atsuki-yamaguchi's picture
Upload folder using huggingface_hub
b2f3b84 verified
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 1.0,
"eval_steps": 500,
"global_step": 1257,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.002386634844868735,
"grad_norm": 27.54071617126465,
"learning_rate": 2.3809523809523808e-06,
"loss": 6.8548,
"step": 3
},
{
"epoch": 0.00477326968973747,
"grad_norm": 16.49394989013672,
"learning_rate": 4.7619047619047615e-06,
"loss": 6.5867,
"step": 6
},
{
"epoch": 0.007159904534606206,
"grad_norm": 19.11440658569336,
"learning_rate": 7.142857142857143e-06,
"loss": 6.2734,
"step": 9
},
{
"epoch": 0.00954653937947494,
"grad_norm": 20.9352970123291,
"learning_rate": 9.523809523809523e-06,
"loss": 5.535,
"step": 12
},
{
"epoch": 0.011933174224343675,
"grad_norm": 11.74897575378418,
"learning_rate": 1.1904761904761905e-05,
"loss": 5.2226,
"step": 15
},
{
"epoch": 0.014319809069212411,
"grad_norm": 15.756311416625977,
"learning_rate": 1.4285714285714285e-05,
"loss": 4.9951,
"step": 18
},
{
"epoch": 0.016706443914081145,
"grad_norm": 9.158567428588867,
"learning_rate": 1.6666666666666667e-05,
"loss": 4.6998,
"step": 21
},
{
"epoch": 0.01909307875894988,
"grad_norm": 10.05123233795166,
"learning_rate": 1.9047619047619046e-05,
"loss": 4.4934,
"step": 24
},
{
"epoch": 0.021479713603818614,
"grad_norm": 8.713202476501465,
"learning_rate": 2.1428571428571428e-05,
"loss": 4.4422,
"step": 27
},
{
"epoch": 0.02386634844868735,
"grad_norm": 6.295734882354736,
"learning_rate": 2.380952380952381e-05,
"loss": 4.307,
"step": 30
},
{
"epoch": 0.026252983293556086,
"grad_norm": 4.285629749298096,
"learning_rate": 2.6190476190476192e-05,
"loss": 3.9384,
"step": 33
},
{
"epoch": 0.028639618138424822,
"grad_norm": 3.6104960441589355,
"learning_rate": 2.857142857142857e-05,
"loss": 4.0232,
"step": 36
},
{
"epoch": 0.031026252983293555,
"grad_norm": 3.1964364051818848,
"learning_rate": 3.095238095238095e-05,
"loss": 4.0479,
"step": 39
},
{
"epoch": 0.03341288782816229,
"grad_norm": 3.879831075668335,
"learning_rate": 3.3333333333333335e-05,
"loss": 3.8802,
"step": 42
},
{
"epoch": 0.03579952267303103,
"grad_norm": 4.037631511688232,
"learning_rate": 3.571428571428572e-05,
"loss": 3.8634,
"step": 45
},
{
"epoch": 0.03818615751789976,
"grad_norm": 2.8461647033691406,
"learning_rate": 3.809523809523809e-05,
"loss": 3.7112,
"step": 48
},
{
"epoch": 0.0405727923627685,
"grad_norm": 3.655538558959961,
"learning_rate": 4.047619047619048e-05,
"loss": 3.6577,
"step": 51
},
{
"epoch": 0.04295942720763723,
"grad_norm": 3.792799711227417,
"learning_rate": 4.2857142857142856e-05,
"loss": 3.6948,
"step": 54
},
{
"epoch": 0.045346062052505964,
"grad_norm": 5.230392932891846,
"learning_rate": 4.523809523809524e-05,
"loss": 3.618,
"step": 57
},
{
"epoch": 0.0477326968973747,
"grad_norm": 5.6567230224609375,
"learning_rate": 4.761904761904762e-05,
"loss": 3.4814,
"step": 60
},
{
"epoch": 0.050119331742243436,
"grad_norm": 3.167008399963379,
"learning_rate": 5e-05,
"loss": 3.3859,
"step": 63
},
{
"epoch": 0.05250596658711217,
"grad_norm": 4.307368278503418,
"learning_rate": 5.2380952380952384e-05,
"loss": 3.2242,
"step": 66
},
{
"epoch": 0.05489260143198091,
"grad_norm": 3.390169382095337,
"learning_rate": 5.4761904761904766e-05,
"loss": 3.0145,
"step": 69
},
{
"epoch": 0.057279236276849645,
"grad_norm": 3.3308513164520264,
"learning_rate": 5.714285714285714e-05,
"loss": 3.0705,
"step": 72
},
{
"epoch": 0.059665871121718374,
"grad_norm": 3.2529304027557373,
"learning_rate": 5.9523809523809524e-05,
"loss": 3.0395,
"step": 75
},
{
"epoch": 0.06205250596658711,
"grad_norm": 2.5623693466186523,
"learning_rate": 6.19047619047619e-05,
"loss": 2.9807,
"step": 78
},
{
"epoch": 0.06443914081145585,
"grad_norm": 2.3922460079193115,
"learning_rate": 6.428571428571429e-05,
"loss": 2.8913,
"step": 81
},
{
"epoch": 0.06682577565632458,
"grad_norm": 3.3650715351104736,
"learning_rate": 6.666666666666667e-05,
"loss": 2.7402,
"step": 84
},
{
"epoch": 0.06921241050119331,
"grad_norm": 2.4342546463012695,
"learning_rate": 6.904761904761905e-05,
"loss": 2.7315,
"step": 87
},
{
"epoch": 0.07159904534606205,
"grad_norm": 2.673886775970459,
"learning_rate": 7.142857142857143e-05,
"loss": 2.7205,
"step": 90
},
{
"epoch": 0.07398568019093078,
"grad_norm": 3.11196231842041,
"learning_rate": 7.380952380952382e-05,
"loss": 2.7564,
"step": 93
},
{
"epoch": 0.07637231503579953,
"grad_norm": 2.593316078186035,
"learning_rate": 7.619047619047618e-05,
"loss": 2.6578,
"step": 96
},
{
"epoch": 0.07875894988066826,
"grad_norm": 24.880268096923828,
"learning_rate": 7.857142857142858e-05,
"loss": 2.6199,
"step": 99
},
{
"epoch": 0.081145584725537,
"grad_norm": 4.8875274658203125,
"learning_rate": 8.095238095238096e-05,
"loss": 2.6157,
"step": 102
},
{
"epoch": 0.08353221957040573,
"grad_norm": 18.756973266601562,
"learning_rate": 8.333333333333334e-05,
"loss": 2.6938,
"step": 105
},
{
"epoch": 0.08591885441527446,
"grad_norm": 5.592845439910889,
"learning_rate": 8.571428571428571e-05,
"loss": 2.6238,
"step": 108
},
{
"epoch": 0.0883054892601432,
"grad_norm": 2.847984790802002,
"learning_rate": 8.80952380952381e-05,
"loss": 2.5633,
"step": 111
},
{
"epoch": 0.09069212410501193,
"grad_norm": 2.497422218322754,
"learning_rate": 9.047619047619048e-05,
"loss": 2.6791,
"step": 114
},
{
"epoch": 0.09307875894988067,
"grad_norm": 3.0090177059173584,
"learning_rate": 9.285714285714286e-05,
"loss": 2.5656,
"step": 117
},
{
"epoch": 0.0954653937947494,
"grad_norm": 4.84130859375,
"learning_rate": 9.523809523809524e-05,
"loss": 2.5623,
"step": 120
},
{
"epoch": 0.09785202863961814,
"grad_norm": 3.03096866607666,
"learning_rate": 9.761904761904762e-05,
"loss": 2.581,
"step": 123
},
{
"epoch": 0.10023866348448687,
"grad_norm": 3.2894864082336426,
"learning_rate": 0.0001,
"loss": 2.3262,
"step": 126
},
{
"epoch": 0.1026252983293556,
"grad_norm": 3.5789895057678223,
"learning_rate": 9.999961058466053e-05,
"loss": 2.5495,
"step": 129
},
{
"epoch": 0.10501193317422435,
"grad_norm": 2.6820790767669678,
"learning_rate": 9.999844234470782e-05,
"loss": 2.4322,
"step": 132
},
{
"epoch": 0.10739856801909307,
"grad_norm": 2.5585029125213623,
"learning_rate": 9.999649529833915e-05,
"loss": 2.4225,
"step": 135
},
{
"epoch": 0.10978520286396182,
"grad_norm": 2.9007081985473633,
"learning_rate": 9.999376947588288e-05,
"loss": 2.5102,
"step": 138
},
{
"epoch": 0.11217183770883055,
"grad_norm": 3.485059976577759,
"learning_rate": 9.999026491979808e-05,
"loss": 2.4618,
"step": 141
},
{
"epoch": 0.11455847255369929,
"grad_norm": 2.1991868019104004,
"learning_rate": 9.99859816846739e-05,
"loss": 2.4147,
"step": 144
},
{
"epoch": 0.11694510739856802,
"grad_norm": 2.4278526306152344,
"learning_rate": 9.998091983722863e-05,
"loss": 2.406,
"step": 147
},
{
"epoch": 0.11933174224343675,
"grad_norm": 2.217151403427124,
"learning_rate": 9.99750794563087e-05,
"loss": 2.2757,
"step": 150
},
{
"epoch": 0.12171837708830549,
"grad_norm": 2.5492608547210693,
"learning_rate": 9.996846063288747e-05,
"loss": 2.3297,
"step": 153
},
{
"epoch": 0.12410501193317422,
"grad_norm": 2.6618294715881348,
"learning_rate": 9.996106347006379e-05,
"loss": 2.4732,
"step": 156
},
{
"epoch": 0.12649164677804295,
"grad_norm": 2.283311605453491,
"learning_rate": 9.99528880830604e-05,
"loss": 2.2756,
"step": 159
},
{
"epoch": 0.1288782816229117,
"grad_norm": 2.560292959213257,
"learning_rate": 9.994393459922218e-05,
"loss": 2.2277,
"step": 162
},
{
"epoch": 0.13126491646778043,
"grad_norm": 4.688173294067383,
"learning_rate": 9.993420315801406e-05,
"loss": 2.2035,
"step": 165
},
{
"epoch": 0.13365155131264916,
"grad_norm": 2.796677350997925,
"learning_rate": 9.992369391101895e-05,
"loss": 2.2856,
"step": 168
},
{
"epoch": 0.1360381861575179,
"grad_norm": 2.2355430126190186,
"learning_rate": 9.991240702193532e-05,
"loss": 2.4203,
"step": 171
},
{
"epoch": 0.13842482100238662,
"grad_norm": 4.946108818054199,
"learning_rate": 9.990034266657467e-05,
"loss": 2.3584,
"step": 174
},
{
"epoch": 0.14081145584725538,
"grad_norm": 2.3821146488189697,
"learning_rate": 9.988750103285883e-05,
"loss": 2.2558,
"step": 177
},
{
"epoch": 0.1431980906921241,
"grad_norm": 2.569561719894409,
"learning_rate": 9.987388232081694e-05,
"loss": 2.2857,
"step": 180
},
{
"epoch": 0.14558472553699284,
"grad_norm": 2.7219672203063965,
"learning_rate": 9.985948674258243e-05,
"loss": 2.1856,
"step": 183
},
{
"epoch": 0.14797136038186157,
"grad_norm": 2.2356839179992676,
"learning_rate": 9.984431452238967e-05,
"loss": 2.2922,
"step": 186
},
{
"epoch": 0.15035799522673032,
"grad_norm": 2.406996488571167,
"learning_rate": 9.982836589657043e-05,
"loss": 2.2222,
"step": 189
},
{
"epoch": 0.15274463007159905,
"grad_norm": 2.0661559104919434,
"learning_rate": 9.981164111355035e-05,
"loss": 2.2842,
"step": 192
},
{
"epoch": 0.15513126491646778,
"grad_norm": 2.405298948287964,
"learning_rate": 9.979414043384485e-05,
"loss": 2.2644,
"step": 195
},
{
"epoch": 0.1575178997613365,
"grad_norm": 2.26259446144104,
"learning_rate": 9.977586413005531e-05,
"loss": 2.3201,
"step": 198
},
{
"epoch": 0.15990453460620524,
"grad_norm": 2.7550411224365234,
"learning_rate": 9.975681248686461e-05,
"loss": 2.3025,
"step": 201
},
{
"epoch": 0.162291169451074,
"grad_norm": 2.2119550704956055,
"learning_rate": 9.973698580103285e-05,
"loss": 2.2355,
"step": 204
},
{
"epoch": 0.16467780429594273,
"grad_norm": 2.1809093952178955,
"learning_rate": 9.971638438139266e-05,
"loss": 2.3667,
"step": 207
},
{
"epoch": 0.16706443914081145,
"grad_norm": 2.4465034008026123,
"learning_rate": 9.96950085488444e-05,
"loss": 2.1225,
"step": 210
},
{
"epoch": 0.16945107398568018,
"grad_norm": 2.1557154655456543,
"learning_rate": 9.967285863635112e-05,
"loss": 2.3583,
"step": 213
},
{
"epoch": 0.1718377088305489,
"grad_norm": 2.4219322204589844,
"learning_rate": 9.964993498893349e-05,
"loss": 2.2503,
"step": 216
},
{
"epoch": 0.17422434367541767,
"grad_norm": 2.533348798751831,
"learning_rate": 9.962623796366429e-05,
"loss": 2.2262,
"step": 219
},
{
"epoch": 0.1766109785202864,
"grad_norm": 2.264911413192749,
"learning_rate": 9.960176792966289e-05,
"loss": 2.202,
"step": 222
},
{
"epoch": 0.17899761336515513,
"grad_norm": 1.754560112953186,
"learning_rate": 9.95765252680896e-05,
"loss": 2.2589,
"step": 225
},
{
"epoch": 0.18138424821002386,
"grad_norm": 1.8504457473754883,
"learning_rate": 9.95505103721396e-05,
"loss": 2.2058,
"step": 228
},
{
"epoch": 0.18377088305489261,
"grad_norm": 2.1537466049194336,
"learning_rate": 9.952372364703687e-05,
"loss": 2.3089,
"step": 231
},
{
"epoch": 0.18615751789976134,
"grad_norm": 3.062028408050537,
"learning_rate": 9.949616551002787e-05,
"loss": 2.195,
"step": 234
},
{
"epoch": 0.18854415274463007,
"grad_norm": 2.369779586791992,
"learning_rate": 9.946783639037504e-05,
"loss": 2.1499,
"step": 237
},
{
"epoch": 0.1909307875894988,
"grad_norm": 2.2154617309570312,
"learning_rate": 9.943873672935014e-05,
"loss": 2.1925,
"step": 240
},
{
"epoch": 0.19331742243436753,
"grad_norm": 2.502539873123169,
"learning_rate": 9.940886698022734e-05,
"loss": 2.0562,
"step": 243
},
{
"epoch": 0.1957040572792363,
"grad_norm": 2.294062614440918,
"learning_rate": 9.93782276082762e-05,
"loss": 2.1681,
"step": 246
},
{
"epoch": 0.19809069212410502,
"grad_norm": 2.0213210582733154,
"learning_rate": 9.934681909075434e-05,
"loss": 2.1332,
"step": 249
},
{
"epoch": 0.20047732696897375,
"grad_norm": 2.3265905380249023,
"learning_rate": 9.931464191690015e-05,
"loss": 2.2054,
"step": 252
},
{
"epoch": 0.20286396181384247,
"grad_norm": 1.9375816583633423,
"learning_rate": 9.928169658792498e-05,
"loss": 2.1961,
"step": 255
},
{
"epoch": 0.2052505966587112,
"grad_norm": 2.081695318222046,
"learning_rate": 9.924798361700553e-05,
"loss": 2.1484,
"step": 258
},
{
"epoch": 0.20763723150357996,
"grad_norm": 2.059936046600342,
"learning_rate": 9.92135035292757e-05,
"loss": 2.1565,
"step": 261
},
{
"epoch": 0.2100238663484487,
"grad_norm": 1.9999018907546997,
"learning_rate": 9.91782568618185e-05,
"loss": 2.2006,
"step": 264
},
{
"epoch": 0.21241050119331742,
"grad_norm": 1.8792791366577148,
"learning_rate": 9.914224416365764e-05,
"loss": 2.2193,
"step": 267
},
{
"epoch": 0.21479713603818615,
"grad_norm": 1.8357597589492798,
"learning_rate": 9.910546599574902e-05,
"loss": 2.3437,
"step": 270
},
{
"epoch": 0.2171837708830549,
"grad_norm": 2.886901378631592,
"learning_rate": 9.906792293097194e-05,
"loss": 2.1577,
"step": 273
},
{
"epoch": 0.21957040572792363,
"grad_norm": 2.1063649654388428,
"learning_rate": 9.90296155541202e-05,
"loss": 2.1812,
"step": 276
},
{
"epoch": 0.22195704057279236,
"grad_norm": 2.063605785369873,
"learning_rate": 9.899054446189304e-05,
"loss": 2.0393,
"step": 279
},
{
"epoch": 0.2243436754176611,
"grad_norm": 2.5164003372192383,
"learning_rate": 9.895071026288574e-05,
"loss": 2.0649,
"step": 282
},
{
"epoch": 0.22673031026252982,
"grad_norm": 2.180410385131836,
"learning_rate": 9.891011357758022e-05,
"loss": 2.0445,
"step": 285
},
{
"epoch": 0.22911694510739858,
"grad_norm": 2.030186176300049,
"learning_rate": 9.886875503833536e-05,
"loss": 2.2003,
"step": 288
},
{
"epoch": 0.2315035799522673,
"grad_norm": 2.293309450149536,
"learning_rate": 9.882663528937717e-05,
"loss": 2.1257,
"step": 291
},
{
"epoch": 0.23389021479713604,
"grad_norm": 1.9527851343154907,
"learning_rate": 9.87837549867887e-05,
"loss": 1.9611,
"step": 294
},
{
"epoch": 0.23627684964200477,
"grad_norm": 2.0777316093444824,
"learning_rate": 9.87401147984998e-05,
"loss": 2.2418,
"step": 297
},
{
"epoch": 0.2386634844868735,
"grad_norm": 3.1081202030181885,
"learning_rate": 9.869571540427689e-05,
"loss": 2.1499,
"step": 300
},
{
"epoch": 0.24105011933174225,
"grad_norm": 1.9104883670806885,
"learning_rate": 9.865055749571215e-05,
"loss": 2.2008,
"step": 303
},
{
"epoch": 0.24343675417661098,
"grad_norm": 1.9581879377365112,
"learning_rate": 9.860464177621284e-05,
"loss": 1.9274,
"step": 306
},
{
"epoch": 0.2458233890214797,
"grad_norm": 2.2974860668182373,
"learning_rate": 9.855796896099045e-05,
"loss": 2.1965,
"step": 309
},
{
"epoch": 0.24821002386634844,
"grad_norm": 3.9561026096343994,
"learning_rate": 9.851053977704931e-05,
"loss": 1.9958,
"step": 312
},
{
"epoch": 0.25059665871121717,
"grad_norm": 1.9768146276474,
"learning_rate": 9.846235496317555e-05,
"loss": 1.9944,
"step": 315
},
{
"epoch": 0.2529832935560859,
"grad_norm": 1.7646844387054443,
"learning_rate": 9.841341526992536e-05,
"loss": 2.114,
"step": 318
},
{
"epoch": 0.2553699284009546,
"grad_norm": 1.744960904121399,
"learning_rate": 9.836372145961345e-05,
"loss": 2.1758,
"step": 321
},
{
"epoch": 0.2577565632458234,
"grad_norm": 1.8891479969024658,
"learning_rate": 9.83132743063011e-05,
"loss": 2.174,
"step": 324
},
{
"epoch": 0.26014319809069214,
"grad_norm": 2.207690477371216,
"learning_rate": 9.826207459578411e-05,
"loss": 1.9944,
"step": 327
},
{
"epoch": 0.26252983293556087,
"grad_norm": 2.369786500930786,
"learning_rate": 9.821012312558058e-05,
"loss": 1.9725,
"step": 330
},
{
"epoch": 0.2649164677804296,
"grad_norm": 1.8267792463302612,
"learning_rate": 9.815742070491852e-05,
"loss": 2.1383,
"step": 333
},
{
"epoch": 0.26730310262529833,
"grad_norm": 2.1697428226470947,
"learning_rate": 9.810396815472314e-05,
"loss": 2.036,
"step": 336
},
{
"epoch": 0.26968973747016706,
"grad_norm": 2.0004632472991943,
"learning_rate": 9.804976630760419e-05,
"loss": 2.077,
"step": 339
},
{
"epoch": 0.2720763723150358,
"grad_norm": 1.9695758819580078,
"learning_rate": 9.799481600784286e-05,
"loss": 1.9859,
"step": 342
},
{
"epoch": 0.2744630071599045,
"grad_norm": 2.275029420852661,
"learning_rate": 9.793911811137875e-05,
"loss": 2.0383,
"step": 345
},
{
"epoch": 0.27684964200477324,
"grad_norm": 2.0272061824798584,
"learning_rate": 9.788267348579648e-05,
"loss": 2.0818,
"step": 348
},
{
"epoch": 0.27923627684964203,
"grad_norm": 1.8712291717529297,
"learning_rate": 9.782548301031217e-05,
"loss": 2.0953,
"step": 351
},
{
"epoch": 0.28162291169451076,
"grad_norm": 2.198537826538086,
"learning_rate": 9.776754757575975e-05,
"loss": 2.0162,
"step": 354
},
{
"epoch": 0.2840095465393795,
"grad_norm": 2.8860132694244385,
"learning_rate": 9.770886808457709e-05,
"loss": 2.1242,
"step": 357
},
{
"epoch": 0.2863961813842482,
"grad_norm": 2.483168125152588,
"learning_rate": 9.764944545079196e-05,
"loss": 2.1847,
"step": 360
},
{
"epoch": 0.28878281622911695,
"grad_norm": 2.067625045776367,
"learning_rate": 9.758928060000778e-05,
"loss": 2.1283,
"step": 363
},
{
"epoch": 0.2911694510739857,
"grad_norm": 1.8211537599563599,
"learning_rate": 9.752837446938915e-05,
"loss": 1.9557,
"step": 366
},
{
"epoch": 0.2935560859188544,
"grad_norm": 1.7822015285491943,
"learning_rate": 9.746672800764735e-05,
"loss": 1.9696,
"step": 369
},
{
"epoch": 0.29594272076372313,
"grad_norm": 2.028804302215576,
"learning_rate": 9.740434217502547e-05,
"loss": 1.9916,
"step": 372
},
{
"epoch": 0.29832935560859186,
"grad_norm": 1.9470692873001099,
"learning_rate": 9.734121794328357e-05,
"loss": 2.0649,
"step": 375
},
{
"epoch": 0.30071599045346065,
"grad_norm": 2.003063201904297,
"learning_rate": 9.727735629568336e-05,
"loss": 2.0997,
"step": 378
},
{
"epoch": 0.3031026252983294,
"grad_norm": 3.469019889831543,
"learning_rate": 9.721275822697306e-05,
"loss": 2.0181,
"step": 381
},
{
"epoch": 0.3054892601431981,
"grad_norm": 1.79829740524292,
"learning_rate": 9.714742474337186e-05,
"loss": 2.1289,
"step": 384
},
{
"epoch": 0.30787589498806683,
"grad_norm": 1.9561500549316406,
"learning_rate": 9.708135686255416e-05,
"loss": 2.2278,
"step": 387
},
{
"epoch": 0.31026252983293556,
"grad_norm": 2.805483102798462,
"learning_rate": 9.701455561363379e-05,
"loss": 1.9664,
"step": 390
},
{
"epoch": 0.3126491646778043,
"grad_norm": 1.7945538759231567,
"learning_rate": 9.6947022037148e-05,
"loss": 2.0011,
"step": 393
},
{
"epoch": 0.315035799522673,
"grad_norm": 2.022679328918457,
"learning_rate": 9.687875718504126e-05,
"loss": 1.9724,
"step": 396
},
{
"epoch": 0.31742243436754175,
"grad_norm": 3.296950340270996,
"learning_rate": 9.680976212064874e-05,
"loss": 2.0533,
"step": 399
},
{
"epoch": 0.3198090692124105,
"grad_norm": 1.849898099899292,
"learning_rate": 9.674003791867991e-05,
"loss": 2.0683,
"step": 402
},
{
"epoch": 0.3221957040572792,
"grad_norm": 2.1924362182617188,
"learning_rate": 9.666958566520174e-05,
"loss": 2.1004,
"step": 405
},
{
"epoch": 0.324582338902148,
"grad_norm": 2.3722035884857178,
"learning_rate": 9.659840645762175e-05,
"loss": 2.131,
"step": 408
},
{
"epoch": 0.3269689737470167,
"grad_norm": 2.3362410068511963,
"learning_rate": 9.652650140467093e-05,
"loss": 2.0556,
"step": 411
},
{
"epoch": 0.32935560859188545,
"grad_norm": 1.6226763725280762,
"learning_rate": 9.645387162638652e-05,
"loss": 2.0418,
"step": 414
},
{
"epoch": 0.3317422434367542,
"grad_norm": 1.9308342933654785,
"learning_rate": 9.638051825409453e-05,
"loss": 2.2221,
"step": 417
},
{
"epoch": 0.3341288782816229,
"grad_norm": 1.530688762664795,
"learning_rate": 9.630644243039207e-05,
"loss": 1.9592,
"step": 420
},
{
"epoch": 0.33651551312649164,
"grad_norm": 1.967718482017517,
"learning_rate": 9.623164530912963e-05,
"loss": 2.1817,
"step": 423
},
{
"epoch": 0.33890214797136037,
"grad_norm": 1.803712248802185,
"learning_rate": 9.615612805539305e-05,
"loss": 1.9614,
"step": 426
},
{
"epoch": 0.3412887828162291,
"grad_norm": 1.9774303436279297,
"learning_rate": 9.607989184548543e-05,
"loss": 2.0609,
"step": 429
},
{
"epoch": 0.3436754176610978,
"grad_norm": 1.9333628416061401,
"learning_rate": 9.600293786690872e-05,
"loss": 2.0487,
"step": 432
},
{
"epoch": 0.3460620525059666,
"grad_norm": 1.785407543182373,
"learning_rate": 9.592526731834537e-05,
"loss": 2.1801,
"step": 435
},
{
"epoch": 0.34844868735083534,
"grad_norm": 2.0003998279571533,
"learning_rate": 9.584688140963944e-05,
"loss": 1.8782,
"step": 438
},
{
"epoch": 0.35083532219570407,
"grad_norm": 1.7324368953704834,
"learning_rate": 9.576778136177798e-05,
"loss": 2.0483,
"step": 441
},
{
"epoch": 0.3532219570405728,
"grad_norm": 1.8628923892974854,
"learning_rate": 9.568796840687184e-05,
"loss": 1.9283,
"step": 444
},
{
"epoch": 0.3556085918854415,
"grad_norm": 1.822113037109375,
"learning_rate": 9.560744378813659e-05,
"loss": 2.0287,
"step": 447
},
{
"epoch": 0.35799522673031026,
"grad_norm": 1.9748612642288208,
"learning_rate": 9.552620875987311e-05,
"loss": 1.9536,
"step": 450
},
{
"epoch": 0.360381861575179,
"grad_norm": 1.875239372253418,
"learning_rate": 9.544426458744804e-05,
"loss": 1.8725,
"step": 453
},
{
"epoch": 0.3627684964200477,
"grad_norm": 1.9234291315078735,
"learning_rate": 9.536161254727408e-05,
"loss": 1.9569,
"step": 456
},
{
"epoch": 0.36515513126491644,
"grad_norm": 1.6600582599639893,
"learning_rate": 9.527825392679012e-05,
"loss": 1.9387,
"step": 459
},
{
"epoch": 0.36754176610978523,
"grad_norm": 1.7599331140518188,
"learning_rate": 9.51941900244412e-05,
"loss": 1.941,
"step": 462
},
{
"epoch": 0.36992840095465396,
"grad_norm": 1.6681755781173706,
"learning_rate": 9.51094221496582e-05,
"loss": 1.9404,
"step": 465
},
{
"epoch": 0.3723150357995227,
"grad_norm": 1.8792699575424194,
"learning_rate": 9.502395162283759e-05,
"loss": 1.8169,
"step": 468
},
{
"epoch": 0.3747016706443914,
"grad_norm": 1.6947994232177734,
"learning_rate": 9.493777977532072e-05,
"loss": 1.9733,
"step": 471
},
{
"epoch": 0.37708830548926014,
"grad_norm": 1.909915804862976,
"learning_rate": 9.485090794937319e-05,
"loss": 1.9364,
"step": 474
},
{
"epoch": 0.3794749403341289,
"grad_norm": 1.6911386251449585,
"learning_rate": 9.476333749816382e-05,
"loss": 1.8554,
"step": 477
},
{
"epoch": 0.3818615751789976,
"grad_norm": 1.8519269227981567,
"learning_rate": 9.467506978574371e-05,
"loss": 1.8662,
"step": 480
},
{
"epoch": 0.38424821002386633,
"grad_norm": 1.8122614622116089,
"learning_rate": 9.45861061870249e-05,
"loss": 2.0183,
"step": 483
},
{
"epoch": 0.38663484486873506,
"grad_norm": 2.058873414993286,
"learning_rate": 9.449644808775902e-05,
"loss": 1.9499,
"step": 486
},
{
"epoch": 0.38902147971360385,
"grad_norm": 1.720900297164917,
"learning_rate": 9.44060968845156e-05,
"loss": 1.9129,
"step": 489
},
{
"epoch": 0.3914081145584726,
"grad_norm": 1.9670952558517456,
"learning_rate": 9.431505398466045e-05,
"loss": 2.0279,
"step": 492
},
{
"epoch": 0.3937947494033413,
"grad_norm": 1.800014853477478,
"learning_rate": 9.42233208063336e-05,
"loss": 1.8313,
"step": 495
},
{
"epoch": 0.39618138424821003,
"grad_norm": 2.6560916900634766,
"learning_rate": 9.413089877842736e-05,
"loss": 2.0691,
"step": 498
},
{
"epoch": 0.39856801909307876,
"grad_norm": 1.8405070304870605,
"learning_rate": 9.403778934056391e-05,
"loss": 2.0424,
"step": 501
},
{
"epoch": 0.4009546539379475,
"grad_norm": 1.798326849937439,
"learning_rate": 9.394399394307303e-05,
"loss": 2.2063,
"step": 504
},
{
"epoch": 0.4033412887828162,
"grad_norm": 1.9824241399765015,
"learning_rate": 9.384951404696933e-05,
"loss": 1.861,
"step": 507
},
{
"epoch": 0.40572792362768495,
"grad_norm": 2.272447109222412,
"learning_rate": 9.375435112392969e-05,
"loss": 2.0633,
"step": 510
},
{
"epoch": 0.4081145584725537,
"grad_norm": 1.8447043895721436,
"learning_rate": 9.365850665627016e-05,
"loss": 1.9229,
"step": 513
},
{
"epoch": 0.4105011933174224,
"grad_norm": 1.7020233869552612,
"learning_rate": 9.356198213692297e-05,
"loss": 1.8778,
"step": 516
},
{
"epoch": 0.4128878281622912,
"grad_norm": 1.7794997692108154,
"learning_rate": 9.346477906941331e-05,
"loss": 1.833,
"step": 519
},
{
"epoch": 0.4152744630071599,
"grad_norm": 1.775090217590332,
"learning_rate": 9.336689896783573e-05,
"loss": 1.8797,
"step": 522
},
{
"epoch": 0.41766109785202865,
"grad_norm": 1.8821420669555664,
"learning_rate": 9.32683433568308e-05,
"loss": 1.9237,
"step": 525
},
{
"epoch": 0.4200477326968974,
"grad_norm": 2.1922411918640137,
"learning_rate": 9.316911377156117e-05,
"loss": 1.9907,
"step": 528
},
{
"epoch": 0.4224343675417661,
"grad_norm": 1.8223942518234253,
"learning_rate": 9.306921175768775e-05,
"loss": 1.924,
"step": 531
},
{
"epoch": 0.42482100238663484,
"grad_norm": 1.8186813592910767,
"learning_rate": 9.29686388713456e-05,
"loss": 2.1474,
"step": 534
},
{
"epoch": 0.42720763723150357,
"grad_norm": 1.6708881855010986,
"learning_rate": 9.286739667911972e-05,
"loss": 1.956,
"step": 537
},
{
"epoch": 0.4295942720763723,
"grad_norm": 1.6984988451004028,
"learning_rate": 9.276548675802059e-05,
"loss": 1.9995,
"step": 540
},
{
"epoch": 0.431980906921241,
"grad_norm": 1.5586644411087036,
"learning_rate": 9.266291069545972e-05,
"loss": 1.9233,
"step": 543
},
{
"epoch": 0.4343675417661098,
"grad_norm": 1.7961502075195312,
"learning_rate": 9.255967008922474e-05,
"loss": 1.9941,
"step": 546
},
{
"epoch": 0.43675417661097854,
"grad_norm": 1.4892523288726807,
"learning_rate": 9.245576654745471e-05,
"loss": 2.0961,
"step": 549
},
{
"epoch": 0.43914081145584727,
"grad_norm": 1.635480523109436,
"learning_rate": 9.235120168861496e-05,
"loss": 1.8031,
"step": 552
},
{
"epoch": 0.441527446300716,
"grad_norm": 2.0391037464141846,
"learning_rate": 9.224597714147186e-05,
"loss": 2.0117,
"step": 555
},
{
"epoch": 0.4439140811455847,
"grad_norm": 1.8993330001831055,
"learning_rate": 9.214009454506753e-05,
"loss": 1.8346,
"step": 558
},
{
"epoch": 0.44630071599045346,
"grad_norm": 1.830111026763916,
"learning_rate": 9.203355554869428e-05,
"loss": 1.9372,
"step": 561
},
{
"epoch": 0.4486873508353222,
"grad_norm": 3.3967881202697754,
"learning_rate": 9.192636181186888e-05,
"loss": 1.7628,
"step": 564
},
{
"epoch": 0.4510739856801909,
"grad_norm": 1.7912521362304688,
"learning_rate": 9.181851500430673e-05,
"loss": 1.8183,
"step": 567
},
{
"epoch": 0.45346062052505964,
"grad_norm": 1.9313652515411377,
"learning_rate": 9.171001680589588e-05,
"loss": 1.8283,
"step": 570
},
{
"epoch": 0.45584725536992843,
"grad_norm": 2.1597900390625,
"learning_rate": 9.160086890667086e-05,
"loss": 1.8924,
"step": 573
},
{
"epoch": 0.45823389021479716,
"grad_norm": 1.6880937814712524,
"learning_rate": 9.14910730067863e-05,
"loss": 1.9477,
"step": 576
},
{
"epoch": 0.4606205250596659,
"grad_norm": 1.6879347562789917,
"learning_rate": 9.138063081649051e-05,
"loss": 2.0598,
"step": 579
},
{
"epoch": 0.4630071599045346,
"grad_norm": 1.8424838781356812,
"learning_rate": 9.126954405609882e-05,
"loss": 1.9193,
"step": 582
},
{
"epoch": 0.46539379474940334,
"grad_norm": 2.013153076171875,
"learning_rate": 9.115781445596676e-05,
"loss": 1.7798,
"step": 585
},
{
"epoch": 0.4677804295942721,
"grad_norm": 2.0057547092437744,
"learning_rate": 9.104544375646313e-05,
"loss": 2.1303,
"step": 588
},
{
"epoch": 0.4701670644391408,
"grad_norm": 1.924677848815918,
"learning_rate": 9.093243370794291e-05,
"loss": 1.9005,
"step": 591
},
{
"epoch": 0.47255369928400953,
"grad_norm": 1.8524523973464966,
"learning_rate": 9.081878607071996e-05,
"loss": 1.9822,
"step": 594
},
{
"epoch": 0.47494033412887826,
"grad_norm": 1.9615447521209717,
"learning_rate": 9.07045026150396e-05,
"loss": 1.9297,
"step": 597
},
{
"epoch": 0.477326968973747,
"grad_norm": 1.671190619468689,
"learning_rate": 9.058958512105104e-05,
"loss": 1.8514,
"step": 600
},
{
"epoch": 0.4797136038186158,
"grad_norm": 3.258155107498169,
"learning_rate": 9.047403537877971e-05,
"loss": 1.8893,
"step": 603
},
{
"epoch": 0.4821002386634845,
"grad_norm": 1.7724066972732544,
"learning_rate": 9.035785518809927e-05,
"loss": 1.9497,
"step": 606
},
{
"epoch": 0.48448687350835323,
"grad_norm": 1.958068609237671,
"learning_rate": 9.024104635870368e-05,
"loss": 1.917,
"step": 609
},
{
"epoch": 0.48687350835322196,
"grad_norm": 1.6382533311843872,
"learning_rate": 9.012361071007891e-05,
"loss": 1.9146,
"step": 612
},
{
"epoch": 0.4892601431980907,
"grad_norm": 3.012364625930786,
"learning_rate": 9.000555007147469e-05,
"loss": 1.9391,
"step": 615
},
{
"epoch": 0.4916467780429594,
"grad_norm": 2.1372387409210205,
"learning_rate": 8.988686628187597e-05,
"loss": 1.9188,
"step": 618
},
{
"epoch": 0.49403341288782815,
"grad_norm": 1.587446689605713,
"learning_rate": 8.976756118997427e-05,
"loss": 1.9319,
"step": 621
},
{
"epoch": 0.4964200477326969,
"grad_norm": 2.177863121032715,
"learning_rate": 8.964763665413893e-05,
"loss": 1.7627,
"step": 624
},
{
"epoch": 0.4988066825775656,
"grad_norm": 1.8430842161178589,
"learning_rate": 8.952709454238808e-05,
"loss": 2.0253,
"step": 627
},
{
"epoch": 0.5011933174224343,
"grad_norm": 1.6200464963912964,
"learning_rate": 8.940593673235962e-05,
"loss": 1.9912,
"step": 630
},
{
"epoch": 0.5035799522673031,
"grad_norm": 1.8540515899658203,
"learning_rate": 8.928416511128195e-05,
"loss": 1.8121,
"step": 633
},
{
"epoch": 0.5059665871121718,
"grad_norm": 1.7102622985839844,
"learning_rate": 8.916178157594453e-05,
"loss": 1.8878,
"step": 636
},
{
"epoch": 0.5083532219570406,
"grad_norm": 1.7976316213607788,
"learning_rate": 8.903878803266841e-05,
"loss": 2.0787,
"step": 639
},
{
"epoch": 0.5107398568019093,
"grad_norm": 1.524535894393921,
"learning_rate": 8.891518639727649e-05,
"loss": 1.772,
"step": 642
},
{
"epoch": 0.513126491646778,
"grad_norm": 1.793609380722046,
"learning_rate": 8.879097859506372e-05,
"loss": 1.9283,
"step": 645
},
{
"epoch": 0.5155131264916468,
"grad_norm": 1.700348973274231,
"learning_rate": 8.866616656076696e-05,
"loss": 1.8155,
"step": 648
},
{
"epoch": 0.5178997613365155,
"grad_norm": 1.611598253250122,
"learning_rate": 8.854075223853508e-05,
"loss": 1.9446,
"step": 651
},
{
"epoch": 0.5202863961813843,
"grad_norm": 1.7585581541061401,
"learning_rate": 8.841473758189854e-05,
"loss": 1.7997,
"step": 654
},
{
"epoch": 0.522673031026253,
"grad_norm": 1.823596715927124,
"learning_rate": 8.828812455373891e-05,
"loss": 2.0054,
"step": 657
},
{
"epoch": 0.5250596658711217,
"grad_norm": 1.7616968154907227,
"learning_rate": 8.816091512625843e-05,
"loss": 1.9281,
"step": 660
},
{
"epoch": 0.5274463007159904,
"grad_norm": 2.0290961265563965,
"learning_rate": 8.803311128094918e-05,
"loss": 1.9559,
"step": 663
},
{
"epoch": 0.5298329355608592,
"grad_norm": 1.957862377166748,
"learning_rate": 8.790471500856228e-05,
"loss": 2.0428,
"step": 666
},
{
"epoch": 0.5322195704057279,
"grad_norm": 1.768784999847412,
"learning_rate": 8.777572830907684e-05,
"loss": 2.1063,
"step": 669
},
{
"epoch": 0.5346062052505967,
"grad_norm": 1.6378014087677002,
"learning_rate": 8.764615319166886e-05,
"loss": 2.0242,
"step": 672
},
{
"epoch": 0.5369928400954654,
"grad_norm": 1.6081739664077759,
"learning_rate": 8.751599167467985e-05,
"loss": 1.7975,
"step": 675
},
{
"epoch": 0.5393794749403341,
"grad_norm": 1.6603517532348633,
"learning_rate": 8.738524578558547e-05,
"loss": 1.8832,
"step": 678
},
{
"epoch": 0.5417661097852029,
"grad_norm": 1.7453463077545166,
"learning_rate": 8.72539175609639e-05,
"loss": 1.7962,
"step": 681
},
{
"epoch": 0.5441527446300716,
"grad_norm": 1.584671139717102,
"learning_rate": 8.712200904646416e-05,
"loss": 1.7748,
"step": 684
},
{
"epoch": 0.5465393794749404,
"grad_norm": 1.787086844444275,
"learning_rate": 8.698952229677422e-05,
"loss": 1.8125,
"step": 687
},
{
"epoch": 0.548926014319809,
"grad_norm": 2.3501312732696533,
"learning_rate": 8.685645937558896e-05,
"loss": 2.0739,
"step": 690
},
{
"epoch": 0.5513126491646778,
"grad_norm": 1.6689587831497192,
"learning_rate": 8.67228223555781e-05,
"loss": 1.823,
"step": 693
},
{
"epoch": 0.5536992840095465,
"grad_norm": 1.7818080186843872,
"learning_rate": 8.658861331835385e-05,
"loss": 1.9059,
"step": 696
},
{
"epoch": 0.5560859188544153,
"grad_norm": 1.8221330642700195,
"learning_rate": 8.645383435443852e-05,
"loss": 1.8294,
"step": 699
},
{
"epoch": 0.5584725536992841,
"grad_norm": 1.7873986959457397,
"learning_rate": 8.631848756323197e-05,
"loss": 1.8954,
"step": 702
},
{
"epoch": 0.5608591885441527,
"grad_norm": 1.6317650079727173,
"learning_rate": 8.618257505297886e-05,
"loss": 1.9043,
"step": 705
},
{
"epoch": 0.5632458233890215,
"grad_norm": 1.8384369611740112,
"learning_rate": 8.604609894073584e-05,
"loss": 1.7076,
"step": 708
},
{
"epoch": 0.5656324582338902,
"grad_norm": 1.737593412399292,
"learning_rate": 8.590906135233854e-05,
"loss": 1.8403,
"step": 711
},
{
"epoch": 0.568019093078759,
"grad_norm": 1.7566007375717163,
"learning_rate": 8.577146442236857e-05,
"loss": 1.9017,
"step": 714
},
{
"epoch": 0.5704057279236276,
"grad_norm": 1.785732626914978,
"learning_rate": 8.563331029412012e-05,
"loss": 1.93,
"step": 717
},
{
"epoch": 0.5727923627684964,
"grad_norm": 2.0541484355926514,
"learning_rate": 8.549460111956664e-05,
"loss": 1.8559,
"step": 720
},
{
"epoch": 0.5751789976133651,
"grad_norm": 2.421417474746704,
"learning_rate": 8.535533905932738e-05,
"loss": 1.8517,
"step": 723
},
{
"epoch": 0.5775656324582339,
"grad_norm": 1.7755558490753174,
"learning_rate": 8.521552628263362e-05,
"loss": 1.8476,
"step": 726
},
{
"epoch": 0.5799522673031027,
"grad_norm": 1.632957100868225,
"learning_rate": 8.507516496729495e-05,
"loss": 1.8007,
"step": 729
},
{
"epoch": 0.5823389021479713,
"grad_norm": 1.681512475013733,
"learning_rate": 8.493425729966534e-05,
"loss": 1.921,
"step": 732
},
{
"epoch": 0.5847255369928401,
"grad_norm": 1.6569448709487915,
"learning_rate": 8.479280547460907e-05,
"loss": 1.8165,
"step": 735
},
{
"epoch": 0.5871121718377088,
"grad_norm": 1.630233645439148,
"learning_rate": 8.465081169546659e-05,
"loss": 1.9824,
"step": 738
},
{
"epoch": 0.5894988066825776,
"grad_norm": 1.7391397953033447,
"learning_rate": 8.450827817402011e-05,
"loss": 1.9757,
"step": 741
},
{
"epoch": 0.5918854415274463,
"grad_norm": 2.036689281463623,
"learning_rate": 8.436520713045922e-05,
"loss": 1.7685,
"step": 744
},
{
"epoch": 0.594272076372315,
"grad_norm": 1.4583113193511963,
"learning_rate": 8.422160079334628e-05,
"loss": 1.8458,
"step": 747
},
{
"epoch": 0.5966587112171837,
"grad_norm": 3.0061111450195312,
"learning_rate": 8.40774613995817e-05,
"loss": 1.895,
"step": 750
},
{
"epoch": 0.5990453460620525,
"grad_norm": 1.791383147239685,
"learning_rate": 8.393279119436912e-05,
"loss": 1.9162,
"step": 753
},
{
"epoch": 0.6014319809069213,
"grad_norm": 2.032726526260376,
"learning_rate": 8.378759243118044e-05,
"loss": 2.0624,
"step": 756
},
{
"epoch": 0.60381861575179,
"grad_norm": 1.6108125448226929,
"learning_rate": 8.364186737172068e-05,
"loss": 1.807,
"step": 759
},
{
"epoch": 0.6062052505966588,
"grad_norm": 1.748790979385376,
"learning_rate": 8.349561828589277e-05,
"loss": 2.0034,
"step": 762
},
{
"epoch": 0.6085918854415274,
"grad_norm": 1.6538299322128296,
"learning_rate": 8.33488474517622e-05,
"loss": 1.8479,
"step": 765
},
{
"epoch": 0.6109785202863962,
"grad_norm": 1.8360040187835693,
"learning_rate": 8.320155715552155e-05,
"loss": 1.7923,
"step": 768
},
{
"epoch": 0.6133651551312649,
"grad_norm": 1.8506301641464233,
"learning_rate": 8.305374969145488e-05,
"loss": 1.9816,
"step": 771
},
{
"epoch": 0.6157517899761337,
"grad_norm": 1.670559048652649,
"learning_rate": 8.290542736190188e-05,
"loss": 1.7492,
"step": 774
},
{
"epoch": 0.6181384248210023,
"grad_norm": 1.654943585395813,
"learning_rate": 8.275659247722222e-05,
"loss": 1.7765,
"step": 777
},
{
"epoch": 0.6205250596658711,
"grad_norm": 1.7641255855560303,
"learning_rate": 8.260724735575933e-05,
"loss": 1.8642,
"step": 780
},
{
"epoch": 0.6229116945107399,
"grad_norm": 2.0606367588043213,
"learning_rate": 8.24573943238045e-05,
"loss": 1.938,
"step": 783
},
{
"epoch": 0.6252983293556086,
"grad_norm": 1.590706706047058,
"learning_rate": 8.230703571556048e-05,
"loss": 1.7473,
"step": 786
},
{
"epoch": 0.6276849642004774,
"grad_norm": 1.6293972730636597,
"learning_rate": 8.215617387310524e-05,
"loss": 1.803,
"step": 789
},
{
"epoch": 0.630071599045346,
"grad_norm": 1.719766616821289,
"learning_rate": 8.200481114635536e-05,
"loss": 1.8393,
"step": 792
},
{
"epoch": 0.6324582338902148,
"grad_norm": 1.6134871244430542,
"learning_rate": 8.185294989302958e-05,
"loss": 1.7842,
"step": 795
},
{
"epoch": 0.6348448687350835,
"grad_norm": 1.6503950357437134,
"learning_rate": 8.170059247861194e-05,
"loss": 1.7826,
"step": 798
},
{
"epoch": 0.6372315035799523,
"grad_norm": 1.8185105323791504,
"learning_rate": 8.154774127631501e-05,
"loss": 1.7388,
"step": 801
},
{
"epoch": 0.639618138424821,
"grad_norm": 1.7190958261489868,
"learning_rate": 8.139439866704293e-05,
"loss": 1.833,
"step": 804
},
{
"epoch": 0.6420047732696897,
"grad_norm": 1.6161839962005615,
"learning_rate": 8.124056703935423e-05,
"loss": 1.801,
"step": 807
},
{
"epoch": 0.6443914081145584,
"grad_norm": 1.5883020162582397,
"learning_rate": 8.108624878942477e-05,
"loss": 1.8195,
"step": 810
},
{
"epoch": 0.6467780429594272,
"grad_norm": 1.6530261039733887,
"learning_rate": 8.093144632101026e-05,
"loss": 1.7564,
"step": 813
},
{
"epoch": 0.649164677804296,
"grad_norm": 1.6084448099136353,
"learning_rate": 8.077616204540897e-05,
"loss": 1.8133,
"step": 816
},
{
"epoch": 0.6515513126491647,
"grad_norm": 1.791277289390564,
"learning_rate": 8.062039838142402e-05,
"loss": 1.7978,
"step": 819
},
{
"epoch": 0.6539379474940334,
"grad_norm": 1.726449728012085,
"learning_rate": 8.046415775532585e-05,
"loss": 1.7988,
"step": 822
},
{
"epoch": 0.6563245823389021,
"grad_norm": 1.9136977195739746,
"learning_rate": 8.030744260081426e-05,
"loss": 1.826,
"step": 825
},
{
"epoch": 0.6587112171837709,
"grad_norm": 1.9753978252410889,
"learning_rate": 8.015025535898073e-05,
"loss": 1.8964,
"step": 828
},
{
"epoch": 0.6610978520286396,
"grad_norm": 1.6367616653442383,
"learning_rate": 7.999259847827015e-05,
"loss": 1.8764,
"step": 831
},
{
"epoch": 0.6634844868735084,
"grad_norm": 1.7397571802139282,
"learning_rate": 7.983447441444281e-05,
"loss": 1.72,
"step": 834
},
{
"epoch": 0.665871121718377,
"grad_norm": 2.1970224380493164,
"learning_rate": 7.967588563053616e-05,
"loss": 1.7779,
"step": 837
},
{
"epoch": 0.6682577565632458,
"grad_norm": 1.6252713203430176,
"learning_rate": 7.951683459682641e-05,
"loss": 1.8043,
"step": 840
},
{
"epoch": 0.6706443914081146,
"grad_norm": 1.6703697443008423,
"learning_rate": 7.935732379079008e-05,
"loss": 2.0202,
"step": 843
},
{
"epoch": 0.6730310262529833,
"grad_norm": 2.3959710597991943,
"learning_rate": 7.919735569706533e-05,
"loss": 1.9039,
"step": 846
},
{
"epoch": 0.6754176610978521,
"grad_norm": 1.7018755674362183,
"learning_rate": 7.903693280741331e-05,
"loss": 1.8505,
"step": 849
},
{
"epoch": 0.6778042959427207,
"grad_norm": 1.9524502754211426,
"learning_rate": 7.887605762067945e-05,
"loss": 1.7577,
"step": 852
},
{
"epoch": 0.6801909307875895,
"grad_norm": 1.9558862447738647,
"learning_rate": 7.871473264275429e-05,
"loss": 1.749,
"step": 855
},
{
"epoch": 0.6825775656324582,
"grad_norm": 1.880624532699585,
"learning_rate": 7.855296038653475e-05,
"loss": 2.0309,
"step": 858
},
{
"epoch": 0.684964200477327,
"grad_norm": 1.5008827447891235,
"learning_rate": 7.83907433718847e-05,
"loss": 1.7858,
"step": 861
},
{
"epoch": 0.6873508353221957,
"grad_norm": 1.7783069610595703,
"learning_rate": 7.82280841255959e-05,
"loss": 1.755,
"step": 864
},
{
"epoch": 0.6897374701670644,
"grad_norm": 1.6861096620559692,
"learning_rate": 7.80649851813486e-05,
"loss": 1.912,
"step": 867
},
{
"epoch": 0.6921241050119332,
"grad_norm": 1.5460323095321655,
"learning_rate": 7.790144907967201e-05,
"loss": 1.7781,
"step": 870
},
{
"epoch": 0.6945107398568019,
"grad_norm": 1.703471302986145,
"learning_rate": 7.773747836790481e-05,
"loss": 1.9725,
"step": 873
},
{
"epoch": 0.6968973747016707,
"grad_norm": 1.86622953414917,
"learning_rate": 7.757307560015538e-05,
"loss": 1.7674,
"step": 876
},
{
"epoch": 0.6992840095465394,
"grad_norm": 1.678232192993164,
"learning_rate": 7.740824333726213e-05,
"loss": 1.731,
"step": 879
},
{
"epoch": 0.7016706443914081,
"grad_norm": 1.9272021055221558,
"learning_rate": 7.724298414675353e-05,
"loss": 1.7698,
"step": 882
},
{
"epoch": 0.7040572792362768,
"grad_norm": 1.73203444480896,
"learning_rate": 7.707730060280812e-05,
"loss": 1.913,
"step": 885
},
{
"epoch": 0.7064439140811456,
"grad_norm": 1.5196224451065063,
"learning_rate": 7.691119528621444e-05,
"loss": 1.8283,
"step": 888
},
{
"epoch": 0.7088305489260143,
"grad_norm": 1.798051357269287,
"learning_rate": 7.674467078433081e-05,
"loss": 1.9543,
"step": 891
},
{
"epoch": 0.711217183770883,
"grad_norm": 1.6706217527389526,
"learning_rate": 7.657772969104508e-05,
"loss": 1.6487,
"step": 894
},
{
"epoch": 0.7136038186157518,
"grad_norm": 1.788098692893982,
"learning_rate": 7.641037460673412e-05,
"loss": 1.6219,
"step": 897
},
{
"epoch": 0.7159904534606205,
"grad_norm": 1.5081647634506226,
"learning_rate": 7.624260813822342e-05,
"loss": 1.7062,
"step": 900
},
{
"epoch": 0.7183770883054893,
"grad_norm": 1.6908079385757446,
"learning_rate": 7.607443289874642e-05,
"loss": 1.8747,
"step": 903
},
{
"epoch": 0.720763723150358,
"grad_norm": 1.4784934520721436,
"learning_rate": 7.590585150790389e-05,
"loss": 1.9055,
"step": 906
},
{
"epoch": 0.7231503579952268,
"grad_norm": 1.4617854356765747,
"learning_rate": 7.573686659162293e-05,
"loss": 1.8485,
"step": 909
},
{
"epoch": 0.7255369928400954,
"grad_norm": 1.7699053287506104,
"learning_rate": 7.556748078211635e-05,
"loss": 1.9475,
"step": 912
},
{
"epoch": 0.7279236276849642,
"grad_norm": 1.409713625907898,
"learning_rate": 7.53976967178414e-05,
"loss": 1.8663,
"step": 915
},
{
"epoch": 0.7303102625298329,
"grad_norm": 1.4845449924468994,
"learning_rate": 7.522751704345887e-05,
"loss": 1.8818,
"step": 918
},
{
"epoch": 0.7326968973747017,
"grad_norm": 1.5580782890319824,
"learning_rate": 7.505694440979178e-05,
"loss": 1.8258,
"step": 921
},
{
"epoch": 0.7350835322195705,
"grad_norm": 1.523646593093872,
"learning_rate": 7.488598147378416e-05,
"loss": 1.6651,
"step": 924
},
{
"epoch": 0.7374701670644391,
"grad_norm": 1.6686230897903442,
"learning_rate": 7.471463089845956e-05,
"loss": 1.7406,
"step": 927
},
{
"epoch": 0.7398568019093079,
"grad_norm": 1.5957698822021484,
"learning_rate": 7.454289535287968e-05,
"loss": 1.7128,
"step": 930
},
{
"epoch": 0.7422434367541766,
"grad_norm": 1.6253639459609985,
"learning_rate": 7.437077751210279e-05,
"loss": 1.9308,
"step": 933
},
{
"epoch": 0.7446300715990454,
"grad_norm": 1.6088500022888184,
"learning_rate": 7.419828005714194e-05,
"loss": 1.8083,
"step": 936
},
{
"epoch": 0.747016706443914,
"grad_norm": 1.4673054218292236,
"learning_rate": 7.402540567492337e-05,
"loss": 1.7764,
"step": 939
},
{
"epoch": 0.7494033412887828,
"grad_norm": 1.5151442289352417,
"learning_rate": 7.385215705824449e-05,
"loss": 1.974,
"step": 942
},
{
"epoch": 0.7517899761336515,
"grad_norm": 1.7416108846664429,
"learning_rate": 7.367853690573208e-05,
"loss": 1.638,
"step": 945
},
{
"epoch": 0.7541766109785203,
"grad_norm": 1.7864446640014648,
"learning_rate": 7.350454792180016e-05,
"loss": 1.7128,
"step": 948
},
{
"epoch": 0.7565632458233891,
"grad_norm": 1.6676292419433594,
"learning_rate": 7.333019281660789e-05,
"loss": 1.9339,
"step": 951
},
{
"epoch": 0.7589498806682577,
"grad_norm": 1.4791321754455566,
"learning_rate": 7.31554743060174e-05,
"loss": 1.7184,
"step": 954
},
{
"epoch": 0.7613365155131265,
"grad_norm": 1.6596827507019043,
"learning_rate": 7.298039511155138e-05,
"loss": 1.7909,
"step": 957
},
{
"epoch": 0.7637231503579952,
"grad_norm": 1.8340952396392822,
"learning_rate": 7.280495796035079e-05,
"loss": 1.8153,
"step": 960
},
{
"epoch": 0.766109785202864,
"grad_norm": 1.6827830076217651,
"learning_rate": 7.262916558513237e-05,
"loss": 1.6396,
"step": 963
},
{
"epoch": 0.7684964200477327,
"grad_norm": 2.0441389083862305,
"learning_rate": 7.245302072414601e-05,
"loss": 1.7888,
"step": 966
},
{
"epoch": 0.7708830548926014,
"grad_norm": 1.436802864074707,
"learning_rate": 7.227652612113213e-05,
"loss": 1.6449,
"step": 969
},
{
"epoch": 0.7732696897374701,
"grad_norm": 1.5358387231826782,
"learning_rate": 7.209968452527896e-05,
"loss": 1.742,
"step": 972
},
{
"epoch": 0.7756563245823389,
"grad_norm": 1.583846092224121,
"learning_rate": 7.192249869117971e-05,
"loss": 1.8354,
"step": 975
},
{
"epoch": 0.7780429594272077,
"grad_norm": 1.6381198167800903,
"learning_rate": 7.174497137878966e-05,
"loss": 1.7394,
"step": 978
},
{
"epoch": 0.7804295942720764,
"grad_norm": 1.707513689994812,
"learning_rate": 7.156710535338312e-05,
"loss": 1.8892,
"step": 981
},
{
"epoch": 0.7828162291169452,
"grad_norm": 1.7260991334915161,
"learning_rate": 7.138890338551048e-05,
"loss": 1.832,
"step": 984
},
{
"epoch": 0.7852028639618138,
"grad_norm": 1.486690878868103,
"learning_rate": 7.121036825095492e-05,
"loss": 1.6662,
"step": 987
},
{
"epoch": 0.7875894988066826,
"grad_norm": 1.551670789718628,
"learning_rate": 7.103150273068921e-05,
"loss": 1.7214,
"step": 990
},
{
"epoch": 0.7899761336515513,
"grad_norm": 1.50273859500885,
"learning_rate": 7.085230961083249e-05,
"loss": 1.842,
"step": 993
},
{
"epoch": 0.7923627684964201,
"grad_norm": 1.8055702447891235,
"learning_rate": 7.067279168260671e-05,
"loss": 1.809,
"step": 996
},
{
"epoch": 0.7947494033412887,
"grad_norm": 1.560795545578003,
"learning_rate": 7.04929517422933e-05,
"loss": 1.7781,
"step": 999
},
{
"epoch": 0.7971360381861575,
"grad_norm": 1.615140438079834,
"learning_rate": 7.031279259118946e-05,
"loss": 1.6276,
"step": 1002
},
{
"epoch": 0.7995226730310262,
"grad_norm": 1.745884656906128,
"learning_rate": 7.013231703556471e-05,
"loss": 1.8841,
"step": 1005
},
{
"epoch": 0.801909307875895,
"grad_norm": 1.553002953529358,
"learning_rate": 6.995152788661705e-05,
"loss": 1.772,
"step": 1008
},
{
"epoch": 0.8042959427207638,
"grad_norm": 1.5847023725509644,
"learning_rate": 6.977042796042917e-05,
"loss": 1.7492,
"step": 1011
},
{
"epoch": 0.8066825775656324,
"grad_norm": 1.5925689935684204,
"learning_rate": 6.958902007792466e-05,
"loss": 1.7523,
"step": 1014
},
{
"epoch": 0.8090692124105012,
"grad_norm": 2.4262924194335938,
"learning_rate": 6.940730706482399e-05,
"loss": 1.8089,
"step": 1017
},
{
"epoch": 0.8114558472553699,
"grad_norm": 1.591212272644043,
"learning_rate": 6.922529175160054e-05,
"loss": 1.6939,
"step": 1020
},
{
"epoch": 0.8138424821002387,
"grad_norm": 1.5190821886062622,
"learning_rate": 6.904297697343655e-05,
"loss": 1.9102,
"step": 1023
},
{
"epoch": 0.8162291169451074,
"grad_norm": 1.556463599205017,
"learning_rate": 6.886036557017881e-05,
"loss": 1.8486,
"step": 1026
},
{
"epoch": 0.8186157517899761,
"grad_norm": 1.6876753568649292,
"learning_rate": 6.867746038629462e-05,
"loss": 1.9492,
"step": 1029
},
{
"epoch": 0.8210023866348448,
"grad_norm": 1.4290738105773926,
"learning_rate": 6.849426427082735e-05,
"loss": 1.7682,
"step": 1032
},
{
"epoch": 0.8233890214797136,
"grad_norm": 1.6473435163497925,
"learning_rate": 6.83107800773521e-05,
"loss": 1.8299,
"step": 1035
},
{
"epoch": 0.8257756563245824,
"grad_norm": 1.6430293321609497,
"learning_rate": 6.812701066393124e-05,
"loss": 1.7187,
"step": 1038
},
{
"epoch": 0.8281622911694511,
"grad_norm": 1.537840485572815,
"learning_rate": 6.79429588930699e-05,
"loss": 1.7417,
"step": 1041
},
{
"epoch": 0.8305489260143198,
"grad_norm": 2.192531108856201,
"learning_rate": 6.775862763167142e-05,
"loss": 1.7398,
"step": 1044
},
{
"epoch": 0.8329355608591885,
"grad_norm": 1.550441861152649,
"learning_rate": 6.757401975099262e-05,
"loss": 1.6647,
"step": 1047
},
{
"epoch": 0.8353221957040573,
"grad_norm": 3.2861685752868652,
"learning_rate": 6.738913812659912e-05,
"loss": 1.8928,
"step": 1050
},
{
"epoch": 0.837708830548926,
"grad_norm": 2.5351181030273438,
"learning_rate": 6.720398563832055e-05,
"loss": 1.7647,
"step": 1053
},
{
"epoch": 0.8400954653937948,
"grad_norm": 1.4341228008270264,
"learning_rate": 6.701856517020565e-05,
"loss": 1.8559,
"step": 1056
},
{
"epoch": 0.8424821002386634,
"grad_norm": 1.5583041906356812,
"learning_rate": 6.683287961047742e-05,
"loss": 1.9764,
"step": 1059
},
{
"epoch": 0.8448687350835322,
"grad_norm": 1.6127870082855225,
"learning_rate": 6.664693185148807e-05,
"loss": 1.7269,
"step": 1062
},
{
"epoch": 0.847255369928401,
"grad_norm": 1.662477731704712,
"learning_rate": 6.646072478967397e-05,
"loss": 1.8924,
"step": 1065
},
{
"epoch": 0.8496420047732697,
"grad_norm": 1.673250675201416,
"learning_rate": 6.627426132551058e-05,
"loss": 1.7657,
"step": 1068
},
{
"epoch": 0.8520286396181385,
"grad_norm": 1.5422923564910889,
"learning_rate": 6.608754436346725e-05,
"loss": 1.689,
"step": 1071
},
{
"epoch": 0.8544152744630071,
"grad_norm": 1.7752854824066162,
"learning_rate": 6.590057681196191e-05,
"loss": 1.6743,
"step": 1074
},
{
"epoch": 0.8568019093078759,
"grad_norm": 1.5732386112213135,
"learning_rate": 6.571336158331589e-05,
"loss": 1.967,
"step": 1077
},
{
"epoch": 0.8591885441527446,
"grad_norm": 1.6915782690048218,
"learning_rate": 6.552590159370844e-05,
"loss": 1.6303,
"step": 1080
},
{
"epoch": 0.8615751789976134,
"grad_norm": 1.5300819873809814,
"learning_rate": 6.53381997631314e-05,
"loss": 1.855,
"step": 1083
},
{
"epoch": 0.863961813842482,
"grad_norm": 1.5509366989135742,
"learning_rate": 6.515025901534364e-05,
"loss": 1.7637,
"step": 1086
},
{
"epoch": 0.8663484486873508,
"grad_norm": 1.4942635297775269,
"learning_rate": 6.496208227782556e-05,
"loss": 1.7456,
"step": 1089
},
{
"epoch": 0.8687350835322196,
"grad_norm": 1.7065314054489136,
"learning_rate": 6.477367248173352e-05,
"loss": 1.7814,
"step": 1092
},
{
"epoch": 0.8711217183770883,
"grad_norm": 1.5171153545379639,
"learning_rate": 6.458503256185404e-05,
"loss": 1.7815,
"step": 1095
},
{
"epoch": 0.8735083532219571,
"grad_norm": 1.507451057434082,
"learning_rate": 6.439616545655834e-05,
"loss": 1.6747,
"step": 1098
},
{
"epoch": 0.8758949880668258,
"grad_norm": 1.4318897724151611,
"learning_rate": 6.420707410775626e-05,
"loss": 1.8103,
"step": 1101
},
{
"epoch": 0.8782816229116945,
"grad_norm": 1.7363917827606201,
"learning_rate": 6.401776146085072e-05,
"loss": 2.0862,
"step": 1104
},
{
"epoch": 0.8806682577565632,
"grad_norm": 1.6215336322784424,
"learning_rate": 6.382823046469167e-05,
"loss": 1.831,
"step": 1107
},
{
"epoch": 0.883054892601432,
"grad_norm": 1.4969305992126465,
"learning_rate": 6.363848407153016e-05,
"loss": 1.6292,
"step": 1110
},
{
"epoch": 0.8854415274463007,
"grad_norm": 1.5934969186782837,
"learning_rate": 6.344852523697247e-05,
"loss": 1.8363,
"step": 1113
},
{
"epoch": 0.8878281622911695,
"grad_norm": 1.6700190305709839,
"learning_rate": 6.325835691993394e-05,
"loss": 1.5579,
"step": 1116
},
{
"epoch": 0.8902147971360382,
"grad_norm": 1.6834436655044556,
"learning_rate": 6.306798208259297e-05,
"loss": 1.7307,
"step": 1119
},
{
"epoch": 0.8926014319809069,
"grad_norm": 1.5143299102783203,
"learning_rate": 6.287740369034485e-05,
"loss": 1.5429,
"step": 1122
},
{
"epoch": 0.8949880668257757,
"grad_norm": 1.6106433868408203,
"learning_rate": 6.26866247117555e-05,
"loss": 1.6108,
"step": 1125
},
{
"epoch": 0.8973747016706444,
"grad_norm": 1.6025004386901855,
"learning_rate": 6.249564811851543e-05,
"loss": 1.8183,
"step": 1128
},
{
"epoch": 0.8997613365155132,
"grad_norm": 1.3973323106765747,
"learning_rate": 6.230447688539316e-05,
"loss": 1.6247,
"step": 1131
},
{
"epoch": 0.9021479713603818,
"grad_norm": 1.7466994524002075,
"learning_rate": 6.211311399018916e-05,
"loss": 1.7925,
"step": 1134
},
{
"epoch": 0.9045346062052506,
"grad_norm": 1.537386178970337,
"learning_rate": 6.192156241368929e-05,
"loss": 1.7553,
"step": 1137
},
{
"epoch": 0.9069212410501193,
"grad_norm": 1.7690049409866333,
"learning_rate": 6.172982513961845e-05,
"loss": 1.7205,
"step": 1140
},
{
"epoch": 0.9093078758949881,
"grad_norm": 1.7017157077789307,
"learning_rate": 6.153790515459404e-05,
"loss": 1.6503,
"step": 1143
},
{
"epoch": 0.9116945107398569,
"grad_norm": 1.6713097095489502,
"learning_rate": 6.13458054480795e-05,
"loss": 1.8414,
"step": 1146
},
{
"epoch": 0.9140811455847255,
"grad_norm": 1.5956037044525146,
"learning_rate": 6.115352901233779e-05,
"loss": 1.8079,
"step": 1149
},
{
"epoch": 0.9164677804295943,
"grad_norm": 1.4910719394683838,
"learning_rate": 6.096107884238458e-05,
"loss": 1.6404,
"step": 1152
},
{
"epoch": 0.918854415274463,
"grad_norm": 1.4896982908248901,
"learning_rate": 6.0768457935941817e-05,
"loss": 1.7845,
"step": 1155
},
{
"epoch": 0.9212410501193318,
"grad_norm": 1.8766003847122192,
"learning_rate": 6.0575669293390954e-05,
"loss": 1.8104,
"step": 1158
},
{
"epoch": 0.9236276849642004,
"grad_norm": 1.8520368337631226,
"learning_rate": 6.038271591772615e-05,
"loss": 1.8249,
"step": 1161
},
{
"epoch": 0.9260143198090692,
"grad_norm": 1.7131208181381226,
"learning_rate": 6.0189600814507604e-05,
"loss": 1.8909,
"step": 1164
},
{
"epoch": 0.9284009546539379,
"grad_norm": 1.3130649328231812,
"learning_rate": 5.9996326991814654e-05,
"loss": 1.5973,
"step": 1167
},
{
"epoch": 0.9307875894988067,
"grad_norm": 1.7194204330444336,
"learning_rate": 5.980289746019892e-05,
"loss": 1.8653,
"step": 1170
},
{
"epoch": 0.9331742243436754,
"grad_norm": 1.6134217977523804,
"learning_rate": 5.9609315232637483e-05,
"loss": 1.5751,
"step": 1173
},
{
"epoch": 0.9355608591885441,
"grad_norm": 1.5093486309051514,
"learning_rate": 5.941558332448589e-05,
"loss": 1.6214,
"step": 1176
},
{
"epoch": 0.9379474940334129,
"grad_norm": 1.4754223823547363,
"learning_rate": 5.922170475343125e-05,
"loss": 1.69,
"step": 1179
},
{
"epoch": 0.9403341288782816,
"grad_norm": 1.7144575119018555,
"learning_rate": 5.9027682539445104e-05,
"loss": 1.6786,
"step": 1182
},
{
"epoch": 0.9427207637231504,
"grad_norm": 1.453788161277771,
"learning_rate": 5.883351970473654e-05,
"loss": 1.8321,
"step": 1185
},
{
"epoch": 0.9451073985680191,
"grad_norm": 1.688596248626709,
"learning_rate": 5.863921927370498e-05,
"loss": 1.7347,
"step": 1188
},
{
"epoch": 0.9474940334128878,
"grad_norm": 1.600243330001831,
"learning_rate": 5.8444784272893175e-05,
"loss": 1.678,
"step": 1191
},
{
"epoch": 0.9498806682577565,
"grad_norm": 1.6450378894805908,
"learning_rate": 5.8250217730939973e-05,
"loss": 1.6785,
"step": 1194
},
{
"epoch": 0.9522673031026253,
"grad_norm": 1.597424030303955,
"learning_rate": 5.8055522678533225e-05,
"loss": 1.6594,
"step": 1197
},
{
"epoch": 0.954653937947494,
"grad_norm": 1.7151094675064087,
"learning_rate": 5.786070214836254e-05,
"loss": 1.6255,
"step": 1200
},
{
"epoch": 0.9570405727923628,
"grad_norm": 1.4792280197143555,
"learning_rate": 5.7665759175072034e-05,
"loss": 1.8538,
"step": 1203
},
{
"epoch": 0.9594272076372315,
"grad_norm": 1.4509261846542358,
"learning_rate": 5.747069679521305e-05,
"loss": 1.7551,
"step": 1206
},
{
"epoch": 0.9618138424821002,
"grad_norm": 1.5576130151748657,
"learning_rate": 5.727551804719693e-05,
"loss": 1.7022,
"step": 1209
},
{
"epoch": 0.964200477326969,
"grad_norm": 1.6218217611312866,
"learning_rate": 5.708022597124758e-05,
"loss": 1.675,
"step": 1212
},
{
"epoch": 0.9665871121718377,
"grad_norm": 1.6137958765029907,
"learning_rate": 5.688482360935423e-05,
"loss": 1.8624,
"step": 1215
},
{
"epoch": 0.9689737470167065,
"grad_norm": 1.4816503524780273,
"learning_rate": 5.668931400522396e-05,
"loss": 1.791,
"step": 1218
},
{
"epoch": 0.9713603818615751,
"grad_norm": 1.543349027633667,
"learning_rate": 5.649370020423431e-05,
"loss": 1.6761,
"step": 1221
},
{
"epoch": 0.9737470167064439,
"grad_norm": 1.5985081195831299,
"learning_rate": 5.629798525338589e-05,
"loss": 1.6904,
"step": 1224
},
{
"epoch": 0.9761336515513126,
"grad_norm": 1.6068087816238403,
"learning_rate": 5.6102172201254835e-05,
"loss": 1.7542,
"step": 1227
},
{
"epoch": 0.9785202863961814,
"grad_norm": 1.3910298347473145,
"learning_rate": 5.5906264097945407e-05,
"loss": 1.7712,
"step": 1230
},
{
"epoch": 0.9809069212410502,
"grad_norm": 1.47352135181427,
"learning_rate": 5.5710263995042434e-05,
"loss": 1.8361,
"step": 1233
},
{
"epoch": 0.9832935560859188,
"grad_norm": 1.5060720443725586,
"learning_rate": 5.551417494556376e-05,
"loss": 1.7454,
"step": 1236
},
{
"epoch": 0.9856801909307876,
"grad_norm": 1.4085941314697266,
"learning_rate": 5.531800000391275e-05,
"loss": 1.7465,
"step": 1239
},
{
"epoch": 0.9880668257756563,
"grad_norm": 1.5695737600326538,
"learning_rate": 5.5121742225830665e-05,
"loss": 1.8416,
"step": 1242
},
{
"epoch": 0.9904534606205251,
"grad_norm": 1.7752211093902588,
"learning_rate": 5.4925404668349076e-05,
"loss": 1.779,
"step": 1245
},
{
"epoch": 0.9928400954653938,
"grad_norm": 1.3961496353149414,
"learning_rate": 5.472899038974225e-05,
"loss": 1.6866,
"step": 1248
},
{
"epoch": 0.9952267303102625,
"grad_norm": 1.5418916940689087,
"learning_rate": 5.45325024494795e-05,
"loss": 1.7653,
"step": 1251
},
{
"epoch": 0.9976133651551312,
"grad_norm": 1.6181163787841797,
"learning_rate": 5.433594390817756e-05,
"loss": 1.8697,
"step": 1254
},
{
"epoch": 1.0,
"grad_norm": 2.188194990158081,
"learning_rate": 5.413931782755283e-05,
"loss": 1.8344,
"step": 1257
}
],
"logging_steps": 3,
"max_steps": 2514,
"num_input_tokens_seen": 0,
"num_train_epochs": 2,
"save_steps": 1257,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": false
},
"attributes": {}
}
},
"total_flos": 2.127117763309732e+17,
"train_batch_size": 8,
"trial_name": null,
"trial_params": null
}