| { | |
| "best_metric": 0.7932864427566528, | |
| "best_model_checkpoint": "data/Mistral-7B_task-1_180-samples_config-2_auto/checkpoint-34", | |
| "epoch": 10.941176470588236, | |
| "eval_steps": 500, | |
| "global_step": 93, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.11764705882352941, | |
| "grad_norm": 2.450850009918213, | |
| "learning_rate": 2.5e-06, | |
| "loss": 1.0532, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.23529411764705882, | |
| "grad_norm": 2.664348602294922, | |
| "learning_rate": 5e-06, | |
| "loss": 1.038, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.47058823529411764, | |
| "grad_norm": 2.440438985824585, | |
| "learning_rate": 1e-05, | |
| "loss": 1.1264, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.7058823529411765, | |
| "grad_norm": 2.530811309814453, | |
| "learning_rate": 1.5e-05, | |
| "loss": 1.0888, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.9411764705882353, | |
| "grad_norm": 2.1098403930664062, | |
| "learning_rate": 2e-05, | |
| "loss": 1.0549, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.9411764705882353, | |
| "eval_loss": 0.9900373816490173, | |
| "eval_runtime": 14.9949, | |
| "eval_samples_per_second": 2.401, | |
| "eval_steps_per_second": 2.401, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 1.1764705882352942, | |
| "grad_norm": 1.6433088779449463, | |
| "learning_rate": 2.5e-05, | |
| "loss": 0.9855, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 1.4117647058823528, | |
| "grad_norm": 1.2828103303909302, | |
| "learning_rate": 3e-05, | |
| "loss": 0.8472, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 1.6470588235294117, | |
| "grad_norm": 1.1351650953292847, | |
| "learning_rate": 3.5e-05, | |
| "loss": 0.8693, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 1.8823529411764706, | |
| "grad_norm": 1.176932692527771, | |
| "learning_rate": 4e-05, | |
| "loss": 0.8222, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_loss": 0.8563865423202515, | |
| "eval_runtime": 14.9839, | |
| "eval_samples_per_second": 2.403, | |
| "eval_steps_per_second": 2.403, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 2.1176470588235294, | |
| "grad_norm": 1.0588655471801758, | |
| "learning_rate": 4.5e-05, | |
| "loss": 0.7559, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 2.3529411764705883, | |
| "grad_norm": 1.0662921667099, | |
| "learning_rate": 5e-05, | |
| "loss": 0.7459, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 2.588235294117647, | |
| "grad_norm": 0.9237774610519409, | |
| "learning_rate": 5.500000000000001e-05, | |
| "loss": 0.7367, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 2.8235294117647056, | |
| "grad_norm": 0.9688161015510559, | |
| "learning_rate": 6e-05, | |
| "loss": 0.7511, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 2.9411764705882355, | |
| "eval_loss": 0.798934280872345, | |
| "eval_runtime": 14.9861, | |
| "eval_samples_per_second": 2.402, | |
| "eval_steps_per_second": 2.402, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 3.0588235294117645, | |
| "grad_norm": 0.9950412511825562, | |
| "learning_rate": 6.500000000000001e-05, | |
| "loss": 0.6961, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 3.2941176470588234, | |
| "grad_norm": 0.9355748295783997, | |
| "learning_rate": 7e-05, | |
| "loss": 0.6115, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 3.5294117647058822, | |
| "grad_norm": 1.1376206874847412, | |
| "learning_rate": 7.500000000000001e-05, | |
| "loss": 0.6316, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 3.764705882352941, | |
| "grad_norm": 1.149610996246338, | |
| "learning_rate": 8e-05, | |
| "loss": 0.595, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 1.1590631008148193, | |
| "learning_rate": 8.5e-05, | |
| "loss": 0.5644, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_loss": 0.7932864427566528, | |
| "eval_runtime": 15.0042, | |
| "eval_samples_per_second": 2.399, | |
| "eval_steps_per_second": 2.399, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 4.235294117647059, | |
| "grad_norm": 1.2311124801635742, | |
| "learning_rate": 9e-05, | |
| "loss": 0.4864, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 4.470588235294118, | |
| "grad_norm": 1.1586334705352783, | |
| "learning_rate": 9.5e-05, | |
| "loss": 0.4053, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 4.705882352941177, | |
| "grad_norm": 1.3652201890945435, | |
| "learning_rate": 0.0001, | |
| "loss": 0.4336, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 4.9411764705882355, | |
| "grad_norm": 1.6470670700073242, | |
| "learning_rate": 9.999238475781957e-05, | |
| "loss": 0.3911, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 4.9411764705882355, | |
| "eval_loss": 0.9402649998664856, | |
| "eval_runtime": 14.9808, | |
| "eval_samples_per_second": 2.403, | |
| "eval_steps_per_second": 2.403, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 5.176470588235294, | |
| "grad_norm": 1.2947176694869995, | |
| "learning_rate": 9.99695413509548e-05, | |
| "loss": 0.2718, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 5.411764705882353, | |
| "grad_norm": 1.3543827533721924, | |
| "learning_rate": 9.99314767377287e-05, | |
| "loss": 0.2345, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 5.647058823529412, | |
| "grad_norm": 1.7274010181427002, | |
| "learning_rate": 9.987820251299122e-05, | |
| "loss": 0.2069, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 5.882352941176471, | |
| "grad_norm": 1.9732545614242554, | |
| "learning_rate": 9.980973490458728e-05, | |
| "loss": 0.2483, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_loss": 1.0406984090805054, | |
| "eval_runtime": 14.9922, | |
| "eval_samples_per_second": 2.401, | |
| "eval_steps_per_second": 2.401, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 6.117647058823529, | |
| "grad_norm": 1.330666422843933, | |
| "learning_rate": 9.972609476841367e-05, | |
| "loss": 0.187, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 6.352941176470588, | |
| "grad_norm": 1.65781831741333, | |
| "learning_rate": 9.962730758206611e-05, | |
| "loss": 0.1408, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 6.588235294117647, | |
| "grad_norm": 1.3344322443008423, | |
| "learning_rate": 9.951340343707852e-05, | |
| "loss": 0.1191, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 6.823529411764706, | |
| "grad_norm": 1.7343159914016724, | |
| "learning_rate": 9.938441702975689e-05, | |
| "loss": 0.1183, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 6.9411764705882355, | |
| "eval_loss": 1.192215085029602, | |
| "eval_runtime": 14.9851, | |
| "eval_samples_per_second": 2.402, | |
| "eval_steps_per_second": 2.402, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 7.0588235294117645, | |
| "grad_norm": 1.408785104751587, | |
| "learning_rate": 9.924038765061042e-05, | |
| "loss": 0.1173, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 7.294117647058823, | |
| "grad_norm": 1.2464698553085327, | |
| "learning_rate": 9.908135917238321e-05, | |
| "loss": 0.0852, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 7.529411764705882, | |
| "grad_norm": 1.4764487743377686, | |
| "learning_rate": 9.890738003669029e-05, | |
| "loss": 0.0745, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 7.764705882352941, | |
| "grad_norm": 2.1253485679626465, | |
| "learning_rate": 9.871850323926177e-05, | |
| "loss": 0.088, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "grad_norm": 1.791509985923767, | |
| "learning_rate": 9.851478631379982e-05, | |
| "loss": 0.0848, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_loss": 1.2871198654174805, | |
| "eval_runtime": 14.983, | |
| "eval_samples_per_second": 2.403, | |
| "eval_steps_per_second": 2.403, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 8.235294117647058, | |
| "grad_norm": 1.1020724773406982, | |
| "learning_rate": 9.829629131445342e-05, | |
| "loss": 0.0543, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 8.470588235294118, | |
| "grad_norm": 1.1706922054290771, | |
| "learning_rate": 9.806308479691595e-05, | |
| "loss": 0.0547, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 8.705882352941176, | |
| "grad_norm": 1.9083949327468872, | |
| "learning_rate": 9.781523779815179e-05, | |
| "loss": 0.0682, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 8.941176470588236, | |
| "grad_norm": 2.0456669330596924, | |
| "learning_rate": 9.755282581475769e-05, | |
| "loss": 0.0657, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 8.941176470588236, | |
| "eval_loss": 1.3606077432632446, | |
| "eval_runtime": 14.982, | |
| "eval_samples_per_second": 2.403, | |
| "eval_steps_per_second": 2.403, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 9.176470588235293, | |
| "grad_norm": 1.3962442874908447, | |
| "learning_rate": 9.727592877996585e-05, | |
| "loss": 0.0433, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 9.411764705882353, | |
| "grad_norm": 1.3246513605117798, | |
| "learning_rate": 9.698463103929542e-05, | |
| "loss": 0.0416, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 9.647058823529411, | |
| "grad_norm": 1.652782917022705, | |
| "learning_rate": 9.667902132486009e-05, | |
| "loss": 0.0564, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 9.882352941176471, | |
| "grad_norm": 1.3555011749267578, | |
| "learning_rate": 9.635919272833938e-05, | |
| "loss": 0.0445, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "eval_loss": 1.36819589138031, | |
| "eval_runtime": 14.9982, | |
| "eval_samples_per_second": 2.4, | |
| "eval_steps_per_second": 2.4, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 10.117647058823529, | |
| "grad_norm": 0.7261105179786682, | |
| "learning_rate": 9.602524267262203e-05, | |
| "loss": 0.0392, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 10.352941176470589, | |
| "grad_norm": 1.1909233331680298, | |
| "learning_rate": 9.567727288213005e-05, | |
| "loss": 0.0336, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 10.588235294117647, | |
| "grad_norm": 1.4687535762786865, | |
| "learning_rate": 9.53153893518325e-05, | |
| "loss": 0.0445, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 10.823529411764707, | |
| "grad_norm": 1.1714824438095093, | |
| "learning_rate": 9.493970231495835e-05, | |
| "loss": 0.0398, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 10.941176470588236, | |
| "eval_loss": 1.3901447057724, | |
| "eval_runtime": 14.9942, | |
| "eval_samples_per_second": 2.401, | |
| "eval_steps_per_second": 2.401, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 10.941176470588236, | |
| "step": 93, | |
| "total_flos": 2.518477265751245e+16, | |
| "train_loss": 0.3923261252141768, | |
| "train_runtime": 1825.5377, | |
| "train_samples_per_second": 3.725, | |
| "train_steps_per_second": 0.219 | |
| } | |
| ], | |
| "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": 2.518477265751245e+16, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |