| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 2.9895227008149012, | |
| "eval_steps": 500, | |
| "global_step": 321, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.09313154831199069, | |
| "grad_norm": 6.410240671476086, | |
| "learning_rate": 5e-06, | |
| "loss": 1.0737, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.18626309662398138, | |
| "grad_norm": 2.110670339648224, | |
| "learning_rate": 5e-06, | |
| "loss": 0.853, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.27939464493597205, | |
| "grad_norm": 2.0454490264727103, | |
| "learning_rate": 5e-06, | |
| "loss": 0.808, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.37252619324796277, | |
| "grad_norm": 1.9173512178087246, | |
| "learning_rate": 5e-06, | |
| "loss": 0.7705, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.46565774155995343, | |
| "grad_norm": 1.9696383768635166, | |
| "learning_rate": 5e-06, | |
| "loss": 0.7493, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.5587892898719441, | |
| "grad_norm": 1.7929306896649821, | |
| "learning_rate": 5e-06, | |
| "loss": 0.723, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.6519208381839348, | |
| "grad_norm": 1.4541049068033436, | |
| "learning_rate": 5e-06, | |
| "loss": 0.7029, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.7450523864959255, | |
| "grad_norm": 1.4546232801341403, | |
| "learning_rate": 5e-06, | |
| "loss": 0.7045, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.8381839348079162, | |
| "grad_norm": 0.9836741916976506, | |
| "learning_rate": 5e-06, | |
| "loss": 0.6913, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.9313154831199069, | |
| "grad_norm": 0.9546295476288479, | |
| "learning_rate": 5e-06, | |
| "loss": 0.6853, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.9965075669383003, | |
| "eval_loss": 0.6684155464172363, | |
| "eval_runtime": 72.6852, | |
| "eval_samples_per_second": 39.788, | |
| "eval_steps_per_second": 0.633, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 1.0244470314318976, | |
| "grad_norm": 1.2182872648581187, | |
| "learning_rate": 5e-06, | |
| "loss": 0.7062, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 1.1175785797438882, | |
| "grad_norm": 1.4541964745779468, | |
| "learning_rate": 5e-06, | |
| "loss": 0.6098, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 1.210710128055879, | |
| "grad_norm": 0.9495507104405151, | |
| "learning_rate": 5e-06, | |
| "loss": 0.6019, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 1.3038416763678695, | |
| "grad_norm": 1.8134473930318298, | |
| "learning_rate": 5e-06, | |
| "loss": 0.6043, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 1.3969732246798603, | |
| "grad_norm": 1.113009214401007, | |
| "learning_rate": 5e-06, | |
| "loss": 0.6023, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 1.490104772991851, | |
| "grad_norm": 2.029132520416773, | |
| "learning_rate": 5e-06, | |
| "loss": 0.6021, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 1.5832363213038416, | |
| "grad_norm": 1.6144281176922932, | |
| "learning_rate": 5e-06, | |
| "loss": 0.6002, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 1.6763678696158324, | |
| "grad_norm": 1.2068997504682133, | |
| "learning_rate": 5e-06, | |
| "loss": 0.5935, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 1.7694994179278232, | |
| "grad_norm": 1.001966561421665, | |
| "learning_rate": 5e-06, | |
| "loss": 0.589, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 1.8626309662398137, | |
| "grad_norm": 0.7426199844782707, | |
| "learning_rate": 5e-06, | |
| "loss": 0.5954, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 1.9557625145518043, | |
| "grad_norm": 1.0472743759335235, | |
| "learning_rate": 5e-06, | |
| "loss": 0.591, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 1.9930151338766007, | |
| "eval_loss": 0.6565061807632446, | |
| "eval_runtime": 76.8161, | |
| "eval_samples_per_second": 37.648, | |
| "eval_steps_per_second": 0.599, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 2.0488940628637953, | |
| "grad_norm": 1.1047685289937579, | |
| "learning_rate": 5e-06, | |
| "loss": 0.5986, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 2.142025611175786, | |
| "grad_norm": 1.214031879234503, | |
| "learning_rate": 5e-06, | |
| "loss": 0.5179, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 2.2351571594877764, | |
| "grad_norm": 0.8187191061599544, | |
| "learning_rate": 5e-06, | |
| "loss": 0.5178, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 2.328288707799767, | |
| "grad_norm": 0.8726515587132341, | |
| "learning_rate": 5e-06, | |
| "loss": 0.513, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 2.421420256111758, | |
| "grad_norm": 0.8422751030451243, | |
| "learning_rate": 5e-06, | |
| "loss": 0.5137, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 2.5145518044237485, | |
| "grad_norm": 0.9018678534732393, | |
| "learning_rate": 5e-06, | |
| "loss": 0.5263, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 2.607683352735739, | |
| "grad_norm": 0.875500810626841, | |
| "learning_rate": 5e-06, | |
| "loss": 0.5201, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 2.70081490104773, | |
| "grad_norm": 0.8624139031471002, | |
| "learning_rate": 5e-06, | |
| "loss": 0.5181, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 2.7939464493597206, | |
| "grad_norm": 0.8654410079756801, | |
| "learning_rate": 5e-06, | |
| "loss": 0.5197, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 2.887077997671711, | |
| "grad_norm": 1.255098018906923, | |
| "learning_rate": 5e-06, | |
| "loss": 0.5274, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 2.980209545983702, | |
| "grad_norm": 0.8384998281941215, | |
| "learning_rate": 5e-06, | |
| "loss": 0.5134, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 2.9895227008149012, | |
| "eval_loss": 0.6689090132713318, | |
| "eval_runtime": 75.5704, | |
| "eval_samples_per_second": 38.269, | |
| "eval_steps_per_second": 0.609, | |
| "step": 321 | |
| }, | |
| { | |
| "epoch": 2.9895227008149012, | |
| "step": 321, | |
| "total_flos": 537477576130560.0, | |
| "train_loss": 0.6323956919607715, | |
| "train_runtime": 10362.5439, | |
| "train_samples_per_second": 15.907, | |
| "train_steps_per_second": 0.031 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 321, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 537477576130560.0, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |