m2m100_418M_twi_en_rel / trainer_state.json
Davlan's picture
add MT model
3726005
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 3.0,
"global_step": 176622,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.01,
"learning_rate": 4.985845477913284e-05,
"loss": 2.8944,
"step": 500
},
{
"epoch": 0.02,
"learning_rate": 4.9716909558265676e-05,
"loss": 2.4395,
"step": 1000
},
{
"epoch": 0.03,
"learning_rate": 4.957536433739852e-05,
"loss": 2.2319,
"step": 1500
},
{
"epoch": 0.03,
"learning_rate": 4.943381911653135e-05,
"loss": 2.1707,
"step": 2000
},
{
"epoch": 0.04,
"learning_rate": 4.929227389566419e-05,
"loss": 2.0619,
"step": 2500
},
{
"epoch": 0.05,
"learning_rate": 4.9150728674797024e-05,
"loss": 2.0174,
"step": 3000
},
{
"epoch": 0.06,
"learning_rate": 4.900918345392986e-05,
"loss": 1.9404,
"step": 3500
},
{
"epoch": 0.07,
"learning_rate": 4.8867638233062705e-05,
"loss": 1.9597,
"step": 4000
},
{
"epoch": 0.08,
"learning_rate": 4.872609301219554e-05,
"loss": 1.8656,
"step": 4500
},
{
"epoch": 0.08,
"learning_rate": 4.858454779132838e-05,
"loss": 1.8542,
"step": 5000
},
{
"epoch": 0.09,
"learning_rate": 4.8443002570461215e-05,
"loss": 1.8505,
"step": 5500
},
{
"epoch": 0.1,
"learning_rate": 4.8301457349594046e-05,
"loss": 1.8365,
"step": 6000
},
{
"epoch": 0.11,
"learning_rate": 4.815991212872689e-05,
"loss": 1.8022,
"step": 6500
},
{
"epoch": 0.12,
"learning_rate": 4.8018366907859726e-05,
"loss": 1.7895,
"step": 7000
},
{
"epoch": 0.13,
"learning_rate": 4.787682168699256e-05,
"loss": 1.7696,
"step": 7500
},
{
"epoch": 0.14,
"learning_rate": 4.77352764661254e-05,
"loss": 1.7488,
"step": 8000
},
{
"epoch": 0.14,
"learning_rate": 4.759373124525824e-05,
"loss": 1.7451,
"step": 8500
},
{
"epoch": 0.15,
"learning_rate": 4.7452186024391074e-05,
"loss": 1.7162,
"step": 9000
},
{
"epoch": 0.16,
"learning_rate": 4.731064080352392e-05,
"loss": 1.7143,
"step": 9500
},
{
"epoch": 0.17,
"learning_rate": 4.716909558265675e-05,
"loss": 1.6986,
"step": 10000
},
{
"epoch": 0.18,
"learning_rate": 4.7027550361789585e-05,
"loss": 1.681,
"step": 10500
},
{
"epoch": 0.19,
"learning_rate": 4.688600514092242e-05,
"loss": 1.6736,
"step": 11000
},
{
"epoch": 0.2,
"learning_rate": 4.674445992005526e-05,
"loss": 1.6508,
"step": 11500
},
{
"epoch": 0.2,
"learning_rate": 4.66029146991881e-05,
"loss": 1.6617,
"step": 12000
},
{
"epoch": 0.21,
"learning_rate": 4.646136947832094e-05,
"loss": 1.6108,
"step": 12500
},
{
"epoch": 0.22,
"learning_rate": 4.6319824257453776e-05,
"loss": 1.6545,
"step": 13000
},
{
"epoch": 0.23,
"learning_rate": 4.617827903658661e-05,
"loss": 1.6464,
"step": 13500
},
{
"epoch": 0.24,
"learning_rate": 4.603673381571944e-05,
"loss": 1.6237,
"step": 14000
},
{
"epoch": 0.25,
"learning_rate": 4.589518859485229e-05,
"loss": 1.6063,
"step": 14500
},
{
"epoch": 0.25,
"learning_rate": 4.5753643373985124e-05,
"loss": 1.588,
"step": 15000
},
{
"epoch": 0.26,
"learning_rate": 4.561209815311796e-05,
"loss": 1.5754,
"step": 15500
},
{
"epoch": 0.27,
"learning_rate": 4.54705529322508e-05,
"loss": 1.5963,
"step": 16000
},
{
"epoch": 0.28,
"learning_rate": 4.5329007711383634e-05,
"loss": 1.5919,
"step": 16500
},
{
"epoch": 0.29,
"learning_rate": 4.518746249051647e-05,
"loss": 1.5754,
"step": 17000
},
{
"epoch": 0.3,
"learning_rate": 4.504591726964931e-05,
"loss": 1.553,
"step": 17500
},
{
"epoch": 0.31,
"learning_rate": 4.4904372048782145e-05,
"loss": 1.5608,
"step": 18000
},
{
"epoch": 0.31,
"learning_rate": 4.476282682791498e-05,
"loss": 1.5672,
"step": 18500
},
{
"epoch": 0.32,
"learning_rate": 4.462128160704782e-05,
"loss": 1.5506,
"step": 19000
},
{
"epoch": 0.33,
"learning_rate": 4.4479736386180656e-05,
"loss": 1.5369,
"step": 19500
},
{
"epoch": 0.34,
"learning_rate": 4.43381911653135e-05,
"loss": 1.5178,
"step": 20000
},
{
"epoch": 0.35,
"learning_rate": 4.419664594444634e-05,
"loss": 1.5118,
"step": 20500
},
{
"epoch": 0.36,
"learning_rate": 4.4055100723579174e-05,
"loss": 1.4995,
"step": 21000
},
{
"epoch": 0.37,
"learning_rate": 4.3913555502712004e-05,
"loss": 1.5198,
"step": 21500
},
{
"epoch": 0.37,
"learning_rate": 4.377201028184485e-05,
"loss": 1.5209,
"step": 22000
},
{
"epoch": 0.38,
"learning_rate": 4.3630465060977684e-05,
"loss": 1.5126,
"step": 22500
},
{
"epoch": 0.39,
"learning_rate": 4.348891984011052e-05,
"loss": 1.4869,
"step": 23000
},
{
"epoch": 0.4,
"learning_rate": 4.334737461924336e-05,
"loss": 1.5059,
"step": 23500
},
{
"epoch": 0.41,
"learning_rate": 4.3205829398376195e-05,
"loss": 1.4804,
"step": 24000
},
{
"epoch": 0.42,
"learning_rate": 4.306428417750903e-05,
"loss": 1.4839,
"step": 24500
},
{
"epoch": 0.42,
"learning_rate": 4.2922738956641876e-05,
"loss": 1.4825,
"step": 25000
},
{
"epoch": 0.43,
"learning_rate": 4.2781193735774706e-05,
"loss": 1.4772,
"step": 25500
},
{
"epoch": 0.44,
"learning_rate": 4.263964851490754e-05,
"loss": 1.473,
"step": 26000
},
{
"epoch": 0.45,
"learning_rate": 4.249810329404038e-05,
"loss": 1.4911,
"step": 26500
},
{
"epoch": 0.46,
"learning_rate": 4.235655807317322e-05,
"loss": 1.4361,
"step": 27000
},
{
"epoch": 0.47,
"learning_rate": 4.221501285230606e-05,
"loss": 1.4403,
"step": 27500
},
{
"epoch": 0.48,
"learning_rate": 4.20734676314389e-05,
"loss": 1.4461,
"step": 28000
},
{
"epoch": 0.48,
"learning_rate": 4.1931922410571734e-05,
"loss": 1.4483,
"step": 28500
},
{
"epoch": 0.49,
"learning_rate": 4.179037718970457e-05,
"loss": 1.4453,
"step": 29000
},
{
"epoch": 0.5,
"learning_rate": 4.16488319688374e-05,
"loss": 1.4422,
"step": 29500
},
{
"epoch": 0.51,
"learning_rate": 4.1507286747970245e-05,
"loss": 1.4719,
"step": 30000
},
{
"epoch": 0.52,
"learning_rate": 4.136574152710308e-05,
"loss": 1.4151,
"step": 30500
},
{
"epoch": 0.53,
"learning_rate": 4.122419630623592e-05,
"loss": 1.4438,
"step": 31000
},
{
"epoch": 0.54,
"learning_rate": 4.1082651085368756e-05,
"loss": 1.4478,
"step": 31500
},
{
"epoch": 0.54,
"learning_rate": 4.094110586450159e-05,
"loss": 1.4229,
"step": 32000
},
{
"epoch": 0.55,
"learning_rate": 4.079956064363443e-05,
"loss": 1.4018,
"step": 32500
},
{
"epoch": 0.56,
"learning_rate": 4.0658015422767267e-05,
"loss": 1.4353,
"step": 33000
},
{
"epoch": 0.57,
"learning_rate": 4.0516470201900104e-05,
"loss": 1.4442,
"step": 33500
},
{
"epoch": 0.58,
"learning_rate": 4.037492498103294e-05,
"loss": 1.4065,
"step": 34000
},
{
"epoch": 0.59,
"learning_rate": 4.023337976016578e-05,
"loss": 1.41,
"step": 34500
},
{
"epoch": 0.59,
"learning_rate": 4.0091834539298614e-05,
"loss": 1.4075,
"step": 35000
},
{
"epoch": 0.6,
"learning_rate": 3.995028931843146e-05,
"loss": 1.4151,
"step": 35500
},
{
"epoch": 0.61,
"learning_rate": 3.9808744097564295e-05,
"loss": 1.4142,
"step": 36000
},
{
"epoch": 0.62,
"learning_rate": 3.966719887669713e-05,
"loss": 1.3852,
"step": 36500
},
{
"epoch": 0.63,
"learning_rate": 3.952565365582996e-05,
"loss": 1.3897,
"step": 37000
},
{
"epoch": 0.64,
"learning_rate": 3.93841084349628e-05,
"loss": 1.3706,
"step": 37500
},
{
"epoch": 0.65,
"learning_rate": 3.924256321409564e-05,
"loss": 1.4128,
"step": 38000
},
{
"epoch": 0.65,
"learning_rate": 3.910101799322848e-05,
"loss": 1.3818,
"step": 38500
},
{
"epoch": 0.66,
"learning_rate": 3.8959472772361316e-05,
"loss": 1.409,
"step": 39000
},
{
"epoch": 0.67,
"learning_rate": 3.8817927551494153e-05,
"loss": 1.375,
"step": 39500
},
{
"epoch": 0.68,
"learning_rate": 3.867638233062699e-05,
"loss": 1.3812,
"step": 40000
},
{
"epoch": 0.69,
"learning_rate": 3.853483710975983e-05,
"loss": 1.4002,
"step": 40500
},
{
"epoch": 0.7,
"learning_rate": 3.8393291888892664e-05,
"loss": 1.3661,
"step": 41000
},
{
"epoch": 0.7,
"learning_rate": 3.82517466680255e-05,
"loss": 1.3859,
"step": 41500
},
{
"epoch": 0.71,
"learning_rate": 3.811020144715834e-05,
"loss": 1.3628,
"step": 42000
},
{
"epoch": 0.72,
"learning_rate": 3.7968656226291175e-05,
"loss": 1.379,
"step": 42500
},
{
"epoch": 0.73,
"learning_rate": 3.782711100542401e-05,
"loss": 1.3489,
"step": 43000
},
{
"epoch": 0.74,
"learning_rate": 3.7685565784556856e-05,
"loss": 1.3694,
"step": 43500
},
{
"epoch": 0.75,
"learning_rate": 3.754402056368969e-05,
"loss": 1.3647,
"step": 44000
},
{
"epoch": 0.76,
"learning_rate": 3.740247534282253e-05,
"loss": 1.3606,
"step": 44500
},
{
"epoch": 0.76,
"learning_rate": 3.726093012195536e-05,
"loss": 1.35,
"step": 45000
},
{
"epoch": 0.77,
"learning_rate": 3.71193849010882e-05,
"loss": 1.3448,
"step": 45500
},
{
"epoch": 0.78,
"learning_rate": 3.697783968022104e-05,
"loss": 1.3484,
"step": 46000
},
{
"epoch": 0.79,
"learning_rate": 3.683629445935388e-05,
"loss": 1.3608,
"step": 46500
},
{
"epoch": 0.8,
"learning_rate": 3.6694749238486714e-05,
"loss": 1.3365,
"step": 47000
},
{
"epoch": 0.81,
"learning_rate": 3.655320401761955e-05,
"loss": 1.3694,
"step": 47500
},
{
"epoch": 0.82,
"learning_rate": 3.641165879675239e-05,
"loss": 1.3421,
"step": 48000
},
{
"epoch": 0.82,
"learning_rate": 3.6270113575885225e-05,
"loss": 1.3557,
"step": 48500
},
{
"epoch": 0.83,
"learning_rate": 3.612856835501806e-05,
"loss": 1.3413,
"step": 49000
},
{
"epoch": 0.84,
"learning_rate": 3.59870231341509e-05,
"loss": 1.3425,
"step": 49500
},
{
"epoch": 0.85,
"learning_rate": 3.5845477913283736e-05,
"loss": 1.3325,
"step": 50000
},
{
"epoch": 0.86,
"learning_rate": 3.570393269241657e-05,
"loss": 1.3516,
"step": 50500
},
{
"epoch": 0.87,
"learning_rate": 3.5562387471549416e-05,
"loss": 1.3072,
"step": 51000
},
{
"epoch": 0.87,
"learning_rate": 3.542084225068225e-05,
"loss": 1.3197,
"step": 51500
},
{
"epoch": 0.88,
"learning_rate": 3.527929702981509e-05,
"loss": 1.337,
"step": 52000
},
{
"epoch": 0.89,
"learning_rate": 3.513775180894792e-05,
"loss": 1.3209,
"step": 52500
},
{
"epoch": 0.9,
"learning_rate": 3.499620658808076e-05,
"loss": 1.3069,
"step": 53000
},
{
"epoch": 0.91,
"learning_rate": 3.48546613672136e-05,
"loss": 1.3077,
"step": 53500
},
{
"epoch": 0.92,
"learning_rate": 3.471311614634644e-05,
"loss": 1.3089,
"step": 54000
},
{
"epoch": 0.93,
"learning_rate": 3.4571570925479275e-05,
"loss": 1.3068,
"step": 54500
},
{
"epoch": 0.93,
"learning_rate": 3.443002570461211e-05,
"loss": 1.2908,
"step": 55000
},
{
"epoch": 0.94,
"learning_rate": 3.428848048374495e-05,
"loss": 1.3157,
"step": 55500
},
{
"epoch": 0.95,
"learning_rate": 3.4146935262877786e-05,
"loss": 1.3011,
"step": 56000
},
{
"epoch": 0.96,
"learning_rate": 3.400539004201062e-05,
"loss": 1.3082,
"step": 56500
},
{
"epoch": 0.97,
"learning_rate": 3.386384482114346e-05,
"loss": 1.2824,
"step": 57000
},
{
"epoch": 0.98,
"learning_rate": 3.3722299600276296e-05,
"loss": 1.3018,
"step": 57500
},
{
"epoch": 0.99,
"learning_rate": 3.358075437940913e-05,
"loss": 1.3023,
"step": 58000
},
{
"epoch": 0.99,
"learning_rate": 3.343920915854197e-05,
"loss": 1.2661,
"step": 58500
},
{
"epoch": 1.0,
"learning_rate": 3.3297663937674814e-05,
"loss": 1.2604,
"step": 59000
},
{
"epoch": 1.01,
"learning_rate": 3.315611871680765e-05,
"loss": 1.1665,
"step": 59500
},
{
"epoch": 1.02,
"learning_rate": 3.301457349594049e-05,
"loss": 1.1594,
"step": 60000
},
{
"epoch": 1.03,
"learning_rate": 3.287302827507332e-05,
"loss": 1.1756,
"step": 60500
},
{
"epoch": 1.04,
"learning_rate": 3.2731483054206155e-05,
"loss": 1.1536,
"step": 61000
},
{
"epoch": 1.04,
"learning_rate": 3.2589937833339e-05,
"loss": 1.174,
"step": 61500
},
{
"epoch": 1.05,
"learning_rate": 3.2448392612471835e-05,
"loss": 1.1602,
"step": 62000
},
{
"epoch": 1.06,
"learning_rate": 3.230684739160467e-05,
"loss": 1.159,
"step": 62500
},
{
"epoch": 1.07,
"learning_rate": 3.216530217073751e-05,
"loss": 1.1501,
"step": 63000
},
{
"epoch": 1.08,
"learning_rate": 3.2023756949870346e-05,
"loss": 1.1679,
"step": 63500
},
{
"epoch": 1.09,
"learning_rate": 3.188221172900318e-05,
"loss": 1.157,
"step": 64000
},
{
"epoch": 1.1,
"learning_rate": 3.174066650813602e-05,
"loss": 1.1697,
"step": 64500
},
{
"epoch": 1.1,
"learning_rate": 3.159912128726886e-05,
"loss": 1.1709,
"step": 65000
},
{
"epoch": 1.11,
"learning_rate": 3.1457576066401694e-05,
"loss": 1.1614,
"step": 65500
},
{
"epoch": 1.12,
"learning_rate": 3.131603084553453e-05,
"loss": 1.1775,
"step": 66000
},
{
"epoch": 1.13,
"learning_rate": 3.117448562466737e-05,
"loss": 1.163,
"step": 66500
},
{
"epoch": 1.14,
"learning_rate": 3.103294040380021e-05,
"loss": 1.1589,
"step": 67000
},
{
"epoch": 1.15,
"learning_rate": 3.089139518293305e-05,
"loss": 1.1636,
"step": 67500
},
{
"epoch": 1.16,
"learning_rate": 3.074984996206588e-05,
"loss": 1.16,
"step": 68000
},
{
"epoch": 1.16,
"learning_rate": 3.0608304741198715e-05,
"loss": 1.1608,
"step": 68500
},
{
"epoch": 1.17,
"learning_rate": 3.0466759520331556e-05,
"loss": 1.1585,
"step": 69000
},
{
"epoch": 1.18,
"learning_rate": 3.0325214299464393e-05,
"loss": 1.1426,
"step": 69500
},
{
"epoch": 1.19,
"learning_rate": 3.0183669078597233e-05,
"loss": 1.1527,
"step": 70000
},
{
"epoch": 1.2,
"learning_rate": 3.004212385773007e-05,
"loss": 1.1633,
"step": 70500
},
{
"epoch": 1.21,
"learning_rate": 2.9900578636862907e-05,
"loss": 1.1517,
"step": 71000
},
{
"epoch": 1.21,
"learning_rate": 2.9759033415995747e-05,
"loss": 1.14,
"step": 71500
},
{
"epoch": 1.22,
"learning_rate": 2.9617488195128577e-05,
"loss": 1.1512,
"step": 72000
},
{
"epoch": 1.23,
"learning_rate": 2.9475942974261418e-05,
"loss": 1.1671,
"step": 72500
},
{
"epoch": 1.24,
"learning_rate": 2.9334397753394255e-05,
"loss": 1.1556,
"step": 73000
},
{
"epoch": 1.25,
"learning_rate": 2.919285253252709e-05,
"loss": 1.1589,
"step": 73500
},
{
"epoch": 1.26,
"learning_rate": 2.9051307311659932e-05,
"loss": 1.1601,
"step": 74000
},
{
"epoch": 1.27,
"learning_rate": 2.890976209079277e-05,
"loss": 1.1532,
"step": 74500
},
{
"epoch": 1.27,
"learning_rate": 2.8768216869925606e-05,
"loss": 1.1598,
"step": 75000
},
{
"epoch": 1.28,
"learning_rate": 2.8626671649058446e-05,
"loss": 1.1276,
"step": 75500
},
{
"epoch": 1.29,
"learning_rate": 2.8485126428191276e-05,
"loss": 1.1504,
"step": 76000
},
{
"epoch": 1.3,
"learning_rate": 2.8343581207324116e-05,
"loss": 1.148,
"step": 76500
},
{
"epoch": 1.31,
"learning_rate": 2.8202035986456953e-05,
"loss": 1.1191,
"step": 77000
},
{
"epoch": 1.32,
"learning_rate": 2.806049076558979e-05,
"loss": 1.1404,
"step": 77500
},
{
"epoch": 1.32,
"learning_rate": 2.791894554472263e-05,
"loss": 1.1588,
"step": 78000
},
{
"epoch": 1.33,
"learning_rate": 2.7777400323855468e-05,
"loss": 1.1469,
"step": 78500
},
{
"epoch": 1.34,
"learning_rate": 2.7635855102988308e-05,
"loss": 1.1506,
"step": 79000
},
{
"epoch": 1.35,
"learning_rate": 2.7494309882121138e-05,
"loss": 1.1439,
"step": 79500
},
{
"epoch": 1.36,
"learning_rate": 2.7352764661253978e-05,
"loss": 1.1454,
"step": 80000
},
{
"epoch": 1.37,
"learning_rate": 2.7211219440386815e-05,
"loss": 1.124,
"step": 80500
},
{
"epoch": 1.38,
"learning_rate": 2.7069674219519652e-05,
"loss": 1.1389,
"step": 81000
},
{
"epoch": 1.38,
"learning_rate": 2.6928128998652492e-05,
"loss": 1.1187,
"step": 81500
},
{
"epoch": 1.39,
"learning_rate": 2.678658377778533e-05,
"loss": 1.137,
"step": 82000
},
{
"epoch": 1.4,
"learning_rate": 2.6645038556918166e-05,
"loss": 1.1282,
"step": 82500
},
{
"epoch": 1.41,
"learning_rate": 2.6503493336051007e-05,
"loss": 1.1296,
"step": 83000
},
{
"epoch": 1.42,
"learning_rate": 2.6361948115183837e-05,
"loss": 1.1243,
"step": 83500
},
{
"epoch": 1.43,
"learning_rate": 2.6220402894316677e-05,
"loss": 1.1032,
"step": 84000
},
{
"epoch": 1.44,
"learning_rate": 2.6078857673449514e-05,
"loss": 1.1147,
"step": 84500
},
{
"epoch": 1.44,
"learning_rate": 2.593731245258235e-05,
"loss": 1.1078,
"step": 85000
},
{
"epoch": 1.45,
"learning_rate": 2.579576723171519e-05,
"loss": 1.1214,
"step": 85500
},
{
"epoch": 1.46,
"learning_rate": 2.5654222010848028e-05,
"loss": 1.1286,
"step": 86000
},
{
"epoch": 1.47,
"learning_rate": 2.5512676789980865e-05,
"loss": 1.1465,
"step": 86500
},
{
"epoch": 1.48,
"learning_rate": 2.5371131569113705e-05,
"loss": 1.1162,
"step": 87000
},
{
"epoch": 1.49,
"learning_rate": 2.5229586348246536e-05,
"loss": 1.1207,
"step": 87500
},
{
"epoch": 1.49,
"learning_rate": 2.5088041127379376e-05,
"loss": 1.1217,
"step": 88000
},
{
"epoch": 1.5,
"learning_rate": 2.4946495906512213e-05,
"loss": 1.1229,
"step": 88500
},
{
"epoch": 1.51,
"learning_rate": 2.480495068564505e-05,
"loss": 1.1155,
"step": 89000
},
{
"epoch": 1.52,
"learning_rate": 2.466340546477789e-05,
"loss": 1.1064,
"step": 89500
},
{
"epoch": 1.53,
"learning_rate": 2.4521860243910727e-05,
"loss": 1.1406,
"step": 90000
},
{
"epoch": 1.54,
"learning_rate": 2.4380315023043564e-05,
"loss": 1.1165,
"step": 90500
},
{
"epoch": 1.55,
"learning_rate": 2.42387698021764e-05,
"loss": 1.1003,
"step": 91000
},
{
"epoch": 1.55,
"learning_rate": 2.4097224581309238e-05,
"loss": 1.138,
"step": 91500
},
{
"epoch": 1.56,
"learning_rate": 2.3955679360442078e-05,
"loss": 1.1322,
"step": 92000
},
{
"epoch": 1.57,
"learning_rate": 2.381413413957491e-05,
"loss": 1.1355,
"step": 92500
},
{
"epoch": 1.58,
"learning_rate": 2.367258891870775e-05,
"loss": 1.1083,
"step": 93000
},
{
"epoch": 1.59,
"learning_rate": 2.353104369784059e-05,
"loss": 1.1357,
"step": 93500
},
{
"epoch": 1.6,
"learning_rate": 2.3389498476973422e-05,
"loss": 1.1202,
"step": 94000
},
{
"epoch": 1.61,
"learning_rate": 2.3247953256106263e-05,
"loss": 1.1429,
"step": 94500
},
{
"epoch": 1.61,
"learning_rate": 2.31064080352391e-05,
"loss": 1.1217,
"step": 95000
},
{
"epoch": 1.62,
"learning_rate": 2.2964862814371937e-05,
"loss": 1.1065,
"step": 95500
},
{
"epoch": 1.63,
"learning_rate": 2.2823317593504773e-05,
"loss": 1.1067,
"step": 96000
},
{
"epoch": 1.64,
"learning_rate": 2.268177237263761e-05,
"loss": 1.1185,
"step": 96500
},
{
"epoch": 1.65,
"learning_rate": 2.2540227151770447e-05,
"loss": 1.1204,
"step": 97000
},
{
"epoch": 1.66,
"learning_rate": 2.2398681930903288e-05,
"loss": 1.1138,
"step": 97500
},
{
"epoch": 1.66,
"learning_rate": 2.225713671003612e-05,
"loss": 1.1063,
"step": 98000
},
{
"epoch": 1.67,
"learning_rate": 2.211559148916896e-05,
"loss": 1.1229,
"step": 98500
},
{
"epoch": 1.68,
"learning_rate": 2.19740462683018e-05,
"loss": 1.1081,
"step": 99000
},
{
"epoch": 1.69,
"learning_rate": 2.1832501047434635e-05,
"loss": 1.1201,
"step": 99500
},
{
"epoch": 1.7,
"learning_rate": 2.1690955826567472e-05,
"loss": 1.1106,
"step": 100000
},
{
"epoch": 1.71,
"learning_rate": 2.154941060570031e-05,
"loss": 1.1125,
"step": 100500
},
{
"epoch": 1.72,
"learning_rate": 2.140786538483315e-05,
"loss": 1.1001,
"step": 101000
},
{
"epoch": 1.72,
"learning_rate": 2.1266320163965986e-05,
"loss": 1.1056,
"step": 101500
},
{
"epoch": 1.73,
"learning_rate": 2.112477494309882e-05,
"loss": 1.1091,
"step": 102000
},
{
"epoch": 1.74,
"learning_rate": 2.098322972223166e-05,
"loss": 1.0946,
"step": 102500
},
{
"epoch": 1.75,
"learning_rate": 2.0841684501364497e-05,
"loss": 1.1076,
"step": 103000
},
{
"epoch": 1.76,
"learning_rate": 2.0700139280497334e-05,
"loss": 1.087,
"step": 103500
},
{
"epoch": 1.77,
"learning_rate": 2.055859405963017e-05,
"loss": 1.0725,
"step": 104000
},
{
"epoch": 1.77,
"learning_rate": 2.0417048838763008e-05,
"loss": 1.0888,
"step": 104500
},
{
"epoch": 1.78,
"learning_rate": 2.0275503617895848e-05,
"loss": 1.0725,
"step": 105000
},
{
"epoch": 1.79,
"learning_rate": 2.0133958397028685e-05,
"loss": 1.1141,
"step": 105500
},
{
"epoch": 1.8,
"learning_rate": 1.999241317616152e-05,
"loss": 1.068,
"step": 106000
},
{
"epoch": 1.81,
"learning_rate": 1.985086795529436e-05,
"loss": 1.1066,
"step": 106500
},
{
"epoch": 1.82,
"learning_rate": 1.9709322734427196e-05,
"loss": 1.0781,
"step": 107000
},
{
"epoch": 1.83,
"learning_rate": 1.9567777513560033e-05,
"loss": 1.1044,
"step": 107500
},
{
"epoch": 1.83,
"learning_rate": 1.942623229269287e-05,
"loss": 1.0917,
"step": 108000
},
{
"epoch": 1.84,
"learning_rate": 1.9284687071825707e-05,
"loss": 1.1,
"step": 108500
},
{
"epoch": 1.85,
"learning_rate": 1.9143141850958547e-05,
"loss": 1.1022,
"step": 109000
},
{
"epoch": 1.86,
"learning_rate": 1.900159663009138e-05,
"loss": 1.0836,
"step": 109500
},
{
"epoch": 1.87,
"learning_rate": 1.8860051409224218e-05,
"loss": 1.0685,
"step": 110000
},
{
"epoch": 1.88,
"learning_rate": 1.8718506188357058e-05,
"loss": 1.0973,
"step": 110500
},
{
"epoch": 1.89,
"learning_rate": 1.8576960967489895e-05,
"loss": 1.0817,
"step": 111000
},
{
"epoch": 1.89,
"learning_rate": 1.8435415746622732e-05,
"loss": 1.0591,
"step": 111500
},
{
"epoch": 1.9,
"learning_rate": 1.829387052575557e-05,
"loss": 1.0618,
"step": 112000
},
{
"epoch": 1.91,
"learning_rate": 1.8152325304888406e-05,
"loss": 1.0897,
"step": 112500
},
{
"epoch": 1.92,
"learning_rate": 1.8010780084021246e-05,
"loss": 1.0807,
"step": 113000
},
{
"epoch": 1.93,
"learning_rate": 1.786923486315408e-05,
"loss": 1.088,
"step": 113500
},
{
"epoch": 1.94,
"learning_rate": 1.772768964228692e-05,
"loss": 1.076,
"step": 114000
},
{
"epoch": 1.94,
"learning_rate": 1.7586144421419757e-05,
"loss": 1.0801,
"step": 114500
},
{
"epoch": 1.95,
"learning_rate": 1.7444599200552594e-05,
"loss": 1.0648,
"step": 115000
},
{
"epoch": 1.96,
"learning_rate": 1.730305397968543e-05,
"loss": 1.0987,
"step": 115500
},
{
"epoch": 1.97,
"learning_rate": 1.7161508758818267e-05,
"loss": 1.0721,
"step": 116000
},
{
"epoch": 1.98,
"learning_rate": 1.7019963537951104e-05,
"loss": 1.0808,
"step": 116500
},
{
"epoch": 1.99,
"learning_rate": 1.6878418317083945e-05,
"loss": 1.0698,
"step": 117000
},
{
"epoch": 2.0,
"learning_rate": 1.6736873096216778e-05,
"loss": 1.0791,
"step": 117500
},
{
"epoch": 2.0,
"learning_rate": 1.659532787534962e-05,
"loss": 1.0071,
"step": 118000
},
{
"epoch": 2.01,
"learning_rate": 1.6453782654482455e-05,
"loss": 0.9496,
"step": 118500
},
{
"epoch": 2.02,
"learning_rate": 1.6312237433615292e-05,
"loss": 0.9152,
"step": 119000
},
{
"epoch": 2.03,
"learning_rate": 1.617069221274813e-05,
"loss": 0.9348,
"step": 119500
},
{
"epoch": 2.04,
"learning_rate": 1.6029146991880966e-05,
"loss": 0.9364,
"step": 120000
},
{
"epoch": 2.05,
"learning_rate": 1.5887601771013803e-05,
"loss": 0.9292,
"step": 120500
},
{
"epoch": 2.06,
"learning_rate": 1.5746056550146643e-05,
"loss": 0.9313,
"step": 121000
},
{
"epoch": 2.06,
"learning_rate": 1.5604511329279477e-05,
"loss": 0.9426,
"step": 121500
},
{
"epoch": 2.07,
"learning_rate": 1.5462966108412317e-05,
"loss": 0.9316,
"step": 122000
},
{
"epoch": 2.08,
"learning_rate": 1.5321420887545154e-05,
"loss": 0.9448,
"step": 122500
},
{
"epoch": 2.09,
"learning_rate": 1.5179875666677993e-05,
"loss": 0.952,
"step": 123000
},
{
"epoch": 2.1,
"learning_rate": 1.5038330445810828e-05,
"loss": 0.9276,
"step": 123500
},
{
"epoch": 2.11,
"learning_rate": 1.4896785224943665e-05,
"loss": 0.9292,
"step": 124000
},
{
"epoch": 2.11,
"learning_rate": 1.4755240004076504e-05,
"loss": 0.9324,
"step": 124500
},
{
"epoch": 2.12,
"learning_rate": 1.4613694783209339e-05,
"loss": 0.9583,
"step": 125000
},
{
"epoch": 2.13,
"learning_rate": 1.4472149562342177e-05,
"loss": 0.9429,
"step": 125500
},
{
"epoch": 2.14,
"learning_rate": 1.4330604341475014e-05,
"loss": 0.9307,
"step": 126000
},
{
"epoch": 2.15,
"learning_rate": 1.4189059120607853e-05,
"loss": 0.9457,
"step": 126500
},
{
"epoch": 2.16,
"learning_rate": 1.4047513899740688e-05,
"loss": 0.9301,
"step": 127000
},
{
"epoch": 2.17,
"learning_rate": 1.3905968678873527e-05,
"loss": 0.9294,
"step": 127500
},
{
"epoch": 2.17,
"learning_rate": 1.3764423458006366e-05,
"loss": 0.9503,
"step": 128000
},
{
"epoch": 2.18,
"learning_rate": 1.3622878237139202e-05,
"loss": 0.9405,
"step": 128500
},
{
"epoch": 2.19,
"learning_rate": 1.3481333016272038e-05,
"loss": 0.9407,
"step": 129000
},
{
"epoch": 2.2,
"learning_rate": 1.3339787795404876e-05,
"loss": 0.9298,
"step": 129500
},
{
"epoch": 2.21,
"learning_rate": 1.3198242574537715e-05,
"loss": 0.9291,
"step": 130000
},
{
"epoch": 2.22,
"learning_rate": 1.3056697353670552e-05,
"loss": 0.9357,
"step": 130500
},
{
"epoch": 2.23,
"learning_rate": 1.2915152132803387e-05,
"loss": 0.9402,
"step": 131000
},
{
"epoch": 2.23,
"learning_rate": 1.2773606911936226e-05,
"loss": 0.946,
"step": 131500
},
{
"epoch": 2.24,
"learning_rate": 1.2632061691069064e-05,
"loss": 0.9283,
"step": 132000
},
{
"epoch": 2.25,
"learning_rate": 1.24905164702019e-05,
"loss": 0.9265,
"step": 132500
},
{
"epoch": 2.26,
"learning_rate": 1.2348971249334738e-05,
"loss": 0.9375,
"step": 133000
},
{
"epoch": 2.27,
"learning_rate": 1.2207426028467575e-05,
"loss": 0.9464,
"step": 133500
},
{
"epoch": 2.28,
"learning_rate": 1.2065880807600414e-05,
"loss": 0.912,
"step": 134000
},
{
"epoch": 2.28,
"learning_rate": 1.192433558673325e-05,
"loss": 0.9494,
"step": 134500
},
{
"epoch": 2.29,
"learning_rate": 1.1782790365866088e-05,
"loss": 0.9518,
"step": 135000
},
{
"epoch": 2.3,
"learning_rate": 1.1641245144998924e-05,
"loss": 0.9338,
"step": 135500
},
{
"epoch": 2.31,
"learning_rate": 1.1499699924131763e-05,
"loss": 0.94,
"step": 136000
},
{
"epoch": 2.32,
"learning_rate": 1.13581547032646e-05,
"loss": 0.9358,
"step": 136500
},
{
"epoch": 2.33,
"learning_rate": 1.1216609482397437e-05,
"loss": 0.909,
"step": 137000
},
{
"epoch": 2.34,
"learning_rate": 1.1075064261530274e-05,
"loss": 0.9448,
"step": 137500
},
{
"epoch": 2.34,
"learning_rate": 1.0933519040663112e-05,
"loss": 0.9145,
"step": 138000
},
{
"epoch": 2.35,
"learning_rate": 1.079197381979595e-05,
"loss": 0.9152,
"step": 138500
},
{
"epoch": 2.36,
"learning_rate": 1.0650428598928786e-05,
"loss": 0.9417,
"step": 139000
},
{
"epoch": 2.37,
"learning_rate": 1.0508883378061623e-05,
"loss": 0.9386,
"step": 139500
},
{
"epoch": 2.38,
"learning_rate": 1.0367338157194462e-05,
"loss": 0.9299,
"step": 140000
},
{
"epoch": 2.39,
"learning_rate": 1.0225792936327299e-05,
"loss": 0.9149,
"step": 140500
},
{
"epoch": 2.39,
"learning_rate": 1.0084247715460136e-05,
"loss": 0.9211,
"step": 141000
},
{
"epoch": 2.4,
"learning_rate": 9.942702494592973e-06,
"loss": 0.9246,
"step": 141500
},
{
"epoch": 2.41,
"learning_rate": 9.80115727372581e-06,
"loss": 0.9388,
"step": 142000
},
{
"epoch": 2.42,
"learning_rate": 9.659612052858648e-06,
"loss": 0.934,
"step": 142500
},
{
"epoch": 2.43,
"learning_rate": 9.518066831991485e-06,
"loss": 0.9138,
"step": 143000
},
{
"epoch": 2.44,
"learning_rate": 9.376521611124322e-06,
"loss": 0.9336,
"step": 143500
},
{
"epoch": 2.45,
"learning_rate": 9.234976390257159e-06,
"loss": 0.9482,
"step": 144000
},
{
"epoch": 2.45,
"learning_rate": 9.093431169389998e-06,
"loss": 0.9201,
"step": 144500
},
{
"epoch": 2.46,
"learning_rate": 8.951885948522835e-06,
"loss": 0.9133,
"step": 145000
},
{
"epoch": 2.47,
"learning_rate": 8.810340727655671e-06,
"loss": 0.9247,
"step": 145500
},
{
"epoch": 2.48,
"learning_rate": 8.668795506788508e-06,
"loss": 0.9263,
"step": 146000
},
{
"epoch": 2.49,
"learning_rate": 8.527250285921347e-06,
"loss": 0.9105,
"step": 146500
},
{
"epoch": 2.5,
"learning_rate": 8.385705065054184e-06,
"loss": 0.9171,
"step": 147000
},
{
"epoch": 2.51,
"learning_rate": 8.244159844187021e-06,
"loss": 0.9121,
"step": 147500
},
{
"epoch": 2.51,
"learning_rate": 8.102614623319858e-06,
"loss": 0.9187,
"step": 148000
},
{
"epoch": 2.52,
"learning_rate": 7.961069402452696e-06,
"loss": 0.9277,
"step": 148500
},
{
"epoch": 2.53,
"learning_rate": 7.819524181585533e-06,
"loss": 0.92,
"step": 149000
},
{
"epoch": 2.54,
"learning_rate": 7.67797896071837e-06,
"loss": 0.9226,
"step": 149500
},
{
"epoch": 2.55,
"learning_rate": 7.536433739851208e-06,
"loss": 0.9193,
"step": 150000
},
{
"epoch": 2.56,
"learning_rate": 7.394888518984046e-06,
"loss": 0.9084,
"step": 150500
},
{
"epoch": 2.56,
"learning_rate": 7.253343298116883e-06,
"loss": 0.9122,
"step": 151000
},
{
"epoch": 2.57,
"learning_rate": 7.1117980772497205e-06,
"loss": 0.9314,
"step": 151500
},
{
"epoch": 2.58,
"learning_rate": 6.9702528563825574e-06,
"loss": 0.9201,
"step": 152000
},
{
"epoch": 2.59,
"learning_rate": 6.828707635515395e-06,
"loss": 0.9192,
"step": 152500
},
{
"epoch": 2.6,
"learning_rate": 6.687162414648232e-06,
"loss": 0.9146,
"step": 153000
},
{
"epoch": 2.61,
"learning_rate": 6.54561719378107e-06,
"loss": 0.9102,
"step": 153500
},
{
"epoch": 2.62,
"learning_rate": 6.404071972913907e-06,
"loss": 0.9089,
"step": 154000
},
{
"epoch": 2.62,
"learning_rate": 6.262526752046745e-06,
"loss": 0.8985,
"step": 154500
},
{
"epoch": 2.63,
"learning_rate": 6.1209815311795815e-06,
"loss": 0.9025,
"step": 155000
},
{
"epoch": 2.64,
"learning_rate": 5.979436310312419e-06,
"loss": 0.9224,
"step": 155500
},
{
"epoch": 2.65,
"learning_rate": 5.837891089445256e-06,
"loss": 0.9171,
"step": 156000
},
{
"epoch": 2.66,
"learning_rate": 5.696345868578094e-06,
"loss": 0.9075,
"step": 156500
},
{
"epoch": 2.67,
"learning_rate": 5.554800647710931e-06,
"loss": 0.9062,
"step": 157000
},
{
"epoch": 2.68,
"learning_rate": 5.413255426843769e-06,
"loss": 0.9235,
"step": 157500
},
{
"epoch": 2.68,
"learning_rate": 5.271710205976606e-06,
"loss": 0.9049,
"step": 158000
},
{
"epoch": 2.69,
"learning_rate": 5.130164985109443e-06,
"loss": 0.8954,
"step": 158500
},
{
"epoch": 2.7,
"learning_rate": 4.98861976424228e-06,
"loss": 0.8983,
"step": 159000
},
{
"epoch": 2.71,
"learning_rate": 4.847074543375118e-06,
"loss": 0.907,
"step": 159500
},
{
"epoch": 2.72,
"learning_rate": 4.705529322507955e-06,
"loss": 0.9124,
"step": 160000
},
{
"epoch": 2.73,
"learning_rate": 4.563984101640793e-06,
"loss": 0.901,
"step": 160500
},
{
"epoch": 2.73,
"learning_rate": 4.42243888077363e-06,
"loss": 0.8981,
"step": 161000
},
{
"epoch": 2.74,
"learning_rate": 4.2808936599064675e-06,
"loss": 0.8991,
"step": 161500
},
{
"epoch": 2.75,
"learning_rate": 4.139348439039304e-06,
"loss": 0.9339,
"step": 162000
},
{
"epoch": 2.76,
"learning_rate": 3.997803218172142e-06,
"loss": 0.8987,
"step": 162500
},
{
"epoch": 2.77,
"learning_rate": 3.856257997304979e-06,
"loss": 0.9018,
"step": 163000
},
{
"epoch": 2.78,
"learning_rate": 3.714712776437817e-06,
"loss": 0.8967,
"step": 163500
},
{
"epoch": 2.79,
"learning_rate": 3.5731675555706534e-06,
"loss": 0.8911,
"step": 164000
},
{
"epoch": 2.79,
"learning_rate": 3.4316223347034907e-06,
"loss": 0.905,
"step": 164500
},
{
"epoch": 2.8,
"learning_rate": 3.2900771138363285e-06,
"loss": 0.9041,
"step": 165000
},
{
"epoch": 2.81,
"learning_rate": 3.148531892969166e-06,
"loss": 0.9099,
"step": 165500
},
{
"epoch": 2.82,
"learning_rate": 3.006986672102003e-06,
"loss": 0.9018,
"step": 166000
},
{
"epoch": 2.83,
"learning_rate": 2.865441451234841e-06,
"loss": 0.8966,
"step": 166500
},
{
"epoch": 2.84,
"learning_rate": 2.7238962303676783e-06,
"loss": 0.9009,
"step": 167000
},
{
"epoch": 2.85,
"learning_rate": 2.5823510095005157e-06,
"loss": 0.8879,
"step": 167500
},
{
"epoch": 2.85,
"learning_rate": 2.4408057886333526e-06,
"loss": 0.8921,
"step": 168000
},
{
"epoch": 2.86,
"learning_rate": 2.29926056776619e-06,
"loss": 0.8915,
"step": 168500
},
{
"epoch": 2.87,
"learning_rate": 2.1577153468990273e-06,
"loss": 0.8915,
"step": 169000
},
{
"epoch": 2.88,
"learning_rate": 2.0161701260318647e-06,
"loss": 0.8864,
"step": 169500
},
{
"epoch": 2.89,
"learning_rate": 1.874624905164702e-06,
"loss": 0.9067,
"step": 170000
},
{
"epoch": 2.9,
"learning_rate": 1.7330796842975394e-06,
"loss": 0.8879,
"step": 170500
},
{
"epoch": 2.9,
"learning_rate": 1.5915344634303767e-06,
"loss": 0.884,
"step": 171000
},
{
"epoch": 2.91,
"learning_rate": 1.4499892425632143e-06,
"loss": 0.9025,
"step": 171500
},
{
"epoch": 2.92,
"learning_rate": 1.3084440216960516e-06,
"loss": 0.8972,
"step": 172000
},
{
"epoch": 2.93,
"learning_rate": 1.166898800828889e-06,
"loss": 0.9099,
"step": 172500
},
{
"epoch": 2.94,
"learning_rate": 1.0253535799617263e-06,
"loss": 0.909,
"step": 173000
},
{
"epoch": 2.95,
"learning_rate": 8.838083590945634e-07,
"loss": 0.8996,
"step": 173500
},
{
"epoch": 2.96,
"learning_rate": 7.422631382274009e-07,
"loss": 0.9133,
"step": 174000
},
{
"epoch": 2.96,
"learning_rate": 6.007179173602383e-07,
"loss": 0.8987,
"step": 174500
},
{
"epoch": 2.97,
"learning_rate": 4.591726964930756e-07,
"loss": 0.9038,
"step": 175000
},
{
"epoch": 2.98,
"learning_rate": 3.17627475625913e-07,
"loss": 0.9018,
"step": 175500
},
{
"epoch": 2.99,
"learning_rate": 1.7608225475875035e-07,
"loss": 0.9129,
"step": 176000
},
{
"epoch": 3.0,
"learning_rate": 3.453703389158769e-08,
"loss": 0.8882,
"step": 176500
},
{
"epoch": 3.0,
"step": 176622,
"total_flos": 3.2538865963524096e+17,
"train_loss": 1.1883046192680158,
"train_runtime": 49689.8332,
"train_samples_per_second": 35.545,
"train_steps_per_second": 3.554
}
],
"max_steps": 176622,
"num_train_epochs": 3,
"total_flos": 3.2538865963524096e+17,
"trial_name": null,
"trial_params": null
}