| { | |
| "best_global_step": null, | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 5.0, | |
| "eval_steps": 500, | |
| "global_step": 215, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.11627906976744186, | |
| "grad_norm": 1.426078351046737, | |
| "learning_rate": 2.272727272727273e-05, | |
| "loss": 1.0374, | |
| "num_tokens": 616397.0, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.23255813953488372, | |
| "grad_norm": 1.4337078601293307, | |
| "learning_rate": 4.545454545454546e-05, | |
| "loss": 0.8459, | |
| "num_tokens": 1218796.0, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.3488372093023256, | |
| "grad_norm": 0.9412483454905438, | |
| "learning_rate": 4.9957324896583496e-05, | |
| "loss": 0.717, | |
| "num_tokens": 1836857.0, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.46511627906976744, | |
| "grad_norm": 1.2421575625153427, | |
| "learning_rate": 4.9784234694692117e-05, | |
| "loss": 0.6697, | |
| "num_tokens": 2480724.0, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.5813953488372093, | |
| "grad_norm": 1.039756996050679, | |
| "learning_rate": 4.9479087149591016e-05, | |
| "loss": 0.6576, | |
| "num_tokens": 3108504.0, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.6976744186046512, | |
| "grad_norm": 0.7843538876272397, | |
| "learning_rate": 4.9043690578893296e-05, | |
| "loss": 0.6104, | |
| "num_tokens": 3730397.0, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.813953488372093, | |
| "grad_norm": 0.7523189873152597, | |
| "learning_rate": 4.8480625161598e-05, | |
| "loss": 0.5963, | |
| "num_tokens": 4370252.0, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.9302325581395349, | |
| "grad_norm": 1.861170722861451, | |
| "learning_rate": 4.779322764783724e-05, | |
| "loss": 0.7153, | |
| "num_tokens": 4903914.0, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 1.0465116279069768, | |
| "grad_norm": 1.6296026123428795, | |
| "learning_rate": 4.6985571585149876e-05, | |
| "loss": 0.5757, | |
| "num_tokens": 5419349.0, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 1.1627906976744187, | |
| "grad_norm": 0.8964132137282615, | |
| "learning_rate": 4.60624431784618e-05, | |
| "loss": 0.519, | |
| "num_tokens": 6028764.0, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 1.2790697674418605, | |
| "grad_norm": 0.7192499555698023, | |
| "learning_rate": 4.502931292682758e-05, | |
| "loss": 0.4979, | |
| "num_tokens": 6625180.0, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 1.3953488372093024, | |
| "grad_norm": 0.8051552030202953, | |
| "learning_rate": 4.389230320501529e-05, | |
| "loss": 0.5165, | |
| "num_tokens": 7255749.0, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 1.5116279069767442, | |
| "grad_norm": 53.24305130149784, | |
| "learning_rate": 4.2658151982047536e-05, | |
| "loss": 0.525, | |
| "num_tokens": 7834089.0, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 1.627906976744186, | |
| "grad_norm": 0.6314644559742063, | |
| "learning_rate": 4.133417289170434e-05, | |
| "loss": 0.5228, | |
| "num_tokens": 8481532.0, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 1.744186046511628, | |
| "grad_norm": 0.7714043941894847, | |
| "learning_rate": 3.992821189161138e-05, | |
| "loss": 0.5209, | |
| "num_tokens": 9110904.0, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 1.8604651162790697, | |
| "grad_norm": 0.793200983051883, | |
| "learning_rate": 3.8448600767754265e-05, | |
| "loss": 0.5047, | |
| "num_tokens": 9725035.0, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 1.9767441860465116, | |
| "grad_norm": 0.784007077545182, | |
| "learning_rate": 3.6904107759952624e-05, | |
| "loss": 0.485, | |
| "num_tokens": 10320646.0, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 2.0930232558139537, | |
| "grad_norm": 0.7267651662469797, | |
| "learning_rate": 3.5303885600889935e-05, | |
| "loss": 0.4358, | |
| "num_tokens": 10877844.0, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 2.2093023255813953, | |
| "grad_norm": 0.7334836153849675, | |
| "learning_rate": 3.365741727662187e-05, | |
| "loss": 0.4085, | |
| "num_tokens": 11472723.0, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 2.3255813953488373, | |
| "grad_norm": 0.9638635507157925, | |
| "learning_rate": 3.1974459829988826e-05, | |
| "loss": 0.411, | |
| "num_tokens": 12095224.0, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 2.441860465116279, | |
| "grad_norm": 0.566412620058458, | |
| "learning_rate": 3.026498653995607e-05, | |
| "loss": 0.3724, | |
| "num_tokens": 12706666.0, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 2.558139534883721, | |
| "grad_norm": 0.7503759786149439, | |
| "learning_rate": 2.8539127819529143e-05, | |
| "loss": 0.3938, | |
| "num_tokens": 13319481.0, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 2.6744186046511627, | |
| "grad_norm": 0.6009758378668209, | |
| "learning_rate": 2.6807111182486174e-05, | |
| "loss": 0.3838, | |
| "num_tokens": 13929481.0, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 2.7906976744186047, | |
| "grad_norm": 0.7071770345433811, | |
| "learning_rate": 2.507920063468675e-05, | |
| "loss": 0.3704, | |
| "num_tokens": 14531032.0, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 2.9069767441860463, | |
| "grad_norm": 0.6933855131506369, | |
| "learning_rate": 2.3365635849127166e-05, | |
| "loss": 0.3946, | |
| "num_tokens": 15156384.0, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 3.0232558139534884, | |
| "grad_norm": 0.657092893175657, | |
| "learning_rate": 2.1676571485193282e-05, | |
| "loss": 0.3853, | |
| "num_tokens": 15707749.0, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 3.13953488372093, | |
| "grad_norm": 0.7324356510004003, | |
| "learning_rate": 2.002201701170766e-05, | |
| "loss": 0.3217, | |
| "num_tokens": 16340606.0, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 3.255813953488372, | |
| "grad_norm": 0.6246116853697837, | |
| "learning_rate": 1.8411777390382367e-05, | |
| "loss": 0.2979, | |
| "num_tokens": 16944484.0, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 3.3720930232558137, | |
| "grad_norm": 0.6458581868017114, | |
| "learning_rate": 1.6855394971189778e-05, | |
| "loss": 0.3229, | |
| "num_tokens": 17553663.0, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 3.488372093023256, | |
| "grad_norm": 0.563906679313201, | |
| "learning_rate": 1.5362092943981944e-05, | |
| "loss": 0.3026, | |
| "num_tokens": 18188173.0, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 3.604651162790698, | |
| "grad_norm": 0.5655342559997661, | |
| "learning_rate": 1.3940720681466734e-05, | |
| "loss": 0.2997, | |
| "num_tokens": 18819640.0, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 3.7209302325581395, | |
| "grad_norm": 0.6061135438357336, | |
| "learning_rate": 1.2599701297440578e-05, | |
| "loss": 0.3158, | |
| "num_tokens": 19430442.0, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 3.8372093023255816, | |
| "grad_norm": 0.5589360223860185, | |
| "learning_rate": 1.1346981731050059e-05, | |
| "loss": 0.2665, | |
| "num_tokens": 20003957.0, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 3.953488372093023, | |
| "grad_norm": 0.5770875534708484, | |
| "learning_rate": 1.0189985652885225e-05, | |
| "loss": 0.3215, | |
| "num_tokens": 20589734.0, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 4.069767441860465, | |
| "grad_norm": 0.501582071030718, | |
| "learning_rate": 9.135569471985015e-06, | |
| "loss": 0.2683, | |
| "num_tokens": 21131356.0, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 4.186046511627907, | |
| "grad_norm": 0.6129991511301848, | |
| "learning_rate": 8.189981704459338e-06, | |
| "loss": 0.2346, | |
| "num_tokens": 21741947.0, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 4.3023255813953485, | |
| "grad_norm": 0.5871856744225346, | |
| "learning_rate": 7.358825944511101e-06, | |
| "loss": 0.2473, | |
| "num_tokens": 22338072.0, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 4.4186046511627906, | |
| "grad_norm": 0.5101346511723733, | |
| "learning_rate": 6.647027657293392e-06, | |
| "loss": 0.2479, | |
| "num_tokens": 22958505.0, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 4.534883720930233, | |
| "grad_norm": 0.6244435714805572, | |
| "learning_rate": 6.0588049903889796e-06, | |
| "loss": 0.2544, | |
| "num_tokens": 23589480.0, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 4.651162790697675, | |
| "grad_norm": 0.5334908287870592, | |
| "learning_rate": 5.597643776884412e-06, | |
| "loss": 0.2294, | |
| "num_tokens": 24195388.0, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 4.767441860465116, | |
| "grad_norm": 0.5107011201150088, | |
| "learning_rate": 5.266276878171664e-06, | |
| "loss": 0.2345, | |
| "num_tokens": 24813280.0, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 4.883720930232558, | |
| "grad_norm": 0.5467832269520996, | |
| "learning_rate": 5.0666679888927605e-06, | |
| "loss": 0.2326, | |
| "num_tokens": 25383723.0, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 0.5165825558969128, | |
| "learning_rate": 5e-06, | |
| "loss": 0.2413, | |
| "num_tokens": 25965455.0, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "step": 215, | |
| "total_flos": 22842265731072.0, | |
| "train_loss": 0.4351522645284963, | |
| "train_runtime": 311.5406, | |
| "train_samples_per_second": 5.457, | |
| "train_steps_per_second": 0.69 | |
| } | |
| ], | |
| "logging_steps": 5, | |
| "max_steps": 215, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 5, | |
| "save_steps": 100, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 22842265731072.0, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |