864508628845215, "learning_rate": 2.7391304347826087e-06, "loss": 1.4908, "step": 64 }, { "epoch": 0.08485639686684072, "grad_norm": 6.967901229858398, "learning_rate": 2.782608695652174e-06, "loss": 2.005, "step": 65 }, { "epoch": 0.08616187989556136, "grad_norm": 6.503068923950195, "learning_rate": 2.8260869565217393e-06, "loss": 1.7264, "step": 66 }, { "epoch": 0.08746736292428199, "grad_norm": 8.195791244506836, "learning_rate": 2.8695652173913046e-06, "loss": 1.8441, "step": 67 }, { "epoch": 0.08877284595300261, "grad_norm": 6.543520927429199, "learning_rate": 2.9130434782608695e-06, "loss": 1.7226, "step": 68 }, { "epoch": 0.09007832898172324, "grad_norm": 6.958765506744385, "learning_rate": 2.956521739130435e-06, "loss": 1.5904, "step": 69 }, { "epoch": 0.09138381201044386, "grad_norm": 7.033638954162598, "learning_rate": 3e-06, "loss": 1.747, "step": 70 }, { "epoch": 0.0926892950391645, "grad_norm": 6.882011890411377, "learning_rate": 3.043478260869566e-06, "loss": 1.529, "step": 71 }, { "epoch": 0.09399477806788512, "grad_norm": 7.19775915145874, "learning_rate": 3.0869565217391307e-06, "loss": 1.614, "step": 72 }, { "epoch": 0.09530026109660575, "grad_norm": 6.953028678894043, "learning_rate": 3.130434782608696e-06, "loss": 1.4833, "step": 73 }, { "epoch": 0.09660574412532637, "grad_norm": 6.080812931060791, "learning_rate": 3.1739130434782613e-06, "loss": 1.711, "step": 74 }, { "epoch": 0.097911227154047, "grad_norm": 5.82197904586792, "learning_rate": 3.217391304347826e-06, "loss": 1.3938, "step": 75 }, { "epoch": 0.09921671018276762, "grad_norm": 8.465730667114258, "learning_rate": 3.2608695652173914e-06, "loss": 1.4442, "step": 76 }, { "epoch": 0.10052219321148825, "grad_norm": 6.07127571105957, "learning_rate": 3.3043478260869567e-06, "loss": 1.5917, "step": 77 }, { "epoch": 0.10182767624020887, "grad_norm": 5.691837310791016, "learning_rate": 3.347826086956522e-06, "loss": 1.5098, "step": 78 }, { "epoch": 0.1031331592689295, "grad_norm": 5.80321741104126, "learning_rate": 3.391304347826087e-06, "loss": 1.4531, "step": 79 }, { "epoch": 0.10443864229765012, "grad_norm": 5.356811046600342, "learning_rate": 3.4347826086956526e-06, "loss": 1.4953, "step": 80 }, { "epoch": 0.10574412532637076, "grad_norm": 5.728530406951904, "learning_rate": 3.4782608695652175e-06, "loss": 1.335, "step": 81 }, { "epoch": 0.10704960835509138, "grad_norm": 5.913639545440674, "learning_rate": 3.5217391304347832e-06, "loss": 1.2438, "step": 82 }, { "epoch": 0.10835509138381201, "grad_norm": 6.109486103057861, "learning_rate": 3.565217391304348e-06, "loss": 1.4674, "step": 83 }, { "epoch": 0.10966057441253264, "grad_norm": 5.304638862609863, "learning_rate": 3.6086956521739134e-06, "loss": 1.3314, "step": 84 }, { "epoch": 0.11096605744125326, "grad_norm": 7.0575947761535645, "learning_rate": 3.6521739130434787e-06, "loss": 1.5562, "step": 85 }, { "epoch": 0.1122715404699739, "grad_norm": 6.692266941070557, "learning_rate": 3.6956521739130436e-06, "loss": 1.4454, "step": 86 }, { "epoch": 0.11357702349869452, "grad_norm": 5.039459228515625, "learning_rate": 3.739130434782609e-06, "loss": 1.4114, "step": 87 }, { "epoch": 0.11488250652741515, "grad_norm": 4.929737567901611, "learning_rate": 3.782608695652174e-06, "loss": 1.474, "step": 88 }, { "epoch": 0.11618798955613577, "grad_norm": 5.124248027801514, "learning_rate": 3.8260869565217395e-06, "loss": 1.4531, "step": 89 }, { "epoch": 0.1174934725848564, "grad_norm": 4.9309539794921875, "learning_rate": 3.869565217391304e-06, "loss": 1.5903, "step": 90 }, { "epoch": 0.11879895561357702, "grad_norm": 7.606995105743408, "learning_rate": 3.91304347826087e-06, "loss": 1.3618, "step": 91 }, { "epoch": 0.12010443864229765, "grad_norm": 4.079611301422119, "learning_rate": 3.956521739130435e-06, "loss": 1.5122, "step": 92 }, { "epoch": 0.12140992167101827, "grad_norm": 5.476497173309326, "learning_rate": 4.000000000000001e-06, "loss": 1.4746, "step": 93 }, { "epoch": 0.1227154046997389, "grad_norm": 5.848145961761475, "learning_rate": 4.0434782608695655e-06, "loss": 1.3182, "step": 94 }, { "epoch": 0.12402088772845953, "grad_norm": 4.475169658660889, "learning_rate": 4.086956521739131e-06, "loss": 1.555, "step": 95 }, { "epoch": 0.12532637075718014, "grad_norm": 5.525176525115967, "learning_rate": 4.130434782608696e-06, "loss": 1.2863, "step": 96 }, { "epoch": 0.12663185378590078, "grad_norm": 4.61910343170166, "learning_rate": 4.173913043478261e-06, "loss": 1.4348, "step": 97 }, { "epoch": 0.1279373368146214, "grad_norm": 4.7762956619262695, "learning_rate": 4.217391304347827e-06, "loss": 1.4197, "step": 98 }, { "epoch": 0.12924281984334204, "grad_norm": 5.008528709411621, "learning_rate": 4.260869565217392e-06, "loss": 1.2634, "step": 99 }, { "epoch": 0.13054830287206268, "grad_norm": 4.664148807525635, "learning_rate": 4.304347826086957e-06, "loss": 1.4136, "step": 100 }, { "epoch": 0.13185378590078328, "grad_norm": 5.766796588897705, "learning_rate": 4.347826086956522e-06, "loss": 1.2996, "step": 101 }, { "epoch": 0.13315926892950392, "grad_norm": 4.839200019836426, "learning_rate": 4.391304347826087e-06, "loss": 1.3461, "step": 102 }, { "epoch": 0.13446475195822455, "grad_norm": 4.718023777008057, "learning_rate": 4.434782608695653e-06, "loss": 1.4692, "step": 103 }, { "epoch": 0.13577023498694518, "grad_norm": 4.648873805999756, "learning_rate": 4.478260869565218e-06, "loss": 1.3517, "step": 104 }, { "epoch": 0.1370757180156658, "grad_norm": 4.951013088226318, "learning_rate": 4.5217391304347826e-06, "loss": 1.3022, "step": 105 }, { "epoch": 0.13838120104438642, "grad_norm": 4.690581321716309, "learning_rate": 4.565217391304348e-06, "loss": 1.3184, "step": 106 }, { "epoch": 0.13968668407310705, "grad_norm": 5.579658508300781, "learning_rate": 4.608695652173913e-06, "loss": 1.4332, "step": 107 }, { "epoch": 0.1409921671018277, "grad_norm": 5.342446327209473, "learning_rate": 4.652173913043478e-06, "loss": 1.1963, "step": 108 }, { "epoch": 0.1422976501305483, "grad_norm": 5.018988132476807, "learning_rate": 4.695652173913044e-06, "loss": 1.4055, "step": 109 }, { "epoch": 0.14360313315926893, "grad_norm": 4.684205532073975, "learning_rate": 4.739130434782609e-06, "loss": 1.4202, "step": 110 }, { "epoch": 0.14490861618798956, "grad_norm": 5.469826698303223, "learning_rate": 4.782608695652174e-06, "loss": 1.2459, "step": 111 }, { "epoch": 0.1462140992167102, "grad_norm": 5.289001941680908, "learning_rate": 4.826086956521739e-06, "loss": 1.3096, "step": 112 }, { "epoch": 0.1475195822454308, "grad_norm": 6.0019307136535645, "learning_rate": 4.869565217391305e-06, "loss": 1.2713, "step": 113 }, { "epoch": 0.14882506527415143, "grad_norm": 5.404791831970215, "learning_rate": 4.91304347826087e-06, "loss": 1.2938, "step": 114 }, { "epoch": 0.15013054830287206, "grad_norm": 5.542120456695557, "learning_rate": 4.9565217391304355e-06, "loss": 1.1835, "step": 115 }, { "epoch": 0.1514360313315927, "grad_norm": 5.147332668304443, "learning_rate": 5e-06, "loss": 1.2996, "step": 116 }, { "epoch": 0.15274151436031333, "grad_norm": 5.533376216888428, "learning_rate": 5.043478260869565e-06, "loss": 1.2003, "step": 117 }, { "epoch": 0.15404699738903394, "grad_norm": 5.008963108062744, "learning_rate": 5.08695652173913e-06, "loss": 1.3246, "step": 118 }, { "epoch": 0.15535248041775457, "grad_norm": 5.555875778198242, "learning_rate": 5.130434782608697e-06, "loss": 1.135, "step": 119 }, { "epoch": 0.1566579634464752, "grad_norm": 4.46732759475708, "learning_rate": 5.173913043478262e-06, "loss": 1.2538, "step": 120 }, { "epoch": 0.15796344647519583, "grad_norm": 4.604671001434326, "learning_rate": 5.2173913043478265e-06, "loss": 1.3874, "step": 121 }, { "epoch": 0.15926892950391644, "grad_norm": 5.569374084472656, "learning_rate": 5.260869565217391e-06, "loss": 1.1093, "step": 122 }, { "epoch": 0.16057441253263707, "grad_norm": 5.53762149810791, "learning_rate": 5.304347826086957e-06, "loss": 1.1716, "step": 123 }, { "epoch": 0.1618798955613577, "grad_norm": 5.3505635261535645, "learning_rate": 5.347826086956523e-06, "loss": 1.2468, "step": 124 }, { "epoch": 0.16318537859007834, "grad_norm": 4.739141941070557, "learning_rate": 5.391304347826088e-06, "loss": 1.2914, "step": 125 }, { "epoch": 0.16449086161879894, "grad_norm": 5.153298377990723, "learning_rate": 5.4347826086956525e-06, "loss": 1.3058, "step": 126 }, { "epoch": 0.16579634464751958, "grad_norm": 5.608656406402588, "learning_rate": 5.478260869565217e-06, "loss": 1.1396, "step": 127 }, { "epoch": 0.1671018276762402, "grad_norm": 4.71812629699707, "learning_rate": 5.521739130434783e-06, "loss": 1.3091, "step": 128 }, { "epoch": 0.16840731070496084, "grad_norm": 5.45646333694458, "learning_rate": 5.565217391304348e-06, "loss": 1.1863, "step": 129 }, { "epoch": 0.16971279373368145, "grad_norm": 4.760155200958252, "learning_rate": 5.608695652173914e-06, "loss": 1.3157, "step": 130 }, { "epoch": 0.17101827676240208, "grad_norm": 4.585694789886475, "learning_rate": 5.652173913043479e-06, "loss": 1.2637, "step": 131 }, { "epoch": 0.17232375979112272, "grad_norm": 4.066949844360352, "learning_rate": 5.695652173913044e-06, "loss": 1.2527, "step": 132 }, { "epoch": 0.17362924281984335, "grad_norm": 4.325253963470459, "learning_rate": 5.739130434782609e-06, "loss": 1.2208, "step": 133 }, { "epoch": 0.17493472584856398, "grad_norm": 4.59459114074707, "learning_rate": 5.782608695652174e-06, "loss": 1.1922, "step": 134 }, { "epoch": 0.1762402088772846, "grad_norm": 5.406242847442627, "learning_rate": 5.826086956521739e-06, "loss": 1.1057, "step": 135 }, { "epoch": 0.17754569190600522, "grad_norm": 6.997798442840576, "learning_rate": 5.8695652173913055e-06, "loss": 1.0743, "step": 136 }, { "epoch": 0.17885117493472585, "grad_norm": 5.093508243560791, "learning_rate": 5.91304347826087e-06, "loss": 1.198, "step": 137 }, { "epoch": 0.1801566579634465, "grad_norm": 5.264876365661621, "learning_rate": 5.956521739130435e-06, "loss": 1.1481, "step": 138 }, { "epoch": 0.1814621409921671, "grad_norm": 5.140352249145508, "learning_rate": 6e-06, "loss": 1.2606, "step": 139 }, { "epoch": 0.18276762402088773, "grad_norm": 5.183084487915039, "learning_rate": 6.043478260869565e-06, "loss": 1.232, "step": 140 }, { "epoch": 0.18407310704960836, "grad_norm": 5.368699550628662, "learning_rate": 6.086956521739132e-06, "loss": 1.184, "step": 141 }, { "epoch": 0.185378590078329, "grad_norm": 5.232748985290527, "learning_rate": 6.1304347826086965e-06, "loss": 1.1276, "step": 142 }, { "epoch": 0.1866840731070496, "grad_norm": 4.855964183807373, "learning_rate": 6.173913043478261e-06, "loss": 1.3108, "step": 143 }, { "epoch": 0.18798955613577023, "grad_norm": 5.073291778564453, "learning_rate": 6.217391304347826e-06, "loss": 1.2408, "step": 144 }, { "epoch": 0.18929503916449086, "grad_norm": 3.892029285430908, "learning_rate": 6.260869565217392e-06, "loss": 1.2822, "step": 145 }, { "epoch": 0.1906005221932115, "grad_norm": 4.541572093963623, "learning_rate": 6.304347826086958e-06, "loss": 1.1654, "step": 146 }, { "epoch": 0.1919060052219321, "grad_norm": 5.486248970031738, "learning_rate": 6.3478260869565225e-06, "loss": 1.1812, "step": 147 }, { "epoch": 0.19321148825065274, "grad_norm": 4.299222946166992, "learning_rate": 6.391304347826087e-06, "loss": 1.2974, "step": 148 }, { "epoch": 0.19451697127937337, "grad_norm": 4.5948920249938965, "learning_rate": 6.434782608695652e-06, "loss": 1.0539, "step": 149 }, { "epoch": 0.195822454308094, "grad_norm": 4.492305278778076, "learning_rate": 6.478260869565218e-06, "loss": 1.2622, "step": 150 }, { "epoch": 0.19712793733681463, "grad_norm": 4.918149471282959, "learning_rate": 6.521739130434783e-06, "loss": 1.153, "step": 151 }, { "epoch": 0.19843342036553524, "grad_norm": 5.152765274047852, "learning_rate": 6.565217391304349e-06, "loss": 1.2198, "step": 152 }, { "epoch": 0.19973890339425587, "grad_norm": 4.229496955871582, "learning_rate": 6.6086956521739135e-06, "loss": 1.3567, "step": 153 }, { "epoch": 0.2010443864229765, "grad_norm": 3.853389263153076, "learning_rate": 6.652173913043479e-06, "loss": 1.3444, "step": 154 }, { "epoch": 0.20234986945169714, "grad_norm": 5.4623541831970215, "learning_rate": 6.695652173913044e-06, "loss": 1.1967, "step": 155 }, { "epoch": 0.20365535248041775, "grad_norm": 4.875569820404053, "learning_rate": 6.739130434782609e-06, "loss": 1.1035, "step": 156 }, { "epoch": 0.20496083550913838, "grad_norm": 6.2351274490356445, "learning_rate": 6.782608695652174e-06, "loss": 1.0545, "step": 157 }, { "epoch": 0.206266318537859, "grad_norm": 4.612958908081055, "learning_rate": 6.8260869565217395e-06, "loss": 1.2488, "step": 158 }, { "epoch": 0.20757180156657964, "grad_norm": 4.716128349304199, "learning_rate": 6.869565217391305e-06, "loss": 1.1488, "step": 159 }, { "epoch": 0.20887728459530025, "grad_norm": 4.888503551483154, "learning_rate": 6.91304347826087e-06, "loss": 1.2064, "step": 160 }, { "epoch": 0.21018276762402088, "grad_norm": 5.530930995941162, "learning_rate": 6.956521739130435e-06, "loss": 1.1556, "step": 161 }, { "epoch": 0.21148825065274152, "grad_norm": 4.6509928703308105, "learning_rate": 7e-06, "loss": 1.1813, "step": 162 }, { "epoch": 0.21279373368146215, "grad_norm": 5.7834014892578125, "learning_rate": 7.0434782608695665e-06, "loss": 1.2709, "step": 163 }, { "epoch": 0.21409921671018275, "grad_norm": 4.15659761428833, "learning_rate": 7.086956521739131e-06, "loss": 1.2317, "step": 164 }, { "epoch": 0.2154046997389034, "grad_norm": 5.3440423011779785, "learning_rate": 7.130434782608696e-06, "loss": 1.0472, "step": 165 }, { "epoch": 0.21671018276762402, "grad_norm": 5.057955741882324, "learning_rate": 7.173913043478261e-06, "loss": 1.1973, "step": 166 }, { "epoch": 0.21801566579634465, "grad_norm": 4.623597621917725, "learning_rate": 7.217391304347827e-06, "loss": 1.0377, "step": 167 }, { "epoch": 0.2193211488250653, "grad_norm": 4.733643531799316, "learning_rate": 7.2608695652173925e-06, "loss": 1.162, "step": 168 }, { "epoch": 0.2206266318537859, "grad_norm": 5.628612518310547, "learning_rate": 7.304347826086957e-06, "loss": 1.1525, "step": 169 }, { "epoch": 0.22193211488250653, "grad_norm": 5.475170135498047, "learning_rate": 7.347826086956522e-06, "loss": 1.0749, "step": 170 }, { "epoch": 0.22323759791122716, "grad_norm": 4.44382905960083, "learning_rate": 7.391304347826087e-06, "loss": 1.2297, "step": 171 }, { "epoch": 0.2245430809399478, "grad_norm": 5.323819160461426, "learning_rate": 7.434782608695653e-06, "loss": 1.1552, "step": 172 }, { "epoch": 0.2258485639686684, "grad_norm": 4.817879676818848, "learning_rate": 7.478260869565218e-06, "loss": 1.1442, "step": 173 }, { "epoch": 0.22715404699738903, "grad_norm": 4.210514068603516, "learning_rate": 7.5217391304347835e-06, "loss": 1.1588, "step": 174 }, { "epoch": 0.22845953002610966, "grad_norm": 5.8715972900390625, "learning_rate": 7.565217391304348e-06, "loss": 1.1381, "step": 175 }, { "epoch": 0.2297650130548303, "grad_norm": 4.18674373626709, "learning_rate": 7.608695652173914e-06, "loss": 1.2203, "step": 176 }, { "epoch": 0.2310704960835509, "grad_norm": 5.350541114807129, "learning_rate": 7.652173913043479e-06, "loss": 1.1423, "step": 177 }, { "epoch": 0.23237597911227154, "grad_norm": 3.548884153366089, "learning_rate": 7.695652173913044e-06, "loss": 1.2836, "step": 178 }, { "epoch": 0.23368146214099217, "grad_norm": 4.7561798095703125, "learning_rate": 7.739130434782609e-06, "loss": 1.1088, "step": 179 }, { "epoch": 0.2349869451697128, "grad_norm": 4.444833755493164, "learning_rate": 7.782608695652174e-06, "loss": 1.2267, "step": 180 }, { "epoch": 0.2362924281984334, "grad_norm": 4.664194583892822, "learning_rate": 7.82608695652174e-06, "loss": 1.0252, "step": 181 }, { "epoch": 0.23759791122715404, "grad_norm": 3.8590571880340576, "learning_rate": 7.869565217391305e-06, "loss": 1.1552, "step": 182 }, { "epoch": 0.23890339425587467, "grad_norm": 4.627655506134033, "learning_rate": 7.91304347826087e-06, "loss": 1.1205, "step": 183 }, { "epoch": 0.2402088772845953, "grad_norm": 4.328120231628418, "learning_rate": 7.956521739130435e-06, "loss": 1.1293, "step": 184 }, { "epoch": 0.24151436031331594, "grad_norm": 4.3291707038879395, "learning_rate": 8.000000000000001e-06, "loss": 1.1649, "step": 185 }, { "epoch": 0.24281984334203655, "grad_norm": 5.821404457092285, "learning_rate": 8.043478260869566e-06, "loss": 1.0572, "step": 186 }, { "epoch": 0.24412532637075718, "grad_norm": 5.093186378479004, "learning_rate": 8.086956521739131e-06, "loss": 1.1317, "step": 187 }, { "epoch": 0.2454308093994778, "grad_norm": 5.693207740783691, "learning_rate": 8.130434782608696e-06, "loss": 1.0729, "step": 188 }, { "epoch": 0.24673629242819844, "grad_norm": 4.409574031829834, "learning_rate": 8.173913043478263e-06, "loss": 1.1985, "step": 189 }, { "epoch": 0.24804177545691905, "grad_norm": 3.981541156768799, "learning_rate": 8.217391304347827e-06, "loss": 1.176, "step": 190 }, { "epoch": 0.24934725848563968, "grad_norm": 5.060450077056885, "learning_rate": 8.260869565217392e-06, "loss": 1.0809, "step": 191 }, { "epoch": 0.2506527415143603, "grad_norm": 4.082695960998535, "learning_rate": 8.304347826086957e-06, "loss": 1.1121, "step": 192 }, { "epoch": 0.2519582245430809, "grad_norm": 5.351813316345215, "learning_rate": 8.347826086956522e-06, "loss": 1.1483, "step": 193 }, { "epoch": 0.25326370757180156, "grad_norm": 5.971114158630371, "learning_rate": 8.391304347826089e-06, "loss": 1.0515, "step": 194 }, { "epoch": 0.2545691906005222, "grad_norm": 4.145455837249756, "learning_rate": 8.434782608695653e-06, "loss": 1.1516, "step": 195 }, { "epoch": 0.2558746736292428, "grad_norm": 4.308621883392334, "learning_rate": 8.478260869565218e-06, "loss": 1.1791, "step": 196 }, { "epoch": 0.25718015665796345, "grad_norm": 4.447066307067871, "learning_rate": 8.521739130434783e-06, "loss": 1.1601, "step": 197 }, { "epoch": 0.2584856396866841, "grad_norm": 4.359739303588867, "learning_rate": 8.56521739130435e-06, "loss": 1.0169, "step": 198 }, { "epoch": 0.2597911227154047, "grad_norm": 4.467184066772461, "learning_rate": 8.608695652173915e-06, "loss": 1.1659, "step": 199 }, { "epoch": 0.26109660574412535, "grad_norm": 4.292892932891846, "learning_rate": 8.65217391304348e-06, "loss": 1.1761, "step": 200 }, { "epoch": 0.26240208877284593, "grad_norm": 4.494922637939453, "learning_rate": 8.695652173913044e-06, "loss": 1.1007, "step": 201 }, { "epoch": 0.26370757180156656, "grad_norm": 6.163149833679199, "learning_rate": 8.73913043478261e-06, "loss": 1.0207, "step": 202 }, { "epoch": 0.2650130548302872, "grad_norm": 4.168249607086182, "learning_rate": 8.782608695652174e-06, "loss": 1.1591, "step": 203 }, { "epoch": 0.26631853785900783, "grad_norm": 5.886813163757324, "learning_rate": 8.82608695652174e-06, "loss": 1.0525, "step": 204 }, { "epoch": 0.26762402088772846, "grad_norm": 4.867691516876221, "learning_rate": 8.869565217391306e-06, "loss": 1.0957, "step": 205 }, { "epoch": 0.2689295039164491, "grad_norm": 4.821184158325195, "learning_rate": 8.91304347826087e-06, "loss": 1.1106, "step": 206 }, { "epoch": 0.27023498694516973, "grad_norm": 5.410760879516602, "learning_rate": 8.956521739130435e-06, "loss": 1.113, "step": 207 }, { "epoch": 0.27154046997389036, "grad_norm": 4.879662036895752, "learning_rate": 9e-06, "loss": 1.0882, "step": 208 }, { "epoch": 0.27284595300261094, "grad_norm": 4.209608554840088, "learning_rate": 9.043478260869565e-06, "loss": 1.2546, "step": 209 }, { "epoch": 0.2741514360313316, "grad_norm": 3.73410964012146, "learning_rate": 9.086956521739132e-06, "loss": 1.1328, "step": 210 }, { "epoch": 0.2754569190600522, "grad_norm": 3.970329523086548, "learning_rate": 9.130434782608697e-06, "loss": 1.1907, "step": 211 }, { "epoch": 0.27676240208877284, "grad_norm": 4.084975242614746, "learning_rate": 9.173913043478261e-06, "loss": 1.1169, "step": 212 }, { "epoch": 0.2780678851174935, "grad_norm": 3.322183132171631, "learning_rate": 9.217391304347826e-06, "loss": 1.1638, "step": 213 }, { "epoch": 0.2793733681462141, "grad_norm": 5.285995006561279, "learning_rate": 9.260869565217391e-06, "loss": 1.0535, "step": 214 }, { "epoch": 0.28067885117493474, "grad_norm": 4.238169193267822, "learning_rate": 9.304347826086956e-06, "loss": 1.1849, "step": 215 }, { "epoch": 0.2819843342036554, "grad_norm": 3.605634927749634, "learning_rate": 9.347826086956523e-06, "loss": 1.0233, "step": 216 }, { "epoch": 0.283289817232376, "grad_norm": 5.667592525482178, "learning_rate": 9.391304347826087e-06, "loss": 1.0373, "step": 217 }, { "epoch": 0.2845953002610966, "grad_norm": 4.301891803741455, "learning_rate": 9.434782608695652e-06, "loss": 1.1323, "step": 218 }, { "epoch": 0.2859007832898172, "grad_norm": 4.9134087562561035, "learning_rate": 9.478260869565217e-06, "loss": 1.0368, "step": 219 }, { "epoch": 0.28720626631853785, "grad_norm": 5.021116256713867, "learning_rate": 9.521739130434784e-06, "loss": 1.0574, "step": 220 }, { "epoch": 0.2885117493472585, "grad_norm": 5.035464286804199, "learning_rate": 9.565217391304349e-06, "loss": 1.021, "step": 221 }, { "epoch": 0.2898172323759791, "grad_norm": 3.801708221435547, "learning_rate": 9.608695652173914e-06, "loss": 1.0053, "step": 222 }, { "epoch": 0.29112271540469975, "grad_norm": 4.671306133270264, "learning_rate": 9.652173913043478e-06, "loss": 1.1239, "step": 223 }, { "epoch": 0.2924281984334204, "grad_norm": 5.1868181228637695, "learning_rate": 9.695652173913043e-06, "loss": 1.1199, "step": 224 }, { "epoch": 0.293733681462141, "grad_norm": 4.014660835266113, "learning_rate": 9.73913043478261e-06, "loss": 1.0605, "step": 225 }, { "epoch": 0.2950391644908616, "grad_norm": 5.288611888885498, "learning_rate": 9.782608695652175e-06, "loss": 1.0914, "step": 226 }, { "epoch": 0.2963446475195822, "grad_norm": 5.0326972007751465, "learning_rate": 9.82608695652174e-06, "loss": 1.0554, "step": 227 }, { "epoch": 0.29765013054830286, "grad_norm": 4.603734970092773, "learning_rate": 9.869565217391304e-06, "loss": 1.0234, "step": 228 }, { "epoch": 0.2989556135770235, "grad_norm": 4.916118621826172, "learning_rate": 9.913043478260871e-06, "loss": 1.1493, "step": 229 }, { "epoch": 0.3002610966057441, "grad_norm": 3.913612127304077, "learning_rate": 9.956521739130436e-06, "loss": 1.1293, "step": 230 }, { "epoch": 0.30156657963446476, "grad_norm": 5.516775131225586, "learning_rate": 1e-05, "loss": 1.0106, "step": 231 }, { "epoch": 0.3028720626631854, "grad_norm": 5.419295787811279, "learning_rate": 9.99999955304673e-06, "loss": 1.1342, "step": 232 }, { "epoch": 0.304177545691906, "grad_norm": 4.708040714263916, "learning_rate": 9.999998212186997e-06, "loss": 1.1284, "step": 233 }, { "epoch": 0.30548302872062666, "grad_norm": 4.938497543334961, "learning_rate": 9.999995977421039e-06, "loss": 0.9507, "step": 234 }, { "epoch": 0.30678851174934724, "grad_norm": 4.604511737823486, "learning_rate": 9.999992848749262e-06, "loss": 1.129, "step": 235 }, { "epoch": 0.30809399477806787, "grad_norm": 4.270117282867432, "learning_rate": 9.99998882617222e-06, "loss": 1.0214, "step": 236 }, { "epoch": 0.3093994778067885, "grad_norm": 5.361505031585693, "learning_rate": 9.999983909690632e-06, "loss": 1.0938, "step": 237 }, { "epoch": 0.31070496083550914, "grad_norm": 4.273087501525879, "learning_rate": 9.99997809930538e-06, "loss": 1.1571, "step": 238 }, { "epoch": 0.31201044386422977, "grad_norm": 3.928175449371338, "learning_rate": 9.999971395017501e-06, "loss": 1.0477, "step": 239 }, { "epoch": 0.3133159268929504, "grad_norm": 4.953639507293701, "learning_rate": 9.999963796828191e-06, "loss": 1.0496, "step": 240 }, { "epoch": 0.31462140992167104, "grad_norm": 5.077392578125, "learning_rate": 9.999955304738815e-06, "loss": 1.0222, "step": 241 }, { "epoch": 0.31592689295039167, "grad_norm": 4.167980670928955, "learning_rate": 9.999945918750887e-06, "loss": 1.0795, "step": 242 }, { "epoch": 0.31723237597911225, "grad_norm": 4.604512691497803, "learning_rate": 9.999935638866084e-06, "loss": 1.1388, "step": 243 }, { "epoch": 0.3185378590078329, "grad_norm": 5.150665760040283, "learning_rate": 9.999924465086246e-06, "loss": 1.0034, "step": 244 }, { "epoch": 0.3198433420365535, "grad_norm": 4.584643363952637, "learning_rate": 9.999912397413371e-06, "loss": 1.0119, "step": 245 }, { "epoch": 0.32114882506527415, "grad_norm": 5.10865592956543, "learning_rate": 9.999899435849613e-06, "loss": 1.0599, "step": 246 }, { "epoch": 0.3224543080939948, "grad_norm": 4.574052810668945, "learning_rate": 9.999885580397295e-06, "loss": 1.067, "step": 247 }, { "epoch": 0.3237597911227154, "grad_norm": 3.7644200325012207, "learning_rate": 9.999870831058889e-06, "loss": 1.0567, "step": 248 }, { "epoch": 0.32506527415143605, "grad_norm": 4.030111312866211, "learning_rate": 9.999855187837036e-06, "loss": 0.9851, "step": 249 }, { "epoch": 0.3263707571801567, "grad_norm": 4.424080848693848, "learning_rate": 9.999838650734528e-06, "loss": 1.0151, "step": 250 }, { "epoch": 0.3276762402088773, "grad_norm": 5.072233200073242, "learning_rate": 9.999821219754324e-06, "loss": 1.1843, "step": 251 }, { "epoch": 0.3289817232375979, "grad_norm": 3.8587546348571777, "learning_rate": 9.999802894899542e-06, "loss": 1.0666, "step": 252 }, { "epoch": 0.3302872062663185, "grad_norm": 5.038973808288574, "learning_rate": 9.999783676173456e-06, "loss": 1.0766, "step": 253 }, { "epoch": 0.33159268929503916, "grad_norm": 4.486608982086182, "learning_rate": 9.999763563579501e-06, "loss": 1.1399, "step": 254 }, { "epoch": 0.3328981723237598, "grad_norm": 5.212406158447266, "learning_rate": 9.999742557121278e-06, "loss": 0.9974, "step": 255 }, { "epoch": 0.3342036553524804, "grad_norm": 3.3768393993377686, "learning_rate": 9.999720656802534e-06, "loss": 1.0553, "step": 256 }, { "epoch": 0.33550913838120106, "grad_norm": 4.206703186035156, "learning_rate": 9.999697862627192e-06, "loss": 0.9499, "step": 257 }, { "epoch": 0.3368146214099217, "grad_norm": 3.6921374797821045, "learning_rate": 9.999674174599324e-06, "loss": 1.1727, "step": 258 }, { "epoch": 0.3381201044386423, "grad_norm": 5.639096736907959, "learning_rate": 9.999649592723166e-06, "loss": 0.9945, "step": 259 }, { "epoch": 0.3394255874673629, "grad_norm": 4.328437328338623, "learning_rate": 9.999624117003111e-06, "loss": 1.0476, "step": 260 }, { "epoch": 0.34073107049608353, "grad_norm": 3.7602615356445312, "learning_rate": 9.999597747443714e-06, "loss": 0.9953, "step": 261 }, { "epoch": 0.34203655352480417, "grad_norm": 4.203237056732178, "learning_rate": 9.999570484049693e-06, "loss": 0.9903, "step": 262 }, { "epoch": 0.3433420365535248, "grad_norm": 4.573807716369629, "learning_rate": 9.999542326825916e-06, "loss": 0.9938, "step": 263 }, { "epoch": 0.34464751958224543, "grad_norm": 4.2973856925964355, "learning_rate": 9.999513275777421e-06, "loss": 1.1379, "step": 264 }, { "epoch": 0.34595300261096606, "grad_norm": 4.387681007385254, "learning_rate": 9.999483330909401e-06, "loss": 1.0351, "step": 265 }, { "epoch": 0.3472584856396867, "grad_norm": 4.294944763183594, "learning_rate": 9.99945249222721e-06, "loss": 0.9896, "step": 266 }, { "epoch": 0.34856396866840733, "grad_norm": 4.054105281829834, "learning_rate": 9.999420759736362e-06, "loss": 0.9732, "step": 267 }, { "epoch": 0.34986945169712796, "grad_norm": 4.987805366516113, "learning_rate": 9.999388133442528e-06, "loss": 1.1409, "step": 268 }, { "epoch": 0.35117493472584854, "grad_norm": 5.079680919647217, "learning_rate": 9.999354613351541e-06, "loss": 1.0208, "step": 269 }, { "epoch": 0.3524804177545692, "grad_norm": 4.577601432800293, "learning_rate": 9.999320199469397e-06, "loss": 1.0265, "step": 270 }, { "epoch": 0.3537859007832898, "grad_norm": 5.179898738861084, "learning_rate": 9.999284891802246e-06, "loss": 1.0038, "step": 271 }, { "epoch": 0.35509138381201044, "grad_norm": 4.7278971672058105, "learning_rate": 9.9992486903564e-06, "loss": 1.0279, "step": 272 }, { "epoch": 0.3563968668407311, "grad_norm": 3.5444467067718506, "learning_rate": 9.999211595138333e-06, "loss": 0.9626, "step": 273 }, { "epoch": 0.3577023498694517, "grad_norm": 4.063403129577637, "learning_rate": 9.999173606154675e-06, "loss": 1.0078, "step": 274 }, { "epoch": 0.35900783289817234, "grad_norm": 4.3961005210876465, "learning_rate": 9.999134723412219e-06, "loss": 1.0466, "step": 275 }, { "epoch": 0.360313315926893, "grad_norm": 5.2766618728637695, "learning_rate": 9.999094946917916e-06, "loss": 1.2242, "step": 276 }, { "epoch": 0.36161879895561355, "grad_norm": 5.00185489654541, "learning_rate": 9.999054276678879e-06, "loss": 0.9887, "step": 277 }, { "epoch": 0.3629242819843342, "grad_norm": 4.1430840492248535, "learning_rate": 9.999012712702373e-06, "loss": 0.9544, "step": 278 }, { "epoch": 0.3642297650130548, "grad_norm": 2.6793410778045654, "learning_rate": 9.998970254995838e-06, "loss": 1.1671, "step": 279 }, { "epoch": 0.36553524804177545, "grad_norm": 4.533927917480469, "learning_rate": 9.998926903566858e-06, "loss": 1.0945, "step": 280 }, { "epoch": 0.3668407310704961, "grad_norm": 4.640816688537598, "learning_rate": 9.998882658423185e-06, "loss": 1.0124, "step": 281 }, { "epoch": 0.3681462140992167, "grad_norm": 5.527533531188965, "learning_rate": 9.99883751957273e-06, "loss": 0.9477, "step": 282 }, { "epoch": 0.36945169712793735, "grad_norm": 4.468657970428467, "learning_rate": 9.998791487023564e-06, "loss": 0.9955, "step": 283 }, { "epoch": 0.370757180156658, "grad_norm": 4.9115309715271, "learning_rate": 9.998744560783912e-06, "loss": 0.9041, "step": 284 }, { "epoch": 0.3720626631853786, "grad_norm": 6.852635383605957, "learning_rate": 9.99869674086217e-06, "loss": 0.9615, "step": 285 }, { "epoch": 0.3733681462140992, "grad_norm": 3.5186474323272705, "learning_rate": 9.998648027266884e-06, "loss": 1.1328, "step": 286 }, { "epoch": 0.37467362924281983, "grad_norm": 4.597840785980225, "learning_rate": 9.998598420006763e-06, "loss": 0.9214, "step": 287 }, { "epoch": 0.37597911227154046, "grad_norm": 2.7896690368652344, "learning_rate": 9.998547919090675e-06, "loss": 1.0209, "step": 288 }, { "epoch": 0.3772845953002611, "grad_norm": 5.183618068695068, "learning_rate": 9.99849652452765e-06, "loss": 1.0206, "step": 289 }, { "epoch": 0.3785900783289817, "grad_norm": 4.944556713104248, "learning_rate": 9.998444236326876e-06, "loss": 1.0389, "step": 290 }, { "epoch": 0.37989556135770236, "grad_norm": 4.514628887176514, "learning_rate": 9.998391054497703e-06, "loss": 0.9837, "step": 291 }, { "epoch": 0.381201044386423, "grad_norm": 5.130414009094238, "learning_rate": 9.998336979049636e-06, "loss": 0.9744, "step": 292 }, { "epoch": 0.3825065274151436, "grad_norm": 4.680412292480469, "learning_rate": 9.998282009992344e-06, "loss": 1.0148, "step": 293 }, { "epoch": 0.3838120104438642, "grad_norm": 4.171786785125732, "learning_rate": 9.998226147335655e-06, "loss": 1.0054, "step": 294 }, { "epoch": 0.38511749347258484, "grad_norm": 4.593602180480957, "learning_rate": 9.998169391089554e-06, "loss": 0.9835, "step": 295 }, { "epoch": 0.38642297650130547, "grad_norm": 4.139461040496826, "learning_rate": 9.998111741264189e-06, "loss": 1.0334, "step": 296 }, { "epoch": 0.3877284595300261, "grad_norm": 5.740220069885254, "learning_rate": 9.99805319786987e-06, "loss": 0.9331, "step": 297 }, { "epoch": 0.38903394255874674, "grad_norm": 4.396590232849121, "learning_rate": 9.997993760917058e-06, "loss": 0.9256, "step": 298 }, { "epoch": 0.39033942558746737, "grad_norm": 3.1165435314178467, "learning_rate": 9.997933430416384e-06, "loss": 1.0585, "step": 299 }, { "epoch": 0.391644908616188, "grad_norm": 5.034556865692139, "learning_rate": 9.99787220637863e-06, "loss": 1.0563, "step": 300 }, { "epoch": 0.39295039164490864, "grad_norm": 5.612239837646484, "learning_rate": 9.997810088814745e-06, "loss": 0.7457, "step": 301 }, { "epoch": 0.39425587467362927, "grad_norm": 4.099153995513916, "learning_rate": 9.997747077735831e-06, "loss": 1.0015, "step": 302 }, { "epoch": 0.39556135770234985, "grad_norm": 4.536993980407715, "learning_rate": 9.997683173153158e-06, "loss": 1.0288, "step": 303 }, { "epoch": 0.3968668407310705, "grad_norm": 5.2137064933776855, "learning_rate": 9.997618375078146e-06, "loss": 1.0029, "step": 304 }, { "epoch": 0.3981723237597911, "grad_norm": 4.79557991027832, "learning_rate": 9.997552683522382e-06, "loss": 1.0322, "step": 305 }, { "epoch": 0.39947780678851175, "grad_norm": 4.451254367828369, "learning_rate": 9.997486098497611e-06, "loss": 0.8052, "step": 306 }, { "epoch": 0.4007832898172324, "grad_norm": 4.53281831741333, "learning_rate": 9.997418620015738e-06, "loss": 1.0408, "step": 307 }, { "epoch": 0.402088772845953, "grad_norm": 4.719001293182373, "learning_rate": 9.997350248088822e-06, "loss": 1.0031, "step": 308 }, { "epoch": 0.40339425587467365, "grad_norm": 4.456696510314941, "learning_rate": 9.997280982729091e-06, "loss": 1.0516, "step": 309 }, { "epoch": 0.4046997389033943, "grad_norm": 4.226430416107178, "learning_rate": 9.997210823948928e-06, "loss": 0.9917, "step": 310 }, { "epoch": 0.40600522193211486, "grad_norm": 4.320338726043701, "learning_rate": 9.997139771760876e-06, "loss": 1.0009, "step": 311 }, { "epoch": 0.4073107049608355, "grad_norm": 3.833554744720459, "learning_rate": 9.997067826177635e-06, "loss": 1.011, "step": 312 }, { "epoch": 0.4086161879895561, "grad_norm": 5.951189041137695, "learning_rate": 9.996994987212072e-06, "loss": 1.0463, "step": 313 }, { "epoch": 0.40992167101827676, "grad_norm": 4.604957103729248, "learning_rate": 9.996921254877206e-06, "loss": 0.9515, "step": 314 }, { "epoch": 0.4112271540469974, "grad_norm": 5.360300540924072, "learning_rate": 9.99684662918622e-06, "loss": 0.8158, "step": 315 }, { "epoch": 0.412532637075718, "grad_norm": 4.976587772369385, "learning_rate": 9.996771110152455e-06, "loss": 0.9627, "step": 316 }, { "epoch": 0.41383812010443866, "grad_norm": 4.313925743103027, "learning_rate": 9.996694697789414e-06, "loss": 0.9227, "step": 317 }, { "epoch": 0.4151436031331593, "grad_norm": 4.289072036743164, "learning_rate": 9.996617392110757e-06, "loss": 0.9657, "step": 318 }, { "epoch": 0.4164490861618799, "grad_norm": 5.662639617919922, "learning_rate": 9.996539193130303e-06, "loss": 0.8754, "step": 319 }, { "epoch": 0.4177545691906005, "grad_norm": 4.364449977874756, "learning_rate": 9.996460100862036e-06, "loss": 0.9871, "step": 320 }, { "epoch": 0.41906005221932113, "grad_norm": 4.841615676879883, "learning_rate": 9.996380115320095e-06, "loss": 0.9476, "step": 321 }, { "epoch": 0.42036553524804177, "grad_norm": 4.1591410636901855, "learning_rate": 9.996299236518779e-06, "loss": 0.9389, "step": 322 }, { "epoch": 0.4216710182767624, "grad_norm": 3.929408550262451, "learning_rate": 9.996217464472549e-06, "loss": 0.9441, "step": 323 }, { "epoch": 0.42297650130548303, "grad_norm": 5.341044902801514, "learning_rate": 9.996134799196022e-06, "loss": 1.0827, "step": 324 }, { "epoch": 0.42428198433420367, "grad_norm": 5.232245922088623, "learning_rate": 9.996051240703978e-06, "loss": 0.9793, "step": 325 }, { "epoch": 0.4255874673629243, "grad_norm": 4.7988200187683105, "learning_rate": 9.995966789011357e-06, "loss": 1.0737, "step": 326 }, { "epoch": 0.42689295039164493, "grad_norm": 3.562227249145508, "learning_rate": 9.995881444133257e-06, "loss": 1.0871, "step": 327 }, { "epoch": 0.4281984334203655, "grad_norm": 5.280902862548828, "learning_rate": 9.995795206084937e-06, "loss": 0.9425, "step": 328 }, { "epoch": 0.42950391644908614, "grad_norm": 4.245062351226807, "learning_rate": 9.99570807488181e-06, "loss": 0.9406, "step": 329 }, { "epoch": 0.4308093994778068, "grad_norm": 3.380915880203247, "learning_rate": 9.99562005053946e-06, "loss": 0.9543, "step": 330 }, { "epoch": 0.4321148825065274, "grad_norm": 5.290715217590332, "learning_rate": 9.995531133073619e-06, "loss": 0.9834, "step": 331 }, { "epoch": 0.43342036553524804, "grad_norm": 5.875310897827148, "learning_rate": 9.995441322500186e-06, "loss": 0.9656, "step": 332 }, { "epoch": 0.4347258485639687, "grad_norm": 5.510551929473877, "learning_rate": 9.995350618835218e-06, "loss": 0.9603, "step": 333 }, { "epoch": 0.4360313315926893, "grad_norm": 3.080106258392334, "learning_rate": 9.995259022094929e-06, "loss": 1.0476, "step": 334 }, { "epoch": 0.43733681462140994, "grad_norm": 4.890894889831543, "learning_rate": 9.995166532295696e-06, "loss": 0.9496, "step": 335 }, { "epoch": 0.4386422976501306, "grad_norm": 4.105751991271973, "learning_rate": 9.995073149454057e-06, "loss": 0.9533, "step": 336 }, { "epoch": 0.43994778067885115, "grad_norm": 4.756442070007324, "learning_rate": 9.994978873586703e-06, "loss": 0.9618, "step": 337 }, { "epoch": 0.4412532637075718, "grad_norm": 6.793393611907959, "learning_rate": 9.99488370471049e-06, "loss": 0.9023, "step": 338 }, { "epoch": 0.4425587467362924, "grad_norm": 2.742494583129883, "learning_rate": 9.994787642842432e-06, "loss": 0.9645, "step": 339 }, { "epoch": 0.44386422976501305, "grad_norm": 4.378128528594971, "learning_rate": 9.994690687999705e-06, "loss": 0.9292, "step": 340 }, { "epoch": 0.4451697127937337, "grad_norm": 4.567923069000244, "learning_rate": 9.994592840199642e-06, "loss": 0.8705, "step": 341 }, { "epoch": 0.4464751958224543, "grad_norm": 5.349486351013184, "learning_rate": 9.994494099459734e-06, "loss": 1.0149, "step": 342 }, { "epoch": 0.44778067885117495, "grad_norm": 5.894613742828369, "learning_rate": 9.994394465797638e-06, "loss": 0.957, "step": 343 }, { "epoch": 0.4490861618798956, "grad_norm": 3.755742073059082, "learning_rate": 9.994293939231163e-06, "loss": 0.967, "step": 344 }, { "epoch": 0.45039164490861616, "grad_norm": 4.413910388946533, "learning_rate": 9.994192519778285e-06, "loss": 0.9415, "step": 345 }, { "epoch": 0.4516971279373368, "grad_norm": 4.756941318511963, "learning_rate": 9.994090207457133e-06, "loss": 0.9394, "step": 346 }, { "epoch": 0.45300261096605743, "grad_norm": 5.714898586273193, "learning_rate": 9.993987002285997e-06, "loss": 0.9778, "step": 347 }, { "epoch": 0.45430809399477806, "grad_norm": 3.8421223163604736, "learning_rate": 9.993882904283333e-06, "loss": 1.0345, "step": 348 }, { "epoch": 0.4556135770234987, "grad_norm": 4.590973854064941, "learning_rate": 9.993777913467748e-06, "loss": 0.943, "step": 349 }, { "epoch": 0.45691906005221933, "grad_norm": 4.788003444671631, "learning_rate": 9.993672029858015e-06, "loss": 0.8779, "step": 350 }, { "epoch": 0.45822454308093996, "grad_norm": 5.632896423339844, "learning_rate": 9.99356525347306e-06, "loss": 1.0183, "step": 351 }, { "epoch": 0.4595300261096606, "grad_norm": 5.488896369934082, "learning_rate": 9.993457584331976e-06, "loss": 0.9345, "step": 352 }, { "epoch": 0.4608355091383812, "grad_norm": 4.878500461578369, "learning_rate": 9.993349022454014e-06, "loss": 0.995, "step": 353 }, { "epoch": 0.4621409921671018, "grad_norm": 4.892000198364258, "learning_rate": 9.993239567858578e-06, "loss": 1.0423, "step": 354 }, { "epoch": 0.46344647519582244, "grad_norm": 4.354773044586182, "learning_rate": 9.993129220565238e-06, "loss": 0.7715, "step": 355 }, { "epoch": 0.46475195822454307, "grad_norm": 4.678215980529785, "learning_rate": 9.993017980593725e-06, "loss": 0.8841, "step": 356 }, { "epoch": 0.4660574412532637, "grad_norm": 4.596961498260498, "learning_rate": 9.992905847963924e-06, "loss": 0.9159, "step": 357 }, { "epoch": 0.46736292428198434, "grad_norm": 4.867253303527832, "learning_rate": 9.992792822695881e-06, "loss": 0.8791, "step": 358 }, { "epoch": 0.46866840731070497, "grad_norm": 5.305583477020264, "learning_rate": 9.992678904809805e-06, "loss": 0.9534, "step": 359 }, { "epoch": 0.4699738903394256, "grad_norm": 4.570066928863525, "learning_rate": 9.992564094326063e-06, "loss": 1.0048, "step": 360 }, { "epoch": 0.47127937336814624, "grad_norm": 4.911692142486572, "learning_rate": 9.99244839126518e-06, "loss": 1.0121, "step": 361 }, { "epoch": 0.4725848563968668, "grad_norm": 3.4557299613952637, "learning_rate": 9.992331795647838e-06, "loss": 1.0568, "step": 362 }, { "epoch": 0.47389033942558745, "grad_norm": 5.287811756134033, "learning_rate": 9.992214307494888e-06, "loss": 0.8997, "step": 363 }, { "epoch": 0.4751958224543081, "grad_norm": 4.57080078125, "learning_rate": 9.992095926827332e-06, "loss": 0.9725, "step": 364 }, { "epoch": 0.4765013054830287, "grad_norm": 5.861119270324707, "learning_rate": 9.991976653666335e-06, "loss": 0.8803, "step": 365 }, { "epoch": 0.47780678851174935, "grad_norm": 5.165677070617676, "learning_rate": 9.991856488033218e-06, "loss": 0.9755, "step": 366 }, { "epoch": 0.47911227154047, "grad_norm": 4.672550678253174, "learning_rate": 9.991735429949469e-06, "loss": 1.0361, "step": 367 }, { "epoch": 0.4804177545691906, "grad_norm": 5.230209827423096, "learning_rate": 9.991613479436726e-06, "loss": 0.9656, "step": 368 }, { "epoch": 0.48172323759791125, "grad_norm": 5.215057373046875, "learning_rate": 9.991490636516796e-06, "loss": 0.9644, "step": 369 }, { "epoch": 0.4830287206266319, "grad_norm": 5.273270606994629, "learning_rate": 9.991366901211639e-06, "loss": 0.9247, "step": 370 }, { "epoch": 0.48433420365535246, "grad_norm": 4.94844388961792, "learning_rate": 9.991242273543376e-06, "loss": 0.9666, "step": 371 }, { "epoch": 0.4856396866840731, "grad_norm": 4.145085334777832, "learning_rate": 9.991116753534288e-06, "loss": 0.9798, "step": 372 }, { "epoch": 0.4869451697127937, "grad_norm": 4.516151428222656, "learning_rate": 9.990990341206817e-06, "loss": 0.8322, "step": 373 }, { "epoch": 0.48825065274151436, "grad_norm": 4.7209930419921875, "learning_rate": 9.990863036583561e-06, "loss": 0.9147, "step": 374 }, { "epoch": 0.489556135770235, "grad_norm": 4.612196445465088, "learning_rate": 9.990734839687283e-06, "loss": 0.952, "step": 375 }, { "epoch": 0.4908616187989556, "grad_norm": 5.50564432144165, "learning_rate": 9.990605750540898e-06, "loss": 0.9028, "step": 376 }, { "epoch": 0.49216710182767626, "grad_norm": 4.446320056915283, "learning_rate": 9.990475769167489e-06, "loss": 0.9539, "step": 377 }, { "epoch": 0.4934725848563969, "grad_norm": 4.967776775360107, "learning_rate": 9.990344895590292e-06, "loss": 0.821, "step": 378 }, { "epoch": 0.49477806788511747, "grad_norm": 4.9003167152404785, "learning_rate": 9.990213129832704e-06, "loss": 0.9007, "step": 379 }, { "epoch": 0.4960835509138381, "grad_norm": 4.007465839385986, "learning_rate": 9.990080471918284e-06, "loss": 1.0887, "step": 380 }, { "epoch": 0.49738903394255873, "grad_norm": 4.532343864440918, "learning_rate": 9.989946921870749e-06, "loss": 0.9612, "step": 381 }, { "epoch": 0.49869451697127937, "grad_norm": 5.862714767456055, "learning_rate": 9.989812479713974e-06, "loss": 0.9728, "step": 382 }, { "epoch": 0.5, "grad_norm": 4.00089168548584, "learning_rate": 9.989677145471994e-06, "loss": 0.9533, "step": 383 }, { "epoch": 0.5013054830287206, "grad_norm": 4.348776340484619, "learning_rate": 9.989540919169005e-06, "loss": 0.9485, "step": 384 }, { "epoch": 0.5026109660574413, "grad_norm": 5.006931304931641, "learning_rate": 9.989403800829363e-06, "loss": 0.9621, "step": 385 }, { "epoch": 0.5039164490861618, "grad_norm": 3.7387444972991943, "learning_rate": 9.989265790477581e-06, "loss": 0.9074, "step": 386 }, { "epoch": 0.5052219321148825, "grad_norm": 2.950324296951294, "learning_rate": 9.989126888138331e-06, "loss": 0.9253, "step": 387 }, { "epoch": 0.5065274151436031, "grad_norm": 3.8281092643737793, "learning_rate": 9.988987093836451e-06, "loss": 0.8782, "step": 388 }, { "epoch": 0.5078328981723238, "grad_norm": 5.167135715484619, "learning_rate": 9.988846407596929e-06, "loss": 0.9167, "step": 389 }, { "epoch": 0.5091383812010444, "grad_norm": 4.9226765632629395, "learning_rate": 9.98870482944492e-06, "loss": 0.9342, "step": 390 }, { "epoch": 0.5104438642297651, "grad_norm": 4.172088623046875, "learning_rate": 9.988562359405733e-06, "loss": 0.93, "step": 391 }, { "epoch": 0.5117493472584856, "grad_norm": 4.942967891693115, "learning_rate": 9.98841899750484e-06, "loss": 1.0177, "step": 392 }, { "epoch": 0.5130548302872062, "grad_norm": 4.532896995544434, "learning_rate": 9.988274743767872e-06, "loss": 0.8956, "step": 393 }, { "epoch": 0.5143603133159269, "grad_norm": 3.684152841567993, "learning_rate": 9.988129598220617e-06, "loss": 0.9665, "step": 394 }, { "epoch": 0.5156657963446475, "grad_norm": 4.732635498046875, "learning_rate": 9.987983560889028e-06, "loss": 0.8872, "step": 395 }, { "epoch": 0.5169712793733682, "grad_norm": 3.0347297191619873, "learning_rate": 9.98783663179921e-06, "loss": 0.9428, "step": 396 }, { "epoch": 0.5182767624020888, "grad_norm": 4.534252166748047, "learning_rate": 9.987688810977434e-06, "loss": 0.9611, "step": 397 }, { "epoch": 0.5195822454308094, "grad_norm": 3.280902147293091, "learning_rate": 9.987540098450126e-06, "loss": 1.0129, "step": 398 }, { "epoch": 0.52088772845953, "grad_norm": 3.574319362640381, "learning_rate": 9.987390494243873e-06, "loss": 0.8527, "step": 399 }, { "epoch": 0.5221932114882507, "grad_norm": 4.57360315322876, "learning_rate": 9.987239998385422e-06, "loss": 0.9663, "step": 400 }, { "epoch": 0.5234986945169713, "grad_norm": 3.9249699115753174, "learning_rate": 9.987088610901678e-06, "loss": 1.0865, "step": 401 }, { "epoch": 0.5248041775456919, "grad_norm": 4.501255989074707, "learning_rate": 9.986936331819708e-06, "loss": 0.873, "step": 402 }, { "epoch": 0.5261096605744126, "grad_norm": 4.1838202476501465, "learning_rate": 9.986783161166733e-06, "loss": 0.8634, "step": 403 }, { "epoch": 0.5274151436031331, "grad_norm": 4.4898295402526855, "learning_rate": 9.986629098970141e-06, "loss": 0.8747, "step": 404 }, { "epoch": 0.5287206266318538, "grad_norm": 4.543522834777832, "learning_rate": 9.986474145257476e-06, "loss": 0.837, "step": 405 }, { "epoch": 0.5300261096605744, "grad_norm": 5.075877666473389, "learning_rate": 9.986318300056436e-06, "loss": 0.928, "step": 406 }, { "epoch": 0.5313315926892951, "grad_norm": 4.821189880371094, "learning_rate": 9.986161563394887e-06, "loss": 0.8881, "step": 407 }, { "epoch": 0.5326370757180157, "grad_norm": 4.528730392456055, "learning_rate": 9.98600393530085e-06, "loss": 0.8719, "step": 408 }, { "epoch": 0.5339425587467362, "grad_norm": 4.315455436706543, "learning_rate": 9.985845415802507e-06, "loss": 0.8702, "step": 409 }, { "epoch": 0.5352480417754569, "grad_norm": 3.3726658821105957, "learning_rate": 9.985686004928195e-06, "loss": 0.9231, "step": 410 }, { "epoch": 0.5365535248041775, "grad_norm": 5.4692063331604, "learning_rate": 9.985525702706415e-06, "loss": 0.8986, "step": 411 }, { "epoch": 0.5378590078328982, "grad_norm": 4.321691989898682, "learning_rate": 9.985364509165829e-06, "loss": 0.806, "step": 412 }, { "epoch": 0.5391644908616188, "grad_norm": 3.1946957111358643, "learning_rate": 9.985202424335249e-06, "loss": 0.8749, "step": 413 }, { "epoch": 0.5404699738903395, "grad_norm": 3.6660807132720947, "learning_rate": 9.985039448243661e-06, "loss": 0.9091, "step": 414 }, { "epoch": 0.54177545691906, "grad_norm": 5.0404052734375, "learning_rate": 9.984875580920196e-06, "loss": 0.9146, "step": 415 }, { "epoch": 0.5430809399477807, "grad_norm": 3.3806302547454834, "learning_rate": 9.984710822394153e-06, "loss": 0.8595, "step": 416 }, { "epoch": 0.5443864229765013, "grad_norm": 4.497293472290039, "learning_rate": 9.984545172694985e-06, "loss": 0.9709, "step": 417 }, { "epoch": 0.5456919060052219, "grad_norm": 4.66550350189209, "learning_rate": 9.984378631852311e-06, "loss": 0.8741, "step": 418 }, { "epoch": 0.5469973890339426, "grad_norm": 6.776769638061523, "learning_rate": 9.984211199895903e-06, "loss": 0.8423, "step": 419 }, { "epoch": 0.5483028720626631, "grad_norm": 4.76498556137085, "learning_rate": 9.984042876855694e-06, "loss": 0.9891, "step": 420 }, { "epoch": 0.5496083550913838, "grad_norm": 5.721564292907715, "learning_rate": 9.98387366276178e-06, "loss": 0.9406, "step": 421 }, { "epoch": 0.5509138381201044, "grad_norm": 3.7162911891937256, "learning_rate": 9.98370355764441e-06, "loss": 0.8156, "step": 422 }, { "epoch": 0.5522193211488251, "grad_norm": 3.821258306503296, "learning_rate": 9.983532561534e-06, "loss": 0.8882, "step": 423 }, { "epoch": 0.5535248041775457, "grad_norm": 5.639698028564453, "learning_rate": 9.983360674461115e-06, "loss": 1.0168, "step": 424 }, { "epoch": 0.5548302872062664, "grad_norm": 3.6814827919006348, "learning_rate": 9.983187896456489e-06, "loss": 0.8499, "step": 425 }, { "epoch": 0.556135770234987, "grad_norm": 15.611799240112305, "learning_rate": 9.983014227551011e-06, "loss": 0.8509, "step": 426 }, { "epoch": 0.5574412532637075, "grad_norm": 4.1461710929870605, "learning_rate": 9.98283966777573e-06, "loss": 0.8681, "step": 427 }, { "epoch": 0.5587467362924282, "grad_norm": 4.177596092224121, "learning_rate": 9.982664217161851e-06, "loss": 0.9751, "step": 428 }, { "epoch": 0.5600522193211488, "grad_norm": 5.541667938232422, "learning_rate": 9.982487875740746e-06, "loss": 0.8738, "step": 429 }, { "epoch": 0.5613577023498695, "grad_norm": 2.9947948455810547, "learning_rate": 9.982310643543939e-06, "loss": 1.0376, "step": 430 }, { "epoch": 0.5626631853785901, "grad_norm": 4.42106819152832, "learning_rate": 9.982132520603115e-06, "loss": 0.8982, "step": 431 }, { "epoch": 0.5639686684073107, "grad_norm": 5.248688697814941, "learning_rate": 9.981953506950122e-06, "loss": 0.8743, "step": 432 }, { "epoch": 0.5652741514360313, "grad_norm": 4.24997615814209, "learning_rate": 9.981773602616961e-06, "loss": 0.8486, "step": 433 }, { "epoch": 0.566579634464752, "grad_norm": 4.115297794342041, "learning_rate": 9.981592807635797e-06, "loss": 0.8685, "step": 434 }, { "epoch": 0.5678851174934726, "grad_norm": 4.6851043701171875, "learning_rate": 9.981411122038954e-06, "loss": 0.9137, "step": 435 }, { "epoch": 0.5691906005221932, "grad_norm": 5.422494411468506, "learning_rate": 9.981228545858913e-06, "loss": 0.8165, "step": 436 }, { "epoch": 0.5704960835509139, "grad_norm": 4.618397235870361, "learning_rate": 9.981045079128313e-06, "loss": 0.9062, "step": 437 }, { "epoch": 0.5718015665796344, "grad_norm": 4.353048801422119, "learning_rate": 9.980860721879959e-06, "loss": 0.8093, "step": 438 }, { "epoch": 0.5731070496083551, "grad_norm": 4.91734504699707, "learning_rate": 9.980675474146807e-06, "loss": 0.9059, "step": 439 }, { "epoch": 0.5744125326370757, "grad_norm": 5.124383926391602, "learning_rate": 9.980489335961973e-06, "loss": 0.9619, "step": 440 }, { "epoch": 0.5757180156657964, "grad_norm": 4.5638556480407715, "learning_rate": 9.980302307358743e-06, "loss": 0.8242, "step": 441 }, { "epoch": 0.577023498694517, "grad_norm": 3.567636251449585, "learning_rate": 9.98011438837055e-06, "loss": 0.9005, "step": 442 }, { "epoch": 0.5783289817232375, "grad_norm": 5.0671539306640625, "learning_rate": 9.979925579030988e-06, "loss": 0.8377, "step": 443 }, { "epoch": 0.5796344647519582, "grad_norm": 4.934540271759033, "learning_rate": 9.979735879373818e-06, "loss": 0.7801, "step": 444 }, { "epoch": 0.5809399477806788, "grad_norm": 4.897219181060791, "learning_rate": 9.97954528943295e-06, "loss": 0.7932, "step": 445 }, { "epoch": 0.5822454308093995, "grad_norm": 4.641899585723877, "learning_rate": 9.97935380924246e-06, "loss": 0.7827, "step": 446 }, { "epoch": 0.5835509138381201, "grad_norm": 5.29866886138916, "learning_rate": 9.979161438836582e-06, "loss": 0.8685, "step": 447 }, { "epoch": 0.5848563968668408, "grad_norm": 4.180384159088135, "learning_rate": 9.978968178249706e-06, "loss": 0.9935, "step": 448 }, { "epoch": 0.5861618798955613, "grad_norm": 5.03877067565918, "learning_rate": 9.978774027516384e-06, "loss": 0.9407, "step": 449 }, { "epoch": 0.587467362924282, "grad_norm": 3.9605135917663574, "learning_rate": 9.978578986671327e-06, "loss": 0.8068, "step": 450 }, { "epoch": 0.5887728459530026, "grad_norm": 5.755600929260254, "learning_rate": 9.978383055749407e-06, "loss": 0.7392, "step": 451 }, { "epoch": 0.5900783289817232, "grad_norm": 4.349735260009766, "learning_rate": 9.978186234785648e-06, "loss": 0.8856, "step": 452 }, { "epoch": 0.5913838120104439, "grad_norm": 3.8118412494659424, "learning_rate": 9.977988523815241e-06, "loss": 0.8995, "step": 453 }, { "epoch": 0.5926892950391645, "grad_norm": 3.4277503490448, "learning_rate": 9.977789922873532e-06, "loss": 0.8932, "step": 454 }, { "epoch": 0.5939947780678851, "grad_norm": 4.093258857727051, "learning_rate": 9.977590431996027e-06, "loss": 0.8452, "step": 455 }, { "epoch": 0.5953002610966057, "grad_norm": 3.7631845474243164, "learning_rate": 9.977390051218393e-06, "loss": 0.7851, "step": 456 }, { "epoch": 0.5966057441253264, "grad_norm": 5.62789249420166, "learning_rate": 9.977188780576452e-06, "loss": 1.0058, "step": 457 }, { "epoch": 0.597911227154047, "grad_norm": 3.8881006240844727, "learning_rate": 9.97698662010619e-06, "loss": 0.7777, "step": 458 }, { "epoch": 0.5992167101827677, "grad_norm": 4.584966659545898, "learning_rate": 9.976783569843747e-06, "loss": 0.7274, "step": 459 }, { "epoch": 0.6005221932114883, "grad_norm": 8.070284843444824, "learning_rate": 9.976579629825425e-06, "loss": 0.8924, "step": 460 }, { "epoch": 0.6018276762402088, "grad_norm": 4.454860687255859, "learning_rate": 9.976374800087687e-06, "loss": 0.8958, "step": 461 }, { "epoch": 0.6031331592689295, "grad_norm": 4.778026580810547, "learning_rate": 9.976169080667149e-06, "loss": 0.8882, "step": 462 }, { "epoch": 0.6044386422976501, "grad_norm": 4.317173004150391, "learning_rate": 9.975962471600593e-06, "loss": 0.8775, "step": 463 }, { "epoch": 0.6057441253263708, "grad_norm": 4.69331169128418, "learning_rate": 9.975754972924953e-06, "loss": 0.8436, "step": 464 }, { "epoch": 0.6070496083550914, "grad_norm": 4.98107385635376, "learning_rate": 9.975546584677332e-06, "loss": 0.9735, "step": 465 }, { "epoch": 0.608355091383812, "grad_norm": 5.0528059005737305, "learning_rate": 9.97533730689498e-06, "loss": 0.8978, "step": 466 }, { "epoch": 0.6096605744125326, "grad_norm": 4.3778557777404785, "learning_rate": 9.975127139615317e-06, "loss": 0.8876, "step": 467 }, { "epoch": 0.6109660574412533, "grad_norm": 4.9897003173828125, "learning_rate": 9.974916082875911e-06, "loss": 0.7858, "step": 468 }, { "epoch": 0.6122715404699739, "grad_norm": 4.299252986907959, "learning_rate": 9.974704136714501e-06, "loss": 0.8254, "step": 469 }, { "epoch": 0.6135770234986945, "grad_norm": 4.887605667114258, "learning_rate": 9.974491301168976e-06, "loss": 0.9087, "step": 470 }, { "epoch": 0.6148825065274152, "grad_norm": 4.3851637840271, "learning_rate": 9.974277576277386e-06, "loss": 0.9424, "step": 471 }, { "epoch": 0.6161879895561357, "grad_norm": 4.069988250732422, "learning_rate": 9.974062962077944e-06, "loss": 0.9063, "step": 472 }, { "epoch": 0.6174934725848564, "grad_norm": 4.141973972320557, "learning_rate": 9.973847458609016e-06, "loss": 0.8593, "step": 473 }, { "epoch": 0.618798955613577, "grad_norm": 4.8528032302856445, "learning_rate": 9.973631065909132e-06, "loss": 0.9804, "step": 474 }, { "epoch": 0.6201044386422977, "grad_norm": 4.980158805847168, "learning_rate": 9.973413784016978e-06, "loss": 0.904, "step": 475 }, { "epoch": 0.6214099216710183, "grad_norm": 5.102324962615967, "learning_rate": 9.9731956129714e-06, "loss": 0.8961, "step": 476 }, { "epoch": 0.6227154046997389, "grad_norm": 4.180404186248779, "learning_rate": 9.972976552811405e-06, "loss": 0.7011, "step": 477 }, { "epoch": 0.6240208877284595, "grad_norm": 4.874111175537109, "learning_rate": 9.972756603576153e-06, "loss": 0.791, "step": 478 }, { "epoch": 0.6253263707571801, "grad_norm": 4.386068344116211, "learning_rate": 9.97253576530497e-06, "loss": 0.8548, "step": 479 }, { "epoch": 0.6266318537859008, "grad_norm": 4.804571628570557, "learning_rate": 9.972314038037337e-06, "loss": 0.8186, "step": 480 }, { "epoch": 0.6279373368146214, "grad_norm": 5.074531078338623, "learning_rate": 9.972091421812893e-06, "loss": 0.8346, "step": 481 }, { "epoch": 0.6292428198433421, "grad_norm": 5.962896347045898, "learning_rate": 9.97186791667144e-06, "loss": 0.9861, "step": 482 }, { "epoch": 0.6305483028720626, "grad_norm": 4.353085041046143, "learning_rate": 9.971643522652936e-06, "loss": 0.8452, "step": 483 }, { "epoch": 0.6318537859007833, "grad_norm": 4.09246826171875, "learning_rate": 9.971418239797497e-06, "loss": 0.707, "step": 484 }, { "epoch": 0.6331592689295039, "grad_norm": 4.096375465393066, "learning_rate": 9.9711920681454e-06, "loss": 0.7713, "step": 485 }, { "epoch": 0.6344647519582245, "grad_norm": 5.573315143585205, "learning_rate": 9.970965007737081e-06, "loss": 0.7645, "step": 486 }, { "epoch": 0.6357702349869452, "grad_norm": 5.026029586791992, "learning_rate": 9.970737058613134e-06, "loss": 0.7885, "step": 487 }, { "epoch": 0.6370757180156658, "grad_norm": 4.524381160736084, "learning_rate": 9.97050822081431e-06, "loss": 0.8752, "step": 488 }, { "epoch": 0.6383812010443864, "grad_norm": 3.946291923522949, "learning_rate": 9.970278494381525e-06, "loss": 0.8366, "step": 489 }, { "epoch": 0.639686684073107, "grad_norm": 4.236980438232422, "learning_rate": 9.970047879355846e-06, "loss": 0.7762, "step": 490 }, { "epoch": 0.6409921671018277, "grad_norm": 5.575711250305176, "learning_rate": 9.969816375778504e-06, "loss": 0.9172, "step": 491 }, { "epoch": 0.6422976501305483, "grad_norm": 4.820695400238037, "learning_rate": 9.96958398369089e-06, "loss": 0.782, "step": 492 }, { "epoch": 0.643603133159269, "grad_norm": 3.5960237979888916, "learning_rate": 9.969350703134548e-06, "loss": 0.8487, "step": 493 }, { "epoch": 0.6449086161879896, "grad_norm": 4.111789703369141, "learning_rate": 9.969116534151184e-06, "loss": 0.9205, "step": 494 }, { "epoch": 0.6462140992167101, "grad_norm": 3.667329788208008, "learning_rate": 9.968881476782666e-06, "loss": 0.7803, "step": 495 }, { "epoch": 0.6475195822454308, "grad_norm": 4.51666784286499, "learning_rate": 9.968645531071014e-06, "loss": 0.8843, "step": 496 }, { "epoch": 0.6488250652741514, "grad_norm": 4.298446178436279, "learning_rate": 9.968408697058413e-06, "loss": 0.8097, "step": 497 }, { "epoch": 0.6501305483028721, "grad_norm": 5.380995273590088, "learning_rate": 9.968170974787207e-06, "loss": 0.8736, "step": 498 }, { "epoch": 0.6514360313315927, "grad_norm": 6.139780521392822, "learning_rate": 9.967932364299891e-06, "loss": 0.8213, "step": 499 }, { "epoch": 0.6527415143603134, "grad_norm": 4.36410665512085, "learning_rate": 9.967692865639128e-06, "loss": 0.8564, "step": 500 }, { "epoch": 0.6540469973890339, "grad_norm": 3.911728620529175, "learning_rate": 9.967452478847735e-06, "loss": 0.9016, "step": 501 }, { "epoch": 0.6553524804177546, "grad_norm": 5.431145191192627, "learning_rate": 9.967211203968686e-06, "loss": 0.8504, "step": 502 }, { "epoch": 0.6566579634464752, "grad_norm": 4.462330341339111, "learning_rate": 9.966969041045121e-06, "loss": 0.8266, "step": 503 }, { "epoch": 0.6579634464751958, "grad_norm": 4.027176380157471, "learning_rate": 9.96672599012033e-06, "loss": 0.8258, "step": 504 }, { "epoch": 0.6592689295039165, "grad_norm": 4.721169471740723, "learning_rate": 9.96648205123777e-06, "loss": 0.8127, "step": 505 }, { "epoch": 0.660574412532637, "grad_norm": 3.9733047485351562, "learning_rate": 9.966237224441048e-06, "loss": 0.8712, "step": 506 }, { "epoch": 0.6618798955613577, "grad_norm": 5.255932331085205, "learning_rate": 9.965991509773938e-06, "loss": 0.8238, "step": 507 }, { "epoch": 0.6631853785900783, "grad_norm": 4.696491718292236, "learning_rate": 9.965744907280368e-06, "loss": 0.6762, "step": 508 }, { "epoch": 0.664490861618799, "grad_norm": 4.199885845184326, "learning_rate": 9.965497417004428e-06, "loss": 0.7417, "step": 509 }, { "epoch": 0.6657963446475196, "grad_norm": 4.241267681121826, "learning_rate": 9.965249038990361e-06, "loss": 0.9501, "step": 510 }, { "epoch": 0.6671018276762402, "grad_norm": 4.862495422363281, "learning_rate": 9.964999773282574e-06, "loss": 0.7853, "step": 511 }, { "epoch": 0.6684073107049608, "grad_norm": 6.180248737335205, "learning_rate": 9.964749619925632e-06, "loss": 1.1, "step": 512 }, { "epoch": 0.6697127937336814, "grad_norm": 4.770278453826904, "learning_rate": 9.964498578964257e-06, "loss": 0.8854, "step": 513 }, { "epoch": 0.6710182767624021, "grad_norm": 5.907410621643066, "learning_rate": 9.96424665044333e-06, "loss": 0.7461, "step": 514 }, { "epoch": 0.6723237597911227, "grad_norm": 4.1516804695129395, "learning_rate": 9.96399383440789e-06, "loss": 0.7384, "step": 515 }, { "epoch": 0.6736292428198434, "grad_norm": 4.970891952514648, "learning_rate": 9.963740130903138e-06, "loss": 0.741, "step": 516 }, { "epoch": 0.674934725848564, "grad_norm": 7.407042503356934, "learning_rate": 9.963485539974432e-06, "loss": 0.917, "step": 517 }, { "epoch": 0.6762402088772846, "grad_norm": 4.518277645111084, "learning_rate": 9.963230061667287e-06, "loss": 0.9001, "step": 518 }, { "epoch": 0.6775456919060052, "grad_norm": 5.061564922332764, "learning_rate": 9.962973696027377e-06, "loss": 0.8005, "step": 519 }, { "epoch": 0.6788511749347258, "grad_norm": 4.848884105682373, "learning_rate": 9.962716443100533e-06, "loss": 0.8589, "step": 520 }, { "epoch": 0.6801566579634465, "grad_norm": 4.297856330871582, "learning_rate": 9.962458302932753e-06, "loss": 0.768, "step": 521 }, { "epoch": 0.6814621409921671, "grad_norm": 4.250111103057861, "learning_rate": 9.962199275570184e-06, "loss": 0.6417, "step": 522 }, { "epoch": 0.6827676240208878, "grad_norm": 4.367997169494629, "learning_rate": 9.961939361059136e-06, "loss": 0.8365, "step": 523 }, { "epoch": 0.6840731070496083, "grad_norm": 4.982820510864258, "learning_rate": 9.961678559446076e-06, "loss": 0.833, "step": 524 }, { "epoch": 0.685378590078329, "grad_norm": 3.7981181144714355, "learning_rate": 9.961416870777632e-06, "loss": 0.8039, "step": 525 }, { "epoch": 0.6866840731070496, "grad_norm": 4.2907562255859375, "learning_rate": 9.961154295100586e-06, "loss": 0.8033, "step": 526 }, { "epoch": 0.6879895561357703, "grad_norm": 4.903866291046143, "learning_rate": 9.960890832461886e-06, "loss": 0.7755, "step": 527 }, { "epoch": 0.6892950391644909, "grad_norm": 7.7080464363098145, "learning_rate": 9.96062648290863e-06, "loss": 0.8294, "step": 528 }, { "epoch": 0.6906005221932114, "grad_norm": 4.788351535797119, "learning_rate": 9.960361246488084e-06, "loss": 0.8044, "step": 529 }, { "epoch": 0.6919060052219321, "grad_norm": 4.185435771942139, "learning_rate": 9.960095123247662e-06, "loss": 0.7702, "step": 530 }, { "epoch": 0.6932114882506527, "grad_norm": 3.919527053833008, "learning_rate": 9.959828113234945e-06, "loss": 0.8074, "step": 531 }, { "epoch": 0.6945169712793734, "grad_norm": 4.983836650848389, "learning_rate": 9.959560216497665e-06, "loss": 0.7025, "step": 532 }, { "epoch": 0.695822454308094, "grad_norm": 5.186157703399658, "learning_rate": 9.959291433083723e-06, "loss": 0.7558, "step": 533 }, { "epoch": 0.6971279373368147, "grad_norm": 4.656023025512695, "learning_rate": 9.959021763041168e-06, "loss": 0.7227, "step": 534 }, { "epoch": 0.6984334203655352, "grad_norm": 5.5350141525268555, "learning_rate": 9.958751206418215e-06, "loss": 0.843, "step": 535 }, { "epoch": 0.6997389033942559, "grad_norm": 4.811600208282471, "learning_rate": 9.958479763263234e-06, "loss": 0.6964, "step": 536 }, { "epoch": 0.7010443864229765, "grad_norm": 5.287219047546387, "learning_rate": 9.958207433624751e-06, "loss": 0.7773, "step": 537 }, { "epoch": 0.7023498694516971, "grad_norm": 4.261974811553955, "learning_rate": 9.957934217551457e-06, "loss": 0.8438, "step": 538 }, { "epoch": 0.7036553524804178, "grad_norm": 4.542464256286621, "learning_rate": 9.957660115092196e-06, "loss": 0.7879, "step": 539 }, { "epoch": 0.7049608355091384, "grad_norm": 4.811707019805908, "learning_rate": 9.957385126295973e-06, "loss": 0.8285, "step": 540 }, { "epoch": 0.706266318537859, "grad_norm": 3.3990023136138916, "learning_rate": 9.95710925121195e-06, "loss": 0.7828, "step": 541 }, { "epoch": 0.7075718015665796, "grad_norm": 4.9082465171813965, "learning_rate": 9.95683248988945e-06, "loss": 0.8391, "step": 542 }, { "epoch": 0.7088772845953003, "grad_norm": 4.565324783325195, "learning_rate": 9.95655484237795e-06, "loss": 0.6178, "step": 543 }, { "epoch": 0.7101827676240209, "grad_norm": 4.190896511077881, "learning_rate": 9.956276308727091e-06, "loss": 0.7128, "step": 544 }, { "epoch": 0.7114882506527415, "grad_norm": 4.144186019897461, "learning_rate": 9.95599688898667e-06, "loss": 0.6682, "step": 545 }, { "epoch": 0.7127937336814621, "grad_norm": 4.983522891998291, "learning_rate": 9.955716583206639e-06, "loss": 0.6032, "step": 546 }, { "epoch": 0.7140992167101827, "grad_norm": 4.6537556648254395, "learning_rate": 9.955435391437114e-06, "loss": 0.6614, "step": 547 }, { "epoch": 0.7154046997389034, "grad_norm": 5.390091896057129, "learning_rate": 9.955153313728365e-06, "loss": 0.7578, "step": 548 }, { "epoch": 0.716710182767624, "grad_norm": 5.474254608154297, "learning_rate": 9.954870350130823e-06, "loss": 0.8781, "step": 549 }, { "epoch": 0.7180156657963447, "grad_norm": 5.731166362762451, "learning_rate": 9.954586500695076e-06, "loss": 0.7256, "step": 550 }, { "epoch": 0.7193211488250653, "grad_norm": 4.953951358795166, "learning_rate": 9.954301765471874e-06, "loss": 0.7795, "step": 551 }, { "epoch": 0.720626631853786, "grad_norm": 6.447653770446777, "learning_rate": 9.954016144512117e-06, "loss": 0.9137, "step": 552 }, { "epoch": 0.7219321148825065, "grad_norm": 4.938129425048828, "learning_rate": 9.953729637866874e-06, "loss": 0.7003, "step": 553 }, { "epoch": 0.7232375979112271, "grad_norm": 4.386180877685547, "learning_rate": 9.953442245587364e-06, "loss": 0.594, "step": 554 }, { "epoch": 0.7245430809399478, "grad_norm": 4.35795783996582, "learning_rate": 9.953153967724967e-06, "loss": 0.7517, "step": 555 }, { "epoch": 0.7258485639686684, "grad_norm": 5.036312580108643, "learning_rate": 9.952864804331225e-06, "loss": 0.7946, "step": 556 }, { "epoch": 0.7271540469973891, "grad_norm": 6.022003173828125, "learning_rate": 9.952574755457831e-06, "loss": 0.8323, "step": 557 }, { "epoch": 0.7284595300261096, "grad_norm": 4.915140628814697, "learning_rate": 9.952283821156643e-06, "loss": 0.6946, "step": 558 }, { "epoch": 0.7297650130548303, "grad_norm": 4.384474754333496, "learning_rate": 9.951992001479674e-06, "loss": 0.7907, "step": 559 }, { "epoch": 0.7310704960835509, "grad_norm": 5.338643550872803, "learning_rate": 9.951699296479094e-06, "loss": 0.7924, "step": 560 }, { "epoch": 0.7323759791122716, "grad_norm": 5.292734622955322, "learning_rate": 9.951405706207236e-06, "loss": 0.6182, "step": 561 }, { "epoch": 0.7336814621409922, "grad_norm": 4.753750801086426, "learning_rate": 9.951111230716588e-06, "loss": 0.6396, "step": 562 }, { "epoch": 0.7349869451697127, "grad_norm": 4.651604652404785, "learning_rate": 9.950815870059793e-06, "loss": 0.59, "step": 563 }, { "epoch": 0.7362924281984334, "grad_norm": 5.738794326782227, "learning_rate": 9.950519624289663e-06, "loss": 0.8649, "step": 564 }, { "epoch": 0.737597911227154, "grad_norm": 4.786860466003418, "learning_rate": 9.950222493459154e-06, "loss": 0.837, "step": 565 }, { "epoch": 0.7389033942558747, "grad_norm": 4.026802062988281, "learning_rate": 9.949924477621391e-06, "loss": 0.7169, "step": 566 }, { "epoch": 0.7402088772845953, "grad_norm": 5.939464569091797, "learning_rate": 9.949625576829654e-06, "loss": 0.8242, "step": 567 }, { "epoch": 0.741514360313316, "grad_norm": 5.163181304931641, "learning_rate": 9.94932579113738e-06, "loss": 0.7026, "step": 568 }, { "epoch": 0.7428198433420365, "grad_norm": 5.3248186111450195, "learning_rate": 9.949025120598167e-06, "loss": 0.6741, "step": 569 }, { "epoch": 0.7441253263707572, "grad_norm": 4.4750776290893555, "learning_rate": 9.948723565265765e-06, "loss": 0.7323, "step": 570 }, { "epoch": 0.7454308093994778, "grad_norm": 3.791274309158325, "learning_rate": 9.948421125194089e-06, "loss": 0.7605, "step": 571 }, { "epoch": 0.7467362924281984, "grad_norm": 5.154573917388916, "learning_rate": 9.94811780043721e-06, "loss": 0.7807, "step": 572 }, { "epoch": 0.7480417754569191, "grad_norm": 4.231841087341309, "learning_rate": 9.947813591049357e-06, "loss": 0.655, "step": 573 }, { "epoch": 0.7493472584856397, "grad_norm": 5.357124328613281, "learning_rate": 9.947508497084915e-06, "loss": 0.8175, "step": 574 }, { "epoch": 0.7506527415143603, "grad_norm": 4.435036659240723, "learning_rate": 9.947202518598432e-06, "loss": 0.6891, "step": 575 }, { "epoch": 0.7519582245430809, "grad_norm": 4.429723739624023, "learning_rate": 9.946895655644609e-06, "loss": 0.7052, "step": 576 }, { "epoch": 0.7532637075718016, "grad_norm": 5.4913530349731445, "learning_rate": 9.946587908278307e-06, "loss": 0.7998, "step": 577 }, { "epoch": 0.7545691906005222, "grad_norm": 4.118889808654785, "learning_rate": 9.946279276554547e-06, "loss": 0.8433, "step": 578 }, { "epoch": 0.7558746736292428, "grad_norm": 5.845130920410156, "learning_rate": 9.945969760528506e-06, "loss": 0.8485, "step": 579 }, { "epoch": 0.7571801566579635, "grad_norm": 4.9625630378723145, "learning_rate": 9.94565936025552e-06, "loss": 0.6822, "step": 580 }, { "epoch": 0.758485639686684, "grad_norm": 4.850309371948242, "learning_rate": 9.945348075791084e-06, "loss": 0.7549, "step": 581 }, { "epoch": 0.7597911227154047, "grad_norm": 4.235795974731445, "learning_rate": 9.945035907190845e-06, "loss": 0.7372, "step": 582 }, { "epoch": 0.7610966057441253, "grad_norm": 5.105690002441406, "learning_rate": 9.944722854510619e-06, "loss": 0.7366, "step": 583 }, { "epoch": 0.762402088772846, "grad_norm": 4.8748064041137695, "learning_rate": 9.94440891780637e-06, "loss": 0.685, "step": 584 }, { "epoch": 0.7637075718015666, "grad_norm": 4.392982006072998, "learning_rate": 9.944094097134225e-06, "loss": 0.8476, "step": 585 }, { "epoch": 0.7650130548302873, "grad_norm": 5.710671424865723, "learning_rate": 9.94377839255047e-06, "loss": 0.7771, "step": 586 }, { "epoch": 0.7663185378590078, "grad_norm": 5.1785993576049805, "learning_rate": 9.943461804111544e-06, "loss": 0.6926, "step": 587 }, { "epoch": 0.7676240208877284, "grad_norm": 4.489890098571777, "learning_rate": 9.943144331874048e-06, "loss": 0.7235, "step": 588 }, { "epoch": 0.7689295039164491, "grad_norm": 5.507998466491699, "learning_rate": 9.94282597589474e-06, "loss": 0.7326, "step": 589 }, { "epoch": 0.7702349869451697, "grad_norm": 5.283398151397705, "learning_rate": 9.94250673623054e-06, "loss": 0.852, "step": 590 }, { "epoch": 0.7715404699738904, "grad_norm": 5.571526050567627, "learning_rate": 9.942186612938515e-06, "loss": 0.7594, "step": 591 }, { "epoch": 0.7728459530026109, "grad_norm": 5.070629119873047, "learning_rate": 9.941865606075904e-06, "loss": 0.664, "step": 592 }, { "epoch": 0.7741514360313316, "grad_norm": 4.749879360198975, "learning_rate": 9.941543715700091e-06, "loss": 0.7738, "step": 593 }, { "epoch": 0.7754569190600522, "grad_norm": 3.4824788570404053, "learning_rate": 9.941220941868628e-06, "loss": 0.7136, "step": 594 }, { "epoch": 0.7767624020887729, "grad_norm": 4.249370574951172, "learning_rate": 9.94089728463922e-06, "loss": 0.6321, "step": 595 }, { "epoch": 0.7780678851174935, "grad_norm": 4.781676292419434, "learning_rate": 9.94057274406973e-06, "loss": 0.6277, "step": 596 }, { "epoch": 0.779373368146214, "grad_norm": 4.233142852783203, "learning_rate": 9.940247320218182e-06, "loss": 0.778, "step": 597 }, { "epoch": 0.7806788511749347, "grad_norm": 4.60373592376709, "learning_rate": 9.939921013142753e-06, "loss": 0.6583, "step": 598 }, { "epoch": 0.7819843342036553, "grad_norm": 4.624236106872559, "learning_rate": 9.939593822901781e-06, "loss": 0.7377, "step": 599 }, { "epoch": 0.783289817232376, "grad_norm": 4.2885637283325195, "learning_rate": 9.939265749553764e-06, "loss": 0.804, "step": 600 }, { "epoch": 0.7845953002610966, "grad_norm": 4.387545108795166, "learning_rate": 9.938936793157351e-06, "loss": 0.7487, "step": 601 }, { "epoch": 0.7859007832898173, "grad_norm": 4.6884236335754395, "learning_rate": 9.938606953771358e-06, "loss": 0.7234, "step": 602 }, { "epoch": 0.7872062663185379, "grad_norm": 3.91796612739563, "learning_rate": 9.938276231454752e-06, "loss": 0.5716, "step": 603 }, { "epoch": 0.7885117493472585, "grad_norm": 4.980195045471191, "learning_rate": 9.937944626266658e-06, "loss": 0.5797, "step": 604 }, { "epoch": 0.7898172323759791, "grad_norm": 6.711381912231445, "learning_rate": 9.937612138266364e-06, "loss": 0.7166, "step": 605 }, { "epoch": 0.7911227154046997, "grad_norm": 4.2306928634643555, "learning_rate": 9.937278767513312e-06, "loss": 0.6649, "step": 606 }, { "epoch": 0.7924281984334204, "grad_norm": 5.685878276824951, "learning_rate": 9.936944514067101e-06, "loss": 0.7274, "step": 607 }, { "epoch": 0.793733681462141, "grad_norm": 4.680644512176514, "learning_rate": 9.936609377987489e-06, "loss": 0.7081, "step": 608 }, { "epoch": 0.7950391644908616, "grad_norm": 4.35481071472168, "learning_rate": 9.936273359334395e-06, "loss": 0.7494, "step": 609 }, { "epoch": 0.7963446475195822, "grad_norm": 5.686800003051758, "learning_rate": 9.93593645816789e-06, "loss": 0.5345, "step": 610 }, { "epoch": 0.7976501305483029, "grad_norm": 5.0050482749938965, "learning_rate": 9.935598674548208e-06, "loss": 0.708, "step": 611 }, { "epoch": 0.7989556135770235, "grad_norm": 3.55810546875, "learning_rate": 9.935260008535736e-06, "loss": 0.557, "step": 612 }, { "epoch": 0.8002610966057441, "grad_norm": 6.2711262702941895, "learning_rate": 9.934920460191022e-06, "loss": 0.76, "step": 613 }, { "epoch": 0.8015665796344648, "grad_norm": 5.1374831199646, "learning_rate": 9.934580029574772e-06, "loss": 0.6926, "step": 614 }, { "epoch": 0.8028720626631853, "grad_norm": 5.445853233337402, "learning_rate": 9.934238716747846e-06, "loss": 0.5818, "step": 615 }, { "epoch": 0.804177545691906, "grad_norm": 5.008171081542969, "learning_rate": 9.933896521771266e-06, "loss": 0.8021, "step": 616 }, { "epoch": 0.8054830287206266, "grad_norm": 5.058992385864258, "learning_rate": 9.933553444706211e-06, "loss": 0.8383, "step": 617 }, { "epoch": 0.8067885117493473, "grad_norm": 5.107936382293701, "learning_rate": 9.933209485614019e-06, "loss": 0.6276, "step": 618 }, { "epoch": 0.8080939947780679, "grad_norm": 4.0084638595581055, "learning_rate": 9.932864644556177e-06, "loss": 0.6035, "step": 619 }, { "epoch": 0.8093994778067886, "grad_norm": 4.290340423583984, "learning_rate": 9.93251892159434e-06, "loss": 0.6911, "step": 620 }, { "epoch": 0.8107049608355091, "grad_norm": 4.1771135330200195, "learning_rate": 9.932172316790317e-06, "loss": 0.7854, "step": 621 }, { "epoch": 0.8120104438642297, "grad_norm": 5.001226902008057, "learning_rate": 9.931824830206073e-06, "loss": 0.7722, "step": 622 }, { "epoch": 0.8133159268929504, "grad_norm": 4.7259039878845215, "learning_rate": 9.931476461903733e-06, "loss": 0.6965, "step": 623 }, { "epoch": 0.814621409921671, "grad_norm": 3.927022933959961, "learning_rate": 9.93112721194558e-06, "loss": 0.6661, "step": 624 }, { "epoch": 0.8159268929503917, "grad_norm": 5.005312919616699, "learning_rate": 9.930777080394049e-06, "loss": 0.7607, "step": 625 }, { "epoch": 0.8172323759791122, "grad_norm": 3.768132209777832, "learning_rate": 9.930426067311743e-06, "loss": 0.6326, "step": 626 }, { "epoch": 0.8185378590078329, "grad_norm": 4.882970333099365, "learning_rate": 9.930074172761411e-06, "loss": 0.7104, "step": 627 }, { "epoch": 0.8198433420365535, "grad_norm": 6.0292134284973145, "learning_rate": 9.929721396805969e-06, "loss": 0.9068, "step": 628 }, { "epoch": 0.8211488250652742, "grad_norm": 5.609174728393555, "learning_rate": 9.929367739508486e-06, "loss": 0.6859, "step": 629 }, { "epoch": 0.8224543080939948, "grad_norm": 4.657751083374023, "learning_rate": 9.929013200932187e-06, "loss": 0.7726, "step": 630 }, { "epoch": 0.8237597911227154, "grad_norm": 3.767678737640381, "learning_rate": 9.928657781140461e-06, "loss": 0.6892, "step": 631 }, { "epoch": 0.825065274151436, "grad_norm": 4.610442161560059, "learning_rate": 9.928301480196848e-06, "loss": 0.6351, "step": 632 }, { "epoch": 0.8263707571801566, "grad_norm": 5.460919380187988, "learning_rate": 9.927944298165046e-06, "loss": 0.7279, "step": 633 }, { "epoch": 0.8276762402088773, "grad_norm": 4.3896260261535645, "learning_rate": 9.927586235108915e-06, "loss": 0.6588, "step": 634 }, { "epoch": 0.8289817232375979, "grad_norm": 5.227085590362549, "learning_rate": 9.92722729109247e-06, "loss": 0.7246, "step": 635 }, { "epoch": 0.8302872062663186, "grad_norm": 6.378992080688477, "learning_rate": 9.926867466179883e-06, "loss": 0.6216, "step": 636 }, { "epoch": 0.8315926892950392, "grad_norm": 4.378927707672119, "learning_rate": 9.926506760435484e-06, "loss": 0.6808, "step": 637 }, { "epoch": 0.8328981723237598, "grad_norm": 6.350910186767578, "learning_rate": 9.92614517392376e-06, "loss": 0.7947, "step": 638 }, { "epoch": 0.8342036553524804, "grad_norm": 4.856624603271484, "learning_rate": 9.925782706709354e-06, "loss": 0.7993, "step": 639 }, { "epoch": 0.835509138381201, "grad_norm": 6.548943042755127, "learning_rate": 9.925419358857074e-06, "loss": 0.7942, "step": 640 }, { "epoch": 0.8368146214099217, "grad_norm": 4.927968978881836, "learning_rate": 9.925055130431876e-06, "loss": 0.763, "step": 641 }, { "epoch": 0.8381201044386423, "grad_norm": 5.082693099975586, "learning_rate": 9.924690021498878e-06, "loss": 0.5639, "step": 642 }, { "epoch": 0.839425587467363, "grad_norm": 5.029530048370361, "learning_rate": 9.924324032123352e-06, "loss": 0.5773, "step": 643 }, { "epoch": 0.8407310704960835, "grad_norm": 5.326099395751953, "learning_rate": 9.923957162370735e-06, "loss": 0.682, "step": 644 }, { "epoch": 0.8420365535248042, "grad_norm": 4.428802490234375, "learning_rate": 9.923589412306612e-06, "loss": 0.7226, "step": 645 }, { "epoch": 0.8433420365535248, "grad_norm": 5.302255153656006, "learning_rate": 9.923220781996732e-06, "loss": 0.7286, "step": 646 }, { "epoch": 0.8446475195822454, "grad_norm": 3.671492576599121, "learning_rate": 9.922851271507e-06, "loss": 0.6649, "step": 647 }, { "epoch": 0.8459530026109661, "grad_norm": 4.688279151916504, "learning_rate": 9.922480880903473e-06, "loss": 0.5919, "step": 648 }, { "epoch": 0.8472584856396866, "grad_norm": 4.232366561889648, "learning_rate": 9.922109610252377e-06, "loss": 0.6814, "step": 649 }, { "epoch": 0.8485639686684073, "grad_norm": 5.093742370605469, "learning_rate": 9.921737459620084e-06, "loss": 0.6567, "step": 650 }, { "epoch": 0.8498694516971279, "grad_norm": 4.772390842437744, "learning_rate": 9.921364429073128e-06, "loss": 0.7615, "step": 651 }, { "epoch": 0.8511749347258486, "grad_norm": 4.386397838592529, "learning_rate": 9.920990518678199e-06, "loss": 0.6197, "step": 652 }, { "epoch": 0.8524804177545692, "grad_norm": 6.18479061126709, "learning_rate": 9.920615728502147e-06, "loss": 0.7016, "step": 653 }, { "epoch": 0.8537859007832899, "grad_norm": 4.941991806030273, "learning_rate": 9.920240058611976e-06, "loss": 0.7988, "step": 654 }, { "epoch": 0.8550913838120104, "grad_norm": 5.674962043762207, "learning_rate": 9.91986350907485e-06, "loss": 0.8119, "step": 655 }, { "epoch": 0.856396866840731, "grad_norm": 5.072239398956299, "learning_rate": 9.91948607995809e-06, "loss": 0.7563, "step": 656 }, { "epoch": 0.8577023498694517, "grad_norm": 4.027274131774902, "learning_rate": 9.919107771329171e-06, "loss": 0.7536, "step": 657 }, { "epoch": 0.8590078328981723, "grad_norm": 3.3328287601470947, "learning_rate": 9.918728583255727e-06, "loss": 0.6897, "step": 658 }, { "epoch": 0.860313315926893, "grad_norm": 4.050899982452393, "learning_rate": 9.918348515805552e-06, "loss": 0.6458, "step": 659 }, { "epoch": 0.8616187989556136, "grad_norm": 5.081582546234131, "learning_rate": 9.917967569046594e-06, "loss": 0.6363, "step": 660 }, { "epoch": 0.8629242819843342, "grad_norm": 3.924694776535034, "learning_rate": 9.91758574304696e-06, "loss": 0.5272, "step": 661 }, { "epoch": 0.8642297650130548, "grad_norm": 5.051509380340576, "learning_rate": 9.917203037874913e-06, "loss": 0.7346, "step": 662 }, { "epoch": 0.8655352480417755, "grad_norm": 4.361501216888428, "learning_rate": 9.916819453598869e-06, "loss": 0.4877, "step": 663 }, { "epoch": 0.8668407310704961, "grad_norm": 5.4952712059021, "learning_rate": 9.916434990287413e-06, "loss": 0.6723, "step": 664 }, { "epoch": 0.8681462140992167, "grad_norm": 5.31600284576416, "learning_rate": 9.916049648009277e-06, "loss": 0.5976, "step": 665 }, { "epoch": 0.8694516971279374, "grad_norm": 4.942739963531494, "learning_rate": 9.91566342683335e-06, "loss": 0.6734, "step": 666 }, { "epoch": 0.8707571801566579, "grad_norm": 5.473392963409424, "learning_rate": 9.915276326828687e-06, "loss": 0.6872, "step": 667 }, { "epoch": 0.8720626631853786, "grad_norm": 4.497568607330322, "learning_rate": 9.914888348064489e-06, "loss": 0.6944, "step": 668 }, { "epoch": 0.8733681462140992, "grad_norm": 4.91173791885376, "learning_rate": 9.914499490610121e-06, "loss": 0.6285, "step": 669 }, { "epoch": 0.8746736292428199, "grad_norm": 5.096947193145752, "learning_rate": 9.914109754535105e-06, "loss": 0.7523, "step": 670 }, { "epoch": 0.8759791122715405, "grad_norm": 4.7527384757995605, "learning_rate": 9.913719139909118e-06, "loss": 0.6989, "step": 671 }, { "epoch": 0.8772845953002611, "grad_norm": 4.501630783081055, "learning_rate": 9.913327646801991e-06, "loss": 0.7088, "step": 672 }, { "epoch": 0.8785900783289817, "grad_norm": 4.643214225769043, "learning_rate": 9.912935275283721e-06, "loss": 0.6707, "step": 673 }, { "epoch": 0.8798955613577023, "grad_norm": 4.930369853973389, "learning_rate": 9.912542025424451e-06, "loss": 0.5605, "step": 674 }, { "epoch": 0.881201044386423, "grad_norm": 5.763865947723389, "learning_rate": 9.912147897294495e-06, "loss": 0.6321, "step": 675 }, { "epoch": 0.8825065274151436, "grad_norm": 4.2301859855651855, "learning_rate": 9.911752890964307e-06, "loss": 0.5098, "step": 676 }, { "epoch": 0.8838120104438643, "grad_norm": 5.032220840454102, "learning_rate": 9.911357006504512e-06, "loss": 0.6642, "step": 677 }, { "epoch": 0.8851174934725848, "grad_norm": 6.185760498046875, "learning_rate": 9.910960243985885e-06, "loss": 0.6732, "step": 678 }, { "epoch": 0.8864229765013055, "grad_norm": 4.283046722412109, "learning_rate": 9.91056260347936e-06, "loss": 0.5549, "step": 679 }, { "epoch": 0.8877284595300261, "grad_norm": 4.3415703773498535, "learning_rate": 9.910164085056027e-06, "loss": 0.6057, "step": 680 }, { "epoch": 0.8890339425587467, "grad_norm": 4.00949764251709, "learning_rate": 9.909764688787133e-06, "loss": 0.5788, "step": 681 }, { "epoch": 0.8903394255874674, "grad_norm": 5.43450403213501, "learning_rate": 9.909364414744086e-06, "loss": 0.6177, "step": 682 }, { "epoch": 0.891644908616188, "grad_norm": 5.2341461181640625, "learning_rate": 9.908963262998443e-06, "loss": 0.6365, "step": 683 }, { "epoch": 0.8929503916449086, "grad_norm": 5.447936058044434, "learning_rate": 9.908561233621928e-06, "loss": 0.5577, "step": 684 }, { "epoch": 0.8942558746736292, "grad_norm": 5.697659015655518, "learning_rate": 9.90815832668641e-06, "loss": 0.7613, "step": 685 }, { "epoch": 0.8955613577023499, "grad_norm": 5.284674644470215, "learning_rate": 9.907754542263926e-06, "loss": 0.7335, "step": 686 }, { "epoch": 0.8968668407310705, "grad_norm": 4.192033290863037, "learning_rate": 9.907349880426664e-06, "loss": 0.6457, "step": 687 }, { "epoch": 0.8981723237597912, "grad_norm": 4.319976806640625, "learning_rate": 9.90694434124697e-06, "loss": 0.6442, "step": 688 }, { "epoch": 0.8994778067885117, "grad_norm": 5.026998996734619, "learning_rate": 9.906537924797343e-06, "loss": 0.5705, "step": 689 }, { "epoch": 0.9007832898172323, "grad_norm": 5.584085464477539, "learning_rate": 9.906130631150448e-06, "loss": 0.6831, "step": 690 }, { "epoch": 0.902088772845953, "grad_norm": 5.847357273101807, "learning_rate": 9.9057224603791e-06, "loss": 0.7069, "step": 691 }, { "epoch": 0.9033942558746736, "grad_norm": 5.536150932312012, "learning_rate": 9.90531341255627e-06, "loss": 0.6625, "step": 692 }, { "epoch": 0.9046997389033943, "grad_norm": 4.548271179199219, "learning_rate": 9.90490348775509e-06, "loss": 0.5618, "step": 693 }, { "epoch": 0.9060052219321149, "grad_norm": 4.848156929016113, "learning_rate": 9.904492686048848e-06, "loss": 0.4544, "step": 694 }, { "epoch": 0.9073107049608355, "grad_norm": 4.376445770263672, "learning_rate": 9.904081007510985e-06, "loss": 0.5424, "step": 695 }, { "epoch": 0.9086161879895561, "grad_norm": 5.1361188888549805, "learning_rate": 9.903668452215106e-06, "loss": 0.757, "step": 696 }, { "epoch": 0.9099216710182768, "grad_norm": 4.635985851287842, "learning_rate": 9.90325502023496e-06, "loss": 0.7151, "step": 697 }, { "epoch": 0.9112271540469974, "grad_norm": 5.051006317138672, "learning_rate": 9.90284071164447e-06, "loss": 0.5903, "step": 698 }, { "epoch": 0.912532637075718, "grad_norm": 3.8899319171905518, "learning_rate": 9.902425526517701e-06, "loss": 0.5417, "step": 699 }, { "epoch": 0.9138381201044387, "grad_norm": 5.900970935821533, "learning_rate": 9.902009464928882e-06, "loss": 0.8008, "step": 700 }, { "epoch": 0.9151436031331592, "grad_norm": 6.042471408843994, "learning_rate": 9.901592526952397e-06, "loss": 0.7795, "step": 701 }, { "epoch": 0.9164490861618799, "grad_norm": 5.104783058166504, "learning_rate": 9.901174712662786e-06, "loss": 0.6771, "step": 702 }, { "epoch": 0.9177545691906005, "grad_norm": 4.828611373901367, "learning_rate": 9.900756022134748e-06, "loss": 0.6427, "step": 703 }, { "epoch": 0.9190600522193212, "grad_norm": 4.073781490325928, "learning_rate": 9.900336455443135e-06, "loss": 0.4977, "step": 704 }, { "epoch": 0.9203655352480418, "grad_norm": 5.358088493347168, "learning_rate": 9.89991601266296e-06, "loss": 0.5105, "step": 705 }, { "epoch": 0.9216710182767625, "grad_norm": 4.909475803375244, "learning_rate": 9.899494693869388e-06, "loss": 0.6264, "step": 706 }, { "epoch": 0.922976501305483, "grad_norm": 5.3793745040893555, "learning_rate": 9.899072499137744e-06, "loss": 0.573, "step": 707 }, { "epoch": 0.9242819843342036, "grad_norm": 6.8178019523620605, "learning_rate": 9.898649428543509e-06, "loss": 0.7121, "step": 708 }, { "epoch": 0.9255874673629243, "grad_norm": 4.192954063415527, "learning_rate": 9.898225482162319e-06, "loss": 0.4991, "step": 709 }, { "epoch": 0.9268929503916449, "grad_norm": 4.895427227020264, "learning_rate": 9.897800660069969e-06, "loss": 0.5957, "step": 710 }, { "epoch": 0.9281984334203656, "grad_norm": 4.833358287811279, "learning_rate": 9.897374962342407e-06, "loss": 0.583, "step": 711 }, { "epoch": 0.9295039164490861, "grad_norm": 4.371520519256592, "learning_rate": 9.896948389055743e-06, "loss": 0.5996, "step": 712 }, { "epoch": 0.9308093994778068, "grad_norm": 3.607424020767212, "learning_rate": 9.896520940286239e-06, "loss": 0.4567, "step": 713 }, { "epoch": 0.9321148825065274, "grad_norm": 5.5188703536987305, "learning_rate": 9.896092616110314e-06, "loss": 0.6042, "step": 714 }, { "epoch": 0.933420365535248, "grad_norm": 4.270289421081543, "learning_rate": 9.895663416604542e-06, "loss": 0.5665, "step": 715 }, { "epoch": 0.9347258485639687, "grad_norm": 5.1854753494262695, "learning_rate": 9.895233341845662e-06, "loss": 0.6264, "step": 716 }, { "epoch": 0.9360313315926893, "grad_norm": 4.048816680908203, "learning_rate": 9.89480239191056e-06, "loss": 0.4198, "step": 717 }, { "epoch": 0.9373368146214099, "grad_norm": 3.7166566848754883, "learning_rate": 9.89437056687628e-06, "loss": 0.5474, "step": 718 }, { "epoch": 0.9386422976501305, "grad_norm": 5.171710968017578, "learning_rate": 9.893937866820026e-06, "loss": 0.6559, "step": 719 }, { "epoch": 0.9399477806788512, "grad_norm": 6.002296447753906, "learning_rate": 9.89350429181916e-06, "loss": 0.7794, "step": 720 }, { "epoch": 0.9412532637075718, "grad_norm": 4.7609663009643555, "learning_rate": 9.89306984195119e-06, "loss": 0.5159, "step": 721 }, { "epoch": 0.9425587467362925, "grad_norm": 5.149716377258301, "learning_rate": 9.892634517293794e-06, "loss": 0.5764, "step": 722 }, { "epoch": 0.943864229765013, "grad_norm": 4.949183940887451, "learning_rate": 9.892198317924797e-06, "loss": 0.615, "step": 723 }, { "epoch": 0.9451697127937336, "grad_norm": 3.691082000732422, "learning_rate": 9.891761243922183e-06, "loss": 0.5226, "step": 724 }, { "epoch": 0.9464751958224543, "grad_norm": 5.445268630981445, "learning_rate": 9.891323295364093e-06, "loss": 0.6273, "step": 725 }, { "epoch": 0.9477806788511749, "grad_norm": 5.309256076812744, "learning_rate": 9.890884472328826e-06, "loss": 0.5535, "step": 726 }, { "epoch": 0.9490861618798956, "grad_norm": 5.907495021820068, "learning_rate": 9.890444774894832e-06, "loss": 0.59, "step": 727 }, { "epoch": 0.9503916449086162, "grad_norm": 4.673529624938965, "learning_rate": 9.890004203140721e-06, "loss": 0.5656, "step": 728 }, { "epoch": 0.9516971279373369, "grad_norm": 4.041966915130615, "learning_rate": 9.889562757145264e-06, "loss": 0.4764, "step": 729 }, { "epoch": 0.9530026109660574, "grad_norm": 5.080813407897949, "learning_rate": 9.889120436987377e-06, "loss": 0.7001, "step": 730 }, { "epoch": 0.9543080939947781, "grad_norm": 5.262834548950195, "learning_rate": 9.88867724274614e-06, "loss": 0.6084, "step": 731 }, { "epoch": 0.9556135770234987, "grad_norm": 4.256019592285156, "learning_rate": 9.888233174500791e-06, "loss": 0.7419, "step": 732 }, { "epoch": 0.9569190600522193, "grad_norm": 4.393291473388672, "learning_rate": 9.88778823233072e-06, "loss": 0.5257, "step": 733 }, { "epoch": 0.95822454308094, "grad_norm": 4.351951599121094, "learning_rate": 9.887342416315471e-06, "loss": 0.6129, "step": 734 }, { "epoch": 0.9595300261096605, "grad_norm": 5.853490352630615, "learning_rate": 9.886895726534751e-06, "loss": 0.5242, "step": 735 }, { "epoch": 0.9608355091383812, "grad_norm": 5.303028583526611, "learning_rate": 9.88644816306842e-06, "loss": 0.6136, "step": 736 }, { "epoch": 0.9621409921671018, "grad_norm": 5.388094425201416, "learning_rate": 9.885999725996492e-06, "loss": 0.5877, "step": 737 }, { "epoch": 0.9634464751958225, "grad_norm": 4.314108371734619, "learning_rate": 9.88555041539914e-06, "loss": 0.4981, "step": 738 }, { "epoch": 0.9647519582245431, "grad_norm": 4.965282440185547, "learning_rate": 9.885100231356692e-06, "loss": 0.4649, "step": 739 }, { "epoch": 0.9660574412532638, "grad_norm": 4.005826950073242, "learning_rate": 9.884649173949633e-06, "loss": 0.3877, "step": 740 }, { "epoch": 0.9673629242819843, "grad_norm": 5.2421875, "learning_rate": 9.884197243258604e-06, "loss": 0.6712, "step": 741 }, { "epoch": 0.9686684073107049, "grad_norm": 5.263722896575928, "learning_rate": 9.8837444393644e-06, "loss": 0.5532, "step": 742 }, { "epoch": 0.9699738903394256, "grad_norm": 5.188045978546143, "learning_rate": 9.883290762347976e-06, "loss": 0.6802, "step": 743 }, { "epoch": 0.9712793733681462, "grad_norm": 3.635991334915161, "learning_rate": 9.88283621229044e-06, "loss": 0.5374, "step": 744 }, { "epoch": 0.9725848563968669, "grad_norm": 5.232844829559326, "learning_rate": 9.882380789273056e-06, "loss": 0.632, "step": 745 }, { "epoch": 0.9738903394255874, "grad_norm": 4.434850215911865, "learning_rate": 9.881924493377247e-06, "loss": 0.4386, "step": 746 }, { "epoch": 0.9751958224543081, "grad_norm": 5.262275218963623, "learning_rate": 9.881467324684591e-06, "loss": 0.5623, "step": 747 }, { "epoch": 0.9765013054830287, "grad_norm": 5.2481842041015625, "learning_rate": 9.881009283276817e-06, "loss": 0.5339, "step": 748 }, { "epoch": 0.9778067885117493, "grad_norm": 5.861098289489746, "learning_rate": 9.880550369235817e-06, "loss": 0.5493, "step": 749 }, { "epoch": 0.97911227154047, "grad_norm": 5.329290866851807, "learning_rate": 9.880090582643637e-06, "loss": 0.5638, "step": 750 }, { "epoch": 0.9804177545691906, "grad_norm": 5.125471115112305, "learning_rate": 9.87962992358248e-06, "loss": 0.5609, "step": 751 }, { "epoch": 0.9817232375979112, "grad_norm": 4.306994915008545, "learning_rate": 9.879168392134695e-06, "loss": 0.4826, "step": 752 }, { "epoch": 0.9830287206266318, "grad_norm": 5.099780082702637, "learning_rate": 9.878705988382805e-06, "loss": 0.537, "step": 753 }, { "epoch": 0.9843342036553525, "grad_norm": 4.647615909576416, "learning_rate": 9.878242712409474e-06, "loss": 0.6142, "step": 754 }, { "epoch": 0.9856396866840731, "grad_norm": 5.4094624519348145, "learning_rate": 9.877778564297527e-06, "loss": 0.5529, "step": 755 }, { "epoch": 0.9869451697127938, "grad_norm": 5.18509578704834, "learning_rate": 9.877313544129947e-06, "loss": 0.5203, "step": 756 }, { "epoch": 0.9882506527415144, "grad_norm": 3.7404799461364746, "learning_rate": 9.87684765198987e-06, "loss": 0.498, "step": 757 }, { "epoch": 0.9895561357702349, "grad_norm": 6.21742057800293, "learning_rate": 9.876380887960588e-06, "loss": 0.6036, "step": 758 }, { "epoch": 0.9908616187989556, "grad_norm": 4.395812034606934, "learning_rate": 9.87591325212555e-06, "loss": 0.5156, "step": 759 }, { "epoch": 0.9921671018276762, "grad_norm": 4.891811370849609, "learning_rate": 9.875444744568362e-06, "loss": 0.4212, "step": 760 }, { "epoch": 0.9934725848563969, "grad_norm": 3.621562957763672, "learning_rate": 9.874975365372782e-06, "loss": 0.4825, "step": 761 }, { "epoch": 0.9947780678851175, "grad_norm": 6.227236270904541, "learning_rate": 9.87450511462273e-06, "loss": 0.6061, "step": 762 }, { "epoch": 0.9960835509138382, "grad_norm": 5.498770713806152, "learning_rate": 9.874033992402274e-06, "loss": 0.4628, "step": 763 }, { "epoch": 0.9973890339425587, "grad_norm": 4.937992572784424, "learning_rate": 9.873561998795644e-06, "loss": 0.5779, "step": 764 }, { "epoch": 0.9986945169712794, "grad_norm": 4.154996871948242, "learning_rate": 9.873089133887224e-06, "loss": 0.4937, "step": 765 }, { "epoch": 1.0, "grad_norm": 5.029611110687256, "learning_rate": 9.87261539776155e-06, "loss": 0.7487, "step": 766 }, { "epoch": 1.0013054830287207, "grad_norm": 4.637132167816162, "learning_rate": 9.872140790503321e-06, "loss": 0.4906, "step": 767 }, { "epoch": 1.0026109660574412, "grad_norm": 3.9247632026672363, "learning_rate": 9.871665312197388e-06, "loss": 0.5269, "step": 768 }, { "epoch": 1.0039164490861618, "grad_norm": 4.640912055969238, "learning_rate": 9.871188962928754e-06, "loss": 0.5082, "step": 769 }, { "epoch": 1.0052219321148825, "grad_norm": 4.986839771270752, "learning_rate": 9.870711742782584e-06, "loss": 0.6138, "step": 770 }, { "epoch": 1.0065274151436032, "grad_norm": 4.93174934387207, "learning_rate": 9.870233651844197e-06, "loss": 0.5109, "step": 771 }, { "epoch": 1.0078328981723237, "grad_norm": 3.777580738067627, "learning_rate": 9.869754690199065e-06, "loss": 0.2581, "step": 772 }, { "epoch": 1.0091383812010444, "grad_norm": 5.1973557472229, "learning_rate": 9.869274857932816e-06, "loss": 0.5863, "step": 773 }, { "epoch": 1.010443864229765, "grad_norm": 4.320672035217285, "learning_rate": 9.868794155131238e-06, "loss": 0.5215, "step": 774 }, { "epoch": 1.0117493472584855, "grad_norm": 4.793182849884033, "learning_rate": 9.86831258188027e-06, "loss": 0.4452, "step": 775 }, { "epoch": 1.0130548302872062, "grad_norm": 5.6731486320495605, "learning_rate": 9.86783013826601e-06, "loss": 0.5818, "step": 776 }, { "epoch": 1.014360313315927, "grad_norm": 4.2367730140686035, "learning_rate": 9.867346824374708e-06, "loss": 0.6639, "step": 777 }, { "epoch": 1.0156657963446476, "grad_norm": 4.460761547088623, "learning_rate": 9.866862640292772e-06, "loss": 0.3694, "step": 778 }, { "epoch": 1.016971279373368, "grad_norm": 4.786682605743408, "learning_rate": 9.866377586106765e-06, "loss": 0.5232, "step": 779 }, { "epoch": 1.0182767624020888, "grad_norm": 3.874709367752075, "learning_rate": 9.865891661903406e-06, "loss": 0.4185, "step": 780 }, { "epoch": 1.0195822454308094, "grad_norm": 4.494345664978027, "learning_rate": 9.865404867769567e-06, "loss": 0.5415, "step": 781 }, { "epoch": 1.0208877284595301, "grad_norm": 4.036777019500732, "learning_rate": 9.864917203792282e-06, "loss": 0.4775, "step": 782 }, { "epoch": 1.0221932114882506, "grad_norm": 4.726285934448242, "learning_rate": 9.864428670058735e-06, "loss": 0.4764, "step": 783 }, { "epoch": 1.0234986945169713, "grad_norm": 4.294616222381592, "learning_rate": 9.863939266656264e-06, "loss": 0.5634, "step": 784 }, { "epoch": 1.024804177545692, "grad_norm": 4.2871856689453125, "learning_rate": 9.863448993672366e-06, "loss": 0.4759, "step": 785 }, { "epoch": 1.0261096605744124, "grad_norm": 3.9311258792877197, "learning_rate": 9.862957851194694e-06, "loss": 0.4316, "step": 786 }, { "epoch": 1.0274151436031331, "grad_norm": 4.423882007598877, "learning_rate": 9.862465839311055e-06, "loss": 0.3238, "step": 787 }, { "epoch": 1.0287206266318538, "grad_norm": 5.30523681640625, "learning_rate": 9.86197295810941e-06, "loss": 0.5849, "step": 788 }, { "epoch": 1.0300261096605745, "grad_norm": 3.751352548599243, "learning_rate": 9.861479207677879e-06, "loss": 0.4122, "step": 789 }, { "epoch": 1.031331592689295, "grad_norm": 4.959531307220459, "learning_rate": 9.860984588104734e-06, "loss": 0.4614, "step": 790 }, { "epoch": 1.0326370757180157, "grad_norm": 5.842597484588623, "learning_rate": 9.860489099478404e-06, "loss": 0.6482, "step": 791 }, { "epoch": 1.0339425587467364, "grad_norm": 3.9046783447265625, "learning_rate": 9.859992741887473e-06, "loss": 0.4241, "step": 792 }, { "epoch": 1.0352480417754568, "grad_norm": 5.108365058898926, "learning_rate": 9.85949551542068e-06, "loss": 0.6234, "step": 793 }, { "epoch": 1.0365535248041775, "grad_norm": 6.527075290679932, "learning_rate": 9.858997420166923e-06, "loss": 0.5641, "step": 794 }, { "epoch": 1.0378590078328982, "grad_norm": 3.830294609069824, "learning_rate": 9.858498456215247e-06, "loss": 0.4179, "step": 795 }, { "epoch": 1.0391644908616189, "grad_norm": 5.400402069091797, "learning_rate": 9.85799862365486e-06, "loss": 0.4586, "step": 796 }, { "epoch": 1.0404699738903394, "grad_norm": 5.075482368469238, "learning_rate": 9.857497922575123e-06, "loss": 0.4399, "step": 797 }, { "epoch": 1.04177545691906, "grad_norm": 4.4987053871154785, "learning_rate": 9.856996353065553e-06, "loss": 0.373, "step": 798 }, { "epoch": 1.0430809399477807, "grad_norm": 5.037891387939453, "learning_rate": 9.856493915215819e-06, "loss": 0.5389, "step": 799 }, { "epoch": 1.0443864229765012, "grad_norm": 5.2861247062683105, "learning_rate": 9.855990609115749e-06, "loss": 0.5381, "step": 800 }, { "epoch": 1.0456919060052219, "grad_norm": 5.369529724121094, "learning_rate": 9.855486434855323e-06, "loss": 0.4829, "step": 801 }, { "epoch": 1.0469973890339426, "grad_norm": 3.6280837059020996, "learning_rate": 9.85498139252468e-06, "loss": 0.3606, "step": 802 }, { "epoch": 1.0483028720626633, "grad_norm": 4.366899013519287, "learning_rate": 9.85447548221411e-06, "loss": 0.38, "step": 803 }, { "epoch": 1.0496083550913837, "grad_norm": 5.451596260070801, "learning_rate": 9.853968704014063e-06, "loss": 0.4226, "step": 804 }, { "epoch": 1.0509138381201044, "grad_norm": 4.752744674682617, "learning_rate": 9.85346105801514e-06, "loss": 0.4745, "step": 805 }, { "epoch": 1.052219321148825, "grad_norm": 5.483562469482422, "learning_rate": 9.852952544308097e-06, "loss": 0.442, "step": 806 }, { "epoch": 1.0535248041775458, "grad_norm": 3.938718795776367, "learning_rate": 9.85244316298385e-06, "loss": 0.5027, "step": 807 }, { "epoch": 1.0548302872062663, "grad_norm": 4.42593240737915, "learning_rate": 9.851932914133465e-06, "loss": 0.4281, "step": 808 }, { "epoch": 1.056135770234987, "grad_norm": 3.9978458881378174, "learning_rate": 9.851421797848164e-06, "loss": 0.4293, "step": 809 }, { "epoch": 1.0574412532637076, "grad_norm": 4.53882360458374, "learning_rate": 9.850909814219327e-06, "loss": 0.3461, "step": 810 }, { "epoch": 1.058746736292428, "grad_norm": 4.179983615875244, "learning_rate": 9.850396963338486e-06, "loss": 0.4834, "step": 811 }, { "epoch": 1.0600522193211488, "grad_norm": 5.095449924468994, "learning_rate": 9.84988324529733e-06, "loss": 0.5181, "step": 812 }, { "epoch": 1.0613577023498695, "grad_norm": 3.0284500122070312, "learning_rate": 9.849368660187701e-06, "loss": 0.3416, "step": 813 }, { "epoch": 1.0626631853785902, "grad_norm": 5.371264934539795, "learning_rate": 9.848853208101599e-06, "loss": 0.5945, "step": 814 }, { "epoch": 1.0639686684073106, "grad_norm": 4.938394546508789, "learning_rate": 9.848336889131176e-06, "loss": 0.4695, "step": 815 }, { "epoch": 1.0652741514360313, "grad_norm": 5.254785060882568, "learning_rate": 9.847819703368739e-06, "loss": 0.5522, "step": 816 }, { "epoch": 1.066579634464752, "grad_norm": 4.8830790519714355, "learning_rate": 9.847301650906753e-06, "loss": 0.456, "step": 817 }, { "epoch": 1.0678851174934727, "grad_norm": 5.0137457847595215, "learning_rate": 9.846782731837835e-06, "loss": 0.4267, "step": 818 }, { "epoch": 1.0691906005221932, "grad_norm": 4.841352462768555, "learning_rate": 9.846262946254758e-06, "loss": 0.433, "step": 819 }, { "epoch": 1.0704960835509139, "grad_norm": 4.191051006317139, "learning_rate": 9.845742294250451e-06, "loss": 0.3965, "step": 820 }, { "epoch": 1.0718015665796345, "grad_norm": 5.601452827453613, "learning_rate": 9.845220775917997e-06, "loss": 0.4968, "step": 821 }, { "epoch": 1.073107049608355, "grad_norm": 4.917871475219727, "learning_rate": 9.844698391350633e-06, "loss": 0.4276, "step": 822 }, { "epoch": 1.0744125326370757, "grad_norm": 4.416004657745361, "learning_rate": 9.844175140641752e-06, "loss": 0.4484, "step": 823 }, { "epoch": 1.0757180156657964, "grad_norm": 4.058265209197998, "learning_rate": 9.843651023884899e-06, "loss": 0.4021, "step": 824 }, { "epoch": 1.0770234986945169, "grad_norm": 3.8106637001037598, "learning_rate": 9.84312604117378e-06, "loss": 0.4211, "step": 825 }, { "epoch": 1.0783289817232375, "grad_norm": 5.096569538116455, "learning_rate": 9.84260019260225e-06, "loss": 0.3695, "step": 826 }, { "epoch": 1.0796344647519582, "grad_norm": 4.563962936401367, "learning_rate": 9.842073478264322e-06, "loss": 0.3911, "step": 827 }, { "epoch": 1.080939947780679, "grad_norm": 3.9786953926086426, "learning_rate": 9.841545898254162e-06, "loss": 0.3363, "step": 828 }, { "epoch": 1.0822454308093994, "grad_norm": 4.931168556213379, "learning_rate": 9.84101745266609e-06, "loss": 0.4832, "step": 829 }, { "epoch": 1.08355091383812, "grad_norm": 4.337825775146484, "learning_rate": 9.840488141594583e-06, "loss": 0.4781, "step": 830 }, { "epoch": 1.0848563968668408, "grad_norm": 4.6602678298950195, "learning_rate": 9.839957965134276e-06, "loss": 0.3952, "step": 831 }, { "epoch": 1.0861618798955615, "grad_norm": 4.231556415557861, "learning_rate": 9.839426923379949e-06, "loss": 0.3375, "step": 832 }, { "epoch": 1.087467362924282, "grad_norm": 5.298889636993408, "learning_rate": 9.838895016426545e-06, "loss": 0.4749, "step": 833 }, { "epoch": 1.0887728459530026, "grad_norm": 4.782133102416992, "learning_rate": 9.838362244369158e-06, "loss": 0.4738, "step": 834 }, { "epoch": 1.0900783289817233, "grad_norm": 4.706536769866943, "learning_rate": 9.837828607303039e-06, "loss": 0.4478, "step": 835 }, { "epoch": 1.0913838120104438, "grad_norm": 4.524285793304443, "learning_rate": 9.837294105323592e-06, "loss": 0.5463, "step": 836 }, { "epoch": 1.0926892950391645, "grad_norm": 3.7831380367279053, "learning_rate": 9.836758738526376e-06, "loss": 0.4411, "step": 837 }, { "epoch": 1.0939947780678851, "grad_norm": 4.965592384338379, "learning_rate": 9.836222507007101e-06, "loss": 0.5259, "step": 838 }, { "epoch": 1.0953002610966058, "grad_norm": 4.313797473907471, "learning_rate": 9.835685410861642e-06, "loss": 0.3525, "step": 839 }, { "epoch": 1.0966057441253263, "grad_norm": 4.741175174713135, "learning_rate": 9.835147450186016e-06, "loss": 0.5578, "step": 840 }, { "epoch": 1.097911227154047, "grad_norm": 4.50051736831665, "learning_rate": 9.834608625076402e-06, "loss": 0.4528, "step": 841 }, { "epoch": 1.0992167101827677, "grad_norm": 4.074885845184326, "learning_rate": 9.834068935629133e-06, "loss": 0.3655, "step": 842 }, { "epoch": 1.1005221932114884, "grad_norm": 4.132756233215332, "learning_rate": 9.833528381940696e-06, "loss": 0.4282, "step": 843 }, { "epoch": 1.1018276762402088, "grad_norm": 4.187598705291748, "learning_rate": 9.832986964107728e-06, "loss": 0.3433, "step": 844 }, { "epoch": 1.1031331592689295, "grad_norm": 4.637507438659668, "learning_rate": 9.832444682227029e-06, "loss": 0.4424, "step": 845 }, { "epoch": 1.1044386422976502, "grad_norm": 5.626105308532715, "learning_rate": 9.831901536395544e-06, "loss": 0.5136, "step": 846 }, { "epoch": 1.1057441253263707, "grad_norm": 6.116469383239746, "learning_rate": 9.831357526710384e-06, "loss": 0.4253, "step": 847 }, { "epoch": 1.1070496083550914, "grad_norm": 5.6301469802856445, "learning_rate": 9.8308126532688e-06, "loss": 0.4361, "step": 848 }, { "epoch": 1.108355091383812, "grad_norm": 5.212791919708252, "learning_rate": 9.830266916168212e-06, "loss": 0.5383, "step": 849 }, { "epoch": 1.1096605744125327, "grad_norm": 4.087728500366211, "learning_rate": 9.829720315506185e-06, "loss": 0.4087, "step": 850 }, { "epoch": 1.1109660574412532, "grad_norm": 4.568334579467773, "learning_rate": 9.829172851380438e-06, "loss": 0.4259, "step": 851 }, { "epoch": 1.112271540469974, "grad_norm": 4.1244072914123535, "learning_rate": 9.82862452388885e-06, "loss": 0.4533, "step": 852 }, { "epoch": 1.1135770234986946, "grad_norm": 5.467619895935059, "learning_rate": 9.828075333129454e-06, "loss": 0.417, "step": 853 }, { "epoch": 1.114882506527415, "grad_norm": 5.856048583984375, "learning_rate": 9.82752527920043e-06, "loss": 0.5299, "step": 854 }, { "epoch": 1.1161879895561357, "grad_norm": 5.483866214752197, "learning_rate": 9.826974362200121e-06, "loss": 0.4655, "step": 855 }, { "epoch": 1.1174934725848564, "grad_norm": 5.497887134552002, "learning_rate": 9.826422582227022e-06, "loss": 0.6048, "step": 856 }, { "epoch": 1.1187989556135771, "grad_norm": 5.896219730377197, "learning_rate": 9.825869939379776e-06, "loss": 0.4517, "step": 857 }, { "epoch": 1.1201044386422976, "grad_norm": 4.422613143920898, "learning_rate": 9.82531643375719e-06, "loss": 0.3979, "step": 858 }, { "epoch": 1.1214099216710183, "grad_norm": 3.7964725494384766, "learning_rate": 9.824762065458217e-06, "loss": 0.3172, "step": 859 }, { "epoch": 1.122715404699739, "grad_norm": 4.699267387390137, "learning_rate": 9.824206834581969e-06, "loss": 0.4254, "step": 860 }, { "epoch": 1.1240208877284594, "grad_norm": 3.6302826404571533, "learning_rate": 9.82365074122771e-06, "loss": 0.4978, "step": 861 }, { "epoch": 1.1253263707571801, "grad_norm": 3.6397194862365723, "learning_rate": 9.823093785494862e-06, "loss": 0.4402, "step": 862 }, { "epoch": 1.1266318537859008, "grad_norm": 3.7338650226593018, "learning_rate": 9.822535967482996e-06, "loss": 0.3978, "step": 863 }, { "epoch": 1.1279373368146215, "grad_norm": 4.297580718994141, "learning_rate": 9.82197728729184e-06, "loss": 0.4549, "step": 864 }, { "epoch": 1.129242819843342, "grad_norm": 5.083653450012207, "learning_rate": 9.821417745021273e-06, "loss": 0.5106, "step": 865 }, { "epoch": 1.1305483028720626, "grad_norm": 4.914961338043213, "learning_rate": 9.820857340771335e-06, "loss": 0.5261, "step": 866 }, { "epoch": 1.1318537859007833, "grad_norm": 4.697208881378174, "learning_rate": 9.820296074642214e-06, "loss": 0.4627, "step": 867 }, { "epoch": 1.133159268929504, "grad_norm": 4.744974613189697, "learning_rate": 9.819733946734253e-06, "loss": 0.3893, "step": 868 }, { "epoch": 1.1344647519582245, "grad_norm": 4.243831157684326, "learning_rate": 9.81917095714795e-06, "loss": 0.3809, "step": 869 }, { "epoch": 1.1357702349869452, "grad_norm": 4.39644193649292, "learning_rate": 9.81860710598396e-06, "loss": 0.3559, "step": 870 }, { "epoch": 1.1370757180156659, "grad_norm": 5.276587963104248, "learning_rate": 9.818042393343086e-06, "loss": 0.5737, "step": 871 }, { "epoch": 1.1383812010443863, "grad_norm": 4.845165252685547, "learning_rate": 9.817476819326288e-06, "loss": 0.4048, "step": 872 }, { "epoch": 1.139686684073107, "grad_norm": 4.105392932891846, "learning_rate": 9.816910384034682e-06, "loss": 0.3686, "step": 873 }, { "epoch": 1.1409921671018277, "grad_norm": 4.9760212898254395, "learning_rate": 9.816343087569533e-06, "loss": 0.4512, "step": 874 }, { "epoch": 1.1422976501305482, "grad_norm": 5.826573848724365, "learning_rate": 9.815774930032266e-06, "loss": 0.4588, "step": 875 }, { "epoch": 1.1436031331592689, "grad_norm": 4.787039279937744, "learning_rate": 9.815205911524456e-06, "loss": 0.4078, "step": 876 }, { "epoch": 1.1449086161879896, "grad_norm": 5.111564636230469, "learning_rate": 9.814636032147832e-06, "loss": 0.3168, "step": 877 }, { "epoch": 1.1462140992167102, "grad_norm": 5.707624912261963, "learning_rate": 9.81406529200428e-06, "loss": 0.4732, "step": 878 }, { "epoch": 1.1475195822454307, "grad_norm": 5.176662921905518, "learning_rate": 9.813493691195835e-06, "loss": 0.4179, "step": 879 }, { "epoch": 1.1488250652741514, "grad_norm": 5.659109115600586, "learning_rate": 9.81292122982469e-06, "loss": 0.457, "step": 880 }, { "epoch": 1.150130548302872, "grad_norm": 5.313989639282227, "learning_rate": 9.812347907993191e-06, "loss": 0.4704, "step": 881 }, { "epoch": 1.1514360313315928, "grad_norm": 4.156805515289307, "learning_rate": 9.811773725803834e-06, "loss": 0.3841, "step": 882 }, { "epoch": 1.1527415143603132, "grad_norm": 4.3481268882751465, "learning_rate": 9.811198683359276e-06, "loss": 0.4849, "step": 883 }, { "epoch": 1.154046997389034, "grad_norm": 4.462480545043945, "learning_rate": 9.810622780762322e-06, "loss": 0.2962, "step": 884 }, { "epoch": 1.1553524804177546, "grad_norm": 5.006856918334961, "learning_rate": 9.810046018115934e-06, "loss": 0.4063, "step": 885 }, { "epoch": 1.156657963446475, "grad_norm": 4.138820648193359, "learning_rate": 9.809468395523225e-06, "loss": 0.3353, "step": 886 }, { "epoch": 1.1579634464751958, "grad_norm": 4.414172172546387, "learning_rate": 9.808889913087463e-06, "loss": 0.3669, "step": 887 }, { "epoch": 1.1592689295039165, "grad_norm": 6.479883193969727, "learning_rate": 9.80831057091207e-06, "loss": 0.5981, "step": 888 }, { "epoch": 1.1605744125326372, "grad_norm": 5.525407314300537, "learning_rate": 9.80773036910062e-06, "loss": 0.5117, "step": 889 }, { "epoch": 1.1618798955613576, "grad_norm": 4.596141815185547, "learning_rate": 9.807149307756847e-06, "loss": 0.4211, "step": 890 }, { "epoch": 1.1631853785900783, "grad_norm": 3.7249648571014404, "learning_rate": 9.806567386984629e-06, "loss": 0.3113, "step": 891 }, { "epoch": 1.164490861618799, "grad_norm": 5.386457920074463, "learning_rate": 9.805984606888006e-06, "loss": 0.4705, "step": 892 }, { "epoch": 1.1657963446475197, "grad_norm": 3.9536068439483643, "learning_rate": 9.805400967571166e-06, "loss": 0.2835, "step": 893 }, { "epoch": 1.1671018276762402, "grad_norm": 3.872854232788086, "learning_rate": 9.804816469138451e-06, "loss": 0.3513, "step": 894 }, { "epoch": 1.1684073107049608, "grad_norm": 4.718245029449463, "learning_rate": 9.804231111694364e-06, "loss": 0.3939, "step": 895 }, { "epoch": 1.1697127937336815, "grad_norm": 5.740579128265381, "learning_rate": 9.803644895343552e-06, "loss": 0.5252, "step": 896 }, { "epoch": 1.171018276762402, "grad_norm": 5.576089859008789, "learning_rate": 9.80305782019082e-06, "loss": 0.5629, "step": 897 }, { "epoch": 1.1723237597911227, "grad_norm": 5.544049263000488, "learning_rate": 9.802469886341127e-06, "loss": 0.5557, "step": 898 }, { "epoch": 1.1736292428198434, "grad_norm": 3.600653648376465, "learning_rate": 9.801881093899584e-06, "loss": 0.4048, "step": 899 }, { "epoch": 1.174934725848564, "grad_norm": 4.285521984100342, "learning_rate": 9.801291442971456e-06, "loss": 0.486, "step": 900 }, { "epoch": 1.1762402088772845, "grad_norm": 5.207653045654297, "learning_rate": 9.800700933662162e-06, "loss": 0.5222, "step": 901 }, { "epoch": 1.1775456919060052, "grad_norm": 4.696203231811523, "learning_rate": 9.800109566077271e-06, "loss": 0.3551, "step": 902 }, { "epoch": 1.178851174934726, "grad_norm": 4.782215118408203, "learning_rate": 9.799517340322515e-06, "loss": 0.469, "step": 903 }, { "epoch": 1.1801566579634466, "grad_norm": 4.36832332611084, "learning_rate": 9.798924256503766e-06, "loss": 0.3972, "step": 904 }, { "epoch": 1.181462140992167, "grad_norm": 4.662067413330078, "learning_rate": 9.798330314727061e-06, "loss": 0.317, "step": 905 }, { "epoch": 1.1827676240208878, "grad_norm": 3.5920846462249756, "learning_rate": 9.797735515098582e-06, "loss": 0.4295, "step": 906 }, { "epoch": 1.1840731070496084, "grad_norm": 5.081185817718506, "learning_rate": 9.797139857724673e-06, "loss": 0.5402, "step": 907 }, { "epoch": 1.185378590078329, "grad_norm": 4.606866359710693, "learning_rate": 9.796543342711824e-06, "loss": 0.4084, "step": 908 }, { "epoch": 1.1866840731070496, "grad_norm": 4.225372314453125, "learning_rate": 9.795945970166678e-06, "loss": 0.443, "step": 909 }, { "epoch": 1.1879895561357703, "grad_norm": 5.047664165496826, "learning_rate": 9.795347740196036e-06, "loss": 0.4907, "step": 910 }, { "epoch": 1.1892950391644908, "grad_norm": 5.364892959594727, "learning_rate": 9.79474865290685e-06, "loss": 0.5173, "step": 911 }, { "epoch": 1.1906005221932114, "grad_norm": 5.186886310577393, "learning_rate": 9.794148708406229e-06, "loss": 0.5371, "step": 912 }, { "epoch": 1.1919060052219321, "grad_norm": 5.127737045288086, "learning_rate": 9.793547906801427e-06, "loss": 0.476, "step": 913 }, { "epoch": 1.1932114882506528, "grad_norm": 5.4123454093933105, "learning_rate": 9.792946248199859e-06, "loss": 0.3516, "step": 914 }, { "epoch": 1.1945169712793733, "grad_norm": 5.023273468017578, "learning_rate": 9.792343732709089e-06, "loss": 0.4311, "step": 915 }, { "epoch": 1.195822454308094, "grad_norm": 5.361569881439209, "learning_rate": 9.791740360436837e-06, "loss": 0.5834, "step": 916 }, { "epoch": 1.1971279373368147, "grad_norm": 4.093585968017578, "learning_rate": 9.791136131490973e-06, "loss": 0.2231, "step": 917 }, { "epoch": 1.1984334203655354, "grad_norm": 3.935002326965332, "learning_rate": 9.79053104597952e-06, "loss": 0.4717, "step": 918 }, { "epoch": 1.1997389033942558, "grad_norm": 3.807569980621338, "learning_rate": 9.789925104010663e-06, "loss": 0.5334, "step": 919 }, { "epoch": 1.2010443864229765, "grad_norm": 3.215235948562622, "learning_rate": 9.789318305692726e-06, "loss": 0.3315, "step": 920 }, { "epoch": 1.2023498694516972, "grad_norm": 4.508053779602051, "learning_rate": 9.788710651134197e-06, "loss": 0.4081, "step": 921 }, { "epoch": 1.2036553524804177, "grad_norm": 4.884442329406738, "learning_rate": 9.788102140443712e-06, "loss": 0.4767, "step": 922 }, { "epoch": 1.2049608355091384, "grad_norm": 3.816801071166992, "learning_rate": 9.787492773730061e-06, "loss": 0.288, "step": 923 }, { "epoch": 1.206266318537859, "grad_norm": 5.29806661605835, "learning_rate": 9.786882551102185e-06, "loss": 0.3748, "step": 924 }, { "epoch": 1.2075718015665797, "grad_norm": 4.9355926513671875, "learning_rate": 9.786271472669186e-06, "loss": 0.5751, "step": 925 }, { "epoch": 1.2088772845953002, "grad_norm": 3.4193480014801025, "learning_rate": 9.78565953854031e-06, "loss": 0.5097, "step": 926 }, { "epoch": 1.2101827676240209, "grad_norm": 3.8810713291168213, "learning_rate": 9.785046748824962e-06, "loss": 0.3844, "step": 927 }, { "epoch": 1.2114882506527416, "grad_norm": 4.186249256134033, "learning_rate": 9.784433103632693e-06, "loss": 0.3249, "step": 928 }, { "epoch": 1.2127937336814623, "grad_norm": 4.94191312789917, "learning_rate": 9.783818603073213e-06, "loss": 0.451, "step": 929 }, { "epoch": 1.2140992167101827, "grad_norm": 4.161808013916016, "learning_rate": 9.783203247256388e-06, "loss": 0.4026, "step": 930 }, { "epoch": 1.2154046997389034, "grad_norm": 4.628011703491211, "learning_rate": 9.782587036292224e-06, "loss": 0.373, "step": 931 }, { "epoch": 1.216710182767624, "grad_norm": 5.061295032501221, "learning_rate": 9.781969970290894e-06, "loss": 0.4965, "step": 932 }, { "epoch": 1.2180156657963446, "grad_norm": 5.058967590332031, "learning_rate": 9.781352049362713e-06, "loss": 0.39, "step": 933 }, { "epoch": 1.2193211488250653, "grad_norm": 5.249612808227539, "learning_rate": 9.780733273618159e-06, "loss": 0.3649, "step": 934 }, { "epoch": 1.220626631853786, "grad_norm": 3.1688971519470215, "learning_rate": 9.780113643167855e-06, "loss": 0.3705, "step": 935 }, { "epoch": 1.2219321148825064, "grad_norm": 4.939879894256592, "learning_rate": 9.779493158122578e-06, "loss": 0.3412, "step": 936 }, { "epoch": 1.223237597911227, "grad_norm": 4.736397743225098, "learning_rate": 9.778871818593262e-06, "loss": 0.473, "step": 937 }, { "epoch": 1.2245430809399478, "grad_norm": 5.240589618682861, "learning_rate": 9.77824962469099e-06, "loss": 0.4302, "step": 938 }, { "epoch": 1.2258485639686685, "grad_norm": 4.780882358551025, "learning_rate": 9.777626576526994e-06, "loss": 0.4466, "step": 939 }, { "epoch": 1.227154046997389, "grad_norm": 5.050500869750977, "learning_rate": 9.777002674212673e-06, "loss": 0.4032, "step": 940 }, { "epoch": 1.2284595300261096, "grad_norm": 6.087315559387207, "learning_rate": 9.77637791785956e-06, "loss": 0.4624, "step": 941 }, { "epoch": 1.2297650130548303, "grad_norm": 4.100681781768799, "learning_rate": 9.775752307579353e-06, "loss": 0.3308, "step": 942 }, { "epoch": 1.231070496083551, "grad_norm": 3.78788161277771, "learning_rate": 9.775125843483902e-06, "loss": 0.3175, "step": 943 }, { "epoch": 1.2323759791122715, "grad_norm": 5.811645984649658, "learning_rate": 9.774498525685204e-06, "loss": 0.4135, "step": 944 }, { "epoch": 1.2336814621409922, "grad_norm": 5.297332286834717, "learning_rate": 9.77387035429541e-06, "loss": 0.5963, "step": 945 }, { "epoch": 1.2349869451697129, "grad_norm": 4.352571487426758, "learning_rate": 9.77324132942683e-06, "loss": 0.3479, "step": 946 }, { "epoch": 1.2362924281984333, "grad_norm": 4.361113548278809, "learning_rate": 9.77261145119192e-06, "loss": 0.4659, "step": 947 }, { "epoch": 1.237597911227154, "grad_norm": 4.750274181365967, "learning_rate": 9.77198071970329e-06, "loss": 0.5314, "step": 948 }, { "epoch": 1.2389033942558747, "grad_norm": 5.038723468780518, "learning_rate": 9.771349135073701e-06, "loss": 0.35, "step": 949 }, { "epoch": 1.2402088772845954, "grad_norm": 4.079617023468018, "learning_rate": 9.770716697416071e-06, "loss": 0.2175, "step": 950 }, { "epoch": 1.2415143603133159, "grad_norm": 4.287713527679443, "learning_rate": 9.77008340684347e-06, "loss": 0.3625, "step": 951 }, { "epoch": 1.2428198433420365, "grad_norm": 4.2200140953063965, "learning_rate": 9.769449263469114e-06, "loss": 0.4226, "step": 952 }, { "epoch": 1.2441253263707572, "grad_norm": 5.344121932983398, "learning_rate": 9.76881426740638e-06, "loss": 0.5048, "step": 953 }, { "epoch": 1.245430809399478, "grad_norm": 4.596002578735352, "learning_rate": 9.768178418768792e-06, "loss": 0.4298, "step": 954 }, { "epoch": 1.2467362924281984, "grad_norm": 4.632657051086426, "learning_rate": 9.767541717670026e-06, "loss": 0.4217, "step": 955 }, { "epoch": 1.248041775456919, "grad_norm": 3.9735774993896484, "learning_rate": 9.766904164223914e-06, "loss": 0.261, "step": 956 }, { "epoch": 1.2493472584856398, "grad_norm": 4.167870044708252, "learning_rate": 9.766265758544438e-06, "loss": 0.2588, "step": 957 }, { "epoch": 1.2506527415143602, "grad_norm": 4.105578899383545, "learning_rate": 9.765626500745732e-06, "loss": 0.3115, "step": 958 }, { "epoch": 1.251958224543081, "grad_norm": 3.8330562114715576, "learning_rate": 9.764986390942087e-06, "loss": 0.3346, "step": 959 }, { "epoch": 1.2532637075718016, "grad_norm": 4.45250940322876, "learning_rate": 9.76434542924794e-06, "loss": 0.4157, "step": 960 }, { "epoch": 1.254569190600522, "grad_norm": 5.148488998413086, "learning_rate": 9.763703615777881e-06, "loss": 0.4253, "step": 961 }, { "epoch": 1.2558746736292428, "grad_norm": 3.722425699234009, "learning_rate": 9.76306095064666e-06, "loss": 0.3851, "step": 962 }, { "epoch": 1.2571801566579635, "grad_norm": 3.3586583137512207, "learning_rate": 9.762417433969168e-06, "loss": 0.2075, "step": 963 }, { "epoch": 1.2584856396866841, "grad_norm": 5.675699234008789, "learning_rate": 9.761773065860457e-06, "loss": 0.4938, "step": 964 }, { "epoch": 1.2597911227154048, "grad_norm": 4.085519790649414, "learning_rate": 9.761127846435726e-06, "loss": 0.336, "step": 965 }, { "epoch": 1.2610966057441253, "grad_norm": 4.089145183563232, "learning_rate": 9.760481775810328e-06, "loss": 0.3656, "step": 966 }, { "epoch": 1.262402088772846, "grad_norm": 3.8827743530273438, "learning_rate": 9.759834854099773e-06, "loss": 0.2802, "step": 967 }, { "epoch": 1.2637075718015667, "grad_norm": 3.1665940284729004, "learning_rate": 9.759187081419711e-06, "loss": 0.2961, "step": 968 }, { "epoch": 1.2650130548302871, "grad_norm": 5.066347122192383, "learning_rate": 9.758538457885957e-06, "loss": 0.3641, "step": 969 }, { "epoch": 1.2663185378590078, "grad_norm": 4.626343250274658, "learning_rate": 9.757888983614471e-06, "loss": 0.4419, "step": 970 }, { "epoch": 1.2676240208877285, "grad_norm": 5.140439987182617, "learning_rate": 9.757238658721368e-06, "loss": 0.4816, "step": 971 }, { "epoch": 1.268929503916449, "grad_norm": 4.335917949676514, "learning_rate": 9.756587483322912e-06, "loss": 0.3329, "step": 972 }, { "epoch": 1.2702349869451697, "grad_norm": 4.044816970825195, "learning_rate": 9.755935457535522e-06, "loss": 0.4982, "step": 973 }, { "epoch": 1.2715404699738904, "grad_norm": 4.603396415710449, "learning_rate": 9.755282581475769e-06, "loss": 0.4274, "step": 974 }, { "epoch": 1.2728459530026108, "grad_norm": 5.62738561630249, "learning_rate": 9.754628855260373e-06, "loss": 0.4924, "step": 975 }, { "epoch": 1.2741514360313315, "grad_norm": 5.222236156463623, "learning_rate": 9.75397427900621e-06, "loss": 0.4363, "step": 976 }, { "epoch": 1.2754569190600522, "grad_norm": 4.63100528717041, "learning_rate": 9.753318852830305e-06, "loss": 0.4231, "step": 977 }, { "epoch": 1.276762402088773, "grad_norm": 4.2783331871032715, "learning_rate": 9.752662576849835e-06, "loss": 0.4377, "step": 978 }, { "epoch": 1.2780678851174936, "grad_norm": 5.0516886711120605, "learning_rate": 9.752005451182131e-06, "loss": 0.4145, "step": 979 }, { "epoch": 1.279373368146214, "grad_norm": 4.937648773193359, "learning_rate": 9.751347475944677e-06, "loss": 0.3683, "step": 980 }, { "epoch": 1.2806788511749347, "grad_norm": 11.149308204650879, "learning_rate": 9.750688651255102e-06, "loss": 0.4675, "step": 981 }, { "epoch": 1.2819843342036554, "grad_norm": 4.1352620124816895, "learning_rate": 9.750028977231196e-06, "loss": 0.3547, "step": 982 }, { "epoch": 1.283289817232376, "grad_norm": 3.195274591445923, "learning_rate": 9.749368453990892e-06, "loss": 0.2324, "step": 983 }, { "epoch": 1.2845953002610966, "grad_norm": 4.986560344696045, "learning_rate": 9.748707081652285e-06, "loss": 0.5102, "step": 984 }, { "epoch": 1.2859007832898173, "grad_norm": 3.88100004196167, "learning_rate": 9.74804486033361e-06, "loss": 0.278, "step": 985 }, { "epoch": 1.2872062663185377, "grad_norm": 5.440158843994141, "learning_rate": 9.747381790153262e-06, "loss": 0.4475, "step": 986 }, { "epoch": 1.2885117493472584, "grad_norm": 4.544145107269287, "learning_rate": 9.746717871229788e-06, "loss": 0.3631, "step": 987 }, { "epoch": 1.2898172323759791, "grad_norm": 4.182274341583252, "learning_rate": 9.746053103681879e-06, "loss": 0.3661, "step": 988 }, { "epoch": 1.2911227154046998, "grad_norm": 4.371706485748291, "learning_rate": 9.745387487628388e-06, "loss": 0.4047, "step": 989 }, { "epoch": 1.2924281984334205, "grad_norm": 5.455784320831299, "learning_rate": 9.744721023188314e-06, "loss": 0.474, "step": 990 }, { "epoch": 1.293733681462141, "grad_norm": 3.266099691390991, "learning_rate": 9.744053710480806e-06, "loss": 0.2109, "step": 991 }, { "epoch": 1.2950391644908616, "grad_norm": 3.6263699531555176, "learning_rate": 9.743385549625168e-06, "loss": 0.2988, "step": 992 }, { "epoch": 1.2963446475195823, "grad_norm": 4.131960391998291, "learning_rate": 9.742716540740856e-06, "loss": 0.328, "step": 993 }, { "epoch": 1.2976501305483028, "grad_norm": 5.587815761566162, "learning_rate": 9.742046683947475e-06, "loss": 0.3925, "step": 994 }, { "epoch": 1.2989556135770235, "grad_norm": 4.439681053161621, "learning_rate": 9.741375979364784e-06, "loss": 0.3241, "step": 995 }, { "epoch": 1.3002610966057442, "grad_norm": 5.0455451011657715, "learning_rate": 9.74070442711269e-06, "loss": 0.347, "step": 996 }, { "epoch": 1.3015665796344646, "grad_norm": 5.311579704284668, "learning_rate": 9.740032027311257e-06, "loss": 0.5104, "step": 997 }, { "epoch": 1.3028720626631853, "grad_norm": 5.186616897583008, "learning_rate": 9.739358780080695e-06, "loss": 0.4497, "step": 998 }, { "epoch": 1.304177545691906, "grad_norm": 5.080558776855469, "learning_rate": 9.73868468554137e-06, "loss": 0.3338, "step": 999 }, { "epoch": 1.3054830287206267, "grad_norm": 4.733298301696777, "learning_rate": 9.738009743813797e-06, "loss": 0.491, "step": 1000 }, { "epoch": 1.3067885117493472, "grad_norm": 3.804081439971924, "learning_rate": 9.737333955018641e-06, "loss": 0.2184, "step": 1001 }, { "epoch": 1.3080939947780679, "grad_norm": 3.0834076404571533, "learning_rate": 9.736657319276723e-06, "loss": 0.2164, "step": 1002 }, { "epoch": 1.3093994778067886, "grad_norm": 3.448531150817871, "learning_rate": 9.735979836709014e-06, "loss": 0.339, "step": 1003 }, { "epoch": 1.3107049608355092, "grad_norm": 5.885087490081787, "learning_rate": 9.73530150743663e-06, "loss": 0.4364, "step": 1004 }, { "epoch": 1.3120104438642297, "grad_norm": 5.453160285949707, "learning_rate": 9.734622331580847e-06, "loss": 0.3824, "step": 1005 }, { "epoch": 1.3133159268929504, "grad_norm": 6.290554523468018, "learning_rate": 9.73394230926309e-06, "loss": 0.5501, "step": 1006 }, { "epoch": 1.314621409921671, "grad_norm": 3.170377254486084, "learning_rate": 9.733261440604932e-06, "loss": 0.2723, "step": 1007 }, { "epoch": 1.3159268929503916, "grad_norm": 5.138110637664795, "learning_rate": 9.7325797257281e-06, "loss": 0.4394, "step": 1008 }, { "epoch": 1.3172323759791122, "grad_norm": 3.1678969860076904, "learning_rate": 9.731897164754474e-06, "loss": 0.4832, "step": 1009 }, { "epoch": 1.318537859007833, "grad_norm": 3.4497666358947754, "learning_rate": 9.73121375780608e-06, "loss": 0.2428, "step": 1010 }, { "epoch": 1.3198433420365534, "grad_norm": 3.605051040649414, "learning_rate": 9.7305295050051e-06, "loss": 0.372, "step": 1011 }, { "epoch": 1.321148825065274, "grad_norm": 3.3613574504852295, "learning_rate": 9.729844406473866e-06, "loss": 0.4428, "step": 1012 }, { "epoch": 1.3224543080939948, "grad_norm": 3.7124581336975098, "learning_rate": 9.72915846233486e-06, "loss": 0.3794, "step": 1013 }, { "epoch": 1.3237597911227155, "grad_norm": 4.8300275802612305, "learning_rate": 9.728471672710718e-06, "loss": 0.4136, "step": 1014 }, { "epoch": 1.3250652741514362, "grad_norm": 4.142878532409668, "learning_rate": 9.72778403772422e-06, "loss": 0.2894, "step": 1015 }, { "epoch": 1.3263707571801566, "grad_norm": 5.318364143371582, "learning_rate": 9.727095557498309e-06, "loss": 0.4484, "step": 1016 }, { "epoch": 1.3276762402088773, "grad_norm": 3.617779493331909, "learning_rate": 9.726406232156068e-06, "loss": 0.4146, "step": 1017 }, { "epoch": 1.328981723237598, "grad_norm": 4.326317310333252, "learning_rate": 9.725716061820736e-06, "loss": 0.3516, "step": 1018 }, { "epoch": 1.3302872062663185, "grad_norm": 4.865210056304932, "learning_rate": 9.725025046615704e-06, "loss": 0.3491, "step": 1019 }, { "epoch": 1.3315926892950392, "grad_norm": 5.737142562866211, "learning_rate": 9.724333186664511e-06, "loss": 0.4288, "step": 1020 }, { "epoch": 1.3328981723237598, "grad_norm": 4.232952117919922, "learning_rate": 9.723640482090849e-06, "loss": 0.2694, "step": 1021 }, { "epoch": 1.3342036553524803, "grad_norm": 5.008326053619385, "learning_rate": 9.722946933018562e-06, "loss": 0.3802, "step": 1022 }, { "epoch": 1.335509138381201, "grad_norm": 4.3083977699279785, "learning_rate": 9.722252539571643e-06, "loss": 0.2998, "step": 1023 }, { "epoch": 1.3368146214099217, "grad_norm": 3.5800437927246094, "learning_rate": 9.721557301874236e-06, "loss": 0.1428, "step": 1024 }, { "epoch": 1.3381201044386424, "grad_norm": 4.447956562042236, "learning_rate": 9.720861220050636e-06, "loss": 0.3954, "step": 1025 }, { "epoch": 1.3394255874673628, "grad_norm": 3.7465293407440186, "learning_rate": 9.72016429422529e-06, "loss": 0.2786, "step": 1026 }, { "epoch": 1.3407310704960835, "grad_norm": 3.6874325275421143, "learning_rate": 9.719466524522795e-06, "loss": 0.3461, "step": 1027 }, { "epoch": 1.3420365535248042, "grad_norm": 3.432619094848633, "learning_rate": 9.718767911067902e-06, "loss": 0.2212, "step": 1028 }, { "epoch": 1.343342036553525, "grad_norm": 4.597324371337891, "learning_rate": 9.718068453985506e-06, "loss": 0.3387, "step": 1029 }, { "epoch": 1.3446475195822454, "grad_norm": 4.920337677001953, "learning_rate": 9.717368153400659e-06, "loss": 0.3636, "step": 1030 }, { "epoch": 1.345953002610966, "grad_norm": 4.65589714050293, "learning_rate": 9.71666700943856e-06, "loss": 0.2906, "step": 1031 }, { "epoch": 1.3472584856396868, "grad_norm": 5.226668357849121, "learning_rate": 9.715965022224563e-06, "loss": 0.4855, "step": 1032 }, { "epoch": 1.3485639686684072, "grad_norm": 4.706356048583984, "learning_rate": 9.71526219188417e-06, "loss": 0.2682, "step": 1033 }, { "epoch": 1.349869451697128, "grad_norm": 3.995593547821045, "learning_rate": 9.714558518543028e-06, "loss": 0.4182, "step": 1034 }, { "epoch": 1.3511749347258486, "grad_norm": 4.559717655181885, "learning_rate": 9.71385400232695e-06, "loss": 0.4687, "step": 1035 }, { "epoch": 1.352480417754569, "grad_norm": 5.000036716461182, "learning_rate": 9.713148643361884e-06, "loss": 0.4255, "step": 1036 }, { "epoch": 1.3537859007832898, "grad_norm": 4.097792625427246, "learning_rate": 9.712442441773937e-06, "loss": 0.2697, "step": 1037 }, { "epoch": 1.3550913838120104, "grad_norm": 3.8433144092559814, "learning_rate": 9.711735397689363e-06, "loss": 0.3423, "step": 1038 }, { "epoch": 1.3563968668407311, "grad_norm": 5.342410087585449, "learning_rate": 9.711027511234572e-06, "loss": 0.4728, "step": 1039 }, { "epoch": 1.3577023498694518, "grad_norm": 4.950399875640869, "learning_rate": 9.710318782536117e-06, "loss": 0.3843, "step": 1040 }, { "epoch": 1.3590078328981723, "grad_norm": 4.276471138000488, "learning_rate": 9.709609211720707e-06, "loss": 0.3356, "step": 1041 }, { "epoch": 1.360313315926893, "grad_norm": 4.815263748168945, "learning_rate": 9.7088987989152e-06, "loss": 0.306, "step": 1042 }, { "epoch": 1.3616187989556137, "grad_norm": 4.875151634216309, "learning_rate": 9.708187544246605e-06, "loss": 0.4256, "step": 1043 }, { "epoch": 1.3629242819843341, "grad_norm": 3.964383125305176, "learning_rate": 9.707475447842079e-06, "loss": 0.3294, "step": 1044 }, { "epoch": 1.3642297650130548, "grad_norm": 4.06675910949707, "learning_rate": 9.706762509828934e-06, "loss": 0.4102, "step": 1045 }, { "epoch": 1.3655352480417755, "grad_norm": 5.499902725219727, "learning_rate": 9.706048730334629e-06, "loss": 0.4525, "step": 1046 }, { "epoch": 1.366840731070496, "grad_norm": 4.503942012786865, "learning_rate": 9.705334109486775e-06, "loss": 0.3193, "step": 1047 }, { "epoch": 1.3681462140992167, "grad_norm": 5.074225902557373, "learning_rate": 9.70461864741313e-06, "loss": 0.4419, "step": 1048 }, { "epoch": 1.3694516971279374, "grad_norm": 4.521297931671143, "learning_rate": 9.70390234424161e-06, "loss": 0.2533, "step": 1049 }, { "epoch": 1.370757180156658, "grad_norm": 4.067261219024658, "learning_rate": 9.703185200100273e-06, "loss": 0.2472, "step": 1050 }, { "epoch": 1.3720626631853787, "grad_norm": 3.8434484004974365, "learning_rate": 9.702467215117333e-06, "loss": 0.4436, "step": 1051 }, { "epoch": 1.3733681462140992, "grad_norm": 4.0190911293029785, "learning_rate": 9.701748389421151e-06, "loss": 0.2061, "step": 1052 }, { "epoch": 1.3746736292428199, "grad_norm": 3.063732147216797, "learning_rate": 9.70102872314024e-06, "loss": 0.185, "step": 1053 }, { "epoch": 1.3759791122715406, "grad_norm": 5.416615009307861, "learning_rate": 9.700308216403262e-06, "loss": 0.4761, "step": 1054 }, { "epoch": 1.377284595300261, "grad_norm": 5.243214130401611, "learning_rate": 9.699586869339033e-06, "loss": 0.3671, "step": 1055 }, { "epoch": 1.3785900783289817, "grad_norm": 4.307839393615723, "learning_rate": 9.698864682076512e-06, "loss": 0.3144, "step": 1056 }, { "epoch": 1.3798955613577024, "grad_norm": 5.059262275695801, "learning_rate": 9.698141654744816e-06, "loss": 0.3386, "step": 1057 }, { "epoch": 1.3812010443864229, "grad_norm": 3.0676486492156982, "learning_rate": 9.697417787473208e-06, "loss": 0.2124, "step": 1058 }, { "epoch": 1.3825065274151436, "grad_norm": 5.714021682739258, "learning_rate": 9.696693080391102e-06, "loss": 0.4805, "step": 1059 }, { "epoch": 1.3838120104438643, "grad_norm": 4.683905124664307, "learning_rate": 9.695967533628061e-06, "loss": 0.3008, "step": 1060 }, { "epoch": 1.3851174934725847, "grad_norm": 3.4699020385742188, "learning_rate": 9.6952411473138e-06, "loss": 0.2589, "step": 1061 }, { "epoch": 1.3864229765013054, "grad_norm": 4.840597152709961, "learning_rate": 9.694513921578183e-06, "loss": 0.3656, "step": 1062 }, { "epoch": 1.387728459530026, "grad_norm": 3.472649574279785, "learning_rate": 9.693785856551225e-06, "loss": 0.2119, "step": 1063 }, { "epoch": 1.3890339425587468, "grad_norm": 3.5497710704803467, "learning_rate": 9.693056952363092e-06, "loss": 0.3421, "step": 1064 }, { "epoch": 1.3903394255874675, "grad_norm": 3.2771267890930176, "learning_rate": 9.692327209144094e-06, "loss": 0.284, "step": 1065 }, { "epoch": 1.391644908616188, "grad_norm": 4.915031433105469, "learning_rate": 9.691596627024699e-06, "loss": 0.3293, "step": 1066 }, { "epoch": 1.3929503916449086, "grad_norm": 5.463450908660889, "learning_rate": 9.69086520613552e-06, "loss": 0.5757, "step": 1067 }, { "epoch": 1.3942558746736293, "grad_norm": 5.233564853668213, "learning_rate": 9.690132946607321e-06, "loss": 0.2312, "step": 1068 }, { "epoch": 1.3955613577023498, "grad_norm": 4.831157207489014, "learning_rate": 9.689399848571018e-06, "loss": 0.3917, "step": 1069 }, { "epoch": 1.3968668407310705, "grad_norm": 4.65952730178833, "learning_rate": 9.688665912157674e-06, "loss": 0.3861, "step": 1070 }, { "epoch": 1.3981723237597912, "grad_norm": 3.5118045806884766, "learning_rate": 9.687931137498505e-06, "loss": 0.3317, "step": 1071 }, { "epoch": 1.3994778067885116, "grad_norm": 3.8822686672210693, "learning_rate": 9.68719552472487e-06, "loss": 0.3688, "step": 1072 }, { "epoch": 1.4007832898172323, "grad_norm": 4.417736053466797, "learning_rate": 9.686459073968288e-06, "loss": 0.3719, "step": 1073 }, { "epoch": 1.402088772845953, "grad_norm": 4.856489181518555, "learning_rate": 9.685721785360423e-06, "loss": 0.408, "step": 1074 }, { "epoch": 1.4033942558746737, "grad_norm": 3.087930917739868, "learning_rate": 9.684983659033083e-06, "loss": 0.3524, "step": 1075 }, { "epoch": 1.4046997389033944, "grad_norm": 3.8722357749938965, "learning_rate": 9.684244695118235e-06, "loss": 0.4037, "step": 1076 }, { "epoch": 1.4060052219321149, "grad_norm": 3.7680118083953857, "learning_rate": 9.683504893747992e-06, "loss": 0.2809, "step": 1077 }, { "epoch": 1.4073107049608355, "grad_norm": 2.9700262546539307, "learning_rate": 9.682764255054617e-06, "loss": 0.1911, "step": 1078 }, { "epoch": 1.4086161879895562, "grad_norm": 4.1383867263793945, "learning_rate": 9.68202277917052e-06, "loss": 0.364, "step": 1079 }, { "epoch": 1.4099216710182767, "grad_norm": 4.199687480926514, "learning_rate": 9.681280466228264e-06, "loss": 0.3582, "step": 1080 }, { "epoch": 1.4112271540469974, "grad_norm": 4.927296161651611, "learning_rate": 9.680537316360563e-06, "loss": 0.4838, "step": 1081 }, { "epoch": 1.412532637075718, "grad_norm": 4.647497653961182, "learning_rate": 9.679793329700276e-06, "loss": 0.3537, "step": 1082 }, { "epoch": 1.4138381201044385, "grad_norm": 3.7429239749908447, "learning_rate": 9.679048506380412e-06, "loss": 0.2443, "step": 1083 }, { "epoch": 1.4151436031331592, "grad_norm": 4.329532623291016, "learning_rate": 9.678302846534137e-06, "loss": 0.2991, "step": 1084 }, { "epoch": 1.41644908616188, "grad_norm": 3.673182725906372, "learning_rate": 9.677556350294756e-06, "loss": 0.1539, "step": 1085 }, { "epoch": 1.4177545691906004, "grad_norm": 3.360703706741333, "learning_rate": 9.67680901779573e-06, "loss": 0.1317, "step": 1086 }, { "epoch": 1.419060052219321, "grad_norm": 4.299886703491211, "learning_rate": 9.67606084917067e-06, "loss": 0.3265, "step": 1087 }, { "epoch": 1.4203655352480418, "grad_norm": 3.9708545207977295, "learning_rate": 9.675311844553332e-06, "loss": 0.2443, "step": 1088 }, { "epoch": 1.4216710182767625, "grad_norm": 4.0286335945129395, "learning_rate": 9.674562004077625e-06, "loss": 0.3854, "step": 1089 }, { "epoch": 1.4229765013054831, "grad_norm": 6.1282501220703125, "learning_rate": 9.673811327877606e-06, "loss": 0.4373, "step": 1090 }, { "epoch": 1.4242819843342036, "grad_norm": 3.824521064758301, "learning_rate": 9.673059816087485e-06, "loss": 0.1459, "step": 1091 }, { "epoch": 1.4255874673629243, "grad_norm": 4.667178630828857, "learning_rate": 9.672307468841613e-06, "loss": 0.3205, "step": 1092 }, { "epoch": 1.426892950391645, "grad_norm": 4.687437057495117, "learning_rate": 9.6715542862745e-06, "loss": 0.3538, "step": 1093 }, { "epoch": 1.4281984334203655, "grad_norm": 3.01175594329834, "learning_rate": 9.670800268520799e-06, "loss": 0.2412, "step": 1094 }, { "epoch": 1.4295039164490861, "grad_norm": 4.177124500274658, "learning_rate": 9.670045415715314e-06, "loss": 0.2438, "step": 1095 }, { "epoch": 1.4308093994778068, "grad_norm": 4.493793964385986, "learning_rate": 9.669289727993e-06, "loss": 0.2351, "step": 1096 }, { "epoch": 1.4321148825065273, "grad_norm": 4.990242004394531, "learning_rate": 9.668533205488957e-06, "loss": 0.4469, "step": 1097 }, { "epoch": 1.433420365535248, "grad_norm": 4.369771480560303, "learning_rate": 9.667775848338441e-06, "loss": 0.3457, "step": 1098 }, { "epoch": 1.4347258485639687, "grad_norm": 3.1890196800231934, "learning_rate": 9.667017656676848e-06, "loss": 0.3291, "step": 1099 }, { "epoch": 1.4360313315926894, "grad_norm": 4.172609329223633, "learning_rate": 9.666258630639737e-06, "loss": 0.4639, "step": 1100 }, { "epoch": 1.43733681462141, "grad_norm": 3.873204469680786, "learning_rate": 9.6654987703628e-06, "loss": 0.35, "step": 1101 }, { "epoch": 1.4386422976501305, "grad_norm": 3.3318424224853516, "learning_rate": 9.664738075981888e-06, "loss": 0.2461, "step": 1102 }, { "epoch": 1.4399477806788512, "grad_norm": 3.3405373096466064, "learning_rate": 9.663976547633e-06, "loss": 0.2687, "step": 1103 }, { "epoch": 1.441253263707572, "grad_norm": 2.939822196960449, "learning_rate": 9.663214185452281e-06, "loss": 0.1645, "step": 1104 }, { "epoch": 1.4425587467362924, "grad_norm": 4.490384101867676, "learning_rate": 9.662450989576028e-06, "loss": 0.2878, "step": 1105 }, { "epoch": 1.443864229765013, "grad_norm": 4.268855094909668, "learning_rate": 9.66168696014069e-06, "loss": 0.3415, "step": 1106 }, { "epoch": 1.4451697127937337, "grad_norm": 4.13908576965332, "learning_rate": 9.660922097282855e-06, "loss": 0.3386, "step": 1107 }, { "epoch": 1.4464751958224542, "grad_norm": 4.860899448394775, "learning_rate": 9.660156401139268e-06, "loss": 0.3255, "step": 1108 }, { "epoch": 1.447780678851175, "grad_norm": 4.61310338973999, "learning_rate": 9.659389871846824e-06, "loss": 0.3432, "step": 1109 }, { "epoch": 1.4490861618798956, "grad_norm": 3.4329512119293213, "learning_rate": 9.658622509542564e-06, "loss": 0.1952, "step": 1110 }, { "epoch": 1.450391644908616, "grad_norm": 4.242414951324463, "learning_rate": 9.657854314363673e-06, "loss": 0.2864, "step": 1111 }, { "epoch": 1.4516971279373367, "grad_norm": 3.648285150527954, "learning_rate": 9.657085286447495e-06, "loss": 0.2271, "step": 1112 }, { "epoch": 1.4530026109660574, "grad_norm": 5.7316412925720215, "learning_rate": 9.656315425931516e-06, "loss": 0.4815, "step": 1113 }, { "epoch": 1.4543080939947781, "grad_norm": 4.452277660369873, "learning_rate": 9.65554473295337e-06, "loss": 0.3393, "step": 1114 }, { "epoch": 1.4556135770234988, "grad_norm": 3.084791898727417, "learning_rate": 9.654773207650848e-06, "loss": 0.2326, "step": 1115 }, { "epoch": 1.4569190600522193, "grad_norm": 4.422718048095703, "learning_rate": 9.65400085016188e-06, "loss": 0.352, "step": 1116 }, { "epoch": 1.45822454308094, "grad_norm": 3.5677566528320312, "learning_rate": 9.653227660624551e-06, "loss": 0.2516, "step": 1117 }, { "epoch": 1.4595300261096606, "grad_norm": 3.9391982555389404, "learning_rate": 9.652453639177092e-06, "loss": 0.2725, "step": 1118 }, { "epoch": 1.4608355091383811, "grad_norm": 4.369596004486084, "learning_rate": 9.651678785957885e-06, "loss": 0.3334, "step": 1119 }, { "epoch": 1.4621409921671018, "grad_norm": 4.064145565032959, "learning_rate": 9.650903101105458e-06, "loss": 0.2499, "step": 1120 }, { "epoch": 1.4634464751958225, "grad_norm": 4.827727317810059, "learning_rate": 9.650126584758488e-06, "loss": 0.3218, "step": 1121 }, { "epoch": 1.464751958224543, "grad_norm": 4.622961521148682, "learning_rate": 9.649349237055806e-06, "loss": 0.4542, "step": 1122 }, { "epoch": 1.4660574412532636, "grad_norm": 4.458795547485352, "learning_rate": 9.648571058136381e-06, "loss": 0.4119, "step": 1123 }, { "epoch": 1.4673629242819843, "grad_norm": 3.80401349067688, "learning_rate": 9.64779204813934e-06, "loss": 0.4519, "step": 1124 }, { "epoch": 1.468668407310705, "grad_norm": 4.090779781341553, "learning_rate": 9.647012207203955e-06, "loss": 0.4265, "step": 1125 }, { "epoch": 1.4699738903394257, "grad_norm": 4.481323719024658, "learning_rate": 9.646231535469647e-06, "loss": 0.3208, "step": 1126 }, { "epoch": 1.4712793733681462, "grad_norm": 4.624547958374023, "learning_rate": 9.645450033075989e-06, "loss": 0.289, "step": 1127 }, { "epoch": 1.4725848563968669, "grad_norm": 3.75881028175354, "learning_rate": 9.644667700162693e-06, "loss": 0.2239, "step": 1128 }, { "epoch": 1.4738903394255876, "grad_norm": 3.308011293411255, "learning_rate": 9.643884536869628e-06, "loss": 0.1955, "step": 1129 }, { "epoch": 1.475195822454308, "grad_norm": 5.003265857696533, "learning_rate": 9.643100543336812e-06, "loss": 0.2663, "step": 1130 }, { "epoch": 1.4765013054830287, "grad_norm": 3.9640092849731445, "learning_rate": 9.642315719704402e-06, "loss": 0.2215, "step": 1131 }, { "epoch": 1.4778067885117494, "grad_norm": 5.283466815948486, "learning_rate": 9.641530066112717e-06, "loss": 0.2995, "step": 1132 }, { "epoch": 1.4791122715404699, "grad_norm": 3.5616977214813232, "learning_rate": 9.640743582702213e-06, "loss": 0.2239, "step": 1133 }, { "epoch": 1.4804177545691906, "grad_norm": 4.375097751617432, "learning_rate": 9.639956269613498e-06, "loss": 0.2365, "step": 1134 }, { "epoch": 1.4817232375979112, "grad_norm": 6.089715957641602, "learning_rate": 9.63916812698733e-06, "loss": 0.5624, "step": 1135 }, { "epoch": 1.483028720626632, "grad_norm": 4.269568920135498, "learning_rate": 9.638379154964615e-06, "loss": 0.3453, "step": 1136 }, { "epoch": 1.4843342036553524, "grad_norm": 4.627082347869873, "learning_rate": 9.637589353686404e-06, "loss": 0.343, "step": 1137 }, { "epoch": 1.485639686684073, "grad_norm": 3.121539354324341, "learning_rate": 9.6367987232939e-06, "loss": 0.2657, "step": 1138 }, { "epoch": 1.4869451697127938, "grad_norm": 3.444214344024658, "learning_rate": 9.636007263928456e-06, "loss": 0.2685, "step": 1139 }, { "epoch": 1.4882506527415145, "grad_norm": 4.403367519378662, "learning_rate": 9.635214975731566e-06, "loss": 0.3274, "step": 1140 }, { "epoch": 1.489556135770235, "grad_norm": 4.295964241027832, "learning_rate": 9.634421858844877e-06, "loss": 0.2898, "step": 1141 }, { "epoch": 1.4908616187989556, "grad_norm": 4.609456539154053, "learning_rate": 9.633627913410187e-06, "loss": 0.3812, "step": 1142 }, { "epoch": 1.4921671018276763, "grad_norm": 4.630237579345703, "learning_rate": 9.632833139569433e-06, "loss": 0.3626, "step": 1143 }, { "epoch": 1.4934725848563968, "grad_norm": 5.529992580413818, "learning_rate": 9.63203753746471e-06, "loss": 0.4628, "step": 1144 }, { "epoch": 1.4947780678851175, "grad_norm": 5.06845235824585, "learning_rate": 9.631241107238255e-06, "loss": 0.3468, "step": 1145 }, { "epoch": 1.4960835509138382, "grad_norm": 3.0546071529388428, "learning_rate": 9.630443849032456e-06, "loss": 0.224, "step": 1146 }, { "epoch": 1.4973890339425586, "grad_norm": 4.287580966949463, "learning_rate": 9.629645762989845e-06, "loss": 0.2835, "step": 1147 }, { "epoch": 1.4986945169712793, "grad_norm": 4.146231174468994, "learning_rate": 9.628846849253108e-06, "loss": 0.2191, "step": 1148 }, { "epoch": 1.5, "grad_norm": 4.416562080383301, "learning_rate": 9.628047107965074e-06, "loss": 0.3143, "step": 1149 }, { "epoch": 1.5013054830287205, "grad_norm": 4.191816329956055, "learning_rate": 9.627246539268724e-06, "loss": 0.3096, "step": 1150 }, { "epoch": 1.5026109660574414, "grad_norm": 3.575880289077759, "learning_rate": 9.626445143307183e-06, "loss": 0.2688, "step": 1151 }, { "epoch": 1.5039164490861618, "grad_norm": 3.67958402633667, "learning_rate": 9.625642920223726e-06, "loss": 0.2903, "step": 1152 }, { "epoch": 1.5052219321148825, "grad_norm": 4.505551815032959, "learning_rate": 9.624839870161775e-06, "loss": 0.3005, "step": 1153 }, { "epoch": 1.5065274151436032, "grad_norm": 3.005070447921753, "learning_rate": 9.6240359932649e-06, "loss": 0.1422, "step": 1154 }, { "epoch": 1.5078328981723237, "grad_norm": 5.226949214935303, "learning_rate": 9.623231289676819e-06, "loss": 0.3499, "step": 1155 }, { "epoch": 1.5091383812010444, "grad_norm": 3.2405426502227783, "learning_rate": 9.622425759541401e-06, "loss": 0.173, "step": 1156 }, { "epoch": 1.510443864229765, "grad_norm": 3.4567465782165527, "learning_rate": 9.621619403002659e-06, "loss": 0.3078, "step": 1157 }, { "epoch": 1.5117493472584855, "grad_norm": 4.628517150878906, "learning_rate": 9.62081222020475e-06, "loss": 0.326, "step": 1158 }, { "epoch": 1.5130548302872062, "grad_norm": 3.1140100955963135, "learning_rate": 9.620004211291986e-06, "loss": 0.158, "step": 1159 }, { "epoch": 1.514360313315927, "grad_norm": 5.226457595825195, "learning_rate": 9.619195376408825e-06, "loss": 0.3766, "step": 1160 }, { "epoch": 1.5156657963446474, "grad_norm": 6.621033191680908, "learning_rate": 9.618385715699872e-06, "loss": 0.4452, "step": 1161 }, { "epoch": 1.5169712793733683, "grad_norm": 4.406672477722168, "learning_rate": 9.617575229309875e-06, "loss": 0.2753, "step": 1162 }, { "epoch": 1.5182767624020888, "grad_norm": 4.463873863220215, "learning_rate": 9.61676391738374e-06, "loss": 0.3085, "step": 1163 }, { "epoch": 1.5195822454308094, "grad_norm": 4.387753963470459, "learning_rate": 9.615951780066507e-06, "loss": 0.3513, "step": 1164 }, { "epoch": 1.5208877284595301, "grad_norm": 3.5790205001831055, "learning_rate": 9.615138817503379e-06, "loss": 0.2719, "step": 1165 }, { "epoch": 1.5221932114882506, "grad_norm": 2.674168109893799, "learning_rate": 9.614325029839694e-06, "loss": 0.1495, "step": 1166 }, { "epoch": 1.5234986945169713, "grad_norm": 2.8852460384368896, "learning_rate": 9.613510417220941e-06, "loss": 0.2015, "step": 1167 }, { "epoch": 1.524804177545692, "grad_norm": 4.173092365264893, "learning_rate": 9.61269497979276e-06, "loss": 0.2624, "step": 1168 }, { "epoch": 1.5261096605744124, "grad_norm": 3.323917865753174, "learning_rate": 9.611878717700935e-06, "loss": 0.1893, "step": 1169 }, { "epoch": 1.5274151436031331, "grad_norm": 3.4185328483581543, "learning_rate": 9.6110616310914e-06, "loss": 0.2316, "step": 1170 }, { "epoch": 1.5287206266318538, "grad_norm": 3.645498037338257, "learning_rate": 9.610243720110231e-06, "loss": 0.2105, "step": 1171 }, { "epoch": 1.5300261096605743, "grad_norm": 3.9646236896514893, "learning_rate": 9.609424984903659e-06, "loss": 0.2952, "step": 1172 }, { "epoch": 1.5313315926892952, "grad_norm": 3.744579315185547, "learning_rate": 9.608605425618056e-06, "loss": 0.2976, "step": 1173 }, { "epoch": 1.5326370757180157, "grad_norm": 3.0853848457336426, "learning_rate": 9.607785042399946e-06, "loss": 0.2081, "step": 1174 }, { "epoch": 1.5339425587467361, "grad_norm": 3.668217420578003, "learning_rate": 9.606963835395997e-06, "loss": 0.1776, "step": 1175 }, { "epoch": 1.535248041775457, "grad_norm": 3.7948992252349854, "learning_rate": 9.606141804753025e-06, "loss": 0.2545, "step": 1176 }, { "epoch": 1.5365535248041775, "grad_norm": 3.53332257270813, "learning_rate": 9.605318950617994e-06, "loss": 0.2639, "step": 1177 }, { "epoch": 1.5378590078328982, "grad_norm": 4.4300336837768555, "learning_rate": 9.604495273138016e-06, "loss": 0.3862, "step": 1178 }, { "epoch": 1.5391644908616189, "grad_norm": 4.893519401550293, "learning_rate": 9.603670772460347e-06, "loss": 0.2592, "step": 1179 }, { "epoch": 1.5404699738903394, "grad_norm": 4.972348213195801, "learning_rate": 9.602845448732396e-06, "loss": 0.3409, "step": 1180 }, { "epoch": 1.54177545691906, "grad_norm": 5.354752063751221, "learning_rate": 9.602019302101711e-06, "loss": 0.3472, "step": 1181 }, { "epoch": 1.5430809399477807, "grad_norm": 5.053654193878174, "learning_rate": 9.601192332715995e-06, "loss": 0.3156, "step": 1182 }, { "epoch": 1.5443864229765012, "grad_norm": 4.161074638366699, "learning_rate": 9.600364540723093e-06, "loss": 0.2769, "step": 1183 }, { "epoch": 1.5456919060052219, "grad_norm": 4.066928863525391, "learning_rate": 9.599535926271e-06, "loss": 0.1716, "step": 1184 }, { "epoch": 1.5469973890339426, "grad_norm": 4.417094707489014, "learning_rate": 9.598706489507855e-06, "loss": 0.2598, "step": 1185 }, { "epoch": 1.548302872062663, "grad_norm": 4.84038782119751, "learning_rate": 9.597876230581946e-06, "loss": 0.2821, "step": 1186 }, { "epoch": 1.549608355091384, "grad_norm": 3.7599995136260986, "learning_rate": 9.597045149641709e-06, "loss": 0.2803, "step": 1187 }, { "epoch": 1.5509138381201044, "grad_norm": 4.110788822174072, "learning_rate": 9.596213246835726e-06, "loss": 0.3433, "step": 1188 }, { "epoch": 1.552219321148825, "grad_norm": 4.171308994293213, "learning_rate": 9.595380522312724e-06, "loss": 0.2665, "step": 1189 }, { "epoch": 1.5535248041775458, "grad_norm": 3.3144521713256836, "learning_rate": 9.59454697622158e-06, "loss": 0.1668, "step": 1190 }, { "epoch": 1.5548302872062663, "grad_norm": 3.881612539291382, "learning_rate": 9.593712608711316e-06, "loss": 0.2905, "step": 1191 }, { "epoch": 1.556135770234987, "grad_norm": 4.195339202880859, "learning_rate": 9.592877419931101e-06, "loss": 0.2764, "step": 1192 }, { "epoch": 1.5574412532637076, "grad_norm": 4.155247211456299, "learning_rate": 9.59204141003025e-06, "loss": 0.2364, "step": 1193 }, { "epoch": 1.558746736292428, "grad_norm": 4.228659152984619, "learning_rate": 9.59120457915823e-06, "loss": 0.326, "step": 1194 }, { "epoch": 1.5600522193211488, "grad_norm": 3.951005697250366, "learning_rate": 9.590366927464647e-06, "loss": 0.2586, "step": 1195 }, { "epoch": 1.5613577023498695, "grad_norm": 4.5771403312683105, "learning_rate": 9.589528455099258e-06, "loss": 0.2881, "step": 1196 }, { "epoch": 1.56266318537859, "grad_norm": 3.9963488578796387, "learning_rate": 9.588689162211968e-06, "loss": 0.2482, "step": 1197 }, { "epoch": 1.5639686684073109, "grad_norm": 3.55670166015625, "learning_rate": 9.587849048952825e-06, "loss": 0.1986, "step": 1198 }, { "epoch": 1.5652741514360313, "grad_norm": 4.272223949432373, "learning_rate": 9.587008115472024e-06, "loss": 0.2734, "step": 1199 }, { "epoch": 1.566579634464752, "grad_norm": 4.600804805755615, "learning_rate": 9.586166361919913e-06, "loss": 0.2579, "step": 1200 }, { "epoch": 1.5678851174934727, "grad_norm": 4.964235782623291, "learning_rate": 9.585323788446977e-06, "loss": 0.3662, "step": 1201 }, { "epoch": 1.5691906005221932, "grad_norm": 4.237382888793945, "learning_rate": 9.584480395203855e-06, "loss": 0.3212, "step": 1202 }, { "epoch": 1.5704960835509139, "grad_norm": 3.887195587158203, "learning_rate": 9.58363618234133e-06, "loss": 0.2512, "step": 1203 }, { "epoch": 1.5718015665796345, "grad_norm": 4.686578273773193, "learning_rate": 9.58279115001033e-06, "loss": 0.2978, "step": 1204 }, { "epoch": 1.573107049608355, "grad_norm": 4.532405853271484, "learning_rate": 9.581945298361935e-06, "loss": 0.2599, "step": 1205 }, { "epoch": 1.5744125326370757, "grad_norm": 5.355201244354248, "learning_rate": 9.581098627547361e-06, "loss": 0.2682, "step": 1206 }, { "epoch": 1.5757180156657964, "grad_norm": 5.071072578430176, "learning_rate": 9.580251137717982e-06, "loss": 0.3486, "step": 1207 }, { "epoch": 1.5770234986945169, "grad_norm": 3.684063673019409, "learning_rate": 9.579402829025312e-06, "loss": 0.2235, "step": 1208 }, { "epoch": 1.5783289817232375, "grad_norm": 4.564535140991211, "learning_rate": 9.578553701621012e-06, "loss": 0.3439, "step": 1209 }, { "epoch": 1.5796344647519582, "grad_norm": 3.1252686977386475, "learning_rate": 9.57770375565689e-06, "loss": 0.1738, "step": 1210 }, { "epoch": 1.5809399477806787, "grad_norm": 5.920300006866455, "learning_rate": 9.5768529912849e-06, "loss": 0.4102, "step": 1211 }, { "epoch": 1.5822454308093996, "grad_norm": 3.921830654144287, "learning_rate": 9.576001408657145e-06, "loss": 0.2056, "step": 1212 }, { "epoch": 1.58355091383812, "grad_norm": 4.719174861907959, "learning_rate": 9.57514900792587e-06, "loss": 0.204, "step": 1213 }, { "epoch": 1.5848563968668408, "grad_norm": 4.428525447845459, "learning_rate": 9.57429578924347e-06, "loss": 0.3124, "step": 1214 }, { "epoch": 1.5861618798955615, "grad_norm": 4.3325886726379395, "learning_rate": 9.573441752762482e-06, "loss": 0.2916, "step": 1215 }, { "epoch": 1.587467362924282, "grad_norm": 3.372246742248535, "learning_rate": 9.572586898635596e-06, "loss": 0.2319, "step": 1216 }, { "epoch": 1.5887728459530026, "grad_norm": 4.874876499176025, "learning_rate": 9.57173122701564e-06, "loss": 0.3189, "step": 1217 }, { "epoch": 1.5900783289817233, "grad_norm": 3.8998963832855225, "learning_rate": 9.570874738055591e-06, "loss": 0.2687, "step": 1218 }, { "epoch": 1.5913838120104438, "grad_norm": 3.8870038986206055, "learning_rate": 9.57001743190858e-06, "loss": 0.2934, "step": 1219 }, { "epoch": 1.5926892950391645, "grad_norm": 5.025107383728027, "learning_rate": 9.56915930872787e-06, "loss": 0.3783, "step": 1220 }, { "epoch": 1.5939947780678851, "grad_norm": 3.221954584121704, "learning_rate": 9.56830036866688e-06, "loss": 0.132, "step": 1221 }, { "epoch": 1.5953002610966056, "grad_norm": 5.265909671783447, "learning_rate": 9.567440611879175e-06, "loss": 0.2677, "step": 1222 }, { "epoch": 1.5966057441253265, "grad_norm": 4.44109582901001, "learning_rate": 9.56658003851846e-06, "loss": 0.313, "step": 1223 }, { "epoch": 1.597911227154047, "grad_norm": 3.4314382076263428, "learning_rate": 9.56571864873859e-06, "loss": 0.2025, "step": 1224 }, { "epoch": 1.5992167101827677, "grad_norm": 4.589463710784912, "learning_rate": 9.564856442693569e-06, "loss": 0.3431, "step": 1225 }, { "epoch": 1.6005221932114884, "grad_norm": 4.6382575035095215, "learning_rate": 9.563993420537538e-06, "loss": 0.3177, "step": 1226 }, { "epoch": 1.6018276762402088, "grad_norm": 3.0748450756073, "learning_rate": 9.56312958242479e-06, "loss": 0.1512, "step": 1227 }, { "epoch": 1.6031331592689295, "grad_norm": 4.324173927307129, "learning_rate": 9.562264928509768e-06, "loss": 0.2774, "step": 1228 }, { "epoch": 1.6044386422976502, "grad_norm": 4.370386123657227, "learning_rate": 9.56139945894705e-06, "loss": 0.2705, "step": 1229 }, { "epoch": 1.6057441253263707, "grad_norm": 3.9873743057250977, "learning_rate": 9.560533173891371e-06, "loss": 0.2265, "step": 1230 }, { "epoch": 1.6070496083550914, "grad_norm": 3.1528074741363525, "learning_rate": 9.559666073497602e-06, "loss": 0.1278, "step": 1231 }, { "epoch": 1.608355091383812, "grad_norm": 4.268191337585449, "learning_rate": 9.558798157920767e-06, "loss": 0.2218, "step": 1232 }, { "epoch": 1.6096605744125325, "grad_norm": 4.865320205688477, "learning_rate": 9.557929427316033e-06, "loss": 0.3154, "step": 1233 }, { "epoch": 1.6109660574412534, "grad_norm": 4.674643039703369, "learning_rate": 9.55705988183871e-06, "loss": 0.2868, "step": 1234 }, { "epoch": 1.612271540469974, "grad_norm": 5.451572895050049, "learning_rate": 9.556189521644261e-06, "loss": 0.3089, "step": 1235 }, { "epoch": 1.6135770234986944, "grad_norm": 4.441591262817383, "learning_rate": 9.555318346888287e-06, "loss": 0.2853, "step": 1236 }, { "epoch": 1.6148825065274153, "grad_norm": 3.651561975479126, "learning_rate": 9.554446357726539e-06, "loss": 0.1787, "step": 1237 }, { "epoch": 1.6161879895561357, "grad_norm": 4.393618583679199, "learning_rate": 9.553573554314912e-06, "loss": 0.2761, "step": 1238 }, { "epoch": 1.6174934725848564, "grad_norm": 2.7502288818359375, "learning_rate": 9.552699936809446e-06, "loss": 0.1348, "step": 1239 }, { "epoch": 1.6187989556135771, "grad_norm": 3.478529691696167, "learning_rate": 9.551825505366328e-06, "loss": 0.2479, "step": 1240 }, { "epoch": 1.6201044386422976, "grad_norm": 5.010476589202881, "learning_rate": 9.550950260141893e-06, "loss": 0.3791, "step": 1241 }, { "epoch": 1.6214099216710183, "grad_norm": 3.138357162475586, "learning_rate": 9.550074201292614e-06, "loss": 0.2197, "step": 1242 }, { "epoch": 1.622715404699739, "grad_norm": 3.1854465007781982, "learning_rate": 9.549197328975115e-06, "loss": 0.2295, "step": 1243 }, { "epoch": 1.6240208877284594, "grad_norm": 3.5180773735046387, "learning_rate": 9.548319643346167e-06, "loss": 0.2733, "step": 1244 }, { "epoch": 1.6253263707571801, "grad_norm": 3.6095945835113525, "learning_rate": 9.547441144562681e-06, "loss": 0.2804, "step": 1245 }, { "epoch": 1.6266318537859008, "grad_norm": 3.292659044265747, "learning_rate": 9.546561832781717e-06, "loss": 0.1653, "step": 1246 }, { "epoch": 1.6279373368146213, "grad_norm": 3.727997064590454, "learning_rate": 9.54568170816048e-06, "loss": 0.2175, "step": 1247 }, { "epoch": 1.6292428198433422, "grad_norm": 4.270644664764404, "learning_rate": 9.544800770856319e-06, "loss": 0.279, "step": 1248 }, { "epoch": 1.6305483028720626, "grad_norm": 3.9997293949127197, "learning_rate": 9.543919021026731e-06, "loss": 0.2587, "step": 1249 }, { "epoch": 1.6318537859007833, "grad_norm": 4.69527006149292, "learning_rate": 9.543036458829354e-06, "loss": 0.2929, "step": 1250 }, { "epoch": 1.633159268929504, "grad_norm": 3.2573726177215576, "learning_rate": 9.542153084421974e-06, "loss": 0.2804, "step": 1251 }, { "epoch": 1.6344647519582245, "grad_norm": 4.35969352722168, "learning_rate": 9.541268897962523e-06, "loss": 0.3463, "step": 1252 }, { "epoch": 1.6357702349869452, "grad_norm": 3.746411085128784, "learning_rate": 9.540383899609078e-06, "loss": 0.266, "step": 1253 }, { "epoch": 1.6370757180156659, "grad_norm": 3.366956949234009, "learning_rate": 9.539498089519856e-06, "loss": 0.2797, "step": 1254 }, { "epoch": 1.6383812010443863, "grad_norm": 4.452497959136963, "learning_rate": 9.538611467853228e-06, "loss": 0.353, "step": 1255 }, { "epoch": 1.639686684073107, "grad_norm": 3.154116153717041, "learning_rate": 9.537724034767703e-06, "loss": 0.2432, "step": 1256 }, { "epoch": 1.6409921671018277, "grad_norm": 5.015351295471191, "learning_rate": 9.536835790421936e-06, "loss": 0.4505, "step": 1257 }, { "epoch": 1.6422976501305482, "grad_norm": 4.6641011238098145, "learning_rate": 9.535946734974732e-06, "loss": 0.396, "step": 1258 }, { "epoch": 1.643603133159269, "grad_norm": 3.3629114627838135, "learning_rate": 9.535056868585036e-06, "loss": 0.1666, "step": 1259 }, { "epoch": 1.6449086161879896, "grad_norm": 3.3876953125, "learning_rate": 9.534166191411938e-06, "loss": 0.2834, "step": 1260 }, { "epoch": 1.64621409921671, "grad_norm": 2.8693339824676514, "learning_rate": 9.533274703614675e-06, "loss": 0.1782, "step": 1261 }, { "epoch": 1.647519582245431, "grad_norm": 3.6841180324554443, "learning_rate": 9.53238240535263e-06, "loss": 0.2535, "step": 1262 }, { "epoch": 1.6488250652741514, "grad_norm": 3.883976459503174, "learning_rate": 9.531489296785329e-06, "loss": 0.2048, "step": 1263 }, { "epoch": 1.650130548302872, "grad_norm": 4.535897254943848, "learning_rate": 9.53059537807244e-06, "loss": 0.3223, "step": 1264 }, { "epoch": 1.6514360313315928, "grad_norm": 3.4636523723602295, "learning_rate": 9.529700649373783e-06, "loss": 0.2076, "step": 1265 }, { "epoch": 1.6527415143603132, "grad_norm": 5.304198741912842, "learning_rate": 9.528805110849318e-06, "loss": 0.3212, "step": 1266 }, { "epoch": 1.654046997389034, "grad_norm": 4.847413539886475, "learning_rate": 9.52790876265915e-06, "loss": 0.3153, "step": 1267 }, { "epoch": 1.6553524804177546, "grad_norm": 5.623435974121094, "learning_rate": 9.527011604963527e-06, "loss": 0.4347, "step": 1268 }, { "epoch": 1.656657963446475, "grad_norm": 3.57401704788208, "learning_rate": 9.526113637922846e-06, "loss": 0.1902, "step": 1269 }, { "epoch": 1.6579634464751958, "grad_norm": 3.2018346786499023, "learning_rate": 9.525214861697647e-06, "loss": 0.162, "step": 1270 }, { "epoch": 1.6592689295039165, "grad_norm": 4.231534957885742, "learning_rate": 9.524315276448615e-06, "loss": 0.2261, "step": 1271 }, { "epoch": 1.660574412532637, "grad_norm": 3.97792649269104, "learning_rate": 9.523414882336577e-06, "loss": 0.265, "step": 1272 }, { "epoch": 1.6618798955613578, "grad_norm": 4.346375465393066, "learning_rate": 9.522513679522508e-06, "loss": 0.2625, "step": 1273 }, { "epoch": 1.6631853785900783, "grad_norm": 4.492981433868408, "learning_rate": 9.521611668167526e-06, "loss": 0.2285, "step": 1274 }, { "epoch": 1.664490861618799, "grad_norm": 4.368302345275879, "learning_rate": 9.520708848432893e-06, "loss": 0.3757, "step": 1275 }, { "epoch": 1.6657963446475197, "grad_norm": 3.8111133575439453, "learning_rate": 9.519805220480018e-06, "loss": 0.3337, "step": 1276 }, { "epoch": 1.6671018276762402, "grad_norm": 4.048853397369385, "learning_rate": 9.518900784470453e-06, "loss": 0.2183, "step": 1277 }, { "epoch": 1.6684073107049608, "grad_norm": 4.813344955444336, "learning_rate": 9.517995540565892e-06, "loss": 0.3809, "step": 1278 }, { "epoch": 1.6697127937336815, "grad_norm": 4.256529331207275, "learning_rate": 9.517089488928175e-06, "loss": 0.2685, "step": 1279 }, { "epoch": 1.671018276762402, "grad_norm": 4.119245529174805, "learning_rate": 9.51618262971929e-06, "loss": 0.3564, "step": 1280 }, { "epoch": 1.6723237597911227, "grad_norm": 5.038973331451416, "learning_rate": 9.515274963101366e-06, "loss": 0.338, "step": 1281 }, { "epoch": 1.6736292428198434, "grad_norm": 4.034912586212158, "learning_rate": 9.514366489236678e-06, "loss": 0.2549, "step": 1282 }, { "epoch": 1.6749347258485638, "grad_norm": 3.629831552505493, "learning_rate": 9.513457208287638e-06, "loss": 0.217, "step": 1283 }, { "epoch": 1.6762402088772848, "grad_norm": 4.17323637008667, "learning_rate": 9.512547120416815e-06, "loss": 0.2444, "step": 1284 }, { "epoch": 1.6775456919060052, "grad_norm": 4.2057905197143555, "learning_rate": 9.511636225786914e-06, "loss": 0.2907, "step": 1285 }, { "epoch": 1.6788511749347257, "grad_norm": 2.633713483810425, "learning_rate": 9.510724524560783e-06, "loss": 0.1277, "step": 1286 }, { "epoch": 1.6801566579634466, "grad_norm": 4.377604007720947, "learning_rate": 9.509812016901422e-06, "loss": 0.3411, "step": 1287 }, { "epoch": 1.681462140992167, "grad_norm": 5.383780002593994, "learning_rate": 9.508898702971966e-06, "loss": 0.3661, "step": 1288 }, { "epoch": 1.6827676240208878, "grad_norm": 2.669163703918457, "learning_rate": 9.507984582935703e-06, "loss": 0.1587, "step": 1289 }, { "epoch": 1.6840731070496084, "grad_norm": 4.691281795501709, "learning_rate": 9.507069656956056e-06, "loss": 0.2945, "step": 1290 }, { "epoch": 1.685378590078329, "grad_norm": 4.319573879241943, "learning_rate": 9.506153925196599e-06, "loss": 0.2883, "step": 1291 }, { "epoch": 1.6866840731070496, "grad_norm": 6.238972187042236, "learning_rate": 9.505237387821046e-06, "loss": 0.3244, "step": 1292 }, { "epoch": 1.6879895561357703, "grad_norm": 5.169991493225098, "learning_rate": 9.50432004499326e-06, "loss": 0.3253, "step": 1293 }, { "epoch": 1.6892950391644908, "grad_norm": 4.070102691650391, "learning_rate": 9.50340189687724e-06, "loss": 0.2761, "step": 1294 }, { "epoch": 1.6906005221932114, "grad_norm": 4.100717544555664, "learning_rate": 9.50248294363714e-06, "loss": 0.2608, "step": 1295 }, { "epoch": 1.6919060052219321, "grad_norm": 3.4220104217529297, "learning_rate": 9.501563185437246e-06, "loss": 0.18, "step": 1296 }, { "epoch": 1.6932114882506526, "grad_norm": 4.16070032119751, "learning_rate": 9.500642622441995e-06, "loss": 0.245, "step": 1297 }, { "epoch": 1.6945169712793735, "grad_norm": 5.069825172424316, "learning_rate": 9.499721254815968e-06, "loss": 0.3487, "step": 1298 }, { "epoch": 1.695822454308094, "grad_norm": 3.1426022052764893, "learning_rate": 9.498799082723887e-06, "loss": 0.1222, "step": 1299 }, { "epoch": 1.6971279373368147, "grad_norm": 4.454225540161133, "learning_rate": 9.49787610633062e-06, "loss": 0.2852, "step": 1300 }, { "epoch": 1.6984334203655354, "grad_norm": 4.89827299118042, "learning_rate": 9.496952325801178e-06, "loss": 0.3553, "step": 1301 }, { "epoch": 1.6997389033942558, "grad_norm": 3.7856082916259766, "learning_rate": 9.496027741300715e-06, "loss": 0.2898, "step": 1302 }, { "epoch": 1.7010443864229765, "grad_norm": 3.2490365505218506, "learning_rate": 9.49510235299453e-06, "loss": 0.1754, "step": 1303 }, { "epoch": 1.7023498694516972, "grad_norm": 5.095042705535889, "learning_rate": 9.494176161048064e-06, "loss": 0.2634, "step": 1304 }, { "epoch": 1.7036553524804177, "grad_norm": 3.54014253616333, "learning_rate": 9.493249165626903e-06, "loss": 0.2229, "step": 1305 }, { "epoch": 1.7049608355091384, "grad_norm": 4.070872783660889, "learning_rate": 9.492321366896778e-06, "loss": 0.284, "step": 1306 }, { "epoch": 1.706266318537859, "grad_norm": 3.8146870136260986, "learning_rate": 9.491392765023561e-06, "loss": 0.2616, "step": 1307 }, { "epoch": 1.7075718015665795, "grad_norm": 3.8669846057891846, "learning_rate": 9.49046336017327e-06, "loss": 0.2611, "step": 1308 }, { "epoch": 1.7088772845953004, "grad_norm": 3.0674612522125244, "learning_rate": 9.489533152512063e-06, "loss": 0.1886, "step": 1309 }, { "epoch": 1.7101827676240209, "grad_norm": 3.918076753616333, "learning_rate": 9.488602142206244e-06, "loss": 0.2262, "step": 1310 }, { "epoch": 1.7114882506527413, "grad_norm": 3.8691439628601074, "learning_rate": 9.487670329422262e-06, "loss": 0.2274, "step": 1311 }, { "epoch": 1.7127937336814623, "grad_norm": 3.6170654296875, "learning_rate": 9.486737714326706e-06, "loss": 0.2317, "step": 1312 }, { "epoch": 1.7140992167101827, "grad_norm": 3.4947280883789062, "learning_rate": 9.48580429708631e-06, "loss": 0.2099, "step": 1313 }, { "epoch": 1.7154046997389034, "grad_norm": 4.313266277313232, "learning_rate": 9.484870077867953e-06, "loss": 0.1694, "step": 1314 }, { "epoch": 1.716710182767624, "grad_norm": 4.592824935913086, "learning_rate": 9.483935056838657e-06, "loss": 0.3394, "step": 1315 }, { "epoch": 1.7180156657963446, "grad_norm": 3.2460267543792725, "learning_rate": 9.482999234165583e-06, "loss": 0.1479, "step": 1316 }, { "epoch": 1.7193211488250653, "grad_norm": 3.3629257678985596, "learning_rate": 9.48206261001604e-06, "loss": 0.1845, "step": 1317 }, { "epoch": 1.720626631853786, "grad_norm": 3.889153242111206, "learning_rate": 9.481125184557481e-06, "loss": 0.1852, "step": 1318 }, { "epoch": 1.7219321148825064, "grad_norm": 3.0594980716705322, "learning_rate": 9.480186957957495e-06, "loss": 0.1352, "step": 1319 }, { "epoch": 1.723237597911227, "grad_norm": 3.958376884460449, "learning_rate": 9.479247930383826e-06, "loss": 0.1648, "step": 1320 }, { "epoch": 1.7245430809399478, "grad_norm": 2.8625872135162354, "learning_rate": 9.47830810200435e-06, "loss": 0.1182, "step": 1321 }, { "epoch": 1.7258485639686683, "grad_norm": 3.7711873054504395, "learning_rate": 9.47736747298709e-06, "loss": 0.2273, "step": 1322 }, { "epoch": 1.7271540469973892, "grad_norm": 3.113983392715454, "learning_rate": 9.476426043500218e-06, "loss": 0.2145, "step": 1323 }, { "epoch": 1.7284595300261096, "grad_norm": 4.320217609405518, "learning_rate": 9.47548381371204e-06, "loss": 0.2731, "step": 1324 }, { "epoch": 1.7297650130548303, "grad_norm": 4.664668083190918, "learning_rate": 9.474540783791008e-06, "loss": 0.3079, "step": 1325 }, { "epoch": 1.731070496083551, "grad_norm": 3.259098768234253, "learning_rate": 9.473596953905721e-06, "loss": 0.232, "step": 1326 }, { "epoch": 1.7323759791122715, "grad_norm": 4.41044282913208, "learning_rate": 9.472652324224917e-06, "loss": 0.2893, "step": 1327 }, { "epoch": 1.7336814621409922, "grad_norm": 4.316563606262207, "learning_rate": 9.471706894917479e-06, "loss": 0.2428, "step": 1328 }, { "epoch": 1.7349869451697129, "grad_norm": 5.342155456542969, "learning_rate": 9.47076066615243e-06, "loss": 0.3294, "step": 1329 }, { "epoch": 1.7362924281984333, "grad_norm": 3.3688833713531494, "learning_rate": 9.469813638098941e-06, "loss": 0.1579, "step": 1330 }, { "epoch": 1.737597911227154, "grad_norm": 3.1883058547973633, "learning_rate": 9.46886581092632e-06, "loss": 0.147, "step": 1331 }, { "epoch": 1.7389033942558747, "grad_norm": 3.3131465911865234, "learning_rate": 9.467917184804023e-06, "loss": 0.1536, "step": 1332 }, { "epoch": 1.7402088772845952, "grad_norm": 3.8605263233184814, "learning_rate": 9.466967759901647e-06, "loss": 0.1593, "step": 1333 }, { "epoch": 1.741514360313316, "grad_norm": 3.3263916969299316, "learning_rate": 9.466017536388927e-06, "loss": 0.189, "step": 1334 }, { "epoch": 1.7428198433420365, "grad_norm": 5.258123874664307, "learning_rate": 9.46506651443575e-06, "loss": 0.3098, "step": 1335 }, { "epoch": 1.7441253263707572, "grad_norm": 5.415931701660156, "learning_rate": 9.46411469421214e-06, "loss": 0.3791, "step": 1336 }, { "epoch": 1.745430809399478, "grad_norm": 4.405088424682617, "learning_rate": 9.463162075888264e-06, "loss": 0.2719, "step": 1337 }, { "epoch": 1.7467362924281984, "grad_norm": 3.227496862411499, "learning_rate": 9.46220865963443e-06, "loss": 0.1337, "step": 1338 }, { "epoch": 1.748041775456919, "grad_norm": 3.825601577758789, "learning_rate": 9.461254445621097e-06, "loss": 0.2579, "step": 1339 }, { "epoch": 1.7493472584856398, "grad_norm": 4.961741924285889, "learning_rate": 9.460299434018854e-06, "loss": 0.2744, "step": 1340 }, { "epoch": 1.7506527415143602, "grad_norm": 7.841126441955566, "learning_rate": 9.459343624998444e-06, "loss": 0.4832, "step": 1341 }, { "epoch": 1.751958224543081, "grad_norm": 4.363831996917725, "learning_rate": 9.458387018730745e-06, "loss": 0.2906, "step": 1342 }, { "epoch": 1.7532637075718016, "grad_norm": 4.5446367263793945, "learning_rate": 9.457429615386781e-06, "loss": 0.3014, "step": 1343 }, { "epoch": 1.754569190600522, "grad_norm": 4.378509998321533, "learning_rate": 9.456471415137718e-06, "loss": 0.2059, "step": 1344 }, { "epoch": 1.7558746736292428, "grad_norm": 4.584950923919678, "learning_rate": 9.455512418154866e-06, "loss": 0.3414, "step": 1345 }, { "epoch": 1.7571801566579635, "grad_norm": 4.655567169189453, "learning_rate": 9.454552624609672e-06, "loss": 0.3374, "step": 1346 }, { "epoch": 1.758485639686684, "grad_norm": 3.564784049987793, "learning_rate": 9.453592034673733e-06, "loss": 0.2292, "step": 1347 }, { "epoch": 1.7597911227154048, "grad_norm": 5.451045036315918, "learning_rate": 9.45263064851878e-06, "loss": 0.3633, "step": 1348 }, { "epoch": 1.7610966057441253, "grad_norm": 4.675060749053955, "learning_rate": 9.451668466316697e-06, "loss": 0.2968, "step": 1349 }, { "epoch": 1.762402088772846, "grad_norm": 4.276337146759033, "learning_rate": 9.4507054882395e-06, "loss": 0.2474, "step": 1350 }, { "epoch": 1.7637075718015667, "grad_norm": 4.238956928253174, "learning_rate": 9.449741714459353e-06, "loss": 0.3053, "step": 1351 }, { "epoch": 1.7650130548302871, "grad_norm": 3.4331254959106445, "learning_rate": 9.448777145148557e-06, "loss": 0.2274, "step": 1352 }, { "epoch": 1.7663185378590078, "grad_norm": 3.1939144134521484, "learning_rate": 9.447811780479566e-06, "loss": 0.1605, "step": 1353 }, { "epoch": 1.7676240208877285, "grad_norm": 3.255837917327881, "learning_rate": 9.446845620624963e-06, "loss": 0.2483, "step": 1354 }, { "epoch": 1.768929503916449, "grad_norm": 4.116135597229004, "learning_rate": 9.445878665757482e-06, "loss": 0.2157, "step": 1355 }, { "epoch": 1.7702349869451697, "grad_norm": 4.317563533782959, "learning_rate": 9.444910916049994e-06, "loss": 0.2214, "step": 1356 }, { "epoch": 1.7715404699738904, "grad_norm": 3.095097303390503, "learning_rate": 9.443942371675519e-06, "loss": 0.1378, "step": 1357 }, { "epoch": 1.7728459530026108, "grad_norm": 4.4663238525390625, "learning_rate": 9.442973032807213e-06, "loss": 0.2669, "step": 1358 }, { "epoch": 1.7741514360313317, "grad_norm": 3.2665798664093018, "learning_rate": 9.442002899618371e-06, "loss": 0.1476, "step": 1359 }, { "epoch": 1.7754569190600522, "grad_norm": 3.9021594524383545, "learning_rate": 9.441031972282442e-06, "loss": 0.2261, "step": 1360 }, { "epoch": 1.776762402088773, "grad_norm": 4.162119388580322, "learning_rate": 9.440060250973006e-06, "loss": 0.3141, "step": 1361 }, { "epoch": 1.7780678851174936, "grad_norm": 4.4652910232543945, "learning_rate": 9.439087735863788e-06, "loss": 0.3101, "step": 1362 }, { "epoch": 1.779373368146214, "grad_norm": 4.6768903732299805, "learning_rate": 9.438114427128654e-06, "loss": 0.3087, "step": 1363 }, { "epoch": 1.7806788511749347, "grad_norm": 4.292434215545654, "learning_rate": 9.43714032494162e-06, "loss": 0.2368, "step": 1364 }, { "epoch": 1.7819843342036554, "grad_norm": 3.703267812728882, "learning_rate": 9.436165429476832e-06, "loss": 0.2631, "step": 1365 }, { "epoch": 1.783289817232376, "grad_norm": 3.000286340713501, "learning_rate": 9.435189740908581e-06, "loss": 0.1788, "step": 1366 }, { "epoch": 1.7845953002610966, "grad_norm": 3.597381591796875, "learning_rate": 9.434213259411308e-06, "loss": 0.206, "step": 1367 }, { "epoch": 1.7859007832898173, "grad_norm": 3.9636778831481934, "learning_rate": 9.433235985159586e-06, "loss": 0.1851, "step": 1368 }, { "epoch": 1.7872062663185377, "grad_norm": 4.220325946807861, "learning_rate": 9.432257918328134e-06, "loss": 0.3468, "step": 1369 }, { "epoch": 1.7885117493472587, "grad_norm": 4.042929649353027, "learning_rate": 9.431279059091812e-06, "loss": 0.1647, "step": 1370 }, { "epoch": 1.7898172323759791, "grad_norm": 4.046924591064453, "learning_rate": 9.430299407625621e-06, "loss": 0.2649, "step": 1371 }, { "epoch": 1.7911227154046996, "grad_norm": 4.241121292114258, "learning_rate": 9.429318964104704e-06, "loss": 0.2159, "step": 1372 }, { "epoch": 1.7924281984334205, "grad_norm": 3.9573957920074463, "learning_rate": 9.428337728704349e-06, "loss": 0.2585, "step": 1373 }, { "epoch": 1.793733681462141, "grad_norm": 3.3471460342407227, "learning_rate": 9.427355701599981e-06, "loss": 0.1505, "step": 1374 }, { "epoch": 1.7950391644908616, "grad_norm": 2.987391471862793, "learning_rate": 9.426372882967167e-06, "loss": 0.1752, "step": 1375 }, { "epoch": 1.7963446475195823, "grad_norm": 4.032235145568848, "learning_rate": 9.425389272981618e-06, "loss": 0.2039, "step": 1376 }, { "epoch": 1.7976501305483028, "grad_norm": 4.948434352874756, "learning_rate": 9.424404871819185e-06, "loss": 0.3552, "step": 1377 }, { "epoch": 1.7989556135770235, "grad_norm": 3.5806620121002197, "learning_rate": 9.423419679655856e-06, "loss": 0.2322, "step": 1378 }, { "epoch": 1.8002610966057442, "grad_norm": 4.46107292175293, "learning_rate": 9.422433696667772e-06, "loss": 0.2972, "step": 1379 }, { "epoch": 1.8015665796344646, "grad_norm": 4.245721340179443, "learning_rate": 9.421446923031205e-06, "loss": 0.2631, "step": 1380 }, { "epoch": 1.8028720626631853, "grad_norm": 3.4600143432617188, "learning_rate": 9.420459358922573e-06, "loss": 0.1926, "step": 1381 }, { "epoch": 1.804177545691906, "grad_norm": 3.2691609859466553, "learning_rate": 9.419471004518432e-06, "loss": 0.1399, "step": 1382 }, { "epoch": 1.8054830287206265, "grad_norm": 5.2512898445129395, "learning_rate": 9.418481859995481e-06, "loss": 0.1731, "step": 1383 }, { "epoch": 1.8067885117493474, "grad_norm": 3.5663323402404785, "learning_rate": 9.417491925530565e-06, "loss": 0.2062, "step": 1384 }, { "epoch": 1.8080939947780679, "grad_norm": 3.2697677612304688, "learning_rate": 9.416501201300658e-06, "loss": 0.1774, "step": 1385 }, { "epoch": 1.8093994778067886, "grad_norm": 3.936838388442993, "learning_rate": 9.415509687482892e-06, "loss": 0.1813, "step": 1386 }, { "epoch": 1.8107049608355092, "grad_norm": 3.1723825931549072, "learning_rate": 9.414517384254524e-06, "loss": 0.1509, "step": 1387 }, { "epoch": 1.8120104438642297, "grad_norm": 3.843472957611084, "learning_rate": 9.413524291792962e-06, "loss": 0.1669, "step": 1388 }, { "epoch": 1.8133159268929504, "grad_norm": 7.408055782318115, "learning_rate": 9.412530410275753e-06, "loss": 0.3545, "step": 1389 }, { "epoch": 1.814621409921671, "grad_norm": 4.9568657875061035, "learning_rate": 9.411535739880585e-06, "loss": 0.3459, "step": 1390 }, { "epoch": 1.8159268929503916, "grad_norm": 3.4355151653289795, "learning_rate": 9.410540280785281e-06, "loss": 0.1762, "step": 1391 }, { "epoch": 1.8172323759791122, "grad_norm": 4.16213321685791, "learning_rate": 9.409544033167819e-06, "loss": 0.2274, "step": 1392 }, { "epoch": 1.818537859007833, "grad_norm": 5.824769020080566, "learning_rate": 9.408546997206302e-06, "loss": 0.4336, "step": 1393 }, { "epoch": 1.8198433420365534, "grad_norm": 3.6296873092651367, "learning_rate": 9.407549173078987e-06, "loss": 0.1852, "step": 1394 }, { "epoch": 1.8211488250652743, "grad_norm": 4.322221279144287, "learning_rate": 9.406550560964263e-06, "loss": 0.2411, "step": 1395 }, { "epoch": 1.8224543080939948, "grad_norm": 4.409387588500977, "learning_rate": 9.405551161040665e-06, "loss": 0.2981, "step": 1396 }, { "epoch": 1.8237597911227152, "grad_norm": 4.173820495605469, "learning_rate": 9.404550973486865e-06, "loss": 0.243, "step": 1397 }, { "epoch": 1.8250652741514362, "grad_norm": 4.020219802856445, "learning_rate": 9.403549998481677e-06, "loss": 0.2739, "step": 1398 }, { "epoch": 1.8263707571801566, "grad_norm": 4.625650882720947, "learning_rate": 9.402548236204061e-06, "loss": 0.28, "step": 1399 }, { "epoch": 1.8276762402088773, "grad_norm": 4.291539669036865, "learning_rate": 9.40154568683311e-06, "loss": 0.2561, "step": 1400 }, { "epoch": 1.828981723237598, "grad_norm": 4.198670864105225, "learning_rate": 9.400542350548064e-06, "loss": 0.1848, "step": 1401 }, { "epoch": 1.8302872062663185, "grad_norm": 5.208884239196777, "learning_rate": 9.399538227528297e-06, "loss": 0.3686, "step": 1402 }, { "epoch": 1.8315926892950392, "grad_norm": 4.034629821777344, "learning_rate": 9.39853331795333e-06, "loss": 0.2535, "step": 1403 }, { "epoch": 1.8328981723237598, "grad_norm": 3.0943620204925537, "learning_rate": 9.39752762200282e-06, "loss": 0.2263, "step": 1404 }, { "epoch": 1.8342036553524803, "grad_norm": 3.952110528945923, "learning_rate": 9.396521139856567e-06, "loss": 0.2801, "step": 1405 }, { "epoch": 1.835509138381201, "grad_norm": 3.7394936084747314, "learning_rate": 9.395513871694515e-06, "loss": 0.178, "step": 1406 }, { "epoch": 1.8368146214099217, "grad_norm": 5.324584007263184, "learning_rate": 9.39450581769674e-06, "loss": 0.3894, "step": 1407 }, { "epoch": 1.8381201044386422, "grad_norm": 2.718724489212036, "learning_rate": 9.393496978043464e-06, "loss": 0.1846, "step": 1408 }, { "epoch": 1.839425587467363, "grad_norm": 3.756744861602783, "learning_rate": 9.392487352915052e-06, "loss": 0.2265, "step": 1409 }, { "epoch": 1.8407310704960835, "grad_norm": 5.497984409332275, "learning_rate": 9.391476942492001e-06, "loss": 0.351, "step": 1410 }, { "epoch": 1.8420365535248042, "grad_norm": 4.627076625823975, "learning_rate": 9.39046574695496e-06, "loss": 0.288, "step": 1411 }, { "epoch": 1.843342036553525, "grad_norm": 3.6233904361724854, "learning_rate": 9.389453766484703e-06, "loss": 0.2077, "step": 1412 }, { "epoch": 1.8446475195822454, "grad_norm": 2.864353656768799, "learning_rate": 9.388441001262162e-06, "loss": 0.2572, "step": 1413 }, { "epoch": 1.845953002610966, "grad_norm": 4.209203720092773, "learning_rate": 9.387427451468396e-06, "loss": 0.2457, "step": 1414 }, { "epoch": 1.8472584856396868, "grad_norm": 3.2222914695739746, "learning_rate": 9.386413117284611e-06, "loss": 0.1602, "step": 1415 }, { "epoch": 1.8485639686684072, "grad_norm": 5.361471652984619, "learning_rate": 9.385397998892148e-06, "loss": 0.2729, "step": 1416 }, { "epoch": 1.849869451697128, "grad_norm": 2.43467378616333, "learning_rate": 9.384382096472493e-06, "loss": 0.1281, "step": 1417 }, { "epoch": 1.8511749347258486, "grad_norm": 4.187060356140137, "learning_rate": 9.38336541020727e-06, "loss": 0.256, "step": 1418 }, { "epoch": 1.852480417754569, "grad_norm": 2.6193201541900635, "learning_rate": 9.382347940278242e-06, "loss": 0.1133, "step": 1419 }, { "epoch": 1.85378590078329, "grad_norm": 3.3628268241882324, "learning_rate": 9.381329686867318e-06, "loss": 0.1774, "step": 1420 }, { "epoch": 1.8550913838120104, "grad_norm": 2.2455718517303467, "learning_rate": 9.380310650156539e-06, "loss": 0.088, "step": 1421 }, { "epoch": 1.856396866840731, "grad_norm": 3.2322423458099365, "learning_rate": 9.37929083032809e-06, "loss": 0.24, "step": 1422 }, { "epoch": 1.8577023498694518, "grad_norm": 4.5297770500183105, "learning_rate": 9.378270227564296e-06, "loss": 0.259, "step": 1423 }, { "epoch": 1.8590078328981723, "grad_norm": 3.0534987449645996, "learning_rate": 9.377248842047623e-06, "loss": 0.1864, "step": 1424 }, { "epoch": 1.860313315926893, "grad_norm": 3.317842960357666, "learning_rate": 9.376226673960675e-06, "loss": 0.1762, "step": 1425 }, { "epoch": 1.8616187989556137, "grad_norm": 4.275114059448242, "learning_rate": 9.375203723486195e-06, "loss": 0.2284, "step": 1426 }, { "epoch": 1.8629242819843341, "grad_norm": 5.84952974319458, "learning_rate": 9.37417999080707e-06, "loss": 0.3973, "step": 1427 }, { "epoch": 1.8642297650130548, "grad_norm": 3.9714572429656982, "learning_rate": 9.373155476106322e-06, "loss": 0.2379, "step": 1428 }, { "epoch": 1.8655352480417755, "grad_norm": 4.49545431137085, "learning_rate": 9.372130179567116e-06, "loss": 0.233, "step": 1429 }, { "epoch": 1.866840731070496, "grad_norm": 3.6145212650299072, "learning_rate": 9.371104101372756e-06, "loss": 0.1705, "step": 1430 }, { "epoch": 1.8681462140992167, "grad_norm": 4.242685317993164, "learning_rate": 9.370077241706685e-06, "loss": 0.2273, "step": 1431 }, { "epoch": 1.8694516971279374, "grad_norm": 4.154048442840576, "learning_rate": 9.369049600752487e-06, "loss": 0.1941, "step": 1432 }, { "epoch": 1.8707571801566578, "grad_norm": 3.9181361198425293, "learning_rate": 9.368021178693885e-06, "loss": 0.2712, "step": 1433 }, { "epoch": 1.8720626631853787, "grad_norm": 2.335494041442871, "learning_rate": 9.366991975714742e-06, "loss": 0.0988, "step": 1434 }, { "epoch": 1.8733681462140992, "grad_norm": 3.900698184967041, "learning_rate": 9.36596199199906e-06, "loss": 0.1871, "step": 1435 }, { "epoch": 1.8746736292428199, "grad_norm": 3.3609836101531982, "learning_rate": 9.364931227730978e-06, "loss": 0.2053, "step": 1436 }, { "epoch": 1.8759791122715406, "grad_norm": 4.846827983856201, "learning_rate": 9.363899683094781e-06, "loss": 0.2762, "step": 1437 }, { "epoch": 1.877284595300261, "grad_norm": 3.5120747089385986, "learning_rate": 9.36286735827489e-06, "loss": 0.2017, "step": 1438 }, { "epoch": 1.8785900783289817, "grad_norm": 5.474160194396973, "learning_rate": 9.361834253455863e-06, "loss": 0.3046, "step": 1439 }, { "epoch": 1.8798955613577024, "grad_norm": 3.1328012943267822, "learning_rate": 9.360800368822403e-06, "loss": 0.1409, "step": 1440 }, { "epoch": 1.8812010443864229, "grad_norm": 4.450156211853027, "learning_rate": 9.359765704559345e-06, "loss": 0.2547, "step": 1441 }, { "epoch": 1.8825065274151436, "grad_norm": 3.2955784797668457, "learning_rate": 9.35873026085167e-06, "loss": 0.172, "step": 1442 }, { "epoch": 1.8838120104438643, "grad_norm": 4.083607196807861, "learning_rate": 9.357694037884496e-06, "loss": 0.2549, "step": 1443 }, { "epoch": 1.8851174934725847, "grad_norm": 3.636730194091797, "learning_rate": 9.35665703584308e-06, "loss": 0.182, "step": 1444 }, { "epoch": 1.8864229765013056, "grad_norm": 4.556021690368652, "learning_rate": 9.35561925491282e-06, "loss": 0.2539, "step": 1445 }, { "epoch": 1.887728459530026, "grad_norm": 2.696518659591675, "learning_rate": 9.35458069527925e-06, "loss": 0.1252, "step": 1446 }, { "epoch": 1.8890339425587466, "grad_norm": 1.7870749235153198, "learning_rate": 9.353541357128044e-06, "loss": 0.0769, "step": 1447 }, { "epoch": 1.8903394255874675, "grad_norm": 3.6830973625183105, "learning_rate": 9.35250124064502e-06, "loss": 0.2055, "step": 1448 }, { "epoch": 1.891644908616188, "grad_norm": 3.6634793281555176, "learning_rate": 9.351460346016128e-06, "loss": 0.1579, "step": 1449 }, { "epoch": 1.8929503916449086, "grad_norm": 4.128732204437256, "learning_rate": 9.350418673427462e-06, "loss": 0.2233, "step": 1450 }, { "epoch": 1.8942558746736293, "grad_norm": 4.5887980461120605, "learning_rate": 9.349376223065254e-06, "loss": 0.2564, "step": 1451 }, { "epoch": 1.8955613577023498, "grad_norm": 5.105954170227051, "learning_rate": 9.348332995115873e-06, "loss": 0.2889, "step": 1452 }, { "epoch": 1.8968668407310705, "grad_norm": 3.2114927768707275, "learning_rate": 9.34728898976583e-06, "loss": 0.1778, "step": 1453 }, { "epoch": 1.8981723237597912, "grad_norm": 3.244920253753662, "learning_rate": 9.346244207201773e-06, "loss": 0.1396, "step": 1454 }, { "epoch": 1.8994778067885116, "grad_norm": 3.8998396396636963, "learning_rate": 9.345198647610491e-06, "loss": 0.209, "step": 1455 }, { "epoch": 1.9007832898172323, "grad_norm": 2.9274861812591553, "learning_rate": 9.344152311178908e-06, "loss": 0.2019, "step": 1456 }, { "epoch": 1.902088772845953, "grad_norm": 2.219942331314087, "learning_rate": 9.343105198094092e-06, "loss": 0.0986, "step": 1457 }, { "epoch": 1.9033942558746735, "grad_norm": 5.1367597579956055, "learning_rate": 9.342057308543244e-06, "loss": 0.4167, "step": 1458 }, { "epoch": 1.9046997389033944, "grad_norm": 3.402559995651245, "learning_rate": 9.34100864271371e-06, "loss": 0.1878, "step": 1459 }, { "epoch": 1.9060052219321149, "grad_norm": 3.082578420639038, "learning_rate": 9.339959200792972e-06, "loss": 0.2108, "step": 1460 }, { "epoch": 1.9073107049608355, "grad_norm": 3.881732702255249, "learning_rate": 9.338908982968648e-06, "loss": 0.1653, "step": 1461 }, { "epoch": 1.9086161879895562, "grad_norm": 3.3911726474761963, "learning_rate": 9.337857989428501e-06, "loss": 0.2204, "step": 1462 }, { "epoch": 1.9099216710182767, "grad_norm": 4.457372188568115, "learning_rate": 9.336806220360425e-06, "loss": 0.2657, "step": 1463 }, { "epoch": 1.9112271540469974, "grad_norm": 4.147392272949219, "learning_rate": 9.335753675952459e-06, "loss": 0.2081, "step": 1464 }, { "epoch": 1.912532637075718, "grad_norm": 3.7387194633483887, "learning_rate": 9.334700356392779e-06, "loss": 0.1903, "step": 1465 }, { "epoch": 1.9138381201044385, "grad_norm": 3.2053980827331543, "learning_rate": 9.333646261869696e-06, "loss": 0.1907, "step": 1466 }, { "epoch": 1.9151436031331592, "grad_norm": 2.8787519931793213, "learning_rate": 9.332591392571665e-06, "loss": 0.1041, "step": 1467 }, { "epoch": 1.91644908616188, "grad_norm": 3.723668098449707, "learning_rate": 9.331535748687273e-06, "loss": 0.1584, "step": 1468 }, { "epoch": 1.9177545691906004, "grad_norm": 2.9837589263916016, "learning_rate": 9.330479330405256e-06, "loss": 0.1493, "step": 1469 }, { "epoch": 1.9190600522193213, "grad_norm": 4.648545265197754, "learning_rate": 9.329422137914476e-06, "loss": 0.2617, "step": 1470 }, { "epoch": 1.9203655352480418, "grad_norm": 3.4173855781555176, "learning_rate": 9.328364171403942e-06, "loss": 0.1986, "step": 1471 }, { "epoch": 1.9216710182767625, "grad_norm": 2.7682647705078125, "learning_rate": 9.327305431062799e-06, "loss": 0.1135, "step": 1472 }, { "epoch": 1.9229765013054831, "grad_norm": 4.819709300994873, "learning_rate": 9.326245917080328e-06, "loss": 0.2643, "step": 1473 }, { "epoch": 1.9242819843342036, "grad_norm": 3.2468020915985107, "learning_rate": 9.325185629645952e-06, "loss": 0.1353, "step": 1474 }, { "epoch": 1.9255874673629243, "grad_norm": 3.9541399478912354, "learning_rate": 9.324124568949229e-06, "loss": 0.2153, "step": 1475 }, { "epoch": 1.926892950391645, "grad_norm": 3.215718984603882, "learning_rate": 9.323062735179858e-06, "loss": 0.1988, "step": 1476 }, { "epoch": 1.9281984334203655, "grad_norm": 3.7686097621917725, "learning_rate": 9.322000128527675e-06, "loss": 0.2032, "step": 1477 }, { "epoch": 1.9295039164490861, "grad_norm": 3.1517016887664795, "learning_rate": 9.320936749182654e-06, "loss": 0.1476, "step": 1478 }, { "epoch": 1.9308093994778068, "grad_norm": 2.3918778896331787, "learning_rate": 9.319872597334907e-06, "loss": 0.1358, "step": 1479 }, { "epoch": 1.9321148825065273, "grad_norm": 2.7829010486602783, "learning_rate": 9.318807673174685e-06, "loss": 0.1546, "step": 1480 }, { "epoch": 1.933420365535248, "grad_norm": 3.953246831893921, "learning_rate": 9.317741976892378e-06, "loss": 0.202, "step": 1481 }, { "epoch": 1.9347258485639687, "grad_norm": 3.8942482471466064, "learning_rate": 9.316675508678508e-06, "loss": 0.1859, "step": 1482 }, { "epoch": 1.9360313315926891, "grad_norm": 3.3161706924438477, "learning_rate": 9.315608268723743e-06, "loss": 0.1717, "step": 1483 }, { "epoch": 1.93733681462141, "grad_norm": 2.361400604248047, "learning_rate": 9.314540257218886e-06, "loss": 0.1067, "step": 1484 }, { "epoch": 1.9386422976501305, "grad_norm": 2.3361382484436035, "learning_rate": 9.313471474354877e-06, "loss": 0.1556, "step": 1485 }, { "epoch": 1.9399477806788512, "grad_norm": 4.937022686004639, "learning_rate": 9.312401920322793e-06, "loss": 0.3199, "step": 1486 }, { "epoch": 1.941253263707572, "grad_norm": 4.757171154022217, "learning_rate": 9.311331595313852e-06, "loss": 0.2321, "step": 1487 }, { "epoch": 1.9425587467362924, "grad_norm": 3.810451030731201, "learning_rate": 9.310260499519405e-06, "loss": 0.2191, "step": 1488 }, { "epoch": 1.943864229765013, "grad_norm": 4.043835163116455, "learning_rate": 9.309188633130949e-06, "loss": 0.2634, "step": 1489 }, { "epoch": 1.9451697127937337, "grad_norm": 4.127200603485107, "learning_rate": 9.30811599634011e-06, "loss": 0.2967, "step": 1490 }, { "epoch": 1.9464751958224542, "grad_norm": 4.264739036560059, "learning_rate": 9.307042589338656e-06, "loss": 0.248, "step": 1491 }, { "epoch": 1.947780678851175, "grad_norm": 4.213417053222656, "learning_rate": 9.305968412318491e-06, "loss": 0.2764, "step": 1492 }, { "epoch": 1.9490861618798956, "grad_norm": 4.823579788208008, "learning_rate": 9.304893465471661e-06, "loss": 0.2771, "step": 1493 }, { "epoch": 1.950391644908616, "grad_norm": 2.686047315597534, "learning_rate": 9.303817748990344e-06, "loss": 0.1003, "step": 1494 }, { "epoch": 1.951697127937337, "grad_norm": 4.658565521240234, "learning_rate": 9.302741263066858e-06, "loss": 0.2678, "step": 1495 }, { "epoch": 1.9530026109660574, "grad_norm": 5.250690460205078, "learning_rate": 9.30166400789366e-06, "loss": 0.3231, "step": 1496 }, { "epoch": 1.9543080939947781, "grad_norm": 2.2009940147399902, "learning_rate": 9.300585983663344e-06, "loss": 0.0843, "step": 1497 }, { "epoch": 1.9556135770234988, "grad_norm": 2.5008387565612793, "learning_rate": 9.299507190568636e-06, "loss": 0.1836, "step": 1498 }, { "epoch": 1.9569190600522193, "grad_norm": 4.361505508422852, "learning_rate": 9.298427628802407e-06, "loss": 0.3007, "step": 1499 }, { "epoch": 1.95822454308094, "grad_norm": 3.704599618911743, "learning_rate": 9.297347298557664e-06, "loss": 0.2055, "step": 1500 }, { "epoch": 1.9595300261096606, "grad_norm": 3.937103509902954, "learning_rate": 9.296266200027547e-06, "loss": 0.1983, "step": 1501 }, { "epoch": 1.9608355091383811, "grad_norm": 3.164963960647583, "learning_rate": 9.295184333405338e-06, "loss": 0.1843, "step": 1502 }, { "epoch": 1.9621409921671018, "grad_norm": 2.597686767578125, "learning_rate": 9.294101698884453e-06, "loss": 0.1679, "step": 1503 }, { "epoch": 1.9634464751958225, "grad_norm": 2.6486363410949707, "learning_rate": 9.293018296658446e-06, "loss": 0.1499, "step": 1504 }, { "epoch": 1.964751958224543, "grad_norm": 2.8976097106933594, "learning_rate": 9.291934126921014e-06, "loss": 0.1642, "step": 1505 }, { "epoch": 1.9660574412532639, "grad_norm": 5.717504024505615, "learning_rate": 9.290849189865982e-06, "loss": 0.3507, "step": 1506 }, { "epoch": 1.9673629242819843, "grad_norm": 3.7029170989990234, "learning_rate": 9.289763485687317e-06, "loss": 0.1944, "step": 1507 }, { "epoch": 1.9686684073107048, "grad_norm": 3.269014358520508, "learning_rate": 9.288677014579124e-06, "loss": 0.1942, "step": 1508 }, { "epoch": 1.9699738903394257, "grad_norm": 4.876411437988281, "learning_rate": 9.287589776735644e-06, "loss": 0.3528, "step": 1509 }, { "epoch": 1.9712793733681462, "grad_norm": 3.272645950317383, "learning_rate": 9.286501772351251e-06, "loss": 0.2118, "step": 1510 }, { "epoch": 1.9725848563968669, "grad_norm": 3.1705691814422607, "learning_rate": 9.285413001620465e-06, "loss": 0.16, "step": 1511 }, { "epoch": 1.9738903394255876, "grad_norm": 3.1545495986938477, "learning_rate": 9.284323464737933e-06, "loss": 0.2051, "step": 1512 }, { "epoch": 1.975195822454308, "grad_norm": 3.1422417163848877, "learning_rate": 9.283233161898447e-06, "loss": 0.1181, "step": 1513 }, { "epoch": 1.9765013054830287, "grad_norm": 5.006057262420654, "learning_rate": 9.282142093296932e-06, "loss": 0.2924, "step": 1514 }, { "epoch": 1.9778067885117494, "grad_norm": 4.311558723449707, "learning_rate": 9.281050259128453e-06, "loss": 0.2417, "step": 1515 }, { "epoch": 1.9791122715404699, "grad_norm": 3.9262735843658447, "learning_rate": 9.279957659588203e-06, "loss": 0.1975, "step": 1516 }, { "epoch": 1.9804177545691906, "grad_norm": 5.625145435333252, "learning_rate": 9.278864294871525e-06, "loss": 0.3777, "step": 1517 }, { "epoch": 1.9817232375979112, "grad_norm": 2.7376890182495117, "learning_rate": 9.27777016517389e-06, "loss": 0.1023, "step": 1518 }, { "epoch": 1.9830287206266317, "grad_norm": 4.1617913246154785, "learning_rate": 9.276675270690907e-06, "loss": 0.2191, "step": 1519 }, { "epoch": 1.9843342036553526, "grad_norm": 4.638853073120117, "learning_rate": 9.275579611618324e-06, "loss": 0.1946, "step": 1520 }, { "epoch": 1.985639686684073, "grad_norm": 4.022986888885498, "learning_rate": 9.274483188152024e-06, "loss": 0.2371, "step": 1521 }, { "epoch": 1.9869451697127938, "grad_norm": 2.5449941158294678, "learning_rate": 9.273386000488025e-06, "loss": 0.1164, "step": 1522 }, { "epoch": 1.9882506527415145, "grad_norm": 3.646543502807617, "learning_rate": 9.272288048822486e-06, "loss": 0.2189, "step": 1523 }, { "epoch": 1.989556135770235, "grad_norm": 4.033478260040283, "learning_rate": 9.2711893333517e-06, "loss": 0.2375, "step": 1524 }, { "epoch": 1.9908616187989556, "grad_norm": 4.8788676261901855, "learning_rate": 9.270089854272097e-06, "loss": 0.2316, "step": 1525 }, { "epoch": 1.9921671018276763, "grad_norm": 3.2990105152130127, "learning_rate": 9.268989611780241e-06, "loss": 0.1856, "step": 1526 }, { "epoch": 1.9934725848563968, "grad_norm": 4.063973903656006, "learning_rate": 9.267888606072838e-06, "loss": 0.1974, "step": 1527 }, { "epoch": 1.9947780678851175, "grad_norm": 3.7050929069519043, "learning_rate": 9.266786837346727e-06, "loss": 0.1924, "step": 1528 }, { "epoch": 1.9960835509138382, "grad_norm": 3.309389352798462, "learning_rate": 9.26568430579888e-06, "loss": 0.1521, "step": 1529 }, { "epoch": 1.9973890339425586, "grad_norm": 2.7831056118011475, "learning_rate": 9.264581011626412e-06, "loss": 0.1091, "step": 1530 }, { "epoch": 1.9986945169712795, "grad_norm": 2.7118186950683594, "learning_rate": 9.263476955026571e-06, "loss": 0.1182, "step": 1531 }, { "epoch": 2.0, "grad_norm": 2.8898229598999023, "learning_rate": 9.262372136196743e-06, "loss": 0.1257, "step": 1532 }, { "epoch": 2.0013054830287205, "grad_norm": 2.6873955726623535, "learning_rate": 9.261266555334446e-06, "loss": 0.1114, "step": 1533 }, { "epoch": 2.0026109660574414, "grad_norm": 2.8446240425109863, "learning_rate": 9.260160212637337e-06, "loss": 0.1193, "step": 1534 }, { "epoch": 2.003916449086162, "grad_norm": 2.1050803661346436, "learning_rate": 9.259053108303214e-06, "loss": 0.1151, "step": 1535 }, { "epoch": 2.0052219321148823, "grad_norm": 2.385373592376709, "learning_rate": 9.257945242530002e-06, "loss": 0.0823, "step": 1536 }, { "epoch": 2.006527415143603, "grad_norm": 3.30935001373291, "learning_rate": 9.256836615515767e-06, "loss": 0.164, "step": 1537 }, { "epoch": 2.0078328981723237, "grad_norm": 2.435448408126831, "learning_rate": 9.255727227458715e-06, "loss": 0.1229, "step": 1538 }, { "epoch": 2.0091383812010446, "grad_norm": 3.483673334121704, "learning_rate": 9.254617078557179e-06, "loss": 0.1421, "step": 1539 }, { "epoch": 2.010443864229765, "grad_norm": 4.021923542022705, "learning_rate": 9.253506169009636e-06, "loss": 0.2587, "step": 1540 }, { "epoch": 2.0117493472584855, "grad_norm": 2.791393756866455, "learning_rate": 9.252394499014694e-06, "loss": 0.1224, "step": 1541 }, { "epoch": 2.0130548302872064, "grad_norm": 2.5473804473876953, "learning_rate": 9.2512820687711e-06, "loss": 0.1504, "step": 1542 }, { "epoch": 2.014360313315927, "grad_norm": 3.65739107131958, "learning_rate": 9.250168878477735e-06, "loss": 0.1545, "step": 1543 }, { "epoch": 2.0156657963446474, "grad_norm": 2.2087366580963135, "learning_rate": 9.249054928333616e-06, "loss": 0.0933, "step": 1544 }, { "epoch": 2.0169712793733683, "grad_norm": 3.5285069942474365, "learning_rate": 9.247940218537896e-06, "loss": 0.1796, "step": 1545 }, { "epoch": 2.0182767624020888, "grad_norm": 3.3223750591278076, "learning_rate": 9.246824749289868e-06, "loss": 0.1818, "step": 1546 }, { "epoch": 2.019582245430809, "grad_norm": 3.9556708335876465, "learning_rate": 9.245708520788953e-06, "loss": 0.2389, "step": 1547 }, { "epoch": 2.02088772845953, "grad_norm": 3.2928121089935303, "learning_rate": 9.244591533234716e-06, "loss": 0.1532, "step": 1548 }, { "epoch": 2.0221932114882506, "grad_norm": 3.915703773498535, "learning_rate": 9.243473786826848e-06, "loss": 0.1818, "step": 1549 }, { "epoch": 2.023498694516971, "grad_norm": 3.1992859840393066, "learning_rate": 9.242355281765188e-06, "loss": 0.1394, "step": 1550 }, { "epoch": 2.024804177545692, "grad_norm": 3.131044864654541, "learning_rate": 9.241236018249697e-06, "loss": 0.1491, "step": 1551 }, { "epoch": 2.0261096605744124, "grad_norm": 3.288909435272217, "learning_rate": 9.24011599648048e-06, "loss": 0.1311, "step": 1552 }, { "epoch": 2.0274151436031334, "grad_norm": 5.110584259033203, "learning_rate": 9.23899521665778e-06, "loss": 0.2897, "step": 1553 }, { "epoch": 2.028720626631854, "grad_norm": 3.5262722969055176, "learning_rate": 9.237873678981968e-06, "loss": 0.1833, "step": 1554 }, { "epoch": 2.0300261096605743, "grad_norm": 4.225942134857178, "learning_rate": 9.236751383653553e-06, "loss": 0.201, "step": 1555 }, { "epoch": 2.031331592689295, "grad_norm": 4.144259452819824, "learning_rate": 9.235628330873184e-06, "loss": 0.1542, "step": 1556 }, { "epoch": 2.0326370757180157, "grad_norm": 3.282069444656372, "learning_rate": 9.234504520841639e-06, "loss": 0.1991, "step": 1557 }, { "epoch": 2.033942558746736, "grad_norm": 3.7449941635131836, "learning_rate": 9.233379953759835e-06, "loss": 0.1875, "step": 1558 }, { "epoch": 2.035248041775457, "grad_norm": 2.694796562194824, "learning_rate": 9.232254629828824e-06, "loss": 0.1047, "step": 1559 }, { "epoch": 2.0365535248041775, "grad_norm": 3.304600238800049, "learning_rate": 9.231128549249794e-06, "loss": 0.182, "step": 1560 }, { "epoch": 2.037859007832898, "grad_norm": 2.981200695037842, "learning_rate": 9.230001712224065e-06, "loss": 0.1589, "step": 1561 }, { "epoch": 2.039164490861619, "grad_norm": 3.559706449508667, "learning_rate": 9.228874118953095e-06, "loss": 0.1837, "step": 1562 }, { "epoch": 2.0404699738903394, "grad_norm": 4.100469589233398, "learning_rate": 9.227745769638476e-06, "loss": 0.1564, "step": 1563 }, { "epoch": 2.0417754569190603, "grad_norm": 3.238297462463379, "learning_rate": 9.226616664481937e-06, "loss": 0.1525, "step": 1564 }, { "epoch": 2.0430809399477807, "grad_norm": 2.982931613922119, "learning_rate": 9.225486803685342e-06, "loss": 0.1144, "step": 1565 }, { "epoch": 2.044386422976501, "grad_norm": 3.76670503616333, "learning_rate": 9.224356187450686e-06, "loss": 0.166, "step": 1566 }, { "epoch": 2.045691906005222, "grad_norm": 2.921903371810913, "learning_rate": 9.223224815980103e-06, "loss": 0.1396, "step": 1567 }, { "epoch": 2.0469973890339426, "grad_norm": 3.8409228324890137, "learning_rate": 9.222092689475863e-06, "loss": 0.1698, "step": 1568 }, { "epoch": 2.048302872062663, "grad_norm": 4.700915813446045, "learning_rate": 9.220959808140367e-06, "loss": 0.2639, "step": 1569 }, { "epoch": 2.049608355091384, "grad_norm": 2.4071404933929443, "learning_rate": 9.219826172176154e-06, "loss": 0.1166, "step": 1570 }, { "epoch": 2.0509138381201044, "grad_norm": 3.7058236598968506, "learning_rate": 9.218691781785895e-06, "loss": 0.2129, "step": 1571 }, { "epoch": 2.052219321148825, "grad_norm": 2.25942325592041, "learning_rate": 9.2175566371724e-06, "loss": 0.0857, "step": 1572 }, { "epoch": 2.053524804177546, "grad_norm": 2.8401098251342773, "learning_rate": 9.216420738538613e-06, "loss": 0.0949, "step": 1573 }, { "epoch": 2.0548302872062663, "grad_norm": 3.7243192195892334, "learning_rate": 9.215284086087608e-06, "loss": 0.1448, "step": 1574 }, { "epoch": 2.056135770234987, "grad_norm": 2.291712760925293, "learning_rate": 9.214146680022598e-06, "loss": 0.0922, "step": 1575 }, { "epoch": 2.0574412532637076, "grad_norm": 3.10800838470459, "learning_rate": 9.213008520546932e-06, "loss": 0.1378, "step": 1576 }, { "epoch": 2.058746736292428, "grad_norm": 3.232893228530884, "learning_rate": 9.21186960786409e-06, "loss": 0.121, "step": 1577 }, { "epoch": 2.060052219321149, "grad_norm": 4.409466743469238, "learning_rate": 9.210729942177687e-06, "loss": 0.2517, "step": 1578 }, { "epoch": 2.0613577023498695, "grad_norm": 3.195516586303711, "learning_rate": 9.209589523691478e-06, "loss": 0.1572, "step": 1579 }, { "epoch": 2.06266318537859, "grad_norm": 2.391859292984009, "learning_rate": 9.208448352609344e-06, "loss": 0.1029, "step": 1580 }, { "epoch": 2.063968668407311, "grad_norm": 2.167759895324707, "learning_rate": 9.207306429135308e-06, "loss": 0.0716, "step": 1581 }, { "epoch": 2.0652741514360313, "grad_norm": 4.594433307647705, "learning_rate": 9.206163753473522e-06, "loss": 0.2241, "step": 1582 }, { "epoch": 2.066579634464752, "grad_norm": 2.5609753131866455, "learning_rate": 9.205020325828278e-06, "loss": 0.1152, "step": 1583 }, { "epoch": 2.0678851174934727, "grad_norm": 3.146400213241577, "learning_rate": 9.203876146403998e-06, "loss": 0.1436, "step": 1584 }, { "epoch": 2.069190600522193, "grad_norm": 3.2570865154266357, "learning_rate": 9.202731215405241e-06, "loss": 0.1211, "step": 1585 }, { "epoch": 2.0704960835509136, "grad_norm": 4.274707794189453, "learning_rate": 9.201585533036696e-06, "loss": 0.1982, "step": 1586 }, { "epoch": 2.0718015665796345, "grad_norm": 2.228811025619507, "learning_rate": 9.200439099503194e-06, "loss": 0.1027, "step": 1587 }, { "epoch": 2.073107049608355, "grad_norm": 2.4780337810516357, "learning_rate": 9.199291915009693e-06, "loss": 0.1102, "step": 1588 }, { "epoch": 2.074412532637076, "grad_norm": 2.8563969135284424, "learning_rate": 9.198143979761289e-06, "loss": 0.124, "step": 1589 }, { "epoch": 2.0757180156657964, "grad_norm": 3.3803446292877197, "learning_rate": 9.196995293963211e-06, "loss": 0.1391, "step": 1590 }, { "epoch": 2.077023498694517, "grad_norm": 3.099701166152954, "learning_rate": 9.195845857820824e-06, "loss": 0.1318, "step": 1591 }, { "epoch": 2.0783289817232378, "grad_norm": 3.8973052501678467, "learning_rate": 9.194695671539623e-06, "loss": 0.1294, "step": 1592 }, { "epoch": 2.0796344647519582, "grad_norm": 4.035362720489502, "learning_rate": 9.193544735325242e-06, "loss": 0.1821, "step": 1593 }, { "epoch": 2.0809399477806787, "grad_norm": 4.289149761199951, "learning_rate": 9.192393049383445e-06, "loss": 0.1885, "step": 1594 }, { "epoch": 2.0822454308093996, "grad_norm": 3.8874850273132324, "learning_rate": 9.191240613920136e-06, "loss": 0.1948, "step": 1595 }, { "epoch": 2.08355091383812, "grad_norm": 3.1654484272003174, "learning_rate": 9.190087429141343e-06, "loss": 0.1266, "step": 1596 }, { "epoch": 2.0848563968668405, "grad_norm": 4.192917823791504, "learning_rate": 9.188933495253239e-06, "loss": 0.1591, "step": 1597 }, { "epoch": 2.0861618798955615, "grad_norm": 2.99303936958313, "learning_rate": 9.187778812462124e-06, "loss": 0.1548, "step": 1598 }, { "epoch": 2.087467362924282, "grad_norm": 2.554405689239502, "learning_rate": 9.186623380974433e-06, "loss": 0.0829, "step": 1599 }, { "epoch": 2.0887728459530024, "grad_norm": 3.0192606449127197, "learning_rate": 9.185467200996734e-06, "loss": 0.1448, "step": 1600 }, { "epoch": 2.0900783289817233, "grad_norm": 1.9988819360733032, "learning_rate": 9.184310272735734e-06, "loss": 0.0803, "step": 1601 }, { "epoch": 2.0913838120104438, "grad_norm": 4.604189872741699, "learning_rate": 9.18315259639827e-06, "loss": 0.2546, "step": 1602 }, { "epoch": 2.0926892950391647, "grad_norm": 3.169858694076538, "learning_rate": 9.18199417219131e-06, "loss": 0.1436, "step": 1603 }, { "epoch": 2.093994778067885, "grad_norm": 3.5824246406555176, "learning_rate": 9.18083500032196e-06, "loss": 0.1675, "step": 1604 }, { "epoch": 2.0953002610966056, "grad_norm": 2.89316987991333, "learning_rate": 9.179675080997459e-06, "loss": 0.1013, "step": 1605 }, { "epoch": 2.0966057441253265, "grad_norm": 3.095094680786133, "learning_rate": 9.178514414425178e-06, "loss": 0.1714, "step": 1606 }, { "epoch": 2.097911227154047, "grad_norm": 3.3552587032318115, "learning_rate": 9.17735300081262e-06, "loss": 0.1589, "step": 1607 }, { "epoch": 2.0992167101827675, "grad_norm": 2.9293715953826904, "learning_rate": 9.17619084036743e-06, "loss": 0.1293, "step": 1608 }, { "epoch": 2.1005221932114884, "grad_norm": 2.8815064430236816, "learning_rate": 9.175027933297376e-06, "loss": 0.1403, "step": 1609 }, { "epoch": 2.101827676240209, "grad_norm": 3.08553409576416, "learning_rate": 9.173864279810367e-06, "loss": 0.1463, "step": 1610 }, { "epoch": 2.1031331592689293, "grad_norm": 2.0239791870117188, "learning_rate": 9.172699880114439e-06, "loss": 0.1212, "step": 1611 }, { "epoch": 2.10443864229765, "grad_norm": 1.6485800743103027, "learning_rate": 9.171534734417767e-06, "loss": 0.0896, "step": 1612 }, { "epoch": 2.1057441253263707, "grad_norm": 3.7918500900268555, "learning_rate": 9.170368842928656e-06, "loss": 0.1824, "step": 1613 }, { "epoch": 2.1070496083550916, "grad_norm": 4.321269512176514, "learning_rate": 9.169202205855548e-06, "loss": 0.21, "step": 1614 }, { "epoch": 2.108355091383812, "grad_norm": 4.30157995223999, "learning_rate": 9.168034823407015e-06, "loss": 0.2376, "step": 1615 }, { "epoch": 2.1096605744125325, "grad_norm": 4.51170015335083, "learning_rate": 9.166866695791761e-06, "loss": 0.2567, "step": 1616 }, { "epoch": 2.1109660574412534, "grad_norm": 2.464839458465576, "learning_rate": 9.165697823218629e-06, "loss": 0.0727, "step": 1617 }, { "epoch": 2.112271540469974, "grad_norm": 3.2326502799987793, "learning_rate": 9.164528205896588e-06, "loss": 0.1426, "step": 1618 }, { "epoch": 2.1135770234986944, "grad_norm": 2.0443615913391113, "learning_rate": 9.163357844034745e-06, "loss": 0.1131, "step": 1619 }, { "epoch": 2.1148825065274153, "grad_norm": 3.0657100677490234, "learning_rate": 9.16218673784234e-06, "loss": 0.1179, "step": 1620 }, { "epoch": 2.1161879895561357, "grad_norm": 2.1923561096191406, "learning_rate": 9.161014887528745e-06, "loss": 0.1158, "step": 1621 }, { "epoch": 2.117493472584856, "grad_norm": 2.491220474243164, "learning_rate": 9.159842293303462e-06, "loss": 0.104, "step": 1622 }, { "epoch": 2.118798955613577, "grad_norm": 5.473113536834717, "learning_rate": 9.158668955376134e-06, "loss": 0.3572, "step": 1623 }, { "epoch": 2.1201044386422976, "grad_norm": 3.585726261138916, "learning_rate": 9.157494873956525e-06, "loss": 0.1244, "step": 1624 }, { "epoch": 2.1214099216710185, "grad_norm": 3.0093865394592285, "learning_rate": 9.156320049254546e-06, "loss": 0.1356, "step": 1625 }, { "epoch": 2.122715404699739, "grad_norm": 2.8502559661865234, "learning_rate": 9.155144481480228e-06, "loss": 0.1195, "step": 1626 }, { "epoch": 2.1240208877284594, "grad_norm": 2.740832805633545, "learning_rate": 9.153968170843743e-06, "loss": 0.1421, "step": 1627 }, { "epoch": 2.1253263707571803, "grad_norm": 3.467851161956787, "learning_rate": 9.152791117555395e-06, "loss": 0.1943, "step": 1628 }, { "epoch": 2.126631853785901, "grad_norm": 3.2789156436920166, "learning_rate": 9.151613321825616e-06, "loss": 0.1658, "step": 1629 }, { "epoch": 2.1279373368146213, "grad_norm": 2.362457513809204, "learning_rate": 9.150434783864975e-06, "loss": 0.0991, "step": 1630 }, { "epoch": 2.129242819843342, "grad_norm": 2.593410015106201, "learning_rate": 9.149255503884173e-06, "loss": 0.094, "step": 1631 }, { "epoch": 2.1305483028720626, "grad_norm": 3.9534928798675537, "learning_rate": 9.148075482094043e-06, "loss": 0.1924, "step": 1632 }, { "epoch": 2.131853785900783, "grad_norm": 2.1914260387420654, "learning_rate": 9.14689471870555e-06, "loss": 0.0986, "step": 1633 }, { "epoch": 2.133159268929504, "grad_norm": 3.4105489253997803, "learning_rate": 9.145713213929796e-06, "loss": 0.1605, "step": 1634 }, { "epoch": 2.1344647519582245, "grad_norm": 2.3692562580108643, "learning_rate": 9.144530967978008e-06, "loss": 0.0838, "step": 1635 }, { "epoch": 2.1357702349869454, "grad_norm": 3.0031371116638184, "learning_rate": 9.14334798106155e-06, "loss": 0.1738, "step": 1636 }, { "epoch": 2.137075718015666, "grad_norm": 3.8378498554229736, "learning_rate": 9.142164253391919e-06, "loss": 0.1758, "step": 1637 }, { "epoch": 2.1383812010443863, "grad_norm": 2.1176819801330566, "learning_rate": 9.140979785180742e-06, "loss": 0.0604, "step": 1638 }, { "epoch": 2.1396866840731072, "grad_norm": 4.318655967712402, "learning_rate": 9.139794576639782e-06, "loss": 0.2605, "step": 1639 }, { "epoch": 2.1409921671018277, "grad_norm": 3.7991878986358643, "learning_rate": 9.13860862798093e-06, "loss": 0.182, "step": 1640 }, { "epoch": 2.142297650130548, "grad_norm": 2.938815116882324, "learning_rate": 9.137421939416213e-06, "loss": 0.1016, "step": 1641 }, { "epoch": 2.143603133159269, "grad_norm": 1.4802241325378418, "learning_rate": 9.136234511157787e-06, "loss": 0.0504, "step": 1642 }, { "epoch": 2.1449086161879896, "grad_norm": 2.625304937362671, "learning_rate": 9.135046343417945e-06, "loss": 0.1046, "step": 1643 }, { "epoch": 2.14621409921671, "grad_norm": 3.819563388824463, "learning_rate": 9.133857436409106e-06, "loss": 0.1613, "step": 1644 }, { "epoch": 2.147519582245431, "grad_norm": 3.242746114730835, "learning_rate": 9.132667790343825e-06, "loss": 0.139, "step": 1645 }, { "epoch": 2.1488250652741514, "grad_norm": 3.036539077758789, "learning_rate": 9.131477405434793e-06, "loss": 0.1302, "step": 1646 }, { "epoch": 2.150130548302872, "grad_norm": 2.1474010944366455, "learning_rate": 9.13028628189482e-06, "loss": 0.0821, "step": 1647 }, { "epoch": 2.151436031331593, "grad_norm": 5.086942672729492, "learning_rate": 9.129094419936864e-06, "loss": 0.3665, "step": 1648 }, { "epoch": 2.1527415143603132, "grad_norm": 1.6741089820861816, "learning_rate": 9.127901819774003e-06, "loss": 0.049, "step": 1649 }, { "epoch": 2.1540469973890337, "grad_norm": 3.8880112171173096, "learning_rate": 9.126708481619455e-06, "loss": 0.1771, "step": 1650 }, { "epoch": 2.1553524804177546, "grad_norm": 2.110764980316162, "learning_rate": 9.125514405686565e-06, "loss": 0.093, "step": 1651 }, { "epoch": 2.156657963446475, "grad_norm": 2.2039613723754883, "learning_rate": 9.124319592188811e-06, "loss": 0.0845, "step": 1652 }, { "epoch": 2.157963446475196, "grad_norm": 2.255340576171875, "learning_rate": 9.123124041339805e-06, "loss": 0.0799, "step": 1653 }, { "epoch": 2.1592689295039165, "grad_norm": 2.1461329460144043, "learning_rate": 9.121927753353287e-06, "loss": 0.0827, "step": 1654 }, { "epoch": 2.160574412532637, "grad_norm": 3.9780495166778564, "learning_rate": 9.120730728443133e-06, "loss": 0.1505, "step": 1655 }, { "epoch": 2.161879895561358, "grad_norm": 3.303396463394165, "learning_rate": 9.119532966823346e-06, "loss": 0.1313, "step": 1656 }, { "epoch": 2.1631853785900783, "grad_norm": 2.1104981899261475, "learning_rate": 9.118334468708065e-06, "loss": 0.0894, "step": 1657 }, { "epoch": 2.164490861618799, "grad_norm": 2.3108856678009033, "learning_rate": 9.11713523431156e-06, "loss": 0.084, "step": 1658 }, { "epoch": 2.1657963446475197, "grad_norm": 3.457564115524292, "learning_rate": 9.115935263848233e-06, "loss": 0.1298, "step": 1659 }, { "epoch": 2.16710182767624, "grad_norm": 3.158428192138672, "learning_rate": 9.114734557532611e-06, "loss": 0.1257, "step": 1660 }, { "epoch": 2.1684073107049606, "grad_norm": 3.082343816757202, "learning_rate": 9.113533115579362e-06, "loss": 0.1383, "step": 1661 }, { "epoch": 2.1697127937336815, "grad_norm": 4.275191307067871, "learning_rate": 9.11233093820328e-06, "loss": 0.1231, "step": 1662 }, { "epoch": 2.171018276762402, "grad_norm": 3.4219651222229004, "learning_rate": 9.111128025619294e-06, "loss": 0.126, "step": 1663 }, { "epoch": 2.172323759791123, "grad_norm": 3.677185297012329, "learning_rate": 9.109924378042457e-06, "loss": 0.1615, "step": 1664 }, { "epoch": 2.1736292428198434, "grad_norm": 4.45488977432251, "learning_rate": 9.108719995687965e-06, "loss": 0.2539, "step": 1665 }, { "epoch": 2.174934725848564, "grad_norm": 4.347939491271973, "learning_rate": 9.107514878771135e-06, "loss": 0.2126, "step": 1666 }, { "epoch": 2.1762402088772848, "grad_norm": 2.3345465660095215, "learning_rate": 9.106309027507421e-06, "loss": 0.0924, "step": 1667 }, { "epoch": 2.177545691906005, "grad_norm": 2.415796995162964, "learning_rate": 9.105102442112405e-06, "loss": 0.0941, "step": 1668 }, { "epoch": 2.1788511749347257, "grad_norm": 2.5291340351104736, "learning_rate": 9.103895122801804e-06, "loss": 0.0981, "step": 1669 }, { "epoch": 2.1801566579634466, "grad_norm": 4.988849639892578, "learning_rate": 9.102687069791464e-06, "loss": 0.2708, "step": 1670 }, { "epoch": 2.181462140992167, "grad_norm": 1.8167088031768799, "learning_rate": 9.10147828329736e-06, "loss": 0.0562, "step": 1671 }, { "epoch": 2.1827676240208875, "grad_norm": 3.1535701751708984, "learning_rate": 9.100268763535603e-06, "loss": 0.1511, "step": 1672 }, { "epoch": 2.1840731070496084, "grad_norm": 3.286993980407715, "learning_rate": 9.099058510722431e-06, "loss": 0.1125, "step": 1673 }, { "epoch": 2.185378590078329, "grad_norm": 1.7948250770568848, "learning_rate": 9.097847525074216e-06, "loss": 0.0708, "step": 1674 }, { "epoch": 2.18668407310705, "grad_norm": 3.209502935409546, "learning_rate": 9.096635806807459e-06, "loss": 0.1446, "step": 1675 }, { "epoch": 2.1879895561357703, "grad_norm": 3.6386568546295166, "learning_rate": 9.095423356138791e-06, "loss": 0.1462, "step": 1676 }, { "epoch": 2.1892950391644908, "grad_norm": 5.789756774902344, "learning_rate": 9.094210173284976e-06, "loss": 0.2512, "step": 1677 }, { "epoch": 2.1906005221932117, "grad_norm": 2.996037006378174, "learning_rate": 9.092996258462912e-06, "loss": 0.1237, "step": 1678 }, { "epoch": 2.191906005221932, "grad_norm": 2.551654815673828, "learning_rate": 9.09178161188962e-06, "loss": 0.0904, "step": 1679 }, { "epoch": 2.1932114882506526, "grad_norm": 2.8869121074676514, "learning_rate": 9.090566233782258e-06, "loss": 0.0964, "step": 1680 }, { "epoch": 2.1945169712793735, "grad_norm": 2.6318106651306152, "learning_rate": 9.089350124358112e-06, "loss": 0.0926, "step": 1681 }, { "epoch": 2.195822454308094, "grad_norm": 2.674509048461914, "learning_rate": 9.088133283834598e-06, "loss": 0.1392, "step": 1682 }, { "epoch": 2.1971279373368144, "grad_norm": 4.447915554046631, "learning_rate": 9.08691571242927e-06, "loss": 0.1579, "step": 1683 }, { "epoch": 2.1984334203655354, "grad_norm": 2.6103620529174805, "learning_rate": 9.085697410359802e-06, "loss": 0.1018, "step": 1684 }, { "epoch": 2.199738903394256, "grad_norm": 3.720806837081909, "learning_rate": 9.084478377844005e-06, "loss": 0.1955, "step": 1685 }, { "epoch": 2.2010443864229767, "grad_norm": 2.938732385635376, "learning_rate": 9.083258615099819e-06, "loss": 0.1526, "step": 1686 }, { "epoch": 2.202349869451697, "grad_norm": 3.9781031608581543, "learning_rate": 9.082038122345316e-06, "loss": 0.1878, "step": 1687 }, { "epoch": 2.2036553524804177, "grad_norm": 2.5145084857940674, "learning_rate": 9.080816899798697e-06, "loss": 0.1039, "step": 1688 }, { "epoch": 2.2049608355091386, "grad_norm": 3.0798821449279785, "learning_rate": 9.079594947678292e-06, "loss": 0.1139, "step": 1689 }, { "epoch": 2.206266318537859, "grad_norm": 3.329089879989624, "learning_rate": 9.078372266202565e-06, "loss": 0.1347, "step": 1690 }, { "epoch": 2.2075718015665795, "grad_norm": 3.8820345401763916, "learning_rate": 9.077148855590107e-06, "loss": 0.1497, "step": 1691 }, { "epoch": 2.2088772845953004, "grad_norm": 2.919802188873291, "learning_rate": 9.075924716059642e-06, "loss": 0.1313, "step": 1692 }, { "epoch": 2.210182767624021, "grad_norm": 2.678307294845581, "learning_rate": 9.074699847830024e-06, "loss": 0.121, "step": 1693 }, { "epoch": 2.2114882506527413, "grad_norm": 2.141226291656494, "learning_rate": 9.073474251120235e-06, "loss": 0.0803, "step": 1694 }, { "epoch": 2.2127937336814623, "grad_norm": 2.4920902252197266, "learning_rate": 9.072247926149388e-06, "loss": 0.1098, "step": 1695 }, { "epoch": 2.2140992167101827, "grad_norm": 3.631624221801758, "learning_rate": 9.071020873136732e-06, "loss": 0.1552, "step": 1696 }, { "epoch": 2.215404699738903, "grad_norm": 2.4036200046539307, "learning_rate": 9.069793092301636e-06, "loss": 0.096, "step": 1697 }, { "epoch": 2.216710182767624, "grad_norm": 2.7188405990600586, "learning_rate": 9.068564583863607e-06, "loss": 0.1278, "step": 1698 }, { "epoch": 2.2180156657963446, "grad_norm": 1.7580981254577637, "learning_rate": 9.067335348042277e-06, "loss": 0.0599, "step": 1699 }, { "epoch": 2.2193211488250655, "grad_norm": 3.6770102977752686, "learning_rate": 9.066105385057412e-06, "loss": 0.1795, "step": 1700 }, { "epoch": 2.220626631853786, "grad_norm": 2.7613043785095215, "learning_rate": 9.064874695128905e-06, "loss": 0.0882, "step": 1701 }, { "epoch": 2.2219321148825064, "grad_norm": 3.080204486846924, "learning_rate": 9.063643278476781e-06, "loss": 0.1057, "step": 1702 }, { "epoch": 2.2232375979112273, "grad_norm": 2.6061699390411377, "learning_rate": 9.062411135321196e-06, "loss": 0.0952, "step": 1703 }, { "epoch": 2.224543080939948, "grad_norm": 2.0860893726348877, "learning_rate": 9.061178265882433e-06, "loss": 0.0845, "step": 1704 }, { "epoch": 2.2258485639686683, "grad_norm": 2.310960292816162, "learning_rate": 9.059944670380906e-06, "loss": 0.0834, "step": 1705 }, { "epoch": 2.227154046997389, "grad_norm": 3.7324085235595703, "learning_rate": 9.058710349037159e-06, "loss": 0.163, "step": 1706 }, { "epoch": 2.2284595300261096, "grad_norm": 3.424032211303711, "learning_rate": 9.057475302071864e-06, "loss": 0.1606, "step": 1707 }, { "epoch": 2.22976501305483, "grad_norm": 2.135648488998413, "learning_rate": 9.056239529705824e-06, "loss": 0.0958, "step": 1708 }, { "epoch": 2.231070496083551, "grad_norm": 2.4260261058807373, "learning_rate": 9.055003032159976e-06, "loss": 0.0921, "step": 1709 }, { "epoch": 2.2323759791122715, "grad_norm": 3.9525516033172607, "learning_rate": 9.053765809655381e-06, "loss": 0.1831, "step": 1710 }, { "epoch": 2.233681462140992, "grad_norm": 3.654252290725708, "learning_rate": 9.052527862413227e-06, "loss": 0.1743, "step": 1711 }, { "epoch": 2.234986945169713, "grad_norm": 2.4661474227905273, "learning_rate": 9.051289190654842e-06, "loss": 0.0896, "step": 1712 }, { "epoch": 2.2362924281984333, "grad_norm": 1.864582896232605, "learning_rate": 9.050049794601676e-06, "loss": 0.099, "step": 1713 }, { "epoch": 2.2375979112271542, "grad_norm": 1.8312435150146484, "learning_rate": 9.048809674475306e-06, "loss": 0.0923, "step": 1714 }, { "epoch": 2.2389033942558747, "grad_norm": 3.293246269226074, "learning_rate": 9.047568830497445e-06, "loss": 0.1202, "step": 1715 }, { "epoch": 2.240208877284595, "grad_norm": 1.7990732192993164, "learning_rate": 9.046327262889934e-06, "loss": 0.0607, "step": 1716 }, { "epoch": 2.241514360313316, "grad_norm": 3.331395149230957, "learning_rate": 9.045084971874738e-06, "loss": 0.1629, "step": 1717 }, { "epoch": 2.2428198433420365, "grad_norm": 2.9150197505950928, "learning_rate": 9.04384195767396e-06, "loss": 0.1117, "step": 1718 }, { "epoch": 2.244125326370757, "grad_norm": 1.9604214429855347, "learning_rate": 9.042598220509824e-06, "loss": 0.0648, "step": 1719 }, { "epoch": 2.245430809399478, "grad_norm": 4.824312210083008, "learning_rate": 9.04135376060469e-06, "loss": 0.1667, "step": 1720 }, { "epoch": 2.2467362924281984, "grad_norm": 2.8939173221588135, "learning_rate": 9.040108578181043e-06, "loss": 0.1287, "step": 1721 }, { "epoch": 2.248041775456919, "grad_norm": 2.525820255279541, "learning_rate": 9.038862673461497e-06, "loss": 0.0797, "step": 1722 }, { "epoch": 2.2493472584856398, "grad_norm": 3.097977876663208, "learning_rate": 9.037616046668798e-06, "loss": 0.1233, "step": 1723 }, { "epoch": 2.2506527415143602, "grad_norm": 2.7518088817596436, "learning_rate": 9.036368698025821e-06, "loss": 0.0955, "step": 1724 }, { "epoch": 2.251958224543081, "grad_norm": 2.577124834060669, "learning_rate": 9.035120627755564e-06, "loss": 0.0965, "step": 1725 }, { "epoch": 2.2532637075718016, "grad_norm": 3.0648093223571777, "learning_rate": 9.033871836081164e-06, "loss": 0.1412, "step": 1726 }, { "epoch": 2.254569190600522, "grad_norm": 4.827856063842773, "learning_rate": 9.032622323225878e-06, "loss": 0.2401, "step": 1727 }, { "epoch": 2.255874673629243, "grad_norm": 2.609083414077759, "learning_rate": 9.031372089413097e-06, "loss": 0.1069, "step": 1728 }, { "epoch": 2.2571801566579635, "grad_norm": 1.8638418912887573, "learning_rate": 9.030121134866338e-06, "loss": 0.0527, "step": 1729 }, { "epoch": 2.258485639686684, "grad_norm": 3.2911181449890137, "learning_rate": 9.028869459809251e-06, "loss": 0.1392, "step": 1730 }, { "epoch": 2.259791122715405, "grad_norm": 2.85219144821167, "learning_rate": 9.027617064465609e-06, "loss": 0.1201, "step": 1731 }, { "epoch": 2.2610966057441253, "grad_norm": 1.744975209236145, "learning_rate": 9.026363949059319e-06, "loss": 0.0513, "step": 1732 }, { "epoch": 2.2624020887728458, "grad_norm": 2.7231245040893555, "learning_rate": 9.025110113814414e-06, "loss": 0.1167, "step": 1733 }, { "epoch": 2.2637075718015667, "grad_norm": 2.351167678833008, "learning_rate": 9.023855558955055e-06, "loss": 0.0796, "step": 1734 }, { "epoch": 2.265013054830287, "grad_norm": 4.254123210906982, "learning_rate": 9.022600284705537e-06, "loss": 0.2114, "step": 1735 }, { "epoch": 2.266318537859008, "grad_norm": 7.259061336517334, "learning_rate": 9.021344291290275e-06, "loss": 0.1383, "step": 1736 }, { "epoch": 2.2676240208877285, "grad_norm": 1.6668890714645386, "learning_rate": 9.020087578933819e-06, "loss": 0.0471, "step": 1737 }, { "epoch": 2.268929503916449, "grad_norm": 3.5843911170959473, "learning_rate": 9.018830147860844e-06, "loss": 0.1138, "step": 1738 }, { "epoch": 2.27023498694517, "grad_norm": 2.656080722808838, "learning_rate": 9.01757199829616e-06, "loss": 0.124, "step": 1739 }, { "epoch": 2.2715404699738904, "grad_norm": 2.722414493560791, "learning_rate": 9.016313130464696e-06, "loss": 0.1117, "step": 1740 }, { "epoch": 2.272845953002611, "grad_norm": 3.6728334426879883, "learning_rate": 9.015053544591515e-06, "loss": 0.1347, "step": 1741 }, { "epoch": 2.2741514360313317, "grad_norm": 3.942530632019043, "learning_rate": 9.013793240901807e-06, "loss": 0.2024, "step": 1742 }, { "epoch": 2.275456919060052, "grad_norm": 4.779889106750488, "learning_rate": 9.012532219620893e-06, "loss": 0.2339, "step": 1743 }, { "epoch": 2.2767624020887727, "grad_norm": 2.738227367401123, "learning_rate": 9.01127048097422e-06, "loss": 0.1223, "step": 1744 }, { "epoch": 2.2780678851174936, "grad_norm": 2.966574192047119, "learning_rate": 9.010008025187359e-06, "loss": 0.1624, "step": 1745 }, { "epoch": 2.279373368146214, "grad_norm": 2.0337517261505127, "learning_rate": 9.008744852486018e-06, "loss": 0.067, "step": 1746 }, { "epoch": 2.280678851174935, "grad_norm": 2.951179027557373, "learning_rate": 9.007480963096028e-06, "loss": 0.093, "step": 1747 }, { "epoch": 2.2819843342036554, "grad_norm": 2.8404319286346436, "learning_rate": 9.006216357243344e-06, "loss": 0.1015, "step": 1748 }, { "epoch": 2.283289817232376, "grad_norm": 3.747481107711792, "learning_rate": 9.004951035154061e-06, "loss": 0.1488, "step": 1749 }, { "epoch": 2.2845953002610964, "grad_norm": 3.792058229446411, "learning_rate": 9.003684997054391e-06, "loss": 0.1666, "step": 1750 }, { "epoch": 2.2859007832898173, "grad_norm": 3.6498606204986572, "learning_rate": 9.00241824317068e-06, "loss": 0.1405, "step": 1751 }, { "epoch": 2.2872062663185377, "grad_norm": 3.5023863315582275, "learning_rate": 9.001150773729397e-06, "loss": 0.1449, "step": 1752 }, { "epoch": 2.2885117493472587, "grad_norm": 3.2590668201446533, "learning_rate": 8.999882588957144e-06, "loss": 0.1219, "step": 1753 }, { "epoch": 2.289817232375979, "grad_norm": 2.785480260848999, "learning_rate": 8.998613689080648e-06, "loss": 0.1139, "step": 1754 }, { "epoch": 2.2911227154046996, "grad_norm": 1.8596402406692505, "learning_rate": 8.997344074326765e-06, "loss": 0.0784, "step": 1755 }, { "epoch": 2.2924281984334205, "grad_norm": 3.213247060775757, "learning_rate": 8.996073744922478e-06, "loss": 0.1238, "step": 1756 }, { "epoch": 2.293733681462141, "grad_norm": 2.7292730808258057, "learning_rate": 8.994802701094898e-06, "loss": 0.0836, "step": 1757 }, { "epoch": 2.2950391644908614, "grad_norm": 2.631747245788574, "learning_rate": 8.993530943071264e-06, "loss": 0.1017, "step": 1758 }, { "epoch": 2.2963446475195823, "grad_norm": 4.304591178894043, "learning_rate": 8.992258471078943e-06, "loss": 0.2056, "step": 1759 }, { "epoch": 2.297650130548303, "grad_norm": 3.8751659393310547, "learning_rate": 8.99098528534543e-06, "loss": 0.1811, "step": 1760 }, { "epoch": 2.2989556135770233, "grad_norm": 1.937435269355774, "learning_rate": 8.989711386098342e-06, "loss": 0.072, "step": 1761 }, { "epoch": 2.300261096605744, "grad_norm": 2.4831888675689697, "learning_rate": 8.988436773565437e-06, "loss": 0.0926, "step": 1762 }, { "epoch": 2.3015665796344646, "grad_norm": 1.6681996583938599, "learning_rate": 8.987161447974583e-06, "loss": 0.0631, "step": 1763 }, { "epoch": 2.3028720626631856, "grad_norm": 3.299065113067627, "learning_rate": 8.985885409553792e-06, "loss": 0.0904, "step": 1764 }, { "epoch": 2.304177545691906, "grad_norm": 2.6789989471435547, "learning_rate": 8.98460865853119e-06, "loss": 0.0881, "step": 1765 }, { "epoch": 2.3054830287206265, "grad_norm": 2.2594377994537354, "learning_rate": 8.983331195135038e-06, "loss": 0.0618, "step": 1766 }, { "epoch": 2.3067885117493474, "grad_norm": 3.576622247695923, "learning_rate": 8.982053019593723e-06, "loss": 0.1259, "step": 1767 }, { "epoch": 2.308093994778068, "grad_norm": 4.096035003662109, "learning_rate": 8.980774132135762e-06, "loss": 0.1587, "step": 1768 }, { "epoch": 2.3093994778067883, "grad_norm": 1.9867472648620605, "learning_rate": 8.97949453298979e-06, "loss": 0.0526, "step": 1769 }, { "epoch": 2.3107049608355092, "grad_norm": 4.265872001647949, "learning_rate": 8.97821422238458e-06, "loss": 0.1857, "step": 1770 }, { "epoch": 2.3120104438642297, "grad_norm": 1.892458200454712, "learning_rate": 8.976933200549025e-06, "loss": 0.0537, "step": 1771 }, { "epoch": 2.31331592689295, "grad_norm": 2.9152920246124268, "learning_rate": 8.97565146771215e-06, "loss": 0.1103, "step": 1772 }, { "epoch": 2.314621409921671, "grad_norm": 3.6105308532714844, "learning_rate": 8.974369024103102e-06, "loss": 0.1732, "step": 1773 }, { "epoch": 2.3159268929503916, "grad_norm": 3.735886573791504, "learning_rate": 8.973085869951161e-06, "loss": 0.1499, "step": 1774 }, { "epoch": 2.3172323759791125, "grad_norm": 4.896224021911621, "learning_rate": 8.971802005485729e-06, "loss": 0.2533, "step": 1775 }, { "epoch": 2.318537859007833, "grad_norm": 3.233171224594116, "learning_rate": 8.970517430936338e-06, "loss": 0.1219, "step": 1776 }, { "epoch": 2.3198433420365534, "grad_norm": 3.1865570545196533, "learning_rate": 8.969232146532646e-06, "loss": 0.1456, "step": 1777 }, { "epoch": 2.3211488250652743, "grad_norm": 2.4336297512054443, "learning_rate": 8.967946152504437e-06, "loss": 0.085, "step": 1778 }, { "epoch": 2.322454308093995, "grad_norm": 2.051788806915283, "learning_rate": 8.966659449081622e-06, "loss": 0.0824, "step": 1779 }, { "epoch": 2.3237597911227152, "grad_norm": 2.7760212421417236, "learning_rate": 8.96537203649424e-06, "loss": 0.1201, "step": 1780 }, { "epoch": 2.325065274151436, "grad_norm": 2.570463180541992, "learning_rate": 8.964083914972458e-06, "loss": 0.1002, "step": 1781 }, { "epoch": 2.3263707571801566, "grad_norm": 1.6691845655441284, "learning_rate": 8.96279508474657e-06, "loss": 0.0672, "step": 1782 }, { "epoch": 2.327676240208877, "grad_norm": 2.477091073989868, "learning_rate": 8.961505546046987e-06, "loss": 0.1025, "step": 1783 }, { "epoch": 2.328981723237598, "grad_norm": 3.7081151008605957, "learning_rate": 8.960215299104263e-06, "loss": 0.171, "step": 1784 }, { "epoch": 2.3302872062663185, "grad_norm": 2.4688382148742676, "learning_rate": 8.958924344149062e-06, "loss": 0.0809, "step": 1785 }, { "epoch": 2.3315926892950394, "grad_norm": 2.640355348587036, "learning_rate": 8.95763268141219e-06, "loss": 0.1004, "step": 1786 }, { "epoch": 2.33289817232376, "grad_norm": 1.8581123352050781, "learning_rate": 8.956340311124567e-06, "loss": 0.0629, "step": 1787 }, { "epoch": 2.3342036553524803, "grad_norm": 2.7072246074676514, "learning_rate": 8.955047233517248e-06, "loss": 0.1403, "step": 1788 }, { "epoch": 2.335509138381201, "grad_norm": 1.8944941759109497, "learning_rate": 8.953753448821409e-06, "loss": 0.0987, "step": 1789 }, { "epoch": 2.3368146214099217, "grad_norm": 2.7462751865386963, "learning_rate": 8.952458957268356e-06, "loss": 0.0961, "step": 1790 }, { "epoch": 2.338120104438642, "grad_norm": 4.539660453796387, "learning_rate": 8.951163759089518e-06, "loss": 0.2035, "step": 1791 }, { "epoch": 2.339425587467363, "grad_norm": 1.88591468334198, "learning_rate": 8.949867854516451e-06, "loss": 0.0691, "step": 1792 }, { "epoch": 2.3407310704960835, "grad_norm": 2.0693674087524414, "learning_rate": 8.948571243780846e-06, "loss": 0.0617, "step": 1793 }, { "epoch": 2.342036553524804, "grad_norm": 3.8292291164398193, "learning_rate": 8.947273927114504e-06, "loss": 0.2062, "step": 1794 }, { "epoch": 2.343342036553525, "grad_norm": 2.093806028366089, "learning_rate": 8.945975904749366e-06, "loss": 0.0633, "step": 1795 }, { "epoch": 2.3446475195822454, "grad_norm": 1.7689400911331177, "learning_rate": 8.944677176917493e-06, "loss": 0.0631, "step": 1796 }, { "epoch": 2.3459530026109663, "grad_norm": 3.4390289783477783, "learning_rate": 8.943377743851072e-06, "loss": 0.1598, "step": 1797 }, { "epoch": 2.3472584856396868, "grad_norm": 2.2308132648468018, "learning_rate": 8.942077605782418e-06, "loss": 0.0896, "step": 1798 }, { "epoch": 2.348563968668407, "grad_norm": 1.8903899192810059, "learning_rate": 8.940776762943972e-06, "loss": 0.0809, "step": 1799 }, { "epoch": 2.349869451697128, "grad_norm": 4.17438268661499, "learning_rate": 8.9394752155683e-06, "loss": 0.1736, "step": 1800 }, { "epoch": 2.3511749347258486, "grad_norm": 3.421846389770508, "learning_rate": 8.938172963888095e-06, "loss": 0.1791, "step": 1801 }, { "epoch": 2.352480417754569, "grad_norm": 3.3672759532928467, "learning_rate": 8.936870008136174e-06, "loss": 0.1252, "step": 1802 }, { "epoch": 2.35378590078329, "grad_norm": 3.014657497406006, "learning_rate": 8.935566348545482e-06, "loss": 0.1203, "step": 1803 }, { "epoch": 2.3550913838120104, "grad_norm": 2.1459901332855225, "learning_rate": 8.93426198534909e-06, "loss": 0.0625, "step": 1804 }, { "epoch": 2.356396866840731, "grad_norm": 3.6439146995544434, "learning_rate": 8.93295691878019e-06, "loss": 0.1641, "step": 1805 }, { "epoch": 2.357702349869452, "grad_norm": 2.8594071865081787, "learning_rate": 8.931651149072109e-06, "loss": 0.1194, "step": 1806 }, { "epoch": 2.3590078328981723, "grad_norm": 1.5997254848480225, "learning_rate": 8.93034467645829e-06, "loss": 0.0652, "step": 1807 }, { "epoch": 2.360313315926893, "grad_norm": 2.4087367057800293, "learning_rate": 8.929037501172307e-06, "loss": 0.0864, "step": 1808 }, { "epoch": 2.3616187989556137, "grad_norm": 3.4050991535186768, "learning_rate": 8.927729623447859e-06, "loss": 0.1287, "step": 1809 }, { "epoch": 2.362924281984334, "grad_norm": 3.38580060005188, "learning_rate": 8.926421043518768e-06, "loss": 0.136, "step": 1810 }, { "epoch": 2.3642297650130546, "grad_norm": 3.8482627868652344, "learning_rate": 8.925111761618987e-06, "loss": 0.1459, "step": 1811 }, { "epoch": 2.3655352480417755, "grad_norm": 3.8344602584838867, "learning_rate": 8.923801777982588e-06, "loss": 0.187, "step": 1812 }, { "epoch": 2.366840731070496, "grad_norm": 2.3408279418945312, "learning_rate": 8.922491092843774e-06, "loss": 0.0868, "step": 1813 }, { "epoch": 2.368146214099217, "grad_norm": 2.208613157272339, "learning_rate": 8.921179706436872e-06, "loss": 0.0893, "step": 1814 }, { "epoch": 2.3694516971279374, "grad_norm": 2.6018736362457275, "learning_rate": 8.91986761899633e-06, "loss": 0.1079, "step": 1815 }, { "epoch": 2.370757180156658, "grad_norm": 4.205713272094727, "learning_rate": 8.918554830756725e-06, "loss": 0.1977, "step": 1816 }, { "epoch": 2.3720626631853787, "grad_norm": 3.466052293777466, "learning_rate": 8.917241341952761e-06, "loss": 0.1306, "step": 1817 }, { "epoch": 2.373368146214099, "grad_norm": 2.4664011001586914, "learning_rate": 8.915927152819266e-06, "loss": 0.0813, "step": 1818 }, { "epoch": 2.3746736292428197, "grad_norm": 3.417567253112793, "learning_rate": 8.914612263591189e-06, "loss": 0.1423, "step": 1819 }, { "epoch": 2.3759791122715406, "grad_norm": 3.3542003631591797, "learning_rate": 8.91329667450361e-06, "loss": 0.171, "step": 1820 }, { "epoch": 2.377284595300261, "grad_norm": 1.8347440958023071, "learning_rate": 8.911980385791731e-06, "loss": 0.0642, "step": 1821 }, { "epoch": 2.3785900783289815, "grad_norm": 2.5314390659332275, "learning_rate": 8.91066339769088e-06, "loss": 0.077, "step": 1822 }, { "epoch": 2.3798955613577024, "grad_norm": 2.9698402881622314, "learning_rate": 8.90934571043651e-06, "loss": 0.0932, "step": 1823 }, { "epoch": 2.381201044386423, "grad_norm": 4.755293846130371, "learning_rate": 8.9080273242642e-06, "loss": 0.2236, "step": 1824 }, { "epoch": 2.382506527415144, "grad_norm": 2.613734722137451, "learning_rate": 8.906708239409649e-06, "loss": 0.0965, "step": 1825 }, { "epoch": 2.3838120104438643, "grad_norm": 3.6638035774230957, "learning_rate": 8.90538845610869e-06, "loss": 0.1935, "step": 1826 }, { "epoch": 2.3851174934725847, "grad_norm": 2.2379989624023438, "learning_rate": 8.904067974597271e-06, "loss": 0.0724, "step": 1827 }, { "epoch": 2.3864229765013056, "grad_norm": 3.640486001968384, "learning_rate": 8.902746795111472e-06, "loss": 0.1038, "step": 1828 }, { "epoch": 2.387728459530026, "grad_norm": 2.291300058364868, "learning_rate": 8.901424917887494e-06, "loss": 0.0897, "step": 1829 }, { "epoch": 2.3890339425587466, "grad_norm": 3.499647855758667, "learning_rate": 8.900102343161665e-06, "loss": 0.1379, "step": 1830 }, { "epoch": 2.3903394255874675, "grad_norm": 2.0270557403564453, "learning_rate": 8.898779071170435e-06, "loss": 0.0742, "step": 1831 }, { "epoch": 2.391644908616188, "grad_norm": 2.0482845306396484, "learning_rate": 8.897455102150384e-06, "loss": 0.0635, "step": 1832 }, { "epoch": 2.3929503916449084, "grad_norm": 2.0639610290527344, "learning_rate": 8.896130436338208e-06, "loss": 0.0627, "step": 1833 }, { "epoch": 2.3942558746736293, "grad_norm": 3.72035813331604, "learning_rate": 8.894805073970735e-06, "loss": 0.2154, "step": 1834 }, { "epoch": 2.39556135770235, "grad_norm": 3.1126859188079834, "learning_rate": 8.893479015284916e-06, "loss": 0.1822, "step": 1835 }, { "epoch": 2.3968668407310707, "grad_norm": 3.7393946647644043, "learning_rate": 8.892152260517824e-06, "loss": 0.1924, "step": 1836 }, { "epoch": 2.398172323759791, "grad_norm": 1.8886784315109253, "learning_rate": 8.890824809906657e-06, "loss": 0.0857, "step": 1837 }, { "epoch": 2.3994778067885116, "grad_norm": 2.359377861022949, "learning_rate": 8.88949666368874e-06, "loss": 0.0966, "step": 1838 }, { "epoch": 2.4007832898172325, "grad_norm": 3.5547661781311035, "learning_rate": 8.88816782210152e-06, "loss": 0.1775, "step": 1839 }, { "epoch": 2.402088772845953, "grad_norm": 3.4607057571411133, "learning_rate": 8.88683828538257e-06, "loss": 0.1739, "step": 1840 }, { "epoch": 2.4033942558746735, "grad_norm": 1.6922246217727661, "learning_rate": 8.885508053769586e-06, "loss": 0.0679, "step": 1841 }, { "epoch": 2.4046997389033944, "grad_norm": 2.1472580432891846, "learning_rate": 8.884177127500386e-06, "loss": 0.0842, "step": 1842 }, { "epoch": 2.406005221932115, "grad_norm": 5.63610315322876, "learning_rate": 8.88284550681292e-06, "loss": 0.3405, "step": 1843 }, { "epoch": 2.4073107049608353, "grad_norm": 2.7999014854431152, "learning_rate": 8.881513191945252e-06, "loss": 0.1008, "step": 1844 }, { "epoch": 2.4086161879895562, "grad_norm": 4.090418815612793, "learning_rate": 8.880180183135578e-06, "loss": 0.1537, "step": 1845 }, { "epoch": 2.4099216710182767, "grad_norm": 2.5789332389831543, "learning_rate": 8.878846480622211e-06, "loss": 0.0996, "step": 1846 }, { "epoch": 2.4112271540469976, "grad_norm": 3.586855411529541, "learning_rate": 8.8775120846436e-06, "loss": 0.1868, "step": 1847 }, { "epoch": 2.412532637075718, "grad_norm": 2.8669939041137695, "learning_rate": 8.876176995438301e-06, "loss": 0.0771, "step": 1848 }, { "epoch": 2.4138381201044385, "grad_norm": 2.2880988121032715, "learning_rate": 8.874841213245009e-06, "loss": 0.0935, "step": 1849 }, { "epoch": 2.4151436031331595, "grad_norm": 2.6612813472747803, "learning_rate": 8.873504738302535e-06, "loss": 0.1165, "step": 1850 }, { "epoch": 2.41644908616188, "grad_norm": 2.2290737628936768, "learning_rate": 8.872167570849814e-06, "loss": 0.0694, "step": 1851 }, { "epoch": 2.4177545691906004, "grad_norm": 3.2927026748657227, "learning_rate": 8.870829711125909e-06, "loss": 0.1413, "step": 1852 }, { "epoch": 2.4190600522193213, "grad_norm": 2.47290301322937, "learning_rate": 8.869491159370004e-06, "loss": 0.0869, "step": 1853 }, { "epoch": 2.4203655352480418, "grad_norm": 2.507963180541992, "learning_rate": 8.868151915821406e-06, "loss": 0.1077, "step": 1854 }, { "epoch": 2.4216710182767622, "grad_norm": 2.4924862384796143, "learning_rate": 8.866811980719547e-06, "loss": 0.0829, "step": 1855 }, { "epoch": 2.422976501305483, "grad_norm": 3.0459964275360107, "learning_rate": 8.865471354303982e-06, "loss": 0.149, "step": 1856 }, { "epoch": 2.4242819843342036, "grad_norm": 2.7658417224884033, "learning_rate": 8.864130036814392e-06, "loss": 0.0986, "step": 1857 }, { "epoch": 2.4255874673629245, "grad_norm": 2.953183889389038, "learning_rate": 8.862788028490578e-06, "loss": 0.069, "step": 1858 }, { "epoch": 2.426892950391645, "grad_norm": 2.9738433361053467, "learning_rate": 8.861445329572466e-06, "loss": 0.0974, "step": 1859 }, { "epoch": 2.4281984334203655, "grad_norm": 3.7113394737243652, "learning_rate": 8.860101940300104e-06, "loss": 0.114, "step": 1860 }, { "epoch": 2.429503916449086, "grad_norm": 3.3439762592315674, "learning_rate": 8.858757860913669e-06, "loss": 0.1345, "step": 1861 }, { "epoch": 2.430809399477807, "grad_norm": 3.1887123584747314, "learning_rate": 8.857413091653452e-06, "loss": 0.1359, "step": 1862 }, { "epoch": 2.4321148825065273, "grad_norm": 4.123287200927734, "learning_rate": 8.856067632759878e-06, "loss": 0.1515, "step": 1863 }, { "epoch": 2.433420365535248, "grad_norm": 3.361614227294922, "learning_rate": 8.854721484473485e-06, "loss": 0.1511, "step": 1864 }, { "epoch": 2.4347258485639687, "grad_norm": 1.8422484397888184, "learning_rate": 8.853374647034942e-06, "loss": 0.0633, "step": 1865 }, { "epoch": 2.436031331592689, "grad_norm": 3.070155143737793, "learning_rate": 8.852027120685037e-06, "loss": 0.1652, "step": 1866 }, { "epoch": 2.43733681462141, "grad_norm": 4.731071472167969, "learning_rate": 8.850678905664684e-06, "loss": 0.2396, "step": 1867 }, { "epoch": 2.4386422976501305, "grad_norm": 3.817735433578491, "learning_rate": 8.849330002214919e-06, "loss": 0.1931, "step": 1868 }, { "epoch": 2.439947780678851, "grad_norm": 2.435957193374634, "learning_rate": 8.847980410576898e-06, "loss": 0.0848, "step": 1869 }, { "epoch": 2.441253263707572, "grad_norm": 4.452448844909668, "learning_rate": 8.846630130991905e-06, "loss": 0.215, "step": 1870 }, { "epoch": 2.4425587467362924, "grad_norm": 2.751843214035034, "learning_rate": 8.845279163701344e-06, "loss": 0.0765, "step": 1871 }, { "epoch": 2.443864229765013, "grad_norm": 3.21321439743042, "learning_rate": 8.843927508946744e-06, "loss": 0.1513, "step": 1872 }, { "epoch": 2.4451697127937337, "grad_norm": 3.680206298828125, "learning_rate": 8.842575166969753e-06, "loss": 0.1683, "step": 1873 }, { "epoch": 2.446475195822454, "grad_norm": 3.255981206893921, "learning_rate": 8.841222138012146e-06, "loss": 0.1411, "step": 1874 }, { "epoch": 2.447780678851175, "grad_norm": 1.5528477430343628, "learning_rate": 8.839868422315819e-06, "loss": 0.0511, "step": 1875 }, { "epoch": 2.4490861618798956, "grad_norm": 1.6969887018203735, "learning_rate": 8.83851402012279e-06, "loss": 0.0556, "step": 1876 }, { "epoch": 2.450391644908616, "grad_norm": 2.9925005435943604, "learning_rate": 8.837158931675206e-06, "loss": 0.1424, "step": 1877 }, { "epoch": 2.451697127937337, "grad_norm": 2.706150531768799, "learning_rate": 8.835803157215323e-06, "loss": 0.1104, "step": 1878 }, { "epoch": 2.4530026109660574, "grad_norm": 2.3051159381866455, "learning_rate": 8.834446696985535e-06, "loss": 0.0739, "step": 1879 }, { "epoch": 2.454308093994778, "grad_norm": 2.0888495445251465, "learning_rate": 8.833089551228349e-06, "loss": 0.0907, "step": 1880 }, { "epoch": 2.455613577023499, "grad_norm": 2.335237503051758, "learning_rate": 8.831731720186398e-06, "loss": 0.0802, "step": 1881 }, { "epoch": 2.4569190600522193, "grad_norm": 3.9661006927490234, "learning_rate": 8.830373204102437e-06, "loss": 0.1776, "step": 1882 }, { "epoch": 2.4582245430809397, "grad_norm": 3.1612586975097656, "learning_rate": 8.829014003219343e-06, "loss": 0.1007, "step": 1883 }, { "epoch": 2.4595300261096606, "grad_norm": 1.8223323822021484, "learning_rate": 8.827654117780114e-06, "loss": 0.0686, "step": 1884 }, { "epoch": 2.460835509138381, "grad_norm": 3.521597146987915, "learning_rate": 8.826293548027874e-06, "loss": 0.1603, "step": 1885 }, { "epoch": 2.462140992167102, "grad_norm": 3.8714842796325684, "learning_rate": 8.824932294205868e-06, "loss": 0.2073, "step": 1886 }, { "epoch": 2.4634464751958225, "grad_norm": 3.082098960876465, "learning_rate": 8.82357035655746e-06, "loss": 0.1199, "step": 1887 }, { "epoch": 2.464751958224543, "grad_norm": 3.302156448364258, "learning_rate": 8.822207735326143e-06, "loss": 0.1104, "step": 1888 }, { "epoch": 2.466057441253264, "grad_norm": 3.4306554794311523, "learning_rate": 8.820844430755525e-06, "loss": 0.147, "step": 1889 }, { "epoch": 2.4673629242819843, "grad_norm": 2.030278205871582, "learning_rate": 8.81948044308934e-06, "loss": 0.0634, "step": 1890 }, { "epoch": 2.468668407310705, "grad_norm": 1.9212332963943481, "learning_rate": 8.818115772571444e-06, "loss": 0.0578, "step": 1891 }, { "epoch": 2.4699738903394257, "grad_norm": 3.3536689281463623, "learning_rate": 8.816750419445813e-06, "loss": 0.112, "step": 1892 }, { "epoch": 2.471279373368146, "grad_norm": 2.3681182861328125, "learning_rate": 8.815384383956548e-06, "loss": 0.084, "step": 1893 }, { "epoch": 2.4725848563968666, "grad_norm": 2.513282299041748, "learning_rate": 8.814017666347874e-06, "loss": 0.1217, "step": 1894 }, { "epoch": 2.4738903394255876, "grad_norm": 3.3103480339050293, "learning_rate": 8.81265026686413e-06, "loss": 0.1564, "step": 1895 }, { "epoch": 2.475195822454308, "grad_norm": 2.8021674156188965, "learning_rate": 8.81128218574978e-06, "loss": 0.1279, "step": 1896 }, { "epoch": 2.476501305483029, "grad_norm": 2.993950366973877, "learning_rate": 8.809913423249417e-06, "loss": 0.0977, "step": 1897 }, { "epoch": 2.4778067885117494, "grad_norm": 2.999570369720459, "learning_rate": 8.808543979607745e-06, "loss": 0.1124, "step": 1898 }, { "epoch": 2.47911227154047, "grad_norm": 2.3185958862304688, "learning_rate": 8.807173855069598e-06, "loss": 0.0932, "step": 1899 }, { "epoch": 2.480417754569191, "grad_norm": 2.337383985519409, "learning_rate": 8.805803049879928e-06, "loss": 0.0962, "step": 1900 }, { "epoch": 2.4817232375979112, "grad_norm": 2.765310764312744, "learning_rate": 8.804431564283808e-06, "loss": 0.1351, "step": 1901 }, { "epoch": 2.4830287206266317, "grad_norm": 2.055757999420166, "learning_rate": 8.803059398526437e-06, "loss": 0.0621, "step": 1902 }, { "epoch": 2.4843342036553526, "grad_norm": 3.5761818885803223, "learning_rate": 8.80168655285313e-06, "loss": 0.1287, "step": 1903 }, { "epoch": 2.485639686684073, "grad_norm": 3.6500232219696045, "learning_rate": 8.800313027509326e-06, "loss": 0.1542, "step": 1904 }, { "epoch": 2.4869451697127936, "grad_norm": 1.8909883499145508, "learning_rate": 8.798938822740588e-06, "loss": 0.059, "step": 1905 }, { "epoch": 2.4882506527415145, "grad_norm": 3.6751198768615723, "learning_rate": 8.797563938792595e-06, "loss": 0.1811, "step": 1906 }, { "epoch": 2.489556135770235, "grad_norm": 3.2179906368255615, "learning_rate": 8.796188375911153e-06, "loss": 0.1291, "step": 1907 }, { "epoch": 2.490861618798956, "grad_norm": 3.515162944793701, "learning_rate": 8.794812134342187e-06, "loss": 0.1173, "step": 1908 }, { "epoch": 2.4921671018276763, "grad_norm": 1.9659826755523682, "learning_rate": 8.79343521433174e-06, "loss": 0.0699, "step": 1909 }, { "epoch": 2.493472584856397, "grad_norm": 2.219184160232544, "learning_rate": 8.792057616125985e-06, "loss": 0.0708, "step": 1910 }, { "epoch": 2.4947780678851172, "grad_norm": 3.5134594440460205, "learning_rate": 8.790679339971205e-06, "loss": 0.1411, "step": 1911 }, { "epoch": 2.496083550913838, "grad_norm": 3.019277811050415, "learning_rate": 8.789300386113813e-06, "loss": 0.1208, "step": 1912 }, { "epoch": 2.4973890339425586, "grad_norm": 2.0646491050720215, "learning_rate": 8.78792075480034e-06, "loss": 0.0823, "step": 1913 }, { "epoch": 2.4986945169712795, "grad_norm": 1.7719666957855225, "learning_rate": 8.786540446277439e-06, "loss": 0.0542, "step": 1914 }, { "epoch": 2.5, "grad_norm": 3.4907751083374023, "learning_rate": 8.785159460791883e-06, "loss": 0.1474, "step": 1915 }, { "epoch": 2.5013054830287205, "grad_norm": 2.219792604446411, "learning_rate": 8.783777798590565e-06, "loss": 0.0566, "step": 1916 }, { "epoch": 2.5026109660574414, "grad_norm": 3.183744192123413, "learning_rate": 8.782395459920502e-06, "loss": 0.1256, "step": 1917 }, { "epoch": 2.503916449086162, "grad_norm": 2.557450771331787, "learning_rate": 8.78101244502883e-06, "loss": 0.1032, "step": 1918 }, { "epoch": 2.5052219321148828, "grad_norm": 1.3601317405700684, "learning_rate": 8.779628754162804e-06, "loss": 0.0467, "step": 1919 }, { "epoch": 2.506527415143603, "grad_norm": 1.814314842224121, "learning_rate": 8.778244387569805e-06, "loss": 0.0643, "step": 1920 }, { "epoch": 2.5078328981723237, "grad_norm": 2.4726481437683105, "learning_rate": 8.776859345497331e-06, "loss": 0.0921, "step": 1921 }, { "epoch": 2.509138381201044, "grad_norm": 3.4266915321350098, "learning_rate": 8.775473628193e-06, "loss": 0.0888, "step": 1922 }, { "epoch": 2.510443864229765, "grad_norm": 2.9883856773376465, "learning_rate": 8.774087235904557e-06, "loss": 0.123, "step": 1923 }, { "epoch": 2.5117493472584855, "grad_norm": 2.269253730773926, "learning_rate": 8.772700168879857e-06, "loss": 0.071, "step": 1924 }, { "epoch": 2.5130548302872064, "grad_norm": 2.945098400115967, "learning_rate": 8.771312427366885e-06, "loss": 0.1047, "step": 1925 }, { "epoch": 2.514360313315927, "grad_norm": 1.635989785194397, "learning_rate": 8.769924011613744e-06, "loss": 0.0569, "step": 1926 }, { "epoch": 2.5156657963446474, "grad_norm": 2.1300253868103027, "learning_rate": 8.768534921868654e-06, "loss": 0.092, "step": 1927 }, { "epoch": 2.5169712793733683, "grad_norm": 2.8139727115631104, "learning_rate": 8.767145158379961e-06, "loss": 0.1182, "step": 1928 }, { "epoch": 2.5182767624020888, "grad_norm": 2.6679182052612305, "learning_rate": 8.765754721396127e-06, "loss": 0.1129, "step": 1929 }, { "epoch": 2.5195822454308097, "grad_norm": 1.3458172082901, "learning_rate": 8.764363611165736e-06, "loss": 0.0501, "step": 1930 }, { "epoch": 2.52088772845953, "grad_norm": 2.971325397491455, "learning_rate": 8.762971827937494e-06, "loss": 0.1294, "step": 1931 }, { "epoch": 2.5221932114882506, "grad_norm": 2.536311149597168, "learning_rate": 8.761579371960224e-06, "loss": 0.0849, "step": 1932 }, { "epoch": 2.523498694516971, "grad_norm": 2.7901575565338135, "learning_rate": 8.760186243482873e-06, "loss": 0.1017, "step": 1933 }, { "epoch": 2.524804177545692, "grad_norm": 3.2359251976013184, "learning_rate": 8.758792442754508e-06, "loss": 0.1325, "step": 1934 }, { "epoch": 2.5261096605744124, "grad_norm": 3.6457395553588867, "learning_rate": 8.757397970024308e-06, "loss": 0.1374, "step": 1935 }, { "epoch": 2.5274151436031334, "grad_norm": 2.150097608566284, "learning_rate": 8.756002825541583e-06, "loss": 0.0743, "step": 1936 }, { "epoch": 2.528720626631854, "grad_norm": 3.173081874847412, "learning_rate": 8.75460700955576e-06, "loss": 0.0843, "step": 1937 }, { "epoch": 2.5300261096605743, "grad_norm": 2.1626477241516113, "learning_rate": 8.753210522316382e-06, "loss": 0.0741, "step": 1938 }, { "epoch": 2.531331592689295, "grad_norm": 2.3605024814605713, "learning_rate": 8.751813364073116e-06, "loss": 0.0729, "step": 1939 }, { "epoch": 2.5326370757180157, "grad_norm": 2.634507656097412, "learning_rate": 8.750415535075749e-06, "loss": 0.0817, "step": 1940 }, { "epoch": 2.533942558746736, "grad_norm": 1.965801477432251, "learning_rate": 8.749017035574183e-06, "loss": 0.0603, "step": 1941 }, { "epoch": 2.535248041775457, "grad_norm": 1.9986258745193481, "learning_rate": 8.74761786581845e-06, "loss": 0.0737, "step": 1942 }, { "epoch": 2.5365535248041775, "grad_norm": 2.0469040870666504, "learning_rate": 8.746218026058688e-06, "loss": 0.0674, "step": 1943 }, { "epoch": 2.537859007832898, "grad_norm": 3.329195261001587, "learning_rate": 8.744817516545167e-06, "loss": 0.1054, "step": 1944 }, { "epoch": 2.539164490861619, "grad_norm": 1.7871750593185425, "learning_rate": 8.74341633752827e-06, "loss": 0.064, "step": 1945 }, { "epoch": 2.5404699738903394, "grad_norm": 2.6456081867218018, "learning_rate": 8.7420144892585e-06, "loss": 0.0925, "step": 1946 }, { "epoch": 2.5417754569190603, "grad_norm": 3.3387372493743896, "learning_rate": 8.740611971986488e-06, "loss": 0.1386, "step": 1947 }, { "epoch": 2.5430809399477807, "grad_norm": 1.947710633277893, "learning_rate": 8.73920878596297e-06, "loss": 0.0739, "step": 1948 }, { "epoch": 2.544386422976501, "grad_norm": 3.2871742248535156, "learning_rate": 8.737804931438815e-06, "loss": 0.1159, "step": 1949 }, { "epoch": 2.5456919060052217, "grad_norm": 2.408215284347534, "learning_rate": 8.736400408665003e-06, "loss": 0.0958, "step": 1950 }, { "epoch": 2.5469973890339426, "grad_norm": 2.576390266418457, "learning_rate": 8.734995217892636e-06, "loss": 0.0993, "step": 1951 }, { "epoch": 2.548302872062663, "grad_norm": 2.300138235092163, "learning_rate": 8.733589359372937e-06, "loss": 0.0719, "step": 1952 }, { "epoch": 2.549608355091384, "grad_norm": 2.7990708351135254, "learning_rate": 8.732182833357248e-06, "loss": 0.0953, "step": 1953 }, { "epoch": 2.5509138381201044, "grad_norm": 3.8019776344299316, "learning_rate": 8.73077564009703e-06, "loss": 0.1332, "step": 1954 }, { "epoch": 2.552219321148825, "grad_norm": 2.2679004669189453, "learning_rate": 8.729367779843857e-06, "loss": 0.0858, "step": 1955 }, { "epoch": 2.553524804177546, "grad_norm": 2.701777219772339, "learning_rate": 8.727959252849438e-06, "loss": 0.1018, "step": 1956 }, { "epoch": 2.5548302872062663, "grad_norm": 2.337507724761963, "learning_rate": 8.726550059365583e-06, "loss": 0.0688, "step": 1957 }, { "epoch": 2.556135770234987, "grad_norm": 3.0712897777557373, "learning_rate": 8.725140199644232e-06, "loss": 0.1027, "step": 1958 }, { "epoch": 2.5574412532637076, "grad_norm": 2.7051119804382324, "learning_rate": 8.723729673937445e-06, "loss": 0.1303, "step": 1959 }, { "epoch": 2.558746736292428, "grad_norm": 2.8558802604675293, "learning_rate": 8.722318482497392e-06, "loss": 0.1221, "step": 1960 }, { "epoch": 2.5600522193211486, "grad_norm": 2.3837850093841553, "learning_rate": 8.72090662557637e-06, "loss": 0.0609, "step": 1961 }, { "epoch": 2.5613577023498695, "grad_norm": 4.080503463745117, "learning_rate": 8.719494103426796e-06, "loss": 0.1519, "step": 1962 }, { "epoch": 2.56266318537859, "grad_norm": 2.0113730430603027, "learning_rate": 8.718080916301197e-06, "loss": 0.0786, "step": 1963 }, { "epoch": 2.563968668407311, "grad_norm": 3.098862886428833, "learning_rate": 8.716667064452227e-06, "loss": 0.1175, "step": 1964 }, { "epoch": 2.5652741514360313, "grad_norm": 2.4988696575164795, "learning_rate": 8.715252548132655e-06, "loss": 0.0819, "step": 1965 }, { "epoch": 2.566579634464752, "grad_norm": 2.392270803451538, "learning_rate": 8.713837367595374e-06, "loss": 0.0731, "step": 1966 }, { "epoch": 2.5678851174934727, "grad_norm": 2.3560080528259277, "learning_rate": 8.712421523093388e-06, "loss": 0.0727, "step": 1967 }, { "epoch": 2.569190600522193, "grad_norm": 3.212507963180542, "learning_rate": 8.711005014879825e-06, "loss": 0.1215, "step": 1968 }, { "epoch": 2.570496083550914, "grad_norm": 1.3185524940490723, "learning_rate": 8.70958784320793e-06, "loss": 0.0581, "step": 1969 }, { "epoch": 2.5718015665796345, "grad_norm": 1.836456537246704, "learning_rate": 8.708170008331066e-06, "loss": 0.0648, "step": 1970 }, { "epoch": 2.573107049608355, "grad_norm": 2.8211543560028076, "learning_rate": 8.706751510502716e-06, "loss": 0.0885, "step": 1971 }, { "epoch": 2.5744125326370755, "grad_norm": 3.4130330085754395, "learning_rate": 8.70533234997648e-06, "loss": 0.1213, "step": 1972 }, { "epoch": 2.5757180156657964, "grad_norm": 1.372713327407837, "learning_rate": 8.70391252700608e-06, "loss": 0.0522, "step": 1973 }, { "epoch": 2.577023498694517, "grad_norm": 3.627730131149292, "learning_rate": 8.70249204184535e-06, "loss": 0.1392, "step": 1974 }, { "epoch": 2.5783289817232378, "grad_norm": 2.766653537750244, "learning_rate": 8.70107089474825e-06, "loss": 0.1113, "step": 1975 }, { "epoch": 2.5796344647519582, "grad_norm": 3.8212313652038574, "learning_rate": 8.699649085968853e-06, "loss": 0.1439, "step": 1976 }, { "epoch": 2.5809399477806787, "grad_norm": 2.123952865600586, "learning_rate": 8.698226615761352e-06, "loss": 0.0679, "step": 1977 }, { "epoch": 2.5822454308093996, "grad_norm": 2.548046827316284, "learning_rate": 8.696803484380056e-06, "loss": 0.075, "step": 1978 }, { "epoch": 2.58355091383812, "grad_norm": 3.445495367050171, "learning_rate": 8.695379692079397e-06, "loss": 0.131, "step": 1979 }, { "epoch": 2.584856396866841, "grad_norm": 2.92989444732666, "learning_rate": 8.693955239113922e-06, "loss": 0.1472, "step": 1980 }, { "epoch": 2.5861618798955615, "grad_norm": 2.5343849658966064, "learning_rate": 8.692530125738294e-06, "loss": 0.0799, "step": 1981 }, { "epoch": 2.587467362924282, "grad_norm": 2.5958268642425537, "learning_rate": 8.6911043522073e-06, "loss": 0.0823, "step": 1982 }, { "epoch": 2.5887728459530024, "grad_norm": 3.627124786376953, "learning_rate": 8.689677918775841e-06, "loss": 0.1572, "step": 1983 }, { "epoch": 2.5900783289817233, "grad_norm": 1.6309524774551392, "learning_rate": 8.688250825698936e-06, "loss": 0.0633, "step": 1984 }, { "epoch": 2.5913838120104438, "grad_norm": 3.285534381866455, "learning_rate": 8.68682307323172e-06, "loss": 0.1334, "step": 1985 }, { "epoch": 2.5926892950391647, "grad_norm": 2.0799856185913086, "learning_rate": 8.685394661629453e-06, "loss": 0.0706, "step": 1986 }, { "epoch": 2.593994778067885, "grad_norm": 1.9476710557937622, "learning_rate": 8.683965591147507e-06, "loss": 0.0565, "step": 1987 }, { "epoch": 2.5953002610966056, "grad_norm": 2.954263925552368, "learning_rate": 8.682535862041373e-06, "loss": 0.0941, "step": 1988 }, { "epoch": 2.5966057441253265, "grad_norm": 5.054914474487305, "learning_rate": 8.681105474566657e-06, "loss": 0.2563, "step": 1989 }, { "epoch": 2.597911227154047, "grad_norm": 1.7596503496170044, "learning_rate": 8.679674428979087e-06, "loss": 0.0541, "step": 1990 }, { "epoch": 2.599216710182768, "grad_norm": 3.4140641689300537, "learning_rate": 8.678242725534512e-06, "loss": 0.1359, "step": 1991 }, { "epoch": 2.6005221932114884, "grad_norm": 3.531031608581543, "learning_rate": 8.676810364488888e-06, "loss": 0.1464, "step": 1992 }, { "epoch": 2.601827676240209, "grad_norm": 2.4681339263916016, "learning_rate": 8.675377346098295e-06, "loss": 0.067, "step": 1993 }, { "epoch": 2.6031331592689293, "grad_norm": 3.0902841091156006, "learning_rate": 8.67394367061893e-06, "loss": 0.1224, "step": 1994 }, { "epoch": 2.60443864229765, "grad_norm": 2.477736234664917, "learning_rate": 8.67250933830711e-06, "loss": 0.0881, "step": 1995 }, { "epoch": 2.6057441253263707, "grad_norm": 2.944396734237671, "learning_rate": 8.671074349419266e-06, "loss": 0.1097, "step": 1996 }, { "epoch": 2.6070496083550916, "grad_norm": 3.362340211868286, "learning_rate": 8.669638704211947e-06, "loss": 0.1674, "step": 1997 }, { "epoch": 2.608355091383812, "grad_norm": 1.3778170347213745, "learning_rate": 8.668202402941816e-06, "loss": 0.048, "step": 1998 }, { "epoch": 2.6096605744125325, "grad_norm": 2.4940762519836426, "learning_rate": 8.666765445865662e-06, "loss": 0.0774, "step": 1999 }, { "epoch": 2.6109660574412534, "grad_norm": 3.175813913345337, "learning_rate": 8.665327833240383e-06, "loss": 0.1229, "step": 2000 }, { "epoch": 2.612271540469974, "grad_norm": 3.230027437210083, "learning_rate": 8.663889565322998e-06, "loss": 0.1142, "step": 2001 }, { "epoch": 2.6135770234986944, "grad_norm": 1.6820976734161377, "learning_rate": 8.662450642370645e-06, "loss": 0.0525, "step": 2002 }, { "epoch": 2.6148825065274153, "grad_norm": 3.5229852199554443, "learning_rate": 8.661011064640571e-06, "loss": 0.1244, "step": 2003 }, { "epoch": 2.6161879895561357, "grad_norm": 2.577991247177124, "learning_rate": 8.65957083239015e-06, "loss": 0.0853, "step": 2004 }, { "epoch": 2.617493472584856, "grad_norm": 1.9553124904632568, "learning_rate": 8.658129945876868e-06, "loss": 0.0664, "step": 2005 }, { "epoch": 2.618798955613577, "grad_norm": 2.848902702331543, "learning_rate": 8.656688405358327e-06, "loss": 0.0905, "step": 2006 }, { "epoch": 2.6201044386422976, "grad_norm": 1.5091824531555176, "learning_rate": 8.655246211092248e-06, "loss": 0.0686, "step": 2007 }, { "epoch": 2.6214099216710185, "grad_norm": 2.26336669921875, "learning_rate": 8.653803363336468e-06, "loss": 0.0787, "step": 2008 }, { "epoch": 2.622715404699739, "grad_norm": 1.9055275917053223, "learning_rate": 8.652359862348944e-06, "loss": 0.0555, "step": 2009 }, { "epoch": 2.6240208877284594, "grad_norm": 1.9863078594207764, "learning_rate": 8.650915708387743e-06, "loss": 0.084, "step": 2010 }, { "epoch": 2.62532637075718, "grad_norm": 2.8224282264709473, "learning_rate": 8.649470901711056e-06, "loss": 0.131, "step": 2011 }, { "epoch": 2.626631853785901, "grad_norm": 3.6200337409973145, "learning_rate": 8.648025442577186e-06, "loss": 0.1244, "step": 2012 }, { "epoch": 2.6279373368146213, "grad_norm": 2.941828727722168, "learning_rate": 8.646579331244554e-06, "loss": 0.0799, "step": 2013 }, { "epoch": 2.629242819843342, "grad_norm": 4.252133846282959, "learning_rate": 8.645132567971697e-06, "loss": 0.1715, "step": 2014 }, { "epoch": 2.6305483028720626, "grad_norm": 1.6941167116165161, "learning_rate": 8.64368515301727e-06, "loss": 0.0574, "step": 2015 }, { "epoch": 2.631853785900783, "grad_norm": 2.769165277481079, "learning_rate": 8.642237086640045e-06, "loss": 0.1172, "step": 2016 }, { "epoch": 2.633159268929504, "grad_norm": 2.338134288787842, "learning_rate": 8.640788369098909e-06, "loss": 0.0852, "step": 2017 }, { "epoch": 2.6344647519582245, "grad_norm": 1.8698315620422363, "learning_rate": 8.639339000652863e-06, "loss": 0.0696, "step": 2018 }, { "epoch": 2.6357702349869454, "grad_norm": 2.3189890384674072, "learning_rate": 8.637888981561029e-06, "loss": 0.0729, "step": 2019 }, { "epoch": 2.637075718015666, "grad_norm": 3.1041266918182373, "learning_rate": 8.636438312082642e-06, "loss": 0.1045, "step": 2020 }, { "epoch": 2.6383812010443863, "grad_norm": 2.179077386856079, "learning_rate": 8.634986992477057e-06, "loss": 0.0812, "step": 2021 }, { "epoch": 2.639686684073107, "grad_norm": 3.921675205230713, "learning_rate": 8.63353502300374e-06, "loss": 0.1517, "step": 2022 }, { "epoch": 2.6409921671018277, "grad_norm": 2.099447011947632, "learning_rate": 8.632082403922278e-06, "loss": 0.0786, "step": 2023 }, { "epoch": 2.642297650130548, "grad_norm": 1.6414105892181396, "learning_rate": 8.630629135492372e-06, "loss": 0.0511, "step": 2024 }, { "epoch": 2.643603133159269, "grad_norm": 2.424699544906616, "learning_rate": 8.629175217973838e-06, "loss": 0.0844, "step": 2025 }, { "epoch": 2.6449086161879896, "grad_norm": 3.793910264968872, "learning_rate": 8.62772065162661e-06, "loss": 0.138, "step": 2026 }, { "epoch": 2.64621409921671, "grad_norm": 2.148981809616089, "learning_rate": 8.626265436710736e-06, "loss": 0.0764, "step": 2027 }, { "epoch": 2.647519582245431, "grad_norm": 3.1640822887420654, "learning_rate": 8.624809573486385e-06, "loss": 0.1011, "step": 2028 }, { "epoch": 2.6488250652741514, "grad_norm": 2.5309691429138184, "learning_rate": 8.623353062213836e-06, "loss": 0.0692, "step": 2029 }, { "epoch": 2.6501305483028723, "grad_norm": 3.5875658988952637, "learning_rate": 8.621895903153484e-06, "loss": 0.1253, "step": 2030 }, { "epoch": 2.651436031331593, "grad_norm": 1.7034248113632202, "learning_rate": 8.620438096565842e-06, "loss": 0.0764, "step": 2031 }, { "epoch": 2.6527415143603132, "grad_norm": 2.8708066940307617, "learning_rate": 8.618979642711542e-06, "loss": 0.1473, "step": 2032 }, { "epoch": 2.6540469973890337, "grad_norm": 4.622320175170898, "learning_rate": 8.617520541851325e-06, "loss": 0.239, "step": 2033 }, { "epoch": 2.6553524804177546, "grad_norm": 3.93156099319458, "learning_rate": 8.616060794246053e-06, "loss": 0.1432, "step": 2034 }, { "epoch": 2.656657963446475, "grad_norm": 3.09747576713562, "learning_rate": 8.614600400156701e-06, "loss": 0.1019, "step": 2035 }, { "epoch": 2.657963446475196, "grad_norm": 2.9008030891418457, "learning_rate": 8.61313935984436e-06, "loss": 0.1089, "step": 2036 }, { "epoch": 2.6592689295039165, "grad_norm": 1.8194265365600586, "learning_rate": 8.611677673570237e-06, "loss": 0.061, "step": 2037 }, { "epoch": 2.660574412532637, "grad_norm": 2.358272075653076, "learning_rate": 8.610215341595654e-06, "loss": 0.086, "step": 2038 }, { "epoch": 2.661879895561358, "grad_norm": 2.398984909057617, "learning_rate": 8.608752364182048e-06, "loss": 0.0902, "step": 2039 }, { "epoch": 2.6631853785900783, "grad_norm": 3.0883820056915283, "learning_rate": 8.607288741590971e-06, "loss": 0.1373, "step": 2040 }, { "epoch": 2.6644908616187992, "grad_norm": 3.663548231124878, "learning_rate": 8.605824474084095e-06, "loss": 0.1295, "step": 2041 }, { "epoch": 2.6657963446475197, "grad_norm": 2.9663846492767334, "learning_rate": 8.604359561923202e-06, "loss": 0.1401, "step": 2042 }, { "epoch": 2.66710182767624, "grad_norm": 2.8870134353637695, "learning_rate": 8.602894005370188e-06, "loss": 0.1327, "step": 2043 }, { "epoch": 2.6684073107049606, "grad_norm": 1.6545377969741821, "learning_rate": 8.60142780468707e-06, "loss": 0.0525, "step": 2044 }, { "epoch": 2.6697127937336815, "grad_norm": 3.9353082180023193, "learning_rate": 8.599960960135977e-06, "loss": 0.1644, "step": 2045 }, { "epoch": 2.671018276762402, "grad_norm": 2.2046661376953125, "learning_rate": 8.598493471979155e-06, "loss": 0.0773, "step": 2046 }, { "epoch": 2.672323759791123, "grad_norm": 2.328528642654419, "learning_rate": 8.597025340478962e-06, "loss": 0.0768, "step": 2047 }, { "epoch": 2.6736292428198434, "grad_norm": 4.104942798614502, "learning_rate": 8.59555656589787e-06, "loss": 0.1741, "step": 2048 }, { "epoch": 2.674934725848564, "grad_norm": 2.4273886680603027, "learning_rate": 8.594087148498471e-06, "loss": 0.0772, "step": 2049 }, { "epoch": 2.6762402088772848, "grad_norm": 3.7192111015319824, "learning_rate": 8.592617088543469e-06, "loss": 0.1698, "step": 2050 }, { "epoch": 2.677545691906005, "grad_norm": 2.7609925270080566, "learning_rate": 8.591146386295685e-06, "loss": 0.0682, "step": 2051 }, { "epoch": 2.6788511749347257, "grad_norm": 2.312673568725586, "learning_rate": 8.589675042018049e-06, "loss": 0.0757, "step": 2052 }, { "epoch": 2.6801566579634466, "grad_norm": 2.276407241821289, "learning_rate": 8.588203055973613e-06, "loss": 0.0674, "step": 2053 }, { "epoch": 2.681462140992167, "grad_norm": 2.1940972805023193, "learning_rate": 8.58673042842554e-06, "loss": 0.0779, "step": 2054 }, { "epoch": 2.6827676240208875, "grad_norm": 3.6715028285980225, "learning_rate": 8.585257159637108e-06, "loss": 0.1487, "step": 2055 }, { "epoch": 2.6840731070496084, "grad_norm": 3.2442731857299805, "learning_rate": 8.583783249871708e-06, "loss": 0.1568, "step": 2056 }, { "epoch": 2.685378590078329, "grad_norm": 4.937743663787842, "learning_rate": 8.582308699392852e-06, "loss": 0.2556, "step": 2057 }, { "epoch": 2.68668407310705, "grad_norm": 2.5229198932647705, "learning_rate": 8.580833508464159e-06, "loss": 0.0772, "step": 2058 }, { "epoch": 2.6879895561357703, "grad_norm": 2.938544988632202, "learning_rate": 8.579357677349365e-06, "loss": 0.1116, "step": 2059 }, { "epoch": 2.6892950391644908, "grad_norm": 4.473315715789795, "learning_rate": 8.577881206312323e-06, "loss": 0.1574, "step": 2060 }, { "epoch": 2.690600522193211, "grad_norm": 2.1343812942504883, "learning_rate": 8.576404095616996e-06, "loss": 0.0644, "step": 2061 }, { "epoch": 2.691906005221932, "grad_norm": 3.3019649982452393, "learning_rate": 8.574926345527468e-06, "loss": 0.1215, "step": 2062 }, { "epoch": 2.6932114882506526, "grad_norm": 1.7263813018798828, "learning_rate": 8.573447956307928e-06, "loss": 0.0596, "step": 2063 }, { "epoch": 2.6945169712793735, "grad_norm": 2.3813891410827637, "learning_rate": 8.571968928222688e-06, "loss": 0.0689, "step": 2064 }, { "epoch": 2.695822454308094, "grad_norm": 2.1659538745880127, "learning_rate": 8.57048926153617e-06, "loss": 0.069, "step": 2065 }, { "epoch": 2.6971279373368144, "grad_norm": 1.6401532888412476, "learning_rate": 8.569008956512907e-06, "loss": 0.058, "step": 2066 }, { "epoch": 2.6984334203655354, "grad_norm": 2.73140025138855, "learning_rate": 8.567528013417556e-06, "loss": 0.0835, "step": 2067 }, { "epoch": 2.699738903394256, "grad_norm": 3.6573851108551025, "learning_rate": 8.566046432514877e-06, "loss": 0.1675, "step": 2068 }, { "epoch": 2.7010443864229767, "grad_norm": 2.683363437652588, "learning_rate": 8.564564214069753e-06, "loss": 0.0772, "step": 2069 }, { "epoch": 2.702349869451697, "grad_norm": 2.2396373748779297, "learning_rate": 8.563081358347172e-06, "loss": 0.0714, "step": 2070 }, { "epoch": 2.7036553524804177, "grad_norm": 1.8404204845428467, "learning_rate": 8.561597865612245e-06, "loss": 0.0603, "step": 2071 }, { "epoch": 2.704960835509138, "grad_norm": 2.374779224395752, "learning_rate": 8.560113736130192e-06, "loss": 0.0734, "step": 2072 }, { "epoch": 2.706266318537859, "grad_norm": 2.335264205932617, "learning_rate": 8.558628970166347e-06, "loss": 0.07, "step": 2073 }, { "epoch": 2.7075718015665795, "grad_norm": 2.212970018386841, "learning_rate": 8.55714356798616e-06, "loss": 0.0701, "step": 2074 }, { "epoch": 2.7088772845953004, "grad_norm": 3.9841485023498535, "learning_rate": 8.555657529855188e-06, "loss": 0.1689, "step": 2075 }, { "epoch": 2.710182767624021, "grad_norm": 2.139437198638916, "learning_rate": 8.554170856039113e-06, "loss": 0.0599, "step": 2076 }, { "epoch": 2.7114882506527413, "grad_norm": 2.983032464981079, "learning_rate": 8.55268354680372e-06, "loss": 0.0746, "step": 2077 }, { "epoch": 2.7127937336814623, "grad_norm": 2.9982521533966064, "learning_rate": 8.551195602414916e-06, "loss": 0.1271, "step": 2078 }, { "epoch": 2.7140992167101827, "grad_norm": 1.5271461009979248, "learning_rate": 8.549707023138715e-06, "loss": 0.0509, "step": 2079 }, { "epoch": 2.7154046997389036, "grad_norm": 2.4998703002929688, "learning_rate": 8.548217809241248e-06, "loss": 0.0751, "step": 2080 }, { "epoch": 2.716710182767624, "grad_norm": 1.913388967514038, "learning_rate": 8.546727960988758e-06, "loss": 0.0539, "step": 2081 }, { "epoch": 2.7180156657963446, "grad_norm": 1.7432955503463745, "learning_rate": 8.545237478647604e-06, "loss": 0.0682, "step": 2082 }, { "epoch": 2.719321148825065, "grad_norm": 2.8741748332977295, "learning_rate": 8.543746362484253e-06, "loss": 0.1051, "step": 2083 }, { "epoch": 2.720626631853786, "grad_norm": 1.871505618095398, "learning_rate": 8.542254612765291e-06, "loss": 0.0721, "step": 2084 }, { "epoch": 2.7219321148825064, "grad_norm": 2.585294723510742, "learning_rate": 8.540762229757415e-06, "loss": 0.0952, "step": 2085 }, { "epoch": 2.7232375979112273, "grad_norm": 1.2840192317962646, "learning_rate": 8.539269213727434e-06, "loss": 0.0457, "step": 2086 }, { "epoch": 2.724543080939948, "grad_norm": 3.9867444038391113, "learning_rate": 8.537775564942273e-06, "loss": 0.1497, "step": 2087 }, { "epoch": 2.7258485639686683, "grad_norm": 2.9377505779266357, "learning_rate": 8.536281283668965e-06, "loss": 0.1128, "step": 2088 }, { "epoch": 2.727154046997389, "grad_norm": 2.4728541374206543, "learning_rate": 8.534786370174665e-06, "loss": 0.0915, "step": 2089 }, { "epoch": 2.7284595300261096, "grad_norm": 1.7982221841812134, "learning_rate": 8.533290824726632e-06, "loss": 0.0596, "step": 2090 }, { "epoch": 2.7297650130548305, "grad_norm": 1.685861587524414, "learning_rate": 8.531794647592242e-06, "loss": 0.0588, "step": 2091 }, { "epoch": 2.731070496083551, "grad_norm": 1.5154930353164673, "learning_rate": 8.530297839038983e-06, "loss": 0.0455, "step": 2092 }, { "epoch": 2.7323759791122715, "grad_norm": 3.8200769424438477, "learning_rate": 8.528800399334458e-06, "loss": 0.1597, "step": 2093 }, { "epoch": 2.733681462140992, "grad_norm": 2.05411958694458, "learning_rate": 8.527302328746381e-06, "loss": 0.0657, "step": 2094 }, { "epoch": 2.734986945169713, "grad_norm": 3.749190092086792, "learning_rate": 8.525803627542577e-06, "loss": 0.1137, "step": 2095 }, { "epoch": 2.7362924281984333, "grad_norm": 2.1258299350738525, "learning_rate": 8.524304295990988e-06, "loss": 0.0678, "step": 2096 }, { "epoch": 2.7375979112271542, "grad_norm": 1.6789865493774414, "learning_rate": 8.522804334359666e-06, "loss": 0.0636, "step": 2097 }, { "epoch": 2.7389033942558747, "grad_norm": 2.956821918487549, "learning_rate": 8.521303742916776e-06, "loss": 0.1097, "step": 2098 }, { "epoch": 2.740208877284595, "grad_norm": 2.79233980178833, "learning_rate": 8.519802521930594e-06, "loss": 0.0896, "step": 2099 }, { "epoch": 2.741514360313316, "grad_norm": 1.527705430984497, "learning_rate": 8.518300671669513e-06, "loss": 0.0535, "step": 2100 }, { "epoch": 2.7428198433420365, "grad_norm": 2.6329376697540283, "learning_rate": 8.516798192402034e-06, "loss": 0.0846, "step": 2101 }, { "epoch": 2.7441253263707575, "grad_norm": 2.646712303161621, "learning_rate": 8.515295084396773e-06, "loss": 0.1063, "step": 2102 }, { "epoch": 2.745430809399478, "grad_norm": 3.1801209449768066, "learning_rate": 8.513791347922456e-06, "loss": 0.1171, "step": 2103 }, { "epoch": 2.7467362924281984, "grad_norm": 2.5624821186065674, "learning_rate": 8.512286983247925e-06, "loss": 0.0863, "step": 2104 }, { "epoch": 2.748041775456919, "grad_norm": 1.968214750289917, "learning_rate": 8.510781990642131e-06, "loss": 0.0624, "step": 2105 }, { "epoch": 2.7493472584856398, "grad_norm": 3.8259692192077637, "learning_rate": 8.50927637037414e-06, "loss": 0.1833, "step": 2106 }, { "epoch": 2.7506527415143602, "grad_norm": 1.6860926151275635, "learning_rate": 8.507770122713129e-06, "loss": 0.0577, "step": 2107 }, { "epoch": 2.751958224543081, "grad_norm": 2.2674262523651123, "learning_rate": 8.506263247928384e-06, "loss": 0.0746, "step": 2108 }, { "epoch": 2.7532637075718016, "grad_norm": 2.933256149291992, "learning_rate": 8.504755746289309e-06, "loss": 0.1219, "step": 2109 }, { "epoch": 2.754569190600522, "grad_norm": 2.120363712310791, "learning_rate": 8.503247618065415e-06, "loss": 0.0702, "step": 2110 }, { "epoch": 2.7558746736292425, "grad_norm": 2.7028074264526367, "learning_rate": 8.501738863526328e-06, "loss": 0.0977, "step": 2111 }, { "epoch": 2.7571801566579635, "grad_norm": 2.507636785507202, "learning_rate": 8.500229482941784e-06, "loss": 0.0773, "step": 2112 }, { "epoch": 2.758485639686684, "grad_norm": 2.5025718212127686, "learning_rate": 8.498719476581637e-06, "loss": 0.0586, "step": 2113 }, { "epoch": 2.759791122715405, "grad_norm": 2.7232747077941895, "learning_rate": 8.49720884471584e-06, "loss": 0.0727, "step": 2114 }, { "epoch": 2.7610966057441253, "grad_norm": 2.6165964603424072, "learning_rate": 8.495697587614471e-06, "loss": 0.0872, "step": 2115 }, { "epoch": 2.7624020887728458, "grad_norm": 1.394669532775879, "learning_rate": 8.494185705547714e-06, "loss": 0.0481, "step": 2116 }, { "epoch": 2.7637075718015667, "grad_norm": 2.9094181060791016, "learning_rate": 8.492673198785865e-06, "loss": 0.0828, "step": 2117 }, { "epoch": 2.765013054830287, "grad_norm": 2.2750561237335205, "learning_rate": 8.491160067599331e-06, "loss": 0.0864, "step": 2118 }, { "epoch": 2.766318537859008, "grad_norm": 4.503406047821045, "learning_rate": 8.489646312258633e-06, "loss": 0.1848, "step": 2119 }, { "epoch": 2.7676240208877285, "grad_norm": 3.7491135597229004, "learning_rate": 8.4881319330344e-06, "loss": 0.1755, "step": 2120 }, { "epoch": 2.768929503916449, "grad_norm": 3.4468116760253906, "learning_rate": 8.486616930197378e-06, "loss": 0.1368, "step": 2121 }, { "epoch": 2.7702349869451695, "grad_norm": 2.214064121246338, "learning_rate": 8.485101304018418e-06, "loss": 0.0693, "step": 2122 }, { "epoch": 2.7715404699738904, "grad_norm": 2.9689111709594727, "learning_rate": 8.483585054768486e-06, "loss": 0.1164, "step": 2123 }, { "epoch": 2.772845953002611, "grad_norm": 2.66496205329895, "learning_rate": 8.48206818271866e-06, "loss": 0.0904, "step": 2124 }, { "epoch": 2.7741514360313317, "grad_norm": 3.5179226398468018, "learning_rate": 8.48055068814013e-06, "loss": 0.1536, "step": 2125 }, { "epoch": 2.775456919060052, "grad_norm": 2.493715286254883, "learning_rate": 8.479032571304193e-06, "loss": 0.0764, "step": 2126 }, { "epoch": 2.7767624020887727, "grad_norm": 1.9859591722488403, "learning_rate": 8.47751383248226e-06, "loss": 0.0598, "step": 2127 }, { "epoch": 2.7780678851174936, "grad_norm": 2.0633251667022705, "learning_rate": 8.475994471945854e-06, "loss": 0.0655, "step": 2128 }, { "epoch": 2.779373368146214, "grad_norm": 1.335663914680481, "learning_rate": 8.474474489966609e-06, "loss": 0.0495, "step": 2129 }, { "epoch": 2.780678851174935, "grad_norm": 1.7759414911270142, "learning_rate": 8.472953886816267e-06, "loss": 0.0752, "step": 2130 }, { "epoch": 2.7819843342036554, "grad_norm": 1.8995656967163086, "learning_rate": 8.471432662766687e-06, "loss": 0.0636, "step": 2131 }, { "epoch": 2.783289817232376, "grad_norm": 2.100449323654175, "learning_rate": 8.469910818089832e-06, "loss": 0.0506, "step": 2132 }, { "epoch": 2.7845953002610964, "grad_norm": 3.1944777965545654, "learning_rate": 8.468388353057782e-06, "loss": 0.1092, "step": 2133 }, { "epoch": 2.7859007832898173, "grad_norm": 3.133293628692627, "learning_rate": 8.466865267942723e-06, "loss": 0.1029, "step": 2134 }, { "epoch": 2.7872062663185377, "grad_norm": 3.042698860168457, "learning_rate": 8.465341563016958e-06, "loss": 0.09, "step": 2135 }, { "epoch": 2.7885117493472587, "grad_norm": 2.5309042930603027, "learning_rate": 8.46381723855289e-06, "loss": 0.0791, "step": 2136 }, { "epoch": 2.789817232375979, "grad_norm": 1.7970690727233887, "learning_rate": 8.462292294823046e-06, "loss": 0.0641, "step": 2137 }, { "epoch": 2.7911227154046996, "grad_norm": 2.6503920555114746, "learning_rate": 8.460766732100055e-06, "loss": 0.106, "step": 2138 }, { "epoch": 2.7924281984334205, "grad_norm": 3.308670997619629, "learning_rate": 8.45924055065666e-06, "loss": 0.1531, "step": 2139 }, { "epoch": 2.793733681462141, "grad_norm": 1.562342882156372, "learning_rate": 8.457713750765712e-06, "loss": 0.0513, "step": 2140 }, { "epoch": 2.795039164490862, "grad_norm": 1.5909620523452759, "learning_rate": 8.456186332700176e-06, "loss": 0.0687, "step": 2141 }, { "epoch": 2.7963446475195823, "grad_norm": 2.646338939666748, "learning_rate": 8.454658296733127e-06, "loss": 0.0773, "step": 2142 }, { "epoch": 2.797650130548303, "grad_norm": 3.334723711013794, "learning_rate": 8.453129643137746e-06, "loss": 0.1207, "step": 2143 }, { "epoch": 2.7989556135770233, "grad_norm": 1.9811062812805176, "learning_rate": 8.451600372187328e-06, "loss": 0.0649, "step": 2144 }, { "epoch": 2.800261096605744, "grad_norm": 3.6174983978271484, "learning_rate": 8.450070484155281e-06, "loss": 0.1562, "step": 2145 }, { "epoch": 2.8015665796344646, "grad_norm": 2.1267902851104736, "learning_rate": 8.448539979315118e-06, "loss": 0.0583, "step": 2146 }, { "epoch": 2.8028720626631856, "grad_norm": 1.4980531930923462, "learning_rate": 8.447008857940466e-06, "loss": 0.0487, "step": 2147 }, { "epoch": 2.804177545691906, "grad_norm": 2.5243451595306396, "learning_rate": 8.44547712030506e-06, "loss": 0.0772, "step": 2148 }, { "epoch": 2.8054830287206265, "grad_norm": 1.8936173915863037, "learning_rate": 8.443944766682744e-06, "loss": 0.0662, "step": 2149 }, { "epoch": 2.8067885117493474, "grad_norm": 3.391498327255249, "learning_rate": 8.442411797347478e-06, "loss": 0.1234, "step": 2150 }, { "epoch": 2.808093994778068, "grad_norm": 2.3970813751220703, "learning_rate": 8.440878212573329e-06, "loss": 0.0903, "step": 2151 }, { "epoch": 2.809399477806789, "grad_norm": 2.732334613800049, "learning_rate": 8.439344012634466e-06, "loss": 0.1365, "step": 2152 }, { "epoch": 2.8107049608355092, "grad_norm": 1.8616224527359009, "learning_rate": 8.437809197805185e-06, "loss": 0.0644, "step": 2153 }, { "epoch": 2.8120104438642297, "grad_norm": 2.471978187561035, "learning_rate": 8.436273768359877e-06, "loss": 0.0621, "step": 2154 }, { "epoch": 2.81331592689295, "grad_norm": 2.179354429244995, "learning_rate": 8.434737724573046e-06, "loss": 0.0874, "step": 2155 }, { "epoch": 2.814621409921671, "grad_norm": 1.2761788368225098, "learning_rate": 8.433201066719312e-06, "loss": 0.0541, "step": 2156 }, { "epoch": 2.8159268929503916, "grad_norm": 3.5824761390686035, "learning_rate": 8.431663795073398e-06, "loss": 0.1621, "step": 2157 }, { "epoch": 2.8172323759791125, "grad_norm": 1.9348326921463013, "learning_rate": 8.430125909910142e-06, "loss": 0.0757, "step": 2158 }, { "epoch": 2.818537859007833, "grad_norm": 2.886075735092163, "learning_rate": 8.428587411504488e-06, "loss": 0.0797, "step": 2159 }, { "epoch": 2.8198433420365534, "grad_norm": 2.352449893951416, "learning_rate": 8.427048300131489e-06, "loss": 0.0837, "step": 2160 }, { "epoch": 2.8211488250652743, "grad_norm": 1.849874496459961, "learning_rate": 8.42550857606631e-06, "loss": 0.0502, "step": 2161 }, { "epoch": 2.822454308093995, "grad_norm": 1.692716360092163, "learning_rate": 8.423968239584228e-06, "loss": 0.0613, "step": 2162 }, { "epoch": 2.8237597911227152, "grad_norm": 2.487128257751465, "learning_rate": 8.422427290960623e-06, "loss": 0.0933, "step": 2163 }, { "epoch": 2.825065274151436, "grad_norm": 2.055490732192993, "learning_rate": 8.420885730470988e-06, "loss": 0.0724, "step": 2164 }, { "epoch": 2.8263707571801566, "grad_norm": 2.4462077617645264, "learning_rate": 8.41934355839093e-06, "loss": 0.0793, "step": 2165 }, { "epoch": 2.827676240208877, "grad_norm": 3.3667430877685547, "learning_rate": 8.417800774996154e-06, "loss": 0.153, "step": 2166 }, { "epoch": 2.828981723237598, "grad_norm": 1.6254128217697144, "learning_rate": 8.416257380562483e-06, "loss": 0.0602, "step": 2167 }, { "epoch": 2.8302872062663185, "grad_norm": 2.8522419929504395, "learning_rate": 8.414713375365847e-06, "loss": 0.1215, "step": 2168 }, { "epoch": 2.8315926892950394, "grad_norm": 1.7526521682739258, "learning_rate": 8.413168759682287e-06, "loss": 0.0493, "step": 2169 }, { "epoch": 2.83289817232376, "grad_norm": 5.267843246459961, "learning_rate": 8.41162353378795e-06, "loss": 0.3262, "step": 2170 }, { "epoch": 2.8342036553524803, "grad_norm": 1.986835241317749, "learning_rate": 8.410077697959094e-06, "loss": 0.0582, "step": 2171 }, { "epoch": 2.8355091383812008, "grad_norm": 2.910958766937256, "learning_rate": 8.408531252472086e-06, "loss": 0.1192, "step": 2172 }, { "epoch": 2.8368146214099217, "grad_norm": 1.682172417640686, "learning_rate": 8.406984197603399e-06, "loss": 0.0434, "step": 2173 }, { "epoch": 2.838120104438642, "grad_norm": 1.800273060798645, "learning_rate": 8.40543653362962e-06, "loss": 0.0705, "step": 2174 }, { "epoch": 2.839425587467363, "grad_norm": 2.5035240650177, "learning_rate": 8.403888260827441e-06, "loss": 0.087, "step": 2175 }, { "epoch": 2.8407310704960835, "grad_norm": 1.539044976234436, "learning_rate": 8.402339379473667e-06, "loss": 0.0608, "step": 2176 }, { "epoch": 2.842036553524804, "grad_norm": 2.191979169845581, "learning_rate": 8.400789889845205e-06, "loss": 0.0768, "step": 2177 }, { "epoch": 2.843342036553525, "grad_norm": 3.7111942768096924, "learning_rate": 8.399239792219078e-06, "loss": 0.134, "step": 2178 }, { "epoch": 2.8446475195822454, "grad_norm": 3.2324347496032715, "learning_rate": 8.397689086872413e-06, "loss": 0.1253, "step": 2179 }, { "epoch": 2.8459530026109663, "grad_norm": 3.805644989013672, "learning_rate": 8.396137774082447e-06, "loss": 0.1777, "step": 2180 }, { "epoch": 2.8472584856396868, "grad_norm": 1.226822018623352, "learning_rate": 8.394585854126526e-06, "loss": 0.0385, "step": 2181 }, { "epoch": 2.848563968668407, "grad_norm": 1.3533477783203125, "learning_rate": 8.393033327282105e-06, "loss": 0.0353, "step": 2182 }, { "epoch": 2.8498694516971277, "grad_norm": 1.883634328842163, "learning_rate": 8.391480193826747e-06, "loss": 0.0671, "step": 2183 }, { "epoch": 2.8511749347258486, "grad_norm": 1.8162428140640259, "learning_rate": 8.389926454038122e-06, "loss": 0.0401, "step": 2184 }, { "epoch": 2.852480417754569, "grad_norm": 2.150308132171631, "learning_rate": 8.38837210819401e-06, "loss": 0.0775, "step": 2185 }, { "epoch": 2.85378590078329, "grad_norm": 3.282965660095215, "learning_rate": 8.386817156572298e-06, "loss": 0.1393, "step": 2186 }, { "epoch": 2.8550913838120104, "grad_norm": 4.240999698638916, "learning_rate": 8.385261599450983e-06, "loss": 0.1829, "step": 2187 }, { "epoch": 2.856396866840731, "grad_norm": 2.5167975425720215, "learning_rate": 8.383705437108172e-06, "loss": 0.0722, "step": 2188 }, { "epoch": 2.857702349869452, "grad_norm": 1.9581526517868042, "learning_rate": 8.382148669822073e-06, "loss": 0.0454, "step": 2189 }, { "epoch": 2.8590078328981723, "grad_norm": 1.9790318012237549, "learning_rate": 8.38059129787101e-06, "loss": 0.0723, "step": 2190 }, { "epoch": 2.860313315926893, "grad_norm": 3.230496883392334, "learning_rate": 8.379033321533415e-06, "loss": 0.0932, "step": 2191 }, { "epoch": 2.8616187989556137, "grad_norm": 2.1444625854492188, "learning_rate": 8.377474741087817e-06, "loss": 0.0687, "step": 2192 }, { "epoch": 2.862924281984334, "grad_norm": 2.460249662399292, "learning_rate": 8.375915556812868e-06, "loss": 0.0734, "step": 2193 }, { "epoch": 2.8642297650130546, "grad_norm": 3.308931589126587, "learning_rate": 8.374355768987317e-06, "loss": 0.1302, "step": 2194 }, { "epoch": 2.8655352480417755, "grad_norm": 1.9418518543243408, "learning_rate": 8.372795377890028e-06, "loss": 0.0785, "step": 2195 }, { "epoch": 2.866840731070496, "grad_norm": 2.8149070739746094, "learning_rate": 8.371234383799968e-06, "loss": 0.0995, "step": 2196 }, { "epoch": 2.868146214099217, "grad_norm": 2.266745090484619, "learning_rate": 8.369672786996213e-06, "loss": 0.0742, "step": 2197 }, { "epoch": 2.8694516971279374, "grad_norm": 2.2715811729431152, "learning_rate": 8.368110587757949e-06, "loss": 0.0663, "step": 2198 }, { "epoch": 2.870757180156658, "grad_norm": 1.5184557437896729, "learning_rate": 8.366547786364465e-06, "loss": 0.0551, "step": 2199 }, { "epoch": 2.8720626631853787, "grad_norm": 2.221792697906494, "learning_rate": 8.364984383095167e-06, "loss": 0.0561, "step": 2200 }, { "epoch": 2.873368146214099, "grad_norm": 3.0731890201568604, "learning_rate": 8.363420378229554e-06, "loss": 0.1024, "step": 2201 }, { "epoch": 2.87467362924282, "grad_norm": 3.598663568496704, "learning_rate": 8.361855772047249e-06, "loss": 0.1145, "step": 2202 }, { "epoch": 2.8759791122715406, "grad_norm": 2.837566375732422, "learning_rate": 8.360290564827966e-06, "loss": 0.096, "step": 2203 }, { "epoch": 2.877284595300261, "grad_norm": 2.4325246810913086, "learning_rate": 8.358724756851542e-06, "loss": 0.0641, "step": 2204 }, { "epoch": 2.8785900783289815, "grad_norm": 3.197420120239258, "learning_rate": 8.357158348397911e-06, "loss": 0.0924, "step": 2205 }, { "epoch": 2.8798955613577024, "grad_norm": 3.446876049041748, "learning_rate": 8.355591339747116e-06, "loss": 0.1342, "step": 2206 }, { "epoch": 2.881201044386423, "grad_norm": 2.633660316467285, "learning_rate": 8.354023731179312e-06, "loss": 0.0661, "step": 2207 }, { "epoch": 2.882506527415144, "grad_norm": 2.1460301876068115, "learning_rate": 8.352455522974759e-06, "loss": 0.0738, "step": 2208 }, { "epoch": 2.8838120104438643, "grad_norm": 1.6464003324508667, "learning_rate": 8.350886715413817e-06, "loss": 0.0445, "step": 2209 }, { "epoch": 2.8851174934725847, "grad_norm": 2.0787220001220703, "learning_rate": 8.349317308776967e-06, "loss": 0.0637, "step": 2210 }, { "epoch": 2.8864229765013056, "grad_norm": 1.8857232332229614, "learning_rate": 8.347747303344784e-06, "loss": 0.0557, "step": 2211 }, { "epoch": 2.887728459530026, "grad_norm": 1.5728907585144043, "learning_rate": 8.346176699397958e-06, "loss": 0.0574, "step": 2212 }, { "epoch": 2.8890339425587466, "grad_norm": 3.203186273574829, "learning_rate": 8.344605497217285e-06, "loss": 0.0897, "step": 2213 }, { "epoch": 2.8903394255874675, "grad_norm": 3.6697518825531006, "learning_rate": 8.343033697083663e-06, "loss": 0.1306, "step": 2214 }, { "epoch": 2.891644908616188, "grad_norm": 1.9825029373168945, "learning_rate": 8.341461299278104e-06, "loss": 0.0598, "step": 2215 }, { "epoch": 2.8929503916449084, "grad_norm": 1.7854955196380615, "learning_rate": 8.339888304081721e-06, "loss": 0.0575, "step": 2216 }, { "epoch": 2.8942558746736293, "grad_norm": 2.5934228897094727, "learning_rate": 8.338314711775737e-06, "loss": 0.0926, "step": 2217 }, { "epoch": 2.89556135770235, "grad_norm": 2.2103497982025146, "learning_rate": 8.33674052264148e-06, "loss": 0.0689, "step": 2218 }, { "epoch": 2.8968668407310707, "grad_norm": 1.991852879524231, "learning_rate": 8.33516573696039e-06, "loss": 0.0553, "step": 2219 }, { "epoch": 2.898172323759791, "grad_norm": 2.353642702102661, "learning_rate": 8.333590355014002e-06, "loss": 0.07, "step": 2220 }, { "epoch": 2.8994778067885116, "grad_norm": 2.07559871673584, "learning_rate": 8.33201437708397e-06, "loss": 0.061, "step": 2221 }, { "epoch": 2.900783289817232, "grad_norm": 3.908768653869629, "learning_rate": 8.330437803452047e-06, "loss": 0.1535, "step": 2222 }, { "epoch": 2.902088772845953, "grad_norm": 2.155795097351074, "learning_rate": 8.328860634400098e-06, "loss": 0.0685, "step": 2223 }, { "epoch": 2.9033942558746735, "grad_norm": 1.6833099126815796, "learning_rate": 8.327282870210088e-06, "loss": 0.0537, "step": 2224 }, { "epoch": 2.9046997389033944, "grad_norm": 1.3552902936935425, "learning_rate": 8.325704511164092e-06, "loss": 0.0432, "step": 2225 }, { "epoch": 2.906005221932115, "grad_norm": 2.193413019180298, "learning_rate": 8.324125557544293e-06, "loss": 0.0799, "step": 2226 }, { "epoch": 2.9073107049608353, "grad_norm": 3.243295907974243, "learning_rate": 8.322546009632977e-06, "loss": 0.1215, "step": 2227 }, { "epoch": 2.9086161879895562, "grad_norm": 1.4329922199249268, "learning_rate": 8.320965867712538e-06, "loss": 0.0434, "step": 2228 }, { "epoch": 2.9099216710182767, "grad_norm": 2.4684040546417236, "learning_rate": 8.319385132065477e-06, "loss": 0.0844, "step": 2229 }, { "epoch": 2.9112271540469976, "grad_norm": 3.486112117767334, "learning_rate": 8.317803802974396e-06, "loss": 0.1512, "step": 2230 }, { "epoch": 2.912532637075718, "grad_norm": 2.7798075675964355, "learning_rate": 8.316221880722013e-06, "loss": 0.1073, "step": 2231 }, { "epoch": 2.9138381201044385, "grad_norm": 2.3139560222625732, "learning_rate": 8.314639365591141e-06, "loss": 0.1012, "step": 2232 }, { "epoch": 2.915143603133159, "grad_norm": 3.9713611602783203, "learning_rate": 8.313056257864708e-06, "loss": 0.1931, "step": 2233 }, { "epoch": 2.91644908616188, "grad_norm": 2.809664249420166, "learning_rate": 8.311472557825741e-06, "loss": 0.1032, "step": 2234 }, { "epoch": 2.9177545691906004, "grad_norm": 1.945120930671692, "learning_rate": 8.309888265757379e-06, "loss": 0.0554, "step": 2235 }, { "epoch": 2.9190600522193213, "grad_norm": 2.4561197757720947, "learning_rate": 8.30830338194286e-06, "loss": 0.0707, "step": 2236 }, { "epoch": 2.9203655352480418, "grad_norm": 1.8780755996704102, "learning_rate": 8.306717906665535e-06, "loss": 0.0508, "step": 2237 }, { "epoch": 2.9216710182767622, "grad_norm": 2.0036020278930664, "learning_rate": 8.305131840208856e-06, "loss": 0.0618, "step": 2238 }, { "epoch": 2.922976501305483, "grad_norm": 2.6845738887786865, "learning_rate": 8.30354518285638e-06, "loss": 0.0909, "step": 2239 }, { "epoch": 2.9242819843342036, "grad_norm": 2.563575029373169, "learning_rate": 8.301957934891777e-06, "loss": 0.0853, "step": 2240 }, { "epoch": 2.9255874673629245, "grad_norm": 1.8821978569030762, "learning_rate": 8.300370096598813e-06, "loss": 0.0671, "step": 2241 }, { "epoch": 2.926892950391645, "grad_norm": 2.180708885192871, "learning_rate": 8.298781668261364e-06, "loss": 0.053, "step": 2242 }, { "epoch": 2.9281984334203655, "grad_norm": 3.640867233276367, "learning_rate": 8.297192650163412e-06, "loss": 0.1189, "step": 2243 }, { "epoch": 2.929503916449086, "grad_norm": 3.387606143951416, "learning_rate": 8.295603042589044e-06, "loss": 0.1396, "step": 2244 }, { "epoch": 2.930809399477807, "grad_norm": 1.0119541883468628, "learning_rate": 8.294012845822453e-06, "loss": 0.037, "step": 2245 }, { "epoch": 2.9321148825065273, "grad_norm": 3.0348920822143555, "learning_rate": 8.292422060147934e-06, "loss": 0.0863, "step": 2246 }, { "epoch": 2.933420365535248, "grad_norm": 1.442053198814392, "learning_rate": 8.290830685849891e-06, "loss": 0.0438, "step": 2247 }, { "epoch": 2.9347258485639687, "grad_norm": 2.1392710208892822, "learning_rate": 8.289238723212833e-06, "loss": 0.0773, "step": 2248 }, { "epoch": 2.936031331592689, "grad_norm": 3.2124977111816406, "learning_rate": 8.287646172521373e-06, "loss": 0.1087, "step": 2249 }, { "epoch": 2.93733681462141, "grad_norm": 2.039524793624878, "learning_rate": 8.286053034060226e-06, "loss": 0.0609, "step": 2250 }, { "epoch": 2.9386422976501305, "grad_norm": 4.226293563842773, "learning_rate": 8.284459308114221e-06, "loss": 0.1944, "step": 2251 }, { "epoch": 2.9399477806788514, "grad_norm": 3.972092628479004, "learning_rate": 8.282864994968283e-06, "loss": 0.1442, "step": 2252 }, { "epoch": 2.941253263707572, "grad_norm": 2.767301559448242, "learning_rate": 8.281270094907444e-06, "loss": 0.0774, "step": 2253 }, { "epoch": 2.9425587467362924, "grad_norm": 2.1094553470611572, "learning_rate": 8.279674608216846e-06, "loss": 0.0719, "step": 2254 }, { "epoch": 2.943864229765013, "grad_norm": 3.418269157409668, "learning_rate": 8.278078535181728e-06, "loss": 0.1181, "step": 2255 }, { "epoch": 2.9451697127937337, "grad_norm": 2.795457601547241, "learning_rate": 8.27648187608744e-06, "loss": 0.0833, "step": 2256 }, { "epoch": 2.946475195822454, "grad_norm": 3.247457981109619, "learning_rate": 8.274884631219439e-06, "loss": 0.1161, "step": 2257 }, { "epoch": 2.947780678851175, "grad_norm": 2.125283718109131, "learning_rate": 8.273286800863276e-06, "loss": 0.0609, "step": 2258 }, { "epoch": 2.9490861618798956, "grad_norm": 1.8053935766220093, "learning_rate": 8.271688385304616e-06, "loss": 0.0594, "step": 2259 }, { "epoch": 2.950391644908616, "grad_norm": 3.011470079421997, "learning_rate": 8.270089384829227e-06, "loss": 0.0693, "step": 2260 }, { "epoch": 2.951697127937337, "grad_norm": 1.5631752014160156, "learning_rate": 8.268489799722978e-06, "loss": 0.0547, "step": 2261 }, { "epoch": 2.9530026109660574, "grad_norm": 1.750220775604248, "learning_rate": 8.266889630271844e-06, "loss": 0.0486, "step": 2262 }, { "epoch": 2.9543080939947783, "grad_norm": 1.2139455080032349, "learning_rate": 8.26528887676191e-06, "loss": 0.0404, "step": 2263 }, { "epoch": 2.955613577023499, "grad_norm": 2.807741165161133, "learning_rate": 8.26368753947936e-06, "loss": 0.1025, "step": 2264 }, { "epoch": 2.9569190600522193, "grad_norm": 2.247982978820801, "learning_rate": 8.262085618710479e-06, "loss": 0.0697, "step": 2265 }, { "epoch": 2.9582245430809397, "grad_norm": 2.6717312335968018, "learning_rate": 8.260483114741662e-06, "loss": 0.0891, "step": 2266 }, { "epoch": 2.9595300261096606, "grad_norm": 1.9721051454544067, "learning_rate": 8.258880027859408e-06, "loss": 0.0612, "step": 2267 }, { "epoch": 2.960835509138381, "grad_norm": 1.7400637865066528, "learning_rate": 8.25727635835032e-06, "loss": 0.0604, "step": 2268 }, { "epoch": 2.962140992167102, "grad_norm": 2.425966501235962, "learning_rate": 8.255672106501101e-06, "loss": 0.0767, "step": 2269 }, { "epoch": 2.9634464751958225, "grad_norm": 1.8829221725463867, "learning_rate": 8.254067272598563e-06, "loss": 0.0548, "step": 2270 }, { "epoch": 2.964751958224543, "grad_norm": 2.204129219055176, "learning_rate": 8.252461856929622e-06, "loss": 0.0543, "step": 2271 }, { "epoch": 2.966057441253264, "grad_norm": 2.6293745040893555, "learning_rate": 8.250855859781292e-06, "loss": 0.0652, "step": 2272 }, { "epoch": 2.9673629242819843, "grad_norm": 2.3721179962158203, "learning_rate": 8.249249281440701e-06, "loss": 0.0683, "step": 2273 }, { "epoch": 2.968668407310705, "grad_norm": 1.862647294998169, "learning_rate": 8.24764212219507e-06, "loss": 0.0571, "step": 2274 }, { "epoch": 2.9699738903394257, "grad_norm": 2.7397074699401855, "learning_rate": 8.246034382331731e-06, "loss": 0.0871, "step": 2275 }, { "epoch": 2.971279373368146, "grad_norm": 1.5643588304519653, "learning_rate": 8.244426062138118e-06, "loss": 0.0518, "step": 2276 }, { "epoch": 2.9725848563968666, "grad_norm": 2.2157416343688965, "learning_rate": 8.24281716190177e-06, "loss": 0.0607, "step": 2277 }, { "epoch": 2.9738903394255876, "grad_norm": 2.0307581424713135, "learning_rate": 8.241207681910326e-06, "loss": 0.0642, "step": 2278 }, { "epoch": 2.975195822454308, "grad_norm": 2.5820107460021973, "learning_rate": 8.239597622451531e-06, "loss": 0.0879, "step": 2279 }, { "epoch": 2.976501305483029, "grad_norm": 1.7196741104125977, "learning_rate": 8.237986983813233e-06, "loss": 0.052, "step": 2280 }, { "epoch": 2.9778067885117494, "grad_norm": 1.3245954513549805, "learning_rate": 8.236375766283388e-06, "loss": 0.0461, "step": 2281 }, { "epoch": 2.97911227154047, "grad_norm": 2.640974283218384, "learning_rate": 8.234763970150048e-06, "loss": 0.068, "step": 2282 }, { "epoch": 2.9804177545691903, "grad_norm": 4.831238746643066, "learning_rate": 8.233151595701374e-06, "loss": 0.2117, "step": 2283 }, { "epoch": 2.9817232375979112, "grad_norm": 3.4821720123291016, "learning_rate": 8.231538643225625e-06, "loss": 0.1356, "step": 2284 }, { "epoch": 2.9830287206266317, "grad_norm": 3.1727187633514404, "learning_rate": 8.22992511301117e-06, "loss": 0.071, "step": 2285 }, { "epoch": 2.9843342036553526, "grad_norm": 3.8315043449401855, "learning_rate": 8.22831100534648e-06, "loss": 0.1638, "step": 2286 }, { "epoch": 2.985639686684073, "grad_norm": 3.883897542953491, "learning_rate": 8.22669632052012e-06, "loss": 0.1406, "step": 2287 }, { "epoch": 2.9869451697127936, "grad_norm": 1.7188642024993896, "learning_rate": 8.225081058820771e-06, "loss": 0.0555, "step": 2288 }, { "epoch": 2.9882506527415145, "grad_norm": 3.694204330444336, "learning_rate": 8.223465220537213e-06, "loss": 0.128, "step": 2289 }, { "epoch": 2.989556135770235, "grad_norm": 2.835907459259033, "learning_rate": 8.221848805958323e-06, "loss": 0.0976, "step": 2290 }, { "epoch": 2.990861618798956, "grad_norm": 2.683298110961914, "learning_rate": 8.220231815373088e-06, "loss": 0.0576, "step": 2291 }, { "epoch": 2.9921671018276763, "grad_norm": 2.10213303565979, "learning_rate": 8.218614249070595e-06, "loss": 0.0604, "step": 2292 }, { "epoch": 2.993472584856397, "grad_norm": 1.2883743047714233, "learning_rate": 8.216996107340037e-06, "loss": 0.0454, "step": 2293 }, { "epoch": 2.9947780678851172, "grad_norm": 1.1272673606872559, "learning_rate": 8.215377390470704e-06, "loss": 0.0388, "step": 2294 }, { "epoch": 2.996083550913838, "grad_norm": 2.796257257461548, "learning_rate": 8.213758098751996e-06, "loss": 0.0839, "step": 2295 }, { "epoch": 2.9973890339425586, "grad_norm": 2.3078725337982178, "learning_rate": 8.212138232473408e-06, "loss": 0.0707, "step": 2296 }, { "epoch": 2.9986945169712795, "grad_norm": 2.7247262001037598, "learning_rate": 8.210517791924545e-06, "loss": 0.1112, "step": 2297 }, { "epoch": 3.0, "grad_norm": 2.040485143661499, "learning_rate": 8.20889677739511e-06, "loss": 0.07, "step": 2298 }, { "epoch": 3.0013054830287205, "grad_norm": 1.2862287759780884, "learning_rate": 8.20727518917491e-06, "loss": 0.0466, "step": 2299 }, { "epoch": 3.0026109660574414, "grad_norm": 2.244676351547241, "learning_rate": 8.205653027553856e-06, "loss": 0.0776, "step": 2300 }, { "epoch": 3.003916449086162, "grad_norm": 2.761383056640625, "learning_rate": 8.20403029282196e-06, "loss": 0.0886, "step": 2301 }, { "epoch": 3.0052219321148823, "grad_norm": 3.754953145980835, "learning_rate": 8.202406985269335e-06, "loss": 0.1422, "step": 2302 }, { "epoch": 3.006527415143603, "grad_norm": 1.7170069217681885, "learning_rate": 8.200783105186201e-06, "loss": 0.0473, "step": 2303 }, { "epoch": 3.0078328981723237, "grad_norm": 1.240556240081787, "learning_rate": 8.199158652862873e-06, "loss": 0.0529, "step": 2304 }, { "epoch": 3.0091383812010446, "grad_norm": 3.0294673442840576, "learning_rate": 8.197533628589775e-06, "loss": 0.1127, "step": 2305 }, { "epoch": 3.010443864229765, "grad_norm": 1.160248041152954, "learning_rate": 8.195908032657432e-06, "loss": 0.0383, "step": 2306 }, { "epoch": 3.0117493472584855, "grad_norm": 3.364778995513916, "learning_rate": 8.19428186535647e-06, "loss": 0.1377, "step": 2307 }, { "epoch": 3.0130548302872064, "grad_norm": 1.389858365058899, "learning_rate": 8.192655126977615e-06, "loss": 0.047, "step": 2308 }, { "epoch": 3.014360313315927, "grad_norm": 1.5841975212097168, "learning_rate": 8.1910278178117e-06, "loss": 0.0467, "step": 2309 }, { "epoch": 3.0156657963446474, "grad_norm": 1.8674511909484863, "learning_rate": 8.189399938149657e-06, "loss": 0.0684, "step": 2310 }, { "epoch": 3.0169712793733683, "grad_norm": 1.8607242107391357, "learning_rate": 8.187771488282519e-06, "loss": 0.0573, "step": 2311 }, { "epoch": 3.0182767624020888, "grad_norm": 1.73256516456604, "learning_rate": 8.186142468501424e-06, "loss": 0.0568, "step": 2312 }, { "epoch": 3.019582245430809, "grad_norm": 2.467102289199829, "learning_rate": 8.184512879097609e-06, "loss": 0.0662, "step": 2313 }, { "epoch": 3.02088772845953, "grad_norm": 1.7477669715881348, "learning_rate": 8.182882720362416e-06, "loss": 0.0714, "step": 2314 }, { "epoch": 3.0221932114882506, "grad_norm": 2.6192827224731445, "learning_rate": 8.181251992587285e-06, "loss": 0.0874, "step": 2315 }, { "epoch": 3.023498694516971, "grad_norm": 2.016500234603882, "learning_rate": 8.17962069606376e-06, "loss": 0.0625, "step": 2316 }, { "epoch": 3.024804177545692, "grad_norm": 1.0755646228790283, "learning_rate": 8.177988831083487e-06, "loss": 0.0338, "step": 2317 }, { "epoch": 3.0261096605744124, "grad_norm": 2.9208855628967285, "learning_rate": 8.176356397938211e-06, "loss": 0.0897, "step": 2318 }, { "epoch": 3.0274151436031334, "grad_norm": 2.4813191890716553, "learning_rate": 8.174723396919784e-06, "loss": 0.0721, "step": 2319 }, { "epoch": 3.028720626631854, "grad_norm": 1.719773530960083, "learning_rate": 8.173089828320153e-06, "loss": 0.0473, "step": 2320 }, { "epoch": 3.0300261096605743, "grad_norm": 2.1758835315704346, "learning_rate": 8.171455692431372e-06, "loss": 0.0798, "step": 2321 }, { "epoch": 3.031331592689295, "grad_norm": 1.8259166479110718, "learning_rate": 8.16982098954559e-06, "loss": 0.0574, "step": 2322 }, { "epoch": 3.0326370757180157, "grad_norm": 1.6219137907028198, "learning_rate": 8.168185719955066e-06, "loss": 0.0472, "step": 2323 }, { "epoch": 3.033942558746736, "grad_norm": 1.4724750518798828, "learning_rate": 8.166549883952153e-06, "loss": 0.0535, "step": 2324 }, { "epoch": 3.035248041775457, "grad_norm": 1.3827118873596191, "learning_rate": 8.164913481829311e-06, "loss": 0.0433, "step": 2325 }, { "epoch": 3.0365535248041775, "grad_norm": 2.2488467693328857, "learning_rate": 8.163276513879093e-06, "loss": 0.0724, "step": 2326 }, { "epoch": 3.037859007832898, "grad_norm": 1.9505590200424194, "learning_rate": 8.161638980394162e-06, "loss": 0.0542, "step": 2327 }, { "epoch": 3.039164490861619, "grad_norm": 2.4893171787261963, "learning_rate": 8.160000881667276e-06, "loss": 0.074, "step": 2328 }, { "epoch": 3.0404699738903394, "grad_norm": 2.675809144973755, "learning_rate": 8.158362217991297e-06, "loss": 0.0578, "step": 2329 }, { "epoch": 3.0417754569190603, "grad_norm": 2.0563251972198486, "learning_rate": 8.156722989659192e-06, "loss": 0.0595, "step": 2330 }, { "epoch": 3.0430809399477807, "grad_norm": 1.8884100914001465, "learning_rate": 8.155083196964018e-06, "loss": 0.0587, "step": 2331 }, { "epoch": 3.044386422976501, "grad_norm": 1.8430163860321045, "learning_rate": 8.153442840198942e-06, "loss": 0.0646, "step": 2332 }, { "epoch": 3.045691906005222, "grad_norm": 1.1142281293869019, "learning_rate": 8.151801919657228e-06, "loss": 0.0443, "step": 2333 }, { "epoch": 3.0469973890339426, "grad_norm": 1.2646218538284302, "learning_rate": 8.150160435632244e-06, "loss": 0.0366, "step": 2334 }, { "epoch": 3.048302872062663, "grad_norm": 4.024001598358154, "learning_rate": 8.148518388417455e-06, "loss": 0.1128, "step": 2335 }, { "epoch": 3.049608355091384, "grad_norm": 1.3490437269210815, "learning_rate": 8.14687577830643e-06, "loss": 0.04, "step": 2336 }, { "epoch": 3.0509138381201044, "grad_norm": 1.7624311447143555, "learning_rate": 8.145232605592835e-06, "loss": 0.0706, "step": 2337 }, { "epoch": 3.052219321148825, "grad_norm": 3.208327531814575, "learning_rate": 8.143588870570441e-06, "loss": 0.1441, "step": 2338 }, { "epoch": 3.053524804177546, "grad_norm": 3.098215103149414, "learning_rate": 8.141944573533113e-06, "loss": 0.1025, "step": 2339 }, { "epoch": 3.0548302872062663, "grad_norm": 1.5849536657333374, "learning_rate": 8.140299714774824e-06, "loss": 0.0687, "step": 2340 }, { "epoch": 3.056135770234987, "grad_norm": 2.8804104328155518, "learning_rate": 8.138654294589644e-06, "loss": 0.0838, "step": 2341 }, { "epoch": 3.0574412532637076, "grad_norm": 2.641188383102417, "learning_rate": 8.137008313271741e-06, "loss": 0.071, "step": 2342 }, { "epoch": 3.058746736292428, "grad_norm": 2.2273764610290527, "learning_rate": 8.135361771115388e-06, "loss": 0.0699, "step": 2343 }, { "epoch": 3.060052219321149, "grad_norm": 2.4337551593780518, "learning_rate": 8.133714668414956e-06, "loss": 0.0753, "step": 2344 }, { "epoch": 3.0613577023498695, "grad_norm": 2.6304075717926025, "learning_rate": 8.132067005464914e-06, "loss": 0.0764, "step": 2345 }, { "epoch": 3.06266318537859, "grad_norm": 1.829929232597351, "learning_rate": 8.130418782559835e-06, "loss": 0.0572, "step": 2346 }, { "epoch": 3.063968668407311, "grad_norm": 2.7290573120117188, "learning_rate": 8.12876999999439e-06, "loss": 0.0731, "step": 2347 }, { "epoch": 3.0652741514360313, "grad_norm": 2.1656594276428223, "learning_rate": 8.127120658063351e-06, "loss": 0.0654, "step": 2348 }, { "epoch": 3.066579634464752, "grad_norm": 1.3606621026992798, "learning_rate": 8.125470757061589e-06, "loss": 0.0398, "step": 2349 }, { "epoch": 3.0678851174934727, "grad_norm": 2.1951255798339844, "learning_rate": 8.123820297284075e-06, "loss": 0.0435, "step": 2350 }, { "epoch": 3.069190600522193, "grad_norm": 1.181632161140442, "learning_rate": 8.122169279025882e-06, "loss": 0.0391, "step": 2351 }, { "epoch": 3.0704960835509136, "grad_norm": 1.5972039699554443, "learning_rate": 8.12051770258218e-06, "loss": 0.05, "step": 2352 }, { "epoch": 3.0718015665796345, "grad_norm": 1.814042568206787, "learning_rate": 8.11886556824824e-06, "loss": 0.0481, "step": 2353 }, { "epoch": 3.073107049608355, "grad_norm": 1.580129623413086, "learning_rate": 8.117212876319432e-06, "loss": 0.0452, "step": 2354 }, { "epoch": 3.074412532637076, "grad_norm": 2.1525285243988037, "learning_rate": 8.115559627091229e-06, "loss": 0.0577, "step": 2355 }, { "epoch": 3.0757180156657964, "grad_norm": 1.8001970052719116, "learning_rate": 8.113905820859196e-06, "loss": 0.0587, "step": 2356 }, { "epoch": 3.077023498694517, "grad_norm": 1.5521970987319946, "learning_rate": 8.11225145791901e-06, "loss": 0.0423, "step": 2357 }, { "epoch": 3.0783289817232378, "grad_norm": 1.0482590198516846, "learning_rate": 8.110596538566434e-06, "loss": 0.0375, "step": 2358 }, { "epoch": 3.0796344647519582, "grad_norm": 2.7500040531158447, "learning_rate": 8.10894106309734e-06, "loss": 0.0791, "step": 2359 }, { "epoch": 3.0809399477806787, "grad_norm": 1.8196872472763062, "learning_rate": 8.107285031807695e-06, "loss": 0.0513, "step": 2360 }, { "epoch": 3.0822454308093996, "grad_norm": 1.4419039487838745, "learning_rate": 8.105628444993566e-06, "loss": 0.0471, "step": 2361 }, { "epoch": 3.08355091383812, "grad_norm": 2.4254279136657715, "learning_rate": 8.103971302951116e-06, "loss": 0.0804, "step": 2362 }, { "epoch": 3.0848563968668405, "grad_norm": 2.993525743484497, "learning_rate": 8.10231360597662e-06, "loss": 0.0871, "step": 2363 }, { "epoch": 3.0861618798955615, "grad_norm": 5.089110374450684, "learning_rate": 8.100655354366437e-06, "loss": 0.1858, "step": 2364 }, { "epoch": 3.087467362924282, "grad_norm": 2.4185092449188232, "learning_rate": 8.098996548417034e-06, "loss": 0.072, "step": 2365 }, { "epoch": 3.0887728459530024, "grad_norm": 1.4787859916687012, "learning_rate": 8.097337188424972e-06, "loss": 0.0409, "step": 2366 }, { "epoch": 3.0900783289817233, "grad_norm": 2.5416836738586426, "learning_rate": 8.095677274686915e-06, "loss": 0.0772, "step": 2367 }, { "epoch": 3.0913838120104438, "grad_norm": 2.6314516067504883, "learning_rate": 8.094016807499624e-06, "loss": 0.0862, "step": 2368 }, { "epoch": 3.0926892950391647, "grad_norm": 1.170101284980774, "learning_rate": 8.092355787159958e-06, "loss": 0.0475, "step": 2369 }, { "epoch": 3.093994778067885, "grad_norm": 1.4015928506851196, "learning_rate": 8.090694213964882e-06, "loss": 0.0413, "step": 2370 }, { "epoch": 3.0953002610966056, "grad_norm": 3.062452554702759, "learning_rate": 8.089032088211449e-06, "loss": 0.1288, "step": 2371 }, { "epoch": 3.0966057441253265, "grad_norm": 3.7774112224578857, "learning_rate": 8.087369410196816e-06, "loss": 0.1459, "step": 2372 }, { "epoch": 3.097911227154047, "grad_norm": 2.559539556503296, "learning_rate": 8.085706180218243e-06, "loss": 0.0684, "step": 2373 }, { "epoch": 3.0992167101827675, "grad_norm": 2.1736948490142822, "learning_rate": 8.08404239857308e-06, "loss": 0.0672, "step": 2374 }, { "epoch": 3.1005221932114884, "grad_norm": 3.2882680892944336, "learning_rate": 8.08237806555878e-06, "loss": 0.1152, "step": 2375 }, { "epoch": 3.101827676240209, "grad_norm": 1.9450832605361938, "learning_rate": 8.080713181472898e-06, "loss": 0.0603, "step": 2376 }, { "epoch": 3.1031331592689293, "grad_norm": 1.6172608137130737, "learning_rate": 8.079047746613082e-06, "loss": 0.0574, "step": 2377 }, { "epoch": 3.10443864229765, "grad_norm": 2.4912679195404053, "learning_rate": 8.077381761277083e-06, "loss": 0.0893, "step": 2378 }, { "epoch": 3.1057441253263707, "grad_norm": 2.514925003051758, "learning_rate": 8.075715225762745e-06, "loss": 0.0497, "step": 2379 }, { "epoch": 3.1070496083550916, "grad_norm": 1.9957292079925537, "learning_rate": 8.074048140368014e-06, "loss": 0.0611, "step": 2380 }, { "epoch": 3.108355091383812, "grad_norm": 1.4828778505325317, "learning_rate": 8.072380505390934e-06, "loss": 0.0491, "step": 2381 }, { "epoch": 3.1096605744125325, "grad_norm": 1.9504576921463013, "learning_rate": 8.070712321129648e-06, "loss": 0.0539, "step": 2382 }, { "epoch": 3.1109660574412534, "grad_norm": 2.738281011581421, "learning_rate": 8.069043587882397e-06, "loss": 0.0989, "step": 2383 }, { "epoch": 3.112271540469974, "grad_norm": 2.49330472946167, "learning_rate": 8.067374305947515e-06, "loss": 0.0779, "step": 2384 }, { "epoch": 3.1135770234986944, "grad_norm": 1.8552166223526, "learning_rate": 8.065704475623442e-06, "loss": 0.056, "step": 2385 }, { "epoch": 3.1148825065274153, "grad_norm": 2.242624521255493, "learning_rate": 8.064034097208712e-06, "loss": 0.0672, "step": 2386 }, { "epoch": 3.1161879895561357, "grad_norm": 1.6343297958374023, "learning_rate": 8.062363171001957e-06, "loss": 0.0474, "step": 2387 }, { "epoch": 3.117493472584856, "grad_norm": 2.6920530796051025, "learning_rate": 8.060691697301908e-06, "loss": 0.0784, "step": 2388 }, { "epoch": 3.118798955613577, "grad_norm": 2.91678786277771, "learning_rate": 8.05901967640739e-06, "loss": 0.1017, "step": 2389 }, { "epoch": 3.1201044386422976, "grad_norm": 2.994342088699341, "learning_rate": 8.057347108617334e-06, "loss": 0.1261, "step": 2390 }, { "epoch": 3.1214099216710185, "grad_norm": 3.0396945476531982, "learning_rate": 8.055673994230762e-06, "loss": 0.1268, "step": 2391 }, { "epoch": 3.122715404699739, "grad_norm": 1.6301774978637695, "learning_rate": 8.054000333546795e-06, "loss": 0.0449, "step": 2392 }, { "epoch": 3.1240208877284594, "grad_norm": 3.3074307441711426, "learning_rate": 8.05232612686465e-06, "loss": 0.1498, "step": 2393 }, { "epoch": 3.1253263707571803, "grad_norm": 2.6727871894836426, "learning_rate": 8.050651374483647e-06, "loss": 0.0661, "step": 2394 }, { "epoch": 3.126631853785901, "grad_norm": 2.5931203365325928, "learning_rate": 8.0489760767032e-06, "loss": 0.0553, "step": 2395 }, { "epoch": 3.1279373368146213, "grad_norm": 1.1976072788238525, "learning_rate": 8.047300233822822e-06, "loss": 0.0403, "step": 2396 }, { "epoch": 3.129242819843342, "grad_norm": 1.5251294374465942, "learning_rate": 8.04562384614212e-06, "loss": 0.0528, "step": 2397 }, { "epoch": 3.1305483028720626, "grad_norm": 4.223265171051025, "learning_rate": 8.043946913960803e-06, "loss": 0.1779, "step": 2398 }, { "epoch": 3.131853785900783, "grad_norm": 5.674642562866211, "learning_rate": 8.042269437578673e-06, "loss": 0.1045, "step": 2399 }, { "epoch": 3.133159268929504, "grad_norm": 1.8073662519454956, "learning_rate": 8.04059141729563e-06, "loss": 0.0497, "step": 2400 }, { "epoch": 3.1344647519582245, "grad_norm": 1.4270871877670288, "learning_rate": 8.03891285341168e-06, "loss": 0.0427, "step": 2401 }, { "epoch": 3.1357702349869454, "grad_norm": 1.5010039806365967, "learning_rate": 8.037233746226911e-06, "loss": 0.0479, "step": 2402 }, { "epoch": 3.137075718015666, "grad_norm": 1.0706130266189575, "learning_rate": 8.035554096041517e-06, "loss": 0.0402, "step": 2403 }, { "epoch": 3.1383812010443863, "grad_norm": 1.7759274244308472, "learning_rate": 8.033873903155794e-06, "loss": 0.0581, "step": 2404 }, { "epoch": 3.1396866840731072, "grad_norm": 1.591680884361267, "learning_rate": 8.032193167870122e-06, "loss": 0.0616, "step": 2405 }, { "epoch": 3.1409921671018277, "grad_norm": 2.9845314025878906, "learning_rate": 8.03051189048499e-06, "loss": 0.1156, "step": 2406 }, { "epoch": 3.142297650130548, "grad_norm": 2.789412260055542, "learning_rate": 8.028830071300974e-06, "loss": 0.096, "step": 2407 }, { "epoch": 3.143603133159269, "grad_norm": 3.248490333557129, "learning_rate": 8.027147710618757e-06, "loss": 0.1112, "step": 2408 }, { "epoch": 3.1449086161879896, "grad_norm": 2.0329337120056152, "learning_rate": 8.02546480873911e-06, "loss": 0.0568, "step": 2409 }, { "epoch": 3.14621409921671, "grad_norm": 2.448091506958008, "learning_rate": 8.023781365962908e-06, "loss": 0.0759, "step": 2410 }, { "epoch": 3.147519582245431, "grad_norm": 1.5375114679336548, "learning_rate": 8.022097382591117e-06, "loss": 0.0467, "step": 2411 }, { "epoch": 3.1488250652741514, "grad_norm": 2.4986464977264404, "learning_rate": 8.0204128589248e-06, "loss": 0.0549, "step": 2412 }, { "epoch": 3.150130548302872, "grad_norm": 1.172946810722351, "learning_rate": 8.01872779526512e-06, "loss": 0.0406, "step": 2413 }, { "epoch": 3.151436031331593, "grad_norm": 2.847184658050537, "learning_rate": 8.017042191913335e-06, "loss": 0.0812, "step": 2414 }, { "epoch": 3.1527415143603132, "grad_norm": 2.179553508758545, "learning_rate": 8.015356049170802e-06, "loss": 0.0809, "step": 2415 }, { "epoch": 3.1540469973890337, "grad_norm": 3.116044759750366, "learning_rate": 8.013669367338966e-06, "loss": 0.1078, "step": 2416 }, { "epoch": 3.1553524804177546, "grad_norm": 1.4515472650527954, "learning_rate": 8.011982146719379e-06, "loss": 0.0457, "step": 2417 }, { "epoch": 3.156657963446475, "grad_norm": 2.177485466003418, "learning_rate": 8.010294387613682e-06, "loss": 0.0604, "step": 2418 }, { "epoch": 3.157963446475196, "grad_norm": 3.1564555168151855, "learning_rate": 8.008606090323615e-06, "loss": 0.0997, "step": 2419 }, { "epoch": 3.1592689295039165, "grad_norm": 2.0732004642486572, "learning_rate": 8.006917255151016e-06, "loss": 0.0754, "step": 2420 }, { "epoch": 3.160574412532637, "grad_norm": 1.7829563617706299, "learning_rate": 8.005227882397814e-06, "loss": 0.0675, "step": 2421 }, { "epoch": 3.161879895561358, "grad_norm": 2.2185959815979004, "learning_rate": 8.003537972366043e-06, "loss": 0.0521, "step": 2422 }, { "epoch": 3.1631853785900783, "grad_norm": 2.621936559677124, "learning_rate": 8.00184752535782e-06, "loss": 0.0916, "step": 2423 }, { "epoch": 3.164490861618799, "grad_norm": 2.013533115386963, "learning_rate": 8.000156541675367e-06, "loss": 0.0521, "step": 2424 }, { "epoch": 3.1657963446475197, "grad_norm": 1.049187183380127, "learning_rate": 7.998465021621005e-06, "loss": 0.0442, "step": 2425 }, { "epoch": 3.16710182767624, "grad_norm": 2.0861594676971436, "learning_rate": 7.996772965497143e-06, "loss": 0.0464, "step": 2426 }, { "epoch": 3.1684073107049606, "grad_norm": 1.2684996128082275, "learning_rate": 7.995080373606288e-06, "loss": 0.0419, "step": 2427 }, { "epoch": 3.1697127937336815, "grad_norm": 2.5054500102996826, "learning_rate": 7.993387246251046e-06, "loss": 0.0698, "step": 2428 }, { "epoch": 3.171018276762402, "grad_norm": 1.6138813495635986, "learning_rate": 7.991693583734115e-06, "loss": 0.0515, "step": 2429 }, { "epoch": 3.172323759791123, "grad_norm": 2.3526663780212402, "learning_rate": 7.989999386358292e-06, "loss": 0.069, "step": 2430 }, { "epoch": 3.1736292428198434, "grad_norm": 1.8785173892974854, "learning_rate": 7.988304654426465e-06, "loss": 0.0547, "step": 2431 }, { "epoch": 3.174934725848564, "grad_norm": 3.161043882369995, "learning_rate": 7.986609388241621e-06, "loss": 0.1018, "step": 2432 }, { "epoch": 3.1762402088772848, "grad_norm": 4.432081699371338, "learning_rate": 7.984913588106843e-06, "loss": 0.2009, "step": 2433 }, { "epoch": 3.177545691906005, "grad_norm": 3.297968864440918, "learning_rate": 7.98321725432531e-06, "loss": 0.1021, "step": 2434 }, { "epoch": 3.1788511749347257, "grad_norm": 1.0894938707351685, "learning_rate": 7.981520387200293e-06, "loss": 0.0385, "step": 2435 }, { "epoch": 3.1801566579634466, "grad_norm": 1.6087491512298584, "learning_rate": 7.979822987035157e-06, "loss": 0.0554, "step": 2436 }, { "epoch": 3.181462140992167, "grad_norm": 1.6627923250198364, "learning_rate": 7.97812505413337e-06, "loss": 0.0533, "step": 2437 }, { "epoch": 3.1827676240208875, "grad_norm": 1.6013997793197632, "learning_rate": 7.976426588798489e-06, "loss": 0.0378, "step": 2438 }, { "epoch": 3.1840731070496084, "grad_norm": 2.0800187587738037, "learning_rate": 7.97472759133417e-06, "loss": 0.0509, "step": 2439 }, { "epoch": 3.185378590078329, "grad_norm": 2.738882303237915, "learning_rate": 7.973028062044158e-06, "loss": 0.0557, "step": 2440 }, { "epoch": 3.18668407310705, "grad_norm": 3.185224771499634, "learning_rate": 7.971328001232298e-06, "loss": 0.1271, "step": 2441 }, { "epoch": 3.1879895561357703, "grad_norm": 2.105640411376953, "learning_rate": 7.969627409202535e-06, "loss": 0.0717, "step": 2442 }, { "epoch": 3.1892950391644908, "grad_norm": 1.358522891998291, "learning_rate": 7.967926286258894e-06, "loss": 0.0442, "step": 2443 }, { "epoch": 3.1906005221932117, "grad_norm": 1.6587517261505127, "learning_rate": 7.966224632705508e-06, "loss": 0.0436, "step": 2444 }, { "epoch": 3.191906005221932, "grad_norm": 1.9983346462249756, "learning_rate": 7.964522448846602e-06, "loss": 0.0645, "step": 2445 }, { "epoch": 3.1932114882506526, "grad_norm": 2.2097291946411133, "learning_rate": 7.962819734986494e-06, "loss": 0.0753, "step": 2446 }, { "epoch": 3.1945169712793735, "grad_norm": 2.677438497543335, "learning_rate": 7.961116491429596e-06, "loss": 0.1108, "step": 2447 }, { "epoch": 3.195822454308094, "grad_norm": 2.094862222671509, "learning_rate": 7.959412718480419e-06, "loss": 0.0542, "step": 2448 }, { "epoch": 3.1971279373368144, "grad_norm": 2.4472603797912598, "learning_rate": 7.957708416443564e-06, "loss": 0.0807, "step": 2449 }, { "epoch": 3.1984334203655354, "grad_norm": 2.486004114151001, "learning_rate": 7.956003585623727e-06, "loss": 0.0654, "step": 2450 }, { "epoch": 3.199738903394256, "grad_norm": 2.041811466217041, "learning_rate": 7.954298226325701e-06, "loss": 0.0603, "step": 2451 }, { "epoch": 3.2010443864229767, "grad_norm": 2.1919572353363037, "learning_rate": 7.952592338854374e-06, "loss": 0.0571, "step": 2452 }, { "epoch": 3.202349869451697, "grad_norm": 2.9395205974578857, "learning_rate": 7.950885923514724e-06, "loss": 0.0963, "step": 2453 }, { "epoch": 3.2036553524804177, "grad_norm": 1.3417216539382935, "learning_rate": 7.949178980611828e-06, "loss": 0.037, "step": 2454 }, { "epoch": 3.2049608355091386, "grad_norm": 1.4055885076522827, "learning_rate": 7.947471510450857e-06, "loss": 0.0381, "step": 2455 }, { "epoch": 3.206266318537859, "grad_norm": 2.3359646797180176, "learning_rate": 7.94576351333707e-06, "loss": 0.0704, "step": 2456 }, { "epoch": 3.2075718015665795, "grad_norm": 2.092123508453369, "learning_rate": 7.944054989575829e-06, "loss": 0.0568, "step": 2457 }, { "epoch": 3.2088772845953004, "grad_norm": 4.299978256225586, "learning_rate": 7.942345939472585e-06, "loss": 0.1305, "step": 2458 }, { "epoch": 3.210182767624021, "grad_norm": 1.2402383089065552, "learning_rate": 7.940636363332884e-06, "loss": 0.0466, "step": 2459 }, { "epoch": 3.2114882506527413, "grad_norm": 1.948562502861023, "learning_rate": 7.938926261462366e-06, "loss": 0.0479, "step": 2460 }, { "epoch": 3.2127937336814623, "grad_norm": 3.306370258331299, "learning_rate": 7.937215634166766e-06, "loss": 0.1047, "step": 2461 }, { "epoch": 3.2140992167101827, "grad_norm": 2.0487399101257324, "learning_rate": 7.935504481751911e-06, "loss": 0.0638, "step": 2462 }, { "epoch": 3.215404699738903, "grad_norm": 1.9900611639022827, "learning_rate": 7.933792804523726e-06, "loss": 0.0621, "step": 2463 }, { "epoch": 3.216710182767624, "grad_norm": 1.1753756999969482, "learning_rate": 7.932080602788222e-06, "loss": 0.0364, "step": 2464 }, { "epoch": 3.2180156657963446, "grad_norm": 1.974670648574829, "learning_rate": 7.930367876851513e-06, "loss": 0.0414, "step": 2465 }, { "epoch": 3.2193211488250655, "grad_norm": 2.6965410709381104, "learning_rate": 7.9286546270198e-06, "loss": 0.0934, "step": 2466 }, { "epoch": 3.220626631853786, "grad_norm": 3.662754774093628, "learning_rate": 7.92694085359938e-06, "loss": 0.1621, "step": 2467 }, { "epoch": 3.2219321148825064, "grad_norm": 2.2270076274871826, "learning_rate": 7.925226556896646e-06, "loss": 0.0735, "step": 2468 }, { "epoch": 3.2232375979112273, "grad_norm": 1.2094484567642212, "learning_rate": 7.92351173721808e-06, "loss": 0.0394, "step": 2469 }, { "epoch": 3.224543080939948, "grad_norm": 1.5824965238571167, "learning_rate": 7.92179639487026e-06, "loss": 0.0433, "step": 2470 }, { "epoch": 3.2258485639686683, "grad_norm": 2.4674224853515625, "learning_rate": 7.920080530159859e-06, "loss": 0.0722, "step": 2471 }, { "epoch": 3.227154046997389, "grad_norm": 1.3987195491790771, "learning_rate": 7.91836414339364e-06, "loss": 0.0414, "step": 2472 }, { "epoch": 3.2284595300261096, "grad_norm": 1.8914343118667603, "learning_rate": 7.91664723487846e-06, "loss": 0.0463, "step": 2473 }, { "epoch": 3.22976501305483, "grad_norm": 1.3807590007781982, "learning_rate": 7.914929804921271e-06, "loss": 0.052, "step": 2474 }, { "epoch": 3.231070496083551, "grad_norm": 1.6758167743682861, "learning_rate": 7.913211853829117e-06, "loss": 0.0452, "step": 2475 }, { "epoch": 3.2323759791122715, "grad_norm": 0.8737969994544983, "learning_rate": 7.911493381909137e-06, "loss": 0.036, "step": 2476 }, { "epoch": 3.233681462140992, "grad_norm": 2.3536150455474854, "learning_rate": 7.90977438946856e-06, "loss": 0.0679, "step": 2477 }, { "epoch": 3.234986945169713, "grad_norm": 1.6790705919265747, "learning_rate": 7.908054876814711e-06, "loss": 0.0497, "step": 2478 }, { "epoch": 3.2362924281984333, "grad_norm": 1.18221914768219, "learning_rate": 7.906334844255007e-06, "loss": 0.0374, "step": 2479 }, { "epoch": 3.2375979112271542, "grad_norm": 1.8799299001693726, "learning_rate": 7.904614292096954e-06, "loss": 0.0535, "step": 2480 }, { "epoch": 3.2389033942558747, "grad_norm": 1.3944048881530762, "learning_rate": 7.90289322064816e-06, "loss": 0.0407, "step": 2481 }, { "epoch": 3.240208877284595, "grad_norm": 1.8899251222610474, "learning_rate": 7.901171630216318e-06, "loss": 0.0414, "step": 2482 }, { "epoch": 3.241514360313316, "grad_norm": 1.824993371963501, "learning_rate": 7.899449521109214e-06, "loss": 0.064, "step": 2483 }, { "epoch": 3.2428198433420365, "grad_norm": 1.72813880443573, "learning_rate": 7.897726893634732e-06, "loss": 0.0498, "step": 2484 }, { "epoch": 3.244125326370757, "grad_norm": 1.8778036832809448, "learning_rate": 7.896003748100844e-06, "loss": 0.064, "step": 2485 }, { "epoch": 3.245430809399478, "grad_norm": 2.881976366043091, "learning_rate": 7.894280084815618e-06, "loss": 0.0907, "step": 2486 }, { "epoch": 3.2467362924281984, "grad_norm": 1.8811758756637573, "learning_rate": 7.89255590408721e-06, "loss": 0.0504, "step": 2487 }, { "epoch": 3.248041775456919, "grad_norm": 1.5723365545272827, "learning_rate": 7.890831206223873e-06, "loss": 0.0536, "step": 2488 }, { "epoch": 3.2493472584856398, "grad_norm": 1.4082533121109009, "learning_rate": 7.88910599153395e-06, "loss": 0.0464, "step": 2489 }, { "epoch": 3.2506527415143602, "grad_norm": 2.82940411567688, "learning_rate": 7.887380260325878e-06, "loss": 0.0813, "step": 2490 }, { "epoch": 3.251958224543081, "grad_norm": 1.8589000701904297, "learning_rate": 7.885654012908186e-06, "loss": 0.0501, "step": 2491 }, { "epoch": 3.2532637075718016, "grad_norm": 1.6726198196411133, "learning_rate": 7.883927249589493e-06, "loss": 0.0398, "step": 2492 }, { "epoch": 3.254569190600522, "grad_norm": 2.5619328022003174, "learning_rate": 7.882199970678512e-06, "loss": 0.0784, "step": 2493 }, { "epoch": 3.255874673629243, "grad_norm": 2.445681571960449, "learning_rate": 7.880472176484051e-06, "loss": 0.059, "step": 2494 }, { "epoch": 3.2571801566579635, "grad_norm": 3.183565855026245, "learning_rate": 7.878743867315004e-06, "loss": 0.112, "step": 2495 }, { "epoch": 3.258485639686684, "grad_norm": 1.9520323276519775, "learning_rate": 7.877015043480363e-06, "loss": 0.0569, "step": 2496 }, { "epoch": 3.259791122715405, "grad_norm": 2.6844663619995117, "learning_rate": 7.875285705289207e-06, "loss": 0.0851, "step": 2497 }, { "epoch": 3.2610966057441253, "grad_norm": 1.2376049757003784, "learning_rate": 7.87355585305071e-06, "loss": 0.0474, "step": 2498 }, { "epoch": 3.2624020887728458, "grad_norm": 1.6593073606491089, "learning_rate": 7.87182548707414e-06, "loss": 0.053, "step": 2499 }, { "epoch": 3.2637075718015667, "grad_norm": 1.5427733659744263, "learning_rate": 7.87009460766885e-06, "loss": 0.0448, "step": 2500 }, { "epoch": 3.265013054830287, "grad_norm": 1.3970359563827515, "learning_rate": 7.86836321514429e-06, "loss": 0.0414, "step": 2501 }, { "epoch": 3.266318537859008, "grad_norm": 1.6662153005599976, "learning_rate": 7.866631309810003e-06, "loss": 0.0578, "step": 2502 }, { "epoch": 3.2676240208877285, "grad_norm": 1.5045440196990967, "learning_rate": 7.86489889197562e-06, "loss": 0.0378, "step": 2503 }, { "epoch": 3.268929503916449, "grad_norm": 1.6150271892547607, "learning_rate": 7.863165961950863e-06, "loss": 0.0387, "step": 2504 }, { "epoch": 3.27023498694517, "grad_norm": 1.1446775197982788, "learning_rate": 7.86143252004555e-06, "loss": 0.0485, "step": 2505 }, { "epoch": 3.2715404699738904, "grad_norm": 2.745053768157959, "learning_rate": 7.859698566569586e-06, "loss": 0.1115, "step": 2506 }, { "epoch": 3.272845953002611, "grad_norm": 1.5964254140853882, "learning_rate": 7.85796410183297e-06, "loss": 0.0464, "step": 2507 }, { "epoch": 3.2741514360313317, "grad_norm": 1.2686421871185303, "learning_rate": 7.856229126145795e-06, "loss": 0.0394, "step": 2508 }, { "epoch": 3.275456919060052, "grad_norm": 1.9726738929748535, "learning_rate": 7.85449363981824e-06, "loss": 0.0413, "step": 2509 }, { "epoch": 3.2767624020887727, "grad_norm": 1.5561535358428955, "learning_rate": 7.852757643160575e-06, "loss": 0.0376, "step": 2510 }, { "epoch": 3.2780678851174936, "grad_norm": 1.8126353025436401, "learning_rate": 7.851021136483165e-06, "loss": 0.0615, "step": 2511 }, { "epoch": 3.279373368146214, "grad_norm": 2.265113353729248, "learning_rate": 7.849284120096467e-06, "loss": 0.0564, "step": 2512 }, { "epoch": 3.280678851174935, "grad_norm": 1.4022085666656494, "learning_rate": 7.847546594311026e-06, "loss": 0.0411, "step": 2513 }, { "epoch": 3.2819843342036554, "grad_norm": 1.740009069442749, "learning_rate": 7.84580855943748e-06, "loss": 0.0482, "step": 2514 }, { "epoch": 3.283289817232376, "grad_norm": 2.8635735511779785, "learning_rate": 7.844070015786556e-06, "loss": 0.1194, "step": 2515 }, { "epoch": 3.2845953002610964, "grad_norm": 1.2145297527313232, "learning_rate": 7.842330963669072e-06, "loss": 0.0379, "step": 2516 }, { "epoch": 3.2859007832898173, "grad_norm": 3.4107160568237305, "learning_rate": 7.84059140339594e-06, "loss": 0.1258, "step": 2517 }, { "epoch": 3.2872062663185377, "grad_norm": 1.8389348983764648, "learning_rate": 7.838851335278161e-06, "loss": 0.0658, "step": 2518 }, { "epoch": 3.2885117493472587, "grad_norm": 2.6536471843719482, "learning_rate": 7.837110759626825e-06, "loss": 0.0623, "step": 2519 }, { "epoch": 3.289817232375979, "grad_norm": 0.9475728869438171, "learning_rate": 7.835369676753116e-06, "loss": 0.0413, "step": 2520 }, { "epoch": 3.2911227154046996, "grad_norm": 1.74058997631073, "learning_rate": 7.833628086968304e-06, "loss": 0.0502, "step": 2521 }, { "epoch": 3.2924281984334205, "grad_norm": 1.2026323080062866, "learning_rate": 7.831885990583757e-06, "loss": 0.0401, "step": 2522 }, { "epoch": 3.293733681462141, "grad_norm": 1.5306260585784912, "learning_rate": 7.830143387910926e-06, "loss": 0.0401, "step": 2523 }, { "epoch": 3.2950391644908614, "grad_norm": 1.7440180778503418, "learning_rate": 7.828400279261358e-06, "loss": 0.0523, "step": 2524 }, { "epoch": 3.2963446475195823, "grad_norm": 1.85578191280365, "learning_rate": 7.826656664946687e-06, "loss": 0.044, "step": 2525 }, { "epoch": 3.297650130548303, "grad_norm": 1.9574214220046997, "learning_rate": 7.824912545278639e-06, "loss": 0.0496, "step": 2526 }, { "epoch": 3.2989556135770233, "grad_norm": 1.4065427780151367, "learning_rate": 7.82316792056903e-06, "loss": 0.0407, "step": 2527 }, { "epoch": 3.300261096605744, "grad_norm": 0.9905129671096802, "learning_rate": 7.821422791129767e-06, "loss": 0.0411, "step": 2528 }, { "epoch": 3.3015665796344646, "grad_norm": 1.611026406288147, "learning_rate": 7.819677157272845e-06, "loss": 0.0407, "step": 2529 }, { "epoch": 3.3028720626631856, "grad_norm": 1.5317682027816772, "learning_rate": 7.817931019310352e-06, "loss": 0.0417, "step": 2530 }, { "epoch": 3.304177545691906, "grad_norm": 1.6986154317855835, "learning_rate": 7.816184377554463e-06, "loss": 0.0473, "step": 2531 }, { "epoch": 3.3054830287206265, "grad_norm": 1.4276055097579956, "learning_rate": 7.814437232317446e-06, "loss": 0.0456, "step": 2532 }, { "epoch": 3.3067885117493474, "grad_norm": 1.9061105251312256, "learning_rate": 7.81268958391166e-06, "loss": 0.0655, "step": 2533 }, { "epoch": 3.308093994778068, "grad_norm": 1.3641295433044434, "learning_rate": 7.810941432649548e-06, "loss": 0.0386, "step": 2534 }, { "epoch": 3.3093994778067883, "grad_norm": 2.79325532913208, "learning_rate": 7.80919277884365e-06, "loss": 0.0855, "step": 2535 }, { "epoch": 3.3107049608355092, "grad_norm": 1.085235834121704, "learning_rate": 7.807443622806592e-06, "loss": 0.0347, "step": 2536 }, { "epoch": 3.3120104438642297, "grad_norm": 3.5964457988739014, "learning_rate": 7.80569396485109e-06, "loss": 0.1181, "step": 2537 }, { "epoch": 3.31331592689295, "grad_norm": 1.1285696029663086, "learning_rate": 7.803943805289946e-06, "loss": 0.0381, "step": 2538 }, { "epoch": 3.314621409921671, "grad_norm": 3.716900587081909, "learning_rate": 7.802193144436064e-06, "loss": 0.1348, "step": 2539 }, { "epoch": 3.3159268929503916, "grad_norm": 0.861534595489502, "learning_rate": 7.800441982602423e-06, "loss": 0.0364, "step": 2540 }, { "epoch": 3.3172323759791125, "grad_norm": 1.1106094121932983, "learning_rate": 7.7986903201021e-06, "loss": 0.0377, "step": 2541 }, { "epoch": 3.318537859007833, "grad_norm": 1.760759711265564, "learning_rate": 7.79693815724826e-06, "loss": 0.0567, "step": 2542 }, { "epoch": 3.3198433420365534, "grad_norm": 1.7553333044052124, "learning_rate": 7.795185494354158e-06, "loss": 0.054, "step": 2543 }, { "epoch": 3.3211488250652743, "grad_norm": 2.547304153442383, "learning_rate": 7.793432331733135e-06, "loss": 0.0584, "step": 2544 }, { "epoch": 3.322454308093995, "grad_norm": 1.300601601600647, "learning_rate": 7.791678669698624e-06, "loss": 0.0373, "step": 2545 }, { "epoch": 3.3237597911227152, "grad_norm": 2.475071907043457, "learning_rate": 7.789924508564149e-06, "loss": 0.0624, "step": 2546 }, { "epoch": 3.325065274151436, "grad_norm": 1.2528338432312012, "learning_rate": 7.788169848643319e-06, "loss": 0.0289, "step": 2547 }, { "epoch": 3.3263707571801566, "grad_norm": 2.2197775840759277, "learning_rate": 7.786414690249834e-06, "loss": 0.0629, "step": 2548 }, { "epoch": 3.327676240208877, "grad_norm": 3.5401663780212402, "learning_rate": 7.784659033697488e-06, "loss": 0.1395, "step": 2549 }, { "epoch": 3.328981723237598, "grad_norm": 1.6441315412521362, "learning_rate": 7.782902879300154e-06, "loss": 0.0436, "step": 2550 }, { "epoch": 3.3302872062663185, "grad_norm": 2.015650987625122, "learning_rate": 7.781146227371803e-06, "loss": 0.0532, "step": 2551 }, { "epoch": 3.3315926892950394, "grad_norm": 1.961761236190796, "learning_rate": 7.779389078226492e-06, "loss": 0.054, "step": 2552 }, { "epoch": 3.33289817232376, "grad_norm": 1.8513764142990112, "learning_rate": 7.777631432178363e-06, "loss": 0.0477, "step": 2553 }, { "epoch": 3.3342036553524803, "grad_norm": 2.4187116622924805, "learning_rate": 7.775873289541653e-06, "loss": 0.0704, "step": 2554 }, { "epoch": 3.335509138381201, "grad_norm": 3.165609359741211, "learning_rate": 7.774114650630683e-06, "loss": 0.0947, "step": 2555 }, { "epoch": 3.3368146214099217, "grad_norm": 1.811149001121521, "learning_rate": 7.772355515759868e-06, "loss": 0.0617, "step": 2556 }, { "epoch": 3.338120104438642, "grad_norm": 0.7971915006637573, "learning_rate": 7.770595885243704e-06, "loss": 0.0354, "step": 2557 }, { "epoch": 3.339425587467363, "grad_norm": 3.1833105087280273, "learning_rate": 7.768835759396784e-06, "loss": 0.1159, "step": 2558 }, { "epoch": 3.3407310704960835, "grad_norm": 1.419297695159912, "learning_rate": 7.767075138533784e-06, "loss": 0.0497, "step": 2559 }, { "epoch": 3.342036553524804, "grad_norm": 1.7927831411361694, "learning_rate": 7.76531402296947e-06, "loss": 0.0502, "step": 2560 }, { "epoch": 3.343342036553525, "grad_norm": 2.288058042526245, "learning_rate": 7.763552413018696e-06, "loss": 0.0684, "step": 2561 }, { "epoch": 3.3446475195822454, "grad_norm": 0.9647486209869385, "learning_rate": 7.761790308996406e-06, "loss": 0.0311, "step": 2562 }, { "epoch": 3.3459530026109663, "grad_norm": 2.0151686668395996, "learning_rate": 7.760027711217632e-06, "loss": 0.0556, "step": 2563 }, { "epoch": 3.3472584856396868, "grad_norm": 1.3200900554656982, "learning_rate": 7.758264619997492e-06, "loss": 0.0351, "step": 2564 }, { "epoch": 3.348563968668407, "grad_norm": 2.560396432876587, "learning_rate": 7.756501035651194e-06, "loss": 0.0924, "step": 2565 }, { "epoch": 3.349869451697128, "grad_norm": 1.8153531551361084, "learning_rate": 7.754736958494035e-06, "loss": 0.0497, "step": 2566 }, { "epoch": 3.3511749347258486, "grad_norm": 1.4534152746200562, "learning_rate": 7.752972388841398e-06, "loss": 0.0455, "step": 2567 }, { "epoch": 3.352480417754569, "grad_norm": 1.8594975471496582, "learning_rate": 7.751207327008754e-06, "loss": 0.0493, "step": 2568 }, { "epoch": 3.35378590078329, "grad_norm": 1.9180917739868164, "learning_rate": 7.749441773311663e-06, "loss": 0.0478, "step": 2569 }, { "epoch": 3.3550913838120104, "grad_norm": 1.4175883531570435, "learning_rate": 7.747675728065778e-06, "loss": 0.0386, "step": 2570 }, { "epoch": 3.356396866840731, "grad_norm": 1.0177656412124634, "learning_rate": 7.74590919158683e-06, "loss": 0.038, "step": 2571 }, { "epoch": 3.357702349869452, "grad_norm": 1.5703392028808594, "learning_rate": 7.744142164190643e-06, "loss": 0.0501, "step": 2572 }, { "epoch": 3.3590078328981723, "grad_norm": 1.5348613262176514, "learning_rate": 7.742374646193129e-06, "loss": 0.0392, "step": 2573 }, { "epoch": 3.360313315926893, "grad_norm": 3.5389914512634277, "learning_rate": 7.740606637910289e-06, "loss": 0.1245, "step": 2574 }, { "epoch": 3.3616187989556137, "grad_norm": 2.451744318008423, "learning_rate": 7.738838139658206e-06, "loss": 0.0752, "step": 2575 }, { "epoch": 3.362924281984334, "grad_norm": 1.7922784090042114, "learning_rate": 7.737069151753058e-06, "loss": 0.0575, "step": 2576 }, { "epoch": 3.3642297650130546, "grad_norm": 1.4721416234970093, "learning_rate": 7.735299674511107e-06, "loss": 0.0588, "step": 2577 }, { "epoch": 3.3655352480417755, "grad_norm": 2.918922185897827, "learning_rate": 7.733529708248699e-06, "loss": 0.0982, "step": 2578 }, { "epoch": 3.366840731070496, "grad_norm": 2.0955660343170166, "learning_rate": 7.731759253282276e-06, "loss": 0.0455, "step": 2579 }, { "epoch": 3.368146214099217, "grad_norm": 2.1960535049438477, "learning_rate": 7.729988309928359e-06, "loss": 0.0549, "step": 2580 }, { "epoch": 3.3694516971279374, "grad_norm": 1.685117483139038, "learning_rate": 7.728216878503557e-06, "loss": 0.0504, "step": 2581 }, { "epoch": 3.370757180156658, "grad_norm": 2.137101888656616, "learning_rate": 7.726444959324572e-06, "loss": 0.0596, "step": 2582 }, { "epoch": 3.3720626631853787, "grad_norm": 1.1969618797302246, "learning_rate": 7.724672552708193e-06, "loss": 0.0386, "step": 2583 }, { "epoch": 3.373368146214099, "grad_norm": 1.3502857685089111, "learning_rate": 7.722899658971288e-06, "loss": 0.0429, "step": 2584 }, { "epoch": 3.3746736292428197, "grad_norm": 1.8704237937927246, "learning_rate": 7.72112627843082e-06, "loss": 0.0576, "step": 2585 }, { "epoch": 3.3759791122715406, "grad_norm": 1.5152759552001953, "learning_rate": 7.719352411403836e-06, "loss": 0.0522, "step": 2586 }, { "epoch": 3.377284595300261, "grad_norm": 2.024496555328369, "learning_rate": 7.717578058207467e-06, "loss": 0.0417, "step": 2587 }, { "epoch": 3.3785900783289815, "grad_norm": 1.47026526927948, "learning_rate": 7.71580321915894e-06, "loss": 0.0381, "step": 2588 }, { "epoch": 3.3798955613577024, "grad_norm": 2.0034494400024414, "learning_rate": 7.714027894575558e-06, "loss": 0.0575, "step": 2589 }, { "epoch": 3.381201044386423, "grad_norm": 2.096081256866455, "learning_rate": 7.712252084774718e-06, "loss": 0.0616, "step": 2590 }, { "epoch": 3.382506527415144, "grad_norm": 1.9239627122879028, "learning_rate": 7.710475790073903e-06, "loss": 0.0634, "step": 2591 }, { "epoch": 3.3838120104438643, "grad_norm": 1.0032705068588257, "learning_rate": 7.708699010790678e-06, "loss": 0.0378, "step": 2592 }, { "epoch": 3.3851174934725847, "grad_norm": 2.893388509750366, "learning_rate": 7.706921747242701e-06, "loss": 0.0566, "step": 2593 }, { "epoch": 3.3864229765013056, "grad_norm": 0.9529011249542236, "learning_rate": 7.705143999747712e-06, "loss": 0.0337, "step": 2594 }, { "epoch": 3.387728459530026, "grad_norm": 1.3210817575454712, "learning_rate": 7.703365768623538e-06, "loss": 0.0499, "step": 2595 }, { "epoch": 3.3890339425587466, "grad_norm": 1.3464261293411255, "learning_rate": 7.701587054188098e-06, "loss": 0.0423, "step": 2596 }, { "epoch": 3.3903394255874675, "grad_norm": 4.668654918670654, "learning_rate": 7.699807856759385e-06, "loss": 0.1909, "step": 2597 }, { "epoch": 3.391644908616188, "grad_norm": 1.68687903881073, "learning_rate": 7.698028176655494e-06, "loss": 0.0571, "step": 2598 }, { "epoch": 3.3929503916449084, "grad_norm": 1.6898947954177856, "learning_rate": 7.696248014194596e-06, "loss": 0.0609, "step": 2599 }, { "epoch": 3.3942558746736293, "grad_norm": 2.2149479389190674, "learning_rate": 7.694467369694948e-06, "loss": 0.0451, "step": 2600 }, { "epoch": 3.39556135770235, "grad_norm": 1.978568434715271, "learning_rate": 7.692686243474898e-06, "loss": 0.0421, "step": 2601 }, { "epoch": 3.3968668407310707, "grad_norm": 1.6070466041564941, "learning_rate": 7.690904635852879e-06, "loss": 0.0391, "step": 2602 }, { "epoch": 3.398172323759791, "grad_norm": 1.4013645648956299, "learning_rate": 7.68912254714741e-06, "loss": 0.0461, "step": 2603 }, { "epoch": 3.3994778067885116, "grad_norm": 1.6206039190292358, "learning_rate": 7.68733997767709e-06, "loss": 0.0462, "step": 2604 }, { "epoch": 3.4007832898172325, "grad_norm": 0.9994009137153625, "learning_rate": 7.685556927760616e-06, "loss": 0.0341, "step": 2605 }, { "epoch": 3.402088772845953, "grad_norm": 1.8120311498641968, "learning_rate": 7.683773397716758e-06, "loss": 0.0481, "step": 2606 }, { "epoch": 3.4033942558746735, "grad_norm": 1.2411257028579712, "learning_rate": 7.681989387864381e-06, "loss": 0.0382, "step": 2607 }, { "epoch": 3.4046997389033944, "grad_norm": 1.5058211088180542, "learning_rate": 7.680204898522433e-06, "loss": 0.0509, "step": 2608 }, { "epoch": 3.406005221932115, "grad_norm": 1.9172269105911255, "learning_rate": 7.678419930009946e-06, "loss": 0.0368, "step": 2609 }, { "epoch": 3.4073107049608353, "grad_norm": 1.5108345746994019, "learning_rate": 7.67663448264604e-06, "loss": 0.0441, "step": 2610 }, { "epoch": 3.4086161879895562, "grad_norm": 1.325252652168274, "learning_rate": 7.674848556749917e-06, "loss": 0.0424, "step": 2611 }, { "epoch": 3.4099216710182767, "grad_norm": 1.674212098121643, "learning_rate": 7.67306215264087e-06, "loss": 0.0433, "step": 2612 }, { "epoch": 3.4112271540469976, "grad_norm": 3.1836140155792236, "learning_rate": 7.671275270638274e-06, "loss": 0.0809, "step": 2613 }, { "epoch": 3.412532637075718, "grad_norm": 1.0253986120224, "learning_rate": 7.66948791106159e-06, "loss": 0.0304, "step": 2614 }, { "epoch": 3.4138381201044385, "grad_norm": 5.695417881011963, "learning_rate": 7.667700074230365e-06, "loss": 0.0845, "step": 2615 }, { "epoch": 3.4151436031331595, "grad_norm": 1.070473551750183, "learning_rate": 7.665911760464231e-06, "loss": 0.0317, "step": 2616 }, { "epoch": 3.41644908616188, "grad_norm": 1.7467083930969238, "learning_rate": 7.664122970082901e-06, "loss": 0.0535, "step": 2617 }, { "epoch": 3.4177545691906004, "grad_norm": 2.9163339138031006, "learning_rate": 7.662333703406185e-06, "loss": 0.082, "step": 2618 }, { "epoch": 3.4190600522193213, "grad_norm": 1.6499325037002563, "learning_rate": 7.660543960753961e-06, "loss": 0.0514, "step": 2619 }, { "epoch": 3.4203655352480418, "grad_norm": 2.039182186126709, "learning_rate": 7.65875374244621e-06, "loss": 0.0456, "step": 2620 }, { "epoch": 3.4216710182767622, "grad_norm": 0.8305462002754211, "learning_rate": 7.656963048802986e-06, "loss": 0.0346, "step": 2621 }, { "epoch": 3.422976501305483, "grad_norm": 1.2585537433624268, "learning_rate": 7.655171880144429e-06, "loss": 0.0456, "step": 2622 }, { "epoch": 3.4242819843342036, "grad_norm": 1.2120765447616577, "learning_rate": 7.653380236790773e-06, "loss": 0.0341, "step": 2623 }, { "epoch": 3.4255874673629245, "grad_norm": 0.8369303941726685, "learning_rate": 7.651588119062324e-06, "loss": 0.0362, "step": 2624 }, { "epoch": 3.426892950391645, "grad_norm": 2.4857537746429443, "learning_rate": 7.649795527279482e-06, "loss": 0.0747, "step": 2625 }, { "epoch": 3.4281984334203655, "grad_norm": 2.156614065170288, "learning_rate": 7.64800246176273e-06, "loss": 0.0568, "step": 2626 }, { "epoch": 3.429503916449086, "grad_norm": 1.9288692474365234, "learning_rate": 7.646208922832631e-06, "loss": 0.0494, "step": 2627 }, { "epoch": 3.430809399477807, "grad_norm": 3.023404836654663, "learning_rate": 7.644414910809838e-06, "loss": 0.1079, "step": 2628 }, { "epoch": 3.4321148825065273, "grad_norm": 2.004993200302124, "learning_rate": 7.64262042601509e-06, "loss": 0.0494, "step": 2629 }, { "epoch": 3.433420365535248, "grad_norm": 1.7470046281814575, "learning_rate": 7.640825468769204e-06, "loss": 0.052, "step": 2630 }, { "epoch": 3.4347258485639687, "grad_norm": 1.3004729747772217, "learning_rate": 7.639030039393083e-06, "loss": 0.0421, "step": 2631 }, { "epoch": 3.436031331592689, "grad_norm": 1.310793161392212, "learning_rate": 7.63723413820772e-06, "loss": 0.0364, "step": 2632 }, { "epoch": 3.43733681462141, "grad_norm": 1.3017972707748413, "learning_rate": 7.635437765534186e-06, "loss": 0.051, "step": 2633 }, { "epoch": 3.4386422976501305, "grad_norm": 1.350611925125122, "learning_rate": 7.633640921693642e-06, "loss": 0.0454, "step": 2634 }, { "epoch": 3.439947780678851, "grad_norm": 3.739849328994751, "learning_rate": 7.631843607007328e-06, "loss": 0.1384, "step": 2635 }, { "epoch": 3.441253263707572, "grad_norm": 1.5101021528244019, "learning_rate": 7.630045821796567e-06, "loss": 0.0431, "step": 2636 }, { "epoch": 3.4425587467362924, "grad_norm": 1.392733097076416, "learning_rate": 7.628247566382776e-06, "loss": 0.0454, "step": 2637 }, { "epoch": 3.443864229765013, "grad_norm": 2.6273722648620605, "learning_rate": 7.626448841087446e-06, "loss": 0.0797, "step": 2638 }, { "epoch": 3.4451697127937337, "grad_norm": 2.8379714488983154, "learning_rate": 7.624649646232154e-06, "loss": 0.0631, "step": 2639 }, { "epoch": 3.446475195822454, "grad_norm": 1.0590031147003174, "learning_rate": 7.622849982138566e-06, "loss": 0.0421, "step": 2640 }, { "epoch": 3.447780678851175, "grad_norm": 1.737034559249878, "learning_rate": 7.621049849128426e-06, "loss": 0.0504, "step": 2641 }, { "epoch": 3.4490861618798956, "grad_norm": 2.6455252170562744, "learning_rate": 7.619249247523563e-06, "loss": 0.0653, "step": 2642 }, { "epoch": 3.450391644908616, "grad_norm": 1.5327024459838867, "learning_rate": 7.617448177645894e-06, "loss": 0.0354, "step": 2643 }, { "epoch": 3.451697127937337, "grad_norm": 1.4662562608718872, "learning_rate": 7.615646639817414e-06, "loss": 0.0475, "step": 2644 }, { "epoch": 3.4530026109660574, "grad_norm": 1.778686761856079, "learning_rate": 7.613844634360205e-06, "loss": 0.0423, "step": 2645 }, { "epoch": 3.454308093994778, "grad_norm": 2.0599887371063232, "learning_rate": 7.612042161596433e-06, "loss": 0.0632, "step": 2646 }, { "epoch": 3.455613577023499, "grad_norm": 1.3452439308166504, "learning_rate": 7.610239221848345e-06, "loss": 0.043, "step": 2647 }, { "epoch": 3.4569190600522193, "grad_norm": 1.1028183698654175, "learning_rate": 7.608435815438275e-06, "loss": 0.0282, "step": 2648 }, { "epoch": 3.4582245430809397, "grad_norm": 2.986908435821533, "learning_rate": 7.606631942688636e-06, "loss": 0.1047, "step": 2649 }, { "epoch": 3.4595300261096606, "grad_norm": 1.027195692062378, "learning_rate": 7.604827603921928e-06, "loss": 0.038, "step": 2650 }, { "epoch": 3.460835509138381, "grad_norm": 1.5456421375274658, "learning_rate": 7.603022799460731e-06, "loss": 0.0398, "step": 2651 }, { "epoch": 3.462140992167102, "grad_norm": 1.1341956853866577, "learning_rate": 7.6012175296277155e-06, "loss": 0.0378, "step": 2652 }, { "epoch": 3.4634464751958225, "grad_norm": 2.748147487640381, "learning_rate": 7.5994117947456235e-06, "loss": 0.0808, "step": 2653 }, { "epoch": 3.464751958224543, "grad_norm": 1.9158810377120972, "learning_rate": 7.5976055951372905e-06, "loss": 0.0483, "step": 2654 }, { "epoch": 3.466057441253264, "grad_norm": 1.358317255973816, "learning_rate": 7.595798931125632e-06, "loss": 0.0362, "step": 2655 }, { "epoch": 3.4673629242819843, "grad_norm": 2.5429093837738037, "learning_rate": 7.593991803033642e-06, "loss": 0.0816, "step": 2656 }, { "epoch": 3.468668407310705, "grad_norm": 1.4463530778884888, "learning_rate": 7.592184211184404e-06, "loss": 0.0463, "step": 2657 }, { "epoch": 3.4699738903394257, "grad_norm": 0.7134208083152771, "learning_rate": 7.5903761559010815e-06, "loss": 0.03, "step": 2658 }, { "epoch": 3.471279373368146, "grad_norm": 1.3389939069747925, "learning_rate": 7.58856763750692e-06, "loss": 0.043, "step": 2659 }, { "epoch": 3.4725848563968666, "grad_norm": 1.883532166481018, "learning_rate": 7.58675865632525e-06, "loss": 0.0701, "step": 2660 }, { "epoch": 3.4738903394255876, "grad_norm": 2.427351474761963, "learning_rate": 7.5849492126794824e-06, "loss": 0.0531, "step": 2661 }, { "epoch": 3.475195822454308, "grad_norm": 1.8123908042907715, "learning_rate": 7.583139306893113e-06, "loss": 0.049, "step": 2662 }, { "epoch": 3.476501305483029, "grad_norm": 2.2688679695129395, "learning_rate": 7.581328939289716e-06, "loss": 0.0674, "step": 2663 }, { "epoch": 3.4778067885117494, "grad_norm": 2.7659173011779785, "learning_rate": 7.579518110192956e-06, "loss": 0.0917, "step": 2664 }, { "epoch": 3.47911227154047, "grad_norm": 1.405255675315857, "learning_rate": 7.577706819926572e-06, "loss": 0.035, "step": 2665 }, { "epoch": 3.480417754569191, "grad_norm": 0.9833861589431763, "learning_rate": 7.575895068814392e-06, "loss": 0.0347, "step": 2666 }, { "epoch": 3.4817232375979112, "grad_norm": 1.713752031326294, "learning_rate": 7.574082857180318e-06, "loss": 0.064, "step": 2667 }, { "epoch": 3.4830287206266317, "grad_norm": 1.2397524118423462, "learning_rate": 7.572270185348345e-06, "loss": 0.0473, "step": 2668 }, { "epoch": 3.4843342036553526, "grad_norm": 2.1409270763397217, "learning_rate": 7.570457053642541e-06, "loss": 0.0488, "step": 2669 }, { "epoch": 3.485639686684073, "grad_norm": 0.9921991229057312, "learning_rate": 7.5686434623870646e-06, "loss": 0.0292, "step": 2670 }, { "epoch": 3.4869451697127936, "grad_norm": 1.4190497398376465, "learning_rate": 7.566829411906146e-06, "loss": 0.0471, "step": 2671 }, { "epoch": 3.4882506527415145, "grad_norm": 2.625422477722168, "learning_rate": 7.565014902524108e-06, "loss": 0.1086, "step": 2672 }, { "epoch": 3.489556135770235, "grad_norm": 2.9068455696105957, "learning_rate": 7.5631999345653494e-06, "loss": 0.0748, "step": 2673 }, { "epoch": 3.490861618798956, "grad_norm": 2.073060989379883, "learning_rate": 7.561384508354352e-06, "loss": 0.0726, "step": 2674 }, { "epoch": 3.4921671018276763, "grad_norm": 1.4105106592178345, "learning_rate": 7.559568624215683e-06, "loss": 0.0431, "step": 2675 }, { "epoch": 3.493472584856397, "grad_norm": 1.7428585290908813, "learning_rate": 7.557752282473985e-06, "loss": 0.0416, "step": 2676 }, { "epoch": 3.4947780678851172, "grad_norm": 1.0364214181900024, "learning_rate": 7.555935483453987e-06, "loss": 0.0382, "step": 2677 }, { "epoch": 3.496083550913838, "grad_norm": 0.9796931147575378, "learning_rate": 7.5541182274804985e-06, "loss": 0.0398, "step": 2678 }, { "epoch": 3.4973890339425586, "grad_norm": 1.7936809062957764, "learning_rate": 7.552300514878413e-06, "loss": 0.0488, "step": 2679 }, { "epoch": 3.4986945169712795, "grad_norm": 1.5549291372299194, "learning_rate": 7.5504823459727e-06, "loss": 0.0432, "step": 2680 }, { "epoch": 3.5, "grad_norm": 1.3106275796890259, "learning_rate": 7.548663721088419e-06, "loss": 0.0501, "step": 2681 }, { "epoch": 3.5013054830287205, "grad_norm": 3.0769362449645996, "learning_rate": 7.546844640550699e-06, "loss": 0.1375, "step": 2682 }, { "epoch": 3.5026109660574414, "grad_norm": 2.30534291267395, "learning_rate": 7.5450251046847645e-06, "loss": 0.0601, "step": 2683 }, { "epoch": 3.503916449086162, "grad_norm": 2.2632734775543213, "learning_rate": 7.5432051138159115e-06, "loss": 0.0891, "step": 2684 }, { "epoch": 3.5052219321148828, "grad_norm": 1.3781017065048218, "learning_rate": 7.54138466826952e-06, "loss": 0.0388, "step": 2685 }, { "epoch": 3.506527415143603, "grad_norm": 0.96246737241745, "learning_rate": 7.5395637683710534e-06, "loss": 0.037, "step": 2686 }, { "epoch": 3.5078328981723237, "grad_norm": 3.8194265365600586, "learning_rate": 7.537742414446052e-06, "loss": 0.1514, "step": 2687 }, { "epoch": 3.509138381201044, "grad_norm": 1.7454472780227661, "learning_rate": 7.535920606820142e-06, "loss": 0.0577, "step": 2688 }, { "epoch": 3.510443864229765, "grad_norm": 1.587623119354248, "learning_rate": 7.534098345819027e-06, "loss": 0.0428, "step": 2689 }, { "epoch": 3.5117493472584855, "grad_norm": 2.223179578781128, "learning_rate": 7.532275631768495e-06, "loss": 0.0632, "step": 2690 }, { "epoch": 3.5130548302872064, "grad_norm": 1.4438499212265015, "learning_rate": 7.530452464994412e-06, "loss": 0.0499, "step": 2691 }, { "epoch": 3.514360313315927, "grad_norm": 1.5025489330291748, "learning_rate": 7.528628845822727e-06, "loss": 0.039, "step": 2692 }, { "epoch": 3.5156657963446474, "grad_norm": 2.1610195636749268, "learning_rate": 7.526804774579468e-06, "loss": 0.0591, "step": 2693 }, { "epoch": 3.5169712793733683, "grad_norm": 1.0608760118484497, "learning_rate": 7.524980251590746e-06, "loss": 0.0311, "step": 2694 }, { "epoch": 3.5182767624020888, "grad_norm": 1.8169070482254028, "learning_rate": 7.523155277182749e-06, "loss": 0.053, "step": 2695 }, { "epoch": 3.5195822454308097, "grad_norm": 3.3435373306274414, "learning_rate": 7.521329851681751e-06, "loss": 0.1168, "step": 2696 }, { "epoch": 3.52088772845953, "grad_norm": 1.791308045387268, "learning_rate": 7.5195039754141045e-06, "loss": 0.0483, "step": 2697 }, { "epoch": 3.5221932114882506, "grad_norm": 1.4367097616195679, "learning_rate": 7.5176776487062386e-06, "loss": 0.0502, "step": 2698 }, { "epoch": 3.523498694516971, "grad_norm": 1.1131166219711304, "learning_rate": 7.515850871884669e-06, "loss": 0.0413, "step": 2699 }, { "epoch": 3.524804177545692, "grad_norm": 1.6143978834152222, "learning_rate": 7.514023645275989e-06, "loss": 0.0479, "step": 2700 }, { "epoch": 3.5261096605744124, "grad_norm": 4.33592414855957, "learning_rate": 7.512195969206872e-06, "loss": 0.1788, "step": 2701 }, { "epoch": 3.5274151436031334, "grad_norm": 2.427454948425293, "learning_rate": 7.510367844004072e-06, "loss": 0.0714, "step": 2702 }, { "epoch": 3.528720626631854, "grad_norm": 5.833173751831055, "learning_rate": 7.508539269994426e-06, "loss": 0.0516, "step": 2703 }, { "epoch": 3.5300261096605743, "grad_norm": 1.7696669101715088, "learning_rate": 7.506710247504844e-06, "loss": 0.0417, "step": 2704 }, { "epoch": 3.531331592689295, "grad_norm": 0.8926332592964172, "learning_rate": 7.504880776862327e-06, "loss": 0.032, "step": 2705 }, { "epoch": 3.5326370757180157, "grad_norm": 3.3265914916992188, "learning_rate": 7.503050858393948e-06, "loss": 0.1229, "step": 2706 }, { "epoch": 3.533942558746736, "grad_norm": 2.6842899322509766, "learning_rate": 7.501220492426857e-06, "loss": 0.0823, "step": 2707 }, { "epoch": 3.535248041775457, "grad_norm": 3.2279436588287354, "learning_rate": 7.499389679288298e-06, "loss": 0.1076, "step": 2708 }, { "epoch": 3.5365535248041775, "grad_norm": 1.6909236907958984, "learning_rate": 7.49755841930558e-06, "loss": 0.0511, "step": 2709 }, { "epoch": 3.537859007832898, "grad_norm": 1.6212776899337769, "learning_rate": 7.4957267128061e-06, "loss": 0.0456, "step": 2710 }, { "epoch": 3.539164490861619, "grad_norm": 2.1441800594329834, "learning_rate": 7.493894560117333e-06, "loss": 0.0586, "step": 2711 }, { "epoch": 3.5404699738903394, "grad_norm": 1.5632457733154297, "learning_rate": 7.492061961566835e-06, "loss": 0.0511, "step": 2712 }, { "epoch": 3.5417754569190603, "grad_norm": 1.6256035566329956, "learning_rate": 7.490228917482235e-06, "loss": 0.0435, "step": 2713 }, { "epoch": 3.5430809399477807, "grad_norm": 2.144834041595459, "learning_rate": 7.488395428191254e-06, "loss": 0.0718, "step": 2714 }, { "epoch": 3.544386422976501, "grad_norm": 1.7672444581985474, "learning_rate": 7.486561494021681e-06, "loss": 0.0396, "step": 2715 }, { "epoch": 3.5456919060052217, "grad_norm": 2.045836925506592, "learning_rate": 7.484727115301391e-06, "loss": 0.0549, "step": 2716 }, { "epoch": 3.5469973890339426, "grad_norm": 1.256052017211914, "learning_rate": 7.482892292358337e-06, "loss": 0.0419, "step": 2717 }, { "epoch": 3.548302872062663, "grad_norm": 3.076176166534424, "learning_rate": 7.48105702552055e-06, "loss": 0.1182, "step": 2718 }, { "epoch": 3.549608355091384, "grad_norm": 1.2551097869873047, "learning_rate": 7.479221315116141e-06, "loss": 0.0488, "step": 2719 }, { "epoch": 3.5509138381201044, "grad_norm": 1.6501590013504028, "learning_rate": 7.4773851614733005e-06, "loss": 0.0475, "step": 2720 }, { "epoch": 3.552219321148825, "grad_norm": 2.5042452812194824, "learning_rate": 7.475548564920301e-06, "loss": 0.0618, "step": 2721 }, { "epoch": 3.553524804177546, "grad_norm": 2.1078152656555176, "learning_rate": 7.4737115257854895e-06, "loss": 0.0846, "step": 2722 }, { "epoch": 3.5548302872062663, "grad_norm": 2.284095287322998, "learning_rate": 7.471874044397295e-06, "loss": 0.0722, "step": 2723 }, { "epoch": 3.556135770234987, "grad_norm": 2.1807103157043457, "learning_rate": 7.470036121084224e-06, "loss": 0.0659, "step": 2724 }, { "epoch": 3.5574412532637076, "grad_norm": 1.2529479265213013, "learning_rate": 7.468197756174864e-06, "loss": 0.04, "step": 2725 }, { "epoch": 3.558746736292428, "grad_norm": 2.8443667888641357, "learning_rate": 7.466358949997879e-06, "loss": 0.0539, "step": 2726 }, { "epoch": 3.5600522193211486, "grad_norm": 1.1553090810775757, "learning_rate": 7.464519702882014e-06, "loss": 0.0389, "step": 2727 }, { "epoch": 3.5613577023498695, "grad_norm": 1.44088613986969, "learning_rate": 7.462680015156092e-06, "loss": 0.0715, "step": 2728 }, { "epoch": 3.56266318537859, "grad_norm": 1.6171338558197021, "learning_rate": 7.460839887149015e-06, "loss": 0.0478, "step": 2729 }, { "epoch": 3.563968668407311, "grad_norm": 1.7254431247711182, "learning_rate": 7.458999319189762e-06, "loss": 0.0549, "step": 2730 }, { "epoch": 3.5652741514360313, "grad_norm": 3.4653937816619873, "learning_rate": 7.457158311607394e-06, "loss": 0.1144, "step": 2731 }, { "epoch": 3.566579634464752, "grad_norm": 2.7869932651519775, "learning_rate": 7.4553168647310485e-06, "loss": 0.0955, "step": 2732 }, { "epoch": 3.5678851174934727, "grad_norm": 4.0640549659729, "learning_rate": 7.453474978889941e-06, "loss": 0.1022, "step": 2733 }, { "epoch": 3.569190600522193, "grad_norm": 1.984526515007019, "learning_rate": 7.4516326544133665e-06, "loss": 0.058, "step": 2734 }, { "epoch": 3.570496083550914, "grad_norm": 1.466545581817627, "learning_rate": 7.4497898916306974e-06, "loss": 0.052, "step": 2735 }, { "epoch": 3.5718015665796345, "grad_norm": 2.372969150543213, "learning_rate": 7.447946690871387e-06, "loss": 0.0671, "step": 2736 }, { "epoch": 3.573107049608355, "grad_norm": 0.8489683270454407, "learning_rate": 7.4461030524649646e-06, "loss": 0.036, "step": 2737 }, { "epoch": 3.5744125326370755, "grad_norm": 1.8781346082687378, "learning_rate": 7.444258976741036e-06, "loss": 0.0424, "step": 2738 }, { "epoch": 3.5757180156657964, "grad_norm": 1.7054009437561035, "learning_rate": 7.4424144640292904e-06, "loss": 0.048, "step": 2739 }, { "epoch": 3.577023498694517, "grad_norm": 1.4577512741088867, "learning_rate": 7.440569514659493e-06, "loss": 0.0479, "step": 2740 }, { "epoch": 3.5783289817232378, "grad_norm": 1.7662408351898193, "learning_rate": 7.438724128961483e-06, "loss": 0.0474, "step": 2741 }, { "epoch": 3.5796344647519582, "grad_norm": 1.3057315349578857, "learning_rate": 7.4368783072651815e-06, "loss": 0.0412, "step": 2742 }, { "epoch": 3.5809399477806787, "grad_norm": 0.8233472108840942, "learning_rate": 7.4350320499005905e-06, "loss": 0.0323, "step": 2743 }, { "epoch": 3.5822454308093996, "grad_norm": 1.6208690404891968, "learning_rate": 7.433185357197781e-06, "loss": 0.0403, "step": 2744 }, { "epoch": 3.58355091383812, "grad_norm": 2.1492369174957275, "learning_rate": 7.431338229486909e-06, "loss": 0.0454, "step": 2745 }, { "epoch": 3.584856396866841, "grad_norm": 1.722753882408142, "learning_rate": 7.429490667098209e-06, "loss": 0.0515, "step": 2746 }, { "epoch": 3.5861618798955615, "grad_norm": 2.6696054935455322, "learning_rate": 7.427642670361988e-06, "loss": 0.057, "step": 2747 }, { "epoch": 3.587467362924282, "grad_norm": 2.4189538955688477, "learning_rate": 7.425794239608635e-06, "loss": 0.079, "step": 2748 }, { "epoch": 3.5887728459530024, "grad_norm": 4.1694207191467285, "learning_rate": 7.423945375168612e-06, "loss": 0.1182, "step": 2749 }, { "epoch": 3.5900783289817233, "grad_norm": 2.0598552227020264, "learning_rate": 7.422096077372464e-06, "loss": 0.0607, "step": 2750 }, { "epoch": 3.5913838120104438, "grad_norm": 3.354088068008423, "learning_rate": 7.42024634655081e-06, "loss": 0.0475, "step": 2751 }, { "epoch": 3.5926892950391647, "grad_norm": 3.8648619651794434, "learning_rate": 7.4183961830343466e-06, "loss": 0.086, "step": 2752 }, { "epoch": 3.593994778067885, "grad_norm": 2.3631482124328613, "learning_rate": 7.41654558715385e-06, "loss": 0.0637, "step": 2753 }, { "epoch": 3.5953002610966056, "grad_norm": 2.9224274158477783, "learning_rate": 7.4146945592401734e-06, "loss": 0.0553, "step": 2754 }, { "epoch": 3.5966057441253265, "grad_norm": 1.2070164680480957, "learning_rate": 7.412843099624241e-06, "loss": 0.0392, "step": 2755 }, { "epoch": 3.597911227154047, "grad_norm": 2.0299971103668213, "learning_rate": 7.410991208637064e-06, "loss": 0.0577, "step": 2756 }, { "epoch": 3.599216710182768, "grad_norm": 1.534315824508667, "learning_rate": 7.409138886609724e-06, "loss": 0.0629, "step": 2757 }, { "epoch": 3.6005221932114884, "grad_norm": 1.8832299709320068, "learning_rate": 7.407286133873381e-06, "loss": 0.055, "step": 2758 }, { "epoch": 3.601827676240209, "grad_norm": 1.539943814277649, "learning_rate": 7.405432950759273e-06, "loss": 0.0381, "step": 2759 }, { "epoch": 3.6031331592689293, "grad_norm": 1.5222864151000977, "learning_rate": 7.403579337598716e-06, "loss": 0.0456, "step": 2760 }, { "epoch": 3.60443864229765, "grad_norm": 2.1525161266326904, "learning_rate": 7.401725294723099e-06, "loss": 0.0541, "step": 2761 }, { "epoch": 3.6057441253263707, "grad_norm": 1.2175272703170776, "learning_rate": 7.3998708224638915e-06, "loss": 0.0431, "step": 2762 }, { "epoch": 3.6070496083550916, "grad_norm": 1.8803014755249023, "learning_rate": 7.398015921152639e-06, "loss": 0.058, "step": 2763 }, { "epoch": 3.608355091383812, "grad_norm": 1.7612565755844116, "learning_rate": 7.396160591120961e-06, "loss": 0.043, "step": 2764 }, { "epoch": 3.6096605744125325, "grad_norm": 2.8729214668273926, "learning_rate": 7.394304832700559e-06, "loss": 0.1184, "step": 2765 }, { "epoch": 3.6109660574412534, "grad_norm": 1.43228018283844, "learning_rate": 7.392448646223206e-06, "loss": 0.0333, "step": 2766 }, { "epoch": 3.612271540469974, "grad_norm": 3.288198709487915, "learning_rate": 7.390592032020752e-06, "loss": 0.1086, "step": 2767 }, { "epoch": 3.6135770234986944, "grad_norm": 1.5293264389038086, "learning_rate": 7.3887349904251285e-06, "loss": 0.0505, "step": 2768 }, { "epoch": 3.6148825065274153, "grad_norm": 1.38770592212677, "learning_rate": 7.386877521768337e-06, "loss": 0.0453, "step": 2769 }, { "epoch": 3.6161879895561357, "grad_norm": 2.767007827758789, "learning_rate": 7.385019626382458e-06, "loss": 0.0503, "step": 2770 }, { "epoch": 3.617493472584856, "grad_norm": 1.4214694499969482, "learning_rate": 7.383161304599652e-06, "loss": 0.0444, "step": 2771 }, { "epoch": 3.618798955613577, "grad_norm": 2.2343409061431885, "learning_rate": 7.381302556752148e-06, "loss": 0.0547, "step": 2772 }, { "epoch": 3.6201044386422976, "grad_norm": 2.635843276977539, "learning_rate": 7.379443383172258e-06, "loss": 0.0654, "step": 2773 }, { "epoch": 3.6214099216710185, "grad_norm": 1.6848995685577393, "learning_rate": 7.377583784192366e-06, "loss": 0.0488, "step": 2774 }, { "epoch": 3.622715404699739, "grad_norm": 3.587494373321533, "learning_rate": 7.375723760144935e-06, "loss": 0.1346, "step": 2775 }, { "epoch": 3.6240208877284594, "grad_norm": 1.3604090213775635, "learning_rate": 7.373863311362499e-06, "loss": 0.0371, "step": 2776 }, { "epoch": 3.62532637075718, "grad_norm": 1.4463979005813599, "learning_rate": 7.372002438177677e-06, "loss": 0.0411, "step": 2777 }, { "epoch": 3.626631853785901, "grad_norm": 0.8545458912849426, "learning_rate": 7.370141140923155e-06, "loss": 0.0339, "step": 2778 }, { "epoch": 3.6279373368146213, "grad_norm": 2.3368895053863525, "learning_rate": 7.3682794199316975e-06, "loss": 0.0663, "step": 2779 }, { "epoch": 3.629242819843342, "grad_norm": 2.3430185317993164, "learning_rate": 7.3664172755361485e-06, "loss": 0.0735, "step": 2780 }, { "epoch": 3.6305483028720626, "grad_norm": 1.926051378250122, "learning_rate": 7.364554708069421e-06, "loss": 0.0688, "step": 2781 }, { "epoch": 3.631853785900783, "grad_norm": 1.3204598426818848, "learning_rate": 7.36269171786451e-06, "loss": 0.0354, "step": 2782 }, { "epoch": 3.633159268929504, "grad_norm": 2.873811721801758, "learning_rate": 7.360828305254481e-06, "loss": 0.1045, "step": 2783 }, { "epoch": 3.6344647519582245, "grad_norm": 1.7884533405303955, "learning_rate": 7.358964470572479e-06, "loss": 0.0502, "step": 2784 }, { "epoch": 3.6357702349869454, "grad_norm": 0.8810755014419556, "learning_rate": 7.3571002141517235e-06, "loss": 0.0366, "step": 2785 }, { "epoch": 3.637075718015666, "grad_norm": 1.4482920169830322, "learning_rate": 7.355235536325507e-06, "loss": 0.0496, "step": 2786 }, { "epoch": 3.6383812010443863, "grad_norm": 1.7614117860794067, "learning_rate": 7.3533704374271985e-06, "loss": 0.0548, "step": 2787 }, { "epoch": 3.639686684073107, "grad_norm": 1.2389672994613647, "learning_rate": 7.351504917790245e-06, "loss": 0.0358, "step": 2788 }, { "epoch": 3.6409921671018277, "grad_norm": 0.9284972548484802, "learning_rate": 7.349638977748166e-06, "loss": 0.0309, "step": 2789 }, { "epoch": 3.642297650130548, "grad_norm": 0.8727552890777588, "learning_rate": 7.347772617634554e-06, "loss": 0.0328, "step": 2790 }, { "epoch": 3.643603133159269, "grad_norm": 1.5267683267593384, "learning_rate": 7.345905837783085e-06, "loss": 0.0495, "step": 2791 }, { "epoch": 3.6449086161879896, "grad_norm": 3.517195701599121, "learning_rate": 7.344038638527498e-06, "loss": 0.1036, "step": 2792 }, { "epoch": 3.64621409921671, "grad_norm": 0.9333078265190125, "learning_rate": 7.342171020201617e-06, "loss": 0.0327, "step": 2793 }, { "epoch": 3.647519582245431, "grad_norm": 1.6711816787719727, "learning_rate": 7.340302983139338e-06, "loss": 0.0554, "step": 2794 }, { "epoch": 3.6488250652741514, "grad_norm": 2.7821030616760254, "learning_rate": 7.338434527674627e-06, "loss": 0.0595, "step": 2795 }, { "epoch": 3.6501305483028723, "grad_norm": 2.113539695739746, "learning_rate": 7.336565654141532e-06, "loss": 0.0361, "step": 2796 }, { "epoch": 3.651436031331593, "grad_norm": 1.987816333770752, "learning_rate": 7.334696362874174e-06, "loss": 0.0519, "step": 2797 }, { "epoch": 3.6527415143603132, "grad_norm": 1.4292112588882446, "learning_rate": 7.332826654206744e-06, "loss": 0.0411, "step": 2798 }, { "epoch": 3.6540469973890337, "grad_norm": 2.956455945968628, "learning_rate": 7.330956528473512e-06, "loss": 0.0929, "step": 2799 }, { "epoch": 3.6553524804177546, "grad_norm": 1.5730481147766113, "learning_rate": 7.329085986008824e-06, "loss": 0.0474, "step": 2800 }, { "epoch": 3.656657963446475, "grad_norm": 3.8949973583221436, "learning_rate": 7.327215027147092e-06, "loss": 0.152, "step": 2801 }, { "epoch": 3.657963446475196, "grad_norm": 2.708848237991333, "learning_rate": 7.325343652222816e-06, "loss": 0.0734, "step": 2802 }, { "epoch": 3.6592689295039165, "grad_norm": 1.2475780248641968, "learning_rate": 7.323471861570558e-06, "loss": 0.035, "step": 2803 }, { "epoch": 3.660574412532637, "grad_norm": 1.882605791091919, "learning_rate": 7.32159965552496e-06, "loss": 0.0509, "step": 2804 }, { "epoch": 3.661879895561358, "grad_norm": 1.2038514614105225, "learning_rate": 7.319727034420741e-06, "loss": 0.0316, "step": 2805 }, { "epoch": 3.6631853785900783, "grad_norm": 1.8357703685760498, "learning_rate": 7.317853998592683e-06, "loss": 0.0473, "step": 2806 }, { "epoch": 3.6644908616187992, "grad_norm": 1.6571769714355469, "learning_rate": 7.3159805483756575e-06, "loss": 0.0478, "step": 2807 }, { "epoch": 3.6657963446475197, "grad_norm": 3.8562753200531006, "learning_rate": 7.3141066841045985e-06, "loss": 0.1142, "step": 2808 }, { "epoch": 3.66710182767624, "grad_norm": 1.6241058111190796, "learning_rate": 7.312232406114518e-06, "loss": 0.0522, "step": 2809 }, { "epoch": 3.6684073107049606, "grad_norm": 1.6905661821365356, "learning_rate": 7.3103577147405025e-06, "loss": 0.0488, "step": 2810 }, { "epoch": 3.6697127937336815, "grad_norm": 1.3344519138336182, "learning_rate": 7.308482610317713e-06, "loss": 0.0384, "step": 2811 }, { "epoch": 3.671018276762402, "grad_norm": 2.202118158340454, "learning_rate": 7.30660709318138e-06, "loss": 0.0529, "step": 2812 }, { "epoch": 3.672323759791123, "grad_norm": 3.428384304046631, "learning_rate": 7.304731163666814e-06, "loss": 0.0661, "step": 2813 }, { "epoch": 3.6736292428198434, "grad_norm": 1.4098567962646484, "learning_rate": 7.302854822109395e-06, "loss": 0.0397, "step": 2814 }, { "epoch": 3.674934725848564, "grad_norm": 3.1560022830963135, "learning_rate": 7.300978068844577e-06, "loss": 0.1008, "step": 2815 }, { "epoch": 3.6762402088772848, "grad_norm": 0.8981575965881348, "learning_rate": 7.299100904207891e-06, "loss": 0.0311, "step": 2816 }, { "epoch": 3.677545691906005, "grad_norm": 1.91708505153656, "learning_rate": 7.297223328534937e-06, "loss": 0.0469, "step": 2817 }, { "epoch": 3.6788511749347257, "grad_norm": 1.8626735210418701, "learning_rate": 7.295345342161388e-06, "loss": 0.0538, "step": 2818 }, { "epoch": 3.6801566579634466, "grad_norm": 1.7225793600082397, "learning_rate": 7.293466945422998e-06, "loss": 0.0495, "step": 2819 }, { "epoch": 3.681462140992167, "grad_norm": 3.603562831878662, "learning_rate": 7.2915881386555855e-06, "loss": 0.0828, "step": 2820 }, { "epoch": 3.6827676240208875, "grad_norm": 0.9403767585754395, "learning_rate": 7.289708922195047e-06, "loss": 0.0401, "step": 2821 }, { "epoch": 3.6840731070496084, "grad_norm": 1.454111099243164, "learning_rate": 7.287829296377353e-06, "loss": 0.0405, "step": 2822 }, { "epoch": 3.685378590078329, "grad_norm": 1.363124966621399, "learning_rate": 7.285949261538542e-06, "loss": 0.0476, "step": 2823 }, { "epoch": 3.68668407310705, "grad_norm": 1.0194761753082275, "learning_rate": 7.284068818014731e-06, "loss": 0.0391, "step": 2824 }, { "epoch": 3.6879895561357703, "grad_norm": 2.1831305027008057, "learning_rate": 7.28218796614211e-06, "loss": 0.0525, "step": 2825 }, { "epoch": 3.6892950391644908, "grad_norm": 1.6580249071121216, "learning_rate": 7.280306706256936e-06, "loss": 0.0369, "step": 2826 }, { "epoch": 3.690600522193211, "grad_norm": 1.9171011447906494, "learning_rate": 7.278425038695547e-06, "loss": 0.0416, "step": 2827 }, { "epoch": 3.691906005221932, "grad_norm": 1.1155180931091309, "learning_rate": 7.276542963794348e-06, "loss": 0.0368, "step": 2828 }, { "epoch": 3.6932114882506526, "grad_norm": 2.0949432849884033, "learning_rate": 7.274660481889819e-06, "loss": 0.0381, "step": 2829 }, { "epoch": 3.6945169712793735, "grad_norm": 2.0748658180236816, "learning_rate": 7.272777593318513e-06, "loss": 0.0482, "step": 2830 }, { "epoch": 3.695822454308094, "grad_norm": 1.1175693273544312, "learning_rate": 7.270894298417056e-06, "loss": 0.041, "step": 2831 }, { "epoch": 3.6971279373368144, "grad_norm": 2.472907066345215, "learning_rate": 7.269010597522144e-06, "loss": 0.0855, "step": 2832 }, { "epoch": 3.6984334203655354, "grad_norm": 2.7181124687194824, "learning_rate": 7.2671264909705486e-06, "loss": 0.0881, "step": 2833 }, { "epoch": 3.699738903394256, "grad_norm": 1.2461200952529907, "learning_rate": 7.2652419790991124e-06, "loss": 0.0327, "step": 2834 }, { "epoch": 3.7010443864229767, "grad_norm": 1.4142348766326904, "learning_rate": 7.263357062244751e-06, "loss": 0.0448, "step": 2835 }, { "epoch": 3.702349869451697, "grad_norm": 1.6473075151443481, "learning_rate": 7.261471740744453e-06, "loss": 0.043, "step": 2836 }, { "epoch": 3.7036553524804177, "grad_norm": 2.0546083450317383, "learning_rate": 7.2595860149352804e-06, "loss": 0.0614, "step": 2837 }, { "epoch": 3.704960835509138, "grad_norm": 2.0542163848876953, "learning_rate": 7.257699885154361e-06, "loss": 0.0614, "step": 2838 }, { "epoch": 3.706266318537859, "grad_norm": 0.8629801869392395, "learning_rate": 7.2558133517389015e-06, "loss": 0.0314, "step": 2839 }, { "epoch": 3.7075718015665795, "grad_norm": 1.8049418926239014, "learning_rate": 7.253926415026181e-06, "loss": 0.0396, "step": 2840 }, { "epoch": 3.7088772845953004, "grad_norm": 1.2120968103408813, "learning_rate": 7.252039075353546e-06, "loss": 0.0328, "step": 2841 }, { "epoch": 3.710182767624021, "grad_norm": 1.0058114528656006, "learning_rate": 7.250151333058419e-06, "loss": 0.0409, "step": 2842 }, { "epoch": 3.7114882506527413, "grad_norm": 1.024260401725769, "learning_rate": 7.248263188478293e-06, "loss": 0.0308, "step": 2843 }, { "epoch": 3.7127937336814623, "grad_norm": 0.885391891002655, "learning_rate": 7.246374641950732e-06, "loss": 0.0313, "step": 2844 }, { "epoch": 3.7140992167101827, "grad_norm": 1.5084320306777954, "learning_rate": 7.244485693813372e-06, "loss": 0.0418, "step": 2845 }, { "epoch": 3.7154046997389036, "grad_norm": 1.7162681818008423, "learning_rate": 7.242596344403925e-06, "loss": 0.0486, "step": 2846 }, { "epoch": 3.716710182767624, "grad_norm": 1.1625709533691406, "learning_rate": 7.240706594060167e-06, "loss": 0.0373, "step": 2847 }, { "epoch": 3.7180156657963446, "grad_norm": 1.2271366119384766, "learning_rate": 7.238816443119955e-06, "loss": 0.0427, "step": 2848 }, { "epoch": 3.719321148825065, "grad_norm": 1.4748166799545288, "learning_rate": 7.236925891921209e-06, "loss": 0.0395, "step": 2849 }, { "epoch": 3.720626631853786, "grad_norm": 2.3930504322052, "learning_rate": 7.235034940801924e-06, "loss": 0.0767, "step": 2850 }, { "epoch": 3.7219321148825064, "grad_norm": 1.7065571546554565, "learning_rate": 7.233143590100168e-06, "loss": 0.0471, "step": 2851 }, { "epoch": 3.7232375979112273, "grad_norm": 1.6911885738372803, "learning_rate": 7.231251840154079e-06, "loss": 0.0414, "step": 2852 }, { "epoch": 3.724543080939948, "grad_norm": 2.1287856101989746, "learning_rate": 7.229359691301868e-06, "loss": 0.0514, "step": 2853 }, { "epoch": 3.7258485639686683, "grad_norm": 3.1316945552825928, "learning_rate": 7.2274671438818136e-06, "loss": 0.0943, "step": 2854 }, { "epoch": 3.727154046997389, "grad_norm": 1.3524831533432007, "learning_rate": 7.225574198232268e-06, "loss": 0.039, "step": 2855 }, { "epoch": 3.7284595300261096, "grad_norm": 1.0944862365722656, "learning_rate": 7.223680854691656e-06, "loss": 0.0326, "step": 2856 }, { "epoch": 3.7297650130548305, "grad_norm": 3.4124562740325928, "learning_rate": 7.221787113598472e-06, "loss": 0.1311, "step": 2857 }, { "epoch": 3.731070496083551, "grad_norm": 1.304380178451538, "learning_rate": 7.219892975291278e-06, "loss": 0.0367, "step": 2858 }, { "epoch": 3.7323759791122715, "grad_norm": 1.3983185291290283, "learning_rate": 7.217998440108716e-06, "loss": 0.0438, "step": 2859 }, { "epoch": 3.733681462140992, "grad_norm": 0.8693439364433289, "learning_rate": 7.21610350838949e-06, "loss": 0.0358, "step": 2860 }, { "epoch": 3.734986945169713, "grad_norm": 1.4176658391952515, "learning_rate": 7.214208180472379e-06, "loss": 0.0413, "step": 2861 }, { "epoch": 3.7362924281984333, "grad_norm": 1.6720134019851685, "learning_rate": 7.212312456696235e-06, "loss": 0.0366, "step": 2862 }, { "epoch": 3.7375979112271542, "grad_norm": 1.0838545560836792, "learning_rate": 7.210416337399971e-06, "loss": 0.0412, "step": 2863 }, { "epoch": 3.7389033942558747, "grad_norm": 0.9491549730300903, "learning_rate": 7.208519822922584e-06, "loss": 0.0276, "step": 2864 }, { "epoch": 3.740208877284595, "grad_norm": 1.6597346067428589, "learning_rate": 7.2066229136031325e-06, "loss": 0.0361, "step": 2865 }, { "epoch": 3.741514360313316, "grad_norm": 1.0679631233215332, "learning_rate": 7.204725609780751e-06, "loss": 0.0307, "step": 2866 }, { "epoch": 3.7428198433420365, "grad_norm": 1.7807809114456177, "learning_rate": 7.202827911794639e-06, "loss": 0.048, "step": 2867 }, { "epoch": 3.7441253263707575, "grad_norm": 1.9504847526550293, "learning_rate": 7.200929819984071e-06, "loss": 0.0561, "step": 2868 }, { "epoch": 3.745430809399478, "grad_norm": 2.871647357940674, "learning_rate": 7.19903133468839e-06, "loss": 0.0973, "step": 2869 }, { "epoch": 3.7467362924281984, "grad_norm": 1.45781672000885, "learning_rate": 7.197132456247009e-06, "loss": 0.0393, "step": 2870 }, { "epoch": 3.748041775456919, "grad_norm": 2.290339708328247, "learning_rate": 7.195233184999413e-06, "loss": 0.047, "step": 2871 }, { "epoch": 3.7493472584856398, "grad_norm": 1.3598802089691162, "learning_rate": 7.193333521285156e-06, "loss": 0.0486, "step": 2872 }, { "epoch": 3.7506527415143602, "grad_norm": 1.5587211847305298, "learning_rate": 7.191433465443862e-06, "loss": 0.0469, "step": 2873 }, { "epoch": 3.751958224543081, "grad_norm": 1.8002444505691528, "learning_rate": 7.189533017815227e-06, "loss": 0.0499, "step": 2874 }, { "epoch": 3.7532637075718016, "grad_norm": 0.9505606293678284, "learning_rate": 7.1876321787390134e-06, "loss": 0.0371, "step": 2875 }, { "epoch": 3.754569190600522, "grad_norm": 1.5045709609985352, "learning_rate": 7.185730948555056e-06, "loss": 0.0302, "step": 2876 }, { "epoch": 3.7558746736292425, "grad_norm": 1.4664204120635986, "learning_rate": 7.18382932760326e-06, "loss": 0.0353, "step": 2877 }, { "epoch": 3.7571801566579635, "grad_norm": 1.7329418659210205, "learning_rate": 7.181927316223599e-06, "loss": 0.0428, "step": 2878 }, { "epoch": 3.758485639686684, "grad_norm": 1.4958951473236084, "learning_rate": 7.180024914756119e-06, "loss": 0.0399, "step": 2879 }, { "epoch": 3.759791122715405, "grad_norm": 1.4900842905044556, "learning_rate": 7.178122123540931e-06, "loss": 0.0627, "step": 2880 }, { "epoch": 3.7610966057441253, "grad_norm": 1.0744500160217285, "learning_rate": 7.1762189429182205e-06, "loss": 0.039, "step": 2881 }, { "epoch": 3.7624020887728458, "grad_norm": 3.11912202835083, "learning_rate": 7.17431537322824e-06, "loss": 0.0674, "step": 2882 }, { "epoch": 3.7637075718015667, "grad_norm": 1.1391531229019165, "learning_rate": 7.17241141481131e-06, "loss": 0.0364, "step": 2883 }, { "epoch": 3.765013054830287, "grad_norm": 1.7637892961502075, "learning_rate": 7.1705070680078274e-06, "loss": 0.0635, "step": 2884 }, { "epoch": 3.766318537859008, "grad_norm": 0.6782556176185608, "learning_rate": 7.16860233315825e-06, "loss": 0.0298, "step": 2885 }, { "epoch": 3.7676240208877285, "grad_norm": 0.7757366895675659, "learning_rate": 7.1666972106031105e-06, "loss": 0.0365, "step": 2886 }, { "epoch": 3.768929503916449, "grad_norm": 0.6889334917068481, "learning_rate": 7.164791700683008e-06, "loss": 0.0275, "step": 2887 }, { "epoch": 3.7702349869451695, "grad_norm": 0.9433267116546631, "learning_rate": 7.162885803738615e-06, "loss": 0.0334, "step": 2888 }, { "epoch": 3.7715404699738904, "grad_norm": 1.127807378768921, "learning_rate": 7.160979520110665e-06, "loss": 0.0344, "step": 2889 }, { "epoch": 3.772845953002611, "grad_norm": 1.4050670862197876, "learning_rate": 7.159072850139971e-06, "loss": 0.0417, "step": 2890 }, { "epoch": 3.7741514360313317, "grad_norm": 1.4956300258636475, "learning_rate": 7.157165794167407e-06, "loss": 0.0387, "step": 2891 }, { "epoch": 3.775456919060052, "grad_norm": 1.6085221767425537, "learning_rate": 7.155258352533919e-06, "loss": 0.0532, "step": 2892 }, { "epoch": 3.7767624020887727, "grad_norm": 0.986240804195404, "learning_rate": 7.153350525580523e-06, "loss": 0.0305, "step": 2893 }, { "epoch": 3.7780678851174936, "grad_norm": 0.9850718975067139, "learning_rate": 7.151442313648305e-06, "loss": 0.0315, "step": 2894 }, { "epoch": 3.779373368146214, "grad_norm": 3.0331549644470215, "learning_rate": 7.149533717078412e-06, "loss": 0.1089, "step": 2895 }, { "epoch": 3.780678851174935, "grad_norm": 1.348091721534729, "learning_rate": 7.14762473621207e-06, "loss": 0.0456, "step": 2896 }, { "epoch": 3.7819843342036554, "grad_norm": 2.7514493465423584, "learning_rate": 7.145715371390567e-06, "loss": 0.0918, "step": 2897 }, { "epoch": 3.783289817232376, "grad_norm": 1.329463005065918, "learning_rate": 7.143805622955262e-06, "loss": 0.0404, "step": 2898 }, { "epoch": 3.7845953002610964, "grad_norm": 1.526049256324768, "learning_rate": 7.141895491247584e-06, "loss": 0.0283, "step": 2899 }, { "epoch": 3.7859007832898173, "grad_norm": 2.664339542388916, "learning_rate": 7.139984976609028e-06, "loss": 0.0916, "step": 2900 }, { "epoch": 3.7872062663185377, "grad_norm": 1.1823829412460327, "learning_rate": 7.138074079381156e-06, "loss": 0.031, "step": 2901 }, { "epoch": 3.7885117493472587, "grad_norm": 1.524946928024292, "learning_rate": 7.136162799905603e-06, "loss": 0.0395, "step": 2902 }, { "epoch": 3.789817232375979, "grad_norm": 1.0898298025131226, "learning_rate": 7.1342511385240695e-06, "loss": 0.0319, "step": 2903 }, { "epoch": 3.7911227154046996, "grad_norm": 3.035219192504883, "learning_rate": 7.132339095578326e-06, "loss": 0.0506, "step": 2904 }, { "epoch": 3.7924281984334205, "grad_norm": 1.275427222251892, "learning_rate": 7.130426671410209e-06, "loss": 0.0364, "step": 2905 }, { "epoch": 3.793733681462141, "grad_norm": 1.6142617464065552, "learning_rate": 7.1285138663616235e-06, "loss": 0.041, "step": 2906 }, { "epoch": 3.795039164490862, "grad_norm": 2.122596263885498, "learning_rate": 7.1266006807745435e-06, "loss": 0.0657, "step": 2907 }, { "epoch": 3.7963446475195823, "grad_norm": 1.4088462591171265, "learning_rate": 7.124687114991011e-06, "loss": 0.0584, "step": 2908 }, { "epoch": 3.797650130548303, "grad_norm": 3.2641799449920654, "learning_rate": 7.1227731693531376e-06, "loss": 0.0528, "step": 2909 }, { "epoch": 3.7989556135770233, "grad_norm": 2.2319374084472656, "learning_rate": 7.120858844203099e-06, "loss": 0.0976, "step": 2910 }, { "epoch": 3.800261096605744, "grad_norm": 3.4871304035186768, "learning_rate": 7.118944139883142e-06, "loss": 0.0978, "step": 2911 }, { "epoch": 3.8015665796344646, "grad_norm": 1.4084550142288208, "learning_rate": 7.117029056735577e-06, "loss": 0.0327, "step": 2912 }, { "epoch": 3.8028720626631856, "grad_norm": 1.234206199645996, "learning_rate": 7.1151135951027884e-06, "loss": 0.0432, "step": 2913 }, { "epoch": 3.804177545691906, "grad_norm": 2.00386381149292, "learning_rate": 7.113197755327224e-06, "loss": 0.06, "step": 2914 }, { "epoch": 3.8054830287206265, "grad_norm": 0.9992315769195557, "learning_rate": 7.111281537751399e-06, "loss": 0.0355, "step": 2915 }, { "epoch": 3.8067885117493474, "grad_norm": 2.7851805686950684, "learning_rate": 7.109364942717899e-06, "loss": 0.099, "step": 2916 }, { "epoch": 3.808093994778068, "grad_norm": 2.225431203842163, "learning_rate": 7.107447970569375e-06, "loss": 0.0756, "step": 2917 }, { "epoch": 3.809399477806789, "grad_norm": 1.9825005531311035, "learning_rate": 7.105530621648544e-06, "loss": 0.0435, "step": 2918 }, { "epoch": 3.8107049608355092, "grad_norm": 1.662176489830017, "learning_rate": 7.103612896298196e-06, "loss": 0.0407, "step": 2919 }, { "epoch": 3.8120104438642297, "grad_norm": 1.6244795322418213, "learning_rate": 7.101694794861178e-06, "loss": 0.0477, "step": 2920 }, { "epoch": 3.81331592689295, "grad_norm": 1.4441050291061401, "learning_rate": 7.099776317680417e-06, "loss": 0.0435, "step": 2921 }, { "epoch": 3.814621409921671, "grad_norm": 7.970920562744141, "learning_rate": 7.097857465098898e-06, "loss": 0.1204, "step": 2922 }, { "epoch": 3.8159268929503916, "grad_norm": 2.3713486194610596, "learning_rate": 7.095938237459678e-06, "loss": 0.0679, "step": 2923 }, { "epoch": 3.8172323759791125, "grad_norm": 3.061779260635376, "learning_rate": 7.094018635105876e-06, "loss": 0.0793, "step": 2924 }, { "epoch": 3.818537859007833, "grad_norm": 1.5968986749649048, "learning_rate": 7.092098658380684e-06, "loss": 0.0369, "step": 2925 }, { "epoch": 3.8198433420365534, "grad_norm": 1.6127498149871826, "learning_rate": 7.0901783076273555e-06, "loss": 0.0428, "step": 2926 }, { "epoch": 3.8211488250652743, "grad_norm": 0.9040641188621521, "learning_rate": 7.088257583189213e-06, "loss": 0.034, "step": 2927 }, { "epoch": 3.822454308093995, "grad_norm": 1.5669926404953003, "learning_rate": 7.086336485409648e-06, "loss": 0.0456, "step": 2928 }, { "epoch": 3.8237597911227152, "grad_norm": 2.7804369926452637, "learning_rate": 7.084415014632118e-06, "loss": 0.0669, "step": 2929 }, { "epoch": 3.825065274151436, "grad_norm": 0.9910932779312134, "learning_rate": 7.082493171200143e-06, "loss": 0.0363, "step": 2930 }, { "epoch": 3.8263707571801566, "grad_norm": 3.0000815391540527, "learning_rate": 7.080570955457316e-06, "loss": 0.0978, "step": 2931 }, { "epoch": 3.827676240208877, "grad_norm": 3.0965521335601807, "learning_rate": 7.078648367747288e-06, "loss": 0.082, "step": 2932 }, { "epoch": 3.828981723237598, "grad_norm": 1.2410626411437988, "learning_rate": 7.076725408413787e-06, "loss": 0.0406, "step": 2933 }, { "epoch": 3.8302872062663185, "grad_norm": 1.241184949874878, "learning_rate": 7.0748020778006e-06, "loss": 0.0396, "step": 2934 }, { "epoch": 3.8315926892950394, "grad_norm": 1.6055136919021606, "learning_rate": 7.0728783762515815e-06, "loss": 0.0455, "step": 2935 }, { "epoch": 3.83289817232376, "grad_norm": 1.2730886936187744, "learning_rate": 7.070954304110656e-06, "loss": 0.0342, "step": 2936 }, { "epoch": 3.8342036553524803, "grad_norm": 2.2328085899353027, "learning_rate": 7.069029861721811e-06, "loss": 0.0699, "step": 2937 }, { "epoch": 3.8355091383812008, "grad_norm": 1.4485605955123901, "learning_rate": 7.0671050494291e-06, "loss": 0.0351, "step": 2938 }, { "epoch": 3.8368146214099217, "grad_norm": 1.3328251838684082, "learning_rate": 7.065179867576641e-06, "loss": 0.0368, "step": 2939 }, { "epoch": 3.838120104438642, "grad_norm": 3.2372419834136963, "learning_rate": 7.063254316508625e-06, "loss": 0.0496, "step": 2940 }, { "epoch": 3.839425587467363, "grad_norm": 2.885582685470581, "learning_rate": 7.0613283965693024e-06, "loss": 0.0922, "step": 2941 }, { "epoch": 3.8407310704960835, "grad_norm": 2.140550136566162, "learning_rate": 7.0594021081029926e-06, "loss": 0.0528, "step": 2942 }, { "epoch": 3.842036553524804, "grad_norm": 2.0487687587738037, "learning_rate": 7.057475451454079e-06, "loss": 0.0567, "step": 2943 }, { "epoch": 3.843342036553525, "grad_norm": 2.3383734226226807, "learning_rate": 7.0555484269670115e-06, "loss": 0.0685, "step": 2944 }, { "epoch": 3.8446475195822454, "grad_norm": 1.1792539358139038, "learning_rate": 7.053621034986308e-06, "loss": 0.0332, "step": 2945 }, { "epoch": 3.8459530026109663, "grad_norm": 1.8526923656463623, "learning_rate": 7.051693275856547e-06, "loss": 0.0535, "step": 2946 }, { "epoch": 3.8472584856396868, "grad_norm": 0.8142361044883728, "learning_rate": 7.049765149922379e-06, "loss": 0.0312, "step": 2947 }, { "epoch": 3.848563968668407, "grad_norm": 1.5989174842834473, "learning_rate": 7.047836657528516e-06, "loss": 0.0379, "step": 2948 }, { "epoch": 3.8498694516971277, "grad_norm": 3.0168654918670654, "learning_rate": 7.045907799019735e-06, "loss": 0.088, "step": 2949 }, { "epoch": 3.8511749347258486, "grad_norm": 1.657706618309021, "learning_rate": 7.043978574740882e-06, "loss": 0.0393, "step": 2950 }, { "epoch": 3.852480417754569, "grad_norm": 2.106475830078125, "learning_rate": 7.042048985036863e-06, "loss": 0.0499, "step": 2951 }, { "epoch": 3.85378590078329, "grad_norm": 1.594003438949585, "learning_rate": 7.040119030252654e-06, "loss": 0.0484, "step": 2952 }, { "epoch": 3.8550913838120104, "grad_norm": 1.2550641298294067, "learning_rate": 7.038188710733298e-06, "loss": 0.0448, "step": 2953 }, { "epoch": 3.856396866840731, "grad_norm": 1.120901346206665, "learning_rate": 7.036258026823895e-06, "loss": 0.0326, "step": 2954 }, { "epoch": 3.857702349869452, "grad_norm": 2.0219497680664062, "learning_rate": 7.034326978869619e-06, "loss": 0.0501, "step": 2955 }, { "epoch": 3.8590078328981723, "grad_norm": 1.633161187171936, "learning_rate": 7.032395567215705e-06, "loss": 0.0524, "step": 2956 }, { "epoch": 3.860313315926893, "grad_norm": 1.450082540512085, "learning_rate": 7.030463792207452e-06, "loss": 0.0426, "step": 2957 }, { "epoch": 3.8616187989556137, "grad_norm": 1.541256308555603, "learning_rate": 7.028531654190223e-06, "loss": 0.0427, "step": 2958 }, { "epoch": 3.862924281984334, "grad_norm": 2.4684300422668457, "learning_rate": 7.026599153509453e-06, "loss": 0.0775, "step": 2959 }, { "epoch": 3.8642297650130546, "grad_norm": 1.327125906944275, "learning_rate": 7.024666290510633e-06, "loss": 0.0331, "step": 2960 }, { "epoch": 3.8655352480417755, "grad_norm": 3.976426124572754, "learning_rate": 7.022733065539326e-06, "loss": 0.1277, "step": 2961 }, { "epoch": 3.866840731070496, "grad_norm": 1.7329140901565552, "learning_rate": 7.020799478941154e-06, "loss": 0.0476, "step": 2962 }, { "epoch": 3.868146214099217, "grad_norm": 1.4875997304916382, "learning_rate": 7.018865531061807e-06, "loss": 0.0442, "step": 2963 }, { "epoch": 3.8694516971279374, "grad_norm": 1.6388853788375854, "learning_rate": 7.0169312222470385e-06, "loss": 0.0351, "step": 2964 }, { "epoch": 3.870757180156658, "grad_norm": 2.0332586765289307, "learning_rate": 7.014996552842668e-06, "loss": 0.0474, "step": 2965 }, { "epoch": 3.8720626631853787, "grad_norm": 1.0187270641326904, "learning_rate": 7.013061523194578e-06, "loss": 0.0349, "step": 2966 }, { "epoch": 3.873368146214099, "grad_norm": 1.70051908493042, "learning_rate": 7.011126133648713e-06, "loss": 0.0434, "step": 2967 }, { "epoch": 3.87467362924282, "grad_norm": 1.1709926128387451, "learning_rate": 7.009190384551088e-06, "loss": 0.0337, "step": 2968 }, { "epoch": 3.8759791122715406, "grad_norm": 2.373490571975708, "learning_rate": 7.007254276247776e-06, "loss": 0.0628, "step": 2969 }, { "epoch": 3.877284595300261, "grad_norm": 1.5209465026855469, "learning_rate": 7.0053178090849195e-06, "loss": 0.0384, "step": 2970 }, { "epoch": 3.8785900783289815, "grad_norm": 0.8723734617233276, "learning_rate": 7.003380983408719e-06, "loss": 0.0327, "step": 2971 }, { "epoch": 3.8798955613577024, "grad_norm": 1.795130729675293, "learning_rate": 7.001443799565446e-06, "loss": 0.0473, "step": 2972 }, { "epoch": 3.881201044386423, "grad_norm": 0.8407188057899475, "learning_rate": 6.999506257901433e-06, "loss": 0.0272, "step": 2973 }, { "epoch": 3.882506527415144, "grad_norm": 3.0510315895080566, "learning_rate": 6.997568358763074e-06, "loss": 0.0808, "step": 2974 }, { "epoch": 3.8838120104438643, "grad_norm": 1.0335032939910889, "learning_rate": 6.995630102496829e-06, "loss": 0.0331, "step": 2975 }, { "epoch": 3.8851174934725847, "grad_norm": 1.0592122077941895, "learning_rate": 6.993691489449226e-06, "loss": 0.0329, "step": 2976 }, { "epoch": 3.8864229765013056, "grad_norm": 1.133744239807129, "learning_rate": 6.9917525199668465e-06, "loss": 0.0355, "step": 2977 }, { "epoch": 3.887728459530026, "grad_norm": 0.8991667032241821, "learning_rate": 6.989813194396348e-06, "loss": 0.0331, "step": 2978 }, { "epoch": 3.8890339425587466, "grad_norm": 0.8250220417976379, "learning_rate": 6.987873513084442e-06, "loss": 0.0299, "step": 2979 }, { "epoch": 3.8903394255874675, "grad_norm": 3.4049949645996094, "learning_rate": 6.985933476377908e-06, "loss": 0.1164, "step": 2980 }, { "epoch": 3.891644908616188, "grad_norm": 2.512533664703369, "learning_rate": 6.98399308462359e-06, "loss": 0.0769, "step": 2981 }, { "epoch": 3.8929503916449084, "grad_norm": 2.548248767852783, "learning_rate": 6.982052338168393e-06, "loss": 0.0719, "step": 2982 }, { "epoch": 3.8942558746736293, "grad_norm": 2.5639240741729736, "learning_rate": 6.980111237359282e-06, "loss": 0.0792, "step": 2983 }, { "epoch": 3.89556135770235, "grad_norm": 1.805325984954834, "learning_rate": 6.978169782543296e-06, "loss": 0.0394, "step": 2984 }, { "epoch": 3.8968668407310707, "grad_norm": 1.286641001701355, "learning_rate": 6.976227974067527e-06, "loss": 0.0362, "step": 2985 }, { "epoch": 3.898172323759791, "grad_norm": 0.9255996942520142, "learning_rate": 6.9742858122791355e-06, "loss": 0.0334, "step": 2986 }, { "epoch": 3.8994778067885116, "grad_norm": 0.8699696063995361, "learning_rate": 6.972343297525343e-06, "loss": 0.032, "step": 2987 }, { "epoch": 3.900783289817232, "grad_norm": 0.9048709273338318, "learning_rate": 6.970400430153435e-06, "loss": 0.0327, "step": 2988 }, { "epoch": 3.902088772845953, "grad_norm": 2.3840649127960205, "learning_rate": 6.968457210510761e-06, "loss": 0.0581, "step": 2989 }, { "epoch": 3.9033942558746735, "grad_norm": 0.8972358107566833, "learning_rate": 6.9665136389447296e-06, "loss": 0.0314, "step": 2990 }, { "epoch": 3.9046997389033944, "grad_norm": 2.083475351333618, "learning_rate": 6.964569715802817e-06, "loss": 0.0543, "step": 2991 }, { "epoch": 3.906005221932115, "grad_norm": 1.5512381792068481, "learning_rate": 6.96262544143256e-06, "loss": 0.0371, "step": 2992 }, { "epoch": 3.9073107049608353, "grad_norm": 1.2466931343078613, "learning_rate": 6.960680816181561e-06, "loss": 0.0432, "step": 2993 }, { "epoch": 3.9086161879895562, "grad_norm": 1.1344798803329468, "learning_rate": 6.958735840397478e-06, "loss": 0.0506, "step": 2994 }, { "epoch": 3.9099216710182767, "grad_norm": 2.057790994644165, "learning_rate": 6.956790514428039e-06, "loss": 0.0605, "step": 2995 }, { "epoch": 3.9112271540469976, "grad_norm": 1.4412521123886108, "learning_rate": 6.954844838621031e-06, "loss": 0.0355, "step": 2996 }, { "epoch": 3.912532637075718, "grad_norm": 2.9695630073547363, "learning_rate": 6.952898813324306e-06, "loss": 0.1082, "step": 2997 }, { "epoch": 3.9138381201044385, "grad_norm": 1.5511432886123657, "learning_rate": 6.950952438885775e-06, "loss": 0.0405, "step": 2998 }, { "epoch": 3.915143603133159, "grad_norm": 3.3651905059814453, "learning_rate": 6.949005715653415e-06, "loss": 0.1396, "step": 2999 }, { "epoch": 3.91644908616188, "grad_norm": 1.1166071891784668, "learning_rate": 6.947058643975263e-06, "loss": 0.0394, "step": 3000 } ], "logging_steps": 1.0, "max_steps": 7660, "num_input_tokens_seen": 0, "num_train_epochs": 10, "save_steps": 1000, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": false }, "attributes": {} } }, "total_flos": 7.002888684747882e+18, "train_batch_size": 1, "trial_name": null, "trial_params": null }