{ "best_metric": null, "best_model_checkpoint": null, "epoch": 1171.4285714285713, "eval_steps": 500, "global_step": 8200, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 7.14, "learning_rate": 9.945238095238095e-05, "loss": 2.1185, "step": 50 }, { "epoch": 14.29, "learning_rate": 9.885714285714286e-05, "loss": 1.3056, "step": 100 }, { "epoch": 21.43, "learning_rate": 9.826190476190476e-05, "loss": 0.9754, "step": 150 }, { "epoch": 28.57, "learning_rate": 9.767857142857144e-05, "loss": 0.7447, "step": 200 }, { "epoch": 35.71, "learning_rate": 9.708333333333334e-05, "loss": 0.6164, "step": 250 }, { "epoch": 42.86, "learning_rate": 9.648809523809524e-05, "loss": 0.5674, "step": 300 }, { "epoch": 50.0, "learning_rate": 9.589285714285715e-05, "loss": 0.5373, "step": 350 }, { "epoch": 57.14, "learning_rate": 9.529761904761905e-05, "loss": 0.5459, "step": 400 }, { "epoch": 64.29, "learning_rate": 9.470238095238096e-05, "loss": 0.4995, "step": 450 }, { "epoch": 71.43, "learning_rate": 9.410714285714286e-05, "loss": 0.4877, "step": 500 }, { "epoch": 78.57, "learning_rate": 9.351190476190477e-05, "loss": 0.5287, "step": 550 }, { "epoch": 85.71, "learning_rate": 9.291666666666667e-05, "loss": 0.4737, "step": 600 }, { "epoch": 92.86, "learning_rate": 9.232142857142858e-05, "loss": 0.4594, "step": 650 }, { "epoch": 100.0, "learning_rate": 9.172619047619048e-05, "loss": 0.4416, "step": 700 }, { "epoch": 107.14, "learning_rate": 9.113095238095238e-05, "loss": 0.446, "step": 750 }, { "epoch": 114.29, "learning_rate": 9.053571428571429e-05, "loss": 0.4313, "step": 800 }, { "epoch": 121.43, "learning_rate": 8.99404761904762e-05, "loss": 0.421, "step": 850 }, { "epoch": 128.57, "learning_rate": 8.93452380952381e-05, "loss": 0.4148, "step": 900 }, { "epoch": 135.71, "learning_rate": 8.875e-05, "loss": 0.4191, "step": 950 }, { "epoch": 142.86, "learning_rate": 8.815476190476191e-05, "loss": 0.4013, "step": 1000 }, { "epoch": 150.0, "learning_rate": 8.755952380952381e-05, "loss": 0.4001, "step": 1050 }, { "epoch": 157.14, "learning_rate": 8.696428571428572e-05, "loss": 0.3944, "step": 1100 }, { "epoch": 164.29, "learning_rate": 8.636904761904763e-05, "loss": 0.3911, "step": 1150 }, { "epoch": 171.43, "learning_rate": 8.577380952380952e-05, "loss": 0.3679, "step": 1200 }, { "epoch": 178.57, "learning_rate": 8.517857142857143e-05, "loss": 0.3749, "step": 1250 }, { "epoch": 185.71, "learning_rate": 8.458333333333333e-05, "loss": 0.3781, "step": 1300 }, { "epoch": 192.86, "learning_rate": 8.398809523809524e-05, "loss": 0.3628, "step": 1350 }, { "epoch": 200.0, "learning_rate": 8.339285714285716e-05, "loss": 0.3614, "step": 1400 }, { "epoch": 207.14, "learning_rate": 8.279761904761905e-05, "loss": 0.3433, "step": 1450 }, { "epoch": 214.29, "learning_rate": 8.220238095238095e-05, "loss": 0.331, "step": 1500 }, { "epoch": 221.43, "learning_rate": 8.160714285714286e-05, "loss": 0.3128, "step": 1550 }, { "epoch": 228.57, "learning_rate": 8.101190476190476e-05, "loss": 0.3186, "step": 1600 }, { "epoch": 235.71, "learning_rate": 8.041666666666668e-05, "loss": 0.2911, "step": 1650 }, { "epoch": 242.86, "learning_rate": 7.982142857142858e-05, "loss": 0.2851, "step": 1700 }, { "epoch": 250.0, "learning_rate": 7.922619047619047e-05, "loss": 0.2748, "step": 1750 }, { "epoch": 257.14, "learning_rate": 7.863095238095238e-05, "loss": 0.2614, "step": 1800 }, { "epoch": 264.29, "learning_rate": 7.803571428571428e-05, "loss": 0.2631, "step": 1850 }, { "epoch": 271.43, "learning_rate": 7.74404761904762e-05, "loss": 0.2467, "step": 1900 }, { "epoch": 278.57, "learning_rate": 7.68452380952381e-05, "loss": 0.2372, "step": 1950 }, { "epoch": 285.71, "learning_rate": 7.625e-05, "loss": 0.23, "step": 2000 }, { "epoch": 292.86, "learning_rate": 7.56547619047619e-05, "loss": 0.2314, "step": 2050 }, { "epoch": 300.0, "learning_rate": 7.505952380952382e-05, "loss": 0.2218, "step": 2100 }, { "epoch": 307.14, "learning_rate": 7.446428571428572e-05, "loss": 0.2165, "step": 2150 }, { "epoch": 314.29, "learning_rate": 7.386904761904763e-05, "loss": 0.2262, "step": 2200 }, { "epoch": 321.43, "learning_rate": 7.327380952380952e-05, "loss": 0.22, "step": 2250 }, { "epoch": 328.57, "learning_rate": 7.267857142857142e-05, "loss": 0.2099, "step": 2300 }, { "epoch": 335.71, "learning_rate": 7.208333333333334e-05, "loss": 0.2138, "step": 2350 }, { "epoch": 342.86, "learning_rate": 7.148809523809525e-05, "loss": 0.2064, "step": 2400 }, { "epoch": 350.0, "learning_rate": 7.089285714285715e-05, "loss": 0.197, "step": 2450 }, { "epoch": 357.14, "learning_rate": 7.030952380952381e-05, "loss": 0.191, "step": 2500 }, { "epoch": 364.29, "learning_rate": 6.971428571428572e-05, "loss": 0.1771, "step": 2550 }, { "epoch": 371.43, "learning_rate": 6.911904761904762e-05, "loss": 0.1912, "step": 2600 }, { "epoch": 378.57, "learning_rate": 6.852380952380953e-05, "loss": 0.1597, "step": 2650 }, { "epoch": 385.71, "learning_rate": 6.792857142857143e-05, "loss": 0.1584, "step": 2700 }, { "epoch": 392.86, "learning_rate": 6.733333333333333e-05, "loss": 0.1518, "step": 2750 }, { "epoch": 400.0, "learning_rate": 6.673809523809524e-05, "loss": 0.1492, "step": 2800 }, { "epoch": 407.14, "learning_rate": 6.614285714285716e-05, "loss": 0.1609, "step": 2850 }, { "epoch": 414.29, "learning_rate": 6.554761904761905e-05, "loss": 0.1424, "step": 2900 }, { "epoch": 421.43, "learning_rate": 6.495238095238095e-05, "loss": 0.1405, "step": 2950 }, { "epoch": 428.57, "learning_rate": 6.435714285714286e-05, "loss": 0.1352, "step": 3000 }, { "epoch": 435.71, "learning_rate": 6.376190476190476e-05, "loss": 0.1286, "step": 3050 }, { "epoch": 442.86, "learning_rate": 6.316666666666668e-05, "loss": 0.1243, "step": 3100 }, { "epoch": 450.0, "learning_rate": 6.257142857142857e-05, "loss": 0.1237, "step": 3150 }, { "epoch": 457.14, "learning_rate": 6.197619047619048e-05, "loss": 0.1232, "step": 3200 }, { "epoch": 464.29, "learning_rate": 6.138095238095238e-05, "loss": 0.1226, "step": 3250 }, { "epoch": 471.43, "learning_rate": 6.078571428571429e-05, "loss": 0.1147, "step": 3300 }, { "epoch": 478.57, "learning_rate": 6.0190476190476195e-05, "loss": 0.1161, "step": 3350 }, { "epoch": 485.71, "learning_rate": 5.959523809523809e-05, "loss": 0.1208, "step": 3400 }, { "epoch": 492.86, "learning_rate": 5.9e-05, "loss": 0.1224, "step": 3450 }, { "epoch": 500.0, "learning_rate": 5.840476190476191e-05, "loss": 0.1116, "step": 3500 }, { "epoch": 507.14, "learning_rate": 5.780952380952381e-05, "loss": 0.1074, "step": 3550 }, { "epoch": 514.29, "learning_rate": 5.721428571428572e-05, "loss": 0.1036, "step": 3600 }, { "epoch": 521.43, "learning_rate": 5.661904761904763e-05, "loss": 0.1074, "step": 3650 }, { "epoch": 528.57, "learning_rate": 5.602380952380952e-05, "loss": 0.1042, "step": 3700 }, { "epoch": 535.71, "learning_rate": 5.542857142857143e-05, "loss": 0.1074, "step": 3750 }, { "epoch": 542.86, "learning_rate": 5.4833333333333336e-05, "loss": 0.1033, "step": 3800 }, { "epoch": 550.0, "learning_rate": 5.423809523809524e-05, "loss": 0.0996, "step": 3850 }, { "epoch": 557.14, "learning_rate": 5.364285714285715e-05, "loss": 0.0969, "step": 3900 }, { "epoch": 564.29, "learning_rate": 5.304761904761905e-05, "loss": 0.0943, "step": 3950 }, { "epoch": 571.43, "learning_rate": 5.2452380952380954e-05, "loss": 0.1012, "step": 4000 }, { "epoch": 578.57, "learning_rate": 5.185714285714286e-05, "loss": 0.1011, "step": 4050 }, { "epoch": 585.71, "learning_rate": 5.126190476190477e-05, "loss": 0.0981, "step": 4100 }, { "epoch": 592.86, "learning_rate": 5.0666666666666674e-05, "loss": 0.0935, "step": 4150 }, { "epoch": 600.0, "learning_rate": 5.007142857142857e-05, "loss": 0.0922, "step": 4200 }, { "epoch": 607.14, "learning_rate": 4.9476190476190476e-05, "loss": 0.0935, "step": 4250 }, { "epoch": 614.29, "learning_rate": 4.888095238095238e-05, "loss": 0.0955, "step": 4300 }, { "epoch": 621.43, "learning_rate": 4.828571428571429e-05, "loss": 0.0909, "step": 4350 }, { "epoch": 628.57, "learning_rate": 4.769047619047619e-05, "loss": 0.0886, "step": 4400 }, { "epoch": 635.71, "learning_rate": 4.70952380952381e-05, "loss": 0.0886, "step": 4450 }, { "epoch": 642.86, "learning_rate": 4.6500000000000005e-05, "loss": 0.0884, "step": 4500 }, { "epoch": 650.0, "learning_rate": 4.591666666666667e-05, "loss": 0.0939, "step": 4550 }, { "epoch": 657.14, "learning_rate": 4.532142857142858e-05, "loss": 0.0879, "step": 4600 }, { "epoch": 664.29, "learning_rate": 4.4726190476190476e-05, "loss": 0.0867, "step": 4650 }, { "epoch": 671.43, "learning_rate": 4.413095238095239e-05, "loss": 0.0858, "step": 4700 }, { "epoch": 678.57, "learning_rate": 4.3535714285714285e-05, "loss": 0.0882, "step": 4750 }, { "epoch": 685.71, "learning_rate": 4.294047619047619e-05, "loss": 0.085, "step": 4800 }, { "epoch": 692.86, "learning_rate": 4.23452380952381e-05, "loss": 0.0787, "step": 4850 }, { "epoch": 700.0, "learning_rate": 4.175e-05, "loss": 0.0798, "step": 4900 }, { "epoch": 707.14, "learning_rate": 4.115476190476191e-05, "loss": 0.0805, "step": 4950 }, { "epoch": 714.29, "learning_rate": 4.0559523809523814e-05, "loss": 0.0796, "step": 5000 }, { "epoch": 721.43, "learning_rate": 3.996428571428571e-05, "loss": 0.084, "step": 5050 }, { "epoch": 728.57, "learning_rate": 3.936904761904762e-05, "loss": 0.0826, "step": 5100 }, { "epoch": 735.71, "learning_rate": 3.877380952380952e-05, "loss": 0.0807, "step": 5150 }, { "epoch": 742.86, "learning_rate": 3.817857142857143e-05, "loss": 0.0744, "step": 5200 }, { "epoch": 750.0, "learning_rate": 3.7583333333333337e-05, "loss": 0.0845, "step": 5250 }, { "epoch": 757.14, "learning_rate": 3.698809523809524e-05, "loss": 0.0785, "step": 5300 }, { "epoch": 764.29, "learning_rate": 3.6392857142857146e-05, "loss": 0.0786, "step": 5350 }, { "epoch": 771.43, "learning_rate": 3.579761904761905e-05, "loss": 0.0743, "step": 5400 }, { "epoch": 778.57, "learning_rate": 3.5202380952380955e-05, "loss": 0.0778, "step": 5450 }, { "epoch": 785.71, "learning_rate": 3.460714285714286e-05, "loss": 0.0706, "step": 5500 }, { "epoch": 792.86, "learning_rate": 3.4011904761904764e-05, "loss": 0.0735, "step": 5550 }, { "epoch": 800.0, "learning_rate": 3.341666666666667e-05, "loss": 0.0727, "step": 5600 }, { "epoch": 807.14, "learning_rate": 3.282142857142857e-05, "loss": 0.0779, "step": 5650 }, { "epoch": 814.29, "learning_rate": 3.222619047619048e-05, "loss": 0.0708, "step": 5700 }, { "epoch": 821.43, "learning_rate": 3.163095238095238e-05, "loss": 0.0708, "step": 5750 }, { "epoch": 828.57, "learning_rate": 3.1035714285714286e-05, "loss": 0.0709, "step": 5800 }, { "epoch": 835.71, "learning_rate": 3.044047619047619e-05, "loss": 0.0719, "step": 5850 }, { "epoch": 842.86, "learning_rate": 2.98452380952381e-05, "loss": 0.0721, "step": 5900 }, { "epoch": 850.0, "learning_rate": 2.925e-05, "loss": 0.0685, "step": 5950 }, { "epoch": 857.14, "learning_rate": 2.8654761904761907e-05, "loss": 0.0711, "step": 6000 }, { "epoch": 864.29, "learning_rate": 2.8059523809523812e-05, "loss": 0.0645, "step": 6050 }, { "epoch": 871.43, "learning_rate": 2.7464285714285713e-05, "loss": 0.0694, "step": 6100 }, { "epoch": 878.57, "learning_rate": 2.686904761904762e-05, "loss": 0.0636, "step": 6150 }, { "epoch": 885.71, "learning_rate": 2.6273809523809522e-05, "loss": 0.0635, "step": 6200 }, { "epoch": 892.86, "learning_rate": 2.567857142857143e-05, "loss": 0.0636, "step": 6250 }, { "epoch": 900.0, "learning_rate": 2.5083333333333338e-05, "loss": 0.0676, "step": 6300 }, { "epoch": 907.14, "learning_rate": 2.4488095238095242e-05, "loss": 0.0614, "step": 6350 }, { "epoch": 914.29, "learning_rate": 2.3892857142857143e-05, "loss": 0.0641, "step": 6400 }, { "epoch": 921.43, "learning_rate": 2.3297619047619048e-05, "loss": 0.0649, "step": 6450 }, { "epoch": 928.57, "learning_rate": 2.2702380952380952e-05, "loss": 0.0627, "step": 6500 }, { "epoch": 935.71, "learning_rate": 2.2107142857142857e-05, "loss": 0.062, "step": 6550 }, { "epoch": 942.86, "learning_rate": 2.1511904761904765e-05, "loss": 0.0621, "step": 6600 }, { "epoch": 950.0, "learning_rate": 2.091666666666667e-05, "loss": 0.061, "step": 6650 }, { "epoch": 957.14, "learning_rate": 2.032142857142857e-05, "loss": 0.0603, "step": 6700 }, { "epoch": 964.29, "learning_rate": 1.9726190476190475e-05, "loss": 0.0586, "step": 6750 }, { "epoch": 971.43, "learning_rate": 1.9130952380952383e-05, "loss": 0.0586, "step": 6800 }, { "epoch": 978.57, "learning_rate": 1.8535714285714287e-05, "loss": 0.0582, "step": 6850 }, { "epoch": 985.71, "learning_rate": 1.794047619047619e-05, "loss": 0.0586, "step": 6900 }, { "epoch": 992.86, "learning_rate": 1.7345238095238096e-05, "loss": 0.0553, "step": 6950 }, { "epoch": 1000.0, "learning_rate": 1.675e-05, "loss": 0.056, "step": 7000 }, { "epoch": 1007.14, "learning_rate": 1.6166666666666665e-05, "loss": 0.0596, "step": 7050 }, { "epoch": 1014.29, "learning_rate": 1.5571428571428573e-05, "loss": 0.055, "step": 7100 }, { "epoch": 1021.43, "learning_rate": 1.4976190476190476e-05, "loss": 0.0545, "step": 7150 }, { "epoch": 1028.57, "learning_rate": 1.438095238095238e-05, "loss": 0.054, "step": 7200 }, { "epoch": 1035.71, "learning_rate": 1.3785714285714285e-05, "loss": 0.0543, "step": 7250 }, { "epoch": 1042.86, "learning_rate": 1.3190476190476191e-05, "loss": 0.0537, "step": 7300 }, { "epoch": 1050.0, "learning_rate": 1.2595238095238096e-05, "loss": 0.0553, "step": 7350 }, { "epoch": 1057.14, "learning_rate": 1.2e-05, "loss": 0.0532, "step": 7400 }, { "epoch": 1064.29, "learning_rate": 1.1404761904761905e-05, "loss": 0.053, "step": 7450 }, { "epoch": 1071.43, "learning_rate": 1.080952380952381e-05, "loss": 0.0531, "step": 7500 }, { "epoch": 1078.57, "learning_rate": 1.0214285714285715e-05, "loss": 0.0546, "step": 7550 }, { "epoch": 1085.71, "learning_rate": 9.61904761904762e-06, "loss": 0.0517, "step": 7600 }, { "epoch": 1092.86, "learning_rate": 9.023809523809524e-06, "loss": 0.0527, "step": 7650 }, { "epoch": 1100.0, "learning_rate": 8.428571428571429e-06, "loss": 0.0533, "step": 7700 }, { "epoch": 1107.14, "learning_rate": 7.833333333333333e-06, "loss": 0.0537, "step": 7750 }, { "epoch": 1114.29, "learning_rate": 7.238095238095238e-06, "loss": 0.0517, "step": 7800 }, { "epoch": 1121.43, "learning_rate": 6.642857142857144e-06, "loss": 0.0517, "step": 7850 }, { "epoch": 1128.57, "learning_rate": 6.047619047619048e-06, "loss": 0.0506, "step": 7900 }, { "epoch": 1135.71, "learning_rate": 5.452380952380952e-06, "loss": 0.0512, "step": 7950 }, { "epoch": 1142.86, "learning_rate": 4.857142857142858e-06, "loss": 0.0482, "step": 8000 }, { "epoch": 1150.0, "learning_rate": 4.261904761904762e-06, "loss": 0.0488, "step": 8050 }, { "epoch": 1157.14, "learning_rate": 3.666666666666667e-06, "loss": 0.0493, "step": 8100 }, { "epoch": 1164.29, "learning_rate": 3.0714285714285715e-06, "loss": 0.0492, "step": 8150 }, { "epoch": 1171.43, "learning_rate": 2.4761904761904764e-06, "loss": 0.0491, "step": 8200 } ], "logging_steps": 50, "max_steps": 8400, "num_train_epochs": 1200, "save_steps": 200, "total_flos": 4.803400934061458e+18, "trial_name": null, "trial_params": null }