test8 / checkpoint-402 /trainer_state.json
frjonah's picture
Upload folder using huggingface_hub
9443ab4 verified
Raw
History Blame Contribute Delete
35.2 kB
{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 28.74074074074074,
"eval_steps": 500,
"global_step": 402,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.14814814814814814,
"grad_norm": 70.21087646484375,
"learning_rate": 1.6666666666666667e-06,
"loss": 6.0132,
"step": 2
},
{
"epoch": 0.2962962962962963,
"grad_norm": 57.58814239501953,
"learning_rate": 5e-06,
"loss": 5.2072,
"step": 4
},
{
"epoch": 0.4444444444444444,
"grad_norm": 46.47439193725586,
"learning_rate": 8.333333333333334e-06,
"loss": 3.0408,
"step": 6
},
{
"epoch": 0.5925925925925926,
"grad_norm": 24.632150650024414,
"learning_rate": 1.1666666666666668e-05,
"loss": 2.3364,
"step": 8
},
{
"epoch": 0.7407407407407407,
"grad_norm": 23.976943969726562,
"learning_rate": 1.5000000000000002e-05,
"loss": 2.185,
"step": 10
},
{
"epoch": 0.8888888888888888,
"grad_norm": 18.078527450561523,
"learning_rate": 1.8333333333333333e-05,
"loss": 1.9432,
"step": 12
},
{
"epoch": 1.0,
"grad_norm": 25.87737464904785,
"learning_rate": 1.9999675557165282e-05,
"loss": 1.5516,
"step": 14
},
{
"epoch": 1.1481481481481481,
"grad_norm": 14.125572204589844,
"learning_rate": 1.9997080140801932e-05,
"loss": 0.8972,
"step": 16
},
{
"epoch": 1.2962962962962963,
"grad_norm": 12.101792335510254,
"learning_rate": 1.9991889981715696e-05,
"loss": 0.8666,
"step": 18
},
{
"epoch": 1.4444444444444444,
"grad_norm": 21.769561767578125,
"learning_rate": 1.9984106427012667e-05,
"loss": 0.9736,
"step": 20
},
{
"epoch": 1.5925925925925926,
"grad_norm": 14.465237617492676,
"learning_rate": 1.9973731496914914e-05,
"loss": 0.8706,
"step": 22
},
{
"epoch": 1.7407407407407407,
"grad_norm": 35.41783142089844,
"learning_rate": 1.9960767884236132e-05,
"loss": 0.7846,
"step": 24
},
{
"epoch": 1.8888888888888888,
"grad_norm": 20.135147094726562,
"learning_rate": 1.9945218953682736e-05,
"loss": 1.1249,
"step": 26
},
{
"epoch": 2.0,
"grad_norm": 20.01813507080078,
"learning_rate": 1.992708874098054e-05,
"loss": 0.8804,
"step": 28
},
{
"epoch": 2.148148148148148,
"grad_norm": 5.577389240264893,
"learning_rate": 1.9906381951827295e-05,
"loss": 0.2036,
"step": 30
},
{
"epoch": 2.2962962962962963,
"grad_norm": 7.592391014099121,
"learning_rate": 1.9883103960671305e-05,
"loss": 0.165,
"step": 32
},
{
"epoch": 2.4444444444444446,
"grad_norm": 9.455511093139648,
"learning_rate": 1.985726080931651e-05,
"loss": 0.1009,
"step": 34
},
{
"epoch": 2.5925925925925926,
"grad_norm": 11.158129692077637,
"learning_rate": 1.9828859205354326e-05,
"loss": 0.1706,
"step": 36
},
{
"epoch": 2.7407407407407405,
"grad_norm": 9.241584777832031,
"learning_rate": 1.979790652042268e-05,
"loss": 0.2829,
"step": 38
},
{
"epoch": 2.888888888888889,
"grad_norm": 9.160713195800781,
"learning_rate": 1.9764410788292724e-05,
"loss": 0.276,
"step": 40
},
{
"epoch": 3.0,
"grad_norm": 9.034798622131348,
"learning_rate": 1.9728380702783644e-05,
"loss": 0.2954,
"step": 42
},
{
"epoch": 3.148148148148148,
"grad_norm": 2.3645989894866943,
"learning_rate": 1.968982561550621e-05,
"loss": 0.0698,
"step": 44
},
{
"epoch": 3.2962962962962963,
"grad_norm": 6.123332500457764,
"learning_rate": 1.9648755533435517e-05,
"loss": 0.0961,
"step": 46
},
{
"epoch": 3.4444444444444446,
"grad_norm": 1.8872700929641724,
"learning_rate": 1.9605181116313725e-05,
"loss": 0.0648,
"step": 48
},
{
"epoch": 3.5925925925925926,
"grad_norm": 6.479188919067383,
"learning_rate": 1.955911367388329e-05,
"loss": 0.1653,
"step": 50
},
{
"epoch": 3.7407407407407405,
"grad_norm": 2.771259069442749,
"learning_rate": 1.9510565162951538e-05,
"loss": 0.057,
"step": 52
},
{
"epoch": 3.888888888888889,
"grad_norm": 10.078364372253418,
"learning_rate": 1.9459548184287254e-05,
"loss": 0.1347,
"step": 54
},
{
"epoch": 4.0,
"grad_norm": 16.32892608642578,
"learning_rate": 1.9406075979350175e-05,
"loss": 0.1828,
"step": 56
},
{
"epoch": 4.148148148148148,
"grad_norm": 0.7437280416488647,
"learning_rate": 1.9350162426854152e-05,
"loss": 0.0483,
"step": 58
},
{
"epoch": 4.296296296296296,
"grad_norm": 4.6985321044921875,
"learning_rate": 1.9291822039164934e-05,
"loss": 0.0657,
"step": 60
},
{
"epoch": 4.444444444444445,
"grad_norm": 3.9528374671936035,
"learning_rate": 1.923106995853349e-05,
"loss": 0.1279,
"step": 62
},
{
"epoch": 4.592592592592593,
"grad_norm": 4.089845180511475,
"learning_rate": 1.9167921953165827e-05,
"loss": 0.0813,
"step": 64
},
{
"epoch": 4.7407407407407405,
"grad_norm": 2.3875203132629395,
"learning_rate": 1.9102394413130348e-05,
"loss": 0.0367,
"step": 66
},
{
"epoch": 4.888888888888889,
"grad_norm": 5.7833967208862305,
"learning_rate": 1.9034504346103825e-05,
"loss": 0.2109,
"step": 68
},
{
"epoch": 5.0,
"grad_norm": 0.2964959740638733,
"learning_rate": 1.896426937295704e-05,
"loss": 0.0383,
"step": 70
},
{
"epoch": 5.148148148148148,
"grad_norm": 4.408813953399658,
"learning_rate": 1.8891707723181294e-05,
"loss": 0.1141,
"step": 72
},
{
"epoch": 5.296296296296296,
"grad_norm": 0.22182708978652954,
"learning_rate": 1.881683823015694e-05,
"loss": 0.0672,
"step": 74
},
{
"epoch": 5.444444444444445,
"grad_norm": 1.0347212553024292,
"learning_rate": 1.873968032626518e-05,
"loss": 0.0131,
"step": 76
},
{
"epoch": 5.592592592592593,
"grad_norm": 0.7128021717071533,
"learning_rate": 1.866025403784439e-05,
"loss": 0.0101,
"step": 78
},
{
"epoch": 5.7407407407407405,
"grad_norm": 0.7934299111366272,
"learning_rate": 1.8578579979992266e-05,
"loss": 0.0266,
"step": 80
},
{
"epoch": 5.888888888888889,
"grad_norm": 13.658126831054688,
"learning_rate": 1.8494679351215212e-05,
"loss": 0.1843,
"step": 82
},
{
"epoch": 6.0,
"grad_norm": 0.41831719875335693,
"learning_rate": 1.8408573927926225e-05,
"loss": 0.0023,
"step": 84
},
{
"epoch": 6.148148148148148,
"grad_norm": 0.07081316411495209,
"learning_rate": 1.8320286058792845e-05,
"loss": 0.0014,
"step": 86
},
{
"epoch": 6.296296296296296,
"grad_norm": 1.6338634490966797,
"learning_rate": 1.8229838658936566e-05,
"loss": 0.0907,
"step": 88
},
{
"epoch": 6.444444444444445,
"grad_norm": 8.390157699584961,
"learning_rate": 1.81372552039852e-05,
"loss": 0.2235,
"step": 90
},
{
"epoch": 6.592592592592593,
"grad_norm": 4.785590648651123,
"learning_rate": 1.804255972397977e-05,
"loss": 0.1009,
"step": 92
},
{
"epoch": 6.7407407407407405,
"grad_norm": 9.069583892822266,
"learning_rate": 1.7945776797137544e-05,
"loss": 0.0224,
"step": 94
},
{
"epoch": 6.888888888888889,
"grad_norm": 0.145456925034523,
"learning_rate": 1.7846931543472722e-05,
"loss": 0.0017,
"step": 96
},
{
"epoch": 7.0,
"grad_norm": 7.696702003479004,
"learning_rate": 1.7746049618276545e-05,
"loss": 0.165,
"step": 98
},
{
"epoch": 7.148148148148148,
"grad_norm": 0.1802881360054016,
"learning_rate": 1.7643157205458483e-05,
"loss": 0.0285,
"step": 100
},
{
"epoch": 7.296296296296296,
"grad_norm": 5.289300918579102,
"learning_rate": 1.753828101075017e-05,
"loss": 0.0386,
"step": 102
},
{
"epoch": 7.444444444444445,
"grad_norm": 0.5074731111526489,
"learning_rate": 1.7431448254773943e-05,
"loss": 0.0149,
"step": 104
},
{
"epoch": 7.592592592592593,
"grad_norm": 4.221737861633301,
"learning_rate": 1.7322686665977738e-05,
"loss": 0.0159,
"step": 106
},
{
"epoch": 7.7407407407407405,
"grad_norm": 2.5934197902679443,
"learning_rate": 1.7212024473438145e-05,
"loss": 0.0519,
"step": 108
},
{
"epoch": 7.888888888888889,
"grad_norm": 0.21167413890361786,
"learning_rate": 1.7099490399533583e-05,
"loss": 0.0142,
"step": 110
},
{
"epoch": 8.0,
"grad_norm": 0.09941914677619934,
"learning_rate": 1.6985113652489374e-05,
"loss": 0.0019,
"step": 112
},
{
"epoch": 8.148148148148149,
"grad_norm": 0.11601820588111877,
"learning_rate": 1.6868923918796753e-05,
"loss": 0.001,
"step": 114
},
{
"epoch": 8.296296296296296,
"grad_norm": 0.048870399594306946,
"learning_rate": 1.6750951355507763e-05,
"loss": 0.004,
"step": 116
},
{
"epoch": 8.444444444444445,
"grad_norm": 10.136484146118164,
"learning_rate": 1.6631226582407954e-05,
"loss": 0.0512,
"step": 118
},
{
"epoch": 8.592592592592592,
"grad_norm": 4.009435653686523,
"learning_rate": 1.650978067406904e-05,
"loss": 0.0569,
"step": 120
},
{
"epoch": 8.74074074074074,
"grad_norm": 0.04615255072712898,
"learning_rate": 1.638664515178348e-05,
"loss": 0.0276,
"step": 122
},
{
"epoch": 8.88888888888889,
"grad_norm": 0.5869151949882507,
"learning_rate": 1.626185197538314e-05,
"loss": 0.0122,
"step": 124
},
{
"epoch": 9.0,
"grad_norm": 0.372694194316864,
"learning_rate": 1.613543353494409e-05,
"loss": 0.004,
"step": 126
},
{
"epoch": 9.148148148148149,
"grad_norm": 0.034844283014535904,
"learning_rate": 1.600742264237979e-05,
"loss": 0.0027,
"step": 128
},
{
"epoch": 9.296296296296296,
"grad_norm": 0.024813378229737282,
"learning_rate": 1.5877852522924733e-05,
"loss": 0.0005,
"step": 130
},
{
"epoch": 9.444444444444445,
"grad_norm": 0.054847851395606995,
"learning_rate": 1.574675680651084e-05,
"loss": 0.0003,
"step": 132
},
{
"epoch": 9.592592592592592,
"grad_norm": 0.012215496972203255,
"learning_rate": 1.561416951903881e-05,
"loss": 0.0411,
"step": 134
},
{
"epoch": 9.74074074074074,
"grad_norm": 0.01614837907254696,
"learning_rate": 1.5480125073546705e-05,
"loss": 0.0055,
"step": 136
},
{
"epoch": 9.88888888888889,
"grad_norm": 0.8813748359680176,
"learning_rate": 1.5344658261278013e-05,
"loss": 0.0283,
"step": 138
},
{
"epoch": 10.0,
"grad_norm": 57.692848205566406,
"learning_rate": 1.5207804242651625e-05,
"loss": 0.3287,
"step": 140
},
{
"epoch": 10.148148148148149,
"grad_norm": 0.01186081487685442,
"learning_rate": 1.5069598538135905e-05,
"loss": 0.0002,
"step": 142
},
{
"epoch": 10.296296296296296,
"grad_norm": 0.03444704785943031,
"learning_rate": 1.4930077019029376e-05,
"loss": 0.0061,
"step": 144
},
{
"epoch": 10.444444444444445,
"grad_norm": 0.05061617121100426,
"learning_rate": 1.4789275898150309e-05,
"loss": 0.0023,
"step": 146
},
{
"epoch": 10.592592592592592,
"grad_norm": 0.19461821019649506,
"learning_rate": 1.4647231720437687e-05,
"loss": 0.0005,
"step": 148
},
{
"epoch": 10.74074074074074,
"grad_norm": 0.5531904697418213,
"learning_rate": 1.450398135346597e-05,
"loss": 0.0225,
"step": 150
},
{
"epoch": 10.88888888888889,
"grad_norm": 0.02902345545589924,
"learning_rate": 1.4359561977876102e-05,
"loss": 0.0572,
"step": 152
},
{
"epoch": 11.0,
"grad_norm": 0.03750229626893997,
"learning_rate": 1.4214011077725293e-05,
"loss": 0.0238,
"step": 154
},
{
"epoch": 11.148148148148149,
"grad_norm": 0.01594114489853382,
"learning_rate": 1.4067366430758004e-05,
"loss": 0.0125,
"step": 156
},
{
"epoch": 11.296296296296296,
"grad_norm": 0.01831846497952938,
"learning_rate": 1.3919666098600753e-05,
"loss": 0.0002,
"step": 158
},
{
"epoch": 11.444444444444445,
"grad_norm": 0.014910983853042126,
"learning_rate": 1.3770948416883205e-05,
"loss": 0.0002,
"step": 160
},
{
"epoch": 11.592592592592592,
"grad_norm": 2.285388946533203,
"learning_rate": 1.362125198528817e-05,
"loss": 0.0251,
"step": 162
},
{
"epoch": 11.74074074074074,
"grad_norm": 0.04311133176088333,
"learning_rate": 1.347061565753303e-05,
"loss": 0.01,
"step": 164
},
{
"epoch": 11.88888888888889,
"grad_norm": 0.014949589967727661,
"learning_rate": 1.3319078531285286e-05,
"loss": 0.0022,
"step": 166
},
{
"epoch": 12.0,
"grad_norm": 0.011438173241913319,
"learning_rate": 1.3166679938014728e-05,
"loss": 0.0003,
"step": 168
},
{
"epoch": 12.148148148148149,
"grad_norm": 0.016681043431162834,
"learning_rate": 1.301345943278496e-05,
"loss": 0.0002,
"step": 170
},
{
"epoch": 12.296296296296296,
"grad_norm": 0.011501719243824482,
"learning_rate": 1.2859456783986892e-05,
"loss": 0.0004,
"step": 172
},
{
"epoch": 12.444444444444445,
"grad_norm": 0.13780571520328522,
"learning_rate": 1.270471196301684e-05,
"loss": 0.0053,
"step": 174
},
{
"epoch": 12.592592592592592,
"grad_norm": 0.15591566264629364,
"learning_rate": 1.2549265133901934e-05,
"loss": 0.0087,
"step": 176
},
{
"epoch": 12.74074074074074,
"grad_norm": 0.07638789713382721,
"learning_rate": 1.2393156642875579e-05,
"loss": 0.0063,
"step": 178
},
{
"epoch": 12.88888888888889,
"grad_norm": 0.009266780689358711,
"learning_rate": 1.2236427007905558e-05,
"loss": 0.0002,
"step": 180
},
{
"epoch": 13.0,
"grad_norm": 0.020039549097418785,
"learning_rate": 1.2079116908177592e-05,
"loss": 0.2838,
"step": 182
},
{
"epoch": 13.148148148148149,
"grad_norm": 0.15673038363456726,
"learning_rate": 1.1921267173537085e-05,
"loss": 0.1877,
"step": 184
},
{
"epoch": 13.296296296296296,
"grad_norm": 0.08017202466726303,
"learning_rate": 1.1762918773891691e-05,
"loss": 0.0042,
"step": 186
},
{
"epoch": 13.444444444444445,
"grad_norm": 0.010346467606723309,
"learning_rate": 1.1604112808577603e-05,
"loss": 0.0002,
"step": 188
},
{
"epoch": 13.592592592592592,
"grad_norm": 0.010894942097365856,
"learning_rate": 1.1444890495692214e-05,
"loss": 0.0001,
"step": 190
},
{
"epoch": 13.74074074074074,
"grad_norm": 0.031103109940886497,
"learning_rate": 1.1285293161395948e-05,
"loss": 0.0068,
"step": 192
},
{
"epoch": 13.88888888888889,
"grad_norm": 0.01819729432463646,
"learning_rate": 1.1125362229186056e-05,
"loss": 0.0002,
"step": 194
},
{
"epoch": 14.0,
"grad_norm": 0.01841341145336628,
"learning_rate": 1.0965139209145153e-05,
"loss": 0.0003,
"step": 196
},
{
"epoch": 14.148148148148149,
"grad_norm": 0.008297973312437534,
"learning_rate": 1.0804665687167262e-05,
"loss": 0.0001,
"step": 198
},
{
"epoch": 14.296296296296296,
"grad_norm": 0.021972045302391052,
"learning_rate": 1.0643983314164195e-05,
"loss": 0.0052,
"step": 200
},
{
"epoch": 14.444444444444445,
"grad_norm": 0.025723164901137352,
"learning_rate": 1.0483133795255072e-05,
"loss": 0.0002,
"step": 202
},
{
"epoch": 14.592592592592592,
"grad_norm": 0.09751021862030029,
"learning_rate": 1.0322158878941733e-05,
"loss": 0.0092,
"step": 204
},
{
"epoch": 14.74074074074074,
"grad_norm": 0.009472132660448551,
"learning_rate": 1.0161100346272913e-05,
"loss": 0.0035,
"step": 206
},
{
"epoch": 14.88888888888889,
"grad_norm": 0.006406329572200775,
"learning_rate": 1e-05,
"loss": 0.0005,
"step": 208
},
{
"epoch": 15.0,
"grad_norm": 0.01075380202382803,
"learning_rate": 9.838899653727088e-06,
"loss": 0.0002,
"step": 210
},
{
"epoch": 15.148148148148149,
"grad_norm": 0.008914481848478317,
"learning_rate": 9.677841121058274e-06,
"loss": 0.0001,
"step": 212
},
{
"epoch": 15.296296296296296,
"grad_norm": 0.13058064877986908,
"learning_rate": 9.516866204744932e-06,
"loss": 0.0042,
"step": 214
},
{
"epoch": 15.444444444444445,
"grad_norm": 0.005530489608645439,
"learning_rate": 9.356016685835807e-06,
"loss": 0.0053,
"step": 216
},
{
"epoch": 15.592592592592592,
"grad_norm": 0.07168114185333252,
"learning_rate": 9.195334312832742e-06,
"loss": 0.0093,
"step": 218
},
{
"epoch": 15.74074074074074,
"grad_norm": 0.005750632844865322,
"learning_rate": 9.034860790854848e-06,
"loss": 0.0001,
"step": 220
},
{
"epoch": 15.88888888888889,
"grad_norm": 0.005580791737884283,
"learning_rate": 8.874637770813947e-06,
"loss": 0.0001,
"step": 222
},
{
"epoch": 16.0,
"grad_norm": 0.008345577865839005,
"learning_rate": 8.714706838604056e-06,
"loss": 0.0001,
"step": 224
},
{
"epoch": 16.14814814814815,
"grad_norm": 0.12289520353078842,
"learning_rate": 8.55510950430779e-06,
"loss": 0.0108,
"step": 226
},
{
"epoch": 16.296296296296298,
"grad_norm": 0.008948876522481441,
"learning_rate": 8.395887191422397e-06,
"loss": 0.0001,
"step": 228
},
{
"epoch": 16.444444444444443,
"grad_norm": 0.005637575872242451,
"learning_rate": 8.23708122610831e-06,
"loss": 0.0074,
"step": 230
},
{
"epoch": 16.59259259259259,
"grad_norm": 0.004851146135479212,
"learning_rate": 8.078732826462917e-06,
"loss": 0.0001,
"step": 232
},
{
"epoch": 16.74074074074074,
"grad_norm": 0.004342212807387114,
"learning_rate": 7.92088309182241e-06,
"loss": 0.0001,
"step": 234
},
{
"epoch": 16.88888888888889,
"grad_norm": 0.0062991357408463955,
"learning_rate": 7.763572992094448e-06,
"loss": 0.0001,
"step": 236
},
{
"epoch": 17.0,
"grad_norm": 0.0036818073131144047,
"learning_rate": 7.606843357124426e-06,
"loss": 0.0001,
"step": 238
},
{
"epoch": 17.14814814814815,
"grad_norm": 0.07259424030780792,
"learning_rate": 7.450734866098066e-06,
"loss": 0.0033,
"step": 240
},
{
"epoch": 17.296296296296298,
"grad_norm": 0.004074610769748688,
"learning_rate": 7.2952880369831635e-06,
"loss": 0.0065,
"step": 242
},
{
"epoch": 17.444444444444443,
"grad_norm": 0.1560247242450714,
"learning_rate": 7.140543216013109e-06,
"loss": 0.0045,
"step": 244
},
{
"epoch": 17.59259259259259,
"grad_norm": 0.004389321897178888,
"learning_rate": 6.986540567215043e-06,
"loss": 0.0001,
"step": 246
},
{
"epoch": 17.74074074074074,
"grad_norm": 0.007073231041431427,
"learning_rate": 6.833320061985278e-06,
"loss": 0.0001,
"step": 248
},
{
"epoch": 17.88888888888889,
"grad_norm": 0.006827354431152344,
"learning_rate": 6.680921468714718e-06,
"loss": 0.0047,
"step": 250
},
{
"epoch": 18.0,
"grad_norm": 0.004763526376336813,
"learning_rate": 6.529384342466971e-06,
"loss": 0.0001,
"step": 252
},
{
"epoch": 18.14814814814815,
"grad_norm": 0.005955778993666172,
"learning_rate": 6.378748014711834e-06,
"loss": 0.0001,
"step": 254
},
{
"epoch": 18.296296296296298,
"grad_norm": 0.003511330345645547,
"learning_rate": 6.229051583116796e-06,
"loss": 0.0001,
"step": 256
},
{
"epoch": 18.444444444444443,
"grad_norm": 0.00399408582597971,
"learning_rate": 6.080333901399252e-06,
"loss": 0.0063,
"step": 258
},
{
"epoch": 18.59259259259259,
"grad_norm": 0.004711697343736887,
"learning_rate": 5.932633569242e-06,
"loss": 0.0001,
"step": 260
},
{
"epoch": 18.74074074074074,
"grad_norm": 0.004290228709578514,
"learning_rate": 5.785988922274711e-06,
"loss": 0.0098,
"step": 262
},
{
"epoch": 18.88888888888889,
"grad_norm": 0.05570642277598381,
"learning_rate": 5.640438022123898e-06,
"loss": 0.0031,
"step": 264
},
{
"epoch": 19.0,
"grad_norm": 0.003965294919908047,
"learning_rate": 5.496018646534032e-06,
"loss": 0.0001,
"step": 266
},
{
"epoch": 19.14814814814815,
"grad_norm": 0.003580670338124037,
"learning_rate": 5.352768279562315e-06,
"loss": 0.0093,
"step": 268
},
{
"epoch": 19.296296296296298,
"grad_norm": 0.004285724833607674,
"learning_rate": 5.210724101849696e-06,
"loss": 0.0001,
"step": 270
},
{
"epoch": 19.444444444444443,
"grad_norm": 0.0037607268895953894,
"learning_rate": 5.069922980970626e-06,
"loss": 0.0001,
"step": 272
},
{
"epoch": 19.59259259259259,
"grad_norm": 0.04302617534995079,
"learning_rate": 4.930401461864099e-06,
"loss": 0.0081,
"step": 274
},
{
"epoch": 19.74074074074074,
"grad_norm": 0.0037424350157380104,
"learning_rate": 4.7921957573483756e-06,
"loss": 0.0001,
"step": 276
},
{
"epoch": 19.88888888888889,
"grad_norm": 0.0032615945674479008,
"learning_rate": 4.655341738721989e-06,
"loss": 0.0001,
"step": 278
},
{
"epoch": 20.0,
"grad_norm": 0.003283739322796464,
"learning_rate": 4.519874926453303e-06,
"loss": 0.0,
"step": 280
},
{
"epoch": 20.14814814814815,
"grad_norm": 0.00264250161126256,
"learning_rate": 4.385830480961192e-06,
"loss": 0.0001,
"step": 282
},
{
"epoch": 20.296296296296298,
"grad_norm": 0.00357725634239614,
"learning_rate": 4.2532431934891646e-06,
"loss": 0.0001,
"step": 284
},
{
"epoch": 20.444444444444443,
"grad_norm": 0.003912975545972586,
"learning_rate": 4.12214747707527e-06,
"loss": 0.0054,
"step": 286
},
{
"epoch": 20.59259259259259,
"grad_norm": 0.0031737168319523335,
"learning_rate": 3.99257735762021e-06,
"loss": 0.0001,
"step": 288
},
{
"epoch": 20.74074074074074,
"grad_norm": 0.004108683671802282,
"learning_rate": 3.864566465055913e-06,
"loss": 0.0043,
"step": 290
},
{
"epoch": 20.88888888888889,
"grad_norm": 0.0033303890377283096,
"learning_rate": 3.738148024616863e-06,
"loss": 0.0001,
"step": 292
},
{
"epoch": 21.0,
"grad_norm": 0.0030161114409565926,
"learning_rate": 3.6133548482165225e-06,
"loss": 0.0098,
"step": 294
},
{
"epoch": 21.14814814814815,
"grad_norm": 0.060281865298748016,
"learning_rate": 3.4902193259309627e-06,
"loss": 0.0033,
"step": 296
},
{
"epoch": 21.296296296296298,
"grad_norm": 0.0027096017729490995,
"learning_rate": 3.3687734175920505e-06,
"loss": 0.0,
"step": 298
},
{
"epoch": 21.444444444444443,
"grad_norm": 0.004156744573265314,
"learning_rate": 3.2490486444922396e-06,
"loss": 0.0001,
"step": 300
},
{
"epoch": 21.59259259259259,
"grad_norm": 0.14578433334827423,
"learning_rate": 3.131076081203247e-06,
"loss": 0.0037,
"step": 302
},
{
"epoch": 21.74074074074074,
"grad_norm": 0.0028298762626945972,
"learning_rate": 3.0148863475106315e-06,
"loss": 0.0062,
"step": 304
},
{
"epoch": 21.88888888888889,
"grad_norm": 0.18271295726299286,
"learning_rate": 2.900509600466418e-06,
"loss": 0.0043,
"step": 306
},
{
"epoch": 22.0,
"grad_norm": 0.00494324741885066,
"learning_rate": 2.7879755265618558e-06,
"loss": 0.0001,
"step": 308
},
{
"epoch": 22.14814814814815,
"grad_norm": 0.004431094974279404,
"learning_rate": 2.6773133340222677e-06,
"loss": 0.0038,
"step": 310
},
{
"epoch": 22.296296296296298,
"grad_norm": 0.002698077354580164,
"learning_rate": 2.5685517452260566e-06,
"loss": 0.0001,
"step": 312
},
{
"epoch": 22.444444444444443,
"grad_norm": 0.0034333046060055494,
"learning_rate": 2.4617189892498326e-06,
"loss": 0.0,
"step": 314
},
{
"epoch": 22.59259259259259,
"grad_norm": 0.07087589800357819,
"learning_rate": 2.3568427945415163e-06,
"loss": 0.0089,
"step": 316
},
{
"epoch": 22.74074074074074,
"grad_norm": 0.0028803746681660414,
"learning_rate": 2.2539503817234553e-06,
"loss": 0.0,
"step": 318
},
{
"epoch": 22.88888888888889,
"grad_norm": 0.0028331754729151726,
"learning_rate": 2.153068456527283e-06,
"loss": 0.0,
"step": 320
},
{
"epoch": 23.0,
"grad_norm": 0.24954956769943237,
"learning_rate": 2.0542232028624585e-06,
"loss": 0.0189,
"step": 322
},
{
"epoch": 23.14814814814815,
"grad_norm": 0.07590242475271225,
"learning_rate": 1.9574402760202315e-06,
"loss": 0.0071,
"step": 324
},
{
"epoch": 23.296296296296298,
"grad_norm": 0.07335228472948074,
"learning_rate": 1.8627447960148036e-06,
"loss": 0.0041,
"step": 326
},
{
"epoch": 23.444444444444443,
"grad_norm": 0.002679693978279829,
"learning_rate": 1.7701613410634367e-06,
"loss": 0.0001,
"step": 328
},
{
"epoch": 23.59259259259259,
"grad_norm": 0.008854858577251434,
"learning_rate": 1.6797139412071583e-06,
"loss": 0.0001,
"step": 330
},
{
"epoch": 23.74074074074074,
"grad_norm": 0.003298202995210886,
"learning_rate": 1.5914260720737796e-06,
"loss": 0.0001,
"step": 332
},
{
"epoch": 23.88888888888889,
"grad_norm": 0.0022927140817046165,
"learning_rate": 1.5053206487847916e-06,
"loss": 0.0047,
"step": 334
},
{
"epoch": 24.0,
"grad_norm": 0.0027822547126561403,
"learning_rate": 1.4214200200077343e-06,
"loss": 0.0,
"step": 336
},
{
"epoch": 24.14814814814815,
"grad_norm": 0.0029783721547573805,
"learning_rate": 1.339745962155613e-06,
"loss": 0.0,
"step": 338
},
{
"epoch": 24.296296296296298,
"grad_norm": 0.0020928133744746447,
"learning_rate": 1.2603196737348211e-06,
"loss": 0.0039,
"step": 340
},
{
"epoch": 24.444444444444443,
"grad_norm": 0.0029716058634221554,
"learning_rate": 1.1831617698430609e-06,
"loss": 0.0001,
"step": 342
},
{
"epoch": 24.59259259259259,
"grad_norm": 0.05463359132409096,
"learning_rate": 1.1082922768187098e-06,
"loss": 0.0058,
"step": 344
},
{
"epoch": 24.74074074074074,
"grad_norm": 0.055414874106645584,
"learning_rate": 1.0357306270429623e-06,
"loss": 0.0041,
"step": 346
},
{
"epoch": 24.88888888888889,
"grad_norm": 0.1706482619047165,
"learning_rate": 9.65495653896179e-07,
"loss": 0.0037,
"step": 348
},
{
"epoch": 25.0,
"grad_norm": 0.004686425905674696,
"learning_rate": 8.976055868696543e-07,
"loss": 0.0,
"step": 350
},
{
"epoch": 25.14814814814815,
"grad_norm": 0.002915605204179883,
"learning_rate": 8.320780468341761e-07,
"loss": 0.0001,
"step": 352
},
{
"epoch": 25.296296296296298,
"grad_norm": 0.177982896566391,
"learning_rate": 7.689300414665124e-07,
"loss": 0.0039,
"step": 354
},
{
"epoch": 25.444444444444443,
"grad_norm": 0.0028289929032325745,
"learning_rate": 7.08177960835068e-07,
"loss": 0.0072,
"step": 356
},
{
"epoch": 25.59259259259259,
"grad_norm": 0.16306143999099731,
"learning_rate": 6.498375731458529e-07,
"loss": 0.0035,
"step": 358
},
{
"epoch": 25.74074074074074,
"grad_norm": 0.002770756371319294,
"learning_rate": 5.939240206498287e-07,
"loss": 0.0001,
"step": 360
},
{
"epoch": 25.88888888888889,
"grad_norm": 0.0030980072915554047,
"learning_rate": 5.404518157127481e-07,
"loss": 0.0041,
"step": 362
},
{
"epoch": 26.0,
"grad_norm": 0.0027522058226168156,
"learning_rate": 4.894348370484648e-07,
"loss": 0.0,
"step": 364
},
{
"epoch": 26.14814814814815,
"grad_norm": 0.0026546705048531294,
"learning_rate": 4.408863261167096e-07,
"loss": 0.0,
"step": 366
},
{
"epoch": 26.296296296296298,
"grad_norm": 0.05512738972902298,
"learning_rate": 3.9481888368627764e-07,
"loss": 0.0042,
"step": 368
},
{
"epoch": 26.444444444444443,
"grad_norm": 0.0037136340979486704,
"learning_rate": 3.5124446656448654e-07,
"loss": 0.0001,
"step": 370
},
{
"epoch": 26.59259259259259,
"grad_norm": 0.002223914721980691,
"learning_rate": 3.101743844937943e-07,
"loss": 0.0106,
"step": 372
},
{
"epoch": 26.74074074074074,
"grad_norm": 0.0037102268543094397,
"learning_rate": 2.716192972163556e-07,
"loss": 0.0001,
"step": 374
},
{
"epoch": 26.88888888888889,
"grad_norm": 0.0032115220092236996,
"learning_rate": 2.355892117072789e-07,
"loss": 0.0,
"step": 376
},
{
"epoch": 27.0,
"grad_norm": 0.003928745165467262,
"learning_rate": 2.0209347957732328e-07,
"loss": 0.0039,
"step": 378
},
{
"epoch": 27.14814814814815,
"grad_norm": 0.0024801164399832487,
"learning_rate": 1.711407946456789e-07,
"loss": 0.0,
"step": 380
},
{
"epoch": 27.296296296296298,
"grad_norm": 0.17018507421016693,
"learning_rate": 1.4273919068349184e-07,
"loss": 0.0076,
"step": 382
},
{
"epoch": 27.444444444444443,
"grad_norm": 0.003076062072068453,
"learning_rate": 1.1689603932869664e-07,
"loss": 0.0,
"step": 384
},
{
"epoch": 27.59259259259259,
"grad_norm": 0.002432670909911394,
"learning_rate": 9.36180481727067e-08,
"loss": 0.0041,
"step": 386
},
{
"epoch": 27.74074074074074,
"grad_norm": 0.003008410567417741,
"learning_rate": 7.291125901946027e-08,
"loss": 0.0,
"step": 388
},
{
"epoch": 27.88888888888889,
"grad_norm": 0.048845987766981125,
"learning_rate": 5.4781046317267103e-08,
"loss": 0.0051,
"step": 390
},
{
"epoch": 28.0,
"grad_norm": 0.006181825418025255,
"learning_rate": 3.923211576387087e-08,
"loss": 0.0001,
"step": 392
},
{
"epoch": 28.14814814814815,
"grad_norm": 0.0037348337937146425,
"learning_rate": 2.6268503085089547e-08,
"loss": 0.0001,
"step": 394
},
{
"epoch": 28.296296296296298,
"grad_norm": 0.0033626286312937737,
"learning_rate": 1.5893572987333293e-08,
"loss": 0.0,
"step": 396
},
{
"epoch": 28.444444444444443,
"grad_norm": 0.0655711367726326,
"learning_rate": 8.110018284304132e-09,
"loss": 0.0049,
"step": 398
},
{
"epoch": 28.59259259259259,
"grad_norm": 0.0024890624918043613,
"learning_rate": 2.9198591980705847e-09,
"loss": 0.0037,
"step": 400
},
{
"epoch": 28.74074074074074,
"grad_norm": 0.1838274598121643,
"learning_rate": 3.244428347204398e-10,
"loss": 0.004,
"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": 2.1706596520820736e+16,
"train_batch_size": 8,
"trial_name": null,
"trial_params": null
}