DIALECT / trainer_state.json
shensq0814's picture
Upload 11 files
184ed3d
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 2.8414472437961735,
"global_step": 75000,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.02,
"learning_rate": 9.936856727915642e-06,
"loss": 1.4639,
"step": 500
},
{
"epoch": 0.04,
"learning_rate": 9.873713455831283e-06,
"loss": 1.1219,
"step": 1000
},
{
"epoch": 0.06,
"learning_rate": 9.810570183746922e-06,
"loss": 1.0672,
"step": 1500
},
{
"epoch": 0.08,
"learning_rate": 9.747426911662563e-06,
"loss": 0.9959,
"step": 2000
},
{
"epoch": 0.09,
"learning_rate": 9.684283639578204e-06,
"loss": 0.9887,
"step": 2500
},
{
"epoch": 0.11,
"learning_rate": 9.621140367493844e-06,
"loss": 0.9706,
"step": 3000
},
{
"epoch": 0.13,
"learning_rate": 9.557997095409485e-06,
"loss": 0.9167,
"step": 3500
},
{
"epoch": 0.15,
"learning_rate": 9.494853823325125e-06,
"loss": 0.9241,
"step": 4000
},
{
"epoch": 0.17,
"learning_rate": 9.431710551240765e-06,
"loss": 0.9152,
"step": 4500
},
{
"epoch": 0.19,
"learning_rate": 9.368567279156406e-06,
"loss": 0.897,
"step": 5000
},
{
"epoch": 0.19,
"eval_loss": 1.8563896417617798,
"eval_runtime": 4.2841,
"eval_samples_per_second": 116.71,
"eval_steps_per_second": 14.706,
"step": 5000
},
{
"epoch": 0.21,
"learning_rate": 9.305424007072047e-06,
"loss": 0.879,
"step": 5500
},
{
"epoch": 0.23,
"learning_rate": 9.242280734987688e-06,
"loss": 0.8759,
"step": 6000
},
{
"epoch": 0.25,
"learning_rate": 9.179137462903327e-06,
"loss": 0.8799,
"step": 6500
},
{
"epoch": 0.27,
"learning_rate": 9.115994190818968e-06,
"loss": 0.8683,
"step": 7000
},
{
"epoch": 0.28,
"learning_rate": 9.052850918734609e-06,
"loss": 0.8598,
"step": 7500
},
{
"epoch": 0.3,
"learning_rate": 8.98970764665025e-06,
"loss": 0.8662,
"step": 8000
},
{
"epoch": 0.32,
"learning_rate": 8.92656437456589e-06,
"loss": 0.8528,
"step": 8500
},
{
"epoch": 0.34,
"learning_rate": 8.863421102481532e-06,
"loss": 0.8444,
"step": 9000
},
{
"epoch": 0.36,
"learning_rate": 8.800277830397171e-06,
"loss": 0.8632,
"step": 9500
},
{
"epoch": 0.38,
"learning_rate": 8.737134558312812e-06,
"loss": 0.852,
"step": 10000
},
{
"epoch": 0.38,
"eval_loss": 1.813950777053833,
"eval_runtime": 4.2949,
"eval_samples_per_second": 116.417,
"eval_steps_per_second": 14.669,
"step": 10000
},
{
"epoch": 0.4,
"learning_rate": 8.673991286228453e-06,
"loss": 0.8545,
"step": 10500
},
{
"epoch": 0.42,
"learning_rate": 8.610848014144094e-06,
"loss": 0.8312,
"step": 11000
},
{
"epoch": 0.44,
"learning_rate": 8.547704742059734e-06,
"loss": 0.84,
"step": 11500
},
{
"epoch": 0.45,
"learning_rate": 8.484561469975375e-06,
"loss": 0.8304,
"step": 12000
},
{
"epoch": 0.47,
"learning_rate": 8.421418197891016e-06,
"loss": 0.83,
"step": 12500
},
{
"epoch": 0.49,
"learning_rate": 8.358274925806657e-06,
"loss": 0.7981,
"step": 13000
},
{
"epoch": 0.51,
"learning_rate": 8.295131653722296e-06,
"loss": 0.8254,
"step": 13500
},
{
"epoch": 0.53,
"learning_rate": 8.231988381637937e-06,
"loss": 0.8275,
"step": 14000
},
{
"epoch": 0.55,
"learning_rate": 8.168845109553578e-06,
"loss": 0.8141,
"step": 14500
},
{
"epoch": 0.57,
"learning_rate": 8.105701837469219e-06,
"loss": 0.8096,
"step": 15000
},
{
"epoch": 0.57,
"eval_loss": 1.7879245281219482,
"eval_runtime": 4.3051,
"eval_samples_per_second": 116.142,
"eval_steps_per_second": 14.634,
"step": 15000
},
{
"epoch": 0.59,
"learning_rate": 8.04255856538486e-06,
"loss": 0.7986,
"step": 15500
},
{
"epoch": 0.61,
"learning_rate": 7.979415293300499e-06,
"loss": 0.8069,
"step": 16000
},
{
"epoch": 0.63,
"learning_rate": 7.91627202121614e-06,
"loss": 0.8343,
"step": 16500
},
{
"epoch": 0.64,
"learning_rate": 7.85312874913178e-06,
"loss": 0.8069,
"step": 17000
},
{
"epoch": 0.66,
"learning_rate": 7.789985477047422e-06,
"loss": 0.8089,
"step": 17500
},
{
"epoch": 0.68,
"learning_rate": 7.726842204963063e-06,
"loss": 0.8144,
"step": 18000
},
{
"epoch": 0.7,
"learning_rate": 7.663698932878702e-06,
"loss": 0.7886,
"step": 18500
},
{
"epoch": 0.72,
"learning_rate": 7.6005556607943435e-06,
"loss": 0.7968,
"step": 19000
},
{
"epoch": 0.74,
"learning_rate": 7.5374123887099835e-06,
"loss": 0.7836,
"step": 19500
},
{
"epoch": 0.76,
"learning_rate": 7.474269116625624e-06,
"loss": 0.8243,
"step": 20000
},
{
"epoch": 0.76,
"eval_loss": 1.7740414142608643,
"eval_runtime": 4.2937,
"eval_samples_per_second": 116.45,
"eval_steps_per_second": 14.673,
"step": 20000
},
{
"epoch": 0.78,
"learning_rate": 7.4111258445412644e-06,
"loss": 0.7716,
"step": 20500
},
{
"epoch": 0.8,
"learning_rate": 7.347982572456905e-06,
"loss": 0.7921,
"step": 21000
},
{
"epoch": 0.81,
"learning_rate": 7.284839300372546e-06,
"loss": 0.7943,
"step": 21500
},
{
"epoch": 0.83,
"learning_rate": 7.221696028288186e-06,
"loss": 0.7644,
"step": 22000
},
{
"epoch": 0.85,
"learning_rate": 7.158552756203827e-06,
"loss": 0.7841,
"step": 22500
},
{
"epoch": 0.87,
"learning_rate": 7.095409484119468e-06,
"loss": 0.7866,
"step": 23000
},
{
"epoch": 0.89,
"learning_rate": 7.032266212035108e-06,
"loss": 0.7967,
"step": 23500
},
{
"epoch": 0.91,
"learning_rate": 6.969122939950749e-06,
"loss": 0.7776,
"step": 24000
},
{
"epoch": 0.93,
"learning_rate": 6.905979667866389e-06,
"loss": 0.7704,
"step": 24500
},
{
"epoch": 0.95,
"learning_rate": 6.84283639578203e-06,
"loss": 0.7536,
"step": 25000
},
{
"epoch": 0.95,
"eval_loss": 1.7646362781524658,
"eval_runtime": 4.2984,
"eval_samples_per_second": 116.323,
"eval_steps_per_second": 14.657,
"step": 25000
},
{
"epoch": 0.97,
"learning_rate": 6.779693123697671e-06,
"loss": 0.7886,
"step": 25500
},
{
"epoch": 0.99,
"learning_rate": 6.716549851613311e-06,
"loss": 0.7812,
"step": 26000
},
{
"epoch": 1.0,
"learning_rate": 6.653406579528952e-06,
"loss": 0.7845,
"step": 26500
},
{
"epoch": 1.02,
"learning_rate": 6.590263307444592e-06,
"loss": 0.7634,
"step": 27000
},
{
"epoch": 1.04,
"learning_rate": 6.527120035360233e-06,
"loss": 0.7592,
"step": 27500
},
{
"epoch": 1.06,
"learning_rate": 6.4639767632758735e-06,
"loss": 0.7494,
"step": 28000
},
{
"epoch": 1.08,
"learning_rate": 6.4008334911915135e-06,
"loss": 0.749,
"step": 28500
},
{
"epoch": 1.1,
"learning_rate": 6.3376902191071544e-06,
"loss": 0.7592,
"step": 29000
},
{
"epoch": 1.12,
"learning_rate": 6.2745469470227945e-06,
"loss": 0.7355,
"step": 29500
},
{
"epoch": 1.14,
"learning_rate": 6.211403674938435e-06,
"loss": 0.7442,
"step": 30000
},
{
"epoch": 1.14,
"eval_loss": 1.762963891029358,
"eval_runtime": 4.2966,
"eval_samples_per_second": 116.372,
"eval_steps_per_second": 14.663,
"step": 30000
},
{
"epoch": 1.16,
"learning_rate": 6.148260402854076e-06,
"loss": 0.7629,
"step": 30500
},
{
"epoch": 1.17,
"learning_rate": 6.085117130769716e-06,
"loss": 0.7227,
"step": 31000
},
{
"epoch": 1.19,
"learning_rate": 6.021973858685357e-06,
"loss": 0.738,
"step": 31500
},
{
"epoch": 1.21,
"learning_rate": 5.958830586600998e-06,
"loss": 0.7293,
"step": 32000
},
{
"epoch": 1.23,
"learning_rate": 5.895687314516638e-06,
"loss": 0.7434,
"step": 32500
},
{
"epoch": 1.25,
"learning_rate": 5.832544042432279e-06,
"loss": 0.7548,
"step": 33000
},
{
"epoch": 1.27,
"learning_rate": 5.769400770347921e-06,
"loss": 0.7304,
"step": 33500
},
{
"epoch": 1.29,
"learning_rate": 5.706257498263561e-06,
"loss": 0.7331,
"step": 34000
},
{
"epoch": 1.31,
"learning_rate": 5.643114226179202e-06,
"loss": 0.7387,
"step": 34500
},
{
"epoch": 1.33,
"learning_rate": 5.5799709540948425e-06,
"loss": 0.7482,
"step": 35000
},
{
"epoch": 1.33,
"eval_loss": 1.7577964067459106,
"eval_runtime": 4.9786,
"eval_samples_per_second": 100.43,
"eval_steps_per_second": 12.654,
"step": 35000
},
{
"epoch": 1.34,
"learning_rate": 5.5168276820104826e-06,
"loss": 0.7177,
"step": 35500
},
{
"epoch": 1.36,
"learning_rate": 5.4536844099261234e-06,
"loss": 0.7227,
"step": 36000
},
{
"epoch": 1.38,
"learning_rate": 5.3905411378417635e-06,
"loss": 0.723,
"step": 36500
},
{
"epoch": 1.4,
"learning_rate": 5.327397865757404e-06,
"loss": 0.7255,
"step": 37000
},
{
"epoch": 1.42,
"learning_rate": 5.264254593673045e-06,
"loss": 0.7269,
"step": 37500
},
{
"epoch": 1.44,
"learning_rate": 5.201111321588685e-06,
"loss": 0.7528,
"step": 38000
},
{
"epoch": 1.46,
"learning_rate": 5.137968049504326e-06,
"loss": 0.7377,
"step": 38500
},
{
"epoch": 1.48,
"learning_rate": 5.074824777419966e-06,
"loss": 0.7317,
"step": 39000
},
{
"epoch": 1.5,
"learning_rate": 5.011681505335607e-06,
"loss": 0.7486,
"step": 39500
},
{
"epoch": 1.52,
"learning_rate": 4.948538233251248e-06,
"loss": 0.7126,
"step": 40000
},
{
"epoch": 1.52,
"eval_loss": 1.7534527778625488,
"eval_runtime": 4.2918,
"eval_samples_per_second": 116.502,
"eval_steps_per_second": 14.679,
"step": 40000
},
{
"epoch": 1.53,
"learning_rate": 4.885394961166888e-06,
"loss": 0.7211,
"step": 40500
},
{
"epoch": 1.55,
"learning_rate": 4.822251689082529e-06,
"loss": 0.7059,
"step": 41000
},
{
"epoch": 1.57,
"learning_rate": 4.75910841699817e-06,
"loss": 0.7331,
"step": 41500
},
{
"epoch": 1.59,
"learning_rate": 4.69596514491381e-06,
"loss": 0.7199,
"step": 42000
},
{
"epoch": 1.61,
"learning_rate": 4.632821872829451e-06,
"loss": 0.7026,
"step": 42500
},
{
"epoch": 1.63,
"learning_rate": 4.569678600745091e-06,
"loss": 0.7181,
"step": 43000
},
{
"epoch": 1.65,
"learning_rate": 4.506535328660732e-06,
"loss": 0.725,
"step": 43500
},
{
"epoch": 1.67,
"learning_rate": 4.4433920565763725e-06,
"loss": 0.7281,
"step": 44000
},
{
"epoch": 1.69,
"learning_rate": 4.380248784492013e-06,
"loss": 0.714,
"step": 44500
},
{
"epoch": 1.7,
"learning_rate": 4.3171055124076535e-06,
"loss": 0.7123,
"step": 45000
},
{
"epoch": 1.7,
"eval_loss": 1.7550365924835205,
"eval_runtime": 4.9628,
"eval_samples_per_second": 100.749,
"eval_steps_per_second": 12.694,
"step": 45000
},
{
"epoch": 1.72,
"learning_rate": 4.2539622403232935e-06,
"loss": 0.7039,
"step": 45500
},
{
"epoch": 1.74,
"learning_rate": 4.190818968238934e-06,
"loss": 0.7207,
"step": 46000
},
{
"epoch": 1.76,
"learning_rate": 4.127675696154575e-06,
"loss": 0.7245,
"step": 46500
},
{
"epoch": 1.78,
"learning_rate": 4.064532424070215e-06,
"loss": 0.7058,
"step": 47000
},
{
"epoch": 1.8,
"learning_rate": 4.001389151985856e-06,
"loss": 0.7216,
"step": 47500
},
{
"epoch": 1.82,
"learning_rate": 3.938245879901496e-06,
"loss": 0.7339,
"step": 48000
},
{
"epoch": 1.84,
"learning_rate": 3.875102607817137e-06,
"loss": 0.7354,
"step": 48500
},
{
"epoch": 1.86,
"learning_rate": 3.8119593357327776e-06,
"loss": 0.7281,
"step": 49000
},
{
"epoch": 1.88,
"learning_rate": 3.7488160636484185e-06,
"loss": 0.7297,
"step": 49500
},
{
"epoch": 1.89,
"learning_rate": 3.685672791564059e-06,
"loss": 0.7211,
"step": 50000
},
{
"epoch": 1.89,
"eval_loss": 1.7511944770812988,
"eval_runtime": 4.9613,
"eval_samples_per_second": 100.78,
"eval_steps_per_second": 12.698,
"step": 50000
},
{
"epoch": 1.91,
"learning_rate": 3.6225295194797e-06,
"loss": 0.7116,
"step": 50500
},
{
"epoch": 1.93,
"learning_rate": 3.5593862473953407e-06,
"loss": 0.7168,
"step": 51000
},
{
"epoch": 1.95,
"learning_rate": 3.496242975310981e-06,
"loss": 0.7073,
"step": 51500
},
{
"epoch": 1.97,
"learning_rate": 3.4330997032266216e-06,
"loss": 0.7085,
"step": 52000
},
{
"epoch": 1.99,
"learning_rate": 3.369956431142262e-06,
"loss": 0.7143,
"step": 52500
},
{
"epoch": 2.01,
"learning_rate": 3.306813159057903e-06,
"loss": 0.7134,
"step": 53000
},
{
"epoch": 2.03,
"learning_rate": 3.2436698869735434e-06,
"loss": 0.6899,
"step": 53500
},
{
"epoch": 2.05,
"learning_rate": 3.180526614889184e-06,
"loss": 0.6816,
"step": 54000
},
{
"epoch": 2.06,
"learning_rate": 3.1173833428048244e-06,
"loss": 0.7157,
"step": 54500
},
{
"epoch": 2.08,
"learning_rate": 3.054240070720465e-06,
"loss": 0.6791,
"step": 55000
},
{
"epoch": 2.08,
"eval_loss": 1.7588887214660645,
"eval_runtime": 4.9575,
"eval_samples_per_second": 100.856,
"eval_steps_per_second": 12.708,
"step": 55000
},
{
"epoch": 2.1,
"learning_rate": 2.9910967986361057e-06,
"loss": 0.693,
"step": 55500
},
{
"epoch": 2.12,
"learning_rate": 2.927953526551746e-06,
"loss": 0.6793,
"step": 56000
},
{
"epoch": 2.14,
"learning_rate": 2.8648102544673866e-06,
"loss": 0.6825,
"step": 56500
},
{
"epoch": 2.16,
"learning_rate": 2.801666982383027e-06,
"loss": 0.6897,
"step": 57000
},
{
"epoch": 2.18,
"learning_rate": 2.738523710298668e-06,
"loss": 0.6834,
"step": 57500
},
{
"epoch": 2.2,
"learning_rate": 2.6753804382143085e-06,
"loss": 0.6959,
"step": 58000
},
{
"epoch": 2.22,
"learning_rate": 2.612237166129949e-06,
"loss": 0.6977,
"step": 58500
},
{
"epoch": 2.24,
"learning_rate": 2.5490938940455894e-06,
"loss": 0.6881,
"step": 59000
},
{
"epoch": 2.25,
"learning_rate": 2.4859506219612303e-06,
"loss": 0.7066,
"step": 59500
},
{
"epoch": 2.27,
"learning_rate": 2.4228073498768707e-06,
"loss": 0.6731,
"step": 60000
},
{
"epoch": 2.27,
"eval_loss": 1.7606250047683716,
"eval_runtime": 4.9678,
"eval_samples_per_second": 100.649,
"eval_steps_per_second": 12.682,
"step": 60000
},
{
"epoch": 2.29,
"learning_rate": 2.3596640777925116e-06,
"loss": 0.6904,
"step": 60500
},
{
"epoch": 2.31,
"learning_rate": 2.296520805708152e-06,
"loss": 0.6873,
"step": 61000
},
{
"epoch": 2.33,
"learning_rate": 2.2333775336237925e-06,
"loss": 0.701,
"step": 61500
},
{
"epoch": 2.35,
"learning_rate": 2.170234261539433e-06,
"loss": 0.6964,
"step": 62000
},
{
"epoch": 2.37,
"learning_rate": 2.1070909894550735e-06,
"loss": 0.7201,
"step": 62500
},
{
"epoch": 2.39,
"learning_rate": 2.0439477173707144e-06,
"loss": 0.6866,
"step": 63000
},
{
"epoch": 2.41,
"learning_rate": 1.980804445286355e-06,
"loss": 0.6929,
"step": 63500
},
{
"epoch": 2.42,
"learning_rate": 1.9176611732019953e-06,
"loss": 0.6939,
"step": 64000
},
{
"epoch": 2.44,
"learning_rate": 1.8545179011176362e-06,
"loss": 0.701,
"step": 64500
},
{
"epoch": 2.46,
"learning_rate": 1.7913746290332768e-06,
"loss": 0.682,
"step": 65000
},
{
"epoch": 2.46,
"eval_loss": 1.7586958408355713,
"eval_runtime": 4.3201,
"eval_samples_per_second": 115.738,
"eval_steps_per_second": 14.583,
"step": 65000
},
{
"epoch": 2.48,
"learning_rate": 1.7282313569489173e-06,
"loss": 0.6694,
"step": 65500
},
{
"epoch": 2.5,
"learning_rate": 1.6650880848645578e-06,
"loss": 0.7003,
"step": 66000
},
{
"epoch": 2.52,
"learning_rate": 1.6019448127801984e-06,
"loss": 0.6837,
"step": 66500
},
{
"epoch": 2.54,
"learning_rate": 1.538801540695839e-06,
"loss": 0.7013,
"step": 67000
},
{
"epoch": 2.56,
"learning_rate": 1.4756582686114796e-06,
"loss": 0.6882,
"step": 67500
},
{
"epoch": 2.58,
"learning_rate": 1.41251499652712e-06,
"loss": 0.6961,
"step": 68000
},
{
"epoch": 2.6,
"learning_rate": 1.3493717244427607e-06,
"loss": 0.6827,
"step": 68500
},
{
"epoch": 2.61,
"learning_rate": 1.2862284523584012e-06,
"loss": 0.6844,
"step": 69000
},
{
"epoch": 2.63,
"learning_rate": 1.2230851802740419e-06,
"loss": 0.6992,
"step": 69500
},
{
"epoch": 2.65,
"learning_rate": 1.1599419081896825e-06,
"loss": 0.6857,
"step": 70000
},
{
"epoch": 2.65,
"eval_loss": 1.7555959224700928,
"eval_runtime": 4.9634,
"eval_samples_per_second": 100.737,
"eval_steps_per_second": 12.693,
"step": 70000
},
{
"epoch": 2.67,
"learning_rate": 1.096798636105323e-06,
"loss": 0.6833,
"step": 70500
},
{
"epoch": 2.69,
"learning_rate": 1.0336553640209637e-06,
"loss": 0.6754,
"step": 71000
},
{
"epoch": 2.71,
"learning_rate": 9.705120919366043e-07,
"loss": 0.675,
"step": 71500
},
{
"epoch": 2.73,
"learning_rate": 9.073688198522448e-07,
"loss": 0.6849,
"step": 72000
},
{
"epoch": 2.75,
"learning_rate": 8.442255477678854e-07,
"loss": 0.6903,
"step": 72500
},
{
"epoch": 2.77,
"learning_rate": 7.810822756835259e-07,
"loss": 0.6832,
"step": 73000
},
{
"epoch": 2.78,
"learning_rate": 7.179390035991665e-07,
"loss": 0.6823,
"step": 73500
},
{
"epoch": 2.8,
"learning_rate": 6.547957315148072e-07,
"loss": 0.6854,
"step": 74000
},
{
"epoch": 2.82,
"learning_rate": 5.916524594304478e-07,
"loss": 0.681,
"step": 74500
},
{
"epoch": 2.84,
"learning_rate": 5.285091873460883e-07,
"loss": 0.6806,
"step": 75000
},
{
"epoch": 2.84,
"eval_loss": 1.7584295272827148,
"eval_runtime": 4.313,
"eval_samples_per_second": 115.929,
"eval_steps_per_second": 14.607,
"step": 75000
}
],
"max_steps": 79185,
"num_train_epochs": 3,
"total_flos": 1.939234887273808e+18,
"trial_name": null,
"trial_params": null
}