| { | |
| "best_global_step": null, | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 9.828478964401295, | |
| "eval_steps": 500, | |
| "global_step": 40, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.20711974110032363, | |
| "grad_norm": 16.311434606415148, | |
| "learning_rate": 2e-05, | |
| "loss": 0.6532, | |
| "mean_token_accuracy": 0.8563957028090954, | |
| "num_tokens": 523828.0, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.41423948220064727, | |
| "grad_norm": 15.736907422857884, | |
| "learning_rate": 1.9e-05, | |
| "loss": 0.4982, | |
| "mean_token_accuracy": 0.8662955602630973, | |
| "num_tokens": 1047723.0, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.6213592233009708, | |
| "grad_norm": 3.447251157289743, | |
| "learning_rate": 1.8e-05, | |
| "loss": 0.4425, | |
| "mean_token_accuracy": 0.873533152975142, | |
| "num_tokens": 1565521.0, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.8284789644012945, | |
| "grad_norm": 4.507555220293444, | |
| "learning_rate": 1.7e-05, | |
| "loss": 0.413, | |
| "mean_token_accuracy": 0.8806356741115451, | |
| "num_tokens": 2089809.0, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 1.2071197411003236, | |
| "grad_norm": 1.9178218589930014, | |
| "learning_rate": 1.6000000000000003e-05, | |
| "loss": 0.8156, | |
| "mean_token_accuracy": 0.8827247977256775, | |
| "num_tokens": 2940027.0, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 1.4142394822006472, | |
| "grad_norm": 0.9419526829426662, | |
| "learning_rate": 1.5000000000000002e-05, | |
| "loss": 0.375, | |
| "mean_token_accuracy": 0.8892204724252224, | |
| "num_tokens": 3459739.0, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 1.6213592233009708, | |
| "grad_norm": 0.73637220317646, | |
| "learning_rate": 1.4e-05, | |
| "loss": 0.3594, | |
| "mean_token_accuracy": 0.8936486886814237, | |
| "num_tokens": 3983567.0, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 1.8284789644012944, | |
| "grad_norm": 0.6243754300549781, | |
| "learning_rate": 1.3000000000000001e-05, | |
| "loss": 0.3622, | |
| "mean_token_accuracy": 0.8927655592560768, | |
| "num_tokens": 4507855.0, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 2.207119741100324, | |
| "grad_norm": 0.5542153802684493, | |
| "learning_rate": 1.2e-05, | |
| "loss": 0.6934, | |
| "mean_token_accuracy": 0.8979568992342267, | |
| "num_tokens": 5364879.0, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 2.414239482200647, | |
| "grad_norm": 0.5137676948302884, | |
| "learning_rate": 1.1000000000000001e-05, | |
| "loss": 0.3317, | |
| "mean_token_accuracy": 0.9016338279470801, | |
| "num_tokens": 5888452.0, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 2.6213592233009706, | |
| "grad_norm": 0.4227012567455179, | |
| "learning_rate": 1e-05, | |
| "loss": 0.3075, | |
| "mean_token_accuracy": 0.9077446917071939, | |
| "num_tokens": 6411678.0, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 2.8284789644012944, | |
| "grad_norm": 0.521003496128672, | |
| "learning_rate": 9e-06, | |
| "loss": 0.3296, | |
| "mean_token_accuracy": 0.9018635489046574, | |
| "num_tokens": 6929430.0, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 3.207119741100324, | |
| "grad_norm": 0.6009267217587392, | |
| "learning_rate": 8.000000000000001e-06, | |
| "loss": 0.5879, | |
| "mean_token_accuracy": 0.9123640724590847, | |
| "num_tokens": 7788482.0, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 3.414239482200647, | |
| "grad_norm": 0.38313891986777304, | |
| "learning_rate": 7e-06, | |
| "loss": 0.2942, | |
| "mean_token_accuracy": 0.9121569795534015, | |
| "num_tokens": 8307992.0, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 3.6213592233009706, | |
| "grad_norm": 0.42965764040529053, | |
| "learning_rate": 6e-06, | |
| "loss": 0.2851, | |
| "mean_token_accuracy": 0.914524182677269, | |
| "num_tokens": 8832280.0, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 3.8284789644012944, | |
| "grad_norm": 0.39075483271320793, | |
| "learning_rate": 5e-06, | |
| "loss": 0.2846, | |
| "mean_token_accuracy": 0.9148006569594145, | |
| "num_tokens": 9350219.0, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 4.207119741100324, | |
| "grad_norm": 0.5041579865544957, | |
| "learning_rate": 4.000000000000001e-06, | |
| "loss": 0.5225, | |
| "mean_token_accuracy": 0.9207058066413516, | |
| "num_tokens": 10206773.0, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 4.414239482200648, | |
| "grad_norm": 0.3226220487928564, | |
| "learning_rate": 3e-06, | |
| "loss": 0.2554, | |
| "mean_token_accuracy": 0.9235557727515697, | |
| "num_tokens": 10731061.0, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 4.621359223300971, | |
| "grad_norm": 0.29138584575319715, | |
| "learning_rate": 2.0000000000000003e-06, | |
| "loss": 0.2589, | |
| "mean_token_accuracy": 0.9219849593937397, | |
| "num_tokens": 11254174.0, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 4.828478964401294, | |
| "grad_norm": 0.3074072577482687, | |
| "learning_rate": 1.0000000000000002e-06, | |
| "loss": 0.2573, | |
| "mean_token_accuracy": 0.9227853054180741, | |
| "num_tokens": 11772003.0, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 5.207119741100324, | |
| "grad_norm": 0.29226028229762824, | |
| "learning_rate": 0.0, | |
| "loss": 0.2581, | |
| "mean_token_accuracy": 0.9223490478470922, | |
| "num_tokens": 520682.0, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 5.414239482200648, | |
| "grad_norm": 0.3039506268459287, | |
| "learning_rate": 9.5e-06, | |
| "loss": 0.2612, | |
| "mean_token_accuracy": 0.921809334307909, | |
| "num_tokens": 1042545.0, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 5.621359223300971, | |
| "grad_norm": 0.3292461938787561, | |
| "learning_rate": 9e-06, | |
| "loss": 0.2338, | |
| "mean_token_accuracy": 0.9292812049388885, | |
| "num_tokens": 1563697.0, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 5.828478964401294, | |
| "grad_norm": 0.46420193174977303, | |
| "learning_rate": 8.5e-06, | |
| "loss": 0.2446, | |
| "mean_token_accuracy": 0.9261427698656917, | |
| "num_tokens": 2087270.0, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 6.207119741100324, | |
| "grad_norm": 0.5152631282390204, | |
| "learning_rate": 8.000000000000001e-06, | |
| "loss": 0.4273, | |
| "mean_token_accuracy": 0.9354684165545872, | |
| "num_tokens": 2942515.0, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 6.414239482200648, | |
| "grad_norm": 0.397171201247213, | |
| "learning_rate": 7.500000000000001e-06, | |
| "loss": 0.1971, | |
| "mean_token_accuracy": 0.9410504633560777, | |
| "num_tokens": 3464980.0, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 6.621359223300971, | |
| "grad_norm": 0.4038092235036345, | |
| "learning_rate": 7e-06, | |
| "loss": 0.1936, | |
| "mean_token_accuracy": 0.9419528925791383, | |
| "num_tokens": 3988553.0, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 6.828478964401294, | |
| "grad_norm": 0.4383287277331565, | |
| "learning_rate": 6.5000000000000004e-06, | |
| "loss": 0.1982, | |
| "mean_token_accuracy": 0.9402873041108251, | |
| "num_tokens": 4512841.0, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 7.207119741100324, | |
| "grad_norm": 0.5819451162543645, | |
| "learning_rate": 6e-06, | |
| "loss": 0.3495, | |
| "mean_token_accuracy": 0.9478773713111878, | |
| "num_tokens": 5363834.0, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 7.414239482200648, | |
| "grad_norm": 0.35512055236444356, | |
| "learning_rate": 5.500000000000001e-06, | |
| "loss": 0.1599, | |
| "mean_token_accuracy": 0.9522946244105697, | |
| "num_tokens": 5887407.0, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 7.621359223300971, | |
| "grad_norm": 0.6014961197056318, | |
| "learning_rate": 5e-06, | |
| "loss": 0.1586, | |
| "mean_token_accuracy": 0.9526052679866552, | |
| "num_tokens": 6408559.0, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 7.828478964401294, | |
| "grad_norm": 0.4044887603769801, | |
| "learning_rate": 4.5e-06, | |
| "loss": 0.1515, | |
| "mean_token_accuracy": 0.9546759780496359, | |
| "num_tokens": 6930887.0, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 8.207119741100323, | |
| "grad_norm": 0.527657049732786, | |
| "learning_rate": 4.000000000000001e-06, | |
| "loss": 0.2954, | |
| "mean_token_accuracy": 0.9571067219688779, | |
| "num_tokens": 7784698.0, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 8.414239482200648, | |
| "grad_norm": 0.5655954589470128, | |
| "learning_rate": 3.5e-06, | |
| "loss": 0.1301, | |
| "mean_token_accuracy": 0.9615885661914945, | |
| "num_tokens": 8307878.0, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 8.62135922330097, | |
| "grad_norm": 0.3905499777914469, | |
| "learning_rate": 3e-06, | |
| "loss": 0.1309, | |
| "mean_token_accuracy": 0.9616770427674055, | |
| "num_tokens": 8831564.0, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 8.828478964401295, | |
| "grad_norm": 0.3609762708880154, | |
| "learning_rate": 2.5e-06, | |
| "loss": 0.1301, | |
| "mean_token_accuracy": 0.961113647557795, | |
| "num_tokens": 9354489.0, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 9.207119741100323, | |
| "grad_norm": 0.6575328899879626, | |
| "learning_rate": 2.0000000000000003e-06, | |
| "loss": 0.2498, | |
| "mean_token_accuracy": 0.9636295852207002, | |
| "num_tokens": 10210406.0, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 9.414239482200648, | |
| "grad_norm": 0.28343130576610354, | |
| "learning_rate": 1.5e-06, | |
| "loss": 0.1245, | |
| "mean_token_accuracy": 0.9638419672846794, | |
| "num_tokens": 10729521.0, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 9.62135922330097, | |
| "grad_norm": 0.3576238970586715, | |
| "learning_rate": 1.0000000000000002e-06, | |
| "loss": 0.1181, | |
| "mean_token_accuracy": 0.965806245803833, | |
| "num_tokens": 11253094.0, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 9.828478964401295, | |
| "grad_norm": 0.38762707391762197, | |
| "learning_rate": 5.000000000000001e-07, | |
| "loss": 0.1138, | |
| "mean_token_accuracy": 0.9672208018600941, | |
| "num_tokens": 11773251.0, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 9.828478964401295, | |
| "step": 40, | |
| "total_flos": 47054728527872.0, | |
| "train_loss": 0.10314766578376293, | |
| "train_runtime": 4354.8726, | |
| "train_samples_per_second": 1.417, | |
| "train_steps_per_second": 0.009 | |
| } | |
| ], | |
| "logging_steps": 1.0, | |
| "max_steps": 40, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 10, | |
| "save_steps": 10, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 47054728527872.0, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |