| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 20.0, | |
| "eval_steps": 500, | |
| "global_step": 5000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.4, | |
| "learning_rate": 5e-05, | |
| "loss": 2.1914, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "learning_rate": 5e-05, | |
| "loss": 1.798, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.6067341772151899, | |
| "eval_loss": 1.8213441371917725, | |
| "eval_runtime": 4.582, | |
| "eval_samples_per_second": 109.122, | |
| "eval_steps_per_second": 13.749, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 1.2, | |
| "learning_rate": 5e-05, | |
| "loss": 1.7534, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 1.6, | |
| "learning_rate": 5e-05, | |
| "loss": 1.7163, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "learning_rate": 5e-05, | |
| "loss": 1.7, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.6077215189873417, | |
| "eval_loss": 1.8046445846557617, | |
| "eval_runtime": 5.1239, | |
| "eval_samples_per_second": 97.582, | |
| "eval_steps_per_second": 12.295, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 2.4, | |
| "learning_rate": 5e-05, | |
| "loss": 1.5854, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 2.8, | |
| "learning_rate": 5e-05, | |
| "loss": 1.5869, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.6071392405063291, | |
| "eval_loss": 1.829284906387329, | |
| "eval_runtime": 5.0393, | |
| "eval_samples_per_second": 99.221, | |
| "eval_steps_per_second": 12.502, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 3.2, | |
| "learning_rate": 5e-05, | |
| "loss": 1.5039, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 3.6, | |
| "learning_rate": 5e-05, | |
| "loss": 1.4165, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "learning_rate": 5e-05, | |
| "loss": 1.4349, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.6042784810126582, | |
| "eval_loss": 1.89736008644104, | |
| "eval_runtime": 4.5584, | |
| "eval_samples_per_second": 109.688, | |
| "eval_steps_per_second": 13.821, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 4.4, | |
| "learning_rate": 5e-05, | |
| "loss": 1.2624, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 4.8, | |
| "learning_rate": 5e-05, | |
| "loss": 1.3111, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_accuracy": 0.601493670886076, | |
| "eval_loss": 1.9769378900527954, | |
| "eval_runtime": 4.4063, | |
| "eval_samples_per_second": 113.474, | |
| "eval_steps_per_second": 14.298, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 5.2, | |
| "learning_rate": 5e-05, | |
| "loss": 1.226, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 5.6, | |
| "learning_rate": 5e-05, | |
| "loss": 1.1608, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "learning_rate": 5e-05, | |
| "loss": 1.197, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_accuracy": 0.599240506329114, | |
| "eval_loss": 2.0634872913360596, | |
| "eval_runtime": 4.8833, | |
| "eval_samples_per_second": 102.389, | |
| "eval_steps_per_second": 12.901, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 6.4, | |
| "learning_rate": 5e-05, | |
| "loss": 1.0336, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 6.8, | |
| "learning_rate": 5e-05, | |
| "loss": 1.0729, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_accuracy": 0.5975443037974684, | |
| "eval_loss": 2.1523237228393555, | |
| "eval_runtime": 4.8877, | |
| "eval_samples_per_second": 102.297, | |
| "eval_steps_per_second": 12.889, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 7.2, | |
| "learning_rate": 5e-05, | |
| "loss": 1.0005, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 7.6, | |
| "learning_rate": 5e-05, | |
| "loss": 0.9437, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "learning_rate": 5e-05, | |
| "loss": 0.9833, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_accuracy": 0.5947341772151898, | |
| "eval_loss": 2.2639756202697754, | |
| "eval_runtime": 4.3964, | |
| "eval_samples_per_second": 113.73, | |
| "eval_steps_per_second": 14.33, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 8.4, | |
| "learning_rate": 5e-05, | |
| "loss": 0.823, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 8.8, | |
| "learning_rate": 5e-05, | |
| "loss": 0.8672, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "eval_accuracy": 0.5924050632911393, | |
| "eval_loss": 2.3642578125, | |
| "eval_runtime": 4.6579, | |
| "eval_samples_per_second": 107.345, | |
| "eval_steps_per_second": 13.526, | |
| "step": 2250 | |
| }, | |
| { | |
| "epoch": 9.2, | |
| "learning_rate": 5e-05, | |
| "loss": 0.8032, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 9.6, | |
| "learning_rate": 5e-05, | |
| "loss": 0.7566, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "learning_rate": 5e-05, | |
| "loss": 0.7883, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "eval_accuracy": 0.5908101265822785, | |
| "eval_loss": 2.4598007202148438, | |
| "eval_runtime": 4.6476, | |
| "eval_samples_per_second": 107.582, | |
| "eval_steps_per_second": 13.555, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 10.4, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6488, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 10.8, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6879, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 11.0, | |
| "eval_accuracy": 0.588987341772152, | |
| "eval_loss": 2.5668857097625732, | |
| "eval_runtime": 4.4085, | |
| "eval_samples_per_second": 113.416, | |
| "eval_steps_per_second": 14.29, | |
| "step": 2750 | |
| }, | |
| { | |
| "epoch": 11.2, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6502, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 11.6, | |
| "learning_rate": 5e-05, | |
| "loss": 0.602, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6295, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "eval_accuracy": 0.5885316455696202, | |
| "eval_loss": 2.700011730194092, | |
| "eval_runtime": 4.8719, | |
| "eval_samples_per_second": 102.63, | |
| "eval_steps_per_second": 12.931, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 12.4, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5228, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 12.8, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5545, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 13.0, | |
| "eval_accuracy": 0.5850886075949367, | |
| "eval_loss": 2.8281185626983643, | |
| "eval_runtime": 4.4033, | |
| "eval_samples_per_second": 113.551, | |
| "eval_steps_per_second": 14.307, | |
| "step": 3250 | |
| }, | |
| { | |
| "epoch": 13.2, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5244, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 13.6, | |
| "learning_rate": 5e-05, | |
| "loss": 0.4924, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 14.0, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5208, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 14.0, | |
| "eval_accuracy": 0.5852911392405064, | |
| "eval_loss": 2.879397392272949, | |
| "eval_runtime": 4.703, | |
| "eval_samples_per_second": 106.315, | |
| "eval_steps_per_second": 13.396, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 14.4, | |
| "learning_rate": 5e-05, | |
| "loss": 0.4405, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 14.8, | |
| "learning_rate": 5e-05, | |
| "loss": 0.4679, | |
| "step": 3700 | |
| }, | |
| { | |
| "epoch": 15.0, | |
| "eval_accuracy": 0.586253164556962, | |
| "eval_loss": 2.9183950424194336, | |
| "eval_runtime": 4.5629, | |
| "eval_samples_per_second": 109.579, | |
| "eval_steps_per_second": 13.807, | |
| "step": 3750 | |
| }, | |
| { | |
| "epoch": 15.2, | |
| "learning_rate": 5e-05, | |
| "loss": 0.4443, | |
| "step": 3800 | |
| }, | |
| { | |
| "epoch": 15.6, | |
| "learning_rate": 5e-05, | |
| "loss": 0.4235, | |
| "step": 3900 | |
| }, | |
| { | |
| "epoch": 16.0, | |
| "learning_rate": 5e-05, | |
| "loss": 0.4464, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 16.0, | |
| "eval_accuracy": 0.5852151898734177, | |
| "eval_loss": 3.0790698528289795, | |
| "eval_runtime": 4.6021, | |
| "eval_samples_per_second": 108.647, | |
| "eval_steps_per_second": 13.689, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 16.4, | |
| "learning_rate": 5e-05, | |
| "loss": 0.3919, | |
| "step": 4100 | |
| }, | |
| { | |
| "epoch": 16.8, | |
| "learning_rate": 5e-05, | |
| "loss": 0.4136, | |
| "step": 4200 | |
| }, | |
| { | |
| "epoch": 17.0, | |
| "eval_accuracy": 0.5855696202531645, | |
| "eval_loss": 3.0832109451293945, | |
| "eval_runtime": 5.1254, | |
| "eval_samples_per_second": 97.553, | |
| "eval_steps_per_second": 12.292, | |
| "step": 4250 | |
| }, | |
| { | |
| "epoch": 17.2, | |
| "learning_rate": 5e-05, | |
| "loss": 0.3902, | |
| "step": 4300 | |
| }, | |
| { | |
| "epoch": 17.6, | |
| "learning_rate": 5e-05, | |
| "loss": 0.3841, | |
| "step": 4400 | |
| }, | |
| { | |
| "epoch": 18.0, | |
| "learning_rate": 5e-05, | |
| "loss": 0.4021, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 18.0, | |
| "eval_accuracy": 0.5846835443037974, | |
| "eval_loss": 3.0944228172302246, | |
| "eval_runtime": 4.5552, | |
| "eval_samples_per_second": 109.765, | |
| "eval_steps_per_second": 13.83, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 18.4, | |
| "learning_rate": 5e-05, | |
| "loss": 0.3577, | |
| "step": 4600 | |
| }, | |
| { | |
| "epoch": 18.8, | |
| "learning_rate": 5e-05, | |
| "loss": 0.3776, | |
| "step": 4700 | |
| }, | |
| { | |
| "epoch": 19.0, | |
| "eval_accuracy": 0.5828354430379746, | |
| "eval_loss": 3.2120306491851807, | |
| "eval_runtime": 4.7023, | |
| "eval_samples_per_second": 106.332, | |
| "eval_steps_per_second": 13.398, | |
| "step": 4750 | |
| }, | |
| { | |
| "epoch": 19.2, | |
| "learning_rate": 5e-05, | |
| "loss": 0.3617, | |
| "step": 4800 | |
| }, | |
| { | |
| "epoch": 19.6, | |
| "learning_rate": 5e-05, | |
| "loss": 0.3575, | |
| "step": 4900 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "learning_rate": 5e-05, | |
| "loss": 0.373, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "eval_accuracy": 0.5839240506329114, | |
| "eval_loss": 3.229841947555542, | |
| "eval_runtime": 4.7155, | |
| "eval_samples_per_second": 106.033, | |
| "eval_steps_per_second": 13.36, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "step": 5000, | |
| "total_flos": 3.1967425075347456e+17, | |
| "train_loss": 0.8756309669494629, | |
| "train_runtime": 3519.8357, | |
| "train_samples_per_second": 45.457, | |
| "train_steps_per_second": 1.421 | |
| } | |
| ], | |
| "logging_steps": 100, | |
| "max_steps": 5000, | |
| "num_train_epochs": 20, | |
| "save_steps": 500, | |
| "total_flos": 3.1967425075347456e+17, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |