| { | |
| "best_metric": 0.7160926461219788, | |
| "best_model_checkpoint": "final_models/glue_models/structroberta_s1_final//finetune/mnli/checkpoint-8000", | |
| "epoch": 4.618937644341801, | |
| "global_step": 10000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.18, | |
| "eval_accuracy": 0.5429747104644775, | |
| "eval_loss": 0.9390358924865723, | |
| "eval_runtime": 12.8367, | |
| "eval_samples_per_second": 511.192, | |
| "eval_steps_per_second": 63.957, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "learning_rate": 4.884526558891455e-05, | |
| "loss": 0.9969, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.37, | |
| "eval_accuracy": 0.584273099899292, | |
| "eval_loss": 0.8898679614067078, | |
| "eval_runtime": 12.8403, | |
| "eval_samples_per_second": 511.046, | |
| "eval_steps_per_second": 63.939, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.46, | |
| "learning_rate": 4.7690531177829104e-05, | |
| "loss": 0.8987, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.55, | |
| "eval_accuracy": 0.6351721882820129, | |
| "eval_loss": 0.8180076479911804, | |
| "eval_runtime": 12.8474, | |
| "eval_samples_per_second": 510.766, | |
| "eval_steps_per_second": 63.904, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 0.69, | |
| "learning_rate": 4.653579676674365e-05, | |
| "loss": 0.8315, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.74, | |
| "eval_accuracy": 0.6568119525909424, | |
| "eval_loss": 0.7814252972602844, | |
| "eval_runtime": 12.834, | |
| "eval_samples_per_second": 511.298, | |
| "eval_steps_per_second": 63.971, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 0.92, | |
| "learning_rate": 4.53810623556582e-05, | |
| "loss": 0.7832, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.92, | |
| "eval_accuracy": 0.6636695861816406, | |
| "eval_loss": 0.7712134718894958, | |
| "eval_runtime": 12.8417, | |
| "eval_samples_per_second": 510.991, | |
| "eval_steps_per_second": 63.932, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 1.11, | |
| "eval_accuracy": 0.6700701117515564, | |
| "eval_loss": 0.7608532905578613, | |
| "eval_runtime": 12.8417, | |
| "eval_samples_per_second": 510.991, | |
| "eval_steps_per_second": 63.932, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 1.15, | |
| "learning_rate": 4.422632794457275e-05, | |
| "loss": 0.7252, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 1.29, | |
| "eval_accuracy": 0.6767753958702087, | |
| "eval_loss": 0.7573378086090088, | |
| "eval_runtime": 12.8419, | |
| "eval_samples_per_second": 510.983, | |
| "eval_steps_per_second": 63.931, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 1.39, | |
| "learning_rate": 4.30715935334873e-05, | |
| "loss": 0.6858, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 1.48, | |
| "eval_accuracy": 0.6862236857414246, | |
| "eval_loss": 0.735822856426239, | |
| "eval_runtime": 12.8397, | |
| "eval_samples_per_second": 511.073, | |
| "eval_steps_per_second": 63.942, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 1.62, | |
| "learning_rate": 4.1916859122401844e-05, | |
| "loss": 0.6767, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 1.66, | |
| "eval_accuracy": 0.6981103420257568, | |
| "eval_loss": 0.7118889093399048, | |
| "eval_runtime": 12.8424, | |
| "eval_samples_per_second": 510.962, | |
| "eval_steps_per_second": 63.929, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 1.85, | |
| "learning_rate": 4.07621247113164e-05, | |
| "loss": 0.6719, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 1.85, | |
| "eval_accuracy": 0.6932337880134583, | |
| "eval_loss": 0.734716534614563, | |
| "eval_runtime": 12.8464, | |
| "eval_samples_per_second": 510.805, | |
| "eval_steps_per_second": 63.909, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 2.03, | |
| "eval_accuracy": 0.7067967057228088, | |
| "eval_loss": 0.7124155163764954, | |
| "eval_runtime": 12.8351, | |
| "eval_samples_per_second": 511.256, | |
| "eval_steps_per_second": 63.965, | |
| "step": 4400 | |
| }, | |
| { | |
| "epoch": 2.08, | |
| "learning_rate": 3.960739030023095e-05, | |
| "loss": 0.6358, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 2.22, | |
| "eval_accuracy": 0.7046632170677185, | |
| "eval_loss": 0.7168723344802856, | |
| "eval_runtime": 12.8893, | |
| "eval_samples_per_second": 509.103, | |
| "eval_steps_per_second": 63.696, | |
| "step": 4800 | |
| }, | |
| { | |
| "epoch": 2.31, | |
| "learning_rate": 3.84526558891455e-05, | |
| "loss": 0.5673, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 2.4, | |
| "eval_accuracy": 0.7063395380973816, | |
| "eval_loss": 0.7214012145996094, | |
| "eval_runtime": 12.8644, | |
| "eval_samples_per_second": 510.09, | |
| "eval_steps_per_second": 63.82, | |
| "step": 5200 | |
| }, | |
| { | |
| "epoch": 2.54, | |
| "learning_rate": 3.729792147806005e-05, | |
| "loss": 0.5681, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 2.59, | |
| "eval_accuracy": 0.7023773193359375, | |
| "eval_loss": 0.7392827272415161, | |
| "eval_runtime": 12.8544, | |
| "eval_samples_per_second": 510.488, | |
| "eval_steps_per_second": 63.869, | |
| "step": 5600 | |
| }, | |
| { | |
| "epoch": 2.77, | |
| "learning_rate": 3.61431870669746e-05, | |
| "loss": 0.5741, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 2.77, | |
| "eval_accuracy": 0.7077110409736633, | |
| "eval_loss": 0.7138980627059937, | |
| "eval_runtime": 12.8474, | |
| "eval_samples_per_second": 510.763, | |
| "eval_steps_per_second": 63.904, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 2.96, | |
| "eval_accuracy": 0.7118256688117981, | |
| "eval_loss": 0.6938605904579163, | |
| "eval_runtime": 12.8504, | |
| "eval_samples_per_second": 510.647, | |
| "eval_steps_per_second": 63.889, | |
| "step": 6400 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "learning_rate": 3.498845265588915e-05, | |
| "loss": 0.5723, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 3.14, | |
| "eval_accuracy": 0.7093873620033264, | |
| "eval_loss": 0.7904877662658691, | |
| "eval_runtime": 12.8645, | |
| "eval_samples_per_second": 510.087, | |
| "eval_steps_per_second": 63.819, | |
| "step": 6800 | |
| }, | |
| { | |
| "epoch": 3.23, | |
| "learning_rate": 3.38337182448037e-05, | |
| "loss": 0.4584, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 3.33, | |
| "eval_accuracy": 0.7139591574668884, | |
| "eval_loss": 0.7631048560142517, | |
| "eval_runtime": 12.84, | |
| "eval_samples_per_second": 511.061, | |
| "eval_steps_per_second": 63.941, | |
| "step": 7200 | |
| }, | |
| { | |
| "epoch": 3.46, | |
| "learning_rate": 3.2678983833718243e-05, | |
| "loss": 0.4606, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 3.51, | |
| "eval_accuracy": 0.7107589244842529, | |
| "eval_loss": 0.7649423480033875, | |
| "eval_runtime": 12.8337, | |
| "eval_samples_per_second": 511.31, | |
| "eval_steps_per_second": 63.972, | |
| "step": 7600 | |
| }, | |
| { | |
| "epoch": 3.7, | |
| "learning_rate": 3.1524249422632794e-05, | |
| "loss": 0.4719, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 3.7, | |
| "eval_accuracy": 0.7160926461219788, | |
| "eval_loss": 0.7716315388679504, | |
| "eval_runtime": 12.849, | |
| "eval_samples_per_second": 510.7, | |
| "eval_steps_per_second": 63.896, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 3.88, | |
| "eval_accuracy": 0.7092350125312805, | |
| "eval_loss": 0.7540602087974548, | |
| "eval_runtime": 12.851, | |
| "eval_samples_per_second": 510.623, | |
| "eval_steps_per_second": 63.886, | |
| "step": 8400 | |
| }, | |
| { | |
| "epoch": 3.93, | |
| "learning_rate": 3.0369515011547345e-05, | |
| "loss": 0.4809, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 4.06, | |
| "eval_accuracy": 0.7098445892333984, | |
| "eval_loss": 0.8660436868667603, | |
| "eval_runtime": 12.8552, | |
| "eval_samples_per_second": 510.455, | |
| "eval_steps_per_second": 63.865, | |
| "step": 8800 | |
| }, | |
| { | |
| "epoch": 4.16, | |
| "learning_rate": 2.9214780600461896e-05, | |
| "loss": 0.4002, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 4.25, | |
| "eval_accuracy": 0.7133495807647705, | |
| "eval_loss": 0.8522323369979858, | |
| "eval_runtime": 12.8555, | |
| "eval_samples_per_second": 510.443, | |
| "eval_steps_per_second": 63.864, | |
| "step": 9200 | |
| }, | |
| { | |
| "epoch": 4.39, | |
| "learning_rate": 2.8060046189376443e-05, | |
| "loss": 0.3682, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 4.43, | |
| "eval_accuracy": 0.7069491147994995, | |
| "eval_loss": 0.8781996965408325, | |
| "eval_runtime": 12.8756, | |
| "eval_samples_per_second": 509.647, | |
| "eval_steps_per_second": 63.764, | |
| "step": 9600 | |
| }, | |
| { | |
| "epoch": 4.62, | |
| "learning_rate": 2.6905311778290994e-05, | |
| "loss": 0.373, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 4.62, | |
| "eval_accuracy": 0.7087777853012085, | |
| "eval_loss": 0.8231818675994873, | |
| "eval_runtime": 12.8953, | |
| "eval_samples_per_second": 508.867, | |
| "eval_steps_per_second": 63.667, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 4.62, | |
| "step": 10000, | |
| "total_flos": 1.0013297089714176e+17, | |
| "train_loss": 0.6100377502441406, | |
| "train_runtime": 5918.4858, | |
| "train_samples_per_second": 438.93, | |
| "train_steps_per_second": 3.658 | |
| } | |
| ], | |
| "max_steps": 21650, | |
| "num_train_epochs": 10, | |
| "total_flos": 1.0013297089714176e+17, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |