| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 1.9534883720930232, | |
| "eval_steps": 11, | |
| "global_step": 86, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 22.073590049775476, | |
| "learning_rate": 2e-08, | |
| "loss": 1.5029, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "eval_loss": 1.5676738023757935, | |
| "eval_runtime": 1.7229, | |
| "eval_samples_per_second": 11.028, | |
| "eval_steps_per_second": 1.741, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 21.339319978585014, | |
| "learning_rate": 4e-08, | |
| "loss": 1.5411, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 21.89138128266215, | |
| "learning_rate": 6e-08, | |
| "loss": 1.535, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 35.68832946989201, | |
| "learning_rate": 8e-08, | |
| "loss": 1.4248, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "grad_norm": 18.922589465764027, | |
| "learning_rate": 1e-07, | |
| "loss": 1.512, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "grad_norm": 22.05229837285426, | |
| "learning_rate": 1.2e-07, | |
| "loss": 1.5849, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 29.050954317842017, | |
| "learning_rate": 1.4e-07, | |
| "loss": 1.4915, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "grad_norm": 16.415909544756897, | |
| "learning_rate": 1.6e-07, | |
| "loss": 1.4711, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "grad_norm": 24.72099068344741, | |
| "learning_rate": 1.8e-07, | |
| "loss": 1.5618, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "grad_norm": 15.495802280155232, | |
| "learning_rate": 2e-07, | |
| "loss": 1.5432, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "grad_norm": 15.385087340525107, | |
| "learning_rate": 2.1999999999999998e-07, | |
| "loss": 1.5892, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "eval_loss": 1.5673789978027344, | |
| "eval_runtime": 0.7198, | |
| "eval_samples_per_second": 26.396, | |
| "eval_steps_per_second": 4.168, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "grad_norm": 18.316263378494824, | |
| "learning_rate": 2.4e-07, | |
| "loss": 1.5068, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "grad_norm": 16.604978508847786, | |
| "learning_rate": 2.6e-07, | |
| "loss": 1.5079, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.33, | |
| "grad_norm": 19.088434906193037, | |
| "learning_rate": 2.8e-07, | |
| "loss": 1.3923, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "grad_norm": 13.367337607829006, | |
| "learning_rate": 3e-07, | |
| "loss": 1.4215, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.37, | |
| "grad_norm": 14.12443554856015, | |
| "learning_rate": 3.2e-07, | |
| "loss": 1.5182, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "grad_norm": 21.547074702210885, | |
| "learning_rate": 3.4000000000000003e-07, | |
| "loss": 1.6407, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.42, | |
| "grad_norm": 19.983618321532845, | |
| "learning_rate": 3.6e-07, | |
| "loss": 1.6124, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.44, | |
| "grad_norm": 14.12870699420598, | |
| "learning_rate": 3.7999999999999996e-07, | |
| "loss": 1.3535, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.47, | |
| "grad_norm": 17.73177207632663, | |
| "learning_rate": 4e-07, | |
| "loss": 1.4723, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.49, | |
| "grad_norm": 24.017411591807985, | |
| "learning_rate": 4.1999999999999995e-07, | |
| "loss": 1.3109, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.51, | |
| "grad_norm": 15.47289342392699, | |
| "learning_rate": 4.3999999999999997e-07, | |
| "loss": 1.2975, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.51, | |
| "eval_loss": 1.5646436214447021, | |
| "eval_runtime": 0.7176, | |
| "eval_samples_per_second": 26.478, | |
| "eval_steps_per_second": 4.181, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.53, | |
| "grad_norm": 21.160644074933437, | |
| "learning_rate": 4.6e-07, | |
| "loss": 1.5044, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.56, | |
| "grad_norm": 15.065995566576719, | |
| "learning_rate": 4.8e-07, | |
| "loss": 1.4756, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.58, | |
| "grad_norm": 17.05786606459713, | |
| "learning_rate": 5e-07, | |
| "loss": 1.4934, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "grad_norm": 15.821412963269017, | |
| "learning_rate": 5.2e-07, | |
| "loss": 1.6885, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.63, | |
| "grad_norm": 17.42199498977738, | |
| "learning_rate": 5.4e-07, | |
| "loss": 1.3696, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.65, | |
| "grad_norm": 17.118541529912182, | |
| "learning_rate": 5.6e-07, | |
| "loss": 1.4963, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.67, | |
| "grad_norm": 16.14285401779511, | |
| "learning_rate": 5.8e-07, | |
| "loss": 1.523, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "grad_norm": 14.413715273595537, | |
| "learning_rate": 6e-07, | |
| "loss": 1.4205, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "grad_norm": 15.696143212367938, | |
| "learning_rate": 6.2e-07, | |
| "loss": 1.417, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.74, | |
| "grad_norm": 18.422902888255102, | |
| "learning_rate": 6.4e-07, | |
| "loss": 1.513, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.77, | |
| "grad_norm": 15.348458059196757, | |
| "learning_rate": 6.6e-07, | |
| "loss": 1.6405, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.77, | |
| "eval_loss": 1.558524489402771, | |
| "eval_runtime": 0.7067, | |
| "eval_samples_per_second": 26.885, | |
| "eval_steps_per_second": 4.245, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.79, | |
| "grad_norm": 31.510016892264296, | |
| "learning_rate": 6.800000000000001e-07, | |
| "loss": 1.395, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.81, | |
| "grad_norm": 15.63296379080965, | |
| "learning_rate": 7e-07, | |
| "loss": 1.5409, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.84, | |
| "grad_norm": 15.627218592491115, | |
| "learning_rate": 7.2e-07, | |
| "loss": 1.3137, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.86, | |
| "grad_norm": 15.337191607694441, | |
| "learning_rate": 7.4e-07, | |
| "loss": 1.7298, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "grad_norm": 16.11020188378557, | |
| "learning_rate": 7.599999999999999e-07, | |
| "loss": 1.403, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.91, | |
| "grad_norm": 19.261102026351743, | |
| "learning_rate": 7.799999999999999e-07, | |
| "loss": 1.5199, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.93, | |
| "grad_norm": 15.249059429242267, | |
| "learning_rate": 8e-07, | |
| "loss": 1.5028, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.95, | |
| "grad_norm": 17.026025037114596, | |
| "learning_rate": 8.199999999999999e-07, | |
| "loss": 1.5746, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.98, | |
| "grad_norm": 20.241389951964763, | |
| "learning_rate": 8.399999999999999e-07, | |
| "loss": 1.5432, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 13.357942060394901, | |
| "learning_rate": 8.599999999999999e-07, | |
| "loss": 1.3017, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 1.02, | |
| "grad_norm": 19.122899956188, | |
| "learning_rate": 8.799999999999999e-07, | |
| "loss": 1.4797, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 1.02, | |
| "eval_loss": 1.5534765720367432, | |
| "eval_runtime": 0.7056, | |
| "eval_samples_per_second": 26.926, | |
| "eval_steps_per_second": 4.251, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 1.05, | |
| "grad_norm": 12.993274416162807, | |
| "learning_rate": 9e-07, | |
| "loss": 1.3812, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 1.02, | |
| "grad_norm": 15.57715477404305, | |
| "learning_rate": 9.2e-07, | |
| "loss": 1.5169, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 1.05, | |
| "grad_norm": 15.047562025949741, | |
| "learning_rate": 9.399999999999999e-07, | |
| "loss": 1.597, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 1.07, | |
| "grad_norm": 13.154118671073913, | |
| "learning_rate": 9.6e-07, | |
| "loss": 1.4144, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 1.09, | |
| "grad_norm": 17.445553030977283, | |
| "learning_rate": 9.8e-07, | |
| "loss": 1.5843, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 1.12, | |
| "grad_norm": 11.552943553949452, | |
| "learning_rate": 1e-06, | |
| "loss": 1.3343, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 1.14, | |
| "grad_norm": 14.151856403767372, | |
| "learning_rate": 1.02e-06, | |
| "loss": 1.428, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 1.16, | |
| "grad_norm": 12.579639490538433, | |
| "learning_rate": 1.04e-06, | |
| "loss": 1.4277, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 1.19, | |
| "grad_norm": 13.03970026338736, | |
| "learning_rate": 1.06e-06, | |
| "loss": 1.3097, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 1.21, | |
| "grad_norm": 18.55159266953034, | |
| "learning_rate": 1.08e-06, | |
| "loss": 1.4071, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 1.23, | |
| "grad_norm": 17.185477508845405, | |
| "learning_rate": 1.1e-06, | |
| "loss": 1.4285, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 1.23, | |
| "eval_loss": 1.5510258674621582, | |
| "eval_runtime": 0.7063, | |
| "eval_samples_per_second": 26.902, | |
| "eval_steps_per_second": 4.248, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 1.26, | |
| "grad_norm": 12.433682723330731, | |
| "learning_rate": 1.12e-06, | |
| "loss": 1.4209, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 1.28, | |
| "grad_norm": 17.533881821169835, | |
| "learning_rate": 1.1399999999999999e-06, | |
| "loss": 1.1525, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 1.3, | |
| "grad_norm": 21.371274746194803, | |
| "learning_rate": 1.16e-06, | |
| "loss": 1.4803, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 1.33, | |
| "grad_norm": 12.91252932992948, | |
| "learning_rate": 1.18e-06, | |
| "loss": 1.342, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 1.35, | |
| "grad_norm": 15.301368123637772, | |
| "learning_rate": 1.2e-06, | |
| "loss": 1.3903, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 1.37, | |
| "grad_norm": 20.669912185962247, | |
| "learning_rate": 1.22e-06, | |
| "loss": 1.3419, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 1.4, | |
| "grad_norm": 14.306797274099731, | |
| "learning_rate": 1.24e-06, | |
| "loss": 1.3697, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 1.42, | |
| "grad_norm": 14.399414559225576, | |
| "learning_rate": 1.26e-06, | |
| "loss": 1.4607, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 1.44, | |
| "grad_norm": 14.679866284169409, | |
| "learning_rate": 1.28e-06, | |
| "loss": 1.4576, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 1.47, | |
| "grad_norm": 15.66686870904599, | |
| "learning_rate": 1.3e-06, | |
| "loss": 1.4088, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 1.49, | |
| "grad_norm": 15.299128708308798, | |
| "learning_rate": 1.32e-06, | |
| "loss": 1.565, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 1.49, | |
| "eval_loss": 1.5496699810028076, | |
| "eval_runtime": 0.7108, | |
| "eval_samples_per_second": 26.732, | |
| "eval_steps_per_second": 4.221, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 1.51, | |
| "grad_norm": 17.72956466975166, | |
| "learning_rate": 1.34e-06, | |
| "loss": 1.5139, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 1.53, | |
| "grad_norm": 12.468931115539235, | |
| "learning_rate": 1.3600000000000001e-06, | |
| "loss": 1.4851, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 1.56, | |
| "grad_norm": 18.51314016158734, | |
| "learning_rate": 1.38e-06, | |
| "loss": 1.3645, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 1.58, | |
| "grad_norm": 14.247910195754642, | |
| "learning_rate": 1.4e-06, | |
| "loss": 1.3199, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 1.6, | |
| "grad_norm": 14.996917407863565, | |
| "learning_rate": 1.42e-06, | |
| "loss": 1.5712, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 1.63, | |
| "grad_norm": 13.685317610155064, | |
| "learning_rate": 1.44e-06, | |
| "loss": 1.37, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 1.65, | |
| "grad_norm": 20.67582737513748, | |
| "learning_rate": 1.46e-06, | |
| "loss": 1.5281, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 1.67, | |
| "grad_norm": 24.21873361894607, | |
| "learning_rate": 1.48e-06, | |
| "loss": 1.5432, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 1.7, | |
| "grad_norm": 18.530551736423014, | |
| "learning_rate": 1.5e-06, | |
| "loss": 1.5514, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 1.72, | |
| "grad_norm": 13.006185380590916, | |
| "learning_rate": 1.5199999999999998e-06, | |
| "loss": 1.3638, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 1.74, | |
| "grad_norm": 16.338789878308003, | |
| "learning_rate": 1.5399999999999999e-06, | |
| "loss": 1.2469, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 1.74, | |
| "eval_loss": 1.5485225915908813, | |
| "eval_runtime": 0.7202, | |
| "eval_samples_per_second": 26.383, | |
| "eval_steps_per_second": 4.166, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 1.77, | |
| "grad_norm": 12.226940015138254, | |
| "learning_rate": 1.5599999999999999e-06, | |
| "loss": 1.2613, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 1.79, | |
| "grad_norm": 16.35044079107655, | |
| "learning_rate": 1.58e-06, | |
| "loss": 1.3039, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 1.81, | |
| "grad_norm": 16.425588490991444, | |
| "learning_rate": 1.6e-06, | |
| "loss": 1.5878, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 1.84, | |
| "grad_norm": 23.31190960420298, | |
| "learning_rate": 1.62e-06, | |
| "loss": 1.1774, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 1.86, | |
| "grad_norm": 12.69134252831033, | |
| "learning_rate": 1.6399999999999998e-06, | |
| "loss": 1.2935, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 1.88, | |
| "grad_norm": 14.81387052352272, | |
| "learning_rate": 1.6599999999999998e-06, | |
| "loss": 1.2247, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 1.91, | |
| "grad_norm": 12.705757722264147, | |
| "learning_rate": 1.6799999999999998e-06, | |
| "loss": 1.4163, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 1.93, | |
| "grad_norm": 15.14994784808861, | |
| "learning_rate": 1.6999999999999998e-06, | |
| "loss": 1.329, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 1.95, | |
| "grad_norm": 25.616427443741422, | |
| "learning_rate": 1.7199999999999998e-06, | |
| "loss": 1.3684, | |
| "step": 86 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 215, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 5, | |
| "save_steps": 43, | |
| "total_flos": 1.2022908877327565e+17, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |