Yaxin's picture
Update from Yaxin
36e38f5
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 3.0,
"global_step": 81789,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.02,
"learning_rate": 4.969433542407904e-05,
"loss": 6.5457,
"step": 500
},
{
"epoch": 0.04,
"learning_rate": 4.938867084815807e-05,
"loss": 5.0387,
"step": 1000
},
{
"epoch": 0.06,
"learning_rate": 4.9083006272237095e-05,
"loss": 4.5967,
"step": 1500
},
{
"epoch": 0.07,
"learning_rate": 4.877734169631613e-05,
"loss": 4.3347,
"step": 2000
},
{
"epoch": 0.09,
"learning_rate": 4.8471677120395165e-05,
"loss": 4.1399,
"step": 2500
},
{
"epoch": 0.11,
"learning_rate": 4.81660125444742e-05,
"loss": 3.9845,
"step": 3000
},
{
"epoch": 0.13,
"learning_rate": 4.786034796855323e-05,
"loss": 3.8614,
"step": 3500
},
{
"epoch": 0.15,
"learning_rate": 4.7554683392632264e-05,
"loss": 3.7864,
"step": 4000
},
{
"epoch": 0.17,
"learning_rate": 4.724901881671129e-05,
"loss": 3.6987,
"step": 4500
},
{
"epoch": 0.18,
"learning_rate": 4.694335424079033e-05,
"loss": 3.6203,
"step": 5000
},
{
"epoch": 0.2,
"learning_rate": 4.663768966486936e-05,
"loss": 3.5568,
"step": 5500
},
{
"epoch": 0.22,
"learning_rate": 4.633202508894839e-05,
"loss": 3.5062,
"step": 6000
},
{
"epoch": 0.24,
"learning_rate": 4.602636051302743e-05,
"loss": 3.4421,
"step": 6500
},
{
"epoch": 0.26,
"learning_rate": 4.572069593710646e-05,
"loss": 3.4067,
"step": 7000
},
{
"epoch": 0.28,
"learning_rate": 4.541503136118549e-05,
"loss": 3.3593,
"step": 7500
},
{
"epoch": 0.29,
"learning_rate": 4.5109366785264526e-05,
"loss": 3.315,
"step": 8000
},
{
"epoch": 0.31,
"learning_rate": 4.4803702209343554e-05,
"loss": 3.2763,
"step": 8500
},
{
"epoch": 0.33,
"learning_rate": 4.449803763342259e-05,
"loss": 3.2345,
"step": 9000
},
{
"epoch": 0.35,
"learning_rate": 4.4192373057501625e-05,
"loss": 3.2031,
"step": 9500
},
{
"epoch": 0.37,
"learning_rate": 4.388670848158066e-05,
"loss": 3.1737,
"step": 10000
},
{
"epoch": 0.39,
"learning_rate": 4.358104390565969e-05,
"loss": 3.14,
"step": 10500
},
{
"epoch": 0.4,
"learning_rate": 4.327537932973872e-05,
"loss": 3.1253,
"step": 11000
},
{
"epoch": 0.42,
"learning_rate": 4.296971475381775e-05,
"loss": 3.089,
"step": 11500
},
{
"epoch": 0.44,
"learning_rate": 4.266405017789679e-05,
"loss": 3.0556,
"step": 12000
},
{
"epoch": 0.46,
"learning_rate": 4.235838560197582e-05,
"loss": 3.0432,
"step": 12500
},
{
"epoch": 0.48,
"learning_rate": 4.205272102605485e-05,
"loss": 3.0195,
"step": 13000
},
{
"epoch": 0.5,
"learning_rate": 4.174705645013388e-05,
"loss": 3.0003,
"step": 13500
},
{
"epoch": 0.51,
"learning_rate": 4.1441391874212915e-05,
"loss": 2.9877,
"step": 14000
},
{
"epoch": 0.53,
"learning_rate": 4.113572729829195e-05,
"loss": 2.951,
"step": 14500
},
{
"epoch": 0.55,
"learning_rate": 4.0830062722370985e-05,
"loss": 2.9401,
"step": 15000
},
{
"epoch": 0.57,
"learning_rate": 4.0524398146450014e-05,
"loss": 2.9238,
"step": 15500
},
{
"epoch": 0.59,
"learning_rate": 4.021873357052904e-05,
"loss": 2.9134,
"step": 16000
},
{
"epoch": 0.61,
"learning_rate": 3.991306899460808e-05,
"loss": 2.8829,
"step": 16500
},
{
"epoch": 0.62,
"learning_rate": 3.960740441868711e-05,
"loss": 2.8803,
"step": 17000
},
{
"epoch": 0.64,
"learning_rate": 3.930173984276614e-05,
"loss": 2.8454,
"step": 17500
},
{
"epoch": 0.66,
"learning_rate": 3.8996075266845176e-05,
"loss": 2.8475,
"step": 18000
},
{
"epoch": 0.68,
"learning_rate": 3.869041069092421e-05,
"loss": 2.838,
"step": 18500
},
{
"epoch": 0.7,
"learning_rate": 3.838474611500324e-05,
"loss": 2.8164,
"step": 19000
},
{
"epoch": 0.72,
"learning_rate": 3.8079081539082275e-05,
"loss": 2.8077,
"step": 19500
},
{
"epoch": 0.73,
"learning_rate": 3.7773416963161304e-05,
"loss": 2.8011,
"step": 20000
},
{
"epoch": 0.75,
"learning_rate": 3.746775238724034e-05,
"loss": 2.7834,
"step": 20500
},
{
"epoch": 0.77,
"learning_rate": 3.7162087811319374e-05,
"loss": 2.7666,
"step": 21000
},
{
"epoch": 0.79,
"learning_rate": 3.685642323539841e-05,
"loss": 2.7591,
"step": 21500
},
{
"epoch": 0.81,
"learning_rate": 3.655075865947744e-05,
"loss": 2.7487,
"step": 22000
},
{
"epoch": 0.83,
"learning_rate": 3.6245094083556466e-05,
"loss": 2.7307,
"step": 22500
},
{
"epoch": 0.84,
"learning_rate": 3.59394295076355e-05,
"loss": 2.7306,
"step": 23000
},
{
"epoch": 0.86,
"learning_rate": 3.563376493171454e-05,
"loss": 2.72,
"step": 23500
},
{
"epoch": 0.88,
"learning_rate": 3.532810035579357e-05,
"loss": 2.7159,
"step": 24000
},
{
"epoch": 0.9,
"learning_rate": 3.50224357798726e-05,
"loss": 2.699,
"step": 24500
},
{
"epoch": 0.92,
"learning_rate": 3.471677120395163e-05,
"loss": 2.6933,
"step": 25000
},
{
"epoch": 0.94,
"learning_rate": 3.4411106628030664e-05,
"loss": 2.6826,
"step": 25500
},
{
"epoch": 0.95,
"learning_rate": 3.41054420521097e-05,
"loss": 2.6856,
"step": 26000
},
{
"epoch": 0.97,
"learning_rate": 3.3799777476188735e-05,
"loss": 2.6802,
"step": 26500
},
{
"epoch": 0.99,
"learning_rate": 3.349411290026776e-05,
"loss": 2.677,
"step": 27000
},
{
"epoch": 1.01,
"learning_rate": 3.318844832434679e-05,
"loss": 2.6471,
"step": 27500
},
{
"epoch": 1.03,
"learning_rate": 3.288278374842583e-05,
"loss": 2.6474,
"step": 28000
},
{
"epoch": 1.05,
"learning_rate": 3.257711917250486e-05,
"loss": 2.636,
"step": 28500
},
{
"epoch": 1.06,
"learning_rate": 3.22714545965839e-05,
"loss": 2.6348,
"step": 29000
},
{
"epoch": 1.08,
"learning_rate": 3.1965790020662926e-05,
"loss": 2.6217,
"step": 29500
},
{
"epoch": 1.1,
"learning_rate": 3.166012544474196e-05,
"loss": 2.6288,
"step": 30000
},
{
"epoch": 1.12,
"learning_rate": 3.135446086882099e-05,
"loss": 2.6043,
"step": 30500
},
{
"epoch": 1.14,
"learning_rate": 3.1048796292900025e-05,
"loss": 2.6075,
"step": 31000
},
{
"epoch": 1.16,
"learning_rate": 3.074313171697906e-05,
"loss": 2.6017,
"step": 31500
},
{
"epoch": 1.17,
"learning_rate": 3.0437467141058092e-05,
"loss": 2.5979,
"step": 32000
},
{
"epoch": 1.19,
"learning_rate": 3.0131802565137124e-05,
"loss": 2.5928,
"step": 32500
},
{
"epoch": 1.21,
"learning_rate": 2.982613798921616e-05,
"loss": 2.5841,
"step": 33000
},
{
"epoch": 1.23,
"learning_rate": 2.9520473413295184e-05,
"loss": 2.5853,
"step": 33500
},
{
"epoch": 1.25,
"learning_rate": 2.921480883737422e-05,
"loss": 2.5731,
"step": 34000
},
{
"epoch": 1.27,
"learning_rate": 2.890914426145325e-05,
"loss": 2.5772,
"step": 34500
},
{
"epoch": 1.28,
"learning_rate": 2.8603479685532286e-05,
"loss": 2.5693,
"step": 35000
},
{
"epoch": 1.3,
"learning_rate": 2.8297815109611318e-05,
"loss": 2.5528,
"step": 35500
},
{
"epoch": 1.32,
"learning_rate": 2.7992150533690353e-05,
"loss": 2.5552,
"step": 36000
},
{
"epoch": 1.34,
"learning_rate": 2.7686485957769382e-05,
"loss": 2.5547,
"step": 36500
},
{
"epoch": 1.36,
"learning_rate": 2.7380821381848414e-05,
"loss": 2.5432,
"step": 37000
},
{
"epoch": 1.38,
"learning_rate": 2.707515680592745e-05,
"loss": 2.5511,
"step": 37500
},
{
"epoch": 1.39,
"learning_rate": 2.676949223000648e-05,
"loss": 2.5405,
"step": 38000
},
{
"epoch": 1.41,
"learning_rate": 2.6463827654085516e-05,
"loss": 2.533,
"step": 38500
},
{
"epoch": 1.43,
"learning_rate": 2.6158163078164548e-05,
"loss": 2.5324,
"step": 39000
},
{
"epoch": 1.45,
"learning_rate": 2.5852498502243576e-05,
"loss": 2.5275,
"step": 39500
},
{
"epoch": 1.47,
"learning_rate": 2.554683392632261e-05,
"loss": 2.5145,
"step": 40000
},
{
"epoch": 1.49,
"learning_rate": 2.5241169350401643e-05,
"loss": 2.5258,
"step": 40500
},
{
"epoch": 1.5,
"learning_rate": 2.493550477448068e-05,
"loss": 2.5094,
"step": 41000
},
{
"epoch": 1.52,
"learning_rate": 2.462984019855971e-05,
"loss": 2.5099,
"step": 41500
},
{
"epoch": 1.54,
"learning_rate": 2.4324175622638742e-05,
"loss": 2.4992,
"step": 42000
},
{
"epoch": 1.56,
"learning_rate": 2.4018511046717778e-05,
"loss": 2.5137,
"step": 42500
},
{
"epoch": 1.58,
"learning_rate": 2.3712846470796806e-05,
"loss": 2.4952,
"step": 43000
},
{
"epoch": 1.6,
"learning_rate": 2.340718189487584e-05,
"loss": 2.4876,
"step": 43500
},
{
"epoch": 1.61,
"learning_rate": 2.3101517318954873e-05,
"loss": 2.4915,
"step": 44000
},
{
"epoch": 1.63,
"learning_rate": 2.2795852743033905e-05,
"loss": 2.4845,
"step": 44500
},
{
"epoch": 1.65,
"learning_rate": 2.2490188167112937e-05,
"loss": 2.4816,
"step": 45000
},
{
"epoch": 1.67,
"learning_rate": 2.2184523591191972e-05,
"loss": 2.482,
"step": 45500
},
{
"epoch": 1.69,
"learning_rate": 2.1878859015271004e-05,
"loss": 2.4739,
"step": 46000
},
{
"epoch": 1.71,
"learning_rate": 2.1573194439350036e-05,
"loss": 2.4801,
"step": 46500
},
{
"epoch": 1.72,
"learning_rate": 2.126752986342907e-05,
"loss": 2.4809,
"step": 47000
},
{
"epoch": 1.74,
"learning_rate": 2.09618652875081e-05,
"loss": 2.4828,
"step": 47500
},
{
"epoch": 1.76,
"learning_rate": 2.0656200711587135e-05,
"loss": 2.4721,
"step": 48000
},
{
"epoch": 1.78,
"learning_rate": 2.0350536135666167e-05,
"loss": 2.4664,
"step": 48500
},
{
"epoch": 1.8,
"learning_rate": 2.00448715597452e-05,
"loss": 2.4725,
"step": 49000
},
{
"epoch": 1.82,
"learning_rate": 1.9739206983824234e-05,
"loss": 2.4721,
"step": 49500
},
{
"epoch": 1.83,
"learning_rate": 1.9433542407903262e-05,
"loss": 2.457,
"step": 50000
},
{
"epoch": 1.85,
"learning_rate": 1.9127877831982297e-05,
"loss": 2.4589,
"step": 50500
},
{
"epoch": 1.87,
"learning_rate": 1.882221325606133e-05,
"loss": 2.452,
"step": 51000
},
{
"epoch": 1.89,
"learning_rate": 1.851654868014036e-05,
"loss": 2.47,
"step": 51500
},
{
"epoch": 1.91,
"learning_rate": 1.8210884104219393e-05,
"loss": 2.4442,
"step": 52000
},
{
"epoch": 1.93,
"learning_rate": 1.7905219528298428e-05,
"loss": 2.4369,
"step": 52500
},
{
"epoch": 1.94,
"learning_rate": 1.759955495237746e-05,
"loss": 2.4525,
"step": 53000
},
{
"epoch": 1.96,
"learning_rate": 1.7293890376456492e-05,
"loss": 2.4357,
"step": 53500
},
{
"epoch": 1.98,
"learning_rate": 1.6988225800535527e-05,
"loss": 2.4435,
"step": 54000
},
{
"epoch": 2.0,
"learning_rate": 1.6682561224614556e-05,
"loss": 2.4401,
"step": 54500
},
{
"epoch": 2.02,
"learning_rate": 1.637689664869359e-05,
"loss": 2.4324,
"step": 55000
},
{
"epoch": 2.04,
"learning_rate": 1.6071232072772623e-05,
"loss": 2.414,
"step": 55500
},
{
"epoch": 2.05,
"learning_rate": 1.5765567496851654e-05,
"loss": 2.4213,
"step": 56000
},
{
"epoch": 2.07,
"learning_rate": 1.545990292093069e-05,
"loss": 2.4357,
"step": 56500
},
{
"epoch": 2.09,
"learning_rate": 1.5154238345009722e-05,
"loss": 2.4265,
"step": 57000
},
{
"epoch": 2.11,
"learning_rate": 1.4848573769088753e-05,
"loss": 2.4244,
"step": 57500
},
{
"epoch": 2.13,
"learning_rate": 1.4542909193167787e-05,
"loss": 2.421,
"step": 58000
},
{
"epoch": 2.15,
"learning_rate": 1.423724461724682e-05,
"loss": 2.416,
"step": 58500
},
{
"epoch": 2.16,
"learning_rate": 1.393158004132585e-05,
"loss": 2.4055,
"step": 59000
},
{
"epoch": 2.18,
"learning_rate": 1.3625915465404884e-05,
"loss": 2.4181,
"step": 59500
},
{
"epoch": 2.2,
"learning_rate": 1.3320250889483918e-05,
"loss": 2.4271,
"step": 60000
},
{
"epoch": 2.22,
"learning_rate": 1.3014586313562948e-05,
"loss": 2.4139,
"step": 60500
},
{
"epoch": 2.24,
"learning_rate": 1.2708921737641981e-05,
"loss": 2.4196,
"step": 61000
},
{
"epoch": 2.26,
"learning_rate": 1.2403257161721015e-05,
"loss": 2.4099,
"step": 61500
},
{
"epoch": 2.27,
"learning_rate": 1.2097592585800047e-05,
"loss": 2.4104,
"step": 62000
},
{
"epoch": 2.29,
"learning_rate": 1.179192800987908e-05,
"loss": 2.4068,
"step": 62500
},
{
"epoch": 2.31,
"learning_rate": 1.1486263433958112e-05,
"loss": 2.3954,
"step": 63000
},
{
"epoch": 2.33,
"learning_rate": 1.1180598858037144e-05,
"loss": 2.4088,
"step": 63500
},
{
"epoch": 2.35,
"learning_rate": 1.0874934282116178e-05,
"loss": 2.3961,
"step": 64000
},
{
"epoch": 2.37,
"learning_rate": 1.056926970619521e-05,
"loss": 2.3877,
"step": 64500
},
{
"epoch": 2.38,
"learning_rate": 1.0263605130274243e-05,
"loss": 2.4076,
"step": 65000
},
{
"epoch": 2.4,
"learning_rate": 9.957940554353275e-06,
"loss": 2.4018,
"step": 65500
},
{
"epoch": 2.42,
"learning_rate": 9.652275978432308e-06,
"loss": 2.3974,
"step": 66000
},
{
"epoch": 2.44,
"learning_rate": 9.34661140251134e-06,
"loss": 2.3806,
"step": 66500
},
{
"epoch": 2.46,
"learning_rate": 9.040946826590372e-06,
"loss": 2.3847,
"step": 67000
},
{
"epoch": 2.48,
"learning_rate": 8.735282250669406e-06,
"loss": 2.3928,
"step": 67500
},
{
"epoch": 2.49,
"learning_rate": 8.42961767474844e-06,
"loss": 2.4031,
"step": 68000
},
{
"epoch": 2.51,
"learning_rate": 8.123953098827471e-06,
"loss": 2.3912,
"step": 68500
},
{
"epoch": 2.53,
"learning_rate": 7.818288522906505e-06,
"loss": 2.389,
"step": 69000
},
{
"epoch": 2.55,
"learning_rate": 7.5126239469855365e-06,
"loss": 2.3909,
"step": 69500
},
{
"epoch": 2.57,
"learning_rate": 7.206959371064568e-06,
"loss": 2.3793,
"step": 70000
},
{
"epoch": 2.59,
"learning_rate": 6.901294795143602e-06,
"loss": 2.4011,
"step": 70500
},
{
"epoch": 2.6,
"learning_rate": 6.595630219222635e-06,
"loss": 2.3929,
"step": 71000
},
{
"epoch": 2.62,
"learning_rate": 6.2899656433016665e-06,
"loss": 2.3841,
"step": 71500
},
{
"epoch": 2.64,
"learning_rate": 5.984301067380699e-06,
"loss": 2.3927,
"step": 72000
},
{
"epoch": 2.66,
"learning_rate": 5.678636491459733e-06,
"loss": 2.3821,
"step": 72500
},
{
"epoch": 2.68,
"learning_rate": 5.3729719155387646e-06,
"loss": 2.3881,
"step": 73000
},
{
"epoch": 2.7,
"learning_rate": 5.067307339617797e-06,
"loss": 2.3934,
"step": 73500
},
{
"epoch": 2.71,
"learning_rate": 4.76164276369683e-06,
"loss": 2.3861,
"step": 74000
},
{
"epoch": 2.73,
"learning_rate": 4.455978187775863e-06,
"loss": 2.3814,
"step": 74500
},
{
"epoch": 2.75,
"learning_rate": 4.150313611854895e-06,
"loss": 2.3782,
"step": 75000
},
{
"epoch": 2.77,
"learning_rate": 3.844649035933927e-06,
"loss": 2.3703,
"step": 75500
},
{
"epoch": 2.79,
"learning_rate": 3.5389844600129603e-06,
"loss": 2.3897,
"step": 76000
},
{
"epoch": 2.81,
"learning_rate": 3.2333198840919934e-06,
"loss": 2.3759,
"step": 76500
},
{
"epoch": 2.82,
"learning_rate": 2.9276553081710253e-06,
"loss": 2.3732,
"step": 77000
},
{
"epoch": 2.84,
"learning_rate": 2.6219907322500584e-06,
"loss": 2.3762,
"step": 77500
},
{
"epoch": 2.86,
"learning_rate": 2.316326156329091e-06,
"loss": 2.388,
"step": 78000
},
{
"epoch": 2.88,
"learning_rate": 2.0106615804081234e-06,
"loss": 2.3819,
"step": 78500
},
{
"epoch": 2.9,
"learning_rate": 1.7049970044871559e-06,
"loss": 2.3834,
"step": 79000
},
{
"epoch": 2.92,
"learning_rate": 1.3993324285661888e-06,
"loss": 2.3727,
"step": 79500
},
{
"epoch": 2.93,
"learning_rate": 1.0936678526452213e-06,
"loss": 2.3974,
"step": 80000
},
{
"epoch": 2.95,
"learning_rate": 7.88003276724254e-07,
"loss": 2.3869,
"step": 80500
},
{
"epoch": 2.97,
"learning_rate": 4.823387008032866e-07,
"loss": 2.3804,
"step": 81000
},
{
"epoch": 2.99,
"learning_rate": 1.7667412488231913e-07,
"loss": 2.3848,
"step": 81500
},
{
"epoch": 3.0,
"step": 81789,
"total_flos": 1.9244068885776384e+16,
"train_loss": 2.7128201870760336,
"train_runtime": 9561.6624,
"train_samples_per_second": 68.43,
"train_steps_per_second": 8.554
}
],
"max_steps": 81789,
"num_train_epochs": 3,
"total_flos": 1.9244068885776384e+16,
"trial_name": null,
"trial_params": null
}