| { |
| "best_global_step": null, |
| "best_metric": null, |
| "best_model_checkpoint": null, |
| "epoch": 0.631578947368421, |
| "eval_steps": 12, |
| "global_step": 30, |
| "is_hyper_param_search": false, |
| "is_local_process_zero": true, |
| "is_world_process_zero": true, |
| "log_history": [ |
| { |
| "epoch": 0, |
| "eval_loss": 1.2786176204681396, |
| "eval_ppl": 3.59167, |
| "eval_runtime": 276.2614, |
| "eval_samples_per_second": 0.576, |
| "eval_steps_per_second": 0.576, |
| "memory/device_reserved (GiB)": 42.45, |
| "memory/max_active (GiB)": 40.26, |
| "memory/max_allocated (GiB)": 38.06, |
| "step": 0 |
| }, |
| { |
| "epoch": 0.10526315789473684, |
| "grad_norm": 4.365630626678467, |
| "learning_rate": 7e-06, |
| "loss": 1.1477760314941405, |
| "memory/device_reserved (GiB)": 71.27, |
| "memory/max_active (GiB)": 53.99, |
| "memory/max_allocated (GiB)": 51.79, |
| "ppl": 3.15118, |
| "step": 5, |
| "tokens/total": 2621440, |
| "tokens/trainable": 489230 |
| }, |
| { |
| "epoch": 0.21052631578947367, |
| "grad_norm": 2.4648349285125732, |
| "learning_rate": 6.952144421988455e-06, |
| "loss": 0.9022371292114257, |
| "memory/device_reserved (GiB)": 71.27, |
| "memory/max_active (GiB)": 53.96, |
| "memory/max_allocated (GiB)": 51.79, |
| "ppl": 2.46511, |
| "step": 10, |
| "tokens/total": 5242880, |
| "tokens/train_per_sec_per_gpu": 225.38, |
| "tokens/trainable": 896733 |
| }, |
| { |
| "epoch": 0.25263157894736843, |
| "eval_loss": 0.7994529604911804, |
| "eval_ppl": 2.22432, |
| "eval_runtime": 276.9872, |
| "eval_samples_per_second": 0.574, |
| "eval_steps_per_second": 0.574, |
| "memory/device_reserved (GiB)": 71.27, |
| "memory/max_active (GiB)": 53.96, |
| "memory/max_allocated (GiB)": 51.79, |
| "step": 12 |
| }, |
| { |
| "epoch": 0.3157894736842105, |
| "grad_norm": 1.6509625911712646, |
| "learning_rate": 6.810031755475611e-06, |
| "loss": 0.7818592548370361, |
| "memory/device_reserved (GiB)": 74.72, |
| "memory/max_active (GiB)": 53.96, |
| "memory/max_allocated (GiB)": 51.79, |
| "ppl": 2.18553, |
| "step": 15, |
| "tokens/total": 7864320, |
| "tokens/train_per_sec_per_gpu": 208.77, |
| "tokens/trainable": 1345756 |
| }, |
| { |
| "epoch": 0.42105263157894735, |
| "grad_norm": 1.7147045135498047, |
| "learning_rate": 6.577980021920982e-06, |
| "loss": 0.7099627494812012, |
| "memory/device_reserved (GiB)": 74.72, |
| "memory/max_active (GiB)": 53.96, |
| "memory/max_allocated (GiB)": 51.79, |
| "ppl": 2.03392, |
| "step": 20, |
| "tokens/total": 10485760, |
| "tokens/train_per_sec_per_gpu": 361.48, |
| "tokens/trainable": 1808219 |
| }, |
| { |
| "epoch": 0.5052631578947369, |
| "eval_loss": 0.6555427312850952, |
| "eval_ppl": 1.92619, |
| "eval_runtime": 248.0295, |
| "eval_samples_per_second": 0.641, |
| "eval_steps_per_second": 0.641, |
| "memory/device_reserved (GiB)": 74.72, |
| "memory/max_active (GiB)": 53.96, |
| "memory/max_allocated (GiB)": 51.79, |
| "step": 24 |
| }, |
| { |
| "epoch": 0.5263157894736842, |
| "grad_norm": 1.5091627836227417, |
| "learning_rate": 6.26303999582478e-06, |
| "loss": 0.6799451828002929, |
| "memory/device_reserved (GiB)": 74.7, |
| "memory/max_active (GiB)": 53.96, |
| "memory/max_allocated (GiB)": 51.79, |
| "ppl": 1.97377, |
| "step": 25, |
| "tokens/total": 13107200, |
| "tokens/train_per_sec_per_gpu": 268.83, |
| "tokens/trainable": 2291286 |
| }, |
| { |
| "epoch": 0.631578947368421, |
| "grad_norm": 1.6102690696716309, |
| "learning_rate": 5.874780970512599e-06, |
| "loss": 0.625995397567749, |
| "memory/device_reserved (GiB)": 74.7, |
| "memory/max_active (GiB)": 53.99, |
| "memory/max_allocated (GiB)": 51.79, |
| "ppl": 1.87011, |
| "step": 30, |
| "tokens/total": 15728640, |
| "tokens/train_per_sec_per_gpu": 380.95, |
| "tokens/trainable": 2755207 |
| } |
| ], |
| "logging_steps": 5, |
| "max_steps": 94, |
| "num_input_tokens_seen": 0, |
| "num_train_epochs": 2, |
| "save_steps": 10, |
| "stateful_callbacks": { |
| "TrainerControl": { |
| "args": { |
| "should_epoch_stop": false, |
| "should_evaluate": false, |
| "should_log": false, |
| "should_save": true, |
| "should_training_stop": false |
| }, |
| "attributes": {} |
| } |
| }, |
| "total_flos": 137398375677952.0, |
| "train_batch_size": 1, |
| "trial_name": null, |
| "trial_params": null |
| } |
|
|
