| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 1.8896447467876039, | |
| "eval_steps": 500, | |
| "global_step": 15000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.06, | |
| "learning_rate": 0.000975, | |
| "loss": 0.9774, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "eval_loss": 0.8144938945770264, | |
| "eval_runtime": 14.6573, | |
| "eval_samples_per_second": 68.225, | |
| "eval_steps_per_second": 34.113, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "learning_rate": 0.00095, | |
| "loss": 0.9617, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "eval_loss": 0.8077166080474854, | |
| "eval_runtime": 15.1765, | |
| "eval_samples_per_second": 65.891, | |
| "eval_steps_per_second": 32.946, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "learning_rate": 0.000925, | |
| "loss": 0.911, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "eval_loss": 0.8064053058624268, | |
| "eval_runtime": 15.3268, | |
| "eval_samples_per_second": 65.245, | |
| "eval_steps_per_second": 32.623, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "learning_rate": 0.0009000000000000001, | |
| "loss": 0.8954, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "eval_loss": 0.8163686990737915, | |
| "eval_runtime": 14.5944, | |
| "eval_samples_per_second": 68.519, | |
| "eval_steps_per_second": 34.26, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "learning_rate": 0.000875, | |
| "loss": 0.883, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "eval_loss": 0.8068735003471375, | |
| "eval_runtime": 14.8614, | |
| "eval_samples_per_second": 67.288, | |
| "eval_steps_per_second": 33.644, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "learning_rate": 0.00085, | |
| "loss": 0.8867, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "eval_loss": 0.7819482088088989, | |
| "eval_runtime": 14.6896, | |
| "eval_samples_per_second": 68.076, | |
| "eval_steps_per_second": 34.038, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.44, | |
| "learning_rate": 0.000825, | |
| "loss": 0.8688, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 0.44, | |
| "eval_loss": 0.8062307238578796, | |
| "eval_runtime": 14.6856, | |
| "eval_samples_per_second": 68.094, | |
| "eval_steps_per_second": 34.047, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "learning_rate": 0.0008, | |
| "loss": 0.8446, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "eval_loss": 0.7707250714302063, | |
| "eval_runtime": 15.0517, | |
| "eval_samples_per_second": 66.438, | |
| "eval_steps_per_second": 33.219, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 0.57, | |
| "learning_rate": 0.0007750000000000001, | |
| "loss": 0.8617, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 0.57, | |
| "eval_loss": 0.7528353333473206, | |
| "eval_runtime": 14.7368, | |
| "eval_samples_per_second": 67.857, | |
| "eval_steps_per_second": 33.929, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 0.63, | |
| "learning_rate": 0.00075, | |
| "loss": 0.8158, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 0.63, | |
| "eval_loss": 0.7551385760307312, | |
| "eval_runtime": 15.3736, | |
| "eval_samples_per_second": 65.047, | |
| "eval_steps_per_second": 32.523, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 0.69, | |
| "learning_rate": 0.000725, | |
| "loss": 0.7889, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 0.69, | |
| "eval_loss": 0.7405046820640564, | |
| "eval_runtime": 15.4488, | |
| "eval_samples_per_second": 64.73, | |
| "eval_steps_per_second": 32.365, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 0.76, | |
| "learning_rate": 0.0007, | |
| "loss": 0.7992, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 0.76, | |
| "eval_loss": 0.7292428016662598, | |
| "eval_runtime": 15.892, | |
| "eval_samples_per_second": 62.925, | |
| "eval_steps_per_second": 31.462, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 0.82, | |
| "learning_rate": 0.000675, | |
| "loss": 0.8051, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 0.82, | |
| "eval_loss": 0.7345249056816101, | |
| "eval_runtime": 14.8049, | |
| "eval_samples_per_second": 67.545, | |
| "eval_steps_per_second": 33.773, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "learning_rate": 0.0006500000000000001, | |
| "loss": 0.7684, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "eval_loss": 0.7357723712921143, | |
| "eval_runtime": 14.7316, | |
| "eval_samples_per_second": 67.881, | |
| "eval_steps_per_second": 33.941, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 0.94, | |
| "learning_rate": 0.000625, | |
| "loss": 0.753, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 0.94, | |
| "eval_loss": 0.7323009371757507, | |
| "eval_runtime": 14.6239, | |
| "eval_samples_per_second": 68.381, | |
| "eval_steps_per_second": 34.191, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 1.01, | |
| "learning_rate": 0.0006, | |
| "loss": 0.7464, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 1.01, | |
| "eval_loss": 0.7275989651679993, | |
| "eval_runtime": 15.2815, | |
| "eval_samples_per_second": 65.439, | |
| "eval_steps_per_second": 32.719, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 1.07, | |
| "learning_rate": 0.000575, | |
| "loss": 0.5429, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 1.07, | |
| "eval_loss": 0.7231326103210449, | |
| "eval_runtime": 15.5099, | |
| "eval_samples_per_second": 64.475, | |
| "eval_steps_per_second": 32.238, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 1.13, | |
| "learning_rate": 0.00055, | |
| "loss": 0.5704, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 1.13, | |
| "eval_loss": 0.717272162437439, | |
| "eval_runtime": 14.9897, | |
| "eval_samples_per_second": 66.712, | |
| "eval_steps_per_second": 33.356, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 1.2, | |
| "learning_rate": 0.0005250000000000001, | |
| "loss": 0.5459, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 1.2, | |
| "eval_loss": 0.7188604474067688, | |
| "eval_runtime": 14.7366, | |
| "eval_samples_per_second": 67.858, | |
| "eval_steps_per_second": 33.929, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 1.26, | |
| "learning_rate": 0.0005, | |
| "loss": 0.5435, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 1.26, | |
| "eval_loss": 0.7037996053695679, | |
| "eval_runtime": 14.5588, | |
| "eval_samples_per_second": 68.687, | |
| "eval_steps_per_second": 34.343, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 1.32, | |
| "learning_rate": 0.000475, | |
| "loss": 0.5429, | |
| "step": 10500 | |
| }, | |
| { | |
| "epoch": 1.32, | |
| "eval_loss": 0.7000067234039307, | |
| "eval_runtime": 14.629, | |
| "eval_samples_per_second": 68.357, | |
| "eval_steps_per_second": 34.179, | |
| "step": 10500 | |
| }, | |
| { | |
| "epoch": 1.39, | |
| "learning_rate": 0.00045000000000000004, | |
| "loss": 0.5363, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 1.39, | |
| "eval_loss": 0.7090610861778259, | |
| "eval_runtime": 15.5146, | |
| "eval_samples_per_second": 64.455, | |
| "eval_steps_per_second": 32.228, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 1.45, | |
| "learning_rate": 0.000425, | |
| "loss": 0.551, | |
| "step": 11500 | |
| }, | |
| { | |
| "epoch": 1.45, | |
| "eval_loss": 0.6937999129295349, | |
| "eval_runtime": 15.2752, | |
| "eval_samples_per_second": 65.466, | |
| "eval_steps_per_second": 32.733, | |
| "step": 11500 | |
| }, | |
| { | |
| "epoch": 1.51, | |
| "learning_rate": 0.0004, | |
| "loss": 0.5345, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 1.51, | |
| "eval_loss": 0.6926913261413574, | |
| "eval_runtime": 14.5585, | |
| "eval_samples_per_second": 68.688, | |
| "eval_steps_per_second": 34.344, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 1.57, | |
| "learning_rate": 0.000375, | |
| "loss": 0.5519, | |
| "step": 12500 | |
| }, | |
| { | |
| "epoch": 1.57, | |
| "eval_loss": 0.6763409972190857, | |
| "eval_runtime": 14.6685, | |
| "eval_samples_per_second": 68.173, | |
| "eval_steps_per_second": 34.087, | |
| "step": 12500 | |
| }, | |
| { | |
| "epoch": 1.64, | |
| "learning_rate": 0.00035, | |
| "loss": 0.5324, | |
| "step": 13000 | |
| }, | |
| { | |
| "epoch": 1.64, | |
| "eval_loss": 0.6778369545936584, | |
| "eval_runtime": 17.3898, | |
| "eval_samples_per_second": 57.505, | |
| "eval_steps_per_second": 28.753, | |
| "step": 13000 | |
| }, | |
| { | |
| "epoch": 1.7, | |
| "learning_rate": 0.00032500000000000004, | |
| "loss": 0.5272, | |
| "step": 13500 | |
| }, | |
| { | |
| "epoch": 1.7, | |
| "eval_loss": 0.6725330948829651, | |
| "eval_runtime": 14.6869, | |
| "eval_samples_per_second": 68.088, | |
| "eval_steps_per_second": 34.044, | |
| "step": 13500 | |
| }, | |
| { | |
| "epoch": 1.76, | |
| "learning_rate": 0.0003, | |
| "loss": 0.5258, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 1.76, | |
| "eval_loss": 0.6668800115585327, | |
| "eval_runtime": 15.0231, | |
| "eval_samples_per_second": 66.564, | |
| "eval_steps_per_second": 33.282, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 1.83, | |
| "learning_rate": 0.000275, | |
| "loss": 0.5229, | |
| "step": 14500 | |
| }, | |
| { | |
| "epoch": 1.83, | |
| "eval_loss": 0.6599903106689453, | |
| "eval_runtime": 14.6661, | |
| "eval_samples_per_second": 68.184, | |
| "eval_steps_per_second": 34.092, | |
| "step": 14500 | |
| }, | |
| { | |
| "epoch": 1.89, | |
| "learning_rate": 0.00025, | |
| "loss": 0.5386, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 1.89, | |
| "eval_loss": 0.659939706325531, | |
| "eval_runtime": 14.6708, | |
| "eval_samples_per_second": 68.163, | |
| "eval_steps_per_second": 34.081, | |
| "step": 15000 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 20000, | |
| "num_train_epochs": 3, | |
| "save_steps": 5000, | |
| "total_flos": 9.418761150170112e+16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |