mT5-small-iwslt2017-calm-ft / trainer_state.json
Sehoon's picture
store
d5309a3
{
"best_metric": 18.4266,
"best_model_checkpoint": "ckpt_mt5_calm/google/mt5-small/iwslt2017_de_en/lr0.0002_e50/checkpoint-309168",
"epoch": 39.0,
"global_step": 502398,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.04,
"learning_rate": 0.0002,
"loss": 5.6576,
"step": 500
},
{
"epoch": 0.08,
"learning_rate": 0.0002,
"loss": 3.6261,
"step": 1000
},
{
"epoch": 0.12,
"learning_rate": 0.0002,
"loss": 3.3328,
"step": 1500
},
{
"epoch": 0.16,
"learning_rate": 0.0002,
"loss": 3.1604,
"step": 2000
},
{
"epoch": 0.19,
"learning_rate": 0.0002,
"loss": 3.0284,
"step": 2500
},
{
"epoch": 0.23,
"learning_rate": 0.0002,
"loss": 2.9396,
"step": 3000
},
{
"epoch": 0.27,
"learning_rate": 0.0002,
"loss": 2.8682,
"step": 3500
},
{
"epoch": 0.31,
"learning_rate": 0.0002,
"loss": 2.7862,
"step": 4000
},
{
"epoch": 0.35,
"learning_rate": 0.0002,
"loss": 2.7437,
"step": 4500
},
{
"epoch": 0.39,
"learning_rate": 0.0002,
"loss": 2.689,
"step": 5000
},
{
"epoch": 0.43,
"learning_rate": 0.0002,
"loss": 2.6371,
"step": 5500
},
{
"epoch": 0.47,
"learning_rate": 0.0002,
"loss": 2.6151,
"step": 6000
},
{
"epoch": 0.5,
"learning_rate": 0.0002,
"loss": 2.5689,
"step": 6500
},
{
"epoch": 0.54,
"learning_rate": 0.0002,
"loss": 2.5371,
"step": 7000
},
{
"epoch": 0.58,
"learning_rate": 0.0002,
"loss": 2.5162,
"step": 7500
},
{
"epoch": 0.62,
"learning_rate": 0.0002,
"loss": 2.4735,
"step": 8000
},
{
"epoch": 0.66,
"learning_rate": 0.0002,
"loss": 2.4349,
"step": 8500
},
{
"epoch": 0.7,
"learning_rate": 0.0002,
"loss": 2.4207,
"step": 9000
},
{
"epoch": 0.74,
"learning_rate": 0.0002,
"loss": 2.3972,
"step": 9500
},
{
"epoch": 0.78,
"learning_rate": 0.0002,
"loss": 2.3552,
"step": 10000
},
{
"epoch": 0.82,
"learning_rate": 0.0002,
"loss": 2.3716,
"step": 10500
},
{
"epoch": 0.85,
"learning_rate": 0.0002,
"loss": 2.3403,
"step": 11000
},
{
"epoch": 0.89,
"learning_rate": 0.0002,
"loss": 2.3245,
"step": 11500
},
{
"epoch": 0.93,
"learning_rate": 0.0002,
"loss": 2.2925,
"step": 12000
},
{
"epoch": 0.97,
"learning_rate": 0.0002,
"loss": 2.284,
"step": 12500
},
{
"epoch": 1.0,
"eval_bleu": 14.1725,
"eval_gen_len": 16.7883,
"eval_runtime": 62.032,
"eval_samples_per_second": 14.315,
"eval_steps_per_second": 1.789,
"step": 12882
},
{
"epoch": 1.01,
"learning_rate": 0.0002,
"loss": 2.2649,
"step": 13000
},
{
"epoch": 1.05,
"learning_rate": 0.0002,
"loss": 2.1906,
"step": 13500
},
{
"epoch": 1.09,
"learning_rate": 0.0002,
"loss": 2.2009,
"step": 14000
},
{
"epoch": 1.13,
"learning_rate": 0.0002,
"loss": 2.1794,
"step": 14500
},
{
"epoch": 1.16,
"learning_rate": 0.0002,
"loss": 2.1502,
"step": 15000
},
{
"epoch": 1.2,
"learning_rate": 0.0002,
"loss": 2.1532,
"step": 15500
},
{
"epoch": 1.24,
"learning_rate": 0.0002,
"loss": 2.1426,
"step": 16000
},
{
"epoch": 1.28,
"learning_rate": 0.0002,
"loss": 2.1453,
"step": 16500
},
{
"epoch": 1.32,
"learning_rate": 0.0002,
"loss": 2.1078,
"step": 17000
},
{
"epoch": 1.36,
"learning_rate": 0.0002,
"loss": 2.1112,
"step": 17500
},
{
"epoch": 1.4,
"learning_rate": 0.0002,
"loss": 2.0962,
"step": 18000
},
{
"epoch": 1.44,
"learning_rate": 0.0002,
"loss": 2.0837,
"step": 18500
},
{
"epoch": 1.47,
"learning_rate": 0.0002,
"loss": 2.0823,
"step": 19000
},
{
"epoch": 1.51,
"learning_rate": 0.0002,
"loss": 2.0657,
"step": 19500
},
{
"epoch": 1.55,
"learning_rate": 0.0002,
"loss": 2.0495,
"step": 20000
},
{
"epoch": 1.59,
"learning_rate": 0.0002,
"loss": 2.0737,
"step": 20500
},
{
"epoch": 1.63,
"learning_rate": 0.0002,
"loss": 2.066,
"step": 21000
},
{
"epoch": 1.67,
"learning_rate": 0.0002,
"loss": 2.0381,
"step": 21500
},
{
"epoch": 1.71,
"learning_rate": 0.0002,
"loss": 2.0529,
"step": 22000
},
{
"epoch": 1.75,
"learning_rate": 0.0002,
"loss": 2.0334,
"step": 22500
},
{
"epoch": 1.79,
"learning_rate": 0.0002,
"loss": 2.0114,
"step": 23000
},
{
"epoch": 1.82,
"learning_rate": 0.0002,
"loss": 2.0103,
"step": 23500
},
{
"epoch": 1.86,
"learning_rate": 0.0002,
"loss": 2.012,
"step": 24000
},
{
"epoch": 1.9,
"learning_rate": 0.0002,
"loss": 1.9964,
"step": 24500
},
{
"epoch": 1.94,
"learning_rate": 0.0002,
"loss": 1.9882,
"step": 25000
},
{
"epoch": 1.98,
"learning_rate": 0.0002,
"loss": 1.983,
"step": 25500
},
{
"epoch": 2.0,
"eval_bleu": 15.2401,
"eval_gen_len": 16.8164,
"eval_runtime": 62.0852,
"eval_samples_per_second": 14.303,
"eval_steps_per_second": 1.788,
"step": 25764
},
{
"epoch": 2.02,
"learning_rate": 0.0002,
"loss": 1.928,
"step": 26000
},
{
"epoch": 2.06,
"learning_rate": 0.0002,
"loss": 1.9137,
"step": 26500
},
{
"epoch": 2.1,
"learning_rate": 0.0002,
"loss": 1.9315,
"step": 27000
},
{
"epoch": 2.13,
"learning_rate": 0.0002,
"loss": 1.9083,
"step": 27500
},
{
"epoch": 2.17,
"learning_rate": 0.0002,
"loss": 1.9119,
"step": 28000
},
{
"epoch": 2.21,
"learning_rate": 0.0002,
"loss": 1.9246,
"step": 28500
},
{
"epoch": 2.25,
"learning_rate": 0.0002,
"loss": 1.913,
"step": 29000
},
{
"epoch": 2.29,
"learning_rate": 0.0002,
"loss": 1.8766,
"step": 29500
},
{
"epoch": 2.33,
"learning_rate": 0.0002,
"loss": 1.8941,
"step": 30000
},
{
"epoch": 2.37,
"learning_rate": 0.0002,
"loss": 1.8892,
"step": 30500
},
{
"epoch": 2.41,
"learning_rate": 0.0002,
"loss": 1.8909,
"step": 31000
},
{
"epoch": 2.45,
"learning_rate": 0.0002,
"loss": 1.8816,
"step": 31500
},
{
"epoch": 2.48,
"learning_rate": 0.0002,
"loss": 1.8775,
"step": 32000
},
{
"epoch": 2.52,
"learning_rate": 0.0002,
"loss": 1.8757,
"step": 32500
},
{
"epoch": 2.56,
"learning_rate": 0.0002,
"loss": 1.8787,
"step": 33000
},
{
"epoch": 2.6,
"learning_rate": 0.0002,
"loss": 1.8606,
"step": 33500
},
{
"epoch": 2.64,
"learning_rate": 0.0002,
"loss": 1.8599,
"step": 34000
},
{
"epoch": 2.68,
"learning_rate": 0.0002,
"loss": 1.8589,
"step": 34500
},
{
"epoch": 2.72,
"learning_rate": 0.0002,
"loss": 1.8403,
"step": 35000
},
{
"epoch": 2.76,
"learning_rate": 0.0002,
"loss": 1.8449,
"step": 35500
},
{
"epoch": 2.79,
"learning_rate": 0.0002,
"loss": 1.8379,
"step": 36000
},
{
"epoch": 2.83,
"learning_rate": 0.0002,
"loss": 1.8604,
"step": 36500
},
{
"epoch": 2.87,
"learning_rate": 0.0002,
"loss": 1.8379,
"step": 37000
},
{
"epoch": 2.91,
"learning_rate": 0.0002,
"loss": 1.8496,
"step": 37500
},
{
"epoch": 2.95,
"learning_rate": 0.0002,
"loss": 1.8231,
"step": 38000
},
{
"epoch": 2.99,
"learning_rate": 0.0002,
"loss": 1.8309,
"step": 38500
},
{
"epoch": 3.0,
"eval_bleu": 0,
"eval_gen_len": 16.8131,
"eval_runtime": 60.7852,
"eval_samples_per_second": 14.609,
"eval_steps_per_second": 1.826,
"step": 38646
},
{
"epoch": 3.03,
"learning_rate": 0.0002,
"loss": 1.7842,
"step": 39000
},
{
"epoch": 3.07,
"learning_rate": 0.0002,
"loss": 1.7731,
"step": 39500
},
{
"epoch": 3.11,
"learning_rate": 0.0002,
"loss": 1.7659,
"step": 40000
},
{
"epoch": 3.14,
"learning_rate": 0.0002,
"loss": 1.7779,
"step": 40500
},
{
"epoch": 3.18,
"learning_rate": 0.0002,
"loss": 1.7633,
"step": 41000
},
{
"epoch": 3.22,
"learning_rate": 0.0002,
"loss": 1.7708,
"step": 41500
},
{
"epoch": 3.26,
"learning_rate": 0.0002,
"loss": 1.7488,
"step": 42000
},
{
"epoch": 3.3,
"learning_rate": 0.0002,
"loss": 1.764,
"step": 42500
},
{
"epoch": 3.34,
"learning_rate": 0.0002,
"loss": 1.7541,
"step": 43000
},
{
"epoch": 3.38,
"learning_rate": 0.0002,
"loss": 1.7493,
"step": 43500
},
{
"epoch": 3.42,
"learning_rate": 0.0002,
"loss": 1.7539,
"step": 44000
},
{
"epoch": 3.45,
"learning_rate": 0.0002,
"loss": 1.7531,
"step": 44500
},
{
"epoch": 3.49,
"learning_rate": 0.0002,
"loss": 1.7575,
"step": 45000
},
{
"epoch": 3.53,
"learning_rate": 0.0002,
"loss": 1.7534,
"step": 45500
},
{
"epoch": 3.57,
"learning_rate": 0.0002,
"loss": 1.7539,
"step": 46000
},
{
"epoch": 3.61,
"learning_rate": 0.0002,
"loss": 1.7335,
"step": 46500
},
{
"epoch": 3.65,
"learning_rate": 0.0002,
"loss": 1.7315,
"step": 47000
},
{
"epoch": 3.69,
"learning_rate": 0.0002,
"loss": 1.7341,
"step": 47500
},
{
"epoch": 3.73,
"learning_rate": 0.0002,
"loss": 1.7319,
"step": 48000
},
{
"epoch": 3.76,
"learning_rate": 0.0002,
"loss": 1.7486,
"step": 48500
},
{
"epoch": 3.8,
"learning_rate": 0.0002,
"loss": 1.7416,
"step": 49000
},
{
"epoch": 3.84,
"learning_rate": 0.0002,
"loss": 1.7299,
"step": 49500
},
{
"epoch": 3.88,
"learning_rate": 0.0002,
"loss": 1.721,
"step": 50000
},
{
"epoch": 3.92,
"learning_rate": 0.0002,
"loss": 1.7332,
"step": 50500
},
{
"epoch": 3.96,
"learning_rate": 0.0002,
"loss": 1.7229,
"step": 51000
},
{
"epoch": 4.0,
"learning_rate": 0.0002,
"loss": 1.7279,
"step": 51500
},
{
"epoch": 4.0,
"eval_bleu": 16.8845,
"eval_gen_len": 16.7917,
"eval_runtime": 62.1153,
"eval_samples_per_second": 14.296,
"eval_steps_per_second": 1.787,
"step": 51528
},
{
"epoch": 4.04,
"learning_rate": 0.0002,
"loss": 1.6815,
"step": 52000
},
{
"epoch": 4.08,
"learning_rate": 0.0002,
"loss": 1.6451,
"step": 52500
},
{
"epoch": 4.11,
"learning_rate": 0.0002,
"loss": 1.673,
"step": 53000
},
{
"epoch": 4.15,
"learning_rate": 0.0002,
"loss": 1.6756,
"step": 53500
},
{
"epoch": 4.19,
"learning_rate": 0.0002,
"loss": 1.6706,
"step": 54000
},
{
"epoch": 4.23,
"learning_rate": 0.0002,
"loss": 1.668,
"step": 54500
},
{
"epoch": 4.27,
"learning_rate": 0.0002,
"loss": 1.6641,
"step": 55000
},
{
"epoch": 4.31,
"learning_rate": 0.0002,
"loss": 1.6641,
"step": 55500
},
{
"epoch": 4.35,
"learning_rate": 0.0002,
"loss": 1.6609,
"step": 56000
},
{
"epoch": 4.39,
"learning_rate": 0.0002,
"loss": 1.6567,
"step": 56500
},
{
"epoch": 4.42,
"learning_rate": 0.0002,
"loss": 1.6656,
"step": 57000
},
{
"epoch": 4.46,
"learning_rate": 0.0002,
"loss": 1.6701,
"step": 57500
},
{
"epoch": 4.5,
"learning_rate": 0.0002,
"loss": 1.6549,
"step": 58000
},
{
"epoch": 4.54,
"learning_rate": 0.0002,
"loss": 1.6589,
"step": 58500
},
{
"epoch": 4.58,
"learning_rate": 0.0002,
"loss": 1.656,
"step": 59000
},
{
"epoch": 4.62,
"learning_rate": 0.0002,
"loss": 1.6644,
"step": 59500
},
{
"epoch": 4.66,
"learning_rate": 0.0002,
"loss": 1.6488,
"step": 60000
},
{
"epoch": 4.7,
"learning_rate": 0.0002,
"loss": 1.6564,
"step": 60500
},
{
"epoch": 4.74,
"learning_rate": 0.0002,
"loss": 1.652,
"step": 61000
},
{
"epoch": 4.77,
"learning_rate": 0.0002,
"loss": 1.647,
"step": 61500
},
{
"epoch": 4.81,
"learning_rate": 0.0002,
"loss": 1.6459,
"step": 62000
},
{
"epoch": 4.85,
"learning_rate": 0.0002,
"loss": 1.6575,
"step": 62500
},
{
"epoch": 4.89,
"learning_rate": 0.0002,
"loss": 1.6462,
"step": 63000
},
{
"epoch": 4.93,
"learning_rate": 0.0002,
"loss": 1.6542,
"step": 63500
},
{
"epoch": 4.97,
"learning_rate": 0.0002,
"loss": 1.6464,
"step": 64000
},
{
"epoch": 5.0,
"eval_bleu": 17.3723,
"eval_gen_len": 16.7883,
"eval_runtime": 62.375,
"eval_samples_per_second": 14.236,
"eval_steps_per_second": 1.78,
"step": 64410
},
{
"epoch": 5.01,
"learning_rate": 0.0002,
"loss": 1.6381,
"step": 64500
},
{
"epoch": 5.05,
"learning_rate": 0.0002,
"loss": 1.59,
"step": 65000
},
{
"epoch": 5.08,
"learning_rate": 0.0002,
"loss": 1.5795,
"step": 65500
},
{
"epoch": 5.12,
"learning_rate": 0.0002,
"loss": 1.5891,
"step": 66000
},
{
"epoch": 5.16,
"learning_rate": 0.0002,
"loss": 1.5908,
"step": 66500
},
{
"epoch": 5.2,
"learning_rate": 0.0002,
"loss": 1.5897,
"step": 67000
},
{
"epoch": 5.24,
"learning_rate": 0.0002,
"loss": 1.5907,
"step": 67500
},
{
"epoch": 5.28,
"learning_rate": 0.0002,
"loss": 1.5953,
"step": 68000
},
{
"epoch": 5.32,
"learning_rate": 0.0002,
"loss": 1.5882,
"step": 68500
},
{
"epoch": 5.36,
"learning_rate": 0.0002,
"loss": 1.5908,
"step": 69000
},
{
"epoch": 5.4,
"learning_rate": 0.0002,
"loss": 1.5877,
"step": 69500
},
{
"epoch": 5.43,
"learning_rate": 0.0002,
"loss": 1.5938,
"step": 70000
},
{
"epoch": 5.47,
"learning_rate": 0.0002,
"loss": 1.5944,
"step": 70500
},
{
"epoch": 5.51,
"learning_rate": 0.0002,
"loss": 1.5955,
"step": 71000
},
{
"epoch": 5.55,
"learning_rate": 0.0002,
"loss": 1.6004,
"step": 71500
},
{
"epoch": 5.59,
"learning_rate": 0.0002,
"loss": 1.5871,
"step": 72000
},
{
"epoch": 5.63,
"learning_rate": 0.0002,
"loss": 1.5908,
"step": 72500
},
{
"epoch": 5.67,
"learning_rate": 0.0002,
"loss": 1.5789,
"step": 73000
},
{
"epoch": 5.71,
"learning_rate": 0.0002,
"loss": 1.5931,
"step": 73500
},
{
"epoch": 5.74,
"learning_rate": 0.0002,
"loss": 1.5884,
"step": 74000
},
{
"epoch": 5.78,
"learning_rate": 0.0002,
"loss": 1.5845,
"step": 74500
},
{
"epoch": 5.82,
"learning_rate": 0.0002,
"loss": 1.5895,
"step": 75000
},
{
"epoch": 5.86,
"learning_rate": 0.0002,
"loss": 1.5705,
"step": 75500
},
{
"epoch": 5.9,
"learning_rate": 0.0002,
"loss": 1.5767,
"step": 76000
},
{
"epoch": 5.94,
"learning_rate": 0.0002,
"loss": 1.5966,
"step": 76500
},
{
"epoch": 5.98,
"learning_rate": 0.0002,
"loss": 1.5795,
"step": 77000
},
{
"epoch": 6.0,
"eval_bleu": 17.5592,
"eval_gen_len": 16.8041,
"eval_runtime": 61.8292,
"eval_samples_per_second": 14.362,
"eval_steps_per_second": 1.795,
"step": 77292
},
{
"epoch": 6.02,
"learning_rate": 0.0002,
"loss": 1.5617,
"step": 77500
},
{
"epoch": 6.05,
"learning_rate": 0.0002,
"loss": 1.5107,
"step": 78000
},
{
"epoch": 6.09,
"learning_rate": 0.0002,
"loss": 1.5298,
"step": 78500
},
{
"epoch": 6.13,
"learning_rate": 0.0002,
"loss": 1.526,
"step": 79000
},
{
"epoch": 6.17,
"learning_rate": 0.0002,
"loss": 1.541,
"step": 79500
},
{
"epoch": 6.21,
"learning_rate": 0.0002,
"loss": 1.5309,
"step": 80000
},
{
"epoch": 6.25,
"learning_rate": 0.0002,
"loss": 1.5233,
"step": 80500
},
{
"epoch": 6.29,
"learning_rate": 0.0002,
"loss": 1.5424,
"step": 81000
},
{
"epoch": 6.33,
"learning_rate": 0.0002,
"loss": 1.5396,
"step": 81500
},
{
"epoch": 6.37,
"learning_rate": 0.0002,
"loss": 1.5217,
"step": 82000
},
{
"epoch": 6.4,
"learning_rate": 0.0002,
"loss": 1.5278,
"step": 82500
},
{
"epoch": 6.44,
"learning_rate": 0.0002,
"loss": 1.5247,
"step": 83000
},
{
"epoch": 6.48,
"learning_rate": 0.0002,
"loss": 1.5293,
"step": 83500
},
{
"epoch": 6.52,
"learning_rate": 0.0002,
"loss": 1.5262,
"step": 84000
},
{
"epoch": 6.56,
"learning_rate": 0.0002,
"loss": 1.5309,
"step": 84500
},
{
"epoch": 6.6,
"learning_rate": 0.0002,
"loss": 1.5257,
"step": 85000
},
{
"epoch": 6.64,
"learning_rate": 0.0002,
"loss": 1.5324,
"step": 85500
},
{
"epoch": 6.68,
"learning_rate": 0.0002,
"loss": 1.5287,
"step": 86000
},
{
"epoch": 6.71,
"learning_rate": 0.0002,
"loss": 1.5425,
"step": 86500
},
{
"epoch": 6.75,
"learning_rate": 0.0002,
"loss": 1.5344,
"step": 87000
},
{
"epoch": 6.79,
"learning_rate": 0.0002,
"loss": 1.5398,
"step": 87500
},
{
"epoch": 6.83,
"learning_rate": 0.0002,
"loss": 1.524,
"step": 88000
},
{
"epoch": 6.87,
"learning_rate": 0.0002,
"loss": 1.5128,
"step": 88500
},
{
"epoch": 6.91,
"learning_rate": 0.0002,
"loss": 1.5298,
"step": 89000
},
{
"epoch": 6.95,
"learning_rate": 0.0002,
"loss": 1.5346,
"step": 89500
},
{
"epoch": 6.99,
"learning_rate": 0.0002,
"loss": 1.5285,
"step": 90000
},
{
"epoch": 7.0,
"eval_bleu": 17.5601,
"eval_gen_len": 16.8041,
"eval_runtime": 62.4423,
"eval_samples_per_second": 14.221,
"eval_steps_per_second": 1.778,
"step": 90174
},
{
"epoch": 7.03,
"learning_rate": 0.0002,
"loss": 1.4934,
"step": 90500
},
{
"epoch": 7.06,
"learning_rate": 0.0002,
"loss": 1.4686,
"step": 91000
},
{
"epoch": 7.1,
"learning_rate": 0.0002,
"loss": 1.4853,
"step": 91500
},
{
"epoch": 7.14,
"learning_rate": 0.0002,
"loss": 1.4662,
"step": 92000
},
{
"epoch": 7.18,
"learning_rate": 0.0002,
"loss": 1.4743,
"step": 92500
},
{
"epoch": 7.22,
"learning_rate": 0.0002,
"loss": 1.4778,
"step": 93000
},
{
"epoch": 7.26,
"learning_rate": 0.0002,
"loss": 1.4742,
"step": 93500
},
{
"epoch": 7.3,
"learning_rate": 0.0002,
"loss": 1.489,
"step": 94000
},
{
"epoch": 7.34,
"learning_rate": 0.0002,
"loss": 1.4726,
"step": 94500
},
{
"epoch": 7.37,
"learning_rate": 0.0002,
"loss": 1.4901,
"step": 95000
},
{
"epoch": 7.41,
"learning_rate": 0.0002,
"loss": 1.4748,
"step": 95500
},
{
"epoch": 7.45,
"learning_rate": 0.0002,
"loss": 1.4775,
"step": 96000
},
{
"epoch": 7.49,
"learning_rate": 0.0002,
"loss": 1.4838,
"step": 96500
},
{
"epoch": 7.53,
"learning_rate": 0.0002,
"loss": 1.4861,
"step": 97000
},
{
"epoch": 7.57,
"learning_rate": 0.0002,
"loss": 1.4815,
"step": 97500
},
{
"epoch": 7.61,
"learning_rate": 0.0002,
"loss": 1.4732,
"step": 98000
},
{
"epoch": 7.65,
"learning_rate": 0.0002,
"loss": 1.4709,
"step": 98500
},
{
"epoch": 7.69,
"learning_rate": 0.0002,
"loss": 1.4873,
"step": 99000
},
{
"epoch": 7.72,
"learning_rate": 0.0002,
"loss": 1.4785,
"step": 99500
},
{
"epoch": 7.76,
"learning_rate": 0.0002,
"loss": 1.4792,
"step": 100000
},
{
"epoch": 7.8,
"learning_rate": 0.0002,
"loss": 1.4881,
"step": 100500
},
{
"epoch": 7.84,
"learning_rate": 0.0002,
"loss": 1.4709,
"step": 101000
},
{
"epoch": 7.88,
"learning_rate": 0.0002,
"loss": 1.4892,
"step": 101500
},
{
"epoch": 7.92,
"learning_rate": 0.0002,
"loss": 1.489,
"step": 102000
},
{
"epoch": 7.96,
"learning_rate": 0.0002,
"loss": 1.4846,
"step": 102500
},
{
"epoch": 8.0,
"learning_rate": 0.0002,
"loss": 1.4847,
"step": 103000
},
{
"epoch": 8.0,
"eval_bleu": 17.8615,
"eval_gen_len": 16.7556,
"eval_runtime": 62.4342,
"eval_samples_per_second": 14.223,
"eval_steps_per_second": 1.778,
"step": 103056
},
{
"epoch": 8.03,
"learning_rate": 0.0002,
"loss": 1.4231,
"step": 103500
},
{
"epoch": 8.07,
"learning_rate": 0.0002,
"loss": 1.4246,
"step": 104000
},
{
"epoch": 8.11,
"learning_rate": 0.0002,
"loss": 1.4217,
"step": 104500
},
{
"epoch": 8.15,
"learning_rate": 0.0002,
"loss": 1.4344,
"step": 105000
},
{
"epoch": 8.19,
"learning_rate": 0.0002,
"loss": 1.4292,
"step": 105500
},
{
"epoch": 8.23,
"learning_rate": 0.0002,
"loss": 1.4268,
"step": 106000
},
{
"epoch": 8.27,
"learning_rate": 0.0002,
"loss": 1.4445,
"step": 106500
},
{
"epoch": 8.31,
"learning_rate": 0.0002,
"loss": 1.4403,
"step": 107000
},
{
"epoch": 8.34,
"learning_rate": 0.0002,
"loss": 1.4356,
"step": 107500
},
{
"epoch": 8.38,
"learning_rate": 0.0002,
"loss": 1.437,
"step": 108000
},
{
"epoch": 8.42,
"learning_rate": 0.0002,
"loss": 1.4413,
"step": 108500
},
{
"epoch": 8.46,
"learning_rate": 0.0002,
"loss": 1.4292,
"step": 109000
},
{
"epoch": 8.5,
"learning_rate": 0.0002,
"loss": 1.4405,
"step": 109500
},
{
"epoch": 8.54,
"learning_rate": 0.0002,
"loss": 1.4414,
"step": 110000
},
{
"epoch": 8.58,
"learning_rate": 0.0002,
"loss": 1.4304,
"step": 110500
},
{
"epoch": 8.62,
"learning_rate": 0.0002,
"loss": 1.4427,
"step": 111000
},
{
"epoch": 8.66,
"learning_rate": 0.0002,
"loss": 1.4404,
"step": 111500
},
{
"epoch": 8.69,
"learning_rate": 0.0002,
"loss": 1.4392,
"step": 112000
},
{
"epoch": 8.73,
"learning_rate": 0.0002,
"loss": 1.4444,
"step": 112500
},
{
"epoch": 8.77,
"learning_rate": 0.0002,
"loss": 1.4355,
"step": 113000
},
{
"epoch": 8.81,
"learning_rate": 0.0002,
"loss": 1.4456,
"step": 113500
},
{
"epoch": 8.85,
"learning_rate": 0.0002,
"loss": 1.4387,
"step": 114000
},
{
"epoch": 8.89,
"learning_rate": 0.0002,
"loss": 1.4412,
"step": 114500
},
{
"epoch": 8.93,
"learning_rate": 0.0002,
"loss": 1.4467,
"step": 115000
},
{
"epoch": 8.97,
"learning_rate": 0.0002,
"loss": 1.434,
"step": 115500
},
{
"epoch": 9.0,
"eval_bleu": 17.6009,
"eval_gen_len": 16.8119,
"eval_runtime": 62.3237,
"eval_samples_per_second": 14.248,
"eval_steps_per_second": 1.781,
"step": 115938
},
{
"epoch": 9.0,
"learning_rate": 0.0002,
"loss": 1.4369,
"step": 116000
},
{
"epoch": 9.04,
"learning_rate": 0.0002,
"loss": 1.3918,
"step": 116500
},
{
"epoch": 9.08,
"learning_rate": 0.0002,
"loss": 1.3817,
"step": 117000
},
{
"epoch": 9.12,
"learning_rate": 0.0002,
"loss": 1.3779,
"step": 117500
},
{
"epoch": 9.16,
"learning_rate": 0.0002,
"loss": 1.4022,
"step": 118000
},
{
"epoch": 9.2,
"learning_rate": 0.0002,
"loss": 1.3981,
"step": 118500
},
{
"epoch": 9.24,
"learning_rate": 0.0002,
"loss": 1.377,
"step": 119000
},
{
"epoch": 9.28,
"learning_rate": 0.0002,
"loss": 1.3755,
"step": 119500
},
{
"epoch": 9.32,
"learning_rate": 0.0002,
"loss": 1.392,
"step": 120000
},
{
"epoch": 9.35,
"learning_rate": 0.0002,
"loss": 1.3988,
"step": 120500
},
{
"epoch": 9.39,
"learning_rate": 0.0002,
"loss": 1.3948,
"step": 121000
},
{
"epoch": 9.43,
"learning_rate": 0.0002,
"loss": 1.4034,
"step": 121500
},
{
"epoch": 9.47,
"learning_rate": 0.0002,
"loss": 1.4071,
"step": 122000
},
{
"epoch": 9.51,
"learning_rate": 0.0002,
"loss": 1.4058,
"step": 122500
},
{
"epoch": 9.55,
"learning_rate": 0.0002,
"loss": 1.4119,
"step": 123000
},
{
"epoch": 9.59,
"learning_rate": 0.0002,
"loss": 1.4018,
"step": 123500
},
{
"epoch": 9.63,
"learning_rate": 0.0002,
"loss": 1.3981,
"step": 124000
},
{
"epoch": 9.66,
"learning_rate": 0.0002,
"loss": 1.3932,
"step": 124500
},
{
"epoch": 9.7,
"learning_rate": 0.0002,
"loss": 1.4005,
"step": 125000
},
{
"epoch": 9.74,
"learning_rate": 0.0002,
"loss": 1.3948,
"step": 125500
},
{
"epoch": 9.78,
"learning_rate": 0.0002,
"loss": 1.4193,
"step": 126000
},
{
"epoch": 9.82,
"learning_rate": 0.0002,
"loss": 1.4066,
"step": 126500
},
{
"epoch": 9.86,
"learning_rate": 0.0002,
"loss": 1.3997,
"step": 127000
},
{
"epoch": 9.9,
"learning_rate": 0.0002,
"loss": 1.4008,
"step": 127500
},
{
"epoch": 9.94,
"learning_rate": 0.0002,
"loss": 1.4135,
"step": 128000
},
{
"epoch": 9.98,
"learning_rate": 0.0002,
"loss": 1.415,
"step": 128500
},
{
"epoch": 10.0,
"eval_bleu": 18.0078,
"eval_gen_len": 16.8097,
"eval_runtime": 62.2078,
"eval_samples_per_second": 14.275,
"eval_steps_per_second": 1.784,
"step": 128820
},
{
"epoch": 10.01,
"learning_rate": 0.0002,
"loss": 1.3898,
"step": 129000
},
{
"epoch": 10.05,
"learning_rate": 0.0002,
"loss": 1.3547,
"step": 129500
},
{
"epoch": 10.09,
"learning_rate": 0.0002,
"loss": 1.3512,
"step": 130000
},
{
"epoch": 10.13,
"learning_rate": 0.0002,
"loss": 1.3527,
"step": 130500
},
{
"epoch": 10.17,
"learning_rate": 0.0002,
"loss": 1.3487,
"step": 131000
},
{
"epoch": 10.21,
"learning_rate": 0.0002,
"loss": 1.3566,
"step": 131500
},
{
"epoch": 10.25,
"learning_rate": 0.0002,
"loss": 1.3476,
"step": 132000
},
{
"epoch": 10.29,
"learning_rate": 0.0002,
"loss": 1.3684,
"step": 132500
},
{
"epoch": 10.32,
"learning_rate": 0.0002,
"loss": 1.3503,
"step": 133000
},
{
"epoch": 10.36,
"learning_rate": 0.0002,
"loss": 1.3608,
"step": 133500
},
{
"epoch": 10.4,
"learning_rate": 0.0002,
"loss": 1.3705,
"step": 134000
},
{
"epoch": 10.44,
"learning_rate": 0.0002,
"loss": 1.3632,
"step": 134500
},
{
"epoch": 10.48,
"learning_rate": 0.0002,
"loss": 1.373,
"step": 135000
},
{
"epoch": 10.52,
"learning_rate": 0.0002,
"loss": 1.3649,
"step": 135500
},
{
"epoch": 10.56,
"learning_rate": 0.0002,
"loss": 1.365,
"step": 136000
},
{
"epoch": 10.6,
"learning_rate": 0.0002,
"loss": 1.3763,
"step": 136500
},
{
"epoch": 10.63,
"learning_rate": 0.0002,
"loss": 1.3545,
"step": 137000
},
{
"epoch": 10.67,
"learning_rate": 0.0002,
"loss": 1.369,
"step": 137500
},
{
"epoch": 10.71,
"learning_rate": 0.0002,
"loss": 1.3709,
"step": 138000
},
{
"epoch": 10.75,
"learning_rate": 0.0002,
"loss": 1.3651,
"step": 138500
},
{
"epoch": 10.79,
"learning_rate": 0.0002,
"loss": 1.3638,
"step": 139000
},
{
"epoch": 10.83,
"learning_rate": 0.0002,
"loss": 1.391,
"step": 139500
},
{
"epoch": 10.87,
"learning_rate": 0.0002,
"loss": 1.3729,
"step": 140000
},
{
"epoch": 10.91,
"learning_rate": 0.0002,
"loss": 1.3748,
"step": 140500
},
{
"epoch": 10.95,
"learning_rate": 0.0002,
"loss": 1.3666,
"step": 141000
},
{
"epoch": 10.98,
"learning_rate": 0.0002,
"loss": 1.3733,
"step": 141500
},
{
"epoch": 11.0,
"eval_bleu": 17.9142,
"eval_gen_len": 16.7815,
"eval_runtime": 62.3004,
"eval_samples_per_second": 14.254,
"eval_steps_per_second": 1.782,
"step": 141702
},
{
"epoch": 11.02,
"learning_rate": 0.0002,
"loss": 1.3412,
"step": 142000
},
{
"epoch": 11.06,
"learning_rate": 0.0002,
"loss": 1.3186,
"step": 142500
},
{
"epoch": 11.1,
"learning_rate": 0.0002,
"loss": 1.327,
"step": 143000
},
{
"epoch": 11.14,
"learning_rate": 0.0002,
"loss": 1.3126,
"step": 143500
},
{
"epoch": 11.18,
"learning_rate": 0.0002,
"loss": 1.3187,
"step": 144000
},
{
"epoch": 11.22,
"learning_rate": 0.0002,
"loss": 1.3258,
"step": 144500
},
{
"epoch": 11.26,
"learning_rate": 0.0002,
"loss": 1.3178,
"step": 145000
},
{
"epoch": 11.29,
"learning_rate": 0.0002,
"loss": 1.3237,
"step": 145500
},
{
"epoch": 11.33,
"learning_rate": 0.0002,
"loss": 1.3379,
"step": 146000
},
{
"epoch": 11.37,
"learning_rate": 0.0002,
"loss": 1.3255,
"step": 146500
},
{
"epoch": 11.41,
"learning_rate": 0.0002,
"loss": 1.3352,
"step": 147000
},
{
"epoch": 11.45,
"learning_rate": 0.0002,
"loss": 1.3338,
"step": 147500
},
{
"epoch": 11.49,
"learning_rate": 0.0002,
"loss": 1.3406,
"step": 148000
},
{
"epoch": 11.53,
"learning_rate": 0.0002,
"loss": 1.3328,
"step": 148500
},
{
"epoch": 11.57,
"learning_rate": 0.0002,
"loss": 1.3434,
"step": 149000
},
{
"epoch": 11.61,
"learning_rate": 0.0002,
"loss": 1.3338,
"step": 149500
},
{
"epoch": 11.64,
"learning_rate": 0.0002,
"loss": 1.349,
"step": 150000
},
{
"epoch": 11.68,
"learning_rate": 0.0002,
"loss": 1.3344,
"step": 150500
},
{
"epoch": 11.72,
"learning_rate": 0.0002,
"loss": 1.3347,
"step": 151000
},
{
"epoch": 11.76,
"learning_rate": 0.0002,
"loss": 1.3376,
"step": 151500
},
{
"epoch": 11.8,
"learning_rate": 0.0002,
"loss": 1.349,
"step": 152000
},
{
"epoch": 11.84,
"learning_rate": 0.0002,
"loss": 1.3412,
"step": 152500
},
{
"epoch": 11.88,
"learning_rate": 0.0002,
"loss": 1.3351,
"step": 153000
},
{
"epoch": 11.92,
"learning_rate": 0.0002,
"loss": 1.3326,
"step": 153500
},
{
"epoch": 11.95,
"learning_rate": 0.0002,
"loss": 1.3479,
"step": 154000
},
{
"epoch": 11.99,
"learning_rate": 0.0002,
"loss": 1.347,
"step": 154500
},
{
"epoch": 12.0,
"eval_bleu": 17.9973,
"eval_gen_len": 16.8029,
"eval_runtime": 62.4045,
"eval_samples_per_second": 14.23,
"eval_steps_per_second": 1.779,
"step": 154584
},
{
"epoch": 12.03,
"learning_rate": 0.0002,
"loss": 1.2939,
"step": 155000
},
{
"epoch": 12.07,
"learning_rate": 0.0002,
"loss": 1.3035,
"step": 155500
},
{
"epoch": 12.11,
"learning_rate": 0.0002,
"loss": 1.2903,
"step": 156000
},
{
"epoch": 12.15,
"learning_rate": 0.0002,
"loss": 1.2951,
"step": 156500
},
{
"epoch": 12.19,
"learning_rate": 0.0002,
"loss": 1.2918,
"step": 157000
},
{
"epoch": 12.23,
"learning_rate": 0.0002,
"loss": 1.3008,
"step": 157500
},
{
"epoch": 12.27,
"learning_rate": 0.0002,
"loss": 1.2957,
"step": 158000
},
{
"epoch": 12.3,
"learning_rate": 0.0002,
"loss": 1.3047,
"step": 158500
},
{
"epoch": 12.34,
"learning_rate": 0.0002,
"loss": 1.3034,
"step": 159000
},
{
"epoch": 12.38,
"learning_rate": 0.0002,
"loss": 1.3015,
"step": 159500
},
{
"epoch": 12.42,
"learning_rate": 0.0002,
"loss": 1.3001,
"step": 160000
},
{
"epoch": 12.46,
"learning_rate": 0.0002,
"loss": 1.3176,
"step": 160500
},
{
"epoch": 12.5,
"learning_rate": 0.0002,
"loss": 1.3133,
"step": 161000
},
{
"epoch": 12.54,
"learning_rate": 0.0002,
"loss": 1.308,
"step": 161500
},
{
"epoch": 12.58,
"learning_rate": 0.0002,
"loss": 1.3032,
"step": 162000
},
{
"epoch": 12.61,
"learning_rate": 0.0002,
"loss": 1.3071,
"step": 162500
},
{
"epoch": 12.65,
"learning_rate": 0.0002,
"loss": 1.3081,
"step": 163000
},
{
"epoch": 12.69,
"learning_rate": 0.0002,
"loss": 1.3174,
"step": 163500
},
{
"epoch": 12.73,
"learning_rate": 0.0002,
"loss": 1.3196,
"step": 164000
},
{
"epoch": 12.77,
"learning_rate": 0.0002,
"loss": 1.3074,
"step": 164500
},
{
"epoch": 12.81,
"learning_rate": 0.0002,
"loss": 1.3075,
"step": 165000
},
{
"epoch": 12.85,
"learning_rate": 0.0002,
"loss": 1.3046,
"step": 165500
},
{
"epoch": 12.89,
"learning_rate": 0.0002,
"loss": 1.3195,
"step": 166000
},
{
"epoch": 12.93,
"learning_rate": 0.0002,
"loss": 1.3124,
"step": 166500
},
{
"epoch": 12.96,
"learning_rate": 0.0002,
"loss": 1.2959,
"step": 167000
},
{
"epoch": 13.0,
"eval_bleu": 18.3822,
"eval_gen_len": 16.7928,
"eval_runtime": 62.3286,
"eval_samples_per_second": 14.247,
"eval_steps_per_second": 1.781,
"step": 167466
},
{
"epoch": 13.0,
"learning_rate": 0.0002,
"loss": 1.3028,
"step": 167500
},
{
"epoch": 13.04,
"learning_rate": 0.0002,
"loss": 1.2535,
"step": 168000
},
{
"epoch": 13.08,
"learning_rate": 0.0002,
"loss": 1.271,
"step": 168500
},
{
"epoch": 13.12,
"learning_rate": 0.0002,
"loss": 1.2735,
"step": 169000
},
{
"epoch": 13.16,
"learning_rate": 0.0002,
"loss": 1.2642,
"step": 169500
},
{
"epoch": 13.2,
"learning_rate": 0.0002,
"loss": 1.2611,
"step": 170000
},
{
"epoch": 13.24,
"learning_rate": 0.0002,
"loss": 1.2789,
"step": 170500
},
{
"epoch": 13.27,
"learning_rate": 0.0002,
"loss": 1.2666,
"step": 171000
},
{
"epoch": 13.31,
"learning_rate": 0.0002,
"loss": 1.2769,
"step": 171500
},
{
"epoch": 13.35,
"learning_rate": 0.0002,
"loss": 1.2782,
"step": 172000
},
{
"epoch": 13.39,
"learning_rate": 0.0002,
"loss": 1.2656,
"step": 172500
},
{
"epoch": 13.43,
"learning_rate": 0.0002,
"loss": 1.2746,
"step": 173000
},
{
"epoch": 13.47,
"learning_rate": 0.0002,
"loss": 1.2687,
"step": 173500
},
{
"epoch": 13.51,
"learning_rate": 0.0002,
"loss": 1.2806,
"step": 174000
},
{
"epoch": 13.55,
"learning_rate": 0.0002,
"loss": 1.2882,
"step": 174500
},
{
"epoch": 13.58,
"learning_rate": 0.0002,
"loss": 1.2799,
"step": 175000
},
{
"epoch": 13.62,
"learning_rate": 0.0002,
"loss": 1.2866,
"step": 175500
},
{
"epoch": 13.66,
"learning_rate": 0.0002,
"loss": 1.293,
"step": 176000
},
{
"epoch": 13.7,
"learning_rate": 0.0002,
"loss": 1.2849,
"step": 176500
},
{
"epoch": 13.74,
"learning_rate": 0.0002,
"loss": 1.2921,
"step": 177000
},
{
"epoch": 13.78,
"learning_rate": 0.0002,
"loss": 1.2758,
"step": 177500
},
{
"epoch": 13.82,
"learning_rate": 0.0002,
"loss": 1.28,
"step": 178000
},
{
"epoch": 13.86,
"learning_rate": 0.0002,
"loss": 1.2912,
"step": 178500
},
{
"epoch": 13.9,
"learning_rate": 0.0002,
"loss": 1.2859,
"step": 179000
},
{
"epoch": 13.93,
"learning_rate": 0.0002,
"loss": 1.284,
"step": 179500
},
{
"epoch": 13.97,
"learning_rate": 0.0002,
"loss": 1.2817,
"step": 180000
},
{
"epoch": 14.0,
"eval_bleu": 18.3689,
"eval_gen_len": 16.8018,
"eval_runtime": 62.0748,
"eval_samples_per_second": 14.305,
"eval_steps_per_second": 1.788,
"step": 180348
},
{
"epoch": 14.01,
"learning_rate": 0.0002,
"loss": 1.278,
"step": 180500
},
{
"epoch": 14.05,
"learning_rate": 0.0002,
"loss": 1.2338,
"step": 181000
},
{
"epoch": 14.09,
"learning_rate": 0.0002,
"loss": 1.2302,
"step": 181500
},
{
"epoch": 14.13,
"learning_rate": 0.0002,
"loss": 1.2266,
"step": 182000
},
{
"epoch": 14.17,
"learning_rate": 0.0002,
"loss": 1.2269,
"step": 182500
},
{
"epoch": 14.21,
"learning_rate": 0.0002,
"loss": 1.2496,
"step": 183000
},
{
"epoch": 14.24,
"learning_rate": 0.0002,
"loss": 1.2438,
"step": 183500
},
{
"epoch": 14.28,
"learning_rate": 0.0002,
"loss": 1.2409,
"step": 184000
},
{
"epoch": 14.32,
"learning_rate": 0.0002,
"loss": 1.2517,
"step": 184500
},
{
"epoch": 14.36,
"learning_rate": 0.0002,
"loss": 1.2554,
"step": 185000
},
{
"epoch": 14.4,
"learning_rate": 0.0002,
"loss": 1.2521,
"step": 185500
},
{
"epoch": 14.44,
"learning_rate": 0.0002,
"loss": 1.2543,
"step": 186000
},
{
"epoch": 14.48,
"learning_rate": 0.0002,
"loss": 1.2601,
"step": 186500
},
{
"epoch": 14.52,
"learning_rate": 0.0002,
"loss": 1.2639,
"step": 187000
},
{
"epoch": 14.56,
"learning_rate": 0.0002,
"loss": 1.2674,
"step": 187500
},
{
"epoch": 14.59,
"learning_rate": 0.0002,
"loss": 1.2606,
"step": 188000
},
{
"epoch": 14.63,
"learning_rate": 0.0002,
"loss": 1.2472,
"step": 188500
},
{
"epoch": 14.67,
"learning_rate": 0.0002,
"loss": 1.2596,
"step": 189000
},
{
"epoch": 14.71,
"learning_rate": 0.0002,
"loss": 1.2596,
"step": 189500
},
{
"epoch": 14.75,
"learning_rate": 0.0002,
"loss": 1.2634,
"step": 190000
},
{
"epoch": 14.79,
"learning_rate": 0.0002,
"loss": 1.2674,
"step": 190500
},
{
"epoch": 14.83,
"learning_rate": 0.0002,
"loss": 1.2622,
"step": 191000
},
{
"epoch": 14.87,
"learning_rate": 0.0002,
"loss": 1.2739,
"step": 191500
},
{
"epoch": 14.9,
"learning_rate": 0.0002,
"loss": 1.2599,
"step": 192000
},
{
"epoch": 14.94,
"learning_rate": 0.0002,
"loss": 1.2661,
"step": 192500
},
{
"epoch": 14.98,
"learning_rate": 0.0002,
"loss": 1.265,
"step": 193000
},
{
"epoch": 15.0,
"eval_bleu": 0,
"eval_gen_len": 16.8029,
"eval_runtime": 60.4599,
"eval_samples_per_second": 14.687,
"eval_steps_per_second": 1.836,
"step": 193230
},
{
"epoch": 15.02,
"learning_rate": 0.0002,
"loss": 1.2297,
"step": 193500
},
{
"epoch": 15.06,
"learning_rate": 0.0002,
"loss": 1.2043,
"step": 194000
},
{
"epoch": 15.1,
"learning_rate": 0.0002,
"loss": 1.2166,
"step": 194500
},
{
"epoch": 15.14,
"learning_rate": 0.0002,
"loss": 1.2201,
"step": 195000
},
{
"epoch": 15.18,
"learning_rate": 0.0002,
"loss": 1.2259,
"step": 195500
},
{
"epoch": 15.22,
"learning_rate": 0.0002,
"loss": 1.2124,
"step": 196000
},
{
"epoch": 15.25,
"learning_rate": 0.0002,
"loss": 1.222,
"step": 196500
},
{
"epoch": 15.29,
"learning_rate": 0.0002,
"loss": 1.2204,
"step": 197000
},
{
"epoch": 15.33,
"learning_rate": 0.0002,
"loss": 1.2207,
"step": 197500
},
{
"epoch": 15.37,
"learning_rate": 0.0002,
"loss": 1.2299,
"step": 198000
},
{
"epoch": 15.41,
"learning_rate": 0.0002,
"loss": 1.2288,
"step": 198500
},
{
"epoch": 15.45,
"learning_rate": 0.0002,
"loss": 1.2288,
"step": 199000
},
{
"epoch": 15.49,
"learning_rate": 0.0002,
"loss": 1.2322,
"step": 199500
},
{
"epoch": 15.53,
"learning_rate": 0.0002,
"loss": 1.2378,
"step": 200000
},
{
"epoch": 15.56,
"learning_rate": 0.0002,
"loss": 1.2242,
"step": 200500
},
{
"epoch": 15.6,
"learning_rate": 0.0002,
"loss": 1.2308,
"step": 201000
},
{
"epoch": 15.64,
"learning_rate": 0.0002,
"loss": 1.2397,
"step": 201500
},
{
"epoch": 15.68,
"learning_rate": 0.0002,
"loss": 1.2477,
"step": 202000
},
{
"epoch": 15.72,
"learning_rate": 0.0002,
"loss": 1.2439,
"step": 202500
},
{
"epoch": 15.76,
"learning_rate": 0.0002,
"loss": 1.2335,
"step": 203000
},
{
"epoch": 15.8,
"learning_rate": 0.0002,
"loss": 1.2402,
"step": 203500
},
{
"epoch": 15.84,
"learning_rate": 0.0002,
"loss": 1.2431,
"step": 204000
},
{
"epoch": 15.87,
"learning_rate": 0.0002,
"loss": 1.2385,
"step": 204500
},
{
"epoch": 15.91,
"learning_rate": 0.0002,
"loss": 1.2461,
"step": 205000
},
{
"epoch": 15.95,
"learning_rate": 0.0002,
"loss": 1.243,
"step": 205500
},
{
"epoch": 15.99,
"learning_rate": 0.0002,
"loss": 1.2405,
"step": 206000
},
{
"epoch": 16.0,
"eval_bleu": 18.2558,
"eval_gen_len": 16.8423,
"eval_runtime": 62.159,
"eval_samples_per_second": 14.286,
"eval_steps_per_second": 1.786,
"step": 206112
},
{
"epoch": 16.03,
"learning_rate": 0.0002,
"loss": 1.1951,
"step": 206500
},
{
"epoch": 16.07,
"learning_rate": 0.0002,
"loss": 1.1899,
"step": 207000
},
{
"epoch": 16.11,
"learning_rate": 0.0002,
"loss": 1.2004,
"step": 207500
},
{
"epoch": 16.15,
"learning_rate": 0.0002,
"loss": 1.1946,
"step": 208000
},
{
"epoch": 16.19,
"learning_rate": 0.0002,
"loss": 1.1996,
"step": 208500
},
{
"epoch": 16.22,
"learning_rate": 0.0002,
"loss": 1.1973,
"step": 209000
},
{
"epoch": 16.26,
"learning_rate": 0.0002,
"loss": 1.1984,
"step": 209500
},
{
"epoch": 16.3,
"learning_rate": 0.0002,
"loss": 1.2094,
"step": 210000
},
{
"epoch": 16.34,
"learning_rate": 0.0002,
"loss": 1.2065,
"step": 210500
},
{
"epoch": 16.38,
"learning_rate": 0.0002,
"loss": 1.191,
"step": 211000
},
{
"epoch": 16.42,
"learning_rate": 0.0002,
"loss": 1.2125,
"step": 211500
},
{
"epoch": 16.46,
"learning_rate": 0.0002,
"loss": 1.2105,
"step": 212000
},
{
"epoch": 16.5,
"learning_rate": 0.0002,
"loss": 1.2144,
"step": 212500
},
{
"epoch": 16.53,
"learning_rate": 0.0002,
"loss": 1.2198,
"step": 213000
},
{
"epoch": 16.57,
"learning_rate": 0.0002,
"loss": 1.2173,
"step": 213500
},
{
"epoch": 16.61,
"learning_rate": 0.0002,
"loss": 1.2166,
"step": 214000
},
{
"epoch": 16.65,
"learning_rate": 0.0002,
"loss": 1.2164,
"step": 214500
},
{
"epoch": 16.69,
"learning_rate": 0.0002,
"loss": 1.2227,
"step": 215000
},
{
"epoch": 16.73,
"learning_rate": 0.0002,
"loss": 1.22,
"step": 215500
},
{
"epoch": 16.77,
"learning_rate": 0.0002,
"loss": 1.2154,
"step": 216000
},
{
"epoch": 16.81,
"learning_rate": 0.0002,
"loss": 1.212,
"step": 216500
},
{
"epoch": 16.85,
"learning_rate": 0.0002,
"loss": 1.2164,
"step": 217000
},
{
"epoch": 16.88,
"learning_rate": 0.0002,
"loss": 1.2179,
"step": 217500
},
{
"epoch": 16.92,
"learning_rate": 0.0002,
"loss": 1.216,
"step": 218000
},
{
"epoch": 16.96,
"learning_rate": 0.0002,
"loss": 1.2367,
"step": 218500
},
{
"epoch": 17.0,
"eval_bleu": 0,
"eval_gen_len": 16.7872,
"eval_runtime": 61.3805,
"eval_samples_per_second": 14.467,
"eval_steps_per_second": 1.808,
"step": 218994
},
{
"epoch": 17.0,
"learning_rate": 0.0002,
"loss": 1.2172,
"step": 219000
},
{
"epoch": 17.04,
"learning_rate": 0.0002,
"loss": 1.1676,
"step": 219500
},
{
"epoch": 17.08,
"learning_rate": 0.0002,
"loss": 1.1779,
"step": 220000
},
{
"epoch": 17.12,
"learning_rate": 0.0002,
"loss": 1.1745,
"step": 220500
},
{
"epoch": 17.16,
"learning_rate": 0.0002,
"loss": 1.1798,
"step": 221000
},
{
"epoch": 17.19,
"learning_rate": 0.0002,
"loss": 1.1735,
"step": 221500
},
{
"epoch": 17.23,
"learning_rate": 0.0002,
"loss": 1.1816,
"step": 222000
},
{
"epoch": 17.27,
"learning_rate": 0.0002,
"loss": 1.1862,
"step": 222500
},
{
"epoch": 17.31,
"learning_rate": 0.0002,
"loss": 1.1802,
"step": 223000
},
{
"epoch": 17.35,
"learning_rate": 0.0002,
"loss": 1.1778,
"step": 223500
},
{
"epoch": 17.39,
"learning_rate": 0.0002,
"loss": 1.1816,
"step": 224000
},
{
"epoch": 17.43,
"learning_rate": 0.0002,
"loss": 1.189,
"step": 224500
},
{
"epoch": 17.47,
"learning_rate": 0.0002,
"loss": 1.1839,
"step": 225000
},
{
"epoch": 17.51,
"learning_rate": 0.0002,
"loss": 1.1898,
"step": 225500
},
{
"epoch": 17.54,
"learning_rate": 0.0002,
"loss": 1.1857,
"step": 226000
},
{
"epoch": 17.58,
"learning_rate": 0.0002,
"loss": 1.1939,
"step": 226500
},
{
"epoch": 17.62,
"learning_rate": 0.0002,
"loss": 1.2107,
"step": 227000
},
{
"epoch": 17.66,
"learning_rate": 0.0002,
"loss": 1.206,
"step": 227500
},
{
"epoch": 17.7,
"learning_rate": 0.0002,
"loss": 1.1943,
"step": 228000
},
{
"epoch": 17.74,
"learning_rate": 0.0002,
"loss": 1.2013,
"step": 228500
},
{
"epoch": 17.78,
"learning_rate": 0.0002,
"loss": 1.2029,
"step": 229000
},
{
"epoch": 17.82,
"learning_rate": 0.0002,
"loss": 1.1914,
"step": 229500
},
{
"epoch": 17.85,
"learning_rate": 0.0002,
"loss": 1.1989,
"step": 230000
},
{
"epoch": 17.89,
"learning_rate": 0.0002,
"loss": 1.1965,
"step": 230500
},
{
"epoch": 17.93,
"learning_rate": 0.0002,
"loss": 1.1966,
"step": 231000
},
{
"epoch": 17.97,
"learning_rate": 0.0002,
"loss": 1.2025,
"step": 231500
},
{
"epoch": 18.0,
"eval_bleu": 18.0334,
"eval_gen_len": 16.8311,
"eval_runtime": 62.1395,
"eval_samples_per_second": 14.29,
"eval_steps_per_second": 1.786,
"step": 231876
},
{
"epoch": 18.01,
"learning_rate": 0.0002,
"loss": 1.1892,
"step": 232000
},
{
"epoch": 18.05,
"learning_rate": 0.0002,
"loss": 1.1473,
"step": 232500
},
{
"epoch": 18.09,
"learning_rate": 0.0002,
"loss": 1.1555,
"step": 233000
},
{
"epoch": 18.13,
"learning_rate": 0.0002,
"loss": 1.1495,
"step": 233500
},
{
"epoch": 18.16,
"learning_rate": 0.0002,
"loss": 1.1551,
"step": 234000
},
{
"epoch": 18.2,
"learning_rate": 0.0002,
"loss": 1.1558,
"step": 234500
},
{
"epoch": 18.24,
"learning_rate": 0.0002,
"loss": 1.1567,
"step": 235000
},
{
"epoch": 18.28,
"learning_rate": 0.0002,
"loss": 1.1623,
"step": 235500
},
{
"epoch": 18.32,
"learning_rate": 0.0002,
"loss": 1.1691,
"step": 236000
},
{
"epoch": 18.36,
"learning_rate": 0.0002,
"loss": 1.1721,
"step": 236500
},
{
"epoch": 18.4,
"learning_rate": 0.0002,
"loss": 1.1776,
"step": 237000
},
{
"epoch": 18.44,
"learning_rate": 0.0002,
"loss": 1.1698,
"step": 237500
},
{
"epoch": 18.48,
"learning_rate": 0.0002,
"loss": 1.1722,
"step": 238000
},
{
"epoch": 18.51,
"learning_rate": 0.0002,
"loss": 1.1708,
"step": 238500
},
{
"epoch": 18.55,
"learning_rate": 0.0002,
"loss": 1.1722,
"step": 239000
},
{
"epoch": 18.59,
"learning_rate": 0.0002,
"loss": 1.1734,
"step": 239500
},
{
"epoch": 18.63,
"learning_rate": 0.0002,
"loss": 1.1813,
"step": 240000
},
{
"epoch": 18.67,
"learning_rate": 0.0002,
"loss": 1.1758,
"step": 240500
},
{
"epoch": 18.71,
"learning_rate": 0.0002,
"loss": 1.1698,
"step": 241000
},
{
"epoch": 18.75,
"learning_rate": 0.0002,
"loss": 1.175,
"step": 241500
},
{
"epoch": 18.79,
"learning_rate": 0.0002,
"loss": 1.1804,
"step": 242000
},
{
"epoch": 18.82,
"learning_rate": 0.0002,
"loss": 1.1825,
"step": 242500
},
{
"epoch": 18.86,
"learning_rate": 0.0002,
"loss": 1.1916,
"step": 243000
},
{
"epoch": 18.9,
"learning_rate": 0.0002,
"loss": 1.1887,
"step": 243500
},
{
"epoch": 18.94,
"learning_rate": 0.0002,
"loss": 1.1731,
"step": 244000
},
{
"epoch": 18.98,
"learning_rate": 0.0002,
"loss": 1.1901,
"step": 244500
},
{
"epoch": 19.0,
"eval_bleu": 18.1719,
"eval_gen_len": 16.8153,
"eval_runtime": 62.3921,
"eval_samples_per_second": 14.233,
"eval_steps_per_second": 1.779,
"step": 244758
},
{
"epoch": 19.02,
"learning_rate": 0.0002,
"loss": 1.1566,
"step": 245000
},
{
"epoch": 19.06,
"learning_rate": 0.0002,
"loss": 1.1361,
"step": 245500
},
{
"epoch": 19.1,
"learning_rate": 0.0002,
"loss": 1.1414,
"step": 246000
},
{
"epoch": 19.14,
"learning_rate": 0.0002,
"loss": 1.1346,
"step": 246500
},
{
"epoch": 19.17,
"learning_rate": 0.0002,
"loss": 1.1397,
"step": 247000
},
{
"epoch": 19.21,
"learning_rate": 0.0002,
"loss": 1.1353,
"step": 247500
},
{
"epoch": 19.25,
"learning_rate": 0.0002,
"loss": 1.146,
"step": 248000
},
{
"epoch": 19.29,
"learning_rate": 0.0002,
"loss": 1.1457,
"step": 248500
},
{
"epoch": 19.33,
"learning_rate": 0.0002,
"loss": 1.1588,
"step": 249000
},
{
"epoch": 19.37,
"learning_rate": 0.0002,
"loss": 1.1394,
"step": 249500
},
{
"epoch": 19.41,
"learning_rate": 0.0002,
"loss": 1.1514,
"step": 250000
},
{
"epoch": 19.45,
"learning_rate": 0.0002,
"loss": 1.164,
"step": 250500
},
{
"epoch": 19.48,
"learning_rate": 0.0002,
"loss": 1.1569,
"step": 251000
},
{
"epoch": 19.52,
"learning_rate": 0.0002,
"loss": 1.1481,
"step": 251500
},
{
"epoch": 19.56,
"learning_rate": 0.0002,
"loss": 1.1536,
"step": 252000
},
{
"epoch": 19.6,
"learning_rate": 0.0002,
"loss": 1.1636,
"step": 252500
},
{
"epoch": 19.64,
"learning_rate": 0.0002,
"loss": 1.1549,
"step": 253000
},
{
"epoch": 19.68,
"learning_rate": 0.0002,
"loss": 1.1441,
"step": 253500
},
{
"epoch": 19.72,
"learning_rate": 0.0002,
"loss": 1.164,
"step": 254000
},
{
"epoch": 19.76,
"learning_rate": 0.0002,
"loss": 1.1622,
"step": 254500
},
{
"epoch": 19.8,
"learning_rate": 0.0002,
"loss": 1.1613,
"step": 255000
},
{
"epoch": 19.83,
"learning_rate": 0.0002,
"loss": 1.1607,
"step": 255500
},
{
"epoch": 19.87,
"learning_rate": 0.0002,
"loss": 1.1599,
"step": 256000
},
{
"epoch": 19.91,
"learning_rate": 0.0002,
"loss": 1.1767,
"step": 256500
},
{
"epoch": 19.95,
"learning_rate": 0.0002,
"loss": 1.1693,
"step": 257000
},
{
"epoch": 19.99,
"learning_rate": 0.0002,
"loss": 1.1617,
"step": 257500
},
{
"epoch": 20.0,
"eval_bleu": 18.0422,
"eval_gen_len": 16.8559,
"eval_runtime": 62.333,
"eval_samples_per_second": 14.246,
"eval_steps_per_second": 1.781,
"step": 257640
},
{
"epoch": 20.03,
"learning_rate": 0.0002,
"loss": 1.1212,
"step": 258000
},
{
"epoch": 20.07,
"learning_rate": 0.0002,
"loss": 1.1118,
"step": 258500
},
{
"epoch": 20.11,
"learning_rate": 0.0002,
"loss": 1.1137,
"step": 259000
},
{
"epoch": 20.14,
"learning_rate": 0.0002,
"loss": 1.1204,
"step": 259500
},
{
"epoch": 20.18,
"learning_rate": 0.0002,
"loss": 1.1269,
"step": 260000
},
{
"epoch": 20.22,
"learning_rate": 0.0002,
"loss": 1.1128,
"step": 260500
},
{
"epoch": 20.26,
"learning_rate": 0.0002,
"loss": 1.1218,
"step": 261000
},
{
"epoch": 20.3,
"learning_rate": 0.0002,
"loss": 1.1237,
"step": 261500
},
{
"epoch": 20.34,
"learning_rate": 0.0002,
"loss": 1.132,
"step": 262000
},
{
"epoch": 20.38,
"learning_rate": 0.0002,
"loss": 1.1367,
"step": 262500
},
{
"epoch": 20.42,
"learning_rate": 0.0002,
"loss": 1.1357,
"step": 263000
},
{
"epoch": 20.45,
"learning_rate": 0.0002,
"loss": 1.1436,
"step": 263500
},
{
"epoch": 20.49,
"learning_rate": 0.0002,
"loss": 1.1336,
"step": 264000
},
{
"epoch": 20.53,
"learning_rate": 0.0002,
"loss": 1.1461,
"step": 264500
},
{
"epoch": 20.57,
"learning_rate": 0.0002,
"loss": 1.1381,
"step": 265000
},
{
"epoch": 20.61,
"learning_rate": 0.0002,
"loss": 1.1387,
"step": 265500
},
{
"epoch": 20.65,
"learning_rate": 0.0002,
"loss": 1.1459,
"step": 266000
},
{
"epoch": 20.69,
"learning_rate": 0.0002,
"loss": 1.136,
"step": 266500
},
{
"epoch": 20.73,
"learning_rate": 0.0002,
"loss": 1.1479,
"step": 267000
},
{
"epoch": 20.77,
"learning_rate": 0.0002,
"loss": 1.1355,
"step": 267500
},
{
"epoch": 20.8,
"learning_rate": 0.0002,
"loss": 1.1478,
"step": 268000
},
{
"epoch": 20.84,
"learning_rate": 0.0002,
"loss": 1.1354,
"step": 268500
},
{
"epoch": 20.88,
"learning_rate": 0.0002,
"loss": 1.1574,
"step": 269000
},
{
"epoch": 20.92,
"learning_rate": 0.0002,
"loss": 1.1544,
"step": 269500
},
{
"epoch": 20.96,
"learning_rate": 0.0002,
"loss": 1.1548,
"step": 270000
},
{
"epoch": 21.0,
"learning_rate": 0.0002,
"loss": 1.1464,
"step": 270500
},
{
"epoch": 21.0,
"eval_bleu": 18.1118,
"eval_gen_len": 16.8176,
"eval_runtime": 61.9099,
"eval_samples_per_second": 14.343,
"eval_steps_per_second": 1.793,
"step": 270522
},
{
"epoch": 21.04,
"learning_rate": 0.0002,
"loss": 1.0956,
"step": 271000
},
{
"epoch": 21.08,
"learning_rate": 0.0002,
"loss": 1.1023,
"step": 271500
},
{
"epoch": 21.11,
"learning_rate": 0.0002,
"loss": 1.0998,
"step": 272000
},
{
"epoch": 21.15,
"learning_rate": 0.0002,
"loss": 1.1041,
"step": 272500
},
{
"epoch": 21.19,
"learning_rate": 0.0002,
"loss": 1.1032,
"step": 273000
},
{
"epoch": 21.23,
"learning_rate": 0.0002,
"loss": 1.1086,
"step": 273500
},
{
"epoch": 21.27,
"learning_rate": 0.0002,
"loss": 1.1115,
"step": 274000
},
{
"epoch": 21.31,
"learning_rate": 0.0002,
"loss": 1.1086,
"step": 274500
},
{
"epoch": 21.35,
"learning_rate": 0.0002,
"loss": 1.1204,
"step": 275000
},
{
"epoch": 21.39,
"learning_rate": 0.0002,
"loss": 1.1094,
"step": 275500
},
{
"epoch": 21.43,
"learning_rate": 0.0002,
"loss": 1.1247,
"step": 276000
},
{
"epoch": 21.46,
"learning_rate": 0.0002,
"loss": 1.119,
"step": 276500
},
{
"epoch": 21.5,
"learning_rate": 0.0002,
"loss": 1.1172,
"step": 277000
},
{
"epoch": 21.54,
"learning_rate": 0.0002,
"loss": 1.1108,
"step": 277500
},
{
"epoch": 21.58,
"learning_rate": 0.0002,
"loss": 1.1237,
"step": 278000
},
{
"epoch": 21.62,
"learning_rate": 0.0002,
"loss": 1.1337,
"step": 278500
},
{
"epoch": 21.66,
"learning_rate": 0.0002,
"loss": 1.1252,
"step": 279000
},
{
"epoch": 21.7,
"learning_rate": 0.0002,
"loss": 1.134,
"step": 279500
},
{
"epoch": 21.74,
"learning_rate": 0.0002,
"loss": 1.1242,
"step": 280000
},
{
"epoch": 21.77,
"learning_rate": 0.0002,
"loss": 1.1321,
"step": 280500
},
{
"epoch": 21.81,
"learning_rate": 0.0002,
"loss": 1.1301,
"step": 281000
},
{
"epoch": 21.85,
"learning_rate": 0.0002,
"loss": 1.1318,
"step": 281500
},
{
"epoch": 21.89,
"learning_rate": 0.0002,
"loss": 1.1373,
"step": 282000
},
{
"epoch": 21.93,
"learning_rate": 0.0002,
"loss": 1.1256,
"step": 282500
},
{
"epoch": 21.97,
"learning_rate": 0.0002,
"loss": 1.1267,
"step": 283000
},
{
"epoch": 22.0,
"eval_bleu": 18.3787,
"eval_gen_len": 16.8986,
"eval_runtime": 62.1828,
"eval_samples_per_second": 14.28,
"eval_steps_per_second": 1.785,
"step": 283404
},
{
"epoch": 22.01,
"learning_rate": 0.0002,
"loss": 1.1261,
"step": 283500
},
{
"epoch": 22.05,
"learning_rate": 0.0002,
"loss": 1.0749,
"step": 284000
},
{
"epoch": 22.09,
"learning_rate": 0.0002,
"loss": 1.0848,
"step": 284500
},
{
"epoch": 22.12,
"learning_rate": 0.0002,
"loss": 1.0883,
"step": 285000
},
{
"epoch": 22.16,
"learning_rate": 0.0002,
"loss": 1.0947,
"step": 285500
},
{
"epoch": 22.2,
"learning_rate": 0.0002,
"loss": 1.0903,
"step": 286000
},
{
"epoch": 22.24,
"learning_rate": 0.0002,
"loss": 1.0951,
"step": 286500
},
{
"epoch": 22.28,
"learning_rate": 0.0002,
"loss": 1.0957,
"step": 287000
},
{
"epoch": 22.32,
"learning_rate": 0.0002,
"loss": 1.0923,
"step": 287500
},
{
"epoch": 22.36,
"learning_rate": 0.0002,
"loss": 1.0955,
"step": 288000
},
{
"epoch": 22.4,
"learning_rate": 0.0002,
"loss": 1.1002,
"step": 288500
},
{
"epoch": 22.43,
"learning_rate": 0.0002,
"loss": 1.1012,
"step": 289000
},
{
"epoch": 22.47,
"learning_rate": 0.0002,
"loss": 1.098,
"step": 289500
},
{
"epoch": 22.51,
"learning_rate": 0.0002,
"loss": 1.1014,
"step": 290000
},
{
"epoch": 22.55,
"learning_rate": 0.0002,
"loss": 1.1282,
"step": 290500
},
{
"epoch": 22.59,
"learning_rate": 0.0002,
"loss": 1.0999,
"step": 291000
},
{
"epoch": 22.63,
"learning_rate": 0.0002,
"loss": 1.116,
"step": 291500
},
{
"epoch": 22.67,
"learning_rate": 0.0002,
"loss": 1.1104,
"step": 292000
},
{
"epoch": 22.71,
"learning_rate": 0.0002,
"loss": 1.1043,
"step": 292500
},
{
"epoch": 22.74,
"learning_rate": 0.0002,
"loss": 1.1052,
"step": 293000
},
{
"epoch": 22.78,
"learning_rate": 0.0002,
"loss": 1.1166,
"step": 293500
},
{
"epoch": 22.82,
"learning_rate": 0.0002,
"loss": 1.1114,
"step": 294000
},
{
"epoch": 22.86,
"learning_rate": 0.0002,
"loss": 1.118,
"step": 294500
},
{
"epoch": 22.9,
"learning_rate": 0.0002,
"loss": 1.1191,
"step": 295000
},
{
"epoch": 22.94,
"learning_rate": 0.0002,
"loss": 1.1195,
"step": 295500
},
{
"epoch": 22.98,
"learning_rate": 0.0002,
"loss": 1.1131,
"step": 296000
},
{
"epoch": 23.0,
"eval_bleu": 18.2406,
"eval_gen_len": 16.857,
"eval_runtime": 61.9634,
"eval_samples_per_second": 14.331,
"eval_steps_per_second": 1.791,
"step": 296286
},
{
"epoch": 23.02,
"learning_rate": 0.0002,
"loss": 1.1018,
"step": 296500
},
{
"epoch": 23.06,
"learning_rate": 0.0002,
"loss": 1.0608,
"step": 297000
},
{
"epoch": 23.09,
"learning_rate": 0.0002,
"loss": 1.0591,
"step": 297500
},
{
"epoch": 23.13,
"learning_rate": 0.0002,
"loss": 1.0785,
"step": 298000
},
{
"epoch": 23.17,
"learning_rate": 0.0002,
"loss": 1.0734,
"step": 298500
},
{
"epoch": 23.21,
"learning_rate": 0.0002,
"loss": 1.0816,
"step": 299000
},
{
"epoch": 23.25,
"learning_rate": 0.0002,
"loss": 1.0689,
"step": 299500
},
{
"epoch": 23.29,
"learning_rate": 0.0002,
"loss": 1.0772,
"step": 300000
},
{
"epoch": 23.33,
"learning_rate": 0.0002,
"loss": 1.0857,
"step": 300500
},
{
"epoch": 23.37,
"learning_rate": 0.0002,
"loss": 1.0898,
"step": 301000
},
{
"epoch": 23.4,
"learning_rate": 0.0002,
"loss": 1.0764,
"step": 301500
},
{
"epoch": 23.44,
"learning_rate": 0.0002,
"loss": 1.0872,
"step": 302000
},
{
"epoch": 23.48,
"learning_rate": 0.0002,
"loss": 1.0924,
"step": 302500
},
{
"epoch": 23.52,
"learning_rate": 0.0002,
"loss": 1.0973,
"step": 303000
},
{
"epoch": 23.56,
"learning_rate": 0.0002,
"loss": 1.0877,
"step": 303500
},
{
"epoch": 23.6,
"learning_rate": 0.0002,
"loss": 1.1021,
"step": 304000
},
{
"epoch": 23.64,
"learning_rate": 0.0002,
"loss": 1.0954,
"step": 304500
},
{
"epoch": 23.68,
"learning_rate": 0.0002,
"loss": 1.0905,
"step": 305000
},
{
"epoch": 23.72,
"learning_rate": 0.0002,
"loss": 1.1017,
"step": 305500
},
{
"epoch": 23.75,
"learning_rate": 0.0002,
"loss": 1.1037,
"step": 306000
},
{
"epoch": 23.79,
"learning_rate": 0.0002,
"loss": 1.098,
"step": 306500
},
{
"epoch": 23.83,
"learning_rate": 0.0002,
"loss": 1.0995,
"step": 307000
},
{
"epoch": 23.87,
"learning_rate": 0.0002,
"loss": 1.0958,
"step": 307500
},
{
"epoch": 23.91,
"learning_rate": 0.0002,
"loss": 1.1006,
"step": 308000
},
{
"epoch": 23.95,
"learning_rate": 0.0002,
"loss": 1.1013,
"step": 308500
},
{
"epoch": 23.99,
"learning_rate": 0.0002,
"loss": 1.1045,
"step": 309000
},
{
"epoch": 24.0,
"eval_bleu": 18.4266,
"eval_gen_len": 16.8491,
"eval_runtime": 61.8264,
"eval_samples_per_second": 14.363,
"eval_steps_per_second": 1.795,
"step": 309168
},
{
"epoch": 24.03,
"learning_rate": 0.0002,
"loss": 1.0672,
"step": 309500
},
{
"epoch": 24.06,
"learning_rate": 0.0002,
"loss": 1.0516,
"step": 310000
},
{
"epoch": 24.1,
"learning_rate": 0.0002,
"loss": 1.0479,
"step": 310500
},
{
"epoch": 24.14,
"learning_rate": 0.0002,
"loss": 1.0559,
"step": 311000
},
{
"epoch": 24.18,
"learning_rate": 0.0002,
"loss": 1.0536,
"step": 311500
},
{
"epoch": 24.22,
"learning_rate": 0.0002,
"loss": 1.0615,
"step": 312000
},
{
"epoch": 24.26,
"learning_rate": 0.0002,
"loss": 1.0728,
"step": 312500
},
{
"epoch": 24.3,
"learning_rate": 0.0002,
"loss": 1.0601,
"step": 313000
},
{
"epoch": 24.34,
"learning_rate": 0.0002,
"loss": 1.0711,
"step": 313500
},
{
"epoch": 24.38,
"learning_rate": 0.0002,
"loss": 1.0615,
"step": 314000
},
{
"epoch": 24.41,
"learning_rate": 0.0002,
"loss": 1.0769,
"step": 314500
},
{
"epoch": 24.45,
"learning_rate": 0.0002,
"loss": 1.0659,
"step": 315000
},
{
"epoch": 24.49,
"learning_rate": 0.0002,
"loss": 1.0734,
"step": 315500
},
{
"epoch": 24.53,
"learning_rate": 0.0002,
"loss": 1.0747,
"step": 316000
},
{
"epoch": 24.57,
"learning_rate": 0.0002,
"loss": 1.0727,
"step": 316500
},
{
"epoch": 24.61,
"learning_rate": 0.0002,
"loss": 1.0737,
"step": 317000
},
{
"epoch": 24.65,
"learning_rate": 0.0002,
"loss": 1.092,
"step": 317500
},
{
"epoch": 24.69,
"learning_rate": 0.0002,
"loss": 1.0764,
"step": 318000
},
{
"epoch": 24.72,
"learning_rate": 0.0002,
"loss": 1.0823,
"step": 318500
},
{
"epoch": 24.76,
"learning_rate": 0.0002,
"loss": 1.0864,
"step": 319000
},
{
"epoch": 24.8,
"learning_rate": 0.0002,
"loss": 1.082,
"step": 319500
},
{
"epoch": 24.84,
"learning_rate": 0.0002,
"loss": 1.0878,
"step": 320000
},
{
"epoch": 24.88,
"learning_rate": 0.0002,
"loss": 1.0978,
"step": 320500
},
{
"epoch": 24.92,
"learning_rate": 0.0002,
"loss": 1.0863,
"step": 321000
},
{
"epoch": 24.96,
"learning_rate": 0.0002,
"loss": 1.0946,
"step": 321500
},
{
"epoch": 25.0,
"learning_rate": 0.0002,
"loss": 1.0897,
"step": 322000
},
{
"epoch": 25.0,
"eval_bleu": 18.2379,
"eval_gen_len": 16.8468,
"eval_runtime": 62.0022,
"eval_samples_per_second": 14.322,
"eval_steps_per_second": 1.79,
"step": 322050
},
{
"epoch": 25.03,
"learning_rate": 0.0002,
"loss": 1.0381,
"step": 322500
},
{
"epoch": 25.07,
"learning_rate": 0.0002,
"loss": 1.0409,
"step": 323000
},
{
"epoch": 25.11,
"learning_rate": 0.0002,
"loss": 1.0445,
"step": 323500
},
{
"epoch": 25.15,
"learning_rate": 0.0002,
"loss": 1.0484,
"step": 324000
},
{
"epoch": 25.19,
"learning_rate": 0.0002,
"loss": 1.0537,
"step": 324500
},
{
"epoch": 25.23,
"learning_rate": 0.0002,
"loss": 1.0482,
"step": 325000
},
{
"epoch": 25.27,
"learning_rate": 0.0002,
"loss": 1.0477,
"step": 325500
},
{
"epoch": 25.31,
"learning_rate": 0.0002,
"loss": 1.048,
"step": 326000
},
{
"epoch": 25.35,
"learning_rate": 0.0002,
"loss": 1.0436,
"step": 326500
},
{
"epoch": 25.38,
"learning_rate": 0.0002,
"loss": 1.053,
"step": 327000
},
{
"epoch": 25.42,
"learning_rate": 0.0002,
"loss": 1.0537,
"step": 327500
},
{
"epoch": 25.46,
"learning_rate": 0.0002,
"loss": 1.0507,
"step": 328000
},
{
"epoch": 25.5,
"learning_rate": 0.0002,
"loss": 1.0679,
"step": 328500
},
{
"epoch": 25.54,
"learning_rate": 0.0002,
"loss": 1.0688,
"step": 329000
},
{
"epoch": 25.58,
"learning_rate": 0.0002,
"loss": 1.0682,
"step": 329500
},
{
"epoch": 25.62,
"learning_rate": 0.0002,
"loss": 1.069,
"step": 330000
},
{
"epoch": 25.66,
"learning_rate": 0.0002,
"loss": 1.0702,
"step": 330500
},
{
"epoch": 25.69,
"learning_rate": 0.0002,
"loss": 1.0752,
"step": 331000
},
{
"epoch": 25.73,
"learning_rate": 0.0002,
"loss": 1.0596,
"step": 331500
},
{
"epoch": 25.77,
"learning_rate": 0.0002,
"loss": 1.0716,
"step": 332000
},
{
"epoch": 25.81,
"learning_rate": 0.0002,
"loss": 1.0713,
"step": 332500
},
{
"epoch": 25.85,
"learning_rate": 0.0002,
"loss": 1.0678,
"step": 333000
},
{
"epoch": 25.89,
"learning_rate": 0.0002,
"loss": 1.0723,
"step": 333500
},
{
"epoch": 25.93,
"learning_rate": 0.0002,
"loss": 1.0662,
"step": 334000
},
{
"epoch": 25.97,
"learning_rate": 0.0002,
"loss": 1.078,
"step": 334500
},
{
"epoch": 26.0,
"eval_bleu": 18.2319,
"eval_gen_len": 16.8277,
"eval_runtime": 61.717,
"eval_samples_per_second": 14.388,
"eval_steps_per_second": 1.799,
"step": 334932
},
{
"epoch": 26.01,
"learning_rate": 0.0002,
"loss": 1.0632,
"step": 335000
},
{
"epoch": 26.04,
"learning_rate": 0.0002,
"loss": 1.0155,
"step": 335500
},
{
"epoch": 26.08,
"learning_rate": 0.0002,
"loss": 1.0247,
"step": 336000
},
{
"epoch": 26.12,
"learning_rate": 0.0002,
"loss": 1.0311,
"step": 336500
},
{
"epoch": 26.16,
"learning_rate": 0.0002,
"loss": 1.0303,
"step": 337000
},
{
"epoch": 26.2,
"learning_rate": 0.0002,
"loss": 1.0398,
"step": 337500
},
{
"epoch": 26.24,
"learning_rate": 0.0002,
"loss": 1.0298,
"step": 338000
},
{
"epoch": 26.28,
"learning_rate": 0.0002,
"loss": 1.0408,
"step": 338500
},
{
"epoch": 26.32,
"learning_rate": 0.0002,
"loss": 1.0297,
"step": 339000
},
{
"epoch": 26.35,
"learning_rate": 0.0002,
"loss": 1.0411,
"step": 339500
},
{
"epoch": 26.39,
"learning_rate": 0.0002,
"loss": 1.0498,
"step": 340000
},
{
"epoch": 26.43,
"learning_rate": 0.0002,
"loss": 1.0527,
"step": 340500
},
{
"epoch": 26.47,
"learning_rate": 0.0002,
"loss": 1.0478,
"step": 341000
},
{
"epoch": 26.51,
"learning_rate": 0.0002,
"loss": 1.0524,
"step": 341500
},
{
"epoch": 26.55,
"learning_rate": 0.0002,
"loss": 1.0468,
"step": 342000
},
{
"epoch": 26.59,
"learning_rate": 0.0002,
"loss": 1.0521,
"step": 342500
},
{
"epoch": 26.63,
"learning_rate": 0.0002,
"loss": 1.0399,
"step": 343000
},
{
"epoch": 26.67,
"learning_rate": 0.0002,
"loss": 1.0444,
"step": 343500
},
{
"epoch": 26.7,
"learning_rate": 0.0002,
"loss": 1.0637,
"step": 344000
},
{
"epoch": 26.74,
"learning_rate": 0.0002,
"loss": 1.0542,
"step": 344500
},
{
"epoch": 26.78,
"learning_rate": 0.0002,
"loss": 1.0607,
"step": 345000
},
{
"epoch": 26.82,
"learning_rate": 0.0002,
"loss": 1.0579,
"step": 345500
},
{
"epoch": 26.86,
"learning_rate": 0.0002,
"loss": 1.063,
"step": 346000
},
{
"epoch": 26.9,
"learning_rate": 0.0002,
"loss": 1.0706,
"step": 346500
},
{
"epoch": 26.94,
"learning_rate": 0.0002,
"loss": 1.0576,
"step": 347000
},
{
"epoch": 26.98,
"learning_rate": 0.0002,
"loss": 1.0585,
"step": 347500
},
{
"epoch": 27.0,
"eval_bleu": 18.1392,
"eval_gen_len": 16.8604,
"eval_runtime": 62.1547,
"eval_samples_per_second": 14.287,
"eval_steps_per_second": 1.786,
"step": 347814
},
{
"epoch": 27.01,
"learning_rate": 0.0002,
"loss": 1.0408,
"step": 348000
},
{
"epoch": 27.05,
"learning_rate": 0.0002,
"loss": 1.0087,
"step": 348500
},
{
"epoch": 27.09,
"learning_rate": 0.0002,
"loss": 1.0131,
"step": 349000
},
{
"epoch": 27.13,
"learning_rate": 0.0002,
"loss": 1.0316,
"step": 349500
},
{
"epoch": 27.17,
"learning_rate": 0.0002,
"loss": 1.0124,
"step": 350000
},
{
"epoch": 27.21,
"learning_rate": 0.0002,
"loss": 1.0223,
"step": 350500
},
{
"epoch": 27.25,
"learning_rate": 0.0002,
"loss": 1.0227,
"step": 351000
},
{
"epoch": 27.29,
"learning_rate": 0.0002,
"loss": 1.0205,
"step": 351500
},
{
"epoch": 27.32,
"learning_rate": 0.0002,
"loss": 1.0371,
"step": 352000
},
{
"epoch": 27.36,
"learning_rate": 0.0002,
"loss": 1.0149,
"step": 352500
},
{
"epoch": 27.4,
"learning_rate": 0.0002,
"loss": 1.0373,
"step": 353000
},
{
"epoch": 27.44,
"learning_rate": 0.0002,
"loss": 1.0349,
"step": 353500
},
{
"epoch": 27.48,
"learning_rate": 0.0002,
"loss": 1.0471,
"step": 354000
},
{
"epoch": 27.52,
"learning_rate": 0.0002,
"loss": 1.0342,
"step": 354500
},
{
"epoch": 27.56,
"learning_rate": 0.0002,
"loss": 1.0354,
"step": 355000
},
{
"epoch": 27.6,
"learning_rate": 0.0002,
"loss": 1.0393,
"step": 355500
},
{
"epoch": 27.64,
"learning_rate": 0.0002,
"loss": 1.0445,
"step": 356000
},
{
"epoch": 27.67,
"learning_rate": 0.0002,
"loss": 1.0316,
"step": 356500
},
{
"epoch": 27.71,
"learning_rate": 0.0002,
"loss": 1.0475,
"step": 357000
},
{
"epoch": 27.75,
"learning_rate": 0.0002,
"loss": 1.0394,
"step": 357500
},
{
"epoch": 27.79,
"learning_rate": 0.0002,
"loss": 1.0384,
"step": 358000
},
{
"epoch": 27.83,
"learning_rate": 0.0002,
"loss": 1.0474,
"step": 358500
},
{
"epoch": 27.87,
"learning_rate": 0.0002,
"loss": 1.0434,
"step": 359000
},
{
"epoch": 27.91,
"learning_rate": 0.0002,
"loss": 1.0488,
"step": 359500
},
{
"epoch": 27.95,
"learning_rate": 0.0002,
"loss": 1.0455,
"step": 360000
},
{
"epoch": 27.98,
"learning_rate": 0.0002,
"loss": 1.0458,
"step": 360500
},
{
"epoch": 28.0,
"eval_bleu": 18.2348,
"eval_gen_len": 16.8153,
"eval_runtime": 62.3003,
"eval_samples_per_second": 14.254,
"eval_steps_per_second": 1.782,
"step": 360696
},
{
"epoch": 28.02,
"learning_rate": 0.0002,
"loss": 1.0191,
"step": 361000
},
{
"epoch": 28.06,
"learning_rate": 0.0002,
"loss": 0.9983,
"step": 361500
},
{
"epoch": 28.1,
"learning_rate": 0.0002,
"loss": 1.0006,
"step": 362000
},
{
"epoch": 28.14,
"learning_rate": 0.0002,
"loss": 1.0131,
"step": 362500
},
{
"epoch": 28.18,
"learning_rate": 0.0002,
"loss": 1.0148,
"step": 363000
},
{
"epoch": 28.22,
"learning_rate": 0.0002,
"loss": 0.993,
"step": 363500
},
{
"epoch": 28.26,
"learning_rate": 0.0002,
"loss": 1.0203,
"step": 364000
},
{
"epoch": 28.3,
"learning_rate": 0.0002,
"loss": 1.006,
"step": 364500
},
{
"epoch": 28.33,
"learning_rate": 0.0002,
"loss": 1.0048,
"step": 365000
},
{
"epoch": 28.37,
"learning_rate": 0.0002,
"loss": 1.0179,
"step": 365500
},
{
"epoch": 28.41,
"learning_rate": 0.0002,
"loss": 1.0171,
"step": 366000
},
{
"epoch": 28.45,
"learning_rate": 0.0002,
"loss": 1.012,
"step": 366500
},
{
"epoch": 28.49,
"learning_rate": 0.0002,
"loss": 1.0296,
"step": 367000
},
{
"epoch": 28.53,
"learning_rate": 0.0002,
"loss": 1.0223,
"step": 367500
},
{
"epoch": 28.57,
"learning_rate": 0.0002,
"loss": 1.0274,
"step": 368000
},
{
"epoch": 28.61,
"learning_rate": 0.0002,
"loss": 1.0249,
"step": 368500
},
{
"epoch": 28.64,
"learning_rate": 0.0002,
"loss": 1.0281,
"step": 369000
},
{
"epoch": 28.68,
"learning_rate": 0.0002,
"loss": 1.0267,
"step": 369500
},
{
"epoch": 28.72,
"learning_rate": 0.0002,
"loss": 1.0269,
"step": 370000
},
{
"epoch": 28.76,
"learning_rate": 0.0002,
"loss": 1.036,
"step": 370500
},
{
"epoch": 28.8,
"learning_rate": 0.0002,
"loss": 1.0278,
"step": 371000
},
{
"epoch": 28.84,
"learning_rate": 0.0002,
"loss": 1.0293,
"step": 371500
},
{
"epoch": 28.88,
"learning_rate": 0.0002,
"loss": 1.0346,
"step": 372000
},
{
"epoch": 28.92,
"learning_rate": 0.0002,
"loss": 1.0296,
"step": 372500
},
{
"epoch": 28.96,
"learning_rate": 0.0002,
"loss": 1.0318,
"step": 373000
},
{
"epoch": 28.99,
"learning_rate": 0.0002,
"loss": 1.0441,
"step": 373500
},
{
"epoch": 29.0,
"eval_bleu": 18.1908,
"eval_gen_len": 16.8097,
"eval_runtime": 61.8572,
"eval_samples_per_second": 14.356,
"eval_steps_per_second": 1.794,
"step": 373578
},
{
"epoch": 29.03,
"learning_rate": 0.0002,
"loss": 0.9882,
"step": 374000
},
{
"epoch": 29.07,
"learning_rate": 0.0002,
"loss": 0.9962,
"step": 374500
},
{
"epoch": 29.11,
"learning_rate": 0.0002,
"loss": 0.9892,
"step": 375000
},
{
"epoch": 29.15,
"learning_rate": 0.0002,
"loss": 0.9931,
"step": 375500
},
{
"epoch": 29.19,
"learning_rate": 0.0002,
"loss": 1.0002,
"step": 376000
},
{
"epoch": 29.23,
"learning_rate": 0.0002,
"loss": 0.9885,
"step": 376500
},
{
"epoch": 29.27,
"learning_rate": 0.0002,
"loss": 1.0042,
"step": 377000
},
{
"epoch": 29.3,
"learning_rate": 0.0002,
"loss": 0.9985,
"step": 377500
},
{
"epoch": 29.34,
"learning_rate": 0.0002,
"loss": 1.0027,
"step": 378000
},
{
"epoch": 29.38,
"learning_rate": 0.0002,
"loss": 0.9939,
"step": 378500
},
{
"epoch": 29.42,
"learning_rate": 0.0002,
"loss": 1.0069,
"step": 379000
},
{
"epoch": 29.46,
"learning_rate": 0.0002,
"loss": 1.0078,
"step": 379500
},
{
"epoch": 29.5,
"learning_rate": 0.0002,
"loss": 0.9913,
"step": 380000
},
{
"epoch": 29.54,
"learning_rate": 0.0002,
"loss": 1.0059,
"step": 380500
},
{
"epoch": 29.58,
"learning_rate": 0.0002,
"loss": 1.0064,
"step": 381000
},
{
"epoch": 29.61,
"learning_rate": 0.0002,
"loss": 1.0171,
"step": 381500
},
{
"epoch": 29.65,
"learning_rate": 0.0002,
"loss": 1.0219,
"step": 382000
},
{
"epoch": 29.69,
"learning_rate": 0.0002,
"loss": 1.0082,
"step": 382500
},
{
"epoch": 29.73,
"learning_rate": 0.0002,
"loss": 1.0228,
"step": 383000
},
{
"epoch": 29.77,
"learning_rate": 0.0002,
"loss": 1.0243,
"step": 383500
},
{
"epoch": 29.81,
"learning_rate": 0.0002,
"loss": 1.0132,
"step": 384000
},
{
"epoch": 29.85,
"learning_rate": 0.0002,
"loss": 1.0348,
"step": 384500
},
{
"epoch": 29.89,
"learning_rate": 0.0002,
"loss": 1.0278,
"step": 385000
},
{
"epoch": 29.93,
"learning_rate": 0.0002,
"loss": 1.0322,
"step": 385500
},
{
"epoch": 29.96,
"learning_rate": 0.0002,
"loss": 1.0239,
"step": 386000
},
{
"epoch": 30.0,
"eval_bleu": 18.053,
"eval_gen_len": 16.8423,
"eval_runtime": 67.9635,
"eval_samples_per_second": 13.066,
"eval_steps_per_second": 1.633,
"step": 386460
},
{
"epoch": 30.0,
"learning_rate": 0.0002,
"loss": 1.0228,
"step": 386500
},
{
"epoch": 30.04,
"learning_rate": 0.0002,
"loss": 0.9594,
"step": 387000
},
{
"epoch": 30.08,
"learning_rate": 0.0002,
"loss": 0.9734,
"step": 387500
},
{
"epoch": 30.12,
"learning_rate": 0.0002,
"loss": 0.9852,
"step": 388000
},
{
"epoch": 30.16,
"learning_rate": 0.0002,
"loss": 0.9749,
"step": 388500
},
{
"epoch": 30.2,
"learning_rate": 0.0002,
"loss": 0.9799,
"step": 389000
},
{
"epoch": 30.24,
"learning_rate": 0.0002,
"loss": 0.9895,
"step": 389500
},
{
"epoch": 30.27,
"learning_rate": 0.0002,
"loss": 0.9901,
"step": 390000
},
{
"epoch": 30.31,
"learning_rate": 0.0002,
"loss": 0.9877,
"step": 390500
},
{
"epoch": 30.35,
"learning_rate": 0.0002,
"loss": 1.0013,
"step": 391000
},
{
"epoch": 30.39,
"learning_rate": 0.0002,
"loss": 0.9998,
"step": 391500
},
{
"epoch": 30.43,
"learning_rate": 0.0002,
"loss": 0.9856,
"step": 392000
},
{
"epoch": 30.47,
"learning_rate": 0.0002,
"loss": 0.9884,
"step": 392500
},
{
"epoch": 30.51,
"learning_rate": 0.0002,
"loss": 0.9962,
"step": 393000
},
{
"epoch": 30.55,
"learning_rate": 0.0002,
"loss": 1.0024,
"step": 393500
},
{
"epoch": 30.59,
"learning_rate": 0.0002,
"loss": 0.9933,
"step": 394000
},
{
"epoch": 30.62,
"learning_rate": 0.0002,
"loss": 0.9983,
"step": 394500
},
{
"epoch": 30.66,
"learning_rate": 0.0002,
"loss": 1.0059,
"step": 395000
},
{
"epoch": 30.7,
"learning_rate": 0.0002,
"loss": 1.0063,
"step": 395500
},
{
"epoch": 30.74,
"learning_rate": 0.0002,
"loss": 1.0029,
"step": 396000
},
{
"epoch": 30.78,
"learning_rate": 0.0002,
"loss": 1.0162,
"step": 396500
},
{
"epoch": 30.82,
"learning_rate": 0.0002,
"loss": 1.0032,
"step": 397000
},
{
"epoch": 30.86,
"learning_rate": 0.0002,
"loss": 1.0047,
"step": 397500
},
{
"epoch": 30.9,
"learning_rate": 0.0002,
"loss": 1.0174,
"step": 398000
},
{
"epoch": 30.93,
"learning_rate": 0.0002,
"loss": 1.016,
"step": 398500
},
{
"epoch": 30.97,
"learning_rate": 0.0002,
"loss": 1.0126,
"step": 399000
},
{
"epoch": 31.0,
"eval_bleu": 17.9637,
"eval_gen_len": 16.8818,
"eval_runtime": 61.6337,
"eval_samples_per_second": 14.408,
"eval_steps_per_second": 1.801,
"step": 399342
},
{
"epoch": 31.01,
"learning_rate": 0.0002,
"loss": 1.0004,
"step": 399500
},
{
"epoch": 31.05,
"learning_rate": 0.0002,
"loss": 0.9761,
"step": 400000
},
{
"epoch": 31.09,
"learning_rate": 0.0002,
"loss": 0.957,
"step": 400500
},
{
"epoch": 31.13,
"learning_rate": 0.0002,
"loss": 0.9687,
"step": 401000
},
{
"epoch": 31.17,
"learning_rate": 0.0002,
"loss": 0.9708,
"step": 401500
},
{
"epoch": 31.21,
"learning_rate": 0.0002,
"loss": 0.9704,
"step": 402000
},
{
"epoch": 31.25,
"learning_rate": 0.0002,
"loss": 0.9664,
"step": 402500
},
{
"epoch": 31.28,
"learning_rate": 0.0002,
"loss": 0.9717,
"step": 403000
},
{
"epoch": 31.32,
"learning_rate": 0.0002,
"loss": 0.9817,
"step": 403500
},
{
"epoch": 31.36,
"learning_rate": 0.0002,
"loss": 0.9783,
"step": 404000
},
{
"epoch": 31.4,
"learning_rate": 0.0002,
"loss": 0.9828,
"step": 404500
},
{
"epoch": 31.44,
"learning_rate": 0.0002,
"loss": 0.9794,
"step": 405000
},
{
"epoch": 31.48,
"learning_rate": 0.0002,
"loss": 0.9803,
"step": 405500
},
{
"epoch": 31.52,
"learning_rate": 0.0002,
"loss": 0.9923,
"step": 406000
},
{
"epoch": 31.56,
"learning_rate": 0.0002,
"loss": 0.9765,
"step": 406500
},
{
"epoch": 31.59,
"learning_rate": 0.0002,
"loss": 0.9881,
"step": 407000
},
{
"epoch": 31.63,
"learning_rate": 0.0002,
"loss": 0.9911,
"step": 407500
},
{
"epoch": 31.67,
"learning_rate": 0.0002,
"loss": 0.9914,
"step": 408000
},
{
"epoch": 31.71,
"learning_rate": 0.0002,
"loss": 0.9898,
"step": 408500
},
{
"epoch": 31.75,
"learning_rate": 0.0002,
"loss": 0.9958,
"step": 409000
},
{
"epoch": 31.79,
"learning_rate": 0.0002,
"loss": 1.0036,
"step": 409500
},
{
"epoch": 31.83,
"learning_rate": 0.0002,
"loss": 0.9993,
"step": 410000
},
{
"epoch": 31.87,
"learning_rate": 0.0002,
"loss": 1.0074,
"step": 410500
},
{
"epoch": 31.9,
"learning_rate": 0.0002,
"loss": 1.0082,
"step": 411000
},
{
"epoch": 31.94,
"learning_rate": 0.0002,
"loss": 1.001,
"step": 411500
},
{
"epoch": 31.98,
"learning_rate": 0.0002,
"loss": 1.0039,
"step": 412000
},
{
"epoch": 32.0,
"eval_bleu": 18.1766,
"eval_gen_len": 16.8908,
"eval_runtime": 61.7857,
"eval_samples_per_second": 14.372,
"eval_steps_per_second": 1.797,
"step": 412224
},
{
"epoch": 32.02,
"learning_rate": 0.0002,
"loss": 0.9791,
"step": 412500
},
{
"epoch": 32.06,
"learning_rate": 0.0002,
"loss": 0.9522,
"step": 413000
},
{
"epoch": 32.1,
"learning_rate": 0.0002,
"loss": 0.9603,
"step": 413500
},
{
"epoch": 32.14,
"learning_rate": 0.0002,
"loss": 0.9552,
"step": 414000
},
{
"epoch": 32.18,
"learning_rate": 0.0002,
"loss": 0.9608,
"step": 414500
},
{
"epoch": 32.22,
"learning_rate": 0.0002,
"loss": 0.9571,
"step": 415000
},
{
"epoch": 32.25,
"learning_rate": 0.0002,
"loss": 0.9775,
"step": 415500
},
{
"epoch": 32.29,
"learning_rate": 0.0002,
"loss": 0.9652,
"step": 416000
},
{
"epoch": 32.33,
"learning_rate": 0.0002,
"loss": 0.9726,
"step": 416500
},
{
"epoch": 32.37,
"learning_rate": 0.0002,
"loss": 0.9688,
"step": 417000
},
{
"epoch": 32.41,
"learning_rate": 0.0002,
"loss": 0.9765,
"step": 417500
},
{
"epoch": 32.45,
"learning_rate": 0.0002,
"loss": 0.9718,
"step": 418000
},
{
"epoch": 32.49,
"learning_rate": 0.0002,
"loss": 0.9796,
"step": 418500
},
{
"epoch": 32.53,
"learning_rate": 0.0002,
"loss": 0.9853,
"step": 419000
},
{
"epoch": 32.56,
"learning_rate": 0.0002,
"loss": 0.9733,
"step": 419500
},
{
"epoch": 32.6,
"learning_rate": 0.0002,
"loss": 0.9786,
"step": 420000
},
{
"epoch": 32.64,
"learning_rate": 0.0002,
"loss": 0.9871,
"step": 420500
},
{
"epoch": 32.68,
"learning_rate": 0.0002,
"loss": 0.9823,
"step": 421000
},
{
"epoch": 32.72,
"learning_rate": 0.0002,
"loss": 0.9776,
"step": 421500
},
{
"epoch": 32.76,
"learning_rate": 0.0002,
"loss": 0.975,
"step": 422000
},
{
"epoch": 32.8,
"learning_rate": 0.0002,
"loss": 0.9788,
"step": 422500
},
{
"epoch": 32.84,
"learning_rate": 0.0002,
"loss": 0.9788,
"step": 423000
},
{
"epoch": 32.88,
"learning_rate": 0.0002,
"loss": 0.9927,
"step": 423500
},
{
"epoch": 32.91,
"learning_rate": 0.0002,
"loss": 0.996,
"step": 424000
},
{
"epoch": 32.95,
"learning_rate": 0.0002,
"loss": 0.9862,
"step": 424500
},
{
"epoch": 32.99,
"learning_rate": 0.0002,
"loss": 0.9898,
"step": 425000
},
{
"epoch": 33.0,
"eval_bleu": 18.1644,
"eval_gen_len": 16.8671,
"eval_runtime": 61.6543,
"eval_samples_per_second": 14.403,
"eval_steps_per_second": 1.8,
"step": 425106
},
{
"epoch": 33.03,
"learning_rate": 0.0002,
"loss": 0.9531,
"step": 425500
},
{
"epoch": 33.07,
"learning_rate": 0.0002,
"loss": 0.942,
"step": 426000
},
{
"epoch": 33.11,
"learning_rate": 0.0002,
"loss": 0.9319,
"step": 426500
},
{
"epoch": 33.15,
"learning_rate": 0.0002,
"loss": 0.9436,
"step": 427000
},
{
"epoch": 33.19,
"learning_rate": 0.0002,
"loss": 0.9437,
"step": 427500
},
{
"epoch": 33.22,
"learning_rate": 0.0002,
"loss": 0.953,
"step": 428000
},
{
"epoch": 33.26,
"learning_rate": 0.0002,
"loss": 0.9571,
"step": 428500
},
{
"epoch": 33.3,
"learning_rate": 0.0002,
"loss": 0.9602,
"step": 429000
},
{
"epoch": 33.34,
"learning_rate": 0.0002,
"loss": 0.9675,
"step": 429500
},
{
"epoch": 33.38,
"learning_rate": 0.0002,
"loss": 0.9566,
"step": 430000
},
{
"epoch": 33.42,
"learning_rate": 0.0002,
"loss": 0.9638,
"step": 430500
},
{
"epoch": 33.46,
"learning_rate": 0.0002,
"loss": 0.9617,
"step": 431000
},
{
"epoch": 33.5,
"learning_rate": 0.0002,
"loss": 0.9765,
"step": 431500
},
{
"epoch": 33.54,
"learning_rate": 0.0002,
"loss": 0.9608,
"step": 432000
},
{
"epoch": 33.57,
"learning_rate": 0.0002,
"loss": 0.9622,
"step": 432500
},
{
"epoch": 33.61,
"learning_rate": 0.0002,
"loss": 0.9713,
"step": 433000
},
{
"epoch": 33.65,
"learning_rate": 0.0002,
"loss": 0.9781,
"step": 433500
},
{
"epoch": 33.69,
"learning_rate": 0.0002,
"loss": 0.9751,
"step": 434000
},
{
"epoch": 33.73,
"learning_rate": 0.0002,
"loss": 0.9765,
"step": 434500
},
{
"epoch": 33.77,
"learning_rate": 0.0002,
"loss": 0.9751,
"step": 435000
},
{
"epoch": 33.81,
"learning_rate": 0.0002,
"loss": 0.9754,
"step": 435500
},
{
"epoch": 33.85,
"learning_rate": 0.0002,
"loss": 0.9695,
"step": 436000
},
{
"epoch": 33.88,
"learning_rate": 0.0002,
"loss": 0.9706,
"step": 436500
},
{
"epoch": 33.92,
"learning_rate": 0.0002,
"loss": 0.9821,
"step": 437000
},
{
"epoch": 33.96,
"learning_rate": 0.0002,
"loss": 0.9688,
"step": 437500
},
{
"epoch": 34.0,
"eval_bleu": 18.1318,
"eval_gen_len": 16.9009,
"eval_runtime": 62.0566,
"eval_samples_per_second": 14.31,
"eval_steps_per_second": 1.789,
"step": 437988
},
{
"epoch": 34.0,
"learning_rate": 0.0002,
"loss": 0.984,
"step": 438000
},
{
"epoch": 34.04,
"learning_rate": 0.0002,
"loss": 0.9307,
"step": 438500
},
{
"epoch": 34.08,
"learning_rate": 0.0002,
"loss": 0.9225,
"step": 439000
},
{
"epoch": 34.12,
"learning_rate": 0.0002,
"loss": 0.9279,
"step": 439500
},
{
"epoch": 34.16,
"learning_rate": 0.0002,
"loss": 0.9382,
"step": 440000
},
{
"epoch": 34.2,
"learning_rate": 0.0002,
"loss": 0.942,
"step": 440500
},
{
"epoch": 34.23,
"learning_rate": 0.0002,
"loss": 0.9408,
"step": 441000
},
{
"epoch": 34.27,
"learning_rate": 0.0002,
"loss": 0.9443,
"step": 441500
},
{
"epoch": 34.31,
"learning_rate": 0.0002,
"loss": 0.9406,
"step": 442000
},
{
"epoch": 34.35,
"learning_rate": 0.0002,
"loss": 0.9473,
"step": 442500
},
{
"epoch": 34.39,
"learning_rate": 0.0002,
"loss": 0.954,
"step": 443000
},
{
"epoch": 34.43,
"learning_rate": 0.0002,
"loss": 0.9582,
"step": 443500
},
{
"epoch": 34.47,
"learning_rate": 0.0002,
"loss": 0.9501,
"step": 444000
},
{
"epoch": 34.51,
"learning_rate": 0.0002,
"loss": 0.9616,
"step": 444500
},
{
"epoch": 34.54,
"learning_rate": 0.0002,
"loss": 0.9538,
"step": 445000
},
{
"epoch": 34.58,
"learning_rate": 0.0002,
"loss": 0.9533,
"step": 445500
},
{
"epoch": 34.62,
"learning_rate": 0.0002,
"loss": 0.9624,
"step": 446000
},
{
"epoch": 34.66,
"learning_rate": 0.0002,
"loss": 0.9515,
"step": 446500
},
{
"epoch": 34.7,
"learning_rate": 0.0002,
"loss": 0.9574,
"step": 447000
},
{
"epoch": 34.74,
"learning_rate": 0.0002,
"loss": 0.96,
"step": 447500
},
{
"epoch": 34.78,
"learning_rate": 0.0002,
"loss": 0.9568,
"step": 448000
},
{
"epoch": 34.82,
"learning_rate": 0.0002,
"loss": 0.9664,
"step": 448500
},
{
"epoch": 34.85,
"learning_rate": 0.0002,
"loss": 0.9747,
"step": 449000
},
{
"epoch": 34.89,
"learning_rate": 0.0002,
"loss": 0.9759,
"step": 449500
},
{
"epoch": 34.93,
"learning_rate": 0.0002,
"loss": 0.9705,
"step": 450000
},
{
"epoch": 34.97,
"learning_rate": 0.0002,
"loss": 0.9766,
"step": 450500
},
{
"epoch": 35.0,
"eval_bleu": 18.323,
"eval_gen_len": 16.866,
"eval_runtime": 68.3726,
"eval_samples_per_second": 12.988,
"eval_steps_per_second": 1.623,
"step": 450870
},
{
"epoch": 35.01,
"learning_rate": 0.0002,
"loss": 0.9624,
"step": 451000
},
{
"epoch": 35.05,
"learning_rate": 0.0002,
"loss": 0.9246,
"step": 451500
},
{
"epoch": 35.09,
"learning_rate": 0.0002,
"loss": 0.9158,
"step": 452000
},
{
"epoch": 35.13,
"learning_rate": 0.0002,
"loss": 0.9284,
"step": 452500
},
{
"epoch": 35.17,
"learning_rate": 0.0002,
"loss": 0.9191,
"step": 453000
},
{
"epoch": 35.2,
"learning_rate": 0.0002,
"loss": 0.9191,
"step": 453500
},
{
"epoch": 35.24,
"learning_rate": 0.0002,
"loss": 0.9344,
"step": 454000
},
{
"epoch": 35.28,
"learning_rate": 0.0002,
"loss": 0.9317,
"step": 454500
},
{
"epoch": 35.32,
"learning_rate": 0.0002,
"loss": 0.933,
"step": 455000
},
{
"epoch": 35.36,
"learning_rate": 0.0002,
"loss": 0.9384,
"step": 455500
},
{
"epoch": 35.4,
"learning_rate": 0.0002,
"loss": 0.9439,
"step": 456000
},
{
"epoch": 35.44,
"learning_rate": 0.0002,
"loss": 0.9408,
"step": 456500
},
{
"epoch": 35.48,
"learning_rate": 0.0002,
"loss": 0.9405,
"step": 457000
},
{
"epoch": 35.51,
"learning_rate": 0.0002,
"loss": 0.9442,
"step": 457500
},
{
"epoch": 35.55,
"learning_rate": 0.0002,
"loss": 0.9438,
"step": 458000
},
{
"epoch": 35.59,
"learning_rate": 0.0002,
"loss": 0.955,
"step": 458500
},
{
"epoch": 35.63,
"learning_rate": 0.0002,
"loss": 0.9492,
"step": 459000
},
{
"epoch": 35.67,
"learning_rate": 0.0002,
"loss": 0.9425,
"step": 459500
},
{
"epoch": 35.71,
"learning_rate": 0.0002,
"loss": 0.9476,
"step": 460000
},
{
"epoch": 35.75,
"learning_rate": 0.0002,
"loss": 0.9579,
"step": 460500
},
{
"epoch": 35.79,
"learning_rate": 0.0002,
"loss": 0.9541,
"step": 461000
},
{
"epoch": 35.83,
"learning_rate": 0.0002,
"loss": 0.9643,
"step": 461500
},
{
"epoch": 35.86,
"learning_rate": 0.0002,
"loss": 0.9594,
"step": 462000
},
{
"epoch": 35.9,
"learning_rate": 0.0002,
"loss": 0.958,
"step": 462500
},
{
"epoch": 35.94,
"learning_rate": 0.0002,
"loss": 0.9654,
"step": 463000
},
{
"epoch": 35.98,
"learning_rate": 0.0002,
"loss": 0.9549,
"step": 463500
},
{
"epoch": 36.0,
"eval_bleu": 18.2868,
"eval_gen_len": 16.8649,
"eval_runtime": 61.8889,
"eval_samples_per_second": 14.348,
"eval_steps_per_second": 1.794,
"step": 463752
},
{
"epoch": 36.02,
"learning_rate": 0.0002,
"loss": 0.9349,
"step": 464000
},
{
"epoch": 36.06,
"learning_rate": 0.0002,
"loss": 0.915,
"step": 464500
},
{
"epoch": 36.1,
"learning_rate": 0.0002,
"loss": 0.9076,
"step": 465000
},
{
"epoch": 36.14,
"learning_rate": 0.0002,
"loss": 0.9159,
"step": 465500
},
{
"epoch": 36.17,
"learning_rate": 0.0002,
"loss": 0.919,
"step": 466000
},
{
"epoch": 36.21,
"learning_rate": 0.0002,
"loss": 0.9191,
"step": 466500
},
{
"epoch": 36.25,
"learning_rate": 0.0002,
"loss": 0.9185,
"step": 467000
},
{
"epoch": 36.29,
"learning_rate": 0.0002,
"loss": 0.919,
"step": 467500
},
{
"epoch": 36.33,
"learning_rate": 0.0002,
"loss": 0.9266,
"step": 468000
},
{
"epoch": 36.37,
"learning_rate": 0.0002,
"loss": 0.926,
"step": 468500
},
{
"epoch": 36.41,
"learning_rate": 0.0002,
"loss": 0.9339,
"step": 469000
},
{
"epoch": 36.45,
"learning_rate": 0.0002,
"loss": 0.931,
"step": 469500
},
{
"epoch": 36.49,
"learning_rate": 0.0002,
"loss": 0.9335,
"step": 470000
},
{
"epoch": 36.52,
"learning_rate": 0.0002,
"loss": 0.9301,
"step": 470500
},
{
"epoch": 36.56,
"learning_rate": 0.0002,
"loss": 0.9267,
"step": 471000
},
{
"epoch": 36.6,
"learning_rate": 0.0002,
"loss": 0.9399,
"step": 471500
},
{
"epoch": 36.64,
"learning_rate": 0.0002,
"loss": 0.9408,
"step": 472000
},
{
"epoch": 36.68,
"learning_rate": 0.0002,
"loss": 0.9327,
"step": 472500
},
{
"epoch": 36.72,
"learning_rate": 0.0002,
"loss": 0.9464,
"step": 473000
},
{
"epoch": 36.76,
"learning_rate": 0.0002,
"loss": 0.9421,
"step": 473500
},
{
"epoch": 36.8,
"learning_rate": 0.0002,
"loss": 0.9483,
"step": 474000
},
{
"epoch": 36.83,
"learning_rate": 0.0002,
"loss": 0.946,
"step": 474500
},
{
"epoch": 36.87,
"learning_rate": 0.0002,
"loss": 0.9619,
"step": 475000
},
{
"epoch": 36.91,
"learning_rate": 0.0002,
"loss": 0.9468,
"step": 475500
},
{
"epoch": 36.95,
"learning_rate": 0.0002,
"loss": 0.9513,
"step": 476000
},
{
"epoch": 36.99,
"learning_rate": 0.0002,
"loss": 0.9443,
"step": 476500
},
{
"epoch": 37.0,
"eval_bleu": 17.956,
"eval_gen_len": 16.8806,
"eval_runtime": 61.3127,
"eval_samples_per_second": 14.483,
"eval_steps_per_second": 1.81,
"step": 476634
},
{
"epoch": 37.03,
"learning_rate": 0.0002,
"loss": 0.909,
"step": 477000
},
{
"epoch": 37.07,
"learning_rate": 0.0002,
"loss": 0.8977,
"step": 477500
},
{
"epoch": 37.11,
"learning_rate": 0.0002,
"loss": 0.8937,
"step": 478000
},
{
"epoch": 37.14,
"learning_rate": 0.0002,
"loss": 0.9056,
"step": 478500
},
{
"epoch": 37.18,
"learning_rate": 0.0002,
"loss": 0.9061,
"step": 479000
},
{
"epoch": 37.22,
"learning_rate": 0.0002,
"loss": 0.9103,
"step": 479500
},
{
"epoch": 37.26,
"learning_rate": 0.0002,
"loss": 0.9094,
"step": 480000
},
{
"epoch": 37.3,
"learning_rate": 0.0002,
"loss": 0.9171,
"step": 480500
},
{
"epoch": 37.34,
"learning_rate": 0.0002,
"loss": 0.9133,
"step": 481000
},
{
"epoch": 37.38,
"learning_rate": 0.0002,
"loss": 0.916,
"step": 481500
},
{
"epoch": 37.42,
"learning_rate": 0.0002,
"loss": 0.9221,
"step": 482000
},
{
"epoch": 37.46,
"learning_rate": 0.0002,
"loss": 0.9297,
"step": 482500
},
{
"epoch": 37.49,
"learning_rate": 0.0002,
"loss": 0.9242,
"step": 483000
},
{
"epoch": 37.53,
"learning_rate": 0.0002,
"loss": 0.9251,
"step": 483500
},
{
"epoch": 37.57,
"learning_rate": 0.0002,
"loss": 0.931,
"step": 484000
},
{
"epoch": 37.61,
"learning_rate": 0.0002,
"loss": 0.9262,
"step": 484500
},
{
"epoch": 37.65,
"learning_rate": 0.0002,
"loss": 0.9254,
"step": 485000
},
{
"epoch": 37.69,
"learning_rate": 0.0002,
"loss": 0.9403,
"step": 485500
},
{
"epoch": 37.73,
"learning_rate": 0.0002,
"loss": 0.9329,
"step": 486000
},
{
"epoch": 37.77,
"learning_rate": 0.0002,
"loss": 0.9434,
"step": 486500
},
{
"epoch": 37.8,
"learning_rate": 0.0002,
"loss": 0.9337,
"step": 487000
},
{
"epoch": 37.84,
"learning_rate": 0.0002,
"loss": 0.9353,
"step": 487500
},
{
"epoch": 37.88,
"learning_rate": 0.0002,
"loss": 0.9366,
"step": 488000
},
{
"epoch": 37.92,
"learning_rate": 0.0002,
"loss": 0.9463,
"step": 488500
},
{
"epoch": 37.96,
"learning_rate": 0.0002,
"loss": 0.9498,
"step": 489000
},
{
"epoch": 38.0,
"learning_rate": 0.0002,
"loss": 0.949,
"step": 489500
},
{
"epoch": 38.0,
"eval_bleu": 18.1049,
"eval_gen_len": 16.8739,
"eval_runtime": 61.2496,
"eval_samples_per_second": 14.498,
"eval_steps_per_second": 1.812,
"step": 489516
},
{
"epoch": 38.04,
"learning_rate": 0.0002,
"loss": 0.8924,
"step": 490000
},
{
"epoch": 38.08,
"learning_rate": 0.0002,
"loss": 0.8943,
"step": 490500
},
{
"epoch": 38.12,
"learning_rate": 0.0002,
"loss": 0.8936,
"step": 491000
},
{
"epoch": 38.15,
"learning_rate": 0.0002,
"loss": 0.9009,
"step": 491500
},
{
"epoch": 38.19,
"learning_rate": 0.0002,
"loss": 0.9069,
"step": 492000
},
{
"epoch": 38.23,
"learning_rate": 0.0002,
"loss": 0.9011,
"step": 492500
},
{
"epoch": 38.27,
"learning_rate": 0.0002,
"loss": 0.9018,
"step": 493000
},
{
"epoch": 38.31,
"learning_rate": 0.0002,
"loss": 0.9103,
"step": 493500
},
{
"epoch": 38.35,
"learning_rate": 0.0002,
"loss": 0.908,
"step": 494000
},
{
"epoch": 38.39,
"learning_rate": 0.0002,
"loss": 0.9002,
"step": 494500
},
{
"epoch": 38.43,
"learning_rate": 0.0002,
"loss": 0.9106,
"step": 495000
},
{
"epoch": 38.46,
"learning_rate": 0.0002,
"loss": 0.9122,
"step": 495500
},
{
"epoch": 38.5,
"learning_rate": 0.0002,
"loss": 0.907,
"step": 496000
},
{
"epoch": 38.54,
"learning_rate": 0.0002,
"loss": 0.9158,
"step": 496500
},
{
"epoch": 38.58,
"learning_rate": 0.0002,
"loss": 0.9171,
"step": 497000
},
{
"epoch": 38.62,
"learning_rate": 0.0002,
"loss": 0.9207,
"step": 497500
},
{
"epoch": 38.66,
"learning_rate": 0.0002,
"loss": 0.9253,
"step": 498000
},
{
"epoch": 38.7,
"learning_rate": 0.0002,
"loss": 0.916,
"step": 498500
},
{
"epoch": 38.74,
"learning_rate": 0.0002,
"loss": 0.9227,
"step": 499000
},
{
"epoch": 38.78,
"learning_rate": 0.0002,
"loss": 0.9294,
"step": 499500
},
{
"epoch": 38.81,
"learning_rate": 0.0002,
"loss": 0.929,
"step": 500000
},
{
"epoch": 38.85,
"learning_rate": 0.0002,
"loss": 0.9228,
"step": 500500
},
{
"epoch": 38.89,
"learning_rate": 0.0002,
"loss": 0.9311,
"step": 501000
},
{
"epoch": 38.93,
"learning_rate": 0.0002,
"loss": 0.9302,
"step": 501500
},
{
"epoch": 38.97,
"learning_rate": 0.0002,
"loss": 0.9284,
"step": 502000
},
{
"epoch": 39.0,
"eval_bleu": 17.9856,
"eval_gen_len": 16.8739,
"eval_runtime": 61.1301,
"eval_samples_per_second": 14.526,
"eval_steps_per_second": 1.816,
"step": 502398
}
],
"max_steps": 644100,
"num_train_epochs": 50,
"total_flos": 5.1575061541075354e+17,
"trial_name": null,
"trial_params": null
}