model_ablation_no_length / trainer_state.json
nancyH's picture
Upload folder using huggingface_hub
377fd78 verified
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 27.19608376393799,
"eval_steps": 500,
"global_step": 100000,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.27,
"learning_rate": 4.000000000000001e-06,
"loss": 7.3758,
"step": 1000
},
{
"epoch": 0.54,
"learning_rate": 8.000000000000001e-06,
"loss": 6.6082,
"step": 2000
},
{
"epoch": 0.82,
"learning_rate": 1.2e-05,
"loss": 6.4181,
"step": 3000
},
{
"epoch": 1.09,
"learning_rate": 1.6000000000000003e-05,
"loss": 6.325,
"step": 4000
},
{
"epoch": 1.36,
"learning_rate": 2e-05,
"loss": 6.2866,
"step": 5000
},
{
"epoch": 1.63,
"learning_rate": 2.4e-05,
"loss": 6.2655,
"step": 6000
},
{
"epoch": 1.9,
"learning_rate": 2.8e-05,
"loss": 6.2543,
"step": 7000
},
{
"epoch": 2.18,
"learning_rate": 3.2000000000000005e-05,
"loss": 6.2456,
"step": 8000
},
{
"epoch": 2.45,
"learning_rate": 3.6e-05,
"loss": 6.2373,
"step": 9000
},
{
"epoch": 2.72,
"learning_rate": 4e-05,
"loss": 6.2309,
"step": 10000
},
{
"epoch": 2.99,
"learning_rate": 3.955555555555556e-05,
"loss": 6.2262,
"step": 11000
},
{
"epoch": 3.26,
"learning_rate": 3.9111111111111115e-05,
"loss": 6.221,
"step": 12000
},
{
"epoch": 3.54,
"learning_rate": 3.866666666666667e-05,
"loss": 6.2168,
"step": 13000
},
{
"epoch": 3.81,
"learning_rate": 3.8222222222222226e-05,
"loss": 6.2153,
"step": 14000
},
{
"epoch": 4.08,
"learning_rate": 3.777777777777778e-05,
"loss": 6.2099,
"step": 15000
},
{
"epoch": 4.35,
"learning_rate": 3.733333333333334e-05,
"loss": 6.209,
"step": 16000
},
{
"epoch": 4.62,
"learning_rate": 3.6888888888888896e-05,
"loss": 6.2097,
"step": 17000
},
{
"epoch": 4.9,
"learning_rate": 3.644444444444445e-05,
"loss": 6.2056,
"step": 18000
},
{
"epoch": 5.17,
"learning_rate": 3.6e-05,
"loss": 6.2036,
"step": 19000
},
{
"epoch": 5.44,
"learning_rate": 3.555555555555555e-05,
"loss": 6.2015,
"step": 20000
},
{
"epoch": 5.71,
"learning_rate": 3.511111111111111e-05,
"loss": 6.2014,
"step": 21000
},
{
"epoch": 5.98,
"learning_rate": 3.466666666666667e-05,
"loss": 6.1992,
"step": 22000
},
{
"epoch": 6.26,
"learning_rate": 3.4222222222222224e-05,
"loss": 6.1994,
"step": 23000
},
{
"epoch": 6.53,
"learning_rate": 3.377777777777778e-05,
"loss": 6.1886,
"step": 24000
},
{
"epoch": 6.8,
"learning_rate": 3.3333333333333335e-05,
"loss": 6.111,
"step": 25000
},
{
"epoch": 7.07,
"learning_rate": 3.288888888888889e-05,
"loss": 6.0575,
"step": 26000
},
{
"epoch": 7.34,
"learning_rate": 3.2444444444444446e-05,
"loss": 6.0078,
"step": 27000
},
{
"epoch": 7.61,
"learning_rate": 3.2000000000000005e-05,
"loss": 5.9631,
"step": 28000
},
{
"epoch": 7.89,
"learning_rate": 3.155555555555556e-05,
"loss": 5.922,
"step": 29000
},
{
"epoch": 8.16,
"learning_rate": 3.111111111111112e-05,
"loss": 5.8761,
"step": 30000
},
{
"epoch": 8.43,
"learning_rate": 3.066666666666667e-05,
"loss": 5.832,
"step": 31000
},
{
"epoch": 8.7,
"learning_rate": 3.0222222222222225e-05,
"loss": 5.7851,
"step": 32000
},
{
"epoch": 8.97,
"learning_rate": 2.977777777777778e-05,
"loss": 5.7381,
"step": 33000
},
{
"epoch": 9.25,
"learning_rate": 2.9333333333333333e-05,
"loss": 5.6705,
"step": 34000
},
{
"epoch": 9.52,
"learning_rate": 2.888888888888889e-05,
"loss": 5.4784,
"step": 35000
},
{
"epoch": 9.79,
"learning_rate": 2.8444444444444447e-05,
"loss": 5.1969,
"step": 36000
},
{
"epoch": 10.06,
"learning_rate": 2.8e-05,
"loss": 5.0725,
"step": 37000
},
{
"epoch": 10.33,
"learning_rate": 2.755555555555556e-05,
"loss": 5.0063,
"step": 38000
},
{
"epoch": 10.61,
"learning_rate": 2.7111111111111114e-05,
"loss": 4.9628,
"step": 39000
},
{
"epoch": 10.88,
"learning_rate": 2.6666666666666667e-05,
"loss": 4.9296,
"step": 40000
},
{
"epoch": 11.15,
"learning_rate": 2.6222222222222226e-05,
"loss": 4.899,
"step": 41000
},
{
"epoch": 11.42,
"learning_rate": 2.577777777777778e-05,
"loss": 4.8777,
"step": 42000
},
{
"epoch": 11.69,
"learning_rate": 2.5333333333333334e-05,
"loss": 4.8581,
"step": 43000
},
{
"epoch": 11.97,
"learning_rate": 2.4888888888888893e-05,
"loss": 4.8393,
"step": 44000
},
{
"epoch": 12.24,
"learning_rate": 2.444444444444445e-05,
"loss": 4.8243,
"step": 45000
},
{
"epoch": 12.51,
"learning_rate": 2.4e-05,
"loss": 4.8118,
"step": 46000
},
{
"epoch": 12.78,
"learning_rate": 2.355555555555556e-05,
"loss": 4.799,
"step": 47000
},
{
"epoch": 13.05,
"learning_rate": 2.3111111111111112e-05,
"loss": 4.7892,
"step": 48000
},
{
"epoch": 13.33,
"learning_rate": 2.2666666666666668e-05,
"loss": 4.7799,
"step": 49000
},
{
"epoch": 13.6,
"learning_rate": 2.2222222222222227e-05,
"loss": 4.7639,
"step": 50000
},
{
"epoch": 13.87,
"learning_rate": 2.177777777777778e-05,
"loss": 4.7586,
"step": 51000
},
{
"epoch": 14.14,
"learning_rate": 2.1333333333333335e-05,
"loss": 4.7515,
"step": 52000
},
{
"epoch": 14.41,
"learning_rate": 2.088888888888889e-05,
"loss": 4.7425,
"step": 53000
},
{
"epoch": 14.69,
"learning_rate": 2.0444444444444446e-05,
"loss": 4.7301,
"step": 54000
},
{
"epoch": 14.96,
"learning_rate": 2e-05,
"loss": 4.7283,
"step": 55000
},
{
"epoch": 15.23,
"learning_rate": 1.9555555555555557e-05,
"loss": 4.7201,
"step": 56000
},
{
"epoch": 15.5,
"learning_rate": 1.9111111111111113e-05,
"loss": 4.7155,
"step": 57000
},
{
"epoch": 15.77,
"learning_rate": 1.866666666666667e-05,
"loss": 4.7073,
"step": 58000
},
{
"epoch": 16.05,
"learning_rate": 1.8222222222222224e-05,
"loss": 4.7043,
"step": 59000
},
{
"epoch": 16.32,
"learning_rate": 1.7777777777777777e-05,
"loss": 4.6977,
"step": 60000
},
{
"epoch": 16.59,
"learning_rate": 1.7333333333333336e-05,
"loss": 4.6925,
"step": 61000
},
{
"epoch": 16.86,
"learning_rate": 1.688888888888889e-05,
"loss": 4.6867,
"step": 62000
},
{
"epoch": 17.13,
"learning_rate": 1.6444444444444444e-05,
"loss": 4.6806,
"step": 63000
},
{
"epoch": 17.41,
"learning_rate": 1.6000000000000003e-05,
"loss": 4.6754,
"step": 64000
},
{
"epoch": 17.68,
"learning_rate": 1.555555555555556e-05,
"loss": 4.6791,
"step": 65000
},
{
"epoch": 17.95,
"learning_rate": 1.5111111111111112e-05,
"loss": 4.6689,
"step": 66000
},
{
"epoch": 18.22,
"learning_rate": 1.4666666666666666e-05,
"loss": 4.6677,
"step": 67000
},
{
"epoch": 18.49,
"learning_rate": 1.4222222222222224e-05,
"loss": 4.6618,
"step": 68000
},
{
"epoch": 18.77,
"learning_rate": 1.377777777777778e-05,
"loss": 4.6565,
"step": 69000
},
{
"epoch": 19.04,
"learning_rate": 1.3333333333333333e-05,
"loss": 4.6544,
"step": 70000
},
{
"epoch": 19.31,
"learning_rate": 1.288888888888889e-05,
"loss": 4.6488,
"step": 71000
},
{
"epoch": 19.58,
"learning_rate": 1.2444444444444446e-05,
"loss": 4.6467,
"step": 72000
},
{
"epoch": 19.85,
"learning_rate": 1.2e-05,
"loss": 4.6478,
"step": 73000
},
{
"epoch": 20.13,
"learning_rate": 1.1555555555555556e-05,
"loss": 4.6429,
"step": 74000
},
{
"epoch": 20.4,
"learning_rate": 1.1111111111111113e-05,
"loss": 4.6436,
"step": 75000
},
{
"epoch": 20.67,
"learning_rate": 1.0666666666666667e-05,
"loss": 4.6352,
"step": 76000
},
{
"epoch": 20.94,
"learning_rate": 1.0222222222222223e-05,
"loss": 4.6338,
"step": 77000
},
{
"epoch": 21.21,
"learning_rate": 9.777777777777779e-06,
"loss": 4.6292,
"step": 78000
},
{
"epoch": 21.48,
"learning_rate": 9.333333333333334e-06,
"loss": 4.6305,
"step": 79000
},
{
"epoch": 21.76,
"learning_rate": 8.888888888888888e-06,
"loss": 4.6269,
"step": 80000
},
{
"epoch": 22.03,
"learning_rate": 8.444444444444446e-06,
"loss": 4.6256,
"step": 81000
},
{
"epoch": 22.3,
"learning_rate": 8.000000000000001e-06,
"loss": 4.6254,
"step": 82000
},
{
"epoch": 22.57,
"learning_rate": 7.555555555555556e-06,
"loss": 4.6209,
"step": 83000
},
{
"epoch": 22.84,
"learning_rate": 7.111111111111112e-06,
"loss": 4.6196,
"step": 84000
},
{
"epoch": 23.12,
"learning_rate": 6.666666666666667e-06,
"loss": 4.6162,
"step": 85000
},
{
"epoch": 23.39,
"learning_rate": 6.222222222222223e-06,
"loss": 4.6128,
"step": 86000
},
{
"epoch": 23.66,
"learning_rate": 5.777777777777778e-06,
"loss": 4.6184,
"step": 87000
},
{
"epoch": 23.93,
"learning_rate": 5.333333333333334e-06,
"loss": 4.6133,
"step": 88000
},
{
"epoch": 24.2,
"learning_rate": 4.888888888888889e-06,
"loss": 4.6098,
"step": 89000
},
{
"epoch": 24.48,
"learning_rate": 4.444444444444444e-06,
"loss": 4.6101,
"step": 90000
},
{
"epoch": 24.75,
"learning_rate": 4.000000000000001e-06,
"loss": 4.6099,
"step": 91000
},
{
"epoch": 25.02,
"learning_rate": 3.555555555555556e-06,
"loss": 4.6092,
"step": 92000
},
{
"epoch": 25.29,
"learning_rate": 3.1111111111111116e-06,
"loss": 4.6061,
"step": 93000
},
{
"epoch": 25.56,
"learning_rate": 2.666666666666667e-06,
"loss": 4.6072,
"step": 94000
},
{
"epoch": 25.84,
"learning_rate": 2.222222222222222e-06,
"loss": 4.6041,
"step": 95000
},
{
"epoch": 26.11,
"learning_rate": 1.777777777777778e-06,
"loss": 4.6047,
"step": 96000
},
{
"epoch": 26.38,
"learning_rate": 1.3333333333333334e-06,
"loss": 4.6079,
"step": 97000
},
{
"epoch": 26.65,
"learning_rate": 8.88888888888889e-07,
"loss": 4.6021,
"step": 98000
},
{
"epoch": 26.92,
"learning_rate": 4.444444444444445e-07,
"loss": 4.6046,
"step": 99000
},
{
"epoch": 27.2,
"learning_rate": 0.0,
"loss": 4.6024,
"step": 100000
}
],
"logging_steps": 1000,
"max_steps": 100000,
"num_train_epochs": 28,
"save_steps": 1000,
"total_flos": 7.577490197651128e+18,
"trial_name": null,
"trial_params": null
}