| { | |
| "best_global_step": null, | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 1.0, | |
| "eval_steps": 170, | |
| "global_step": 170, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0058823529411764705, | |
| "grad_norm": 0.740064799785614, | |
| "learning_rate": 1e-05, | |
| "loss": 2.4395, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.011764705882352941, | |
| "grad_norm": 0.7219232320785522, | |
| "learning_rate": 9.941176470588236e-06, | |
| "loss": 2.3902, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.01764705882352941, | |
| "grad_norm": 0.77315753698349, | |
| "learning_rate": 9.882352941176472e-06, | |
| "loss": 2.4516, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.023529411764705882, | |
| "grad_norm": 0.7578166127204895, | |
| "learning_rate": 9.823529411764706e-06, | |
| "loss": 2.4148, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.029411764705882353, | |
| "grad_norm": 0.7205833196640015, | |
| "learning_rate": 9.764705882352942e-06, | |
| "loss": 2.3372, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.03529411764705882, | |
| "grad_norm": 0.7160035967826843, | |
| "learning_rate": 9.705882352941177e-06, | |
| "loss": 2.2849, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.041176470588235294, | |
| "grad_norm": 0.8269237875938416, | |
| "learning_rate": 9.647058823529412e-06, | |
| "loss": 2.3719, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.047058823529411764, | |
| "grad_norm": 0.7316713333129883, | |
| "learning_rate": 9.588235294117649e-06, | |
| "loss": 2.2175, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.052941176470588235, | |
| "grad_norm": 0.7852907776832581, | |
| "learning_rate": 9.529411764705882e-06, | |
| "loss": 2.2489, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.058823529411764705, | |
| "grad_norm": 0.7100040316581726, | |
| "learning_rate": 9.470588235294119e-06, | |
| "loss": 2.1828, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.06470588235294118, | |
| "grad_norm": 0.6905198693275452, | |
| "learning_rate": 9.411764705882354e-06, | |
| "loss": 2.1709, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.07058823529411765, | |
| "grad_norm": 0.6189457774162292, | |
| "learning_rate": 9.352941176470589e-06, | |
| "loss": 2.1152, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.07647058823529412, | |
| "grad_norm": 0.5859349370002747, | |
| "learning_rate": 9.294117647058824e-06, | |
| "loss": 2.0362, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.08235294117647059, | |
| "grad_norm": 0.6242568492889404, | |
| "learning_rate": 9.23529411764706e-06, | |
| "loss": 2.0808, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.08823529411764706, | |
| "grad_norm": 0.6139904856681824, | |
| "learning_rate": 9.176470588235294e-06, | |
| "loss": 2.017, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.09411764705882353, | |
| "grad_norm": 0.6155012249946594, | |
| "learning_rate": 9.11764705882353e-06, | |
| "loss": 2.0315, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 0.6213613152503967, | |
| "learning_rate": 9.058823529411765e-06, | |
| "loss": 1.9902, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.10588235294117647, | |
| "grad_norm": 0.584740936756134, | |
| "learning_rate": 9e-06, | |
| "loss": 1.9679, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.11176470588235295, | |
| "grad_norm": 0.5694301128387451, | |
| "learning_rate": 8.941176470588237e-06, | |
| "loss": 1.9416, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.11764705882352941, | |
| "grad_norm": 0.5494748950004578, | |
| "learning_rate": 8.88235294117647e-06, | |
| "loss": 1.9129, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.12352941176470589, | |
| "grad_norm": 0.5430072546005249, | |
| "learning_rate": 8.823529411764707e-06, | |
| "loss": 1.89, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.12941176470588237, | |
| "grad_norm": 0.5303496718406677, | |
| "learning_rate": 8.764705882352942e-06, | |
| "loss": 1.8751, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.13529411764705881, | |
| "grad_norm": 0.5339208841323853, | |
| "learning_rate": 8.705882352941177e-06, | |
| "loss": 1.8598, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.1411764705882353, | |
| "grad_norm": 0.5348221659660339, | |
| "learning_rate": 8.647058823529413e-06, | |
| "loss": 1.8426, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.14705882352941177, | |
| "grad_norm": 0.4850575923919678, | |
| "learning_rate": 8.588235294117647e-06, | |
| "loss": 1.8126, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.15294117647058825, | |
| "grad_norm": 0.5005661845207214, | |
| "learning_rate": 8.529411764705883e-06, | |
| "loss": 1.8054, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.1588235294117647, | |
| "grad_norm": 0.47416189312934875, | |
| "learning_rate": 8.470588235294118e-06, | |
| "loss": 1.7775, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.16470588235294117, | |
| "grad_norm": 0.49917134642601013, | |
| "learning_rate": 8.411764705882353e-06, | |
| "loss": 1.7834, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.17058823529411765, | |
| "grad_norm": 0.4690726101398468, | |
| "learning_rate": 8.35294117647059e-06, | |
| "loss": 1.769, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.17647058823529413, | |
| "grad_norm": 0.4899074137210846, | |
| "learning_rate": 8.294117647058825e-06, | |
| "loss": 1.7534, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.18235294117647058, | |
| "grad_norm": 0.4322926104068756, | |
| "learning_rate": 8.23529411764706e-06, | |
| "loss": 1.7127, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.18823529411764706, | |
| "grad_norm": 0.4963333010673523, | |
| "learning_rate": 8.176470588235295e-06, | |
| "loss": 1.7316, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.19411764705882353, | |
| "grad_norm": 0.4416678547859192, | |
| "learning_rate": 8.11764705882353e-06, | |
| "loss": 1.6911, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "grad_norm": 0.44732019305229187, | |
| "learning_rate": 8.058823529411766e-06, | |
| "loss": 1.6832, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.20588235294117646, | |
| "grad_norm": 0.4325319528579712, | |
| "learning_rate": 8.000000000000001e-06, | |
| "loss": 1.6849, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.21176470588235294, | |
| "grad_norm": 0.4243956506252289, | |
| "learning_rate": 7.941176470588236e-06, | |
| "loss": 1.6471, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.21764705882352942, | |
| "grad_norm": 0.41187071800231934, | |
| "learning_rate": 7.882352941176471e-06, | |
| "loss": 1.654, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.2235294117647059, | |
| "grad_norm": 0.40401241183280945, | |
| "learning_rate": 7.823529411764706e-06, | |
| "loss": 1.644, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.22941176470588234, | |
| "grad_norm": 0.4079605042934418, | |
| "learning_rate": 7.764705882352941e-06, | |
| "loss": 1.6209, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.23529411764705882, | |
| "grad_norm": 0.37295785546302795, | |
| "learning_rate": 7.705882352941178e-06, | |
| "loss": 1.6111, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.2411764705882353, | |
| "grad_norm": 0.37890729308128357, | |
| "learning_rate": 7.647058823529411e-06, | |
| "loss": 1.6122, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.24705882352941178, | |
| "grad_norm": 0.3897000849246979, | |
| "learning_rate": 7.588235294117648e-06, | |
| "loss": 1.594, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.2529411764705882, | |
| "grad_norm": 0.37150734663009644, | |
| "learning_rate": 7.529411764705883e-06, | |
| "loss": 1.5683, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.25882352941176473, | |
| "grad_norm": 0.3686462342739105, | |
| "learning_rate": 7.4705882352941185e-06, | |
| "loss": 1.5578, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.2647058823529412, | |
| "grad_norm": 0.3615223467350006, | |
| "learning_rate": 7.4117647058823535e-06, | |
| "loss": 1.5553, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.27058823529411763, | |
| "grad_norm": 0.341239333152771, | |
| "learning_rate": 7.352941176470589e-06, | |
| "loss": 1.5504, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.27647058823529413, | |
| "grad_norm": 0.32972443103790283, | |
| "learning_rate": 7.294117647058823e-06, | |
| "loss": 1.5523, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.2823529411764706, | |
| "grad_norm": 0.3313795328140259, | |
| "learning_rate": 7.235294117647059e-06, | |
| "loss": 1.5367, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.28823529411764703, | |
| "grad_norm": 0.3319094479084015, | |
| "learning_rate": 7.176470588235295e-06, | |
| "loss": 1.5233, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.29411764705882354, | |
| "grad_norm": 0.3231871426105499, | |
| "learning_rate": 7.11764705882353e-06, | |
| "loss": 1.5064, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "grad_norm": 0.3074081838130951, | |
| "learning_rate": 7.058823529411766e-06, | |
| "loss": 1.4804, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.3058823529411765, | |
| "grad_norm": 0.329453706741333, | |
| "learning_rate": 7e-06, | |
| "loss": 1.5033, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.31176470588235294, | |
| "grad_norm": 0.3119613826274872, | |
| "learning_rate": 6.941176470588236e-06, | |
| "loss": 1.4898, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.3176470588235294, | |
| "grad_norm": 0.31654036045074463, | |
| "learning_rate": 6.8823529411764715e-06, | |
| "loss": 1.4599, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.3235294117647059, | |
| "grad_norm": 0.29753053188323975, | |
| "learning_rate": 6.8235294117647065e-06, | |
| "loss": 1.4625, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.32941176470588235, | |
| "grad_norm": 0.30820533633232117, | |
| "learning_rate": 6.764705882352942e-06, | |
| "loss": 1.4759, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.3352941176470588, | |
| "grad_norm": 0.29135259985923767, | |
| "learning_rate": 6.705882352941176e-06, | |
| "loss": 1.4699, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.3411764705882353, | |
| "grad_norm": 0.2927163243293762, | |
| "learning_rate": 6.647058823529412e-06, | |
| "loss": 1.4428, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.34705882352941175, | |
| "grad_norm": 0.3006676137447357, | |
| "learning_rate": 6.588235294117647e-06, | |
| "loss": 1.451, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.35294117647058826, | |
| "grad_norm": 0.29078030586242676, | |
| "learning_rate": 6.529411764705883e-06, | |
| "loss": 1.4352, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.3588235294117647, | |
| "grad_norm": 0.28280261158943176, | |
| "learning_rate": 6.470588235294119e-06, | |
| "loss": 1.4295, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.36470588235294116, | |
| "grad_norm": 0.3001053035259247, | |
| "learning_rate": 6.411764705882354e-06, | |
| "loss": 1.4375, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.37058823529411766, | |
| "grad_norm": 0.28294065594673157, | |
| "learning_rate": 6.352941176470589e-06, | |
| "loss": 1.4144, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.3764705882352941, | |
| "grad_norm": 0.2832286059856415, | |
| "learning_rate": 6.294117647058824e-06, | |
| "loss": 1.4207, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.38235294117647056, | |
| "grad_norm": 0.2754327952861786, | |
| "learning_rate": 6.2352941176470595e-06, | |
| "loss": 1.4362, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.38823529411764707, | |
| "grad_norm": 0.28400981426239014, | |
| "learning_rate": 6.176470588235295e-06, | |
| "loss": 1.382, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.3941176470588235, | |
| "grad_norm": 0.2783932387828827, | |
| "learning_rate": 6.11764705882353e-06, | |
| "loss": 1.4018, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "grad_norm": 0.270181268453598, | |
| "learning_rate": 6.058823529411765e-06, | |
| "loss": 1.4002, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.40588235294117647, | |
| "grad_norm": 0.28010931611061096, | |
| "learning_rate": 6e-06, | |
| "loss": 1.3927, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.4117647058823529, | |
| "grad_norm": 0.28210070729255676, | |
| "learning_rate": 5.941176470588236e-06, | |
| "loss": 1.3775, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.4176470588235294, | |
| "grad_norm": 0.26174265146255493, | |
| "learning_rate": 5.882352941176471e-06, | |
| "loss": 1.3791, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.4235294117647059, | |
| "grad_norm": 0.2730426788330078, | |
| "learning_rate": 5.823529411764707e-06, | |
| "loss": 1.3865, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.4294117647058823, | |
| "grad_norm": 0.25816625356674194, | |
| "learning_rate": 5.764705882352941e-06, | |
| "loss": 1.357, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.43529411764705883, | |
| "grad_norm": 0.25862398743629456, | |
| "learning_rate": 5.705882352941177e-06, | |
| "loss": 1.3597, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.4411764705882353, | |
| "grad_norm": 0.2514458894729614, | |
| "learning_rate": 5.6470588235294125e-06, | |
| "loss": 1.3971, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.4470588235294118, | |
| "grad_norm": 0.2639279067516327, | |
| "learning_rate": 5.588235294117647e-06, | |
| "loss": 1.3693, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.45294117647058824, | |
| "grad_norm": 0.26090630888938904, | |
| "learning_rate": 5.529411764705883e-06, | |
| "loss": 1.3681, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.4588235294117647, | |
| "grad_norm": 0.2618473470211029, | |
| "learning_rate": 5.470588235294119e-06, | |
| "loss": 1.3568, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.4647058823529412, | |
| "grad_norm": 0.25189754366874695, | |
| "learning_rate": 5.411764705882353e-06, | |
| "loss": 1.3628, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.47058823529411764, | |
| "grad_norm": 0.2481844574213028, | |
| "learning_rate": 5.352941176470589e-06, | |
| "loss": 1.3382, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.4764705882352941, | |
| "grad_norm": 0.24728593230247498, | |
| "learning_rate": 5.294117647058824e-06, | |
| "loss": 1.3288, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.4823529411764706, | |
| "grad_norm": 0.25381624698638916, | |
| "learning_rate": 5.23529411764706e-06, | |
| "loss": 1.3215, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.48823529411764705, | |
| "grad_norm": 0.2516557276248932, | |
| "learning_rate": 5.176470588235295e-06, | |
| "loss": 1.3264, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.49411764705882355, | |
| "grad_norm": 0.24683943390846252, | |
| "learning_rate": 5.11764705882353e-06, | |
| "loss": 1.3244, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "grad_norm": 0.24650059640407562, | |
| "learning_rate": 5.058823529411765e-06, | |
| "loss": 1.3259, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.5058823529411764, | |
| "grad_norm": 0.2529411017894745, | |
| "learning_rate": 5e-06, | |
| "loss": 1.3313, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.5117647058823529, | |
| "grad_norm": 0.2540332078933716, | |
| "learning_rate": 4.941176470588236e-06, | |
| "loss": 1.33, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.5176470588235295, | |
| "grad_norm": 0.25214681029319763, | |
| "learning_rate": 4.882352941176471e-06, | |
| "loss": 1.2992, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.5235294117647059, | |
| "grad_norm": 0.27215129137039185, | |
| "learning_rate": 4.823529411764706e-06, | |
| "loss": 1.3119, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.5294117647058824, | |
| "grad_norm": 0.2611463665962219, | |
| "learning_rate": 4.764705882352941e-06, | |
| "loss": 1.3265, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.5352941176470588, | |
| "grad_norm": 0.2502508759498596, | |
| "learning_rate": 4.705882352941177e-06, | |
| "loss": 1.2926, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.5411764705882353, | |
| "grad_norm": 0.26345929503440857, | |
| "learning_rate": 4.647058823529412e-06, | |
| "loss": 1.2975, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.5470588235294118, | |
| "grad_norm": 0.2609890401363373, | |
| "learning_rate": 4.588235294117647e-06, | |
| "loss": 1.2921, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.5529411764705883, | |
| "grad_norm": 0.2622078061103821, | |
| "learning_rate": 4.529411764705883e-06, | |
| "loss": 1.3016, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.5588235294117647, | |
| "grad_norm": 0.2562355101108551, | |
| "learning_rate": 4.4705882352941184e-06, | |
| "loss": 1.2908, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.5647058823529412, | |
| "grad_norm": 0.25484997034072876, | |
| "learning_rate": 4.411764705882353e-06, | |
| "loss": 1.3199, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.5705882352941176, | |
| "grad_norm": 0.25862494111061096, | |
| "learning_rate": 4.352941176470588e-06, | |
| "loss": 1.2855, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.5764705882352941, | |
| "grad_norm": 0.27047714591026306, | |
| "learning_rate": 4.294117647058823e-06, | |
| "loss": 1.3165, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.5823529411764706, | |
| "grad_norm": 0.2632170021533966, | |
| "learning_rate": 4.235294117647059e-06, | |
| "loss": 1.2912, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.5882352941176471, | |
| "grad_norm": 0.25326305627822876, | |
| "learning_rate": 4.176470588235295e-06, | |
| "loss": 1.3053, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.5941176470588235, | |
| "grad_norm": 0.26147395372390747, | |
| "learning_rate": 4.11764705882353e-06, | |
| "loss": 1.2973, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "grad_norm": 0.26799634099006653, | |
| "learning_rate": 4.058823529411765e-06, | |
| "loss": 1.2794, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.6058823529411764, | |
| "grad_norm": 0.2632071077823639, | |
| "learning_rate": 4.000000000000001e-06, | |
| "loss": 1.2867, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.611764705882353, | |
| "grad_norm": 0.27080872654914856, | |
| "learning_rate": 3.941176470588236e-06, | |
| "loss": 1.277, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.6176470588235294, | |
| "grad_norm": 0.2697356939315796, | |
| "learning_rate": 3.882352941176471e-06, | |
| "loss": 1.2697, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.6235294117647059, | |
| "grad_norm": 0.27979159355163574, | |
| "learning_rate": 3.8235294117647055e-06, | |
| "loss": 1.2746, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.6294117647058823, | |
| "grad_norm": 0.2690213620662689, | |
| "learning_rate": 3.7647058823529414e-06, | |
| "loss": 1.2734, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.6352941176470588, | |
| "grad_norm": 0.27870768308639526, | |
| "learning_rate": 3.7058823529411767e-06, | |
| "loss": 1.2707, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.6411764705882353, | |
| "grad_norm": 0.29579660296440125, | |
| "learning_rate": 3.6470588235294117e-06, | |
| "loss": 1.2616, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.6470588235294118, | |
| "grad_norm": 0.2851077914237976, | |
| "learning_rate": 3.5882352941176475e-06, | |
| "loss": 1.2591, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.6529411764705882, | |
| "grad_norm": 0.307041198015213, | |
| "learning_rate": 3.529411764705883e-06, | |
| "loss": 1.2522, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.6588235294117647, | |
| "grad_norm": 0.29607197642326355, | |
| "learning_rate": 3.470588235294118e-06, | |
| "loss": 1.2831, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.6647058823529411, | |
| "grad_norm": 0.29029569029808044, | |
| "learning_rate": 3.4117647058823532e-06, | |
| "loss": 1.2539, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.6705882352941176, | |
| "grad_norm": 0.28268927335739136, | |
| "learning_rate": 3.352941176470588e-06, | |
| "loss": 1.2652, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.6764705882352942, | |
| "grad_norm": 0.28747496008872986, | |
| "learning_rate": 3.2941176470588236e-06, | |
| "loss": 1.2394, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.6823529411764706, | |
| "grad_norm": 0.2939983904361725, | |
| "learning_rate": 3.2352941176470594e-06, | |
| "loss": 1.2639, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.6882352941176471, | |
| "grad_norm": 0.2975703179836273, | |
| "learning_rate": 3.1764705882352943e-06, | |
| "loss": 1.2762, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.6941176470588235, | |
| "grad_norm": 0.2900603413581848, | |
| "learning_rate": 3.1176470588235297e-06, | |
| "loss": 1.2623, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "grad_norm": 0.2925064265727997, | |
| "learning_rate": 3.058823529411765e-06, | |
| "loss": 1.27, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.7058823529411765, | |
| "grad_norm": 0.2913402318954468, | |
| "learning_rate": 3e-06, | |
| "loss": 1.2558, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.711764705882353, | |
| "grad_norm": 0.3211301863193512, | |
| "learning_rate": 2.9411764705882355e-06, | |
| "loss": 1.2397, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.7176470588235294, | |
| "grad_norm": 0.3004200756549835, | |
| "learning_rate": 2.8823529411764704e-06, | |
| "loss": 1.2627, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.7235294117647059, | |
| "grad_norm": 0.3165768086910248, | |
| "learning_rate": 2.8235294117647062e-06, | |
| "loss": 1.2388, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.7294117647058823, | |
| "grad_norm": 0.29654860496520996, | |
| "learning_rate": 2.7647058823529416e-06, | |
| "loss": 1.2332, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.7352941176470589, | |
| "grad_norm": 0.3117150068283081, | |
| "learning_rate": 2.7058823529411766e-06, | |
| "loss": 1.2588, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.7411764705882353, | |
| "grad_norm": 0.33643701672554016, | |
| "learning_rate": 2.647058823529412e-06, | |
| "loss": 1.2289, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.7470588235294118, | |
| "grad_norm": 0.3130914270877838, | |
| "learning_rate": 2.5882352941176473e-06, | |
| "loss": 1.263, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.7529411764705882, | |
| "grad_norm": 0.3396664559841156, | |
| "learning_rate": 2.5294117647058823e-06, | |
| "loss": 1.2592, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.7588235294117647, | |
| "grad_norm": 0.30291828513145447, | |
| "learning_rate": 2.470588235294118e-06, | |
| "loss": 1.2577, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.7647058823529411, | |
| "grad_norm": 0.32175707817077637, | |
| "learning_rate": 2.411764705882353e-06, | |
| "loss": 1.247, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.7705882352941177, | |
| "grad_norm": 0.346138596534729, | |
| "learning_rate": 2.3529411764705885e-06, | |
| "loss": 1.2476, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.7764705882352941, | |
| "grad_norm": 0.3127652406692505, | |
| "learning_rate": 2.2941176470588234e-06, | |
| "loss": 1.2392, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.7823529411764706, | |
| "grad_norm": 0.349590927362442, | |
| "learning_rate": 2.2352941176470592e-06, | |
| "loss": 1.2377, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.788235294117647, | |
| "grad_norm": 0.3107239305973053, | |
| "learning_rate": 2.176470588235294e-06, | |
| "loss": 1.239, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.7941176470588235, | |
| "grad_norm": 0.33791080117225647, | |
| "learning_rate": 2.1176470588235296e-06, | |
| "loss": 1.246, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 0.3314568102359772, | |
| "learning_rate": 2.058823529411765e-06, | |
| "loss": 1.2402, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.8058823529411765, | |
| "grad_norm": 0.3275523781776428, | |
| "learning_rate": 2.0000000000000003e-06, | |
| "loss": 1.2348, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.8117647058823529, | |
| "grad_norm": 0.33062854409217834, | |
| "learning_rate": 1.9411764705882353e-06, | |
| "loss": 1.2427, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.8176470588235294, | |
| "grad_norm": 0.35148942470550537, | |
| "learning_rate": 1.8823529411764707e-06, | |
| "loss": 1.2261, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.8235294117647058, | |
| "grad_norm": 0.3389197289943695, | |
| "learning_rate": 1.8235294117647058e-06, | |
| "loss": 1.2362, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.8294117647058824, | |
| "grad_norm": 0.3360951244831085, | |
| "learning_rate": 1.7647058823529414e-06, | |
| "loss": 1.2302, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.8352941176470589, | |
| "grad_norm": 0.34131404757499695, | |
| "learning_rate": 1.7058823529411766e-06, | |
| "loss": 1.2266, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.8411764705882353, | |
| "grad_norm": 0.328914076089859, | |
| "learning_rate": 1.6470588235294118e-06, | |
| "loss": 1.2308, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.8470588235294118, | |
| "grad_norm": 0.34804269671440125, | |
| "learning_rate": 1.5882352941176472e-06, | |
| "loss": 1.2212, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.8529411764705882, | |
| "grad_norm": 0.35386762022972107, | |
| "learning_rate": 1.5294117647058826e-06, | |
| "loss": 1.229, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.8588235294117647, | |
| "grad_norm": 0.33942756056785583, | |
| "learning_rate": 1.4705882352941177e-06, | |
| "loss": 1.2434, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.8647058823529412, | |
| "grad_norm": 0.32963618636131287, | |
| "learning_rate": 1.4117647058823531e-06, | |
| "loss": 1.2385, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.8705882352941177, | |
| "grad_norm": 0.3417942225933075, | |
| "learning_rate": 1.3529411764705883e-06, | |
| "loss": 1.2242, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.8764705882352941, | |
| "grad_norm": 0.33753451704978943, | |
| "learning_rate": 1.2941176470588237e-06, | |
| "loss": 1.2234, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.8823529411764706, | |
| "grad_norm": 0.3514120876789093, | |
| "learning_rate": 1.235294117647059e-06, | |
| "loss": 1.2241, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.888235294117647, | |
| "grad_norm": 0.35951969027519226, | |
| "learning_rate": 1.1764705882352942e-06, | |
| "loss": 1.2347, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 0.8941176470588236, | |
| "grad_norm": 0.3717687726020813, | |
| "learning_rate": 1.1176470588235296e-06, | |
| "loss": 1.2224, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "grad_norm": 0.3542497754096985, | |
| "learning_rate": 1.0588235294117648e-06, | |
| "loss": 1.2236, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 0.9058823529411765, | |
| "grad_norm": 0.3436025083065033, | |
| "learning_rate": 1.0000000000000002e-06, | |
| "loss": 1.2294, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 0.9117647058823529, | |
| "grad_norm": 0.37331488728523254, | |
| "learning_rate": 9.411764705882353e-07, | |
| "loss": 1.2046, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.9176470588235294, | |
| "grad_norm": 0.34907183051109314, | |
| "learning_rate": 8.823529411764707e-07, | |
| "loss": 1.2213, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 0.9235294117647059, | |
| "grad_norm": 0.36500322818756104, | |
| "learning_rate": 8.235294117647059e-07, | |
| "loss": 1.2091, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 0.9294117647058824, | |
| "grad_norm": 0.38440433144569397, | |
| "learning_rate": 7.647058823529413e-07, | |
| "loss": 1.2249, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 0.9352941176470588, | |
| "grad_norm": 0.3387817144393921, | |
| "learning_rate": 7.058823529411766e-07, | |
| "loss": 1.2154, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 0.9411764705882353, | |
| "grad_norm": 0.34928762912750244, | |
| "learning_rate": 6.470588235294118e-07, | |
| "loss": 1.2227, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.9470588235294117, | |
| "grad_norm": 0.36257097125053406, | |
| "learning_rate": 5.882352941176471e-07, | |
| "loss": 1.2211, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 0.9529411764705882, | |
| "grad_norm": 0.3723115026950836, | |
| "learning_rate": 5.294117647058824e-07, | |
| "loss": 1.2283, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 0.9588235294117647, | |
| "grad_norm": 0.3423607349395752, | |
| "learning_rate": 4.7058823529411767e-07, | |
| "loss": 1.2295, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 0.9647058823529412, | |
| "grad_norm": 0.3787173628807068, | |
| "learning_rate": 4.1176470588235295e-07, | |
| "loss": 1.2201, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 0.9705882352941176, | |
| "grad_norm": 0.36642688512802124, | |
| "learning_rate": 3.529411764705883e-07, | |
| "loss": 1.2313, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.9764705882352941, | |
| "grad_norm": 0.3594622313976288, | |
| "learning_rate": 2.9411764705882356e-07, | |
| "loss": 1.2128, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 0.9823529411764705, | |
| "grad_norm": 0.3701726496219635, | |
| "learning_rate": 2.3529411764705883e-07, | |
| "loss": 1.2324, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 0.9882352941176471, | |
| "grad_norm": 0.34158623218536377, | |
| "learning_rate": 1.7647058823529414e-07, | |
| "loss": 1.2364, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 0.9941176470588236, | |
| "grad_norm": 0.3631001114845276, | |
| "learning_rate": 1.1764705882352942e-07, | |
| "loss": 1.2191, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 0.40616530179977417, | |
| "learning_rate": 5.882352941176471e-08, | |
| "loss": 1.2029, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_loss": 1.2080979347229004, | |
| "eval_runtime": 4.1103, | |
| "eval_samples_per_second": 4.379, | |
| "eval_steps_per_second": 0.73, | |
| "step": 170 | |
| } | |
| ], | |
| "logging_steps": 1.0, | |
| "max_steps": 170, | |
| "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": 2.7061227320088986e+17, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |