m2m100_418M_swa_en_rel_news / trainer_state.json
Davlan's picture
add MT model
f59c1de
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 3.0,
"global_step": 135420,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.01,
"learning_rate": 4.9815389159651454e-05,
"loss": 2.0517,
"step": 500
},
{
"epoch": 0.02,
"learning_rate": 4.963077831930291e-05,
"loss": 1.8571,
"step": 1000
},
{
"epoch": 0.03,
"learning_rate": 4.9446167478954364e-05,
"loss": 1.7584,
"step": 1500
},
{
"epoch": 0.04,
"learning_rate": 4.926155663860582e-05,
"loss": 1.716,
"step": 2000
},
{
"epoch": 0.06,
"learning_rate": 4.9076945798257274e-05,
"loss": 1.6634,
"step": 2500
},
{
"epoch": 0.07,
"learning_rate": 4.889233495790873e-05,
"loss": 1.6163,
"step": 3000
},
{
"epoch": 0.08,
"learning_rate": 4.8707724117560184e-05,
"loss": 1.5747,
"step": 3500
},
{
"epoch": 0.09,
"learning_rate": 4.852311327721164e-05,
"loss": 1.5611,
"step": 4000
},
{
"epoch": 0.1,
"learning_rate": 4.8338502436863094e-05,
"loss": 1.5445,
"step": 4500
},
{
"epoch": 0.11,
"learning_rate": 4.815389159651455e-05,
"loss": 1.5054,
"step": 5000
},
{
"epoch": 0.12,
"learning_rate": 4.7969280756166004e-05,
"loss": 1.4937,
"step": 5500
},
{
"epoch": 0.13,
"learning_rate": 4.778466991581746e-05,
"loss": 1.4762,
"step": 6000
},
{
"epoch": 0.14,
"learning_rate": 4.7600059075468914e-05,
"loss": 1.4681,
"step": 6500
},
{
"epoch": 0.16,
"learning_rate": 4.7415448235120366e-05,
"loss": 1.4336,
"step": 7000
},
{
"epoch": 0.17,
"learning_rate": 4.7230837394771824e-05,
"loss": 1.4241,
"step": 7500
},
{
"epoch": 0.18,
"learning_rate": 4.7046226554423276e-05,
"loss": 1.4174,
"step": 8000
},
{
"epoch": 0.19,
"learning_rate": 4.686161571407473e-05,
"loss": 1.42,
"step": 8500
},
{
"epoch": 0.2,
"learning_rate": 4.6677004873726186e-05,
"loss": 1.3927,
"step": 9000
},
{
"epoch": 0.21,
"learning_rate": 4.649239403337764e-05,
"loss": 1.3908,
"step": 9500
},
{
"epoch": 0.22,
"learning_rate": 4.6307783193029096e-05,
"loss": 1.3729,
"step": 10000
},
{
"epoch": 0.23,
"learning_rate": 4.612317235268055e-05,
"loss": 1.3742,
"step": 10500
},
{
"epoch": 0.24,
"learning_rate": 4.5938561512332006e-05,
"loss": 1.3511,
"step": 11000
},
{
"epoch": 0.25,
"learning_rate": 4.575395067198346e-05,
"loss": 1.3396,
"step": 11500
},
{
"epoch": 0.27,
"learning_rate": 4.5569339831634916e-05,
"loss": 1.3579,
"step": 12000
},
{
"epoch": 0.28,
"learning_rate": 4.538472899128637e-05,
"loss": 1.3267,
"step": 12500
},
{
"epoch": 0.29,
"learning_rate": 4.5200118150937826e-05,
"loss": 1.3323,
"step": 13000
},
{
"epoch": 0.3,
"learning_rate": 4.501550731058928e-05,
"loss": 1.3223,
"step": 13500
},
{
"epoch": 0.31,
"learning_rate": 4.4830896470240736e-05,
"loss": 1.3195,
"step": 14000
},
{
"epoch": 0.32,
"learning_rate": 4.464628562989219e-05,
"loss": 1.3044,
"step": 14500
},
{
"epoch": 0.33,
"learning_rate": 4.4461674789543646e-05,
"loss": 1.3149,
"step": 15000
},
{
"epoch": 0.34,
"learning_rate": 4.42770639491951e-05,
"loss": 1.2981,
"step": 15500
},
{
"epoch": 0.35,
"learning_rate": 4.4092453108846556e-05,
"loss": 1.2765,
"step": 16000
},
{
"epoch": 0.37,
"learning_rate": 4.390784226849801e-05,
"loss": 1.28,
"step": 16500
},
{
"epoch": 0.38,
"learning_rate": 4.3723231428149466e-05,
"loss": 1.2679,
"step": 17000
},
{
"epoch": 0.39,
"learning_rate": 4.353862058780092e-05,
"loss": 1.2642,
"step": 17500
},
{
"epoch": 0.4,
"learning_rate": 4.3354009747452376e-05,
"loss": 1.2611,
"step": 18000
},
{
"epoch": 0.41,
"learning_rate": 4.316939890710383e-05,
"loss": 1.2533,
"step": 18500
},
{
"epoch": 0.42,
"learning_rate": 4.2984788066755286e-05,
"loss": 1.2679,
"step": 19000
},
{
"epoch": 0.43,
"learning_rate": 4.280017722640674e-05,
"loss": 1.2508,
"step": 19500
},
{
"epoch": 0.44,
"learning_rate": 4.261556638605819e-05,
"loss": 1.2394,
"step": 20000
},
{
"epoch": 0.45,
"learning_rate": 4.243095554570964e-05,
"loss": 1.2401,
"step": 20500
},
{
"epoch": 0.47,
"learning_rate": 4.22463447053611e-05,
"loss": 1.2592,
"step": 21000
},
{
"epoch": 0.48,
"learning_rate": 4.206173386501255e-05,
"loss": 1.2119,
"step": 21500
},
{
"epoch": 0.49,
"learning_rate": 4.187712302466401e-05,
"loss": 1.2365,
"step": 22000
},
{
"epoch": 0.5,
"learning_rate": 4.169251218431546e-05,
"loss": 1.2346,
"step": 22500
},
{
"epoch": 0.51,
"learning_rate": 4.150790134396692e-05,
"loss": 1.201,
"step": 23000
},
{
"epoch": 0.52,
"learning_rate": 4.132329050361837e-05,
"loss": 1.2217,
"step": 23500
},
{
"epoch": 0.53,
"learning_rate": 4.113867966326983e-05,
"loss": 1.2158,
"step": 24000
},
{
"epoch": 0.54,
"learning_rate": 4.095406882292128e-05,
"loss": 1.2141,
"step": 24500
},
{
"epoch": 0.55,
"learning_rate": 4.076945798257274e-05,
"loss": 1.2004,
"step": 25000
},
{
"epoch": 0.56,
"learning_rate": 4.058484714222419e-05,
"loss": 1.1878,
"step": 25500
},
{
"epoch": 0.58,
"learning_rate": 4.040023630187565e-05,
"loss": 1.2181,
"step": 26000
},
{
"epoch": 0.59,
"learning_rate": 4.02156254615271e-05,
"loss": 1.1957,
"step": 26500
},
{
"epoch": 0.6,
"learning_rate": 4.003101462117856e-05,
"loss": 1.2023,
"step": 27000
},
{
"epoch": 0.61,
"learning_rate": 3.984640378083001e-05,
"loss": 1.1952,
"step": 27500
},
{
"epoch": 0.62,
"learning_rate": 3.966179294048147e-05,
"loss": 1.1945,
"step": 28000
},
{
"epoch": 0.63,
"learning_rate": 3.947718210013292e-05,
"loss": 1.1837,
"step": 28500
},
{
"epoch": 0.64,
"learning_rate": 3.929257125978438e-05,
"loss": 1.1754,
"step": 29000
},
{
"epoch": 0.65,
"learning_rate": 3.910796041943583e-05,
"loss": 1.1882,
"step": 29500
},
{
"epoch": 0.66,
"learning_rate": 3.892334957908729e-05,
"loss": 1.1748,
"step": 30000
},
{
"epoch": 0.68,
"learning_rate": 3.873873873873874e-05,
"loss": 1.1479,
"step": 30500
},
{
"epoch": 0.69,
"learning_rate": 3.85541278983902e-05,
"loss": 1.1779,
"step": 31000
},
{
"epoch": 0.7,
"learning_rate": 3.836951705804165e-05,
"loss": 1.1589,
"step": 31500
},
{
"epoch": 0.71,
"learning_rate": 3.818490621769311e-05,
"loss": 1.1643,
"step": 32000
},
{
"epoch": 0.72,
"learning_rate": 3.800029537734456e-05,
"loss": 1.1636,
"step": 32500
},
{
"epoch": 0.73,
"learning_rate": 3.781568453699601e-05,
"loss": 1.1633,
"step": 33000
},
{
"epoch": 0.74,
"learning_rate": 3.7631073696647464e-05,
"loss": 1.1588,
"step": 33500
},
{
"epoch": 0.75,
"learning_rate": 3.744646285629892e-05,
"loss": 1.1486,
"step": 34000
},
{
"epoch": 0.76,
"learning_rate": 3.7261852015950374e-05,
"loss": 1.1578,
"step": 34500
},
{
"epoch": 0.78,
"learning_rate": 3.707724117560183e-05,
"loss": 1.1462,
"step": 35000
},
{
"epoch": 0.79,
"learning_rate": 3.6892630335253284e-05,
"loss": 1.1571,
"step": 35500
},
{
"epoch": 0.8,
"learning_rate": 3.670801949490474e-05,
"loss": 1.1379,
"step": 36000
},
{
"epoch": 0.81,
"learning_rate": 3.6523408654556194e-05,
"loss": 1.1421,
"step": 36500
},
{
"epoch": 0.82,
"learning_rate": 3.633879781420765e-05,
"loss": 1.1423,
"step": 37000
},
{
"epoch": 0.83,
"learning_rate": 3.6154186973859104e-05,
"loss": 1.1377,
"step": 37500
},
{
"epoch": 0.84,
"learning_rate": 3.596957613351056e-05,
"loss": 1.1335,
"step": 38000
},
{
"epoch": 0.85,
"learning_rate": 3.5784965293162014e-05,
"loss": 1.128,
"step": 38500
},
{
"epoch": 0.86,
"learning_rate": 3.560035445281347e-05,
"loss": 1.1401,
"step": 39000
},
{
"epoch": 0.88,
"learning_rate": 3.5415743612464924e-05,
"loss": 1.1311,
"step": 39500
},
{
"epoch": 0.89,
"learning_rate": 3.523113277211638e-05,
"loss": 1.1107,
"step": 40000
},
{
"epoch": 0.9,
"learning_rate": 3.5046521931767834e-05,
"loss": 1.1295,
"step": 40500
},
{
"epoch": 0.91,
"learning_rate": 3.486191109141929e-05,
"loss": 1.1126,
"step": 41000
},
{
"epoch": 0.92,
"learning_rate": 3.4677300251070744e-05,
"loss": 1.1161,
"step": 41500
},
{
"epoch": 0.93,
"learning_rate": 3.44926894107222e-05,
"loss": 1.1058,
"step": 42000
},
{
"epoch": 0.94,
"learning_rate": 3.4308078570373654e-05,
"loss": 1.1165,
"step": 42500
},
{
"epoch": 0.95,
"learning_rate": 3.412346773002511e-05,
"loss": 1.1174,
"step": 43000
},
{
"epoch": 0.96,
"learning_rate": 3.3938856889676564e-05,
"loss": 1.1234,
"step": 43500
},
{
"epoch": 0.97,
"learning_rate": 3.375424604932802e-05,
"loss": 1.1048,
"step": 44000
},
{
"epoch": 0.99,
"learning_rate": 3.3569635208979474e-05,
"loss": 1.1074,
"step": 44500
},
{
"epoch": 1.0,
"learning_rate": 3.3385024368630926e-05,
"loss": 1.1096,
"step": 45000
},
{
"epoch": 1.01,
"learning_rate": 3.3200413528282384e-05,
"loss": 1.0264,
"step": 45500
},
{
"epoch": 1.02,
"learning_rate": 3.3015802687933836e-05,
"loss": 1.004,
"step": 46000
},
{
"epoch": 1.03,
"learning_rate": 3.283119184758529e-05,
"loss": 1.0024,
"step": 46500
},
{
"epoch": 1.04,
"learning_rate": 3.2646581007236746e-05,
"loss": 1.0019,
"step": 47000
},
{
"epoch": 1.05,
"learning_rate": 3.24619701668882e-05,
"loss": 1.0017,
"step": 47500
},
{
"epoch": 1.06,
"learning_rate": 3.2277359326539656e-05,
"loss": 1.0006,
"step": 48000
},
{
"epoch": 1.07,
"learning_rate": 3.209274848619111e-05,
"loss": 1.0011,
"step": 48500
},
{
"epoch": 1.09,
"learning_rate": 3.1908137645842566e-05,
"loss": 0.9957,
"step": 49000
},
{
"epoch": 1.1,
"learning_rate": 3.172352680549402e-05,
"loss": 0.9965,
"step": 49500
},
{
"epoch": 1.11,
"learning_rate": 3.1538915965145476e-05,
"loss": 0.9937,
"step": 50000
},
{
"epoch": 1.12,
"learning_rate": 3.135430512479693e-05,
"loss": 0.9965,
"step": 50500
},
{
"epoch": 1.13,
"learning_rate": 3.1169694284448386e-05,
"loss": 0.9894,
"step": 51000
},
{
"epoch": 1.14,
"learning_rate": 3.098508344409984e-05,
"loss": 1.0062,
"step": 51500
},
{
"epoch": 1.15,
"learning_rate": 3.0800472603751296e-05,
"loss": 0.998,
"step": 52000
},
{
"epoch": 1.16,
"learning_rate": 3.061586176340275e-05,
"loss": 1.006,
"step": 52500
},
{
"epoch": 1.17,
"learning_rate": 3.0431250923054206e-05,
"loss": 0.9896,
"step": 53000
},
{
"epoch": 1.19,
"learning_rate": 3.0246640082705658e-05,
"loss": 0.9818,
"step": 53500
},
{
"epoch": 1.2,
"learning_rate": 3.0062029242357116e-05,
"loss": 0.9861,
"step": 54000
},
{
"epoch": 1.21,
"learning_rate": 2.9877418402008568e-05,
"loss": 0.9894,
"step": 54500
},
{
"epoch": 1.22,
"learning_rate": 2.9692807561660023e-05,
"loss": 0.9931,
"step": 55000
},
{
"epoch": 1.23,
"learning_rate": 2.9508196721311478e-05,
"loss": 0.9847,
"step": 55500
},
{
"epoch": 1.24,
"learning_rate": 2.9323585880962933e-05,
"loss": 0.982,
"step": 56000
},
{
"epoch": 1.25,
"learning_rate": 2.9138975040614384e-05,
"loss": 0.9919,
"step": 56500
},
{
"epoch": 1.26,
"learning_rate": 2.8954364200265843e-05,
"loss": 0.9575,
"step": 57000
},
{
"epoch": 1.27,
"learning_rate": 2.8769753359917294e-05,
"loss": 0.9857,
"step": 57500
},
{
"epoch": 1.28,
"learning_rate": 2.8585142519568753e-05,
"loss": 0.9926,
"step": 58000
},
{
"epoch": 1.3,
"learning_rate": 2.8400531679220204e-05,
"loss": 0.9974,
"step": 58500
},
{
"epoch": 1.31,
"learning_rate": 2.8215920838871663e-05,
"loss": 0.9904,
"step": 59000
},
{
"epoch": 1.32,
"learning_rate": 2.8031309998523115e-05,
"loss": 0.9795,
"step": 59500
},
{
"epoch": 1.33,
"learning_rate": 2.7846699158174573e-05,
"loss": 0.9782,
"step": 60000
},
{
"epoch": 1.34,
"learning_rate": 2.7662088317826025e-05,
"loss": 0.9867,
"step": 60500
},
{
"epoch": 1.35,
"learning_rate": 2.7477477477477483e-05,
"loss": 0.9693,
"step": 61000
},
{
"epoch": 1.36,
"learning_rate": 2.7292866637128935e-05,
"loss": 0.9734,
"step": 61500
},
{
"epoch": 1.37,
"learning_rate": 2.710825579678039e-05,
"loss": 0.9769,
"step": 62000
},
{
"epoch": 1.38,
"learning_rate": 2.692364495643184e-05,
"loss": 0.9698,
"step": 62500
},
{
"epoch": 1.4,
"learning_rate": 2.67390341160833e-05,
"loss": 0.967,
"step": 63000
},
{
"epoch": 1.41,
"learning_rate": 2.655442327573475e-05,
"loss": 0.9702,
"step": 63500
},
{
"epoch": 1.42,
"learning_rate": 2.636981243538621e-05,
"loss": 0.9626,
"step": 64000
},
{
"epoch": 1.43,
"learning_rate": 2.618520159503766e-05,
"loss": 0.9773,
"step": 64500
},
{
"epoch": 1.44,
"learning_rate": 2.600059075468912e-05,
"loss": 0.9641,
"step": 65000
},
{
"epoch": 1.45,
"learning_rate": 2.581597991434057e-05,
"loss": 0.9649,
"step": 65500
},
{
"epoch": 1.46,
"learning_rate": 2.563136907399203e-05,
"loss": 0.9756,
"step": 66000
},
{
"epoch": 1.47,
"learning_rate": 2.544675823364348e-05,
"loss": 0.9709,
"step": 66500
},
{
"epoch": 1.48,
"learning_rate": 2.526214739329494e-05,
"loss": 0.9645,
"step": 67000
},
{
"epoch": 1.5,
"learning_rate": 2.507753655294639e-05,
"loss": 0.9836,
"step": 67500
},
{
"epoch": 1.51,
"learning_rate": 2.4892925712597846e-05,
"loss": 0.9709,
"step": 68000
},
{
"epoch": 1.52,
"learning_rate": 2.47083148722493e-05,
"loss": 0.9866,
"step": 68500
},
{
"epoch": 1.53,
"learning_rate": 2.4523704031900753e-05,
"loss": 0.965,
"step": 69000
},
{
"epoch": 1.54,
"learning_rate": 2.4339093191552208e-05,
"loss": 0.954,
"step": 69500
},
{
"epoch": 1.55,
"learning_rate": 2.4154482351203663e-05,
"loss": 0.9618,
"step": 70000
},
{
"epoch": 1.56,
"learning_rate": 2.3969871510855118e-05,
"loss": 0.953,
"step": 70500
},
{
"epoch": 1.57,
"learning_rate": 2.3785260670506573e-05,
"loss": 0.958,
"step": 71000
},
{
"epoch": 1.58,
"learning_rate": 2.3600649830158028e-05,
"loss": 0.9615,
"step": 71500
},
{
"epoch": 1.6,
"learning_rate": 2.3416038989809483e-05,
"loss": 0.9483,
"step": 72000
},
{
"epoch": 1.61,
"learning_rate": 2.3231428149460938e-05,
"loss": 0.943,
"step": 72500
},
{
"epoch": 1.62,
"learning_rate": 2.3046817309112393e-05,
"loss": 0.9649,
"step": 73000
},
{
"epoch": 1.63,
"learning_rate": 2.2862206468763848e-05,
"loss": 0.9615,
"step": 73500
},
{
"epoch": 1.64,
"learning_rate": 2.2677595628415303e-05,
"loss": 0.9523,
"step": 74000
},
{
"epoch": 1.65,
"learning_rate": 2.2492984788066758e-05,
"loss": 0.9596,
"step": 74500
},
{
"epoch": 1.66,
"learning_rate": 2.2308373947718213e-05,
"loss": 0.958,
"step": 75000
},
{
"epoch": 1.67,
"learning_rate": 2.2123763107369665e-05,
"loss": 0.9438,
"step": 75500
},
{
"epoch": 1.68,
"learning_rate": 2.193915226702112e-05,
"loss": 0.9594,
"step": 76000
},
{
"epoch": 1.69,
"learning_rate": 2.1754541426672575e-05,
"loss": 0.9671,
"step": 76500
},
{
"epoch": 1.71,
"learning_rate": 2.156993058632403e-05,
"loss": 0.9495,
"step": 77000
},
{
"epoch": 1.72,
"learning_rate": 2.1385319745975485e-05,
"loss": 0.9601,
"step": 77500
},
{
"epoch": 1.73,
"learning_rate": 2.120070890562694e-05,
"loss": 0.9369,
"step": 78000
},
{
"epoch": 1.74,
"learning_rate": 2.1016098065278395e-05,
"loss": 0.9433,
"step": 78500
},
{
"epoch": 1.75,
"learning_rate": 2.083148722492985e-05,
"loss": 0.9548,
"step": 79000
},
{
"epoch": 1.76,
"learning_rate": 2.0646876384581305e-05,
"loss": 0.9358,
"step": 79500
},
{
"epoch": 1.77,
"learning_rate": 2.046226554423276e-05,
"loss": 0.9431,
"step": 80000
},
{
"epoch": 1.78,
"learning_rate": 2.0277654703884215e-05,
"loss": 0.9318,
"step": 80500
},
{
"epoch": 1.79,
"learning_rate": 2.009304386353567e-05,
"loss": 0.9319,
"step": 81000
},
{
"epoch": 1.81,
"learning_rate": 1.990843302318712e-05,
"loss": 0.9415,
"step": 81500
},
{
"epoch": 1.82,
"learning_rate": 1.9723822182838576e-05,
"loss": 0.9328,
"step": 82000
},
{
"epoch": 1.83,
"learning_rate": 1.953921134249003e-05,
"loss": 0.9291,
"step": 82500
},
{
"epoch": 1.84,
"learning_rate": 1.9354600502141486e-05,
"loss": 0.957,
"step": 83000
},
{
"epoch": 1.85,
"learning_rate": 1.916998966179294e-05,
"loss": 0.9336,
"step": 83500
},
{
"epoch": 1.86,
"learning_rate": 1.8985378821444396e-05,
"loss": 0.9365,
"step": 84000
},
{
"epoch": 1.87,
"learning_rate": 1.880076798109585e-05,
"loss": 0.9335,
"step": 84500
},
{
"epoch": 1.88,
"learning_rate": 1.8616157140747306e-05,
"loss": 0.9275,
"step": 85000
},
{
"epoch": 1.89,
"learning_rate": 1.843154630039876e-05,
"loss": 0.927,
"step": 85500
},
{
"epoch": 1.91,
"learning_rate": 1.8246935460050216e-05,
"loss": 0.9325,
"step": 86000
},
{
"epoch": 1.92,
"learning_rate": 1.806232461970167e-05,
"loss": 0.925,
"step": 86500
},
{
"epoch": 1.93,
"learning_rate": 1.7877713779353126e-05,
"loss": 0.912,
"step": 87000
},
{
"epoch": 1.94,
"learning_rate": 1.769310293900458e-05,
"loss": 0.9082,
"step": 87500
},
{
"epoch": 1.95,
"learning_rate": 1.7508492098656033e-05,
"loss": 0.9339,
"step": 88000
},
{
"epoch": 1.96,
"learning_rate": 1.7323881258307488e-05,
"loss": 0.9345,
"step": 88500
},
{
"epoch": 1.97,
"learning_rate": 1.7139270417958943e-05,
"loss": 0.9295,
"step": 89000
},
{
"epoch": 1.98,
"learning_rate": 1.6954659577610398e-05,
"loss": 0.908,
"step": 89500
},
{
"epoch": 1.99,
"learning_rate": 1.6770048737261853e-05,
"loss": 0.9205,
"step": 90000
},
{
"epoch": 2.0,
"learning_rate": 1.6585437896913308e-05,
"loss": 0.8905,
"step": 90500
},
{
"epoch": 2.02,
"learning_rate": 1.6400827056564763e-05,
"loss": 0.8215,
"step": 91000
},
{
"epoch": 2.03,
"learning_rate": 1.6216216216216218e-05,
"loss": 0.8212,
"step": 91500
},
{
"epoch": 2.04,
"learning_rate": 1.6031605375867673e-05,
"loss": 0.818,
"step": 92000
},
{
"epoch": 2.05,
"learning_rate": 1.5846994535519128e-05,
"loss": 0.8175,
"step": 92500
},
{
"epoch": 2.06,
"learning_rate": 1.5662383695170583e-05,
"loss": 0.8372,
"step": 93000
},
{
"epoch": 2.07,
"learning_rate": 1.5477772854822038e-05,
"loss": 0.8227,
"step": 93500
},
{
"epoch": 2.08,
"learning_rate": 1.5293162014473493e-05,
"loss": 0.8239,
"step": 94000
},
{
"epoch": 2.09,
"learning_rate": 1.5108551174124946e-05,
"loss": 0.8244,
"step": 94500
},
{
"epoch": 2.1,
"learning_rate": 1.4923940333776401e-05,
"loss": 0.8349,
"step": 95000
},
{
"epoch": 2.12,
"learning_rate": 1.4739329493427856e-05,
"loss": 0.8196,
"step": 95500
},
{
"epoch": 2.13,
"learning_rate": 1.455471865307931e-05,
"loss": 0.8313,
"step": 96000
},
{
"epoch": 2.14,
"learning_rate": 1.4370107812730765e-05,
"loss": 0.824,
"step": 96500
},
{
"epoch": 2.15,
"learning_rate": 1.418549697238222e-05,
"loss": 0.8168,
"step": 97000
},
{
"epoch": 2.16,
"learning_rate": 1.4000886132033675e-05,
"loss": 0.8297,
"step": 97500
},
{
"epoch": 2.17,
"learning_rate": 1.381627529168513e-05,
"loss": 0.8143,
"step": 98000
},
{
"epoch": 2.18,
"learning_rate": 1.3631664451336585e-05,
"loss": 0.815,
"step": 98500
},
{
"epoch": 2.19,
"learning_rate": 1.344705361098804e-05,
"loss": 0.8245,
"step": 99000
},
{
"epoch": 2.2,
"learning_rate": 1.3262442770639493e-05,
"loss": 0.8204,
"step": 99500
},
{
"epoch": 2.22,
"learning_rate": 1.3077831930290948e-05,
"loss": 0.8103,
"step": 100000
},
{
"epoch": 2.23,
"learning_rate": 1.2893221089942403e-05,
"loss": 0.8161,
"step": 100500
},
{
"epoch": 2.24,
"learning_rate": 1.2708610249593858e-05,
"loss": 0.8201,
"step": 101000
},
{
"epoch": 2.25,
"learning_rate": 1.2523999409245313e-05,
"loss": 0.8101,
"step": 101500
},
{
"epoch": 2.26,
"learning_rate": 1.2339388568896766e-05,
"loss": 0.8275,
"step": 102000
},
{
"epoch": 2.27,
"learning_rate": 1.2154777728548221e-05,
"loss": 0.8231,
"step": 102500
},
{
"epoch": 2.28,
"learning_rate": 1.1970166888199675e-05,
"loss": 0.8271,
"step": 103000
},
{
"epoch": 2.29,
"learning_rate": 1.178555604785113e-05,
"loss": 0.8237,
"step": 103500
},
{
"epoch": 2.3,
"learning_rate": 1.1600945207502585e-05,
"loss": 0.8142,
"step": 104000
},
{
"epoch": 2.32,
"learning_rate": 1.141633436715404e-05,
"loss": 0.8055,
"step": 104500
},
{
"epoch": 2.33,
"learning_rate": 1.1231723526805495e-05,
"loss": 0.8083,
"step": 105000
},
{
"epoch": 2.34,
"learning_rate": 1.104711268645695e-05,
"loss": 0.8079,
"step": 105500
},
{
"epoch": 2.35,
"learning_rate": 1.0862501846108405e-05,
"loss": 0.8091,
"step": 106000
},
{
"epoch": 2.36,
"learning_rate": 1.0677891005759858e-05,
"loss": 0.8191,
"step": 106500
},
{
"epoch": 2.37,
"learning_rate": 1.0493280165411313e-05,
"loss": 0.8086,
"step": 107000
},
{
"epoch": 2.38,
"learning_rate": 1.0308669325062768e-05,
"loss": 0.8108,
"step": 107500
},
{
"epoch": 2.39,
"learning_rate": 1.0124058484714223e-05,
"loss": 0.8091,
"step": 108000
},
{
"epoch": 2.4,
"learning_rate": 9.939447644365678e-06,
"loss": 0.8117,
"step": 108500
},
{
"epoch": 2.41,
"learning_rate": 9.754836804017133e-06,
"loss": 0.8033,
"step": 109000
},
{
"epoch": 2.43,
"learning_rate": 9.570225963668587e-06,
"loss": 0.8002,
"step": 109500
},
{
"epoch": 2.44,
"learning_rate": 9.385615123320042e-06,
"loss": 0.8096,
"step": 110000
},
{
"epoch": 2.45,
"learning_rate": 9.201004282971497e-06,
"loss": 0.8101,
"step": 110500
},
{
"epoch": 2.46,
"learning_rate": 9.016393442622952e-06,
"loss": 0.8114,
"step": 111000
},
{
"epoch": 2.47,
"learning_rate": 8.831782602274407e-06,
"loss": 0.8058,
"step": 111500
},
{
"epoch": 2.48,
"learning_rate": 8.647171761925862e-06,
"loss": 0.8021,
"step": 112000
},
{
"epoch": 2.49,
"learning_rate": 8.462560921577317e-06,
"loss": 0.7998,
"step": 112500
},
{
"epoch": 2.5,
"learning_rate": 8.27795008122877e-06,
"loss": 0.8056,
"step": 113000
},
{
"epoch": 2.51,
"learning_rate": 8.093339240880225e-06,
"loss": 0.809,
"step": 113500
},
{
"epoch": 2.53,
"learning_rate": 7.90872840053168e-06,
"loss": 0.8068,
"step": 114000
},
{
"epoch": 2.54,
"learning_rate": 7.724117560183135e-06,
"loss": 0.7999,
"step": 114500
},
{
"epoch": 2.55,
"learning_rate": 7.539506719834589e-06,
"loss": 0.8014,
"step": 115000
},
{
"epoch": 2.56,
"learning_rate": 7.354895879486044e-06,
"loss": 0.7997,
"step": 115500
},
{
"epoch": 2.57,
"learning_rate": 7.170285039137499e-06,
"loss": 0.8028,
"step": 116000
},
{
"epoch": 2.58,
"learning_rate": 6.985674198788954e-06,
"loss": 0.8104,
"step": 116500
},
{
"epoch": 2.59,
"learning_rate": 6.801063358440408e-06,
"loss": 0.7978,
"step": 117000
},
{
"epoch": 2.6,
"learning_rate": 6.616452518091863e-06,
"loss": 0.7932,
"step": 117500
},
{
"epoch": 2.61,
"learning_rate": 6.431841677743318e-06,
"loss": 0.799,
"step": 118000
},
{
"epoch": 2.63,
"learning_rate": 6.2472308373947724e-06,
"loss": 0.7945,
"step": 118500
},
{
"epoch": 2.64,
"learning_rate": 6.062619997046227e-06,
"loss": 0.7987,
"step": 119000
},
{
"epoch": 2.65,
"learning_rate": 5.878009156697682e-06,
"loss": 0.7974,
"step": 119500
},
{
"epoch": 2.66,
"learning_rate": 5.693398316349137e-06,
"loss": 0.8064,
"step": 120000
},
{
"epoch": 2.67,
"learning_rate": 5.508787476000591e-06,
"loss": 0.7996,
"step": 120500
},
{
"epoch": 2.68,
"learning_rate": 5.324176635652046e-06,
"loss": 0.795,
"step": 121000
},
{
"epoch": 2.69,
"learning_rate": 5.139565795303501e-06,
"loss": 0.7941,
"step": 121500
},
{
"epoch": 2.7,
"learning_rate": 4.954954954954955e-06,
"loss": 0.8007,
"step": 122000
},
{
"epoch": 2.71,
"learning_rate": 4.77034411460641e-06,
"loss": 0.8166,
"step": 122500
},
{
"epoch": 2.72,
"learning_rate": 4.585733274257865e-06,
"loss": 0.801,
"step": 123000
},
{
"epoch": 2.74,
"learning_rate": 4.401122433909319e-06,
"loss": 0.7959,
"step": 123500
},
{
"epoch": 2.75,
"learning_rate": 4.216511593560774e-06,
"loss": 0.7991,
"step": 124000
},
{
"epoch": 2.76,
"learning_rate": 4.031900753212229e-06,
"loss": 0.7994,
"step": 124500
},
{
"epoch": 2.77,
"learning_rate": 3.847289912863683e-06,
"loss": 0.7976,
"step": 125000
},
{
"epoch": 2.78,
"learning_rate": 3.6626790725151383e-06,
"loss": 0.805,
"step": 125500
},
{
"epoch": 2.79,
"learning_rate": 3.4780682321665933e-06,
"loss": 0.8001,
"step": 126000
},
{
"epoch": 2.8,
"learning_rate": 3.293457391818048e-06,
"loss": 0.8004,
"step": 126500
},
{
"epoch": 2.81,
"learning_rate": 3.1088465514695025e-06,
"loss": 0.791,
"step": 127000
},
{
"epoch": 2.82,
"learning_rate": 2.924235711120957e-06,
"loss": 0.7863,
"step": 127500
},
{
"epoch": 2.84,
"learning_rate": 2.7396248707724117e-06,
"loss": 0.7936,
"step": 128000
},
{
"epoch": 2.85,
"learning_rate": 2.5550140304238667e-06,
"loss": 0.7943,
"step": 128500
},
{
"epoch": 2.86,
"learning_rate": 2.3704031900753213e-06,
"loss": 0.7987,
"step": 129000
},
{
"epoch": 2.87,
"learning_rate": 2.185792349726776e-06,
"loss": 0.7963,
"step": 129500
},
{
"epoch": 2.88,
"learning_rate": 2.001181509378231e-06,
"loss": 0.7948,
"step": 130000
},
{
"epoch": 2.89,
"learning_rate": 1.8165706690296854e-06,
"loss": 0.7938,
"step": 130500
},
{
"epoch": 2.9,
"learning_rate": 1.6319598286811402e-06,
"loss": 0.7924,
"step": 131000
},
{
"epoch": 2.91,
"learning_rate": 1.447348988332595e-06,
"loss": 0.7914,
"step": 131500
},
{
"epoch": 2.92,
"learning_rate": 1.2627381479840496e-06,
"loss": 0.797,
"step": 132000
},
{
"epoch": 2.94,
"learning_rate": 1.0781273076355044e-06,
"loss": 0.7809,
"step": 132500
},
{
"epoch": 2.95,
"learning_rate": 8.935164672869591e-07,
"loss": 0.7967,
"step": 133000
},
{
"epoch": 2.96,
"learning_rate": 7.089056269384138e-07,
"loss": 0.7986,
"step": 133500
},
{
"epoch": 2.97,
"learning_rate": 5.242947865898686e-07,
"loss": 0.7883,
"step": 134000
},
{
"epoch": 2.98,
"learning_rate": 3.396839462413233e-07,
"loss": 0.7814,
"step": 134500
},
{
"epoch": 2.99,
"learning_rate": 1.5507310589277803e-07,
"loss": 0.7805,
"step": 135000
},
{
"epoch": 3.0,
"step": 135420,
"total_flos": 4.1954280130740224e+17,
"train_loss": 1.017597695056136,
"train_runtime": 46985.3753,
"train_samples_per_second": 57.643,
"train_steps_per_second": 2.882
}
],
"max_steps": 135420,
"num_train_epochs": 3,
"total_flos": 4.1954280130740224e+17,
"trial_name": null,
"trial_params": null
}