| { | |
| "best_metric": 0.08551913499832153, | |
| "best_model_checkpoint": "./fine-tuned/checkpoint-3000", | |
| "epoch": 0.48, | |
| "eval_steps": 500, | |
| "global_step": 3000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.008, | |
| "grad_norm": 14499.107421875, | |
| "learning_rate": 2.9919999999999998e-05, | |
| "loss": 0.3351, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.016, | |
| "grad_norm": 9562.1748046875, | |
| "learning_rate": 2.9840000000000002e-05, | |
| "loss": 0.0964, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.024, | |
| "grad_norm": 11098.59375, | |
| "learning_rate": 2.976e-05, | |
| "loss": 0.0895, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.032, | |
| "grad_norm": 9281.0146484375, | |
| "learning_rate": 2.968e-05, | |
| "loss": 0.0797, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 10050.3623046875, | |
| "learning_rate": 2.96e-05, | |
| "loss": 0.0812, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.048, | |
| "grad_norm": 7611.0849609375, | |
| "learning_rate": 2.9520000000000002e-05, | |
| "loss": 0.0755, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.056, | |
| "grad_norm": 9915.1259765625, | |
| "learning_rate": 2.944e-05, | |
| "loss": 0.0793, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.064, | |
| "grad_norm": 10182.263671875, | |
| "learning_rate": 2.936e-05, | |
| "loss": 0.0775, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.072, | |
| "grad_norm": 11287.8271484375, | |
| "learning_rate": 2.928e-05, | |
| "loss": 0.0782, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 6672.08251953125, | |
| "learning_rate": 2.92e-05, | |
| "loss": 0.0811, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "eval_loss": 0.09235642850399017, | |
| "eval_runtime": 109.274, | |
| "eval_samples_per_second": 18.303, | |
| "eval_steps_per_second": 2.288, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.088, | |
| "grad_norm": 6587.6513671875, | |
| "learning_rate": 2.9120000000000002e-05, | |
| "loss": 0.0815, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.096, | |
| "grad_norm": 6632.0947265625, | |
| "learning_rate": 2.904e-05, | |
| "loss": 0.0794, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.104, | |
| "grad_norm": 9301.228515625, | |
| "learning_rate": 2.896e-05, | |
| "loss": 0.076, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.112, | |
| "grad_norm": 10575.0791015625, | |
| "learning_rate": 2.888e-05, | |
| "loss": 0.0791, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "grad_norm": 8609.86328125, | |
| "learning_rate": 2.88e-05, | |
| "loss": 0.0799, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.128, | |
| "grad_norm": 11379.4423828125, | |
| "learning_rate": 2.8720000000000003e-05, | |
| "loss": 0.0759, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.136, | |
| "grad_norm": 8489.6904296875, | |
| "learning_rate": 2.864e-05, | |
| "loss": 0.0753, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 0.144, | |
| "grad_norm": 12353.6279296875, | |
| "learning_rate": 2.856e-05, | |
| "loss": 0.075, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.152, | |
| "grad_norm": 11535.3994140625, | |
| "learning_rate": 2.8480000000000002e-05, | |
| "loss": 0.0757, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 8291.939453125, | |
| "learning_rate": 2.84e-05, | |
| "loss": 0.0753, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "eval_loss": 0.08949962258338928, | |
| "eval_runtime": 109.2536, | |
| "eval_samples_per_second": 18.306, | |
| "eval_steps_per_second": 2.288, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.168, | |
| "grad_norm": 8266.658203125, | |
| "learning_rate": 2.832e-05, | |
| "loss": 0.0767, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 0.176, | |
| "grad_norm": 6160.548828125, | |
| "learning_rate": 2.824e-05, | |
| "loss": 0.067, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 0.184, | |
| "grad_norm": 7343.408203125, | |
| "learning_rate": 2.816e-05, | |
| "loss": 0.0717, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 0.192, | |
| "grad_norm": 5661.76318359375, | |
| "learning_rate": 2.8080000000000002e-05, | |
| "loss": 0.0733, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "grad_norm": 8678.46484375, | |
| "learning_rate": 2.8e-05, | |
| "loss": 0.0737, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 0.208, | |
| "grad_norm": 6331.21533203125, | |
| "learning_rate": 2.792e-05, | |
| "loss": 0.0696, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 0.216, | |
| "grad_norm": 10563.5400390625, | |
| "learning_rate": 2.784e-05, | |
| "loss": 0.0747, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 0.224, | |
| "grad_norm": 7221.74365234375, | |
| "learning_rate": 2.7760000000000002e-05, | |
| "loss": 0.0716, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 0.232, | |
| "grad_norm": 6486.46142578125, | |
| "learning_rate": 2.768e-05, | |
| "loss": 0.0711, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "grad_norm": 6838.505859375, | |
| "learning_rate": 2.7600000000000003e-05, | |
| "loss": 0.0703, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "eval_loss": 0.08808805048465729, | |
| "eval_runtime": 109.2355, | |
| "eval_samples_per_second": 18.309, | |
| "eval_steps_per_second": 2.289, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.248, | |
| "grad_norm": 6751.6494140625, | |
| "learning_rate": 2.752e-05, | |
| "loss": 0.0781, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 0.256, | |
| "grad_norm": 5040.9033203125, | |
| "learning_rate": 2.7439999999999998e-05, | |
| "loss": 0.0686, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 0.264, | |
| "grad_norm": 8748.07421875, | |
| "learning_rate": 2.7360000000000002e-05, | |
| "loss": 0.0689, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 0.272, | |
| "grad_norm": 5971.705078125, | |
| "learning_rate": 2.728e-05, | |
| "loss": 0.0671, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "grad_norm": 10833.1357421875, | |
| "learning_rate": 2.72e-05, | |
| "loss": 0.0734, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 0.288, | |
| "grad_norm": 10036.919921875, | |
| "learning_rate": 2.712e-05, | |
| "loss": 0.0715, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 0.296, | |
| "grad_norm": 7755.1669921875, | |
| "learning_rate": 2.704e-05, | |
| "loss": 0.0669, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 0.304, | |
| "grad_norm": 7584.822265625, | |
| "learning_rate": 2.696e-05, | |
| "loss": 0.0699, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 0.312, | |
| "grad_norm": 10103.142578125, | |
| "learning_rate": 2.688e-05, | |
| "loss": 0.07, | |
| "step": 1950 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "grad_norm": 5768.24267578125, | |
| "learning_rate": 2.68e-05, | |
| "loss": 0.0709, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "eval_loss": 0.08704760670661926, | |
| "eval_runtime": 109.4348, | |
| "eval_samples_per_second": 18.276, | |
| "eval_steps_per_second": 2.284, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.328, | |
| "grad_norm": 6016.46826171875, | |
| "learning_rate": 2.672e-05, | |
| "loss": 0.0663, | |
| "step": 2050 | |
| }, | |
| { | |
| "epoch": 0.336, | |
| "grad_norm": 6869.53076171875, | |
| "learning_rate": 2.6640000000000002e-05, | |
| "loss": 0.073, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 0.344, | |
| "grad_norm": 6099.595703125, | |
| "learning_rate": 2.656e-05, | |
| "loss": 0.0667, | |
| "step": 2150 | |
| }, | |
| { | |
| "epoch": 0.352, | |
| "grad_norm": 6923.919921875, | |
| "learning_rate": 2.648e-05, | |
| "loss": 0.0653, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 0.36, | |
| "grad_norm": 8005.85595703125, | |
| "learning_rate": 2.64e-05, | |
| "loss": 0.0685, | |
| "step": 2250 | |
| }, | |
| { | |
| "epoch": 0.368, | |
| "grad_norm": 6473.466796875, | |
| "learning_rate": 2.632e-05, | |
| "loss": 0.0678, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 0.376, | |
| "grad_norm": 7177.6328125, | |
| "learning_rate": 2.6240000000000003e-05, | |
| "loss": 0.0637, | |
| "step": 2350 | |
| }, | |
| { | |
| "epoch": 0.384, | |
| "grad_norm": 5574.75439453125, | |
| "learning_rate": 2.616e-05, | |
| "loss": 0.0698, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 0.392, | |
| "grad_norm": 6910.39599609375, | |
| "learning_rate": 2.608e-05, | |
| "loss": 0.0645, | |
| "step": 2450 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "grad_norm": 5913.9775390625, | |
| "learning_rate": 2.6000000000000002e-05, | |
| "loss": 0.068, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "eval_loss": 0.08615937829017639, | |
| "eval_runtime": 109.2621, | |
| "eval_samples_per_second": 18.305, | |
| "eval_steps_per_second": 2.288, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 0.408, | |
| "grad_norm": 7447.5625, | |
| "learning_rate": 2.592e-05, | |
| "loss": 0.0672, | |
| "step": 2550 | |
| }, | |
| { | |
| "epoch": 0.416, | |
| "grad_norm": 7057.10009765625, | |
| "learning_rate": 2.584e-05, | |
| "loss": 0.0683, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 0.424, | |
| "grad_norm": 8279.7392578125, | |
| "learning_rate": 2.576e-05, | |
| "loss": 0.0631, | |
| "step": 2650 | |
| }, | |
| { | |
| "epoch": 0.432, | |
| "grad_norm": 7663.275390625, | |
| "learning_rate": 2.568e-05, | |
| "loss": 0.0698, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 0.44, | |
| "grad_norm": 7116.74609375, | |
| "learning_rate": 2.5600000000000002e-05, | |
| "loss": 0.0703, | |
| "step": 2750 | |
| }, | |
| { | |
| "epoch": 0.448, | |
| "grad_norm": 8839.5986328125, | |
| "learning_rate": 2.552e-05, | |
| "loss": 0.0654, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 0.456, | |
| "grad_norm": 7157.17333984375, | |
| "learning_rate": 2.544e-05, | |
| "loss": 0.0628, | |
| "step": 2850 | |
| }, | |
| { | |
| "epoch": 0.464, | |
| "grad_norm": 7690.267578125, | |
| "learning_rate": 2.536e-05, | |
| "loss": 0.0694, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 0.472, | |
| "grad_norm": 5030.39501953125, | |
| "learning_rate": 2.5280000000000002e-05, | |
| "loss": 0.0654, | |
| "step": 2950 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "grad_norm": 7269.51171875, | |
| "learning_rate": 2.52e-05, | |
| "loss": 0.0732, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "eval_loss": 0.08551913499832153, | |
| "eval_runtime": 109.2626, | |
| "eval_samples_per_second": 18.305, | |
| "eval_steps_per_second": 2.288, | |
| "step": 3000 | |
| } | |
| ], | |
| "logging_steps": 50, | |
| "max_steps": 18750, | |
| "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": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 1.461498937344e+16, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |