pythia-125M-Summarization-sft / trainer_state.json
dmayhem93's picture
Upload with huggingface_hub
dd98158
raw
history blame contribute delete
No virus
15.6 kB
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 1.0,
"global_step": 14299,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.01,
"learning_rate": 9.488135456452206e-05,
"loss": 3.4925,
"step": 100
},
{
"epoch": 0.01,
"learning_rate": 0.0001,
"loss": 1.9277,
"step": 200
},
{
"epoch": 0.02,
"learning_rate": 0.0001,
"loss": 1.9662,
"step": 300
},
{
"epoch": 0.03,
"learning_rate": 0.0001,
"loss": 1.9422,
"step": 400
},
{
"epoch": 0.03,
"learning_rate": 0.0001,
"loss": 1.9676,
"step": 500
},
{
"epoch": 0.04,
"learning_rate": 0.0001,
"loss": 1.9137,
"step": 600
},
{
"epoch": 0.05,
"learning_rate": 0.0001,
"loss": 1.9491,
"step": 700
},
{
"epoch": 0.06,
"learning_rate": 0.0001,
"loss": 1.9413,
"step": 800
},
{
"epoch": 0.06,
"learning_rate": 0.0001,
"loss": 1.9424,
"step": 900
},
{
"epoch": 0.07,
"learning_rate": 0.0001,
"loss": 1.916,
"step": 1000
},
{
"epoch": 0.08,
"learning_rate": 0.0001,
"loss": 1.8919,
"step": 1100
},
{
"epoch": 0.08,
"learning_rate": 0.0001,
"loss": 1.9228,
"step": 1200
},
{
"epoch": 0.09,
"learning_rate": 0.0001,
"loss": 1.9008,
"step": 1300
},
{
"epoch": 0.1,
"learning_rate": 0.0001,
"loss": 1.9189,
"step": 1400
},
{
"epoch": 0.1,
"learning_rate": 0.0001,
"loss": 1.8898,
"step": 1500
},
{
"epoch": 0.11,
"learning_rate": 0.0001,
"loss": 1.8724,
"step": 1600
},
{
"epoch": 0.12,
"learning_rate": 0.0001,
"loss": 1.9123,
"step": 1700
},
{
"epoch": 0.13,
"learning_rate": 0.0001,
"loss": 1.934,
"step": 1800
},
{
"epoch": 0.13,
"learning_rate": 0.0001,
"loss": 1.904,
"step": 1900
},
{
"epoch": 0.14,
"learning_rate": 0.0001,
"loss": 1.9295,
"step": 2000
},
{
"epoch": 0.15,
"learning_rate": 0.0001,
"loss": 1.926,
"step": 2100
},
{
"epoch": 0.15,
"learning_rate": 0.0001,
"loss": 1.8874,
"step": 2200
},
{
"epoch": 0.16,
"learning_rate": 0.0001,
"loss": 1.9097,
"step": 2300
},
{
"epoch": 0.17,
"learning_rate": 0.0001,
"loss": 1.9119,
"step": 2400
},
{
"epoch": 0.17,
"learning_rate": 0.0001,
"loss": 1.8833,
"step": 2500
},
{
"epoch": 0.18,
"learning_rate": 0.0001,
"loss": 1.894,
"step": 2600
},
{
"epoch": 0.19,
"learning_rate": 0.0001,
"loss": 1.9345,
"step": 2700
},
{
"epoch": 0.2,
"learning_rate": 0.0001,
"loss": 1.8902,
"step": 2800
},
{
"epoch": 0.2,
"learning_rate": 0.0001,
"loss": 1.8798,
"step": 2900
},
{
"epoch": 0.21,
"learning_rate": 0.0001,
"loss": 1.8891,
"step": 3000
},
{
"epoch": 0.22,
"learning_rate": 0.0001,
"loss": 1.8853,
"step": 3100
},
{
"epoch": 0.22,
"learning_rate": 0.0001,
"loss": 1.865,
"step": 3200
},
{
"epoch": 0.23,
"learning_rate": 0.0001,
"loss": 1.9037,
"step": 3300
},
{
"epoch": 0.24,
"learning_rate": 0.0001,
"loss": 1.8672,
"step": 3400
},
{
"epoch": 0.24,
"learning_rate": 0.0001,
"loss": 1.868,
"step": 3500
},
{
"epoch": 0.25,
"learning_rate": 0.0001,
"loss": 1.8933,
"step": 3600
},
{
"epoch": 0.26,
"learning_rate": 0.0001,
"loss": 1.8709,
"step": 3700
},
{
"epoch": 0.27,
"learning_rate": 0.0001,
"loss": 1.8901,
"step": 3800
},
{
"epoch": 0.27,
"learning_rate": 0.0001,
"loss": 1.87,
"step": 3900
},
{
"epoch": 0.28,
"learning_rate": 0.0001,
"loss": 1.9241,
"step": 4000
},
{
"epoch": 0.29,
"learning_rate": 0.0001,
"loss": 1.8981,
"step": 4100
},
{
"epoch": 0.29,
"learning_rate": 0.0001,
"loss": 1.8564,
"step": 4200
},
{
"epoch": 0.3,
"learning_rate": 0.0001,
"loss": 1.8906,
"step": 4300
},
{
"epoch": 0.31,
"learning_rate": 0.0001,
"loss": 1.8751,
"step": 4400
},
{
"epoch": 0.31,
"learning_rate": 0.0001,
"loss": 1.8649,
"step": 4500
},
{
"epoch": 0.32,
"learning_rate": 0.0001,
"loss": 1.8901,
"step": 4600
},
{
"epoch": 0.33,
"learning_rate": 0.0001,
"loss": 1.8843,
"step": 4700
},
{
"epoch": 0.34,
"learning_rate": 0.0001,
"loss": 1.876,
"step": 4800
},
{
"epoch": 0.34,
"learning_rate": 0.0001,
"loss": 1.8486,
"step": 4900
},
{
"epoch": 0.35,
"learning_rate": 0.0001,
"loss": 1.8679,
"step": 5000
},
{
"epoch": 0.36,
"learning_rate": 0.0001,
"loss": 1.8295,
"step": 5100
},
{
"epoch": 0.36,
"learning_rate": 0.0001,
"loss": 1.8675,
"step": 5200
},
{
"epoch": 0.37,
"learning_rate": 0.0001,
"loss": 1.8501,
"step": 5300
},
{
"epoch": 0.38,
"learning_rate": 0.0001,
"loss": 1.861,
"step": 5400
},
{
"epoch": 0.38,
"learning_rate": 0.0001,
"loss": 1.8825,
"step": 5500
},
{
"epoch": 0.39,
"learning_rate": 0.0001,
"loss": 1.8431,
"step": 5600
},
{
"epoch": 0.4,
"learning_rate": 0.0001,
"loss": 1.8278,
"step": 5700
},
{
"epoch": 0.41,
"learning_rate": 0.0001,
"loss": 1.8671,
"step": 5800
},
{
"epoch": 0.41,
"learning_rate": 0.0001,
"loss": 1.8763,
"step": 5900
},
{
"epoch": 0.42,
"learning_rate": 0.0001,
"loss": 1.8645,
"step": 6000
},
{
"epoch": 0.43,
"learning_rate": 0.0001,
"loss": 1.8367,
"step": 6100
},
{
"epoch": 0.43,
"learning_rate": 0.0001,
"loss": 1.8838,
"step": 6200
},
{
"epoch": 0.44,
"learning_rate": 0.0001,
"loss": 1.8699,
"step": 6300
},
{
"epoch": 0.45,
"learning_rate": 0.0001,
"loss": 1.87,
"step": 6400
},
{
"epoch": 0.45,
"learning_rate": 0.0001,
"loss": 1.843,
"step": 6500
},
{
"epoch": 0.46,
"learning_rate": 0.0001,
"loss": 1.8441,
"step": 6600
},
{
"epoch": 0.47,
"learning_rate": 0.0001,
"loss": 1.8656,
"step": 6700
},
{
"epoch": 0.48,
"learning_rate": 0.0001,
"loss": 1.8527,
"step": 6800
},
{
"epoch": 0.48,
"learning_rate": 0.0001,
"loss": 1.8294,
"step": 6900
},
{
"epoch": 0.49,
"learning_rate": 0.0001,
"loss": 1.871,
"step": 7000
},
{
"epoch": 0.5,
"learning_rate": 0.0001,
"loss": 1.8634,
"step": 7100
},
{
"epoch": 0.5,
"learning_rate": 0.0001,
"loss": 1.8614,
"step": 7200
},
{
"epoch": 0.51,
"learning_rate": 0.0001,
"loss": 1.8229,
"step": 7300
},
{
"epoch": 0.52,
"learning_rate": 0.0001,
"loss": 1.8955,
"step": 7400
},
{
"epoch": 0.52,
"learning_rate": 0.0001,
"loss": 1.8379,
"step": 7500
},
{
"epoch": 0.53,
"learning_rate": 0.0001,
"loss": 1.8533,
"step": 7600
},
{
"epoch": 0.54,
"learning_rate": 0.0001,
"loss": 1.8454,
"step": 7700
},
{
"epoch": 0.55,
"learning_rate": 0.0001,
"loss": 1.8168,
"step": 7800
},
{
"epoch": 0.55,
"learning_rate": 0.0001,
"loss": 1.8283,
"step": 7900
},
{
"epoch": 0.56,
"learning_rate": 0.0001,
"loss": 1.8545,
"step": 8000
},
{
"epoch": 0.57,
"learning_rate": 0.0001,
"loss": 1.8133,
"step": 8100
},
{
"epoch": 0.57,
"learning_rate": 0.0001,
"loss": 1.8389,
"step": 8200
},
{
"epoch": 0.58,
"learning_rate": 0.0001,
"loss": 1.8246,
"step": 8300
},
{
"epoch": 0.59,
"learning_rate": 0.0001,
"loss": 1.8351,
"step": 8400
},
{
"epoch": 0.59,
"learning_rate": 0.0001,
"loss": 1.8244,
"step": 8500
},
{
"epoch": 0.6,
"learning_rate": 0.0001,
"loss": 1.8137,
"step": 8600
},
{
"epoch": 0.61,
"learning_rate": 0.0001,
"loss": 1.8016,
"step": 8700
},
{
"epoch": 0.62,
"learning_rate": 0.0001,
"loss": 1.8237,
"step": 8800
},
{
"epoch": 0.62,
"learning_rate": 0.0001,
"loss": 1.812,
"step": 8900
},
{
"epoch": 0.63,
"learning_rate": 0.0001,
"loss": 1.8345,
"step": 9000
},
{
"epoch": 0.64,
"learning_rate": 0.0001,
"loss": 1.8411,
"step": 9100
},
{
"epoch": 0.64,
"learning_rate": 0.0001,
"loss": 1.8077,
"step": 9200
},
{
"epoch": 0.65,
"learning_rate": 0.0001,
"loss": 1.8581,
"step": 9300
},
{
"epoch": 0.66,
"learning_rate": 0.0001,
"loss": 1.8242,
"step": 9400
},
{
"epoch": 0.66,
"learning_rate": 0.0001,
"loss": 1.8166,
"step": 9500
},
{
"epoch": 0.67,
"learning_rate": 0.0001,
"loss": 1.8227,
"step": 9600
},
{
"epoch": 0.68,
"learning_rate": 0.0001,
"loss": 1.8011,
"step": 9700
},
{
"epoch": 0.69,
"learning_rate": 0.0001,
"loss": 1.8253,
"step": 9800
},
{
"epoch": 0.69,
"learning_rate": 0.0001,
"loss": 1.7924,
"step": 9900
},
{
"epoch": 0.7,
"learning_rate": 0.0001,
"loss": 1.8091,
"step": 10000
},
{
"epoch": 0.71,
"learning_rate": 0.0001,
"loss": 1.8178,
"step": 10100
},
{
"epoch": 0.71,
"learning_rate": 0.0001,
"loss": 1.811,
"step": 10200
},
{
"epoch": 0.72,
"learning_rate": 0.0001,
"loss": 1.8134,
"step": 10300
},
{
"epoch": 0.73,
"learning_rate": 0.0001,
"loss": 1.8311,
"step": 10400
},
{
"epoch": 0.73,
"learning_rate": 0.0001,
"loss": 1.8269,
"step": 10500
},
{
"epoch": 0.74,
"learning_rate": 0.0001,
"loss": 1.8286,
"step": 10600
},
{
"epoch": 0.75,
"learning_rate": 0.0001,
"loss": 1.805,
"step": 10700
},
{
"epoch": 0.76,
"learning_rate": 0.0001,
"loss": 1.7735,
"step": 10800
},
{
"epoch": 0.76,
"learning_rate": 0.0001,
"loss": 1.761,
"step": 10900
},
{
"epoch": 0.77,
"learning_rate": 0.0001,
"loss": 1.7869,
"step": 11000
},
{
"epoch": 0.78,
"learning_rate": 0.0001,
"loss": 1.8054,
"step": 11100
},
{
"epoch": 0.78,
"learning_rate": 0.0001,
"loss": 1.7958,
"step": 11200
},
{
"epoch": 0.79,
"learning_rate": 0.0001,
"loss": 1.8049,
"step": 11300
},
{
"epoch": 0.8,
"learning_rate": 0.0001,
"loss": 1.7943,
"step": 11400
},
{
"epoch": 0.8,
"learning_rate": 0.0001,
"loss": 1.7771,
"step": 11500
},
{
"epoch": 0.81,
"learning_rate": 0.0001,
"loss": 1.7981,
"step": 11600
},
{
"epoch": 0.82,
"learning_rate": 0.0001,
"loss": 1.8178,
"step": 11700
},
{
"epoch": 0.83,
"learning_rate": 0.0001,
"loss": 1.7732,
"step": 11800
},
{
"epoch": 0.83,
"learning_rate": 0.0001,
"loss": 1.7991,
"step": 11900
},
{
"epoch": 0.84,
"learning_rate": 0.0001,
"loss": 1.7949,
"step": 12000
},
{
"epoch": 0.85,
"learning_rate": 0.0001,
"loss": 1.7829,
"step": 12100
},
{
"epoch": 0.85,
"learning_rate": 0.0001,
"loss": 1.7816,
"step": 12200
},
{
"epoch": 0.86,
"learning_rate": 0.0001,
"loss": 1.7896,
"step": 12300
},
{
"epoch": 0.87,
"learning_rate": 0.0001,
"loss": 1.7797,
"step": 12400
},
{
"epoch": 0.87,
"learning_rate": 0.0001,
"loss": 1.7965,
"step": 12500
},
{
"epoch": 0.88,
"learning_rate": 0.0001,
"loss": 1.7855,
"step": 12600
},
{
"epoch": 0.89,
"learning_rate": 0.0001,
"loss": 1.7883,
"step": 12700
},
{
"epoch": 0.9,
"learning_rate": 0.0001,
"loss": 1.7606,
"step": 12800
},
{
"epoch": 0.9,
"learning_rate": 0.0001,
"loss": 1.7792,
"step": 12900
},
{
"epoch": 0.91,
"learning_rate": 0.0001,
"loss": 1.7773,
"step": 13000
},
{
"epoch": 0.92,
"learning_rate": 0.0001,
"loss": 1.7914,
"step": 13100
},
{
"epoch": 0.92,
"learning_rate": 0.0001,
"loss": 1.787,
"step": 13200
},
{
"epoch": 0.93,
"learning_rate": 0.0001,
"loss": 1.7892,
"step": 13300
},
{
"epoch": 0.94,
"learning_rate": 0.0001,
"loss": 1.7676,
"step": 13400
},
{
"epoch": 0.94,
"learning_rate": 0.0001,
"loss": 1.7775,
"step": 13500
},
{
"epoch": 0.95,
"learning_rate": 0.0001,
"loss": 1.7874,
"step": 13600
},
{
"epoch": 0.96,
"learning_rate": 0.0001,
"loss": 1.7517,
"step": 13700
},
{
"epoch": 0.97,
"learning_rate": 0.0001,
"loss": 1.7669,
"step": 13800
},
{
"epoch": 0.97,
"learning_rate": 0.0001,
"loss": 1.7647,
"step": 13900
},
{
"epoch": 0.98,
"learning_rate": 0.0001,
"loss": 1.7818,
"step": 14000
},
{
"epoch": 0.99,
"learning_rate": 0.0001,
"loss": 1.7826,
"step": 14100
},
{
"epoch": 0.99,
"learning_rate": 0.0001,
"loss": 1.7929,
"step": 14200
},
{
"epoch": 1.0,
"eval_loss": 1.7685546875,
"eval_runtime": 3.0679,
"eval_samples_per_second": 761.101,
"eval_steps_per_second": 95.178,
"step": 14299
}
],
"max_steps": 14299,
"num_train_epochs": 1,
"total_flos": 4.847476972453888e+16,
"trial_name": null,
"trial_params": null
}