zuberta / trainer_state.json
leonzong's picture
Upload folder using huggingface_hub
f4ee18a verified
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 2.218982703597179,
"eval_steps": 500,
"global_step": 112640,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0,
"learning_rate": 4.9917917602406e-05,
"loss": 9.4912,
"step": 250
},
{
"epoch": 0.01,
"learning_rate": 4.9835835204812e-05,
"loss": 9.0253,
"step": 500
},
{
"epoch": 0.01,
"learning_rate": 4.9753752807218e-05,
"loss": 8.889,
"step": 750
},
{
"epoch": 0.02,
"learning_rate": 4.9671670409624e-05,
"loss": 8.7297,
"step": 1000
},
{
"epoch": 0.02,
"learning_rate": 4.958958801203e-05,
"loss": 8.6079,
"step": 1250
},
{
"epoch": 0.03,
"learning_rate": 4.9507505614436e-05,
"loss": 8.525,
"step": 1500
},
{
"epoch": 0.03,
"learning_rate": 4.9425423216842e-05,
"loss": 8.4223,
"step": 1750
},
{
"epoch": 0.04,
"learning_rate": 4.9343340819248e-05,
"loss": 8.3692,
"step": 2000
},
{
"epoch": 0.04,
"learning_rate": 4.926125842165399e-05,
"loss": 8.332,
"step": 2250
},
{
"epoch": 0.05,
"learning_rate": 4.917917602406e-05,
"loss": 8.2672,
"step": 2500
},
{
"epoch": 0.05,
"learning_rate": 4.9097093626466e-05,
"loss": 8.1887,
"step": 2750
},
{
"epoch": 0.06,
"learning_rate": 4.9015011228871996e-05,
"loss": 8.1799,
"step": 3000
},
{
"epoch": 0.06,
"learning_rate": 4.893292883127799e-05,
"loss": 8.1428,
"step": 3250
},
{
"epoch": 0.07,
"learning_rate": 4.8850846433683995e-05,
"loss": 8.1542,
"step": 3500
},
{
"epoch": 0.07,
"learning_rate": 4.8768764036089995e-05,
"loss": 8.0947,
"step": 3750
},
{
"epoch": 0.08,
"learning_rate": 4.868668163849599e-05,
"loss": 8.0386,
"step": 4000
},
{
"epoch": 0.08,
"learning_rate": 4.860459924090199e-05,
"loss": 7.9978,
"step": 4250
},
{
"epoch": 0.09,
"learning_rate": 4.852251684330799e-05,
"loss": 7.9489,
"step": 4500
},
{
"epoch": 0.09,
"learning_rate": 4.8440434445713986e-05,
"loss": 7.9363,
"step": 4750
},
{
"epoch": 0.1,
"learning_rate": 4.8358352048119985e-05,
"loss": 7.9265,
"step": 5000
},
{
"epoch": 0.1,
"learning_rate": 4.8276269650525985e-05,
"loss": 7.9027,
"step": 5250
},
{
"epoch": 0.11,
"learning_rate": 4.819418725293199e-05,
"loss": 7.8451,
"step": 5500
},
{
"epoch": 0.11,
"learning_rate": 4.8112104855337984e-05,
"loss": 7.7337,
"step": 5750
},
{
"epoch": 0.12,
"learning_rate": 4.8030022457743983e-05,
"loss": 7.7837,
"step": 6000
},
{
"epoch": 0.12,
"learning_rate": 4.794794006014998e-05,
"loss": 7.7527,
"step": 6250
},
{
"epoch": 0.13,
"learning_rate": 4.786585766255598e-05,
"loss": 7.7434,
"step": 6500
},
{
"epoch": 0.13,
"learning_rate": 4.778377526496198e-05,
"loss": 7.6981,
"step": 6750
},
{
"epoch": 0.14,
"learning_rate": 4.770169286736798e-05,
"loss": 7.6708,
"step": 7000
},
{
"epoch": 0.14,
"learning_rate": 4.761961046977398e-05,
"loss": 7.5951,
"step": 7250
},
{
"epoch": 0.15,
"learning_rate": 4.753752807217998e-05,
"loss": 7.5975,
"step": 7500
},
{
"epoch": 0.15,
"learning_rate": 4.745544567458598e-05,
"loss": 7.5186,
"step": 7750
},
{
"epoch": 0.16,
"learning_rate": 4.737336327699198e-05,
"loss": 7.4844,
"step": 8000
},
{
"epoch": 0.16,
"learning_rate": 4.729128087939798e-05,
"loss": 7.4021,
"step": 8250
},
{
"epoch": 0.17,
"learning_rate": 4.720919848180398e-05,
"loss": 7.4089,
"step": 8500
},
{
"epoch": 0.17,
"learning_rate": 4.712711608420998e-05,
"loss": 7.3132,
"step": 8750
},
{
"epoch": 0.18,
"learning_rate": 4.704503368661598e-05,
"loss": 7.3552,
"step": 9000
},
{
"epoch": 0.18,
"learning_rate": 4.696295128902197e-05,
"loss": 7.3327,
"step": 9250
},
{
"epoch": 0.19,
"learning_rate": 4.6880868891427976e-05,
"loss": 7.2294,
"step": 9500
},
{
"epoch": 0.19,
"learning_rate": 4.6798786493833976e-05,
"loss": 7.2296,
"step": 9750
},
{
"epoch": 0.2,
"learning_rate": 4.671670409623997e-05,
"loss": 7.1595,
"step": 10000
},
{
"epoch": 0.2,
"learning_rate": 4.663462169864597e-05,
"loss": 7.1434,
"step": 10250
},
{
"epoch": 0.21,
"learning_rate": 4.6552539301051974e-05,
"loss": 7.1707,
"step": 10500
},
{
"epoch": 0.21,
"learning_rate": 4.6470456903457974e-05,
"loss": 7.0756,
"step": 10750
},
{
"epoch": 0.22,
"learning_rate": 4.6388374505863967e-05,
"loss": 7.0205,
"step": 11000
},
{
"epoch": 0.22,
"learning_rate": 4.6306292108269966e-05,
"loss": 7.0475,
"step": 11250
},
{
"epoch": 0.23,
"learning_rate": 4.622420971067597e-05,
"loss": 7.0355,
"step": 11500
},
{
"epoch": 0.23,
"learning_rate": 4.6142127313081965e-05,
"loss": 6.9695,
"step": 11750
},
{
"epoch": 0.24,
"learning_rate": 4.6060044915487965e-05,
"loss": 6.962,
"step": 12000
},
{
"epoch": 0.24,
"learning_rate": 4.5977962517893964e-05,
"loss": 6.9281,
"step": 12250
},
{
"epoch": 0.25,
"learning_rate": 4.5895880120299964e-05,
"loss": 6.9421,
"step": 12500
},
{
"epoch": 0.25,
"learning_rate": 4.581379772270596e-05,
"loss": 6.9498,
"step": 12750
},
{
"epoch": 0.26,
"learning_rate": 4.573171532511196e-05,
"loss": 6.8833,
"step": 13000
},
{
"epoch": 0.26,
"learning_rate": 4.564963292751796e-05,
"loss": 6.8878,
"step": 13250
},
{
"epoch": 0.27,
"learning_rate": 4.556755052992396e-05,
"loss": 6.845,
"step": 13500
},
{
"epoch": 0.27,
"learning_rate": 4.548546813232996e-05,
"loss": 6.833,
"step": 13750
},
{
"epoch": 0.28,
"learning_rate": 4.540338573473596e-05,
"loss": 6.8062,
"step": 14000
},
{
"epoch": 0.28,
"learning_rate": 4.532130333714196e-05,
"loss": 6.8126,
"step": 14250
},
{
"epoch": 0.29,
"learning_rate": 4.523922093954796e-05,
"loss": 6.7655,
"step": 14500
},
{
"epoch": 0.29,
"learning_rate": 4.515713854195396e-05,
"loss": 6.7613,
"step": 14750
},
{
"epoch": 0.3,
"learning_rate": 4.507505614435996e-05,
"loss": 6.6869,
"step": 15000
},
{
"epoch": 0.3,
"learning_rate": 4.499297374676595e-05,
"loss": 6.6834,
"step": 15250
},
{
"epoch": 0.31,
"learning_rate": 4.491089134917196e-05,
"loss": 6.7371,
"step": 15500
},
{
"epoch": 0.31,
"learning_rate": 4.482880895157796e-05,
"loss": 6.6962,
"step": 15750
},
{
"epoch": 0.32,
"learning_rate": 4.4746726553983956e-05,
"loss": 6.6554,
"step": 16000
},
{
"epoch": 0.32,
"learning_rate": 4.466464415638995e-05,
"loss": 6.6898,
"step": 16250
},
{
"epoch": 0.33,
"learning_rate": 4.4582561758795955e-05,
"loss": 6.6349,
"step": 16500
},
{
"epoch": 0.33,
"learning_rate": 4.4500479361201955e-05,
"loss": 6.6075,
"step": 16750
},
{
"epoch": 0.33,
"learning_rate": 4.441839696360795e-05,
"loss": 6.6128,
"step": 17000
},
{
"epoch": 0.34,
"learning_rate": 4.433631456601395e-05,
"loss": 6.5883,
"step": 17250
},
{
"epoch": 0.34,
"learning_rate": 4.4254232168419953e-05,
"loss": 6.5475,
"step": 17500
},
{
"epoch": 0.35,
"learning_rate": 4.4172149770825946e-05,
"loss": 6.4996,
"step": 17750
},
{
"epoch": 0.35,
"learning_rate": 4.4090067373231946e-05,
"loss": 6.5243,
"step": 18000
},
{
"epoch": 0.36,
"learning_rate": 4.4007984975637945e-05,
"loss": 6.5113,
"step": 18250
},
{
"epoch": 0.36,
"learning_rate": 4.392590257804395e-05,
"loss": 6.4859,
"step": 18500
},
{
"epoch": 0.37,
"learning_rate": 4.3843820180449944e-05,
"loss": 6.473,
"step": 18750
},
{
"epoch": 0.37,
"learning_rate": 4.3761737782855944e-05,
"loss": 6.5033,
"step": 19000
},
{
"epoch": 0.38,
"learning_rate": 4.367965538526194e-05,
"loss": 6.4411,
"step": 19250
},
{
"epoch": 0.38,
"learning_rate": 4.359757298766794e-05,
"loss": 6.4374,
"step": 19500
},
{
"epoch": 0.39,
"learning_rate": 4.351549059007394e-05,
"loss": 6.4449,
"step": 19750
},
{
"epoch": 0.39,
"learning_rate": 4.343340819247994e-05,
"loss": 6.4043,
"step": 20000
},
{
"epoch": 0.4,
"learning_rate": 4.335132579488594e-05,
"loss": 6.3784,
"step": 20250
},
{
"epoch": 0.4,
"learning_rate": 4.326924339729194e-05,
"loss": 6.3888,
"step": 20500
},
{
"epoch": 0.41,
"learning_rate": 4.318716099969794e-05,
"loss": 6.3818,
"step": 20750
},
{
"epoch": 0.41,
"learning_rate": 4.310507860210394e-05,
"loss": 6.3793,
"step": 21000
},
{
"epoch": 0.42,
"learning_rate": 4.302299620450993e-05,
"loss": 6.3635,
"step": 21250
},
{
"epoch": 0.42,
"learning_rate": 4.294091380691594e-05,
"loss": 6.3109,
"step": 21500
},
{
"epoch": 0.43,
"learning_rate": 4.285883140932194e-05,
"loss": 6.3004,
"step": 21750
},
{
"epoch": 0.43,
"learning_rate": 4.277674901172794e-05,
"loss": 6.3036,
"step": 22000
},
{
"epoch": 0.44,
"learning_rate": 4.269466661413393e-05,
"loss": 6.233,
"step": 22250
},
{
"epoch": 0.44,
"learning_rate": 4.2612584216539936e-05,
"loss": 6.242,
"step": 22500
},
{
"epoch": 0.45,
"learning_rate": 4.2530501818945936e-05,
"loss": 6.2686,
"step": 22750
},
{
"epoch": 0.45,
"learning_rate": 4.244841942135193e-05,
"loss": 6.2183,
"step": 23000
},
{
"epoch": 0.46,
"learning_rate": 4.236633702375793e-05,
"loss": 6.2428,
"step": 23250
},
{
"epoch": 0.46,
"learning_rate": 4.2284254626163934e-05,
"loss": 6.1933,
"step": 23500
},
{
"epoch": 0.47,
"learning_rate": 4.2202172228569934e-05,
"loss": 6.1957,
"step": 23750
},
{
"epoch": 0.47,
"learning_rate": 4.212008983097593e-05,
"loss": 6.2449,
"step": 24000
},
{
"epoch": 0.48,
"learning_rate": 4.2038007433381926e-05,
"loss": 6.186,
"step": 24250
},
{
"epoch": 0.48,
"learning_rate": 4.195592503578793e-05,
"loss": 6.1722,
"step": 24500
},
{
"epoch": 0.49,
"learning_rate": 4.1873842638193925e-05,
"loss": 6.193,
"step": 24750
},
{
"epoch": 0.49,
"learning_rate": 4.1791760240599925e-05,
"loss": 6.1829,
"step": 25000
},
{
"epoch": 0.5,
"learning_rate": 4.1709677843005924e-05,
"loss": 6.1587,
"step": 25250
},
{
"epoch": 0.5,
"learning_rate": 4.1627595445411924e-05,
"loss": 6.1058,
"step": 25500
},
{
"epoch": 0.51,
"learning_rate": 4.154551304781792e-05,
"loss": 6.1346,
"step": 25750
},
{
"epoch": 0.51,
"learning_rate": 4.146343065022392e-05,
"loss": 6.0702,
"step": 26000
},
{
"epoch": 0.52,
"learning_rate": 4.138134825262992e-05,
"loss": 6.0994,
"step": 26250
},
{
"epoch": 0.52,
"learning_rate": 4.129926585503592e-05,
"loss": 6.0871,
"step": 26500
},
{
"epoch": 0.53,
"learning_rate": 4.121718345744192e-05,
"loss": 6.0275,
"step": 26750
},
{
"epoch": 0.53,
"learning_rate": 4.113510105984792e-05,
"loss": 6.0883,
"step": 27000
},
{
"epoch": 0.54,
"learning_rate": 4.105301866225392e-05,
"loss": 6.0602,
"step": 27250
},
{
"epoch": 0.54,
"learning_rate": 4.097093626465992e-05,
"loss": 5.9898,
"step": 27500
},
{
"epoch": 0.55,
"learning_rate": 4.088885386706592e-05,
"loss": 6.0722,
"step": 27750
},
{
"epoch": 0.55,
"learning_rate": 4.080677146947192e-05,
"loss": 5.987,
"step": 28000
},
{
"epoch": 0.56,
"learning_rate": 4.072468907187791e-05,
"loss": 6.0631,
"step": 28250
},
{
"epoch": 0.56,
"learning_rate": 4.064260667428392e-05,
"loss": 5.9497,
"step": 28500
},
{
"epoch": 0.57,
"learning_rate": 4.056052427668992e-05,
"loss": 5.9776,
"step": 28750
},
{
"epoch": 0.57,
"learning_rate": 4.047844187909591e-05,
"loss": 5.9793,
"step": 29000
},
{
"epoch": 0.58,
"learning_rate": 4.039635948150191e-05,
"loss": 6.004,
"step": 29250
},
{
"epoch": 0.58,
"learning_rate": 4.0314277083907916e-05,
"loss": 5.9236,
"step": 29500
},
{
"epoch": 0.59,
"learning_rate": 4.0232194686313915e-05,
"loss": 5.9507,
"step": 29750
},
{
"epoch": 0.59,
"learning_rate": 4.015011228871991e-05,
"loss": 5.9362,
"step": 30000
},
{
"epoch": 0.6,
"learning_rate": 4.006802989112591e-05,
"loss": 5.9727,
"step": 30250
},
{
"epoch": 0.6,
"learning_rate": 3.9985947493531914e-05,
"loss": 5.8949,
"step": 30500
},
{
"epoch": 0.61,
"learning_rate": 3.9903865095937906e-05,
"loss": 5.9043,
"step": 30750
},
{
"epoch": 0.61,
"learning_rate": 3.9821782698343906e-05,
"loss": 5.8899,
"step": 31000
},
{
"epoch": 0.62,
"learning_rate": 3.9739700300749905e-05,
"loss": 5.8585,
"step": 31250
},
{
"epoch": 0.62,
"learning_rate": 3.9657617903155905e-05,
"loss": 5.8793,
"step": 31500
},
{
"epoch": 0.63,
"learning_rate": 3.9575535505561904e-05,
"loss": 5.8836,
"step": 31750
},
{
"epoch": 0.63,
"learning_rate": 3.9493453107967904e-05,
"loss": 5.8597,
"step": 32000
},
{
"epoch": 0.64,
"learning_rate": 3.94113707103739e-05,
"loss": 5.8795,
"step": 32250
},
{
"epoch": 0.64,
"learning_rate": 3.93292883127799e-05,
"loss": 5.8787,
"step": 32500
},
{
"epoch": 0.65,
"learning_rate": 3.92472059151859e-05,
"loss": 5.8603,
"step": 32750
},
{
"epoch": 0.65,
"learning_rate": 3.91651235175919e-05,
"loss": 5.8186,
"step": 33000
},
{
"epoch": 0.66,
"learning_rate": 3.90830411199979e-05,
"loss": 5.7699,
"step": 33250
},
{
"epoch": 0.66,
"learning_rate": 3.90009587224039e-05,
"loss": 5.7936,
"step": 33500
},
{
"epoch": 0.66,
"learning_rate": 3.89188763248099e-05,
"loss": 5.7771,
"step": 33750
},
{
"epoch": 0.67,
"learning_rate": 3.88367939272159e-05,
"loss": 5.7897,
"step": 34000
},
{
"epoch": 0.67,
"learning_rate": 3.875471152962189e-05,
"loss": 5.7541,
"step": 34250
},
{
"epoch": 0.68,
"learning_rate": 3.86726291320279e-05,
"loss": 5.7193,
"step": 34500
},
{
"epoch": 0.68,
"learning_rate": 3.85905467344339e-05,
"loss": 5.7607,
"step": 34750
},
{
"epoch": 0.69,
"learning_rate": 3.85084643368399e-05,
"loss": 5.743,
"step": 35000
},
{
"epoch": 0.69,
"learning_rate": 3.842638193924589e-05,
"loss": 5.7274,
"step": 35250
},
{
"epoch": 0.7,
"learning_rate": 3.8344299541651897e-05,
"loss": 5.7183,
"step": 35500
},
{
"epoch": 0.7,
"learning_rate": 3.8262217144057896e-05,
"loss": 5.6701,
"step": 35750
},
{
"epoch": 0.71,
"learning_rate": 3.818013474646389e-05,
"loss": 5.7014,
"step": 36000
},
{
"epoch": 0.71,
"learning_rate": 3.809805234886989e-05,
"loss": 5.684,
"step": 36250
},
{
"epoch": 0.72,
"learning_rate": 3.8015969951275895e-05,
"loss": 5.697,
"step": 36500
},
{
"epoch": 0.72,
"learning_rate": 3.793388755368189e-05,
"loss": 5.6821,
"step": 36750
},
{
"epoch": 0.73,
"learning_rate": 3.785180515608789e-05,
"loss": 5.739,
"step": 37000
},
{
"epoch": 0.73,
"learning_rate": 3.7769722758493886e-05,
"loss": 5.6367,
"step": 37250
},
{
"epoch": 0.74,
"learning_rate": 3.768764036089989e-05,
"loss": 5.625,
"step": 37500
},
{
"epoch": 0.74,
"learning_rate": 3.7605557963305885e-05,
"loss": 5.6101,
"step": 37750
},
{
"epoch": 0.75,
"learning_rate": 3.7523475565711885e-05,
"loss": 5.6124,
"step": 38000
},
{
"epoch": 0.75,
"learning_rate": 3.7441393168117884e-05,
"loss": 5.6286,
"step": 38250
},
{
"epoch": 0.76,
"learning_rate": 3.7359310770523884e-05,
"loss": 5.6069,
"step": 38500
},
{
"epoch": 0.76,
"learning_rate": 3.727722837292988e-05,
"loss": 5.6131,
"step": 38750
},
{
"epoch": 0.77,
"learning_rate": 3.719514597533588e-05,
"loss": 5.588,
"step": 39000
},
{
"epoch": 0.77,
"learning_rate": 3.711306357774188e-05,
"loss": 5.5794,
"step": 39250
},
{
"epoch": 0.78,
"learning_rate": 3.703098118014788e-05,
"loss": 5.6147,
"step": 39500
},
{
"epoch": 0.78,
"learning_rate": 3.694889878255388e-05,
"loss": 5.5989,
"step": 39750
},
{
"epoch": 0.79,
"learning_rate": 3.686681638495988e-05,
"loss": 5.5775,
"step": 40000
},
{
"epoch": 0.79,
"learning_rate": 3.678473398736588e-05,
"loss": 5.554,
"step": 40250
},
{
"epoch": 0.8,
"learning_rate": 3.670265158977188e-05,
"loss": 5.6252,
"step": 40500
},
{
"epoch": 0.8,
"learning_rate": 3.662056919217788e-05,
"loss": 5.575,
"step": 40750
},
{
"epoch": 0.81,
"learning_rate": 3.653848679458388e-05,
"loss": 5.5285,
"step": 41000
},
{
"epoch": 0.81,
"learning_rate": 3.645640439698987e-05,
"loss": 5.5481,
"step": 41250
},
{
"epoch": 0.82,
"learning_rate": 3.637432199939588e-05,
"loss": 5.4834,
"step": 41500
},
{
"epoch": 0.82,
"learning_rate": 3.629223960180188e-05,
"loss": 5.5461,
"step": 41750
},
{
"epoch": 0.83,
"learning_rate": 3.621015720420787e-05,
"loss": 5.5474,
"step": 42000
},
{
"epoch": 0.83,
"learning_rate": 3.612807480661387e-05,
"loss": 5.53,
"step": 42250
},
{
"epoch": 0.84,
"learning_rate": 3.6045992409019876e-05,
"loss": 5.4902,
"step": 42500
},
{
"epoch": 0.84,
"learning_rate": 3.5963910011425875e-05,
"loss": 5.4589,
"step": 42750
},
{
"epoch": 0.85,
"learning_rate": 3.588182761383187e-05,
"loss": 5.4909,
"step": 43000
},
{
"epoch": 0.85,
"learning_rate": 3.579974521623787e-05,
"loss": 5.4721,
"step": 43250
},
{
"epoch": 0.86,
"learning_rate": 3.5717662818643874e-05,
"loss": 5.5191,
"step": 43500
},
{
"epoch": 0.86,
"learning_rate": 3.5635580421049866e-05,
"loss": 5.4646,
"step": 43750
},
{
"epoch": 0.87,
"learning_rate": 3.5553498023455866e-05,
"loss": 5.4786,
"step": 44000
},
{
"epoch": 0.87,
"learning_rate": 3.5471415625861865e-05,
"loss": 5.4943,
"step": 44250
},
{
"epoch": 0.88,
"learning_rate": 3.5389333228267865e-05,
"loss": 5.443,
"step": 44500
},
{
"epoch": 0.88,
"learning_rate": 3.5307250830673864e-05,
"loss": 5.4313,
"step": 44750
},
{
"epoch": 0.89,
"learning_rate": 3.5225168433079864e-05,
"loss": 5.4594,
"step": 45000
},
{
"epoch": 0.89,
"learning_rate": 3.514308603548586e-05,
"loss": 5.4203,
"step": 45250
},
{
"epoch": 0.9,
"learning_rate": 3.506100363789186e-05,
"loss": 5.3763,
"step": 45500
},
{
"epoch": 0.9,
"learning_rate": 3.497892124029786e-05,
"loss": 5.445,
"step": 45750
},
{
"epoch": 0.91,
"learning_rate": 3.489683884270386e-05,
"loss": 5.4139,
"step": 46000
},
{
"epoch": 0.91,
"learning_rate": 3.481475644510986e-05,
"loss": 5.3907,
"step": 46250
},
{
"epoch": 0.92,
"learning_rate": 3.473267404751586e-05,
"loss": 5.4021,
"step": 46500
},
{
"epoch": 0.92,
"learning_rate": 3.465059164992186e-05,
"loss": 5.3955,
"step": 46750
},
{
"epoch": 0.93,
"learning_rate": 3.456850925232786e-05,
"loss": 5.3653,
"step": 47000
},
{
"epoch": 0.93,
"learning_rate": 3.448642685473385e-05,
"loss": 5.3789,
"step": 47250
},
{
"epoch": 0.94,
"learning_rate": 3.440434445713986e-05,
"loss": 5.3561,
"step": 47500
},
{
"epoch": 0.94,
"learning_rate": 3.432226205954586e-05,
"loss": 5.3199,
"step": 47750
},
{
"epoch": 0.95,
"learning_rate": 3.424017966195186e-05,
"loss": 5.3322,
"step": 48000
},
{
"epoch": 0.95,
"learning_rate": 3.415809726435785e-05,
"loss": 5.3303,
"step": 48250
},
{
"epoch": 0.96,
"learning_rate": 3.407601486676386e-05,
"loss": 5.2998,
"step": 48500
},
{
"epoch": 0.96,
"learning_rate": 3.3993932469169856e-05,
"loss": 5.3534,
"step": 48750
},
{
"epoch": 0.97,
"learning_rate": 3.391185007157585e-05,
"loss": 5.3316,
"step": 49000
},
{
"epoch": 0.97,
"learning_rate": 3.382976767398185e-05,
"loss": 5.3047,
"step": 49250
},
{
"epoch": 0.98,
"learning_rate": 3.3747685276387855e-05,
"loss": 5.2876,
"step": 49500
},
{
"epoch": 0.98,
"learning_rate": 3.366560287879385e-05,
"loss": 5.3265,
"step": 49750
},
{
"epoch": 0.98,
"learning_rate": 3.358352048119985e-05,
"loss": 5.2812,
"step": 50000
},
{
"epoch": 0.99,
"learning_rate": 3.3501438083605846e-05,
"loss": 5.2702,
"step": 50250
},
{
"epoch": 0.99,
"learning_rate": 3.341935568601185e-05,
"loss": 5.2525,
"step": 50500
},
{
"epoch": 1.0,
"learning_rate": 3.3337273288417845e-05,
"loss": 5.293,
"step": 50750
},
{
"epoch": 1.0,
"eval_loss": 5.185632228851318,
"eval_runtime": 624.5274,
"eval_samples_per_second": 289.038,
"eval_steps_per_second": 9.032,
"step": 50762
},
{
"epoch": 1.0,
"learning_rate": 3.3255190890823845e-05,
"loss": 5.252,
"step": 51000
},
{
"epoch": 1.01,
"learning_rate": 3.3173108493229844e-05,
"loss": 5.2646,
"step": 51250
},
{
"epoch": 1.01,
"learning_rate": 3.3091026095635844e-05,
"loss": 5.2291,
"step": 51500
},
{
"epoch": 1.02,
"learning_rate": 3.300894369804184e-05,
"loss": 5.2391,
"step": 51750
},
{
"epoch": 1.02,
"learning_rate": 3.292686130044784e-05,
"loss": 5.1969,
"step": 52000
},
{
"epoch": 1.03,
"learning_rate": 3.284477890285384e-05,
"loss": 5.2636,
"step": 52250
},
{
"epoch": 1.03,
"learning_rate": 3.276269650525984e-05,
"loss": 5.23,
"step": 52500
},
{
"epoch": 1.04,
"learning_rate": 3.268061410766584e-05,
"loss": 5.2188,
"step": 52750
},
{
"epoch": 1.04,
"learning_rate": 3.259853171007184e-05,
"loss": 5.2196,
"step": 53000
},
{
"epoch": 1.05,
"learning_rate": 3.2516449312477834e-05,
"loss": 5.1822,
"step": 53250
},
{
"epoch": 1.05,
"learning_rate": 3.243436691488384e-05,
"loss": 5.1899,
"step": 53500
},
{
"epoch": 1.06,
"learning_rate": 3.235228451728984e-05,
"loss": 5.1678,
"step": 53750
},
{
"epoch": 1.06,
"learning_rate": 3.227020211969584e-05,
"loss": 5.1901,
"step": 54000
},
{
"epoch": 1.07,
"learning_rate": 3.218811972210183e-05,
"loss": 5.2002,
"step": 54250
},
{
"epoch": 1.07,
"learning_rate": 3.210603732450784e-05,
"loss": 5.1926,
"step": 54500
},
{
"epoch": 1.08,
"learning_rate": 3.202395492691384e-05,
"loss": 5.1494,
"step": 54750
},
{
"epoch": 1.08,
"learning_rate": 3.194187252931983e-05,
"loss": 5.1964,
"step": 55000
},
{
"epoch": 1.09,
"learning_rate": 3.185979013172583e-05,
"loss": 5.1541,
"step": 55250
},
{
"epoch": 1.09,
"learning_rate": 3.1777707734131836e-05,
"loss": 5.1612,
"step": 55500
},
{
"epoch": 1.1,
"learning_rate": 3.1695625336537835e-05,
"loss": 5.1887,
"step": 55750
},
{
"epoch": 1.1,
"learning_rate": 3.161354293894383e-05,
"loss": 5.16,
"step": 56000
},
{
"epoch": 1.11,
"learning_rate": 3.153146054134983e-05,
"loss": 5.1533,
"step": 56250
},
{
"epoch": 1.11,
"learning_rate": 3.1449378143755834e-05,
"loss": 5.1851,
"step": 56500
},
{
"epoch": 1.12,
"learning_rate": 3.1367295746161826e-05,
"loss": 5.1552,
"step": 56750
},
{
"epoch": 1.12,
"learning_rate": 3.1285213348567826e-05,
"loss": 5.1378,
"step": 57000
},
{
"epoch": 1.13,
"learning_rate": 3.1203130950973825e-05,
"loss": 5.118,
"step": 57250
},
{
"epoch": 1.13,
"learning_rate": 3.1121048553379825e-05,
"loss": 5.1494,
"step": 57500
},
{
"epoch": 1.14,
"learning_rate": 3.1038966155785824e-05,
"loss": 5.0843,
"step": 57750
},
{
"epoch": 1.14,
"learning_rate": 3.0956883758191824e-05,
"loss": 5.097,
"step": 58000
},
{
"epoch": 1.15,
"learning_rate": 3.0874801360597823e-05,
"loss": 5.103,
"step": 58250
},
{
"epoch": 1.15,
"learning_rate": 3.079271896300382e-05,
"loss": 5.0778,
"step": 58500
},
{
"epoch": 1.16,
"learning_rate": 3.071063656540982e-05,
"loss": 5.0565,
"step": 58750
},
{
"epoch": 1.16,
"learning_rate": 3.062855416781582e-05,
"loss": 5.1443,
"step": 59000
},
{
"epoch": 1.17,
"learning_rate": 3.054647177022182e-05,
"loss": 5.0715,
"step": 59250
},
{
"epoch": 1.17,
"learning_rate": 3.046438937262782e-05,
"loss": 5.0871,
"step": 59500
},
{
"epoch": 1.18,
"learning_rate": 3.038230697503382e-05,
"loss": 5.0992,
"step": 59750
},
{
"epoch": 1.18,
"learning_rate": 3.0300224577439816e-05,
"loss": 5.0874,
"step": 60000
},
{
"epoch": 1.19,
"learning_rate": 3.0218142179845816e-05,
"loss": 5.1127,
"step": 60250
},
{
"epoch": 1.19,
"learning_rate": 3.013605978225182e-05,
"loss": 5.0684,
"step": 60500
},
{
"epoch": 1.2,
"learning_rate": 3.0053977384657815e-05,
"loss": 5.0856,
"step": 60750
},
{
"epoch": 1.2,
"learning_rate": 2.9971894987063814e-05,
"loss": 5.062,
"step": 61000
},
{
"epoch": 1.21,
"learning_rate": 2.988981258946981e-05,
"loss": 5.0753,
"step": 61250
},
{
"epoch": 1.21,
"learning_rate": 2.9807730191875817e-05,
"loss": 5.055,
"step": 61500
},
{
"epoch": 1.22,
"learning_rate": 2.9725647794281813e-05,
"loss": 5.0444,
"step": 61750
},
{
"epoch": 1.22,
"learning_rate": 2.9643565396687812e-05,
"loss": 5.0642,
"step": 62000
},
{
"epoch": 1.23,
"learning_rate": 2.956148299909381e-05,
"loss": 5.0566,
"step": 62250
},
{
"epoch": 1.23,
"learning_rate": 2.947940060149981e-05,
"loss": 5.0445,
"step": 62500
},
{
"epoch": 1.24,
"learning_rate": 2.939731820390581e-05,
"loss": 5.0221,
"step": 62750
},
{
"epoch": 1.24,
"learning_rate": 2.9315235806311807e-05,
"loss": 5.0624,
"step": 63000
},
{
"epoch": 1.25,
"learning_rate": 2.9233153408717807e-05,
"loss": 5.0271,
"step": 63250
},
{
"epoch": 1.25,
"learning_rate": 2.915107101112381e-05,
"loss": 5.0478,
"step": 63500
},
{
"epoch": 1.26,
"learning_rate": 2.906898861352981e-05,
"loss": 5.0316,
"step": 63750
},
{
"epoch": 1.26,
"learning_rate": 2.8986906215935805e-05,
"loss": 4.9997,
"step": 64000
},
{
"epoch": 1.27,
"learning_rate": 2.8904823818341804e-05,
"loss": 4.9997,
"step": 64250
},
{
"epoch": 1.27,
"learning_rate": 2.8822741420747807e-05,
"loss": 5.0058,
"step": 64500
},
{
"epoch": 1.28,
"learning_rate": 2.8740659023153803e-05,
"loss": 4.9975,
"step": 64750
},
{
"epoch": 1.28,
"learning_rate": 2.8658576625559803e-05,
"loss": 4.9933,
"step": 65000
},
{
"epoch": 1.29,
"learning_rate": 2.85764942279658e-05,
"loss": 4.9868,
"step": 65250
},
{
"epoch": 1.29,
"learning_rate": 2.8494411830371802e-05,
"loss": 4.981,
"step": 65500
},
{
"epoch": 1.3,
"learning_rate": 2.84123294327778e-05,
"loss": 5.0185,
"step": 65750
},
{
"epoch": 1.3,
"learning_rate": 2.8330247035183798e-05,
"loss": 4.9715,
"step": 66000
},
{
"epoch": 1.31,
"learning_rate": 2.8248164637589797e-05,
"loss": 5.0356,
"step": 66250
},
{
"epoch": 1.31,
"learning_rate": 2.81660822399958e-05,
"loss": 4.9427,
"step": 66500
},
{
"epoch": 1.31,
"learning_rate": 2.80839998424018e-05,
"loss": 4.8855,
"step": 66750
},
{
"epoch": 1.32,
"learning_rate": 2.8001917444807796e-05,
"loss": 4.9676,
"step": 67000
},
{
"epoch": 1.32,
"learning_rate": 2.7919835047213795e-05,
"loss": 4.9374,
"step": 67250
},
{
"epoch": 1.33,
"learning_rate": 2.7837752649619798e-05,
"loss": 4.9648,
"step": 67500
},
{
"epoch": 1.33,
"learning_rate": 2.7755670252025794e-05,
"loss": 4.9769,
"step": 67750
},
{
"epoch": 1.34,
"learning_rate": 2.7673587854431793e-05,
"loss": 4.8962,
"step": 68000
},
{
"epoch": 1.34,
"learning_rate": 2.759150545683779e-05,
"loss": 4.9559,
"step": 68250
},
{
"epoch": 1.35,
"learning_rate": 2.7509423059243792e-05,
"loss": 4.929,
"step": 68500
},
{
"epoch": 1.35,
"learning_rate": 2.7427340661649792e-05,
"loss": 4.946,
"step": 68750
},
{
"epoch": 1.36,
"learning_rate": 2.7345258264055788e-05,
"loss": 4.9562,
"step": 69000
},
{
"epoch": 1.36,
"learning_rate": 2.7263175866461788e-05,
"loss": 4.9309,
"step": 69250
},
{
"epoch": 1.37,
"learning_rate": 2.718109346886779e-05,
"loss": 4.9074,
"step": 69500
},
{
"epoch": 1.37,
"learning_rate": 2.709901107127379e-05,
"loss": 4.8918,
"step": 69750
},
{
"epoch": 1.38,
"learning_rate": 2.7016928673679786e-05,
"loss": 4.9262,
"step": 70000
},
{
"epoch": 1.38,
"learning_rate": 2.6934846276085786e-05,
"loss": 4.8864,
"step": 70250
},
{
"epoch": 1.39,
"learning_rate": 2.685276387849179e-05,
"loss": 4.9013,
"step": 70500
},
{
"epoch": 1.39,
"learning_rate": 2.6770681480897785e-05,
"loss": 4.8744,
"step": 70750
},
{
"epoch": 1.4,
"learning_rate": 2.6688599083303784e-05,
"loss": 4.881,
"step": 71000
},
{
"epoch": 1.4,
"learning_rate": 2.6606516685709787e-05,
"loss": 4.911,
"step": 71250
},
{
"epoch": 1.41,
"learning_rate": 2.6524434288115786e-05,
"loss": 4.8839,
"step": 71500
},
{
"epoch": 1.41,
"learning_rate": 2.6442351890521783e-05,
"loss": 4.8996,
"step": 71750
},
{
"epoch": 1.42,
"learning_rate": 2.6360269492927782e-05,
"loss": 4.9019,
"step": 72000
},
{
"epoch": 1.42,
"learning_rate": 2.6278187095333785e-05,
"loss": 4.8942,
"step": 72250
},
{
"epoch": 1.43,
"learning_rate": 2.619610469773978e-05,
"loss": 4.8242,
"step": 72500
},
{
"epoch": 1.43,
"learning_rate": 2.611402230014578e-05,
"loss": 4.8776,
"step": 72750
},
{
"epoch": 1.44,
"learning_rate": 2.6031939902551777e-05,
"loss": 4.8732,
"step": 73000
},
{
"epoch": 1.44,
"learning_rate": 2.594985750495778e-05,
"loss": 4.8902,
"step": 73250
},
{
"epoch": 1.45,
"learning_rate": 2.586777510736378e-05,
"loss": 4.8638,
"step": 73500
},
{
"epoch": 1.45,
"learning_rate": 2.5785692709769775e-05,
"loss": 4.8418,
"step": 73750
},
{
"epoch": 1.46,
"learning_rate": 2.5703610312175775e-05,
"loss": 4.8541,
"step": 74000
},
{
"epoch": 1.46,
"learning_rate": 2.5621527914581777e-05,
"loss": 4.8397,
"step": 74250
},
{
"epoch": 1.47,
"learning_rate": 2.5539445516987777e-05,
"loss": 4.8609,
"step": 74500
},
{
"epoch": 1.47,
"learning_rate": 2.5457363119393773e-05,
"loss": 4.8698,
"step": 74750
},
{
"epoch": 1.48,
"learning_rate": 2.5375280721799773e-05,
"loss": 4.88,
"step": 75000
},
{
"epoch": 1.48,
"learning_rate": 2.5293198324205775e-05,
"loss": 4.8273,
"step": 75250
},
{
"epoch": 1.49,
"learning_rate": 2.521111592661177e-05,
"loss": 4.8285,
"step": 75500
},
{
"epoch": 1.49,
"learning_rate": 2.512903352901777e-05,
"loss": 4.853,
"step": 75750
},
{
"epoch": 1.5,
"learning_rate": 2.5046951131423767e-05,
"loss": 4.8384,
"step": 76000
},
{
"epoch": 1.5,
"learning_rate": 2.496486873382977e-05,
"loss": 4.7944,
"step": 76250
},
{
"epoch": 1.51,
"learning_rate": 2.488278633623577e-05,
"loss": 4.8423,
"step": 76500
},
{
"epoch": 1.51,
"learning_rate": 2.4800703938641766e-05,
"loss": 4.797,
"step": 76750
},
{
"epoch": 1.52,
"learning_rate": 2.471862154104777e-05,
"loss": 4.7585,
"step": 77000
},
{
"epoch": 1.52,
"learning_rate": 2.4636539143453765e-05,
"loss": 4.8479,
"step": 77250
},
{
"epoch": 1.53,
"learning_rate": 2.4554456745859767e-05,
"loss": 4.7687,
"step": 77500
},
{
"epoch": 1.53,
"learning_rate": 2.4472374348265764e-05,
"loss": 4.793,
"step": 77750
},
{
"epoch": 1.54,
"learning_rate": 2.4390291950671763e-05,
"loss": 4.8033,
"step": 78000
},
{
"epoch": 1.54,
"learning_rate": 2.4308209553077763e-05,
"loss": 4.8058,
"step": 78250
},
{
"epoch": 1.55,
"learning_rate": 2.4226127155483762e-05,
"loss": 4.8193,
"step": 78500
},
{
"epoch": 1.55,
"learning_rate": 2.414404475788976e-05,
"loss": 4.8197,
"step": 78750
},
{
"epoch": 1.56,
"learning_rate": 2.406196236029576e-05,
"loss": 4.817,
"step": 79000
},
{
"epoch": 1.56,
"learning_rate": 2.397987996270176e-05,
"loss": 4.7666,
"step": 79250
},
{
"epoch": 1.57,
"learning_rate": 2.389779756510776e-05,
"loss": 4.779,
"step": 79500
},
{
"epoch": 1.57,
"learning_rate": 2.381571516751376e-05,
"loss": 4.7867,
"step": 79750
},
{
"epoch": 1.58,
"learning_rate": 2.373363276991976e-05,
"loss": 4.711,
"step": 80000
},
{
"epoch": 1.58,
"learning_rate": 2.3651550372325755e-05,
"loss": 4.764,
"step": 80250
},
{
"epoch": 1.59,
"learning_rate": 2.3569467974731758e-05,
"loss": 4.7442,
"step": 80500
},
{
"epoch": 1.59,
"learning_rate": 2.3487385577137754e-05,
"loss": 4.7251,
"step": 80750
},
{
"epoch": 1.6,
"learning_rate": 2.3405303179543757e-05,
"loss": 4.7445,
"step": 81000
},
{
"epoch": 1.6,
"learning_rate": 2.3323220781949753e-05,
"loss": 4.7737,
"step": 81250
},
{
"epoch": 1.61,
"learning_rate": 2.3241138384355753e-05,
"loss": 4.7479,
"step": 81500
},
{
"epoch": 1.61,
"learning_rate": 2.3159055986761752e-05,
"loss": 4.7482,
"step": 81750
},
{
"epoch": 1.62,
"learning_rate": 2.307697358916775e-05,
"loss": 4.7181,
"step": 82000
},
{
"epoch": 1.62,
"learning_rate": 2.299489119157375e-05,
"loss": 4.7784,
"step": 82250
},
{
"epoch": 1.63,
"learning_rate": 2.291280879397975e-05,
"loss": 4.7506,
"step": 82500
},
{
"epoch": 1.63,
"learning_rate": 2.283072639638575e-05,
"loss": 4.7269,
"step": 82750
},
{
"epoch": 1.64,
"learning_rate": 2.274864399879175e-05,
"loss": 4.7102,
"step": 83000
},
{
"epoch": 1.64,
"learning_rate": 2.2666561601197746e-05,
"loss": 4.7706,
"step": 83250
},
{
"epoch": 1.64,
"learning_rate": 2.258447920360375e-05,
"loss": 4.7343,
"step": 83500
},
{
"epoch": 1.65,
"learning_rate": 2.2502396806009745e-05,
"loss": 4.6911,
"step": 83750
},
{
"epoch": 1.65,
"learning_rate": 2.2420314408415748e-05,
"loss": 4.7585,
"step": 84000
},
{
"epoch": 1.66,
"learning_rate": 2.2338232010821744e-05,
"loss": 4.6949,
"step": 84250
},
{
"epoch": 1.66,
"learning_rate": 2.2256149613227743e-05,
"loss": 4.7423,
"step": 84500
},
{
"epoch": 1.67,
"learning_rate": 2.2174067215633743e-05,
"loss": 4.7317,
"step": 84750
},
{
"epoch": 1.67,
"learning_rate": 2.2091984818039742e-05,
"loss": 4.7393,
"step": 85000
},
{
"epoch": 1.68,
"learning_rate": 2.200990242044574e-05,
"loss": 4.7046,
"step": 85250
},
{
"epoch": 1.68,
"learning_rate": 2.192782002285174e-05,
"loss": 4.77,
"step": 85500
},
{
"epoch": 1.69,
"learning_rate": 2.184573762525774e-05,
"loss": 4.7457,
"step": 85750
},
{
"epoch": 1.69,
"learning_rate": 2.176365522766374e-05,
"loss": 4.6749,
"step": 86000
},
{
"epoch": 1.7,
"learning_rate": 2.168157283006974e-05,
"loss": 4.6972,
"step": 86250
},
{
"epoch": 1.7,
"learning_rate": 2.159949043247574e-05,
"loss": 4.6846,
"step": 86500
},
{
"epoch": 1.71,
"learning_rate": 2.1517408034881735e-05,
"loss": 4.7132,
"step": 86750
},
{
"epoch": 1.71,
"learning_rate": 2.1435325637287738e-05,
"loss": 4.6606,
"step": 87000
},
{
"epoch": 1.72,
"learning_rate": 2.1353243239693734e-05,
"loss": 4.6989,
"step": 87250
},
{
"epoch": 1.72,
"learning_rate": 2.1271160842099737e-05,
"loss": 4.7223,
"step": 87500
},
{
"epoch": 1.73,
"learning_rate": 2.1189078444505733e-05,
"loss": 4.6806,
"step": 87750
},
{
"epoch": 1.73,
"learning_rate": 2.1106996046911733e-05,
"loss": 4.6947,
"step": 88000
},
{
"epoch": 1.74,
"learning_rate": 2.1024913649317732e-05,
"loss": 4.712,
"step": 88250
},
{
"epoch": 1.74,
"learning_rate": 2.094283125172373e-05,
"loss": 4.7026,
"step": 88500
},
{
"epoch": 1.75,
"learning_rate": 2.086074885412973e-05,
"loss": 4.714,
"step": 88750
},
{
"epoch": 1.75,
"learning_rate": 2.077866645653573e-05,
"loss": 4.6463,
"step": 89000
},
{
"epoch": 1.76,
"learning_rate": 2.069658405894173e-05,
"loss": 4.6939,
"step": 89250
},
{
"epoch": 1.76,
"learning_rate": 2.061450166134773e-05,
"loss": 4.6927,
"step": 89500
},
{
"epoch": 1.77,
"learning_rate": 2.0532419263753726e-05,
"loss": 4.6733,
"step": 89750
},
{
"epoch": 1.77,
"learning_rate": 2.045033686615973e-05,
"loss": 4.6651,
"step": 90000
},
{
"epoch": 1.78,
"learning_rate": 2.0368254468565725e-05,
"loss": 4.6826,
"step": 90250
},
{
"epoch": 1.78,
"learning_rate": 2.0286172070971728e-05,
"loss": 4.6593,
"step": 90500
},
{
"epoch": 1.79,
"learning_rate": 2.0204089673377724e-05,
"loss": 4.6748,
"step": 90750
},
{
"epoch": 1.79,
"learning_rate": 2.0122007275783723e-05,
"loss": 4.6453,
"step": 91000
},
{
"epoch": 1.8,
"learning_rate": 2.0039924878189723e-05,
"loss": 4.6764,
"step": 91250
},
{
"epoch": 1.8,
"learning_rate": 1.9957842480595722e-05,
"loss": 4.6282,
"step": 91500
},
{
"epoch": 1.81,
"learning_rate": 1.987576008300172e-05,
"loss": 4.6437,
"step": 91750
},
{
"epoch": 1.81,
"learning_rate": 1.979367768540772e-05,
"loss": 4.6456,
"step": 92000
},
{
"epoch": 1.82,
"learning_rate": 1.971159528781372e-05,
"loss": 4.6155,
"step": 92250
},
{
"epoch": 1.82,
"learning_rate": 1.962951289021972e-05,
"loss": 4.609,
"step": 92500
},
{
"epoch": 1.83,
"learning_rate": 1.9547430492625716e-05,
"loss": 4.6056,
"step": 92750
},
{
"epoch": 1.83,
"learning_rate": 1.946534809503172e-05,
"loss": 4.6138,
"step": 93000
},
{
"epoch": 1.84,
"learning_rate": 1.9383265697437715e-05,
"loss": 4.6259,
"step": 93250
},
{
"epoch": 1.84,
"learning_rate": 1.9301183299843718e-05,
"loss": 4.651,
"step": 93500
},
{
"epoch": 1.85,
"learning_rate": 1.9219100902249714e-05,
"loss": 4.6287,
"step": 93750
},
{
"epoch": 1.85,
"learning_rate": 1.9137018504655717e-05,
"loss": 4.6163,
"step": 94000
},
{
"epoch": 1.86,
"learning_rate": 1.9054936107061713e-05,
"loss": 4.5989,
"step": 94250
},
{
"epoch": 1.86,
"learning_rate": 1.8972853709467713e-05,
"loss": 4.6224,
"step": 94500
},
{
"epoch": 1.87,
"learning_rate": 1.8890771311873712e-05,
"loss": 4.5913,
"step": 94750
},
{
"epoch": 1.87,
"learning_rate": 1.8808688914279712e-05,
"loss": 4.6267,
"step": 95000
},
{
"epoch": 1.88,
"learning_rate": 1.872660651668571e-05,
"loss": 4.5906,
"step": 95250
},
{
"epoch": 1.88,
"learning_rate": 1.864452411909171e-05,
"loss": 4.6433,
"step": 95500
},
{
"epoch": 1.89,
"learning_rate": 1.856244172149771e-05,
"loss": 4.6444,
"step": 95750
},
{
"epoch": 1.89,
"learning_rate": 1.848035932390371e-05,
"loss": 4.6358,
"step": 96000
},
{
"epoch": 1.9,
"learning_rate": 1.8398276926309706e-05,
"loss": 4.577,
"step": 96250
},
{
"epoch": 1.9,
"learning_rate": 1.831619452871571e-05,
"loss": 4.6281,
"step": 96500
},
{
"epoch": 1.91,
"learning_rate": 1.8234112131121705e-05,
"loss": 4.613,
"step": 96750
},
{
"epoch": 1.91,
"learning_rate": 1.8152029733527708e-05,
"loss": 4.6085,
"step": 97000
},
{
"epoch": 1.92,
"learning_rate": 1.8069947335933704e-05,
"loss": 4.6347,
"step": 97250
},
{
"epoch": 1.92,
"learning_rate": 1.7987864938339703e-05,
"loss": 4.6094,
"step": 97500
},
{
"epoch": 1.93,
"learning_rate": 1.7905782540745703e-05,
"loss": 4.6255,
"step": 97750
},
{
"epoch": 1.93,
"learning_rate": 1.7823700143151702e-05,
"loss": 4.5426,
"step": 98000
},
{
"epoch": 1.94,
"learning_rate": 1.7741617745557702e-05,
"loss": 4.594,
"step": 98250
},
{
"epoch": 1.94,
"learning_rate": 1.76595353479637e-05,
"loss": 4.5371,
"step": 98500
},
{
"epoch": 1.95,
"learning_rate": 1.75774529503697e-05,
"loss": 4.6085,
"step": 98750
},
{
"epoch": 1.95,
"learning_rate": 1.74953705527757e-05,
"loss": 4.6097,
"step": 99000
},
{
"epoch": 1.96,
"learning_rate": 1.7413288155181696e-05,
"loss": 4.5315,
"step": 99250
},
{
"epoch": 1.96,
"learning_rate": 1.73312057575877e-05,
"loss": 4.5781,
"step": 99500
},
{
"epoch": 1.97,
"learning_rate": 1.7249123359993695e-05,
"loss": 4.5609,
"step": 99750
},
{
"epoch": 1.97,
"learning_rate": 1.7167040962399698e-05,
"loss": 4.5971,
"step": 100000
},
{
"epoch": 1.97,
"learning_rate": 1.7084958564805694e-05,
"loss": 4.5534,
"step": 100250
},
{
"epoch": 1.98,
"learning_rate": 1.7002876167211694e-05,
"loss": 4.5275,
"step": 100500
},
{
"epoch": 1.98,
"learning_rate": 1.6920793769617693e-05,
"loss": 4.5443,
"step": 100750
},
{
"epoch": 1.99,
"learning_rate": 1.6838711372023693e-05,
"loss": 4.5559,
"step": 101000
},
{
"epoch": 1.99,
"learning_rate": 1.6756628974429692e-05,
"loss": 4.5289,
"step": 101250
},
{
"epoch": 2.0,
"learning_rate": 1.6674546576835692e-05,
"loss": 4.574,
"step": 101500
},
{
"epoch": 2.0,
"eval_loss": 4.450372695922852,
"eval_runtime": 619.9479,
"eval_samples_per_second": 291.173,
"eval_steps_per_second": 9.099,
"step": 101524
},
{
"epoch": 2.0,
"learning_rate": 1.659246417924169e-05,
"loss": 4.4787,
"step": 101750
},
{
"epoch": 2.01,
"learning_rate": 1.651038178164769e-05,
"loss": 4.5425,
"step": 102000
},
{
"epoch": 2.01,
"learning_rate": 1.642829938405369e-05,
"loss": 4.5304,
"step": 102250
},
{
"epoch": 2.02,
"learning_rate": 1.634621698645969e-05,
"loss": 4.5141,
"step": 102500
},
{
"epoch": 2.02,
"learning_rate": 1.6264134588865686e-05,
"loss": 4.5189,
"step": 102750
},
{
"epoch": 2.03,
"learning_rate": 1.618205219127169e-05,
"loss": 4.5433,
"step": 103000
},
{
"epoch": 2.03,
"learning_rate": 1.6099969793677685e-05,
"loss": 4.561,
"step": 103250
},
{
"epoch": 2.04,
"learning_rate": 1.6017887396083688e-05,
"loss": 4.5514,
"step": 103500
},
{
"epoch": 2.04,
"learning_rate": 1.5935804998489684e-05,
"loss": 4.5083,
"step": 103750
},
{
"epoch": 2.05,
"learning_rate": 1.5853722600895683e-05,
"loss": 4.524,
"step": 104000
},
{
"epoch": 2.05,
"learning_rate": 1.5771640203301683e-05,
"loss": 4.5076,
"step": 104250
},
{
"epoch": 2.06,
"learning_rate": 1.5689557805707682e-05,
"loss": 4.5146,
"step": 104500
},
{
"epoch": 2.06,
"learning_rate": 1.5607475408113682e-05,
"loss": 4.4952,
"step": 104750
},
{
"epoch": 2.07,
"learning_rate": 1.552539301051968e-05,
"loss": 4.513,
"step": 105000
},
{
"epoch": 2.07,
"learning_rate": 1.544331061292568e-05,
"loss": 4.5001,
"step": 105250
},
{
"epoch": 2.08,
"learning_rate": 1.536122821533168e-05,
"loss": 4.5537,
"step": 105500
},
{
"epoch": 2.08,
"learning_rate": 1.5279145817737676e-05,
"loss": 4.5184,
"step": 105750
},
{
"epoch": 2.09,
"learning_rate": 1.519706342014368e-05,
"loss": 4.4982,
"step": 106000
},
{
"epoch": 2.09,
"learning_rate": 1.5114981022549677e-05,
"loss": 4.497,
"step": 106250
},
{
"epoch": 2.1,
"learning_rate": 1.5032898624955677e-05,
"loss": 4.4993,
"step": 106500
},
{
"epoch": 2.1,
"learning_rate": 1.4950816227361674e-05,
"loss": 4.5091,
"step": 106750
},
{
"epoch": 2.11,
"learning_rate": 1.4868733829767676e-05,
"loss": 4.5167,
"step": 107000
},
{
"epoch": 2.11,
"learning_rate": 1.4786651432173673e-05,
"loss": 4.4561,
"step": 107250
},
{
"epoch": 2.12,
"learning_rate": 1.4704569034579675e-05,
"loss": 4.5323,
"step": 107500
},
{
"epoch": 2.12,
"learning_rate": 1.4622486636985672e-05,
"loss": 4.5174,
"step": 107750
},
{
"epoch": 2.13,
"learning_rate": 1.4540404239391672e-05,
"loss": 4.5102,
"step": 108000
},
{
"epoch": 2.13,
"learning_rate": 1.445832184179767e-05,
"loss": 4.4722,
"step": 108250
},
{
"epoch": 2.14,
"learning_rate": 1.437623944420367e-05,
"loss": 4.4973,
"step": 108500
},
{
"epoch": 2.14,
"learning_rate": 1.4294157046609669e-05,
"loss": 4.4885,
"step": 108750
},
{
"epoch": 2.15,
"learning_rate": 1.421207464901567e-05,
"loss": 4.4767,
"step": 109000
},
{
"epoch": 2.15,
"learning_rate": 1.4129992251421668e-05,
"loss": 4.4758,
"step": 109250
},
{
"epoch": 2.16,
"learning_rate": 1.4047909853827667e-05,
"loss": 4.4632,
"step": 109500
},
{
"epoch": 2.16,
"learning_rate": 1.3965827456233665e-05,
"loss": 4.4848,
"step": 109750
},
{
"epoch": 2.17,
"learning_rate": 1.3883745058639666e-05,
"loss": 4.4566,
"step": 110000
},
{
"epoch": 2.17,
"learning_rate": 1.3801662661045664e-05,
"loss": 4.4524,
"step": 110250
},
{
"epoch": 2.18,
"learning_rate": 1.3719580263451665e-05,
"loss": 4.448,
"step": 110500
},
{
"epoch": 2.18,
"learning_rate": 1.3637497865857663e-05,
"loss": 4.458,
"step": 110750
},
{
"epoch": 2.19,
"learning_rate": 1.3555415468263662e-05,
"loss": 4.4448,
"step": 111000
},
{
"epoch": 2.19,
"learning_rate": 1.347333307066966e-05,
"loss": 4.4719,
"step": 111250
},
{
"epoch": 2.2,
"learning_rate": 1.3391250673075661e-05,
"loss": 4.4535,
"step": 111500
},
{
"epoch": 2.2,
"learning_rate": 1.3309168275481659e-05,
"loss": 4.4429,
"step": 111750
},
{
"epoch": 2.21,
"learning_rate": 1.322708587788766e-05,
"loss": 4.4854,
"step": 112000
},
{
"epoch": 2.21,
"learning_rate": 1.3145003480293658e-05,
"loss": 4.409,
"step": 112250
},
{
"epoch": 2.22,
"learning_rate": 1.3062921082699658e-05,
"loss": 4.4335,
"step": 112500
}
],
"logging_steps": 250,
"max_steps": 152286,
"num_train_epochs": 3,
"save_steps": 2048,
"total_flos": 4.66845448223232e+17,
"trial_name": null,
"trial_params": null
}