| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 10.0, | |
| "eval_steps": 500, | |
| "global_step": 130, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.07692307692307693, | |
| "grad_norm": 480.0, | |
| "learning_rate": 1.5384615384615387e-05, | |
| "loss": 49.667, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.38461538461538464, | |
| "grad_norm": 260.0, | |
| "learning_rate": 7.692307692307693e-05, | |
| "loss": 40.4711, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.7692307692307693, | |
| "grad_norm": 26.75, | |
| "learning_rate": 0.00015384615384615385, | |
| "loss": 23.4809, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_loss": 8.250589370727539, | |
| "eval_runtime": 2.0471, | |
| "eval_samples_per_second": 4.885, | |
| "eval_steps_per_second": 0.977, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 1.1538461538461537, | |
| "grad_norm": 15.25, | |
| "learning_rate": 0.00019985583705641418, | |
| "loss": 19.4813, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 1.5384615384615383, | |
| "grad_norm": 34.75, | |
| "learning_rate": 0.00019823877374156647, | |
| "loss": 16.6271, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 1.9230769230769231, | |
| "grad_norm": 21.875, | |
| "learning_rate": 0.00019485364419471454, | |
| "loss": 7.0034, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_loss": 3.5644798278808594, | |
| "eval_runtime": 2.0643, | |
| "eval_samples_per_second": 4.844, | |
| "eval_steps_per_second": 0.969, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 2.3076923076923075, | |
| "grad_norm": 5.5625, | |
| "learning_rate": 0.0001897613727639014, | |
| "loss": 2.3031, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 2.6923076923076925, | |
| "grad_norm": 3.09375, | |
| "learning_rate": 0.00018305360832480117, | |
| "loss": 1.9278, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_loss": 2.7830893993377686, | |
| "eval_runtime": 2.0487, | |
| "eval_samples_per_second": 4.881, | |
| "eval_steps_per_second": 0.976, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 3.076923076923077, | |
| "grad_norm": 1.8671875, | |
| "learning_rate": 0.00017485107481711012, | |
| "loss": 1.6366, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 3.4615384615384617, | |
| "grad_norm": 5.84375, | |
| "learning_rate": 0.0001653013984983585, | |
| "loss": 1.4332, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 3.8461538461538463, | |
| "grad_norm": 1.5703125, | |
| "learning_rate": 0.00015457645101945046, | |
| "loss": 1.3284, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_loss": 2.7295384407043457, | |
| "eval_runtime": 2.0468, | |
| "eval_samples_per_second": 4.886, | |
| "eval_steps_per_second": 0.977, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 4.230769230769231, | |
| "grad_norm": 2.0625, | |
| "learning_rate": 0.00014286925614030542, | |
| "loss": 1.1714, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 4.615384615384615, | |
| "grad_norm": 1.7578125, | |
| "learning_rate": 0.0001303905157574247, | |
| "loss": 1.0989, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 2.921875, | |
| "learning_rate": 0.00011736481776669306, | |
| "loss": 1.046, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_loss": 2.7646915912628174, | |
| "eval_runtime": 2.047, | |
| "eval_samples_per_second": 4.885, | |
| "eval_steps_per_second": 0.977, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 5.384615384615385, | |
| "grad_norm": 2.65625, | |
| "learning_rate": 0.00010402659401094152, | |
| "loss": 0.9141, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 5.769230769230769, | |
| "grad_norm": 1.7734375, | |
| "learning_rate": 9.061590105968208e-05, | |
| "loss": 0.8611, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_loss": 2.862928867340088, | |
| "eval_runtime": 2.0464, | |
| "eval_samples_per_second": 4.887, | |
| "eval_steps_per_second": 0.977, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 6.153846153846154, | |
| "grad_norm": 11.625, | |
| "learning_rate": 7.73740997570278e-05, | |
| "loss": 0.7981, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 6.538461538461538, | |
| "grad_norm": 2.09375, | |
| "learning_rate": 6.453951129574644e-05, | |
| "loss": 0.7228, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 6.923076923076923, | |
| "grad_norm": 2.078125, | |
| "learning_rate": 5.234312799786921e-05, | |
| "loss": 0.7242, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_loss": 3.040492296218872, | |
| "eval_runtime": 2.0484, | |
| "eval_samples_per_second": 4.882, | |
| "eval_steps_per_second": 0.976, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 7.3076923076923075, | |
| "grad_norm": 1.421875, | |
| "learning_rate": 4.100445599768774e-05, | |
| "loss": 0.6348, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 7.6923076923076925, | |
| "grad_norm": 2.25, | |
| "learning_rate": 3.072756464904006e-05, | |
| "loss": 0.6192, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_loss": 3.1549479961395264, | |
| "eval_runtime": 2.0433, | |
| "eval_samples_per_second": 4.894, | |
| "eval_steps_per_second": 0.979, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 8.076923076923077, | |
| "grad_norm": 1.3359375, | |
| "learning_rate": 2.1697413758237784e-05, | |
| "loss": 0.5742, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 8.461538461538462, | |
| "grad_norm": 1.5703125, | |
| "learning_rate": 1.4076524743778319e-05, | |
| "loss": 0.5514, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 8.846153846153847, | |
| "grad_norm": 1.40625, | |
| "learning_rate": 8.002055634117578e-06, | |
| "loss": 0.5312, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "eval_loss": 3.3040664196014404, | |
| "eval_runtime": 2.0464, | |
| "eval_samples_per_second": 4.887, | |
| "eval_steps_per_second": 0.977, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 9.23076923076923, | |
| "grad_norm": 1.265625, | |
| "learning_rate": 3.5833325466437694e-06, | |
| "loss": 0.5132, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 9.615384615384615, | |
| "grad_norm": 1.3984375, | |
| "learning_rate": 8.998820754091531e-07, | |
| "loss": 0.4966, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "grad_norm": 1.2109375, | |
| "learning_rate": 0.0, | |
| "loss": 0.5202, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "eval_loss": 3.298276901245117, | |
| "eval_runtime": 2.0482, | |
| "eval_samples_per_second": 4.882, | |
| "eval_steps_per_second": 0.976, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "step": 130, | |
| "total_flos": 9.937675599478784e+16, | |
| "train_loss": 4.9734316147290745, | |
| "train_runtime": 870.6795, | |
| "train_samples_per_second": 2.366, | |
| "train_steps_per_second": 0.149 | |
| } | |
| ], | |
| "logging_steps": 5, | |
| "max_steps": 130, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 10, | |
| "save_steps": 100, | |
| "total_flos": 9.937675599478784e+16, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |