LEIA-LM-large / trainer_state.json
saroyehun's picture
Upload with huggingface_hub
b222698
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 55.74102448027301,
"global_step": 100000,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.28,
"learning_rate": 4.975e-05,
"loss": 1.2516,
"step": 500
},
{
"epoch": 0.56,
"learning_rate": 4.9500000000000004e-05,
"loss": 1.2053,
"step": 1000
},
{
"epoch": 0.84,
"learning_rate": 4.9250000000000004e-05,
"loss": 1.1918,
"step": 1500
},
{
"epoch": 1.11,
"learning_rate": 4.9e-05,
"loss": 1.1834,
"step": 2000
},
{
"epoch": 1.39,
"learning_rate": 4.875e-05,
"loss": 1.1709,
"step": 2500
},
{
"epoch": 1.67,
"learning_rate": 4.85e-05,
"loss": 1.1644,
"step": 3000
},
{
"epoch": 1.95,
"learning_rate": 4.825e-05,
"loss": 1.16,
"step": 3500
},
{
"epoch": 2.23,
"learning_rate": 4.8e-05,
"loss": 1.1519,
"step": 4000
},
{
"epoch": 2.51,
"learning_rate": 4.775e-05,
"loss": 1.1457,
"step": 4500
},
{
"epoch": 2.79,
"learning_rate": 4.75e-05,
"loss": 1.1413,
"step": 5000
},
{
"epoch": 3.07,
"learning_rate": 4.7249999999999997e-05,
"loss": 1.1355,
"step": 5500
},
{
"epoch": 3.34,
"learning_rate": 4.7e-05,
"loss": 1.1261,
"step": 6000
},
{
"epoch": 3.62,
"learning_rate": 4.6750000000000005e-05,
"loss": 1.1246,
"step": 6500
},
{
"epoch": 3.9,
"learning_rate": 4.6500000000000005e-05,
"loss": 1.1243,
"step": 7000
},
{
"epoch": 4.18,
"learning_rate": 4.6250000000000006e-05,
"loss": 1.1197,
"step": 7500
},
{
"epoch": 4.46,
"learning_rate": 4.600000000000001e-05,
"loss": 1.115,
"step": 8000
},
{
"epoch": 4.74,
"learning_rate": 4.575e-05,
"loss": 1.1122,
"step": 8500
},
{
"epoch": 5.02,
"learning_rate": 4.55e-05,
"loss": 1.1142,
"step": 9000
},
{
"epoch": 5.3,
"learning_rate": 4.525e-05,
"loss": 1.1065,
"step": 9500
},
{
"epoch": 5.57,
"learning_rate": 4.5e-05,
"loss": 1.1047,
"step": 10000
},
{
"epoch": 5.85,
"learning_rate": 4.4750000000000004e-05,
"loss": 1.1034,
"step": 10500
},
{
"epoch": 6.13,
"learning_rate": 4.4500000000000004e-05,
"loss": 1.1027,
"step": 11000
},
{
"epoch": 6.41,
"learning_rate": 4.4250000000000005e-05,
"loss": 1.0974,
"step": 11500
},
{
"epoch": 6.69,
"learning_rate": 4.4000000000000006e-05,
"loss": 1.096,
"step": 12000
},
{
"epoch": 6.97,
"learning_rate": 4.375e-05,
"loss": 1.0961,
"step": 12500
},
{
"epoch": 7.25,
"learning_rate": 4.35e-05,
"loss": 1.0912,
"step": 13000
},
{
"epoch": 7.52,
"learning_rate": 4.325e-05,
"loss": 1.0892,
"step": 13500
},
{
"epoch": 7.8,
"learning_rate": 4.3e-05,
"loss": 1.0898,
"step": 14000
},
{
"epoch": 8.08,
"learning_rate": 4.275e-05,
"loss": 1.0875,
"step": 14500
},
{
"epoch": 8.36,
"learning_rate": 4.25e-05,
"loss": 1.0839,
"step": 15000
},
{
"epoch": 8.64,
"learning_rate": 4.2250000000000004e-05,
"loss": 1.0814,
"step": 15500
},
{
"epoch": 8.92,
"learning_rate": 4.2e-05,
"loss": 1.0824,
"step": 16000
},
{
"epoch": 9.2,
"learning_rate": 4.175e-05,
"loss": 1.0795,
"step": 16500
},
{
"epoch": 9.48,
"learning_rate": 4.15e-05,
"loss": 1.076,
"step": 17000
},
{
"epoch": 9.75,
"learning_rate": 4.125e-05,
"loss": 1.0759,
"step": 17500
},
{
"epoch": 10.03,
"learning_rate": 4.1e-05,
"loss": 1.0764,
"step": 18000
},
{
"epoch": 10.31,
"learning_rate": 4.075e-05,
"loss": 1.0688,
"step": 18500
},
{
"epoch": 10.59,
"learning_rate": 4.05e-05,
"loss": 1.0715,
"step": 19000
},
{
"epoch": 10.87,
"learning_rate": 4.025e-05,
"loss": 1.0712,
"step": 19500
},
{
"epoch": 11.15,
"learning_rate": 4e-05,
"loss": 1.0691,
"step": 20000
},
{
"epoch": 11.43,
"learning_rate": 3.9750000000000004e-05,
"loss": 1.0657,
"step": 20500
},
{
"epoch": 11.71,
"learning_rate": 3.9500000000000005e-05,
"loss": 1.0637,
"step": 21000
},
{
"epoch": 11.98,
"learning_rate": 3.9250000000000005e-05,
"loss": 1.0664,
"step": 21500
},
{
"epoch": 12.26,
"learning_rate": 3.9000000000000006e-05,
"loss": 1.0629,
"step": 22000
},
{
"epoch": 12.54,
"learning_rate": 3.875e-05,
"loss": 1.061,
"step": 22500
},
{
"epoch": 12.82,
"learning_rate": 3.85e-05,
"loss": 1.0616,
"step": 23000
},
{
"epoch": 13.1,
"learning_rate": 3.825e-05,
"loss": 1.0612,
"step": 23500
},
{
"epoch": 13.38,
"learning_rate": 3.8e-05,
"loss": 1.0563,
"step": 24000
},
{
"epoch": 13.66,
"learning_rate": 3.775e-05,
"loss": 1.0574,
"step": 24500
},
{
"epoch": 13.93,
"learning_rate": 3.7500000000000003e-05,
"loss": 1.0574,
"step": 25000
},
{
"epoch": 14.21,
"learning_rate": 3.7250000000000004e-05,
"loss": 1.0525,
"step": 25500
},
{
"epoch": 14.49,
"learning_rate": 3.7e-05,
"loss": 1.0518,
"step": 26000
},
{
"epoch": 14.77,
"learning_rate": 3.675e-05,
"loss": 1.0511,
"step": 26500
},
{
"epoch": 15.05,
"learning_rate": 3.65e-05,
"loss": 1.0529,
"step": 27000
},
{
"epoch": 15.33,
"learning_rate": 3.625e-05,
"loss": 1.0452,
"step": 27500
},
{
"epoch": 15.61,
"learning_rate": 3.6e-05,
"loss": 1.0464,
"step": 28000
},
{
"epoch": 15.89,
"learning_rate": 3.575e-05,
"loss": 1.047,
"step": 28500
},
{
"epoch": 16.16,
"learning_rate": 3.55e-05,
"loss": 1.0449,
"step": 29000
},
{
"epoch": 16.44,
"learning_rate": 3.525e-05,
"loss": 1.0425,
"step": 29500
},
{
"epoch": 16.72,
"learning_rate": 3.5e-05,
"loss": 1.0428,
"step": 30000
},
{
"epoch": 17.0,
"learning_rate": 3.475e-05,
"loss": 1.0445,
"step": 30500
},
{
"epoch": 17.28,
"learning_rate": 3.45e-05,
"loss": 1.0382,
"step": 31000
},
{
"epoch": 17.56,
"learning_rate": 3.4250000000000006e-05,
"loss": 1.0376,
"step": 31500
},
{
"epoch": 17.84,
"learning_rate": 3.4000000000000007e-05,
"loss": 1.0384,
"step": 32000
},
{
"epoch": 18.12,
"learning_rate": 3.375000000000001e-05,
"loss": 1.0392,
"step": 32500
},
{
"epoch": 18.39,
"learning_rate": 3.35e-05,
"loss": 1.0337,
"step": 33000
},
{
"epoch": 18.67,
"learning_rate": 3.325e-05,
"loss": 1.0333,
"step": 33500
},
{
"epoch": 18.95,
"learning_rate": 3.3e-05,
"loss": 1.0357,
"step": 34000
},
{
"epoch": 19.23,
"learning_rate": 3.275e-05,
"loss": 1.033,
"step": 34500
},
{
"epoch": 19.51,
"learning_rate": 3.2500000000000004e-05,
"loss": 1.0312,
"step": 35000
},
{
"epoch": 19.79,
"learning_rate": 3.2250000000000005e-05,
"loss": 1.0299,
"step": 35500
},
{
"epoch": 20.07,
"learning_rate": 3.2000000000000005e-05,
"loss": 1.0293,
"step": 36000
},
{
"epoch": 20.35,
"learning_rate": 3.175e-05,
"loss": 1.0262,
"step": 36500
},
{
"epoch": 20.62,
"learning_rate": 3.15e-05,
"loss": 1.0268,
"step": 37000
},
{
"epoch": 20.9,
"learning_rate": 3.125e-05,
"loss": 1.0271,
"step": 37500
},
{
"epoch": 21.18,
"learning_rate": 3.1e-05,
"loss": 1.0236,
"step": 38000
},
{
"epoch": 21.46,
"learning_rate": 3.075e-05,
"loss": 1.0245,
"step": 38500
},
{
"epoch": 21.74,
"learning_rate": 3.05e-05,
"loss": 1.0223,
"step": 39000
},
{
"epoch": 22.02,
"learning_rate": 3.025e-05,
"loss": 1.0244,
"step": 39500
},
{
"epoch": 22.3,
"learning_rate": 3e-05,
"loss": 1.0186,
"step": 40000
},
{
"epoch": 22.57,
"learning_rate": 2.975e-05,
"loss": 1.0195,
"step": 40500
},
{
"epoch": 22.85,
"learning_rate": 2.95e-05,
"loss": 1.0192,
"step": 41000
},
{
"epoch": 23.13,
"learning_rate": 2.925e-05,
"loss": 1.018,
"step": 41500
},
{
"epoch": 23.41,
"learning_rate": 2.9e-05,
"loss": 1.0169,
"step": 42000
},
{
"epoch": 23.69,
"learning_rate": 2.8749999999999997e-05,
"loss": 1.0155,
"step": 42500
},
{
"epoch": 23.97,
"learning_rate": 2.8499999999999998e-05,
"loss": 1.0165,
"step": 43000
},
{
"epoch": 24.25,
"learning_rate": 2.825e-05,
"loss": 1.014,
"step": 43500
},
{
"epoch": 24.53,
"learning_rate": 2.8000000000000003e-05,
"loss": 1.0134,
"step": 44000
},
{
"epoch": 24.8,
"learning_rate": 2.7750000000000004e-05,
"loss": 1.0121,
"step": 44500
},
{
"epoch": 25.08,
"learning_rate": 2.7500000000000004e-05,
"loss": 1.0139,
"step": 45000
},
{
"epoch": 25.36,
"learning_rate": 2.725e-05,
"loss": 1.0115,
"step": 45500
},
{
"epoch": 25.64,
"learning_rate": 2.7000000000000002e-05,
"loss": 1.0096,
"step": 46000
},
{
"epoch": 25.92,
"learning_rate": 2.6750000000000003e-05,
"loss": 1.0114,
"step": 46500
},
{
"epoch": 26.2,
"learning_rate": 2.6500000000000004e-05,
"loss": 1.0101,
"step": 47000
},
{
"epoch": 26.48,
"learning_rate": 2.625e-05,
"loss": 1.0067,
"step": 47500
},
{
"epoch": 26.76,
"learning_rate": 2.6000000000000002e-05,
"loss": 1.0067,
"step": 48000
},
{
"epoch": 27.03,
"learning_rate": 2.5750000000000002e-05,
"loss": 1.0097,
"step": 48500
},
{
"epoch": 27.31,
"learning_rate": 2.5500000000000003e-05,
"loss": 1.003,
"step": 49000
},
{
"epoch": 27.59,
"learning_rate": 2.525e-05,
"loss": 1.0046,
"step": 49500
},
{
"epoch": 27.87,
"learning_rate": 2.5e-05,
"loss": 1.0031,
"step": 50000
},
{
"epoch": 28.15,
"learning_rate": 2.4750000000000002e-05,
"loss": 1.0033,
"step": 50500
},
{
"epoch": 28.43,
"learning_rate": 2.45e-05,
"loss": 1.0001,
"step": 51000
},
{
"epoch": 28.71,
"learning_rate": 2.425e-05,
"loss": 0.9997,
"step": 51500
},
{
"epoch": 28.99,
"learning_rate": 2.4e-05,
"loss": 1.0009,
"step": 52000
},
{
"epoch": 29.26,
"learning_rate": 2.375e-05,
"loss": 0.9986,
"step": 52500
},
{
"epoch": 29.54,
"learning_rate": 2.35e-05,
"loss": 0.9972,
"step": 53000
},
{
"epoch": 29.82,
"learning_rate": 2.3250000000000003e-05,
"loss": 0.9999,
"step": 53500
},
{
"epoch": 30.1,
"learning_rate": 2.3000000000000003e-05,
"loss": 0.9984,
"step": 54000
},
{
"epoch": 30.38,
"learning_rate": 2.275e-05,
"loss": 0.994,
"step": 54500
},
{
"epoch": 30.66,
"learning_rate": 2.25e-05,
"loss": 0.9948,
"step": 55000
},
{
"epoch": 30.94,
"learning_rate": 2.2250000000000002e-05,
"loss": 0.9953,
"step": 55500
},
{
"epoch": 31.22,
"learning_rate": 2.2000000000000003e-05,
"loss": 0.9958,
"step": 56000
},
{
"epoch": 31.49,
"learning_rate": 2.175e-05,
"loss": 0.9939,
"step": 56500
},
{
"epoch": 31.77,
"learning_rate": 2.15e-05,
"loss": 0.9933,
"step": 57000
},
{
"epoch": 32.05,
"learning_rate": 2.125e-05,
"loss": 0.9942,
"step": 57500
},
{
"epoch": 32.33,
"learning_rate": 2.1e-05,
"loss": 0.9909,
"step": 58000
},
{
"epoch": 32.61,
"learning_rate": 2.075e-05,
"loss": 0.9909,
"step": 58500
},
{
"epoch": 32.89,
"learning_rate": 2.05e-05,
"loss": 0.9911,
"step": 59000
},
{
"epoch": 33.17,
"learning_rate": 2.025e-05,
"loss": 0.9895,
"step": 59500
},
{
"epoch": 33.44,
"learning_rate": 2e-05,
"loss": 0.9877,
"step": 60000
},
{
"epoch": 33.72,
"learning_rate": 1.9750000000000002e-05,
"loss": 0.9886,
"step": 60500
},
{
"epoch": 34.0,
"learning_rate": 1.9500000000000003e-05,
"loss": 0.9905,
"step": 61000
},
{
"epoch": 34.28,
"learning_rate": 1.925e-05,
"loss": 0.9851,
"step": 61500
},
{
"epoch": 34.56,
"learning_rate": 1.9e-05,
"loss": 0.987,
"step": 62000
},
{
"epoch": 34.84,
"learning_rate": 1.8750000000000002e-05,
"loss": 0.9852,
"step": 62500
},
{
"epoch": 35.12,
"learning_rate": 1.85e-05,
"loss": 0.9863,
"step": 63000
},
{
"epoch": 35.4,
"learning_rate": 1.825e-05,
"loss": 0.9814,
"step": 63500
},
{
"epoch": 35.67,
"learning_rate": 1.8e-05,
"loss": 0.9819,
"step": 64000
},
{
"epoch": 35.95,
"learning_rate": 1.775e-05,
"loss": 0.9824,
"step": 64500
},
{
"epoch": 36.23,
"learning_rate": 1.75e-05,
"loss": 0.9808,
"step": 65000
},
{
"epoch": 36.51,
"learning_rate": 1.725e-05,
"loss": 0.9798,
"step": 65500
},
{
"epoch": 36.79,
"learning_rate": 1.7000000000000003e-05,
"loss": 0.9803,
"step": 66000
},
{
"epoch": 37.07,
"learning_rate": 1.675e-05,
"loss": 0.9815,
"step": 66500
},
{
"epoch": 37.35,
"learning_rate": 1.65e-05,
"loss": 0.9783,
"step": 67000
},
{
"epoch": 37.63,
"learning_rate": 1.6250000000000002e-05,
"loss": 0.9787,
"step": 67500
},
{
"epoch": 37.9,
"learning_rate": 1.6000000000000003e-05,
"loss": 0.9788,
"step": 68000
},
{
"epoch": 38.18,
"learning_rate": 1.575e-05,
"loss": 0.9775,
"step": 68500
},
{
"epoch": 38.46,
"learning_rate": 1.55e-05,
"loss": 0.9759,
"step": 69000
},
{
"epoch": 38.74,
"learning_rate": 1.525e-05,
"loss": 0.9758,
"step": 69500
},
{
"epoch": 39.02,
"learning_rate": 1.5e-05,
"loss": 0.9774,
"step": 70000
},
{
"epoch": 39.3,
"learning_rate": 1.475e-05,
"loss": 0.9723,
"step": 70500
},
{
"epoch": 39.58,
"learning_rate": 1.45e-05,
"loss": 0.9725,
"step": 71000
},
{
"epoch": 39.85,
"learning_rate": 1.4249999999999999e-05,
"loss": 0.9747,
"step": 71500
},
{
"epoch": 40.13,
"learning_rate": 1.4000000000000001e-05,
"loss": 0.9743,
"step": 72000
},
{
"epoch": 40.41,
"learning_rate": 1.3750000000000002e-05,
"loss": 0.9709,
"step": 72500
},
{
"epoch": 40.69,
"learning_rate": 1.3500000000000001e-05,
"loss": 0.9715,
"step": 73000
},
{
"epoch": 40.97,
"learning_rate": 1.3250000000000002e-05,
"loss": 0.9707,
"step": 73500
},
{
"epoch": 41.25,
"learning_rate": 1.3000000000000001e-05,
"loss": 0.9729,
"step": 74000
},
{
"epoch": 41.53,
"learning_rate": 1.2750000000000002e-05,
"loss": 0.9701,
"step": 74500
},
{
"epoch": 41.81,
"learning_rate": 1.25e-05,
"loss": 0.9691,
"step": 75000
},
{
"epoch": 42.08,
"learning_rate": 1.225e-05,
"loss": 0.9708,
"step": 75500
},
{
"epoch": 42.36,
"learning_rate": 1.2e-05,
"loss": 0.9678,
"step": 76000
},
{
"epoch": 42.64,
"learning_rate": 1.175e-05,
"loss": 0.9665,
"step": 76500
},
{
"epoch": 42.92,
"learning_rate": 1.1500000000000002e-05,
"loss": 0.9691,
"step": 77000
},
{
"epoch": 43.2,
"learning_rate": 1.125e-05,
"loss": 0.968,
"step": 77500
},
{
"epoch": 43.48,
"learning_rate": 1.1000000000000001e-05,
"loss": 0.9662,
"step": 78000
},
{
"epoch": 43.76,
"learning_rate": 1.075e-05,
"loss": 0.9672,
"step": 78500
},
{
"epoch": 44.04,
"learning_rate": 1.05e-05,
"loss": 0.9666,
"step": 79000
},
{
"epoch": 44.31,
"learning_rate": 1.025e-05,
"loss": 0.965,
"step": 79500
},
{
"epoch": 44.59,
"learning_rate": 1e-05,
"loss": 0.9644,
"step": 80000
},
{
"epoch": 44.87,
"learning_rate": 9.750000000000002e-06,
"loss": 0.9624,
"step": 80500
},
{
"epoch": 45.15,
"learning_rate": 9.5e-06,
"loss": 0.9652,
"step": 81000
},
{
"epoch": 45.43,
"learning_rate": 9.25e-06,
"loss": 0.9633,
"step": 81500
},
{
"epoch": 45.71,
"learning_rate": 9e-06,
"loss": 0.9621,
"step": 82000
},
{
"epoch": 45.99,
"learning_rate": 8.75e-06,
"loss": 0.9613,
"step": 82500
},
{
"epoch": 46.27,
"learning_rate": 8.500000000000002e-06,
"loss": 0.9628,
"step": 83000
},
{
"epoch": 46.54,
"learning_rate": 8.25e-06,
"loss": 0.9596,
"step": 83500
},
{
"epoch": 46.82,
"learning_rate": 8.000000000000001e-06,
"loss": 0.9597,
"step": 84000
},
{
"epoch": 47.1,
"learning_rate": 7.75e-06,
"loss": 0.9618,
"step": 84500
},
{
"epoch": 47.38,
"learning_rate": 7.5e-06,
"loss": 0.958,
"step": 85000
},
{
"epoch": 47.66,
"learning_rate": 7.25e-06,
"loss": 0.9585,
"step": 85500
},
{
"epoch": 47.94,
"learning_rate": 7.000000000000001e-06,
"loss": 0.9594,
"step": 86000
},
{
"epoch": 48.22,
"learning_rate": 6.750000000000001e-06,
"loss": 0.9583,
"step": 86500
},
{
"epoch": 48.49,
"learning_rate": 6.5000000000000004e-06,
"loss": 0.9576,
"step": 87000
},
{
"epoch": 48.77,
"learning_rate": 6.25e-06,
"loss": 0.9572,
"step": 87500
},
{
"epoch": 49.05,
"learning_rate": 6e-06,
"loss": 0.9592,
"step": 88000
},
{
"epoch": 49.33,
"learning_rate": 5.750000000000001e-06,
"loss": 0.9576,
"step": 88500
},
{
"epoch": 49.61,
"learning_rate": 5.500000000000001e-06,
"loss": 0.9553,
"step": 89000
},
{
"epoch": 49.89,
"learning_rate": 5.25e-06,
"loss": 0.9553,
"step": 89500
},
{
"epoch": 50.17,
"learning_rate": 5e-06,
"loss": 0.9577,
"step": 90000
},
{
"epoch": 50.45,
"learning_rate": 4.75e-06,
"loss": 0.9542,
"step": 90500
},
{
"epoch": 50.72,
"learning_rate": 4.5e-06,
"loss": 0.953,
"step": 91000
},
{
"epoch": 51.0,
"learning_rate": 4.250000000000001e-06,
"loss": 0.956,
"step": 91500
},
{
"epoch": 51.28,
"learning_rate": 4.000000000000001e-06,
"loss": 0.9544,
"step": 92000
},
{
"epoch": 51.56,
"learning_rate": 3.75e-06,
"loss": 0.953,
"step": 92500
},
{
"epoch": 51.84,
"learning_rate": 3.5000000000000004e-06,
"loss": 0.953,
"step": 93000
},
{
"epoch": 52.12,
"learning_rate": 3.2500000000000002e-06,
"loss": 0.9542,
"step": 93500
},
{
"epoch": 52.4,
"learning_rate": 3e-06,
"loss": 0.9516,
"step": 94000
},
{
"epoch": 52.68,
"learning_rate": 2.7500000000000004e-06,
"loss": 0.9508,
"step": 94500
},
{
"epoch": 52.95,
"learning_rate": 2.5e-06,
"loss": 0.9506,
"step": 95000
},
{
"epoch": 53.23,
"learning_rate": 2.25e-06,
"loss": 0.9524,
"step": 95500
},
{
"epoch": 53.51,
"learning_rate": 2.0000000000000003e-06,
"loss": 0.951,
"step": 96000
},
{
"epoch": 53.79,
"learning_rate": 1.7500000000000002e-06,
"loss": 0.9494,
"step": 96500
},
{
"epoch": 54.07,
"learning_rate": 1.5e-06,
"loss": 0.9523,
"step": 97000
},
{
"epoch": 54.35,
"learning_rate": 1.25e-06,
"loss": 0.9509,
"step": 97500
},
{
"epoch": 54.63,
"learning_rate": 1.0000000000000002e-06,
"loss": 0.9502,
"step": 98000
},
{
"epoch": 54.9,
"learning_rate": 7.5e-07,
"loss": 0.9489,
"step": 98500
},
{
"epoch": 55.18,
"learning_rate": 5.000000000000001e-07,
"loss": 0.951,
"step": 99000
},
{
"epoch": 55.46,
"learning_rate": 2.5000000000000004e-07,
"loss": 0.9484,
"step": 99500
},
{
"epoch": 55.74,
"learning_rate": 0.0,
"loss": 0.9494,
"step": 100000
}
],
"max_steps": 100000,
"num_train_epochs": 56,
"total_flos": 2.38719562761468e+19,
"trial_name": null,
"trial_params": null
}