| { | |
| "best_metric": 0.28984853625297546, | |
| "best_model_checkpoint": "data/Mistral-7B_task-3_180-samples_config-2_auto/checkpoint-42", | |
| "epoch": 12.0, | |
| "eval_steps": 500, | |
| "global_step": 102, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.11764705882352941, | |
| "grad_norm": 14.107588768005371, | |
| "learning_rate": 2.5e-06, | |
| "loss": 3.0371, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.23529411764705882, | |
| "grad_norm": 20.320606231689453, | |
| "learning_rate": 5e-06, | |
| "loss": 4.1482, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.47058823529411764, | |
| "grad_norm": 15.692017555236816, | |
| "learning_rate": 1e-05, | |
| "loss": 3.41, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.7058823529411765, | |
| "grad_norm": 10.95263671875, | |
| "learning_rate": 1.5e-05, | |
| "loss": 2.8065, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.9411764705882353, | |
| "grad_norm": 8.83460807800293, | |
| "learning_rate": 2e-05, | |
| "loss": 2.4492, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.9411764705882353, | |
| "eval_loss": 1.9220411777496338, | |
| "eval_runtime": 39.2353, | |
| "eval_samples_per_second": 0.918, | |
| "eval_steps_per_second": 0.918, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 1.1764705882352942, | |
| "grad_norm": 7.487240791320801, | |
| "learning_rate": 2.5e-05, | |
| "loss": 1.7411, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 1.4117647058823528, | |
| "grad_norm": 7.128180027008057, | |
| "learning_rate": 3e-05, | |
| "loss": 1.3658, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 1.6470588235294117, | |
| "grad_norm": 4.592349529266357, | |
| "learning_rate": 3.5e-05, | |
| "loss": 0.9112, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 1.8823529411764706, | |
| "grad_norm": 2.225466251373291, | |
| "learning_rate": 4e-05, | |
| "loss": 0.637, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_loss": 0.5030009150505066, | |
| "eval_runtime": 39.2353, | |
| "eval_samples_per_second": 0.918, | |
| "eval_steps_per_second": 0.918, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 2.1176470588235294, | |
| "grad_norm": 1.4838582277297974, | |
| "learning_rate": 4.5e-05, | |
| "loss": 0.5194, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 2.3529411764705883, | |
| "grad_norm": 1.0014225244522095, | |
| "learning_rate": 5e-05, | |
| "loss": 0.4796, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 2.588235294117647, | |
| "grad_norm": 0.7945687174797058, | |
| "learning_rate": 5.500000000000001e-05, | |
| "loss": 0.3369, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 2.8235294117647056, | |
| "grad_norm": 0.7629735469818115, | |
| "learning_rate": 6e-05, | |
| "loss": 0.3581, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 2.9411764705882355, | |
| "eval_loss": 0.33218392729759216, | |
| "eval_runtime": 39.245, | |
| "eval_samples_per_second": 0.917, | |
| "eval_steps_per_second": 0.917, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 3.0588235294117645, | |
| "grad_norm": 0.5562334060668945, | |
| "learning_rate": 6.500000000000001e-05, | |
| "loss": 0.2747, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 3.2941176470588234, | |
| "grad_norm": 0.7018760442733765, | |
| "learning_rate": 7e-05, | |
| "loss": 0.3232, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 3.5294117647058822, | |
| "grad_norm": 0.6312670707702637, | |
| "learning_rate": 7.500000000000001e-05, | |
| "loss": 0.2575, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 3.764705882352941, | |
| "grad_norm": 0.777540922164917, | |
| "learning_rate": 8e-05, | |
| "loss": 0.2877, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 0.6104640960693359, | |
| "learning_rate": 8.5e-05, | |
| "loss": 0.2132, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_loss": 0.29338082671165466, | |
| "eval_runtime": 39.2385, | |
| "eval_samples_per_second": 0.917, | |
| "eval_steps_per_second": 0.917, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 4.235294117647059, | |
| "grad_norm": 0.6385318636894226, | |
| "learning_rate": 9e-05, | |
| "loss": 0.2021, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 4.470588235294118, | |
| "grad_norm": 0.8091533184051514, | |
| "learning_rate": 9.5e-05, | |
| "loss": 0.2344, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 4.705882352941177, | |
| "grad_norm": 0.6098454594612122, | |
| "learning_rate": 0.0001, | |
| "loss": 0.2088, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 4.9411764705882355, | |
| "grad_norm": 0.6987226605415344, | |
| "learning_rate": 9.999238475781957e-05, | |
| "loss": 0.1909, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 4.9411764705882355, | |
| "eval_loss": 0.28984853625297546, | |
| "eval_runtime": 39.2314, | |
| "eval_samples_per_second": 0.918, | |
| "eval_steps_per_second": 0.918, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 5.176470588235294, | |
| "grad_norm": 0.594110369682312, | |
| "learning_rate": 9.99695413509548e-05, | |
| "loss": 0.1752, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 5.411764705882353, | |
| "grad_norm": 0.6123061776161194, | |
| "learning_rate": 9.99314767377287e-05, | |
| "loss": 0.0941, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 5.647058823529412, | |
| "grad_norm": 0.7264482378959656, | |
| "learning_rate": 9.987820251299122e-05, | |
| "loss": 0.1267, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 5.882352941176471, | |
| "grad_norm": 0.7877674102783203, | |
| "learning_rate": 9.980973490458728e-05, | |
| "loss": 0.1449, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_loss": 0.31277716159820557, | |
| "eval_runtime": 39.2378, | |
| "eval_samples_per_second": 0.917, | |
| "eval_steps_per_second": 0.917, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 6.117647058823529, | |
| "grad_norm": 0.4139578938484192, | |
| "learning_rate": 9.972609476841367e-05, | |
| "loss": 0.1387, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 6.352941176470588, | |
| "grad_norm": 0.5420397520065308, | |
| "learning_rate": 9.962730758206611e-05, | |
| "loss": 0.0973, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 6.588235294117647, | |
| "grad_norm": 6.396479606628418, | |
| "learning_rate": 9.951340343707852e-05, | |
| "loss": 0.0752, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 6.823529411764706, | |
| "grad_norm": 0.803831160068512, | |
| "learning_rate": 9.938441702975689e-05, | |
| "loss": 0.0826, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 6.9411764705882355, | |
| "eval_loss": 0.34875136613845825, | |
| "eval_runtime": 39.2256, | |
| "eval_samples_per_second": 0.918, | |
| "eval_steps_per_second": 0.918, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 7.0588235294117645, | |
| "grad_norm": 0.6163623332977295, | |
| "learning_rate": 9.924038765061042e-05, | |
| "loss": 0.1003, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 7.294117647058823, | |
| "grad_norm": 0.4413851797580719, | |
| "learning_rate": 9.908135917238321e-05, | |
| "loss": 0.0529, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 7.529411764705882, | |
| "grad_norm": 0.7398694157600403, | |
| "learning_rate": 9.890738003669029e-05, | |
| "loss": 0.0379, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 7.764705882352941, | |
| "grad_norm": 0.6616209745407104, | |
| "learning_rate": 9.871850323926177e-05, | |
| "loss": 0.0363, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "grad_norm": 0.9595809578895569, | |
| "learning_rate": 9.851478631379982e-05, | |
| "loss": 0.0486, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_loss": 0.42040765285491943, | |
| "eval_runtime": 39.2367, | |
| "eval_samples_per_second": 0.918, | |
| "eval_steps_per_second": 0.918, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 8.235294117647058, | |
| "grad_norm": 0.6897940635681152, | |
| "learning_rate": 9.829629131445342e-05, | |
| "loss": 0.0229, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 8.470588235294118, | |
| "grad_norm": 0.6147724390029907, | |
| "learning_rate": 9.806308479691595e-05, | |
| "loss": 0.0376, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 8.705882352941176, | |
| "grad_norm": 0.28395959734916687, | |
| "learning_rate": 9.781523779815179e-05, | |
| "loss": 0.0216, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 8.941176470588236, | |
| "grad_norm": 1.163219690322876, | |
| "learning_rate": 9.755282581475769e-05, | |
| "loss": 0.0329, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 8.941176470588236, | |
| "eval_loss": 0.43057769536972046, | |
| "eval_runtime": 39.2334, | |
| "eval_samples_per_second": 0.918, | |
| "eval_steps_per_second": 0.918, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 9.176470588235293, | |
| "grad_norm": 0.35383060574531555, | |
| "learning_rate": 9.727592877996585e-05, | |
| "loss": 0.0175, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 9.411764705882353, | |
| "grad_norm": 0.37475332617759705, | |
| "learning_rate": 9.698463103929542e-05, | |
| "loss": 0.013, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 9.647058823529411, | |
| "grad_norm": 0.8736576437950134, | |
| "learning_rate": 9.667902132486009e-05, | |
| "loss": 0.0179, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 9.882352941176471, | |
| "grad_norm": 0.35307109355926514, | |
| "learning_rate": 9.635919272833938e-05, | |
| "loss": 0.0163, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "eval_loss": 0.46002960205078125, | |
| "eval_runtime": 39.2371, | |
| "eval_samples_per_second": 0.917, | |
| "eval_steps_per_second": 0.917, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 10.117647058823529, | |
| "grad_norm": 0.2860690653324127, | |
| "learning_rate": 9.602524267262203e-05, | |
| "loss": 0.015, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 10.352941176470589, | |
| "grad_norm": 0.24865588545799255, | |
| "learning_rate": 9.567727288213005e-05, | |
| "loss": 0.0072, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 10.588235294117647, | |
| "grad_norm": 0.6196215152740479, | |
| "learning_rate": 9.53153893518325e-05, | |
| "loss": 0.0139, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 10.823529411764707, | |
| "grad_norm": 0.7884829640388489, | |
| "learning_rate": 9.493970231495835e-05, | |
| "loss": 0.0215, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 10.941176470588236, | |
| "eval_loss": 0.44994524121284485, | |
| "eval_runtime": 39.2456, | |
| "eval_samples_per_second": 0.917, | |
| "eval_steps_per_second": 0.917, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 11.058823529411764, | |
| "grad_norm": 0.14393794536590576, | |
| "learning_rate": 9.45503262094184e-05, | |
| "loss": 0.01, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 11.294117647058824, | |
| "grad_norm": 0.354295015335083, | |
| "learning_rate": 9.414737964294636e-05, | |
| "loss": 0.0125, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 11.529411764705882, | |
| "grad_norm": 0.40489304065704346, | |
| "learning_rate": 9.373098535696979e-05, | |
| "loss": 0.0082, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 11.764705882352942, | |
| "grad_norm": 0.4602890610694885, | |
| "learning_rate": 9.330127018922194e-05, | |
| "loss": 0.0055, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "grad_norm": 0.10772852599620819, | |
| "learning_rate": 9.285836503510562e-05, | |
| "loss": 0.0048, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "eval_loss": 0.4926491379737854, | |
| "eval_runtime": 39.3825, | |
| "eval_samples_per_second": 0.914, | |
| "eval_steps_per_second": 0.914, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "step": 102, | |
| "total_flos": 1.6007026740874445e+17, | |
| "train_loss": 0.43703304879440397, | |
| "train_runtime": 5927.738, | |
| "train_samples_per_second": 1.147, | |
| "train_steps_per_second": 0.067 | |
| } | |
| ], | |
| "logging_steps": 2, | |
| "max_steps": 400, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 50, | |
| "save_steps": 25, | |
| "stateful_callbacks": { | |
| "EarlyStoppingCallback": { | |
| "args": { | |
| "early_stopping_patience": 7, | |
| "early_stopping_threshold": 0.0 | |
| }, | |
| "attributes": { | |
| "early_stopping_patience_counter": 0 | |
| } | |
| }, | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 1.6007026740874445e+17, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |