GeoChrono / trainer_state.json
Davidup1's picture
Add GeoChrono weights (LoRA adapter + TempEnc) and model card
27385c3 verified
Raw
History Blame Contribute Delete
285 kB
{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 1.0,
"eval_steps": 500,
"global_step": 1640,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0006098490623570666,
"grad_norm": 7.8219733238220215,
"learning_rate": 0.0,
"loss": 1.5698,
"step": 1
},
{
"epoch": 0.0012196981247141333,
"grad_norm": 8.10168743133545,
"learning_rate": 2.0000000000000003e-06,
"loss": 1.4899,
"step": 2
},
{
"epoch": 0.0018295471870711998,
"grad_norm": 7.603939056396484,
"learning_rate": 4.000000000000001e-06,
"loss": 1.2716,
"step": 3
},
{
"epoch": 0.0024393962494282666,
"grad_norm": 9.248464584350586,
"learning_rate": 6e-06,
"loss": 1.8027,
"step": 4
},
{
"epoch": 0.003049245311785333,
"grad_norm": 13.36034107208252,
"learning_rate": 8.000000000000001e-06,
"loss": 1.3079,
"step": 5
},
{
"epoch": 0.0036590943741423997,
"grad_norm": 12.134799003601074,
"learning_rate": 1e-05,
"loss": 1.4236,
"step": 6
},
{
"epoch": 0.004268943436499467,
"grad_norm": 8.766926765441895,
"learning_rate": 1.2e-05,
"loss": 1.6485,
"step": 7
},
{
"epoch": 0.004878792498856533,
"grad_norm": 6.477776527404785,
"learning_rate": 1.4000000000000001e-05,
"loss": 1.1011,
"step": 8
},
{
"epoch": 0.0054886415612136,
"grad_norm": 7.021245002746582,
"learning_rate": 1.6000000000000003e-05,
"loss": 1.0687,
"step": 9
},
{
"epoch": 0.006098490623570666,
"grad_norm": 7.125764846801758,
"learning_rate": 1.8e-05,
"loss": 1.2423,
"step": 10
},
{
"epoch": 0.006708339685927733,
"grad_norm": 4.024333477020264,
"learning_rate": 2e-05,
"loss": 0.786,
"step": 11
},
{
"epoch": 0.007318188748284799,
"grad_norm": 4.602781295776367,
"learning_rate": 2.2000000000000003e-05,
"loss": 0.832,
"step": 12
},
{
"epoch": 0.007928037810641867,
"grad_norm": 5.284022331237793,
"learning_rate": 2.4e-05,
"loss": 0.9077,
"step": 13
},
{
"epoch": 0.008537886872998933,
"grad_norm": 4.996356964111328,
"learning_rate": 2.6000000000000002e-05,
"loss": 0.7465,
"step": 14
},
{
"epoch": 0.009147735935356,
"grad_norm": 3.5360984802246094,
"learning_rate": 2.8000000000000003e-05,
"loss": 0.6977,
"step": 15
},
{
"epoch": 0.009757584997713066,
"grad_norm": 2.2018723487854004,
"learning_rate": 3e-05,
"loss": 0.5227,
"step": 16
},
{
"epoch": 0.010367434060070133,
"grad_norm": 2.3838095664978027,
"learning_rate": 3.2000000000000005e-05,
"loss": 0.5368,
"step": 17
},
{
"epoch": 0.0109772831224272,
"grad_norm": 1.957159399986267,
"learning_rate": 3.4000000000000007e-05,
"loss": 0.5218,
"step": 18
},
{
"epoch": 0.011587132184784266,
"grad_norm": 2.302069664001465,
"learning_rate": 3.6e-05,
"loss": 0.5013,
"step": 19
},
{
"epoch": 0.012196981247141333,
"grad_norm": 1.1447867155075073,
"learning_rate": 3.8e-05,
"loss": 0.3673,
"step": 20
},
{
"epoch": 0.012806830309498399,
"grad_norm": 1.151745319366455,
"learning_rate": 4e-05,
"loss": 0.3867,
"step": 21
},
{
"epoch": 0.013416679371855466,
"grad_norm": 1.0864030122756958,
"learning_rate": 4.2e-05,
"loss": 0.3799,
"step": 22
},
{
"epoch": 0.014026528434212532,
"grad_norm": 1.0294002294540405,
"learning_rate": 4.4000000000000006e-05,
"loss": 0.3445,
"step": 23
},
{
"epoch": 0.014636377496569599,
"grad_norm": 0.8172827959060669,
"learning_rate": 4.600000000000001e-05,
"loss": 0.3636,
"step": 24
},
{
"epoch": 0.015246226558926665,
"grad_norm": 0.9214715957641602,
"learning_rate": 4.8e-05,
"loss": 0.3412,
"step": 25
},
{
"epoch": 0.015856075621283734,
"grad_norm": 1.0714569091796875,
"learning_rate": 5e-05,
"loss": 0.3569,
"step": 26
},
{
"epoch": 0.0164659246836408,
"grad_norm": 0.8058624267578125,
"learning_rate": 5.2000000000000004e-05,
"loss": 0.3483,
"step": 27
},
{
"epoch": 0.017075773745997867,
"grad_norm": 0.8464831709861755,
"learning_rate": 5.4000000000000005e-05,
"loss": 0.3407,
"step": 28
},
{
"epoch": 0.017685622808354933,
"grad_norm": 1.0150617361068726,
"learning_rate": 5.6000000000000006e-05,
"loss": 0.3378,
"step": 29
},
{
"epoch": 0.018295471870712,
"grad_norm": 0.9095039367675781,
"learning_rate": 5.8e-05,
"loss": 0.3577,
"step": 30
},
{
"epoch": 0.018905320933069066,
"grad_norm": 0.7577035427093506,
"learning_rate": 6e-05,
"loss": 0.3282,
"step": 31
},
{
"epoch": 0.019515169995426133,
"grad_norm": 0.9115796089172363,
"learning_rate": 6.2e-05,
"loss": 0.2846,
"step": 32
},
{
"epoch": 0.0201250190577832,
"grad_norm": 0.707153856754303,
"learning_rate": 6.400000000000001e-05,
"loss": 0.2617,
"step": 33
},
{
"epoch": 0.020734868120140266,
"grad_norm": 1.957801103591919,
"learning_rate": 6.6e-05,
"loss": 0.2795,
"step": 34
},
{
"epoch": 0.021344717182497332,
"grad_norm": 0.7698462605476379,
"learning_rate": 6.800000000000001e-05,
"loss": 0.2299,
"step": 35
},
{
"epoch": 0.0219545662448544,
"grad_norm": 0.6001272201538086,
"learning_rate": 7e-05,
"loss": 0.286,
"step": 36
},
{
"epoch": 0.022564415307211465,
"grad_norm": 0.742516040802002,
"learning_rate": 7.2e-05,
"loss": 0.3008,
"step": 37
},
{
"epoch": 0.023174264369568532,
"grad_norm": 0.5710421204566956,
"learning_rate": 7.4e-05,
"loss": 0.258,
"step": 38
},
{
"epoch": 0.0237841134319256,
"grad_norm": 0.5841881632804871,
"learning_rate": 7.6e-05,
"loss": 0.2183,
"step": 39
},
{
"epoch": 0.024393962494282665,
"grad_norm": 0.7163995504379272,
"learning_rate": 7.800000000000001e-05,
"loss": 0.2727,
"step": 40
},
{
"epoch": 0.02500381155663973,
"grad_norm": 0.806366503238678,
"learning_rate": 8e-05,
"loss": 0.3226,
"step": 41
},
{
"epoch": 0.025613660618996798,
"grad_norm": 0.6190110445022583,
"learning_rate": 8.2e-05,
"loss": 0.2479,
"step": 42
},
{
"epoch": 0.026223509681353865,
"grad_norm": 0.6145042181015015,
"learning_rate": 8.4e-05,
"loss": 0.2239,
"step": 43
},
{
"epoch": 0.02683335874371093,
"grad_norm": 0.7031387090682983,
"learning_rate": 8.6e-05,
"loss": 0.2373,
"step": 44
},
{
"epoch": 0.027443207806067998,
"grad_norm": 0.6843944787979126,
"learning_rate": 8.800000000000001e-05,
"loss": 0.2592,
"step": 45
},
{
"epoch": 0.028053056868425064,
"grad_norm": 0.5825070738792419,
"learning_rate": 9e-05,
"loss": 0.2313,
"step": 46
},
{
"epoch": 0.02866290593078213,
"grad_norm": 0.5755884647369385,
"learning_rate": 9.200000000000001e-05,
"loss": 0.2205,
"step": 47
},
{
"epoch": 0.029272754993139197,
"grad_norm": 0.7310126423835754,
"learning_rate": 9.4e-05,
"loss": 0.2076,
"step": 48
},
{
"epoch": 0.029882604055496264,
"grad_norm": 1.1359120607376099,
"learning_rate": 9.6e-05,
"loss": 0.3037,
"step": 49
},
{
"epoch": 0.03049245311785333,
"grad_norm": 1.4995906352996826,
"learning_rate": 9.8e-05,
"loss": 0.2876,
"step": 50
},
{
"epoch": 0.031102302180210397,
"grad_norm": 1.1709572076797485,
"learning_rate": 0.0001,
"loss": 0.2122,
"step": 51
},
{
"epoch": 0.03171215124256747,
"grad_norm": 0.56809002161026,
"learning_rate": 9.999990240100085e-05,
"loss": 0.1943,
"step": 52
},
{
"epoch": 0.032322000304924534,
"grad_norm": 0.7915175557136536,
"learning_rate": 9.999960960438442e-05,
"loss": 0.2691,
"step": 53
},
{
"epoch": 0.0329318493672816,
"grad_norm": 1.079586148262024,
"learning_rate": 9.999912161129376e-05,
"loss": 0.2715,
"step": 54
},
{
"epoch": 0.03354169842963867,
"grad_norm": 0.5489707589149475,
"learning_rate": 9.999843842363401e-05,
"loss": 0.1916,
"step": 55
},
{
"epoch": 0.03415154749199573,
"grad_norm": 0.70656418800354,
"learning_rate": 9.999756004407229e-05,
"loss": 0.226,
"step": 56
},
{
"epoch": 0.0347613965543528,
"grad_norm": 1.0411139726638794,
"learning_rate": 9.999648647603774e-05,
"loss": 0.213,
"step": 57
},
{
"epoch": 0.035371245616709866,
"grad_norm": 0.9722528457641602,
"learning_rate": 9.999521772372156e-05,
"loss": 0.2693,
"step": 58
},
{
"epoch": 0.03598109467906693,
"grad_norm": 0.8676877617835999,
"learning_rate": 9.99937537920769e-05,
"loss": 0.2275,
"step": 59
},
{
"epoch": 0.036590943741424,
"grad_norm": 0.8274833559989929,
"learning_rate": 9.999209468681885e-05,
"loss": 0.29,
"step": 60
},
{
"epoch": 0.037200792803781066,
"grad_norm": 0.5366308689117432,
"learning_rate": 9.999024041442456e-05,
"loss": 0.2021,
"step": 61
},
{
"epoch": 0.03781064186613813,
"grad_norm": 0.6157708764076233,
"learning_rate": 9.998819098213297e-05,
"loss": 0.2215,
"step": 62
},
{
"epoch": 0.0384204909284952,
"grad_norm": 0.8965314030647278,
"learning_rate": 9.998594639794501e-05,
"loss": 0.2394,
"step": 63
},
{
"epoch": 0.039030339990852266,
"grad_norm": 0.5548007488250732,
"learning_rate": 9.998350667062346e-05,
"loss": 0.1848,
"step": 64
},
{
"epoch": 0.03964018905320933,
"grad_norm": 0.536480724811554,
"learning_rate": 9.998087180969289e-05,
"loss": 0.2138,
"step": 65
},
{
"epoch": 0.0402500381155664,
"grad_norm": 0.6392802596092224,
"learning_rate": 9.997804182543973e-05,
"loss": 0.2032,
"step": 66
},
{
"epoch": 0.040859887177923465,
"grad_norm": 0.5861805081367493,
"learning_rate": 9.997501672891207e-05,
"loss": 0.2672,
"step": 67
},
{
"epoch": 0.04146973624028053,
"grad_norm": 0.8933752179145813,
"learning_rate": 9.997179653191982e-05,
"loss": 0.247,
"step": 68
},
{
"epoch": 0.0420795853026376,
"grad_norm": 0.5527284741401672,
"learning_rate": 9.996838124703447e-05,
"loss": 0.2126,
"step": 69
},
{
"epoch": 0.042689434364994665,
"grad_norm": 0.5914862751960754,
"learning_rate": 9.996477088758916e-05,
"loss": 0.2442,
"step": 70
},
{
"epoch": 0.04329928342735173,
"grad_norm": 0.6851914525032043,
"learning_rate": 9.99609654676786e-05,
"loss": 0.2155,
"step": 71
},
{
"epoch": 0.0439091324897088,
"grad_norm": 0.6642138957977295,
"learning_rate": 9.995696500215898e-05,
"loss": 0.222,
"step": 72
},
{
"epoch": 0.044518981552065864,
"grad_norm": 0.6638391017913818,
"learning_rate": 9.995276950664796e-05,
"loss": 0.1715,
"step": 73
},
{
"epoch": 0.04512883061442293,
"grad_norm": 0.8532464504241943,
"learning_rate": 9.99483789975246e-05,
"loss": 0.2424,
"step": 74
},
{
"epoch": 0.04573867967678,
"grad_norm": 0.6275679469108582,
"learning_rate": 9.994379349192926e-05,
"loss": 0.2011,
"step": 75
},
{
"epoch": 0.046348528739137064,
"grad_norm": 0.5084139108657837,
"learning_rate": 9.993901300776359e-05,
"loss": 0.1855,
"step": 76
},
{
"epoch": 0.04695837780149413,
"grad_norm": 0.7881125807762146,
"learning_rate": 9.993403756369038e-05,
"loss": 0.2189,
"step": 77
},
{
"epoch": 0.0475682268638512,
"grad_norm": 0.56772780418396,
"learning_rate": 9.992886717913356e-05,
"loss": 0.1822,
"step": 78
},
{
"epoch": 0.048178075926208264,
"grad_norm": 0.7930513024330139,
"learning_rate": 9.992350187427815e-05,
"loss": 0.196,
"step": 79
},
{
"epoch": 0.04878792498856533,
"grad_norm": 0.885890543460846,
"learning_rate": 9.991794167007004e-05,
"loss": 0.218,
"step": 80
},
{
"epoch": 0.0493977740509224,
"grad_norm": 0.604590654373169,
"learning_rate": 9.991218658821608e-05,
"loss": 0.1659,
"step": 81
},
{
"epoch": 0.05000762311327946,
"grad_norm": 0.7208395600318909,
"learning_rate": 9.990623665118384e-05,
"loss": 0.1731,
"step": 82
},
{
"epoch": 0.05061747217563653,
"grad_norm": 0.5281884074211121,
"learning_rate": 9.990009188220167e-05,
"loss": 0.1749,
"step": 83
},
{
"epoch": 0.051227321237993596,
"grad_norm": 0.6611881256103516,
"learning_rate": 9.989375230525848e-05,
"loss": 0.1986,
"step": 84
},
{
"epoch": 0.05183717030035066,
"grad_norm": 0.5471932291984558,
"learning_rate": 9.988721794510374e-05,
"loss": 0.174,
"step": 85
},
{
"epoch": 0.05244701936270773,
"grad_norm": 0.7030887007713318,
"learning_rate": 9.988048882724732e-05,
"loss": 0.2158,
"step": 86
},
{
"epoch": 0.053056868425064796,
"grad_norm": 0.6978226900100708,
"learning_rate": 9.987356497795943e-05,
"loss": 0.2102,
"step": 87
},
{
"epoch": 0.05366671748742186,
"grad_norm": 0.6604421734809875,
"learning_rate": 9.986644642427051e-05,
"loss": 0.2105,
"step": 88
},
{
"epoch": 0.05427656654977893,
"grad_norm": 0.7822443246841431,
"learning_rate": 9.985913319397109e-05,
"loss": 0.231,
"step": 89
},
{
"epoch": 0.054886415612135996,
"grad_norm": 0.644214928150177,
"learning_rate": 9.985162531561174e-05,
"loss": 0.1743,
"step": 90
},
{
"epoch": 0.05549626467449306,
"grad_norm": 0.6288518905639648,
"learning_rate": 9.984392281850293e-05,
"loss": 0.21,
"step": 91
},
{
"epoch": 0.05610611373685013,
"grad_norm": 0.6551951169967651,
"learning_rate": 9.983602573271485e-05,
"loss": 0.2292,
"step": 92
},
{
"epoch": 0.056715962799207195,
"grad_norm": 0.9332770109176636,
"learning_rate": 9.982793408907747e-05,
"loss": 0.2198,
"step": 93
},
{
"epoch": 0.05732581186156426,
"grad_norm": 0.801500678062439,
"learning_rate": 9.981964791918021e-05,
"loss": 0.1698,
"step": 94
},
{
"epoch": 0.05793566092392133,
"grad_norm": 0.5751333832740784,
"learning_rate": 9.981116725537194e-05,
"loss": 0.1856,
"step": 95
},
{
"epoch": 0.058545509986278395,
"grad_norm": 0.6595227718353271,
"learning_rate": 9.980249213076084e-05,
"loss": 0.183,
"step": 96
},
{
"epoch": 0.05915535904863546,
"grad_norm": 0.5908553600311279,
"learning_rate": 9.979362257921427e-05,
"loss": 0.165,
"step": 97
},
{
"epoch": 0.05976520811099253,
"grad_norm": 0.5509040355682373,
"learning_rate": 9.978455863535859e-05,
"loss": 0.1613,
"step": 98
},
{
"epoch": 0.060375057173349594,
"grad_norm": 0.8269245028495789,
"learning_rate": 9.977530033457905e-05,
"loss": 0.2185,
"step": 99
},
{
"epoch": 0.06098490623570666,
"grad_norm": 0.8192322850227356,
"learning_rate": 9.97658477130197e-05,
"loss": 0.2298,
"step": 100
},
{
"epoch": 0.06159475529806373,
"grad_norm": 0.6920491456985474,
"learning_rate": 9.97562008075832e-05,
"loss": 0.19,
"step": 101
},
{
"epoch": 0.062204604360420794,
"grad_norm": 0.5310209393501282,
"learning_rate": 9.97463596559307e-05,
"loss": 0.1521,
"step": 102
},
{
"epoch": 0.06281445342277786,
"grad_norm": 0.750678300857544,
"learning_rate": 9.973632429648165e-05,
"loss": 0.2227,
"step": 103
},
{
"epoch": 0.06342430248513493,
"grad_norm": 0.7221071124076843,
"learning_rate": 9.972609476841367e-05,
"loss": 0.1662,
"step": 104
},
{
"epoch": 0.064034151547492,
"grad_norm": 0.8504064679145813,
"learning_rate": 9.971567111166246e-05,
"loss": 0.1817,
"step": 105
},
{
"epoch": 0.06464400060984907,
"grad_norm": 0.5753822326660156,
"learning_rate": 9.970505336692153e-05,
"loss": 0.1713,
"step": 106
},
{
"epoch": 0.06525384967220613,
"grad_norm": 1.5662932395935059,
"learning_rate": 9.969424157564215e-05,
"loss": 0.1973,
"step": 107
},
{
"epoch": 0.0658636987345632,
"grad_norm": 0.5891379117965698,
"learning_rate": 9.968323578003311e-05,
"loss": 0.1452,
"step": 108
},
{
"epoch": 0.06647354779692026,
"grad_norm": 0.714789867401123,
"learning_rate": 9.967203602306061e-05,
"loss": 0.1555,
"step": 109
},
{
"epoch": 0.06708339685927733,
"grad_norm": 1.1253595352172852,
"learning_rate": 9.966064234844804e-05,
"loss": 0.1717,
"step": 110
},
{
"epoch": 0.06769324592163439,
"grad_norm": 0.6432620882987976,
"learning_rate": 9.964905480067586e-05,
"loss": 0.1538,
"step": 111
},
{
"epoch": 0.06830309498399147,
"grad_norm": 0.6219606995582581,
"learning_rate": 9.963727342498137e-05,
"loss": 0.1164,
"step": 112
},
{
"epoch": 0.06891294404634853,
"grad_norm": 0.6670606136322021,
"learning_rate": 9.96252982673586e-05,
"loss": 0.2202,
"step": 113
},
{
"epoch": 0.0695227931087056,
"grad_norm": 0.709822416305542,
"learning_rate": 9.961312937455811e-05,
"loss": 0.1908,
"step": 114
},
{
"epoch": 0.07013264217106266,
"grad_norm": 0.6181677579879761,
"learning_rate": 9.960076679408674e-05,
"loss": 0.1355,
"step": 115
},
{
"epoch": 0.07074249123341973,
"grad_norm": 0.6124622225761414,
"learning_rate": 9.958821057420754e-05,
"loss": 0.1693,
"step": 116
},
{
"epoch": 0.07135234029577679,
"grad_norm": 0.6326810121536255,
"learning_rate": 9.957546076393943e-05,
"loss": 0.1983,
"step": 117
},
{
"epoch": 0.07196218935813387,
"grad_norm": 0.6671566963195801,
"learning_rate": 9.956251741305722e-05,
"loss": 0.1552,
"step": 118
},
{
"epoch": 0.07257203842049093,
"grad_norm": 0.7035337686538696,
"learning_rate": 9.954938057209121e-05,
"loss": 0.1456,
"step": 119
},
{
"epoch": 0.073181887482848,
"grad_norm": 0.8739194869995117,
"learning_rate": 9.953605029232711e-05,
"loss": 0.179,
"step": 120
},
{
"epoch": 0.07379173654520506,
"grad_norm": 0.7865602374076843,
"learning_rate": 9.952252662580579e-05,
"loss": 0.16,
"step": 121
},
{
"epoch": 0.07440158560756213,
"grad_norm": 0.8371756672859192,
"learning_rate": 9.950880962532309e-05,
"loss": 0.1845,
"step": 122
},
{
"epoch": 0.07501143466991919,
"grad_norm": 0.6414897441864014,
"learning_rate": 9.949489934442966e-05,
"loss": 0.1318,
"step": 123
},
{
"epoch": 0.07562128373227626,
"grad_norm": 0.8123409748077393,
"learning_rate": 9.948079583743067e-05,
"loss": 0.1715,
"step": 124
},
{
"epoch": 0.07623113279463332,
"grad_norm": 0.696685254573822,
"learning_rate": 9.946649915938562e-05,
"loss": 0.1771,
"step": 125
},
{
"epoch": 0.0768409818569904,
"grad_norm": 0.6860955953598022,
"learning_rate": 9.94520093661082e-05,
"loss": 0.1928,
"step": 126
},
{
"epoch": 0.07745083091934746,
"grad_norm": 0.6983869075775146,
"learning_rate": 9.943732651416597e-05,
"loss": 0.1657,
"step": 127
},
{
"epoch": 0.07806067998170453,
"grad_norm": 0.5797171592712402,
"learning_rate": 9.942245066088021e-05,
"loss": 0.1588,
"step": 128
},
{
"epoch": 0.07867052904406159,
"grad_norm": 0.6848919987678528,
"learning_rate": 9.940738186432565e-05,
"loss": 0.1578,
"step": 129
},
{
"epoch": 0.07928037810641866,
"grad_norm": 0.599567174911499,
"learning_rate": 9.939212018333023e-05,
"loss": 0.1565,
"step": 130
},
{
"epoch": 0.07989022716877572,
"grad_norm": 0.8321368098258972,
"learning_rate": 9.937666567747501e-05,
"loss": 0.2055,
"step": 131
},
{
"epoch": 0.0805000762311328,
"grad_norm": 0.9468752145767212,
"learning_rate": 9.936101840709372e-05,
"loss": 0.183,
"step": 132
},
{
"epoch": 0.08110992529348986,
"grad_norm": 0.595227062702179,
"learning_rate": 9.934517843327269e-05,
"loss": 0.1307,
"step": 133
},
{
"epoch": 0.08171977435584693,
"grad_norm": 0.5828656554222107,
"learning_rate": 9.932914581785052e-05,
"loss": 0.1684,
"step": 134
},
{
"epoch": 0.08232962341820399,
"grad_norm": 0.5222749710083008,
"learning_rate": 9.931292062341793e-05,
"loss": 0.1527,
"step": 135
},
{
"epoch": 0.08293947248056106,
"grad_norm": 0.7175277471542358,
"learning_rate": 9.92965029133174e-05,
"loss": 0.145,
"step": 136
},
{
"epoch": 0.08354932154291812,
"grad_norm": 0.8291264176368713,
"learning_rate": 9.927989275164305e-05,
"loss": 0.1877,
"step": 137
},
{
"epoch": 0.0841591706052752,
"grad_norm": 1.054272174835205,
"learning_rate": 9.926309020324025e-05,
"loss": 0.239,
"step": 138
},
{
"epoch": 0.08476901966763226,
"grad_norm": 0.7082330584526062,
"learning_rate": 9.924609533370551e-05,
"loss": 0.149,
"step": 139
},
{
"epoch": 0.08537886872998933,
"grad_norm": 0.6745434403419495,
"learning_rate": 9.922890820938608e-05,
"loss": 0.1965,
"step": 140
},
{
"epoch": 0.08598871779234639,
"grad_norm": 0.5474583506584167,
"learning_rate": 9.921152889737984e-05,
"loss": 0.1358,
"step": 141
},
{
"epoch": 0.08659856685470346,
"grad_norm": 0.6209768652915955,
"learning_rate": 9.919395746553493e-05,
"loss": 0.1706,
"step": 142
},
{
"epoch": 0.08720841591706052,
"grad_norm": 0.7166369557380676,
"learning_rate": 9.917619398244949e-05,
"loss": 0.1479,
"step": 143
},
{
"epoch": 0.0878182649794176,
"grad_norm": 0.5569589138031006,
"learning_rate": 9.915823851747144e-05,
"loss": 0.1566,
"step": 144
},
{
"epoch": 0.08842811404177466,
"grad_norm": 0.570838451385498,
"learning_rate": 9.914009114069824e-05,
"loss": 0.1723,
"step": 145
},
{
"epoch": 0.08903796310413173,
"grad_norm": 0.7048601508140564,
"learning_rate": 9.912175192297648e-05,
"loss": 0.1812,
"step": 146
},
{
"epoch": 0.08964781216648879,
"grad_norm": 0.5294231176376343,
"learning_rate": 9.910322093590177e-05,
"loss": 0.1309,
"step": 147
},
{
"epoch": 0.09025766122884586,
"grad_norm": 0.7757968306541443,
"learning_rate": 9.908449825181829e-05,
"loss": 0.1597,
"step": 148
},
{
"epoch": 0.09086751029120292,
"grad_norm": 0.6745859980583191,
"learning_rate": 9.90655839438187e-05,
"loss": 0.1689,
"step": 149
},
{
"epoch": 0.09147735935356,
"grad_norm": 0.9369275569915771,
"learning_rate": 9.90464780857437e-05,
"loss": 0.2189,
"step": 150
},
{
"epoch": 0.09208720841591705,
"grad_norm": 0.6415305137634277,
"learning_rate": 9.902718075218176e-05,
"loss": 0.1709,
"step": 151
},
{
"epoch": 0.09269705747827413,
"grad_norm": 0.7711929082870483,
"learning_rate": 9.90076920184689e-05,
"loss": 0.1713,
"step": 152
},
{
"epoch": 0.09330690654063119,
"grad_norm": 0.6208341121673584,
"learning_rate": 9.898801196068839e-05,
"loss": 0.1771,
"step": 153
},
{
"epoch": 0.09391675560298826,
"grad_norm": 0.743375837802887,
"learning_rate": 9.896814065567036e-05,
"loss": 0.1845,
"step": 154
},
{
"epoch": 0.09452660466534532,
"grad_norm": 0.7383750677108765,
"learning_rate": 9.89480781809916e-05,
"loss": 0.1385,
"step": 155
},
{
"epoch": 0.0951364537277024,
"grad_norm": 0.6448755264282227,
"learning_rate": 9.89278246149752e-05,
"loss": 0.1536,
"step": 156
},
{
"epoch": 0.09574630279005945,
"grad_norm": 0.7740316987037659,
"learning_rate": 9.890738003669029e-05,
"loss": 0.1792,
"step": 157
},
{
"epoch": 0.09635615185241653,
"grad_norm": 0.6544023156166077,
"learning_rate": 9.888674452595166e-05,
"loss": 0.1743,
"step": 158
},
{
"epoch": 0.09696600091477359,
"grad_norm": 0.4671114683151245,
"learning_rate": 9.886591816331954e-05,
"loss": 0.1134,
"step": 159
},
{
"epoch": 0.09757584997713066,
"grad_norm": 0.9692893028259277,
"learning_rate": 9.88449010300992e-05,
"loss": 0.1387,
"step": 160
},
{
"epoch": 0.09818569903948772,
"grad_norm": 0.5491570830345154,
"learning_rate": 9.882369320834069e-05,
"loss": 0.1745,
"step": 161
},
{
"epoch": 0.0987955481018448,
"grad_norm": 0.7274909019470215,
"learning_rate": 9.880229478083849e-05,
"loss": 0.1828,
"step": 162
},
{
"epoch": 0.09940539716420187,
"grad_norm": 0.9059796333312988,
"learning_rate": 9.878070583113123e-05,
"loss": 0.1733,
"step": 163
},
{
"epoch": 0.10001524622655893,
"grad_norm": 0.6845661401748657,
"learning_rate": 9.875892644350128e-05,
"loss": 0.1491,
"step": 164
},
{
"epoch": 0.100625095288916,
"grad_norm": 0.5804721713066101,
"learning_rate": 9.87369567029745e-05,
"loss": 0.1735,
"step": 165
},
{
"epoch": 0.10123494435127306,
"grad_norm": 0.7379568219184875,
"learning_rate": 9.87147966953199e-05,
"loss": 0.1564,
"step": 166
},
{
"epoch": 0.10184479341363013,
"grad_norm": 0.5581180453300476,
"learning_rate": 9.869244650704923e-05,
"loss": 0.1462,
"step": 167
},
{
"epoch": 0.10245464247598719,
"grad_norm": 0.5281278491020203,
"learning_rate": 9.866990622541677e-05,
"loss": 0.1568,
"step": 168
},
{
"epoch": 0.10306449153834427,
"grad_norm": 0.6497232913970947,
"learning_rate": 9.864717593841883e-05,
"loss": 0.1718,
"step": 169
},
{
"epoch": 0.10367434060070133,
"grad_norm": 0.613874077796936,
"learning_rate": 9.862425573479357e-05,
"loss": 0.1906,
"step": 170
},
{
"epoch": 0.1042841896630584,
"grad_norm": 0.4820992648601532,
"learning_rate": 9.860114570402054e-05,
"loss": 0.1364,
"step": 171
},
{
"epoch": 0.10489403872541546,
"grad_norm": 0.46377211809158325,
"learning_rate": 9.857784593632038e-05,
"loss": 0.1503,
"step": 172
},
{
"epoch": 0.10550388778777253,
"grad_norm": 0.6748605370521545,
"learning_rate": 9.855435652265446e-05,
"loss": 0.2078,
"step": 173
},
{
"epoch": 0.10611373685012959,
"grad_norm": 0.6897035241127014,
"learning_rate": 9.853067755472446e-05,
"loss": 0.2135,
"step": 174
},
{
"epoch": 0.10672358591248667,
"grad_norm": 0.6089524030685425,
"learning_rate": 9.85068091249722e-05,
"loss": 0.148,
"step": 175
},
{
"epoch": 0.10733343497484372,
"grad_norm": 0.6201213598251343,
"learning_rate": 9.848275132657903e-05,
"loss": 0.1608,
"step": 176
},
{
"epoch": 0.1079432840372008,
"grad_norm": 0.6302520632743835,
"learning_rate": 9.845850425346563e-05,
"loss": 0.165,
"step": 177
},
{
"epoch": 0.10855313309955786,
"grad_norm": 0.8897461891174316,
"learning_rate": 9.84340680002916e-05,
"loss": 0.1849,
"step": 178
},
{
"epoch": 0.10916298216191493,
"grad_norm": 0.6461714506149292,
"learning_rate": 9.840944266245511e-05,
"loss": 0.1514,
"step": 179
},
{
"epoch": 0.10977283122427199,
"grad_norm": 3.0706777572631836,
"learning_rate": 9.838462833609248e-05,
"loss": 0.1704,
"step": 180
},
{
"epoch": 0.11038268028662906,
"grad_norm": 0.5703252553939819,
"learning_rate": 9.835962511807786e-05,
"loss": 0.1302,
"step": 181
},
{
"epoch": 0.11099252934898612,
"grad_norm": 0.5591781735420227,
"learning_rate": 9.83344331060228e-05,
"loss": 0.1391,
"step": 182
},
{
"epoch": 0.1116023784113432,
"grad_norm": 0.5740253925323486,
"learning_rate": 9.830905239827593e-05,
"loss": 0.15,
"step": 183
},
{
"epoch": 0.11221222747370026,
"grad_norm": 0.4672653377056122,
"learning_rate": 9.828348309392247e-05,
"loss": 0.1268,
"step": 184
},
{
"epoch": 0.11282207653605733,
"grad_norm": 0.6052584052085876,
"learning_rate": 9.825772529278401e-05,
"loss": 0.1546,
"step": 185
},
{
"epoch": 0.11343192559841439,
"grad_norm": 0.7341001033782959,
"learning_rate": 9.823177909541794e-05,
"loss": 0.1517,
"step": 186
},
{
"epoch": 0.11404177466077146,
"grad_norm": 0.6068568229675293,
"learning_rate": 9.820564460311718e-05,
"loss": 0.128,
"step": 187
},
{
"epoch": 0.11465162372312852,
"grad_norm": 0.5739975571632385,
"learning_rate": 9.817932191790978e-05,
"loss": 0.1292,
"step": 188
},
{
"epoch": 0.1152614727854856,
"grad_norm": 0.5718744397163391,
"learning_rate": 9.815281114255841e-05,
"loss": 0.1525,
"step": 189
},
{
"epoch": 0.11587132184784266,
"grad_norm": 0.45995739102363586,
"learning_rate": 9.812611238056009e-05,
"loss": 0.1146,
"step": 190
},
{
"epoch": 0.11648117091019973,
"grad_norm": 1.0611133575439453,
"learning_rate": 9.809922573614569e-05,
"loss": 0.1542,
"step": 191
},
{
"epoch": 0.11709101997255679,
"grad_norm": 0.7767623066902161,
"learning_rate": 9.807215131427965e-05,
"loss": 0.192,
"step": 192
},
{
"epoch": 0.11770086903491386,
"grad_norm": 0.8179883360862732,
"learning_rate": 9.804488922065937e-05,
"loss": 0.1387,
"step": 193
},
{
"epoch": 0.11831071809727092,
"grad_norm": 0.8066700100898743,
"learning_rate": 9.801743956171501e-05,
"loss": 0.2009,
"step": 194
},
{
"epoch": 0.118920567159628,
"grad_norm": 0.9031579494476318,
"learning_rate": 9.798980244460893e-05,
"loss": 0.1903,
"step": 195
},
{
"epoch": 0.11953041622198506,
"grad_norm": 0.6633782982826233,
"learning_rate": 9.796197797723532e-05,
"loss": 0.1238,
"step": 196
},
{
"epoch": 0.12014026528434213,
"grad_norm": 0.7691205739974976,
"learning_rate": 9.79339662682198e-05,
"loss": 0.1662,
"step": 197
},
{
"epoch": 0.12075011434669919,
"grad_norm": 0.5919376611709595,
"learning_rate": 9.790576742691895e-05,
"loss": 0.122,
"step": 198
},
{
"epoch": 0.12135996340905626,
"grad_norm": 0.6353310942649841,
"learning_rate": 9.787738156341992e-05,
"loss": 0.1418,
"step": 199
},
{
"epoch": 0.12196981247141332,
"grad_norm": 0.5124211311340332,
"learning_rate": 9.784880878854e-05,
"loss": 0.1029,
"step": 200
},
{
"epoch": 0.1225796615337704,
"grad_norm": 0.7390457987785339,
"learning_rate": 9.782004921382612e-05,
"loss": 0.1166,
"step": 201
},
{
"epoch": 0.12318951059612746,
"grad_norm": 0.6599264740943909,
"learning_rate": 9.779110295155456e-05,
"loss": 0.1292,
"step": 202
},
{
"epoch": 0.12379935965848453,
"grad_norm": 0.7947335839271545,
"learning_rate": 9.776197011473033e-05,
"loss": 0.1755,
"step": 203
},
{
"epoch": 0.12440920872084159,
"grad_norm": 0.7999613881111145,
"learning_rate": 9.773265081708687e-05,
"loss": 0.1776,
"step": 204
},
{
"epoch": 0.12501905778319866,
"grad_norm": 0.7566004395484924,
"learning_rate": 9.770314517308554e-05,
"loss": 0.1664,
"step": 205
},
{
"epoch": 0.12562890684555572,
"grad_norm": 0.6981890201568604,
"learning_rate": 9.767345329791522e-05,
"loss": 0.1452,
"step": 206
},
{
"epoch": 0.12623875590791278,
"grad_norm": 0.6032444834709167,
"learning_rate": 9.764357530749178e-05,
"loss": 0.1253,
"step": 207
},
{
"epoch": 0.12684860497026987,
"grad_norm": 0.7930732369422913,
"learning_rate": 9.761351131845768e-05,
"loss": 0.1707,
"step": 208
},
{
"epoch": 0.12745845403262693,
"grad_norm": 0.6833851337432861,
"learning_rate": 9.758326144818155e-05,
"loss": 0.1569,
"step": 209
},
{
"epoch": 0.128068303094984,
"grad_norm": 0.5903204679489136,
"learning_rate": 9.755282581475769e-05,
"loss": 0.148,
"step": 210
},
{
"epoch": 0.12867815215734105,
"grad_norm": 0.6545243263244629,
"learning_rate": 9.752220453700556e-05,
"loss": 0.1565,
"step": 211
},
{
"epoch": 0.12928800121969813,
"grad_norm": 0.5270501375198364,
"learning_rate": 9.749139773446943e-05,
"loss": 0.1467,
"step": 212
},
{
"epoch": 0.1298978502820552,
"grad_norm": 0.5369648337364197,
"learning_rate": 9.74604055274178e-05,
"loss": 0.148,
"step": 213
},
{
"epoch": 0.13050769934441225,
"grad_norm": 0.615350604057312,
"learning_rate": 9.742922803684302e-05,
"loss": 0.1886,
"step": 214
},
{
"epoch": 0.1311175484067693,
"grad_norm": 0.5637040734291077,
"learning_rate": 9.739786538446076e-05,
"loss": 0.1587,
"step": 215
},
{
"epoch": 0.1317273974691264,
"grad_norm": 0.5875048041343689,
"learning_rate": 9.736631769270957e-05,
"loss": 0.1898,
"step": 216
},
{
"epoch": 0.13233724653148346,
"grad_norm": 0.7337537407875061,
"learning_rate": 9.733458508475037e-05,
"loss": 0.1434,
"step": 217
},
{
"epoch": 0.13294709559384052,
"grad_norm": 0.48101916909217834,
"learning_rate": 9.730266768446598e-05,
"loss": 0.1374,
"step": 218
},
{
"epoch": 0.13355694465619758,
"grad_norm": 0.6084973812103271,
"learning_rate": 9.727056561646066e-05,
"loss": 0.1808,
"step": 219
},
{
"epoch": 0.13416679371855467,
"grad_norm": 0.5653364062309265,
"learning_rate": 9.723827900605962e-05,
"loss": 0.1898,
"step": 220
},
{
"epoch": 0.13477664278091173,
"grad_norm": 0.9579099416732788,
"learning_rate": 9.720580797930845e-05,
"loss": 0.2098,
"step": 221
},
{
"epoch": 0.13538649184326879,
"grad_norm": 0.5323110222816467,
"learning_rate": 9.717315266297277e-05,
"loss": 0.1585,
"step": 222
},
{
"epoch": 0.13599634090562585,
"grad_norm": 0.5441873669624329,
"learning_rate": 9.714031318453764e-05,
"loss": 0.1267,
"step": 223
},
{
"epoch": 0.13660618996798293,
"grad_norm": 0.5249777436256409,
"learning_rate": 9.710728967220704e-05,
"loss": 0.1556,
"step": 224
},
{
"epoch": 0.13721603903034,
"grad_norm": 0.6340452432632446,
"learning_rate": 9.707408225490344e-05,
"loss": 0.1208,
"step": 225
},
{
"epoch": 0.13782588809269705,
"grad_norm": 0.5065081119537354,
"learning_rate": 9.704069106226727e-05,
"loss": 0.1377,
"step": 226
},
{
"epoch": 0.1384357371550541,
"grad_norm": 0.47786062955856323,
"learning_rate": 9.700711622465643e-05,
"loss": 0.125,
"step": 227
},
{
"epoch": 0.1390455862174112,
"grad_norm": 0.7114598751068115,
"learning_rate": 9.697335787314573e-05,
"loss": 0.1988,
"step": 228
},
{
"epoch": 0.13965543527976826,
"grad_norm": 0.6802954077720642,
"learning_rate": 9.693941613952642e-05,
"loss": 0.1772,
"step": 229
},
{
"epoch": 0.14026528434212532,
"grad_norm": 0.6225735545158386,
"learning_rate": 9.690529115630567e-05,
"loss": 0.1837,
"step": 230
},
{
"epoch": 0.14087513340448238,
"grad_norm": 0.574621856212616,
"learning_rate": 9.687098305670605e-05,
"loss": 0.1344,
"step": 231
},
{
"epoch": 0.14148498246683947,
"grad_norm": 0.6660715341567993,
"learning_rate": 9.6836491974665e-05,
"loss": 0.1612,
"step": 232
},
{
"epoch": 0.14209483152919652,
"grad_norm": 0.6641862392425537,
"learning_rate": 9.680181804483434e-05,
"loss": 0.1506,
"step": 233
},
{
"epoch": 0.14270468059155358,
"grad_norm": 0.5605201125144958,
"learning_rate": 9.676696140257969e-05,
"loss": 0.1369,
"step": 234
},
{
"epoch": 0.14331452965391064,
"grad_norm": 0.6830882430076599,
"learning_rate": 9.673192218398e-05,
"loss": 0.1568,
"step": 235
},
{
"epoch": 0.14392437871626773,
"grad_norm": 0.48301899433135986,
"learning_rate": 9.669670052582695e-05,
"loss": 0.1344,
"step": 236
},
{
"epoch": 0.1445342277786248,
"grad_norm": 0.6020053029060364,
"learning_rate": 9.66612965656245e-05,
"loss": 0.1535,
"step": 237
},
{
"epoch": 0.14514407684098185,
"grad_norm": 0.4740085005760193,
"learning_rate": 9.662571044158831e-05,
"loss": 0.1359,
"step": 238
},
{
"epoch": 0.1457539259033389,
"grad_norm": 0.793387770652771,
"learning_rate": 9.658994229264514e-05,
"loss": 0.1702,
"step": 239
},
{
"epoch": 0.146363774965696,
"grad_norm": 0.6762639880180359,
"learning_rate": 9.655399225843245e-05,
"loss": 0.1604,
"step": 240
},
{
"epoch": 0.14697362402805306,
"grad_norm": 0.5262989401817322,
"learning_rate": 9.651786047929773e-05,
"loss": 0.1574,
"step": 241
},
{
"epoch": 0.14758347309041012,
"grad_norm": 0.6588354706764221,
"learning_rate": 9.648154709629798e-05,
"loss": 0.1512,
"step": 242
},
{
"epoch": 0.1481933221527672,
"grad_norm": 0.5401725172996521,
"learning_rate": 9.644505225119922e-05,
"loss": 0.1443,
"step": 243
},
{
"epoch": 0.14880317121512426,
"grad_norm": 0.6582406163215637,
"learning_rate": 9.640837608647583e-05,
"loss": 0.1401,
"step": 244
},
{
"epoch": 0.14941302027748132,
"grad_norm": 0.4655030369758606,
"learning_rate": 9.637151874531014e-05,
"loss": 0.1424,
"step": 245
},
{
"epoch": 0.15002286933983838,
"grad_norm": 0.6858920454978943,
"learning_rate": 9.633448037159167e-05,
"loss": 0.1851,
"step": 246
},
{
"epoch": 0.15063271840219547,
"grad_norm": 0.9004249572753906,
"learning_rate": 9.62972611099168e-05,
"loss": 0.1774,
"step": 247
},
{
"epoch": 0.15124256746455253,
"grad_norm": 0.5213025212287903,
"learning_rate": 9.625986110558801e-05,
"loss": 0.1331,
"step": 248
},
{
"epoch": 0.1518524165269096,
"grad_norm": 0.6229725480079651,
"learning_rate": 9.622228050461343e-05,
"loss": 0.1259,
"step": 249
},
{
"epoch": 0.15246226558926665,
"grad_norm": 0.5384789705276489,
"learning_rate": 9.618451945370622e-05,
"loss": 0.1458,
"step": 250
},
{
"epoch": 0.15307211465162374,
"grad_norm": 0.5865654349327087,
"learning_rate": 9.614657810028402e-05,
"loss": 0.1293,
"step": 251
},
{
"epoch": 0.1536819637139808,
"grad_norm": 0.5768997073173523,
"learning_rate": 9.610845659246834e-05,
"loss": 0.1461,
"step": 252
},
{
"epoch": 0.15429181277633786,
"grad_norm": 0.5526010990142822,
"learning_rate": 9.607015507908401e-05,
"loss": 0.1633,
"step": 253
},
{
"epoch": 0.15490166183869492,
"grad_norm": 0.6163491010665894,
"learning_rate": 9.603167370965865e-05,
"loss": 0.1427,
"step": 254
},
{
"epoch": 0.155511510901052,
"grad_norm": 0.9269428849220276,
"learning_rate": 9.599301263442192e-05,
"loss": 0.1763,
"step": 255
},
{
"epoch": 0.15612135996340906,
"grad_norm": 0.5760687589645386,
"learning_rate": 9.595417200430516e-05,
"loss": 0.1449,
"step": 256
},
{
"epoch": 0.15673120902576612,
"grad_norm": 0.7990864515304565,
"learning_rate": 9.591515197094064e-05,
"loss": 0.15,
"step": 257
},
{
"epoch": 0.15734105808812318,
"grad_norm": 0.8166128396987915,
"learning_rate": 9.587595268666099e-05,
"loss": 0.1754,
"step": 258
},
{
"epoch": 0.15795090715048027,
"grad_norm": 0.6038745045661926,
"learning_rate": 9.583657430449862e-05,
"loss": 0.1222,
"step": 259
},
{
"epoch": 0.15856075621283733,
"grad_norm": 0.6139023900032043,
"learning_rate": 9.579701697818519e-05,
"loss": 0.143,
"step": 260
},
{
"epoch": 0.1591706052751944,
"grad_norm": 0.7803279757499695,
"learning_rate": 9.575728086215092e-05,
"loss": 0.1752,
"step": 261
},
{
"epoch": 0.15978045433755145,
"grad_norm": 0.5674574375152588,
"learning_rate": 9.571736611152402e-05,
"loss": 0.1396,
"step": 262
},
{
"epoch": 0.16039030339990853,
"grad_norm": 0.6714860796928406,
"learning_rate": 9.567727288213005e-05,
"loss": 0.1281,
"step": 263
},
{
"epoch": 0.1610001524622656,
"grad_norm": 0.5477873086929321,
"learning_rate": 9.563700133049139e-05,
"loss": 0.1355,
"step": 264
},
{
"epoch": 0.16161000152462265,
"grad_norm": 0.6249256730079651,
"learning_rate": 9.559655161382657e-05,
"loss": 0.1398,
"step": 265
},
{
"epoch": 0.1622198505869797,
"grad_norm": 0.689401388168335,
"learning_rate": 9.555592389004966e-05,
"loss": 0.155,
"step": 266
},
{
"epoch": 0.1628296996493368,
"grad_norm": 0.5395768284797668,
"learning_rate": 9.551511831776965e-05,
"loss": 0.1519,
"step": 267
},
{
"epoch": 0.16343954871169386,
"grad_norm": 0.6177000999450684,
"learning_rate": 9.547413505628991e-05,
"loss": 0.1366,
"step": 268
},
{
"epoch": 0.16404939777405092,
"grad_norm": 0.7658178210258484,
"learning_rate": 9.543297426560739e-05,
"loss": 0.1359,
"step": 269
},
{
"epoch": 0.16465924683640798,
"grad_norm": 0.5808216333389282,
"learning_rate": 9.53916361064122e-05,
"loss": 0.148,
"step": 270
},
{
"epoch": 0.16526909589876507,
"grad_norm": 0.7378920316696167,
"learning_rate": 9.535012074008687e-05,
"loss": 0.1395,
"step": 271
},
{
"epoch": 0.16587894496112213,
"grad_norm": 0.7066279649734497,
"learning_rate": 9.53084283287057e-05,
"loss": 0.1425,
"step": 272
},
{
"epoch": 0.1664887940234792,
"grad_norm": 0.6681056022644043,
"learning_rate": 9.526655903503423e-05,
"loss": 0.1551,
"step": 273
},
{
"epoch": 0.16709864308583625,
"grad_norm": 0.5979323387145996,
"learning_rate": 9.522451302252847e-05,
"loss": 0.1423,
"step": 274
},
{
"epoch": 0.16770849214819333,
"grad_norm": 0.5763559341430664,
"learning_rate": 9.518229045533438e-05,
"loss": 0.1364,
"step": 275
},
{
"epoch": 0.1683183412105504,
"grad_norm": 0.5659794211387634,
"learning_rate": 9.513989149828718e-05,
"loss": 0.158,
"step": 276
},
{
"epoch": 0.16892819027290745,
"grad_norm": 0.5583329200744629,
"learning_rate": 9.50973163169107e-05,
"loss": 0.1521,
"step": 277
},
{
"epoch": 0.1695380393352645,
"grad_norm": 0.6379377245903015,
"learning_rate": 9.505456507741675e-05,
"loss": 0.1635,
"step": 278
},
{
"epoch": 0.1701478883976216,
"grad_norm": 0.5981450080871582,
"learning_rate": 9.501163794670444e-05,
"loss": 0.1383,
"step": 279
},
{
"epoch": 0.17075773745997866,
"grad_norm": 0.44199758768081665,
"learning_rate": 9.496853509235958e-05,
"loss": 0.1453,
"step": 280
},
{
"epoch": 0.17136758652233572,
"grad_norm": 0.522309422492981,
"learning_rate": 9.492525668265399e-05,
"loss": 0.1599,
"step": 281
},
{
"epoch": 0.17197743558469278,
"grad_norm": 0.5427623391151428,
"learning_rate": 9.488180288654485e-05,
"loss": 0.1758,
"step": 282
},
{
"epoch": 0.17258728464704987,
"grad_norm": 0.45050814747810364,
"learning_rate": 9.483817387367403e-05,
"loss": 0.1317,
"step": 283
},
{
"epoch": 0.17319713370940693,
"grad_norm": 0.5131423473358154,
"learning_rate": 9.479436981436746e-05,
"loss": 0.1351,
"step": 284
},
{
"epoch": 0.17380698277176398,
"grad_norm": 0.6341767311096191,
"learning_rate": 9.475039087963442e-05,
"loss": 0.1636,
"step": 285
},
{
"epoch": 0.17441683183412104,
"grad_norm": 0.5274122357368469,
"learning_rate": 9.470623724116692e-05,
"loss": 0.1149,
"step": 286
},
{
"epoch": 0.17502668089647813,
"grad_norm": 0.5990244150161743,
"learning_rate": 9.4661909071339e-05,
"loss": 0.1415,
"step": 287
},
{
"epoch": 0.1756365299588352,
"grad_norm": 0.49017491936683655,
"learning_rate": 9.461740654320607e-05,
"loss": 0.1152,
"step": 288
},
{
"epoch": 0.17624637902119225,
"grad_norm": 1.1685532331466675,
"learning_rate": 9.45727298305042e-05,
"loss": 0.149,
"step": 289
},
{
"epoch": 0.1768562280835493,
"grad_norm": 1.0425652265548706,
"learning_rate": 9.452787910764948e-05,
"loss": 0.15,
"step": 290
},
{
"epoch": 0.1774660771459064,
"grad_norm": 0.471015989780426,
"learning_rate": 9.448285454973738e-05,
"loss": 0.1192,
"step": 291
},
{
"epoch": 0.17807592620826346,
"grad_norm": 0.4285735785961151,
"learning_rate": 9.443765633254191e-05,
"loss": 0.1064,
"step": 292
},
{
"epoch": 0.17868577527062052,
"grad_norm": 0.7212364673614502,
"learning_rate": 9.439228463251515e-05,
"loss": 0.1688,
"step": 293
},
{
"epoch": 0.17929562433297758,
"grad_norm": 0.5714042782783508,
"learning_rate": 9.434673962678638e-05,
"loss": 0.1146,
"step": 294
},
{
"epoch": 0.17990547339533466,
"grad_norm": 0.6169041991233826,
"learning_rate": 9.430102149316146e-05,
"loss": 0.145,
"step": 295
},
{
"epoch": 0.18051532245769172,
"grad_norm": 0.6639889478683472,
"learning_rate": 9.425513041012219e-05,
"loss": 0.1495,
"step": 296
},
{
"epoch": 0.18112517152004878,
"grad_norm": 0.5344529747962952,
"learning_rate": 9.420906655682553e-05,
"loss": 0.1247,
"step": 297
},
{
"epoch": 0.18173502058240584,
"grad_norm": 0.6271331906318665,
"learning_rate": 9.416283011310286e-05,
"loss": 0.1359,
"step": 298
},
{
"epoch": 0.18234486964476293,
"grad_norm": 0.9338291883468628,
"learning_rate": 9.411642125945945e-05,
"loss": 0.1911,
"step": 299
},
{
"epoch": 0.18295471870712,
"grad_norm": 0.6480358839035034,
"learning_rate": 9.406984017707361e-05,
"loss": 0.1738,
"step": 300
},
{
"epoch": 0.18356456776947705,
"grad_norm": 0.48333096504211426,
"learning_rate": 9.402308704779599e-05,
"loss": 0.1053,
"step": 301
},
{
"epoch": 0.1841744168318341,
"grad_norm": 0.4472801983356476,
"learning_rate": 9.397616205414896e-05,
"loss": 0.1228,
"step": 302
},
{
"epoch": 0.1847842658941912,
"grad_norm": 0.4834698438644409,
"learning_rate": 9.392906537932582e-05,
"loss": 0.1287,
"step": 303
},
{
"epoch": 0.18539411495654826,
"grad_norm": 0.6475247144699097,
"learning_rate": 9.388179720719008e-05,
"loss": 0.1782,
"step": 304
},
{
"epoch": 0.18600396401890532,
"grad_norm": 0.4615669250488281,
"learning_rate": 9.38343577222748e-05,
"loss": 0.098,
"step": 305
},
{
"epoch": 0.18661381308126238,
"grad_norm": 0.4329100549221039,
"learning_rate": 9.378674710978185e-05,
"loss": 0.1301,
"step": 306
},
{
"epoch": 0.18722366214361946,
"grad_norm": 0.44199681282043457,
"learning_rate": 9.373896555558113e-05,
"loss": 0.1314,
"step": 307
},
{
"epoch": 0.18783351120597652,
"grad_norm": 0.5693058371543884,
"learning_rate": 9.369101324620992e-05,
"loss": 0.1428,
"step": 308
},
{
"epoch": 0.18844336026833358,
"grad_norm": 0.6422009468078613,
"learning_rate": 9.364289036887213e-05,
"loss": 0.128,
"step": 309
},
{
"epoch": 0.18905320933069064,
"grad_norm": 0.6876850724220276,
"learning_rate": 9.359459711143752e-05,
"loss": 0.1598,
"step": 310
},
{
"epoch": 0.18966305839304773,
"grad_norm": 0.4797840118408203,
"learning_rate": 9.354613366244108e-05,
"loss": 0.1328,
"step": 311
},
{
"epoch": 0.1902729074554048,
"grad_norm": 0.6175855994224548,
"learning_rate": 9.349750021108211e-05,
"loss": 0.146,
"step": 312
},
{
"epoch": 0.19088275651776185,
"grad_norm": 0.4904415011405945,
"learning_rate": 9.344869694722372e-05,
"loss": 0.129,
"step": 313
},
{
"epoch": 0.1914926055801189,
"grad_norm": 0.8877284526824951,
"learning_rate": 9.339972406139185e-05,
"loss": 0.1431,
"step": 314
},
{
"epoch": 0.192102454642476,
"grad_norm": 0.6473569273948669,
"learning_rate": 9.335058174477471e-05,
"loss": 0.1641,
"step": 315
},
{
"epoch": 0.19271230370483305,
"grad_norm": 0.6872034072875977,
"learning_rate": 9.330127018922194e-05,
"loss": 0.1454,
"step": 316
},
{
"epoch": 0.19332215276719011,
"grad_norm": 0.6833274364471436,
"learning_rate": 9.325178958724386e-05,
"loss": 0.1537,
"step": 317
},
{
"epoch": 0.19393200182954717,
"grad_norm": 0.7237617373466492,
"learning_rate": 9.320214013201078e-05,
"loss": 0.1602,
"step": 318
},
{
"epoch": 0.19454185089190426,
"grad_norm": 0.7009502053260803,
"learning_rate": 9.315232201735217e-05,
"loss": 0.1207,
"step": 319
},
{
"epoch": 0.19515169995426132,
"grad_norm": 0.8232503533363342,
"learning_rate": 9.310233543775597e-05,
"loss": 0.1911,
"step": 320
},
{
"epoch": 0.19576154901661838,
"grad_norm": 0.4722892940044403,
"learning_rate": 9.305218058836778e-05,
"loss": 0.0987,
"step": 321
},
{
"epoch": 0.19637139807897544,
"grad_norm": 0.5606905221939087,
"learning_rate": 9.300185766499012e-05,
"loss": 0.1363,
"step": 322
},
{
"epoch": 0.19698124714133253,
"grad_norm": 0.6053794026374817,
"learning_rate": 9.295136686408166e-05,
"loss": 0.1538,
"step": 323
},
{
"epoch": 0.1975910962036896,
"grad_norm": 0.4831788241863251,
"learning_rate": 9.290070838275649e-05,
"loss": 0.1294,
"step": 324
},
{
"epoch": 0.19820094526604665,
"grad_norm": 0.5479839444160461,
"learning_rate": 9.284988241878326e-05,
"loss": 0.1284,
"step": 325
},
{
"epoch": 0.19881079432840373,
"grad_norm": 0.5481233596801758,
"learning_rate": 9.279888917058452e-05,
"loss": 0.1237,
"step": 326
},
{
"epoch": 0.1994206433907608,
"grad_norm": 0.6046422123908997,
"learning_rate": 9.274772883723587e-05,
"loss": 0.1365,
"step": 327
},
{
"epoch": 0.20003049245311785,
"grad_norm": 0.47926750779151917,
"learning_rate": 9.26964016184652e-05,
"loss": 0.1087,
"step": 328
},
{
"epoch": 0.2006403415154749,
"grad_norm": 0.6542229056358337,
"learning_rate": 9.264490771465191e-05,
"loss": 0.1502,
"step": 329
},
{
"epoch": 0.201250190577832,
"grad_norm": 0.6066355109214783,
"learning_rate": 9.259324732682615e-05,
"loss": 0.137,
"step": 330
},
{
"epoch": 0.20186003964018906,
"grad_norm": 0.54039466381073,
"learning_rate": 9.254142065666801e-05,
"loss": 0.1331,
"step": 331
},
{
"epoch": 0.20246988870254612,
"grad_norm": 0.4951845705509186,
"learning_rate": 9.248942790650672e-05,
"loss": 0.113,
"step": 332
},
{
"epoch": 0.20307973776490318,
"grad_norm": 0.7850728631019592,
"learning_rate": 9.243726927931991e-05,
"loss": 0.1144,
"step": 333
},
{
"epoch": 0.20368958682726027,
"grad_norm": 0.5985704064369202,
"learning_rate": 9.238494497873273e-05,
"loss": 0.1159,
"step": 334
},
{
"epoch": 0.20429943588961733,
"grad_norm": 0.6223556399345398,
"learning_rate": 9.233245520901723e-05,
"loss": 0.1249,
"step": 335
},
{
"epoch": 0.20490928495197439,
"grad_norm": 0.5848407745361328,
"learning_rate": 9.22798001750913e-05,
"loss": 0.115,
"step": 336
},
{
"epoch": 0.20551913401433144,
"grad_norm": 0.7041256427764893,
"learning_rate": 9.222698008251813e-05,
"loss": 0.153,
"step": 337
},
{
"epoch": 0.20612898307668853,
"grad_norm": 0.9106941223144531,
"learning_rate": 9.217399513750524e-05,
"loss": 0.1669,
"step": 338
},
{
"epoch": 0.2067388321390456,
"grad_norm": 0.616457998752594,
"learning_rate": 9.21208455469037e-05,
"loss": 0.1389,
"step": 339
},
{
"epoch": 0.20734868120140265,
"grad_norm": 0.6155223250389099,
"learning_rate": 9.206753151820742e-05,
"loss": 0.1086,
"step": 340
},
{
"epoch": 0.2079585302637597,
"grad_norm": 0.500391960144043,
"learning_rate": 9.201405325955221e-05,
"loss": 0.0977,
"step": 341
},
{
"epoch": 0.2085683793261168,
"grad_norm": 0.5170807242393494,
"learning_rate": 9.196041097971508e-05,
"loss": 0.1176,
"step": 342
},
{
"epoch": 0.20917822838847386,
"grad_norm": 0.5956445336341858,
"learning_rate": 9.190660488811331e-05,
"loss": 0.1382,
"step": 343
},
{
"epoch": 0.20978807745083092,
"grad_norm": 0.8480910658836365,
"learning_rate": 9.185263519480377e-05,
"loss": 0.1782,
"step": 344
},
{
"epoch": 0.21039792651318798,
"grad_norm": 0.72772616147995,
"learning_rate": 9.179850211048193e-05,
"loss": 0.148,
"step": 345
},
{
"epoch": 0.21100777557554506,
"grad_norm": 0.5668420791625977,
"learning_rate": 9.174420584648123e-05,
"loss": 0.1181,
"step": 346
},
{
"epoch": 0.21161762463790212,
"grad_norm": 0.4397040903568268,
"learning_rate": 9.168974661477205e-05,
"loss": 0.1016,
"step": 347
},
{
"epoch": 0.21222747370025918,
"grad_norm": 0.5080854296684265,
"learning_rate": 9.163512462796114e-05,
"loss": 0.1194,
"step": 348
},
{
"epoch": 0.21283732276261624,
"grad_norm": 0.5484637022018433,
"learning_rate": 9.158034009929046e-05,
"loss": 0.127,
"step": 349
},
{
"epoch": 0.21344717182497333,
"grad_norm": 0.5635702013969421,
"learning_rate": 9.152539324263667e-05,
"loss": 0.1314,
"step": 350
},
{
"epoch": 0.2140570208873304,
"grad_norm": 0.7773858308792114,
"learning_rate": 9.14702842725101e-05,
"loss": 0.1671,
"step": 351
},
{
"epoch": 0.21466686994968745,
"grad_norm": 0.4813694357872009,
"learning_rate": 9.141501340405394e-05,
"loss": 0.1378,
"step": 352
},
{
"epoch": 0.2152767190120445,
"grad_norm": 0.5751268863677979,
"learning_rate": 9.135958085304344e-05,
"loss": 0.1248,
"step": 353
},
{
"epoch": 0.2158865680744016,
"grad_norm": 0.5828167200088501,
"learning_rate": 9.13039868358851e-05,
"loss": 0.1319,
"step": 354
},
{
"epoch": 0.21649641713675866,
"grad_norm": 0.6154289841651917,
"learning_rate": 9.12482315696157e-05,
"loss": 0.1649,
"step": 355
},
{
"epoch": 0.21710626619911572,
"grad_norm": 0.579248309135437,
"learning_rate": 9.119231527190158e-05,
"loss": 0.154,
"step": 356
},
{
"epoch": 0.21771611526147278,
"grad_norm": 0.5655073523521423,
"learning_rate": 9.113623816103773e-05,
"loss": 0.1245,
"step": 357
},
{
"epoch": 0.21832596432382986,
"grad_norm": 0.4601236879825592,
"learning_rate": 9.108000045594695e-05,
"loss": 0.1118,
"step": 358
},
{
"epoch": 0.21893581338618692,
"grad_norm": 0.6339589357376099,
"learning_rate": 9.102360237617899e-05,
"loss": 0.139,
"step": 359
},
{
"epoch": 0.21954566244854398,
"grad_norm": 0.5259086489677429,
"learning_rate": 9.096704414190969e-05,
"loss": 0.1157,
"step": 360
},
{
"epoch": 0.22015551151090104,
"grad_norm": 0.5676952004432678,
"learning_rate": 9.091032597394012e-05,
"loss": 0.1204,
"step": 361
},
{
"epoch": 0.22076536057325813,
"grad_norm": 0.5581339001655579,
"learning_rate": 9.085344809369575e-05,
"loss": 0.1296,
"step": 362
},
{
"epoch": 0.2213752096356152,
"grad_norm": 0.6574373245239258,
"learning_rate": 9.079641072322556e-05,
"loss": 0.1406,
"step": 363
},
{
"epoch": 0.22198505869797225,
"grad_norm": 0.9715606570243835,
"learning_rate": 9.073921408520115e-05,
"loss": 0.2091,
"step": 364
},
{
"epoch": 0.2225949077603293,
"grad_norm": 0.4806126356124878,
"learning_rate": 9.068185840291588e-05,
"loss": 0.1111,
"step": 365
},
{
"epoch": 0.2232047568226864,
"grad_norm": 0.4505021572113037,
"learning_rate": 9.062434390028407e-05,
"loss": 0.1141,
"step": 366
},
{
"epoch": 0.22381460588504345,
"grad_norm": 0.4007764160633087,
"learning_rate": 9.056667080184003e-05,
"loss": 0.0971,
"step": 367
},
{
"epoch": 0.22442445494740051,
"grad_norm": 0.45892658829689026,
"learning_rate": 9.050883933273722e-05,
"loss": 0.1113,
"step": 368
},
{
"epoch": 0.22503430400975757,
"grad_norm": 0.47205668687820435,
"learning_rate": 9.045084971874738e-05,
"loss": 0.1106,
"step": 369
},
{
"epoch": 0.22564415307211466,
"grad_norm": 0.6841827630996704,
"learning_rate": 9.039270218625964e-05,
"loss": 0.1299,
"step": 370
},
{
"epoch": 0.22625400213447172,
"grad_norm": 0.6656270623207092,
"learning_rate": 9.033439696227965e-05,
"loss": 0.1596,
"step": 371
},
{
"epoch": 0.22686385119682878,
"grad_norm": 0.5883314609527588,
"learning_rate": 9.027593427442867e-05,
"loss": 0.1299,
"step": 372
},
{
"epoch": 0.22747370025918584,
"grad_norm": 0.6299435496330261,
"learning_rate": 9.021731435094268e-05,
"loss": 0.1516,
"step": 373
},
{
"epoch": 0.22808354932154293,
"grad_norm": 0.469434529542923,
"learning_rate": 9.015853742067152e-05,
"loss": 0.0748,
"step": 374
},
{
"epoch": 0.2286933983839,
"grad_norm": 0.5710963606834412,
"learning_rate": 9.009960371307798e-05,
"loss": 0.1279,
"step": 375
},
{
"epoch": 0.22930324744625705,
"grad_norm": 0.83963942527771,
"learning_rate": 9.004051345823689e-05,
"loss": 0.1476,
"step": 376
},
{
"epoch": 0.2299130965086141,
"grad_norm": 0.5791646242141724,
"learning_rate": 8.998126688683422e-05,
"loss": 0.1265,
"step": 377
},
{
"epoch": 0.2305229455709712,
"grad_norm": 0.49047380685806274,
"learning_rate": 8.992186423016625e-05,
"loss": 0.1078,
"step": 378
},
{
"epoch": 0.23113279463332825,
"grad_norm": 0.7132998108863831,
"learning_rate": 8.986230572013855e-05,
"loss": 0.1355,
"step": 379
},
{
"epoch": 0.2317426436956853,
"grad_norm": 0.5791451334953308,
"learning_rate": 8.980259158926516e-05,
"loss": 0.1373,
"step": 380
},
{
"epoch": 0.23235249275804237,
"grad_norm": 0.6328429579734802,
"learning_rate": 8.974272207066767e-05,
"loss": 0.1508,
"step": 381
},
{
"epoch": 0.23296234182039946,
"grad_norm": 0.6854023337364197,
"learning_rate": 8.968269739807427e-05,
"loss": 0.1389,
"step": 382
},
{
"epoch": 0.23357219088275652,
"grad_norm": 0.5477676391601562,
"learning_rate": 8.962251780581887e-05,
"loss": 0.1281,
"step": 383
},
{
"epoch": 0.23418203994511358,
"grad_norm": 0.9283621907234192,
"learning_rate": 8.956218352884022e-05,
"loss": 0.1881,
"step": 384
},
{
"epoch": 0.23479188900747064,
"grad_norm": 0.5491249561309814,
"learning_rate": 8.95016948026809e-05,
"loss": 0.1512,
"step": 385
},
{
"epoch": 0.23540173806982773,
"grad_norm": 0.5348832011222839,
"learning_rate": 8.944105186348646e-05,
"loss": 0.1403,
"step": 386
},
{
"epoch": 0.23601158713218479,
"grad_norm": 0.46851012110710144,
"learning_rate": 8.938025494800454e-05,
"loss": 0.1229,
"step": 387
},
{
"epoch": 0.23662143619454185,
"grad_norm": 0.46080854535102844,
"learning_rate": 8.931930429358384e-05,
"loss": 0.162,
"step": 388
},
{
"epoch": 0.2372312852568989,
"grad_norm": 0.8089540004730225,
"learning_rate": 8.925820013817329e-05,
"loss": 0.1317,
"step": 389
},
{
"epoch": 0.237841134319256,
"grad_norm": 0.40738773345947266,
"learning_rate": 8.919694272032107e-05,
"loss": 0.1008,
"step": 390
},
{
"epoch": 0.23845098338161305,
"grad_norm": 0.5190976858139038,
"learning_rate": 8.913553227917367e-05,
"loss": 0.1339,
"step": 391
},
{
"epoch": 0.2390608324439701,
"grad_norm": 0.7443719506263733,
"learning_rate": 8.9073969054475e-05,
"loss": 0.1275,
"step": 392
},
{
"epoch": 0.23967068150632717,
"grad_norm": 0.5426774024963379,
"learning_rate": 8.901225328656542e-05,
"loss": 0.1343,
"step": 393
},
{
"epoch": 0.24028053056868426,
"grad_norm": 0.6413959264755249,
"learning_rate": 8.895038521638084e-05,
"loss": 0.1717,
"step": 394
},
{
"epoch": 0.24089037963104132,
"grad_norm": 0.5257138013839722,
"learning_rate": 8.88883650854517e-05,
"loss": 0.1234,
"step": 395
},
{
"epoch": 0.24150022869339838,
"grad_norm": 0.6546505093574524,
"learning_rate": 8.882619313590212e-05,
"loss": 0.1454,
"step": 396
},
{
"epoch": 0.24211007775575544,
"grad_norm": 0.7410861253738403,
"learning_rate": 8.876386961044891e-05,
"loss": 0.1613,
"step": 397
},
{
"epoch": 0.24271992681811252,
"grad_norm": 0.42969658970832825,
"learning_rate": 8.87013947524006e-05,
"loss": 0.1042,
"step": 398
},
{
"epoch": 0.24332977588046958,
"grad_norm": 0.6237689852714539,
"learning_rate": 8.863876880565656e-05,
"loss": 0.1485,
"step": 399
},
{
"epoch": 0.24393962494282664,
"grad_norm": 0.6199328303337097,
"learning_rate": 8.857599201470596e-05,
"loss": 0.0994,
"step": 400
},
{
"epoch": 0.2445494740051837,
"grad_norm": 0.5510534048080444,
"learning_rate": 8.851306462462688e-05,
"loss": 0.1185,
"step": 401
},
{
"epoch": 0.2451593230675408,
"grad_norm": 0.5354424715042114,
"learning_rate": 8.844998688108535e-05,
"loss": 0.1431,
"step": 402
},
{
"epoch": 0.24576917212989785,
"grad_norm": 0.6038997173309326,
"learning_rate": 8.83867590303343e-05,
"loss": 0.1372,
"step": 403
},
{
"epoch": 0.2463790211922549,
"grad_norm": 0.6790061593055725,
"learning_rate": 8.83233813192128e-05,
"loss": 0.1292,
"step": 404
},
{
"epoch": 0.246988870254612,
"grad_norm": 0.4462777078151703,
"learning_rate": 8.825985399514487e-05,
"loss": 0.1034,
"step": 405
},
{
"epoch": 0.24759871931696906,
"grad_norm": 0.7352806925773621,
"learning_rate": 8.819617730613862e-05,
"loss": 0.1389,
"step": 406
},
{
"epoch": 0.24820856837932612,
"grad_norm": 0.5056217908859253,
"learning_rate": 8.813235150078531e-05,
"loss": 0.1023,
"step": 407
},
{
"epoch": 0.24881841744168318,
"grad_norm": 0.4696578085422516,
"learning_rate": 8.806837682825835e-05,
"loss": 0.1206,
"step": 408
},
{
"epoch": 0.24942826650404026,
"grad_norm": 0.7777018547058105,
"learning_rate": 8.800425353831226e-05,
"loss": 0.1485,
"step": 409
},
{
"epoch": 0.2500381155663973,
"grad_norm": 0.5746772289276123,
"learning_rate": 8.793998188128183e-05,
"loss": 0.1566,
"step": 410
},
{
"epoch": 0.2506479646287544,
"grad_norm": 0.7811577916145325,
"learning_rate": 8.787556210808101e-05,
"loss": 0.1812,
"step": 411
},
{
"epoch": 0.25125781369111144,
"grad_norm": 0.5522210001945496,
"learning_rate": 8.781099447020203e-05,
"loss": 0.1338,
"step": 412
},
{
"epoch": 0.2518676627534685,
"grad_norm": 0.6656386852264404,
"learning_rate": 8.774627921971436e-05,
"loss": 0.1519,
"step": 413
},
{
"epoch": 0.25247751181582556,
"grad_norm": 0.5977262258529663,
"learning_rate": 8.768141660926375e-05,
"loss": 0.1361,
"step": 414
},
{
"epoch": 0.2530873608781827,
"grad_norm": 0.4966430068016052,
"learning_rate": 8.761640689207123e-05,
"loss": 0.1119,
"step": 415
},
{
"epoch": 0.25369720994053974,
"grad_norm": 0.4720788598060608,
"learning_rate": 8.755125032193214e-05,
"loss": 0.1397,
"step": 416
},
{
"epoch": 0.2543070590028968,
"grad_norm": 0.6315327286720276,
"learning_rate": 8.748594715321512e-05,
"loss": 0.1411,
"step": 417
},
{
"epoch": 0.25491690806525386,
"grad_norm": 0.455010324716568,
"learning_rate": 8.742049764086114e-05,
"loss": 0.1257,
"step": 418
},
{
"epoch": 0.2555267571276109,
"grad_norm": 0.40056300163269043,
"learning_rate": 8.735490204038243e-05,
"loss": 0.0902,
"step": 419
},
{
"epoch": 0.256136606189968,
"grad_norm": 0.4765312075614929,
"learning_rate": 8.728916060786162e-05,
"loss": 0.1275,
"step": 420
},
{
"epoch": 0.25674645525232503,
"grad_norm": 0.45411917567253113,
"learning_rate": 8.722327359995064e-05,
"loss": 0.1227,
"step": 421
},
{
"epoch": 0.2573563043146821,
"grad_norm": 0.47428813576698303,
"learning_rate": 8.715724127386972e-05,
"loss": 0.1243,
"step": 422
},
{
"epoch": 0.2579661533770392,
"grad_norm": 0.4526023864746094,
"learning_rate": 8.709106388740642e-05,
"loss": 0.1104,
"step": 423
},
{
"epoch": 0.25857600243939627,
"grad_norm": 0.510002851486206,
"learning_rate": 8.702474169891459e-05,
"loss": 0.1187,
"step": 424
},
{
"epoch": 0.25918585150175333,
"grad_norm": 0.5239256620407104,
"learning_rate": 8.695827496731342e-05,
"loss": 0.1068,
"step": 425
},
{
"epoch": 0.2597957005641104,
"grad_norm": 0.8652831315994263,
"learning_rate": 8.689166395208636e-05,
"loss": 0.202,
"step": 426
},
{
"epoch": 0.26040554962646745,
"grad_norm": 0.76209557056427,
"learning_rate": 8.682490891328016e-05,
"loss": 0.1535,
"step": 427
},
{
"epoch": 0.2610153986888245,
"grad_norm": 0.720308244228363,
"learning_rate": 8.675801011150381e-05,
"loss": 0.1587,
"step": 428
},
{
"epoch": 0.26162524775118157,
"grad_norm": 0.5442304611206055,
"learning_rate": 8.669096780792753e-05,
"loss": 0.1303,
"step": 429
},
{
"epoch": 0.2622350968135386,
"grad_norm": 0.5355377793312073,
"learning_rate": 8.662378226428183e-05,
"loss": 0.1551,
"step": 430
},
{
"epoch": 0.26284494587589574,
"grad_norm": 0.641097366809845,
"learning_rate": 8.655645374285637e-05,
"loss": 0.1304,
"step": 431
},
{
"epoch": 0.2634547949382528,
"grad_norm": 0.6802278757095337,
"learning_rate": 8.6488982506499e-05,
"loss": 0.1391,
"step": 432
},
{
"epoch": 0.26406464400060986,
"grad_norm": 0.4821889400482178,
"learning_rate": 8.64213688186147e-05,
"loss": 0.1137,
"step": 433
},
{
"epoch": 0.2646744930629669,
"grad_norm": 0.5178048014640808,
"learning_rate": 8.635361294316464e-05,
"loss": 0.1298,
"step": 434
},
{
"epoch": 0.265284342125324,
"grad_norm": 0.5639777183532715,
"learning_rate": 8.628571514466501e-05,
"loss": 0.1219,
"step": 435
},
{
"epoch": 0.26589419118768104,
"grad_norm": 0.3960331380367279,
"learning_rate": 8.621767568818613e-05,
"loss": 0.1092,
"step": 436
},
{
"epoch": 0.2665040402500381,
"grad_norm": 0.6051567196846008,
"learning_rate": 8.61494948393513e-05,
"loss": 0.1385,
"step": 437
},
{
"epoch": 0.26711388931239516,
"grad_norm": 0.6859558820724487,
"learning_rate": 8.608117286433582e-05,
"loss": 0.1392,
"step": 438
},
{
"epoch": 0.2677237383747523,
"grad_norm": 0.39182791113853455,
"learning_rate": 8.601271002986595e-05,
"loss": 0.1148,
"step": 439
},
{
"epoch": 0.26833358743710933,
"grad_norm": 0.452949196100235,
"learning_rate": 8.594410660321786e-05,
"loss": 0.1117,
"step": 440
},
{
"epoch": 0.2689434364994664,
"grad_norm": 0.4267730414867401,
"learning_rate": 8.587536285221656e-05,
"loss": 0.1075,
"step": 441
},
{
"epoch": 0.26955328556182345,
"grad_norm": 0.4756946265697479,
"learning_rate": 8.580647904523493e-05,
"loss": 0.1185,
"step": 442
},
{
"epoch": 0.2701631346241805,
"grad_norm": 0.4810199439525604,
"learning_rate": 8.573745545119257e-05,
"loss": 0.1224,
"step": 443
},
{
"epoch": 0.27077298368653757,
"grad_norm": 0.4866829216480255,
"learning_rate": 8.566829233955485e-05,
"loss": 0.1401,
"step": 444
},
{
"epoch": 0.27138283274889463,
"grad_norm": 0.4917653203010559,
"learning_rate": 8.559898998033178e-05,
"loss": 0.1034,
"step": 445
},
{
"epoch": 0.2719926818112517,
"grad_norm": 0.6910008788108826,
"learning_rate": 8.552954864407699e-05,
"loss": 0.1644,
"step": 446
},
{
"epoch": 0.2726025308736088,
"grad_norm": 0.7964531183242798,
"learning_rate": 8.545996860188668e-05,
"loss": 0.1794,
"step": 447
},
{
"epoch": 0.27321237993596587,
"grad_norm": 0.6004226207733154,
"learning_rate": 8.539025012539854e-05,
"loss": 0.145,
"step": 448
},
{
"epoch": 0.2738222289983229,
"grad_norm": 0.6408104300498962,
"learning_rate": 8.532039348679073e-05,
"loss": 0.1274,
"step": 449
},
{
"epoch": 0.27443207806068,
"grad_norm": 0.6084713339805603,
"learning_rate": 8.525039895878078e-05,
"loss": 0.1531,
"step": 450
},
{
"epoch": 0.27504192712303704,
"grad_norm": 0.6172600984573364,
"learning_rate": 8.518026681462448e-05,
"loss": 0.1108,
"step": 451
},
{
"epoch": 0.2756517761853941,
"grad_norm": 0.41464439034461975,
"learning_rate": 8.510999732811495e-05,
"loss": 0.122,
"step": 452
},
{
"epoch": 0.27626162524775116,
"grad_norm": 0.4926139712333679,
"learning_rate": 8.503959077358143e-05,
"loss": 0.0989,
"step": 453
},
{
"epoch": 0.2768714743101082,
"grad_norm": 0.5940062999725342,
"learning_rate": 8.496904742588832e-05,
"loss": 0.119,
"step": 454
},
{
"epoch": 0.27748132337246534,
"grad_norm": 0.6266347765922546,
"learning_rate": 8.4898367560434e-05,
"loss": 0.1414,
"step": 455
},
{
"epoch": 0.2780911724348224,
"grad_norm": 0.6943479180335999,
"learning_rate": 8.482755145314986e-05,
"loss": 0.1586,
"step": 456
},
{
"epoch": 0.27870102149717946,
"grad_norm": 0.588623583316803,
"learning_rate": 8.475659938049912e-05,
"loss": 0.1394,
"step": 457
},
{
"epoch": 0.2793108705595365,
"grad_norm": 0.4774170219898224,
"learning_rate": 8.468551161947584e-05,
"loss": 0.1057,
"step": 458
},
{
"epoch": 0.2799207196218936,
"grad_norm": 0.40607884526252747,
"learning_rate": 8.46142884476038e-05,
"loss": 0.1197,
"step": 459
},
{
"epoch": 0.28053056868425064,
"grad_norm": 0.5247285962104797,
"learning_rate": 8.454293014293539e-05,
"loss": 0.1151,
"step": 460
},
{
"epoch": 0.2811404177466077,
"grad_norm": 0.4830247163772583,
"learning_rate": 8.44714369840506e-05,
"loss": 0.1138,
"step": 461
},
{
"epoch": 0.28175026680896476,
"grad_norm": 0.6262061595916748,
"learning_rate": 8.439980925005586e-05,
"loss": 0.1522,
"step": 462
},
{
"epoch": 0.28236011587132187,
"grad_norm": 0.3681347072124481,
"learning_rate": 8.432804722058296e-05,
"loss": 0.0883,
"step": 463
},
{
"epoch": 0.28296996493367893,
"grad_norm": 0.5847671031951904,
"learning_rate": 8.4256151175788e-05,
"loss": 0.1164,
"step": 464
},
{
"epoch": 0.283579813996036,
"grad_norm": 0.479976087808609,
"learning_rate": 8.418412139635025e-05,
"loss": 0.1031,
"step": 465
},
{
"epoch": 0.28418966305839305,
"grad_norm": 0.5320427417755127,
"learning_rate": 8.41119581634711e-05,
"loss": 0.1323,
"step": 466
},
{
"epoch": 0.2847995121207501,
"grad_norm": 0.4357379376888275,
"learning_rate": 8.403966175887292e-05,
"loss": 0.1172,
"step": 467
},
{
"epoch": 0.28540936118310717,
"grad_norm": 0.42172175645828247,
"learning_rate": 8.396723246479797e-05,
"loss": 0.0888,
"step": 468
},
{
"epoch": 0.28601921024546423,
"grad_norm": 0.3643954396247864,
"learning_rate": 8.389467056400732e-05,
"loss": 0.0842,
"step": 469
},
{
"epoch": 0.2866290593078213,
"grad_norm": 0.4871020019054413,
"learning_rate": 8.382197633977973e-05,
"loss": 0.1291,
"step": 470
},
{
"epoch": 0.2872389083701784,
"grad_norm": 0.6621859669685364,
"learning_rate": 8.374915007591053e-05,
"loss": 0.1451,
"step": 471
},
{
"epoch": 0.28784875743253546,
"grad_norm": 0.4751411974430084,
"learning_rate": 8.367619205671054e-05,
"loss": 0.0996,
"step": 472
},
{
"epoch": 0.2884586064948925,
"grad_norm": 0.6847923398017883,
"learning_rate": 8.360310256700497e-05,
"loss": 0.1369,
"step": 473
},
{
"epoch": 0.2890684555572496,
"grad_norm": 0.6670457720756531,
"learning_rate": 8.352988189213222e-05,
"loss": 0.1638,
"step": 474
},
{
"epoch": 0.28967830461960664,
"grad_norm": 0.7312725186347961,
"learning_rate": 8.345653031794292e-05,
"loss": 0.1401,
"step": 475
},
{
"epoch": 0.2902881536819637,
"grad_norm": 0.7312857508659363,
"learning_rate": 8.338304813079865e-05,
"loss": 0.167,
"step": 476
},
{
"epoch": 0.29089800274432076,
"grad_norm": 0.5596891045570374,
"learning_rate": 8.330943561757092e-05,
"loss": 0.1242,
"step": 477
},
{
"epoch": 0.2915078518066778,
"grad_norm": 0.782714307308197,
"learning_rate": 8.323569306564005e-05,
"loss": 0.1722,
"step": 478
},
{
"epoch": 0.29211770086903494,
"grad_norm": 0.5147573351860046,
"learning_rate": 8.316182076289401e-05,
"loss": 0.1155,
"step": 479
},
{
"epoch": 0.292727549931392,
"grad_norm": 0.5077108144760132,
"learning_rate": 8.30878189977273e-05,
"loss": 0.149,
"step": 480
},
{
"epoch": 0.29333739899374905,
"grad_norm": 0.541913628578186,
"learning_rate": 8.301368805903988e-05,
"loss": 0.1302,
"step": 481
},
{
"epoch": 0.2939472480561061,
"grad_norm": 0.540307343006134,
"learning_rate": 8.293942823623593e-05,
"loss": 0.1059,
"step": 482
},
{
"epoch": 0.2945570971184632,
"grad_norm": 0.5217983722686768,
"learning_rate": 8.286503981922283e-05,
"loss": 0.1016,
"step": 483
},
{
"epoch": 0.29516694618082023,
"grad_norm": 0.44219234585762024,
"learning_rate": 8.279052309841e-05,
"loss": 0.0936,
"step": 484
},
{
"epoch": 0.2957767952431773,
"grad_norm": 0.5337596535682678,
"learning_rate": 8.271587836470775e-05,
"loss": 0.1285,
"step": 485
},
{
"epoch": 0.2963866443055344,
"grad_norm": 0.6421526074409485,
"learning_rate": 8.264110590952609e-05,
"loss": 0.1435,
"step": 486
},
{
"epoch": 0.29699649336789147,
"grad_norm": 0.4665091931819916,
"learning_rate": 8.256620602477372e-05,
"loss": 0.1102,
"step": 487
},
{
"epoch": 0.2976063424302485,
"grad_norm": 0.61270672082901,
"learning_rate": 8.249117900285676e-05,
"loss": 0.1102,
"step": 488
},
{
"epoch": 0.2982161914926056,
"grad_norm": 0.49927446246147156,
"learning_rate": 8.241602513667774e-05,
"loss": 0.1193,
"step": 489
},
{
"epoch": 0.29882604055496265,
"grad_norm": 0.525644063949585,
"learning_rate": 8.234074471963434e-05,
"loss": 0.108,
"step": 490
},
{
"epoch": 0.2994358896173197,
"grad_norm": 0.7441113591194153,
"learning_rate": 8.226533804561827e-05,
"loss": 0.1816,
"step": 491
},
{
"epoch": 0.30004573867967677,
"grad_norm": 0.5911127328872681,
"learning_rate": 8.218980540901418e-05,
"loss": 0.1166,
"step": 492
},
{
"epoch": 0.3006555877420338,
"grad_norm": 0.6734939217567444,
"learning_rate": 8.211414710469845e-05,
"loss": 0.1127,
"step": 493
},
{
"epoch": 0.30126543680439094,
"grad_norm": 0.6848216652870178,
"learning_rate": 8.203836342803807e-05,
"loss": 0.1361,
"step": 494
},
{
"epoch": 0.301875285866748,
"grad_norm": 0.5680652856826782,
"learning_rate": 8.19624546748895e-05,
"loss": 0.1269,
"step": 495
},
{
"epoch": 0.30248513492910506,
"grad_norm": 0.5631565451622009,
"learning_rate": 8.188642114159747e-05,
"loss": 0.1341,
"step": 496
},
{
"epoch": 0.3030949839914621,
"grad_norm": 0.5553311109542847,
"learning_rate": 8.181026312499383e-05,
"loss": 0.137,
"step": 497
},
{
"epoch": 0.3037048330538192,
"grad_norm": 0.8692022562026978,
"learning_rate": 8.173398092239646e-05,
"loss": 0.1186,
"step": 498
},
{
"epoch": 0.30431468211617624,
"grad_norm": 0.6839601397514343,
"learning_rate": 8.165757483160798e-05,
"loss": 0.1553,
"step": 499
},
{
"epoch": 0.3049245311785333,
"grad_norm": 0.9205794334411621,
"learning_rate": 8.158104515091476e-05,
"loss": 0.167,
"step": 500
},
{
"epoch": 0.30553438024089036,
"grad_norm": 0.6153144240379333,
"learning_rate": 8.150439217908556e-05,
"loss": 0.1266,
"step": 501
},
{
"epoch": 0.3061442293032475,
"grad_norm": 0.47289010882377625,
"learning_rate": 8.142761621537055e-05,
"loss": 0.124,
"step": 502
},
{
"epoch": 0.30675407836560453,
"grad_norm": 0.5337610244750977,
"learning_rate": 8.13507175595e-05,
"loss": 0.1154,
"step": 503
},
{
"epoch": 0.3073639274279616,
"grad_norm": 0.5720561146736145,
"learning_rate": 8.12736965116832e-05,
"loss": 0.1393,
"step": 504
},
{
"epoch": 0.30797377649031865,
"grad_norm": 0.552126944065094,
"learning_rate": 8.11965533726072e-05,
"loss": 0.1404,
"step": 505
},
{
"epoch": 0.3085836255526757,
"grad_norm": 0.47788533568382263,
"learning_rate": 8.11192884434358e-05,
"loss": 0.1193,
"step": 506
},
{
"epoch": 0.30919347461503277,
"grad_norm": 0.468497097492218,
"learning_rate": 8.104190202580812e-05,
"loss": 0.1405,
"step": 507
},
{
"epoch": 0.30980332367738983,
"grad_norm": 0.5368284583091736,
"learning_rate": 8.096439442183766e-05,
"loss": 0.1474,
"step": 508
},
{
"epoch": 0.3104131727397469,
"grad_norm": 0.38195329904556274,
"learning_rate": 8.0886765934111e-05,
"loss": 0.1005,
"step": 509
},
{
"epoch": 0.311023021802104,
"grad_norm": 0.35298991203308105,
"learning_rate": 8.080901686568665e-05,
"loss": 0.0953,
"step": 510
},
{
"epoch": 0.31163287086446106,
"grad_norm": 0.5385501980781555,
"learning_rate": 8.073114752009387e-05,
"loss": 0.1364,
"step": 511
},
{
"epoch": 0.3122427199268181,
"grad_norm": 0.3665884733200073,
"learning_rate": 8.065315820133146e-05,
"loss": 0.0914,
"step": 512
},
{
"epoch": 0.3128525689891752,
"grad_norm": 0.5298899412155151,
"learning_rate": 8.05750492138666e-05,
"loss": 0.1102,
"step": 513
},
{
"epoch": 0.31346241805153224,
"grad_norm": 0.49108877778053284,
"learning_rate": 8.049682086263365e-05,
"loss": 0.1392,
"step": 514
},
{
"epoch": 0.3140722671138893,
"grad_norm": 0.5653191804885864,
"learning_rate": 8.041847345303297e-05,
"loss": 0.136,
"step": 515
},
{
"epoch": 0.31468211617624636,
"grad_norm": 0.5892064571380615,
"learning_rate": 8.034000729092968e-05,
"loss": 0.1137,
"step": 516
},
{
"epoch": 0.3152919652386034,
"grad_norm": 0.7107273936271667,
"learning_rate": 8.026142268265256e-05,
"loss": 0.1272,
"step": 517
},
{
"epoch": 0.31590181430096054,
"grad_norm": 0.46203091740608215,
"learning_rate": 8.018271993499278e-05,
"loss": 0.1152,
"step": 518
},
{
"epoch": 0.3165116633633176,
"grad_norm": 0.5683470368385315,
"learning_rate": 8.01038993552027e-05,
"loss": 0.1125,
"step": 519
},
{
"epoch": 0.31712151242567466,
"grad_norm": 0.40482738614082336,
"learning_rate": 8.00249612509947e-05,
"loss": 0.089,
"step": 520
},
{
"epoch": 0.3177313614880317,
"grad_norm": 0.5423827767372131,
"learning_rate": 7.994590593054001e-05,
"loss": 0.1436,
"step": 521
},
{
"epoch": 0.3183412105503888,
"grad_norm": 0.8095924258232117,
"learning_rate": 7.986673370246742e-05,
"loss": 0.1768,
"step": 522
},
{
"epoch": 0.31895105961274584,
"grad_norm": 0.4192773401737213,
"learning_rate": 7.978744487586214e-05,
"loss": 0.1019,
"step": 523
},
{
"epoch": 0.3195609086751029,
"grad_norm": 0.5442997813224792,
"learning_rate": 7.970803976026457e-05,
"loss": 0.1274,
"step": 524
},
{
"epoch": 0.32017075773745995,
"grad_norm": 0.9472556710243225,
"learning_rate": 7.962851866566912e-05,
"loss": 0.1862,
"step": 525
},
{
"epoch": 0.32078060679981707,
"grad_norm": 0.6008622646331787,
"learning_rate": 7.954888190252292e-05,
"loss": 0.1281,
"step": 526
},
{
"epoch": 0.32139045586217413,
"grad_norm": 0.5782567262649536,
"learning_rate": 7.946912978172474e-05,
"loss": 0.1107,
"step": 527
},
{
"epoch": 0.3220003049245312,
"grad_norm": 0.4934207797050476,
"learning_rate": 7.938926261462366e-05,
"loss": 0.1085,
"step": 528
},
{
"epoch": 0.32261015398688825,
"grad_norm": 0.5315726399421692,
"learning_rate": 7.93092807130179e-05,
"loss": 0.1533,
"step": 529
},
{
"epoch": 0.3232200030492453,
"grad_norm": 0.555449903011322,
"learning_rate": 7.922918438915361e-05,
"loss": 0.1172,
"step": 530
},
{
"epoch": 0.32382985211160237,
"grad_norm": 0.4190885126590729,
"learning_rate": 7.91489739557236e-05,
"loss": 0.1301,
"step": 531
},
{
"epoch": 0.3244397011739594,
"grad_norm": 0.56365567445755,
"learning_rate": 7.906864972586624e-05,
"loss": 0.1414,
"step": 532
},
{
"epoch": 0.3250495502363165,
"grad_norm": 0.4952602982521057,
"learning_rate": 7.898821201316407e-05,
"loss": 0.1112,
"step": 533
},
{
"epoch": 0.3256593992986736,
"grad_norm": 0.42723169922828674,
"learning_rate": 7.890766113164271e-05,
"loss": 0.111,
"step": 534
},
{
"epoch": 0.32626924836103066,
"grad_norm": 0.4798721969127655,
"learning_rate": 7.882699739576959e-05,
"loss": 0.1118,
"step": 535
},
{
"epoch": 0.3268790974233877,
"grad_norm": 0.4729415476322174,
"learning_rate": 7.874622112045269e-05,
"loss": 0.1496,
"step": 536
},
{
"epoch": 0.3274889464857448,
"grad_norm": 0.5551421046257019,
"learning_rate": 7.866533262103936e-05,
"loss": 0.1237,
"step": 537
},
{
"epoch": 0.32809879554810184,
"grad_norm": 0.5174462199211121,
"learning_rate": 7.858433221331508e-05,
"loss": 0.1176,
"step": 538
},
{
"epoch": 0.3287086446104589,
"grad_norm": 0.5168452858924866,
"learning_rate": 7.850322021350215e-05,
"loss": 0.1424,
"step": 539
},
{
"epoch": 0.32931849367281596,
"grad_norm": 0.429595947265625,
"learning_rate": 7.842199693825862e-05,
"loss": 0.0991,
"step": 540
},
{
"epoch": 0.329928342735173,
"grad_norm": 0.4393986761569977,
"learning_rate": 7.83406627046769e-05,
"loss": 0.1237,
"step": 541
},
{
"epoch": 0.33053819179753013,
"grad_norm": 0.5737545490264893,
"learning_rate": 7.825921783028255e-05,
"loss": 0.1618,
"step": 542
},
{
"epoch": 0.3311480408598872,
"grad_norm": 0.44508734345436096,
"learning_rate": 7.817766263303313e-05,
"loss": 0.1184,
"step": 543
},
{
"epoch": 0.33175788992224425,
"grad_norm": 0.4665411114692688,
"learning_rate": 7.809599743131684e-05,
"loss": 0.1041,
"step": 544
},
{
"epoch": 0.3323677389846013,
"grad_norm": 0.508711576461792,
"learning_rate": 7.801422254395138e-05,
"loss": 0.1097,
"step": 545
},
{
"epoch": 0.3329775880469584,
"grad_norm": 0.8179792165756226,
"learning_rate": 7.793233829018262e-05,
"loss": 0.1873,
"step": 546
},
{
"epoch": 0.33358743710931543,
"grad_norm": 0.3753645420074463,
"learning_rate": 7.785034498968344e-05,
"loss": 0.0761,
"step": 547
},
{
"epoch": 0.3341972861716725,
"grad_norm": 0.8901389837265015,
"learning_rate": 7.776824296255236e-05,
"loss": 0.1377,
"step": 548
},
{
"epoch": 0.33480713523402955,
"grad_norm": 0.5930473804473877,
"learning_rate": 7.768603252931243e-05,
"loss": 0.1216,
"step": 549
},
{
"epoch": 0.33541698429638667,
"grad_norm": 0.6265133619308472,
"learning_rate": 7.760371401090988e-05,
"loss": 0.1429,
"step": 550
},
{
"epoch": 0.3360268333587437,
"grad_norm": 0.5446604490280151,
"learning_rate": 7.752128772871292e-05,
"loss": 0.1084,
"step": 551
},
{
"epoch": 0.3366366824211008,
"grad_norm": 0.6148346066474915,
"learning_rate": 7.743875400451047e-05,
"loss": 0.1339,
"step": 552
},
{
"epoch": 0.33724653148345785,
"grad_norm": 0.5988903641700745,
"learning_rate": 7.735611316051084e-05,
"loss": 0.1319,
"step": 553
},
{
"epoch": 0.3378563805458149,
"grad_norm": 0.46344494819641113,
"learning_rate": 7.727336551934062e-05,
"loss": 0.1019,
"step": 554
},
{
"epoch": 0.33846622960817196,
"grad_norm": 0.5027177333831787,
"learning_rate": 7.719051140404327e-05,
"loss": 0.1357,
"step": 555
},
{
"epoch": 0.339076078670529,
"grad_norm": 0.5424127578735352,
"learning_rate": 7.710755113807794e-05,
"loss": 0.1254,
"step": 556
},
{
"epoch": 0.3396859277328861,
"grad_norm": 0.3771025538444519,
"learning_rate": 7.702448504531819e-05,
"loss": 0.0861,
"step": 557
},
{
"epoch": 0.3402957767952432,
"grad_norm": 0.4751587212085724,
"learning_rate": 7.694131345005071e-05,
"loss": 0.1134,
"step": 558
},
{
"epoch": 0.34090562585760026,
"grad_norm": 0.48964354395866394,
"learning_rate": 7.685803667697411e-05,
"loss": 0.1068,
"step": 559
},
{
"epoch": 0.3415154749199573,
"grad_norm": 0.545700192451477,
"learning_rate": 7.677465505119754e-05,
"loss": 0.1026,
"step": 560
},
{
"epoch": 0.3421253239823144,
"grad_norm": 0.5670740008354187,
"learning_rate": 7.669116889823955e-05,
"loss": 0.1424,
"step": 561
},
{
"epoch": 0.34273517304467144,
"grad_norm": 0.770907998085022,
"learning_rate": 7.660757854402671e-05,
"loss": 0.143,
"step": 562
},
{
"epoch": 0.3433450221070285,
"grad_norm": 1.103381633758545,
"learning_rate": 7.652388431489248e-05,
"loss": 0.145,
"step": 563
},
{
"epoch": 0.34395487116938556,
"grad_norm": 0.5053901076316833,
"learning_rate": 7.644008653757571e-05,
"loss": 0.1088,
"step": 564
},
{
"epoch": 0.3445647202317426,
"grad_norm": 0.5928632020950317,
"learning_rate": 7.635618553921962e-05,
"loss": 0.1304,
"step": 565
},
{
"epoch": 0.34517456929409973,
"grad_norm": 0.4465699791908264,
"learning_rate": 7.627218164737031e-05,
"loss": 0.1341,
"step": 566
},
{
"epoch": 0.3457844183564568,
"grad_norm": 0.580589771270752,
"learning_rate": 7.618807518997563e-05,
"loss": 0.1377,
"step": 567
},
{
"epoch": 0.34639426741881385,
"grad_norm": 0.630957841873169,
"learning_rate": 7.610386649538384e-05,
"loss": 0.1528,
"step": 568
},
{
"epoch": 0.3470041164811709,
"grad_norm": 0.4806077778339386,
"learning_rate": 7.601955589234227e-05,
"loss": 0.1056,
"step": 569
},
{
"epoch": 0.34761396554352797,
"grad_norm": 0.43553707003593445,
"learning_rate": 7.593514370999617e-05,
"loss": 0.0799,
"step": 570
},
{
"epoch": 0.34822381460588503,
"grad_norm": 0.5793433785438538,
"learning_rate": 7.585063027788731e-05,
"loss": 0.1149,
"step": 571
},
{
"epoch": 0.3488336636682421,
"grad_norm": 0.8291622996330261,
"learning_rate": 7.576601592595274e-05,
"loss": 0.1511,
"step": 572
},
{
"epoch": 0.3494435127305992,
"grad_norm": 0.6264448761940002,
"learning_rate": 7.568130098452351e-05,
"loss": 0.1114,
"step": 573
},
{
"epoch": 0.35005336179295626,
"grad_norm": 0.8981962203979492,
"learning_rate": 7.559648578432338e-05,
"loss": 0.1584,
"step": 574
},
{
"epoch": 0.3506632108553133,
"grad_norm": 0.5635583400726318,
"learning_rate": 7.551157065646746e-05,
"loss": 0.1161,
"step": 575
},
{
"epoch": 0.3512730599176704,
"grad_norm": 0.40898945927619934,
"learning_rate": 7.542655593246103e-05,
"loss": 0.0709,
"step": 576
},
{
"epoch": 0.35188290898002744,
"grad_norm": 0.6884272694587708,
"learning_rate": 7.534144194419817e-05,
"loss": 0.1212,
"step": 577
},
{
"epoch": 0.3524927580423845,
"grad_norm": 0.6667736172676086,
"learning_rate": 7.525622902396047e-05,
"loss": 0.1626,
"step": 578
},
{
"epoch": 0.35310260710474156,
"grad_norm": 0.6642476320266724,
"learning_rate": 7.517091750441576e-05,
"loss": 0.1439,
"step": 579
},
{
"epoch": 0.3537124561670986,
"grad_norm": 0.4169001877307892,
"learning_rate": 7.508550771861683e-05,
"loss": 0.0886,
"step": 580
},
{
"epoch": 0.35432230522945574,
"grad_norm": 0.5685600638389587,
"learning_rate": 7.500000000000001e-05,
"loss": 0.1272,
"step": 581
},
{
"epoch": 0.3549321542918128,
"grad_norm": 0.4451376795768738,
"learning_rate": 7.491439468238403e-05,
"loss": 0.1001,
"step": 582
},
{
"epoch": 0.35554200335416986,
"grad_norm": 0.7579962015151978,
"learning_rate": 7.482869209996867e-05,
"loss": 0.1677,
"step": 583
},
{
"epoch": 0.3561518524165269,
"grad_norm": 0.4882470965385437,
"learning_rate": 7.474289258733332e-05,
"loss": 0.1564,
"step": 584
},
{
"epoch": 0.356761701478884,
"grad_norm": 0.4724157750606537,
"learning_rate": 7.465699647943586e-05,
"loss": 0.1362,
"step": 585
},
{
"epoch": 0.35737155054124103,
"grad_norm": 0.489364892244339,
"learning_rate": 7.457100411161128e-05,
"loss": 0.0986,
"step": 586
},
{
"epoch": 0.3579813996035981,
"grad_norm": 0.5231732726097107,
"learning_rate": 7.44849158195703e-05,
"loss": 0.1163,
"step": 587
},
{
"epoch": 0.35859124866595515,
"grad_norm": 0.3738481104373932,
"learning_rate": 7.43987319393982e-05,
"loss": 0.0832,
"step": 588
},
{
"epoch": 0.35920109772831227,
"grad_norm": 0.47332149744033813,
"learning_rate": 7.431245280755336e-05,
"loss": 0.1265,
"step": 589
},
{
"epoch": 0.35981094679066933,
"grad_norm": 0.3875868022441864,
"learning_rate": 7.422607876086612e-05,
"loss": 0.1089,
"step": 590
},
{
"epoch": 0.3604207958530264,
"grad_norm": 0.4316892921924591,
"learning_rate": 7.413961013653726e-05,
"loss": 0.103,
"step": 591
},
{
"epoch": 0.36103064491538345,
"grad_norm": 0.6579156517982483,
"learning_rate": 7.405304727213681e-05,
"loss": 0.1141,
"step": 592
},
{
"epoch": 0.3616404939777405,
"grad_norm": 0.5627062320709229,
"learning_rate": 7.396639050560275e-05,
"loss": 0.1045,
"step": 593
},
{
"epoch": 0.36225034304009757,
"grad_norm": 0.5660955905914307,
"learning_rate": 7.387964017523964e-05,
"loss": 0.1117,
"step": 594
},
{
"epoch": 0.3628601921024546,
"grad_norm": 0.6846521496772766,
"learning_rate": 7.379279661971727e-05,
"loss": 0.0974,
"step": 595
},
{
"epoch": 0.3634700411648117,
"grad_norm": 0.5044648051261902,
"learning_rate": 7.370586017806942e-05,
"loss": 0.1072,
"step": 596
},
{
"epoch": 0.3640798902271688,
"grad_norm": 0.42822015285491943,
"learning_rate": 7.361883118969247e-05,
"loss": 0.1075,
"step": 597
},
{
"epoch": 0.36468973928952586,
"grad_norm": 0.5746731758117676,
"learning_rate": 7.353170999434412e-05,
"loss": 0.1243,
"step": 598
},
{
"epoch": 0.3652995883518829,
"grad_norm": 0.6602866053581238,
"learning_rate": 7.3444496932142e-05,
"loss": 0.1081,
"step": 599
},
{
"epoch": 0.36590943741424,
"grad_norm": 0.6254577040672302,
"learning_rate": 7.335719234356246e-05,
"loss": 0.1071,
"step": 600
},
{
"epoch": 0.36651928647659704,
"grad_norm": 0.5332274436950684,
"learning_rate": 7.326979656943906e-05,
"loss": 0.0993,
"step": 601
},
{
"epoch": 0.3671291355389541,
"grad_norm": 0.5956599712371826,
"learning_rate": 7.318230995096145e-05,
"loss": 0.118,
"step": 602
},
{
"epoch": 0.36773898460131116,
"grad_norm": 0.45792531967163086,
"learning_rate": 7.309473282967387e-05,
"loss": 0.0998,
"step": 603
},
{
"epoch": 0.3683488336636682,
"grad_norm": 0.3908950686454773,
"learning_rate": 7.30070655474739e-05,
"loss": 0.0951,
"step": 604
},
{
"epoch": 0.36895868272602533,
"grad_norm": 0.48248735070228577,
"learning_rate": 7.291930844661109e-05,
"loss": 0.1003,
"step": 605
},
{
"epoch": 0.3695685317883824,
"grad_norm": 0.6083711981773376,
"learning_rate": 7.283146186968565e-05,
"loss": 0.1252,
"step": 606
},
{
"epoch": 0.37017838085073945,
"grad_norm": 0.5875548720359802,
"learning_rate": 7.274352615964712e-05,
"loss": 0.1156,
"step": 607
},
{
"epoch": 0.3707882299130965,
"grad_norm": 0.6106542348861694,
"learning_rate": 7.265550165979297e-05,
"loss": 0.1268,
"step": 608
},
{
"epoch": 0.37139807897545357,
"grad_norm": 0.512162983417511,
"learning_rate": 7.256738871376732e-05,
"loss": 0.104,
"step": 609
},
{
"epoch": 0.37200792803781063,
"grad_norm": 0.5985874533653259,
"learning_rate": 7.24791876655596e-05,
"loss": 0.1373,
"step": 610
},
{
"epoch": 0.3726177771001677,
"grad_norm": 0.6054688692092896,
"learning_rate": 7.239089885950316e-05,
"loss": 0.1272,
"step": 611
},
{
"epoch": 0.37322762616252475,
"grad_norm": 0.5596964955329895,
"learning_rate": 7.230252264027397e-05,
"loss": 0.1047,
"step": 612
},
{
"epoch": 0.37383747522488187,
"grad_norm": 0.5471627116203308,
"learning_rate": 7.221405935288925e-05,
"loss": 0.116,
"step": 613
},
{
"epoch": 0.3744473242872389,
"grad_norm": 0.9812818765640259,
"learning_rate": 7.212550934270613e-05,
"loss": 0.164,
"step": 614
},
{
"epoch": 0.375057173349596,
"grad_norm": 0.5124916434288025,
"learning_rate": 7.203687295542032e-05,
"loss": 0.0891,
"step": 615
},
{
"epoch": 0.37566702241195304,
"grad_norm": 0.5081048011779785,
"learning_rate": 7.19481505370647e-05,
"loss": 0.0939,
"step": 616
},
{
"epoch": 0.3762768714743101,
"grad_norm": 0.42085763812065125,
"learning_rate": 7.185934243400806e-05,
"loss": 0.0763,
"step": 617
},
{
"epoch": 0.37688672053666716,
"grad_norm": 0.6093540191650391,
"learning_rate": 7.177044899295368e-05,
"loss": 0.1339,
"step": 618
},
{
"epoch": 0.3774965695990242,
"grad_norm": 0.5402736663818359,
"learning_rate": 7.168147056093797e-05,
"loss": 0.1101,
"step": 619
},
{
"epoch": 0.3781064186613813,
"grad_norm": 0.4964229464530945,
"learning_rate": 7.15924074853292e-05,
"loss": 0.0945,
"step": 620
},
{
"epoch": 0.3787162677237384,
"grad_norm": 0.510540246963501,
"learning_rate": 7.150326011382604e-05,
"loss": 0.1282,
"step": 621
},
{
"epoch": 0.37932611678609546,
"grad_norm": 0.5624058842658997,
"learning_rate": 7.141402879445624e-05,
"loss": 0.1111,
"step": 622
},
{
"epoch": 0.3799359658484525,
"grad_norm": 0.6607561707496643,
"learning_rate": 7.132471387557532e-05,
"loss": 0.1069,
"step": 623
},
{
"epoch": 0.3805458149108096,
"grad_norm": 0.4943874180316925,
"learning_rate": 7.123531570586514e-05,
"loss": 0.1198,
"step": 624
},
{
"epoch": 0.38115566397316664,
"grad_norm": 0.41025540232658386,
"learning_rate": 7.114583463433259e-05,
"loss": 0.102,
"step": 625
},
{
"epoch": 0.3817655130355237,
"grad_norm": 0.7535240650177002,
"learning_rate": 7.105627101030817e-05,
"loss": 0.1407,
"step": 626
},
{
"epoch": 0.38237536209788076,
"grad_norm": 0.46041688323020935,
"learning_rate": 7.096662518344468e-05,
"loss": 0.1037,
"step": 627
},
{
"epoch": 0.3829852111602378,
"grad_norm": 0.5211940407752991,
"learning_rate": 7.087689750371587e-05,
"loss": 0.1018,
"step": 628
},
{
"epoch": 0.38359506022259493,
"grad_norm": 0.4839499294757843,
"learning_rate": 7.078708832141497e-05,
"loss": 0.1031,
"step": 629
},
{
"epoch": 0.384204909284952,
"grad_norm": 0.6028909087181091,
"learning_rate": 7.069719798715346e-05,
"loss": 0.1424,
"step": 630
},
{
"epoch": 0.38481475834730905,
"grad_norm": 0.5010972619056702,
"learning_rate": 7.060722685185961e-05,
"loss": 0.1031,
"step": 631
},
{
"epoch": 0.3854246074096661,
"grad_norm": 0.570289671421051,
"learning_rate": 7.051717526677711e-05,
"loss": 0.1104,
"step": 632
},
{
"epoch": 0.38603445647202317,
"grad_norm": 0.6728094220161438,
"learning_rate": 7.042704358346375e-05,
"loss": 0.1625,
"step": 633
},
{
"epoch": 0.38664430553438023,
"grad_norm": 0.4294874966144562,
"learning_rate": 7.033683215379002e-05,
"loss": 0.1039,
"step": 634
},
{
"epoch": 0.3872541545967373,
"grad_norm": 0.4257156550884247,
"learning_rate": 7.024654132993772e-05,
"loss": 0.0983,
"step": 635
},
{
"epoch": 0.38786400365909435,
"grad_norm": 0.5675230026245117,
"learning_rate": 7.015617146439863e-05,
"loss": 0.1298,
"step": 636
},
{
"epoch": 0.38847385272145146,
"grad_norm": 0.6300134658813477,
"learning_rate": 7.006572290997304e-05,
"loss": 0.1472,
"step": 637
},
{
"epoch": 0.3890837017838085,
"grad_norm": 0.8972352147102356,
"learning_rate": 6.997519601976855e-05,
"loss": 0.2066,
"step": 638
},
{
"epoch": 0.3896935508461656,
"grad_norm": 0.7109792232513428,
"learning_rate": 6.988459114719849e-05,
"loss": 0.1199,
"step": 639
},
{
"epoch": 0.39030339990852264,
"grad_norm": 0.6873154044151306,
"learning_rate": 6.979390864598062e-05,
"loss": 0.1567,
"step": 640
},
{
"epoch": 0.3909132489708797,
"grad_norm": 0.5286821126937866,
"learning_rate": 6.970314887013584e-05,
"loss": 0.108,
"step": 641
},
{
"epoch": 0.39152309803323676,
"grad_norm": 0.5313868522644043,
"learning_rate": 6.961231217398668e-05,
"loss": 0.1375,
"step": 642
},
{
"epoch": 0.3921329470955938,
"grad_norm": 0.3940941393375397,
"learning_rate": 6.952139891215593e-05,
"loss": 0.0881,
"step": 643
},
{
"epoch": 0.3927427961579509,
"grad_norm": 0.4487074017524719,
"learning_rate": 6.943040943956535e-05,
"loss": 0.1095,
"step": 644
},
{
"epoch": 0.393352645220308,
"grad_norm": 0.6489326357841492,
"learning_rate": 6.93393441114342e-05,
"loss": 0.1263,
"step": 645
},
{
"epoch": 0.39396249428266505,
"grad_norm": 0.4232451915740967,
"learning_rate": 6.924820328327786e-05,
"loss": 0.1216,
"step": 646
},
{
"epoch": 0.3945723433450221,
"grad_norm": 0.5296417474746704,
"learning_rate": 6.915698731090648e-05,
"loss": 0.1085,
"step": 647
},
{
"epoch": 0.3951821924073792,
"grad_norm": 0.4138530492782593,
"learning_rate": 6.906569655042357e-05,
"loss": 0.1185,
"step": 648
},
{
"epoch": 0.39579204146973623,
"grad_norm": 0.48320767283439636,
"learning_rate": 6.897433135822461e-05,
"loss": 0.1392,
"step": 649
},
{
"epoch": 0.3964018905320933,
"grad_norm": 0.4069627523422241,
"learning_rate": 6.888289209099565e-05,
"loss": 0.0906,
"step": 650
},
{
"epoch": 0.39701173959445035,
"grad_norm": 0.5048792958259583,
"learning_rate": 6.879137910571191e-05,
"loss": 0.1389,
"step": 651
},
{
"epoch": 0.39762158865680747,
"grad_norm": 0.575352668762207,
"learning_rate": 6.869979275963643e-05,
"loss": 0.1306,
"step": 652
},
{
"epoch": 0.3982314377191645,
"grad_norm": 0.4008471369743347,
"learning_rate": 6.860813341031866e-05,
"loss": 0.106,
"step": 653
},
{
"epoch": 0.3988412867815216,
"grad_norm": 0.3759486675262451,
"learning_rate": 6.8516401415593e-05,
"loss": 0.1009,
"step": 654
},
{
"epoch": 0.39945113584387865,
"grad_norm": 0.41545742750167847,
"learning_rate": 6.842459713357752e-05,
"loss": 0.0786,
"step": 655
},
{
"epoch": 0.4000609849062357,
"grad_norm": 0.5301156640052795,
"learning_rate": 6.833272092267241e-05,
"loss": 0.0994,
"step": 656
},
{
"epoch": 0.40067083396859277,
"grad_norm": 0.4906249940395355,
"learning_rate": 6.824077314155877e-05,
"loss": 0.0975,
"step": 657
},
{
"epoch": 0.4012806830309498,
"grad_norm": 0.6189441680908203,
"learning_rate": 6.814875414919701e-05,
"loss": 0.1302,
"step": 658
},
{
"epoch": 0.4018905320933069,
"grad_norm": 0.613351583480835,
"learning_rate": 6.805666430482564e-05,
"loss": 0.1155,
"step": 659
},
{
"epoch": 0.402500381155664,
"grad_norm": 0.3712262213230133,
"learning_rate": 6.79645039679597e-05,
"loss": 0.0808,
"step": 660
},
{
"epoch": 0.40311023021802106,
"grad_norm": 0.5155539512634277,
"learning_rate": 6.787227349838947e-05,
"loss": 0.1137,
"step": 661
},
{
"epoch": 0.4037200792803781,
"grad_norm": 0.3875385522842407,
"learning_rate": 6.777997325617898e-05,
"loss": 0.0665,
"step": 662
},
{
"epoch": 0.4043299283427352,
"grad_norm": 0.5062624216079712,
"learning_rate": 6.768760360166471e-05,
"loss": 0.0847,
"step": 663
},
{
"epoch": 0.40493977740509224,
"grad_norm": 0.4611454904079437,
"learning_rate": 6.759516489545408e-05,
"loss": 0.1039,
"step": 664
},
{
"epoch": 0.4055496264674493,
"grad_norm": 0.8415635824203491,
"learning_rate": 6.750265749842409e-05,
"loss": 0.1326,
"step": 665
},
{
"epoch": 0.40615947552980636,
"grad_norm": 0.579078197479248,
"learning_rate": 6.741008177171995e-05,
"loss": 0.1141,
"step": 666
},
{
"epoch": 0.4067693245921634,
"grad_norm": 0.5113241672515869,
"learning_rate": 6.731743807675355e-05,
"loss": 0.0902,
"step": 667
},
{
"epoch": 0.40737917365452053,
"grad_norm": 0.7772948145866394,
"learning_rate": 6.722472677520219e-05,
"loss": 0.1451,
"step": 668
},
{
"epoch": 0.4079890227168776,
"grad_norm": 0.5634645819664001,
"learning_rate": 6.713194822900706e-05,
"loss": 0.1096,
"step": 669
},
{
"epoch": 0.40859887177923465,
"grad_norm": 0.6254982352256775,
"learning_rate": 6.703910280037193e-05,
"loss": 0.1058,
"step": 670
},
{
"epoch": 0.4092087208415917,
"grad_norm": 0.49209198355674744,
"learning_rate": 6.694619085176159e-05,
"loss": 0.1101,
"step": 671
},
{
"epoch": 0.40981856990394877,
"grad_norm": 0.651881217956543,
"learning_rate": 6.685321274590059e-05,
"loss": 0.0931,
"step": 672
},
{
"epoch": 0.41042841896630583,
"grad_norm": 0.4484933912754059,
"learning_rate": 6.676016884577173e-05,
"loss": 0.0871,
"step": 673
},
{
"epoch": 0.4110382680286629,
"grad_norm": 0.5280038714408875,
"learning_rate": 6.666705951461468e-05,
"loss": 0.1056,
"step": 674
},
{
"epoch": 0.41164811709101995,
"grad_norm": 0.47604578733444214,
"learning_rate": 6.657388511592452e-05,
"loss": 0.095,
"step": 675
},
{
"epoch": 0.41225796615337706,
"grad_norm": 0.45174434781074524,
"learning_rate": 6.64806460134504e-05,
"loss": 0.0985,
"step": 676
},
{
"epoch": 0.4128678152157341,
"grad_norm": 0.7718876004219055,
"learning_rate": 6.638734257119401e-05,
"loss": 0.1397,
"step": 677
},
{
"epoch": 0.4134776642780912,
"grad_norm": 0.6108860373497009,
"learning_rate": 6.62939751534083e-05,
"loss": 0.1041,
"step": 678
},
{
"epoch": 0.41408751334044824,
"grad_norm": 0.4118732511997223,
"learning_rate": 6.620054412459588e-05,
"loss": 0.1009,
"step": 679
},
{
"epoch": 0.4146973624028053,
"grad_norm": 0.6372990012168884,
"learning_rate": 6.610704984950778e-05,
"loss": 0.1101,
"step": 680
},
{
"epoch": 0.41530721146516236,
"grad_norm": 0.582733690738678,
"learning_rate": 6.601349269314188e-05,
"loss": 0.0966,
"step": 681
},
{
"epoch": 0.4159170605275194,
"grad_norm": 0.515181303024292,
"learning_rate": 6.591987302074161e-05,
"loss": 0.0992,
"step": 682
},
{
"epoch": 0.4165269095898765,
"grad_norm": 0.6335199475288391,
"learning_rate": 6.582619119779439e-05,
"loss": 0.1333,
"step": 683
},
{
"epoch": 0.4171367586522336,
"grad_norm": 0.5660648345947266,
"learning_rate": 6.573244759003032e-05,
"loss": 0.1096,
"step": 684
},
{
"epoch": 0.41774660771459066,
"grad_norm": 0.7536609768867493,
"learning_rate": 6.56386425634207e-05,
"loss": 0.1099,
"step": 685
},
{
"epoch": 0.4183564567769477,
"grad_norm": 0.7920714020729065,
"learning_rate": 6.554477648417657e-05,
"loss": 0.1488,
"step": 686
},
{
"epoch": 0.4189663058393048,
"grad_norm": 0.5822864174842834,
"learning_rate": 6.545084971874738e-05,
"loss": 0.0938,
"step": 687
},
{
"epoch": 0.41957615490166184,
"grad_norm": 0.6719368696212769,
"learning_rate": 6.535686263381944e-05,
"loss": 0.1268,
"step": 688
},
{
"epoch": 0.4201860039640189,
"grad_norm": 0.5888537764549255,
"learning_rate": 6.526281559631458e-05,
"loss": 0.1116,
"step": 689
},
{
"epoch": 0.42079585302637595,
"grad_norm": 0.456779807806015,
"learning_rate": 6.516870897338863e-05,
"loss": 0.1008,
"step": 690
},
{
"epoch": 0.421405702088733,
"grad_norm": 0.46741586923599243,
"learning_rate": 6.507454313243015e-05,
"loss": 0.119,
"step": 691
},
{
"epoch": 0.42201555115109013,
"grad_norm": 0.45705685019493103,
"learning_rate": 6.498031844105876e-05,
"loss": 0.0862,
"step": 692
},
{
"epoch": 0.4226254002134472,
"grad_norm": 0.49153685569763184,
"learning_rate": 6.48860352671239e-05,
"loss": 0.1015,
"step": 693
},
{
"epoch": 0.42323524927580425,
"grad_norm": 0.5189732909202576,
"learning_rate": 6.47916939787033e-05,
"loss": 0.1098,
"step": 694
},
{
"epoch": 0.4238450983381613,
"grad_norm": 0.5602004528045654,
"learning_rate": 6.469729494410158e-05,
"loss": 0.1126,
"step": 695
},
{
"epoch": 0.42445494740051837,
"grad_norm": 0.7254711389541626,
"learning_rate": 6.460283853184879e-05,
"loss": 0.1059,
"step": 696
},
{
"epoch": 0.4250647964628754,
"grad_norm": 0.563317596912384,
"learning_rate": 6.450832511069897e-05,
"loss": 0.1252,
"step": 697
},
{
"epoch": 0.4256746455252325,
"grad_norm": 0.5111974477767944,
"learning_rate": 6.441375504962877e-05,
"loss": 0.1213,
"step": 698
},
{
"epoch": 0.42628449458758955,
"grad_norm": 0.6136337518692017,
"learning_rate": 6.431912871783586e-05,
"loss": 0.1554,
"step": 699
},
{
"epoch": 0.42689434364994666,
"grad_norm": 0.5464217066764832,
"learning_rate": 6.42244464847377e-05,
"loss": 0.1206,
"step": 700
},
{
"epoch": 0.4275041927123037,
"grad_norm": 0.41383665800094604,
"learning_rate": 6.412970871996995e-05,
"loss": 0.0889,
"step": 701
},
{
"epoch": 0.4281140417746608,
"grad_norm": 0.8681038022041321,
"learning_rate": 6.403491579338499e-05,
"loss": 0.116,
"step": 702
},
{
"epoch": 0.42872389083701784,
"grad_norm": 0.4806249439716339,
"learning_rate": 6.394006807505067e-05,
"loss": 0.0949,
"step": 703
},
{
"epoch": 0.4293337398993749,
"grad_norm": 0.35108646750450134,
"learning_rate": 6.384516593524865e-05,
"loss": 0.0661,
"step": 704
},
{
"epoch": 0.42994358896173196,
"grad_norm": 0.5926347374916077,
"learning_rate": 6.37502097444731e-05,
"loss": 0.1182,
"step": 705
},
{
"epoch": 0.430553438024089,
"grad_norm": 0.5471864938735962,
"learning_rate": 6.365519987342917e-05,
"loss": 0.1084,
"step": 706
},
{
"epoch": 0.4311632870864461,
"grad_norm": 0.5175963640213013,
"learning_rate": 6.356013669303162e-05,
"loss": 0.1059,
"step": 707
},
{
"epoch": 0.4317731361488032,
"grad_norm": 0.4707430899143219,
"learning_rate": 6.346502057440327e-05,
"loss": 0.1122,
"step": 708
},
{
"epoch": 0.43238298521116025,
"grad_norm": 0.5045830607414246,
"learning_rate": 6.336985188887366e-05,
"loss": 0.0827,
"step": 709
},
{
"epoch": 0.4329928342735173,
"grad_norm": 0.5156387686729431,
"learning_rate": 6.327463100797755e-05,
"loss": 0.1035,
"step": 710
},
{
"epoch": 0.4336026833358744,
"grad_norm": 0.4540978968143463,
"learning_rate": 6.317935830345338e-05,
"loss": 0.1015,
"step": 711
},
{
"epoch": 0.43421253239823143,
"grad_norm": 0.4323015809059143,
"learning_rate": 6.308403414724205e-05,
"loss": 0.0935,
"step": 712
},
{
"epoch": 0.4348223814605885,
"grad_norm": 0.6293617486953735,
"learning_rate": 6.298865891148518e-05,
"loss": 0.1208,
"step": 713
},
{
"epoch": 0.43543223052294555,
"grad_norm": 0.5659090876579285,
"learning_rate": 6.289323296852393e-05,
"loss": 0.0989,
"step": 714
},
{
"epoch": 0.4360420795853026,
"grad_norm": 0.45548877120018005,
"learning_rate": 6.279775669089733e-05,
"loss": 0.1088,
"step": 715
},
{
"epoch": 0.4366519286476597,
"grad_norm": 0.4256194531917572,
"learning_rate": 6.270223045134096e-05,
"loss": 0.096,
"step": 716
},
{
"epoch": 0.4372617777100168,
"grad_norm": 0.44176793098449707,
"learning_rate": 6.260665462278544e-05,
"loss": 0.0719,
"step": 717
},
{
"epoch": 0.43787162677237385,
"grad_norm": 0.741546630859375,
"learning_rate": 6.251102957835493e-05,
"loss": 0.1138,
"step": 718
},
{
"epoch": 0.4384814758347309,
"grad_norm": 0.4579002261161804,
"learning_rate": 6.241535569136584e-05,
"loss": 0.1121,
"step": 719
},
{
"epoch": 0.43909132489708796,
"grad_norm": 0.5200890302658081,
"learning_rate": 6.231963333532515e-05,
"loss": 0.1021,
"step": 720
},
{
"epoch": 0.439701173959445,
"grad_norm": 0.5439337491989136,
"learning_rate": 6.222386288392913e-05,
"loss": 0.1037,
"step": 721
},
{
"epoch": 0.4403110230218021,
"grad_norm": 0.6618098020553589,
"learning_rate": 6.212804471106178e-05,
"loss": 0.1097,
"step": 722
},
{
"epoch": 0.44092087208415914,
"grad_norm": 0.6067342758178711,
"learning_rate": 6.203217919079342e-05,
"loss": 0.123,
"step": 723
},
{
"epoch": 0.44153072114651626,
"grad_norm": 0.5829437375068665,
"learning_rate": 6.193626669737919e-05,
"loss": 0.1242,
"step": 724
},
{
"epoch": 0.4421405702088733,
"grad_norm": 0.5808911919593811,
"learning_rate": 6.184030760525764e-05,
"loss": 0.1141,
"step": 725
},
{
"epoch": 0.4427504192712304,
"grad_norm": 0.5883690118789673,
"learning_rate": 6.174430228904919e-05,
"loss": 0.1297,
"step": 726
},
{
"epoch": 0.44336026833358744,
"grad_norm": 0.9864974617958069,
"learning_rate": 6.164825112355477e-05,
"loss": 0.1127,
"step": 727
},
{
"epoch": 0.4439701173959445,
"grad_norm": 0.5406239032745361,
"learning_rate": 6.155215448375432e-05,
"loss": 0.1078,
"step": 728
},
{
"epoch": 0.44457996645830156,
"grad_norm": 0.6124576926231384,
"learning_rate": 6.145601274480521e-05,
"loss": 0.1341,
"step": 729
},
{
"epoch": 0.4451898155206586,
"grad_norm": 0.5239964127540588,
"learning_rate": 6.1359826282041e-05,
"loss": 0.1086,
"step": 730
},
{
"epoch": 0.44579966458301573,
"grad_norm": 0.6284562349319458,
"learning_rate": 6.126359547096975e-05,
"loss": 0.1194,
"step": 731
},
{
"epoch": 0.4464095136453728,
"grad_norm": 0.4746851921081543,
"learning_rate": 6.116732068727271e-05,
"loss": 0.0857,
"step": 732
},
{
"epoch": 0.44701936270772985,
"grad_norm": 0.45579567551612854,
"learning_rate": 6.107100230680279e-05,
"loss": 0.08,
"step": 733
},
{
"epoch": 0.4476292117700869,
"grad_norm": 0.38993462920188904,
"learning_rate": 6.0974640705583067e-05,
"loss": 0.0772,
"step": 734
},
{
"epoch": 0.44823906083244397,
"grad_norm": 0.5083535313606262,
"learning_rate": 6.0878236259805396e-05,
"loss": 0.1108,
"step": 735
},
{
"epoch": 0.44884890989480103,
"grad_norm": 0.4697110056877136,
"learning_rate": 6.078178934582885e-05,
"loss": 0.07,
"step": 736
},
{
"epoch": 0.4494587589571581,
"grad_norm": 0.5669456720352173,
"learning_rate": 6.068530034017835e-05,
"loss": 0.1208,
"step": 737
},
{
"epoch": 0.45006860801951515,
"grad_norm": 0.44945263862609863,
"learning_rate": 6.0588769619543086e-05,
"loss": 0.0852,
"step": 738
},
{
"epoch": 0.45067845708187226,
"grad_norm": 0.45248860120773315,
"learning_rate": 6.049219756077514e-05,
"loss": 0.0971,
"step": 739
},
{
"epoch": 0.4512883061442293,
"grad_norm": 0.5336349010467529,
"learning_rate": 6.0395584540887963e-05,
"loss": 0.1082,
"step": 740
},
{
"epoch": 0.4518981552065864,
"grad_norm": 0.5328031778335571,
"learning_rate": 6.029893093705492e-05,
"loss": 0.0894,
"step": 741
},
{
"epoch": 0.45250800426894344,
"grad_norm": 0.4532652497291565,
"learning_rate": 6.020223712660781e-05,
"loss": 0.0784,
"step": 742
},
{
"epoch": 0.4531178533313005,
"grad_norm": 0.4273979961872101,
"learning_rate": 6.010550348703538e-05,
"loss": 0.0714,
"step": 743
},
{
"epoch": 0.45372770239365756,
"grad_norm": 0.7462209463119507,
"learning_rate": 6.00087303959819e-05,
"loss": 0.1264,
"step": 744
},
{
"epoch": 0.4543375514560146,
"grad_norm": 0.465087354183197,
"learning_rate": 5.991191823124565e-05,
"loss": 0.0923,
"step": 745
},
{
"epoch": 0.4549474005183717,
"grad_norm": 0.7796549797058105,
"learning_rate": 5.981506737077744e-05,
"loss": 0.1313,
"step": 746
},
{
"epoch": 0.4555572495807288,
"grad_norm": 0.5143921971321106,
"learning_rate": 5.971817819267913e-05,
"loss": 0.0589,
"step": 747
},
{
"epoch": 0.45616709864308586,
"grad_norm": 0.39878177642822266,
"learning_rate": 5.962125107520223e-05,
"loss": 0.0767,
"step": 748
},
{
"epoch": 0.4567769477054429,
"grad_norm": 0.524658203125,
"learning_rate": 5.952428639674632e-05,
"loss": 0.0806,
"step": 749
},
{
"epoch": 0.4573867967678,
"grad_norm": 0.46519771218299866,
"learning_rate": 5.942728453585759e-05,
"loss": 0.0937,
"step": 750
},
{
"epoch": 0.45799664583015703,
"grad_norm": 0.44922301173210144,
"learning_rate": 5.9330245871227454e-05,
"loss": 0.0844,
"step": 751
},
{
"epoch": 0.4586064948925141,
"grad_norm": 0.48858335614204407,
"learning_rate": 5.923317078169096e-05,
"loss": 0.0947,
"step": 752
},
{
"epoch": 0.45921634395487115,
"grad_norm": 0.5561586618423462,
"learning_rate": 5.9136059646225375e-05,
"loss": 0.1093,
"step": 753
},
{
"epoch": 0.4598261930172282,
"grad_norm": 0.5627870559692383,
"learning_rate": 5.903891284394868e-05,
"loss": 0.1078,
"step": 754
},
{
"epoch": 0.46043604207958533,
"grad_norm": 0.709014892578125,
"learning_rate": 5.8941730754118116e-05,
"loss": 0.1276,
"step": 755
},
{
"epoch": 0.4610458911419424,
"grad_norm": 0.5438176393508911,
"learning_rate": 5.884451375612865e-05,
"loss": 0.0921,
"step": 756
},
{
"epoch": 0.46165574020429945,
"grad_norm": 0.48858410120010376,
"learning_rate": 5.874726222951157e-05,
"loss": 0.0938,
"step": 757
},
{
"epoch": 0.4622655892666565,
"grad_norm": 0.5165119767189026,
"learning_rate": 5.864997655393293e-05,
"loss": 0.0772,
"step": 758
},
{
"epoch": 0.46287543832901357,
"grad_norm": 0.7340338826179504,
"learning_rate": 5.855265710919211e-05,
"loss": 0.1437,
"step": 759
},
{
"epoch": 0.4634852873913706,
"grad_norm": 0.5283742547035217,
"learning_rate": 5.845530427522034e-05,
"loss": 0.0783,
"step": 760
},
{
"epoch": 0.4640951364537277,
"grad_norm": 0.5049426555633545,
"learning_rate": 5.835791843207916e-05,
"loss": 0.1039,
"step": 761
},
{
"epoch": 0.46470498551608475,
"grad_norm": 0.6662763953208923,
"learning_rate": 5.826049995995905e-05,
"loss": 0.0856,
"step": 762
},
{
"epoch": 0.46531483457844186,
"grad_norm": 0.7156078219413757,
"learning_rate": 5.8163049239177784e-05,
"loss": 0.0891,
"step": 763
},
{
"epoch": 0.4659246836407989,
"grad_norm": 0.6114723086357117,
"learning_rate": 5.8065566650179084e-05,
"loss": 0.147,
"step": 764
},
{
"epoch": 0.466534532703156,
"grad_norm": 0.5334054231643677,
"learning_rate": 5.7968052573531084e-05,
"loss": 0.1118,
"step": 765
},
{
"epoch": 0.46714438176551304,
"grad_norm": 0.5489366054534912,
"learning_rate": 5.787050738992482e-05,
"loss": 0.1097,
"step": 766
},
{
"epoch": 0.4677542308278701,
"grad_norm": 0.5465648770332336,
"learning_rate": 5.7772931480172795e-05,
"loss": 0.1104,
"step": 767
},
{
"epoch": 0.46836407989022716,
"grad_norm": 0.5904105305671692,
"learning_rate": 5.7675325225207455e-05,
"loss": 0.1487,
"step": 768
},
{
"epoch": 0.4689739289525842,
"grad_norm": 0.610356867313385,
"learning_rate": 5.757768900607972e-05,
"loss": 0.117,
"step": 769
},
{
"epoch": 0.4695837780149413,
"grad_norm": 0.5449678301811218,
"learning_rate": 5.748002320395746e-05,
"loss": 0.1006,
"step": 770
},
{
"epoch": 0.4701936270772984,
"grad_norm": 0.42668890953063965,
"learning_rate": 5.738232820012407e-05,
"loss": 0.0988,
"step": 771
},
{
"epoch": 0.47080347613965545,
"grad_norm": 0.41906148195266724,
"learning_rate": 5.7284604375976913e-05,
"loss": 0.0924,
"step": 772
},
{
"epoch": 0.4714133252020125,
"grad_norm": 0.44049015641212463,
"learning_rate": 5.718685211302592e-05,
"loss": 0.0788,
"step": 773
},
{
"epoch": 0.47202317426436957,
"grad_norm": 0.5249271988868713,
"learning_rate": 5.7089071792892e-05,
"loss": 0.0954,
"step": 774
},
{
"epoch": 0.47263302332672663,
"grad_norm": 0.4835070073604584,
"learning_rate": 5.699126379730559e-05,
"loss": 0.094,
"step": 775
},
{
"epoch": 0.4732428723890837,
"grad_norm": 0.5688073635101318,
"learning_rate": 5.6893428508105225e-05,
"loss": 0.0933,
"step": 776
},
{
"epoch": 0.47385272145144075,
"grad_norm": 0.4810085892677307,
"learning_rate": 5.6795566307235915e-05,
"loss": 0.11,
"step": 777
},
{
"epoch": 0.4744625705137978,
"grad_norm": 0.5228220820426941,
"learning_rate": 5.669767757674781e-05,
"loss": 0.1047,
"step": 778
},
{
"epoch": 0.4750724195761549,
"grad_norm": 0.8499104380607605,
"learning_rate": 5.6599762698794554e-05,
"loss": 0.1078,
"step": 779
},
{
"epoch": 0.475682268638512,
"grad_norm": 0.373967707157135,
"learning_rate": 5.650182205563197e-05,
"loss": 0.077,
"step": 780
},
{
"epoch": 0.47629211770086904,
"grad_norm": 0.7530614733695984,
"learning_rate": 5.640385602961634e-05,
"loss": 0.1696,
"step": 781
},
{
"epoch": 0.4769019667632261,
"grad_norm": 0.5162388682365417,
"learning_rate": 5.630586500320317e-05,
"loss": 0.1204,
"step": 782
},
{
"epoch": 0.47751181582558316,
"grad_norm": 0.3680694103240967,
"learning_rate": 5.620784935894547e-05,
"loss": 0.084,
"step": 783
},
{
"epoch": 0.4781216648879402,
"grad_norm": 0.5393168926239014,
"learning_rate": 5.610980947949239e-05,
"loss": 0.0939,
"step": 784
},
{
"epoch": 0.4787315139502973,
"grad_norm": 0.6208060383796692,
"learning_rate": 5.601174574758771e-05,
"loss": 0.1027,
"step": 785
},
{
"epoch": 0.47934136301265434,
"grad_norm": 0.5323257446289062,
"learning_rate": 5.5913658546068295e-05,
"loss": 0.0839,
"step": 786
},
{
"epoch": 0.47995121207501146,
"grad_norm": 0.5011024475097656,
"learning_rate": 5.581554825786267e-05,
"loss": 0.0995,
"step": 787
},
{
"epoch": 0.4805610611373685,
"grad_norm": 0.5328444838523865,
"learning_rate": 5.571741526598946e-05,
"loss": 0.1031,
"step": 788
},
{
"epoch": 0.4811709101997256,
"grad_norm": 0.32099008560180664,
"learning_rate": 5.5619259953555945e-05,
"loss": 0.0754,
"step": 789
},
{
"epoch": 0.48178075926208264,
"grad_norm": 0.48089420795440674,
"learning_rate": 5.552108270375654e-05,
"loss": 0.103,
"step": 790
},
{
"epoch": 0.4823906083244397,
"grad_norm": 0.6017656326293945,
"learning_rate": 5.5422883899871284e-05,
"loss": 0.0976,
"step": 791
},
{
"epoch": 0.48300045738679676,
"grad_norm": 0.5539811849594116,
"learning_rate": 5.532466392526438e-05,
"loss": 0.1237,
"step": 792
},
{
"epoch": 0.4836103064491538,
"grad_norm": 0.5454038977622986,
"learning_rate": 5.522642316338268e-05,
"loss": 0.103,
"step": 793
},
{
"epoch": 0.4842201555115109,
"grad_norm": 0.6915159821510315,
"learning_rate": 5.512816199775419e-05,
"loss": 0.1127,
"step": 794
},
{
"epoch": 0.484830004573868,
"grad_norm": 0.4176887571811676,
"learning_rate": 5.5029880811986544e-05,
"loss": 0.0899,
"step": 795
},
{
"epoch": 0.48543985363622505,
"grad_norm": 0.4416688084602356,
"learning_rate": 5.493157998976559e-05,
"loss": 0.0845,
"step": 796
},
{
"epoch": 0.4860497026985821,
"grad_norm": 0.8192511200904846,
"learning_rate": 5.483325991485379e-05,
"loss": 0.1184,
"step": 797
},
{
"epoch": 0.48665955176093917,
"grad_norm": 0.6466198563575745,
"learning_rate": 5.473492097108877e-05,
"loss": 0.1062,
"step": 798
},
{
"epoch": 0.48726940082329623,
"grad_norm": 0.4846915304660797,
"learning_rate": 5.463656354238184e-05,
"loss": 0.088,
"step": 799
},
{
"epoch": 0.4878792498856533,
"grad_norm": 0.5764984488487244,
"learning_rate": 5.453818801271645e-05,
"loss": 0.0818,
"step": 800
},
{
"epoch": 0.48848909894801035,
"grad_norm": 0.9716119766235352,
"learning_rate": 5.4439794766146746e-05,
"loss": 0.0989,
"step": 801
},
{
"epoch": 0.4890989480103674,
"grad_norm": 0.49489036202430725,
"learning_rate": 5.434138418679602e-05,
"loss": 0.09,
"step": 802
},
{
"epoch": 0.4897087970727245,
"grad_norm": 0.5157510042190552,
"learning_rate": 5.424295665885523e-05,
"loss": 0.0863,
"step": 803
},
{
"epoch": 0.4903186461350816,
"grad_norm": 0.7257834076881409,
"learning_rate": 5.414451256658149e-05,
"loss": 0.1027,
"step": 804
},
{
"epoch": 0.49092849519743864,
"grad_norm": 0.6347923278808594,
"learning_rate": 5.404605229429664e-05,
"loss": 0.1195,
"step": 805
},
{
"epoch": 0.4915383442597957,
"grad_norm": 0.7614789605140686,
"learning_rate": 5.39475762263856e-05,
"loss": 0.1166,
"step": 806
},
{
"epoch": 0.49214819332215276,
"grad_norm": 0.5669092535972595,
"learning_rate": 5.384908474729501e-05,
"loss": 0.111,
"step": 807
},
{
"epoch": 0.4927580423845098,
"grad_norm": 1.0184087753295898,
"learning_rate": 5.3750578241531655e-05,
"loss": 0.1315,
"step": 808
},
{
"epoch": 0.4933678914468669,
"grad_norm": 0.5959277153015137,
"learning_rate": 5.365205709366099e-05,
"loss": 0.1031,
"step": 809
},
{
"epoch": 0.493977740509224,
"grad_norm": 0.5251980423927307,
"learning_rate": 5.355352168830565e-05,
"loss": 0.0773,
"step": 810
},
{
"epoch": 0.49458758957158105,
"grad_norm": 0.6087118983268738,
"learning_rate": 5.34549724101439e-05,
"loss": 0.1011,
"step": 811
},
{
"epoch": 0.4951974386339381,
"grad_norm": 0.5269709229469299,
"learning_rate": 5.335640964390818e-05,
"loss": 0.0921,
"step": 812
},
{
"epoch": 0.4958072876962952,
"grad_norm": 0.3719567656517029,
"learning_rate": 5.325783377438357e-05,
"loss": 0.0738,
"step": 813
},
{
"epoch": 0.49641713675865223,
"grad_norm": 0.46428948640823364,
"learning_rate": 5.315924518640634e-05,
"loss": 0.0613,
"step": 814
},
{
"epoch": 0.4970269858210093,
"grad_norm": 0.47667044401168823,
"learning_rate": 5.306064426486237e-05,
"loss": 0.1025,
"step": 815
},
{
"epoch": 0.49763683488336635,
"grad_norm": 0.6388903856277466,
"learning_rate": 5.296203139468572e-05,
"loss": 0.0996,
"step": 816
},
{
"epoch": 0.4982466839457234,
"grad_norm": 0.40715262293815613,
"learning_rate": 5.286340696085709e-05,
"loss": 0.0759,
"step": 817
},
{
"epoch": 0.4988565330080805,
"grad_norm": 0.6345905661582947,
"learning_rate": 5.276477134840231e-05,
"loss": 0.1452,
"step": 818
},
{
"epoch": 0.4994663820704376,
"grad_norm": 0.467983603477478,
"learning_rate": 5.266612494239088e-05,
"loss": 0.0765,
"step": 819
},
{
"epoch": 0.5000762311327946,
"grad_norm": 0.4643031060695648,
"learning_rate": 5.2567468127934406e-05,
"loss": 0.0961,
"step": 820
},
{
"epoch": 0.5006860801951517,
"grad_norm": 0.4682145118713379,
"learning_rate": 5.246880129018516e-05,
"loss": 0.0973,
"step": 821
},
{
"epoch": 0.5012959292575088,
"grad_norm": 0.5168704986572266,
"learning_rate": 5.2370124814334495e-05,
"loss": 0.0976,
"step": 822
},
{
"epoch": 0.5019057783198658,
"grad_norm": 0.4242720901966095,
"learning_rate": 5.227143908561145e-05,
"loss": 0.0837,
"step": 823
},
{
"epoch": 0.5025156273822229,
"grad_norm": 0.3759913742542267,
"learning_rate": 5.2172744489281145e-05,
"loss": 0.0619,
"step": 824
},
{
"epoch": 0.5031254764445799,
"grad_norm": 0.5110340714454651,
"learning_rate": 5.207404141064334e-05,
"loss": 0.0996,
"step": 825
},
{
"epoch": 0.503735325506937,
"grad_norm": 0.5003123879432678,
"learning_rate": 5.197533023503089e-05,
"loss": 0.083,
"step": 826
},
{
"epoch": 0.5043451745692941,
"grad_norm": 0.598993718624115,
"learning_rate": 5.187661134780829e-05,
"loss": 0.1122,
"step": 827
},
{
"epoch": 0.5049550236316511,
"grad_norm": 0.49572473764419556,
"learning_rate": 5.177788513437013e-05,
"loss": 0.0628,
"step": 828
},
{
"epoch": 0.5055648726940082,
"grad_norm": 0.7073934674263,
"learning_rate": 5.1679151980139564e-05,
"loss": 0.1073,
"step": 829
},
{
"epoch": 0.5061747217563654,
"grad_norm": 0.6053644418716431,
"learning_rate": 5.158041227056689e-05,
"loss": 0.1043,
"step": 830
},
{
"epoch": 0.5067845708187224,
"grad_norm": 0.4864380359649658,
"learning_rate": 5.148166639112799e-05,
"loss": 0.0941,
"step": 831
},
{
"epoch": 0.5073944198810795,
"grad_norm": 0.4678514003753662,
"learning_rate": 5.1382914727322815e-05,
"loss": 0.0668,
"step": 832
},
{
"epoch": 0.5080042689434365,
"grad_norm": 0.8016685247421265,
"learning_rate": 5.128415766467392e-05,
"loss": 0.1224,
"step": 833
},
{
"epoch": 0.5086141180057936,
"grad_norm": 0.6746135950088501,
"learning_rate": 5.118539558872488e-05,
"loss": 0.1259,
"step": 834
},
{
"epoch": 0.5092239670681507,
"grad_norm": 0.6684184074401855,
"learning_rate": 5.1086628885038946e-05,
"loss": 0.1296,
"step": 835
},
{
"epoch": 0.5098338161305077,
"grad_norm": 0.5928740501403809,
"learning_rate": 5.0987857939197324e-05,
"loss": 0.1053,
"step": 836
},
{
"epoch": 0.5104436651928648,
"grad_norm": 0.41154876351356506,
"learning_rate": 5.0889083136797875e-05,
"loss": 0.0731,
"step": 837
},
{
"epoch": 0.5110535142552218,
"grad_norm": 0.3903616666793823,
"learning_rate": 5.079030486345345e-05,
"loss": 0.0554,
"step": 838
},
{
"epoch": 0.5116633633175789,
"grad_norm": 0.5829324126243591,
"learning_rate": 5.0691523504790474e-05,
"loss": 0.0902,
"step": 839
},
{
"epoch": 0.512273212379936,
"grad_norm": 0.42772719264030457,
"learning_rate": 5.0592739446447425e-05,
"loss": 0.1054,
"step": 840
},
{
"epoch": 0.512883061442293,
"grad_norm": 0.729825496673584,
"learning_rate": 5.049395307407329e-05,
"loss": 0.1002,
"step": 841
},
{
"epoch": 0.5134929105046501,
"grad_norm": 0.4020238220691681,
"learning_rate": 5.039516477332614e-05,
"loss": 0.0909,
"step": 842
},
{
"epoch": 0.5141027595670071,
"grad_norm": 0.9202263951301575,
"learning_rate": 5.029637492987153e-05,
"loss": 0.1014,
"step": 843
},
{
"epoch": 0.5147126086293642,
"grad_norm": 0.5091873407363892,
"learning_rate": 5.0197583929381066e-05,
"loss": 0.1107,
"step": 844
},
{
"epoch": 0.5153224576917212,
"grad_norm": 0.48788222670555115,
"learning_rate": 5.009879215753085e-05,
"loss": 0.0657,
"step": 845
},
{
"epoch": 0.5159323067540784,
"grad_norm": 0.5596025586128235,
"learning_rate": 5e-05,
"loss": 0.1414,
"step": 846
},
{
"epoch": 0.5165421558164355,
"grad_norm": 0.4461548924446106,
"learning_rate": 4.990120784246917e-05,
"loss": 0.0918,
"step": 847
},
{
"epoch": 0.5171520048787925,
"grad_norm": 0.614068865776062,
"learning_rate": 4.980241607061894e-05,
"loss": 0.103,
"step": 848
},
{
"epoch": 0.5177618539411496,
"grad_norm": 0.4434625804424286,
"learning_rate": 4.970362507012848e-05,
"loss": 0.1009,
"step": 849
},
{
"epoch": 0.5183717030035067,
"grad_norm": 0.47529685497283936,
"learning_rate": 4.960483522667387e-05,
"loss": 0.0921,
"step": 850
},
{
"epoch": 0.5189815520658637,
"grad_norm": 0.6131975054740906,
"learning_rate": 4.950604692592672e-05,
"loss": 0.1081,
"step": 851
},
{
"epoch": 0.5195914011282208,
"grad_norm": 0.8078132271766663,
"learning_rate": 4.940726055355259e-05,
"loss": 0.1256,
"step": 852
},
{
"epoch": 0.5202012501905778,
"grad_norm": 0.591907799243927,
"learning_rate": 4.9308476495209544e-05,
"loss": 0.1078,
"step": 853
},
{
"epoch": 0.5208110992529349,
"grad_norm": 0.4097784459590912,
"learning_rate": 4.920969513654656e-05,
"loss": 0.0671,
"step": 854
},
{
"epoch": 0.521420948315292,
"grad_norm": 0.5087611675262451,
"learning_rate": 4.911091686320213e-05,
"loss": 0.1013,
"step": 855
},
{
"epoch": 0.522030797377649,
"grad_norm": 0.8746770024299622,
"learning_rate": 4.901214206080269e-05,
"loss": 0.1127,
"step": 856
},
{
"epoch": 0.5226406464400061,
"grad_norm": 0.5117619037628174,
"learning_rate": 4.891337111496107e-05,
"loss": 0.1049,
"step": 857
},
{
"epoch": 0.5232504955023631,
"grad_norm": 0.5372995734214783,
"learning_rate": 4.881460441127513e-05,
"loss": 0.1537,
"step": 858
},
{
"epoch": 0.5238603445647202,
"grad_norm": 0.7196261286735535,
"learning_rate": 4.87158423353261e-05,
"loss": 0.1126,
"step": 859
},
{
"epoch": 0.5244701936270773,
"grad_norm": 0.4715578854084015,
"learning_rate": 4.86170852726772e-05,
"loss": 0.084,
"step": 860
},
{
"epoch": 0.5250800426894343,
"grad_norm": 0.5337904691696167,
"learning_rate": 4.851833360887201e-05,
"loss": 0.0975,
"step": 861
},
{
"epoch": 0.5256898917517915,
"grad_norm": 0.4751030206680298,
"learning_rate": 4.841958772943311e-05,
"loss": 0.0825,
"step": 862
},
{
"epoch": 0.5262997408141485,
"grad_norm": 0.34388238191604614,
"learning_rate": 4.8320848019860454e-05,
"loss": 0.0675,
"step": 863
},
{
"epoch": 0.5269095898765056,
"grad_norm": 0.44842541217803955,
"learning_rate": 4.8222114865629886e-05,
"loss": 0.0845,
"step": 864
},
{
"epoch": 0.5275194389388627,
"grad_norm": 0.4816577732563019,
"learning_rate": 4.8123388652191715e-05,
"loss": 0.1009,
"step": 865
},
{
"epoch": 0.5281292880012197,
"grad_norm": 0.49896806478500366,
"learning_rate": 4.802466976496911e-05,
"loss": 0.1055,
"step": 866
},
{
"epoch": 0.5287391370635768,
"grad_norm": 0.7189498543739319,
"learning_rate": 4.7925958589356675e-05,
"loss": 0.1086,
"step": 867
},
{
"epoch": 0.5293489861259338,
"grad_norm": 0.6172751188278198,
"learning_rate": 4.782725551071886e-05,
"loss": 0.0865,
"step": 868
},
{
"epoch": 0.5299588351882909,
"grad_norm": 0.431423157453537,
"learning_rate": 4.7728560914388566e-05,
"loss": 0.1016,
"step": 869
},
{
"epoch": 0.530568684250648,
"grad_norm": 0.9568158388137817,
"learning_rate": 4.762987518566551e-05,
"loss": 0.1108,
"step": 870
},
{
"epoch": 0.531178533313005,
"grad_norm": 0.6039659380912781,
"learning_rate": 4.7531198709814854e-05,
"loss": 0.097,
"step": 871
},
{
"epoch": 0.5317883823753621,
"grad_norm": 0.5661967396736145,
"learning_rate": 4.7432531872065605e-05,
"loss": 0.0817,
"step": 872
},
{
"epoch": 0.5323982314377191,
"grad_norm": 0.6298314332962036,
"learning_rate": 4.7333875057609126e-05,
"loss": 0.0873,
"step": 873
},
{
"epoch": 0.5330080805000762,
"grad_norm": 0.7972177863121033,
"learning_rate": 4.7235228651597694e-05,
"loss": 0.0818,
"step": 874
},
{
"epoch": 0.5336179295624333,
"grad_norm": 0.591971218585968,
"learning_rate": 4.713659303914292e-05,
"loss": 0.0831,
"step": 875
},
{
"epoch": 0.5342277786247903,
"grad_norm": 0.4230268597602844,
"learning_rate": 4.703796860531429e-05,
"loss": 0.0709,
"step": 876
},
{
"epoch": 0.5348376276871474,
"grad_norm": 0.8520568609237671,
"learning_rate": 4.693935573513764e-05,
"loss": 0.096,
"step": 877
},
{
"epoch": 0.5354474767495045,
"grad_norm": 0.3330819606781006,
"learning_rate": 4.684075481359368e-05,
"loss": 0.0623,
"step": 878
},
{
"epoch": 0.5360573258118616,
"grad_norm": 0.49577683210372925,
"learning_rate": 4.674216622561644e-05,
"loss": 0.0859,
"step": 879
},
{
"epoch": 0.5366671748742187,
"grad_norm": 0.8136767745018005,
"learning_rate": 4.664359035609183e-05,
"loss": 0.1332,
"step": 880
},
{
"epoch": 0.5372770239365757,
"grad_norm": 0.5900670289993286,
"learning_rate": 4.654502758985611e-05,
"loss": 0.0708,
"step": 881
},
{
"epoch": 0.5378868729989328,
"grad_norm": 0.6791537404060364,
"learning_rate": 4.644647831169435e-05,
"loss": 0.1083,
"step": 882
},
{
"epoch": 0.5384967220612898,
"grad_norm": 0.7151476144790649,
"learning_rate": 4.6347942906339015e-05,
"loss": 0.1339,
"step": 883
},
{
"epoch": 0.5391065711236469,
"grad_norm": 0.931968092918396,
"learning_rate": 4.624942175846835e-05,
"loss": 0.1169,
"step": 884
},
{
"epoch": 0.539716420186004,
"grad_norm": 0.47680747509002686,
"learning_rate": 4.615091525270501e-05,
"loss": 0.0858,
"step": 885
},
{
"epoch": 0.540326269248361,
"grad_norm": 0.48191407322883606,
"learning_rate": 4.6052423773614404e-05,
"loss": 0.1005,
"step": 886
},
{
"epoch": 0.5409361183107181,
"grad_norm": 0.7821782827377319,
"learning_rate": 4.595394770570337e-05,
"loss": 0.1004,
"step": 887
},
{
"epoch": 0.5415459673730751,
"grad_norm": 0.6783169507980347,
"learning_rate": 4.585548743341851e-05,
"loss": 0.111,
"step": 888
},
{
"epoch": 0.5421558164354322,
"grad_norm": 0.5506354570388794,
"learning_rate": 4.575704334114478e-05,
"loss": 0.1143,
"step": 889
},
{
"epoch": 0.5427656654977893,
"grad_norm": 0.3778802752494812,
"learning_rate": 4.5658615813204e-05,
"loss": 0.0884,
"step": 890
},
{
"epoch": 0.5433755145601463,
"grad_norm": 0.48291489481925964,
"learning_rate": 4.5560205233853266e-05,
"loss": 0.1099,
"step": 891
},
{
"epoch": 0.5439853636225034,
"grad_norm": 0.5400805473327637,
"learning_rate": 4.546181198728357e-05,
"loss": 0.1082,
"step": 892
},
{
"epoch": 0.5445952126848606,
"grad_norm": 0.4781759977340698,
"learning_rate": 4.5363436457618174e-05,
"loss": 0.0883,
"step": 893
},
{
"epoch": 0.5452050617472176,
"grad_norm": 0.5502304434776306,
"learning_rate": 4.5265079028911244e-05,
"loss": 0.1174,
"step": 894
},
{
"epoch": 0.5458149108095747,
"grad_norm": 0.5287635326385498,
"learning_rate": 4.516674008514623e-05,
"loss": 0.1009,
"step": 895
},
{
"epoch": 0.5464247598719317,
"grad_norm": 0.5179880261421204,
"learning_rate": 4.506842001023442e-05,
"loss": 0.0988,
"step": 896
},
{
"epoch": 0.5470346089342888,
"grad_norm": 0.5177373290061951,
"learning_rate": 4.497011918801347e-05,
"loss": 0.1005,
"step": 897
},
{
"epoch": 0.5476444579966459,
"grad_norm": 0.505449652671814,
"learning_rate": 4.487183800224583e-05,
"loss": 0.1079,
"step": 898
},
{
"epoch": 0.5482543070590029,
"grad_norm": 0.6120879650115967,
"learning_rate": 4.477357683661734e-05,
"loss": 0.1275,
"step": 899
},
{
"epoch": 0.54886415612136,
"grad_norm": 0.4165956377983093,
"learning_rate": 4.467533607473563e-05,
"loss": 0.0996,
"step": 900
},
{
"epoch": 0.549474005183717,
"grad_norm": 0.32662227749824524,
"learning_rate": 4.4577116100128735e-05,
"loss": 0.0746,
"step": 901
},
{
"epoch": 0.5500838542460741,
"grad_norm": 0.5906914472579956,
"learning_rate": 4.447891729624347e-05,
"loss": 0.1108,
"step": 902
},
{
"epoch": 0.5506937033084311,
"grad_norm": 0.5592126846313477,
"learning_rate": 4.4380740046444066e-05,
"loss": 0.1085,
"step": 903
},
{
"epoch": 0.5513035523707882,
"grad_norm": 0.4679374098777771,
"learning_rate": 4.4282584734010554e-05,
"loss": 0.0925,
"step": 904
},
{
"epoch": 0.5519134014331453,
"grad_norm": 0.5998743772506714,
"learning_rate": 4.418445174213734e-05,
"loss": 0.1086,
"step": 905
},
{
"epoch": 0.5525232504955023,
"grad_norm": 0.4834987223148346,
"learning_rate": 4.4086341453931716e-05,
"loss": 0.115,
"step": 906
},
{
"epoch": 0.5531330995578594,
"grad_norm": 0.637567400932312,
"learning_rate": 4.39882542524123e-05,
"loss": 0.106,
"step": 907
},
{
"epoch": 0.5537429486202164,
"grad_norm": 0.47768768668174744,
"learning_rate": 4.389019052050762e-05,
"loss": 0.1063,
"step": 908
},
{
"epoch": 0.5543527976825736,
"grad_norm": 0.5357100367546082,
"learning_rate": 4.379215064105454e-05,
"loss": 0.0951,
"step": 909
},
{
"epoch": 0.5549626467449307,
"grad_norm": 0.4694567322731018,
"learning_rate": 4.369413499679684e-05,
"loss": 0.117,
"step": 910
},
{
"epoch": 0.5555724958072877,
"grad_norm": 0.5599071979522705,
"learning_rate": 4.3596143970383664e-05,
"loss": 0.0995,
"step": 911
},
{
"epoch": 0.5561823448696448,
"grad_norm": 0.42838895320892334,
"learning_rate": 4.3498177944368044e-05,
"loss": 0.0881,
"step": 912
},
{
"epoch": 0.5567921939320019,
"grad_norm": 0.4186185300350189,
"learning_rate": 4.340023730120545e-05,
"loss": 0.07,
"step": 913
},
{
"epoch": 0.5574020429943589,
"grad_norm": 0.34970325231552124,
"learning_rate": 4.330232242325221e-05,
"loss": 0.0779,
"step": 914
},
{
"epoch": 0.558011892056716,
"grad_norm": 0.3880801796913147,
"learning_rate": 4.3204433692764096e-05,
"loss": 0.0922,
"step": 915
},
{
"epoch": 0.558621741119073,
"grad_norm": 0.5633447766304016,
"learning_rate": 4.3106571491894786e-05,
"loss": 0.1038,
"step": 916
},
{
"epoch": 0.5592315901814301,
"grad_norm": 0.4482801854610443,
"learning_rate": 4.3008736202694414e-05,
"loss": 0.0986,
"step": 917
},
{
"epoch": 0.5598414392437872,
"grad_norm": 0.47528374195098877,
"learning_rate": 4.2910928207108005e-05,
"loss": 0.0942,
"step": 918
},
{
"epoch": 0.5604512883061442,
"grad_norm": 0.48216068744659424,
"learning_rate": 4.281314788697408e-05,
"loss": 0.085,
"step": 919
},
{
"epoch": 0.5610611373685013,
"grad_norm": 0.6034498810768127,
"learning_rate": 4.271539562402309e-05,
"loss": 0.1065,
"step": 920
},
{
"epoch": 0.5616709864308583,
"grad_norm": 0.9733392000198364,
"learning_rate": 4.2617671799875944e-05,
"loss": 0.1291,
"step": 921
},
{
"epoch": 0.5622808354932154,
"grad_norm": 0.46398019790649414,
"learning_rate": 4.2519976796042554e-05,
"loss": 0.1117,
"step": 922
},
{
"epoch": 0.5628906845555725,
"grad_norm": 0.3748071789741516,
"learning_rate": 4.242231099392029e-05,
"loss": 0.0846,
"step": 923
},
{
"epoch": 0.5635005336179295,
"grad_norm": 0.468289315700531,
"learning_rate": 4.232467477479255e-05,
"loss": 0.0777,
"step": 924
},
{
"epoch": 0.5641103826802867,
"grad_norm": 0.4760587513446808,
"learning_rate": 4.222706851982721e-05,
"loss": 0.1192,
"step": 925
},
{
"epoch": 0.5647202317426437,
"grad_norm": 0.47677165269851685,
"learning_rate": 4.212949261007519e-05,
"loss": 0.0864,
"step": 926
},
{
"epoch": 0.5653300808050008,
"grad_norm": 0.49228742718696594,
"learning_rate": 4.203194742646893e-05,
"loss": 0.0896,
"step": 927
},
{
"epoch": 0.5659399298673579,
"grad_norm": 0.7319192886352539,
"learning_rate": 4.193443334982093e-05,
"loss": 0.0914,
"step": 928
},
{
"epoch": 0.5665497789297149,
"grad_norm": 0.4847294092178345,
"learning_rate": 4.1836950760822235e-05,
"loss": 0.0873,
"step": 929
},
{
"epoch": 0.567159627992072,
"grad_norm": 0.705895721912384,
"learning_rate": 4.1739500040040964e-05,
"loss": 0.1172,
"step": 930
},
{
"epoch": 0.567769477054429,
"grad_norm": 0.6366685032844543,
"learning_rate": 4.1642081567920846e-05,
"loss": 0.1112,
"step": 931
},
{
"epoch": 0.5683793261167861,
"grad_norm": 0.7112983465194702,
"learning_rate": 4.154469572477967e-05,
"loss": 0.1664,
"step": 932
},
{
"epoch": 0.5689891751791432,
"grad_norm": 0.48351553082466125,
"learning_rate": 4.144734289080791e-05,
"loss": 0.0989,
"step": 933
},
{
"epoch": 0.5695990242415002,
"grad_norm": 0.5432968139648438,
"learning_rate": 4.1350023446067084e-05,
"loss": 0.0673,
"step": 934
},
{
"epoch": 0.5702088733038573,
"grad_norm": 0.5261378884315491,
"learning_rate": 4.125273777048845e-05,
"loss": 0.0701,
"step": 935
},
{
"epoch": 0.5708187223662143,
"grad_norm": 0.6279370188713074,
"learning_rate": 4.115548624387137e-05,
"loss": 0.0871,
"step": 936
},
{
"epoch": 0.5714285714285714,
"grad_norm": 0.5255637168884277,
"learning_rate": 4.1058269245881896e-05,
"loss": 0.0954,
"step": 937
},
{
"epoch": 0.5720384204909285,
"grad_norm": 0.5804142951965332,
"learning_rate": 4.0961087156051334e-05,
"loss": 0.0721,
"step": 938
},
{
"epoch": 0.5726482695532855,
"grad_norm": 0.5322815179824829,
"learning_rate": 4.086394035377463e-05,
"loss": 0.1005,
"step": 939
},
{
"epoch": 0.5732581186156426,
"grad_norm": 0.6627009510993958,
"learning_rate": 4.076682921830906e-05,
"loss": 0.1171,
"step": 940
},
{
"epoch": 0.5738679676779997,
"grad_norm": 0.6268516778945923,
"learning_rate": 4.066975412877255e-05,
"loss": 0.1156,
"step": 941
},
{
"epoch": 0.5744778167403568,
"grad_norm": 0.722302258014679,
"learning_rate": 4.0572715464142416e-05,
"loss": 0.0915,
"step": 942
},
{
"epoch": 0.5750876658027139,
"grad_norm": 0.6084862351417542,
"learning_rate": 4.04757136032537e-05,
"loss": 0.0947,
"step": 943
},
{
"epoch": 0.5756975148650709,
"grad_norm": 0.5269582271575928,
"learning_rate": 4.037874892479778e-05,
"loss": 0.0845,
"step": 944
},
{
"epoch": 0.576307363927428,
"grad_norm": 0.49478936195373535,
"learning_rate": 4.028182180732088e-05,
"loss": 0.0894,
"step": 945
},
{
"epoch": 0.576917212989785,
"grad_norm": 0.6067556738853455,
"learning_rate": 4.0184932629222575e-05,
"loss": 0.1107,
"step": 946
},
{
"epoch": 0.5775270620521421,
"grad_norm": 0.6461120843887329,
"learning_rate": 4.0088081768754365e-05,
"loss": 0.1432,
"step": 947
},
{
"epoch": 0.5781369111144992,
"grad_norm": 0.4849057197570801,
"learning_rate": 3.99912696040181e-05,
"loss": 0.0886,
"step": 948
},
{
"epoch": 0.5787467601768562,
"grad_norm": 0.4269167482852936,
"learning_rate": 3.9894496512964635e-05,
"loss": 0.073,
"step": 949
},
{
"epoch": 0.5793566092392133,
"grad_norm": 0.5539390444755554,
"learning_rate": 3.97977628733922e-05,
"loss": 0.0906,
"step": 950
},
{
"epoch": 0.5799664583015703,
"grad_norm": 0.5726113319396973,
"learning_rate": 3.970106906294509e-05,
"loss": 0.1105,
"step": 951
},
{
"epoch": 0.5805763073639274,
"grad_norm": 0.5241560339927673,
"learning_rate": 3.960441545911204e-05,
"loss": 0.0815,
"step": 952
},
{
"epoch": 0.5811861564262845,
"grad_norm": 0.6997159719467163,
"learning_rate": 3.950780243922486e-05,
"loss": 0.0663,
"step": 953
},
{
"epoch": 0.5817960054886415,
"grad_norm": 0.4698490500450134,
"learning_rate": 3.9411230380456925e-05,
"loss": 0.0735,
"step": 954
},
{
"epoch": 0.5824058545509986,
"grad_norm": 0.6257011294364929,
"learning_rate": 3.9314699659821666e-05,
"loss": 0.0939,
"step": 955
},
{
"epoch": 0.5830157036133556,
"grad_norm": 0.4785114824771881,
"learning_rate": 3.921821065417116e-05,
"loss": 0.076,
"step": 956
},
{
"epoch": 0.5836255526757128,
"grad_norm": 0.6592893004417419,
"learning_rate": 3.9121763740194616e-05,
"loss": 0.0932,
"step": 957
},
{
"epoch": 0.5842354017380699,
"grad_norm": 0.5891658067703247,
"learning_rate": 3.9025359294416945e-05,
"loss": 0.0933,
"step": 958
},
{
"epoch": 0.5848452508004269,
"grad_norm": 0.7100630402565002,
"learning_rate": 3.892899769319722e-05,
"loss": 0.141,
"step": 959
},
{
"epoch": 0.585455099862784,
"grad_norm": 0.5067220330238342,
"learning_rate": 3.8832679312727296e-05,
"loss": 0.0783,
"step": 960
},
{
"epoch": 0.586064948925141,
"grad_norm": 0.40470728278160095,
"learning_rate": 3.873640452903026e-05,
"loss": 0.0809,
"step": 961
},
{
"epoch": 0.5866747979874981,
"grad_norm": 0.5892449617385864,
"learning_rate": 3.8640173717959007e-05,
"loss": 0.1326,
"step": 962
},
{
"epoch": 0.5872846470498552,
"grad_norm": 0.8093147873878479,
"learning_rate": 3.854398725519479e-05,
"loss": 0.1123,
"step": 963
},
{
"epoch": 0.5878944961122122,
"grad_norm": 0.5357198119163513,
"learning_rate": 3.8447845516245695e-05,
"loss": 0.0791,
"step": 964
},
{
"epoch": 0.5885043451745693,
"grad_norm": 0.5130123496055603,
"learning_rate": 3.835174887644523e-05,
"loss": 0.1112,
"step": 965
},
{
"epoch": 0.5891141942369263,
"grad_norm": 0.61879563331604,
"learning_rate": 3.825569771095082e-05,
"loss": 0.1099,
"step": 966
},
{
"epoch": 0.5897240432992834,
"grad_norm": 0.3868868052959442,
"learning_rate": 3.815969239474238e-05,
"loss": 0.0591,
"step": 967
},
{
"epoch": 0.5903338923616405,
"grad_norm": 0.7583017945289612,
"learning_rate": 3.8063733302620826e-05,
"loss": 0.1271,
"step": 968
},
{
"epoch": 0.5909437414239975,
"grad_norm": 0.6140808463096619,
"learning_rate": 3.796782080920659e-05,
"loss": 0.0776,
"step": 969
},
{
"epoch": 0.5915535904863546,
"grad_norm": 0.3586467206478119,
"learning_rate": 3.787195528893823e-05,
"loss": 0.0671,
"step": 970
},
{
"epoch": 0.5921634395487116,
"grad_norm": 0.556140124797821,
"learning_rate": 3.777613711607087e-05,
"loss": 0.1058,
"step": 971
},
{
"epoch": 0.5927732886110688,
"grad_norm": 0.4077626168727875,
"learning_rate": 3.768036666467486e-05,
"loss": 0.0682,
"step": 972
},
{
"epoch": 0.5933831376734259,
"grad_norm": 0.3156653642654419,
"learning_rate": 3.758464430863417e-05,
"loss": 0.0659,
"step": 973
},
{
"epoch": 0.5939929867357829,
"grad_norm": 0.490740567445755,
"learning_rate": 3.748897042164508e-05,
"loss": 0.082,
"step": 974
},
{
"epoch": 0.59460283579814,
"grad_norm": 0.539813220500946,
"learning_rate": 3.7393345377214586e-05,
"loss": 0.0854,
"step": 975
},
{
"epoch": 0.595212684860497,
"grad_norm": 0.5377735495567322,
"learning_rate": 3.729776954865905e-05,
"loss": 0.0937,
"step": 976
},
{
"epoch": 0.5958225339228541,
"grad_norm": 0.5617977380752563,
"learning_rate": 3.720224330910268e-05,
"loss": 0.1059,
"step": 977
},
{
"epoch": 0.5964323829852112,
"grad_norm": 0.38793593645095825,
"learning_rate": 3.7106767031476075e-05,
"loss": 0.0676,
"step": 978
},
{
"epoch": 0.5970422320475682,
"grad_norm": 0.5439491868019104,
"learning_rate": 3.701134108851483e-05,
"loss": 0.0938,
"step": 979
},
{
"epoch": 0.5976520811099253,
"grad_norm": 0.478755921125412,
"learning_rate": 3.691596585275797e-05,
"loss": 0.1056,
"step": 980
},
{
"epoch": 0.5982619301722824,
"grad_norm": 0.6683340072631836,
"learning_rate": 3.682064169654663e-05,
"loss": 0.1018,
"step": 981
},
{
"epoch": 0.5988717792346394,
"grad_norm": 0.4465368390083313,
"learning_rate": 3.672536899202247e-05,
"loss": 0.0842,
"step": 982
},
{
"epoch": 0.5994816282969965,
"grad_norm": 0.4302210807800293,
"learning_rate": 3.663014811112634e-05,
"loss": 0.0838,
"step": 983
},
{
"epoch": 0.6000914773593535,
"grad_norm": 0.4858892261981964,
"learning_rate": 3.6534979425596743e-05,
"loss": 0.0733,
"step": 984
},
{
"epoch": 0.6007013264217106,
"grad_norm": 0.5444077849388123,
"learning_rate": 3.6439863306968395e-05,
"loss": 0.1004,
"step": 985
},
{
"epoch": 0.6013111754840677,
"grad_norm": 0.4020465612411499,
"learning_rate": 3.6344800126570844e-05,
"loss": 0.0581,
"step": 986
},
{
"epoch": 0.6019210245464247,
"grad_norm": 0.8379632830619812,
"learning_rate": 3.6249790255526915e-05,
"loss": 0.1026,
"step": 987
},
{
"epoch": 0.6025308736087819,
"grad_norm": 0.7171007990837097,
"learning_rate": 3.615483406475137e-05,
"loss": 0.13,
"step": 988
},
{
"epoch": 0.6031407226711389,
"grad_norm": 0.3926325738430023,
"learning_rate": 3.605993192494934e-05,
"loss": 0.0719,
"step": 989
},
{
"epoch": 0.603750571733496,
"grad_norm": 0.5274457931518555,
"learning_rate": 3.5965084206615014e-05,
"loss": 0.0746,
"step": 990
},
{
"epoch": 0.6043604207958531,
"grad_norm": 0.4578186869621277,
"learning_rate": 3.587029128003006e-05,
"loss": 0.1015,
"step": 991
},
{
"epoch": 0.6049702698582101,
"grad_norm": 0.41254541277885437,
"learning_rate": 3.577555351526229e-05,
"loss": 0.0791,
"step": 992
},
{
"epoch": 0.6055801189205672,
"grad_norm": 0.43652525544166565,
"learning_rate": 3.5680871282164144e-05,
"loss": 0.0686,
"step": 993
},
{
"epoch": 0.6061899679829242,
"grad_norm": 0.4345393776893616,
"learning_rate": 3.5586244950371247e-05,
"loss": 0.0874,
"step": 994
},
{
"epoch": 0.6067998170452813,
"grad_norm": 0.36000826954841614,
"learning_rate": 3.549167488930103e-05,
"loss": 0.0857,
"step": 995
},
{
"epoch": 0.6074096661076384,
"grad_norm": 0.6321271061897278,
"learning_rate": 3.539716146815122e-05,
"loss": 0.0809,
"step": 996
},
{
"epoch": 0.6080195151699954,
"grad_norm": 0.6639065742492676,
"learning_rate": 3.5302705055898425e-05,
"loss": 0.0858,
"step": 997
},
{
"epoch": 0.6086293642323525,
"grad_norm": 0.4803217947483063,
"learning_rate": 3.52083060212967e-05,
"loss": 0.0774,
"step": 998
},
{
"epoch": 0.6092392132947095,
"grad_norm": 0.7727973461151123,
"learning_rate": 3.5113964732876106e-05,
"loss": 0.1447,
"step": 999
},
{
"epoch": 0.6098490623570666,
"grad_norm": 0.5914996862411499,
"learning_rate": 3.5019681558941254e-05,
"loss": 0.0898,
"step": 1000
},
{
"epoch": 0.6104589114194237,
"grad_norm": 0.6371393799781799,
"learning_rate": 3.492545686756986e-05,
"loss": 0.1283,
"step": 1001
},
{
"epoch": 0.6110687604817807,
"grad_norm": 0.6423852443695068,
"learning_rate": 3.4831291026611365e-05,
"loss": 0.1087,
"step": 1002
},
{
"epoch": 0.6116786095441378,
"grad_norm": 0.6141225695610046,
"learning_rate": 3.473718440368544e-05,
"loss": 0.1261,
"step": 1003
},
{
"epoch": 0.612288458606495,
"grad_norm": 0.4114612936973572,
"learning_rate": 3.464313736618058e-05,
"loss": 0.0652,
"step": 1004
},
{
"epoch": 0.612898307668852,
"grad_norm": 0.5977374911308289,
"learning_rate": 3.4549150281252636e-05,
"loss": 0.1165,
"step": 1005
},
{
"epoch": 0.6135081567312091,
"grad_norm": 0.6169608235359192,
"learning_rate": 3.4455223515823446e-05,
"loss": 0.0817,
"step": 1006
},
{
"epoch": 0.6141180057935661,
"grad_norm": 0.4477900266647339,
"learning_rate": 3.4361357436579316e-05,
"loss": 0.0979,
"step": 1007
},
{
"epoch": 0.6147278548559232,
"grad_norm": 0.4506249725818634,
"learning_rate": 3.426755240996969e-05,
"loss": 0.0857,
"step": 1008
},
{
"epoch": 0.6153377039182802,
"grad_norm": 0.3944280743598938,
"learning_rate": 3.417380880220563e-05,
"loss": 0.0657,
"step": 1009
},
{
"epoch": 0.6159475529806373,
"grad_norm": 0.5278686285018921,
"learning_rate": 3.4080126979258406e-05,
"loss": 0.1111,
"step": 1010
},
{
"epoch": 0.6165574020429944,
"grad_norm": 0.5839405655860901,
"learning_rate": 3.3986507306858125e-05,
"loss": 0.0841,
"step": 1011
},
{
"epoch": 0.6171672511053514,
"grad_norm": 0.35346266627311707,
"learning_rate": 3.3892950150492234e-05,
"loss": 0.0724,
"step": 1012
},
{
"epoch": 0.6177771001677085,
"grad_norm": 0.6193497776985168,
"learning_rate": 3.379945587540414e-05,
"loss": 0.0818,
"step": 1013
},
{
"epoch": 0.6183869492300655,
"grad_norm": 0.5301856398582458,
"learning_rate": 3.3706024846591715e-05,
"loss": 0.0912,
"step": 1014
},
{
"epoch": 0.6189967982924226,
"grad_norm": 0.7754840850830078,
"learning_rate": 3.361265742880599e-05,
"loss": 0.0969,
"step": 1015
},
{
"epoch": 0.6196066473547797,
"grad_norm": 0.5716555714607239,
"learning_rate": 3.351935398654961e-05,
"loss": 0.1295,
"step": 1016
},
{
"epoch": 0.6202164964171367,
"grad_norm": 0.829879105091095,
"learning_rate": 3.342611488407549e-05,
"loss": 0.0972,
"step": 1017
},
{
"epoch": 0.6208263454794938,
"grad_norm": 0.4706299602985382,
"learning_rate": 3.3332940485385336e-05,
"loss": 0.0984,
"step": 1018
},
{
"epoch": 0.6214361945418508,
"grad_norm": 0.566028892993927,
"learning_rate": 3.323983115422827e-05,
"loss": 0.0865,
"step": 1019
},
{
"epoch": 0.622046043604208,
"grad_norm": 0.4814322590827942,
"learning_rate": 3.314678725409942e-05,
"loss": 0.0898,
"step": 1020
},
{
"epoch": 0.6226558926665651,
"grad_norm": 1.2092725038528442,
"learning_rate": 3.3053809148238426e-05,
"loss": 0.0919,
"step": 1021
},
{
"epoch": 0.6232657417289221,
"grad_norm": 0.6150869131088257,
"learning_rate": 3.296089719962809e-05,
"loss": 0.1157,
"step": 1022
},
{
"epoch": 0.6238755907912792,
"grad_norm": 0.5328996181488037,
"learning_rate": 3.286805177099293e-05,
"loss": 0.1031,
"step": 1023
},
{
"epoch": 0.6244854398536362,
"grad_norm": 0.6576485633850098,
"learning_rate": 3.2775273224797825e-05,
"loss": 0.096,
"step": 1024
},
{
"epoch": 0.6250952889159933,
"grad_norm": 0.6489903926849365,
"learning_rate": 3.268256192324647e-05,
"loss": 0.0781,
"step": 1025
},
{
"epoch": 0.6257051379783504,
"grad_norm": 0.8015328049659729,
"learning_rate": 3.258991822828007e-05,
"loss": 0.1349,
"step": 1026
},
{
"epoch": 0.6263149870407074,
"grad_norm": 0.6585468649864197,
"learning_rate": 3.249734250157592e-05,
"loss": 0.1104,
"step": 1027
},
{
"epoch": 0.6269248361030645,
"grad_norm": 0.42796945571899414,
"learning_rate": 3.240483510454594e-05,
"loss": 0.0777,
"step": 1028
},
{
"epoch": 0.6275346851654215,
"grad_norm": 0.8731417059898376,
"learning_rate": 3.231239639833531e-05,
"loss": 0.1192,
"step": 1029
},
{
"epoch": 0.6281445342277786,
"grad_norm": 0.5331026315689087,
"learning_rate": 3.222002674382103e-05,
"loss": 0.0815,
"step": 1030
},
{
"epoch": 0.6287543832901357,
"grad_norm": 0.5337620973587036,
"learning_rate": 3.212772650161056e-05,
"loss": 0.106,
"step": 1031
},
{
"epoch": 0.6293642323524927,
"grad_norm": 0.5502578616142273,
"learning_rate": 3.20354960320403e-05,
"loss": 0.0854,
"step": 1032
},
{
"epoch": 0.6299740814148498,
"grad_norm": 0.4488840401172638,
"learning_rate": 3.1943335695174365e-05,
"loss": 0.0733,
"step": 1033
},
{
"epoch": 0.6305839304772068,
"grad_norm": 0.43602392077445984,
"learning_rate": 3.185124585080299e-05,
"loss": 0.053,
"step": 1034
},
{
"epoch": 0.6311937795395639,
"grad_norm": 0.4926394522190094,
"learning_rate": 3.175922685844125e-05,
"loss": 0.0937,
"step": 1035
},
{
"epoch": 0.6318036286019211,
"grad_norm": 0.7370961904525757,
"learning_rate": 3.16672790773276e-05,
"loss": 0.1112,
"step": 1036
},
{
"epoch": 0.6324134776642781,
"grad_norm": 0.6153743267059326,
"learning_rate": 3.15754028664225e-05,
"loss": 0.1195,
"step": 1037
},
{
"epoch": 0.6330233267266352,
"grad_norm": 0.5078874230384827,
"learning_rate": 3.1483598584407006e-05,
"loss": 0.0812,
"step": 1038
},
{
"epoch": 0.6336331757889923,
"grad_norm": 0.5889832973480225,
"learning_rate": 3.1391866589681346e-05,
"loss": 0.0773,
"step": 1039
},
{
"epoch": 0.6342430248513493,
"grad_norm": 0.5485363602638245,
"learning_rate": 3.130020724036357e-05,
"loss": 0.1199,
"step": 1040
},
{
"epoch": 0.6348528739137064,
"grad_norm": 0.4648326635360718,
"learning_rate": 3.12086208942881e-05,
"loss": 0.0944,
"step": 1041
},
{
"epoch": 0.6354627229760634,
"grad_norm": 0.5292264223098755,
"learning_rate": 3.1117107909004364e-05,
"loss": 0.1093,
"step": 1042
},
{
"epoch": 0.6360725720384205,
"grad_norm": 0.6185947060585022,
"learning_rate": 3.10256686417754e-05,
"loss": 0.1058,
"step": 1043
},
{
"epoch": 0.6366824211007776,
"grad_norm": 0.6592523455619812,
"learning_rate": 3.093430344957643e-05,
"loss": 0.0842,
"step": 1044
},
{
"epoch": 0.6372922701631346,
"grad_norm": 0.4584198594093323,
"learning_rate": 3.084301268909353e-05,
"loss": 0.0892,
"step": 1045
},
{
"epoch": 0.6379021192254917,
"grad_norm": 0.42501208186149597,
"learning_rate": 3.075179671672216e-05,
"loss": 0.0853,
"step": 1046
},
{
"epoch": 0.6385119682878487,
"grad_norm": 0.5462825894355774,
"learning_rate": 3.0660655888565825e-05,
"loss": 0.0993,
"step": 1047
},
{
"epoch": 0.6391218173502058,
"grad_norm": 0.4431364834308624,
"learning_rate": 3.0569590560434666e-05,
"loss": 0.074,
"step": 1048
},
{
"epoch": 0.6397316664125628,
"grad_norm": 0.3417268991470337,
"learning_rate": 3.0478601087844096e-05,
"loss": 0.0616,
"step": 1049
},
{
"epoch": 0.6403415154749199,
"grad_norm": 0.3934859037399292,
"learning_rate": 3.038768782601335e-05,
"loss": 0.0638,
"step": 1050
},
{
"epoch": 0.640951364537277,
"grad_norm": 0.5025169253349304,
"learning_rate": 3.0296851129864168e-05,
"loss": 0.1082,
"step": 1051
},
{
"epoch": 0.6415612135996341,
"grad_norm": 0.7204434871673584,
"learning_rate": 3.0206091354019393e-05,
"loss": 0.0851,
"step": 1052
},
{
"epoch": 0.6421710626619912,
"grad_norm": 0.42415502667427063,
"learning_rate": 3.0115408852801535e-05,
"loss": 0.07,
"step": 1053
},
{
"epoch": 0.6427809117243483,
"grad_norm": 0.586955189704895,
"learning_rate": 3.0024803980231462e-05,
"loss": 0.1007,
"step": 1054
},
{
"epoch": 0.6433907607867053,
"grad_norm": 0.49578601121902466,
"learning_rate": 2.9934277090026964e-05,
"loss": 0.0988,
"step": 1055
},
{
"epoch": 0.6440006098490624,
"grad_norm": 0.4772571325302124,
"learning_rate": 2.98438285356014e-05,
"loss": 0.1009,
"step": 1056
},
{
"epoch": 0.6446104589114194,
"grad_norm": 0.34304940700531006,
"learning_rate": 2.97534586700623e-05,
"loss": 0.0668,
"step": 1057
},
{
"epoch": 0.6452203079737765,
"grad_norm": 1.111816167831421,
"learning_rate": 2.9663167846209998e-05,
"loss": 0.1169,
"step": 1058
},
{
"epoch": 0.6458301570361336,
"grad_norm": 0.41902774572372437,
"learning_rate": 2.9572956416536267e-05,
"loss": 0.073,
"step": 1059
},
{
"epoch": 0.6464400060984906,
"grad_norm": 0.6246191263198853,
"learning_rate": 2.94828247332229e-05,
"loss": 0.1135,
"step": 1060
},
{
"epoch": 0.6470498551608477,
"grad_norm": 0.41872212290763855,
"learning_rate": 2.9392773148140408e-05,
"loss": 0.0845,
"step": 1061
},
{
"epoch": 0.6476597042232047,
"grad_norm": 0.43361425399780273,
"learning_rate": 2.930280201284654e-05,
"loss": 0.0737,
"step": 1062
},
{
"epoch": 0.6482695532855618,
"grad_norm": 0.8737676739692688,
"learning_rate": 2.9212911678585043e-05,
"loss": 0.0884,
"step": 1063
},
{
"epoch": 0.6488794023479189,
"grad_norm": 0.9231886863708496,
"learning_rate": 2.912310249628415e-05,
"loss": 0.1001,
"step": 1064
},
{
"epoch": 0.6494892514102759,
"grad_norm": 0.42312076687812805,
"learning_rate": 2.9033374816555338e-05,
"loss": 0.0769,
"step": 1065
},
{
"epoch": 0.650099100472633,
"grad_norm": 0.48789942264556885,
"learning_rate": 2.894372898969186e-05,
"loss": 0.0781,
"step": 1066
},
{
"epoch": 0.6507089495349901,
"grad_norm": 0.5557985901832581,
"learning_rate": 2.885416536566744e-05,
"loss": 0.0638,
"step": 1067
},
{
"epoch": 0.6513187985973472,
"grad_norm": 0.4608156979084015,
"learning_rate": 2.876468429413487e-05,
"loss": 0.0693,
"step": 1068
},
{
"epoch": 0.6519286476597043,
"grad_norm": 0.37635406851768494,
"learning_rate": 2.8675286124424693e-05,
"loss": 0.0718,
"step": 1069
},
{
"epoch": 0.6525384967220613,
"grad_norm": 0.7597861289978027,
"learning_rate": 2.858597120554378e-05,
"loss": 0.1138,
"step": 1070
},
{
"epoch": 0.6531483457844184,
"grad_norm": 0.42954444885253906,
"learning_rate": 2.8496739886173995e-05,
"loss": 0.0701,
"step": 1071
},
{
"epoch": 0.6537581948467754,
"grad_norm": 0.5570069551467896,
"learning_rate": 2.8407592514670812e-05,
"loss": 0.0961,
"step": 1072
},
{
"epoch": 0.6543680439091325,
"grad_norm": 0.5138557553291321,
"learning_rate": 2.8318529439062035e-05,
"loss": 0.0996,
"step": 1073
},
{
"epoch": 0.6549778929714896,
"grad_norm": 0.48275506496429443,
"learning_rate": 2.822955100704634e-05,
"loss": 0.0823,
"step": 1074
},
{
"epoch": 0.6555877420338466,
"grad_norm": 0.5181817412376404,
"learning_rate": 2.8140657565991958e-05,
"loss": 0.0741,
"step": 1075
},
{
"epoch": 0.6561975910962037,
"grad_norm": 1.271902322769165,
"learning_rate": 2.805184946293532e-05,
"loss": 0.1112,
"step": 1076
},
{
"epoch": 0.6568074401585607,
"grad_norm": 0.7653577923774719,
"learning_rate": 2.7963127044579697e-05,
"loss": 0.0889,
"step": 1077
},
{
"epoch": 0.6574172892209178,
"grad_norm": 0.6154307723045349,
"learning_rate": 2.787449065729388e-05,
"loss": 0.1197,
"step": 1078
},
{
"epoch": 0.6580271382832749,
"grad_norm": 0.5012805461883545,
"learning_rate": 2.7785940647110763e-05,
"loss": 0.0871,
"step": 1079
},
{
"epoch": 0.6586369873456319,
"grad_norm": 0.5342831015586853,
"learning_rate": 2.7697477359726047e-05,
"loss": 0.0993,
"step": 1080
},
{
"epoch": 0.659246836407989,
"grad_norm": 0.7462201118469238,
"learning_rate": 2.7609101140496863e-05,
"loss": 0.0885,
"step": 1081
},
{
"epoch": 0.659856685470346,
"grad_norm": 0.4900333881378174,
"learning_rate": 2.752081233444041e-05,
"loss": 0.0688,
"step": 1082
},
{
"epoch": 0.6604665345327032,
"grad_norm": 1.1281187534332275,
"learning_rate": 2.743261128623269e-05,
"loss": 0.1572,
"step": 1083
},
{
"epoch": 0.6610763835950603,
"grad_norm": 0.521187424659729,
"learning_rate": 2.7344498340207053e-05,
"loss": 0.0927,
"step": 1084
},
{
"epoch": 0.6616862326574173,
"grad_norm": 0.6113362908363342,
"learning_rate": 2.72564738403529e-05,
"loss": 0.1127,
"step": 1085
},
{
"epoch": 0.6622960817197744,
"grad_norm": 0.5153065919876099,
"learning_rate": 2.716853813031435e-05,
"loss": 0.0638,
"step": 1086
},
{
"epoch": 0.6629059307821314,
"grad_norm": 0.4519743323326111,
"learning_rate": 2.708069155338892e-05,
"loss": 0.0907,
"step": 1087
},
{
"epoch": 0.6635157798444885,
"grad_norm": 0.413540244102478,
"learning_rate": 2.699293445252612e-05,
"loss": 0.0573,
"step": 1088
},
{
"epoch": 0.6641256289068456,
"grad_norm": 0.40838727355003357,
"learning_rate": 2.6905267170326143e-05,
"loss": 0.0755,
"step": 1089
},
{
"epoch": 0.6647354779692026,
"grad_norm": 0.4339296817779541,
"learning_rate": 2.6817690049038568e-05,
"loss": 0.0763,
"step": 1090
},
{
"epoch": 0.6653453270315597,
"grad_norm": 0.4559721350669861,
"learning_rate": 2.6730203430560947e-05,
"loss": 0.0863,
"step": 1091
},
{
"epoch": 0.6659551760939167,
"grad_norm": 0.5837572813034058,
"learning_rate": 2.6642807656437562e-05,
"loss": 0.057,
"step": 1092
},
{
"epoch": 0.6665650251562738,
"grad_norm": 0.8145636320114136,
"learning_rate": 2.6555503067858013e-05,
"loss": 0.1265,
"step": 1093
},
{
"epoch": 0.6671748742186309,
"grad_norm": 0.5761756300926208,
"learning_rate": 2.646829000565591e-05,
"loss": 0.0755,
"step": 1094
},
{
"epoch": 0.6677847232809879,
"grad_norm": 0.8050199747085571,
"learning_rate": 2.6381168810307533e-05,
"loss": 0.1228,
"step": 1095
},
{
"epoch": 0.668394572343345,
"grad_norm": 0.4040774703025818,
"learning_rate": 2.6294139821930597e-05,
"loss": 0.0631,
"step": 1096
},
{
"epoch": 0.669004421405702,
"grad_norm": 0.3464125394821167,
"learning_rate": 2.6207203380282747e-05,
"loss": 0.0714,
"step": 1097
},
{
"epoch": 0.6696142704680591,
"grad_norm": 0.4505070745944977,
"learning_rate": 2.6120359824760388e-05,
"loss": 0.0752,
"step": 1098
},
{
"epoch": 0.6702241195304163,
"grad_norm": 0.6385988593101501,
"learning_rate": 2.603360949439727e-05,
"loss": 0.0734,
"step": 1099
},
{
"epoch": 0.6708339685927733,
"grad_norm": 0.42046037316322327,
"learning_rate": 2.59469527278632e-05,
"loss": 0.0732,
"step": 1100
},
{
"epoch": 0.6714438176551304,
"grad_norm": 0.6027272343635559,
"learning_rate": 2.5860389863462765e-05,
"loss": 0.0936,
"step": 1101
},
{
"epoch": 0.6720536667174875,
"grad_norm": 0.6541715860366821,
"learning_rate": 2.5773921239133896e-05,
"loss": 0.1291,
"step": 1102
},
{
"epoch": 0.6726635157798445,
"grad_norm": 0.5477768182754517,
"learning_rate": 2.5687547192446647e-05,
"loss": 0.0818,
"step": 1103
},
{
"epoch": 0.6732733648422016,
"grad_norm": 0.6563355326652527,
"learning_rate": 2.5601268060601813e-05,
"loss": 0.0845,
"step": 1104
},
{
"epoch": 0.6738832139045586,
"grad_norm": 0.5657819509506226,
"learning_rate": 2.5515084180429716e-05,
"loss": 0.0905,
"step": 1105
},
{
"epoch": 0.6744930629669157,
"grad_norm": 0.5672662854194641,
"learning_rate": 2.542899588838875e-05,
"loss": 0.0791,
"step": 1106
},
{
"epoch": 0.6751029120292728,
"grad_norm": 0.46479132771492004,
"learning_rate": 2.5343003520564158e-05,
"loss": 0.0732,
"step": 1107
},
{
"epoch": 0.6757127610916298,
"grad_norm": 0.48830583691596985,
"learning_rate": 2.5257107412666708e-05,
"loss": 0.0984,
"step": 1108
},
{
"epoch": 0.6763226101539869,
"grad_norm": 0.9312662482261658,
"learning_rate": 2.5171307900031345e-05,
"loss": 0.099,
"step": 1109
},
{
"epoch": 0.6769324592163439,
"grad_norm": 0.7440546154975891,
"learning_rate": 2.508560531761597e-05,
"loss": 0.1067,
"step": 1110
},
{
"epoch": 0.677542308278701,
"grad_norm": 0.6724880933761597,
"learning_rate": 2.500000000000001e-05,
"loss": 0.0905,
"step": 1111
},
{
"epoch": 0.678152157341058,
"grad_norm": 0.6264297962188721,
"learning_rate": 2.4914492281383205e-05,
"loss": 0.0708,
"step": 1112
},
{
"epoch": 0.6787620064034151,
"grad_norm": 0.344620943069458,
"learning_rate": 2.4829082495584242e-05,
"loss": 0.0604,
"step": 1113
},
{
"epoch": 0.6793718554657722,
"grad_norm": 0.6882098913192749,
"learning_rate": 2.4743770976039544e-05,
"loss": 0.1171,
"step": 1114
},
{
"epoch": 0.6799817045281293,
"grad_norm": 0.5530173182487488,
"learning_rate": 2.465855805580185e-05,
"loss": 0.0809,
"step": 1115
},
{
"epoch": 0.6805915535904864,
"grad_norm": 0.5042785406112671,
"learning_rate": 2.4573444067538986e-05,
"loss": 0.0952,
"step": 1116
},
{
"epoch": 0.6812014026528435,
"grad_norm": 0.5109937191009521,
"learning_rate": 2.448842934353256e-05,
"loss": 0.0702,
"step": 1117
},
{
"epoch": 0.6818112517152005,
"grad_norm": 0.596640408039093,
"learning_rate": 2.440351421567663e-05,
"loss": 0.0852,
"step": 1118
},
{
"epoch": 0.6824211007775576,
"grad_norm": 0.584658145904541,
"learning_rate": 2.4318699015476493e-05,
"loss": 0.0963,
"step": 1119
},
{
"epoch": 0.6830309498399146,
"grad_norm": 0.5948860049247742,
"learning_rate": 2.4233984074047273e-05,
"loss": 0.0867,
"step": 1120
},
{
"epoch": 0.6836407989022717,
"grad_norm": 0.7172620296478271,
"learning_rate": 2.414936972211272e-05,
"loss": 0.1236,
"step": 1121
},
{
"epoch": 0.6842506479646288,
"grad_norm": 0.7591496706008911,
"learning_rate": 2.406485629000386e-05,
"loss": 0.1127,
"step": 1122
},
{
"epoch": 0.6848604970269858,
"grad_norm": 0.517919659614563,
"learning_rate": 2.3980444107657747e-05,
"loss": 0.1116,
"step": 1123
},
{
"epoch": 0.6854703460893429,
"grad_norm": 0.7248156070709229,
"learning_rate": 2.3896133504616185e-05,
"loss": 0.1026,
"step": 1124
},
{
"epoch": 0.6860801951516999,
"grad_norm": 0.4589289724826813,
"learning_rate": 2.3811924810024384e-05,
"loss": 0.0755,
"step": 1125
},
{
"epoch": 0.686690044214057,
"grad_norm": 0.4241478443145752,
"learning_rate": 2.3727818352629712e-05,
"loss": 0.0598,
"step": 1126
},
{
"epoch": 0.687299893276414,
"grad_norm": 0.6549184918403625,
"learning_rate": 2.3643814460780394e-05,
"loss": 0.0706,
"step": 1127
},
{
"epoch": 0.6879097423387711,
"grad_norm": 0.37013790011405945,
"learning_rate": 2.35599134624243e-05,
"loss": 0.0798,
"step": 1128
},
{
"epoch": 0.6885195914011282,
"grad_norm": 0.5305967330932617,
"learning_rate": 2.347611568510754e-05,
"loss": 0.0665,
"step": 1129
},
{
"epoch": 0.6891294404634852,
"grad_norm": 0.6524767875671387,
"learning_rate": 2.33924214559733e-05,
"loss": 0.0848,
"step": 1130
},
{
"epoch": 0.6897392895258424,
"grad_norm": 0.7240805625915527,
"learning_rate": 2.3308831101760486e-05,
"loss": 0.1027,
"step": 1131
},
{
"epoch": 0.6903491385881995,
"grad_norm": 0.4233236610889435,
"learning_rate": 2.322534494880248e-05,
"loss": 0.0922,
"step": 1132
},
{
"epoch": 0.6909589876505565,
"grad_norm": 0.36252397298812866,
"learning_rate": 2.3141963323025916e-05,
"loss": 0.0553,
"step": 1133
},
{
"epoch": 0.6915688367129136,
"grad_norm": 0.5424758791923523,
"learning_rate": 2.3058686549949306e-05,
"loss": 0.0779,
"step": 1134
},
{
"epoch": 0.6921786857752706,
"grad_norm": 0.45313170552253723,
"learning_rate": 2.2975514954681838e-05,
"loss": 0.0547,
"step": 1135
},
{
"epoch": 0.6927885348376277,
"grad_norm": 0.5699437856674194,
"learning_rate": 2.289244886192207e-05,
"loss": 0.1141,
"step": 1136
},
{
"epoch": 0.6933983838999848,
"grad_norm": 0.5247575044631958,
"learning_rate": 2.2809488595956745e-05,
"loss": 0.1114,
"step": 1137
},
{
"epoch": 0.6940082329623418,
"grad_norm": 0.6782296299934387,
"learning_rate": 2.2726634480659403e-05,
"loss": 0.113,
"step": 1138
},
{
"epoch": 0.6946180820246989,
"grad_norm": 0.5767406821250916,
"learning_rate": 2.264388683948918e-05,
"loss": 0.1008,
"step": 1139
},
{
"epoch": 0.6952279310870559,
"grad_norm": 0.46866175532341003,
"learning_rate": 2.2561245995489566e-05,
"loss": 0.0705,
"step": 1140
},
{
"epoch": 0.695837780149413,
"grad_norm": 0.5739036202430725,
"learning_rate": 2.247871227128709e-05,
"loss": 0.096,
"step": 1141
},
{
"epoch": 0.6964476292117701,
"grad_norm": 0.6201688051223755,
"learning_rate": 2.2396285989090134e-05,
"loss": 0.1022,
"step": 1142
},
{
"epoch": 0.6970574782741271,
"grad_norm": 0.6683262586593628,
"learning_rate": 2.2313967470687593e-05,
"loss": 0.1031,
"step": 1143
},
{
"epoch": 0.6976673273364842,
"grad_norm": 0.4578685760498047,
"learning_rate": 2.2231757037447664e-05,
"loss": 0.0768,
"step": 1144
},
{
"epoch": 0.6982771763988412,
"grad_norm": 0.3859054744243622,
"learning_rate": 2.2149655010316573e-05,
"loss": 0.0767,
"step": 1145
},
{
"epoch": 0.6988870254611984,
"grad_norm": 0.6152511239051819,
"learning_rate": 2.2067661709817383e-05,
"loss": 0.0851,
"step": 1146
},
{
"epoch": 0.6994968745235555,
"grad_norm": 0.6137005090713501,
"learning_rate": 2.1985777456048633e-05,
"loss": 0.1057,
"step": 1147
},
{
"epoch": 0.7001067235859125,
"grad_norm": 0.8141719698905945,
"learning_rate": 2.1904002568683173e-05,
"loss": 0.0833,
"step": 1148
},
{
"epoch": 0.7007165726482696,
"grad_norm": 0.5515589118003845,
"learning_rate": 2.1822337366966898e-05,
"loss": 0.1192,
"step": 1149
},
{
"epoch": 0.7013264217106266,
"grad_norm": 0.6049636006355286,
"learning_rate": 2.1740782169717455e-05,
"loss": 0.0937,
"step": 1150
},
{
"epoch": 0.7019362707729837,
"grad_norm": 0.4691774547100067,
"learning_rate": 2.1659337295323118e-05,
"loss": 0.0805,
"step": 1151
},
{
"epoch": 0.7025461198353408,
"grad_norm": 0.418965607881546,
"learning_rate": 2.1578003061741388e-05,
"loss": 0.0651,
"step": 1152
},
{
"epoch": 0.7031559688976978,
"grad_norm": 0.4960416555404663,
"learning_rate": 2.149677978649786e-05,
"loss": 0.0869,
"step": 1153
},
{
"epoch": 0.7037658179600549,
"grad_norm": 0.5811144709587097,
"learning_rate": 2.1415667786684952e-05,
"loss": 0.0839,
"step": 1154
},
{
"epoch": 0.7043756670224119,
"grad_norm": 0.4969620108604431,
"learning_rate": 2.1334667378960644e-05,
"loss": 0.0737,
"step": 1155
},
{
"epoch": 0.704985516084769,
"grad_norm": 0.7334999442100525,
"learning_rate": 2.125377887954732e-05,
"loss": 0.0735,
"step": 1156
},
{
"epoch": 0.7055953651471261,
"grad_norm": 0.9087291955947876,
"learning_rate": 2.1173002604230425e-05,
"loss": 0.1183,
"step": 1157
},
{
"epoch": 0.7062052142094831,
"grad_norm": 0.7997822761535645,
"learning_rate": 2.1092338868357302e-05,
"loss": 0.1012,
"step": 1158
},
{
"epoch": 0.7068150632718402,
"grad_norm": 0.43627968430519104,
"learning_rate": 2.1011787986835934e-05,
"loss": 0.0748,
"step": 1159
},
{
"epoch": 0.7074249123341972,
"grad_norm": 0.5334597826004028,
"learning_rate": 2.093135027413377e-05,
"loss": 0.1024,
"step": 1160
},
{
"epoch": 0.7080347613965543,
"grad_norm": 0.3994443714618683,
"learning_rate": 2.0851026044276406e-05,
"loss": 0.0586,
"step": 1161
},
{
"epoch": 0.7086446104589115,
"grad_norm": 0.4374444782733917,
"learning_rate": 2.0770815610846418e-05,
"loss": 0.0802,
"step": 1162
},
{
"epoch": 0.7092544595212685,
"grad_norm": 0.506949245929718,
"learning_rate": 2.0690719286982125e-05,
"loss": 0.0713,
"step": 1163
},
{
"epoch": 0.7098643085836256,
"grad_norm": 0.8575443029403687,
"learning_rate": 2.061073738537635e-05,
"loss": 0.1609,
"step": 1164
},
{
"epoch": 0.7104741576459827,
"grad_norm": 0.4683690369129181,
"learning_rate": 2.0530870218275273e-05,
"loss": 0.0838,
"step": 1165
},
{
"epoch": 0.7110840067083397,
"grad_norm": 0.6132832765579224,
"learning_rate": 2.0451118097477094e-05,
"loss": 0.073,
"step": 1166
},
{
"epoch": 0.7116938557706968,
"grad_norm": 0.4818894863128662,
"learning_rate": 2.0371481334330912e-05,
"loss": 0.0577,
"step": 1167
},
{
"epoch": 0.7123037048330538,
"grad_norm": 0.4278302788734436,
"learning_rate": 2.0291960239735436e-05,
"loss": 0.1042,
"step": 1168
},
{
"epoch": 0.7129135538954109,
"grad_norm": 0.7337779402732849,
"learning_rate": 2.0212555124137866e-05,
"loss": 0.0929,
"step": 1169
},
{
"epoch": 0.713523402957768,
"grad_norm": 0.4221004545688629,
"learning_rate": 2.0133266297532587e-05,
"loss": 0.0974,
"step": 1170
},
{
"epoch": 0.714133252020125,
"grad_norm": 1.0733360052108765,
"learning_rate": 2.005409406946e-05,
"loss": 0.1153,
"step": 1171
},
{
"epoch": 0.7147431010824821,
"grad_norm": 0.4175255000591278,
"learning_rate": 1.9975038749005314e-05,
"loss": 0.0741,
"step": 1172
},
{
"epoch": 0.7153529501448391,
"grad_norm": 0.34019866585731506,
"learning_rate": 1.9896100644797317e-05,
"loss": 0.0575,
"step": 1173
},
{
"epoch": 0.7159627992071962,
"grad_norm": 0.3449137210845947,
"learning_rate": 1.9817280065007244e-05,
"loss": 0.083,
"step": 1174
},
{
"epoch": 0.7165726482695532,
"grad_norm": 1.0228346586227417,
"learning_rate": 1.973857731734746e-05,
"loss": 0.0807,
"step": 1175
},
{
"epoch": 0.7171824973319103,
"grad_norm": 0.7103592753410339,
"learning_rate": 1.9659992709070345e-05,
"loss": 0.1338,
"step": 1176
},
{
"epoch": 0.7177923463942674,
"grad_norm": 0.7471181154251099,
"learning_rate": 1.958152654696705e-05,
"loss": 0.1055,
"step": 1177
},
{
"epoch": 0.7184021954566245,
"grad_norm": 1.3090336322784424,
"learning_rate": 1.950317913736636e-05,
"loss": 0.0644,
"step": 1178
},
{
"epoch": 0.7190120445189816,
"grad_norm": 0.452388197183609,
"learning_rate": 1.942495078613341e-05,
"loss": 0.0864,
"step": 1179
},
{
"epoch": 0.7196218935813387,
"grad_norm": 0.34744614362716675,
"learning_rate": 1.9346841798668553e-05,
"loss": 0.0563,
"step": 1180
},
{
"epoch": 0.7202317426436957,
"grad_norm": 0.5133278369903564,
"learning_rate": 1.9268852479906147e-05,
"loss": 0.0958,
"step": 1181
},
{
"epoch": 0.7208415917060528,
"grad_norm": 0.6480668783187866,
"learning_rate": 1.919098313431335e-05,
"loss": 0.1086,
"step": 1182
},
{
"epoch": 0.7214514407684098,
"grad_norm": 0.593155562877655,
"learning_rate": 1.9113234065889014e-05,
"loss": 0.0888,
"step": 1183
},
{
"epoch": 0.7220612898307669,
"grad_norm": 0.5548598170280457,
"learning_rate": 1.9035605578162364e-05,
"loss": 0.0842,
"step": 1184
},
{
"epoch": 0.722671138893124,
"grad_norm": 0.49532800912857056,
"learning_rate": 1.8958097974191907e-05,
"loss": 0.0822,
"step": 1185
},
{
"epoch": 0.723280987955481,
"grad_norm": 0.36185330152511597,
"learning_rate": 1.888071155656421e-05,
"loss": 0.0718,
"step": 1186
},
{
"epoch": 0.7238908370178381,
"grad_norm": 0.598893404006958,
"learning_rate": 1.8803446627392797e-05,
"loss": 0.0667,
"step": 1187
},
{
"epoch": 0.7245006860801951,
"grad_norm": 0.6224693059921265,
"learning_rate": 1.872630348831682e-05,
"loss": 0.113,
"step": 1188
},
{
"epoch": 0.7251105351425522,
"grad_norm": 0.6309583187103271,
"learning_rate": 1.8649282440500015e-05,
"loss": 0.0865,
"step": 1189
},
{
"epoch": 0.7257203842049093,
"grad_norm": 0.6631870269775391,
"learning_rate": 1.8572383784629477e-05,
"loss": 0.0892,
"step": 1190
},
{
"epoch": 0.7263302332672663,
"grad_norm": 0.5936716198921204,
"learning_rate": 1.849560782091445e-05,
"loss": 0.0895,
"step": 1191
},
{
"epoch": 0.7269400823296234,
"grad_norm": 0.46353256702423096,
"learning_rate": 1.8418954849085263e-05,
"loss": 0.083,
"step": 1192
},
{
"epoch": 0.7275499313919804,
"grad_norm": 0.5421329736709595,
"learning_rate": 1.834242516839203e-05,
"loss": 0.0696,
"step": 1193
},
{
"epoch": 0.7281597804543376,
"grad_norm": 0.6117755770683289,
"learning_rate": 1.826601907760357e-05,
"loss": 0.0749,
"step": 1194
},
{
"epoch": 0.7287696295166947,
"grad_norm": 0.5190672278404236,
"learning_rate": 1.8189736875006185e-05,
"loss": 0.0868,
"step": 1195
},
{
"epoch": 0.7293794785790517,
"grad_norm": 0.49801555275917053,
"learning_rate": 1.811357885840254e-05,
"loss": 0.0577,
"step": 1196
},
{
"epoch": 0.7299893276414088,
"grad_norm": 0.5652840733528137,
"learning_rate": 1.8037545325110504e-05,
"loss": 0.102,
"step": 1197
},
{
"epoch": 0.7305991767037658,
"grad_norm": 0.4611136019229889,
"learning_rate": 1.796163657196194e-05,
"loss": 0.0729,
"step": 1198
},
{
"epoch": 0.7312090257661229,
"grad_norm": 0.48652029037475586,
"learning_rate": 1.788585289530158e-05,
"loss": 0.0769,
"step": 1199
},
{
"epoch": 0.73181887482848,
"grad_norm": 0.5864943861961365,
"learning_rate": 1.781019459098584e-05,
"loss": 0.0826,
"step": 1200
},
{
"epoch": 0.732428723890837,
"grad_norm": 0.5369698405265808,
"learning_rate": 1.7734661954381754e-05,
"loss": 0.0959,
"step": 1201
},
{
"epoch": 0.7330385729531941,
"grad_norm": 0.6008253693580627,
"learning_rate": 1.7659255280365682e-05,
"loss": 0.0879,
"step": 1202
},
{
"epoch": 0.7336484220155511,
"grad_norm": 1.049195408821106,
"learning_rate": 1.7583974863322274e-05,
"loss": 0.0967,
"step": 1203
},
{
"epoch": 0.7342582710779082,
"grad_norm": 0.716066300868988,
"learning_rate": 1.7508820997143254e-05,
"loss": 0.117,
"step": 1204
},
{
"epoch": 0.7348681201402653,
"grad_norm": 1.32161283493042,
"learning_rate": 1.7433793975226298e-05,
"loss": 0.0952,
"step": 1205
},
{
"epoch": 0.7354779692026223,
"grad_norm": 0.44295260310173035,
"learning_rate": 1.7358894090473925e-05,
"loss": 0.087,
"step": 1206
},
{
"epoch": 0.7360878182649794,
"grad_norm": 0.5871211886405945,
"learning_rate": 1.728412163529227e-05,
"loss": 0.075,
"step": 1207
},
{
"epoch": 0.7366976673273364,
"grad_norm": 0.46307092905044556,
"learning_rate": 1.720947690159001e-05,
"loss": 0.0687,
"step": 1208
},
{
"epoch": 0.7373075163896935,
"grad_norm": 0.6493417620658875,
"learning_rate": 1.713496018077717e-05,
"loss": 0.1167,
"step": 1209
},
{
"epoch": 0.7379173654520507,
"grad_norm": 0.48898711800575256,
"learning_rate": 1.7060571763764087e-05,
"loss": 0.074,
"step": 1210
},
{
"epoch": 0.7385272145144077,
"grad_norm": 0.3575912117958069,
"learning_rate": 1.6986311940960147e-05,
"loss": 0.0511,
"step": 1211
},
{
"epoch": 0.7391370635767648,
"grad_norm": 0.40686553716659546,
"learning_rate": 1.6912181002272713e-05,
"loss": 0.0812,
"step": 1212
},
{
"epoch": 0.7397469126391218,
"grad_norm": 0.400678426027298,
"learning_rate": 1.6838179237106016e-05,
"loss": 0.087,
"step": 1213
},
{
"epoch": 0.7403567617014789,
"grad_norm": 0.6721731424331665,
"learning_rate": 1.6764306934359965e-05,
"loss": 0.1012,
"step": 1214
},
{
"epoch": 0.740966610763836,
"grad_norm": 0.5827667117118835,
"learning_rate": 1.66905643824291e-05,
"loss": 0.058,
"step": 1215
},
{
"epoch": 0.741576459826193,
"grad_norm": 0.5858330130577087,
"learning_rate": 1.661695186920138e-05,
"loss": 0.1007,
"step": 1216
},
{
"epoch": 0.7421863088885501,
"grad_norm": 0.4252987802028656,
"learning_rate": 1.6543469682057106e-05,
"loss": 0.0697,
"step": 1217
},
{
"epoch": 0.7427961579509071,
"grad_norm": 0.7781935334205627,
"learning_rate": 1.6470118107867778e-05,
"loss": 0.0955,
"step": 1218
},
{
"epoch": 0.7434060070132642,
"grad_norm": 0.569746732711792,
"learning_rate": 1.6396897432995044e-05,
"loss": 0.0951,
"step": 1219
},
{
"epoch": 0.7440158560756213,
"grad_norm": 0.6230528950691223,
"learning_rate": 1.6323807943289467e-05,
"loss": 0.0805,
"step": 1220
},
{
"epoch": 0.7446257051379783,
"grad_norm": 0.5333548784255981,
"learning_rate": 1.6250849924089484e-05,
"loss": 0.0758,
"step": 1221
},
{
"epoch": 0.7452355542003354,
"grad_norm": 0.899002730846405,
"learning_rate": 1.6178023660220294e-05,
"loss": 0.095,
"step": 1222
},
{
"epoch": 0.7458454032626924,
"grad_norm": 0.9524864554405212,
"learning_rate": 1.6105329435992682e-05,
"loss": 0.0939,
"step": 1223
},
{
"epoch": 0.7464552523250495,
"grad_norm": 0.5031396150588989,
"learning_rate": 1.6032767535202043e-05,
"loss": 0.0889,
"step": 1224
},
{
"epoch": 0.7470651013874067,
"grad_norm": 0.6588827967643738,
"learning_rate": 1.5960338241127093e-05,
"loss": 0.0871,
"step": 1225
},
{
"epoch": 0.7476749504497637,
"grad_norm": 0.49696049094200134,
"learning_rate": 1.5888041836528915e-05,
"loss": 0.0922,
"step": 1226
},
{
"epoch": 0.7482847995121208,
"grad_norm": 0.6000326871871948,
"learning_rate": 1.581587860364977e-05,
"loss": 0.1008,
"step": 1227
},
{
"epoch": 0.7488946485744779,
"grad_norm": 0.6604710817337036,
"learning_rate": 1.5743848824212014e-05,
"loss": 0.0783,
"step": 1228
},
{
"epoch": 0.7495044976368349,
"grad_norm": 0.7499123811721802,
"learning_rate": 1.567195277941706e-05,
"loss": 0.0821,
"step": 1229
},
{
"epoch": 0.750114346699192,
"grad_norm": 0.5564082264900208,
"learning_rate": 1.560019074994416e-05,
"loss": 0.0968,
"step": 1230
},
{
"epoch": 0.750724195761549,
"grad_norm": 0.5198761224746704,
"learning_rate": 1.552856301594942e-05,
"loss": 0.0858,
"step": 1231
},
{
"epoch": 0.7513340448239061,
"grad_norm": 0.44238951802253723,
"learning_rate": 1.5457069857064623e-05,
"loss": 0.0656,
"step": 1232
},
{
"epoch": 0.7519438938862631,
"grad_norm": 0.3478459417819977,
"learning_rate": 1.5385711552396227e-05,
"loss": 0.045,
"step": 1233
},
{
"epoch": 0.7525537429486202,
"grad_norm": 0.37783607840538025,
"learning_rate": 1.5314488380524182e-05,
"loss": 0.0798,
"step": 1234
},
{
"epoch": 0.7531635920109773,
"grad_norm": 0.4698028564453125,
"learning_rate": 1.5243400619500903e-05,
"loss": 0.0889,
"step": 1235
},
{
"epoch": 0.7537734410733343,
"grad_norm": 0.6131026744842529,
"learning_rate": 1.5172448546850165e-05,
"loss": 0.0777,
"step": 1236
},
{
"epoch": 0.7543832901356914,
"grad_norm": 0.529146134853363,
"learning_rate": 1.5101632439565998e-05,
"loss": 0.0715,
"step": 1237
},
{
"epoch": 0.7549931391980484,
"grad_norm": 0.7306383848190308,
"learning_rate": 1.503095257411169e-05,
"loss": 0.0638,
"step": 1238
},
{
"epoch": 0.7556029882604055,
"grad_norm": 0.9812265038490295,
"learning_rate": 1.4960409226418576e-05,
"loss": 0.0819,
"step": 1239
},
{
"epoch": 0.7562128373227626,
"grad_norm": 0.5744767189025879,
"learning_rate": 1.4890002671885072e-05,
"loss": 0.1196,
"step": 1240
},
{
"epoch": 0.7568226863851197,
"grad_norm": 0.5726638436317444,
"learning_rate": 1.4819733185375534e-05,
"loss": 0.1161,
"step": 1241
},
{
"epoch": 0.7574325354474768,
"grad_norm": 0.44419267773628235,
"learning_rate": 1.4749601041219247e-05,
"loss": 0.07,
"step": 1242
},
{
"epoch": 0.7580423845098339,
"grad_norm": 0.4185696244239807,
"learning_rate": 1.4679606513209283e-05,
"loss": 0.0562,
"step": 1243
},
{
"epoch": 0.7586522335721909,
"grad_norm": 0.6230916380882263,
"learning_rate": 1.4609749874601463e-05,
"loss": 0.1092,
"step": 1244
},
{
"epoch": 0.759262082634548,
"grad_norm": 0.43910908699035645,
"learning_rate": 1.4540031398113335e-05,
"loss": 0.0847,
"step": 1245
},
{
"epoch": 0.759871931696905,
"grad_norm": 0.4537461996078491,
"learning_rate": 1.4470451355923027e-05,
"loss": 0.0735,
"step": 1246
},
{
"epoch": 0.7604817807592621,
"grad_norm": 0.6280653476715088,
"learning_rate": 1.4401010019668226e-05,
"loss": 0.106,
"step": 1247
},
{
"epoch": 0.7610916298216192,
"grad_norm": 0.47639888525009155,
"learning_rate": 1.4331707660445155e-05,
"loss": 0.0727,
"step": 1248
},
{
"epoch": 0.7617014788839762,
"grad_norm": 0.5208837985992432,
"learning_rate": 1.4262544548807432e-05,
"loss": 0.0915,
"step": 1249
},
{
"epoch": 0.7623113279463333,
"grad_norm": 0.6564781069755554,
"learning_rate": 1.4193520954765083e-05,
"loss": 0.0971,
"step": 1250
},
{
"epoch": 0.7629211770086903,
"grad_norm": 0.4653925597667694,
"learning_rate": 1.4124637147783432e-05,
"loss": 0.091,
"step": 1251
},
{
"epoch": 0.7635310260710474,
"grad_norm": 0.6703091263771057,
"learning_rate": 1.4055893396782143e-05,
"loss": 0.087,
"step": 1252
},
{
"epoch": 0.7641408751334045,
"grad_norm": 0.6103700995445251,
"learning_rate": 1.3987289970134049e-05,
"loss": 0.0661,
"step": 1253
},
{
"epoch": 0.7647507241957615,
"grad_norm": 0.8154565691947937,
"learning_rate": 1.3918827135664187e-05,
"loss": 0.1014,
"step": 1254
},
{
"epoch": 0.7653605732581186,
"grad_norm": 0.4761587977409363,
"learning_rate": 1.3850505160648709e-05,
"loss": 0.0893,
"step": 1255
},
{
"epoch": 0.7659704223204756,
"grad_norm": 0.45273202657699585,
"learning_rate": 1.378232431181386e-05,
"loss": 0.0661,
"step": 1256
},
{
"epoch": 0.7665802713828328,
"grad_norm": 0.5135424733161926,
"learning_rate": 1.371428485533498e-05,
"loss": 0.0717,
"step": 1257
},
{
"epoch": 0.7671901204451899,
"grad_norm": 0.6998366713523865,
"learning_rate": 1.3646387056835368e-05,
"loss": 0.0879,
"step": 1258
},
{
"epoch": 0.7677999695075469,
"grad_norm": 0.754704475402832,
"learning_rate": 1.3578631181385305e-05,
"loss": 0.0685,
"step": 1259
},
{
"epoch": 0.768409818569904,
"grad_norm": 0.5335888266563416,
"learning_rate": 1.3511017493501005e-05,
"loss": 0.0781,
"step": 1260
},
{
"epoch": 0.769019667632261,
"grad_norm": 0.39155882596969604,
"learning_rate": 1.3443546257143624e-05,
"loss": 0.0561,
"step": 1261
},
{
"epoch": 0.7696295166946181,
"grad_norm": 0.335376113653183,
"learning_rate": 1.337621773571816e-05,
"loss": 0.0486,
"step": 1262
},
{
"epoch": 0.7702393657569752,
"grad_norm": 0.47816699743270874,
"learning_rate": 1.3309032192072463e-05,
"loss": 0.0707,
"step": 1263
},
{
"epoch": 0.7708492148193322,
"grad_norm": 0.7821969389915466,
"learning_rate": 1.3241989888496204e-05,
"loss": 0.0677,
"step": 1264
},
{
"epoch": 0.7714590638816893,
"grad_norm": 0.4009250998497009,
"learning_rate": 1.3175091086719832e-05,
"loss": 0.0673,
"step": 1265
},
{
"epoch": 0.7720689129440463,
"grad_norm": 0.7203315496444702,
"learning_rate": 1.3108336047913633e-05,
"loss": 0.1077,
"step": 1266
},
{
"epoch": 0.7726787620064034,
"grad_norm": 0.48224198818206787,
"learning_rate": 1.304172503268658e-05,
"loss": 0.0652,
"step": 1267
},
{
"epoch": 0.7732886110687605,
"grad_norm": 0.8973504900932312,
"learning_rate": 1.297525830108542e-05,
"loss": 0.0994,
"step": 1268
},
{
"epoch": 0.7738984601311175,
"grad_norm": 0.6743061542510986,
"learning_rate": 1.29089361125936e-05,
"loss": 0.0814,
"step": 1269
},
{
"epoch": 0.7745083091934746,
"grad_norm": 0.7337107062339783,
"learning_rate": 1.2842758726130283e-05,
"loss": 0.1032,
"step": 1270
},
{
"epoch": 0.7751181582558316,
"grad_norm": 0.7888491749763489,
"learning_rate": 1.277672640004936e-05,
"loss": 0.0704,
"step": 1271
},
{
"epoch": 0.7757280073181887,
"grad_norm": 0.5230604410171509,
"learning_rate": 1.2710839392138386e-05,
"loss": 0.0974,
"step": 1272
},
{
"epoch": 0.7763378563805459,
"grad_norm": 0.70344477891922,
"learning_rate": 1.2645097959617585e-05,
"loss": 0.0882,
"step": 1273
},
{
"epoch": 0.7769477054429029,
"grad_norm": 0.7866513133049011,
"learning_rate": 1.2579502359138872e-05,
"loss": 0.08,
"step": 1274
},
{
"epoch": 0.77755755450526,
"grad_norm": 0.7914690971374512,
"learning_rate": 1.251405284678488e-05,
"loss": 0.1021,
"step": 1275
},
{
"epoch": 0.778167403567617,
"grad_norm": 1.0091532468795776,
"learning_rate": 1.2448749678067856e-05,
"loss": 0.1226,
"step": 1276
},
{
"epoch": 0.7787772526299741,
"grad_norm": 0.4124462902545929,
"learning_rate": 1.238359310792877e-05,
"loss": 0.0734,
"step": 1277
},
{
"epoch": 0.7793871016923312,
"grad_norm": 0.6571420431137085,
"learning_rate": 1.2318583390736254e-05,
"loss": 0.0716,
"step": 1278
},
{
"epoch": 0.7799969507546882,
"grad_norm": 0.7664232850074768,
"learning_rate": 1.2253720780285639e-05,
"loss": 0.0882,
"step": 1279
},
{
"epoch": 0.7806067998170453,
"grad_norm": 0.4563249945640564,
"learning_rate": 1.218900552979797e-05,
"loss": 0.104,
"step": 1280
},
{
"epoch": 0.7812166488794023,
"grad_norm": 0.4493783414363861,
"learning_rate": 1.2124437891918993e-05,
"loss": 0.0701,
"step": 1281
},
{
"epoch": 0.7818264979417594,
"grad_norm": 0.5224953889846802,
"learning_rate": 1.206001811871818e-05,
"loss": 0.0646,
"step": 1282
},
{
"epoch": 0.7824363470041165,
"grad_norm": 0.7538795471191406,
"learning_rate": 1.1995746461687734e-05,
"loss": 0.119,
"step": 1283
},
{
"epoch": 0.7830461960664735,
"grad_norm": 0.6143316030502319,
"learning_rate": 1.1931623171741652e-05,
"loss": 0.0882,
"step": 1284
},
{
"epoch": 0.7836560451288306,
"grad_norm": 0.49068811535835266,
"learning_rate": 1.186764849921468e-05,
"loss": 0.0671,
"step": 1285
},
{
"epoch": 0.7842658941911876,
"grad_norm": 0.5120086073875427,
"learning_rate": 1.1803822693861378e-05,
"loss": 0.0895,
"step": 1286
},
{
"epoch": 0.7848757432535447,
"grad_norm": 0.9712346196174622,
"learning_rate": 1.174014600485514e-05,
"loss": 0.1107,
"step": 1287
},
{
"epoch": 0.7854855923159018,
"grad_norm": 0.47291091084480286,
"learning_rate": 1.1676618680787189e-05,
"loss": 0.1075,
"step": 1288
},
{
"epoch": 0.7860954413782589,
"grad_norm": 0.5939892530441284,
"learning_rate": 1.1613240969665685e-05,
"loss": 0.1104,
"step": 1289
},
{
"epoch": 0.786705290440616,
"grad_norm": 0.45562052726745605,
"learning_rate": 1.1550013118914666e-05,
"loss": 0.0645,
"step": 1290
},
{
"epoch": 0.787315139502973,
"grad_norm": 0.5815163850784302,
"learning_rate": 1.1486935375373126e-05,
"loss": 0.1086,
"step": 1291
},
{
"epoch": 0.7879249885653301,
"grad_norm": 0.43459656834602356,
"learning_rate": 1.1424007985294032e-05,
"loss": 0.0647,
"step": 1292
},
{
"epoch": 0.7885348376276872,
"grad_norm": 0.4165402948856354,
"learning_rate": 1.1361231194343436e-05,
"loss": 0.0594,
"step": 1293
},
{
"epoch": 0.7891446866900442,
"grad_norm": 0.562037467956543,
"learning_rate": 1.1298605247599392e-05,
"loss": 0.0707,
"step": 1294
},
{
"epoch": 0.7897545357524013,
"grad_norm": 0.5341572761535645,
"learning_rate": 1.1236130389551092e-05,
"loss": 0.0769,
"step": 1295
},
{
"epoch": 0.7903643848147583,
"grad_norm": 0.7066879868507385,
"learning_rate": 1.1173806864097886e-05,
"loss": 0.0938,
"step": 1296
},
{
"epoch": 0.7909742338771154,
"grad_norm": 0.5032780766487122,
"learning_rate": 1.1111634914548297e-05,
"loss": 0.0884,
"step": 1297
},
{
"epoch": 0.7915840829394725,
"grad_norm": 0.5331841707229614,
"learning_rate": 1.1049614783619162e-05,
"loss": 0.1095,
"step": 1298
},
{
"epoch": 0.7921939320018295,
"grad_norm": 0.35556507110595703,
"learning_rate": 1.0987746713434576e-05,
"loss": 0.0779,
"step": 1299
},
{
"epoch": 0.7928037810641866,
"grad_norm": 0.35171961784362793,
"learning_rate": 1.0926030945525007e-05,
"loss": 0.065,
"step": 1300
},
{
"epoch": 0.7934136301265436,
"grad_norm": 0.5984443426132202,
"learning_rate": 1.0864467720826343e-05,
"loss": 0.1008,
"step": 1301
},
{
"epoch": 0.7940234791889007,
"grad_norm": 0.6116414070129395,
"learning_rate": 1.080305727967893e-05,
"loss": 0.1122,
"step": 1302
},
{
"epoch": 0.7946333282512578,
"grad_norm": 0.6918548345565796,
"learning_rate": 1.0741799861826706e-05,
"loss": 0.0982,
"step": 1303
},
{
"epoch": 0.7952431773136149,
"grad_norm": 0.4391896724700928,
"learning_rate": 1.0680695706416161e-05,
"loss": 0.0707,
"step": 1304
},
{
"epoch": 0.795853026375972,
"grad_norm": 0.38541465997695923,
"learning_rate": 1.0619745051995472e-05,
"loss": 0.075,
"step": 1305
},
{
"epoch": 0.796462875438329,
"grad_norm": 0.668455958366394,
"learning_rate": 1.0558948136513535e-05,
"loss": 0.0855,
"step": 1306
},
{
"epoch": 0.7970727245006861,
"grad_norm": 0.4920218586921692,
"learning_rate": 1.0498305197319115e-05,
"loss": 0.0799,
"step": 1307
},
{
"epoch": 0.7976825735630432,
"grad_norm": 0.8307283520698547,
"learning_rate": 1.0437816471159789e-05,
"loss": 0.1091,
"step": 1308
},
{
"epoch": 0.7982924226254002,
"grad_norm": 0.6896660923957825,
"learning_rate": 1.0377482194181132e-05,
"loss": 0.0951,
"step": 1309
},
{
"epoch": 0.7989022716877573,
"grad_norm": 0.6967519521713257,
"learning_rate": 1.0317302601925743e-05,
"loss": 0.1083,
"step": 1310
},
{
"epoch": 0.7995121207501144,
"grad_norm": 0.48772314190864563,
"learning_rate": 1.0257277929332332e-05,
"loss": 0.0863,
"step": 1311
},
{
"epoch": 0.8001219698124714,
"grad_norm": 0.44613274931907654,
"learning_rate": 1.0197408410734838e-05,
"loss": 0.0783,
"step": 1312
},
{
"epoch": 0.8007318188748285,
"grad_norm": 0.755577027797699,
"learning_rate": 1.0137694279861454e-05,
"loss": 0.0889,
"step": 1313
},
{
"epoch": 0.8013416679371855,
"grad_norm": 0.5232129693031311,
"learning_rate": 1.0078135769833758e-05,
"loss": 0.0682,
"step": 1314
},
{
"epoch": 0.8019515169995426,
"grad_norm": 0.4053167402744293,
"learning_rate": 1.0018733113165773e-05,
"loss": 0.0751,
"step": 1315
},
{
"epoch": 0.8025613660618997,
"grad_norm": 0.6713986992835999,
"learning_rate": 9.959486541763119e-06,
"loss": 0.0743,
"step": 1316
},
{
"epoch": 0.8031712151242567,
"grad_norm": 0.4893331825733185,
"learning_rate": 9.900396286922026e-06,
"loss": 0.0803,
"step": 1317
},
{
"epoch": 0.8037810641866138,
"grad_norm": 0.4546436071395874,
"learning_rate": 9.841462579328486e-06,
"loss": 0.0764,
"step": 1318
},
{
"epoch": 0.8043909132489708,
"grad_norm": 0.41673123836517334,
"learning_rate": 9.782685649057333e-06,
"loss": 0.0756,
"step": 1319
},
{
"epoch": 0.805000762311328,
"grad_norm": 0.5227165818214417,
"learning_rate": 9.72406572557133e-06,
"loss": 0.0685,
"step": 1320
},
{
"epoch": 0.8056106113736851,
"grad_norm": 0.4759485423564911,
"learning_rate": 9.66560303772035e-06,
"loss": 0.0905,
"step": 1321
},
{
"epoch": 0.8062204604360421,
"grad_norm": 0.4120442271232605,
"learning_rate": 9.607297813740362e-06,
"loss": 0.0557,
"step": 1322
},
{
"epoch": 0.8068303094983992,
"grad_norm": 0.7001487016677856,
"learning_rate": 9.549150281252633e-06,
"loss": 0.1053,
"step": 1323
},
{
"epoch": 0.8074401585607562,
"grad_norm": 0.37015411257743835,
"learning_rate": 9.491160667262783e-06,
"loss": 0.0545,
"step": 1324
},
{
"epoch": 0.8080500076231133,
"grad_norm": 0.5629733800888062,
"learning_rate": 9.433329198159974e-06,
"loss": 0.1014,
"step": 1325
},
{
"epoch": 0.8086598566854704,
"grad_norm": 0.5162930488586426,
"learning_rate": 9.375656099715934e-06,
"loss": 0.08,
"step": 1326
},
{
"epoch": 0.8092697057478274,
"grad_norm": 0.7029132843017578,
"learning_rate": 9.31814159708413e-06,
"loss": 0.0848,
"step": 1327
},
{
"epoch": 0.8098795548101845,
"grad_norm": 0.40555059909820557,
"learning_rate": 9.26078591479887e-06,
"loss": 0.0879,
"step": 1328
},
{
"epoch": 0.8104894038725415,
"grad_norm": 0.49707451462745667,
"learning_rate": 9.203589276774439e-06,
"loss": 0.0836,
"step": 1329
},
{
"epoch": 0.8110992529348986,
"grad_norm": 0.8126875758171082,
"learning_rate": 9.146551906304241e-06,
"loss": 0.0971,
"step": 1330
},
{
"epoch": 0.8117091019972557,
"grad_norm": 0.472047358751297,
"learning_rate": 9.08967402605988e-06,
"loss": 0.0752,
"step": 1331
},
{
"epoch": 0.8123189510596127,
"grad_norm": 0.5368672609329224,
"learning_rate": 9.03295585809032e-06,
"loss": 0.0871,
"step": 1332
},
{
"epoch": 0.8129288001219698,
"grad_norm": 0.3740680515766144,
"learning_rate": 8.976397623821003e-06,
"loss": 0.0788,
"step": 1333
},
{
"epoch": 0.8135386491843268,
"grad_norm": 0.4937654435634613,
"learning_rate": 8.919999544053036e-06,
"loss": 0.0635,
"step": 1334
},
{
"epoch": 0.8141484982466839,
"grad_norm": 0.6364428997039795,
"learning_rate": 8.86376183896226e-06,
"loss": 0.0753,
"step": 1335
},
{
"epoch": 0.8147583473090411,
"grad_norm": 0.5690184831619263,
"learning_rate": 8.80768472809842e-06,
"loss": 0.0796,
"step": 1336
},
{
"epoch": 0.8153681963713981,
"grad_norm": 0.4338169991970062,
"learning_rate": 8.751768430384305e-06,
"loss": 0.0646,
"step": 1337
},
{
"epoch": 0.8159780454337552,
"grad_norm": 0.39896783232688904,
"learning_rate": 8.696013164114902e-06,
"loss": 0.083,
"step": 1338
},
{
"epoch": 0.8165878944961122,
"grad_norm": 0.5548663139343262,
"learning_rate": 8.640419146956557e-06,
"loss": 0.0937,
"step": 1339
},
{
"epoch": 0.8171977435584693,
"grad_norm": 0.4257078766822815,
"learning_rate": 8.584986595946071e-06,
"loss": 0.0661,
"step": 1340
},
{
"epoch": 0.8178075926208264,
"grad_norm": 0.46818113327026367,
"learning_rate": 8.529715727489912e-06,
"loss": 0.0775,
"step": 1341
},
{
"epoch": 0.8184174416831834,
"grad_norm": 0.568156898021698,
"learning_rate": 8.474606757363334e-06,
"loss": 0.074,
"step": 1342
},
{
"epoch": 0.8190272907455405,
"grad_norm": 0.822402834892273,
"learning_rate": 8.419659900709536e-06,
"loss": 0.074,
"step": 1343
},
{
"epoch": 0.8196371398078975,
"grad_norm": 0.47720807790756226,
"learning_rate": 8.364875372038876e-06,
"loss": 0.077,
"step": 1344
},
{
"epoch": 0.8202469888702546,
"grad_norm": 0.6677123308181763,
"learning_rate": 8.310253385227946e-06,
"loss": 0.0839,
"step": 1345
},
{
"epoch": 0.8208568379326117,
"grad_norm": 0.5275459885597229,
"learning_rate": 8.255794153518798e-06,
"loss": 0.1045,
"step": 1346
},
{
"epoch": 0.8214666869949687,
"grad_norm": 0.9287574291229248,
"learning_rate": 8.201497889518073e-06,
"loss": 0.1235,
"step": 1347
},
{
"epoch": 0.8220765360573258,
"grad_norm": 0.5663622617721558,
"learning_rate": 8.147364805196239e-06,
"loss": 0.0974,
"step": 1348
},
{
"epoch": 0.8226863851196828,
"grad_norm": 0.45999595522880554,
"learning_rate": 8.093395111886687e-06,
"loss": 0.1135,
"step": 1349
},
{
"epoch": 0.8232962341820399,
"grad_norm": 0.37675905227661133,
"learning_rate": 8.039589020284926e-06,
"loss": 0.0614,
"step": 1350
},
{
"epoch": 0.823906083244397,
"grad_norm": 0.6055320501327515,
"learning_rate": 7.985946740447791e-06,
"loss": 0.095,
"step": 1351
},
{
"epoch": 0.8245159323067541,
"grad_norm": 0.5386014580726624,
"learning_rate": 7.932468481792582e-06,
"loss": 0.093,
"step": 1352
},
{
"epoch": 0.8251257813691112,
"grad_norm": 0.698286235332489,
"learning_rate": 7.879154453096304e-06,
"loss": 0.0853,
"step": 1353
},
{
"epoch": 0.8257356304314682,
"grad_norm": 0.48650923371315,
"learning_rate": 7.826004862494773e-06,
"loss": 0.0918,
"step": 1354
},
{
"epoch": 0.8263454794938253,
"grad_norm": 0.35784631967544556,
"learning_rate": 7.773019917481872e-06,
"loss": 0.0769,
"step": 1355
},
{
"epoch": 0.8269553285561824,
"grad_norm": 0.7131590247154236,
"learning_rate": 7.720199824908692e-06,
"loss": 0.0943,
"step": 1356
},
{
"epoch": 0.8275651776185394,
"grad_norm": 0.4979296624660492,
"learning_rate": 7.667544790982778e-06,
"loss": 0.0788,
"step": 1357
},
{
"epoch": 0.8281750266808965,
"grad_norm": 0.7505434155464172,
"learning_rate": 7.615055021267265e-06,
"loss": 0.1067,
"step": 1358
},
{
"epoch": 0.8287848757432535,
"grad_norm": 0.43658870458602905,
"learning_rate": 7.562730720680112e-06,
"loss": 0.0536,
"step": 1359
},
{
"epoch": 0.8293947248056106,
"grad_norm": 0.49229925870895386,
"learning_rate": 7.510572093493295e-06,
"loss": 0.0595,
"step": 1360
},
{
"epoch": 0.8300045738679677,
"grad_norm": 0.5588842034339905,
"learning_rate": 7.458579343331995e-06,
"loss": 0.0774,
"step": 1361
},
{
"epoch": 0.8306144229303247,
"grad_norm": 0.43612658977508545,
"learning_rate": 7.40675267317385e-06,
"loss": 0.0743,
"step": 1362
},
{
"epoch": 0.8312242719926818,
"grad_norm": 0.4666178524494171,
"learning_rate": 7.3550922853480915e-06,
"loss": 0.0695,
"step": 1363
},
{
"epoch": 0.8318341210550388,
"grad_norm": 0.37131235003471375,
"learning_rate": 7.30359838153481e-06,
"loss": 0.0668,
"step": 1364
},
{
"epoch": 0.8324439701173959,
"grad_norm": 0.4440286457538605,
"learning_rate": 7.252271162764129e-06,
"loss": 0.0681,
"step": 1365
},
{
"epoch": 0.833053819179753,
"grad_norm": 0.686726450920105,
"learning_rate": 7.2011108294154804e-06,
"loss": 0.0805,
"step": 1366
},
{
"epoch": 0.83366366824211,
"grad_norm": 0.6650717854499817,
"learning_rate": 7.150117581216748e-06,
"loss": 0.0915,
"step": 1367
},
{
"epoch": 0.8342735173044672,
"grad_norm": 0.4983760118484497,
"learning_rate": 7.099291617243526e-06,
"loss": 0.0634,
"step": 1368
},
{
"epoch": 0.8348833663668243,
"grad_norm": 0.45572003722190857,
"learning_rate": 7.048633135918347e-06,
"loss": 0.0811,
"step": 1369
},
{
"epoch": 0.8354932154291813,
"grad_norm": 0.511346161365509,
"learning_rate": 6.998142335009883e-06,
"loss": 0.0954,
"step": 1370
},
{
"epoch": 0.8361030644915384,
"grad_norm": 0.7270820140838623,
"learning_rate": 6.947819411632223e-06,
"loss": 0.0797,
"step": 1371
},
{
"epoch": 0.8367129135538954,
"grad_norm": 0.535539984703064,
"learning_rate": 6.897664562244027e-06,
"loss": 0.0564,
"step": 1372
},
{
"epoch": 0.8373227626162525,
"grad_norm": 0.7058387398719788,
"learning_rate": 6.8476779826478265e-06,
"loss": 0.1044,
"step": 1373
},
{
"epoch": 0.8379326116786096,
"grad_norm": 0.4986245632171631,
"learning_rate": 6.797859867989226e-06,
"loss": 0.0743,
"step": 1374
},
{
"epoch": 0.8385424607409666,
"grad_norm": 0.6708939075469971,
"learning_rate": 6.748210412756134e-06,
"loss": 0.0975,
"step": 1375
},
{
"epoch": 0.8391523098033237,
"grad_norm": 0.41141071915626526,
"learning_rate": 6.698729810778065e-06,
"loss": 0.0622,
"step": 1376
},
{
"epoch": 0.8397621588656807,
"grad_norm": 0.47120988368988037,
"learning_rate": 6.649418255225298e-06,
"loss": 0.0725,
"step": 1377
},
{
"epoch": 0.8403720079280378,
"grad_norm": 0.5629366040229797,
"learning_rate": 6.600275938608164e-06,
"loss": 0.0768,
"step": 1378
},
{
"epoch": 0.8409818569903948,
"grad_norm": 0.47137877345085144,
"learning_rate": 6.551303052776292e-06,
"loss": 0.0686,
"step": 1379
},
{
"epoch": 0.8415917060527519,
"grad_norm": 0.47057226300239563,
"learning_rate": 6.502499788917893e-06,
"loss": 0.0613,
"step": 1380
},
{
"epoch": 0.842201555115109,
"grad_norm": 0.607197642326355,
"learning_rate": 6.45386633755894e-06,
"loss": 0.0619,
"step": 1381
},
{
"epoch": 0.842811404177466,
"grad_norm": 0.5288928151130676,
"learning_rate": 6.405402888562484e-06,
"loss": 0.0684,
"step": 1382
},
{
"epoch": 0.8434212532398232,
"grad_norm": 0.5057141184806824,
"learning_rate": 6.357109631127889e-06,
"loss": 0.0934,
"step": 1383
},
{
"epoch": 0.8440311023021803,
"grad_norm": 0.48722875118255615,
"learning_rate": 6.308986753790081e-06,
"loss": 0.064,
"step": 1384
},
{
"epoch": 0.8446409513645373,
"grad_norm": 0.40020257234573364,
"learning_rate": 6.261034444418879e-06,
"loss": 0.0704,
"step": 1385
},
{
"epoch": 0.8452508004268944,
"grad_norm": 0.5671085715293884,
"learning_rate": 6.213252890218163e-06,
"loss": 0.0909,
"step": 1386
},
{
"epoch": 0.8458606494892514,
"grad_norm": 0.8954089283943176,
"learning_rate": 6.165642277725203e-06,
"loss": 0.0839,
"step": 1387
},
{
"epoch": 0.8464704985516085,
"grad_norm": 0.5945757031440735,
"learning_rate": 6.118202792809924e-06,
"loss": 0.0858,
"step": 1388
},
{
"epoch": 0.8470803476139656,
"grad_norm": 0.9332445859909058,
"learning_rate": 6.07093462067419e-06,
"loss": 0.094,
"step": 1389
},
{
"epoch": 0.8476901966763226,
"grad_norm": 0.5933021903038025,
"learning_rate": 6.023837945851041e-06,
"loss": 0.0954,
"step": 1390
},
{
"epoch": 0.8483000457386797,
"grad_norm": 0.39101532101631165,
"learning_rate": 5.976912952204017e-06,
"loss": 0.0741,
"step": 1391
},
{
"epoch": 0.8489098948010367,
"grad_norm": 0.5083923935890198,
"learning_rate": 5.9301598229264064e-06,
"loss": 0.0907,
"step": 1392
},
{
"epoch": 0.8495197438633938,
"grad_norm": 0.6381988525390625,
"learning_rate": 5.883578740540546e-06,
"loss": 0.0884,
"step": 1393
},
{
"epoch": 0.8501295929257509,
"grad_norm": 0.5870521068572998,
"learning_rate": 5.837169886897143e-06,
"loss": 0.0648,
"step": 1394
},
{
"epoch": 0.8507394419881079,
"grad_norm": 1.0675225257873535,
"learning_rate": 5.79093344317449e-06,
"loss": 0.0964,
"step": 1395
},
{
"epoch": 0.851349291050465,
"grad_norm": 0.403812050819397,
"learning_rate": 5.74486958987781e-06,
"loss": 0.0764,
"step": 1396
},
{
"epoch": 0.851959140112822,
"grad_norm": 0.48853600025177,
"learning_rate": 5.698978506838532e-06,
"loss": 0.0967,
"step": 1397
},
{
"epoch": 0.8525689891751791,
"grad_norm": 0.904900848865509,
"learning_rate": 5.653260373213632e-06,
"loss": 0.103,
"step": 1398
},
{
"epoch": 0.8531788382375363,
"grad_norm": 0.6909879446029663,
"learning_rate": 5.607715367484861e-06,
"loss": 0.0728,
"step": 1399
},
{
"epoch": 0.8537886872998933,
"grad_norm": 0.6291486024856567,
"learning_rate": 5.562343667458098e-06,
"loss": 0.0858,
"step": 1400
},
{
"epoch": 0.8543985363622504,
"grad_norm": 0.747890293598175,
"learning_rate": 5.51714545026264e-06,
"loss": 0.0913,
"step": 1401
},
{
"epoch": 0.8550083854246074,
"grad_norm": 0.46804454922676086,
"learning_rate": 5.472120892350513e-06,
"loss": 0.067,
"step": 1402
},
{
"epoch": 0.8556182344869645,
"grad_norm": 0.6729840040206909,
"learning_rate": 5.4272701694958076e-06,
"loss": 0.0967,
"step": 1403
},
{
"epoch": 0.8562280835493216,
"grad_norm": 0.5981701016426086,
"learning_rate": 5.382593456793933e-06,
"loss": 0.0991,
"step": 1404
},
{
"epoch": 0.8568379326116786,
"grad_norm": 1.0369232892990112,
"learning_rate": 5.338090928660999e-06,
"loss": 0.1206,
"step": 1405
},
{
"epoch": 0.8574477816740357,
"grad_norm": 0.35695895552635193,
"learning_rate": 5.293762758833071e-06,
"loss": 0.089,
"step": 1406
},
{
"epoch": 0.8580576307363927,
"grad_norm": 0.5548967719078064,
"learning_rate": 5.249609120365578e-06,
"loss": 0.0873,
"step": 1407
},
{
"epoch": 0.8586674797987498,
"grad_norm": 0.5450605154037476,
"learning_rate": 5.205630185632548e-06,
"loss": 0.0995,
"step": 1408
},
{
"epoch": 0.8592773288611069,
"grad_norm": 0.8446047902107239,
"learning_rate": 5.16182612632598e-06,
"loss": 0.0922,
"step": 1409
},
{
"epoch": 0.8598871779234639,
"grad_norm": 0.3984534740447998,
"learning_rate": 5.1181971134551646e-06,
"loss": 0.0738,
"step": 1410
},
{
"epoch": 0.860497026985821,
"grad_norm": 0.42127400636672974,
"learning_rate": 5.074743317346009e-06,
"loss": 0.066,
"step": 1411
},
{
"epoch": 0.861106876048178,
"grad_norm": 0.8325464725494385,
"learning_rate": 5.031464907640421e-06,
"loss": 0.0969,
"step": 1412
},
{
"epoch": 0.8617167251105351,
"grad_norm": 0.6694953441619873,
"learning_rate": 4.988362053295564e-06,
"loss": 0.1122,
"step": 1413
},
{
"epoch": 0.8623265741728922,
"grad_norm": 0.6972560882568359,
"learning_rate": 4.945434922583259e-06,
"loss": 0.0874,
"step": 1414
},
{
"epoch": 0.8629364232352493,
"grad_norm": 0.6007323265075684,
"learning_rate": 4.902683683089304e-06,
"loss": 0.0716,
"step": 1415
},
{
"epoch": 0.8635462722976064,
"grad_norm": 0.5268117785453796,
"learning_rate": 4.860108501712824e-06,
"loss": 0.0727,
"step": 1416
},
{
"epoch": 0.8641561213599634,
"grad_norm": 0.6263404488563538,
"learning_rate": 4.817709544665628e-06,
"loss": 0.0795,
"step": 1417
},
{
"epoch": 0.8647659704223205,
"grad_norm": 0.36302539706230164,
"learning_rate": 4.775486977471549e-06,
"loss": 0.0513,
"step": 1418
},
{
"epoch": 0.8653758194846776,
"grad_norm": 0.49809274077415466,
"learning_rate": 4.733440964965791e-06,
"loss": 0.051,
"step": 1419
},
{
"epoch": 0.8659856685470346,
"grad_norm": 0.34820839762687683,
"learning_rate": 4.691571671294298e-06,
"loss": 0.06,
"step": 1420
},
{
"epoch": 0.8665955176093917,
"grad_norm": 0.4885585904121399,
"learning_rate": 4.649879259913137e-06,
"loss": 0.0786,
"step": 1421
},
{
"epoch": 0.8672053666717487,
"grad_norm": 0.40935587882995605,
"learning_rate": 4.608363893587803e-06,
"loss": 0.0777,
"step": 1422
},
{
"epoch": 0.8678152157341058,
"grad_norm": 1.1883491277694702,
"learning_rate": 4.567025734392622e-06,
"loss": 0.0922,
"step": 1423
},
{
"epoch": 0.8684250647964629,
"grad_norm": 0.3187713027000427,
"learning_rate": 4.525864943710112e-06,
"loss": 0.0539,
"step": 1424
},
{
"epoch": 0.8690349138588199,
"grad_norm": 0.7005350589752197,
"learning_rate": 4.484881682230341e-06,
"loss": 0.0772,
"step": 1425
},
{
"epoch": 0.869644762921177,
"grad_norm": 0.3911183178424835,
"learning_rate": 4.4440761099503455e-06,
"loss": 0.0643,
"step": 1426
},
{
"epoch": 0.870254611983534,
"grad_norm": 0.44154277443885803,
"learning_rate": 4.403448386173437e-06,
"loss": 0.078,
"step": 1427
},
{
"epoch": 0.8708644610458911,
"grad_norm": 0.41577833890914917,
"learning_rate": 4.362998669508617e-06,
"loss": 0.0579,
"step": 1428
},
{
"epoch": 0.8714743101082482,
"grad_norm": 0.9702714681625366,
"learning_rate": 4.322727117869951e-06,
"loss": 0.1058,
"step": 1429
},
{
"epoch": 0.8720841591706052,
"grad_norm": 0.7547863125801086,
"learning_rate": 4.28263388847599e-06,
"loss": 0.0827,
"step": 1430
},
{
"epoch": 0.8726940082329624,
"grad_norm": 0.7339526414871216,
"learning_rate": 4.242719137849077e-06,
"loss": 0.0671,
"step": 1431
},
{
"epoch": 0.8733038572953195,
"grad_norm": 0.6482008695602417,
"learning_rate": 4.2029830218148105e-06,
"loss": 0.0981,
"step": 1432
},
{
"epoch": 0.8739137063576765,
"grad_norm": 0.6121746301651001,
"learning_rate": 4.163425695501388e-06,
"loss": 0.0761,
"step": 1433
},
{
"epoch": 0.8745235554200336,
"grad_norm": 0.6097187399864197,
"learning_rate": 4.124047313339025e-06,
"loss": 0.0905,
"step": 1434
},
{
"epoch": 0.8751334044823906,
"grad_norm": 0.41008633375167847,
"learning_rate": 4.0848480290593625e-06,
"loss": 0.0545,
"step": 1435
},
{
"epoch": 0.8757432535447477,
"grad_norm": 0.501014232635498,
"learning_rate": 4.045827995694834e-06,
"loss": 0.0695,
"step": 1436
},
{
"epoch": 0.8763531026071048,
"grad_norm": 0.7358724474906921,
"learning_rate": 4.00698736557808e-06,
"loss": 0.0715,
"step": 1437
},
{
"epoch": 0.8769629516694618,
"grad_norm": 0.8394466638565063,
"learning_rate": 3.968326290341362e-06,
"loss": 0.1284,
"step": 1438
},
{
"epoch": 0.8775728007318189,
"grad_norm": 0.560092031955719,
"learning_rate": 3.929844920915987e-06,
"loss": 0.0813,
"step": 1439
},
{
"epoch": 0.8781826497941759,
"grad_norm": 0.5677507519721985,
"learning_rate": 3.891543407531673e-06,
"loss": 0.0878,
"step": 1440
},
{
"epoch": 0.878792498856533,
"grad_norm": 0.5624688863754272,
"learning_rate": 3.853421899715992e-06,
"loss": 0.0805,
"step": 1441
},
{
"epoch": 0.87940234791889,
"grad_norm": 0.6462926864624023,
"learning_rate": 3.815480546293787e-06,
"loss": 0.0752,
"step": 1442
},
{
"epoch": 0.8800121969812471,
"grad_norm": 0.4995611011981964,
"learning_rate": 3.7777194953865667e-06,
"loss": 0.093,
"step": 1443
},
{
"epoch": 0.8806220460436042,
"grad_norm": 0.4873253405094147,
"learning_rate": 3.740138894411993e-06,
"loss": 0.0694,
"step": 1444
},
{
"epoch": 0.8812318951059612,
"grad_norm": 0.6330360174179077,
"learning_rate": 3.702738890083207e-06,
"loss": 0.0931,
"step": 1445
},
{
"epoch": 0.8818417441683183,
"grad_norm": 0.5129160284996033,
"learning_rate": 3.6655196284083317e-06,
"loss": 0.0931,
"step": 1446
},
{
"epoch": 0.8824515932306755,
"grad_norm": 0.6182073354721069,
"learning_rate": 3.628481254689875e-06,
"loss": 0.0869,
"step": 1447
},
{
"epoch": 0.8830614422930325,
"grad_norm": 0.5169357061386108,
"learning_rate": 3.5916239135241612e-06,
"loss": 0.0586,
"step": 1448
},
{
"epoch": 0.8836712913553896,
"grad_norm": 0.6112831830978394,
"learning_rate": 3.5549477488007854e-06,
"loss": 0.0693,
"step": 1449
},
{
"epoch": 0.8842811404177466,
"grad_norm": 0.5941155552864075,
"learning_rate": 3.5184529037020186e-06,
"loss": 0.0778,
"step": 1450
},
{
"epoch": 0.8848909894801037,
"grad_norm": 0.42443686723709106,
"learning_rate": 3.4821395207022766e-06,
"loss": 0.0712,
"step": 1451
},
{
"epoch": 0.8855008385424608,
"grad_norm": 0.4274662137031555,
"learning_rate": 3.4460077415675474e-06,
"loss": 0.0746,
"step": 1452
},
{
"epoch": 0.8861106876048178,
"grad_norm": 0.553561270236969,
"learning_rate": 3.4100577073548634e-06,
"loss": 0.1208,
"step": 1453
},
{
"epoch": 0.8867205366671749,
"grad_norm": 0.45995670557022095,
"learning_rate": 3.3742895584117085e-06,
"loss": 0.0737,
"step": 1454
},
{
"epoch": 0.8873303857295319,
"grad_norm": 0.9046350121498108,
"learning_rate": 3.3387034343755065e-06,
"loss": 0.1108,
"step": 1455
},
{
"epoch": 0.887940234791889,
"grad_norm": 0.5415409207344055,
"learning_rate": 3.303299474173066e-06,
"loss": 0.0608,
"step": 1456
},
{
"epoch": 0.888550083854246,
"grad_norm": 0.9230696558952332,
"learning_rate": 3.2680778160200155e-06,
"loss": 0.0916,
"step": 1457
},
{
"epoch": 0.8891599329166031,
"grad_norm": 0.6655247807502747,
"learning_rate": 3.233038597420318e-06,
"loss": 0.0772,
"step": 1458
},
{
"epoch": 0.8897697819789602,
"grad_norm": 0.6361746191978455,
"learning_rate": 3.198181955165669e-06,
"loss": 0.1065,
"step": 1459
},
{
"epoch": 0.8903796310413172,
"grad_norm": 0.5962023735046387,
"learning_rate": 3.1635080253350046e-06,
"loss": 0.0974,
"step": 1460
},
{
"epoch": 0.8909894801036743,
"grad_norm": 0.6026719212532043,
"learning_rate": 3.1290169432939553e-06,
"loss": 0.0899,
"step": 1461
},
{
"epoch": 0.8915993291660315,
"grad_norm": 0.394766241312027,
"learning_rate": 3.0947088436943326e-06,
"loss": 0.0822,
"step": 1462
},
{
"epoch": 0.8922091782283885,
"grad_norm": 0.535057544708252,
"learning_rate": 3.060583860473587e-06,
"loss": 0.0686,
"step": 1463
},
{
"epoch": 0.8928190272907456,
"grad_norm": 0.5572860240936279,
"learning_rate": 3.0266421268542735e-06,
"loss": 0.0768,
"step": 1464
},
{
"epoch": 0.8934288763531026,
"grad_norm": 0.2924409508705139,
"learning_rate": 2.9928837753435746e-06,
"loss": 0.047,
"step": 1465
},
{
"epoch": 0.8940387254154597,
"grad_norm": 0.6642356514930725,
"learning_rate": 2.9593089377327245e-06,
"loss": 0.0788,
"step": 1466
},
{
"epoch": 0.8946485744778168,
"grad_norm": 0.48515382409095764,
"learning_rate": 2.9259177450965682e-06,
"loss": 0.0995,
"step": 1467
},
{
"epoch": 0.8952584235401738,
"grad_norm": 0.5138064026832581,
"learning_rate": 2.8927103277929746e-06,
"loss": 0.0774,
"step": 1468
},
{
"epoch": 0.8958682726025309,
"grad_norm": 0.529265820980072,
"learning_rate": 2.8596868154623703e-06,
"loss": 0.0801,
"step": 1469
},
{
"epoch": 0.8964781216648879,
"grad_norm": 0.6112221479415894,
"learning_rate": 2.826847337027222e-06,
"loss": 0.0847,
"step": 1470
},
{
"epoch": 0.897087970727245,
"grad_norm": 0.6569098234176636,
"learning_rate": 2.794192020691544e-06,
"loss": 0.0777,
"step": 1471
},
{
"epoch": 0.8976978197896021,
"grad_norm": 0.5918464660644531,
"learning_rate": 2.7617209939403866e-06,
"loss": 0.1031,
"step": 1472
},
{
"epoch": 0.8983076688519591,
"grad_norm": 0.7132411003112793,
"learning_rate": 2.7294343835393368e-06,
"loss": 0.1003,
"step": 1473
},
{
"epoch": 0.8989175179143162,
"grad_norm": 0.5176982283592224,
"learning_rate": 2.6973323155340234e-06,
"loss": 0.0788,
"step": 1474
},
{
"epoch": 0.8995273669766732,
"grad_norm": 0.6047642827033997,
"learning_rate": 2.665414915249631e-06,
"loss": 0.0959,
"step": 1475
},
{
"epoch": 0.9001372160390303,
"grad_norm": 0.67558354139328,
"learning_rate": 2.6336823072904304e-06,
"loss": 0.0894,
"step": 1476
},
{
"epoch": 0.9007470651013874,
"grad_norm": 0.4685373902320862,
"learning_rate": 2.6021346155392423e-06,
"loss": 0.0808,
"step": 1477
},
{
"epoch": 0.9013569141637445,
"grad_norm": 0.4643312990665436,
"learning_rate": 2.570771963156987e-06,
"loss": 0.0662,
"step": 1478
},
{
"epoch": 0.9019667632261016,
"grad_norm": 0.43990299105644226,
"learning_rate": 2.539594472582213e-06,
"loss": 0.0804,
"step": 1479
},
{
"epoch": 0.9025766122884586,
"grad_norm": 0.5891377329826355,
"learning_rate": 2.50860226553058e-06,
"loss": 0.1186,
"step": 1480
},
{
"epoch": 0.9031864613508157,
"grad_norm": 0.4054217040538788,
"learning_rate": 2.4777954629944477e-06,
"loss": 0.0707,
"step": 1481
},
{
"epoch": 0.9037963104131728,
"grad_norm": 0.598389208316803,
"learning_rate": 2.4471741852423237e-06,
"loss": 0.0841,
"step": 1482
},
{
"epoch": 0.9044061594755298,
"grad_norm": 0.4020671248435974,
"learning_rate": 2.416738551818454e-06,
"loss": 0.0604,
"step": 1483
},
{
"epoch": 0.9050160085378869,
"grad_norm": 0.6247240900993347,
"learning_rate": 2.386488681542326e-06,
"loss": 0.0855,
"step": 1484
},
{
"epoch": 0.9056258576002439,
"grad_norm": 0.5398787260055542,
"learning_rate": 2.3564246925082357e-06,
"loss": 0.062,
"step": 1485
},
{
"epoch": 0.906235706662601,
"grad_norm": 0.39306768774986267,
"learning_rate": 2.3265467020847866e-06,
"loss": 0.0751,
"step": 1486
},
{
"epoch": 0.9068455557249581,
"grad_norm": 0.46908843517303467,
"learning_rate": 2.2968548269144574e-06,
"loss": 0.0817,
"step": 1487
},
{
"epoch": 0.9074554047873151,
"grad_norm": 1.40064537525177,
"learning_rate": 2.2673491829131364e-06,
"loss": 0.0674,
"step": 1488
},
{
"epoch": 0.9080652538496722,
"grad_norm": 0.6439633965492249,
"learning_rate": 2.238029885269677e-06,
"loss": 0.101,
"step": 1489
},
{
"epoch": 0.9086751029120292,
"grad_norm": 0.9020304083824158,
"learning_rate": 2.2088970484454517e-06,
"loss": 0.0804,
"step": 1490
},
{
"epoch": 0.9092849519743863,
"grad_norm": 0.4058304727077484,
"learning_rate": 2.179950786173879e-06,
"loss": 0.0675,
"step": 1491
},
{
"epoch": 0.9098948010367434,
"grad_norm": 0.3557364344596863,
"learning_rate": 2.1511912114600188e-06,
"loss": 0.0651,
"step": 1492
},
{
"epoch": 0.9105046500991004,
"grad_norm": 0.5541882514953613,
"learning_rate": 2.122618436580082e-06,
"loss": 0.074,
"step": 1493
},
{
"epoch": 0.9111144991614576,
"grad_norm": 0.6702840328216553,
"learning_rate": 2.0942325730810565e-06,
"loss": 0.0946,
"step": 1494
},
{
"epoch": 0.9117243482238147,
"grad_norm": 0.4067557454109192,
"learning_rate": 2.066033731780209e-06,
"loss": 0.0858,
"step": 1495
},
{
"epoch": 0.9123341972861717,
"grad_norm": 0.4685390889644623,
"learning_rate": 2.038022022764685e-06,
"loss": 0.0694,
"step": 1496
},
{
"epoch": 0.9129440463485288,
"grad_norm": 0.5509610176086426,
"learning_rate": 2.01019755539108e-06,
"loss": 0.091,
"step": 1497
},
{
"epoch": 0.9135538954108858,
"grad_norm": 0.5147364139556885,
"learning_rate": 1.9825604382849916e-06,
"loss": 0.0755,
"step": 1498
},
{
"epoch": 0.9141637444732429,
"grad_norm": 0.5612689852714539,
"learning_rate": 1.9551107793406355e-06,
"loss": 0.0866,
"step": 1499
},
{
"epoch": 0.9147735935356,
"grad_norm": 0.5147980451583862,
"learning_rate": 1.927848685720368e-06,
"loss": 0.0691,
"step": 1500
},
{
"epoch": 0.915383442597957,
"grad_norm": 0.4939705431461334,
"learning_rate": 1.9007742638543102e-06,
"loss": 0.0749,
"step": 1501
},
{
"epoch": 0.9159932916603141,
"grad_norm": 0.3338252604007721,
"learning_rate": 1.8738876194399236e-06,
"loss": 0.0496,
"step": 1502
},
{
"epoch": 0.9166031407226711,
"grad_norm": 0.43770214915275574,
"learning_rate": 1.8471888574415951e-06,
"loss": 0.078,
"step": 1503
},
{
"epoch": 0.9172129897850282,
"grad_norm": 0.359861820936203,
"learning_rate": 1.820678082090227e-06,
"loss": 0.0569,
"step": 1504
},
{
"epoch": 0.9178228388473852,
"grad_norm": 0.5050402879714966,
"learning_rate": 1.794355396882813e-06,
"loss": 0.0456,
"step": 1505
},
{
"epoch": 0.9184326879097423,
"grad_norm": 0.39417606592178345,
"learning_rate": 1.7682209045820686e-06,
"loss": 0.0743,
"step": 1506
},
{
"epoch": 0.9190425369720994,
"grad_norm": 0.5741586089134216,
"learning_rate": 1.7422747072160017e-06,
"loss": 0.0717,
"step": 1507
},
{
"epoch": 0.9196523860344564,
"grad_norm": 0.498136967420578,
"learning_rate": 1.7165169060775365e-06,
"loss": 0.0657,
"step": 1508
},
{
"epoch": 0.9202622350968135,
"grad_norm": 0.4638533294200897,
"learning_rate": 1.6909476017240912e-06,
"loss": 0.0768,
"step": 1509
},
{
"epoch": 0.9208720841591707,
"grad_norm": 0.6988344192504883,
"learning_rate": 1.665566893977205e-06,
"loss": 0.0946,
"step": 1510
},
{
"epoch": 0.9214819332215277,
"grad_norm": 0.49192455410957336,
"learning_rate": 1.6403748819221466e-06,
"loss": 0.084,
"step": 1511
},
{
"epoch": 0.9220917822838848,
"grad_norm": 0.3809088170528412,
"learning_rate": 1.6153716639075222e-06,
"loss": 0.0638,
"step": 1512
},
{
"epoch": 0.9227016313462418,
"grad_norm": 0.4600454568862915,
"learning_rate": 1.5905573375449012e-06,
"loss": 0.0998,
"step": 1513
},
{
"epoch": 0.9233114804085989,
"grad_norm": 0.5057666301727295,
"learning_rate": 1.5659319997084089e-06,
"loss": 0.0895,
"step": 1514
},
{
"epoch": 0.923921329470956,
"grad_norm": 0.5334294438362122,
"learning_rate": 1.5414957465343882e-06,
"loss": 0.0642,
"step": 1515
},
{
"epoch": 0.924531178533313,
"grad_norm": 0.4860777258872986,
"learning_rate": 1.5172486734209789e-06,
"loss": 0.0714,
"step": 1516
},
{
"epoch": 0.9251410275956701,
"grad_norm": 0.6882983446121216,
"learning_rate": 1.4931908750278e-06,
"loss": 0.0856,
"step": 1517
},
{
"epoch": 0.9257508766580271,
"grad_norm": 0.41912010312080383,
"learning_rate": 1.4693224452755282e-06,
"loss": 0.0723,
"step": 1518
},
{
"epoch": 0.9263607257203842,
"grad_norm": 0.690753698348999,
"learning_rate": 1.4456434773455541e-06,
"loss": 0.1038,
"step": 1519
},
{
"epoch": 0.9269705747827413,
"grad_norm": 0.45494991540908813,
"learning_rate": 1.4221540636796205e-06,
"loss": 0.0535,
"step": 1520
},
{
"epoch": 0.9275804238450983,
"grad_norm": 0.4833679497241974,
"learning_rate": 1.3988542959794627e-06,
"loss": 0.0917,
"step": 1521
},
{
"epoch": 0.9281902729074554,
"grad_norm": 0.6435315012931824,
"learning_rate": 1.3757442652064357e-06,
"loss": 0.0878,
"step": 1522
},
{
"epoch": 0.9288001219698124,
"grad_norm": 0.4793548285961151,
"learning_rate": 1.3528240615811816e-06,
"loss": 0.0997,
"step": 1523
},
{
"epoch": 0.9294099710321695,
"grad_norm": 0.5330131649971008,
"learning_rate": 1.3300937745832521e-06,
"loss": 0.062,
"step": 1524
},
{
"epoch": 0.9300198200945266,
"grad_norm": 0.5057196617126465,
"learning_rate": 1.3075534929507693e-06,
"loss": 0.079,
"step": 1525
},
{
"epoch": 0.9306296691568837,
"grad_norm": 0.41491448879241943,
"learning_rate": 1.2852033046801104e-06,
"loss": 0.0683,
"step": 1526
},
{
"epoch": 0.9312395182192408,
"grad_norm": 0.4796571433544159,
"learning_rate": 1.2630432970255013e-06,
"loss": 0.0832,
"step": 1527
},
{
"epoch": 0.9318493672815978,
"grad_norm": 0.8676946759223938,
"learning_rate": 1.2410735564987341e-06,
"loss": 0.0984,
"step": 1528
},
{
"epoch": 0.9324592163439549,
"grad_norm": 0.848152220249176,
"learning_rate": 1.2192941688687843e-06,
"loss": 0.1079,
"step": 1529
},
{
"epoch": 0.933069065406312,
"grad_norm": 0.5840216279029846,
"learning_rate": 1.1977052191615156e-06,
"loss": 0.081,
"step": 1530
},
{
"epoch": 0.933678914468669,
"grad_norm": 0.6824812889099121,
"learning_rate": 1.1763067916593262e-06,
"loss": 0.082,
"step": 1531
},
{
"epoch": 0.9342887635310261,
"grad_norm": 0.661533534526825,
"learning_rate": 1.1550989699008142e-06,
"loss": 0.0784,
"step": 1532
},
{
"epoch": 0.9348986125933831,
"grad_norm": 0.6074123978614807,
"learning_rate": 1.1340818366804729e-06,
"loss": 0.1214,
"step": 1533
},
{
"epoch": 0.9355084616557402,
"grad_norm": 0.40318530797958374,
"learning_rate": 1.1132554740483415e-06,
"loss": 0.0519,
"step": 1534
},
{
"epoch": 0.9361183107180973,
"grad_norm": 0.6060242056846619,
"learning_rate": 1.0926199633097157e-06,
"loss": 0.0902,
"step": 1535
},
{
"epoch": 0.9367281597804543,
"grad_norm": 0.3710693418979645,
"learning_rate": 1.0721753850247984e-06,
"loss": 0.0634,
"step": 1536
},
{
"epoch": 0.9373380088428114,
"grad_norm": 0.4415774643421173,
"learning_rate": 1.0519218190084056e-06,
"loss": 0.058,
"step": 1537
},
{
"epoch": 0.9379478579051684,
"grad_norm": 1.086997151374817,
"learning_rate": 1.03185934432965e-06,
"loss": 0.077,
"step": 1538
},
{
"epoch": 0.9385577069675255,
"grad_norm": 0.47956815361976624,
"learning_rate": 1.0119880393116176e-06,
"loss": 0.0617,
"step": 1539
},
{
"epoch": 0.9391675560298826,
"grad_norm": 0.5265284776687622,
"learning_rate": 9.923079815310988e-07,
"loss": 0.0941,
"step": 1540
},
{
"epoch": 0.9397774050922397,
"grad_norm": 0.3481796979904175,
"learning_rate": 9.728192478182574e-07,
"loss": 0.0473,
"step": 1541
},
{
"epoch": 0.9403872541545968,
"grad_norm": 0.581805944442749,
"learning_rate": 9.535219142563168e-07,
"loss": 0.0859,
"step": 1542
},
{
"epoch": 0.9409971032169538,
"grad_norm": 0.5024018883705139,
"learning_rate": 9.344160561812921e-07,
"loss": 0.0887,
"step": 1543
},
{
"epoch": 0.9416069522793109,
"grad_norm": 0.9359962344169617,
"learning_rate": 9.155017481817075e-07,
"loss": 0.0938,
"step": 1544
},
{
"epoch": 0.942216801341668,
"grad_norm": 0.40007781982421875,
"learning_rate": 8.967790640982465e-07,
"loss": 0.0731,
"step": 1545
},
{
"epoch": 0.942826650404025,
"grad_norm": 0.41537192463874817,
"learning_rate": 8.782480770235247e-07,
"loss": 0.0635,
"step": 1546
},
{
"epoch": 0.9434364994663821,
"grad_norm": 0.4418744146823883,
"learning_rate": 8.599088593017723e-07,
"loss": 0.0888,
"step": 1547
},
{
"epoch": 0.9440463485287391,
"grad_norm": 0.5433666706085205,
"learning_rate": 8.417614825285636e-07,
"loss": 0.086,
"step": 1548
},
{
"epoch": 0.9446561975910962,
"grad_norm": 0.491115003824234,
"learning_rate": 8.238060175505269e-07,
"loss": 0.0726,
"step": 1549
},
{
"epoch": 0.9452660466534533,
"grad_norm": 0.611727774143219,
"learning_rate": 8.060425344650846e-07,
"loss": 0.0841,
"step": 1550
},
{
"epoch": 0.9458758957158103,
"grad_norm": 0.6789376139640808,
"learning_rate": 7.884711026201585e-07,
"loss": 0.0886,
"step": 1551
},
{
"epoch": 0.9464857447781674,
"grad_norm": 0.4292146563529968,
"learning_rate": 7.710917906139204e-07,
"loss": 0.0661,
"step": 1552
},
{
"epoch": 0.9470955938405244,
"grad_norm": 0.7551100850105286,
"learning_rate": 7.53904666294497e-07,
"loss": 0.0686,
"step": 1553
},
{
"epoch": 0.9477054429028815,
"grad_norm": 0.6248593330383301,
"learning_rate": 7.369097967597493e-07,
"loss": 0.1278,
"step": 1554
},
{
"epoch": 0.9483152919652386,
"grad_norm": 0.45670968294143677,
"learning_rate": 7.201072483569549e-07,
"loss": 0.0708,
"step": 1555
},
{
"epoch": 0.9489251410275956,
"grad_norm": 0.4589419364929199,
"learning_rate": 7.034970866825974e-07,
"loss": 0.0795,
"step": 1556
},
{
"epoch": 0.9495349900899528,
"grad_norm": 0.6680642366409302,
"learning_rate": 6.870793765820782e-07,
"loss": 0.0923,
"step": 1557
},
{
"epoch": 0.9501448391523099,
"grad_norm": 0.5196901559829712,
"learning_rate": 6.708541821494829e-07,
"loss": 0.082,
"step": 1558
},
{
"epoch": 0.9507546882146669,
"grad_norm": 0.7338657975196838,
"learning_rate": 6.548215667273206e-07,
"loss": 0.089,
"step": 1559
},
{
"epoch": 0.951364537277024,
"grad_norm": 0.4829563796520233,
"learning_rate": 6.389815929062848e-07,
"loss": 0.0812,
"step": 1560
},
{
"epoch": 0.951974386339381,
"grad_norm": 0.7363712787628174,
"learning_rate": 6.233343225249933e-07,
"loss": 0.116,
"step": 1561
},
{
"epoch": 0.9525842354017381,
"grad_norm": 0.5185093879699707,
"learning_rate": 6.078798166697541e-07,
"loss": 0.0693,
"step": 1562
},
{
"epoch": 0.9531940844640951,
"grad_norm": 0.4027321934700012,
"learning_rate": 5.92618135674361e-07,
"loss": 0.0701,
"step": 1563
},
{
"epoch": 0.9538039335264522,
"grad_norm": 0.4655270278453827,
"learning_rate": 5.775493391197928e-07,
"loss": 0.0612,
"step": 1564
},
{
"epoch": 0.9544137825888093,
"grad_norm": 0.4907931983470917,
"learning_rate": 5.626734858340255e-07,
"loss": 0.0921,
"step": 1565
},
{
"epoch": 0.9550236316511663,
"grad_norm": 0.6910971403121948,
"learning_rate": 5.479906338917984e-07,
"loss": 0.1208,
"step": 1566
},
{
"epoch": 0.9556334807135234,
"grad_norm": 0.5560339093208313,
"learning_rate": 5.335008406143815e-07,
"loss": 0.068,
"step": 1567
},
{
"epoch": 0.9562433297758804,
"grad_norm": 0.5279018878936768,
"learning_rate": 5.192041625693478e-07,
"loss": 0.102,
"step": 1568
},
{
"epoch": 0.9568531788382375,
"grad_norm": 0.4401240348815918,
"learning_rate": 5.051006555703453e-07,
"loss": 0.0573,
"step": 1569
},
{
"epoch": 0.9574630279005946,
"grad_norm": 0.595805823802948,
"learning_rate": 4.911903746769142e-07,
"loss": 0.0645,
"step": 1570
},
{
"epoch": 0.9580728769629516,
"grad_norm": 0.7749236822128296,
"learning_rate": 4.774733741942206e-07,
"loss": 0.0654,
"step": 1571
},
{
"epoch": 0.9586827260253087,
"grad_norm": 0.5895031094551086,
"learning_rate": 4.639497076728949e-07,
"loss": 0.0637,
"step": 1572
},
{
"epoch": 0.9592925750876659,
"grad_norm": 0.516834557056427,
"learning_rate": 4.5061942790879386e-07,
"loss": 0.0778,
"step": 1573
},
{
"epoch": 0.9599024241500229,
"grad_norm": 0.48434749245643616,
"learning_rate": 4.3748258694278344e-07,
"loss": 0.0653,
"step": 1574
},
{
"epoch": 0.96051227321238,
"grad_norm": 0.4496859312057495,
"learning_rate": 4.2453923606057265e-07,
"loss": 0.0888,
"step": 1575
},
{
"epoch": 0.961122122274737,
"grad_norm": 0.5107975602149963,
"learning_rate": 4.1178942579248036e-07,
"loss": 0.0852,
"step": 1576
},
{
"epoch": 0.9617319713370941,
"grad_norm": 0.5424916744232178,
"learning_rate": 3.992332059132631e-07,
"loss": 0.0983,
"step": 1577
},
{
"epoch": 0.9623418203994512,
"grad_norm": 0.41522082686424255,
"learning_rate": 3.86870625441893e-07,
"loss": 0.0572,
"step": 1578
},
{
"epoch": 0.9629516694618082,
"grad_norm": 0.5182363390922546,
"learning_rate": 3.747017326413971e-07,
"loss": 0.0709,
"step": 1579
},
{
"epoch": 0.9635615185241653,
"grad_norm": 0.6186760663986206,
"learning_rate": 3.627265750186348e-07,
"loss": 0.1035,
"step": 1580
},
{
"epoch": 0.9641713675865223,
"grad_norm": 0.797123908996582,
"learning_rate": 3.5094519932415417e-07,
"loss": 0.076,
"step": 1581
},
{
"epoch": 0.9647812166488794,
"grad_norm": 0.5571187138557434,
"learning_rate": 3.3935765155196366e-07,
"loss": 0.0855,
"step": 1582
},
{
"epoch": 0.9653910657112365,
"grad_norm": 0.4230852425098419,
"learning_rate": 3.2796397693939385e-07,
"loss": 0.065,
"step": 1583
},
{
"epoch": 0.9660009147735935,
"grad_norm": 0.430282860994339,
"learning_rate": 3.167642199668863e-07,
"loss": 0.0801,
"step": 1584
},
{
"epoch": 0.9666107638359506,
"grad_norm": 0.48358258605003357,
"learning_rate": 3.0575842435785486e-07,
"loss": 0.0902,
"step": 1585
},
{
"epoch": 0.9672206128983076,
"grad_norm": 0.6248466372489929,
"learning_rate": 2.9494663307847447e-07,
"loss": 0.0859,
"step": 1586
},
{
"epoch": 0.9678304619606647,
"grad_norm": 0.6716404557228088,
"learning_rate": 2.843288883375539e-07,
"loss": 0.1027,
"step": 1587
},
{
"epoch": 0.9684403110230217,
"grad_norm": 0.30844345688819885,
"learning_rate": 2.7390523158633554e-07,
"loss": 0.0563,
"step": 1588
},
{
"epoch": 0.9690501600853789,
"grad_norm": 0.5965458154678345,
"learning_rate": 2.6367570351836237e-07,
"loss": 0.0685,
"step": 1589
},
{
"epoch": 0.969660009147736,
"grad_norm": 0.43928834795951843,
"learning_rate": 2.536403440693003e-07,
"loss": 0.0771,
"step": 1590
},
{
"epoch": 0.970269858210093,
"grad_norm": 0.34640204906463623,
"learning_rate": 2.437991924167937e-07,
"loss": 0.0609,
"step": 1591
},
{
"epoch": 0.9708797072724501,
"grad_norm": 0.4636484682559967,
"learning_rate": 2.341522869803048e-07,
"loss": 0.0648,
"step": 1592
},
{
"epoch": 0.9714895563348072,
"grad_norm": 0.5436096787452698,
"learning_rate": 2.2469966542096322e-07,
"loss": 0.078,
"step": 1593
},
{
"epoch": 0.9720994053971642,
"grad_norm": 0.35389089584350586,
"learning_rate": 2.1544136464142772e-07,
"loss": 0.0494,
"step": 1594
},
{
"epoch": 0.9727092544595213,
"grad_norm": 1.458207607269287,
"learning_rate": 2.0637742078573607e-07,
"loss": 0.1179,
"step": 1595
},
{
"epoch": 0.9733191035218783,
"grad_norm": 0.3609294593334198,
"learning_rate": 1.975078692391552e-07,
"loss": 0.0591,
"step": 1596
},
{
"epoch": 0.9739289525842354,
"grad_norm": 0.520239531993866,
"learning_rate": 1.8883274462806467e-07,
"loss": 0.0894,
"step": 1597
},
{
"epoch": 0.9745388016465925,
"grad_norm": 0.6305734515190125,
"learning_rate": 1.8035208081980115e-07,
"loss": 0.0519,
"step": 1598
},
{
"epoch": 0.9751486507089495,
"grad_norm": 0.8180693984031677,
"learning_rate": 1.7206591092253642e-07,
"loss": 0.117,
"step": 1599
},
{
"epoch": 0.9757584997713066,
"grad_norm": 0.3084523379802704,
"learning_rate": 1.6397426728514964e-07,
"loss": 0.052,
"step": 1600
},
{
"epoch": 0.9763683488336636,
"grad_norm": 0.3350369930267334,
"learning_rate": 1.560771814970885e-07,
"loss": 0.0517,
"step": 1601
},
{
"epoch": 0.9769781978960207,
"grad_norm": 0.5770084261894226,
"learning_rate": 1.4837468438826386e-07,
"loss": 0.0998,
"step": 1602
},
{
"epoch": 0.9775880469583778,
"grad_norm": 0.7938170433044434,
"learning_rate": 1.4086680602891643e-07,
"loss": 0.0719,
"step": 1603
},
{
"epoch": 0.9781978960207348,
"grad_norm": 0.5285010933876038,
"learning_rate": 1.335535757294948e-07,
"loss": 0.0655,
"step": 1604
},
{
"epoch": 0.978807745083092,
"grad_norm": 0.4537477493286133,
"learning_rate": 1.264350220405719e-07,
"loss": 0.0528,
"step": 1605
},
{
"epoch": 0.979417594145449,
"grad_norm": 0.3776329755783081,
"learning_rate": 1.195111727526843e-07,
"loss": 0.0579,
"step": 1606
},
{
"epoch": 0.9800274432078061,
"grad_norm": 0.3889644742012024,
"learning_rate": 1.1278205489626547e-07,
"loss": 0.0461,
"step": 1607
},
{
"epoch": 0.9806372922701632,
"grad_norm": 0.5790485739707947,
"learning_rate": 1.0624769474152363e-07,
"loss": 0.0848,
"step": 1608
},
{
"epoch": 0.9812471413325202,
"grad_norm": 0.6506345272064209,
"learning_rate": 9.99081177983363e-08,
"loss": 0.0901,
"step": 1609
},
{
"epoch": 0.9818569903948773,
"grad_norm": 0.42821401357650757,
"learning_rate": 9.376334881616156e-08,
"loss": 0.0582,
"step": 1610
},
{
"epoch": 0.9824668394572343,
"grad_norm": 0.48001283407211304,
"learning_rate": 8.781341178393244e-08,
"loss": 0.0811,
"step": 1611
},
{
"epoch": 0.9830766885195914,
"grad_norm": 0.4542602002620697,
"learning_rate": 8.20583299299571e-08,
"loss": 0.0561,
"step": 1612
},
{
"epoch": 0.9836865375819485,
"grad_norm": 0.670384407043457,
"learning_rate": 7.649812572185222e-08,
"loss": 0.0881,
"step": 1613
},
{
"epoch": 0.9842963866443055,
"grad_norm": 0.7463256120681763,
"learning_rate": 7.113282086643191e-08,
"loss": 0.0868,
"step": 1614
},
{
"epoch": 0.9849062357066626,
"grad_norm": 0.5491529107093811,
"learning_rate": 6.596243630963006e-08,
"loss": 0.0714,
"step": 1615
},
{
"epoch": 0.9855160847690196,
"grad_norm": 0.5644745230674744,
"learning_rate": 6.098699223641702e-08,
"loss": 0.0901,
"step": 1616
},
{
"epoch": 0.9861259338313767,
"grad_norm": 0.4111098349094391,
"learning_rate": 5.620650807073857e-08,
"loss": 0.0663,
"step": 1617
},
{
"epoch": 0.9867357828937338,
"grad_norm": 0.5614528059959412,
"learning_rate": 5.162100247539936e-08,
"loss": 0.0982,
"step": 1618
},
{
"epoch": 0.9873456319560908,
"grad_norm": 0.40404701232910156,
"learning_rate": 4.723049335204066e-08,
"loss": 0.0626,
"step": 1619
},
{
"epoch": 0.987955481018448,
"grad_norm": 0.5119568109512329,
"learning_rate": 4.303499784102383e-08,
"loss": 0.0946,
"step": 1620
},
{
"epoch": 0.988565330080805,
"grad_norm": 0.6171373724937439,
"learning_rate": 3.9034532321408076e-08,
"loss": 0.0868,
"step": 1621
},
{
"epoch": 0.9891751791431621,
"grad_norm": 0.36349228024482727,
"learning_rate": 3.522911241083948e-08,
"loss": 0.0671,
"step": 1622
},
{
"epoch": 0.9897850282055192,
"grad_norm": 0.4687579572200775,
"learning_rate": 3.161875296553429e-08,
"loss": 0.0804,
"step": 1623
},
{
"epoch": 0.9903948772678762,
"grad_norm": 0.5183506011962891,
"learning_rate": 2.820346808018459e-08,
"loss": 0.0587,
"step": 1624
},
{
"epoch": 0.9910047263302333,
"grad_norm": 0.3921150863170624,
"learning_rate": 2.4983271087924974e-08,
"loss": 0.0627,
"step": 1625
},
{
"epoch": 0.9916145753925903,
"grad_norm": 0.5260083675384521,
"learning_rate": 2.1958174560282595e-08,
"loss": 0.0788,
"step": 1626
},
{
"epoch": 0.9922244244549474,
"grad_norm": 0.5869860649108887,
"learning_rate": 1.9128190307105e-08,
"loss": 0.0656,
"step": 1627
},
{
"epoch": 0.9928342735173045,
"grad_norm": 0.5787657499313354,
"learning_rate": 1.6493329376549016e-08,
"loss": 0.0713,
"step": 1628
},
{
"epoch": 0.9934441225796615,
"grad_norm": 0.7932971715927124,
"learning_rate": 1.4053602054991955e-08,
"loss": 0.0885,
"step": 1629
},
{
"epoch": 0.9940539716420186,
"grad_norm": 0.6366103887557983,
"learning_rate": 1.1809017867037141e-08,
"loss": 0.0731,
"step": 1630
},
{
"epoch": 0.9946638207043756,
"grad_norm": 0.49447983503341675,
"learning_rate": 9.75958557545842e-09,
"loss": 0.0608,
"step": 1631
},
{
"epoch": 0.9952736697667327,
"grad_norm": 0.3736961781978607,
"learning_rate": 7.905313181150176e-09,
"loss": 0.052,
"step": 1632
},
{
"epoch": 0.9958835188290898,
"grad_norm": 0.42686665058135986,
"learning_rate": 6.246207923116254e-09,
"loss": 0.0744,
"step": 1633
},
{
"epoch": 0.9964933678914468,
"grad_norm": 0.4884948432445526,
"learning_rate": 4.782276278442188e-09,
"loss": 0.053,
"step": 1634
},
{
"epoch": 0.9971032169538039,
"grad_norm": 0.5585065484046936,
"learning_rate": 3.513523962256349e-09,
"loss": 0.0616,
"step": 1635
},
{
"epoch": 0.997713066016161,
"grad_norm": 0.9017805457115173,
"learning_rate": 2.4399559277132888e-09,
"loss": 0.0922,
"step": 1636
},
{
"epoch": 0.9983229150785181,
"grad_norm": 0.452959805727005,
"learning_rate": 1.5615763659881933e-09,
"loss": 0.0899,
"step": 1637
},
{
"epoch": 0.9989327641408752,
"grad_norm": 0.735152542591095,
"learning_rate": 8.783887062324692e-10,
"loss": 0.0805,
"step": 1638
},
{
"epoch": 0.9995426132032322,
"grad_norm": 0.3932315707206726,
"learning_rate": 3.903956155848487e-10,
"loss": 0.0596,
"step": 1639
},
{
"epoch": 1.0,
"grad_norm": 0.496868371963501,
"learning_rate": 9.759899914918435e-11,
"loss": 0.0778,
"step": 1640
},
{
"epoch": 1.0,
"step": 1640,
"total_flos": 1.986746635089294e+19,
"train_loss": 0.12436322811990977,
"train_runtime": 37071.6055,
"train_samples_per_second": 2.831,
"train_steps_per_second": 0.044
}
],
"logging_steps": 1.0,
"max_steps": 1640,
"num_input_tokens_seen": 0,
"num_train_epochs": 1,
"save_steps": 500,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 1.986746635089294e+19,
"train_batch_size": 1,
"trial_name": null,
"trial_params": null
}