afro-xlmr-large-76L_script / trainer_state.json
Davlan's picture
Upload 12 files
798d533 verified
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 2.9999330508717734,
"eval_steps": 500,
"global_step": 117624,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.01,
"learning_rate": 4.9787458341835e-05,
"loss": 4.4684,
"step": 500
},
{
"epoch": 0.03,
"learning_rate": 4.957491668367e-05,
"loss": 3.4213,
"step": 1000
},
{
"epoch": 0.04,
"learning_rate": 4.9362375025505e-05,
"loss": 2.9387,
"step": 1500
},
{
"epoch": 0.05,
"learning_rate": 4.914983336734e-05,
"loss": 2.6616,
"step": 2000
},
{
"epoch": 0.06,
"learning_rate": 4.8937291709175e-05,
"loss": 2.4563,
"step": 2500
},
{
"epoch": 0.08,
"learning_rate": 4.872475005101e-05,
"loss": 2.2849,
"step": 3000
},
{
"epoch": 0.09,
"learning_rate": 4.8512208392845e-05,
"loss": 2.1594,
"step": 3500
},
{
"epoch": 0.1,
"learning_rate": 4.829966673468e-05,
"loss": 2.0474,
"step": 4000
},
{
"epoch": 0.11,
"learning_rate": 4.8087125076515e-05,
"loss": 1.9558,
"step": 4500
},
{
"epoch": 0.13,
"learning_rate": 4.787458341835e-05,
"loss": 1.877,
"step": 5000
},
{
"epoch": 0.14,
"learning_rate": 4.7662041760185e-05,
"loss": 1.8105,
"step": 5500
},
{
"epoch": 0.15,
"learning_rate": 4.744950010202e-05,
"loss": 1.7536,
"step": 6000
},
{
"epoch": 0.17,
"learning_rate": 4.7236958443855e-05,
"loss": 1.7081,
"step": 6500
},
{
"epoch": 0.18,
"learning_rate": 4.7024416785689995e-05,
"loss": 1.6638,
"step": 7000
},
{
"epoch": 0.19,
"learning_rate": 4.6811875127524994e-05,
"loss": 1.6277,
"step": 7500
},
{
"epoch": 0.2,
"learning_rate": 4.6599333469359994e-05,
"loss": 1.5962,
"step": 8000
},
{
"epoch": 0.22,
"learning_rate": 4.638679181119499e-05,
"loss": 1.5655,
"step": 8500
},
{
"epoch": 0.23,
"learning_rate": 4.617425015303e-05,
"loss": 1.54,
"step": 9000
},
{
"epoch": 0.24,
"learning_rate": 4.5961708494865e-05,
"loss": 1.5168,
"step": 9500
},
{
"epoch": 0.26,
"learning_rate": 4.57491668367e-05,
"loss": 1.4923,
"step": 10000
},
{
"epoch": 0.27,
"learning_rate": 4.5536625178535e-05,
"loss": 1.475,
"step": 10500
},
{
"epoch": 0.28,
"learning_rate": 4.532408352036999e-05,
"loss": 1.4551,
"step": 11000
},
{
"epoch": 0.29,
"learning_rate": 4.511154186220499e-05,
"loss": 1.4404,
"step": 11500
},
{
"epoch": 0.31,
"learning_rate": 4.489900020403999e-05,
"loss": 1.4239,
"step": 12000
},
{
"epoch": 0.32,
"learning_rate": 4.4686458545874996e-05,
"loss": 1.4138,
"step": 12500
},
{
"epoch": 0.33,
"learning_rate": 4.4473916887709995e-05,
"loss": 1.3989,
"step": 13000
},
{
"epoch": 0.34,
"learning_rate": 4.4261375229544995e-05,
"loss": 1.3857,
"step": 13500
},
{
"epoch": 0.36,
"learning_rate": 4.4048833571379994e-05,
"loss": 1.3811,
"step": 14000
},
{
"epoch": 0.37,
"learning_rate": 4.3836291913214994e-05,
"loss": 1.3652,
"step": 14500
},
{
"epoch": 0.38,
"learning_rate": 4.362375025504999e-05,
"loss": 1.3575,
"step": 15000
},
{
"epoch": 0.4,
"learning_rate": 4.3411208596884986e-05,
"loss": 1.3437,
"step": 15500
},
{
"epoch": 0.41,
"learning_rate": 4.319866693871999e-05,
"loss": 1.3325,
"step": 16000
},
{
"epoch": 0.42,
"learning_rate": 4.298612528055499e-05,
"loss": 1.3273,
"step": 16500
},
{
"epoch": 0.43,
"learning_rate": 4.277358362238999e-05,
"loss": 1.3199,
"step": 17000
},
{
"epoch": 0.45,
"learning_rate": 4.256104196422499e-05,
"loss": 1.3086,
"step": 17500
},
{
"epoch": 0.46,
"learning_rate": 4.234850030605999e-05,
"loss": 1.2987,
"step": 18000
},
{
"epoch": 0.47,
"learning_rate": 4.213595864789499e-05,
"loss": 1.2919,
"step": 18500
},
{
"epoch": 0.48,
"learning_rate": 4.192341698972999e-05,
"loss": 1.2866,
"step": 19000
},
{
"epoch": 0.5,
"learning_rate": 4.171087533156499e-05,
"loss": 1.2785,
"step": 19500
},
{
"epoch": 0.51,
"learning_rate": 4.149833367339999e-05,
"loss": 1.2691,
"step": 20000
},
{
"epoch": 0.52,
"learning_rate": 4.128579201523499e-05,
"loss": 1.2641,
"step": 20500
},
{
"epoch": 0.54,
"learning_rate": 4.1073250357069987e-05,
"loss": 1.2592,
"step": 21000
},
{
"epoch": 0.55,
"learning_rate": 4.0860708698904986e-05,
"loss": 1.2557,
"step": 21500
},
{
"epoch": 0.56,
"learning_rate": 4.0648167040739986e-05,
"loss": 1.2464,
"step": 22000
},
{
"epoch": 0.57,
"learning_rate": 4.0435625382574985e-05,
"loss": 1.2397,
"step": 22500
},
{
"epoch": 0.59,
"learning_rate": 4.022308372440999e-05,
"loss": 1.2366,
"step": 23000
},
{
"epoch": 0.6,
"learning_rate": 4.001054206624499e-05,
"loss": 1.2306,
"step": 23500
},
{
"epoch": 0.61,
"learning_rate": 3.979800040807998e-05,
"loss": 1.2231,
"step": 24000
},
{
"epoch": 0.62,
"learning_rate": 3.958545874991498e-05,
"loss": 1.2238,
"step": 24500
},
{
"epoch": 0.64,
"learning_rate": 3.937291709174998e-05,
"loss": 1.2159,
"step": 25000
},
{
"epoch": 0.65,
"learning_rate": 3.916037543358498e-05,
"loss": 1.213,
"step": 25500
},
{
"epoch": 0.66,
"learning_rate": 3.894783377541998e-05,
"loss": 1.2077,
"step": 26000
},
{
"epoch": 0.68,
"learning_rate": 3.873529211725499e-05,
"loss": 1.1987,
"step": 26500
},
{
"epoch": 0.69,
"learning_rate": 3.852275045908999e-05,
"loss": 1.1973,
"step": 27000
},
{
"epoch": 0.7,
"learning_rate": 3.8310208800924986e-05,
"loss": 1.1994,
"step": 27500
},
{
"epoch": 0.71,
"learning_rate": 3.8097667142759986e-05,
"loss": 1.1927,
"step": 28000
},
{
"epoch": 0.73,
"learning_rate": 3.788512548459498e-05,
"loss": 1.1876,
"step": 28500
},
{
"epoch": 0.74,
"learning_rate": 3.767258382642998e-05,
"loss": 1.1826,
"step": 29000
},
{
"epoch": 0.75,
"learning_rate": 3.746004216826498e-05,
"loss": 1.182,
"step": 29500
},
{
"epoch": 0.77,
"learning_rate": 3.7247500510099984e-05,
"loss": 1.1743,
"step": 30000
},
{
"epoch": 0.78,
"learning_rate": 3.703495885193498e-05,
"loss": 1.17,
"step": 30500
},
{
"epoch": 0.79,
"learning_rate": 3.682241719376998e-05,
"loss": 1.1697,
"step": 31000
},
{
"epoch": 0.8,
"learning_rate": 3.660987553560498e-05,
"loss": 1.1643,
"step": 31500
},
{
"epoch": 0.82,
"learning_rate": 3.639733387743998e-05,
"loss": 1.1603,
"step": 32000
},
{
"epoch": 0.83,
"learning_rate": 3.618479221927498e-05,
"loss": 1.1545,
"step": 32500
},
{
"epoch": 0.84,
"learning_rate": 3.5972250561109974e-05,
"loss": 1.1541,
"step": 33000
},
{
"epoch": 0.85,
"learning_rate": 3.575970890294498e-05,
"loss": 1.1526,
"step": 33500
},
{
"epoch": 0.87,
"learning_rate": 3.554716724477998e-05,
"loss": 1.1493,
"step": 34000
},
{
"epoch": 0.88,
"learning_rate": 3.533462558661498e-05,
"loss": 1.1445,
"step": 34500
},
{
"epoch": 0.89,
"learning_rate": 3.512208392844998e-05,
"loss": 1.1448,
"step": 35000
},
{
"epoch": 0.91,
"learning_rate": 3.490954227028498e-05,
"loss": 1.1402,
"step": 35500
},
{
"epoch": 0.92,
"learning_rate": 3.469700061211998e-05,
"loss": 1.1383,
"step": 36000
},
{
"epoch": 0.93,
"learning_rate": 3.448445895395498e-05,
"loss": 1.1321,
"step": 36500
},
{
"epoch": 0.94,
"learning_rate": 3.4271917295789976e-05,
"loss": 1.1289,
"step": 37000
},
{
"epoch": 0.96,
"learning_rate": 3.4059375637624976e-05,
"loss": 1.1267,
"step": 37500
},
{
"epoch": 0.97,
"learning_rate": 3.3846833979459975e-05,
"loss": 1.1277,
"step": 38000
},
{
"epoch": 0.98,
"learning_rate": 3.3634292321294975e-05,
"loss": 1.1262,
"step": 38500
},
{
"epoch": 0.99,
"learning_rate": 3.3421750663129974e-05,
"loss": 1.1191,
"step": 39000
},
{
"epoch": 1.01,
"learning_rate": 3.3209209004964973e-05,
"loss": 1.1184,
"step": 39500
},
{
"epoch": 1.02,
"learning_rate": 3.299666734679997e-05,
"loss": 1.1128,
"step": 40000
},
{
"epoch": 1.03,
"learning_rate": 3.278412568863497e-05,
"loss": 1.1132,
"step": 40500
},
{
"epoch": 1.05,
"learning_rate": 3.257158403046998e-05,
"loss": 1.1112,
"step": 41000
},
{
"epoch": 1.06,
"learning_rate": 3.235904237230497e-05,
"loss": 1.1079,
"step": 41500
},
{
"epoch": 1.07,
"learning_rate": 3.214650071413997e-05,
"loss": 1.104,
"step": 42000
},
{
"epoch": 1.08,
"learning_rate": 3.193395905597497e-05,
"loss": 1.1002,
"step": 42500
},
{
"epoch": 1.1,
"learning_rate": 3.172141739780997e-05,
"loss": 1.1009,
"step": 43000
},
{
"epoch": 1.11,
"learning_rate": 3.150887573964497e-05,
"loss": 1.0992,
"step": 43500
},
{
"epoch": 1.12,
"learning_rate": 3.1296334081479975e-05,
"loss": 1.0965,
"step": 44000
},
{
"epoch": 1.13,
"learning_rate": 3.1083792423314975e-05,
"loss": 1.0962,
"step": 44500
},
{
"epoch": 1.15,
"learning_rate": 3.0871250765149974e-05,
"loss": 1.0914,
"step": 45000
},
{
"epoch": 1.16,
"learning_rate": 3.0658709106984974e-05,
"loss": 1.0928,
"step": 45500
},
{
"epoch": 1.17,
"learning_rate": 3.0446167448819966e-05,
"loss": 1.0898,
"step": 46000
},
{
"epoch": 1.19,
"learning_rate": 3.023362579065497e-05,
"loss": 1.0877,
"step": 46500
},
{
"epoch": 1.2,
"learning_rate": 3.002108413248997e-05,
"loss": 1.0868,
"step": 47000
},
{
"epoch": 1.21,
"learning_rate": 2.9808542474324968e-05,
"loss": 1.0834,
"step": 47500
},
{
"epoch": 1.22,
"learning_rate": 2.9596000816159968e-05,
"loss": 1.0795,
"step": 48000
},
{
"epoch": 1.24,
"learning_rate": 2.938345915799497e-05,
"loss": 1.0769,
"step": 48500
},
{
"epoch": 1.25,
"learning_rate": 2.917091749982997e-05,
"loss": 1.0741,
"step": 49000
},
{
"epoch": 1.26,
"learning_rate": 2.895837584166497e-05,
"loss": 1.0746,
"step": 49500
},
{
"epoch": 1.28,
"learning_rate": 2.874583418349997e-05,
"loss": 1.0726,
"step": 50000
},
{
"epoch": 1.29,
"learning_rate": 2.8533292525334965e-05,
"loss": 1.0739,
"step": 50500
},
{
"epoch": 1.3,
"learning_rate": 2.8320750867169965e-05,
"loss": 1.0666,
"step": 51000
},
{
"epoch": 1.31,
"learning_rate": 2.8108209209004964e-05,
"loss": 1.0668,
"step": 51500
},
{
"epoch": 1.33,
"learning_rate": 2.7895667550839967e-05,
"loss": 1.0654,
"step": 52000
},
{
"epoch": 1.34,
"learning_rate": 2.7683125892674966e-05,
"loss": 1.0652,
"step": 52500
},
{
"epoch": 1.35,
"learning_rate": 2.7470584234509966e-05,
"loss": 1.065,
"step": 53000
},
{
"epoch": 1.36,
"learning_rate": 2.7258042576344965e-05,
"loss": 1.0675,
"step": 53500
},
{
"epoch": 1.38,
"learning_rate": 2.7045500918179968e-05,
"loss": 1.0609,
"step": 54000
},
{
"epoch": 1.39,
"learning_rate": 2.683295926001496e-05,
"loss": 1.0625,
"step": 54500
},
{
"epoch": 1.4,
"learning_rate": 2.662041760184996e-05,
"loss": 1.058,
"step": 55000
},
{
"epoch": 1.42,
"learning_rate": 2.6407875943684963e-05,
"loss": 1.0542,
"step": 55500
},
{
"epoch": 1.43,
"learning_rate": 2.6195334285519963e-05,
"loss": 1.0533,
"step": 56000
},
{
"epoch": 1.44,
"learning_rate": 2.5982792627354962e-05,
"loss": 1.0482,
"step": 56500
},
{
"epoch": 1.45,
"learning_rate": 2.577025096918996e-05,
"loss": 1.052,
"step": 57000
},
{
"epoch": 1.47,
"learning_rate": 2.5557709311024964e-05,
"loss": 1.0514,
"step": 57500
},
{
"epoch": 1.48,
"learning_rate": 2.5345167652859964e-05,
"loss": 1.0481,
"step": 58000
},
{
"epoch": 1.49,
"learning_rate": 2.5132625994694963e-05,
"loss": 1.0474,
"step": 58500
},
{
"epoch": 1.5,
"learning_rate": 2.4920084336529963e-05,
"loss": 1.0426,
"step": 59000
},
{
"epoch": 1.52,
"learning_rate": 2.4707542678364962e-05,
"loss": 1.042,
"step": 59500
},
{
"epoch": 1.53,
"learning_rate": 2.4495001020199958e-05,
"loss": 1.0408,
"step": 60000
},
{
"epoch": 1.54,
"learning_rate": 2.428245936203496e-05,
"loss": 1.043,
"step": 60500
},
{
"epoch": 1.56,
"learning_rate": 2.406991770386996e-05,
"loss": 1.0384,
"step": 61000
},
{
"epoch": 1.57,
"learning_rate": 2.385737604570496e-05,
"loss": 1.0373,
"step": 61500
},
{
"epoch": 1.58,
"learning_rate": 2.364483438753996e-05,
"loss": 1.0343,
"step": 62000
},
{
"epoch": 1.59,
"learning_rate": 2.343229272937496e-05,
"loss": 1.0348,
"step": 62500
},
{
"epoch": 1.61,
"learning_rate": 2.321975107120996e-05,
"loss": 1.0339,
"step": 63000
},
{
"epoch": 1.62,
"learning_rate": 2.3007209413044958e-05,
"loss": 1.0326,
"step": 63500
},
{
"epoch": 1.63,
"learning_rate": 2.2794667754879957e-05,
"loss": 1.0329,
"step": 64000
},
{
"epoch": 1.65,
"learning_rate": 2.2582126096714957e-05,
"loss": 1.0273,
"step": 64500
},
{
"epoch": 1.66,
"learning_rate": 2.2369584438549956e-05,
"loss": 1.027,
"step": 65000
},
{
"epoch": 1.67,
"learning_rate": 2.2157042780384956e-05,
"loss": 1.0304,
"step": 65500
},
{
"epoch": 1.68,
"learning_rate": 2.194450112221996e-05,
"loss": 1.0279,
"step": 66000
},
{
"epoch": 1.7,
"learning_rate": 2.1731959464054955e-05,
"loss": 1.0274,
"step": 66500
},
{
"epoch": 1.71,
"learning_rate": 2.1519417805889954e-05,
"loss": 1.0244,
"step": 67000
},
{
"epoch": 1.72,
"learning_rate": 2.1306876147724954e-05,
"loss": 1.023,
"step": 67500
},
{
"epoch": 1.73,
"learning_rate": 2.1094334489559956e-05,
"loss": 1.0218,
"step": 68000
},
{
"epoch": 1.75,
"learning_rate": 2.0881792831394956e-05,
"loss": 1.0233,
"step": 68500
},
{
"epoch": 1.76,
"learning_rate": 2.0669251173229952e-05,
"loss": 1.0195,
"step": 69000
},
{
"epoch": 1.77,
"learning_rate": 2.0456709515064955e-05,
"loss": 1.0172,
"step": 69500
},
{
"epoch": 1.79,
"learning_rate": 2.0244167856899954e-05,
"loss": 1.0144,
"step": 70000
},
{
"epoch": 1.8,
"learning_rate": 2.0031626198734954e-05,
"loss": 1.0188,
"step": 70500
},
{
"epoch": 1.81,
"learning_rate": 1.9819084540569953e-05,
"loss": 1.019,
"step": 71000
},
{
"epoch": 1.82,
"learning_rate": 1.9606542882404953e-05,
"loss": 1.014,
"step": 71500
},
{
"epoch": 1.84,
"learning_rate": 1.9394001224239952e-05,
"loss": 1.0112,
"step": 72000
},
{
"epoch": 1.85,
"learning_rate": 1.918145956607495e-05,
"loss": 1.012,
"step": 72500
},
{
"epoch": 1.86,
"learning_rate": 1.896891790790995e-05,
"loss": 1.0117,
"step": 73000
},
{
"epoch": 1.87,
"learning_rate": 1.875637624974495e-05,
"loss": 1.0103,
"step": 73500
},
{
"epoch": 1.89,
"learning_rate": 1.854383459157995e-05,
"loss": 1.0099,
"step": 74000
},
{
"epoch": 1.9,
"learning_rate": 1.833129293341495e-05,
"loss": 1.0073,
"step": 74500
},
{
"epoch": 1.91,
"learning_rate": 1.8118751275249952e-05,
"loss": 1.0091,
"step": 75000
},
{
"epoch": 1.93,
"learning_rate": 1.790620961708495e-05,
"loss": 1.0059,
"step": 75500
},
{
"epoch": 1.94,
"learning_rate": 1.7693667958919948e-05,
"loss": 1.001,
"step": 76000
},
{
"epoch": 1.95,
"learning_rate": 1.748112630075495e-05,
"loss": 1.0034,
"step": 76500
},
{
"epoch": 1.96,
"learning_rate": 1.726858464258995e-05,
"loss": 1.0015,
"step": 77000
},
{
"epoch": 1.98,
"learning_rate": 1.7056042984424946e-05,
"loss": 1.0036,
"step": 77500
},
{
"epoch": 1.99,
"learning_rate": 1.6843501326259946e-05,
"loss": 1.0014,
"step": 78000
},
{
"epoch": 2.0,
"learning_rate": 1.663095966809495e-05,
"loss": 1.0012,
"step": 78500
},
{
"epoch": 2.01,
"learning_rate": 1.6418418009929948e-05,
"loss": 1.0002,
"step": 79000
},
{
"epoch": 2.03,
"learning_rate": 1.6205876351764944e-05,
"loss": 0.9992,
"step": 79500
},
{
"epoch": 2.04,
"learning_rate": 1.5993334693599947e-05,
"loss": 0.996,
"step": 80000
},
{
"epoch": 2.05,
"learning_rate": 1.5780793035434946e-05,
"loss": 0.9908,
"step": 80500
},
{
"epoch": 2.07,
"learning_rate": 1.5568251377269946e-05,
"loss": 0.9953,
"step": 81000
},
{
"epoch": 2.08,
"learning_rate": 1.5355709719104945e-05,
"loss": 0.9921,
"step": 81500
},
{
"epoch": 2.09,
"learning_rate": 1.5143168060939945e-05,
"loss": 0.9906,
"step": 82000
},
{
"epoch": 2.1,
"learning_rate": 1.4930626402774944e-05,
"loss": 0.9919,
"step": 82500
},
{
"epoch": 2.12,
"learning_rate": 1.4718084744609945e-05,
"loss": 0.9903,
"step": 83000
},
{
"epoch": 2.13,
"learning_rate": 1.4505543086444945e-05,
"loss": 0.9882,
"step": 83500
},
{
"epoch": 2.14,
"learning_rate": 1.4293001428279943e-05,
"loss": 0.9889,
"step": 84000
},
{
"epoch": 2.16,
"learning_rate": 1.4080459770114942e-05,
"loss": 0.9855,
"step": 84500
},
{
"epoch": 2.17,
"learning_rate": 1.3867918111949943e-05,
"loss": 0.9882,
"step": 85000
},
{
"epoch": 2.18,
"learning_rate": 1.3655376453784943e-05,
"loss": 0.9851,
"step": 85500
},
{
"epoch": 2.19,
"learning_rate": 1.3442834795619944e-05,
"loss": 0.9853,
"step": 86000
},
{
"epoch": 2.21,
"learning_rate": 1.3230293137454942e-05,
"loss": 0.986,
"step": 86500
},
{
"epoch": 2.22,
"learning_rate": 1.3017751479289941e-05,
"loss": 0.9861,
"step": 87000
},
{
"epoch": 2.23,
"learning_rate": 1.2805209821124942e-05,
"loss": 0.9837,
"step": 87500
},
{
"epoch": 2.24,
"learning_rate": 1.2592668162959942e-05,
"loss": 0.9819,
"step": 88000
},
{
"epoch": 2.26,
"learning_rate": 1.2380126504794941e-05,
"loss": 0.983,
"step": 88500
},
{
"epoch": 2.27,
"learning_rate": 1.2167584846629939e-05,
"loss": 0.9832,
"step": 89000
},
{
"epoch": 2.28,
"learning_rate": 1.195504318846494e-05,
"loss": 0.9788,
"step": 89500
},
{
"epoch": 2.3,
"learning_rate": 1.174250153029994e-05,
"loss": 0.982,
"step": 90000
},
{
"epoch": 2.31,
"learning_rate": 1.1529959872134939e-05,
"loss": 0.9786,
"step": 90500
},
{
"epoch": 2.32,
"learning_rate": 1.1317418213969938e-05,
"loss": 0.98,
"step": 91000
},
{
"epoch": 2.33,
"learning_rate": 1.1104876555804938e-05,
"loss": 0.9789,
"step": 91500
},
{
"epoch": 2.35,
"learning_rate": 1.0892334897639939e-05,
"loss": 0.9767,
"step": 92000
},
{
"epoch": 2.36,
"learning_rate": 1.0679793239474937e-05,
"loss": 0.9762,
"step": 92500
},
{
"epoch": 2.37,
"learning_rate": 1.0467251581309938e-05,
"loss": 0.9766,
"step": 93000
},
{
"epoch": 2.38,
"learning_rate": 1.0254709923144937e-05,
"loss": 0.9749,
"step": 93500
},
{
"epoch": 2.4,
"learning_rate": 1.0042168264979937e-05,
"loss": 0.9752,
"step": 94000
},
{
"epoch": 2.41,
"learning_rate": 9.829626606814936e-06,
"loss": 0.9751,
"step": 94500
},
{
"epoch": 2.42,
"learning_rate": 9.617084948649936e-06,
"loss": 0.9702,
"step": 95000
},
{
"epoch": 2.44,
"learning_rate": 9.404543290484935e-06,
"loss": 0.9744,
"step": 95500
},
{
"epoch": 2.45,
"learning_rate": 9.192001632319935e-06,
"loss": 0.9701,
"step": 96000
},
{
"epoch": 2.46,
"learning_rate": 8.979459974154934e-06,
"loss": 0.9705,
"step": 96500
},
{
"epoch": 2.47,
"learning_rate": 8.766918315989935e-06,
"loss": 0.9708,
"step": 97000
},
{
"epoch": 2.49,
"learning_rate": 8.554376657824933e-06,
"loss": 0.9712,
"step": 97500
},
{
"epoch": 2.5,
"learning_rate": 8.341834999659934e-06,
"loss": 0.9689,
"step": 98000
},
{
"epoch": 2.51,
"learning_rate": 8.129293341494934e-06,
"loss": 0.9672,
"step": 98500
},
{
"epoch": 2.52,
"learning_rate": 7.916751683329933e-06,
"loss": 0.967,
"step": 99000
},
{
"epoch": 2.54,
"learning_rate": 7.704210025164933e-06,
"loss": 0.9665,
"step": 99500
},
{
"epoch": 2.55,
"learning_rate": 7.491668366999932e-06,
"loss": 0.9654,
"step": 100000
},
{
"epoch": 2.56,
"learning_rate": 7.2791267088349325e-06,
"loss": 0.967,
"step": 100500
},
{
"epoch": 2.58,
"learning_rate": 7.066585050669931e-06,
"loss": 0.9666,
"step": 101000
},
{
"epoch": 2.59,
"learning_rate": 6.854043392504931e-06,
"loss": 0.9635,
"step": 101500
},
{
"epoch": 2.6,
"learning_rate": 6.641501734339931e-06,
"loss": 0.9681,
"step": 102000
},
{
"epoch": 2.61,
"learning_rate": 6.428960076174931e-06,
"loss": 0.9631,
"step": 102500
},
{
"epoch": 2.63,
"learning_rate": 6.216418418009931e-06,
"loss": 0.9653,
"step": 103000
},
{
"epoch": 2.64,
"learning_rate": 6.00387675984493e-06,
"loss": 0.9612,
"step": 103500
},
{
"epoch": 2.65,
"learning_rate": 5.7913351016799295e-06,
"loss": 0.9633,
"step": 104000
},
{
"epoch": 2.67,
"learning_rate": 5.578793443514929e-06,
"loss": 0.9595,
"step": 104500
},
{
"epoch": 2.68,
"learning_rate": 5.3662517853499285e-06,
"loss": 0.9617,
"step": 105000
},
{
"epoch": 2.69,
"learning_rate": 5.153710127184928e-06,
"loss": 0.9609,
"step": 105500
},
{
"epoch": 2.7,
"learning_rate": 4.941168469019928e-06,
"loss": 0.9588,
"step": 106000
},
{
"epoch": 2.72,
"learning_rate": 4.728626810854928e-06,
"loss": 0.9613,
"step": 106500
},
{
"epoch": 2.73,
"learning_rate": 4.516085152689927e-06,
"loss": 0.9595,
"step": 107000
},
{
"epoch": 2.74,
"learning_rate": 4.3035434945249275e-06,
"loss": 0.9592,
"step": 107500
},
{
"epoch": 2.75,
"learning_rate": 4.091001836359927e-06,
"loss": 0.959,
"step": 108000
},
{
"epoch": 2.77,
"learning_rate": 3.878460178194926e-06,
"loss": 0.9579,
"step": 108500
},
{
"epoch": 2.78,
"learning_rate": 3.6659185200299263e-06,
"loss": 0.9581,
"step": 109000
},
{
"epoch": 2.79,
"learning_rate": 3.4533768618649257e-06,
"loss": 0.9574,
"step": 109500
},
{
"epoch": 2.81,
"learning_rate": 3.240835203699925e-06,
"loss": 0.9567,
"step": 110000
},
{
"epoch": 2.82,
"learning_rate": 3.028293545534925e-06,
"loss": 0.9584,
"step": 110500
},
{
"epoch": 2.83,
"learning_rate": 2.8157518873699246e-06,
"loss": 0.9519,
"step": 111000
},
{
"epoch": 2.84,
"learning_rate": 2.603210229204924e-06,
"loss": 0.9539,
"step": 111500
},
{
"epoch": 2.86,
"learning_rate": 2.390668571039924e-06,
"loss": 0.9568,
"step": 112000
},
{
"epoch": 2.87,
"learning_rate": 2.1781269128749238e-06,
"loss": 0.9525,
"step": 112500
},
{
"epoch": 2.88,
"learning_rate": 1.9655852547099233e-06,
"loss": 0.958,
"step": 113000
},
{
"epoch": 2.89,
"learning_rate": 1.753043596544923e-06,
"loss": 0.9556,
"step": 113500
},
{
"epoch": 2.91,
"learning_rate": 1.5405019383799224e-06,
"loss": 0.9521,
"step": 114000
},
{
"epoch": 2.92,
"learning_rate": 1.3279602802149223e-06,
"loss": 0.9542,
"step": 114500
},
{
"epoch": 2.93,
"learning_rate": 1.1154186220499217e-06,
"loss": 0.953,
"step": 115000
},
{
"epoch": 2.95,
"learning_rate": 9.028769638849214e-07,
"loss": 0.9548,
"step": 115500
},
{
"epoch": 2.96,
"learning_rate": 6.903353057199211e-07,
"loss": 0.9539,
"step": 116000
},
{
"epoch": 2.97,
"learning_rate": 4.777936475549208e-07,
"loss": 0.9532,
"step": 116500
},
{
"epoch": 2.98,
"learning_rate": 2.6525198938992043e-07,
"loss": 0.9539,
"step": 117000
},
{
"epoch": 3.0,
"learning_rate": 5.271033122492008e-08,
"loss": 0.9543,
"step": 117500
},
{
"epoch": 3.0,
"step": 117624,
"total_flos": 3.510933473606687e+19,
"train_loss": 0.762959968830496,
"train_runtime": 1022407.089,
"train_samples_per_second": 36.815,
"train_steps_per_second": 0.115
}
],
"logging_steps": 500,
"max_steps": 117624,
"num_train_epochs": 3,
"save_steps": 10000,
"total_flos": 3.510933473606687e+19,
"trial_name": null,
"trial_params": null
}