| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 2.9647058823529413, | |
| "eval_steps": 500, | |
| "global_step": 189, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 17.495570135172613, | |
| "learning_rate": 1.0526315789473685e-06, | |
| "loss": 1.5715, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 10.76150785270408, | |
| "learning_rate": 5.263157894736842e-06, | |
| "loss": 1.4595, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 4.725613013527171, | |
| "learning_rate": 1.0526315789473684e-05, | |
| "loss": 1.0131, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "grad_norm": 1.9709789387558339, | |
| "learning_rate": 1.578947368421053e-05, | |
| "loss": 0.9452, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "grad_norm": 1.5718916258677555, | |
| "learning_rate": 1.9998292504580528e-05, | |
| "loss": 0.8657, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.39, | |
| "grad_norm": 1.2587381103664048, | |
| "learning_rate": 1.993859136895274e-05, | |
| "loss": 0.7733, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.47, | |
| "grad_norm": 1.0676841590173272, | |
| "learning_rate": 1.979409767601366e-05, | |
| "loss": 0.7396, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.55, | |
| "grad_norm": 0.9830974732786681, | |
| "learning_rate": 1.956604419500441e-05, | |
| "loss": 0.712, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.63, | |
| "grad_norm": 0.9898127673571485, | |
| "learning_rate": 1.9256376597815565e-05, | |
| "loss": 0.6854, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.71, | |
| "grad_norm": 0.9235108049390474, | |
| "learning_rate": 1.886773685920062e-05, | |
| "loss": 0.6959, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.78, | |
| "grad_norm": 0.9176451831476561, | |
| "learning_rate": 1.840344071637893e-05, | |
| "loss": 0.6707, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.86, | |
| "grad_norm": 1.0570214869530683, | |
| "learning_rate": 1.7867449380334834e-05, | |
| "loss": 0.6703, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.94, | |
| "grad_norm": 0.8971457648885479, | |
| "learning_rate": 1.7264335740162244e-05, | |
| "loss": 0.6791, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.99, | |
| "eval_loss": 0.675278902053833, | |
| "eval_runtime": 7.6872, | |
| "eval_samples_per_second": 19.773, | |
| "eval_steps_per_second": 2.472, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 1.02, | |
| "grad_norm": 1.0643549643394743, | |
| "learning_rate": 1.659924534878723e-05, | |
| "loss": 0.6235, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 1.1, | |
| "grad_norm": 0.9490599850362894, | |
| "learning_rate": 1.5877852522924733e-05, | |
| "loss": 0.4562, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 1.18, | |
| "grad_norm": 0.9144117495898926, | |
| "learning_rate": 1.510631193180907e-05, | |
| "loss": 0.4195, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 1.25, | |
| "grad_norm": 0.8786872768630574, | |
| "learning_rate": 1.429120608772609e-05, | |
| "loss": 0.4257, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 1.33, | |
| "grad_norm": 0.815215205518059, | |
| "learning_rate": 1.3439489186339283e-05, | |
| "loss": 0.4233, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 1.41, | |
| "grad_norm": 0.8699270815065504, | |
| "learning_rate": 1.2558427775944357e-05, | |
| "loss": 0.4208, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 1.49, | |
| "grad_norm": 0.8685370301630462, | |
| "learning_rate": 1.16555387618413e-05, | |
| "loss": 0.4242, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 1.57, | |
| "grad_norm": 1.1593148629393575, | |
| "learning_rate": 1.073852527474874e-05, | |
| "loss": 0.4409, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.65, | |
| "grad_norm": 0.8148155791935645, | |
| "learning_rate": 9.815210950408703e-06, | |
| "loss": 0.4196, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 1.73, | |
| "grad_norm": 0.8416381894160022, | |
| "learning_rate": 8.893473181084993e-06, | |
| "loss": 0.4375, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 1.8, | |
| "grad_norm": 0.8570128253866367, | |
| "learning_rate": 7.9811759084299e-06, | |
| "loss": 0.42, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 1.88, | |
| "grad_norm": 0.9019501174530336, | |
| "learning_rate": 7.086102531106755e-06, | |
| "loss": 0.4289, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 1.96, | |
| "grad_norm": 0.8041684983988516, | |
| "learning_rate": 6.215889499576898e-06, | |
| "loss": 0.4164, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 1.99, | |
| "eval_loss": 0.7019588351249695, | |
| "eval_runtime": 7.6577, | |
| "eval_samples_per_second": 19.849, | |
| "eval_steps_per_second": 2.481, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 2.04, | |
| "grad_norm": 0.8404401588139048, | |
| "learning_rate": 5.3779611645968696e-06, | |
| "loss": 0.3314, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 2.12, | |
| "grad_norm": 1.0503604307814172, | |
| "learning_rate": 4.579466435275506e-06, | |
| "loss": 0.2445, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 2.2, | |
| "grad_norm": 0.7976145439951375, | |
| "learning_rate": 3.827217787102072e-06, | |
| "loss": 0.2222, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 2.27, | |
| "grad_norm": 0.7577793168033735, | |
| "learning_rate": 3.1276331403073733e-06, | |
| "loss": 0.2256, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 2.35, | |
| "grad_norm": 0.7242853721207333, | |
| "learning_rate": 2.4866811044312667e-06, | |
| "loss": 0.2153, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 2.43, | |
| "grad_norm": 0.7007575307345968, | |
| "learning_rate": 1.9098300562505266e-06, | |
| "loss": 0.2148, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 2.51, | |
| "grad_norm": 0.7109375761250729, | |
| "learning_rate": 1.4020014855162755e-06, | |
| "loss": 0.2203, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 2.59, | |
| "grad_norm": 0.7041755633099551, | |
| "learning_rate": 9.675280065387117e-07, | |
| "loss": 0.2123, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 2.67, | |
| "grad_norm": 0.6924970223941338, | |
| "learning_rate": 6.101163938494359e-07, | |
| "loss": 0.2207, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 2.75, | |
| "grad_norm": 0.6779904104237012, | |
| "learning_rate": 3.328159573081258e-07, | |
| "loss": 0.2098, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 2.82, | |
| "grad_norm": 0.717318798866986, | |
| "learning_rate": 1.3799252646597428e-07, | |
| "loss": 0.2168, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 2.9, | |
| "grad_norm": 0.7202766403388026, | |
| "learning_rate": 2.7308266142119788e-08, | |
| "loss": 0.2065, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 2.96, | |
| "eval_loss": 0.8261664509773254, | |
| "eval_runtime": 7.6483, | |
| "eval_samples_per_second": 19.874, | |
| "eval_steps_per_second": 2.484, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 2.96, | |
| "step": 189, | |
| "total_flos": 64372667842560.0, | |
| "train_loss": 0.4923532526329081, | |
| "train_runtime": 4087.1741, | |
| "train_samples_per_second": 1.494, | |
| "train_steps_per_second": 0.046 | |
| } | |
| ], | |
| "logging_steps": 5, | |
| "max_steps": 189, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "save_steps": 500, | |
| "total_flos": 64372667842560.0, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |