| { | |
| "best_metric": 0.7669743299484253, | |
| "best_model_checkpoint": "data/Mistral-7B_task-2_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": 0.937061607837677, | |
| "learning_rate": 2.5e-06, | |
| "loss": 1.146, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.23529411764705882, | |
| "grad_norm": 1.0251848697662354, | |
| "learning_rate": 5e-06, | |
| "loss": 1.1622, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.47058823529411764, | |
| "grad_norm": 0.8865615725517273, | |
| "learning_rate": 1e-05, | |
| "loss": 1.1525, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.7058823529411765, | |
| "grad_norm": 0.7962722182273865, | |
| "learning_rate": 1.5e-05, | |
| "loss": 1.0981, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.9411764705882353, | |
| "grad_norm": 0.6907485723495483, | |
| "learning_rate": 2e-05, | |
| "loss": 1.0907, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.9411764705882353, | |
| "eval_loss": 1.068891167640686, | |
| "eval_runtime": 46.7721, | |
| "eval_samples_per_second": 0.77, | |
| "eval_steps_per_second": 0.77, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 1.1764705882352942, | |
| "grad_norm": 0.5811184644699097, | |
| "learning_rate": 2.5e-05, | |
| "loss": 1.0607, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 1.4117647058823528, | |
| "grad_norm": 0.39460939168930054, | |
| "learning_rate": 3e-05, | |
| "loss": 1.0086, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 1.6470588235294117, | |
| "grad_norm": 0.31468281149864197, | |
| "learning_rate": 3.5e-05, | |
| "loss": 0.9817, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 1.8823529411764706, | |
| "grad_norm": 0.41152703762054443, | |
| "learning_rate": 4e-05, | |
| "loss": 0.9681, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_loss": 0.9508872628211975, | |
| "eval_runtime": 46.7752, | |
| "eval_samples_per_second": 0.77, | |
| "eval_steps_per_second": 0.77, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 2.1176470588235294, | |
| "grad_norm": 0.3482557237148285, | |
| "learning_rate": 4.5e-05, | |
| "loss": 0.9542, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 2.3529411764705883, | |
| "grad_norm": 0.339621901512146, | |
| "learning_rate": 5e-05, | |
| "loss": 0.9026, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 2.588235294117647, | |
| "grad_norm": 0.42149099707603455, | |
| "learning_rate": 5.500000000000001e-05, | |
| "loss": 0.8801, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 2.8235294117647056, | |
| "grad_norm": 0.44959190487861633, | |
| "learning_rate": 6e-05, | |
| "loss": 0.8795, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 2.9411764705882355, | |
| "eval_loss": 0.8435218930244446, | |
| "eval_runtime": 46.764, | |
| "eval_samples_per_second": 0.77, | |
| "eval_steps_per_second": 0.77, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 3.0588235294117645, | |
| "grad_norm": 0.3488735258579254, | |
| "learning_rate": 6.500000000000001e-05, | |
| "loss": 0.8186, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 3.2941176470588234, | |
| "grad_norm": 0.24542245268821716, | |
| "learning_rate": 7e-05, | |
| "loss": 0.7967, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 3.5294117647058822, | |
| "grad_norm": 0.2552747130393982, | |
| "learning_rate": 7.500000000000001e-05, | |
| "loss": 0.765, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 3.764705882352941, | |
| "grad_norm": 0.22961358726024628, | |
| "learning_rate": 8e-05, | |
| "loss": 0.7888, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 0.20904940366744995, | |
| "learning_rate": 8.5e-05, | |
| "loss": 0.7349, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_loss": 0.7955379486083984, | |
| "eval_runtime": 46.7684, | |
| "eval_samples_per_second": 0.77, | |
| "eval_steps_per_second": 0.77, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 4.235294117647059, | |
| "grad_norm": 0.20834381878376007, | |
| "learning_rate": 9e-05, | |
| "loss": 0.7501, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 4.470588235294118, | |
| "grad_norm": 0.19870524108409882, | |
| "learning_rate": 9.5e-05, | |
| "loss": 0.7405, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 4.705882352941177, | |
| "grad_norm": 0.18882936239242554, | |
| "learning_rate": 0.0001, | |
| "loss": 0.7413, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 4.9411764705882355, | |
| "grad_norm": 0.21823865175247192, | |
| "learning_rate": 9.999238475781957e-05, | |
| "loss": 0.721, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 4.9411764705882355, | |
| "eval_loss": 0.7799076437950134, | |
| "eval_runtime": 46.7669, | |
| "eval_samples_per_second": 0.77, | |
| "eval_steps_per_second": 0.77, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 5.176470588235294, | |
| "grad_norm": 0.19386404752731323, | |
| "learning_rate": 9.99695413509548e-05, | |
| "loss": 0.6906, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 5.411764705882353, | |
| "grad_norm": 0.1739526093006134, | |
| "learning_rate": 9.99314767377287e-05, | |
| "loss": 0.7208, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 5.647058823529412, | |
| "grad_norm": 0.19978782534599304, | |
| "learning_rate": 9.987820251299122e-05, | |
| "loss": 0.6942, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 5.882352941176471, | |
| "grad_norm": 0.2283729910850525, | |
| "learning_rate": 9.980973490458728e-05, | |
| "loss": 0.7181, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_loss": 0.7700158357620239, | |
| "eval_runtime": 46.7722, | |
| "eval_samples_per_second": 0.77, | |
| "eval_steps_per_second": 0.77, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 6.117647058823529, | |
| "grad_norm": 0.19452933967113495, | |
| "learning_rate": 9.972609476841367e-05, | |
| "loss": 0.6724, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 6.352941176470588, | |
| "grad_norm": 0.20678135752677917, | |
| "learning_rate": 9.962730758206611e-05, | |
| "loss": 0.6501, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 6.588235294117647, | |
| "grad_norm": 0.22347362339496613, | |
| "learning_rate": 9.951340343707852e-05, | |
| "loss": 0.6705, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 6.823529411764706, | |
| "grad_norm": 0.2423417866230011, | |
| "learning_rate": 9.938441702975689e-05, | |
| "loss": 0.7032, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 6.9411764705882355, | |
| "eval_loss": 0.7669743299484253, | |
| "eval_runtime": 46.7781, | |
| "eval_samples_per_second": 0.77, | |
| "eval_steps_per_second": 0.77, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 7.0588235294117645, | |
| "grad_norm": 0.2525961101055145, | |
| "learning_rate": 9.924038765061042e-05, | |
| "loss": 0.6666, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 7.294117647058823, | |
| "grad_norm": 0.2671414911746979, | |
| "learning_rate": 9.908135917238321e-05, | |
| "loss": 0.6424, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 7.529411764705882, | |
| "grad_norm": 0.3437277674674988, | |
| "learning_rate": 9.890738003669029e-05, | |
| "loss": 0.6145, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 7.764705882352941, | |
| "grad_norm": 0.2997751533985138, | |
| "learning_rate": 9.871850323926177e-05, | |
| "loss": 0.6099, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "grad_norm": 0.3468629717826843, | |
| "learning_rate": 9.851478631379982e-05, | |
| "loss": 0.6292, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_loss": 0.773787796497345, | |
| "eval_runtime": 46.7904, | |
| "eval_samples_per_second": 0.769, | |
| "eval_steps_per_second": 0.769, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 8.235294117647058, | |
| "grad_norm": 0.3100046217441559, | |
| "learning_rate": 9.829629131445342e-05, | |
| "loss": 0.5788, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 8.470588235294118, | |
| "grad_norm": 0.3753797709941864, | |
| "learning_rate": 9.806308479691595e-05, | |
| "loss": 0.5965, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 8.705882352941176, | |
| "grad_norm": 0.33154118061065674, | |
| "learning_rate": 9.781523779815179e-05, | |
| "loss": 0.5952, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 8.941176470588236, | |
| "grad_norm": 0.38849934935569763, | |
| "learning_rate": 9.755282581475769e-05, | |
| "loss": 0.5437, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 8.941176470588236, | |
| "eval_loss": 0.7969802021980286, | |
| "eval_runtime": 46.7758, | |
| "eval_samples_per_second": 0.77, | |
| "eval_steps_per_second": 0.77, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 9.176470588235293, | |
| "grad_norm": 0.39084765315055847, | |
| "learning_rate": 9.727592877996585e-05, | |
| "loss": 0.5428, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 9.411764705882353, | |
| "grad_norm": 0.4236640930175781, | |
| "learning_rate": 9.698463103929542e-05, | |
| "loss": 0.5392, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 9.647058823529411, | |
| "grad_norm": 0.4983026683330536, | |
| "learning_rate": 9.667902132486009e-05, | |
| "loss": 0.5229, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 9.882352941176471, | |
| "grad_norm": 0.5346341133117676, | |
| "learning_rate": 9.635919272833938e-05, | |
| "loss": 0.5035, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "eval_loss": 0.821323573589325, | |
| "eval_runtime": 46.7877, | |
| "eval_samples_per_second": 0.769, | |
| "eval_steps_per_second": 0.769, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 10.117647058823529, | |
| "grad_norm": 0.5391852259635925, | |
| "learning_rate": 9.602524267262203e-05, | |
| "loss": 0.4754, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 10.352941176470589, | |
| "grad_norm": 0.8235352635383606, | |
| "learning_rate": 9.567727288213005e-05, | |
| "loss": 0.484, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 10.588235294117647, | |
| "grad_norm": 0.5596646666526794, | |
| "learning_rate": 9.53153893518325e-05, | |
| "loss": 0.4491, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 10.823529411764707, | |
| "grad_norm": 0.7218156456947327, | |
| "learning_rate": 9.493970231495835e-05, | |
| "loss": 0.4565, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 10.941176470588236, | |
| "eval_loss": 0.8528595566749573, | |
| "eval_runtime": 46.7845, | |
| "eval_samples_per_second": 0.769, | |
| "eval_steps_per_second": 0.769, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 11.058823529411764, | |
| "grad_norm": 0.728357195854187, | |
| "learning_rate": 9.45503262094184e-05, | |
| "loss": 0.4766, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 11.294117647058824, | |
| "grad_norm": 0.7212379574775696, | |
| "learning_rate": 9.414737964294636e-05, | |
| "loss": 0.4424, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 11.529411764705882, | |
| "grad_norm": 0.7260857820510864, | |
| "learning_rate": 9.373098535696979e-05, | |
| "loss": 0.3814, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 11.764705882352942, | |
| "grad_norm": 0.818840503692627, | |
| "learning_rate": 9.330127018922194e-05, | |
| "loss": 0.4105, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "grad_norm": 0.6266764998435974, | |
| "learning_rate": 9.285836503510562e-05, | |
| "loss": 0.3822, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "eval_loss": 0.8877852559089661, | |
| "eval_runtime": 46.7752, | |
| "eval_samples_per_second": 0.77, | |
| "eval_steps_per_second": 0.77, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 12.235294117647058, | |
| "grad_norm": 0.853114664554596, | |
| "learning_rate": 9.24024048078213e-05, | |
| "loss": 0.3609, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 12.470588235294118, | |
| "grad_norm": 0.6987264752388, | |
| "learning_rate": 9.193352839727121e-05, | |
| "loss": 0.3427, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 12.705882352941176, | |
| "grad_norm": 0.8503619432449341, | |
| "learning_rate": 9.145187862775209e-05, | |
| "loss": 0.3577, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 12.941176470588236, | |
| "grad_norm": 0.6981789469718933, | |
| "learning_rate": 9.09576022144496e-05, | |
| "loss": 0.3563, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 12.941176470588236, | |
| "eval_loss": 0.9017508625984192, | |
| "eval_runtime": 46.7688, | |
| "eval_samples_per_second": 0.77, | |
| "eval_steps_per_second": 0.77, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 13.176470588235293, | |
| "grad_norm": 0.6991474032402039, | |
| "learning_rate": 9.045084971874738e-05, | |
| "loss": 0.3363, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 13.411764705882353, | |
| "grad_norm": 0.9836378693580627, | |
| "learning_rate": 8.993177550236464e-05, | |
| "loss": 0.2837, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 13.647058823529411, | |
| "grad_norm": 0.844856321811676, | |
| "learning_rate": 8.940053768033609e-05, | |
| "loss": 0.3085, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 13.882352941176471, | |
| "grad_norm": 0.8222668170928955, | |
| "learning_rate": 8.885729807284856e-05, | |
| "loss": 0.3062, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 14.0, | |
| "eval_loss": 0.9818692803382874, | |
| "eval_runtime": 46.7826, | |
| "eval_samples_per_second": 0.77, | |
| "eval_steps_per_second": 0.77, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 14.0, | |
| "step": 119, | |
| "total_flos": 2.3291782290315674e+17, | |
| "train_loss": 0.6533061255927847, | |
| "train_runtime": 8413.199, | |
| "train_samples_per_second": 0.808, | |
| "train_steps_per_second": 0.048 | |
| } | |
| ], | |
| "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": 2.3291782290315674e+17, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |