| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 2.0, | |
| "eval_steps": 500, | |
| "global_step": 228, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.008771929824561403, | |
| "grad_norm": 39.56407165527344, | |
| "learning_rate": 5.0000000000000004e-08, | |
| "loss": 5.1375, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.017543859649122806, | |
| "grad_norm": 40.30452346801758, | |
| "learning_rate": 1.0000000000000001e-07, | |
| "loss": 5.1185, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.02631578947368421, | |
| "grad_norm": 40.062313079833984, | |
| "learning_rate": 1.5000000000000002e-07, | |
| "loss": 5.0762, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.03508771929824561, | |
| "grad_norm": 39.17148208618164, | |
| "learning_rate": 2.0000000000000002e-07, | |
| "loss": 5.016, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.043859649122807015, | |
| "grad_norm": 40.67367172241211, | |
| "learning_rate": 2.5000000000000004e-07, | |
| "loss": 5.0428, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.05263157894736842, | |
| "grad_norm": 38.18095016479492, | |
| "learning_rate": 3.0000000000000004e-07, | |
| "loss": 5.2025, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.06140350877192982, | |
| "grad_norm": 39.12940979003906, | |
| "learning_rate": 3.5000000000000004e-07, | |
| "loss": 4.9896, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.07017543859649122, | |
| "grad_norm": 38.84568405151367, | |
| "learning_rate": 4.0000000000000003e-07, | |
| "loss": 5.1078, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.07894736842105263, | |
| "grad_norm": 39.38333511352539, | |
| "learning_rate": 4.5000000000000003e-07, | |
| "loss": 5.0808, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.08771929824561403, | |
| "grad_norm": 39.427650451660156, | |
| "learning_rate": 5.000000000000001e-07, | |
| "loss": 5.0534, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.09649122807017543, | |
| "grad_norm": 39.29513168334961, | |
| "learning_rate": 5.5e-07, | |
| "loss": 5.058, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.10526315789473684, | |
| "grad_norm": 39.641231536865234, | |
| "learning_rate": 6.000000000000001e-07, | |
| "loss": 5.0317, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.11403508771929824, | |
| "grad_norm": 37.91259765625, | |
| "learning_rate": 6.5e-07, | |
| "loss": 4.912, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.12280701754385964, | |
| "grad_norm": 38.203548431396484, | |
| "learning_rate": 7.000000000000001e-07, | |
| "loss": 4.9705, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.13157894736842105, | |
| "grad_norm": 39.15998840332031, | |
| "learning_rate": 7.5e-07, | |
| "loss": 4.6962, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.14035087719298245, | |
| "grad_norm": 37.754669189453125, | |
| "learning_rate": 8.000000000000001e-07, | |
| "loss": 4.6262, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.14912280701754385, | |
| "grad_norm": 35.871490478515625, | |
| "learning_rate": 8.500000000000001e-07, | |
| "loss": 4.5422, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.15789473684210525, | |
| "grad_norm": 36.16888427734375, | |
| "learning_rate": 9.000000000000001e-07, | |
| "loss": 4.664, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.16666666666666666, | |
| "grad_norm": 33.520118713378906, | |
| "learning_rate": 9.500000000000001e-07, | |
| "loss": 4.4697, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.17543859649122806, | |
| "grad_norm": 30.896282196044922, | |
| "learning_rate": 1.0000000000000002e-06, | |
| "loss": 4.3568, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.18421052631578946, | |
| "grad_norm": 29.944643020629883, | |
| "learning_rate": 1.0500000000000001e-06, | |
| "loss": 4.2269, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.19298245614035087, | |
| "grad_norm": 25.224485397338867, | |
| "learning_rate": 1.1e-06, | |
| "loss": 4.1272, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.20175438596491227, | |
| "grad_norm": 24.410480499267578, | |
| "learning_rate": 1.1500000000000002e-06, | |
| "loss": 4.0585, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.21052631578947367, | |
| "grad_norm": 21.480648040771484, | |
| "learning_rate": 1.2000000000000002e-06, | |
| "loss": 3.9472, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.21929824561403508, | |
| "grad_norm": 20.61946678161621, | |
| "learning_rate": 1.25e-06, | |
| "loss": 3.8879, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.22807017543859648, | |
| "grad_norm": 19.578271865844727, | |
| "learning_rate": 1.3e-06, | |
| "loss": 3.6783, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.23684210526315788, | |
| "grad_norm": 17.418983459472656, | |
| "learning_rate": 1.3500000000000002e-06, | |
| "loss": 3.6826, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.24561403508771928, | |
| "grad_norm": 18.160301208496094, | |
| "learning_rate": 1.4000000000000001e-06, | |
| "loss": 3.478, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.2543859649122807, | |
| "grad_norm": 17.573204040527344, | |
| "learning_rate": 1.45e-06, | |
| "loss": 3.459, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.2631578947368421, | |
| "grad_norm": 17.1265869140625, | |
| "learning_rate": 1.5e-06, | |
| "loss": 3.3999, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.2719298245614035, | |
| "grad_norm": 15.527145385742188, | |
| "learning_rate": 1.5500000000000002e-06, | |
| "loss": 3.2817, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.2807017543859649, | |
| "grad_norm": 14.773847579956055, | |
| "learning_rate": 1.6000000000000001e-06, | |
| "loss": 3.234, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.2894736842105263, | |
| "grad_norm": 12.039301872253418, | |
| "learning_rate": 1.6500000000000003e-06, | |
| "loss": 3.132, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.2982456140350877, | |
| "grad_norm": 9.217979431152344, | |
| "learning_rate": 1.7000000000000002e-06, | |
| "loss": 3.0548, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.30701754385964913, | |
| "grad_norm": 7.575639724731445, | |
| "learning_rate": 1.75e-06, | |
| "loss": 2.9529, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.3157894736842105, | |
| "grad_norm": 7.496004104614258, | |
| "learning_rate": 1.8000000000000001e-06, | |
| "loss": 2.8967, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.32456140350877194, | |
| "grad_norm": 7.45414924621582, | |
| "learning_rate": 1.85e-06, | |
| "loss": 2.8837, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.3333333333333333, | |
| "grad_norm": 8.555658340454102, | |
| "learning_rate": 1.9000000000000002e-06, | |
| "loss": 2.7473, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.34210526315789475, | |
| "grad_norm": 10.03805160522461, | |
| "learning_rate": 1.9500000000000004e-06, | |
| "loss": 2.7355, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.3508771929824561, | |
| "grad_norm": 9.30649471282959, | |
| "learning_rate": 2.0000000000000003e-06, | |
| "loss": 2.6587, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.35964912280701755, | |
| "grad_norm": 8.510339736938477, | |
| "learning_rate": 2.05e-06, | |
| "loss": 2.5977, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.3684210526315789, | |
| "grad_norm": 4.709080696105957, | |
| "learning_rate": 2.1000000000000002e-06, | |
| "loss": 2.6286, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.37719298245614036, | |
| "grad_norm": 5.128961086273193, | |
| "learning_rate": 2.15e-06, | |
| "loss": 2.4558, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.38596491228070173, | |
| "grad_norm": 5.190136432647705, | |
| "learning_rate": 2.2e-06, | |
| "loss": 2.4432, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.39473684210526316, | |
| "grad_norm": 4.893551349639893, | |
| "learning_rate": 2.25e-06, | |
| "loss": 2.4939, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.40350877192982454, | |
| "grad_norm": 5.2434983253479, | |
| "learning_rate": 2.3000000000000004e-06, | |
| "loss": 2.3381, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.41228070175438597, | |
| "grad_norm": 5.122412204742432, | |
| "learning_rate": 2.35e-06, | |
| "loss": 2.313, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.42105263157894735, | |
| "grad_norm": 4.577274799346924, | |
| "learning_rate": 2.4000000000000003e-06, | |
| "loss": 2.2236, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.4298245614035088, | |
| "grad_norm": 4.722769737243652, | |
| "learning_rate": 2.4500000000000003e-06, | |
| "loss": 2.1987, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.43859649122807015, | |
| "grad_norm": 5.059235095977783, | |
| "learning_rate": 2.5e-06, | |
| "loss": 2.1415, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.4473684210526316, | |
| "grad_norm": 4.454439640045166, | |
| "learning_rate": 2.55e-06, | |
| "loss": 2.0466, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.45614035087719296, | |
| "grad_norm": 4.94586706161499, | |
| "learning_rate": 2.6e-06, | |
| "loss": 1.8762, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.4649122807017544, | |
| "grad_norm": 4.704402446746826, | |
| "learning_rate": 2.6500000000000005e-06, | |
| "loss": 1.8012, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.47368421052631576, | |
| "grad_norm": 6.125903129577637, | |
| "learning_rate": 2.7000000000000004e-06, | |
| "loss": 1.7669, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.4824561403508772, | |
| "grad_norm": 4.5356059074401855, | |
| "learning_rate": 2.7500000000000004e-06, | |
| "loss": 1.6607, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.49122807017543857, | |
| "grad_norm": 6.56803035736084, | |
| "learning_rate": 2.8000000000000003e-06, | |
| "loss": 1.6291, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "grad_norm": 4.910050392150879, | |
| "learning_rate": 2.85e-06, | |
| "loss": 1.5545, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.5087719298245614, | |
| "grad_norm": 8.733433723449707, | |
| "learning_rate": 2.9e-06, | |
| "loss": 1.4206, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.5175438596491229, | |
| "grad_norm": 8.582486152648926, | |
| "learning_rate": 2.95e-06, | |
| "loss": 1.3912, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.5263157894736842, | |
| "grad_norm": 13.710689544677734, | |
| "learning_rate": 3e-06, | |
| "loss": 1.3297, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.5350877192982456, | |
| "grad_norm": 23.400312423706055, | |
| "learning_rate": 3.05e-06, | |
| "loss": 1.296, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.543859649122807, | |
| "grad_norm": 5.678805351257324, | |
| "learning_rate": 3.1000000000000004e-06, | |
| "loss": 1.2259, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.5526315789473685, | |
| "grad_norm": 14.700899124145508, | |
| "learning_rate": 3.1500000000000003e-06, | |
| "loss": 1.1087, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.5614035087719298, | |
| "grad_norm": 19.38919448852539, | |
| "learning_rate": 3.2000000000000003e-06, | |
| "loss": 1.1805, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.5701754385964912, | |
| "grad_norm": 8.460039138793945, | |
| "learning_rate": 3.2500000000000002e-06, | |
| "loss": 1.0963, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.5789473684210527, | |
| "grad_norm": 13.371014595031738, | |
| "learning_rate": 3.3000000000000006e-06, | |
| "loss": 1.0627, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.5877192982456141, | |
| "grad_norm": 22.380569458007812, | |
| "learning_rate": 3.3500000000000005e-06, | |
| "loss": 1.0869, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.5964912280701754, | |
| "grad_norm": 5.780513286590576, | |
| "learning_rate": 3.4000000000000005e-06, | |
| "loss": 0.9991, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.6052631578947368, | |
| "grad_norm": 19.850841522216797, | |
| "learning_rate": 3.45e-06, | |
| "loss": 0.9683, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.6140350877192983, | |
| "grad_norm": 17.160703659057617, | |
| "learning_rate": 3.5e-06, | |
| "loss": 0.845, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.6228070175438597, | |
| "grad_norm": 14.264311790466309, | |
| "learning_rate": 3.5500000000000003e-06, | |
| "loss": 0.8059, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.631578947368421, | |
| "grad_norm": 26.39459991455078, | |
| "learning_rate": 3.6000000000000003e-06, | |
| "loss": 0.85, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.6403508771929824, | |
| "grad_norm": 51.10348892211914, | |
| "learning_rate": 3.65e-06, | |
| "loss": 0.9755, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.6491228070175439, | |
| "grad_norm": 28.795856475830078, | |
| "learning_rate": 3.7e-06, | |
| "loss": 0.8966, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.6578947368421053, | |
| "grad_norm": 4.6617937088012695, | |
| "learning_rate": 3.7500000000000005e-06, | |
| "loss": 0.7716, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.6666666666666666, | |
| "grad_norm": 15.729666709899902, | |
| "learning_rate": 3.8000000000000005e-06, | |
| "loss": 0.7578, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.6754385964912281, | |
| "grad_norm": 7.109970569610596, | |
| "learning_rate": 3.85e-06, | |
| "loss": 0.7055, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.6842105263157895, | |
| "grad_norm": 20.84659194946289, | |
| "learning_rate": 3.900000000000001e-06, | |
| "loss": 0.7458, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.6929824561403509, | |
| "grad_norm": 21.601303100585938, | |
| "learning_rate": 3.95e-06, | |
| "loss": 0.6879, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.7017543859649122, | |
| "grad_norm": 3.6914751529693604, | |
| "learning_rate": 4.000000000000001e-06, | |
| "loss": 0.6179, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.7105263157894737, | |
| "grad_norm": 16.539325714111328, | |
| "learning_rate": 4.05e-06, | |
| "loss": 0.5716, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.7192982456140351, | |
| "grad_norm": 13.931925773620605, | |
| "learning_rate": 4.1e-06, | |
| "loss": 0.558, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.7280701754385965, | |
| "grad_norm": 10.52951717376709, | |
| "learning_rate": 4.15e-06, | |
| "loss": 0.6018, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.7368421052631579, | |
| "grad_norm": 17.337060928344727, | |
| "learning_rate": 4.2000000000000004e-06, | |
| "loss": 0.5501, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.7456140350877193, | |
| "grad_norm": 13.500468254089355, | |
| "learning_rate": 4.25e-06, | |
| "loss": 0.5214, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.7543859649122807, | |
| "grad_norm": 10.290645599365234, | |
| "learning_rate": 4.3e-06, | |
| "loss": 0.4996, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.7631578947368421, | |
| "grad_norm": 9.757556915283203, | |
| "learning_rate": 4.350000000000001e-06, | |
| "loss": 0.498, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.7719298245614035, | |
| "grad_norm": 9.325140953063965, | |
| "learning_rate": 4.4e-06, | |
| "loss": 0.4721, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.7807017543859649, | |
| "grad_norm": 2.9322128295898438, | |
| "learning_rate": 4.450000000000001e-06, | |
| "loss": 0.4528, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.7894736842105263, | |
| "grad_norm": 10.484073638916016, | |
| "learning_rate": 4.5e-06, | |
| "loss": 0.445, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.7982456140350878, | |
| "grad_norm": 32.7827262878418, | |
| "learning_rate": 4.5500000000000005e-06, | |
| "loss": 0.5105, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.8070175438596491, | |
| "grad_norm": 2.8477306365966797, | |
| "learning_rate": 4.600000000000001e-06, | |
| "loss": 0.4117, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.8157894736842105, | |
| "grad_norm": 2.7680225372314453, | |
| "learning_rate": 4.65e-06, | |
| "loss": 0.3653, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.8245614035087719, | |
| "grad_norm": 2.6512742042541504, | |
| "learning_rate": 4.7e-06, | |
| "loss": 0.3878, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.8333333333333334, | |
| "grad_norm": 6.453914165496826, | |
| "learning_rate": 4.75e-06, | |
| "loss": 0.3611, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.8421052631578947, | |
| "grad_norm": 3.4594080448150635, | |
| "learning_rate": 4.800000000000001e-06, | |
| "loss": 0.3817, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.8508771929824561, | |
| "grad_norm": 3.6144917011260986, | |
| "learning_rate": 4.85e-06, | |
| "loss": 0.3618, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.8596491228070176, | |
| "grad_norm": 5.349407196044922, | |
| "learning_rate": 4.9000000000000005e-06, | |
| "loss": 0.3218, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.868421052631579, | |
| "grad_norm": 13.671236991882324, | |
| "learning_rate": 4.95e-06, | |
| "loss": 0.3329, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.8771929824561403, | |
| "grad_norm": 5.84046745300293, | |
| "learning_rate": 5e-06, | |
| "loss": 0.2967, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.8859649122807017, | |
| "grad_norm": 14.005338668823242, | |
| "learning_rate": 4.999963827125897e-06, | |
| "loss": 0.303, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.8947368421052632, | |
| "grad_norm": 9.18114185333252, | |
| "learning_rate": 4.999855309550366e-06, | |
| "loss": 0.2762, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.9035087719298246, | |
| "grad_norm": 3.0800487995147705, | |
| "learning_rate": 4.999674450413725e-06, | |
| "loss": 0.2628, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.9122807017543859, | |
| "grad_norm": 82.03578186035156, | |
| "learning_rate": 4.999421254949728e-06, | |
| "loss": 0.4065, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.9210526315789473, | |
| "grad_norm": 77.66315460205078, | |
| "learning_rate": 4.99909573048542e-06, | |
| "loss": 0.4307, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.9298245614035088, | |
| "grad_norm": 18.28767967224121, | |
| "learning_rate": 4.998697886440927e-06, | |
| "loss": 0.2571, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.9385964912280702, | |
| "grad_norm": 5.960445880889893, | |
| "learning_rate": 4.998227734329177e-06, | |
| "loss": 0.2847, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.9473684210526315, | |
| "grad_norm": 5.437699794769287, | |
| "learning_rate": 4.9976852877555755e-06, | |
| "loss": 0.2728, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.956140350877193, | |
| "grad_norm": 3.379631280899048, | |
| "learning_rate": 4.997070562417602e-06, | |
| "loss": 0.2467, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.9649122807017544, | |
| "grad_norm": 3.1625075340270996, | |
| "learning_rate": 4.996383576104362e-06, | |
| "loss": 0.2273, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.9736842105263158, | |
| "grad_norm": 15.588600158691406, | |
| "learning_rate": 4.995624348696071e-06, | |
| "loss": 0.2486, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.9824561403508771, | |
| "grad_norm": 2.631044387817383, | |
| "learning_rate": 4.9947929021634815e-06, | |
| "loss": 0.1964, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.9912280701754386, | |
| "grad_norm": 4.706504821777344, | |
| "learning_rate": 4.993889260567239e-06, | |
| "loss": 0.1901, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 10.368465423583984, | |
| "learning_rate": 4.9929134500571954e-06, | |
| "loss": 0.1996, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 1.0087719298245614, | |
| "grad_norm": 30.44986343383789, | |
| "learning_rate": 4.991865498871647e-06, | |
| "loss": 0.2606, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 1.0175438596491229, | |
| "grad_norm": 14.421515464782715, | |
| "learning_rate": 4.99074543733652e-06, | |
| "loss": 0.2394, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 1.0263157894736843, | |
| "grad_norm": 14.072005271911621, | |
| "learning_rate": 4.989553297864489e-06, | |
| "loss": 0.2288, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 1.0350877192982457, | |
| "grad_norm": 4.395325660705566, | |
| "learning_rate": 4.988289114954045e-06, | |
| "loss": 0.2129, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 1.043859649122807, | |
| "grad_norm": 7.286703586578369, | |
| "learning_rate": 4.986952925188489e-06, | |
| "loss": 0.186, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 1.0526315789473684, | |
| "grad_norm": 8.332784652709961, | |
| "learning_rate": 4.98554476723488e-06, | |
| "loss": 0.178, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 1.0614035087719298, | |
| "grad_norm": 1.3646447658538818, | |
| "learning_rate": 4.984064681842917e-06, | |
| "loss": 0.1687, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 1.0701754385964912, | |
| "grad_norm": 4.494940757751465, | |
| "learning_rate": 4.982512711843753e-06, | |
| "loss": 0.1881, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 1.0789473684210527, | |
| "grad_norm": 3.3929836750030518, | |
| "learning_rate": 4.980888902148757e-06, | |
| "loss": 0.1764, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 1.087719298245614, | |
| "grad_norm": 1.8281155824661255, | |
| "learning_rate": 4.979193299748225e-06, | |
| "loss": 0.1602, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 1.0964912280701755, | |
| "grad_norm": 3.494239568710327, | |
| "learning_rate": 4.977425953710005e-06, | |
| "loss": 0.1729, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 1.1052631578947367, | |
| "grad_norm": 1.500410556793213, | |
| "learning_rate": 4.975586915178084e-06, | |
| "loss": 0.1666, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 1.1140350877192982, | |
| "grad_norm": 1.4680222272872925, | |
| "learning_rate": 4.973676237371111e-06, | |
| "loss": 0.159, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 1.1228070175438596, | |
| "grad_norm": 3.0383460521698, | |
| "learning_rate": 4.971693975580851e-06, | |
| "loss": 0.1484, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 1.131578947368421, | |
| "grad_norm": 3.74821138381958, | |
| "learning_rate": 4.969640187170591e-06, | |
| "loss": 0.1586, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 1.1403508771929824, | |
| "grad_norm": 4.682602405548096, | |
| "learning_rate": 4.967514931573473e-06, | |
| "loss": 0.1619, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 1.1491228070175439, | |
| "grad_norm": 3.90673565864563, | |
| "learning_rate": 4.965318270290779e-06, | |
| "loss": 0.164, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 1.1578947368421053, | |
| "grad_norm": 2.2017388343811035, | |
| "learning_rate": 4.963050266890152e-06, | |
| "loss": 0.1499, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 1.1666666666666667, | |
| "grad_norm": 2.4211816787719727, | |
| "learning_rate": 4.960710987003753e-06, | |
| "loss": 0.1387, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 1.1754385964912282, | |
| "grad_norm": 1.7753759622573853, | |
| "learning_rate": 4.958300498326363e-06, | |
| "loss": 0.1441, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 1.1842105263157894, | |
| "grad_norm": 1.5529910326004028, | |
| "learning_rate": 4.955818870613425e-06, | |
| "loss": 0.1304, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 1.1929824561403508, | |
| "grad_norm": 2.090593099594116, | |
| "learning_rate": 4.953266175679023e-06, | |
| "loss": 0.1419, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 1.2017543859649122, | |
| "grad_norm": 2.7141878604888916, | |
| "learning_rate": 4.95064248739381e-06, | |
| "loss": 0.1444, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 1.2105263157894737, | |
| "grad_norm": 2.3690481185913086, | |
| "learning_rate": 4.947947881682861e-06, | |
| "loss": 0.1383, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 1.219298245614035, | |
| "grad_norm": 2.2403147220611572, | |
| "learning_rate": 4.945182436523482e-06, | |
| "loss": 0.1418, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 1.2280701754385965, | |
| "grad_norm": 1.3939160108566284, | |
| "learning_rate": 4.942346231942955e-06, | |
| "loss": 0.1307, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 1.236842105263158, | |
| "grad_norm": 11.276732444763184, | |
| "learning_rate": 4.939439350016214e-06, | |
| "loss": 0.1397, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 1.2456140350877192, | |
| "grad_norm": 8.260516166687012, | |
| "learning_rate": 4.9364618748634794e-06, | |
| "loss": 0.1426, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 1.2543859649122808, | |
| "grad_norm": 2.09720516204834, | |
| "learning_rate": 4.933413892647819e-06, | |
| "loss": 0.1323, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 1.263157894736842, | |
| "grad_norm": 1.802125334739685, | |
| "learning_rate": 4.9302954915726535e-06, | |
| "loss": 0.1304, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 1.2719298245614035, | |
| "grad_norm": 1.7151471376419067, | |
| "learning_rate": 4.927106761879207e-06, | |
| "loss": 0.1264, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 1.280701754385965, | |
| "grad_norm": 1.6970336437225342, | |
| "learning_rate": 4.923847795843894e-06, | |
| "loss": 0.1227, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 1.2894736842105263, | |
| "grad_norm": 16.60441017150879, | |
| "learning_rate": 4.920518687775647e-06, | |
| "loss": 0.1606, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 1.2982456140350878, | |
| "grad_norm": 6.470354080200195, | |
| "learning_rate": 4.917119534013194e-06, | |
| "loss": 0.1447, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 1.3070175438596492, | |
| "grad_norm": 1.4908231496810913, | |
| "learning_rate": 4.913650432922264e-06, | |
| "loss": 0.1343, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 1.3157894736842106, | |
| "grad_norm": 3.19964861869812, | |
| "learning_rate": 4.91011148489274e-06, | |
| "loss": 0.1354, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 1.3245614035087718, | |
| "grad_norm": 2.6052839756011963, | |
| "learning_rate": 4.906502792335761e-06, | |
| "loss": 0.1342, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 1.3333333333333333, | |
| "grad_norm": 2.0719165802001953, | |
| "learning_rate": 4.9028244596807525e-06, | |
| "loss": 0.1359, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 1.3421052631578947, | |
| "grad_norm": 0.8086919784545898, | |
| "learning_rate": 4.899076593372405e-06, | |
| "loss": 0.1279, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 1.3508771929824561, | |
| "grad_norm": 1.0056848526000977, | |
| "learning_rate": 4.8952593018675955e-06, | |
| "loss": 0.1162, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 1.3596491228070176, | |
| "grad_norm": 5.72553014755249, | |
| "learning_rate": 4.891372695632249e-06, | |
| "loss": 0.1315, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 1.368421052631579, | |
| "grad_norm": 1.522894024848938, | |
| "learning_rate": 4.887416887138139e-06, | |
| "loss": 0.1266, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 1.3771929824561404, | |
| "grad_norm": 2.019472122192383, | |
| "learning_rate": 4.883391990859635e-06, | |
| "loss": 0.1262, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 1.3859649122807016, | |
| "grad_norm": 1.8594422340393066, | |
| "learning_rate": 4.879298123270391e-06, | |
| "loss": 0.125, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 1.3947368421052633, | |
| "grad_norm": 1.365377426147461, | |
| "learning_rate": 4.8751354028399725e-06, | |
| "loss": 0.1218, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 1.4035087719298245, | |
| "grad_norm": 3.553309917449951, | |
| "learning_rate": 4.870903950030429e-06, | |
| "loss": 0.1272, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 1.412280701754386, | |
| "grad_norm": 2.1770920753479004, | |
| "learning_rate": 4.866603887292809e-06, | |
| "loss": 0.1213, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 1.4210526315789473, | |
| "grad_norm": 1.6058955192565918, | |
| "learning_rate": 4.862235339063613e-06, | |
| "loss": 0.1173, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 1.4298245614035088, | |
| "grad_norm": 1.3208314180374146, | |
| "learning_rate": 4.857798431761199e-06, | |
| "loss": 0.1183, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 1.4385964912280702, | |
| "grad_norm": 1.282729983329773, | |
| "learning_rate": 4.853293293782118e-06, | |
| "loss": 0.1209, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 1.4473684210526316, | |
| "grad_norm": 1.3838152885437012, | |
| "learning_rate": 4.848720055497401e-06, | |
| "loss": 0.1198, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 1.456140350877193, | |
| "grad_norm": 1.2930737733840942, | |
| "learning_rate": 4.844078849248785e-06, | |
| "loss": 0.1268, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 1.4649122807017543, | |
| "grad_norm": 1.7022266387939453, | |
| "learning_rate": 4.839369809344888e-06, | |
| "loss": 0.1198, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 1.4736842105263157, | |
| "grad_norm": 1.0927815437316895, | |
| "learning_rate": 4.834593072057313e-06, | |
| "loss": 0.1132, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 1.4824561403508771, | |
| "grad_norm": 0.9326333999633789, | |
| "learning_rate": 4.829748775616716e-06, | |
| "loss": 0.1193, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 1.4912280701754386, | |
| "grad_norm": 1.3564742803573608, | |
| "learning_rate": 4.8248370602087954e-06, | |
| "loss": 0.118, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 1.5, | |
| "grad_norm": 1.19778573513031, | |
| "learning_rate": 4.819858067970243e-06, | |
| "loss": 0.1122, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 1.5087719298245614, | |
| "grad_norm": 2.8438351154327393, | |
| "learning_rate": 4.814811942984625e-06, | |
| "loss": 0.1217, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 1.5175438596491229, | |
| "grad_norm": 1.0701063871383667, | |
| "learning_rate": 4.809698831278217e-06, | |
| "loss": 0.1114, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 1.526315789473684, | |
| "grad_norm": 0.9053553938865662, | |
| "learning_rate": 4.804518880815776e-06, | |
| "loss": 0.1178, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 1.5350877192982457, | |
| "grad_norm": 0.42274603247642517, | |
| "learning_rate": 4.799272241496259e-06, | |
| "loss": 0.1091, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 1.543859649122807, | |
| "grad_norm": 0.8576470017433167, | |
| "learning_rate": 4.793959065148484e-06, | |
| "loss": 0.1134, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 1.5526315789473686, | |
| "grad_norm": 0.5910662412643433, | |
| "learning_rate": 4.78857950552674e-06, | |
| "loss": 0.1148, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 1.5614035087719298, | |
| "grad_norm": 0.8761632442474365, | |
| "learning_rate": 4.783133718306331e-06, | |
| "loss": 0.1125, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 1.5701754385964912, | |
| "grad_norm": 1.9190795421600342, | |
| "learning_rate": 4.777621861079079e-06, | |
| "loss": 0.1148, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 1.5789473684210527, | |
| "grad_norm": 0.6199957728385925, | |
| "learning_rate": 4.772044093348757e-06, | |
| "loss": 0.1097, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 1.587719298245614, | |
| "grad_norm": 1.562089443206787, | |
| "learning_rate": 4.766400576526479e-06, | |
| "loss": 0.1097, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 1.5964912280701755, | |
| "grad_norm": 1.4957091808319092, | |
| "learning_rate": 4.760691473926021e-06, | |
| "loss": 0.1216, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 1.6052631578947367, | |
| "grad_norm": 0.9863570332527161, | |
| "learning_rate": 4.754916950759105e-06, | |
| "loss": 0.1122, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 1.6140350877192984, | |
| "grad_norm": 0.5803346633911133, | |
| "learning_rate": 4.749077174130609e-06, | |
| "loss": 0.1103, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 1.6228070175438596, | |
| "grad_norm": 1.8789891004562378, | |
| "learning_rate": 4.743172313033738e-06, | |
| "loss": 0.1191, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 1.631578947368421, | |
| "grad_norm": 0.8731380105018616, | |
| "learning_rate": 4.7372025383451285e-06, | |
| "loss": 0.1154, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 1.6403508771929824, | |
| "grad_norm": 1.3535627126693726, | |
| "learning_rate": 4.7311680228199075e-06, | |
| "loss": 0.1123, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 1.6491228070175439, | |
| "grad_norm": 0.7211089134216309, | |
| "learning_rate": 4.725068941086693e-06, | |
| "loss": 0.1134, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 1.6578947368421053, | |
| "grad_norm": 1.4752328395843506, | |
| "learning_rate": 4.718905469642534e-06, | |
| "loss": 0.1185, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 1.6666666666666665, | |
| "grad_norm": 0.9822680354118347, | |
| "learning_rate": 4.712677786847814e-06, | |
| "loss": 0.1146, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 1.6754385964912282, | |
| "grad_norm": 1.1308330297470093, | |
| "learning_rate": 4.706386072921083e-06, | |
| "loss": 0.1061, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 1.6842105263157894, | |
| "grad_norm": 5.331939697265625, | |
| "learning_rate": 4.70003050993384e-06, | |
| "loss": 0.1153, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 1.692982456140351, | |
| "grad_norm": 0.6911673545837402, | |
| "learning_rate": 4.6936112818052674e-06, | |
| "loss": 0.1098, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 1.7017543859649122, | |
| "grad_norm": 0.5160980224609375, | |
| "learning_rate": 4.687128574296912e-06, | |
| "loss": 0.1073, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 1.7105263157894737, | |
| "grad_norm": 1.5724798440933228, | |
| "learning_rate": 4.680582575007303e-06, | |
| "loss": 0.121, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 1.719298245614035, | |
| "grad_norm": 1.3960011005401611, | |
| "learning_rate": 4.6739734733665275e-06, | |
| "loss": 0.1145, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 1.7280701754385965, | |
| "grad_norm": 1.4949183464050293, | |
| "learning_rate": 4.6673014606307465e-06, | |
| "loss": 0.1166, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 1.736842105263158, | |
| "grad_norm": 1.6873422861099243, | |
| "learning_rate": 4.660566729876661e-06, | |
| "loss": 0.1115, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 1.7456140350877192, | |
| "grad_norm": 1.3443641662597656, | |
| "learning_rate": 4.653769475995926e-06, | |
| "loss": 0.1119, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 1.7543859649122808, | |
| "grad_norm": 0.807525098323822, | |
| "learning_rate": 4.646909895689508e-06, | |
| "loss": 0.1059, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 1.763157894736842, | |
| "grad_norm": 1.589316964149475, | |
| "learning_rate": 4.639988187461995e-06, | |
| "loss": 0.1151, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 1.7719298245614035, | |
| "grad_norm": 2.474756956100464, | |
| "learning_rate": 4.633004551615851e-06, | |
| "loss": 0.116, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 1.780701754385965, | |
| "grad_norm": 0.6210195422172546, | |
| "learning_rate": 4.62595919024562e-06, | |
| "loss": 0.1097, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 1.7894736842105263, | |
| "grad_norm": 0.7217905521392822, | |
| "learning_rate": 4.618852307232078e-06, | |
| "loss": 0.1117, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 1.7982456140350878, | |
| "grad_norm": 1.551251769065857, | |
| "learning_rate": 4.611684108236334e-06, | |
| "loss": 0.113, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 1.807017543859649, | |
| "grad_norm": 0.6619828939437866, | |
| "learning_rate": 4.604454800693874e-06, | |
| "loss": 0.113, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 1.8157894736842106, | |
| "grad_norm": 0.9461805820465088, | |
| "learning_rate": 4.597164593808564e-06, | |
| "loss": 0.1093, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 1.8245614035087718, | |
| "grad_norm": 1.2926547527313232, | |
| "learning_rate": 4.589813698546592e-06, | |
| "loss": 0.1128, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 1.8333333333333335, | |
| "grad_norm": 0.8754212856292725, | |
| "learning_rate": 4.582402327630368e-06, | |
| "loss": 0.1104, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 1.8421052631578947, | |
| "grad_norm": 0.846051812171936, | |
| "learning_rate": 4.574930695532357e-06, | |
| "loss": 0.1105, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 1.8508771929824561, | |
| "grad_norm": 1.3332515954971313, | |
| "learning_rate": 4.567399018468889e-06, | |
| "loss": 0.1101, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 1.8596491228070176, | |
| "grad_norm": 0.8729192614555359, | |
| "learning_rate": 4.5598075143938855e-06, | |
| "loss": 0.1081, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 1.868421052631579, | |
| "grad_norm": 0.8618345260620117, | |
| "learning_rate": 4.552156402992567e-06, | |
| "loss": 0.1059, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 1.8771929824561404, | |
| "grad_norm": 1.2135930061340332, | |
| "learning_rate": 4.544445905675082e-06, | |
| "loss": 0.1105, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 1.8859649122807016, | |
| "grad_norm": 0.8405666351318359, | |
| "learning_rate": 4.536676245570111e-06, | |
| "loss": 0.1118, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 1.8947368421052633, | |
| "grad_norm": 0.42860639095306396, | |
| "learning_rate": 4.528847647518403e-06, | |
| "loss": 0.1093, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 1.9035087719298245, | |
| "grad_norm": 1.1538206338882446, | |
| "learning_rate": 4.520960338066271e-06, | |
| "loss": 0.1088, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 1.912280701754386, | |
| "grad_norm": 0.5870749354362488, | |
| "learning_rate": 4.513014545459038e-06, | |
| "loss": 0.1061, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 1.9210526315789473, | |
| "grad_norm": 0.7279748916625977, | |
| "learning_rate": 4.505010499634427e-06, | |
| "loss": 0.1032, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 1.9298245614035088, | |
| "grad_norm": 0.6331414580345154, | |
| "learning_rate": 4.4969484322159125e-06, | |
| "loss": 0.1109, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 1.9385964912280702, | |
| "grad_norm": 0.9024543166160583, | |
| "learning_rate": 4.488828576506014e-06, | |
| "loss": 0.1094, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 1.9473684210526314, | |
| "grad_norm": 3.540376901626587, | |
| "learning_rate": 4.480651167479545e-06, | |
| "loss": 0.1154, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 1.956140350877193, | |
| "grad_norm": 0.9506739377975464, | |
| "learning_rate": 4.472416441776817e-06, | |
| "loss": 0.108, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 1.9649122807017543, | |
| "grad_norm": 0.6585081815719604, | |
| "learning_rate": 4.464124637696786e-06, | |
| "loss": 0.1033, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 1.973684210526316, | |
| "grad_norm": 1.143038034439087, | |
| "learning_rate": 4.455775995190161e-06, | |
| "loss": 0.1092, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 1.9824561403508771, | |
| "grad_norm": 1.148261547088623, | |
| "learning_rate": 4.4473707558524555e-06, | |
| "loss": 0.1076, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 1.9912280701754386, | |
| "grad_norm": 0.7375811338424683, | |
| "learning_rate": 4.438909162917003e-06, | |
| "loss": 0.108, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 0.5254591703414917, | |
| "learning_rate": 4.430391461247911e-06, | |
| "loss": 0.1079, | |
| "step": 228 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 684, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 6, | |
| "save_steps": 114, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 6.91998014829704e+18, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |