| { | |
| "best_metric": 0.8525590896606445, | |
| "best_model_checkpoint": "data/Mistral-7B_task-3_180-samples_config-2_full_auto/checkpoint-59", | |
| "epoch": 14.0, | |
| "eval_steps": 500, | |
| "global_step": 119, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.11764705882352941, | |
| "grad_norm": 0.9888336062431335, | |
| "learning_rate": 2.5e-06, | |
| "loss": 1.242, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.23529411764705882, | |
| "grad_norm": 1.031571626663208, | |
| "learning_rate": 5e-06, | |
| "loss": 1.2501, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.47058823529411764, | |
| "grad_norm": 0.8961756825447083, | |
| "learning_rate": 1e-05, | |
| "loss": 1.2536, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.7058823529411765, | |
| "grad_norm": 0.8075014352798462, | |
| "learning_rate": 1.5e-05, | |
| "loss": 1.1921, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.9411764705882353, | |
| "grad_norm": 0.6888613700866699, | |
| "learning_rate": 2e-05, | |
| "loss": 1.2028, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.9411764705882353, | |
| "eval_loss": 1.1676641702651978, | |
| "eval_runtime": 39.2678, | |
| "eval_samples_per_second": 0.917, | |
| "eval_steps_per_second": 0.917, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 1.1764705882352942, | |
| "grad_norm": 0.5692386031150818, | |
| "learning_rate": 2.5e-05, | |
| "loss": 1.1736, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 1.4117647058823528, | |
| "grad_norm": 0.4184214174747467, | |
| "learning_rate": 3e-05, | |
| "loss": 1.1096, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 1.6470588235294117, | |
| "grad_norm": 0.34876593947410583, | |
| "learning_rate": 3.5e-05, | |
| "loss": 1.0825, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 1.8823529411764706, | |
| "grad_norm": 0.34462854266166687, | |
| "learning_rate": 4e-05, | |
| "loss": 1.0682, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_loss": 1.0442286729812622, | |
| "eval_runtime": 39.2563, | |
| "eval_samples_per_second": 0.917, | |
| "eval_steps_per_second": 0.917, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 2.1176470588235294, | |
| "grad_norm": 0.3480481207370758, | |
| "learning_rate": 4.5e-05, | |
| "loss": 1.0475, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 2.3529411764705883, | |
| "grad_norm": 0.35165414214134216, | |
| "learning_rate": 5e-05, | |
| "loss": 0.9996, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 2.588235294117647, | |
| "grad_norm": 0.476762980222702, | |
| "learning_rate": 5.500000000000001e-05, | |
| "loss": 0.9701, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 2.8235294117647056, | |
| "grad_norm": 0.4732675552368164, | |
| "learning_rate": 6e-05, | |
| "loss": 0.9635, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 2.9411764705882355, | |
| "eval_loss": 0.9280573725700378, | |
| "eval_runtime": 39.2473, | |
| "eval_samples_per_second": 0.917, | |
| "eval_steps_per_second": 0.917, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 3.0588235294117645, | |
| "grad_norm": 0.3753867745399475, | |
| "learning_rate": 6.500000000000001e-05, | |
| "loss": 0.9033, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 3.2941176470588234, | |
| "grad_norm": 0.2556222975254059, | |
| "learning_rate": 7e-05, | |
| "loss": 0.8736, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 3.5294117647058822, | |
| "grad_norm": 0.24436281621456146, | |
| "learning_rate": 7.500000000000001e-05, | |
| "loss": 0.841, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 3.764705882352941, | |
| "grad_norm": 0.2638224959373474, | |
| "learning_rate": 8e-05, | |
| "loss": 0.8726, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 0.23370446264743805, | |
| "learning_rate": 8.5e-05, | |
| "loss": 0.8271, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_loss": 0.8805874586105347, | |
| "eval_runtime": 39.2547, | |
| "eval_samples_per_second": 0.917, | |
| "eval_steps_per_second": 0.917, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 4.235294117647059, | |
| "grad_norm": 0.20400899648666382, | |
| "learning_rate": 9e-05, | |
| "loss": 0.8304, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 4.470588235294118, | |
| "grad_norm": 0.20834453403949738, | |
| "learning_rate": 9.5e-05, | |
| "loss": 0.8156, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 4.705882352941177, | |
| "grad_norm": 0.20543907582759857, | |
| "learning_rate": 0.0001, | |
| "loss": 0.8225, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 4.9411764705882355, | |
| "grad_norm": 0.22563892602920532, | |
| "learning_rate": 9.999238475781957e-05, | |
| "loss": 0.8008, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 4.9411764705882355, | |
| "eval_loss": 0.8643261790275574, | |
| "eval_runtime": 39.25, | |
| "eval_samples_per_second": 0.917, | |
| "eval_steps_per_second": 0.917, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 5.176470588235294, | |
| "grad_norm": 0.24295645952224731, | |
| "learning_rate": 9.99695413509548e-05, | |
| "loss": 0.7693, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 5.411764705882353, | |
| "grad_norm": 0.19249586760997772, | |
| "learning_rate": 9.99314767377287e-05, | |
| "loss": 0.8029, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 5.647058823529412, | |
| "grad_norm": 0.2038288563489914, | |
| "learning_rate": 9.987820251299122e-05, | |
| "loss": 0.778, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 5.882352941176471, | |
| "grad_norm": 0.21831390261650085, | |
| "learning_rate": 9.980973490458728e-05, | |
| "loss": 0.7975, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_loss": 0.854880690574646, | |
| "eval_runtime": 39.259, | |
| "eval_samples_per_second": 0.917, | |
| "eval_steps_per_second": 0.917, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 6.117647058823529, | |
| "grad_norm": 0.2200762927532196, | |
| "learning_rate": 9.972609476841367e-05, | |
| "loss": 0.7418, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 6.352941176470588, | |
| "grad_norm": 0.2441394180059433, | |
| "learning_rate": 9.962730758206611e-05, | |
| "loss": 0.7239, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 6.588235294117647, | |
| "grad_norm": 0.24432151019573212, | |
| "learning_rate": 9.951340343707852e-05, | |
| "loss": 0.7485, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 6.823529411764706, | |
| "grad_norm": 0.2585461139678955, | |
| "learning_rate": 9.938441702975689e-05, | |
| "loss": 0.7857, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 6.9411764705882355, | |
| "eval_loss": 0.8525590896606445, | |
| "eval_runtime": 39.2437, | |
| "eval_samples_per_second": 0.917, | |
| "eval_steps_per_second": 0.917, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 7.0588235294117645, | |
| "grad_norm": 0.27814191579818726, | |
| "learning_rate": 9.924038765061042e-05, | |
| "loss": 0.7329, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 7.294117647058823, | |
| "grad_norm": 0.28424516320228577, | |
| "learning_rate": 9.908135917238321e-05, | |
| "loss": 0.713, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 7.529411764705882, | |
| "grad_norm": 0.4684115946292877, | |
| "learning_rate": 9.890738003669029e-05, | |
| "loss": 0.6823, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 7.764705882352941, | |
| "grad_norm": 0.3626362681388855, | |
| "learning_rate": 9.871850323926177e-05, | |
| "loss": 0.6824, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "grad_norm": 0.41377174854278564, | |
| "learning_rate": 9.851478631379982e-05, | |
| "loss": 0.6998, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_loss": 0.8610419631004333, | |
| "eval_runtime": 39.2588, | |
| "eval_samples_per_second": 0.917, | |
| "eval_steps_per_second": 0.917, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 8.235294117647058, | |
| "grad_norm": 0.41785913705825806, | |
| "learning_rate": 9.829629131445342e-05, | |
| "loss": 0.6432, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 8.470588235294118, | |
| "grad_norm": 0.48981037735939026, | |
| "learning_rate": 9.806308479691595e-05, | |
| "loss": 0.6646, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 8.705882352941176, | |
| "grad_norm": 0.45041418075561523, | |
| "learning_rate": 9.781523779815179e-05, | |
| "loss": 0.661, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 8.941176470588236, | |
| "grad_norm": 0.476771742105484, | |
| "learning_rate": 9.755282581475769e-05, | |
| "loss": 0.5987, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 8.941176470588236, | |
| "eval_loss": 0.8878731727600098, | |
| "eval_runtime": 39.2513, | |
| "eval_samples_per_second": 0.917, | |
| "eval_steps_per_second": 0.917, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 9.176470588235293, | |
| "grad_norm": 0.49439936876296997, | |
| "learning_rate": 9.727592877996585e-05, | |
| "loss": 0.5998, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 9.411764705882353, | |
| "grad_norm": 0.6181315183639526, | |
| "learning_rate": 9.698463103929542e-05, | |
| "loss": 0.5946, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 9.647058823529411, | |
| "grad_norm": 0.5851460695266724, | |
| "learning_rate": 9.667902132486009e-05, | |
| "loss": 0.5762, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 9.882352941176471, | |
| "grad_norm": 0.7100406289100647, | |
| "learning_rate": 9.635919272833938e-05, | |
| "loss": 0.5553, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "eval_loss": 0.9138666987419128, | |
| "eval_runtime": 39.2496, | |
| "eval_samples_per_second": 0.917, | |
| "eval_steps_per_second": 0.917, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 10.117647058823529, | |
| "grad_norm": 0.7029377818107605, | |
| "learning_rate": 9.602524267262203e-05, | |
| "loss": 0.5222, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 10.352941176470589, | |
| "grad_norm": 0.7276427745819092, | |
| "learning_rate": 9.567727288213005e-05, | |
| "loss": 0.5263, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 10.588235294117647, | |
| "grad_norm": 0.7112120985984802, | |
| "learning_rate": 9.53153893518325e-05, | |
| "loss": 0.4885, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 10.823529411764707, | |
| "grad_norm": 0.7985062003135681, | |
| "learning_rate": 9.493970231495835e-05, | |
| "loss": 0.4959, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 10.941176470588236, | |
| "eval_loss": 0.9456067681312561, | |
| "eval_runtime": 39.2558, | |
| "eval_samples_per_second": 0.917, | |
| "eval_steps_per_second": 0.917, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 11.058823529411764, | |
| "grad_norm": 0.7065971493721008, | |
| "learning_rate": 9.45503262094184e-05, | |
| "loss": 0.5183, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 11.294117647058824, | |
| "grad_norm": 1.4326319694519043, | |
| "learning_rate": 9.414737964294636e-05, | |
| "loss": 0.4754, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 11.529411764705882, | |
| "grad_norm": 1.0542569160461426, | |
| "learning_rate": 9.373098535696979e-05, | |
| "loss": 0.4067, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 11.764705882352942, | |
| "grad_norm": 0.9199151992797852, | |
| "learning_rate": 9.330127018922194e-05, | |
| "loss": 0.4442, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "grad_norm": 1.002302646636963, | |
| "learning_rate": 9.285836503510562e-05, | |
| "loss": 0.4148, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "eval_loss": 0.9964881539344788, | |
| "eval_runtime": 39.2547, | |
| "eval_samples_per_second": 0.917, | |
| "eval_steps_per_second": 0.917, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 12.235294117647058, | |
| "grad_norm": 1.159452199935913, | |
| "learning_rate": 9.24024048078213e-05, | |
| "loss": 0.3856, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 12.470588235294118, | |
| "grad_norm": 1.2748922109603882, | |
| "learning_rate": 9.193352839727121e-05, | |
| "loss": 0.3609, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 12.705882352941176, | |
| "grad_norm": 1.052121639251709, | |
| "learning_rate": 9.145187862775209e-05, | |
| "loss": 0.3814, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 12.941176470588236, | |
| "grad_norm": 0.992569088935852, | |
| "learning_rate": 9.09576022144496e-05, | |
| "loss": 0.3734, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 12.941176470588236, | |
| "eval_loss": 1.0426888465881348, | |
| "eval_runtime": 39.2515, | |
| "eval_samples_per_second": 0.917, | |
| "eval_steps_per_second": 0.917, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 13.176470588235293, | |
| "grad_norm": 0.9615302681922913, | |
| "learning_rate": 9.045084971874738e-05, | |
| "loss": 0.3507, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 13.411764705882353, | |
| "grad_norm": 0.9408590197563171, | |
| "learning_rate": 8.993177550236464e-05, | |
| "loss": 0.2905, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 13.647058823529411, | |
| "grad_norm": 1.0201396942138672, | |
| "learning_rate": 8.940053768033609e-05, | |
| "loss": 0.3232, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 13.882352941176471, | |
| "grad_norm": 1.023605227470398, | |
| "learning_rate": 8.885729807284856e-05, | |
| "loss": 0.3239, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 14.0, | |
| "eval_loss": 1.0871130228042603, | |
| "eval_runtime": 39.2585, | |
| "eval_samples_per_second": 0.917, | |
| "eval_steps_per_second": 0.917, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 14.0, | |
| "step": 119, | |
| "total_flos": 1.867486456793006e+17, | |
| "train_loss": 0.7177910867358456, | |
| "train_runtime": 6920.2527, | |
| "train_samples_per_second": 0.983, | |
| "train_steps_per_second": 0.058 | |
| } | |
| ], | |
| "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": 1.867486456793006e+17, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |