test7 / checkpoint-402 /trainer_state.json
frjonah's picture
Upload folder using huggingface_hub
a91aa16 verified
Raw
History Blame Contribute Delete
35.3 kB
{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 28.714285714285715,
"eval_steps": 500,
"global_step": 402,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.14285714285714285,
"grad_norm": 49.90940856933594,
"learning_rate": 1.6666666666666667e-06,
"loss": 6.0076,
"step": 2
},
{
"epoch": 0.2857142857142857,
"grad_norm": 44.15715789794922,
"learning_rate": 5e-06,
"loss": 5.6559,
"step": 4
},
{
"epoch": 0.42857142857142855,
"grad_norm": 56.16670608520508,
"learning_rate": 8.333333333333334e-06,
"loss": 4.4368,
"step": 6
},
{
"epoch": 0.5714285714285714,
"grad_norm": 23.919437408447266,
"learning_rate": 1.1666666666666668e-05,
"loss": 2.7657,
"step": 8
},
{
"epoch": 0.7142857142857143,
"grad_norm": 20.757640838623047,
"learning_rate": 1.5000000000000002e-05,
"loss": 2.3404,
"step": 10
},
{
"epoch": 0.8571428571428571,
"grad_norm": 16.067575454711914,
"learning_rate": 1.8333333333333333e-05,
"loss": 2.038,
"step": 12
},
{
"epoch": 1.0,
"grad_norm": 23.458139419555664,
"learning_rate": 1.9999675557165282e-05,
"loss": 1.5809,
"step": 14
},
{
"epoch": 1.1428571428571428,
"grad_norm": 11.104408264160156,
"learning_rate": 1.9997080140801932e-05,
"loss": 1.1148,
"step": 16
},
{
"epoch": 1.2857142857142856,
"grad_norm": 11.328217506408691,
"learning_rate": 1.9991889981715696e-05,
"loss": 1.2528,
"step": 18
},
{
"epoch": 1.4285714285714286,
"grad_norm": 12.313949584960938,
"learning_rate": 1.9984106427012667e-05,
"loss": 1.2324,
"step": 20
},
{
"epoch": 1.5714285714285714,
"grad_norm": 10.425058364868164,
"learning_rate": 1.9973731496914914e-05,
"loss": 1.0203,
"step": 22
},
{
"epoch": 1.7142857142857144,
"grad_norm": 12.36324691772461,
"learning_rate": 1.9960767884236132e-05,
"loss": 1.0997,
"step": 24
},
{
"epoch": 1.8571428571428572,
"grad_norm": 14.280921936035156,
"learning_rate": 1.9945218953682736e-05,
"loss": 1.43,
"step": 26
},
{
"epoch": 2.0,
"grad_norm": 20.082475662231445,
"learning_rate": 1.992708874098054e-05,
"loss": 1.0748,
"step": 28
},
{
"epoch": 2.142857142857143,
"grad_norm": 5.667484283447266,
"learning_rate": 1.9906381951827295e-05,
"loss": 0.3794,
"step": 30
},
{
"epoch": 2.2857142857142856,
"grad_norm": 6.214536666870117,
"learning_rate": 1.9883103960671305e-05,
"loss": 0.2747,
"step": 32
},
{
"epoch": 2.4285714285714284,
"grad_norm": 6.836840629577637,
"learning_rate": 1.985726080931651e-05,
"loss": 0.2568,
"step": 34
},
{
"epoch": 2.571428571428571,
"grad_norm": 13.044551849365234,
"learning_rate": 1.9828859205354326e-05,
"loss": 0.2937,
"step": 36
},
{
"epoch": 2.7142857142857144,
"grad_norm": 8.866482734680176,
"learning_rate": 1.979790652042268e-05,
"loss": 0.4252,
"step": 38
},
{
"epoch": 2.857142857142857,
"grad_norm": 8.709479331970215,
"learning_rate": 1.9764410788292724e-05,
"loss": 0.3599,
"step": 40
},
{
"epoch": 3.0,
"grad_norm": 9.43436050415039,
"learning_rate": 1.9728380702783644e-05,
"loss": 0.3578,
"step": 42
},
{
"epoch": 3.142857142857143,
"grad_norm": 1.2470844984054565,
"learning_rate": 1.968982561550621e-05,
"loss": 0.1155,
"step": 44
},
{
"epoch": 3.2857142857142856,
"grad_norm": 5.716163635253906,
"learning_rate": 1.9648755533435517e-05,
"loss": 0.0708,
"step": 46
},
{
"epoch": 3.4285714285714284,
"grad_norm": 6.77197790145874,
"learning_rate": 1.9605181116313725e-05,
"loss": 0.081,
"step": 48
},
{
"epoch": 3.571428571428571,
"grad_norm": 5.336516380310059,
"learning_rate": 1.955911367388329e-05,
"loss": 0.112,
"step": 50
},
{
"epoch": 3.7142857142857144,
"grad_norm": 10.314866065979004,
"learning_rate": 1.9510565162951538e-05,
"loss": 0.0958,
"step": 52
},
{
"epoch": 3.857142857142857,
"grad_norm": 5.841635704040527,
"learning_rate": 1.9459548184287254e-05,
"loss": 0.0593,
"step": 54
},
{
"epoch": 4.0,
"grad_norm": 10.58973503112793,
"learning_rate": 1.9406075979350175e-05,
"loss": 0.2013,
"step": 56
},
{
"epoch": 4.142857142857143,
"grad_norm": 3.521483898162842,
"learning_rate": 1.9350162426854152e-05,
"loss": 0.0563,
"step": 58
},
{
"epoch": 4.285714285714286,
"grad_norm": 1.0017471313476562,
"learning_rate": 1.9291822039164934e-05,
"loss": 0.1029,
"step": 60
},
{
"epoch": 4.428571428571429,
"grad_norm": 0.33161500096321106,
"learning_rate": 1.923106995853349e-05,
"loss": 0.024,
"step": 62
},
{
"epoch": 4.571428571428571,
"grad_norm": 6.284483432769775,
"learning_rate": 1.9167921953165827e-05,
"loss": 0.1092,
"step": 64
},
{
"epoch": 4.714285714285714,
"grad_norm": 3.3121423721313477,
"learning_rate": 1.9102394413130348e-05,
"loss": 0.0401,
"step": 66
},
{
"epoch": 4.857142857142857,
"grad_norm": 8.295348167419434,
"learning_rate": 1.9034504346103825e-05,
"loss": 0.3237,
"step": 68
},
{
"epoch": 5.0,
"grad_norm": 2.1199071407318115,
"learning_rate": 1.896426937295704e-05,
"loss": 0.0375,
"step": 70
},
{
"epoch": 5.142857142857143,
"grad_norm": 0.45601171255111694,
"learning_rate": 1.8891707723181294e-05,
"loss": 0.0319,
"step": 72
},
{
"epoch": 5.285714285714286,
"grad_norm": 1.0054666996002197,
"learning_rate": 1.881683823015694e-05,
"loss": 0.0097,
"step": 74
},
{
"epoch": 5.428571428571429,
"grad_norm": 0.8899504542350769,
"learning_rate": 1.873968032626518e-05,
"loss": 0.0769,
"step": 76
},
{
"epoch": 5.571428571428571,
"grad_norm": 0.49617600440979004,
"learning_rate": 1.866025403784439e-05,
"loss": 0.0479,
"step": 78
},
{
"epoch": 5.714285714285714,
"grad_norm": 0.16562028229236603,
"learning_rate": 1.8578579979992266e-05,
"loss": 0.0051,
"step": 80
},
{
"epoch": 5.857142857142857,
"grad_norm": 11.048959732055664,
"learning_rate": 1.8494679351215212e-05,
"loss": 0.0984,
"step": 82
},
{
"epoch": 6.0,
"grad_norm": 7.846388816833496,
"learning_rate": 1.8408573927926225e-05,
"loss": 0.116,
"step": 84
},
{
"epoch": 6.142857142857143,
"grad_norm": 0.11113756895065308,
"learning_rate": 1.8320286058792845e-05,
"loss": 0.0027,
"step": 86
},
{
"epoch": 6.285714285714286,
"grad_norm": 3.4302117824554443,
"learning_rate": 1.8229838658936566e-05,
"loss": 0.0197,
"step": 88
},
{
"epoch": 6.428571428571429,
"grad_norm": 4.105597496032715,
"learning_rate": 1.81372552039852e-05,
"loss": 0.1183,
"step": 90
},
{
"epoch": 6.571428571428571,
"grad_norm": 3.2223429679870605,
"learning_rate": 1.804255972397977e-05,
"loss": 0.006,
"step": 92
},
{
"epoch": 6.714285714285714,
"grad_norm": 4.856196403503418,
"learning_rate": 1.7945776797137544e-05,
"loss": 0.0062,
"step": 94
},
{
"epoch": 6.857142857142857,
"grad_norm": 1.22345769405365,
"learning_rate": 1.7846931543472722e-05,
"loss": 0.0037,
"step": 96
},
{
"epoch": 7.0,
"grad_norm": 8.841476440429688,
"learning_rate": 1.7746049618276545e-05,
"loss": 0.0885,
"step": 98
},
{
"epoch": 7.142857142857143,
"grad_norm": 1.7171941995620728,
"learning_rate": 1.7643157205458483e-05,
"loss": 0.0732,
"step": 100
},
{
"epoch": 7.285714285714286,
"grad_norm": 0.2902078628540039,
"learning_rate": 1.753828101075017e-05,
"loss": 0.0019,
"step": 102
},
{
"epoch": 7.428571428571429,
"grad_norm": 6.003970623016357,
"learning_rate": 1.7431448254773943e-05,
"loss": 0.0219,
"step": 104
},
{
"epoch": 7.571428571428571,
"grad_norm": 0.211229145526886,
"learning_rate": 1.7322686665977738e-05,
"loss": 0.0014,
"step": 106
},
{
"epoch": 7.714285714285714,
"grad_norm": 14.436793327331543,
"learning_rate": 1.7212024473438145e-05,
"loss": 0.1415,
"step": 108
},
{
"epoch": 7.857142857142857,
"grad_norm": 0.05790628120303154,
"learning_rate": 1.7099490399533583e-05,
"loss": 0.0008,
"step": 110
},
{
"epoch": 8.0,
"grad_norm": 0.17556001245975494,
"learning_rate": 1.6985113652489374e-05,
"loss": 0.0044,
"step": 112
},
{
"epoch": 8.142857142857142,
"grad_norm": 0.38069918751716614,
"learning_rate": 1.6868923918796753e-05,
"loss": 0.001,
"step": 114
},
{
"epoch": 8.285714285714286,
"grad_norm": 0.093246690928936,
"learning_rate": 1.6750951355507763e-05,
"loss": 0.0321,
"step": 116
},
{
"epoch": 8.428571428571429,
"grad_norm": 0.15035825967788696,
"learning_rate": 1.6631226582407954e-05,
"loss": 0.0036,
"step": 118
},
{
"epoch": 8.571428571428571,
"grad_norm": 0.06252682954072952,
"learning_rate": 1.650978067406904e-05,
"loss": 0.0035,
"step": 120
},
{
"epoch": 8.714285714285714,
"grad_norm": 6.229085922241211,
"learning_rate": 1.638664515178348e-05,
"loss": 0.1034,
"step": 122
},
{
"epoch": 8.857142857142858,
"grad_norm": 2.403806686401367,
"learning_rate": 1.626185197538314e-05,
"loss": 0.036,
"step": 124
},
{
"epoch": 9.0,
"grad_norm": 0.07740367949008942,
"learning_rate": 1.613543353494409e-05,
"loss": 0.0006,
"step": 126
},
{
"epoch": 9.142857142857142,
"grad_norm": 0.322050541639328,
"learning_rate": 1.600742264237979e-05,
"loss": 0.0013,
"step": 128
},
{
"epoch": 9.285714285714286,
"grad_norm": 0.1652839183807373,
"learning_rate": 1.5877852522924733e-05,
"loss": 0.0008,
"step": 130
},
{
"epoch": 9.428571428571429,
"grad_norm": 0.048877034336328506,
"learning_rate": 1.574675680651084e-05,
"loss": 0.0008,
"step": 132
},
{
"epoch": 9.571428571428571,
"grad_norm": 0.06231057271361351,
"learning_rate": 1.561416951903881e-05,
"loss": 0.0022,
"step": 134
},
{
"epoch": 9.714285714285714,
"grad_norm": 0.03611210733652115,
"learning_rate": 1.5480125073546705e-05,
"loss": 0.0389,
"step": 136
},
{
"epoch": 9.857142857142858,
"grad_norm": 0.52008056640625,
"learning_rate": 1.5344658261278013e-05,
"loss": 0.0089,
"step": 138
},
{
"epoch": 10.0,
"grad_norm": 0.42384108901023865,
"learning_rate": 1.5207804242651625e-05,
"loss": 0.0015,
"step": 140
},
{
"epoch": 10.142857142857142,
"grad_norm": 0.055933400988578796,
"learning_rate": 1.5069598538135905e-05,
"loss": 0.0004,
"step": 142
},
{
"epoch": 10.285714285714286,
"grad_norm": 0.0212051160633564,
"learning_rate": 1.4930077019029376e-05,
"loss": 0.0072,
"step": 144
},
{
"epoch": 10.428571428571429,
"grad_norm": 0.03581853583455086,
"learning_rate": 1.4789275898150309e-05,
"loss": 0.0006,
"step": 146
},
{
"epoch": 10.571428571428571,
"grad_norm": 0.021400820463895798,
"learning_rate": 1.4647231720437687e-05,
"loss": 0.0005,
"step": 148
},
{
"epoch": 10.714285714285714,
"grad_norm": 1.3550739288330078,
"learning_rate": 1.450398135346597e-05,
"loss": 0.0332,
"step": 150
},
{
"epoch": 10.857142857142858,
"grad_norm": 0.014044591225683689,
"learning_rate": 1.4359561977876102e-05,
"loss": 0.0003,
"step": 152
},
{
"epoch": 11.0,
"grad_norm": 0.056435056030750275,
"learning_rate": 1.4214011077725293e-05,
"loss": 0.0004,
"step": 154
},
{
"epoch": 11.142857142857142,
"grad_norm": 0.012104465626180172,
"learning_rate": 1.4067366430758004e-05,
"loss": 0.0214,
"step": 156
},
{
"epoch": 11.285714285714286,
"grad_norm": 0.022569801658391953,
"learning_rate": 1.3919666098600753e-05,
"loss": 0.0003,
"step": 158
},
{
"epoch": 11.428571428571429,
"grad_norm": 0.019581535831093788,
"learning_rate": 1.3770948416883205e-05,
"loss": 0.0003,
"step": 160
},
{
"epoch": 11.571428571428571,
"grad_norm": 0.012900936417281628,
"learning_rate": 1.362125198528817e-05,
"loss": 0.0003,
"step": 162
},
{
"epoch": 11.714285714285714,
"grad_norm": 0.283460795879364,
"learning_rate": 1.347061565753303e-05,
"loss": 0.0151,
"step": 164
},
{
"epoch": 11.857142857142858,
"grad_norm": 0.02773939073085785,
"learning_rate": 1.3319078531285286e-05,
"loss": 0.0002,
"step": 166
},
{
"epoch": 12.0,
"grad_norm": 0.015381716191768646,
"learning_rate": 1.3166679938014728e-05,
"loss": 0.0002,
"step": 168
},
{
"epoch": 12.142857142857142,
"grad_norm": 0.013684879057109356,
"learning_rate": 1.301345943278496e-05,
"loss": 0.0002,
"step": 170
},
{
"epoch": 12.285714285714286,
"grad_norm": 0.011124438606202602,
"learning_rate": 1.2859456783986892e-05,
"loss": 0.0002,
"step": 172
},
{
"epoch": 12.428571428571429,
"grad_norm": 0.2694200873374939,
"learning_rate": 1.270471196301684e-05,
"loss": 0.0059,
"step": 174
},
{
"epoch": 12.571428571428571,
"grad_norm": 0.25839900970458984,
"learning_rate": 1.2549265133901934e-05,
"loss": 0.007,
"step": 176
},
{
"epoch": 12.714285714285714,
"grad_norm": 0.28028616309165955,
"learning_rate": 1.2393156642875579e-05,
"loss": 0.0081,
"step": 178
},
{
"epoch": 12.857142857142858,
"grad_norm": 0.007691177539527416,
"learning_rate": 1.2236427007905558e-05,
"loss": 0.0004,
"step": 180
},
{
"epoch": 13.0,
"grad_norm": 0.03421185538172722,
"learning_rate": 1.2079116908177592e-05,
"loss": 0.0002,
"step": 182
},
{
"epoch": 13.142857142857142,
"grad_norm": 0.21886347234249115,
"learning_rate": 1.1921267173537085e-05,
"loss": 0.0106,
"step": 184
},
{
"epoch": 13.285714285714286,
"grad_norm": 0.12066204100847244,
"learning_rate": 1.1762918773891691e-05,
"loss": 0.0039,
"step": 186
},
{
"epoch": 13.428571428571429,
"grad_norm": 0.009790458716452122,
"learning_rate": 1.1604112808577603e-05,
"loss": 0.0001,
"step": 188
},
{
"epoch": 13.571428571428571,
"grad_norm": 0.006134828552603722,
"learning_rate": 1.1444890495692214e-05,
"loss": 0.0001,
"step": 190
},
{
"epoch": 13.714285714285714,
"grad_norm": 0.007279029581695795,
"learning_rate": 1.1285293161395948e-05,
"loss": 0.0082,
"step": 192
},
{
"epoch": 13.857142857142858,
"grad_norm": 0.007834681309759617,
"learning_rate": 1.1125362229186056e-05,
"loss": 0.0001,
"step": 194
},
{
"epoch": 14.0,
"grad_norm": 0.011400113813579082,
"learning_rate": 1.0965139209145153e-05,
"loss": 0.0001,
"step": 196
},
{
"epoch": 14.142857142857142,
"grad_norm": 0.005645982455462217,
"learning_rate": 1.0804665687167262e-05,
"loss": 0.0001,
"step": 198
},
{
"epoch": 14.285714285714286,
"grad_norm": 0.007113787345588207,
"learning_rate": 1.0643983314164195e-05,
"loss": 0.0046,
"step": 200
},
{
"epoch": 14.428571428571429,
"grad_norm": 0.00671082129701972,
"learning_rate": 1.0483133795255072e-05,
"loss": 0.0001,
"step": 202
},
{
"epoch": 14.571428571428571,
"grad_norm": 0.10411105304956436,
"learning_rate": 1.0322158878941733e-05,
"loss": 0.0092,
"step": 204
},
{
"epoch": 14.714285714285714,
"grad_norm": 0.007047938648611307,
"learning_rate": 1.0161100346272913e-05,
"loss": 0.0046,
"step": 206
},
{
"epoch": 14.857142857142858,
"grad_norm": 0.006609996780753136,
"learning_rate": 1e-05,
"loss": 0.0001,
"step": 208
},
{
"epoch": 15.0,
"grad_norm": 0.00515047088265419,
"learning_rate": 9.838899653727088e-06,
"loss": 0.0001,
"step": 210
},
{
"epoch": 15.142857142857142,
"grad_norm": 0.005954743828624487,
"learning_rate": 9.677841121058274e-06,
"loss": 0.0001,
"step": 212
},
{
"epoch": 15.285714285714286,
"grad_norm": 0.19431306421756744,
"learning_rate": 9.516866204744932e-06,
"loss": 0.0046,
"step": 214
},
{
"epoch": 15.428571428571429,
"grad_norm": 0.007638772018253803,
"learning_rate": 9.356016685835807e-06,
"loss": 0.0045,
"step": 216
},
{
"epoch": 15.571428571428571,
"grad_norm": 0.10914219170808792,
"learning_rate": 9.195334312832742e-06,
"loss": 0.0092,
"step": 218
},
{
"epoch": 15.714285714285714,
"grad_norm": 0.005956027191132307,
"learning_rate": 9.034860790854848e-06,
"loss": 0.0001,
"step": 220
},
{
"epoch": 15.857142857142858,
"grad_norm": 0.006127700675278902,
"learning_rate": 8.874637770813947e-06,
"loss": 0.0001,
"step": 222
},
{
"epoch": 16.0,
"grad_norm": 0.006942708045244217,
"learning_rate": 8.714706838604056e-06,
"loss": 0.0001,
"step": 224
},
{
"epoch": 16.142857142857142,
"grad_norm": 0.17752425372600555,
"learning_rate": 8.55510950430779e-06,
"loss": 0.0098,
"step": 226
},
{
"epoch": 16.285714285714285,
"grad_norm": 0.004300012718886137,
"learning_rate": 8.395887191422397e-06,
"loss": 0.0001,
"step": 228
},
{
"epoch": 16.428571428571427,
"grad_norm": 0.0063268570229411125,
"learning_rate": 8.23708122610831e-06,
"loss": 0.0075,
"step": 230
},
{
"epoch": 16.571428571428573,
"grad_norm": 0.005733097437769175,
"learning_rate": 8.078732826462917e-06,
"loss": 0.0001,
"step": 232
},
{
"epoch": 16.714285714285715,
"grad_norm": 0.0077501749619841576,
"learning_rate": 7.92088309182241e-06,
"loss": 0.0001,
"step": 234
},
{
"epoch": 16.857142857142858,
"grad_norm": 0.005072280764579773,
"learning_rate": 7.763572992094448e-06,
"loss": 0.0001,
"step": 236
},
{
"epoch": 17.0,
"grad_norm": 0.007300873752683401,
"learning_rate": 7.606843357124426e-06,
"loss": 0.0001,
"step": 238
},
{
"epoch": 17.142857142857142,
"grad_norm": 0.10978183150291443,
"learning_rate": 7.450734866098066e-06,
"loss": 0.0036,
"step": 240
},
{
"epoch": 17.285714285714285,
"grad_norm": 0.005480973981320858,
"learning_rate": 7.2952880369831635e-06,
"loss": 0.0061,
"step": 242
},
{
"epoch": 17.428571428571427,
"grad_norm": 0.2027699053287506,
"learning_rate": 7.140543216013109e-06,
"loss": 0.0035,
"step": 244
},
{
"epoch": 17.571428571428573,
"grad_norm": 0.0048587811179459095,
"learning_rate": 6.986540567215043e-06,
"loss": 0.0001,
"step": 246
},
{
"epoch": 17.714285714285715,
"grad_norm": 0.005718125496059656,
"learning_rate": 6.833320061985278e-06,
"loss": 0.0001,
"step": 248
},
{
"epoch": 17.857142857142858,
"grad_norm": 0.0035847134422510862,
"learning_rate": 6.680921468714718e-06,
"loss": 0.0052,
"step": 250
},
{
"epoch": 18.0,
"grad_norm": 0.0035746165085583925,
"learning_rate": 6.529384342466971e-06,
"loss": 0.0001,
"step": 252
},
{
"epoch": 18.142857142857142,
"grad_norm": 0.00451216846704483,
"learning_rate": 6.378748014711834e-06,
"loss": 0.0001,
"step": 254
},
{
"epoch": 18.285714285714285,
"grad_norm": 0.004890494979918003,
"learning_rate": 6.229051583116796e-06,
"loss": 0.0001,
"step": 256
},
{
"epoch": 18.428571428571427,
"grad_norm": 0.004625991452485323,
"learning_rate": 6.080333901399252e-06,
"loss": 0.0058,
"step": 258
},
{
"epoch": 18.571428571428573,
"grad_norm": 0.005847033113241196,
"learning_rate": 5.932633569242e-06,
"loss": 0.0001,
"step": 260
},
{
"epoch": 18.714285714285715,
"grad_norm": 0.006245823111385107,
"learning_rate": 5.785988922274711e-06,
"loss": 0.0072,
"step": 262
},
{
"epoch": 18.857142857142858,
"grad_norm": 0.08823557943105698,
"learning_rate": 5.640438022123898e-06,
"loss": 0.0034,
"step": 264
},
{
"epoch": 19.0,
"grad_norm": 0.004038456827402115,
"learning_rate": 5.496018646534032e-06,
"loss": 0.0001,
"step": 266
},
{
"epoch": 19.142857142857142,
"grad_norm": 0.005825538653880358,
"learning_rate": 5.352768279562315e-06,
"loss": 0.0056,
"step": 268
},
{
"epoch": 19.285714285714285,
"grad_norm": 0.004666970111429691,
"learning_rate": 5.210724101849696e-06,
"loss": 0.0001,
"step": 270
},
{
"epoch": 19.428571428571427,
"grad_norm": 0.004332971293479204,
"learning_rate": 5.069922980970626e-06,
"loss": 0.0001,
"step": 272
},
{
"epoch": 19.571428571428573,
"grad_norm": 0.09540804475545883,
"learning_rate": 4.930401461864099e-06,
"loss": 0.0085,
"step": 274
},
{
"epoch": 19.714285714285715,
"grad_norm": 0.004319526720792055,
"learning_rate": 4.7921957573483756e-06,
"loss": 0.0001,
"step": 276
},
{
"epoch": 19.857142857142858,
"grad_norm": 0.00582715030759573,
"learning_rate": 4.655341738721989e-06,
"loss": 0.0001,
"step": 278
},
{
"epoch": 20.0,
"grad_norm": 0.005314113572239876,
"learning_rate": 4.519874926453303e-06,
"loss": 0.0001,
"step": 280
},
{
"epoch": 20.142857142857142,
"grad_norm": 0.006195093505084515,
"learning_rate": 4.385830480961192e-06,
"loss": 0.0001,
"step": 282
},
{
"epoch": 20.285714285714285,
"grad_norm": 0.004151402972638607,
"learning_rate": 4.2532431934891646e-06,
"loss": 0.0001,
"step": 284
},
{
"epoch": 20.428571428571427,
"grad_norm": 0.004392064176499844,
"learning_rate": 4.12214747707527e-06,
"loss": 0.0047,
"step": 286
},
{
"epoch": 20.571428571428573,
"grad_norm": 0.004756106063723564,
"learning_rate": 3.99257735762021e-06,
"loss": 0.0001,
"step": 288
},
{
"epoch": 20.714285714285715,
"grad_norm": 0.004611567594110966,
"learning_rate": 3.864566465055913e-06,
"loss": 0.0022,
"step": 290
},
{
"epoch": 20.857142857142858,
"grad_norm": 0.004500643815845251,
"learning_rate": 3.738148024616863e-06,
"loss": 0.0001,
"step": 292
},
{
"epoch": 21.0,
"grad_norm": 0.0058865793980658054,
"learning_rate": 3.6133548482165225e-06,
"loss": 0.0081,
"step": 294
},
{
"epoch": 21.142857142857142,
"grad_norm": 0.096845842897892,
"learning_rate": 3.4902193259309627e-06,
"loss": 0.0037,
"step": 296
},
{
"epoch": 21.285714285714285,
"grad_norm": 0.0038972089532762766,
"learning_rate": 3.3687734175920505e-06,
"loss": 0.0001,
"step": 298
},
{
"epoch": 21.428571428571427,
"grad_norm": 0.004685294348746538,
"learning_rate": 3.2490486444922396e-06,
"loss": 0.0001,
"step": 300
},
{
"epoch": 21.571428571428573,
"grad_norm": 0.20793084800243378,
"learning_rate": 3.131076081203247e-06,
"loss": 0.0019,
"step": 302
},
{
"epoch": 21.714285714285715,
"grad_norm": 0.0035532894544303417,
"learning_rate": 3.0148863475106315e-06,
"loss": 0.0058,
"step": 304
},
{
"epoch": 21.857142857142858,
"grad_norm": 0.2300863116979599,
"learning_rate": 2.900509600466418e-06,
"loss": 0.0025,
"step": 306
},
{
"epoch": 22.0,
"grad_norm": 0.007288255728781223,
"learning_rate": 2.7879755265618558e-06,
"loss": 0.0001,
"step": 308
},
{
"epoch": 22.142857142857142,
"grad_norm": 0.005309649743139744,
"learning_rate": 2.6773133340222677e-06,
"loss": 0.0016,
"step": 310
},
{
"epoch": 22.285714285714285,
"grad_norm": 0.003941751550883055,
"learning_rate": 2.5685517452260566e-06,
"loss": 0.0001,
"step": 312
},
{
"epoch": 22.428571428571427,
"grad_norm": 0.004016744904220104,
"learning_rate": 2.4617189892498326e-06,
"loss": 0.0001,
"step": 314
},
{
"epoch": 22.571428571428573,
"grad_norm": 0.13147950172424316,
"learning_rate": 2.3568427945415163e-06,
"loss": 0.0064,
"step": 316
},
{
"epoch": 22.714285714285715,
"grad_norm": 0.003918132279068232,
"learning_rate": 2.2539503817234553e-06,
"loss": 0.0001,
"step": 318
},
{
"epoch": 22.857142857142858,
"grad_norm": 0.004725574050098658,
"learning_rate": 2.153068456527283e-06,
"loss": 0.0001,
"step": 320
},
{
"epoch": 23.0,
"grad_norm": 0.3711271584033966,
"learning_rate": 2.0542232028624585e-06,
"loss": 0.0175,
"step": 322
},
{
"epoch": 23.142857142857142,
"grad_norm": 0.18150761723518372,
"learning_rate": 1.9574402760202315e-06,
"loss": 0.0023,
"step": 324
},
{
"epoch": 23.285714285714285,
"grad_norm": 0.13696324825286865,
"learning_rate": 1.8627447960148036e-06,
"loss": 0.0045,
"step": 326
},
{
"epoch": 23.428571428571427,
"grad_norm": 0.004304251167923212,
"learning_rate": 1.7701613410634367e-06,
"loss": 0.0001,
"step": 328
},
{
"epoch": 23.571428571428573,
"grad_norm": 0.005392234306782484,
"learning_rate": 1.6797139412071583e-06,
"loss": 0.0001,
"step": 330
},
{
"epoch": 23.714285714285715,
"grad_norm": 0.004501474089920521,
"learning_rate": 1.5914260720737796e-06,
"loss": 0.0001,
"step": 332
},
{
"epoch": 23.857142857142858,
"grad_norm": 0.0033914605155587196,
"learning_rate": 1.5053206487847916e-06,
"loss": 0.0047,
"step": 334
},
{
"epoch": 24.0,
"grad_norm": 0.005722349975258112,
"learning_rate": 1.4214200200077343e-06,
"loss": 0.0001,
"step": 336
},
{
"epoch": 24.142857142857142,
"grad_norm": 0.0028757653199136257,
"learning_rate": 1.339745962155613e-06,
"loss": 0.0001,
"step": 338
},
{
"epoch": 24.285714285714285,
"grad_norm": 0.004367240238934755,
"learning_rate": 1.2603196737348211e-06,
"loss": 0.0012,
"step": 340
},
{
"epoch": 24.428571428571427,
"grad_norm": 0.004103364888578653,
"learning_rate": 1.1831617698430609e-06,
"loss": 0.0001,
"step": 342
},
{
"epoch": 24.571428571428573,
"grad_norm": 0.12380176037549973,
"learning_rate": 1.1082922768187098e-06,
"loss": 0.0053,
"step": 344
},
{
"epoch": 24.714285714285715,
"grad_norm": 0.11978839337825775,
"learning_rate": 1.0357306270429623e-06,
"loss": 0.0045,
"step": 346
},
{
"epoch": 24.857142857142858,
"grad_norm": 0.15920782089233398,
"learning_rate": 9.65495653896179e-07,
"loss": 0.0008,
"step": 348
},
{
"epoch": 25.0,
"grad_norm": 0.005539704114198685,
"learning_rate": 8.976055868696543e-07,
"loss": 0.0001,
"step": 350
},
{
"epoch": 25.142857142857142,
"grad_norm": 0.004604020155966282,
"learning_rate": 8.320780468341761e-07,
"loss": 0.0001,
"step": 352
},
{
"epoch": 25.285714285714285,
"grad_norm": 0.19128715991973877,
"learning_rate": 7.689300414665124e-07,
"loss": 0.0012,
"step": 354
},
{
"epoch": 25.428571428571427,
"grad_norm": 0.0033012081403285265,
"learning_rate": 7.08177960835068e-07,
"loss": 0.0067,
"step": 356
},
{
"epoch": 25.571428571428573,
"grad_norm": 0.09938196837902069,
"learning_rate": 6.498375731458529e-07,
"loss": 0.0005,
"step": 358
},
{
"epoch": 25.714285714285715,
"grad_norm": 0.0049908701330423355,
"learning_rate": 5.939240206498287e-07,
"loss": 0.0001,
"step": 360
},
{
"epoch": 25.857142857142858,
"grad_norm": 0.0035182549618184566,
"learning_rate": 5.404518157127481e-07,
"loss": 0.0046,
"step": 362
},
{
"epoch": 26.0,
"grad_norm": 0.0068074483424425125,
"learning_rate": 4.894348370484648e-07,
"loss": 0.0001,
"step": 364
},
{
"epoch": 26.142857142857142,
"grad_norm": 0.004269490484148264,
"learning_rate": 4.408863261167096e-07,
"loss": 0.0001,
"step": 366
},
{
"epoch": 26.285714285714285,
"grad_norm": 0.12823165953159332,
"learning_rate": 3.9481888368627764e-07,
"loss": 0.0042,
"step": 368
},
{
"epoch": 26.428571428571427,
"grad_norm": 0.0035812160931527615,
"learning_rate": 3.5124446656448654e-07,
"loss": 0.0001,
"step": 370
},
{
"epoch": 26.571428571428573,
"grad_norm": 0.00431048171594739,
"learning_rate": 3.101743844937943e-07,
"loss": 0.0059,
"step": 372
},
{
"epoch": 26.714285714285715,
"grad_norm": 0.004997176583856344,
"learning_rate": 2.716192972163556e-07,
"loss": 0.0001,
"step": 374
},
{
"epoch": 26.857142857142858,
"grad_norm": 0.004221776034682989,
"learning_rate": 2.355892117072789e-07,
"loss": 0.0001,
"step": 376
},
{
"epoch": 27.0,
"grad_norm": 0.007233808748424053,
"learning_rate": 2.0209347957732328e-07,
"loss": 0.0008,
"step": 378
},
{
"epoch": 27.142857142857142,
"grad_norm": 0.0030983760952949524,
"learning_rate": 1.711407946456789e-07,
"loss": 0.0001,
"step": 380
},
{
"epoch": 27.285714285714285,
"grad_norm": 0.13253678381443024,
"learning_rate": 1.4273919068349184e-07,
"loss": 0.0014,
"step": 382
},
{
"epoch": 27.428571428571427,
"grad_norm": 0.003873329609632492,
"learning_rate": 1.1689603932869664e-07,
"loss": 0.0001,
"step": 384
},
{
"epoch": 27.571428571428573,
"grad_norm": 0.005466313101351261,
"learning_rate": 9.36180481727067e-08,
"loss": 0.0046,
"step": 386
},
{
"epoch": 27.714285714285715,
"grad_norm": 0.004763293080031872,
"learning_rate": 7.291125901946027e-08,
"loss": 0.0001,
"step": 388
},
{
"epoch": 27.857142857142858,
"grad_norm": 0.11048487573862076,
"learning_rate": 5.4781046317267103e-08,
"loss": 0.0048,
"step": 390
},
{
"epoch": 28.0,
"grad_norm": 0.007729329634457827,
"learning_rate": 3.923211576387087e-08,
"loss": 0.0001,
"step": 392
},
{
"epoch": 28.142857142857142,
"grad_norm": 0.003292845096439123,
"learning_rate": 2.6268503085089547e-08,
"loss": 0.0001,
"step": 394
},
{
"epoch": 28.285714285714285,
"grad_norm": 0.00781972985714674,
"learning_rate": 1.5893572987333293e-08,
"loss": 0.0001,
"step": 396
},
{
"epoch": 28.428571428571427,
"grad_norm": 0.14954733848571777,
"learning_rate": 8.110018284304132e-09,
"loss": 0.0049,
"step": 398
},
{
"epoch": 28.571428571428573,
"grad_norm": 0.005883893929421902,
"learning_rate": 2.9198591980705847e-09,
"loss": 0.0007,
"step": 400
},
{
"epoch": 28.714285714285715,
"grad_norm": 0.13555946946144104,
"learning_rate": 3.244428347204398e-10,
"loss": 0.0008,
"step": 402
}
],
"logging_steps": 2,
"max_steps": 402,
"num_input_tokens_seen": 0,
"num_train_epochs": 29,
"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": 2.0685712456679424e+16,
"train_batch_size": 16,
"trial_name": null,
"trial_params": null
}