m2m100_418M_ibo_en_rel_news / trainer_state.json
Davlan's picture
add MT model
a141b9b
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 3.0,
"global_step": 126582,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.01,
"learning_rate": 4.9802499565499043e-05,
"loss": 2.8425,
"step": 500
},
{
"epoch": 0.02,
"learning_rate": 4.960499913099809e-05,
"loss": 2.4736,
"step": 1000
},
{
"epoch": 0.04,
"learning_rate": 4.940749869649713e-05,
"loss": 2.298,
"step": 1500
},
{
"epoch": 0.05,
"learning_rate": 4.9209998261996174e-05,
"loss": 2.235,
"step": 2000
},
{
"epoch": 0.06,
"learning_rate": 4.901249782749522e-05,
"loss": 2.1605,
"step": 2500
},
{
"epoch": 0.07,
"learning_rate": 4.881499739299427e-05,
"loss": 2.083,
"step": 3000
},
{
"epoch": 0.08,
"learning_rate": 4.861749695849331e-05,
"loss": 2.0588,
"step": 3500
},
{
"epoch": 0.09,
"learning_rate": 4.841999652399236e-05,
"loss": 2.0555,
"step": 4000
},
{
"epoch": 0.11,
"learning_rate": 4.82224960894914e-05,
"loss": 2.0182,
"step": 4500
},
{
"epoch": 0.12,
"learning_rate": 4.802499565499045e-05,
"loss": 1.9641,
"step": 5000
},
{
"epoch": 0.13,
"learning_rate": 4.782749522048949e-05,
"loss": 1.9383,
"step": 5500
},
{
"epoch": 0.14,
"learning_rate": 4.762999478598853e-05,
"loss": 1.9275,
"step": 6000
},
{
"epoch": 0.15,
"learning_rate": 4.743249435148758e-05,
"loss": 1.8923,
"step": 6500
},
{
"epoch": 0.17,
"learning_rate": 4.723499391698662e-05,
"loss": 1.8957,
"step": 7000
},
{
"epoch": 0.18,
"learning_rate": 4.703749348248566e-05,
"loss": 1.8787,
"step": 7500
},
{
"epoch": 0.19,
"learning_rate": 4.683999304798471e-05,
"loss": 1.8516,
"step": 8000
},
{
"epoch": 0.2,
"learning_rate": 4.664249261348375e-05,
"loss": 1.8373,
"step": 8500
},
{
"epoch": 0.21,
"learning_rate": 4.6444992178982796e-05,
"loss": 1.7909,
"step": 9000
},
{
"epoch": 0.23,
"learning_rate": 4.6247491744481844e-05,
"loss": 1.8034,
"step": 9500
},
{
"epoch": 0.24,
"learning_rate": 4.6049991309980885e-05,
"loss": 1.8101,
"step": 10000
},
{
"epoch": 0.25,
"learning_rate": 4.585249087547993e-05,
"loss": 1.7702,
"step": 10500
},
{
"epoch": 0.26,
"learning_rate": 4.5654990440978974e-05,
"loss": 1.7727,
"step": 11000
},
{
"epoch": 0.27,
"learning_rate": 4.5457490006478016e-05,
"loss": 1.7602,
"step": 11500
},
{
"epoch": 0.28,
"learning_rate": 4.5259989571977063e-05,
"loss": 1.761,
"step": 12000
},
{
"epoch": 0.3,
"learning_rate": 4.5062489137476105e-05,
"loss": 1.7219,
"step": 12500
},
{
"epoch": 0.31,
"learning_rate": 4.4864988702975146e-05,
"loss": 1.7497,
"step": 13000
},
{
"epoch": 0.32,
"learning_rate": 4.4667488268474193e-05,
"loss": 1.7244,
"step": 13500
},
{
"epoch": 0.33,
"learning_rate": 4.4469987833973235e-05,
"loss": 1.7144,
"step": 14000
},
{
"epoch": 0.34,
"learning_rate": 4.4272487399472276e-05,
"loss": 1.6983,
"step": 14500
},
{
"epoch": 0.36,
"learning_rate": 4.4074986964971324e-05,
"loss": 1.6779,
"step": 15000
},
{
"epoch": 0.37,
"learning_rate": 4.387748653047037e-05,
"loss": 1.676,
"step": 15500
},
{
"epoch": 0.38,
"learning_rate": 4.367998609596941e-05,
"loss": 1.6737,
"step": 16000
},
{
"epoch": 0.39,
"learning_rate": 4.348248566146846e-05,
"loss": 1.6605,
"step": 16500
},
{
"epoch": 0.4,
"learning_rate": 4.32849852269675e-05,
"loss": 1.6591,
"step": 17000
},
{
"epoch": 0.41,
"learning_rate": 4.308748479246655e-05,
"loss": 1.6598,
"step": 17500
},
{
"epoch": 0.43,
"learning_rate": 4.288998435796559e-05,
"loss": 1.6579,
"step": 18000
},
{
"epoch": 0.44,
"learning_rate": 4.269248392346463e-05,
"loss": 1.6587,
"step": 18500
},
{
"epoch": 0.45,
"learning_rate": 4.249498348896368e-05,
"loss": 1.65,
"step": 19000
},
{
"epoch": 0.46,
"learning_rate": 4.229748305446272e-05,
"loss": 1.6399,
"step": 19500
},
{
"epoch": 0.47,
"learning_rate": 4.209998261996176e-05,
"loss": 1.6409,
"step": 20000
},
{
"epoch": 0.49,
"learning_rate": 4.190248218546081e-05,
"loss": 1.6167,
"step": 20500
},
{
"epoch": 0.5,
"learning_rate": 4.170498175095985e-05,
"loss": 1.6468,
"step": 21000
},
{
"epoch": 0.51,
"learning_rate": 4.15074813164589e-05,
"loss": 1.6084,
"step": 21500
},
{
"epoch": 0.52,
"learning_rate": 4.1309980881957946e-05,
"loss": 1.6214,
"step": 22000
},
{
"epoch": 0.53,
"learning_rate": 4.111248044745699e-05,
"loss": 1.6254,
"step": 22500
},
{
"epoch": 0.55,
"learning_rate": 4.091498001295603e-05,
"loss": 1.5892,
"step": 23000
},
{
"epoch": 0.56,
"learning_rate": 4.0717479578455077e-05,
"loss": 1.5746,
"step": 23500
},
{
"epoch": 0.57,
"learning_rate": 4.051997914395412e-05,
"loss": 1.5701,
"step": 24000
},
{
"epoch": 0.58,
"learning_rate": 4.0322478709453166e-05,
"loss": 1.6088,
"step": 24500
},
{
"epoch": 0.59,
"learning_rate": 4.012497827495221e-05,
"loss": 1.5952,
"step": 25000
},
{
"epoch": 0.6,
"learning_rate": 3.992747784045125e-05,
"loss": 1.5561,
"step": 25500
},
{
"epoch": 0.62,
"learning_rate": 3.9729977405950296e-05,
"loss": 1.5707,
"step": 26000
},
{
"epoch": 0.63,
"learning_rate": 3.953247697144934e-05,
"loss": 1.5582,
"step": 26500
},
{
"epoch": 0.64,
"learning_rate": 3.9334976536948385e-05,
"loss": 1.5612,
"step": 27000
},
{
"epoch": 0.65,
"learning_rate": 3.9137476102447426e-05,
"loss": 1.5552,
"step": 27500
},
{
"epoch": 0.66,
"learning_rate": 3.8939975667946474e-05,
"loss": 1.5632,
"step": 28000
},
{
"epoch": 0.68,
"learning_rate": 3.8742475233445515e-05,
"loss": 1.538,
"step": 28500
},
{
"epoch": 0.69,
"learning_rate": 3.854497479894456e-05,
"loss": 1.5257,
"step": 29000
},
{
"epoch": 0.7,
"learning_rate": 3.8347474364443604e-05,
"loss": 1.5355,
"step": 29500
},
{
"epoch": 0.71,
"learning_rate": 3.8149973929942645e-05,
"loss": 1.5233,
"step": 30000
},
{
"epoch": 0.72,
"learning_rate": 3.795247349544169e-05,
"loss": 1.5154,
"step": 30500
},
{
"epoch": 0.73,
"learning_rate": 3.7754973060940734e-05,
"loss": 1.5493,
"step": 31000
},
{
"epoch": 0.75,
"learning_rate": 3.755747262643978e-05,
"loss": 1.5327,
"step": 31500
},
{
"epoch": 0.76,
"learning_rate": 3.735997219193882e-05,
"loss": 1.5255,
"step": 32000
},
{
"epoch": 0.77,
"learning_rate": 3.7162471757437864e-05,
"loss": 1.51,
"step": 32500
},
{
"epoch": 0.78,
"learning_rate": 3.696497132293691e-05,
"loss": 1.5122,
"step": 33000
},
{
"epoch": 0.79,
"learning_rate": 3.676747088843596e-05,
"loss": 1.5006,
"step": 33500
},
{
"epoch": 0.81,
"learning_rate": 3.6569970453935e-05,
"loss": 1.5106,
"step": 34000
},
{
"epoch": 0.82,
"learning_rate": 3.637247001943405e-05,
"loss": 1.5031,
"step": 34500
},
{
"epoch": 0.83,
"learning_rate": 3.617496958493309e-05,
"loss": 1.5024,
"step": 35000
},
{
"epoch": 0.84,
"learning_rate": 3.597746915043213e-05,
"loss": 1.5231,
"step": 35500
},
{
"epoch": 0.85,
"learning_rate": 3.577996871593118e-05,
"loss": 1.5271,
"step": 36000
},
{
"epoch": 0.87,
"learning_rate": 3.558246828143022e-05,
"loss": 1.508,
"step": 36500
},
{
"epoch": 0.88,
"learning_rate": 3.538496784692926e-05,
"loss": 1.4859,
"step": 37000
},
{
"epoch": 0.89,
"learning_rate": 3.518746741242831e-05,
"loss": 1.4864,
"step": 37500
},
{
"epoch": 0.9,
"learning_rate": 3.498996697792735e-05,
"loss": 1.5045,
"step": 38000
},
{
"epoch": 0.91,
"learning_rate": 3.47924665434264e-05,
"loss": 1.4773,
"step": 38500
},
{
"epoch": 0.92,
"learning_rate": 3.459496610892544e-05,
"loss": 1.4704,
"step": 39000
},
{
"epoch": 0.94,
"learning_rate": 3.439746567442449e-05,
"loss": 1.4816,
"step": 39500
},
{
"epoch": 0.95,
"learning_rate": 3.4199965239923535e-05,
"loss": 1.4899,
"step": 40000
},
{
"epoch": 0.96,
"learning_rate": 3.4002464805422576e-05,
"loss": 1.463,
"step": 40500
},
{
"epoch": 0.97,
"learning_rate": 3.380496437092162e-05,
"loss": 1.4847,
"step": 41000
},
{
"epoch": 0.98,
"learning_rate": 3.3607463936420665e-05,
"loss": 1.478,
"step": 41500
},
{
"epoch": 1.0,
"learning_rate": 3.3409963501919706e-05,
"loss": 1.4617,
"step": 42000
},
{
"epoch": 1.01,
"learning_rate": 3.321246306741875e-05,
"loss": 1.3742,
"step": 42500
},
{
"epoch": 1.02,
"learning_rate": 3.3014962632917795e-05,
"loss": 1.3036,
"step": 43000
},
{
"epoch": 1.03,
"learning_rate": 3.2817462198416836e-05,
"loss": 1.3172,
"step": 43500
},
{
"epoch": 1.04,
"learning_rate": 3.261996176391588e-05,
"loss": 1.2944,
"step": 44000
},
{
"epoch": 1.05,
"learning_rate": 3.2422461329414925e-05,
"loss": 1.319,
"step": 44500
},
{
"epoch": 1.07,
"learning_rate": 3.2224960894913966e-05,
"loss": 1.3015,
"step": 45000
},
{
"epoch": 1.08,
"learning_rate": 3.2027460460413014e-05,
"loss": 1.3011,
"step": 45500
},
{
"epoch": 1.09,
"learning_rate": 3.182996002591206e-05,
"loss": 1.295,
"step": 46000
},
{
"epoch": 1.1,
"learning_rate": 3.16324595914111e-05,
"loss": 1.2994,
"step": 46500
},
{
"epoch": 1.11,
"learning_rate": 3.143495915691015e-05,
"loss": 1.3215,
"step": 47000
},
{
"epoch": 1.13,
"learning_rate": 3.123745872240919e-05,
"loss": 1.3197,
"step": 47500
},
{
"epoch": 1.14,
"learning_rate": 3.103995828790823e-05,
"loss": 1.3114,
"step": 48000
},
{
"epoch": 1.15,
"learning_rate": 3.084245785340728e-05,
"loss": 1.2988,
"step": 48500
},
{
"epoch": 1.16,
"learning_rate": 3.064495741890632e-05,
"loss": 1.3004,
"step": 49000
},
{
"epoch": 1.17,
"learning_rate": 3.0447456984405366e-05,
"loss": 1.275,
"step": 49500
},
{
"epoch": 1.19,
"learning_rate": 3.024995654990441e-05,
"loss": 1.2822,
"step": 50000
},
{
"epoch": 1.2,
"learning_rate": 3.0052456115403455e-05,
"loss": 1.3113,
"step": 50500
},
{
"epoch": 1.21,
"learning_rate": 2.9854955680902503e-05,
"loss": 1.3087,
"step": 51000
},
{
"epoch": 1.22,
"learning_rate": 2.9657455246401544e-05,
"loss": 1.3039,
"step": 51500
},
{
"epoch": 1.23,
"learning_rate": 2.9459954811900585e-05,
"loss": 1.2772,
"step": 52000
},
{
"epoch": 1.24,
"learning_rate": 2.9262454377399633e-05,
"loss": 1.2918,
"step": 52500
},
{
"epoch": 1.26,
"learning_rate": 2.9064953942898678e-05,
"loss": 1.2815,
"step": 53000
},
{
"epoch": 1.27,
"learning_rate": 2.886745350839772e-05,
"loss": 1.3004,
"step": 53500
},
{
"epoch": 1.28,
"learning_rate": 2.8669953073896767e-05,
"loss": 1.2761,
"step": 54000
},
{
"epoch": 1.29,
"learning_rate": 2.8472452639395808e-05,
"loss": 1.2927,
"step": 54500
},
{
"epoch": 1.3,
"learning_rate": 2.827495220489485e-05,
"loss": 1.2815,
"step": 55000
},
{
"epoch": 1.32,
"learning_rate": 2.8077451770393897e-05,
"loss": 1.27,
"step": 55500
},
{
"epoch": 1.33,
"learning_rate": 2.787995133589294e-05,
"loss": 1.2928,
"step": 56000
},
{
"epoch": 1.34,
"learning_rate": 2.7682450901391982e-05,
"loss": 1.2752,
"step": 56500
},
{
"epoch": 1.35,
"learning_rate": 2.748495046689103e-05,
"loss": 1.2935,
"step": 57000
},
{
"epoch": 1.36,
"learning_rate": 2.728745003239007e-05,
"loss": 1.2794,
"step": 57500
},
{
"epoch": 1.37,
"learning_rate": 2.708994959788912e-05,
"loss": 1.2956,
"step": 58000
},
{
"epoch": 1.39,
"learning_rate": 2.689244916338816e-05,
"loss": 1.2856,
"step": 58500
},
{
"epoch": 1.4,
"learning_rate": 2.6694948728887205e-05,
"loss": 1.2743,
"step": 59000
},
{
"epoch": 1.41,
"learning_rate": 2.6497448294386253e-05,
"loss": 1.2743,
"step": 59500
},
{
"epoch": 1.42,
"learning_rate": 2.6299947859885294e-05,
"loss": 1.2753,
"step": 60000
},
{
"epoch": 1.43,
"learning_rate": 2.6102447425384335e-05,
"loss": 1.2653,
"step": 60500
},
{
"epoch": 1.45,
"learning_rate": 2.5904946990883383e-05,
"loss": 1.2778,
"step": 61000
},
{
"epoch": 1.46,
"learning_rate": 2.5707446556382424e-05,
"loss": 1.2726,
"step": 61500
},
{
"epoch": 1.47,
"learning_rate": 2.550994612188147e-05,
"loss": 1.2773,
"step": 62000
},
{
"epoch": 1.48,
"learning_rate": 2.5312445687380516e-05,
"loss": 1.2632,
"step": 62500
},
{
"epoch": 1.49,
"learning_rate": 2.5114945252879557e-05,
"loss": 1.2751,
"step": 63000
},
{
"epoch": 1.5,
"learning_rate": 2.4917444818378602e-05,
"loss": 1.2607,
"step": 63500
},
{
"epoch": 1.52,
"learning_rate": 2.4719944383877646e-05,
"loss": 1.2556,
"step": 64000
},
{
"epoch": 1.53,
"learning_rate": 2.4522443949376688e-05,
"loss": 1.235,
"step": 64500
},
{
"epoch": 1.54,
"learning_rate": 2.4324943514875735e-05,
"loss": 1.2502,
"step": 65000
},
{
"epoch": 1.55,
"learning_rate": 2.412744308037478e-05,
"loss": 1.2633,
"step": 65500
},
{
"epoch": 1.56,
"learning_rate": 2.392994264587382e-05,
"loss": 1.2558,
"step": 66000
},
{
"epoch": 1.58,
"learning_rate": 2.3732442211372865e-05,
"loss": 1.2659,
"step": 66500
},
{
"epoch": 1.59,
"learning_rate": 2.353494177687191e-05,
"loss": 1.258,
"step": 67000
},
{
"epoch": 1.6,
"learning_rate": 2.3337441342370954e-05,
"loss": 1.2609,
"step": 67500
},
{
"epoch": 1.61,
"learning_rate": 2.313994090787e-05,
"loss": 1.2485,
"step": 68000
},
{
"epoch": 1.62,
"learning_rate": 2.2942440473369043e-05,
"loss": 1.2563,
"step": 68500
},
{
"epoch": 1.64,
"learning_rate": 2.2744940038868088e-05,
"loss": 1.2621,
"step": 69000
},
{
"epoch": 1.65,
"learning_rate": 2.254743960436713e-05,
"loss": 1.246,
"step": 69500
},
{
"epoch": 1.66,
"learning_rate": 2.2349939169866174e-05,
"loss": 1.2327,
"step": 70000
},
{
"epoch": 1.67,
"learning_rate": 2.2152438735365218e-05,
"loss": 1.2388,
"step": 70500
},
{
"epoch": 1.68,
"learning_rate": 2.1954938300864263e-05,
"loss": 1.2443,
"step": 71000
},
{
"epoch": 1.69,
"learning_rate": 2.1757437866363307e-05,
"loss": 1.2585,
"step": 71500
},
{
"epoch": 1.71,
"learning_rate": 2.155993743186235e-05,
"loss": 1.2376,
"step": 72000
},
{
"epoch": 1.72,
"learning_rate": 2.1362436997361396e-05,
"loss": 1.2317,
"step": 72500
},
{
"epoch": 1.73,
"learning_rate": 2.116493656286044e-05,
"loss": 1.2279,
"step": 73000
},
{
"epoch": 1.74,
"learning_rate": 2.096743612835948e-05,
"loss": 1.2444,
"step": 73500
},
{
"epoch": 1.75,
"learning_rate": 2.0769935693858526e-05,
"loss": 1.2488,
"step": 74000
},
{
"epoch": 1.77,
"learning_rate": 2.0572435259357574e-05,
"loss": 1.2314,
"step": 74500
},
{
"epoch": 1.78,
"learning_rate": 2.0374934824856615e-05,
"loss": 1.2385,
"step": 75000
},
{
"epoch": 1.79,
"learning_rate": 2.017743439035566e-05,
"loss": 1.2219,
"step": 75500
},
{
"epoch": 1.8,
"learning_rate": 1.9979933955854704e-05,
"loss": 1.2491,
"step": 76000
},
{
"epoch": 1.81,
"learning_rate": 1.978243352135375e-05,
"loss": 1.229,
"step": 76500
},
{
"epoch": 1.82,
"learning_rate": 1.9584933086852793e-05,
"loss": 1.2233,
"step": 77000
},
{
"epoch": 1.84,
"learning_rate": 1.9387432652351838e-05,
"loss": 1.2521,
"step": 77500
},
{
"epoch": 1.85,
"learning_rate": 1.9189932217850882e-05,
"loss": 1.2214,
"step": 78000
},
{
"epoch": 1.86,
"learning_rate": 1.8992431783349923e-05,
"loss": 1.2162,
"step": 78500
},
{
"epoch": 1.87,
"learning_rate": 1.8794931348848968e-05,
"loss": 1.2178,
"step": 79000
},
{
"epoch": 1.88,
"learning_rate": 1.8597430914348012e-05,
"loss": 1.2418,
"step": 79500
},
{
"epoch": 1.9,
"learning_rate": 1.8399930479847057e-05,
"loss": 1.2453,
"step": 80000
},
{
"epoch": 1.91,
"learning_rate": 1.82024300453461e-05,
"loss": 1.228,
"step": 80500
},
{
"epoch": 1.92,
"learning_rate": 1.8004929610845146e-05,
"loss": 1.2036,
"step": 81000
},
{
"epoch": 1.93,
"learning_rate": 1.780742917634419e-05,
"loss": 1.2078,
"step": 81500
},
{
"epoch": 1.94,
"learning_rate": 1.760992874184323e-05,
"loss": 1.2106,
"step": 82000
},
{
"epoch": 1.96,
"learning_rate": 1.7412428307342276e-05,
"loss": 1.2187,
"step": 82500
},
{
"epoch": 1.97,
"learning_rate": 1.721492787284132e-05,
"loss": 1.2267,
"step": 83000
},
{
"epoch": 1.98,
"learning_rate": 1.7017427438340368e-05,
"loss": 1.2262,
"step": 83500
},
{
"epoch": 1.99,
"learning_rate": 1.681992700383941e-05,
"loss": 1.2185,
"step": 84000
},
{
"epoch": 2.0,
"learning_rate": 1.6622426569338454e-05,
"loss": 1.1781,
"step": 84500
},
{
"epoch": 2.01,
"learning_rate": 1.6424926134837498e-05,
"loss": 1.0483,
"step": 85000
},
{
"epoch": 2.03,
"learning_rate": 1.622742570033654e-05,
"loss": 1.0659,
"step": 85500
},
{
"epoch": 2.04,
"learning_rate": 1.6029925265835584e-05,
"loss": 1.0643,
"step": 86000
},
{
"epoch": 2.05,
"learning_rate": 1.583242483133463e-05,
"loss": 1.0453,
"step": 86500
},
{
"epoch": 2.06,
"learning_rate": 1.5634924396833676e-05,
"loss": 1.0541,
"step": 87000
},
{
"epoch": 2.07,
"learning_rate": 1.5437423962332717e-05,
"loss": 1.0534,
"step": 87500
},
{
"epoch": 2.09,
"learning_rate": 1.5239923527831762e-05,
"loss": 1.0641,
"step": 88000
},
{
"epoch": 2.1,
"learning_rate": 1.5042423093330806e-05,
"loss": 1.0771,
"step": 88500
},
{
"epoch": 2.11,
"learning_rate": 1.4844922658829849e-05,
"loss": 1.0638,
"step": 89000
},
{
"epoch": 2.12,
"learning_rate": 1.4647422224328893e-05,
"loss": 1.0677,
"step": 89500
},
{
"epoch": 2.13,
"learning_rate": 1.4449921789827938e-05,
"loss": 1.0776,
"step": 90000
},
{
"epoch": 2.14,
"learning_rate": 1.4252421355326984e-05,
"loss": 1.064,
"step": 90500
},
{
"epoch": 2.16,
"learning_rate": 1.4054920920826025e-05,
"loss": 1.0548,
"step": 91000
},
{
"epoch": 2.17,
"learning_rate": 1.3857420486325071e-05,
"loss": 1.0606,
"step": 91500
},
{
"epoch": 2.18,
"learning_rate": 1.3659920051824116e-05,
"loss": 1.0699,
"step": 92000
},
{
"epoch": 2.19,
"learning_rate": 1.3462419617323157e-05,
"loss": 1.0721,
"step": 92500
},
{
"epoch": 2.2,
"learning_rate": 1.3264919182822203e-05,
"loss": 1.0438,
"step": 93000
},
{
"epoch": 2.22,
"learning_rate": 1.3067418748321248e-05,
"loss": 1.0425,
"step": 93500
},
{
"epoch": 2.23,
"learning_rate": 1.2869918313820292e-05,
"loss": 1.0501,
"step": 94000
},
{
"epoch": 2.24,
"learning_rate": 1.2672417879319335e-05,
"loss": 1.0307,
"step": 94500
},
{
"epoch": 2.25,
"learning_rate": 1.247491744481838e-05,
"loss": 1.037,
"step": 95000
},
{
"epoch": 2.26,
"learning_rate": 1.2277417010317422e-05,
"loss": 1.0401,
"step": 95500
},
{
"epoch": 2.28,
"learning_rate": 1.2079916575816467e-05,
"loss": 1.0504,
"step": 96000
},
{
"epoch": 2.29,
"learning_rate": 1.1882416141315511e-05,
"loss": 1.0381,
"step": 96500
},
{
"epoch": 2.3,
"learning_rate": 1.1684915706814556e-05,
"loss": 1.0302,
"step": 97000
},
{
"epoch": 2.31,
"learning_rate": 1.14874152723136e-05,
"loss": 1.0474,
"step": 97500
},
{
"epoch": 2.32,
"learning_rate": 1.1289914837812645e-05,
"loss": 1.0454,
"step": 98000
},
{
"epoch": 2.33,
"learning_rate": 1.1092414403311688e-05,
"loss": 1.0531,
"step": 98500
},
{
"epoch": 2.35,
"learning_rate": 1.0894913968810732e-05,
"loss": 1.0448,
"step": 99000
},
{
"epoch": 2.36,
"learning_rate": 1.0697413534309776e-05,
"loss": 1.0562,
"step": 99500
},
{
"epoch": 2.37,
"learning_rate": 1.049991309980882e-05,
"loss": 1.0432,
"step": 100000
},
{
"epoch": 2.38,
"learning_rate": 1.0302412665307864e-05,
"loss": 1.0436,
"step": 100500
},
{
"epoch": 2.39,
"learning_rate": 1.0104912230806908e-05,
"loss": 1.0419,
"step": 101000
},
{
"epoch": 2.41,
"learning_rate": 9.907411796305953e-06,
"loss": 1.0457,
"step": 101500
},
{
"epoch": 2.42,
"learning_rate": 9.709911361804996e-06,
"loss": 1.0409,
"step": 102000
},
{
"epoch": 2.43,
"learning_rate": 9.51241092730404e-06,
"loss": 1.031,
"step": 102500
},
{
"epoch": 2.44,
"learning_rate": 9.314910492803085e-06,
"loss": 1.0356,
"step": 103000
},
{
"epoch": 2.45,
"learning_rate": 9.117410058302129e-06,
"loss": 1.0606,
"step": 103500
},
{
"epoch": 2.46,
"learning_rate": 8.919909623801174e-06,
"loss": 1.0432,
"step": 104000
},
{
"epoch": 2.48,
"learning_rate": 8.722409189300216e-06,
"loss": 1.0254,
"step": 104500
},
{
"epoch": 2.49,
"learning_rate": 8.52490875479926e-06,
"loss": 1.0482,
"step": 105000
},
{
"epoch": 2.5,
"learning_rate": 8.327408320298305e-06,
"loss": 1.0522,
"step": 105500
},
{
"epoch": 2.51,
"learning_rate": 8.129907885797348e-06,
"loss": 1.0263,
"step": 106000
},
{
"epoch": 2.52,
"learning_rate": 7.932407451296393e-06,
"loss": 1.0546,
"step": 106500
},
{
"epoch": 2.54,
"learning_rate": 7.734907016795437e-06,
"loss": 1.0475,
"step": 107000
},
{
"epoch": 2.55,
"learning_rate": 7.5374065822944816e-06,
"loss": 1.0293,
"step": 107500
},
{
"epoch": 2.56,
"learning_rate": 7.339906147793525e-06,
"loss": 1.0497,
"step": 108000
},
{
"epoch": 2.57,
"learning_rate": 7.14240571329257e-06,
"loss": 1.0448,
"step": 108500
},
{
"epoch": 2.58,
"learning_rate": 6.944905278791613e-06,
"loss": 1.0275,
"step": 109000
},
{
"epoch": 2.6,
"learning_rate": 6.747404844290659e-06,
"loss": 1.0281,
"step": 109500
},
{
"epoch": 2.61,
"learning_rate": 6.549904409789702e-06,
"loss": 1.0295,
"step": 110000
},
{
"epoch": 2.62,
"learning_rate": 6.352403975288745e-06,
"loss": 1.0143,
"step": 110500
},
{
"epoch": 2.63,
"learning_rate": 6.1549035407877905e-06,
"loss": 1.0277,
"step": 111000
},
{
"epoch": 2.64,
"learning_rate": 5.957403106286834e-06,
"loss": 1.0266,
"step": 111500
},
{
"epoch": 2.65,
"learning_rate": 5.759902671785878e-06,
"loss": 1.0502,
"step": 112000
},
{
"epoch": 2.67,
"learning_rate": 5.562402237284922e-06,
"loss": 1.0312,
"step": 112500
},
{
"epoch": 2.68,
"learning_rate": 5.364901802783967e-06,
"loss": 1.0444,
"step": 113000
},
{
"epoch": 2.69,
"learning_rate": 5.16740136828301e-06,
"loss": 1.0257,
"step": 113500
},
{
"epoch": 2.7,
"learning_rate": 4.969900933782055e-06,
"loss": 1.0199,
"step": 114000
},
{
"epoch": 2.71,
"learning_rate": 4.7724004992810985e-06,
"loss": 1.0178,
"step": 114500
},
{
"epoch": 2.73,
"learning_rate": 4.574900064780143e-06,
"loss": 1.0197,
"step": 115000
},
{
"epoch": 2.74,
"learning_rate": 4.377399630279187e-06,
"loss": 1.0181,
"step": 115500
},
{
"epoch": 2.75,
"learning_rate": 4.179899195778231e-06,
"loss": 1.0492,
"step": 116000
},
{
"epoch": 2.76,
"learning_rate": 3.982398761277275e-06,
"loss": 1.035,
"step": 116500
},
{
"epoch": 2.77,
"learning_rate": 3.7848983267763193e-06,
"loss": 1.0256,
"step": 117000
},
{
"epoch": 2.78,
"learning_rate": 3.5873978922753633e-06,
"loss": 1.0031,
"step": 117500
},
{
"epoch": 2.8,
"learning_rate": 3.3898974577744074e-06,
"loss": 1.0208,
"step": 118000
},
{
"epoch": 2.81,
"learning_rate": 3.1923970232734515e-06,
"loss": 1.0215,
"step": 118500
},
{
"epoch": 2.82,
"learning_rate": 2.9948965887724956e-06,
"loss": 1.0196,
"step": 119000
},
{
"epoch": 2.83,
"learning_rate": 2.7973961542715396e-06,
"loss": 1.0366,
"step": 119500
},
{
"epoch": 2.84,
"learning_rate": 2.5998957197705837e-06,
"loss": 1.0268,
"step": 120000
},
{
"epoch": 2.86,
"learning_rate": 2.4023952852696278e-06,
"loss": 1.0141,
"step": 120500
},
{
"epoch": 2.87,
"learning_rate": 2.204894850768672e-06,
"loss": 1.0083,
"step": 121000
},
{
"epoch": 2.88,
"learning_rate": 2.007394416267716e-06,
"loss": 1.0266,
"step": 121500
},
{
"epoch": 2.89,
"learning_rate": 1.80989398176676e-06,
"loss": 1.0216,
"step": 122000
},
{
"epoch": 2.9,
"learning_rate": 1.612393547265804e-06,
"loss": 1.0168,
"step": 122500
},
{
"epoch": 2.92,
"learning_rate": 1.414893112764848e-06,
"loss": 1.0004,
"step": 123000
},
{
"epoch": 2.93,
"learning_rate": 1.2173926782638924e-06,
"loss": 1.026,
"step": 123500
},
{
"epoch": 2.94,
"learning_rate": 1.0198922437629362e-06,
"loss": 1.0042,
"step": 124000
},
{
"epoch": 2.95,
"learning_rate": 8.223918092619805e-07,
"loss": 0.9972,
"step": 124500
},
{
"epoch": 2.96,
"learning_rate": 6.248913747610245e-07,
"loss": 1.0221,
"step": 125000
},
{
"epoch": 2.97,
"learning_rate": 4.273909402600686e-07,
"loss": 1.02,
"step": 125500
},
{
"epoch": 2.99,
"learning_rate": 2.2989050575911268e-07,
"loss": 1.0402,
"step": 126000
},
{
"epoch": 3.0,
"learning_rate": 3.239007125815677e-08,
"loss": 1.0188,
"step": 126500
},
{
"epoch": 3.0,
"step": 126582,
"total_flos": 1.9038716367888384e+17,
"train_loss": 1.3316941024496511,
"train_runtime": 32427.393,
"train_samples_per_second": 39.035,
"train_steps_per_second": 3.904
}
],
"max_steps": 126582,
"num_train_epochs": 3,
"total_flos": 1.9038716367888384e+17,
"trial_name": null,
"trial_params": null
}