| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.05653710247349823, | |
| "eval_steps": 5, | |
| "global_step": 20, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0028268551236749115, | |
| "grad_norm": NaN, | |
| "learning_rate": 0.0, | |
| "loss": 7.6743, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0028268551236749115, | |
| "eval_loss": 7.129258632659912, | |
| "eval_runtime": 13.6697, | |
| "eval_samples_per_second": 43.6, | |
| "eval_steps_per_second": 5.487, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.005653710247349823, | |
| "grad_norm": NaN, | |
| "learning_rate": 0.0, | |
| "loss": 7.3349, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.008480565371024734, | |
| "grad_norm": 5.114774703979492, | |
| "learning_rate": 2e-05, | |
| "loss": 7.4669, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.011307420494699646, | |
| "grad_norm": 4.991275310516357, | |
| "learning_rate": 4e-05, | |
| "loss": 7.1521, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.014134275618374558, | |
| "grad_norm": NaN, | |
| "learning_rate": 4e-05, | |
| "loss": 7.7429, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.014134275618374558, | |
| "eval_loss": 7.097643852233887, | |
| "eval_runtime": 20.7137, | |
| "eval_samples_per_second": 28.773, | |
| "eval_steps_per_second": 3.621, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.01696113074204947, | |
| "grad_norm": 4.559197902679443, | |
| "learning_rate": 6e-05, | |
| "loss": 7.1861, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.019787985865724382, | |
| "grad_norm": 3.6284019947052, | |
| "learning_rate": 8e-05, | |
| "loss": 6.9673, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.022614840989399292, | |
| "grad_norm": 4.3491082191467285, | |
| "learning_rate": 0.0001, | |
| "loss": 7.2919, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.025441696113074206, | |
| "grad_norm": 3.08682918548584, | |
| "learning_rate": 0.00012, | |
| "loss": 7.258, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.028268551236749116, | |
| "grad_norm": 3.268092393875122, | |
| "learning_rate": 0.00014, | |
| "loss": 7.0434, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.028268551236749116, | |
| "eval_loss": 6.333721160888672, | |
| "eval_runtime": 20.9199, | |
| "eval_samples_per_second": 28.49, | |
| "eval_steps_per_second": 3.585, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.03109540636042403, | |
| "grad_norm": 2.161461114883423, | |
| "learning_rate": 0.00016, | |
| "loss": 6.1538, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.03392226148409894, | |
| "grad_norm": 2.0395100116729736, | |
| "learning_rate": 0.00018, | |
| "loss": 6.6144, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.03674911660777385, | |
| "grad_norm": 2.263904094696045, | |
| "learning_rate": 0.0002, | |
| "loss": 6.1712, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.039575971731448764, | |
| "grad_norm": 1.9835869073867798, | |
| "learning_rate": 0.00019510565162951537, | |
| "loss": 6.0672, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.04240282685512368, | |
| "grad_norm": 1.9458274841308594, | |
| "learning_rate": 0.00018090169943749476, | |
| "loss": 5.7719, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.04240282685512368, | |
| "eval_loss": 5.498767375946045, | |
| "eval_runtime": 20.8896, | |
| "eval_samples_per_second": 28.531, | |
| "eval_steps_per_second": 3.59, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.045229681978798585, | |
| "grad_norm": 1.6456574201583862, | |
| "learning_rate": 0.00015877852522924732, | |
| "loss": 5.6195, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.0480565371024735, | |
| "grad_norm": 1.413881540298462, | |
| "learning_rate": 0.00013090169943749476, | |
| "loss": 5.3649, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.05088339222614841, | |
| "grad_norm": 1.8699934482574463, | |
| "learning_rate": 0.0001, | |
| "loss": 5.6544, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.05371024734982332, | |
| "grad_norm": 2.320040464401245, | |
| "learning_rate": 6.909830056250527e-05, | |
| "loss": 5.3966, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.05653710247349823, | |
| "grad_norm": 1.5797550678253174, | |
| "learning_rate": 4.12214747707527e-05, | |
| "loss": 5.695, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.05653710247349823, | |
| "eval_loss": 5.102807998657227, | |
| "eval_runtime": 20.9456, | |
| "eval_samples_per_second": 28.455, | |
| "eval_steps_per_second": 3.581, | |
| "step": 20 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 20, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 5, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 2.638283774571315e+16, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |