{ "best_metric": null, "best_model_checkpoint": null, "epoch": 18.293569925912376, "global_step": 250000, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.04, "learning_rate": 2.0000000000000003e-06, "loss": 8.4479, "step": 500 }, { "epoch": 0.07, "learning_rate": 4.000000000000001e-06, "loss": 7.3835, "step": 1000 }, { "epoch": 0.11, "learning_rate": 6e-06, "loss": 6.7134, "step": 1500 }, { "epoch": 0.15, "learning_rate": 8.000000000000001e-06, "loss": 6.2752, "step": 2000 }, { "epoch": 0.18, "learning_rate": 1e-05, "loss": 5.9464, "step": 2500 }, { "epoch": 0.22, "learning_rate": 1.2e-05, "loss": 5.643, "step": 3000 }, { "epoch": 0.26, "learning_rate": 1.4000000000000001e-05, "loss": 5.3728, "step": 3500 }, { "epoch": 0.29, "learning_rate": 1.6000000000000003e-05, "loss": 5.1453, "step": 4000 }, { "epoch": 0.33, "learning_rate": 1.8e-05, "loss": 4.9466, "step": 4500 }, { "epoch": 0.37, "learning_rate": 2e-05, "loss": 4.7762, "step": 5000 }, { "epoch": 0.4, "learning_rate": 2.2000000000000003e-05, "loss": 4.6241, "step": 5500 }, { "epoch": 0.44, "learning_rate": 2.4e-05, "loss": 4.4917, "step": 6000 }, { "epoch": 0.48, "learning_rate": 2.6000000000000002e-05, "loss": 4.3683, "step": 6500 }, { "epoch": 0.51, "learning_rate": 2.8000000000000003e-05, "loss": 4.2491, "step": 7000 }, { "epoch": 0.55, "learning_rate": 3e-05, "loss": 4.1498, "step": 7500 }, { "epoch": 0.59, "learning_rate": 3.2000000000000005e-05, "loss": 4.0539, "step": 8000 }, { "epoch": 0.62, "learning_rate": 3.4000000000000007e-05, "loss": 3.968, "step": 8500 }, { "epoch": 0.66, "learning_rate": 3.6e-05, "loss": 3.8855, "step": 9000 }, { "epoch": 0.7, "learning_rate": 3.8e-05, "loss": 3.818, "step": 9500 }, { "epoch": 0.73, "learning_rate": 4e-05, "loss": 3.7477, "step": 10000 }, { "epoch": 0.77, "learning_rate": 4.2e-05, "loss": 3.6863, "step": 10500 }, { "epoch": 0.8, "learning_rate": 4.4000000000000006e-05, "loss": 3.628, "step": 11000 }, { "epoch": 0.84, "learning_rate": 4.600000000000001e-05, "loss": 3.5747, "step": 11500 }, { "epoch": 0.88, "learning_rate": 4.8e-05, "loss": 3.5235, "step": 12000 }, { "epoch": 0.91, "learning_rate": 5e-05, "loss": 3.4808, "step": 12500 }, { "epoch": 0.95, "learning_rate": 5.2000000000000004e-05, "loss": 3.4356, "step": 13000 }, { "epoch": 0.99, "learning_rate": 5.4000000000000005e-05, "loss": 3.3986, "step": 13500 }, { "epoch": 1.02, "learning_rate": 5.6000000000000006e-05, "loss": 3.364, "step": 14000 }, { "epoch": 1.06, "learning_rate": 5.8e-05, "loss": 3.3247, "step": 14500 }, { "epoch": 1.1, "learning_rate": 6e-05, "loss": 3.2949, "step": 15000 }, { "epoch": 1.13, "learning_rate": 6.2e-05, "loss": 3.2666, "step": 15500 }, { "epoch": 1.17, "learning_rate": 6.400000000000001e-05, "loss": 3.2398, "step": 16000 }, { "epoch": 1.21, "learning_rate": 6.6e-05, "loss": 3.2131, "step": 16500 }, { "epoch": 1.24, "learning_rate": 6.800000000000001e-05, "loss": 3.1887, "step": 17000 }, { "epoch": 1.28, "learning_rate": 7e-05, "loss": 3.1665, "step": 17500 }, { "epoch": 1.32, "learning_rate": 7.2e-05, "loss": 3.1422, "step": 18000 }, { "epoch": 1.35, "learning_rate": 7.4e-05, "loss": 3.1238, "step": 18500 }, { "epoch": 1.39, "learning_rate": 7.6e-05, "loss": 3.1004, "step": 19000 }, { "epoch": 1.43, "learning_rate": 7.800000000000001e-05, "loss": 3.0856, "step": 19500 }, { "epoch": 1.46, "learning_rate": 8e-05, "loss": 3.0682, "step": 20000 }, { "epoch": 1.5, "learning_rate": 8.2e-05, "loss": 3.0488, "step": 20500 }, { "epoch": 1.54, "learning_rate": 8.4e-05, "loss": 3.0314, "step": 21000 }, { "epoch": 1.57, "learning_rate": 8.6e-05, "loss": 3.0153, "step": 21500 }, { "epoch": 1.61, "learning_rate": 8.800000000000001e-05, "loss": 3.0039, "step": 22000 }, { "epoch": 1.65, "learning_rate": 9e-05, "loss": 2.9861, "step": 22500 }, { "epoch": 1.68, "learning_rate": 9.200000000000001e-05, "loss": 2.9759, "step": 23000 }, { "epoch": 1.72, "learning_rate": 9.4e-05, "loss": 2.9631, "step": 23500 }, { "epoch": 1.76, "learning_rate": 9.6e-05, "loss": 2.9523, "step": 24000 }, { "epoch": 1.79, "learning_rate": 9.8e-05, "loss": 2.9401, "step": 24500 }, { "epoch": 1.83, "learning_rate": 0.0001, "loss": 2.9267, "step": 25000 }, { "epoch": 1.87, "learning_rate": 9.977777777777779e-05, "loss": 2.9162, "step": 25500 }, { "epoch": 1.9, "learning_rate": 9.955555555555556e-05, "loss": 2.9032, "step": 26000 }, { "epoch": 1.94, "learning_rate": 9.933333333333334e-05, "loss": 2.8907, "step": 26500 }, { "epoch": 1.98, "learning_rate": 9.911111111111112e-05, "loss": 2.8838, "step": 27000 }, { "epoch": 2.01, "learning_rate": 9.888888888888889e-05, "loss": 2.8742, "step": 27500 }, { "epoch": 2.05, "learning_rate": 9.866666666666668e-05, "loss": 2.8601, "step": 28000 }, { "epoch": 2.09, "learning_rate": 9.844444444444444e-05, "loss": 2.8534, "step": 28500 }, { "epoch": 2.12, "learning_rate": 9.822222222222223e-05, "loss": 2.8455, "step": 29000 }, { "epoch": 2.16, "learning_rate": 9.8e-05, "loss": 2.8346, "step": 29500 }, { "epoch": 2.2, "learning_rate": 9.777777777777778e-05, "loss": 2.8289, "step": 30000 }, { "epoch": 2.23, "learning_rate": 9.755555555555555e-05, "loss": 2.8203, "step": 30500 }, { "epoch": 2.27, "learning_rate": 9.733333333333335e-05, "loss": 2.8138, "step": 31000 }, { "epoch": 2.3, "learning_rate": 9.711111111111111e-05, "loss": 2.8068, "step": 31500 }, { "epoch": 2.34, "learning_rate": 9.68888888888889e-05, "loss": 2.7989, "step": 32000 }, { "epoch": 2.38, "learning_rate": 9.666666666666667e-05, "loss": 2.7923, "step": 32500 }, { "epoch": 2.41, "learning_rate": 9.644444444444445e-05, "loss": 2.7847, "step": 33000 }, { "epoch": 2.45, "learning_rate": 9.622222222222222e-05, "loss": 2.7818, "step": 33500 }, { "epoch": 2.49, "learning_rate": 9.6e-05, "loss": 2.7735, "step": 34000 }, { "epoch": 2.52, "learning_rate": 9.577777777777777e-05, "loss": 2.7691, "step": 34500 }, { "epoch": 2.56, "learning_rate": 9.555555555555557e-05, "loss": 2.7585, "step": 35000 }, { "epoch": 2.6, "learning_rate": 9.533333333333334e-05, "loss": 2.7609, "step": 35500 }, { "epoch": 2.63, "learning_rate": 9.511111111111112e-05, "loss": 2.752, "step": 36000 }, { "epoch": 2.67, "learning_rate": 9.488888888888889e-05, "loss": 2.7452, "step": 36500 }, { "epoch": 2.71, "learning_rate": 9.466666666666667e-05, "loss": 2.741, "step": 37000 }, { "epoch": 2.74, "learning_rate": 9.444444444444444e-05, "loss": 2.7384, "step": 37500 }, { "epoch": 2.78, "learning_rate": 9.422222222222223e-05, "loss": 2.7332, "step": 38000 }, { "epoch": 2.82, "learning_rate": 9.4e-05, "loss": 2.726, "step": 38500 }, { "epoch": 2.85, "learning_rate": 9.377777777777779e-05, "loss": 2.7218, "step": 39000 }, { "epoch": 2.89, "learning_rate": 9.355555555555556e-05, "loss": 2.7198, "step": 39500 }, { "epoch": 2.93, "learning_rate": 9.333333333333334e-05, "loss": 2.7156, "step": 40000 }, { "epoch": 2.96, "learning_rate": 9.311111111111111e-05, "loss": 2.71, "step": 40500 }, { "epoch": 3.0, "learning_rate": 9.28888888888889e-05, "loss": 2.7099, "step": 41000 }, { "epoch": 3.04, "learning_rate": 9.266666666666666e-05, "loss": 2.7006, "step": 41500 }, { "epoch": 3.07, "learning_rate": 9.244444444444445e-05, "loss": 2.694, "step": 42000 }, { "epoch": 3.11, "learning_rate": 9.222222222222223e-05, "loss": 2.6938, "step": 42500 }, { "epoch": 3.15, "learning_rate": 9.200000000000001e-05, "loss": 2.6878, "step": 43000 }, { "epoch": 3.18, "learning_rate": 9.177777777777778e-05, "loss": 2.6865, "step": 43500 }, { "epoch": 3.22, "learning_rate": 9.155555555555557e-05, "loss": 2.6847, "step": 44000 }, { "epoch": 3.26, "learning_rate": 9.133333333333334e-05, "loss": 2.6811, "step": 44500 }, { "epoch": 3.29, "learning_rate": 9.111111111111112e-05, "loss": 2.6748, "step": 45000 }, { "epoch": 3.33, "learning_rate": 9.088888888888889e-05, "loss": 2.6733, "step": 45500 }, { "epoch": 3.37, "learning_rate": 9.066666666666667e-05, "loss": 2.6708, "step": 46000 }, { "epoch": 3.4, "learning_rate": 9.044444444444445e-05, "loss": 2.6657, "step": 46500 }, { "epoch": 3.44, "learning_rate": 9.022222222222224e-05, "loss": 2.6616, "step": 47000 }, { "epoch": 3.48, "learning_rate": 9e-05, "loss": 2.6555, "step": 47500 }, { "epoch": 3.51, "learning_rate": 8.977777777777779e-05, "loss": 2.6568, "step": 48000 }, { "epoch": 3.55, "learning_rate": 8.955555555555556e-05, "loss": 2.654, "step": 48500 }, { "epoch": 3.59, "learning_rate": 8.933333333333334e-05, "loss": 2.6504, "step": 49000 }, { "epoch": 3.62, "learning_rate": 8.911111111111111e-05, "loss": 2.6498, "step": 49500 }, { "epoch": 3.66, "learning_rate": 8.888888888888889e-05, "loss": 2.646, "step": 50000 }, { "epoch": 3.7, "learning_rate": 8.866666666666668e-05, "loss": 2.6448, "step": 50500 }, { "epoch": 3.73, "learning_rate": 8.844444444444445e-05, "loss": 2.6415, "step": 51000 }, { "epoch": 3.77, "learning_rate": 8.822222222222223e-05, "loss": 2.6395, "step": 51500 }, { "epoch": 3.81, "learning_rate": 8.800000000000001e-05, "loss": 2.6366, "step": 52000 }, { "epoch": 3.84, "learning_rate": 8.777777777777778e-05, "loss": 2.6376, "step": 52500 }, { "epoch": 3.88, "learning_rate": 8.755555555555556e-05, "loss": 2.6319, "step": 53000 }, { "epoch": 3.91, "learning_rate": 8.733333333333333e-05, "loss": 2.6286, "step": 53500 }, { "epoch": 3.95, "learning_rate": 8.711111111111112e-05, "loss": 2.6267, "step": 54000 }, { "epoch": 3.99, "learning_rate": 8.68888888888889e-05, "loss": 2.6256, "step": 54500 }, { "epoch": 4.02, "learning_rate": 8.666666666666667e-05, "loss": 2.6223, "step": 55000 }, { "epoch": 4.06, "learning_rate": 8.644444444444445e-05, "loss": 2.6161, "step": 55500 }, { "epoch": 4.1, "learning_rate": 8.622222222222222e-05, "loss": 2.6128, "step": 56000 }, { "epoch": 4.13, "learning_rate": 8.6e-05, "loss": 2.6145, "step": 56500 }, { "epoch": 4.17, "learning_rate": 8.577777777777777e-05, "loss": 2.6097, "step": 57000 }, { "epoch": 4.21, "learning_rate": 8.555555555555556e-05, "loss": 2.6107, "step": 57500 }, { "epoch": 4.24, "learning_rate": 8.533333333333334e-05, "loss": 2.6069, "step": 58000 }, { "epoch": 4.28, "learning_rate": 8.511111111111112e-05, "loss": 2.605, "step": 58500 }, { "epoch": 4.32, "learning_rate": 8.488888888888889e-05, "loss": 2.6049, "step": 59000 }, { "epoch": 4.35, "learning_rate": 8.466666666666667e-05, "loss": 2.6003, "step": 59500 }, { "epoch": 4.39, "learning_rate": 8.444444444444444e-05, "loss": 2.5996, "step": 60000 }, { "epoch": 4.43, "learning_rate": 8.422222222222223e-05, "loss": 2.5991, "step": 60500 }, { "epoch": 4.46, "learning_rate": 8.4e-05, "loss": 2.5926, "step": 61000 }, { "epoch": 4.5, "learning_rate": 8.377777777777778e-05, "loss": 2.5933, "step": 61500 }, { "epoch": 4.54, "learning_rate": 8.355555555555556e-05, "loss": 2.5903, "step": 62000 }, { "epoch": 4.57, "learning_rate": 8.333333333333334e-05, "loss": 2.5906, "step": 62500 }, { "epoch": 4.61, "learning_rate": 8.311111111111111e-05, "loss": 2.5894, "step": 63000 }, { "epoch": 4.65, "learning_rate": 8.28888888888889e-05, "loss": 2.5848, "step": 63500 }, { "epoch": 4.68, "learning_rate": 8.266666666666667e-05, "loss": 2.5861, "step": 64000 }, { "epoch": 4.72, "learning_rate": 8.244444444444445e-05, "loss": 2.5817, "step": 64500 }, { "epoch": 4.76, "learning_rate": 8.222222222222222e-05, "loss": 2.5794, "step": 65000 }, { "epoch": 4.79, "learning_rate": 8.2e-05, "loss": 2.5794, "step": 65500 }, { "epoch": 4.83, "learning_rate": 8.177777777777778e-05, "loss": 2.5779, "step": 66000 }, { "epoch": 4.87, "learning_rate": 8.155555555555557e-05, "loss": 2.5759, "step": 66500 }, { "epoch": 4.9, "learning_rate": 8.133333333333334e-05, "loss": 2.5747, "step": 67000 }, { "epoch": 4.94, "learning_rate": 8.111111111111112e-05, "loss": 2.5728, "step": 67500 }, { "epoch": 4.98, "learning_rate": 8.088888888888889e-05, "loss": 2.5727, "step": 68000 }, { "epoch": 5.01, "learning_rate": 8.066666666666667e-05, "loss": 2.5702, "step": 68500 }, { "epoch": 5.05, "learning_rate": 8.044444444444444e-05, "loss": 2.5676, "step": 69000 }, { "epoch": 5.09, "learning_rate": 8.022222222222222e-05, "loss": 2.5671, "step": 69500 }, { "epoch": 5.12, "learning_rate": 8e-05, "loss": 2.563, "step": 70000 }, { "epoch": 5.16, "learning_rate": 7.977777777777779e-05, "loss": 2.5645, "step": 70500 }, { "epoch": 5.2, "learning_rate": 7.955555555555556e-05, "loss": 2.5598, "step": 71000 }, { "epoch": 5.23, "learning_rate": 7.933333333333334e-05, "loss": 2.5602, "step": 71500 }, { "epoch": 5.27, "learning_rate": 7.911111111111111e-05, "loss": 2.559, "step": 72000 }, { "epoch": 5.31, "learning_rate": 7.88888888888889e-05, "loss": 2.5575, "step": 72500 }, { "epoch": 5.34, "learning_rate": 7.866666666666666e-05, "loss": 2.5551, "step": 73000 }, { "epoch": 5.38, "learning_rate": 7.844444444444446e-05, "loss": 2.5545, "step": 73500 }, { "epoch": 5.41, "learning_rate": 7.822222222222223e-05, "loss": 2.5522, "step": 74000 }, { "epoch": 5.45, "learning_rate": 7.800000000000001e-05, "loss": 2.5522, "step": 74500 }, { "epoch": 5.49, "learning_rate": 7.777777777777778e-05, "loss": 2.5505, "step": 75000 }, { "epoch": 5.52, "learning_rate": 7.755555555555556e-05, "loss": 2.5484, "step": 75500 }, { "epoch": 5.56, "learning_rate": 7.733333333333333e-05, "loss": 2.549, "step": 76000 }, { "epoch": 5.6, "learning_rate": 7.711111111111112e-05, "loss": 2.5435, "step": 76500 }, { "epoch": 5.63, "learning_rate": 7.688888888888889e-05, "loss": 2.5482, "step": 77000 }, { "epoch": 5.67, "learning_rate": 7.666666666666667e-05, "loss": 2.5448, "step": 77500 }, { "epoch": 5.71, "learning_rate": 7.644444444444445e-05, "loss": 2.5444, "step": 78000 }, { "epoch": 5.74, "learning_rate": 7.622222222222223e-05, "loss": 2.5409, "step": 78500 }, { "epoch": 5.78, "learning_rate": 7.6e-05, "loss": 2.54, "step": 79000 }, { "epoch": 5.82, "learning_rate": 7.577777777777779e-05, "loss": 2.5387, "step": 79500 }, { "epoch": 5.85, "learning_rate": 7.555555555555556e-05, "loss": 2.5386, "step": 80000 }, { "epoch": 5.89, "learning_rate": 7.533333333333334e-05, "loss": 2.5383, "step": 80500 }, { "epoch": 5.93, "learning_rate": 7.511111111111111e-05, "loss": 2.5369, "step": 81000 }, { "epoch": 5.96, "learning_rate": 7.488888888888889e-05, "loss": 2.5345, "step": 81500 }, { "epoch": 6.0, "learning_rate": 7.466666666666667e-05, "loss": 2.5339, "step": 82000 }, { "epoch": 6.04, "learning_rate": 7.444444444444444e-05, "loss": 2.5291, "step": 82500 }, { "epoch": 6.07, "learning_rate": 7.422222222222223e-05, "loss": 2.5299, "step": 83000 }, { "epoch": 6.11, "learning_rate": 7.4e-05, "loss": 2.5299, "step": 83500 }, { "epoch": 6.15, "learning_rate": 7.377777777777778e-05, "loss": 2.5257, "step": 84000 }, { "epoch": 6.18, "learning_rate": 7.355555555555556e-05, "loss": 2.5279, "step": 84500 }, { "epoch": 6.22, "learning_rate": 7.333333333333333e-05, "loss": 2.5258, "step": 85000 }, { "epoch": 6.26, "learning_rate": 7.311111111111111e-05, "loss": 2.5221, "step": 85500 }, { "epoch": 6.29, "learning_rate": 7.28888888888889e-05, "loss": 2.5254, "step": 86000 }, { "epoch": 6.33, "learning_rate": 7.266666666666667e-05, "loss": 2.5215, "step": 86500 }, { "epoch": 6.37, "learning_rate": 7.244444444444445e-05, "loss": 2.5213, "step": 87000 }, { "epoch": 6.4, "learning_rate": 7.222222222222222e-05, "loss": 2.5234, "step": 87500 }, { "epoch": 6.44, "learning_rate": 7.2e-05, "loss": 2.522, "step": 88000 }, { "epoch": 6.48, "learning_rate": 7.177777777777777e-05, "loss": 2.5178, "step": 88500 }, { "epoch": 6.51, "learning_rate": 7.155555555555555e-05, "loss": 2.5207, "step": 89000 }, { "epoch": 6.55, "learning_rate": 7.133333333333334e-05, "loss": 2.5131, "step": 89500 }, { "epoch": 6.59, "learning_rate": 7.111111111111112e-05, "loss": 2.5124, "step": 90000 }, { "epoch": 6.62, "learning_rate": 7.088888888888889e-05, "loss": 2.5151, "step": 90500 }, { "epoch": 6.66, "learning_rate": 7.066666666666667e-05, "loss": 2.5164, "step": 91000 }, { "epoch": 6.7, "learning_rate": 7.044444444444444e-05, "loss": 2.5125, "step": 91500 }, { "epoch": 6.73, "learning_rate": 7.022222222222222e-05, "loss": 2.5123, "step": 92000 }, { "epoch": 6.77, "learning_rate": 7e-05, "loss": 2.5142, "step": 92500 }, { "epoch": 6.81, "learning_rate": 6.977777777777779e-05, "loss": 2.51, "step": 93000 }, { "epoch": 6.84, "learning_rate": 6.955555555555556e-05, "loss": 2.5102, "step": 93500 }, { "epoch": 6.88, "learning_rate": 6.933333333333334e-05, "loss": 2.5092, "step": 94000 }, { "epoch": 6.91, "learning_rate": 6.911111111111111e-05, "loss": 2.5068, "step": 94500 }, { "epoch": 6.95, "learning_rate": 6.88888888888889e-05, "loss": 2.5085, "step": 95000 }, { "epoch": 6.99, "learning_rate": 6.866666666666666e-05, "loss": 2.5045, "step": 95500 }, { "epoch": 7.02, "learning_rate": 6.844444444444445e-05, "loss": 2.5029, "step": 96000 }, { "epoch": 7.06, "learning_rate": 6.822222222222222e-05, "loss": 2.5023, "step": 96500 }, { "epoch": 7.1, "learning_rate": 6.800000000000001e-05, "loss": 2.5015, "step": 97000 }, { "epoch": 7.13, "learning_rate": 6.777777777777778e-05, "loss": 2.5016, "step": 97500 }, { "epoch": 7.17, "learning_rate": 6.755555555555557e-05, "loss": 2.4991, "step": 98000 }, { "epoch": 7.21, "learning_rate": 6.733333333333333e-05, "loss": 2.4965, "step": 98500 }, { "epoch": 7.24, "learning_rate": 6.711111111111112e-05, "loss": 2.4995, "step": 99000 }, { "epoch": 7.28, "learning_rate": 6.688888888888889e-05, "loss": 2.4981, "step": 99500 }, { "epoch": 7.32, "learning_rate": 6.666666666666667e-05, "loss": 2.4986, "step": 100000 }, { "epoch": 7.35, "learning_rate": 6.644444444444444e-05, "loss": 2.4964, "step": 100500 }, { "epoch": 7.39, "learning_rate": 6.622222222222224e-05, "loss": 2.4955, "step": 101000 }, { "epoch": 7.43, "learning_rate": 6.6e-05, "loss": 2.4908, "step": 101500 }, { "epoch": 7.46, "learning_rate": 6.577777777777779e-05, "loss": 2.4937, "step": 102000 }, { "epoch": 7.5, "learning_rate": 6.555555555555556e-05, "loss": 2.4944, "step": 102500 }, { "epoch": 7.54, "learning_rate": 6.533333333333334e-05, "loss": 2.4918, "step": 103000 }, { "epoch": 7.57, "learning_rate": 6.511111111111111e-05, "loss": 2.4925, "step": 103500 }, { "epoch": 7.61, "learning_rate": 6.488888888888889e-05, "loss": 2.4915, "step": 104000 }, { "epoch": 7.65, "learning_rate": 6.466666666666666e-05, "loss": 2.4896, "step": 104500 }, { "epoch": 7.68, "learning_rate": 6.444444444444446e-05, "loss": 2.49, "step": 105000 }, { "epoch": 7.72, "learning_rate": 6.422222222222223e-05, "loss": 2.489, "step": 105500 }, { "epoch": 7.76, "learning_rate": 6.400000000000001e-05, "loss": 2.4854, "step": 106000 }, { "epoch": 7.79, "learning_rate": 6.377777777777778e-05, "loss": 2.4873, "step": 106500 }, { "epoch": 7.83, "learning_rate": 6.355555555555556e-05, "loss": 2.4864, "step": 107000 }, { "epoch": 7.87, "learning_rate": 6.333333333333333e-05, "loss": 2.4879, "step": 107500 }, { "epoch": 7.9, "learning_rate": 6.311111111111112e-05, "loss": 2.486, "step": 108000 }, { "epoch": 7.94, "learning_rate": 6.28888888888889e-05, "loss": 2.4845, "step": 108500 }, { "epoch": 7.98, "learning_rate": 6.266666666666667e-05, "loss": 2.484, "step": 109000 }, { "epoch": 8.01, "learning_rate": 6.244444444444445e-05, "loss": 2.4848, "step": 109500 }, { "epoch": 8.05, "learning_rate": 6.222222222222222e-05, "loss": 2.4794, "step": 110000 }, { "epoch": 8.09, "learning_rate": 6.2e-05, "loss": 2.4809, "step": 110500 }, { "epoch": 8.12, "learning_rate": 6.177777777777779e-05, "loss": 2.4799, "step": 111000 }, { "epoch": 8.16, "learning_rate": 6.155555555555555e-05, "loss": 2.4771, "step": 111500 }, { "epoch": 8.2, "learning_rate": 6.133333333333334e-05, "loss": 2.477, "step": 112000 }, { "epoch": 8.23, "learning_rate": 6.111111111111112e-05, "loss": 2.4769, "step": 112500 }, { "epoch": 8.27, "learning_rate": 6.08888888888889e-05, "loss": 2.4734, "step": 113000 }, { "epoch": 8.31, "learning_rate": 6.066666666666667e-05, "loss": 2.4773, "step": 113500 }, { "epoch": 8.34, "learning_rate": 6.044444444444445e-05, "loss": 2.4742, "step": 114000 }, { "epoch": 8.38, "learning_rate": 6.0222222222222225e-05, "loss": 2.4734, "step": 114500 }, { "epoch": 8.42, "learning_rate": 6e-05, "loss": 2.4734, "step": 115000 }, { "epoch": 8.45, "learning_rate": 5.977777777777778e-05, "loss": 2.4724, "step": 115500 }, { "epoch": 8.49, "learning_rate": 5.9555555555555554e-05, "loss": 2.4764, "step": 116000 }, { "epoch": 8.52, "learning_rate": 5.9333333333333343e-05, "loss": 2.4712, "step": 116500 }, { "epoch": 8.56, "learning_rate": 5.911111111111112e-05, "loss": 2.4706, "step": 117000 }, { "epoch": 8.6, "learning_rate": 5.8888888888888896e-05, "loss": 2.4697, "step": 117500 }, { "epoch": 8.63, "learning_rate": 5.866666666666667e-05, "loss": 2.473, "step": 118000 }, { "epoch": 8.67, "learning_rate": 5.844444444444445e-05, "loss": 2.4684, "step": 118500 }, { "epoch": 8.71, "learning_rate": 5.8222222222222224e-05, "loss": 2.47, "step": 119000 }, { "epoch": 8.74, "learning_rate": 5.8e-05, "loss": 2.4672, "step": 119500 }, { "epoch": 8.78, "learning_rate": 5.7777777777777776e-05, "loss": 2.4677, "step": 120000 }, { "epoch": 8.82, "learning_rate": 5.755555555555556e-05, "loss": 2.4688, "step": 120500 }, { "epoch": 8.85, "learning_rate": 5.7333333333333336e-05, "loss": 2.4667, "step": 121000 }, { "epoch": 8.89, "learning_rate": 5.711111111111112e-05, "loss": 2.4659, "step": 121500 }, { "epoch": 8.93, "learning_rate": 5.6888888888888895e-05, "loss": 2.4671, "step": 122000 }, { "epoch": 8.96, "learning_rate": 5.666666666666667e-05, "loss": 2.4643, "step": 122500 }, { "epoch": 9.0, "learning_rate": 5.644444444444445e-05, "loss": 2.4693, "step": 123000 }, { "epoch": 9.04, "learning_rate": 5.622222222222222e-05, "loss": 2.4633, "step": 123500 }, { "epoch": 9.07, "learning_rate": 5.6000000000000006e-05, "loss": 2.4625, "step": 124000 }, { "epoch": 9.11, "learning_rate": 5.577777777777778e-05, "loss": 2.4604, "step": 124500 }, { "epoch": 9.15, "learning_rate": 5.555555555555556e-05, "loss": 2.4615, "step": 125000 }, { "epoch": 9.18, "learning_rate": 5.5333333333333334e-05, "loss": 2.4598, "step": 125500 }, { "epoch": 9.22, "learning_rate": 5.511111111111111e-05, "loss": 2.4594, "step": 126000 }, { "epoch": 9.26, "learning_rate": 5.488888888888889e-05, "loss": 2.4614, "step": 126500 }, { "epoch": 9.29, "learning_rate": 5.466666666666666e-05, "loss": 2.4586, "step": 127000 }, { "epoch": 9.33, "learning_rate": 5.4444444444444446e-05, "loss": 2.4582, "step": 127500 }, { "epoch": 9.37, "learning_rate": 5.422222222222223e-05, "loss": 2.4575, "step": 128000 }, { "epoch": 9.4, "learning_rate": 5.4000000000000005e-05, "loss": 2.4573, "step": 128500 }, { "epoch": 9.44, "learning_rate": 5.377777777777778e-05, "loss": 2.4561, "step": 129000 }, { "epoch": 9.48, "learning_rate": 5.355555555555556e-05, "loss": 2.4577, "step": 129500 }, { "epoch": 9.51, "learning_rate": 5.333333333333333e-05, "loss": 2.4561, "step": 130000 }, { "epoch": 9.55, "learning_rate": 5.311111111111111e-05, "loss": 2.4537, "step": 130500 }, { "epoch": 9.59, "learning_rate": 5.2888888888888885e-05, "loss": 2.4535, "step": 131000 }, { "epoch": 9.62, "learning_rate": 5.266666666666666e-05, "loss": 2.4538, "step": 131500 }, { "epoch": 9.66, "learning_rate": 5.244444444444445e-05, "loss": 2.4522, "step": 132000 }, { "epoch": 9.7, "learning_rate": 5.222222222222223e-05, "loss": 2.4503, "step": 132500 }, { "epoch": 9.73, "learning_rate": 5.2000000000000004e-05, "loss": 2.4536, "step": 133000 }, { "epoch": 9.77, "learning_rate": 5.177777777777778e-05, "loss": 2.4518, "step": 133500 }, { "epoch": 9.81, "learning_rate": 5.1555555555555556e-05, "loss": 2.4516, "step": 134000 }, { "epoch": 9.84, "learning_rate": 5.133333333333333e-05, "loss": 2.4519, "step": 134500 }, { "epoch": 9.88, "learning_rate": 5.111111111111111e-05, "loss": 2.4496, "step": 135000 }, { "epoch": 9.92, "learning_rate": 5.0888888888888884e-05, "loss": 2.4468, "step": 135500 }, { "epoch": 9.95, "learning_rate": 5.0666666666666674e-05, "loss": 2.4493, "step": 136000 }, { "epoch": 9.99, "learning_rate": 5.044444444444445e-05, "loss": 2.4487, "step": 136500 }, { "epoch": 10.02, "learning_rate": 5.0222222222222226e-05, "loss": 2.4481, "step": 137000 }, { "epoch": 10.06, "learning_rate": 5e-05, "loss": 2.448, "step": 137500 }, { "epoch": 10.1, "learning_rate": 4.977777777777778e-05, "loss": 2.4451, "step": 138000 }, { "epoch": 10.13, "learning_rate": 4.955555555555556e-05, "loss": 2.4455, "step": 138500 }, { "epoch": 10.17, "learning_rate": 4.933333333333334e-05, "loss": 2.4434, "step": 139000 }, { "epoch": 10.21, "learning_rate": 4.9111111111111114e-05, "loss": 2.4422, "step": 139500 }, { "epoch": 10.24, "learning_rate": 4.888888888888889e-05, "loss": 2.4451, "step": 140000 }, { "epoch": 10.28, "learning_rate": 4.866666666666667e-05, "loss": 2.4431, "step": 140500 }, { "epoch": 10.32, "learning_rate": 4.844444444444445e-05, "loss": 2.4455, "step": 141000 }, { "epoch": 10.35, "learning_rate": 4.8222222222222225e-05, "loss": 2.4408, "step": 141500 }, { "epoch": 10.39, "learning_rate": 4.8e-05, "loss": 2.442, "step": 142000 }, { "epoch": 10.43, "learning_rate": 4.7777777777777784e-05, "loss": 2.4447, "step": 142500 }, { "epoch": 10.46, "learning_rate": 4.755555555555556e-05, "loss": 2.4418, "step": 143000 }, { "epoch": 10.5, "learning_rate": 4.7333333333333336e-05, "loss": 2.4407, "step": 143500 }, { "epoch": 10.54, "learning_rate": 4.711111111111111e-05, "loss": 2.4401, "step": 144000 }, { "epoch": 10.57, "learning_rate": 4.6888888888888895e-05, "loss": 2.4397, "step": 144500 }, { "epoch": 10.61, "learning_rate": 4.666666666666667e-05, "loss": 2.4409, "step": 145000 }, { "epoch": 10.65, "learning_rate": 4.644444444444445e-05, "loss": 2.4409, "step": 145500 }, { "epoch": 10.68, "learning_rate": 4.6222222222222224e-05, "loss": 2.4411, "step": 146000 }, { "epoch": 10.72, "learning_rate": 4.600000000000001e-05, "loss": 2.4361, "step": 146500 }, { "epoch": 10.76, "learning_rate": 4.577777777777778e-05, "loss": 2.4354, "step": 147000 }, { "epoch": 10.79, "learning_rate": 4.555555555555556e-05, "loss": 2.4363, "step": 147500 }, { "epoch": 10.83, "learning_rate": 4.5333333333333335e-05, "loss": 2.435, "step": 148000 }, { "epoch": 10.87, "learning_rate": 4.511111111111112e-05, "loss": 2.4403, "step": 148500 }, { "epoch": 10.9, "learning_rate": 4.4888888888888894e-05, "loss": 2.4357, "step": 149000 }, { "epoch": 10.94, "learning_rate": 4.466666666666667e-05, "loss": 2.4357, "step": 149500 }, { "epoch": 10.98, "learning_rate": 4.4444444444444447e-05, "loss": 2.4365, "step": 150000 }, { "epoch": 11.01, "learning_rate": 4.422222222222222e-05, "loss": 2.4343, "step": 150500 }, { "epoch": 11.05, "learning_rate": 4.4000000000000006e-05, "loss": 2.4338, "step": 151000 }, { "epoch": 11.09, "learning_rate": 4.377777777777778e-05, "loss": 2.4312, "step": 151500 }, { "epoch": 11.12, "learning_rate": 4.355555555555556e-05, "loss": 2.4311, "step": 152000 }, { "epoch": 11.16, "learning_rate": 4.3333333333333334e-05, "loss": 2.4311, "step": 152500 }, { "epoch": 11.2, "learning_rate": 4.311111111111111e-05, "loss": 2.4309, "step": 153000 }, { "epoch": 11.23, "learning_rate": 4.2888888888888886e-05, "loss": 2.4309, "step": 153500 }, { "epoch": 11.27, "learning_rate": 4.266666666666667e-05, "loss": 2.4308, "step": 154000 }, { "epoch": 11.31, "learning_rate": 4.2444444444444445e-05, "loss": 2.433, "step": 154500 }, { "epoch": 11.34, "learning_rate": 4.222222222222222e-05, "loss": 2.427, "step": 155000 }, { "epoch": 11.38, "learning_rate": 4.2e-05, "loss": 2.4294, "step": 155500 }, { "epoch": 11.42, "learning_rate": 4.177777777777778e-05, "loss": 2.4279, "step": 156000 }, { "epoch": 11.45, "learning_rate": 4.155555555555556e-05, "loss": 2.4292, "step": 156500 }, { "epoch": 11.49, "learning_rate": 4.133333333333333e-05, "loss": 2.4267, "step": 157000 }, { "epoch": 11.52, "learning_rate": 4.111111111111111e-05, "loss": 2.4288, "step": 157500 }, { "epoch": 11.56, "learning_rate": 4.088888888888889e-05, "loss": 2.4275, "step": 158000 }, { "epoch": 11.6, "learning_rate": 4.066666666666667e-05, "loss": 2.4297, "step": 158500 }, { "epoch": 11.63, "learning_rate": 4.0444444444444444e-05, "loss": 2.4271, "step": 159000 }, { "epoch": 11.67, "learning_rate": 4.022222222222222e-05, "loss": 2.4256, "step": 159500 }, { "epoch": 11.71, "learning_rate": 4e-05, "loss": 2.4283, "step": 160000 }, { "epoch": 11.74, "learning_rate": 3.977777777777778e-05, "loss": 2.4277, "step": 160500 }, { "epoch": 11.78, "learning_rate": 3.9555555555555556e-05, "loss": 2.4232, "step": 161000 }, { "epoch": 11.82, "learning_rate": 3.933333333333333e-05, "loss": 2.4236, "step": 161500 }, { "epoch": 11.85, "learning_rate": 3.9111111111111115e-05, "loss": 2.4236, "step": 162000 }, { "epoch": 11.89, "learning_rate": 3.888888888888889e-05, "loss": 2.4226, "step": 162500 }, { "epoch": 11.93, "learning_rate": 3.866666666666667e-05, "loss": 2.422, "step": 163000 }, { "epoch": 11.96, "learning_rate": 3.844444444444444e-05, "loss": 2.4236, "step": 163500 }, { "epoch": 12.0, "learning_rate": 3.8222222222222226e-05, "loss": 2.423, "step": 164000 }, { "epoch": 12.04, "learning_rate": 3.8e-05, "loss": 2.4208, "step": 164500 }, { "epoch": 12.07, "learning_rate": 3.777777777777778e-05, "loss": 2.4214, "step": 165000 }, { "epoch": 12.11, "learning_rate": 3.7555555555555554e-05, "loss": 2.4203, "step": 165500 }, { "epoch": 12.15, "learning_rate": 3.733333333333334e-05, "loss": 2.4207, "step": 166000 }, { "epoch": 12.18, "learning_rate": 3.7111111111111113e-05, "loss": 2.4177, "step": 166500 }, { "epoch": 12.22, "learning_rate": 3.688888888888889e-05, "loss": 2.4196, "step": 167000 }, { "epoch": 12.26, "learning_rate": 3.6666666666666666e-05, "loss": 2.4174, "step": 167500 }, { "epoch": 12.29, "learning_rate": 3.644444444444445e-05, "loss": 2.42, "step": 168000 }, { "epoch": 12.33, "learning_rate": 3.6222222222222225e-05, "loss": 2.4194, "step": 168500 }, { "epoch": 12.37, "learning_rate": 3.6e-05, "loss": 2.4155, "step": 169000 }, { "epoch": 12.4, "learning_rate": 3.577777777777778e-05, "loss": 2.4149, "step": 169500 }, { "epoch": 12.44, "learning_rate": 3.555555555555556e-05, "loss": 2.418, "step": 170000 }, { "epoch": 12.48, "learning_rate": 3.5333333333333336e-05, "loss": 2.4199, "step": 170500 }, { "epoch": 12.51, "learning_rate": 3.511111111111111e-05, "loss": 2.4162, "step": 171000 }, { "epoch": 12.55, "learning_rate": 3.4888888888888895e-05, "loss": 2.4188, "step": 171500 }, { "epoch": 12.59, "learning_rate": 3.466666666666667e-05, "loss": 2.4146, "step": 172000 }, { "epoch": 12.62, "learning_rate": 3.444444444444445e-05, "loss": 2.4167, "step": 172500 }, { "epoch": 12.66, "learning_rate": 3.4222222222222224e-05, "loss": 2.414, "step": 173000 }, { "epoch": 12.7, "learning_rate": 3.4000000000000007e-05, "loss": 2.4156, "step": 173500 }, { "epoch": 12.73, "learning_rate": 3.377777777777778e-05, "loss": 2.4155, "step": 174000 }, { "epoch": 12.77, "learning_rate": 3.355555555555556e-05, "loss": 2.4139, "step": 174500 }, { "epoch": 12.81, "learning_rate": 3.3333333333333335e-05, "loss": 2.4154, "step": 175000 }, { "epoch": 12.84, "learning_rate": 3.311111111111112e-05, "loss": 2.415, "step": 175500 }, { "epoch": 12.88, "learning_rate": 3.2888888888888894e-05, "loss": 2.4137, "step": 176000 }, { "epoch": 12.92, "learning_rate": 3.266666666666667e-05, "loss": 2.4126, "step": 176500 }, { "epoch": 12.95, "learning_rate": 3.2444444444444446e-05, "loss": 2.4139, "step": 177000 }, { "epoch": 12.99, "learning_rate": 3.222222222222223e-05, "loss": 2.4129, "step": 177500 }, { "epoch": 13.03, "learning_rate": 3.2000000000000005e-05, "loss": 2.4151, "step": 178000 }, { "epoch": 13.06, "learning_rate": 3.177777777777778e-05, "loss": 2.4109, "step": 178500 }, { "epoch": 13.1, "learning_rate": 3.155555555555556e-05, "loss": 2.4114, "step": 179000 }, { "epoch": 13.13, "learning_rate": 3.1333333333333334e-05, "loss": 2.4106, "step": 179500 }, { "epoch": 13.17, "learning_rate": 3.111111111111111e-05, "loss": 2.4085, "step": 180000 }, { "epoch": 13.21, "learning_rate": 3.088888888888889e-05, "loss": 2.4076, "step": 180500 }, { "epoch": 13.24, "learning_rate": 3.066666666666667e-05, "loss": 2.4094, "step": 181000 }, { "epoch": 13.28, "learning_rate": 3.044444444444445e-05, "loss": 2.4089, "step": 181500 }, { "epoch": 13.32, "learning_rate": 3.0222222222222225e-05, "loss": 2.4082, "step": 182000 }, { "epoch": 13.35, "learning_rate": 3e-05, "loss": 2.4052, "step": 182500 }, { "epoch": 13.39, "learning_rate": 2.9777777777777777e-05, "loss": 2.4079, "step": 183000 }, { "epoch": 13.43, "learning_rate": 2.955555555555556e-05, "loss": 2.405, "step": 183500 }, { "epoch": 13.46, "learning_rate": 2.9333333333333336e-05, "loss": 2.41, "step": 184000 }, { "epoch": 13.5, "learning_rate": 2.9111111111111112e-05, "loss": 2.4085, "step": 184500 }, { "epoch": 13.54, "learning_rate": 2.8888888888888888e-05, "loss": 2.4067, "step": 185000 }, { "epoch": 13.57, "learning_rate": 2.8666666666666668e-05, "loss": 2.4065, "step": 185500 }, { "epoch": 13.61, "learning_rate": 2.8444444444444447e-05, "loss": 2.4059, "step": 186000 }, { "epoch": 13.65, "learning_rate": 2.8222222222222223e-05, "loss": 2.4049, "step": 186500 }, { "epoch": 13.68, "learning_rate": 2.8000000000000003e-05, "loss": 2.4061, "step": 187000 }, { "epoch": 13.72, "learning_rate": 2.777777777777778e-05, "loss": 2.4067, "step": 187500 }, { "epoch": 13.76, "learning_rate": 2.7555555555555555e-05, "loss": 2.4047, "step": 188000 }, { "epoch": 13.79, "learning_rate": 2.733333333333333e-05, "loss": 2.4066, "step": 188500 }, { "epoch": 13.83, "learning_rate": 2.7111111111111114e-05, "loss": 2.4035, "step": 189000 }, { "epoch": 13.87, "learning_rate": 2.688888888888889e-05, "loss": 2.4038, "step": 189500 }, { "epoch": 13.9, "learning_rate": 2.6666666666666667e-05, "loss": 2.4012, "step": 190000 }, { "epoch": 13.94, "learning_rate": 2.6444444444444443e-05, "loss": 2.4037, "step": 190500 }, { "epoch": 13.98, "learning_rate": 2.6222222222222226e-05, "loss": 2.4028, "step": 191000 }, { "epoch": 14.01, "learning_rate": 2.6000000000000002e-05, "loss": 2.4041, "step": 191500 }, { "epoch": 14.05, "learning_rate": 2.5777777777777778e-05, "loss": 2.4024, "step": 192000 }, { "epoch": 14.09, "learning_rate": 2.5555555555555554e-05, "loss": 2.3971, "step": 192500 }, { "epoch": 14.12, "learning_rate": 2.5333333333333337e-05, "loss": 2.3994, "step": 193000 }, { "epoch": 14.16, "learning_rate": 2.5111111111111113e-05, "loss": 2.4016, "step": 193500 }, { "epoch": 14.2, "learning_rate": 2.488888888888889e-05, "loss": 2.3994, "step": 194000 }, { "epoch": 14.23, "learning_rate": 2.466666666666667e-05, "loss": 2.4007, "step": 194500 }, { "epoch": 14.27, "learning_rate": 2.4444444444444445e-05, "loss": 2.3988, "step": 195000 }, { "epoch": 14.31, "learning_rate": 2.4222222222222224e-05, "loss": 2.4009, "step": 195500 }, { "epoch": 14.34, "learning_rate": 2.4e-05, "loss": 2.3996, "step": 196000 }, { "epoch": 14.38, "learning_rate": 2.377777777777778e-05, "loss": 2.3971, "step": 196500 }, { "epoch": 14.42, "learning_rate": 2.3555555555555556e-05, "loss": 2.3975, "step": 197000 }, { "epoch": 14.45, "learning_rate": 2.3333333333333336e-05, "loss": 2.3999, "step": 197500 }, { "epoch": 14.49, "learning_rate": 2.3111111111111112e-05, "loss": 2.4001, "step": 198000 }, { "epoch": 14.53, "learning_rate": 2.288888888888889e-05, "loss": 2.397, "step": 198500 }, { "epoch": 14.56, "learning_rate": 2.2666666666666668e-05, "loss": 2.3955, "step": 199000 }, { "epoch": 14.6, "learning_rate": 2.2444444444444447e-05, "loss": 2.4004, "step": 199500 }, { "epoch": 14.63, "learning_rate": 2.2222222222222223e-05, "loss": 2.3956, "step": 200000 }, { "epoch": 14.67, "learning_rate": 2.2000000000000003e-05, "loss": 2.3958, "step": 200500 }, { "epoch": 14.71, "learning_rate": 2.177777777777778e-05, "loss": 2.3969, "step": 201000 }, { "epoch": 14.74, "learning_rate": 2.1555555555555555e-05, "loss": 2.3975, "step": 201500 }, { "epoch": 14.78, "learning_rate": 2.1333333333333335e-05, "loss": 2.3942, "step": 202000 }, { "epoch": 14.82, "learning_rate": 2.111111111111111e-05, "loss": 2.3981, "step": 202500 }, { "epoch": 14.85, "learning_rate": 2.088888888888889e-05, "loss": 2.3963, "step": 203000 }, { "epoch": 14.89, "learning_rate": 2.0666666666666666e-05, "loss": 2.3942, "step": 203500 }, { "epoch": 14.93, "learning_rate": 2.0444444444444446e-05, "loss": 2.3972, "step": 204000 }, { "epoch": 14.96, "learning_rate": 2.0222222222222222e-05, "loss": 2.3932, "step": 204500 }, { "epoch": 15.0, "learning_rate": 2e-05, "loss": 2.3976, "step": 205000 }, { "epoch": 15.04, "learning_rate": 1.9777777777777778e-05, "loss": 2.3939, "step": 205500 }, { "epoch": 15.07, "learning_rate": 1.9555555555555557e-05, "loss": 2.3921, "step": 206000 }, { "epoch": 15.11, "learning_rate": 1.9333333333333333e-05, "loss": 2.3936, "step": 206500 }, { "epoch": 15.15, "learning_rate": 1.9111111111111113e-05, "loss": 2.3935, "step": 207000 }, { "epoch": 15.18, "learning_rate": 1.888888888888889e-05, "loss": 2.3946, "step": 207500 }, { "epoch": 15.22, "learning_rate": 1.866666666666667e-05, "loss": 2.3919, "step": 208000 }, { "epoch": 15.26, "learning_rate": 1.8444444444444445e-05, "loss": 2.3925, "step": 208500 }, { "epoch": 15.29, "learning_rate": 1.8222222222222224e-05, "loss": 2.3918, "step": 209000 }, { "epoch": 15.33, "learning_rate": 1.8e-05, "loss": 2.3935, "step": 209500 }, { "epoch": 15.37, "learning_rate": 1.777777777777778e-05, "loss": 2.3916, "step": 210000 }, { "epoch": 15.4, "learning_rate": 1.7555555555555556e-05, "loss": 2.3878, "step": 210500 }, { "epoch": 15.44, "learning_rate": 1.7333333333333336e-05, "loss": 2.3892, "step": 211000 }, { "epoch": 15.48, "learning_rate": 1.7111111111111112e-05, "loss": 2.3902, "step": 211500 }, { "epoch": 15.51, "learning_rate": 1.688888888888889e-05, "loss": 2.389, "step": 212000 }, { "epoch": 15.55, "learning_rate": 1.6666666666666667e-05, "loss": 2.39, "step": 212500 }, { "epoch": 15.59, "learning_rate": 1.6444444444444447e-05, "loss": 2.3869, "step": 213000 }, { "epoch": 15.62, "learning_rate": 1.6222222222222223e-05, "loss": 2.3899, "step": 213500 }, { "epoch": 15.66, "learning_rate": 1.6000000000000003e-05, "loss": 2.3903, "step": 214000 }, { "epoch": 15.7, "learning_rate": 1.577777777777778e-05, "loss": 2.3861, "step": 214500 }, { "epoch": 15.73, "learning_rate": 1.5555555555555555e-05, "loss": 2.3861, "step": 215000 }, { "epoch": 15.77, "learning_rate": 1.5333333333333334e-05, "loss": 2.3904, "step": 215500 }, { "epoch": 15.81, "learning_rate": 1.5111111111111112e-05, "loss": 2.3883, "step": 216000 }, { "epoch": 15.84, "learning_rate": 1.4888888888888888e-05, "loss": 2.3879, "step": 216500 }, { "epoch": 15.88, "learning_rate": 1.4666666666666668e-05, "loss": 2.3867, "step": 217000 }, { "epoch": 15.92, "learning_rate": 1.4444444444444444e-05, "loss": 2.3898, "step": 217500 }, { "epoch": 15.95, "learning_rate": 1.4222222222222224e-05, "loss": 2.3882, "step": 218000 }, { "epoch": 15.99, "learning_rate": 1.4000000000000001e-05, "loss": 2.385, "step": 218500 }, { "epoch": 16.03, "learning_rate": 1.3777777777777778e-05, "loss": 2.3853, "step": 219000 }, { "epoch": 16.06, "learning_rate": 1.3555555555555557e-05, "loss": 2.3855, "step": 219500 }, { "epoch": 16.1, "learning_rate": 1.3333333333333333e-05, "loss": 2.3869, "step": 220000 }, { "epoch": 16.13, "learning_rate": 1.3111111111111113e-05, "loss": 2.3859, "step": 220500 }, { "epoch": 16.17, "learning_rate": 1.2888888888888889e-05, "loss": 2.3855, "step": 221000 }, { "epoch": 16.21, "learning_rate": 1.2666666666666668e-05, "loss": 2.3858, "step": 221500 }, { "epoch": 16.24, "learning_rate": 1.2444444444444445e-05, "loss": 2.3846, "step": 222000 }, { "epoch": 16.28, "learning_rate": 1.2222222222222222e-05, "loss": 2.3842, "step": 222500 }, { "epoch": 16.32, "learning_rate": 1.2e-05, "loss": 2.3814, "step": 223000 }, { "epoch": 16.35, "learning_rate": 1.1777777777777778e-05, "loss": 2.3837, "step": 223500 }, { "epoch": 16.39, "learning_rate": 1.1555555555555556e-05, "loss": 2.3841, "step": 224000 }, { "epoch": 16.43, "learning_rate": 1.1333333333333334e-05, "loss": 2.3829, "step": 224500 }, { "epoch": 16.46, "learning_rate": 1.1111111111111112e-05, "loss": 2.3849, "step": 225000 }, { "epoch": 16.5, "learning_rate": 1.088888888888889e-05, "loss": 2.383, "step": 225500 }, { "epoch": 16.54, "learning_rate": 1.0666666666666667e-05, "loss": 2.3795, "step": 226000 }, { "epoch": 16.57, "learning_rate": 1.0444444444444445e-05, "loss": 2.3817, "step": 226500 }, { "epoch": 16.61, "learning_rate": 1.0222222222222223e-05, "loss": 2.3826, "step": 227000 }, { "epoch": 16.65, "learning_rate": 1e-05, "loss": 2.3825, "step": 227500 }, { "epoch": 16.68, "learning_rate": 9.777777777777779e-06, "loss": 2.3789, "step": 228000 }, { "epoch": 16.72, "learning_rate": 9.555555555555556e-06, "loss": 2.3865, "step": 228500 }, { "epoch": 16.76, "learning_rate": 9.333333333333334e-06, "loss": 2.3825, "step": 229000 }, { "epoch": 16.79, "learning_rate": 9.111111111111112e-06, "loss": 2.3808, "step": 229500 }, { "epoch": 16.83, "learning_rate": 8.88888888888889e-06, "loss": 2.3807, "step": 230000 }, { "epoch": 16.87, "learning_rate": 8.666666666666668e-06, "loss": 2.3813, "step": 230500 }, { "epoch": 16.9, "learning_rate": 8.444444444444446e-06, "loss": 2.3817, "step": 231000 }, { "epoch": 16.94, "learning_rate": 8.222222222222223e-06, "loss": 2.3824, "step": 231500 }, { "epoch": 16.98, "learning_rate": 8.000000000000001e-06, "loss": 2.3809, "step": 232000 }, { "epoch": 17.01, "learning_rate": 7.777777777777777e-06, "loss": 2.3798, "step": 232500 }, { "epoch": 17.05, "learning_rate": 7.555555555555556e-06, "loss": 2.377, "step": 233000 }, { "epoch": 17.09, "learning_rate": 7.333333333333334e-06, "loss": 2.3801, "step": 233500 }, { "epoch": 17.12, "learning_rate": 7.111111111111112e-06, "loss": 2.3801, "step": 234000 }, { "epoch": 17.16, "learning_rate": 6.888888888888889e-06, "loss": 2.3816, "step": 234500 }, { "epoch": 17.2, "learning_rate": 6.666666666666667e-06, "loss": 2.3772, "step": 235000 }, { "epoch": 17.23, "learning_rate": 6.4444444444444445e-06, "loss": 2.3818, "step": 235500 }, { "epoch": 17.27, "learning_rate": 6.222222222222222e-06, "loss": 2.3786, "step": 236000 }, { "epoch": 17.31, "learning_rate": 6e-06, "loss": 2.3777, "step": 236500 }, { "epoch": 17.34, "learning_rate": 5.777777777777778e-06, "loss": 2.3774, "step": 237000 }, { "epoch": 17.38, "learning_rate": 5.555555555555556e-06, "loss": 2.3787, "step": 237500 }, { "epoch": 17.42, "learning_rate": 5.333333333333334e-06, "loss": 2.3772, "step": 238000 }, { "epoch": 17.45, "learning_rate": 5.1111111111111115e-06, "loss": 2.3776, "step": 238500 }, { "epoch": 17.49, "learning_rate": 4.888888888888889e-06, "loss": 2.3776, "step": 239000 }, { "epoch": 17.53, "learning_rate": 4.666666666666667e-06, "loss": 2.3782, "step": 239500 }, { "epoch": 17.56, "learning_rate": 4.444444444444445e-06, "loss": 2.3755, "step": 240000 }, { "epoch": 17.6, "learning_rate": 4.222222222222223e-06, "loss": 2.3766, "step": 240500 }, { "epoch": 17.63, "learning_rate": 4.000000000000001e-06, "loss": 2.3774, "step": 241000 }, { "epoch": 17.67, "learning_rate": 3.777777777777778e-06, "loss": 2.3754, "step": 241500 }, { "epoch": 17.71, "learning_rate": 3.555555555555556e-06, "loss": 2.3785, "step": 242000 }, { "epoch": 17.74, "learning_rate": 3.3333333333333333e-06, "loss": 2.3746, "step": 242500 }, { "epoch": 17.78, "learning_rate": 3.111111111111111e-06, "loss": 2.3788, "step": 243000 }, { "epoch": 17.82, "learning_rate": 2.888888888888889e-06, "loss": 2.3751, "step": 243500 }, { "epoch": 17.85, "learning_rate": 2.666666666666667e-06, "loss": 2.3747, "step": 244000 }, { "epoch": 17.89, "learning_rate": 2.4444444444444447e-06, "loss": 2.3758, "step": 244500 }, { "epoch": 17.93, "learning_rate": 2.2222222222222225e-06, "loss": 2.3738, "step": 245000 }, { "epoch": 17.96, "learning_rate": 2.0000000000000003e-06, "loss": 2.3761, "step": 245500 }, { "epoch": 18.0, "learning_rate": 1.777777777777778e-06, "loss": 2.3765, "step": 246000 }, { "epoch": 18.04, "learning_rate": 1.5555555555555556e-06, "loss": 2.3742, "step": 246500 }, { "epoch": 18.07, "learning_rate": 1.3333333333333334e-06, "loss": 2.3765, "step": 247000 }, { "epoch": 18.11, "learning_rate": 1.1111111111111112e-06, "loss": 2.3735, "step": 247500 }, { "epoch": 18.15, "learning_rate": 8.88888888888889e-07, "loss": 2.3758, "step": 248000 }, { "epoch": 18.18, "learning_rate": 6.666666666666667e-07, "loss": 2.3729, "step": 248500 }, { "epoch": 18.22, "learning_rate": 4.444444444444445e-07, "loss": 2.3758, "step": 249000 }, { "epoch": 18.26, "learning_rate": 2.2222222222222224e-07, "loss": 2.3745, "step": 249500 }, { "epoch": 18.29, "learning_rate": 0.0, "loss": 2.3726, "step": 250000 }, { "epoch": 18.29, "step": 250000, "total_flos": 4.212323233189429e+18, "train_loss": 2.6346825883789062, "train_runtime": 359539.1522, "train_samples_per_second": 356.011, "train_steps_per_second": 0.695 } ], "max_steps": 250000, "num_train_epochs": 19, "total_flos": 4.212323233189429e+18, "trial_name": null, "trial_params": null }