m2m100_418M_yor_en_rel / trainer_state.json
Davlan's picture
add MT model
6bb6aad
raw
history blame
34.3 kB
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 3.0,
"global_step": 137964,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.01,
"learning_rate": 4.9818793308399294e-05,
"loss": 3.0397,
"step": 500
},
{
"epoch": 0.02,
"learning_rate": 4.9637586616798586e-05,
"loss": 2.5493,
"step": 1000
},
{
"epoch": 0.03,
"learning_rate": 4.9456379925197885e-05,
"loss": 2.3748,
"step": 1500
},
{
"epoch": 0.04,
"learning_rate": 4.927517323359717e-05,
"loss": 2.2477,
"step": 2000
},
{
"epoch": 0.05,
"learning_rate": 4.909396654199647e-05,
"loss": 2.1911,
"step": 2500
},
{
"epoch": 0.07,
"learning_rate": 4.8912759850395754e-05,
"loss": 2.1259,
"step": 3000
},
{
"epoch": 0.08,
"learning_rate": 4.873155315879505e-05,
"loss": 2.075,
"step": 3500
},
{
"epoch": 0.09,
"learning_rate": 4.855034646719434e-05,
"loss": 2.0118,
"step": 4000
},
{
"epoch": 0.1,
"learning_rate": 4.836913977559364e-05,
"loss": 2.0433,
"step": 4500
},
{
"epoch": 0.11,
"learning_rate": 4.818793308399292e-05,
"loss": 2.0111,
"step": 5000
},
{
"epoch": 0.12,
"learning_rate": 4.800672639239222e-05,
"loss": 1.9897,
"step": 5500
},
{
"epoch": 0.13,
"learning_rate": 4.782551970079151e-05,
"loss": 1.9487,
"step": 6000
},
{
"epoch": 0.14,
"learning_rate": 4.7644313009190805e-05,
"loss": 1.9383,
"step": 6500
},
{
"epoch": 0.15,
"learning_rate": 4.74631063175901e-05,
"loss": 1.9352,
"step": 7000
},
{
"epoch": 0.16,
"learning_rate": 4.728189962598939e-05,
"loss": 1.9031,
"step": 7500
},
{
"epoch": 0.17,
"learning_rate": 4.710069293438869e-05,
"loss": 1.8487,
"step": 8000
},
{
"epoch": 0.18,
"learning_rate": 4.691948624278797e-05,
"loss": 1.8532,
"step": 8500
},
{
"epoch": 0.2,
"learning_rate": 4.673827955118727e-05,
"loss": 1.8631,
"step": 9000
},
{
"epoch": 0.21,
"learning_rate": 4.655707285958656e-05,
"loss": 1.8319,
"step": 9500
},
{
"epoch": 0.22,
"learning_rate": 4.6375866167985856e-05,
"loss": 1.8194,
"step": 10000
},
{
"epoch": 0.23,
"learning_rate": 4.619465947638514e-05,
"loss": 1.8026,
"step": 10500
},
{
"epoch": 0.24,
"learning_rate": 4.601345278478444e-05,
"loss": 1.7736,
"step": 11000
},
{
"epoch": 0.25,
"learning_rate": 4.583224609318373e-05,
"loss": 1.7913,
"step": 11500
},
{
"epoch": 0.26,
"learning_rate": 4.5651039401583024e-05,
"loss": 1.758,
"step": 12000
},
{
"epoch": 0.27,
"learning_rate": 4.5469832709982316e-05,
"loss": 1.7428,
"step": 12500
},
{
"epoch": 0.28,
"learning_rate": 4.528862601838161e-05,
"loss": 1.7607,
"step": 13000
},
{
"epoch": 0.29,
"learning_rate": 4.5107419326780906e-05,
"loss": 1.7452,
"step": 13500
},
{
"epoch": 0.3,
"learning_rate": 4.492621263518019e-05,
"loss": 1.7378,
"step": 14000
},
{
"epoch": 0.32,
"learning_rate": 4.474500594357949e-05,
"loss": 1.7311,
"step": 14500
},
{
"epoch": 0.33,
"learning_rate": 4.4563799251978775e-05,
"loss": 1.6998,
"step": 15000
},
{
"epoch": 0.34,
"learning_rate": 4.4382592560378074e-05,
"loss": 1.708,
"step": 15500
},
{
"epoch": 0.35,
"learning_rate": 4.420138586877736e-05,
"loss": 1.7042,
"step": 16000
},
{
"epoch": 0.36,
"learning_rate": 4.402017917717666e-05,
"loss": 1.6764,
"step": 16500
},
{
"epoch": 0.37,
"learning_rate": 4.383897248557595e-05,
"loss": 1.674,
"step": 17000
},
{
"epoch": 0.38,
"learning_rate": 4.365776579397524e-05,
"loss": 1.6876,
"step": 17500
},
{
"epoch": 0.39,
"learning_rate": 4.3476559102374534e-05,
"loss": 1.6641,
"step": 18000
},
{
"epoch": 0.4,
"learning_rate": 4.3295352410773826e-05,
"loss": 1.6679,
"step": 18500
},
{
"epoch": 0.41,
"learning_rate": 4.311414571917312e-05,
"loss": 1.6863,
"step": 19000
},
{
"epoch": 0.42,
"learning_rate": 4.293293902757241e-05,
"loss": 1.6857,
"step": 19500
},
{
"epoch": 0.43,
"learning_rate": 4.275173233597171e-05,
"loss": 1.6609,
"step": 20000
},
{
"epoch": 0.45,
"learning_rate": 4.2570525644370994e-05,
"loss": 1.6381,
"step": 20500
},
{
"epoch": 0.46,
"learning_rate": 4.238931895277029e-05,
"loss": 1.6463,
"step": 21000
},
{
"epoch": 0.47,
"learning_rate": 4.220811226116958e-05,
"loss": 1.6395,
"step": 21500
},
{
"epoch": 0.48,
"learning_rate": 4.202690556956888e-05,
"loss": 1.6471,
"step": 22000
},
{
"epoch": 0.49,
"learning_rate": 4.184569887796817e-05,
"loss": 1.6154,
"step": 22500
},
{
"epoch": 0.5,
"learning_rate": 4.166449218636746e-05,
"loss": 1.6226,
"step": 23000
},
{
"epoch": 0.51,
"learning_rate": 4.148328549476675e-05,
"loss": 1.6032,
"step": 23500
},
{
"epoch": 0.52,
"learning_rate": 4.1302078803166045e-05,
"loss": 1.6041,
"step": 24000
},
{
"epoch": 0.53,
"learning_rate": 4.112087211156534e-05,
"loss": 1.5877,
"step": 24500
},
{
"epoch": 0.54,
"learning_rate": 4.093966541996463e-05,
"loss": 1.5951,
"step": 25000
},
{
"epoch": 0.55,
"learning_rate": 4.075845872836393e-05,
"loss": 1.584,
"step": 25500
},
{
"epoch": 0.57,
"learning_rate": 4.057725203676321e-05,
"loss": 1.5974,
"step": 26000
},
{
"epoch": 0.58,
"learning_rate": 4.039604534516251e-05,
"loss": 1.627,
"step": 26500
},
{
"epoch": 0.59,
"learning_rate": 4.0214838653561797e-05,
"loss": 1.5767,
"step": 27000
},
{
"epoch": 0.6,
"learning_rate": 4.0033631961961095e-05,
"loss": 1.5782,
"step": 27500
},
{
"epoch": 0.61,
"learning_rate": 3.985242527036039e-05,
"loss": 1.5793,
"step": 28000
},
{
"epoch": 0.62,
"learning_rate": 3.967121857875968e-05,
"loss": 1.5726,
"step": 28500
},
{
"epoch": 0.63,
"learning_rate": 3.949001188715897e-05,
"loss": 1.5635,
"step": 29000
},
{
"epoch": 0.64,
"learning_rate": 3.930880519555826e-05,
"loss": 1.5677,
"step": 29500
},
{
"epoch": 0.65,
"learning_rate": 3.9127598503957555e-05,
"loss": 1.5466,
"step": 30000
},
{
"epoch": 0.66,
"learning_rate": 3.894639181235685e-05,
"loss": 1.5574,
"step": 30500
},
{
"epoch": 0.67,
"learning_rate": 3.876518512075614e-05,
"loss": 1.5535,
"step": 31000
},
{
"epoch": 0.68,
"learning_rate": 3.858397842915543e-05,
"loss": 1.5533,
"step": 31500
},
{
"epoch": 0.7,
"learning_rate": 3.840277173755473e-05,
"loss": 1.5524,
"step": 32000
},
{
"epoch": 0.71,
"learning_rate": 3.8221565045954015e-05,
"loss": 1.5348,
"step": 32500
},
{
"epoch": 0.72,
"learning_rate": 3.8040358354353314e-05,
"loss": 1.5382,
"step": 33000
},
{
"epoch": 0.73,
"learning_rate": 3.7859151662752606e-05,
"loss": 1.5507,
"step": 33500
},
{
"epoch": 0.74,
"learning_rate": 3.76779449711519e-05,
"loss": 1.5265,
"step": 34000
},
{
"epoch": 0.75,
"learning_rate": 3.749673827955119e-05,
"loss": 1.5142,
"step": 34500
},
{
"epoch": 0.76,
"learning_rate": 3.731553158795048e-05,
"loss": 1.5225,
"step": 35000
},
{
"epoch": 0.77,
"learning_rate": 3.7134324896349774e-05,
"loss": 1.5099,
"step": 35500
},
{
"epoch": 0.78,
"learning_rate": 3.6953118204749066e-05,
"loss": 1.5242,
"step": 36000
},
{
"epoch": 0.79,
"learning_rate": 3.677191151314836e-05,
"loss": 1.4989,
"step": 36500
},
{
"epoch": 0.8,
"learning_rate": 3.659070482154765e-05,
"loss": 1.4969,
"step": 37000
},
{
"epoch": 0.82,
"learning_rate": 3.640949812994694e-05,
"loss": 1.5224,
"step": 37500
},
{
"epoch": 0.83,
"learning_rate": 3.6228291438346234e-05,
"loss": 1.5303,
"step": 38000
},
{
"epoch": 0.84,
"learning_rate": 3.604708474674553e-05,
"loss": 1.4989,
"step": 38500
},
{
"epoch": 0.85,
"learning_rate": 3.5865878055144824e-05,
"loss": 1.4963,
"step": 39000
},
{
"epoch": 0.86,
"learning_rate": 3.5684671363544116e-05,
"loss": 1.5205,
"step": 39500
},
{
"epoch": 0.87,
"learning_rate": 3.550346467194341e-05,
"loss": 1.4928,
"step": 40000
},
{
"epoch": 0.88,
"learning_rate": 3.53222579803427e-05,
"loss": 1.4704,
"step": 40500
},
{
"epoch": 0.89,
"learning_rate": 3.514105128874199e-05,
"loss": 1.4674,
"step": 41000
},
{
"epoch": 0.9,
"learning_rate": 3.4959844597141284e-05,
"loss": 1.4732,
"step": 41500
},
{
"epoch": 0.91,
"learning_rate": 3.4778637905540576e-05,
"loss": 1.4808,
"step": 42000
},
{
"epoch": 0.92,
"learning_rate": 3.459743121393987e-05,
"loss": 1.4538,
"step": 42500
},
{
"epoch": 0.94,
"learning_rate": 3.441622452233916e-05,
"loss": 1.4721,
"step": 43000
},
{
"epoch": 0.95,
"learning_rate": 3.423501783073845e-05,
"loss": 1.4908,
"step": 43500
},
{
"epoch": 0.96,
"learning_rate": 3.405381113913775e-05,
"loss": 1.461,
"step": 44000
},
{
"epoch": 0.97,
"learning_rate": 3.387260444753704e-05,
"loss": 1.4764,
"step": 44500
},
{
"epoch": 0.98,
"learning_rate": 3.3691397755936335e-05,
"loss": 1.472,
"step": 45000
},
{
"epoch": 0.99,
"learning_rate": 3.351019106433563e-05,
"loss": 1.4864,
"step": 45500
},
{
"epoch": 1.0,
"learning_rate": 3.332898437273492e-05,
"loss": 1.4507,
"step": 46000
},
{
"epoch": 1.01,
"learning_rate": 3.314777768113421e-05,
"loss": 1.3207,
"step": 46500
},
{
"epoch": 1.02,
"learning_rate": 3.29665709895335e-05,
"loss": 1.3122,
"step": 47000
},
{
"epoch": 1.03,
"learning_rate": 3.2785364297932795e-05,
"loss": 1.312,
"step": 47500
},
{
"epoch": 1.04,
"learning_rate": 3.260415760633209e-05,
"loss": 1.3318,
"step": 48000
},
{
"epoch": 1.05,
"learning_rate": 3.242295091473138e-05,
"loss": 1.3366,
"step": 48500
},
{
"epoch": 1.07,
"learning_rate": 3.224174422313067e-05,
"loss": 1.3078,
"step": 49000
},
{
"epoch": 1.08,
"learning_rate": 3.206053753152996e-05,
"loss": 1.329,
"step": 49500
},
{
"epoch": 1.09,
"learning_rate": 3.187933083992926e-05,
"loss": 1.3233,
"step": 50000
},
{
"epoch": 1.1,
"learning_rate": 3.1698124148328554e-05,
"loss": 1.3229,
"step": 50500
},
{
"epoch": 1.11,
"learning_rate": 3.1516917456727846e-05,
"loss": 1.3238,
"step": 51000
},
{
"epoch": 1.12,
"learning_rate": 3.133571076512714e-05,
"loss": 1.289,
"step": 51500
},
{
"epoch": 1.13,
"learning_rate": 3.115450407352643e-05,
"loss": 1.3107,
"step": 52000
},
{
"epoch": 1.14,
"learning_rate": 3.097329738192572e-05,
"loss": 1.315,
"step": 52500
},
{
"epoch": 1.15,
"learning_rate": 3.0792090690325014e-05,
"loss": 1.304,
"step": 53000
},
{
"epoch": 1.16,
"learning_rate": 3.0610883998724306e-05,
"loss": 1.2727,
"step": 53500
},
{
"epoch": 1.17,
"learning_rate": 3.04296773071236e-05,
"loss": 1.3072,
"step": 54000
},
{
"epoch": 1.19,
"learning_rate": 3.024847061552289e-05,
"loss": 1.3186,
"step": 54500
},
{
"epoch": 1.2,
"learning_rate": 3.0067263923922185e-05,
"loss": 1.2975,
"step": 55000
},
{
"epoch": 1.21,
"learning_rate": 2.988605723232148e-05,
"loss": 1.2986,
"step": 55500
},
{
"epoch": 1.22,
"learning_rate": 2.970485054072077e-05,
"loss": 1.3181,
"step": 56000
},
{
"epoch": 1.23,
"learning_rate": 2.9523643849120064e-05,
"loss": 1.3272,
"step": 56500
},
{
"epoch": 1.24,
"learning_rate": 2.9342437157519353e-05,
"loss": 1.2984,
"step": 57000
},
{
"epoch": 1.25,
"learning_rate": 2.9161230465918648e-05,
"loss": 1.3133,
"step": 57500
},
{
"epoch": 1.26,
"learning_rate": 2.8980023774317937e-05,
"loss": 1.2954,
"step": 58000
},
{
"epoch": 1.27,
"learning_rate": 2.8798817082717232e-05,
"loss": 1.2929,
"step": 58500
},
{
"epoch": 1.28,
"learning_rate": 2.8617610391116524e-05,
"loss": 1.3,
"step": 59000
},
{
"epoch": 1.29,
"learning_rate": 2.843640369951582e-05,
"loss": 1.3084,
"step": 59500
},
{
"epoch": 1.3,
"learning_rate": 2.8255197007915108e-05,
"loss": 1.3143,
"step": 60000
},
{
"epoch": 1.32,
"learning_rate": 2.8073990316314403e-05,
"loss": 1.2952,
"step": 60500
},
{
"epoch": 1.33,
"learning_rate": 2.7892783624713692e-05,
"loss": 1.2866,
"step": 61000
},
{
"epoch": 1.34,
"learning_rate": 2.7711576933112987e-05,
"loss": 1.2821,
"step": 61500
},
{
"epoch": 1.35,
"learning_rate": 2.7530370241512283e-05,
"loss": 1.2699,
"step": 62000
},
{
"epoch": 1.36,
"learning_rate": 2.734916354991157e-05,
"loss": 1.2864,
"step": 62500
},
{
"epoch": 1.37,
"learning_rate": 2.7167956858310867e-05,
"loss": 1.3023,
"step": 63000
},
{
"epoch": 1.38,
"learning_rate": 2.6986750166710155e-05,
"loss": 1.291,
"step": 63500
},
{
"epoch": 1.39,
"learning_rate": 2.680554347510945e-05,
"loss": 1.2995,
"step": 64000
},
{
"epoch": 1.4,
"learning_rate": 2.662433678350874e-05,
"loss": 1.3033,
"step": 64500
},
{
"epoch": 1.41,
"learning_rate": 2.6443130091908035e-05,
"loss": 1.2868,
"step": 65000
},
{
"epoch": 1.42,
"learning_rate": 2.6261923400307327e-05,
"loss": 1.2849,
"step": 65500
},
{
"epoch": 1.44,
"learning_rate": 2.6080716708706622e-05,
"loss": 1.2687,
"step": 66000
},
{
"epoch": 1.45,
"learning_rate": 2.589951001710591e-05,
"loss": 1.2862,
"step": 66500
},
{
"epoch": 1.46,
"learning_rate": 2.5718303325505206e-05,
"loss": 1.2823,
"step": 67000
},
{
"epoch": 1.47,
"learning_rate": 2.55370966339045e-05,
"loss": 1.2911,
"step": 67500
},
{
"epoch": 1.48,
"learning_rate": 2.535588994230379e-05,
"loss": 1.2775,
"step": 68000
},
{
"epoch": 1.49,
"learning_rate": 2.5174683250703085e-05,
"loss": 1.2778,
"step": 68500
},
{
"epoch": 1.5,
"learning_rate": 2.4993476559102374e-05,
"loss": 1.2963,
"step": 69000
},
{
"epoch": 1.51,
"learning_rate": 2.481226986750167e-05,
"loss": 1.2694,
"step": 69500
},
{
"epoch": 1.52,
"learning_rate": 2.463106317590096e-05,
"loss": 1.2634,
"step": 70000
},
{
"epoch": 1.53,
"learning_rate": 2.4449856484300253e-05,
"loss": 1.2821,
"step": 70500
},
{
"epoch": 1.54,
"learning_rate": 2.4268649792699545e-05,
"loss": 1.2781,
"step": 71000
},
{
"epoch": 1.55,
"learning_rate": 2.4087443101098837e-05,
"loss": 1.2473,
"step": 71500
},
{
"epoch": 1.57,
"learning_rate": 2.3906236409498133e-05,
"loss": 1.2575,
"step": 72000
},
{
"epoch": 1.58,
"learning_rate": 2.3725029717897425e-05,
"loss": 1.2691,
"step": 72500
},
{
"epoch": 1.59,
"learning_rate": 2.3543823026296717e-05,
"loss": 1.2604,
"step": 73000
},
{
"epoch": 1.6,
"learning_rate": 2.336261633469601e-05,
"loss": 1.2459,
"step": 73500
},
{
"epoch": 1.61,
"learning_rate": 2.31814096430953e-05,
"loss": 1.2452,
"step": 74000
},
{
"epoch": 1.62,
"learning_rate": 2.3000202951494593e-05,
"loss": 1.2439,
"step": 74500
},
{
"epoch": 1.63,
"learning_rate": 2.2818996259893888e-05,
"loss": 1.2652,
"step": 75000
},
{
"epoch": 1.64,
"learning_rate": 2.263778956829318e-05,
"loss": 1.2783,
"step": 75500
},
{
"epoch": 1.65,
"learning_rate": 2.2456582876692472e-05,
"loss": 1.2636,
"step": 76000
},
{
"epoch": 1.66,
"learning_rate": 2.2275376185091764e-05,
"loss": 1.2613,
"step": 76500
},
{
"epoch": 1.67,
"learning_rate": 2.2094169493491056e-05,
"loss": 1.2371,
"step": 77000
},
{
"epoch": 1.69,
"learning_rate": 2.1912962801890348e-05,
"loss": 1.2711,
"step": 77500
},
{
"epoch": 1.7,
"learning_rate": 2.1731756110289643e-05,
"loss": 1.2697,
"step": 78000
},
{
"epoch": 1.71,
"learning_rate": 2.1550549418688935e-05,
"loss": 1.2448,
"step": 78500
},
{
"epoch": 1.72,
"learning_rate": 2.1369342727088227e-05,
"loss": 1.2617,
"step": 79000
},
{
"epoch": 1.73,
"learning_rate": 2.118813603548752e-05,
"loss": 1.2717,
"step": 79500
},
{
"epoch": 1.74,
"learning_rate": 2.100692934388681e-05,
"loss": 1.2543,
"step": 80000
},
{
"epoch": 1.75,
"learning_rate": 2.0825722652286106e-05,
"loss": 1.2457,
"step": 80500
},
{
"epoch": 1.76,
"learning_rate": 2.06445159606854e-05,
"loss": 1.2377,
"step": 81000
},
{
"epoch": 1.77,
"learning_rate": 2.046330926908469e-05,
"loss": 1.2651,
"step": 81500
},
{
"epoch": 1.78,
"learning_rate": 2.0282102577483982e-05,
"loss": 1.2534,
"step": 82000
},
{
"epoch": 1.79,
"learning_rate": 2.0100895885883274e-05,
"loss": 1.2338,
"step": 82500
},
{
"epoch": 1.8,
"learning_rate": 1.9919689194282566e-05,
"loss": 1.2658,
"step": 83000
},
{
"epoch": 1.82,
"learning_rate": 1.973848250268186e-05,
"loss": 1.2088,
"step": 83500
},
{
"epoch": 1.83,
"learning_rate": 1.9557275811081154e-05,
"loss": 1.2322,
"step": 84000
},
{
"epoch": 1.84,
"learning_rate": 1.9376069119480446e-05,
"loss": 1.2474,
"step": 84500
},
{
"epoch": 1.85,
"learning_rate": 1.9194862427879738e-05,
"loss": 1.2271,
"step": 85000
},
{
"epoch": 1.86,
"learning_rate": 1.901365573627903e-05,
"loss": 1.2463,
"step": 85500
},
{
"epoch": 1.87,
"learning_rate": 1.8832449044678325e-05,
"loss": 1.2355,
"step": 86000
},
{
"epoch": 1.88,
"learning_rate": 1.8651242353077617e-05,
"loss": 1.2466,
"step": 86500
},
{
"epoch": 1.89,
"learning_rate": 1.847003566147691e-05,
"loss": 1.21,
"step": 87000
},
{
"epoch": 1.9,
"learning_rate": 1.82888289698762e-05,
"loss": 1.2379,
"step": 87500
},
{
"epoch": 1.91,
"learning_rate": 1.8107622278275493e-05,
"loss": 1.2444,
"step": 88000
},
{
"epoch": 1.92,
"learning_rate": 1.7926415586674785e-05,
"loss": 1.2213,
"step": 88500
},
{
"epoch": 1.94,
"learning_rate": 1.7745208895074077e-05,
"loss": 1.225,
"step": 89000
},
{
"epoch": 1.95,
"learning_rate": 1.756400220347337e-05,
"loss": 1.2214,
"step": 89500
},
{
"epoch": 1.96,
"learning_rate": 1.7382795511872664e-05,
"loss": 1.233,
"step": 90000
},
{
"epoch": 1.97,
"learning_rate": 1.7201588820271956e-05,
"loss": 1.2158,
"step": 90500
},
{
"epoch": 1.98,
"learning_rate": 1.7020382128671248e-05,
"loss": 1.2242,
"step": 91000
},
{
"epoch": 1.99,
"learning_rate": 1.683917543707054e-05,
"loss": 1.2174,
"step": 91500
},
{
"epoch": 2.0,
"learning_rate": 1.6657968745469836e-05,
"loss": 1.1991,
"step": 92000
},
{
"epoch": 2.01,
"learning_rate": 1.6476762053869128e-05,
"loss": 1.0718,
"step": 92500
},
{
"epoch": 2.02,
"learning_rate": 1.629555536226842e-05,
"loss": 1.0846,
"step": 93000
},
{
"epoch": 2.03,
"learning_rate": 1.611434867066771e-05,
"loss": 1.075,
"step": 93500
},
{
"epoch": 2.04,
"learning_rate": 1.5933141979067004e-05,
"loss": 1.0593,
"step": 94000
},
{
"epoch": 2.05,
"learning_rate": 1.5751935287466296e-05,
"loss": 1.0881,
"step": 94500
},
{
"epoch": 2.07,
"learning_rate": 1.5570728595865588e-05,
"loss": 1.0761,
"step": 95000
},
{
"epoch": 2.08,
"learning_rate": 1.538952190426488e-05,
"loss": 1.0514,
"step": 95500
},
{
"epoch": 2.09,
"learning_rate": 1.5208315212664173e-05,
"loss": 1.1007,
"step": 96000
},
{
"epoch": 2.1,
"learning_rate": 1.5027108521063465e-05,
"loss": 1.0728,
"step": 96500
},
{
"epoch": 2.11,
"learning_rate": 1.4845901829462757e-05,
"loss": 1.0777,
"step": 97000
},
{
"epoch": 2.12,
"learning_rate": 1.4664695137862053e-05,
"loss": 1.1008,
"step": 97500
},
{
"epoch": 2.13,
"learning_rate": 1.4483488446261345e-05,
"loss": 1.0805,
"step": 98000
},
{
"epoch": 2.14,
"learning_rate": 1.4302281754660638e-05,
"loss": 1.0883,
"step": 98500
},
{
"epoch": 2.15,
"learning_rate": 1.412107506305993e-05,
"loss": 1.0727,
"step": 99000
},
{
"epoch": 2.16,
"learning_rate": 1.3939868371459222e-05,
"loss": 1.087,
"step": 99500
},
{
"epoch": 2.17,
"learning_rate": 1.3758661679858514e-05,
"loss": 1.0686,
"step": 100000
},
{
"epoch": 2.19,
"learning_rate": 1.3577454988257806e-05,
"loss": 1.0503,
"step": 100500
},
{
"epoch": 2.2,
"learning_rate": 1.33962482966571e-05,
"loss": 1.0566,
"step": 101000
},
{
"epoch": 2.21,
"learning_rate": 1.3215041605056392e-05,
"loss": 1.0934,
"step": 101500
},
{
"epoch": 2.22,
"learning_rate": 1.3033834913455684e-05,
"loss": 1.0645,
"step": 102000
},
{
"epoch": 2.23,
"learning_rate": 1.2852628221854976e-05,
"loss": 1.0752,
"step": 102500
},
{
"epoch": 2.24,
"learning_rate": 1.2671421530254271e-05,
"loss": 1.0898,
"step": 103000
},
{
"epoch": 2.25,
"learning_rate": 1.2490214838653561e-05,
"loss": 1.0628,
"step": 103500
},
{
"epoch": 2.26,
"learning_rate": 1.2309008147052855e-05,
"loss": 1.0629,
"step": 104000
},
{
"epoch": 2.27,
"learning_rate": 1.2127801455452149e-05,
"loss": 1.058,
"step": 104500
},
{
"epoch": 2.28,
"learning_rate": 1.194659476385144e-05,
"loss": 1.0519,
"step": 105000
},
{
"epoch": 2.29,
"learning_rate": 1.1765388072250733e-05,
"loss": 1.0669,
"step": 105500
},
{
"epoch": 2.3,
"learning_rate": 1.1584181380650025e-05,
"loss": 1.0334,
"step": 106000
},
{
"epoch": 2.32,
"learning_rate": 1.1402974689049317e-05,
"loss": 1.0668,
"step": 106500
},
{
"epoch": 2.33,
"learning_rate": 1.122176799744861e-05,
"loss": 1.0295,
"step": 107000
},
{
"epoch": 2.34,
"learning_rate": 1.1040561305847904e-05,
"loss": 1.07,
"step": 107500
},
{
"epoch": 2.35,
"learning_rate": 1.0859354614247196e-05,
"loss": 1.0617,
"step": 108000
},
{
"epoch": 2.36,
"learning_rate": 1.0678147922646488e-05,
"loss": 1.0757,
"step": 108500
},
{
"epoch": 2.37,
"learning_rate": 1.049694123104578e-05,
"loss": 1.0509,
"step": 109000
},
{
"epoch": 2.38,
"learning_rate": 1.0315734539445072e-05,
"loss": 1.0558,
"step": 109500
},
{
"epoch": 2.39,
"learning_rate": 1.0134527847844366e-05,
"loss": 1.0498,
"step": 110000
},
{
"epoch": 2.4,
"learning_rate": 9.953321156243658e-06,
"loss": 1.0357,
"step": 110500
},
{
"epoch": 2.41,
"learning_rate": 9.772114464642951e-06,
"loss": 1.0562,
"step": 111000
},
{
"epoch": 2.42,
"learning_rate": 9.590907773042243e-06,
"loss": 1.0517,
"step": 111500
},
{
"epoch": 2.44,
"learning_rate": 9.409701081441535e-06,
"loss": 1.0461,
"step": 112000
},
{
"epoch": 2.45,
"learning_rate": 9.228494389840829e-06,
"loss": 1.0622,
"step": 112500
},
{
"epoch": 2.46,
"learning_rate": 9.047287698240121e-06,
"loss": 1.0476,
"step": 113000
},
{
"epoch": 2.47,
"learning_rate": 8.866081006639413e-06,
"loss": 1.0741,
"step": 113500
},
{
"epoch": 2.48,
"learning_rate": 8.684874315038707e-06,
"loss": 1.0664,
"step": 114000
},
{
"epoch": 2.49,
"learning_rate": 8.503667623437999e-06,
"loss": 1.0422,
"step": 114500
},
{
"epoch": 2.5,
"learning_rate": 8.32246093183729e-06,
"loss": 1.0535,
"step": 115000
},
{
"epoch": 2.51,
"learning_rate": 8.141254240236584e-06,
"loss": 1.0415,
"step": 115500
},
{
"epoch": 2.52,
"learning_rate": 7.960047548635876e-06,
"loss": 1.0399,
"step": 116000
},
{
"epoch": 2.53,
"learning_rate": 7.778840857035168e-06,
"loss": 1.0326,
"step": 116500
},
{
"epoch": 2.54,
"learning_rate": 7.597634165434461e-06,
"loss": 1.0545,
"step": 117000
},
{
"epoch": 2.56,
"learning_rate": 7.416427473833754e-06,
"loss": 1.0602,
"step": 117500
},
{
"epoch": 2.57,
"learning_rate": 7.2352207822330476e-06,
"loss": 1.0437,
"step": 118000
},
{
"epoch": 2.58,
"learning_rate": 7.0540140906323395e-06,
"loss": 1.0547,
"step": 118500
},
{
"epoch": 2.59,
"learning_rate": 6.8728073990316315e-06,
"loss": 1.033,
"step": 119000
},
{
"epoch": 2.6,
"learning_rate": 6.691600707430924e-06,
"loss": 1.0492,
"step": 119500
},
{
"epoch": 2.61,
"learning_rate": 6.510394015830216e-06,
"loss": 1.0498,
"step": 120000
},
{
"epoch": 2.62,
"learning_rate": 6.329187324229509e-06,
"loss": 1.0349,
"step": 120500
},
{
"epoch": 2.63,
"learning_rate": 6.147980632628802e-06,
"loss": 1.0501,
"step": 121000
},
{
"epoch": 2.64,
"learning_rate": 5.966773941028095e-06,
"loss": 1.0443,
"step": 121500
},
{
"epoch": 2.65,
"learning_rate": 5.785567249427387e-06,
"loss": 1.0534,
"step": 122000
},
{
"epoch": 2.66,
"learning_rate": 5.60436055782668e-06,
"loss": 1.0447,
"step": 122500
},
{
"epoch": 2.67,
"learning_rate": 5.4231538662259725e-06,
"loss": 1.0397,
"step": 123000
},
{
"epoch": 2.69,
"learning_rate": 5.2419471746252644e-06,
"loss": 1.0453,
"step": 123500
},
{
"epoch": 2.7,
"learning_rate": 5.060740483024557e-06,
"loss": 1.0256,
"step": 124000
},
{
"epoch": 2.71,
"learning_rate": 4.87953379142385e-06,
"loss": 1.0469,
"step": 124500
},
{
"epoch": 2.72,
"learning_rate": 4.698327099823142e-06,
"loss": 1.0383,
"step": 125000
},
{
"epoch": 2.73,
"learning_rate": 4.517120408222436e-06,
"loss": 1.0396,
"step": 125500
},
{
"epoch": 2.74,
"learning_rate": 4.335913716621728e-06,
"loss": 1.0476,
"step": 126000
},
{
"epoch": 2.75,
"learning_rate": 4.15470702502102e-06,
"loss": 1.0232,
"step": 126500
},
{
"epoch": 2.76,
"learning_rate": 3.973500333420313e-06,
"loss": 1.0357,
"step": 127000
},
{
"epoch": 2.77,
"learning_rate": 3.7922936418196054e-06,
"loss": 1.0649,
"step": 127500
},
{
"epoch": 2.78,
"learning_rate": 3.611086950218898e-06,
"loss": 1.0359,
"step": 128000
},
{
"epoch": 2.79,
"learning_rate": 3.4298802586181906e-06,
"loss": 1.0458,
"step": 128500
},
{
"epoch": 2.81,
"learning_rate": 3.248673567017483e-06,
"loss": 1.0207,
"step": 129000
},
{
"epoch": 2.82,
"learning_rate": 3.0674668754167754e-06,
"loss": 1.0144,
"step": 129500
},
{
"epoch": 2.83,
"learning_rate": 2.886260183816068e-06,
"loss": 1.0444,
"step": 130000
},
{
"epoch": 2.84,
"learning_rate": 2.7050534922153607e-06,
"loss": 1.033,
"step": 130500
},
{
"epoch": 2.85,
"learning_rate": 2.523846800614653e-06,
"loss": 1.0244,
"step": 131000
},
{
"epoch": 2.86,
"learning_rate": 2.342640109013946e-06,
"loss": 1.0452,
"step": 131500
},
{
"epoch": 2.87,
"learning_rate": 2.1614334174132383e-06,
"loss": 1.0257,
"step": 132000
},
{
"epoch": 2.88,
"learning_rate": 1.9802267258125307e-06,
"loss": 1.0365,
"step": 132500
},
{
"epoch": 2.89,
"learning_rate": 1.7990200342118236e-06,
"loss": 1.0227,
"step": 133000
},
{
"epoch": 2.9,
"learning_rate": 1.6178133426111162e-06,
"loss": 1.0362,
"step": 133500
},
{
"epoch": 2.91,
"learning_rate": 1.4366066510104086e-06,
"loss": 1.0261,
"step": 134000
},
{
"epoch": 2.92,
"learning_rate": 1.2553999594097012e-06,
"loss": 1.0349,
"step": 134500
},
{
"epoch": 2.94,
"learning_rate": 1.0741932678089936e-06,
"loss": 1.0244,
"step": 135000
},
{
"epoch": 2.95,
"learning_rate": 8.929865762082863e-07,
"loss": 1.0356,
"step": 135500
},
{
"epoch": 2.96,
"learning_rate": 7.117798846075788e-07,
"loss": 1.0498,
"step": 136000
},
{
"epoch": 2.97,
"learning_rate": 5.305731930068714e-07,
"loss": 1.036,
"step": 136500
},
{
"epoch": 2.98,
"learning_rate": 3.4936650140616397e-07,
"loss": 1.0156,
"step": 137000
},
{
"epoch": 2.99,
"learning_rate": 1.681598098054565e-07,
"loss": 1.0577,
"step": 137500
},
{
"epoch": 3.0,
"step": 137964,
"total_flos": 2.6751009830151782e+17,
"train_loss": 1.3410481518223418,
"train_runtime": 43433.1591,
"train_samples_per_second": 31.764,
"train_steps_per_second": 3.176
}
],
"max_steps": 137964,
"num_train_epochs": 3,
"total_flos": 2.6751009830151782e+17,
"trial_name": null,
"trial_params": null
}