{ "best_metric": 0.7895187139511108, "best_model_checkpoint": "data/Mistral-7B_task-2_180-samples_config-2_full/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": 1.0456639528274536, "learning_rate": 2.5e-06, "loss": 1.2021, "step": 1 }, { "epoch": 0.23529411764705882, "grad_norm": 1.160530686378479, "learning_rate": 5e-06, "loss": 1.2279, "step": 2 }, { "epoch": 0.47058823529411764, "grad_norm": 1.0452324151992798, "learning_rate": 1e-05, "loss": 1.2202, "step": 4 }, { "epoch": 0.7058823529411765, "grad_norm": 0.9364596009254456, "learning_rate": 1.5e-05, "loss": 1.1454, "step": 6 }, { "epoch": 0.9411764705882353, "grad_norm": 0.8268067836761475, "learning_rate": 2e-05, "loss": 1.1409, "step": 8 }, { "epoch": 0.9411764705882353, "eval_loss": 1.1128188371658325, "eval_runtime": 45.5858, "eval_samples_per_second": 0.79, "eval_steps_per_second": 0.79, "step": 8 }, { "epoch": 1.1764705882352942, "grad_norm": 0.7106637358665466, "learning_rate": 2.5e-05, "loss": 1.1137, "step": 10 }, { "epoch": 1.4117647058823528, "grad_norm": 0.4322061836719513, "learning_rate": 3e-05, "loss": 1.0466, "step": 12 }, { "epoch": 1.6470588235294117, "grad_norm": 0.33850380778312683, "learning_rate": 3.5e-05, "loss": 1.0067, "step": 14 }, { "epoch": 1.8823529411764706, "grad_norm": 0.33021923899650574, "learning_rate": 4e-05, "loss": 1.0026, "step": 16 }, { "epoch": 2.0, "eval_loss": 0.9789136052131653, "eval_runtime": 45.5832, "eval_samples_per_second": 0.79, "eval_steps_per_second": 0.79, "step": 17 }, { "epoch": 2.1176470588235294, "grad_norm": 0.32015666365623474, "learning_rate": 4.5e-05, "loss": 0.984, "step": 18 }, { "epoch": 2.3529411764705883, "grad_norm": 0.32732608914375305, "learning_rate": 5e-05, "loss": 0.9313, "step": 20 }, { "epoch": 2.588235294117647, "grad_norm": 0.3938484787940979, "learning_rate": 5.500000000000001e-05, "loss": 0.9128, "step": 22 }, { "epoch": 2.8235294117647056, "grad_norm": 0.4519818127155304, "learning_rate": 6e-05, "loss": 0.9165, "step": 24 }, { "epoch": 2.9411764705882355, "eval_loss": 0.8753034472465515, "eval_runtime": 45.5804, "eval_samples_per_second": 0.79, "eval_steps_per_second": 0.79, "step": 25 }, { "epoch": 3.0588235294117645, "grad_norm": 0.34580570459365845, "learning_rate": 6.500000000000001e-05, "loss": 0.853, "step": 26 }, { "epoch": 3.2941176470588234, "grad_norm": 0.2205946445465088, "learning_rate": 7e-05, "loss": 0.8267, "step": 28 }, { "epoch": 3.5294117647058822, "grad_norm": 0.2162715494632721, "learning_rate": 7.500000000000001e-05, "loss": 0.7938, "step": 30 }, { "epoch": 3.764705882352941, "grad_norm": 0.22695396840572357, "learning_rate": 8e-05, "loss": 0.8199, "step": 32 }, { "epoch": 4.0, "grad_norm": 0.1880607008934021, "learning_rate": 8.5e-05, "loss": 0.7642, "step": 34 }, { "epoch": 4.0, "eval_loss": 0.8235906958580017, "eval_runtime": 45.5855, "eval_samples_per_second": 0.79, "eval_steps_per_second": 0.79, "step": 34 }, { "epoch": 4.235294117647059, "grad_norm": 0.1942065954208374, "learning_rate": 9e-05, "loss": 0.7806, "step": 36 }, { "epoch": 4.470588235294118, "grad_norm": 0.20216228067874908, "learning_rate": 9.5e-05, "loss": 0.7715, "step": 38 }, { "epoch": 4.705882352941177, "grad_norm": 0.16827207803726196, "learning_rate": 0.0001, "loss": 0.7682, "step": 40 }, { "epoch": 4.9411764705882355, "grad_norm": 0.18430998921394348, "learning_rate": 9.999238475781957e-05, "loss": 0.7468, "step": 42 }, { "epoch": 4.9411764705882355, "eval_loss": 0.8065404891967773, "eval_runtime": 45.5736, "eval_samples_per_second": 0.79, "eval_steps_per_second": 0.79, "step": 42 }, { "epoch": 5.176470588235294, "grad_norm": 0.1839626580476761, "learning_rate": 9.99695413509548e-05, "loss": 0.7233, "step": 44 }, { "epoch": 5.411764705882353, "grad_norm": 0.16396723687648773, "learning_rate": 9.99314767377287e-05, "loss": 0.752, "step": 46 }, { "epoch": 5.647058823529412, "grad_norm": 0.17820850014686584, "learning_rate": 9.987820251299122e-05, "loss": 0.7214, "step": 48 }, { "epoch": 5.882352941176471, "grad_norm": 0.1832844614982605, "learning_rate": 9.980973490458728e-05, "loss": 0.7471, "step": 50 }, { "epoch": 6.0, "eval_loss": 0.7945607304573059, "eval_runtime": 45.632, "eval_samples_per_second": 0.789, "eval_steps_per_second": 0.789, "step": 51 }, { "epoch": 6.117647058823529, "grad_norm": 0.18397541344165802, "learning_rate": 9.972609476841367e-05, "loss": 0.7011, "step": 52 }, { "epoch": 6.352941176470588, "grad_norm": 0.1946660727262497, "learning_rate": 9.962730758206611e-05, "loss": 0.6777, "step": 54 }, { "epoch": 6.588235294117647, "grad_norm": 0.20609663426876068, "learning_rate": 9.951340343707852e-05, "loss": 0.7041, "step": 56 }, { "epoch": 6.823529411764706, "grad_norm": 0.20319028198719025, "learning_rate": 9.938441702975689e-05, "loss": 0.7355, "step": 58 }, { "epoch": 6.9411764705882355, "eval_loss": 0.7895187139511108, "eval_runtime": 45.6, "eval_samples_per_second": 0.789, "eval_steps_per_second": 0.789, "step": 59 }, { "epoch": 7.0588235294117645, "grad_norm": 0.2320169359445572, "learning_rate": 9.924038765061042e-05, "loss": 0.697, "step": 60 }, { "epoch": 7.294117647058823, "grad_norm": 0.24622970819473267, "learning_rate": 9.908135917238321e-05, "loss": 0.6762, "step": 62 }, { "epoch": 7.529411764705882, "grad_norm": 0.3519413471221924, "learning_rate": 9.890738003669029e-05, "loss": 0.6447, "step": 64 }, { "epoch": 7.764705882352941, "grad_norm": 0.29206714034080505, "learning_rate": 9.871850323926177e-05, "loss": 0.64, "step": 66 }, { "epoch": 8.0, "grad_norm": 0.339593768119812, "learning_rate": 9.851478631379982e-05, "loss": 0.6638, "step": 68 }, { "epoch": 8.0, "eval_loss": 0.7944325804710388, "eval_runtime": 45.5919, "eval_samples_per_second": 0.79, "eval_steps_per_second": 0.79, "step": 68 }, { "epoch": 8.235294117647058, "grad_norm": 0.34579718112945557, "learning_rate": 9.829629131445342e-05, "loss": 0.6154, "step": 70 }, { "epoch": 8.470588235294118, "grad_norm": 0.41827669739723206, "learning_rate": 9.806308479691595e-05, "loss": 0.6323, "step": 72 }, { "epoch": 8.705882352941176, "grad_norm": 0.3427964150905609, "learning_rate": 9.781523779815179e-05, "loss": 0.6285, "step": 74 }, { "epoch": 8.941176470588236, "grad_norm": 0.4158247709274292, "learning_rate": 9.755282581475769e-05, "loss": 0.5766, "step": 76 }, { "epoch": 8.941176470588236, "eval_loss": 0.8071392178535461, "eval_runtime": 45.605, "eval_samples_per_second": 0.789, "eval_steps_per_second": 0.789, "step": 76 }, { "epoch": 9.176470588235293, "grad_norm": 0.42728355526924133, "learning_rate": 9.727592877996585e-05, "loss": 0.5802, "step": 78 }, { "epoch": 9.411764705882353, "grad_norm": 0.6478773951530457, "learning_rate": 9.698463103929542e-05, "loss": 0.5721, "step": 80 }, { "epoch": 9.647058823529411, "grad_norm": 0.49921339750289917, "learning_rate": 9.667902132486009e-05, "loss": 0.5561, "step": 82 }, { "epoch": 9.882352941176471, "grad_norm": 0.554269552230835, "learning_rate": 9.635919272833938e-05, "loss": 0.542, "step": 84 }, { "epoch": 10.0, "eval_loss": 0.8386737704277039, "eval_runtime": 45.5905, "eval_samples_per_second": 0.79, "eval_steps_per_second": 0.79, "step": 85 }, { "epoch": 10.117647058823529, "grad_norm": 0.5353624820709229, "learning_rate": 9.602524267262203e-05, "loss": 0.5079, "step": 86 }, { "epoch": 10.352941176470589, "grad_norm": 0.7397503852844238, "learning_rate": 9.567727288213005e-05, "loss": 0.5127, "step": 88 }, { "epoch": 10.588235294117647, "grad_norm": 0.61458420753479, "learning_rate": 9.53153893518325e-05, "loss": 0.479, "step": 90 }, { "epoch": 10.823529411764707, "grad_norm": 0.7933974862098694, "learning_rate": 9.493970231495835e-05, "loss": 0.4866, "step": 92 }, { "epoch": 10.941176470588236, "eval_loss": 0.8615274429321289, "eval_runtime": 45.5929, "eval_samples_per_second": 0.79, "eval_steps_per_second": 0.79, "step": 93 }, { "epoch": 11.058823529411764, "grad_norm": 0.5533405542373657, "learning_rate": 9.45503262094184e-05, "loss": 0.5055, "step": 94 }, { "epoch": 11.294117647058824, "grad_norm": 1.2706093788146973, "learning_rate": 9.414737964294636e-05, "loss": 0.471, "step": 96 }, { "epoch": 11.529411764705882, "grad_norm": 0.9889309406280518, "learning_rate": 9.373098535696979e-05, "loss": 0.4058, "step": 98 }, { "epoch": 11.764705882352942, "grad_norm": 0.9496291875839233, "learning_rate": 9.330127018922194e-05, "loss": 0.4304, "step": 100 }, { "epoch": 12.0, "grad_norm": 0.6629786491394043, "learning_rate": 9.285836503510562e-05, "loss": 0.4049, "step": 102 }, { "epoch": 12.0, "eval_loss": 0.8862788677215576, "eval_runtime": 45.5933, "eval_samples_per_second": 0.79, "eval_steps_per_second": 0.79, "step": 102 }, { "epoch": 12.235294117647058, "grad_norm": 0.7883285284042358, "learning_rate": 9.24024048078213e-05, "loss": 0.3752, "step": 104 }, { "epoch": 12.470588235294118, "grad_norm": 0.7288637161254883, "learning_rate": 9.193352839727121e-05, "loss": 0.3598, "step": 106 }, { "epoch": 12.705882352941176, "grad_norm": 0.8219139575958252, "learning_rate": 9.145187862775209e-05, "loss": 0.3776, "step": 108 }, { "epoch": 12.941176470588236, "grad_norm": 0.9806149005889893, "learning_rate": 9.09576022144496e-05, "loss": 0.3762, "step": 110 }, { "epoch": 12.941176470588236, "eval_loss": 0.9318940043449402, "eval_runtime": 45.6019, "eval_samples_per_second": 0.789, "eval_steps_per_second": 0.789, "step": 110 }, { "epoch": 13.176470588235293, "grad_norm": 0.7644268274307251, "learning_rate": 9.045084971874738e-05, "loss": 0.3491, "step": 112 }, { "epoch": 13.411764705882353, "grad_norm": 1.0329344272613525, "learning_rate": 8.993177550236464e-05, "loss": 0.2992, "step": 114 }, { "epoch": 13.647058823529411, "grad_norm": 0.8051264882087708, "learning_rate": 8.940053768033609e-05, "loss": 0.3213, "step": 116 }, { "epoch": 13.882352941176471, "grad_norm": 0.882315456867218, "learning_rate": 8.885729807284856e-05, "loss": 0.3261, "step": 118 }, { "epoch": 14.0, "eval_loss": 1.010295033454895, "eval_runtime": 45.5925, "eval_samples_per_second": 0.79, "eval_steps_per_second": 0.79, "step": 119 }, { "epoch": 14.0, "step": 119, "total_flos": 2.2561346208294502e+17, "train_loss": 0.6842404608966923, "train_runtime": 8185.4036, "train_samples_per_second": 0.831, "train_steps_per_second": 0.049 } ], "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": 2.2561346208294502e+17, "train_batch_size": 1, "trial_name": null, "trial_params": null }