{ "best_metric": null, "best_model_checkpoint": null, "epoch": 499.0004, "global_step": 5000, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 12.0, "learning_rate": 4.2000000000000006e-07, "loss": 3.8007, "step": 25 }, { "epoch": 24.0, "learning_rate": 9.200000000000001e-07, "loss": 3.2956, "step": 50 }, { "epoch": 37.0, "learning_rate": 1.42e-06, "loss": 2.4056, "step": 75 }, { "epoch": 49.0, "learning_rate": 1.9200000000000003e-06, "loss": 1.8024, "step": 100 }, { "epoch": 62.0, "learning_rate": 2.42e-06, "loss": 1.5318, "step": 125 }, { "epoch": 74.0, "learning_rate": 2.92e-06, "loss": 1.3252, "step": 150 }, { "epoch": 87.0, "learning_rate": 3.4200000000000007e-06, "loss": 1.1945, "step": 175 }, { "epoch": 99.0, "learning_rate": 3.920000000000001e-06, "loss": 1.064, "step": 200 }, { "epoch": 112.0, "learning_rate": 4.42e-06, "loss": 0.9908, "step": 225 }, { "epoch": 124.0, "learning_rate": 4.92e-06, "loss": 0.9655, "step": 250 }, { "epoch": 137.0, "learning_rate": 5.420000000000001e-06, "loss": 0.8882, "step": 275 }, { "epoch": 149.0, "learning_rate": 5.92e-06, "loss": 0.8713, "step": 300 }, { "epoch": 162.0, "learning_rate": 6.42e-06, "loss": 0.8381, "step": 325 }, { "epoch": 174.0, "learning_rate": 6.92e-06, "loss": 0.8135, "step": 350 }, { "epoch": 187.0, "learning_rate": 7.420000000000001e-06, "loss": 0.7839, "step": 375 }, { "epoch": 199.0, "learning_rate": 7.92e-06, "loss": 0.7653, "step": 400 }, { "epoch": 212.0, "learning_rate": 8.42e-06, "loss": 0.7302, "step": 425 }, { "epoch": 224.0, "learning_rate": 8.920000000000001e-06, "loss": 0.6973, "step": 450 }, { "epoch": 237.0, "learning_rate": 9.42e-06, "loss": 0.6878, "step": 475 }, { "epoch": 249.0, "learning_rate": 9.920000000000002e-06, "loss": 0.6724, "step": 500 }, { "epoch": 262.0, "learning_rate": 9.953333333333333e-06, "loss": 0.6495, "step": 525 }, { "epoch": 274.0, "learning_rate": 9.89777777777778e-06, "loss": 0.6637, "step": 550 }, { "epoch": 287.0, "learning_rate": 9.842222222222223e-06, "loss": 0.6024, "step": 575 }, { "epoch": 299.0, "learning_rate": 9.786666666666667e-06, "loss": 0.6149, "step": 600 }, { "epoch": 312.0, "learning_rate": 9.731111111111113e-06, "loss": 0.5797, "step": 625 }, { "epoch": 324.0, "learning_rate": 9.675555555555555e-06, "loss": 0.6173, "step": 650 }, { "epoch": 337.0, "learning_rate": 9.620000000000001e-06, "loss": 0.5977, "step": 675 }, { "epoch": 349.0, "learning_rate": 9.564444444444445e-06, "loss": 0.5626, "step": 700 }, { "epoch": 362.0, "learning_rate": 9.508888888888889e-06, "loss": 0.536, "step": 725 }, { "epoch": 374.0, "learning_rate": 9.453333333333335e-06, "loss": 0.5508, "step": 750 }, { "epoch": 387.0, "learning_rate": 9.397777777777779e-06, "loss": 0.5738, "step": 775 }, { "epoch": 399.0, "learning_rate": 9.342222222222223e-06, "loss": 0.54, "step": 800 }, { "epoch": 412.0, "learning_rate": 9.286666666666667e-06, "loss": 0.5628, "step": 825 }, { "epoch": 424.0, "learning_rate": 9.231111111111111e-06, "loss": 0.5323, "step": 850 }, { "epoch": 437.0, "learning_rate": 9.175555555555557e-06, "loss": 0.5457, "step": 875 }, { "epoch": 449.0, "learning_rate": 9.12e-06, "loss": 0.5267, "step": 900 }, { "epoch": 462.0, "learning_rate": 9.064444444444447e-06, "loss": 0.4893, "step": 925 }, { "epoch": 474.0, "learning_rate": 9.008888888888889e-06, "loss": 0.5081, "step": 950 }, { "epoch": 487.0, "learning_rate": 8.953333333333335e-06, "loss": 0.5135, "step": 975 }, { "epoch": 499.0, "learning_rate": 8.897777777777779e-06, "loss": 0.5068, "step": 1000 }, { "epoch": 512.0, "learning_rate": 8.842222222222223e-06, "loss": 0.5039, "step": 1025 }, { "epoch": 524.0, "learning_rate": 8.786666666666668e-06, "loss": 0.4582, "step": 1050 }, { "epoch": 537.0, "learning_rate": 8.73111111111111e-06, "loss": 0.4997, "step": 1075 }, { "epoch": 549.0, "learning_rate": 8.675555555555556e-06, "loss": 0.4574, "step": 1100 }, { "epoch": 562.0, "learning_rate": 8.62e-06, "loss": 0.4683, "step": 1125 }, { "epoch": 574.0, "learning_rate": 8.564444444444445e-06, "loss": 0.4714, "step": 1150 }, { "epoch": 587.0, "learning_rate": 8.50888888888889e-06, "loss": 0.4753, "step": 1175 }, { "epoch": 599.0, "learning_rate": 8.453333333333334e-06, "loss": 0.4683, "step": 1200 }, { "epoch": 612.0, "learning_rate": 8.397777777777778e-06, "loss": 0.4389, "step": 1225 }, { "epoch": 624.0, "learning_rate": 8.342222222222222e-06, "loss": 0.4476, "step": 1250 }, { "epoch": 637.0, "learning_rate": 8.286666666666668e-06, "loss": 0.4423, "step": 1275 }, { "epoch": 649.0, "learning_rate": 8.231111111111112e-06, "loss": 0.4633, "step": 1300 }, { "epoch": 662.0, "learning_rate": 8.175555555555556e-06, "loss": 0.4397, "step": 1325 }, { "epoch": 674.0, "learning_rate": 8.120000000000002e-06, "loss": 0.4454, "step": 1350 }, { "epoch": 687.0, "learning_rate": 8.064444444444444e-06, "loss": 0.4511, "step": 1375 }, { "epoch": 699.0, "learning_rate": 8.00888888888889e-06, "loss": 0.4445, "step": 1400 }, { "epoch": 712.0, "learning_rate": 7.953333333333334e-06, "loss": 0.4502, "step": 1425 }, { "epoch": 724.0, "learning_rate": 7.897777777777778e-06, "loss": 0.4365, "step": 1450 }, { "epoch": 737.0, "learning_rate": 7.842222222222224e-06, "loss": 0.4103, "step": 1475 }, { "epoch": 749.0, "learning_rate": 7.786666666666666e-06, "loss": 0.418, "step": 1500 }, { "epoch": 762.0, "learning_rate": 7.731111111111112e-06, "loss": 0.4277, "step": 1525 }, { "epoch": 774.0, "learning_rate": 7.675555555555556e-06, "loss": 0.4084, "step": 1550 }, { "epoch": 787.0, "learning_rate": 7.620000000000001e-06, "loss": 0.4162, "step": 1575 }, { "epoch": 799.0, "learning_rate": 7.564444444444446e-06, "loss": 0.4212, "step": 1600 }, { "epoch": 812.0, "learning_rate": 7.50888888888889e-06, "loss": 0.431, "step": 1625 }, { "epoch": 824.0, "learning_rate": 7.453333333333334e-06, "loss": 0.4143, "step": 1650 }, { "epoch": 837.0, "learning_rate": 7.3977777777777786e-06, "loss": 0.411, "step": 1675 }, { "epoch": 849.0, "learning_rate": 7.342222222222223e-06, "loss": 0.393, "step": 1700 }, { "epoch": 862.0, "learning_rate": 7.2866666666666675e-06, "loss": 0.3908, "step": 1725 }, { "epoch": 874.0, "learning_rate": 7.231111111111112e-06, "loss": 0.4116, "step": 1750 }, { "epoch": 887.0, "learning_rate": 7.1755555555555556e-06, "loss": 0.3817, "step": 1775 }, { "epoch": 899.0, "learning_rate": 7.1200000000000004e-06, "loss": 0.43, "step": 1800 }, { "epoch": 912.0, "learning_rate": 7.0644444444444445e-06, "loss": 0.391, "step": 1825 }, { "epoch": 924.0, "learning_rate": 7.008888888888889e-06, "loss": 0.396, "step": 1850 }, { "epoch": 937.0, "learning_rate": 6.953333333333334e-06, "loss": 0.4004, "step": 1875 }, { "epoch": 949.0, "learning_rate": 6.897777777777779e-06, "loss": 0.4315, "step": 1900 }, { "epoch": 962.0, "learning_rate": 6.842222222222222e-06, "loss": 0.3872, "step": 1925 }, { "epoch": 974.0, "learning_rate": 6.786666666666667e-06, "loss": 0.3935, "step": 1950 }, { "epoch": 987.0, "learning_rate": 6.731111111111111e-06, "loss": 0.3885, "step": 1975 }, { "epoch": 999.0, "learning_rate": 6.675555555555556e-06, "loss": 0.3568, "step": 2000 }, { "epoch": 12.0, "learning_rate": 6.6244444444444445e-06, "loss": 0.3769, "step": 2025 }, { "epoch": 24.0, "learning_rate": 6.568888888888889e-06, "loss": 0.3873, "step": 2050 }, { "epoch": 37.0, "learning_rate": 6.513333333333333e-06, "loss": 0.3741, "step": 2075 }, { "epoch": 49.0, "learning_rate": 6.457777777777778e-06, "loss": 0.3796, "step": 2100 }, { "epoch": 62.0, "learning_rate": 6.402222222222223e-06, "loss": 0.3795, "step": 2125 }, { "epoch": 74.0, "learning_rate": 6.346666666666668e-06, "loss": 0.3596, "step": 2150 }, { "epoch": 87.0, "learning_rate": 6.291111111111111e-06, "loss": 0.3611, "step": 2175 }, { "epoch": 99.0, "learning_rate": 6.235555555555556e-06, "loss": 0.3374, "step": 2200 }, { "epoch": 112.0, "learning_rate": 6.18e-06, "loss": 0.342, "step": 2225 }, { "epoch": 124.0, "learning_rate": 6.124444444444445e-06, "loss": 0.354, "step": 2250 }, { "epoch": 137.0, "learning_rate": 6.06888888888889e-06, "loss": 0.3356, "step": 2275 }, { "epoch": 149.0, "learning_rate": 6.013333333333335e-06, "loss": 0.3599, "step": 2300 }, { "epoch": 162.0, "learning_rate": 5.957777777777778e-06, "loss": 0.3402, "step": 2325 }, { "epoch": 174.0, "learning_rate": 5.902222222222223e-06, "loss": 0.346, "step": 2350 }, { "epoch": 187.0, "learning_rate": 5.846666666666667e-06, "loss": 0.3334, "step": 2375 }, { "epoch": 199.0, "learning_rate": 5.791111111111112e-06, "loss": 0.3321, "step": 2400 }, { "epoch": 212.0, "learning_rate": 5.735555555555557e-06, "loss": 0.3228, "step": 2425 }, { "epoch": 224.0, "learning_rate": 5.68e-06, "loss": 0.3038, "step": 2450 }, { "epoch": 237.0, "learning_rate": 5.624444444444445e-06, "loss": 0.3015, "step": 2475 }, { "epoch": 249.0, "learning_rate": 5.56888888888889e-06, "loss": 0.2926, "step": 2500 }, { "epoch": 262.0, "learning_rate": 5.513333333333334e-06, "loss": 0.2883, "step": 2525 }, { "epoch": 274.0, "learning_rate": 5.4577777777777785e-06, "loss": 0.3021, "step": 2550 }, { "epoch": 287.0, "learning_rate": 5.402222222222223e-06, "loss": 0.2654, "step": 2575 }, { "epoch": 299.0, "learning_rate": 5.346666666666667e-06, "loss": 0.285, "step": 2600 }, { "epoch": 312.0, "learning_rate": 5.2911111111111115e-06, "loss": 0.2615, "step": 2625 }, { "epoch": 324.0, "learning_rate": 5.235555555555556e-06, "loss": 0.2955, "step": 2650 }, { "epoch": 337.0, "learning_rate": 5.18e-06, "loss": 0.287, "step": 2675 }, { "epoch": 349.0, "learning_rate": 5.124444444444445e-06, "loss": 0.2647, "step": 2700 }, { "epoch": 362.0, "learning_rate": 5.06888888888889e-06, "loss": 0.2599, "step": 2725 }, { "epoch": 374.0, "learning_rate": 5.013333333333333e-06, "loss": 0.2649, "step": 2750 }, { "epoch": 387.0, "learning_rate": 4.957777777777778e-06, "loss": 0.2918, "step": 2775 }, { "epoch": 399.0, "learning_rate": 4.902222222222222e-06, "loss": 0.2609, "step": 2800 }, { "epoch": 412.0, "learning_rate": 4.846666666666667e-06, "loss": 0.273, "step": 2825 }, { "epoch": 424.0, "learning_rate": 4.791111111111111e-06, "loss": 0.271, "step": 2850 }, { "epoch": 437.0, "learning_rate": 4.735555555555556e-06, "loss": 0.2772, "step": 2875 }, { "epoch": 449.0, "learning_rate": 4.680000000000001e-06, "loss": 0.2594, "step": 2900 }, { "epoch": 462.0, "learning_rate": 4.624444444444445e-06, "loss": 0.2344, "step": 2925 }, { "epoch": 474.0, "learning_rate": 4.568888888888889e-06, "loss": 0.2488, "step": 2950 }, { "epoch": 487.0, "learning_rate": 4.513333333333333e-06, "loss": 0.2609, "step": 2975 }, { "epoch": 499.0, "learning_rate": 4.457777777777778e-06, "loss": 0.2522, "step": 3000 }, { "epoch": 512.0, "learning_rate": 4.402222222222223e-06, "loss": 0.2601, "step": 3025 }, { "epoch": 524.0, "learning_rate": 4.346666666666667e-06, "loss": 0.2294, "step": 3050 }, { "epoch": 537.0, "learning_rate": 4.291111111111112e-06, "loss": 0.2679, "step": 3075 }, { "epoch": 549.0, "learning_rate": 4.235555555555556e-06, "loss": 0.2329, "step": 3100 }, { "epoch": 562.0, "learning_rate": 4.18e-06, "loss": 0.2355, "step": 3125 }, { "epoch": 574.0, "learning_rate": 4.124444444444445e-06, "loss": 0.2384, "step": 3150 }, { "epoch": 587.0, "learning_rate": 4.0688888888888896e-06, "loss": 0.2418, "step": 3175 }, { "epoch": 599.0, "learning_rate": 4.013333333333334e-06, "loss": 0.2478, "step": 3200 }, { "epoch": 612.0, "learning_rate": 3.9577777777777785e-06, "loss": 0.2278, "step": 3225 }, { "epoch": 624.0, "learning_rate": 3.9022222222222225e-06, "loss": 0.2321, "step": 3250 }, { "epoch": 637.0, "learning_rate": 3.8466666666666665e-06, "loss": 0.2323, "step": 3275 }, { "epoch": 649.0, "learning_rate": 3.7911111111111114e-06, "loss": 0.2405, "step": 3300 }, { "epoch": 662.0, "learning_rate": 3.7355555555555555e-06, "loss": 0.2312, "step": 3325 }, { "epoch": 674.0, "learning_rate": 3.6800000000000003e-06, "loss": 0.2362, "step": 3350 }, { "epoch": 687.0, "learning_rate": 3.624444444444445e-06, "loss": 0.2467, "step": 3375 }, { "epoch": 699.0, "learning_rate": 3.568888888888889e-06, "loss": 0.2489, "step": 3400 }, { "epoch": 712.0, "learning_rate": 3.5133333333333337e-06, "loss": 0.2493, "step": 3425 }, { "epoch": 724.0, "learning_rate": 3.457777777777778e-06, "loss": 0.2391, "step": 3450 }, { "epoch": 737.0, "learning_rate": 3.4022222222222222e-06, "loss": 0.2163, "step": 3475 }, { "epoch": 749.0, "learning_rate": 3.346666666666667e-06, "loss": 0.2142, "step": 3500 }, { "epoch": 762.0, "learning_rate": 3.2911111111111116e-06, "loss": 0.2288, "step": 3525 }, { "epoch": 774.0, "learning_rate": 3.2355555555555556e-06, "loss": 0.2183, "step": 3550 }, { "epoch": 787.0, "learning_rate": 3.1800000000000005e-06, "loss": 0.2307, "step": 3575 }, { "epoch": 799.0, "learning_rate": 3.124444444444445e-06, "loss": 0.225, "step": 3600 }, { "epoch": 812.0, "learning_rate": 3.068888888888889e-06, "loss": 0.237, "step": 3625 }, { "epoch": 824.0, "learning_rate": 3.013333333333334e-06, "loss": 0.2174, "step": 3650 }, { "epoch": 837.0, "learning_rate": 2.957777777777778e-06, "loss": 0.2191, "step": 3675 }, { "epoch": 849.0, "learning_rate": 2.9022222222222223e-06, "loss": 0.21, "step": 3700 }, { "epoch": 862.0, "learning_rate": 2.8466666666666672e-06, "loss": 0.2138, "step": 3725 }, { "epoch": 874.0, "learning_rate": 2.7911111111111113e-06, "loss": 0.2369, "step": 3750 }, { "epoch": 887.0, "learning_rate": 2.7355555555555557e-06, "loss": 0.2073, "step": 3775 }, { "epoch": 899.0, "learning_rate": 2.68e-06, "loss": 0.2411, "step": 3800 }, { "epoch": 912.0, "learning_rate": 2.6244444444444446e-06, "loss": 0.2131, "step": 3825 }, { "epoch": 924.0, "learning_rate": 2.568888888888889e-06, "loss": 0.2209, "step": 3850 }, { "epoch": 937.0, "learning_rate": 2.5133333333333336e-06, "loss": 0.2271, "step": 3875 }, { "epoch": 949.0, "learning_rate": 2.457777777777778e-06, "loss": 0.2488, "step": 3900 }, { "epoch": 962.0, "learning_rate": 2.4044444444444446e-06, "loss": 0.2218, "step": 3925 }, { "epoch": 974.0, "learning_rate": 2.348888888888889e-06, "loss": 0.2298, "step": 3950 }, { "epoch": 987.0, "learning_rate": 2.2933333333333335e-06, "loss": 0.2166, "step": 3975 }, { "epoch": 999.0, "learning_rate": 2.237777777777778e-06, "loss": 0.1946, "step": 4000 }, { "epoch": 12.0, "learning_rate": 2.1844444444444446e-06, "loss": 0.3156, "step": 4025 }, { "epoch": 24.0, "learning_rate": 2.1311111111111112e-06, "loss": 0.3136, "step": 4050 }, { "epoch": 37.0, "learning_rate": 2.0755555555555557e-06, "loss": 0.3256, "step": 4075 }, { "epoch": 49.0, "learning_rate": 2.02e-06, "loss": 0.3073, "step": 4100 }, { "epoch": 62.0, "learning_rate": 1.9644444444444446e-06, "loss": 0.299, "step": 4125 }, { "epoch": 74.0, "learning_rate": 1.908888888888889e-06, "loss": 0.2892, "step": 4150 }, { "epoch": 87.0, "learning_rate": 1.8533333333333333e-06, "loss": 0.3021, "step": 4175 }, { "epoch": 99.0, "learning_rate": 1.797777777777778e-06, "loss": 0.31, "step": 4200 }, { "epoch": 112.0, "learning_rate": 1.7422222222222224e-06, "loss": 0.2901, "step": 4225 }, { "epoch": 124.0, "learning_rate": 1.6866666666666667e-06, "loss": 0.3048, "step": 4250 }, { "epoch": 137.0, "learning_rate": 1.6311111111111114e-06, "loss": 0.2897, "step": 4275 }, { "epoch": 149.0, "learning_rate": 1.5755555555555558e-06, "loss": 0.3163, "step": 4300 }, { "epoch": 162.0, "learning_rate": 1.52e-06, "loss": 0.3215, "step": 4325 }, { "epoch": 174.0, "learning_rate": 1.4644444444444445e-06, "loss": 0.3146, "step": 4350 }, { "epoch": 187.0, "learning_rate": 1.4088888888888892e-06, "loss": 0.2884, "step": 4375 }, { "epoch": 199.0, "learning_rate": 1.3533333333333334e-06, "loss": 0.3121, "step": 4400 }, { "epoch": 212.0, "learning_rate": 1.2977777777777779e-06, "loss": 0.3078, "step": 4425 }, { "epoch": 224.0, "learning_rate": 1.2422222222222224e-06, "loss": 0.3164, "step": 4450 }, { "epoch": 237.0, "learning_rate": 1.1866666666666668e-06, "loss": 0.3028, "step": 4475 }, { "epoch": 249.0, "learning_rate": 1.131111111111111e-06, "loss": 0.3065, "step": 4500 }, { "epoch": 262.0, "learning_rate": 1.0755555555555557e-06, "loss": 0.2862, "step": 4525 }, { "epoch": 274.0, "learning_rate": 1.02e-06, "loss": 0.3, "step": 4550 }, { "epoch": 287.0, "learning_rate": 9.644444444444444e-07, "loss": 0.2896, "step": 4575 }, { "epoch": 299.0, "learning_rate": 9.08888888888889e-07, "loss": 0.2749, "step": 4600 }, { "epoch": 312.0, "learning_rate": 8.533333333333334e-07, "loss": 0.2895, "step": 4625 }, { "epoch": 324.0, "learning_rate": 7.977777777777779e-07, "loss": 0.2751, "step": 4650 }, { "epoch": 337.0, "learning_rate": 7.422222222222223e-07, "loss": 0.2987, "step": 4675 }, { "epoch": 349.0, "learning_rate": 6.866666666666667e-07, "loss": 0.2928, "step": 4700 }, { "epoch": 362.0, "learning_rate": 6.311111111111112e-07, "loss": 0.2992, "step": 4725 }, { "epoch": 374.0, "learning_rate": 5.755555555555555e-07, "loss": 0.2988, "step": 4750 }, { "epoch": 387.0, "learning_rate": 5.2e-07, "loss": 0.2959, "step": 4775 }, { "epoch": 399.0, "learning_rate": 4.6444444444444446e-07, "loss": 0.2812, "step": 4800 }, { "epoch": 412.0, "learning_rate": 4.0888888888888897e-07, "loss": 0.3056, "step": 4825 }, { "epoch": 424.0, "learning_rate": 3.533333333333334e-07, "loss": 0.2853, "step": 4850 }, { "epoch": 437.0, "learning_rate": 2.977777777777778e-07, "loss": 0.2939, "step": 4875 }, { "epoch": 449.0, "learning_rate": 2.4222222222222224e-07, "loss": 0.2945, "step": 4900 }, { "epoch": 462.0, "learning_rate": 1.866666666666667e-07, "loss": 0.3002, "step": 4925 }, { "epoch": 474.0, "learning_rate": 1.3111111111111113e-07, "loss": 0.2956, "step": 4950 }, { "epoch": 487.0, "learning_rate": 7.555555555555556e-08, "loss": 0.2949, "step": 4975 }, { "epoch": 499.0, "learning_rate": 2e-08, "loss": 0.291, "step": 5000 } ], "max_steps": 5000, "num_train_epochs": 9223372036854775807, "total_flos": 9.234732878187725e+19, "trial_name": null, "trial_params": null }