| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 34.285714285714285, | |
| "eval_steps": 10, | |
| "global_step": 60, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 1.0, | |
| "learning_rate": 0, | |
| "loss": 1.3796, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 1.1428571428571428, | |
| "grad_norm": 2.2001447677612305, | |
| "learning_rate": 0.0, | |
| "loss": 0.2105, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 2.2001447677612305, | |
| "learning_rate": 0.0, | |
| "loss": 0.962, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 2.2857142857142856, | |
| "grad_norm": 0.8828794360160828, | |
| "learning_rate": 0.0002, | |
| "loss": 0.2299, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 0.8828794360160828, | |
| "learning_rate": 0.0002, | |
| "loss": 0.8788, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 3.4285714285714284, | |
| "grad_norm": 0.7549806237220764, | |
| "learning_rate": 0.0002, | |
| "loss": 0.5161, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 0.7549806237220764, | |
| "learning_rate": 0.0002, | |
| "loss": 0.5018, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 4.571428571428571, | |
| "grad_norm": 0.5643095970153809, | |
| "learning_rate": 0.0002, | |
| "loss": 0.5318, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 0.5643095970153809, | |
| "learning_rate": 0.0002, | |
| "loss": 0.3574, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 5.714285714285714, | |
| "grad_norm": 0.6527697443962097, | |
| "learning_rate": 0.0002, | |
| "loss": 0.6326, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 5.714285714285714, | |
| "eval_loss": 0.7977780103683472, | |
| "eval_runtime": 1.1094, | |
| "eval_samples_per_second": 6.31, | |
| "eval_steps_per_second": 6.31, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "grad_norm": 0.6527697443962097, | |
| "learning_rate": 0.0002, | |
| "loss": 0.1602, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 6.857142857142857, | |
| "grad_norm": 0.6531093716621399, | |
| "learning_rate": 0.0002, | |
| "loss": 0.5492, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "grad_norm": 0.6531093716621399, | |
| "learning_rate": 0.0002, | |
| "loss": 0.1361, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "grad_norm": 0.5385488867759705, | |
| "learning_rate": 0.0002, | |
| "loss": 0.5893, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "grad_norm": 0.5385488867759705, | |
| "learning_rate": 0.0002, | |
| "loss": 0.475, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 9.142857142857142, | |
| "grad_norm": 0.5094980001449585, | |
| "learning_rate": 0.0002, | |
| "loss": 0.0501, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "grad_norm": 0.5094980001449585, | |
| "learning_rate": 0.0002, | |
| "loss": 0.3318, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 10.285714285714286, | |
| "grad_norm": 0.5739837884902954, | |
| "learning_rate": 0.0002, | |
| "loss": 0.154, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 11.0, | |
| "grad_norm": 0.5739837884902954, | |
| "learning_rate": 0.0002, | |
| "loss": 0.1339, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 11.428571428571429, | |
| "grad_norm": 0.5782333016395569, | |
| "learning_rate": 0.0002, | |
| "loss": 0.1008, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 11.428571428571429, | |
| "eval_loss": 0.1741267442703247, | |
| "eval_runtime": 1.0921, | |
| "eval_samples_per_second": 6.41, | |
| "eval_steps_per_second": 6.41, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "grad_norm": 0.5782333016395569, | |
| "learning_rate": 0.0002, | |
| "loss": 0.1043, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 12.571428571428571, | |
| "grad_norm": 0.558684229850769, | |
| "learning_rate": 0.0002, | |
| "loss": 0.1308, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 13.0, | |
| "grad_norm": 0.558684229850769, | |
| "learning_rate": 0.0002, | |
| "loss": 0.0082, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 13.714285714285714, | |
| "grad_norm": 0.3258131444454193, | |
| "learning_rate": 0.0002, | |
| "loss": 0.0663, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 14.0, | |
| "grad_norm": 0.3258131444454193, | |
| "learning_rate": 0.0002, | |
| "loss": 0.0114, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 14.857142857142858, | |
| "grad_norm": 0.3965975046157837, | |
| "learning_rate": 0.0002, | |
| "loss": 0.0503, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 15.0, | |
| "grad_norm": 0.3965975046157837, | |
| "learning_rate": 0.0002, | |
| "loss": 0.0026, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 16.0, | |
| "grad_norm": 0.6329053044319153, | |
| "learning_rate": 0.0002, | |
| "loss": 0.0271, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 17.0, | |
| "grad_norm": 0.6329053044319153, | |
| "learning_rate": 0.0002, | |
| "loss": 0.0145, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 17.142857142857142, | |
| "grad_norm": 0.1660483181476593, | |
| "learning_rate": 0.0002, | |
| "loss": 0.0003, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 17.142857142857142, | |
| "eval_loss": 0.008966931141912937, | |
| "eval_runtime": 1.1273, | |
| "eval_samples_per_second": 6.21, | |
| "eval_steps_per_second": 6.21, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 18.0, | |
| "grad_norm": 0.1660483181476593, | |
| "learning_rate": 0.0002, | |
| "loss": 0.0076, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 18.285714285714285, | |
| "grad_norm": 0.16035296022891998, | |
| "learning_rate": 0.0002, | |
| "loss": 0.0006, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 19.0, | |
| "grad_norm": 0.16035296022891998, | |
| "learning_rate": 0.0002, | |
| "loss": 0.0027, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 19.428571428571427, | |
| "grad_norm": 0.22989660501480103, | |
| "learning_rate": 0.0002, | |
| "loss": 0.0022, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "grad_norm": 0.22989660501480103, | |
| "learning_rate": 0.0002, | |
| "loss": 0.0006, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 20.571428571428573, | |
| "grad_norm": 0.025392575189471245, | |
| "learning_rate": 0.0002, | |
| "loss": 0.0004, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 21.0, | |
| "grad_norm": 0.025392575189471245, | |
| "learning_rate": 0.0002, | |
| "loss": 0.0037, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 21.714285714285715, | |
| "grad_norm": 0.7075601816177368, | |
| "learning_rate": 0.0002, | |
| "loss": 0.0039, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 22.0, | |
| "grad_norm": 0.7075601816177368, | |
| "learning_rate": 0.0002, | |
| "loss": 0.0002, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 22.857142857142858, | |
| "grad_norm": 0.057493291795253754, | |
| "learning_rate": 0.0002, | |
| "loss": 0.0009, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 22.857142857142858, | |
| "eval_loss": 0.001083305454812944, | |
| "eval_runtime": 1.1213, | |
| "eval_samples_per_second": 6.243, | |
| "eval_steps_per_second": 6.243, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 23.0, | |
| "grad_norm": 0.057493291795253754, | |
| "learning_rate": 0.0002, | |
| "loss": 0.0001, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 24.0, | |
| "grad_norm": 0.043593935668468475, | |
| "learning_rate": 0.0002, | |
| "loss": 0.001, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 25.0, | |
| "grad_norm": 0.043593935668468475, | |
| "learning_rate": 0.0002, | |
| "loss": 0.0017, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 25.142857142857142, | |
| "grad_norm": 0.1038336306810379, | |
| "learning_rate": 0.0002, | |
| "loss": 0.0001, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 26.0, | |
| "grad_norm": 0.1038336306810379, | |
| "learning_rate": 0.0002, | |
| "loss": 0.0013, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 26.285714285714285, | |
| "grad_norm": 0.1458902508020401, | |
| "learning_rate": 0.0002, | |
| "loss": 0.0011, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 27.0, | |
| "grad_norm": 0.1458902508020401, | |
| "learning_rate": 0.0002, | |
| "loss": 0.0003, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 27.428571428571427, | |
| "grad_norm": 0.014756185002624989, | |
| "learning_rate": 0.0002, | |
| "loss": 0.0001, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 28.0, | |
| "grad_norm": 0.014756185002624989, | |
| "learning_rate": 0.0002, | |
| "loss": 0.0005, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 28.571428571428573, | |
| "grad_norm": 0.032949432730674744, | |
| "learning_rate": 0.0002, | |
| "loss": 0.0004, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 28.571428571428573, | |
| "eval_loss": 0.0005236685974523425, | |
| "eval_runtime": 1.1017, | |
| "eval_samples_per_second": 6.354, | |
| "eval_steps_per_second": 6.354, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 29.0, | |
| "grad_norm": 0.032949432730674744, | |
| "learning_rate": 0.0002, | |
| "loss": 0.0001, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 29.714285714285715, | |
| "grad_norm": 0.015912260860204697, | |
| "learning_rate": 0.0002, | |
| "loss": 0.0004, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 30.0, | |
| "grad_norm": 0.015912260860204697, | |
| "learning_rate": 0.0002, | |
| "loss": 0.0001, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 30.857142857142858, | |
| "grad_norm": 0.012292955070734024, | |
| "learning_rate": 0.0002, | |
| "loss": 0.0004, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 31.0, | |
| "grad_norm": 0.012292955070734024, | |
| "learning_rate": 0.0002, | |
| "loss": 0.0, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 32.0, | |
| "grad_norm": 0.009588208049535751, | |
| "learning_rate": 0.0002, | |
| "loss": 0.0003, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 33.0, | |
| "grad_norm": 0.009588208049535751, | |
| "learning_rate": 0.0002, | |
| "loss": 0.0003, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 33.142857142857146, | |
| "grad_norm": 0.007272060494869947, | |
| "learning_rate": 0.0002, | |
| "loss": 0.0, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 34.0, | |
| "grad_norm": 0.007272060494869947, | |
| "learning_rate": 0.0002, | |
| "loss": 0.0002, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 34.285714285714285, | |
| "grad_norm": 0.006634233985096216, | |
| "learning_rate": 0.0002, | |
| "loss": 0.0, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 34.285714285714285, | |
| "eval_loss": 0.00021448711049742997, | |
| "eval_runtime": 1.0979, | |
| "eval_samples_per_second": 6.376, | |
| "eval_steps_per_second": 6.376, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 34.285714285714285, | |
| "step": 60, | |
| "total_flos": 4016214881861632.0, | |
| "train_loss": 0.1554629008491247, | |
| "train_runtime": 98.8483, | |
| "train_samples_per_second": 4.856, | |
| "train_steps_per_second": 0.607 | |
| } | |
| ], | |
| "logging_steps": 1.0, | |
| "max_steps": 60, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 60, | |
| "save_steps": 1000, | |
| "total_flos": 4016214881861632.0, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |