| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 1.0, | |
| "eval_steps": 8, | |
| "global_step": 32, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 0.5859375, | |
| "learning_rate": 2e-05, | |
| "loss": 3.9295, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "eval_loss": 3.907318115234375, | |
| "eval_runtime": 1.2713, | |
| "eval_samples_per_second": 11.012, | |
| "eval_steps_per_second": 5.506, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 0.5078125, | |
| "learning_rate": 4e-05, | |
| "loss": 3.805, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 0.63671875, | |
| "learning_rate": 6e-05, | |
| "loss": 3.8521, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "grad_norm": 0.609375, | |
| "learning_rate": 8e-05, | |
| "loss": 3.8947, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 0.546875, | |
| "learning_rate": 0.0001, | |
| "loss": 3.6494, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "grad_norm": 0.50390625, | |
| "learning_rate": 0.00012, | |
| "loss": 3.6457, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.22, | |
| "grad_norm": 0.640625, | |
| "learning_rate": 0.00014, | |
| "loss": 3.967, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "grad_norm": 0.494140625, | |
| "learning_rate": 0.00016, | |
| "loss": 3.5364, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "eval_loss": 3.6198840141296387, | |
| "eval_runtime": 1.2681, | |
| "eval_samples_per_second": 11.04, | |
| "eval_steps_per_second": 5.52, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "grad_norm": 0.51171875, | |
| "learning_rate": 0.00018, | |
| "loss": 3.5216, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "grad_norm": 0.6171875, | |
| "learning_rate": 0.0002, | |
| "loss": 3.677, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.34, | |
| "grad_norm": 0.734375, | |
| "learning_rate": 0.00019996456111234527, | |
| "loss": 3.058, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "grad_norm": 0.66796875, | |
| "learning_rate": 0.0001998582695676762, | |
| "loss": 3.1333, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.41, | |
| "grad_norm": 0.8359375, | |
| "learning_rate": 0.000199681200703075, | |
| "loss": 3.38, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.44, | |
| "grad_norm": 0.62109375, | |
| "learning_rate": 0.00019943348002101371, | |
| "loss": 3.1371, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.47, | |
| "grad_norm": 0.65625, | |
| "learning_rate": 0.00019911528310040074, | |
| "loss": 3.1479, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "grad_norm": 0.6484375, | |
| "learning_rate": 0.00019872683547213446, | |
| "loss": 3.263, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "eval_loss": 3.1820719242095947, | |
| "eval_runtime": 1.2692, | |
| "eval_samples_per_second": 11.031, | |
| "eval_steps_per_second": 5.515, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.53, | |
| "grad_norm": 0.53125, | |
| "learning_rate": 0.00019826841245925212, | |
| "loss": 2.9833, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.56, | |
| "grad_norm": 0.5703125, | |
| "learning_rate": 0.00019774033898178667, | |
| "loss": 3.0787, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.59, | |
| "grad_norm": 0.71484375, | |
| "learning_rate": 0.00019714298932647098, | |
| "loss": 3.4132, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.62, | |
| "grad_norm": 0.73046875, | |
| "learning_rate": 0.0001964767868814516, | |
| "loss": 2.7304, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.66, | |
| "grad_norm": 0.55078125, | |
| "learning_rate": 0.00019574220383620055, | |
| "loss": 3.0116, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.69, | |
| "grad_norm": 0.6171875, | |
| "learning_rate": 0.00019493976084683813, | |
| "loss": 2.9474, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "grad_norm": 0.61328125, | |
| "learning_rate": 0.00019407002666710336, | |
| "loss": 2.9415, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "grad_norm": 0.6328125, | |
| "learning_rate": 0.00019313361774523385, | |
| "loss": 2.798, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "eval_loss": 2.896176815032959, | |
| "eval_runtime": 1.2765, | |
| "eval_samples_per_second": 10.967, | |
| "eval_steps_per_second": 5.484, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.78, | |
| "grad_norm": 0.8046875, | |
| "learning_rate": 0.00019213119778704128, | |
| "loss": 3.2157, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.81, | |
| "grad_norm": 0.7109375, | |
| "learning_rate": 0.00019106347728549135, | |
| "loss": 3.0666, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.84, | |
| "grad_norm": 0.59765625, | |
| "learning_rate": 0.00018993121301712193, | |
| "loss": 2.8219, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "grad_norm": 0.75390625, | |
| "learning_rate": 0.00018873520750565718, | |
| "loss": 3.1164, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.91, | |
| "grad_norm": 0.67578125, | |
| "learning_rate": 0.00018747630845319612, | |
| "loss": 2.7154, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.94, | |
| "grad_norm": 0.625, | |
| "learning_rate": 0.0001861554081393806, | |
| "loss": 2.7395, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.97, | |
| "grad_norm": 0.8125, | |
| "learning_rate": 0.0001847734427889671, | |
| "loss": 2.8282, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 0.86328125, | |
| "learning_rate": 0.0001833313919082515, | |
| "loss": 2.7787, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_loss": 2.67726731300354, | |
| "eval_runtime": 1.2769, | |
| "eval_samples_per_second": 10.964, | |
| "eval_steps_per_second": 5.482, | |
| "step": 32 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 128, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 4, | |
| "save_steps": 32, | |
| "total_flos": 6667331136651264.0, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |