| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 1.0, | |
| "eval_steps": 0, | |
| "global_step": 93, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.010752688172043012, | |
| "grad_norm": 0.443359375, | |
| "learning_rate": 9.89247311827957e-06, | |
| "loss": 1.6759, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.021505376344086023, | |
| "grad_norm": 0.44921875, | |
| "learning_rate": 9.78494623655914e-06, | |
| "loss": 1.7112, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.03225806451612903, | |
| "grad_norm": 0.4453125, | |
| "learning_rate": 9.67741935483871e-06, | |
| "loss": 1.6568, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.043010752688172046, | |
| "grad_norm": 0.4296875, | |
| "learning_rate": 9.56989247311828e-06, | |
| "loss": 1.6349, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.053763440860215055, | |
| "grad_norm": 0.400390625, | |
| "learning_rate": 9.46236559139785e-06, | |
| "loss": 1.6579, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.06451612903225806, | |
| "grad_norm": 0.390625, | |
| "learning_rate": 9.35483870967742e-06, | |
| "loss": 1.6134, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.07526881720430108, | |
| "grad_norm": 0.37890625, | |
| "learning_rate": 9.24731182795699e-06, | |
| "loss": 1.6222, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.08602150537634409, | |
| "grad_norm": 0.353515625, | |
| "learning_rate": 9.13978494623656e-06, | |
| "loss": 1.6115, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.0967741935483871, | |
| "grad_norm": 0.33984375, | |
| "learning_rate": 9.03225806451613e-06, | |
| "loss": 1.6084, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.10752688172043011, | |
| "grad_norm": 0.310546875, | |
| "learning_rate": 8.9247311827957e-06, | |
| "loss": 1.5482, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.11827956989247312, | |
| "grad_norm": 0.28515625, | |
| "learning_rate": 8.81720430107527e-06, | |
| "loss": 1.5641, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.12903225806451613, | |
| "grad_norm": 0.279296875, | |
| "learning_rate": 8.70967741935484e-06, | |
| "loss": 1.5567, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.13978494623655913, | |
| "grad_norm": 0.26171875, | |
| "learning_rate": 8.602150537634409e-06, | |
| "loss": 1.5356, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.15053763440860216, | |
| "grad_norm": 0.255859375, | |
| "learning_rate": 8.494623655913979e-06, | |
| "loss": 1.5166, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.16129032258064516, | |
| "grad_norm": 0.2734375, | |
| "learning_rate": 8.387096774193549e-06, | |
| "loss": 1.5688, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.17204301075268819, | |
| "grad_norm": 0.271484375, | |
| "learning_rate": 8.279569892473119e-06, | |
| "loss": 1.5558, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.1827956989247312, | |
| "grad_norm": 0.255859375, | |
| "learning_rate": 8.172043010752689e-06, | |
| "loss": 1.4873, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.1935483870967742, | |
| "grad_norm": 0.2578125, | |
| "learning_rate": 8.064516129032258e-06, | |
| "loss": 1.5371, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.20430107526881722, | |
| "grad_norm": 0.455078125, | |
| "learning_rate": 7.956989247311828e-06, | |
| "loss": 1.461, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.21505376344086022, | |
| "grad_norm": 0.224609375, | |
| "learning_rate": 7.849462365591398e-06, | |
| "loss": 1.5093, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.22580645161290322, | |
| "grad_norm": 0.2158203125, | |
| "learning_rate": 7.741935483870968e-06, | |
| "loss": 1.5567, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.23655913978494625, | |
| "grad_norm": 0.208984375, | |
| "learning_rate": 7.634408602150538e-06, | |
| "loss": 1.5329, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.24731182795698925, | |
| "grad_norm": 0.19921875, | |
| "learning_rate": 7.526881720430108e-06, | |
| "loss": 1.4974, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.25806451612903225, | |
| "grad_norm": 0.201171875, | |
| "learning_rate": 7.4193548387096784e-06, | |
| "loss": 1.463, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.26881720430107525, | |
| "grad_norm": 0.2060546875, | |
| "learning_rate": 7.311827956989248e-06, | |
| "loss": 1.4733, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.27956989247311825, | |
| "grad_norm": 0.1923828125, | |
| "learning_rate": 7.204301075268818e-06, | |
| "loss": 1.4257, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.2903225806451613, | |
| "grad_norm": 0.1748046875, | |
| "learning_rate": 7.096774193548388e-06, | |
| "loss": 1.438, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.3010752688172043, | |
| "grad_norm": 0.2021484375, | |
| "learning_rate": 6.989247311827958e-06, | |
| "loss": 1.5012, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.3118279569892473, | |
| "grad_norm": 0.181640625, | |
| "learning_rate": 6.881720430107528e-06, | |
| "loss": 1.4081, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.3225806451612903, | |
| "grad_norm": 0.166015625, | |
| "learning_rate": 6.774193548387097e-06, | |
| "loss": 1.4379, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.3333333333333333, | |
| "grad_norm": 0.1875, | |
| "learning_rate": 6.666666666666667e-06, | |
| "loss": 1.4352, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.34408602150537637, | |
| "grad_norm": 0.197265625, | |
| "learning_rate": 6.5591397849462365e-06, | |
| "loss": 1.4166, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.3548387096774194, | |
| "grad_norm": 0.1787109375, | |
| "learning_rate": 6.451612903225806e-06, | |
| "loss": 1.42, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.3655913978494624, | |
| "grad_norm": 0.1767578125, | |
| "learning_rate": 6.344086021505377e-06, | |
| "loss": 1.4004, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.3763440860215054, | |
| "grad_norm": 0.1728515625, | |
| "learning_rate": 6.236559139784947e-06, | |
| "loss": 1.4403, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.3870967741935484, | |
| "grad_norm": 0.1865234375, | |
| "learning_rate": 6.129032258064517e-06, | |
| "loss": 1.377, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.3978494623655914, | |
| "grad_norm": 0.1640625, | |
| "learning_rate": 6.021505376344087e-06, | |
| "loss": 1.3673, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.40860215053763443, | |
| "grad_norm": 0.1923828125, | |
| "learning_rate": 5.9139784946236566e-06, | |
| "loss": 1.411, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.41935483870967744, | |
| "grad_norm": 0.1650390625, | |
| "learning_rate": 5.806451612903226e-06, | |
| "loss": 1.3724, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.43010752688172044, | |
| "grad_norm": 0.1640625, | |
| "learning_rate": 5.698924731182796e-06, | |
| "loss": 1.3524, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.44086021505376344, | |
| "grad_norm": 0.166015625, | |
| "learning_rate": 5.591397849462365e-06, | |
| "loss": 1.3806, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.45161290322580644, | |
| "grad_norm": 0.162109375, | |
| "learning_rate": 5.483870967741935e-06, | |
| "loss": 1.3775, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.46236559139784944, | |
| "grad_norm": 0.1796875, | |
| "learning_rate": 5.376344086021506e-06, | |
| "loss": 1.4078, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.4731182795698925, | |
| "grad_norm": 0.2451171875, | |
| "learning_rate": 5.268817204301076e-06, | |
| "loss": 1.5332, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.4838709677419355, | |
| "grad_norm": 0.1806640625, | |
| "learning_rate": 5.161290322580646e-06, | |
| "loss": 1.3685, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.4946236559139785, | |
| "grad_norm": 0.181640625, | |
| "learning_rate": 5.0537634408602155e-06, | |
| "loss": 1.3775, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.5053763440860215, | |
| "grad_norm": 0.15234375, | |
| "learning_rate": 4.946236559139785e-06, | |
| "loss": 1.3448, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.5161290322580645, | |
| "grad_norm": 0.1748046875, | |
| "learning_rate": 4.838709677419355e-06, | |
| "loss": 1.3899, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.5268817204301075, | |
| "grad_norm": 0.1708984375, | |
| "learning_rate": 4.731182795698925e-06, | |
| "loss": 1.3537, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.5376344086021505, | |
| "grad_norm": 0.15625, | |
| "learning_rate": 4.623655913978495e-06, | |
| "loss": 1.3664, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.5483870967741935, | |
| "grad_norm": 0.1689453125, | |
| "learning_rate": 4.516129032258065e-06, | |
| "loss": 1.3599, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.5591397849462365, | |
| "grad_norm": 0.15234375, | |
| "learning_rate": 4.408602150537635e-06, | |
| "loss": 1.3479, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.5698924731182796, | |
| "grad_norm": 0.14453125, | |
| "learning_rate": 4.3010752688172045e-06, | |
| "loss": 1.2423, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.5806451612903226, | |
| "grad_norm": 0.1494140625, | |
| "learning_rate": 4.193548387096774e-06, | |
| "loss": 1.3378, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.5913978494623656, | |
| "grad_norm": 0.1552734375, | |
| "learning_rate": 4.086021505376344e-06, | |
| "loss": 1.3556, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.6021505376344086, | |
| "grad_norm": 0.150390625, | |
| "learning_rate": 3.978494623655914e-06, | |
| "loss": 1.3107, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.6129032258064516, | |
| "grad_norm": 0.2412109375, | |
| "learning_rate": 3.870967741935484e-06, | |
| "loss": 1.3446, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.6236559139784946, | |
| "grad_norm": 0.1728515625, | |
| "learning_rate": 3.763440860215054e-06, | |
| "loss": 1.3565, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.6344086021505376, | |
| "grad_norm": 0.1552734375, | |
| "learning_rate": 3.655913978494624e-06, | |
| "loss": 1.3629, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.6451612903225806, | |
| "grad_norm": 0.1484375, | |
| "learning_rate": 3.548387096774194e-06, | |
| "loss": 1.3317, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.6559139784946236, | |
| "grad_norm": 0.1474609375, | |
| "learning_rate": 3.440860215053764e-06, | |
| "loss": 1.3082, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.6666666666666666, | |
| "grad_norm": 0.1435546875, | |
| "learning_rate": 3.3333333333333333e-06, | |
| "loss": 1.3066, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.6774193548387096, | |
| "grad_norm": 0.1572265625, | |
| "learning_rate": 3.225806451612903e-06, | |
| "loss": 1.3483, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.6881720430107527, | |
| "grad_norm": 0.15625, | |
| "learning_rate": 3.1182795698924735e-06, | |
| "loss": 1.3344, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.6989247311827957, | |
| "grad_norm": 0.1533203125, | |
| "learning_rate": 3.0107526881720433e-06, | |
| "loss": 1.3201, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.7096774193548387, | |
| "grad_norm": 0.18359375, | |
| "learning_rate": 2.903225806451613e-06, | |
| "loss": 1.3539, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.7204301075268817, | |
| "grad_norm": 0.1552734375, | |
| "learning_rate": 2.7956989247311827e-06, | |
| "loss": 1.3277, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.7311827956989247, | |
| "grad_norm": 0.154296875, | |
| "learning_rate": 2.688172043010753e-06, | |
| "loss": 1.2985, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.7419354838709677, | |
| "grad_norm": 0.1748046875, | |
| "learning_rate": 2.580645161290323e-06, | |
| "loss": 1.3343, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.7526881720430108, | |
| "grad_norm": 0.1552734375, | |
| "learning_rate": 2.4731182795698927e-06, | |
| "loss": 1.2658, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.7634408602150538, | |
| "grad_norm": 0.1533203125, | |
| "learning_rate": 2.3655913978494625e-06, | |
| "loss": 1.3027, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.7741935483870968, | |
| "grad_norm": 0.16796875, | |
| "learning_rate": 2.2580645161290324e-06, | |
| "loss": 1.3223, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.7849462365591398, | |
| "grad_norm": 0.146484375, | |
| "learning_rate": 2.1505376344086023e-06, | |
| "loss": 1.3234, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.7956989247311828, | |
| "grad_norm": 0.140625, | |
| "learning_rate": 2.043010752688172e-06, | |
| "loss": 1.24, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.8064516129032258, | |
| "grad_norm": 0.166015625, | |
| "learning_rate": 1.935483870967742e-06, | |
| "loss": 1.3627, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.8172043010752689, | |
| "grad_norm": 0.1455078125, | |
| "learning_rate": 1.827956989247312e-06, | |
| "loss": 1.2964, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.8279569892473119, | |
| "grad_norm": 0.15234375, | |
| "learning_rate": 1.720430107526882e-06, | |
| "loss": 1.3288, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.8387096774193549, | |
| "grad_norm": 0.1455078125, | |
| "learning_rate": 1.6129032258064516e-06, | |
| "loss": 1.2812, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.8494623655913979, | |
| "grad_norm": 0.1943359375, | |
| "learning_rate": 1.5053763440860217e-06, | |
| "loss": 1.3724, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.8602150537634409, | |
| "grad_norm": 0.150390625, | |
| "learning_rate": 1.3978494623655913e-06, | |
| "loss": 1.3243, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.8709677419354839, | |
| "grad_norm": 0.1435546875, | |
| "learning_rate": 1.2903225806451614e-06, | |
| "loss": 1.2953, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.8817204301075269, | |
| "grad_norm": 0.1484375, | |
| "learning_rate": 1.1827956989247313e-06, | |
| "loss": 1.3312, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.8924731182795699, | |
| "grad_norm": 0.19921875, | |
| "learning_rate": 1.0752688172043011e-06, | |
| "loss": 1.2928, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.9032258064516129, | |
| "grad_norm": 0.154296875, | |
| "learning_rate": 9.67741935483871e-07, | |
| "loss": 1.3089, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.9139784946236559, | |
| "grad_norm": 0.150390625, | |
| "learning_rate": 8.60215053763441e-07, | |
| "loss": 1.297, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.9247311827956989, | |
| "grad_norm": 0.1484375, | |
| "learning_rate": 7.526881720430108e-07, | |
| "loss": 1.3467, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.9354838709677419, | |
| "grad_norm": 0.15625, | |
| "learning_rate": 6.451612903225807e-07, | |
| "loss": 1.3116, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.946236559139785, | |
| "grad_norm": 0.146484375, | |
| "learning_rate": 5.376344086021506e-07, | |
| "loss": 1.3005, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.956989247311828, | |
| "grad_norm": 0.1728515625, | |
| "learning_rate": 4.301075268817205e-07, | |
| "loss": 1.35, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.967741935483871, | |
| "grad_norm": 0.1484375, | |
| "learning_rate": 3.2258064516129035e-07, | |
| "loss": 1.2488, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.978494623655914, | |
| "grad_norm": 0.1689453125, | |
| "learning_rate": 2.1505376344086024e-07, | |
| "loss": 1.3261, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.989247311827957, | |
| "grad_norm": 0.1435546875, | |
| "learning_rate": 1.0752688172043012e-07, | |
| "loss": 1.324, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 0.162109375, | |
| "learning_rate": 0.0, | |
| "loss": 1.3184, | |
| "step": 93 | |
| } | |
| ], | |
| "logging_steps": 1.0, | |
| "max_steps": 93, | |
| "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": 2.5783842722178662e+17, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |