| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 9.473684210526315, | |
| "eval_steps": 500, | |
| "global_step": 90, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.10526315789473684, | |
| "grad_norm": 2.171875, | |
| "learning_rate": 2.2222222222222223e-05, | |
| "loss": 3.0858, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.5263157894736842, | |
| "grad_norm": 1.84375, | |
| "learning_rate": 0.00011111111111111112, | |
| "loss": 3.0546, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.9473684210526315, | |
| "eval_loss": 2.956000804901123, | |
| "eval_runtime": 0.5468, | |
| "eval_samples_per_second": 18.288, | |
| "eval_steps_per_second": 1.829, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 1.0526315789473684, | |
| "grad_norm": 1.8359375, | |
| "learning_rate": 0.00019992479525042303, | |
| "loss": 2.6964, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 1.5789473684210527, | |
| "grad_norm": 1.3828125, | |
| "learning_rate": 0.00019730448705798239, | |
| "loss": 2.2821, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_loss": 2.696643829345703, | |
| "eval_runtime": 0.5537, | |
| "eval_samples_per_second": 18.061, | |
| "eval_steps_per_second": 1.806, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 2.1052631578947367, | |
| "grad_norm": 0.87890625, | |
| "learning_rate": 0.0001910362940966147, | |
| "loss": 2.0545, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 2.6315789473684212, | |
| "grad_norm": 0.9921875, | |
| "learning_rate": 0.00018135520702629675, | |
| "loss": 1.8527, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 2.9473684210526314, | |
| "eval_loss": 2.6030924320220947, | |
| "eval_runtime": 0.5536, | |
| "eval_samples_per_second": 18.063, | |
| "eval_steps_per_second": 1.806, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 3.1578947368421053, | |
| "grad_norm": 0.431640625, | |
| "learning_rate": 0.0001686241637868734, | |
| "loss": 1.6976, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 3.6842105263157894, | |
| "grad_norm": 0.5, | |
| "learning_rate": 0.00015332044328016914, | |
| "loss": 1.5789, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_loss": 2.58021879196167, | |
| "eval_runtime": 0.5505, | |
| "eval_samples_per_second": 18.165, | |
| "eval_steps_per_second": 1.817, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 4.2105263157894735, | |
| "grad_norm": 0.43359375, | |
| "learning_rate": 0.00013601777248047105, | |
| "loss": 1.5127, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 4.7368421052631575, | |
| "grad_norm": 0.3984375, | |
| "learning_rate": 0.00011736481776669306, | |
| "loss": 1.4487, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 4.947368421052632, | |
| "eval_loss": 2.5688371658325195, | |
| "eval_runtime": 0.555, | |
| "eval_samples_per_second": 18.017, | |
| "eval_steps_per_second": 1.802, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 5.2631578947368425, | |
| "grad_norm": 0.271484375, | |
| "learning_rate": 9.806086682281758e-05, | |
| "loss": 1.4087, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 5.7894736842105265, | |
| "grad_norm": 0.265625, | |
| "learning_rate": 7.882961277705895e-05, | |
| "loss": 1.3708, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_loss": 2.5638375282287598, | |
| "eval_runtime": 0.5463, | |
| "eval_samples_per_second": 18.305, | |
| "eval_steps_per_second": 1.831, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 6.315789473684211, | |
| "grad_norm": 0.310546875, | |
| "learning_rate": 6.039202339608432e-05, | |
| "loss": 1.3437, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 6.842105263157895, | |
| "grad_norm": 0.25390625, | |
| "learning_rate": 4.343931245134616e-05, | |
| "loss": 1.3286, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 6.947368421052632, | |
| "eval_loss": 2.5667476654052734, | |
| "eval_runtime": 0.5585, | |
| "eval_samples_per_second": 17.905, | |
| "eval_steps_per_second": 1.79, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 7.368421052631579, | |
| "grad_norm": 0.2275390625, | |
| "learning_rate": 2.8607026544210114e-05, | |
| "loss": 1.3139, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 7.894736842105263, | |
| "grad_norm": 0.2216796875, | |
| "learning_rate": 1.6451218858706374e-05, | |
| "loss": 1.3104, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_loss": 2.5672240257263184, | |
| "eval_runtime": 0.5487, | |
| "eval_samples_per_second": 18.224, | |
| "eval_steps_per_second": 1.822, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 8.421052631578947, | |
| "grad_norm": 0.208984375, | |
| "learning_rate": 7.427603073110967e-06, | |
| "loss": 1.3054, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 8.947368421052632, | |
| "grad_norm": 0.2138671875, | |
| "learning_rate": 1.874468937261531e-06, | |
| "loss": 1.3046, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 8.947368421052632, | |
| "eval_loss": 2.567439556121826, | |
| "eval_runtime": 0.561, | |
| "eval_samples_per_second": 17.827, | |
| "eval_steps_per_second": 1.783, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 9.473684210526315, | |
| "grad_norm": 0.212890625, | |
| "learning_rate": 0.0, | |
| "loss": 1.3032, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 9.473684210526315, | |
| "eval_loss": 2.5678367614746094, | |
| "eval_runtime": 0.5511, | |
| "eval_samples_per_second": 18.145, | |
| "eval_steps_per_second": 1.814, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 9.473684210526315, | |
| "step": 90, | |
| "total_flos": 5.286360054444851e+16, | |
| "train_loss": 1.6763220760557387, | |
| "train_runtime": 490.2927, | |
| "train_samples_per_second": 8.933, | |
| "train_steps_per_second": 0.184 | |
| } | |
| ], | |
| "logging_steps": 5, | |
| "max_steps": 90, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 10, | |
| "save_steps": 100, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": false, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 5.286360054444851e+16, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |