bart-large-entity / trainer_state.json
Xmm's picture
Upload 14 files
f81cb74
{
"best_metric": 63.1011,
"best_model_checkpoint": "output_train_bart_large_local/checkpoint-48000",
"epoch": 2.0123153700647967,
"global_step": 50000,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.04,
"learning_rate": 3.9842240824211213e-07,
"loss": 1.8309,
"step": 500
},
{
"epoch": 0.08,
"learning_rate": 8.008692852543465e-07,
"loss": 1.0322,
"step": 1000
},
{
"epoch": 0.08,
"eval_bertscore/f1": 0.7264,
"eval_bertscore/precision": 0.7345,
"eval_bertscore/recall": 0.7214,
"eval_mean_prediction_length_characters": 779.505,
"eval_mean_prediction_length_tokens": 167.969,
"eval_num_predicted": 1000,
"eval_rouge/geometric_mean": 45.7604,
"eval_rouge/rouge1": 60.4328,
"eval_rouge/rouge2": 39.8155,
"eval_rouge/rougeL": 39.824,
"eval_rouge/rougeLsum": 57.577,
"eval_runtime": 2363.9821,
"eval_samples_per_second": 0.423,
"eval_steps_per_second": 0.423,
"step": 1000
},
{
"epoch": 0.12,
"learning_rate": 1.203316162266581e-06,
"loss": 0.8078,
"step": 1500
},
{
"epoch": 0.16,
"learning_rate": 1.6057630392788153e-06,
"loss": 0.7401,
"step": 2000
},
{
"epoch": 0.16,
"eval_bertscore/f1": 0.767,
"eval_bertscore/precision": 0.7735,
"eval_bertscore/recall": 0.7627,
"eval_mean_prediction_length_characters": 746.987,
"eval_mean_prediction_length_tokens": 167.637,
"eval_num_predicted": 1000,
"eval_rouge/geometric_mean": 49.7632,
"eval_rouge/rouge1": 64.5596,
"eval_rouge/rouge2": 43.5883,
"eval_rouge/rougeL": 43.7919,
"eval_rouge/rougeLsum": 62.268,
"eval_runtime": 2390.2778,
"eval_samples_per_second": 0.418,
"eval_steps_per_second": 0.418,
"step": 2000
},
{
"epoch": 0.2,
"learning_rate": 2.00820991629105e-06,
"loss": 0.6952,
"step": 2500
},
{
"epoch": 0.24,
"learning_rate": 2.4106567933032843e-06,
"loss": 0.6654,
"step": 3000
},
{
"epoch": 0.24,
"eval_bertscore/f1": 0.775,
"eval_bertscore/precision": 0.7801,
"eval_bertscore/recall": 0.7719,
"eval_mean_prediction_length_characters": 750.739,
"eval_mean_prediction_length_tokens": 167.441,
"eval_num_predicted": 1000,
"eval_rouge/geometric_mean": 51.1255,
"eval_rouge/rouge1": 65.8167,
"eval_rouge/rouge2": 45.0215,
"eval_rouge/rougeL": 45.0979,
"eval_rouge/rougeLsum": 63.4694,
"eval_runtime": 2400.8698,
"eval_samples_per_second": 0.417,
"eval_steps_per_second": 0.417,
"step": 3000
},
{
"epoch": 0.28,
"learning_rate": 2.8131036703155183e-06,
"loss": 0.6532,
"step": 3500
},
{
"epoch": 0.32,
"learning_rate": 3.215550547327753e-06,
"loss": 0.6421,
"step": 4000
},
{
"epoch": 0.32,
"eval_bertscore/f1": 0.778,
"eval_bertscore/precision": 0.7803,
"eval_bertscore/recall": 0.7778,
"eval_mean_prediction_length_characters": 787.437,
"eval_mean_prediction_length_tokens": 175.145,
"eval_num_predicted": 1000,
"eval_rouge/geometric_mean": 51.6325,
"eval_rouge/rouge1": 66.6546,
"eval_rouge/rouge2": 45.5613,
"eval_rouge/rougeL": 45.3256,
"eval_rouge/rougeLsum": 64.2944,
"eval_runtime": 2500.2502,
"eval_samples_per_second": 0.4,
"eval_steps_per_second": 0.4,
"step": 4000
},
{
"epoch": 0.36,
"learning_rate": 3.6179974243399875e-06,
"loss": 0.6253,
"step": 4500
},
{
"epoch": 0.4,
"learning_rate": 4.0196394075981975e-06,
"loss": 0.62,
"step": 5000
},
{
"epoch": 0.4,
"eval_bertscore/f1": 0.7774,
"eval_bertscore/precision": 0.7652,
"eval_bertscore/recall": 0.7921,
"eval_mean_prediction_length_characters": 948.673,
"eval_mean_prediction_length_tokens": 206.93,
"eval_num_predicted": 1000,
"eval_rouge/geometric_mean": 51.4967,
"eval_rouge/rouge1": 67.0983,
"eval_rouge/rouge2": 45.6949,
"eval_rouge/rougeL": 44.541,
"eval_rouge/rougeLsum": 64.5765,
"eval_runtime": 2980.0062,
"eval_samples_per_second": 0.336,
"eval_steps_per_second": 0.336,
"step": 5000
},
{
"epoch": 0.44,
"learning_rate": 4.4220862846104314e-06,
"loss": 0.6041,
"step": 5500
},
{
"epoch": 0.48,
"learning_rate": 4.824533161622666e-06,
"loss": 0.5949,
"step": 6000
},
{
"epoch": 0.48,
"eval_bertscore/f1": 0.7803,
"eval_bertscore/precision": 0.7732,
"eval_bertscore/recall": 0.7895,
"eval_mean_prediction_length_characters": 879.72,
"eval_mean_prediction_length_tokens": 192.817,
"eval_num_predicted": 1000,
"eval_rouge/geometric_mean": 52.1495,
"eval_rouge/rouge1": 67.3707,
"eval_rouge/rouge2": 46.4072,
"eval_rouge/rougeL": 45.3622,
"eval_rouge/rougeLsum": 65.0424,
"eval_runtime": 2666.0858,
"eval_samples_per_second": 0.375,
"eval_steps_per_second": 0.375,
"step": 6000
},
{
"epoch": 0.52,
"learning_rate": 5.2269800386349e-06,
"loss": 0.5762,
"step": 6500
},
{
"epoch": 0.56,
"learning_rate": 5.629426915647135e-06,
"loss": 0.5719,
"step": 7000
},
{
"epoch": 0.56,
"eval_bertscore/f1": 0.7838,
"eval_bertscore/precision": 0.7832,
"eval_bertscore/recall": 0.7866,
"eval_mean_prediction_length_characters": 832.624,
"eval_mean_prediction_length_tokens": 183.52,
"eval_num_predicted": 1000,
"eval_rouge/geometric_mean": 52.9627,
"eval_rouge/rouge1": 67.8554,
"eval_rouge/rouge2": 47.0328,
"eval_rouge/rougeL": 46.5505,
"eval_rouge/rougeLsum": 65.5148,
"eval_runtime": 2493.7707,
"eval_samples_per_second": 0.401,
"eval_steps_per_second": 0.401,
"step": 7000
},
{
"epoch": 0.6,
"learning_rate": 6.031068898905345e-06,
"loss": 0.5718,
"step": 7500
},
{
"epoch": 0.64,
"learning_rate": 6.433515775917579e-06,
"loss": 0.5541,
"step": 8000
},
{
"epoch": 0.64,
"eval_bertscore/f1": 0.7856,
"eval_bertscore/precision": 0.7835,
"eval_bertscore/recall": 0.7897,
"eval_mean_prediction_length_characters": 831.437,
"eval_mean_prediction_length_tokens": 184.28,
"eval_num_predicted": 1000,
"eval_rouge/geometric_mean": 53.234,
"eval_rouge/rouge1": 68.2648,
"eval_rouge/rouge2": 47.5788,
"eval_rouge/rougeL": 46.447,
"eval_rouge/rougeLsum": 65.9056,
"eval_runtime": 2476.0861,
"eval_samples_per_second": 0.404,
"eval_steps_per_second": 0.404,
"step": 8000
},
{
"epoch": 0.68,
"learning_rate": 6.835962652929814e-06,
"loss": 0.5621,
"step": 8500
},
{
"epoch": 0.72,
"learning_rate": 7.238409529942049e-06,
"loss": 0.5509,
"step": 9000
},
{
"epoch": 0.72,
"eval_bertscore/f1": 0.7876,
"eval_bertscore/precision": 0.787,
"eval_bertscore/recall": 0.79,
"eval_mean_prediction_length_characters": 825.034,
"eval_mean_prediction_length_tokens": 181.78,
"eval_num_predicted": 1000,
"eval_rouge/geometric_mean": 53.6596,
"eval_rouge/rouge1": 68.4619,
"eval_rouge/rouge2": 47.9553,
"eval_rouge/rougeL": 47.0605,
"eval_rouge/rougeLsum": 66.229,
"eval_runtime": 2413.4877,
"eval_samples_per_second": 0.414,
"eval_steps_per_second": 0.414,
"step": 9000
},
{
"epoch": 0.76,
"learning_rate": 7.640856406954282e-06,
"loss": 0.5319,
"step": 9500
},
{
"epoch": 0.8,
"learning_rate": 8.042498390212493e-06,
"loss": 0.5317,
"step": 10000
},
{
"epoch": 0.8,
"eval_bertscore/f1": 0.7801,
"eval_bertscore/precision": 0.7653,
"eval_bertscore/recall": 0.7976,
"eval_mean_prediction_length_characters": 983.333,
"eval_mean_prediction_length_tokens": 212.132,
"eval_num_predicted": 1000,
"eval_rouge/geometric_mean": 52.2204,
"eval_rouge/rouge1": 67.4114,
"eval_rouge/rouge2": 46.8366,
"eval_rouge/rougeL": 45.1025,
"eval_rouge/rougeLsum": 64.9737,
"eval_runtime": 2850.8965,
"eval_samples_per_second": 0.351,
"eval_steps_per_second": 0.351,
"step": 10000
},
{
"epoch": 0.85,
"learning_rate": 8.444945267224727e-06,
"loss": 0.5246,
"step": 10500
},
{
"epoch": 0.89,
"learning_rate": 8.847392144236962e-06,
"loss": 0.5306,
"step": 11000
},
{
"epoch": 0.89,
"eval_bertscore/f1": 0.7812,
"eval_bertscore/precision": 0.7727,
"eval_bertscore/recall": 0.7923,
"eval_mean_prediction_length_characters": 929.756,
"eval_mean_prediction_length_tokens": 198.992,
"eval_num_predicted": 1000,
"eval_rouge/geometric_mean": 52.9021,
"eval_rouge/rouge1": 67.7205,
"eval_rouge/rouge2": 47.5736,
"eval_rouge/rougeL": 45.955,
"eval_rouge/rougeLsum": 65.2918,
"eval_runtime": 2641.4953,
"eval_samples_per_second": 0.379,
"eval_steps_per_second": 0.379,
"step": 11000
},
{
"epoch": 0.93,
"learning_rate": 9.249839021249196e-06,
"loss": 0.5118,
"step": 11500
},
{
"epoch": 0.97,
"learning_rate": 9.65228589826143e-06,
"loss": 0.5125,
"step": 12000
},
{
"epoch": 0.97,
"eval_bertscore/f1": 0.7924,
"eval_bertscore/precision": 0.8025,
"eval_bertscore/recall": 0.7847,
"eval_mean_prediction_length_characters": 739.515,
"eval_mean_prediction_length_tokens": 166.617,
"eval_num_predicted": 1000,
"eval_rouge/geometric_mean": 54.3827,
"eval_rouge/rouge1": 68.5657,
"eval_rouge/rouge2": 48.8496,
"eval_rouge/rougeL": 48.0192,
"eval_rouge/rougeLsum": 66.6464,
"eval_runtime": 2173.7087,
"eval_samples_per_second": 0.46,
"eval_steps_per_second": 0.46,
"step": 12000
},
{
"epoch": 1.01,
"learning_rate": 9.993918580525149e-06,
"loss": 0.4953,
"step": 12500
},
{
"epoch": 1.05,
"learning_rate": 9.949202260857123e-06,
"loss": 0.4559,
"step": 13000
},
{
"epoch": 1.05,
"eval_bertscore/f1": 0.7876,
"eval_bertscore/precision": 0.782,
"eval_bertscore/recall": 0.7953,
"eval_mean_prediction_length_characters": 872.104,
"eval_mean_prediction_length_tokens": 188.836,
"eval_num_predicted": 1000,
"eval_rouge/geometric_mean": 54.0286,
"eval_rouge/rouge1": 68.9929,
"eval_rouge/rouge2": 48.683,
"eval_rouge/rougeL": 46.9557,
"eval_rouge/rougeLsum": 66.5892,
"eval_runtime": 2531.5352,
"eval_samples_per_second": 0.395,
"eval_steps_per_second": 0.395,
"step": 13000
},
{
"epoch": 1.09,
"learning_rate": 9.904575373828433e-06,
"loss": 0.4572,
"step": 13500
},
{
"epoch": 1.13,
"learning_rate": 9.859859054160406e-06,
"loss": 0.4455,
"step": 14000
},
{
"epoch": 1.13,
"eval_bertscore/f1": 0.792,
"eval_bertscore/precision": 0.7832,
"eval_bertscore/recall": 0.8031,
"eval_mean_prediction_length_characters": 921.871,
"eval_mean_prediction_length_tokens": 200.026,
"eval_num_predicted": 1000,
"eval_rouge/geometric_mean": 54.5933,
"eval_rouge/rouge1": 69.3876,
"eval_rouge/rouge2": 49.1971,
"eval_rouge/rougeL": 47.6645,
"eval_rouge/rougeLsum": 67.1636,
"eval_runtime": 2729.0713,
"eval_samples_per_second": 0.366,
"eval_steps_per_second": 0.366,
"step": 14000
},
{
"epoch": 1.17,
"learning_rate": 9.81514273449238e-06,
"loss": 0.446,
"step": 14500
},
{
"epoch": 1.21,
"learning_rate": 9.770426414824355e-06,
"loss": 0.4376,
"step": 15000
},
{
"epoch": 1.21,
"eval_bertscore/f1": 0.7943,
"eval_bertscore/precision": 0.7909,
"eval_bertscore/recall": 0.7996,
"eval_mean_prediction_length_characters": 856.751,
"eval_mean_prediction_length_tokens": 186.271,
"eval_num_predicted": 1000,
"eval_rouge/geometric_mean": 55.0336,
"eval_rouge/rouge1": 69.6177,
"eval_rouge/rouge2": 49.6842,
"eval_rouge/rougeL": 48.1889,
"eval_rouge/rougeLsum": 67.4597,
"eval_runtime": 2554.3498,
"eval_samples_per_second": 0.391,
"eval_steps_per_second": 0.391,
"step": 15000
},
{
"epoch": 1.25,
"learning_rate": 9.725710095156328e-06,
"loss": 0.4338,
"step": 15500
},
{
"epoch": 1.29,
"learning_rate": 9.680993775488303e-06,
"loss": 0.4333,
"step": 16000
},
{
"epoch": 1.29,
"eval_bertscore/f1": 0.7941,
"eval_bertscore/precision": 0.7901,
"eval_bertscore/recall": 0.8003,
"eval_mean_prediction_length_characters": 857.762,
"eval_mean_prediction_length_tokens": 185.042,
"eval_num_predicted": 1000,
"eval_rouge/geometric_mean": 55.3969,
"eval_rouge/rouge1": 69.5818,
"eval_rouge/rouge2": 50.0439,
"eval_rouge/rougeL": 48.8213,
"eval_rouge/rougeLsum": 67.3532,
"eval_runtime": 2568.5908,
"eval_samples_per_second": 0.389,
"eval_steps_per_second": 0.389,
"step": 16000
},
{
"epoch": 1.33,
"learning_rate": 9.636366888459613e-06,
"loss": 0.4224,
"step": 16500
},
{
"epoch": 1.37,
"learning_rate": 9.591650568791587e-06,
"loss": 0.4229,
"step": 17000
},
{
"epoch": 1.37,
"eval_bertscore/f1": 0.7968,
"eval_bertscore/precision": 0.7936,
"eval_bertscore/recall": 0.802,
"eval_mean_prediction_length_characters": 850.064,
"eval_mean_prediction_length_tokens": 186.438,
"eval_num_predicted": 1000,
"eval_rouge/geometric_mean": 56.2479,
"eval_rouge/rouge1": 70.2657,
"eval_rouge/rouge2": 50.9461,
"eval_rouge/rougeL": 49.7123,
"eval_rouge/rougeLsum": 68.1758,
"eval_runtime": 2596.5556,
"eval_samples_per_second": 0.385,
"eval_steps_per_second": 0.385,
"step": 17000
},
{
"epoch": 1.41,
"learning_rate": 9.54693424912356e-06,
"loss": 0.4215,
"step": 17500
},
{
"epoch": 1.45,
"learning_rate": 9.502217929455535e-06,
"loss": 0.4145,
"step": 18000
},
{
"epoch": 1.45,
"eval_bertscore/f1": 0.7982,
"eval_bertscore/precision": 0.7974,
"eval_bertscore/recall": 0.801,
"eval_mean_prediction_length_characters": 821.211,
"eval_mean_prediction_length_tokens": 179.198,
"eval_num_predicted": 1000,
"eval_rouge/geometric_mean": 56.2997,
"eval_rouge/rouge1": 70.2035,
"eval_rouge/rouge2": 51.1438,
"eval_rouge/rougeL": 49.7011,
"eval_rouge/rougeLsum": 68.1054,
"eval_runtime": 2500.5637,
"eval_samples_per_second": 0.4,
"eval_steps_per_second": 0.4,
"step": 18000
},
{
"epoch": 1.49,
"learning_rate": 9.457501609787508e-06,
"loss": 0.4108,
"step": 18500
},
{
"epoch": 1.53,
"learning_rate": 9.41287472275882e-06,
"loss": 0.413,
"step": 19000
},
{
"epoch": 1.53,
"eval_bertscore/f1": 0.7981,
"eval_bertscore/precision": 0.7992,
"eval_bertscore/recall": 0.799,
"eval_mean_prediction_length_characters": 806.834,
"eval_mean_prediction_length_tokens": 176.24,
"eval_num_predicted": 1000,
"eval_rouge/geometric_mean": 56.5708,
"eval_rouge/rouge1": 70.2439,
"eval_rouge/rouge2": 51.4876,
"eval_rouge/rougeL": 50.0572,
"eval_rouge/rougeLsum": 68.1941,
"eval_runtime": 2443.1073,
"eval_samples_per_second": 0.409,
"eval_steps_per_second": 0.409,
"step": 19000
},
{
"epoch": 1.57,
"learning_rate": 9.368158403090792e-06,
"loss": 0.4064,
"step": 19500
},
{
"epoch": 1.61,
"learning_rate": 9.323442083422767e-06,
"loss": 0.4004,
"step": 20000
},
{
"epoch": 1.61,
"eval_bertscore/f1": 0.8022,
"eval_bertscore/precision": 0.7971,
"eval_bertscore/recall": 0.8093,
"eval_mean_prediction_length_characters": 860.077,
"eval_mean_prediction_length_tokens": 188.397,
"eval_num_predicted": 1000,
"eval_rouge/geometric_mean": 57.2862,
"eval_rouge/rouge1": 71.0513,
"eval_rouge/rouge2": 52.0715,
"eval_rouge/rougeL": 50.8134,
"eval_rouge/rougeLsum": 69.0639,
"eval_runtime": 2596.2685,
"eval_samples_per_second": 0.385,
"eval_steps_per_second": 0.385,
"step": 20000
},
{
"epoch": 1.65,
"learning_rate": 9.278725763754741e-06,
"loss": 0.3953,
"step": 20500
},
{
"epoch": 1.69,
"learning_rate": 9.234009444086714e-06,
"loss": 0.3905,
"step": 21000
},
{
"epoch": 1.69,
"eval_bertscore/f1": 0.8009,
"eval_bertscore/precision": 0.792,
"eval_bertscore/recall": 0.812,
"eval_mean_prediction_length_characters": 922.278,
"eval_mean_prediction_length_tokens": 199.619,
"eval_num_predicted": 1000,
"eval_rouge/geometric_mean": 56.8514,
"eval_rouge/rouge1": 70.8436,
"eval_rouge/rouge2": 51.8156,
"eval_rouge/rougeL": 50.0567,
"eval_rouge/rougeLsum": 68.8517,
"eval_runtime": 2741.6372,
"eval_samples_per_second": 0.365,
"eval_steps_per_second": 0.365,
"step": 21000
},
{
"epoch": 1.73,
"learning_rate": 9.189382557058024e-06,
"loss": 0.3834,
"step": 21500
},
{
"epoch": 1.77,
"learning_rate": 9.144666237389999e-06,
"loss": 0.3772,
"step": 22000
},
{
"epoch": 1.77,
"eval_bertscore/f1": 0.802,
"eval_bertscore/precision": 0.8014,
"eval_bertscore/recall": 0.8048,
"eval_mean_prediction_length_characters": 821.195,
"eval_mean_prediction_length_tokens": 179.279,
"eval_num_predicted": 1000,
"eval_rouge/geometric_mean": 57.3719,
"eval_rouge/rouge1": 70.4754,
"eval_rouge/rouge2": 52.2981,
"eval_rouge/rougeL": 51.2358,
"eval_rouge/rougeLsum": 68.5129,
"eval_runtime": 2484.708,
"eval_samples_per_second": 0.402,
"eval_steps_per_second": 0.402,
"step": 22000
},
{
"epoch": 1.81,
"learning_rate": 9.099949917721973e-06,
"loss": 0.3796,
"step": 22500
},
{
"epoch": 1.85,
"learning_rate": 9.055233598053946e-06,
"loss": 0.3745,
"step": 23000
},
{
"epoch": 1.85,
"eval_bertscore/f1": 0.8077,
"eval_bertscore/precision": 0.8096,
"eval_bertscore/recall": 0.8076,
"eval_mean_prediction_length_characters": 797.37,
"eval_mean_prediction_length_tokens": 176.263,
"eval_num_predicted": 1000,
"eval_rouge/geometric_mean": 58.2667,
"eval_rouge/rouge1": 71.7377,
"eval_rouge/rouge2": 53.2732,
"eval_rouge/rougeL": 51.7614,
"eval_rouge/rougeLsum": 69.9972,
"eval_runtime": 2452.4348,
"eval_samples_per_second": 0.408,
"eval_steps_per_second": 0.408,
"step": 23000
},
{
"epoch": 1.89,
"learning_rate": 9.01051727838592e-06,
"loss": 0.3705,
"step": 23500
},
{
"epoch": 1.93,
"learning_rate": 8.96589039135723e-06,
"loss": 0.3709,
"step": 24000
},
{
"epoch": 1.93,
"eval_bertscore/f1": 0.8108,
"eval_bertscore/precision": 0.8141,
"eval_bertscore/recall": 0.8094,
"eval_mean_prediction_length_characters": 798.81,
"eval_mean_prediction_length_tokens": 176.217,
"eval_num_predicted": 1000,
"eval_rouge/geometric_mean": 58.6206,
"eval_rouge/rouge1": 72.0193,
"eval_rouge/rouge2": 53.6884,
"eval_rouge/rougeL": 52.098,
"eval_rouge/rougeLsum": 70.3419,
"eval_runtime": 2449.0615,
"eval_samples_per_second": 0.408,
"eval_steps_per_second": 0.408,
"step": 24000
},
{
"epoch": 1.97,
"learning_rate": 8.92126350432854e-06,
"loss": 0.3638,
"step": 24500
},
{
"epoch": 2.01,
"learning_rate": 8.876547184660515e-06,
"loss": 0.3461,
"step": 25000
},
{
"epoch": 2.01,
"eval_bertscore/f1": 0.8092,
"eval_bertscore/precision": 0.8022,
"eval_bertscore/recall": 0.8182,
"eval_mean_prediction_length_characters": 891.59,
"eval_mean_prediction_length_tokens": 193.579,
"eval_num_predicted": 1000,
"eval_rouge/geometric_mean": 58.5159,
"eval_rouge/rouge1": 72.4344,
"eval_rouge/rouge2": 54.0016,
"eval_rouge/rougeL": 51.2235,
"eval_rouge/rougeLsum": 70.5176,
"eval_runtime": 2696.2641,
"eval_samples_per_second": 0.371,
"eval_steps_per_second": 0.371,
"step": 25000
},
{
"epoch": 2.05,
"learning_rate": 8.83183086499249e-06,
"loss": 0.2926,
"step": 25500
},
{
"epoch": 2.09,
"learning_rate": 8.787114545324463e-06,
"loss": 0.2935,
"step": 26000
},
{
"epoch": 2.09,
"eval_bertscore/f1": 0.8099,
"eval_bertscore/precision": 0.806,
"eval_bertscore/recall": 0.8158,
"eval_mean_prediction_length_characters": 858.323,
"eval_mean_prediction_length_tokens": 186.494,
"eval_num_predicted": 1000,
"eval_rouge/geometric_mean": 59.0206,
"eval_rouge/rouge1": 72.3233,
"eval_rouge/rouge2": 54.4823,
"eval_rouge/rougeL": 52.1769,
"eval_rouge/rougeLsum": 70.5191,
"eval_runtime": 2616.4872,
"eval_samples_per_second": 0.382,
"eval_steps_per_second": 0.382,
"step": 26000
},
{
"epoch": 2.13,
"learning_rate": 8.742398225656437e-06,
"loss": 0.2945,
"step": 26500
},
{
"epoch": 2.17,
"learning_rate": 8.69768190598841e-06,
"loss": 0.2859,
"step": 27000
},
{
"epoch": 2.17,
"eval_bertscore/f1": 0.8021,
"eval_bertscore/precision": 0.7967,
"eval_bertscore/recall": 0.8098,
"eval_mean_prediction_length_characters": 862.623,
"eval_mean_prediction_length_tokens": 187.108,
"eval_num_predicted": 1000,
"eval_rouge/geometric_mean": 58.0972,
"eval_rouge/rouge1": 71.3169,
"eval_rouge/rouge2": 53.5502,
"eval_rouge/rougeL": 51.3466,
"eval_rouge/rougeLsum": 69.328,
"eval_runtime": 2591.3634,
"eval_samples_per_second": 0.386,
"eval_steps_per_second": 0.386,
"step": 27000
},
{
"epoch": 2.21,
"learning_rate": 8.65305501895972e-06,
"loss": 0.2913,
"step": 27500
},
{
"epoch": 2.25,
"learning_rate": 8.608338699291695e-06,
"loss": 0.2871,
"step": 28000
},
{
"epoch": 2.25,
"eval_bertscore/f1": 0.8125,
"eval_bertscore/precision": 0.81,
"eval_bertscore/recall": 0.8167,
"eval_mean_prediction_length_characters": 839.348,
"eval_mean_prediction_length_tokens": 182.08,
"eval_num_predicted": 1000,
"eval_rouge/geometric_mean": 59.5943,
"eval_rouge/rouge1": 72.7757,
"eval_rouge/rouge2": 55.0591,
"eval_rouge/rougeL": 52.82,
"eval_rouge/rougeLsum": 70.9482,
"eval_runtime": 2411.5587,
"eval_samples_per_second": 0.415,
"eval_steps_per_second": 0.415,
"step": 28000
},
{
"epoch": 2.29,
"learning_rate": 8.563622379623669e-06,
"loss": 0.2858,
"step": 28500
},
{
"epoch": 2.33,
"learning_rate": 8.518906059955642e-06,
"loss": 0.2849,
"step": 29000
},
{
"epoch": 2.33,
"eval_bertscore/f1": 0.8151,
"eval_bertscore/precision": 0.8177,
"eval_bertscore/recall": 0.8143,
"eval_mean_prediction_length_characters": 801.734,
"eval_mean_prediction_length_tokens": 175.958,
"eval_num_predicted": 1000,
"eval_rouge/geometric_mean": 60.3823,
"eval_rouge/rouge1": 73.1311,
"eval_rouge/rouge2": 56.0185,
"eval_rouge/rougeL": 53.7397,
"eval_rouge/rougeLsum": 71.4246,
"eval_runtime": 2330.5152,
"eval_samples_per_second": 0.429,
"eval_steps_per_second": 0.429,
"step": 29000
},
{
"epoch": 2.37,
"learning_rate": 8.474189740287617e-06,
"loss": 0.2787,
"step": 29500
},
{
"epoch": 2.41,
"learning_rate": 8.429473420619591e-06,
"loss": 0.2819,
"step": 30000
},
{
"epoch": 2.41,
"eval_bertscore/f1": 0.8142,
"eval_bertscore/precision": 0.8077,
"eval_bertscore/recall": 0.8228,
"eval_mean_prediction_length_characters": 894.063,
"eval_mean_prediction_length_tokens": 192.74,
"eval_num_predicted": 1000,
"eval_rouge/geometric_mean": 60.1592,
"eval_rouge/rouge1": 73.1791,
"eval_rouge/rouge2": 55.8507,
"eval_rouge/rougeL": 53.2711,
"eval_rouge/rougeLsum": 71.4419,
"eval_runtime": 2537.8153,
"eval_samples_per_second": 0.394,
"eval_steps_per_second": 0.394,
"step": 30000
},
{
"epoch": 2.45,
"learning_rate": 8.384757100951564e-06,
"loss": 0.2799,
"step": 30500
},
{
"epoch": 2.5,
"learning_rate": 8.340040781283539e-06,
"loss": 0.2776,
"step": 31000
},
{
"epoch": 2.5,
"eval_bertscore/f1": 0.8054,
"eval_bertscore/precision": 0.7925,
"eval_bertscore/recall": 0.821,
"eval_mean_prediction_length_characters": 967.121,
"eval_mean_prediction_length_tokens": 206.944,
"eval_num_predicted": 1000,
"eval_rouge/geometric_mean": 58.7521,
"eval_rouge/rouge1": 71.9074,
"eval_rouge/rouge2": 54.5099,
"eval_rouge/rougeL": 51.7393,
"eval_rouge/rougeLsum": 69.9728,
"eval_runtime": 2713.6577,
"eval_samples_per_second": 0.369,
"eval_steps_per_second": 0.369,
"step": 31000
},
{
"epoch": 2.54,
"learning_rate": 8.295413894254849e-06,
"loss": 0.2761,
"step": 31500
},
{
"epoch": 2.58,
"learning_rate": 8.250697574586823e-06,
"loss": 0.2742,
"step": 32000
},
{
"epoch": 2.58,
"eval_bertscore/f1": 0.8187,
"eval_bertscore/precision": 0.8125,
"eval_bertscore/recall": 0.827,
"eval_mean_prediction_length_characters": 880.74,
"eval_mean_prediction_length_tokens": 191.221,
"eval_num_predicted": 1000,
"eval_rouge/geometric_mean": 60.9358,
"eval_rouge/rouge1": 73.8216,
"eval_rouge/rouge2": 56.8015,
"eval_rouge/rougeL": 53.9603,
"eval_rouge/rougeLsum": 72.1193,
"eval_runtime": 2491.9164,
"eval_samples_per_second": 0.401,
"eval_steps_per_second": 0.401,
"step": 32000
},
{
"epoch": 2.62,
"learning_rate": 8.205981254918796e-06,
"loss": 0.2737,
"step": 32500
},
{
"epoch": 2.66,
"learning_rate": 8.16126493525077e-06,
"loss": 0.2663,
"step": 33000
},
{
"epoch": 2.66,
"eval_bertscore/f1": 0.8215,
"eval_bertscore/precision": 0.8174,
"eval_bertscore/recall": 0.8272,
"eval_mean_prediction_length_characters": 862.393,
"eval_mean_prediction_length_tokens": 187.552,
"eval_num_predicted": 1000,
"eval_rouge/geometric_mean": 61.7551,
"eval_rouge/rouge1": 74.5811,
"eval_rouge/rouge2": 57.6565,
"eval_rouge/rougeL": 54.7699,
"eval_rouge/rougeLsum": 72.9841,
"eval_runtime": 2430.2491,
"eval_samples_per_second": 0.411,
"eval_steps_per_second": 0.411,
"step": 33000
},
{
"epoch": 2.7,
"learning_rate": 8.116548615582743e-06,
"loss": 0.2636,
"step": 33500
},
{
"epoch": 2.74,
"learning_rate": 8.071921728554053e-06,
"loss": 0.2672,
"step": 34000
},
{
"epoch": 2.74,
"eval_bertscore/f1": 0.8194,
"eval_bertscore/precision": 0.8133,
"eval_bertscore/recall": 0.8274,
"eval_mean_prediction_length_characters": 893.688,
"eval_mean_prediction_length_tokens": 193.593,
"eval_num_predicted": 1000,
"eval_rouge/geometric_mean": 61.3651,
"eval_rouge/rouge1": 74.2269,
"eval_rouge/rouge2": 57.4497,
"eval_rouge/rougeL": 54.1897,
"eval_rouge/rougeLsum": 72.5834,
"eval_runtime": 2499.9317,
"eval_samples_per_second": 0.4,
"eval_steps_per_second": 0.4,
"step": 34000
},
{
"epoch": 1.39,
"learning_rate": 9.569096202090125e-06,
"loss": 0.2366,
"step": 34500
},
{
"epoch": 1.41,
"learning_rate": 9.54673714242274e-06,
"loss": 0.2405,
"step": 35000
},
{
"epoch": 1.41,
"eval_bertscore/f1": 0.8145,
"eval_bertscore/precision": 0.8054,
"eval_bertscore/recall": 0.8256,
"eval_mean_prediction_length_characters": 927.422,
"eval_mean_prediction_length_tokens": 200.478,
"eval_num_predicted": 1000,
"eval_rouge/geometric_mean": 60.0655,
"eval_rouge/rouge1": 73.2741,
"eval_rouge/rouge2": 55.8854,
"eval_rouge/rougeL": 52.9208,
"eval_rouge/rougeLsum": 71.6192,
"eval_runtime": 2620.9045,
"eval_samples_per_second": 0.382,
"eval_steps_per_second": 0.382,
"step": 35000
},
{
"epoch": 1.43,
"learning_rate": 9.524378082755352e-06,
"loss": 0.2489,
"step": 35500
},
{
"epoch": 1.45,
"learning_rate": 9.502019023087967e-06,
"loss": 0.2428,
"step": 36000
},
{
"epoch": 1.45,
"eval_bertscore/f1": 0.8171,
"eval_bertscore/precision": 0.8136,
"eval_bertscore/recall": 0.8226,
"eval_mean_prediction_length_characters": 847.398,
"eval_mean_prediction_length_tokens": 183.753,
"eval_num_predicted": 1000,
"eval_rouge/geometric_mean": 60.8081,
"eval_rouge/rouge1": 73.4983,
"eval_rouge/rouge2": 56.5892,
"eval_rouge/rougeL": 54.0596,
"eval_rouge/rougeLsum": 71.9013,
"eval_runtime": 2373.5033,
"eval_samples_per_second": 0.421,
"eval_steps_per_second": 0.421,
"step": 36000
},
{
"epoch": 1.47,
"learning_rate": 9.47965996342058e-06,
"loss": 0.2511,
"step": 36500
},
{
"epoch": 1.49,
"learning_rate": 9.457300903753192e-06,
"loss": 0.2416,
"step": 37000
},
{
"epoch": 1.49,
"eval_bertscore/f1": 0.8155,
"eval_bertscore/precision": 0.8154,
"eval_bertscore/recall": 0.8179,
"eval_mean_prediction_length_characters": 850.697,
"eval_mean_prediction_length_tokens": 185.038,
"eval_num_predicted": 1000,
"eval_rouge/geometric_mean": 60.5988,
"eval_rouge/rouge1": 72.9601,
"eval_rouge/rouge2": 56.3526,
"eval_rouge/rougeL": 54.1244,
"eval_rouge/rougeLsum": 71.3786,
"eval_runtime": 2440.9193,
"eval_samples_per_second": 0.41,
"eval_steps_per_second": 0.41,
"step": 37000
},
{
"epoch": 1.51,
"learning_rate": 9.43498656220514e-06,
"loss": 0.2404,
"step": 37500
},
{
"epoch": 1.53,
"learning_rate": 9.412627502537755e-06,
"loss": 0.2597,
"step": 38000
},
{
"epoch": 1.53,
"eval_bertscore/f1": 0.8146,
"eval_bertscore/precision": 0.8118,
"eval_bertscore/recall": 0.8194,
"eval_mean_prediction_length_characters": 856.43,
"eval_mean_prediction_length_tokens": 186.628,
"eval_num_predicted": 1000,
"eval_rouge/geometric_mean": 60.5122,
"eval_rouge/rouge1": 73.1873,
"eval_rouge/rouge2": 56.3588,
"eval_rouge/rougeL": 53.7195,
"eval_rouge/rougeLsum": 71.5458,
"eval_runtime": 2434.8964,
"eval_samples_per_second": 0.411,
"eval_steps_per_second": 0.411,
"step": 38000
},
{
"epoch": 1.55,
"learning_rate": 9.390268442870366e-06,
"loss": 0.2436,
"step": 38500
},
{
"epoch": 1.57,
"learning_rate": 9.367954101322315e-06,
"loss": 0.2526,
"step": 39000
},
{
"epoch": 1.57,
"eval_bertscore/f1": 0.8206,
"eval_bertscore/precision": 0.8186,
"eval_bertscore/recall": 0.8245,
"eval_mean_prediction_length_characters": 843.759,
"eval_mean_prediction_length_tokens": 183.492,
"eval_num_predicted": 1000,
"eval_rouge/geometric_mean": 61.4692,
"eval_rouge/rouge1": 74.1108,
"eval_rouge/rouge2": 57.3353,
"eval_rouge/rougeL": 54.6598,
"eval_rouge/rougeLsum": 72.5858,
"eval_runtime": 2441.5715,
"eval_samples_per_second": 0.41,
"eval_steps_per_second": 0.41,
"step": 39000
},
{
"epoch": 1.59,
"learning_rate": 9.34559504165493e-06,
"loss": 0.248,
"step": 39500
},
{
"epoch": 1.61,
"learning_rate": 9.323235981987542e-06,
"loss": 0.2431,
"step": 40000
},
{
"epoch": 1.61,
"eval_bertscore/f1": 0.8214,
"eval_bertscore/precision": 0.8186,
"eval_bertscore/recall": 0.826,
"eval_mean_prediction_length_characters": 854.807,
"eval_mean_prediction_length_tokens": 186.06,
"eval_num_predicted": 1000,
"eval_rouge/geometric_mean": 61.6121,
"eval_rouge/rouge1": 74.3063,
"eval_rouge/rouge2": 57.6124,
"eval_rouge/rougeL": 54.6331,
"eval_rouge/rougeLsum": 72.7609,
"eval_runtime": 2483.1491,
"eval_samples_per_second": 0.403,
"eval_steps_per_second": 0.403,
"step": 40000
},
{
"epoch": 1.63,
"learning_rate": 9.300876922320155e-06,
"loss": 0.2413,
"step": 40500
},
{
"epoch": 1.65,
"learning_rate": 9.27851786265277e-06,
"loss": 0.2556,
"step": 41000
},
{
"epoch": 1.65,
"eval_bertscore/f1": 0.8212,
"eval_bertscore/precision": 0.8166,
"eval_bertscore/recall": 0.8277,
"eval_mean_prediction_length_characters": 871.94,
"eval_mean_prediction_length_tokens": 189.412,
"eval_num_predicted": 1000,
"eval_rouge/geometric_mean": 61.6287,
"eval_rouge/rouge1": 74.4291,
"eval_rouge/rouge2": 57.5628,
"eval_rouge/rougeL": 54.6343,
"eval_rouge/rougeLsum": 72.7972,
"eval_runtime": 2504.8302,
"eval_samples_per_second": 0.399,
"eval_steps_per_second": 0.399,
"step": 41000
},
{
"epoch": 1.67,
"learning_rate": 9.256158802985382e-06,
"loss": 0.2414,
"step": 41500
},
{
"epoch": 1.69,
"learning_rate": 9.233799743317997e-06,
"loss": 0.2493,
"step": 42000
},
{
"epoch": 1.69,
"eval_bertscore/f1": 0.8202,
"eval_bertscore/precision": 0.8115,
"eval_bertscore/recall": 0.831,
"eval_mean_prediction_length_characters": 919.152,
"eval_mean_prediction_length_tokens": 199.285,
"eval_num_predicted": 1000,
"eval_rouge/geometric_mean": 61.6474,
"eval_rouge/rouge1": 74.2897,
"eval_rouge/rouge2": 57.7571,
"eval_rouge/rougeL": 54.6022,
"eval_rouge/rougeLsum": 72.7195,
"eval_runtime": 2638.78,
"eval_samples_per_second": 0.379,
"eval_steps_per_second": 0.379,
"step": 42000
},
{
"epoch": 1.71,
"learning_rate": 9.21144068365061e-06,
"loss": 0.2467,
"step": 42500
},
{
"epoch": 1.73,
"learning_rate": 9.189081623983222e-06,
"loss": 0.2421,
"step": 43000
},
{
"epoch": 1.73,
"eval_bertscore/f1": 0.823,
"eval_bertscore/precision": 0.8258,
"eval_bertscore/recall": 0.8222,
"eval_mean_prediction_length_characters": 803.212,
"eval_mean_prediction_length_tokens": 176.172,
"eval_num_predicted": 1000,
"eval_rouge/geometric_mean": 61.9161,
"eval_rouge/rouge1": 74.3746,
"eval_rouge/rouge2": 57.9399,
"eval_rouge/rougeL": 55.0818,
"eval_rouge/rougeLsum": 72.8967,
"eval_runtime": 2327.757,
"eval_samples_per_second": 0.43,
"eval_steps_per_second": 0.43,
"step": 43000
},
{
"epoch": 1.75,
"learning_rate": 9.166722564315837e-06,
"loss": 0.2478,
"step": 43500
},
{
"epoch": 1.77,
"learning_rate": 9.14436350464845e-06,
"loss": 0.229,
"step": 44000
},
{
"epoch": 1.77,
"eval_bertscore/f1": 0.8212,
"eval_bertscore/precision": 0.822,
"eval_bertscore/recall": 0.8227,
"eval_mean_prediction_length_characters": 819.677,
"eval_mean_prediction_length_tokens": 178.389,
"eval_num_predicted": 1000,
"eval_rouge/geometric_mean": 61.8804,
"eval_rouge/rouge1": 74.0116,
"eval_rouge/rouge2": 57.9086,
"eval_rouge/rougeL": 55.2862,
"eval_rouge/rougeLsum": 72.466,
"eval_runtime": 2366.2663,
"eval_samples_per_second": 0.423,
"eval_steps_per_second": 0.423,
"step": 44000
},
{
"epoch": 1.79,
"learning_rate": 9.122049163100398e-06,
"loss": 0.2398,
"step": 44500
},
{
"epoch": 1.81,
"learning_rate": 9.09969010343301e-06,
"loss": 0.2429,
"step": 45000
},
{
"epoch": 1.81,
"eval_bertscore/f1": 0.8243,
"eval_bertscore/precision": 0.8174,
"eval_bertscore/recall": 0.8333,
"eval_mean_prediction_length_characters": 896.966,
"eval_mean_prediction_length_tokens": 195.349,
"eval_num_predicted": 1000,
"eval_rouge/geometric_mean": 62.2879,
"eval_rouge/rouge1": 74.8367,
"eval_rouge/rouge2": 58.4326,
"eval_rouge/rougeL": 55.2638,
"eval_rouge/rougeLsum": 73.3944,
"eval_runtime": 2574.2161,
"eval_samples_per_second": 0.388,
"eval_steps_per_second": 0.388,
"step": 45000
},
{
"epoch": 1.83,
"learning_rate": 9.077331043765623e-06,
"loss": 0.2355,
"step": 45500
},
{
"epoch": 1.85,
"learning_rate": 9.054971984098238e-06,
"loss": 0.2375,
"step": 46000
},
{
"epoch": 1.85,
"eval_bertscore/f1": 0.824,
"eval_bertscore/precision": 0.82,
"eval_bertscore/recall": 0.8303,
"eval_mean_prediction_length_characters": 873.321,
"eval_mean_prediction_length_tokens": 190.408,
"eval_num_predicted": 1000,
"eval_rouge/geometric_mean": 62.1689,
"eval_rouge/rouge1": 74.7062,
"eval_rouge/rouge2": 58.2978,
"eval_rouge/rougeL": 55.1711,
"eval_rouge/rougeLsum": 73.2347,
"eval_runtime": 2618.1437,
"eval_samples_per_second": 0.382,
"eval_steps_per_second": 0.382,
"step": 46000
},
{
"epoch": 1.87,
"learning_rate": 9.032657642550186e-06,
"loss": 0.2425,
"step": 46500
},
{
"epoch": 1.89,
"learning_rate": 9.010298582882799e-06,
"loss": 0.228,
"step": 47000
},
{
"epoch": 1.89,
"eval_bertscore/f1": 0.8258,
"eval_bertscore/precision": 0.8189,
"eval_bertscore/recall": 0.8348,
"eval_mean_prediction_length_characters": 904.242,
"eval_mean_prediction_length_tokens": 196.813,
"eval_num_predicted": 1000,
"eval_rouge/geometric_mean": 62.6657,
"eval_rouge/rouge1": 75.0484,
"eval_rouge/rouge2": 59.0138,
"eval_rouge/rougeL": 55.5642,
"eval_rouge/rougeLsum": 73.6606,
"eval_runtime": 2709.9214,
"eval_samples_per_second": 0.369,
"eval_steps_per_second": 0.369,
"step": 47000
},
{
"epoch": 1.91,
"learning_rate": 8.987939523215412e-06,
"loss": 0.2388,
"step": 47500
},
{
"epoch": 1.93,
"learning_rate": 8.965580463548026e-06,
"loss": 0.2398,
"step": 48000
},
{
"epoch": 1.93,
"eval_bertscore/f1": 0.8287,
"eval_bertscore/precision": 0.8255,
"eval_bertscore/recall": 0.8338,
"eval_mean_prediction_length_characters": 861.415,
"eval_mean_prediction_length_tokens": 187.846,
"eval_num_predicted": 1000,
"eval_rouge/geometric_mean": 63.1011,
"eval_rouge/rouge1": 75.5807,
"eval_rouge/rouge2": 59.457,
"eval_rouge/rougeL": 55.9108,
"eval_rouge/rougeLsum": 74.1699,
"eval_runtime": 2601.0783,
"eval_samples_per_second": 0.384,
"eval_steps_per_second": 0.384,
"step": 48000
},
{
"epoch": 1.95,
"learning_rate": 8.943221403880639e-06,
"loss": 0.2399,
"step": 48500
},
{
"epoch": 1.97,
"learning_rate": 8.920862344213252e-06,
"loss": 0.2373,
"step": 49000
},
{
"epoch": 1.97,
"eval_bertscore/f1": 0.8253,
"eval_bertscore/precision": 0.8244,
"eval_bertscore/recall": 0.828,
"eval_mean_prediction_length_characters": 840.263,
"eval_mean_prediction_length_tokens": 182.473,
"eval_num_predicted": 1000,
"eval_rouge/geometric_mean": 62.7461,
"eval_rouge/rouge1": 74.8987,
"eval_rouge/rouge2": 59.0142,
"eval_rouge/rougeL": 55.8894,
"eval_rouge/rougeLsum": 73.4189,
"eval_runtime": 2549.6944,
"eval_samples_per_second": 0.392,
"eval_steps_per_second": 0.392,
"step": 49000
},
{
"epoch": 1.99,
"learning_rate": 8.8985480026652e-06,
"loss": 0.2397,
"step": 49500
},
{
"epoch": 2.01,
"learning_rate": 8.876233661117149e-06,
"loss": 0.2057,
"step": 50000
},
{
"epoch": 2.01,
"eval_bertscore/f1": 0.8258,
"eval_bertscore/precision": 0.8222,
"eval_bertscore/recall": 0.8313,
"eval_mean_prediction_length_characters": 857.49,
"eval_mean_prediction_length_tokens": 186.547,
"eval_num_predicted": 1000,
"eval_rouge/geometric_mean": 62.8166,
"eval_rouge/rouge1": 75.1042,
"eval_rouge/rouge2": 59.0815,
"eval_rouge/rougeL": 55.8608,
"eval_rouge/rougeLsum": 73.5713,
"eval_runtime": 2602.9106,
"eval_samples_per_second": 0.384,
"eval_steps_per_second": 0.384,
"step": 50000
}
],
"max_steps": 248470,
"num_train_epochs": 10,
"total_flos": 6.239894115827712e+16,
"trial_name": null,
"trial_params": null
}