| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 3.0, | |
| "eval_steps": 10, | |
| "global_step": 81, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.04, | |
| "learning_rate": 0.0, | |
| "loss": 1.0172, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "learning_rate": 3.1546487678572864e-05, | |
| "loss": 2.3198, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "learning_rate": 5e-05, | |
| "loss": 2.4795, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "learning_rate": 5e-05, | |
| "loss": 1.551, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "learning_rate": 5e-05, | |
| "loss": 1.2034, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.22, | |
| "learning_rate": 5e-05, | |
| "loss": 1.0527, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "learning_rate": 5e-05, | |
| "loss": 0.9743, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "learning_rate": 5e-05, | |
| "loss": 0.9738, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.33, | |
| "learning_rate": 5e-05, | |
| "loss": 0.939, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.37, | |
| "learning_rate": 5e-05, | |
| "loss": 0.955, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.37, | |
| "eval_accuracy": 0.2181755564165483, | |
| "eval_loss": 0.9061336517333984, | |
| "eval_runtime": 44.217, | |
| "eval_samples_per_second": 0.814, | |
| "eval_steps_per_second": 0.045, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.41, | |
| "learning_rate": 5e-05, | |
| "loss": 0.9513, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.44, | |
| "learning_rate": 5e-05, | |
| "loss": 0.94, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "learning_rate": 5e-05, | |
| "loss": 0.9211, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.52, | |
| "learning_rate": 5e-05, | |
| "loss": 0.8762, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.56, | |
| "learning_rate": 5e-05, | |
| "loss": 0.9013, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.59, | |
| "learning_rate": 5e-05, | |
| "loss": 0.8758, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.63, | |
| "learning_rate": 5e-05, | |
| "loss": 0.8919, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.67, | |
| "learning_rate": 5e-05, | |
| "loss": 0.8589, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "learning_rate": 5e-05, | |
| "loss": 0.8775, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.74, | |
| "learning_rate": 5e-05, | |
| "loss": 0.8899, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.74, | |
| "eval_accuracy": 0.21962848163933016, | |
| "eval_loss": 0.8734426498413086, | |
| "eval_runtime": 44.8031, | |
| "eval_samples_per_second": 0.804, | |
| "eval_steps_per_second": 0.045, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.78, | |
| "learning_rate": 5e-05, | |
| "loss": 0.9011, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.81, | |
| "learning_rate": 5e-05, | |
| "loss": 0.9195, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.85, | |
| "learning_rate": 5e-05, | |
| "loss": 0.815, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.89, | |
| "learning_rate": 5e-05, | |
| "loss": 0.8216, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.93, | |
| "learning_rate": 5e-05, | |
| "loss": 0.8559, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "learning_rate": 5e-05, | |
| "loss": 0.9071, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6228, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 1.04, | |
| "learning_rate": 5e-05, | |
| "loss": 0.3934, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 1.07, | |
| "learning_rate": 5e-05, | |
| "loss": 0.8846, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 1.11, | |
| "learning_rate": 5e-05, | |
| "loss": 0.879, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 1.11, | |
| "eval_accuracy": 0.2173899005553403, | |
| "eval_loss": 0.9090863466262817, | |
| "eval_runtime": 45.3228, | |
| "eval_samples_per_second": 0.794, | |
| "eval_steps_per_second": 0.044, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 1.15, | |
| "learning_rate": 5e-05, | |
| "loss": 0.8277, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 1.19, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6593, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 1.22, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5255, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 1.26, | |
| "learning_rate": 5e-05, | |
| "loss": 0.4102, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 1.3, | |
| "learning_rate": 5e-05, | |
| "loss": 0.5839, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 1.33, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6197, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 1.37, | |
| "learning_rate": 5e-05, | |
| "loss": 0.4026, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 1.41, | |
| "learning_rate": 5e-05, | |
| "loss": 0.3835, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 1.44, | |
| "learning_rate": 5e-05, | |
| "loss": 0.3433, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 1.48, | |
| "learning_rate": 5e-05, | |
| "loss": 0.3295, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 1.48, | |
| "eval_accuracy": 0.21730380128287916, | |
| "eval_loss": 0.9803322553634644, | |
| "eval_runtime": 44.8946, | |
| "eval_samples_per_second": 0.802, | |
| "eval_steps_per_second": 0.045, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 1.52, | |
| "learning_rate": 5e-05, | |
| "loss": 0.3101, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 1.56, | |
| "learning_rate": 5e-05, | |
| "loss": 0.3492, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 1.59, | |
| "learning_rate": 5e-05, | |
| "loss": 0.2991, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 1.63, | |
| "learning_rate": 5e-05, | |
| "loss": 0.3412, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 1.67, | |
| "learning_rate": 5e-05, | |
| "loss": 0.3824, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 1.7, | |
| "learning_rate": 5e-05, | |
| "loss": 0.3694, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 1.74, | |
| "learning_rate": 5e-05, | |
| "loss": 0.3926, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 1.78, | |
| "learning_rate": 5e-05, | |
| "loss": 0.3993, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 1.81, | |
| "learning_rate": 5e-05, | |
| "loss": 0.3678, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 1.85, | |
| "learning_rate": 5e-05, | |
| "loss": 0.3711, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 1.85, | |
| "eval_accuracy": 0.2173576133281674, | |
| "eval_loss": 0.9820255637168884, | |
| "eval_runtime": 44.7843, | |
| "eval_samples_per_second": 0.804, | |
| "eval_steps_per_second": 0.045, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 1.89, | |
| "learning_rate": 5e-05, | |
| "loss": 0.3825, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 1.93, | |
| "learning_rate": 5e-05, | |
| "loss": 0.3914, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 1.96, | |
| "learning_rate": 5e-05, | |
| "loss": 0.3962, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "learning_rate": 5e-05, | |
| "loss": 0.2342, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 2.04, | |
| "learning_rate": 5e-05, | |
| "loss": 0.1572, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 2.07, | |
| "learning_rate": 5e-05, | |
| "loss": 0.4409, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 2.11, | |
| "learning_rate": 5e-05, | |
| "loss": 0.4072, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 2.15, | |
| "learning_rate": 5e-05, | |
| "loss": 0.4306, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 2.19, | |
| "learning_rate": 5e-05, | |
| "loss": 0.338, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 2.22, | |
| "learning_rate": 5e-05, | |
| "loss": 0.2927, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 2.22, | |
| "eval_accuracy": 0.2152804683800422, | |
| "eval_loss": 1.0269818305969238, | |
| "eval_runtime": 44.7279, | |
| "eval_samples_per_second": 0.805, | |
| "eval_steps_per_second": 0.045, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 2.26, | |
| "learning_rate": 5e-05, | |
| "loss": 0.3043, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 2.3, | |
| "learning_rate": 5e-05, | |
| "loss": 0.3134, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 2.33, | |
| "learning_rate": 5e-05, | |
| "loss": 0.2565, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 2.37, | |
| "learning_rate": 5e-05, | |
| "loss": 0.2333, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 2.41, | |
| "learning_rate": 5e-05, | |
| "loss": 0.2291, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 2.44, | |
| "learning_rate": 5e-05, | |
| "loss": 0.1783, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 2.48, | |
| "learning_rate": 5e-05, | |
| "loss": 0.2058, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 2.52, | |
| "learning_rate": 5e-05, | |
| "loss": 0.208, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 2.56, | |
| "learning_rate": 5e-05, | |
| "loss": 0.1745, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 2.59, | |
| "learning_rate": 5e-05, | |
| "loss": 0.1703, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 2.59, | |
| "eval_accuracy": 0.2131279865685135, | |
| "eval_loss": 1.0966472625732422, | |
| "eval_runtime": 44.8433, | |
| "eval_samples_per_second": 0.803, | |
| "eval_steps_per_second": 0.045, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 2.63, | |
| "learning_rate": 5e-05, | |
| "loss": 0.2235, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 2.67, | |
| "learning_rate": 5e-05, | |
| "loss": 0.2062, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 2.7, | |
| "learning_rate": 5e-05, | |
| "loss": 0.2094, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 2.74, | |
| "learning_rate": 5e-05, | |
| "loss": 0.2123, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 2.78, | |
| "learning_rate": 5e-05, | |
| "loss": 0.2144, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 2.81, | |
| "learning_rate": 5e-05, | |
| "loss": 0.2091, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 2.85, | |
| "learning_rate": 5e-05, | |
| "loss": 0.2116, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 2.89, | |
| "learning_rate": 5e-05, | |
| "loss": 0.2123, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 2.93, | |
| "learning_rate": 5e-05, | |
| "loss": 0.1891, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 2.96, | |
| "learning_rate": 5e-05, | |
| "loss": 0.2011, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 2.96, | |
| "eval_accuracy": 0.2144840501097766, | |
| "eval_loss": 1.1488189697265625, | |
| "eval_runtime": 44.8519, | |
| "eval_samples_per_second": 0.803, | |
| "eval_steps_per_second": 0.045, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "learning_rate": 5e-05, | |
| "loss": 0.1493, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "step": 81, | |
| "total_flos": 10840779079680.0, | |
| "train_loss": 0.5814771251178082, | |
| "train_runtime": 6697.7067, | |
| "train_samples_per_second": 0.377, | |
| "train_steps_per_second": 0.012 | |
| } | |
| ], | |
| "logging_steps": 1.0, | |
| "max_steps": 81, | |
| "num_train_epochs": 3, | |
| "save_steps": 500, | |
| "total_flos": 10840779079680.0, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |