| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 1.0, | |
| "eval_steps": 0, | |
| "global_step": 325, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.003076923076923077, | |
| "grad_norm": 0.384765625, | |
| "learning_rate": 9.96923076923077e-06, | |
| "loss": 1.3452, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.006153846153846154, | |
| "grad_norm": 0.392578125, | |
| "learning_rate": 9.93846153846154e-06, | |
| "loss": 1.3737, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.009230769230769232, | |
| "grad_norm": 0.392578125, | |
| "learning_rate": 9.907692307692309e-06, | |
| "loss": 1.3808, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.012307692307692308, | |
| "grad_norm": 0.3671875, | |
| "learning_rate": 9.876923076923077e-06, | |
| "loss": 1.2991, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.015384615384615385, | |
| "grad_norm": 0.408203125, | |
| "learning_rate": 9.846153846153848e-06, | |
| "loss": 1.3539, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.018461538461538463, | |
| "grad_norm": 0.380859375, | |
| "learning_rate": 9.815384615384617e-06, | |
| "loss": 1.3463, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.021538461538461538, | |
| "grad_norm": 0.357421875, | |
| "learning_rate": 9.784615384615387e-06, | |
| "loss": 1.2862, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.024615384615384615, | |
| "grad_norm": 0.369140625, | |
| "learning_rate": 9.753846153846154e-06, | |
| "loss": 1.3459, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.027692307692307693, | |
| "grad_norm": 0.341796875, | |
| "learning_rate": 9.723076923076924e-06, | |
| "loss": 1.181, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.03076923076923077, | |
| "grad_norm": 0.34375, | |
| "learning_rate": 9.692307692307693e-06, | |
| "loss": 1.2576, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.033846153846153845, | |
| "grad_norm": 0.328125, | |
| "learning_rate": 9.661538461538462e-06, | |
| "loss": 1.2593, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.036923076923076927, | |
| "grad_norm": 0.29296875, | |
| "learning_rate": 9.630769230769232e-06, | |
| "loss": 1.2153, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 0.291015625, | |
| "learning_rate": 9.600000000000001e-06, | |
| "loss": 1.2342, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.043076923076923075, | |
| "grad_norm": 0.30078125, | |
| "learning_rate": 9.569230769230769e-06, | |
| "loss": 1.283, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.046153846153846156, | |
| "grad_norm": 0.26953125, | |
| "learning_rate": 9.53846153846154e-06, | |
| "loss": 1.2321, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.04923076923076923, | |
| "grad_norm": 0.310546875, | |
| "learning_rate": 9.50769230769231e-06, | |
| "loss": 1.2809, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.052307692307692305, | |
| "grad_norm": 0.236328125, | |
| "learning_rate": 9.476923076923079e-06, | |
| "loss": 1.0721, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.055384615384615386, | |
| "grad_norm": 0.384765625, | |
| "learning_rate": 9.446153846153847e-06, | |
| "loss": 1.1137, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.05846153846153846, | |
| "grad_norm": 0.248046875, | |
| "learning_rate": 9.415384615384616e-06, | |
| "loss": 1.1455, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.06153846153846154, | |
| "grad_norm": 0.267578125, | |
| "learning_rate": 9.384615384615385e-06, | |
| "loss": 1.144, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.06461538461538462, | |
| "grad_norm": 0.228515625, | |
| "learning_rate": 9.353846153846155e-06, | |
| "loss": 1.0787, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.06769230769230769, | |
| "grad_norm": 0.240234375, | |
| "learning_rate": 9.323076923076924e-06, | |
| "loss": 1.196, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.07076923076923076, | |
| "grad_norm": 0.2099609375, | |
| "learning_rate": 9.292307692307694e-06, | |
| "loss": 1.0622, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.07384615384615385, | |
| "grad_norm": 0.2421875, | |
| "learning_rate": 9.261538461538461e-06, | |
| "loss": 1.2206, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.07692307692307693, | |
| "grad_norm": 0.2021484375, | |
| "learning_rate": 9.230769230769232e-06, | |
| "loss": 1.0532, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 0.2197265625, | |
| "learning_rate": 9.200000000000002e-06, | |
| "loss": 1.0809, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.08307692307692308, | |
| "grad_norm": 0.201171875, | |
| "learning_rate": 9.169230769230771e-06, | |
| "loss": 1.1008, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.08615384615384615, | |
| "grad_norm": 0.2119140625, | |
| "learning_rate": 9.138461538461539e-06, | |
| "loss": 1.1231, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.08923076923076922, | |
| "grad_norm": 0.1884765625, | |
| "learning_rate": 9.107692307692308e-06, | |
| "loss": 1.0706, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.09230769230769231, | |
| "grad_norm": 0.197265625, | |
| "learning_rate": 9.076923076923078e-06, | |
| "loss": 1.078, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.09538461538461539, | |
| "grad_norm": 0.1884765625, | |
| "learning_rate": 9.046153846153847e-06, | |
| "loss": 1.0354, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.09846153846153846, | |
| "grad_norm": 0.1943359375, | |
| "learning_rate": 9.015384615384616e-06, | |
| "loss": 1.082, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.10153846153846154, | |
| "grad_norm": 0.1630859375, | |
| "learning_rate": 8.984615384615386e-06, | |
| "loss": 0.9989, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.10461538461538461, | |
| "grad_norm": 0.162109375, | |
| "learning_rate": 8.953846153846153e-06, | |
| "loss": 1.0107, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.1076923076923077, | |
| "grad_norm": 0.1845703125, | |
| "learning_rate": 8.923076923076925e-06, | |
| "loss": 1.0773, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.11076923076923077, | |
| "grad_norm": 0.1640625, | |
| "learning_rate": 8.892307692307694e-06, | |
| "loss": 1.0405, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.11384615384615385, | |
| "grad_norm": 0.1884765625, | |
| "learning_rate": 8.861538461538463e-06, | |
| "loss": 1.0293, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.11692307692307692, | |
| "grad_norm": 0.1953125, | |
| "learning_rate": 8.830769230769231e-06, | |
| "loss": 1.1177, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "grad_norm": 0.1689453125, | |
| "learning_rate": 8.8e-06, | |
| "loss": 1.0002, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.12307692307692308, | |
| "grad_norm": 0.1572265625, | |
| "learning_rate": 8.76923076923077e-06, | |
| "loss": 0.985, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.12615384615384614, | |
| "grad_norm": 0.1572265625, | |
| "learning_rate": 8.73846153846154e-06, | |
| "loss": 0.9892, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.12923076923076923, | |
| "grad_norm": 0.162109375, | |
| "learning_rate": 8.707692307692309e-06, | |
| "loss": 0.9789, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.13230769230769232, | |
| "grad_norm": 0.16796875, | |
| "learning_rate": 8.676923076923078e-06, | |
| "loss": 0.9622, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.13538461538461538, | |
| "grad_norm": 0.16796875, | |
| "learning_rate": 8.646153846153846e-06, | |
| "loss": 0.9506, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.13846153846153847, | |
| "grad_norm": 0.19140625, | |
| "learning_rate": 8.615384615384617e-06, | |
| "loss": 0.9658, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.14153846153846153, | |
| "grad_norm": 0.1708984375, | |
| "learning_rate": 8.584615384615386e-06, | |
| "loss": 0.9896, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.14461538461538462, | |
| "grad_norm": 0.169921875, | |
| "learning_rate": 8.553846153846156e-06, | |
| "loss": 1.0076, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.1476923076923077, | |
| "grad_norm": 0.146484375, | |
| "learning_rate": 8.523076923076923e-06, | |
| "loss": 0.9404, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.15076923076923077, | |
| "grad_norm": 0.1435546875, | |
| "learning_rate": 8.492307692307693e-06, | |
| "loss": 0.9393, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.15384615384615385, | |
| "grad_norm": 0.166015625, | |
| "learning_rate": 8.461538461538462e-06, | |
| "loss": 0.9861, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.15692307692307692, | |
| "grad_norm": 0.2109375, | |
| "learning_rate": 8.430769230769231e-06, | |
| "loss": 0.9313, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 0.16796875, | |
| "learning_rate": 8.400000000000001e-06, | |
| "loss": 0.985, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.16307692307692306, | |
| "grad_norm": 0.1572265625, | |
| "learning_rate": 8.36923076923077e-06, | |
| "loss": 0.9624, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.16615384615384615, | |
| "grad_norm": 0.1474609375, | |
| "learning_rate": 8.338461538461538e-06, | |
| "loss": 0.9172, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.16923076923076924, | |
| "grad_norm": 0.1455078125, | |
| "learning_rate": 8.307692307692309e-06, | |
| "loss": 0.9278, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.1723076923076923, | |
| "grad_norm": 0.154296875, | |
| "learning_rate": 8.276923076923078e-06, | |
| "loss": 0.9424, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.1753846153846154, | |
| "grad_norm": 0.162109375, | |
| "learning_rate": 8.246153846153848e-06, | |
| "loss": 0.92, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.17846153846153845, | |
| "grad_norm": 0.1572265625, | |
| "learning_rate": 8.215384615384616e-06, | |
| "loss": 0.9518, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.18153846153846154, | |
| "grad_norm": 0.1484375, | |
| "learning_rate": 8.184615384615385e-06, | |
| "loss": 0.9611, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.18461538461538463, | |
| "grad_norm": 0.13671875, | |
| "learning_rate": 8.153846153846154e-06, | |
| "loss": 0.9184, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.18769230769230769, | |
| "grad_norm": 0.130859375, | |
| "learning_rate": 8.123076923076924e-06, | |
| "loss": 0.8353, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.19076923076923077, | |
| "grad_norm": 0.1650390625, | |
| "learning_rate": 8.092307692307693e-06, | |
| "loss": 0.8753, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.19384615384615383, | |
| "grad_norm": 0.1689453125, | |
| "learning_rate": 8.061538461538463e-06, | |
| "loss": 0.9661, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.19692307692307692, | |
| "grad_norm": 0.1279296875, | |
| "learning_rate": 8.03076923076923e-06, | |
| "loss": 0.8651, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "grad_norm": 0.181640625, | |
| "learning_rate": 8.000000000000001e-06, | |
| "loss": 0.8682, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.20307692307692307, | |
| "grad_norm": 0.14453125, | |
| "learning_rate": 7.96923076923077e-06, | |
| "loss": 0.8992, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.20615384615384616, | |
| "grad_norm": 0.1611328125, | |
| "learning_rate": 7.93846153846154e-06, | |
| "loss": 0.9302, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.20923076923076922, | |
| "grad_norm": 0.1455078125, | |
| "learning_rate": 7.907692307692308e-06, | |
| "loss": 0.9019, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.2123076923076923, | |
| "grad_norm": 0.15625, | |
| "learning_rate": 7.876923076923077e-06, | |
| "loss": 0.8349, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.2153846153846154, | |
| "grad_norm": 0.1630859375, | |
| "learning_rate": 7.846153846153847e-06, | |
| "loss": 0.9397, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.21846153846153846, | |
| "grad_norm": 0.14453125, | |
| "learning_rate": 7.815384615384616e-06, | |
| "loss": 0.8954, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.22153846153846155, | |
| "grad_norm": 0.1533203125, | |
| "learning_rate": 7.784615384615385e-06, | |
| "loss": 0.8975, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.2246153846153846, | |
| "grad_norm": 0.15625, | |
| "learning_rate": 7.753846153846155e-06, | |
| "loss": 0.8709, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.2276923076923077, | |
| "grad_norm": 0.255859375, | |
| "learning_rate": 7.723076923076924e-06, | |
| "loss": 0.9431, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.23076923076923078, | |
| "grad_norm": 0.2236328125, | |
| "learning_rate": 7.692307692307694e-06, | |
| "loss": 0.9374, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.23384615384615384, | |
| "grad_norm": 0.1728515625, | |
| "learning_rate": 7.661538461538463e-06, | |
| "loss": 0.8611, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.23692307692307693, | |
| "grad_norm": 0.1787109375, | |
| "learning_rate": 7.630769230769232e-06, | |
| "loss": 0.866, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "grad_norm": 0.154296875, | |
| "learning_rate": 7.600000000000001e-06, | |
| "loss": 0.8449, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.24307692307692308, | |
| "grad_norm": 0.154296875, | |
| "learning_rate": 7.5692307692307695e-06, | |
| "loss": 0.9007, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.24615384615384617, | |
| "grad_norm": 0.1796875, | |
| "learning_rate": 7.538461538461539e-06, | |
| "loss": 0.8798, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.24923076923076923, | |
| "grad_norm": 0.1474609375, | |
| "learning_rate": 7.507692307692308e-06, | |
| "loss": 0.875, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.2523076923076923, | |
| "grad_norm": 0.1396484375, | |
| "learning_rate": 7.476923076923077e-06, | |
| "loss": 0.8476, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.2553846153846154, | |
| "grad_norm": 0.1396484375, | |
| "learning_rate": 7.446153846153846e-06, | |
| "loss": 0.895, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.25846153846153846, | |
| "grad_norm": 0.16015625, | |
| "learning_rate": 7.4153846153846164e-06, | |
| "loss": 0.8767, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.26153846153846155, | |
| "grad_norm": 0.1533203125, | |
| "learning_rate": 7.384615384615386e-06, | |
| "loss": 0.8622, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.26461538461538464, | |
| "grad_norm": 0.1474609375, | |
| "learning_rate": 7.353846153846154e-06, | |
| "loss": 0.873, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.2676923076923077, | |
| "grad_norm": 0.154296875, | |
| "learning_rate": 7.323076923076924e-06, | |
| "loss": 0.8939, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.27076923076923076, | |
| "grad_norm": 0.224609375, | |
| "learning_rate": 7.292307692307693e-06, | |
| "loss": 0.9088, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.27384615384615385, | |
| "grad_norm": 0.14453125, | |
| "learning_rate": 7.261538461538462e-06, | |
| "loss": 0.8623, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.27692307692307694, | |
| "grad_norm": 0.13671875, | |
| "learning_rate": 7.230769230769231e-06, | |
| "loss": 0.8181, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "grad_norm": 0.1484375, | |
| "learning_rate": 7.2000000000000005e-06, | |
| "loss": 0.8414, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.28307692307692306, | |
| "grad_norm": 0.1591796875, | |
| "learning_rate": 7.169230769230769e-06, | |
| "loss": 0.7819, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.28615384615384615, | |
| "grad_norm": 0.1640625, | |
| "learning_rate": 7.1384615384615385e-06, | |
| "loss": 0.868, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.28923076923076924, | |
| "grad_norm": 0.1787109375, | |
| "learning_rate": 7.107692307692309e-06, | |
| "loss": 0.8873, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.2923076923076923, | |
| "grad_norm": 0.16015625, | |
| "learning_rate": 7.076923076923078e-06, | |
| "loss": 0.9212, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.2953846153846154, | |
| "grad_norm": 0.146484375, | |
| "learning_rate": 7.046153846153847e-06, | |
| "loss": 0.8661, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.29846153846153844, | |
| "grad_norm": 0.2099609375, | |
| "learning_rate": 7.015384615384616e-06, | |
| "loss": 0.807, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.30153846153846153, | |
| "grad_norm": 0.158203125, | |
| "learning_rate": 6.9846153846153854e-06, | |
| "loss": 0.8279, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.3046153846153846, | |
| "grad_norm": 0.2021484375, | |
| "learning_rate": 6.953846153846154e-06, | |
| "loss": 0.8157, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.3076923076923077, | |
| "grad_norm": 0.177734375, | |
| "learning_rate": 6.923076923076923e-06, | |
| "loss": 0.8808, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.31076923076923074, | |
| "grad_norm": 0.1357421875, | |
| "learning_rate": 6.892307692307693e-06, | |
| "loss": 0.835, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.31384615384615383, | |
| "grad_norm": 0.13671875, | |
| "learning_rate": 6.861538461538461e-06, | |
| "loss": 0.7776, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.3169230769230769, | |
| "grad_norm": 0.1318359375, | |
| "learning_rate": 6.830769230769231e-06, | |
| "loss": 0.7889, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "grad_norm": 0.1455078125, | |
| "learning_rate": 6.800000000000001e-06, | |
| "loss": 0.8522, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.3230769230769231, | |
| "grad_norm": 0.14453125, | |
| "learning_rate": 6.76923076923077e-06, | |
| "loss": 0.8448, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.3261538461538461, | |
| "grad_norm": 0.130859375, | |
| "learning_rate": 6.738461538461539e-06, | |
| "loss": 0.8584, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.3292307692307692, | |
| "grad_norm": 0.14453125, | |
| "learning_rate": 6.707692307692308e-06, | |
| "loss": 0.8863, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.3323076923076923, | |
| "grad_norm": 0.1376953125, | |
| "learning_rate": 6.676923076923078e-06, | |
| "loss": 0.7854, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.3353846153846154, | |
| "grad_norm": 0.1416015625, | |
| "learning_rate": 6.646153846153846e-06, | |
| "loss": 0.9035, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.3384615384615385, | |
| "grad_norm": 0.1513671875, | |
| "learning_rate": 6.615384615384616e-06, | |
| "loss": 0.8527, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.3415384615384615, | |
| "grad_norm": 0.30859375, | |
| "learning_rate": 6.584615384615385e-06, | |
| "loss": 0.8383, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.3446153846153846, | |
| "grad_norm": 0.1435546875, | |
| "learning_rate": 6.553846153846154e-06, | |
| "loss": 0.8617, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.3476923076923077, | |
| "grad_norm": 0.154296875, | |
| "learning_rate": 6.523076923076923e-06, | |
| "loss": 0.8182, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.3507692307692308, | |
| "grad_norm": 0.1845703125, | |
| "learning_rate": 6.492307692307693e-06, | |
| "loss": 0.9757, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.35384615384615387, | |
| "grad_norm": 0.1513671875, | |
| "learning_rate": 6.461538461538463e-06, | |
| "loss": 0.8425, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.3569230769230769, | |
| "grad_norm": 0.142578125, | |
| "learning_rate": 6.430769230769231e-06, | |
| "loss": 0.8469, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.36, | |
| "grad_norm": 0.1328125, | |
| "learning_rate": 6.4000000000000006e-06, | |
| "loss": 0.8157, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.3630769230769231, | |
| "grad_norm": 0.150390625, | |
| "learning_rate": 6.36923076923077e-06, | |
| "loss": 0.8123, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.36615384615384616, | |
| "grad_norm": 0.1474609375, | |
| "learning_rate": 6.3384615384615385e-06, | |
| "loss": 0.8003, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.36923076923076925, | |
| "grad_norm": 0.1513671875, | |
| "learning_rate": 6.307692307692308e-06, | |
| "loss": 0.7949, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.3723076923076923, | |
| "grad_norm": 0.1728515625, | |
| "learning_rate": 6.276923076923077e-06, | |
| "loss": 0.8058, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.37538461538461537, | |
| "grad_norm": 0.265625, | |
| "learning_rate": 6.246153846153846e-06, | |
| "loss": 0.8341, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.37846153846153846, | |
| "grad_norm": 0.1552734375, | |
| "learning_rate": 6.215384615384615e-06, | |
| "loss": 0.8021, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.38153846153846155, | |
| "grad_norm": 0.1337890625, | |
| "learning_rate": 6.1846153846153855e-06, | |
| "loss": 0.7892, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.38461538461538464, | |
| "grad_norm": 0.1435546875, | |
| "learning_rate": 6.153846153846155e-06, | |
| "loss": 0.792, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.38769230769230767, | |
| "grad_norm": 0.166015625, | |
| "learning_rate": 6.123076923076923e-06, | |
| "loss": 0.8384, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.39076923076923076, | |
| "grad_norm": 0.2109375, | |
| "learning_rate": 6.092307692307693e-06, | |
| "loss": 0.791, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.39384615384615385, | |
| "grad_norm": 0.1435546875, | |
| "learning_rate": 6.061538461538462e-06, | |
| "loss": 0.847, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.39692307692307693, | |
| "grad_norm": 0.328125, | |
| "learning_rate": 6.030769230769231e-06, | |
| "loss": 0.8738, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "grad_norm": 0.1318359375, | |
| "learning_rate": 6e-06, | |
| "loss": 0.7992, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.40307692307692305, | |
| "grad_norm": 0.173828125, | |
| "learning_rate": 5.9692307692307695e-06, | |
| "loss": 0.8204, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.40615384615384614, | |
| "grad_norm": 0.1611328125, | |
| "learning_rate": 5.938461538461538e-06, | |
| "loss": 0.8229, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.40923076923076923, | |
| "grad_norm": 0.1630859375, | |
| "learning_rate": 5.907692307692308e-06, | |
| "loss": 0.8281, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.4123076923076923, | |
| "grad_norm": 0.134765625, | |
| "learning_rate": 5.876923076923078e-06, | |
| "loss": 0.795, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.4153846153846154, | |
| "grad_norm": 0.1455078125, | |
| "learning_rate": 5.846153846153847e-06, | |
| "loss": 0.8039, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.41846153846153844, | |
| "grad_norm": 0.1455078125, | |
| "learning_rate": 5.815384615384616e-06, | |
| "loss": 0.838, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.42153846153846153, | |
| "grad_norm": 0.38671875, | |
| "learning_rate": 5.784615384615385e-06, | |
| "loss": 0.84, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.4246153846153846, | |
| "grad_norm": 0.1337890625, | |
| "learning_rate": 5.7538461538461545e-06, | |
| "loss": 0.751, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.4276923076923077, | |
| "grad_norm": 0.1376953125, | |
| "learning_rate": 5.723076923076923e-06, | |
| "loss": 0.7599, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.4307692307692308, | |
| "grad_norm": 0.1669921875, | |
| "learning_rate": 5.692307692307692e-06, | |
| "loss": 0.8528, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.4338461538461538, | |
| "grad_norm": 0.158203125, | |
| "learning_rate": 5.661538461538462e-06, | |
| "loss": 0.8013, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.4369230769230769, | |
| "grad_norm": 0.15625, | |
| "learning_rate": 5.63076923076923e-06, | |
| "loss": 0.8377, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.44, | |
| "grad_norm": 0.1474609375, | |
| "learning_rate": 5.600000000000001e-06, | |
| "loss": 0.7798, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.4430769230769231, | |
| "grad_norm": 0.1318359375, | |
| "learning_rate": 5.56923076923077e-06, | |
| "loss": 0.7973, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.4461538461538462, | |
| "grad_norm": 0.169921875, | |
| "learning_rate": 5.538461538461539e-06, | |
| "loss": 0.7647, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.4492307692307692, | |
| "grad_norm": 0.14453125, | |
| "learning_rate": 5.507692307692308e-06, | |
| "loss": 0.8505, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.4523076923076923, | |
| "grad_norm": 0.154296875, | |
| "learning_rate": 5.476923076923077e-06, | |
| "loss": 0.7937, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.4553846153846154, | |
| "grad_norm": 0.142578125, | |
| "learning_rate": 5.446153846153847e-06, | |
| "loss": 0.8358, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.4584615384615385, | |
| "grad_norm": 0.1474609375, | |
| "learning_rate": 5.415384615384615e-06, | |
| "loss": 0.7735, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.46153846153846156, | |
| "grad_norm": 0.15625, | |
| "learning_rate": 5.384615384615385e-06, | |
| "loss": 0.8102, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.4646153846153846, | |
| "grad_norm": 0.146484375, | |
| "learning_rate": 5.353846153846154e-06, | |
| "loss": 0.8282, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 0.4676923076923077, | |
| "grad_norm": 0.16015625, | |
| "learning_rate": 5.323076923076923e-06, | |
| "loss": 0.7942, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 0.4707692307692308, | |
| "grad_norm": 0.15625, | |
| "learning_rate": 5.292307692307693e-06, | |
| "loss": 0.8058, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 0.47384615384615386, | |
| "grad_norm": 0.1904296875, | |
| "learning_rate": 5.261538461538462e-06, | |
| "loss": 0.7791, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 0.47692307692307695, | |
| "grad_norm": 0.1376953125, | |
| "learning_rate": 5.230769230769232e-06, | |
| "loss": 0.7829, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "grad_norm": 0.1435546875, | |
| "learning_rate": 5.2e-06, | |
| "loss": 0.7682, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 0.48307692307692307, | |
| "grad_norm": 0.150390625, | |
| "learning_rate": 5.16923076923077e-06, | |
| "loss": 0.8276, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 0.48615384615384616, | |
| "grad_norm": 0.1494140625, | |
| "learning_rate": 5.138461538461539e-06, | |
| "loss": 0.8172, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 0.48923076923076925, | |
| "grad_norm": 0.1806640625, | |
| "learning_rate": 5.1076923076923075e-06, | |
| "loss": 0.7524, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 0.49230769230769234, | |
| "grad_norm": 0.14453125, | |
| "learning_rate": 5.076923076923077e-06, | |
| "loss": 0.8516, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.49538461538461537, | |
| "grad_norm": 0.142578125, | |
| "learning_rate": 5.046153846153846e-06, | |
| "loss": 0.7736, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 0.49846153846153846, | |
| "grad_norm": 0.1484375, | |
| "learning_rate": 5.015384615384616e-06, | |
| "loss": 0.762, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 0.5015384615384615, | |
| "grad_norm": 0.1474609375, | |
| "learning_rate": 4.984615384615385e-06, | |
| "loss": 0.8315, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 0.5046153846153846, | |
| "grad_norm": 0.1376953125, | |
| "learning_rate": 4.9538461538461545e-06, | |
| "loss": 0.8148, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 0.5076923076923077, | |
| "grad_norm": 0.1435546875, | |
| "learning_rate": 4.923076923076924e-06, | |
| "loss": 0.834, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.5107692307692308, | |
| "grad_norm": 0.1552734375, | |
| "learning_rate": 4.892307692307693e-06, | |
| "loss": 0.8133, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 0.5138461538461538, | |
| "grad_norm": 0.146484375, | |
| "learning_rate": 4.861538461538462e-06, | |
| "loss": 0.7675, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 0.5169230769230769, | |
| "grad_norm": 0.2119140625, | |
| "learning_rate": 4.830769230769231e-06, | |
| "loss": 0.8251, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 0.52, | |
| "grad_norm": 0.1552734375, | |
| "learning_rate": 4.800000000000001e-06, | |
| "loss": 0.8289, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 0.5230769230769231, | |
| "grad_norm": 0.1416015625, | |
| "learning_rate": 4.76923076923077e-06, | |
| "loss": 0.7701, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.5261538461538462, | |
| "grad_norm": 0.1884765625, | |
| "learning_rate": 4.738461538461539e-06, | |
| "loss": 0.8718, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 0.5292307692307693, | |
| "grad_norm": 0.1435546875, | |
| "learning_rate": 4.707692307692308e-06, | |
| "loss": 0.7889, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 0.5323076923076923, | |
| "grad_norm": 0.1513671875, | |
| "learning_rate": 4.676923076923077e-06, | |
| "loss": 0.8025, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 0.5353846153846153, | |
| "grad_norm": 0.158203125, | |
| "learning_rate": 4.646153846153847e-06, | |
| "loss": 0.8177, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 0.5384615384615384, | |
| "grad_norm": 0.1416015625, | |
| "learning_rate": 4.615384615384616e-06, | |
| "loss": 0.7924, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.5415384615384615, | |
| "grad_norm": 0.1396484375, | |
| "learning_rate": 4.5846153846153855e-06, | |
| "loss": 0.7829, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 0.5446153846153846, | |
| "grad_norm": 0.14453125, | |
| "learning_rate": 4.553846153846154e-06, | |
| "loss": 0.7985, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 0.5476923076923077, | |
| "grad_norm": 0.1484375, | |
| "learning_rate": 4.5230769230769235e-06, | |
| "loss": 0.78, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 0.5507692307692308, | |
| "grad_norm": 0.2197265625, | |
| "learning_rate": 4.492307692307693e-06, | |
| "loss": 0.7914, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 0.5538461538461539, | |
| "grad_norm": 0.16796875, | |
| "learning_rate": 4.461538461538462e-06, | |
| "loss": 0.7994, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.556923076923077, | |
| "grad_norm": 0.1416015625, | |
| "learning_rate": 4.430769230769232e-06, | |
| "loss": 0.7736, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 0.56, | |
| "grad_norm": 0.1513671875, | |
| "learning_rate": 4.4e-06, | |
| "loss": 0.8243, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 0.563076923076923, | |
| "grad_norm": 0.1474609375, | |
| "learning_rate": 4.36923076923077e-06, | |
| "loss": 0.7979, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 0.5661538461538461, | |
| "grad_norm": 0.13671875, | |
| "learning_rate": 4.338461538461539e-06, | |
| "loss": 0.7717, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 0.5692307692307692, | |
| "grad_norm": 0.1474609375, | |
| "learning_rate": 4.307692307692308e-06, | |
| "loss": 0.7903, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.5723076923076923, | |
| "grad_norm": 0.1982421875, | |
| "learning_rate": 4.276923076923078e-06, | |
| "loss": 0.8233, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 0.5753846153846154, | |
| "grad_norm": 0.14453125, | |
| "learning_rate": 4.246153846153846e-06, | |
| "loss": 0.7867, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 0.5784615384615385, | |
| "grad_norm": 0.140625, | |
| "learning_rate": 4.215384615384616e-06, | |
| "loss": 0.7673, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 0.5815384615384616, | |
| "grad_norm": 0.1484375, | |
| "learning_rate": 4.184615384615385e-06, | |
| "loss": 0.8351, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 0.5846153846153846, | |
| "grad_norm": 0.1572265625, | |
| "learning_rate": 4.1538461538461545e-06, | |
| "loss": 0.7998, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.5876923076923077, | |
| "grad_norm": 0.18359375, | |
| "learning_rate": 4.123076923076924e-06, | |
| "loss": 0.7791, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 0.5907692307692308, | |
| "grad_norm": 0.1455078125, | |
| "learning_rate": 4.0923076923076925e-06, | |
| "loss": 0.8079, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 0.5938461538461538, | |
| "grad_norm": 0.1435546875, | |
| "learning_rate": 4.061538461538462e-06, | |
| "loss": 0.8053, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 0.5969230769230769, | |
| "grad_norm": 0.14453125, | |
| "learning_rate": 4.030769230769231e-06, | |
| "loss": 0.7641, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "grad_norm": 0.1845703125, | |
| "learning_rate": 4.000000000000001e-06, | |
| "loss": 0.763, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.6030769230769231, | |
| "grad_norm": 0.283203125, | |
| "learning_rate": 3.96923076923077e-06, | |
| "loss": 0.8159, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 0.6061538461538462, | |
| "grad_norm": 0.158203125, | |
| "learning_rate": 3.938461538461539e-06, | |
| "loss": 0.7612, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 0.6092307692307692, | |
| "grad_norm": 0.1396484375, | |
| "learning_rate": 3.907692307692308e-06, | |
| "loss": 0.767, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 0.6123076923076923, | |
| "grad_norm": 0.1474609375, | |
| "learning_rate": 3.876923076923077e-06, | |
| "loss": 0.7698, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 0.6153846153846154, | |
| "grad_norm": 0.2001953125, | |
| "learning_rate": 3.846153846153847e-06, | |
| "loss": 0.7535, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.6184615384615385, | |
| "grad_norm": 0.1875, | |
| "learning_rate": 3.815384615384616e-06, | |
| "loss": 0.7282, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 0.6215384615384615, | |
| "grad_norm": 0.1416015625, | |
| "learning_rate": 3.7846153846153847e-06, | |
| "loss": 0.7595, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 0.6246153846153846, | |
| "grad_norm": 0.1591796875, | |
| "learning_rate": 3.753846153846154e-06, | |
| "loss": 0.7814, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 0.6276923076923077, | |
| "grad_norm": 0.1591796875, | |
| "learning_rate": 3.723076923076923e-06, | |
| "loss": 0.7525, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 0.6307692307692307, | |
| "grad_norm": 0.146484375, | |
| "learning_rate": 3.692307692307693e-06, | |
| "loss": 0.7936, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.6338461538461538, | |
| "grad_norm": 0.2041015625, | |
| "learning_rate": 3.661538461538462e-06, | |
| "loss": 0.8274, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 0.6369230769230769, | |
| "grad_norm": 0.15625, | |
| "learning_rate": 3.630769230769231e-06, | |
| "loss": 0.7306, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "grad_norm": 0.1474609375, | |
| "learning_rate": 3.6000000000000003e-06, | |
| "loss": 0.7726, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 0.6430769230769231, | |
| "grad_norm": 0.13671875, | |
| "learning_rate": 3.5692307692307692e-06, | |
| "loss": 0.759, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 0.6461538461538462, | |
| "grad_norm": 0.154296875, | |
| "learning_rate": 3.538461538461539e-06, | |
| "loss": 0.8068, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.6492307692307693, | |
| "grad_norm": 0.1552734375, | |
| "learning_rate": 3.507692307692308e-06, | |
| "loss": 0.794, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 0.6523076923076923, | |
| "grad_norm": 0.1484375, | |
| "learning_rate": 3.476923076923077e-06, | |
| "loss": 0.7725, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 0.6553846153846153, | |
| "grad_norm": 0.1591796875, | |
| "learning_rate": 3.4461538461538464e-06, | |
| "loss": 0.8455, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 0.6584615384615384, | |
| "grad_norm": 0.1640625, | |
| "learning_rate": 3.4153846153846154e-06, | |
| "loss": 0.7585, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 0.6615384615384615, | |
| "grad_norm": 0.2177734375, | |
| "learning_rate": 3.384615384615385e-06, | |
| "loss": 0.8085, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.6646153846153846, | |
| "grad_norm": 0.1396484375, | |
| "learning_rate": 3.353846153846154e-06, | |
| "loss": 0.7508, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 0.6676923076923077, | |
| "grad_norm": 0.138671875, | |
| "learning_rate": 3.323076923076923e-06, | |
| "loss": 0.7185, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 0.6707692307692308, | |
| "grad_norm": 0.134765625, | |
| "learning_rate": 3.2923076923076925e-06, | |
| "loss": 0.7809, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 0.6738461538461539, | |
| "grad_norm": 0.1416015625, | |
| "learning_rate": 3.2615384615384615e-06, | |
| "loss": 0.7663, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 0.676923076923077, | |
| "grad_norm": 0.1640625, | |
| "learning_rate": 3.2307692307692313e-06, | |
| "loss": 0.7731, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.68, | |
| "grad_norm": 0.1474609375, | |
| "learning_rate": 3.2000000000000003e-06, | |
| "loss": 0.7454, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 0.683076923076923, | |
| "grad_norm": 0.1533203125, | |
| "learning_rate": 3.1692307692307693e-06, | |
| "loss": 0.7649, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 0.6861538461538461, | |
| "grad_norm": 0.13671875, | |
| "learning_rate": 3.1384615384615386e-06, | |
| "loss": 0.7566, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 0.6892307692307692, | |
| "grad_norm": 0.1435546875, | |
| "learning_rate": 3.1076923076923076e-06, | |
| "loss": 0.7649, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 0.6923076923076923, | |
| "grad_norm": 0.208984375, | |
| "learning_rate": 3.0769230769230774e-06, | |
| "loss": 0.7344, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.6953846153846154, | |
| "grad_norm": 0.142578125, | |
| "learning_rate": 3.0461538461538464e-06, | |
| "loss": 0.7808, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 0.6984615384615385, | |
| "grad_norm": 0.1630859375, | |
| "learning_rate": 3.0153846153846154e-06, | |
| "loss": 0.734, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 0.7015384615384616, | |
| "grad_norm": 0.2314453125, | |
| "learning_rate": 2.9846153846153848e-06, | |
| "loss": 0.8274, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 0.7046153846153846, | |
| "grad_norm": 0.150390625, | |
| "learning_rate": 2.953846153846154e-06, | |
| "loss": 0.8302, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 0.7076923076923077, | |
| "grad_norm": 0.244140625, | |
| "learning_rate": 2.9230769230769236e-06, | |
| "loss": 0.7678, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.7107692307692308, | |
| "grad_norm": 0.1533203125, | |
| "learning_rate": 2.8923076923076925e-06, | |
| "loss": 0.7217, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 0.7138461538461538, | |
| "grad_norm": 0.1533203125, | |
| "learning_rate": 2.8615384615384615e-06, | |
| "loss": 0.8025, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 0.7169230769230769, | |
| "grad_norm": 0.1396484375, | |
| "learning_rate": 2.830769230769231e-06, | |
| "loss": 0.7433, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "grad_norm": 0.18359375, | |
| "learning_rate": 2.8000000000000003e-06, | |
| "loss": 0.8146, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 0.7230769230769231, | |
| "grad_norm": 0.1611328125, | |
| "learning_rate": 2.7692307692307697e-06, | |
| "loss": 0.7946, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.7261538461538461, | |
| "grad_norm": 0.142578125, | |
| "learning_rate": 2.7384615384615387e-06, | |
| "loss": 0.7694, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 0.7292307692307692, | |
| "grad_norm": 0.15234375, | |
| "learning_rate": 2.7076923076923076e-06, | |
| "loss": 0.7995, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 0.7323076923076923, | |
| "grad_norm": 0.1533203125, | |
| "learning_rate": 2.676923076923077e-06, | |
| "loss": 0.8061, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 0.7353846153846154, | |
| "grad_norm": 0.1689453125, | |
| "learning_rate": 2.6461538461538464e-06, | |
| "loss": 0.8169, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 0.7384615384615385, | |
| "grad_norm": 0.1728515625, | |
| "learning_rate": 2.615384615384616e-06, | |
| "loss": 0.8631, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.7415384615384616, | |
| "grad_norm": 0.1533203125, | |
| "learning_rate": 2.584615384615385e-06, | |
| "loss": 0.7987, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 0.7446153846153846, | |
| "grad_norm": 0.185546875, | |
| "learning_rate": 2.5538461538461538e-06, | |
| "loss": 0.751, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 0.7476923076923077, | |
| "grad_norm": 0.1513671875, | |
| "learning_rate": 2.523076923076923e-06, | |
| "loss": 0.7463, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 0.7507692307692307, | |
| "grad_norm": 0.150390625, | |
| "learning_rate": 2.4923076923076926e-06, | |
| "loss": 0.7859, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 0.7538461538461538, | |
| "grad_norm": 0.1640625, | |
| "learning_rate": 2.461538461538462e-06, | |
| "loss": 0.7743, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.7569230769230769, | |
| "grad_norm": 0.15234375, | |
| "learning_rate": 2.430769230769231e-06, | |
| "loss": 0.7394, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 0.76, | |
| "grad_norm": 0.1484375, | |
| "learning_rate": 2.4000000000000003e-06, | |
| "loss": 0.778, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 0.7630769230769231, | |
| "grad_norm": 0.25390625, | |
| "learning_rate": 2.3692307692307697e-06, | |
| "loss": 0.8181, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 0.7661538461538462, | |
| "grad_norm": 0.1591796875, | |
| "learning_rate": 2.3384615384615387e-06, | |
| "loss": 0.7685, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 0.7692307692307693, | |
| "grad_norm": 0.1796875, | |
| "learning_rate": 2.307692307692308e-06, | |
| "loss": 0.7564, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.7723076923076924, | |
| "grad_norm": 0.1435546875, | |
| "learning_rate": 2.276923076923077e-06, | |
| "loss": 0.7658, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 0.7753846153846153, | |
| "grad_norm": 0.169921875, | |
| "learning_rate": 2.2461538461538464e-06, | |
| "loss": 0.8001, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 0.7784615384615384, | |
| "grad_norm": 0.1435546875, | |
| "learning_rate": 2.215384615384616e-06, | |
| "loss": 0.7584, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 0.7815384615384615, | |
| "grad_norm": 0.142578125, | |
| "learning_rate": 2.184615384615385e-06, | |
| "loss": 0.7376, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 0.7846153846153846, | |
| "grad_norm": 0.138671875, | |
| "learning_rate": 2.153846153846154e-06, | |
| "loss": 0.7896, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.7876923076923077, | |
| "grad_norm": 0.154296875, | |
| "learning_rate": 2.123076923076923e-06, | |
| "loss": 0.7725, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 0.7907692307692308, | |
| "grad_norm": 0.1591796875, | |
| "learning_rate": 2.0923076923076926e-06, | |
| "loss": 0.7605, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 0.7938461538461539, | |
| "grad_norm": 0.1513671875, | |
| "learning_rate": 2.061538461538462e-06, | |
| "loss": 0.7652, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 0.796923076923077, | |
| "grad_norm": 0.16796875, | |
| "learning_rate": 2.030769230769231e-06, | |
| "loss": 0.7545, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 0.140625, | |
| "learning_rate": 2.0000000000000003e-06, | |
| "loss": 0.7487, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.803076923076923, | |
| "grad_norm": 0.1494140625, | |
| "learning_rate": 1.9692307692307693e-06, | |
| "loss": 0.7608, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 0.8061538461538461, | |
| "grad_norm": 0.1728515625, | |
| "learning_rate": 1.9384615384615387e-06, | |
| "loss": 0.7894, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 0.8092307692307692, | |
| "grad_norm": 0.1416015625, | |
| "learning_rate": 1.907692307692308e-06, | |
| "loss": 0.7624, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 0.8123076923076923, | |
| "grad_norm": 0.162109375, | |
| "learning_rate": 1.876923076923077e-06, | |
| "loss": 0.7798, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 0.8153846153846154, | |
| "grad_norm": 0.166015625, | |
| "learning_rate": 1.8461538461538465e-06, | |
| "loss": 0.8119, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.8184615384615385, | |
| "grad_norm": 0.27734375, | |
| "learning_rate": 1.8153846153846154e-06, | |
| "loss": 0.747, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 0.8215384615384616, | |
| "grad_norm": 0.1591796875, | |
| "learning_rate": 1.7846153846153846e-06, | |
| "loss": 0.7795, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 0.8246153846153846, | |
| "grad_norm": 0.1748046875, | |
| "learning_rate": 1.753846153846154e-06, | |
| "loss": 0.7808, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 0.8276923076923077, | |
| "grad_norm": 0.2177734375, | |
| "learning_rate": 1.7230769230769232e-06, | |
| "loss": 0.763, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 0.8307692307692308, | |
| "grad_norm": 0.1416015625, | |
| "learning_rate": 1.6923076923076926e-06, | |
| "loss": 0.7718, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.8338461538461538, | |
| "grad_norm": 0.1455078125, | |
| "learning_rate": 1.6615384615384616e-06, | |
| "loss": 0.7739, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 0.8369230769230769, | |
| "grad_norm": 0.21484375, | |
| "learning_rate": 1.6307692307692307e-06, | |
| "loss": 0.8157, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 0.84, | |
| "grad_norm": 0.15625, | |
| "learning_rate": 1.6000000000000001e-06, | |
| "loss": 0.7615, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 0.8430769230769231, | |
| "grad_norm": 0.185546875, | |
| "learning_rate": 1.5692307692307693e-06, | |
| "loss": 0.7827, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 0.8461538461538461, | |
| "grad_norm": 0.162109375, | |
| "learning_rate": 1.5384615384615387e-06, | |
| "loss": 0.7695, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.8492307692307692, | |
| "grad_norm": 0.1513671875, | |
| "learning_rate": 1.5076923076923077e-06, | |
| "loss": 0.7875, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 0.8523076923076923, | |
| "grad_norm": 0.14453125, | |
| "learning_rate": 1.476923076923077e-06, | |
| "loss": 0.78, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 0.8553846153846154, | |
| "grad_norm": 0.1494140625, | |
| "learning_rate": 1.4461538461538463e-06, | |
| "loss": 0.7878, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 0.8584615384615385, | |
| "grad_norm": 0.1611328125, | |
| "learning_rate": 1.4153846153846155e-06, | |
| "loss": 0.7788, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 0.8615384615384616, | |
| "grad_norm": 0.146484375, | |
| "learning_rate": 1.3846153846153848e-06, | |
| "loss": 0.789, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.8646153846153846, | |
| "grad_norm": 0.1552734375, | |
| "learning_rate": 1.3538461538461538e-06, | |
| "loss": 0.7298, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 0.8676923076923077, | |
| "grad_norm": 0.216796875, | |
| "learning_rate": 1.3230769230769232e-06, | |
| "loss": 0.7712, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 0.8707692307692307, | |
| "grad_norm": 0.2216796875, | |
| "learning_rate": 1.2923076923076924e-06, | |
| "loss": 0.8052, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 0.8738461538461538, | |
| "grad_norm": 0.1611328125, | |
| "learning_rate": 1.2615384615384616e-06, | |
| "loss": 0.7784, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 0.8769230769230769, | |
| "grad_norm": 0.1748046875, | |
| "learning_rate": 1.230769230769231e-06, | |
| "loss": 0.7561, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "grad_norm": 0.1552734375, | |
| "learning_rate": 1.2000000000000002e-06, | |
| "loss": 0.8075, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 0.8830769230769231, | |
| "grad_norm": 0.14453125, | |
| "learning_rate": 1.1692307692307693e-06, | |
| "loss": 0.7709, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 0.8861538461538462, | |
| "grad_norm": 0.1689453125, | |
| "learning_rate": 1.1384615384615385e-06, | |
| "loss": 0.772, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 0.8892307692307693, | |
| "grad_norm": 0.1396484375, | |
| "learning_rate": 1.107692307692308e-06, | |
| "loss": 0.7596, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 0.8923076923076924, | |
| "grad_norm": 0.24609375, | |
| "learning_rate": 1.076923076923077e-06, | |
| "loss": 0.7566, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.8953846153846153, | |
| "grad_norm": 0.1650390625, | |
| "learning_rate": 1.0461538461538463e-06, | |
| "loss": 0.7683, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 0.8984615384615384, | |
| "grad_norm": 0.15625, | |
| "learning_rate": 1.0153846153846155e-06, | |
| "loss": 0.7895, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 0.9015384615384615, | |
| "grad_norm": 0.1513671875, | |
| "learning_rate": 9.846153846153847e-07, | |
| "loss": 0.7817, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 0.9046153846153846, | |
| "grad_norm": 0.1396484375, | |
| "learning_rate": 9.53846153846154e-07, | |
| "loss": 0.7426, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 0.9076923076923077, | |
| "grad_norm": 0.1552734375, | |
| "learning_rate": 9.230769230769232e-07, | |
| "loss": 0.7586, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 0.9107692307692308, | |
| "grad_norm": 0.158203125, | |
| "learning_rate": 8.923076923076923e-07, | |
| "loss": 0.8118, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 0.9138461538461539, | |
| "grad_norm": 0.1552734375, | |
| "learning_rate": 8.615384615384616e-07, | |
| "loss": 0.7822, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 0.916923076923077, | |
| "grad_norm": 0.154296875, | |
| "learning_rate": 8.307692307692308e-07, | |
| "loss": 0.7686, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 0.92, | |
| "grad_norm": 0.181640625, | |
| "learning_rate": 8.000000000000001e-07, | |
| "loss": 0.7584, | |
| "step": 299 | |
| }, | |
| { | |
| "epoch": 0.9230769230769231, | |
| "grad_norm": 0.142578125, | |
| "learning_rate": 7.692307692307694e-07, | |
| "loss": 0.7538, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.9261538461538461, | |
| "grad_norm": 0.140625, | |
| "learning_rate": 7.384615384615385e-07, | |
| "loss": 0.7849, | |
| "step": 301 | |
| }, | |
| { | |
| "epoch": 0.9292307692307692, | |
| "grad_norm": 0.1455078125, | |
| "learning_rate": 7.076923076923077e-07, | |
| "loss": 0.762, | |
| "step": 302 | |
| }, | |
| { | |
| "epoch": 0.9323076923076923, | |
| "grad_norm": 0.205078125, | |
| "learning_rate": 6.769230769230769e-07, | |
| "loss": 0.7198, | |
| "step": 303 | |
| }, | |
| { | |
| "epoch": 0.9353846153846154, | |
| "grad_norm": 0.171875, | |
| "learning_rate": 6.461538461538462e-07, | |
| "loss": 0.7777, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 0.9384615384615385, | |
| "grad_norm": 0.1943359375, | |
| "learning_rate": 6.153846153846155e-07, | |
| "loss": 0.7475, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 0.9415384615384615, | |
| "grad_norm": 0.166015625, | |
| "learning_rate": 5.846153846153847e-07, | |
| "loss": 0.829, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 0.9446153846153846, | |
| "grad_norm": 0.15234375, | |
| "learning_rate": 5.53846153846154e-07, | |
| "loss": 0.7912, | |
| "step": 307 | |
| }, | |
| { | |
| "epoch": 0.9476923076923077, | |
| "grad_norm": 0.1748046875, | |
| "learning_rate": 5.230769230769231e-07, | |
| "loss": 0.8133, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 0.9507692307692308, | |
| "grad_norm": 0.1513671875, | |
| "learning_rate": 4.923076923076923e-07, | |
| "loss": 0.7625, | |
| "step": 309 | |
| }, | |
| { | |
| "epoch": 0.9538461538461539, | |
| "grad_norm": 0.3828125, | |
| "learning_rate": 4.615384615384616e-07, | |
| "loss": 0.7512, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.9569230769230769, | |
| "grad_norm": 0.1484375, | |
| "learning_rate": 4.307692307692308e-07, | |
| "loss": 0.7716, | |
| "step": 311 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "grad_norm": 0.14453125, | |
| "learning_rate": 4.0000000000000003e-07, | |
| "loss": 0.7843, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 0.963076923076923, | |
| "grad_norm": 0.15234375, | |
| "learning_rate": 3.6923076923076927e-07, | |
| "loss": 0.7108, | |
| "step": 313 | |
| }, | |
| { | |
| "epoch": 0.9661538461538461, | |
| "grad_norm": 0.1455078125, | |
| "learning_rate": 3.3846153846153845e-07, | |
| "loss": 0.7817, | |
| "step": 314 | |
| }, | |
| { | |
| "epoch": 0.9692307692307692, | |
| "grad_norm": 0.166015625, | |
| "learning_rate": 3.0769230769230774e-07, | |
| "loss": 0.7913, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 0.9723076923076923, | |
| "grad_norm": 0.1484375, | |
| "learning_rate": 2.76923076923077e-07, | |
| "loss": 0.7867, | |
| "step": 316 | |
| }, | |
| { | |
| "epoch": 0.9753846153846154, | |
| "grad_norm": 0.16015625, | |
| "learning_rate": 2.4615384615384616e-07, | |
| "loss": 0.8405, | |
| "step": 317 | |
| }, | |
| { | |
| "epoch": 0.9784615384615385, | |
| "grad_norm": 0.25390625, | |
| "learning_rate": 2.153846153846154e-07, | |
| "loss": 0.7689, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 0.9815384615384616, | |
| "grad_norm": 0.1669921875, | |
| "learning_rate": 1.8461538461538464e-07, | |
| "loss": 0.7648, | |
| "step": 319 | |
| }, | |
| { | |
| "epoch": 0.9846153846153847, | |
| "grad_norm": 0.1650390625, | |
| "learning_rate": 1.5384615384615387e-07, | |
| "loss": 0.7832, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.9876923076923076, | |
| "grad_norm": 0.14453125, | |
| "learning_rate": 1.2307692307692308e-07, | |
| "loss": 0.7985, | |
| "step": 321 | |
| }, | |
| { | |
| "epoch": 0.9907692307692307, | |
| "grad_norm": 0.1396484375, | |
| "learning_rate": 9.230769230769232e-08, | |
| "loss": 0.7759, | |
| "step": 322 | |
| }, | |
| { | |
| "epoch": 0.9938461538461538, | |
| "grad_norm": 0.150390625, | |
| "learning_rate": 6.153846153846154e-08, | |
| "loss": 0.775, | |
| "step": 323 | |
| }, | |
| { | |
| "epoch": 0.9969230769230769, | |
| "grad_norm": 0.150390625, | |
| "learning_rate": 3.076923076923077e-08, | |
| "loss": 0.7612, | |
| "step": 324 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 0.1435546875, | |
| "learning_rate": 0.0, | |
| "loss": 0.7615, | |
| "step": 325 | |
| } | |
| ], | |
| "logging_steps": 1.0, | |
| "max_steps": 325, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 0, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 1.0306514624068977e+18, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |