| { |
| "best_metric": 0.8588420748710632, |
| "best_model_checkpoint": "tiny_bert_rand_50_v1_mnli/checkpoint-7670", |
| "epoch": 10.0, |
| "eval_steps": 500, |
| "global_step": 15340, |
| "is_hyper_param_search": false, |
| "is_local_process_zero": true, |
| "is_world_process_zero": true, |
| "log_history": [ |
| { |
| "epoch": 1.0, |
| "grad_norm": 0.98807293176651, |
| "learning_rate": 4.9e-05, |
| "loss": 0.9908, |
| "step": 1534 |
| }, |
| { |
| "epoch": 1.0, |
| "eval_accuracy": 0.5470198675496689, |
| "eval_loss": 0.9352031350135803, |
| "eval_runtime": 3.4922, |
| "eval_samples_per_second": 2810.576, |
| "eval_steps_per_second": 11.168, |
| "step": 1534 |
| }, |
| { |
| "epoch": 2.0, |
| "grad_norm": 1.4757229089736938, |
| "learning_rate": 4.8e-05, |
| "loss": 0.9036, |
| "step": 3068 |
| }, |
| { |
| "epoch": 2.0, |
| "eval_accuracy": 0.587366276107998, |
| "eval_loss": 0.8885725736618042, |
| "eval_runtime": 3.4612, |
| "eval_samples_per_second": 2835.702, |
| "eval_steps_per_second": 11.268, |
| "step": 3068 |
| }, |
| { |
| "epoch": 3.0, |
| "grad_norm": 1.3523396253585815, |
| "learning_rate": 4.7e-05, |
| "loss": 0.8527, |
| "step": 4602 |
| }, |
| { |
| "epoch": 3.0, |
| "eval_accuracy": 0.6013245033112583, |
| "eval_loss": 0.8672203421592712, |
| "eval_runtime": 3.4595, |
| "eval_samples_per_second": 2837.112, |
| "eval_steps_per_second": 11.273, |
| "step": 4602 |
| }, |
| { |
| "epoch": 4.0, |
| "grad_norm": 1.7900631427764893, |
| "learning_rate": 4.600000000000001e-05, |
| "loss": 0.8101, |
| "step": 6136 |
| }, |
| { |
| "epoch": 4.0, |
| "eval_accuracy": 0.6128374936321956, |
| "eval_loss": 0.8600946664810181, |
| "eval_runtime": 3.4549, |
| "eval_samples_per_second": 2840.867, |
| "eval_steps_per_second": 11.288, |
| "step": 6136 |
| }, |
| { |
| "epoch": 5.0, |
| "grad_norm": 1.7126315832138062, |
| "learning_rate": 4.5e-05, |
| "loss": 0.7693, |
| "step": 7670 |
| }, |
| { |
| "epoch": 5.0, |
| "eval_accuracy": 0.6097809475292919, |
| "eval_loss": 0.8588420748710632, |
| "eval_runtime": 3.4464, |
| "eval_samples_per_second": 2847.91, |
| "eval_steps_per_second": 11.316, |
| "step": 7670 |
| }, |
| { |
| "epoch": 6.0, |
| "grad_norm": 2.388108730316162, |
| "learning_rate": 4.4000000000000006e-05, |
| "loss": 0.7296, |
| "step": 9204 |
| }, |
| { |
| "epoch": 6.0, |
| "eval_accuracy": 0.621395822720326, |
| "eval_loss": 0.8736900091171265, |
| "eval_runtime": 3.48, |
| "eval_samples_per_second": 2820.376, |
| "eval_steps_per_second": 11.207, |
| "step": 9204 |
| }, |
| { |
| "epoch": 7.0, |
| "grad_norm": 1.9555968046188354, |
| "learning_rate": 4.3e-05, |
| "loss": 0.6889, |
| "step": 10738 |
| }, |
| { |
| "epoch": 7.0, |
| "eval_accuracy": 0.6221090168110036, |
| "eval_loss": 0.8842542171478271, |
| "eval_runtime": 3.6137, |
| "eval_samples_per_second": 2716.024, |
| "eval_steps_per_second": 10.792, |
| "step": 10738 |
| }, |
| { |
| "epoch": 8.0, |
| "grad_norm": 2.0965774059295654, |
| "learning_rate": 4.2e-05, |
| "loss": 0.6504, |
| "step": 12272 |
| }, |
| { |
| "epoch": 8.0, |
| "eval_accuracy": 0.6101884870096791, |
| "eval_loss": 0.9531626105308533, |
| "eval_runtime": 3.4632, |
| "eval_samples_per_second": 2834.092, |
| "eval_steps_per_second": 11.261, |
| "step": 12272 |
| }, |
| { |
| "epoch": 9.0, |
| "grad_norm": 2.935882806777954, |
| "learning_rate": 4.1e-05, |
| "loss": 0.6127, |
| "step": 13806 |
| }, |
| { |
| "epoch": 9.0, |
| "eval_accuracy": 0.6200713194090678, |
| "eval_loss": 0.9572077393531799, |
| "eval_runtime": 3.4388, |
| "eval_samples_per_second": 2854.167, |
| "eval_steps_per_second": 11.341, |
| "step": 13806 |
| }, |
| { |
| "epoch": 10.0, |
| "grad_norm": 2.946934461593628, |
| "learning_rate": 4e-05, |
| "loss": 0.5741, |
| "step": 15340 |
| }, |
| { |
| "epoch": 10.0, |
| "eval_accuracy": 0.618033622007132, |
| "eval_loss": 0.9933791756629944, |
| "eval_runtime": 3.4268, |
| "eval_samples_per_second": 2864.188, |
| "eval_steps_per_second": 11.381, |
| "step": 15340 |
| }, |
| { |
| "epoch": 10.0, |
| "step": 15340, |
| "total_flos": 1.029836100834816e+17, |
| "train_loss": 0.7582102311792087, |
| "train_runtime": 2632.1731, |
| "train_samples_per_second": 7459.654, |
| "train_steps_per_second": 29.139 |
| } |
| ], |
| "logging_steps": 1, |
| "max_steps": 76700, |
| "num_input_tokens_seen": 0, |
| "num_train_epochs": 50, |
| "save_steps": 500, |
| "total_flos": 1.029836100834816e+17, |
| "train_batch_size": 256, |
| "trial_name": null, |
| "trial_params": null |
| } |
|
|
