| { | |
| "best_metric": 0.72835773229599, | |
| "best_model_checkpoint": "final_models/glue_models/structroberta_s2_final//finetune/mnli-mm/checkpoint-6400", | |
| "epoch": 3.695150115473441, | |
| "global_step": 8000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.18, | |
| "eval_accuracy": 0.5728834867477417, | |
| "eval_loss": 0.9009019136428833, | |
| "eval_runtime": 12.2589, | |
| "eval_samples_per_second": 512.606, | |
| "eval_steps_per_second": 64.117, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "learning_rate": 4.884526558891455e-05, | |
| "loss": 0.986, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.37, | |
| "eval_accuracy": 0.6513367295265198, | |
| "eval_loss": 0.7975157499313354, | |
| "eval_runtime": 12.2206, | |
| "eval_samples_per_second": 514.215, | |
| "eval_steps_per_second": 64.318, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.46, | |
| "learning_rate": 4.7690531177829104e-05, | |
| "loss": 0.8465, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.55, | |
| "eval_accuracy": 0.6755251288414001, | |
| "eval_loss": 0.7505028247833252, | |
| "eval_runtime": 12.2225, | |
| "eval_samples_per_second": 514.134, | |
| "eval_steps_per_second": 64.308, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 0.69, | |
| "learning_rate": 4.653579676674365e-05, | |
| "loss": 0.7931, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.74, | |
| "eval_accuracy": 0.682208776473999, | |
| "eval_loss": 0.7299364805221558, | |
| "eval_runtime": 12.2349, | |
| "eval_samples_per_second": 513.615, | |
| "eval_steps_per_second": 64.243, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 0.92, | |
| "learning_rate": 4.53810623556582e-05, | |
| "loss": 0.7619, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.92, | |
| "eval_accuracy": 0.7059198021888733, | |
| "eval_loss": 0.6974023580551147, | |
| "eval_runtime": 12.2223, | |
| "eval_samples_per_second": 514.143, | |
| "eval_steps_per_second": 64.309, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 1.11, | |
| "eval_accuracy": 0.708625078201294, | |
| "eval_loss": 0.6957436203956604, | |
| "eval_runtime": 12.2195, | |
| "eval_samples_per_second": 514.259, | |
| "eval_steps_per_second": 64.323, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 1.15, | |
| "learning_rate": 4.422632794457275e-05, | |
| "loss": 0.7076, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 1.29, | |
| "eval_accuracy": 0.7118077874183655, | |
| "eval_loss": 0.6871402263641357, | |
| "eval_runtime": 12.2168, | |
| "eval_samples_per_second": 514.374, | |
| "eval_steps_per_second": 64.338, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 1.39, | |
| "learning_rate": 4.30715935334873e-05, | |
| "loss": 0.6707, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 1.48, | |
| "eval_accuracy": 0.713876485824585, | |
| "eval_loss": 0.6786448955535889, | |
| "eval_runtime": 12.228, | |
| "eval_samples_per_second": 513.903, | |
| "eval_steps_per_second": 64.279, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 1.62, | |
| "learning_rate": 4.1916859122401844e-05, | |
| "loss": 0.6685, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 1.66, | |
| "eval_accuracy": 0.7175366282463074, | |
| "eval_loss": 0.6627858281135559, | |
| "eval_runtime": 12.2175, | |
| "eval_samples_per_second": 514.343, | |
| "eval_steps_per_second": 64.334, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 1.85, | |
| "learning_rate": 4.07621247113164e-05, | |
| "loss": 0.6631, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 1.85, | |
| "eval_accuracy": 0.7172183394432068, | |
| "eval_loss": 0.6724812388420105, | |
| "eval_runtime": 12.2271, | |
| "eval_samples_per_second": 513.942, | |
| "eval_steps_per_second": 64.284, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 2.03, | |
| "eval_accuracy": 0.7267664074897766, | |
| "eval_loss": 0.6734680533409119, | |
| "eval_runtime": 12.2328, | |
| "eval_samples_per_second": 513.699, | |
| "eval_steps_per_second": 64.253, | |
| "step": 4400 | |
| }, | |
| { | |
| "epoch": 2.08, | |
| "learning_rate": 3.960739030023095e-05, | |
| "loss": 0.6259, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 2.22, | |
| "eval_accuracy": 0.7243793606758118, | |
| "eval_loss": 0.6886034607887268, | |
| "eval_runtime": 12.2355, | |
| "eval_samples_per_second": 513.587, | |
| "eval_steps_per_second": 64.239, | |
| "step": 4800 | |
| }, | |
| { | |
| "epoch": 2.31, | |
| "learning_rate": 3.84526558891455e-05, | |
| "loss": 0.5623, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 2.4, | |
| "eval_accuracy": 0.7200827598571777, | |
| "eval_loss": 0.6721735596656799, | |
| "eval_runtime": 12.2114, | |
| "eval_samples_per_second": 514.601, | |
| "eval_steps_per_second": 64.366, | |
| "step": 5200 | |
| }, | |
| { | |
| "epoch": 2.54, | |
| "learning_rate": 3.729792147806005e-05, | |
| "loss": 0.5625, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 2.59, | |
| "eval_accuracy": 0.722151517868042, | |
| "eval_loss": 0.6832764148712158, | |
| "eval_runtime": 12.2235, | |
| "eval_samples_per_second": 514.09, | |
| "eval_steps_per_second": 64.302, | |
| "step": 5600 | |
| }, | |
| { | |
| "epoch": 2.77, | |
| "learning_rate": 3.61431870669746e-05, | |
| "loss": 0.5726, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 2.77, | |
| "eval_accuracy": 0.7278803586959839, | |
| "eval_loss": 0.6740399599075317, | |
| "eval_runtime": 12.2331, | |
| "eval_samples_per_second": 513.686, | |
| "eval_steps_per_second": 64.252, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 2.96, | |
| "eval_accuracy": 0.72835773229599, | |
| "eval_loss": 0.6698330044746399, | |
| "eval_runtime": 12.2158, | |
| "eval_samples_per_second": 514.416, | |
| "eval_steps_per_second": 64.343, | |
| "step": 6400 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "learning_rate": 3.498845265588915e-05, | |
| "loss": 0.5712, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 3.14, | |
| "eval_accuracy": 0.7272437810897827, | |
| "eval_loss": 0.7273808717727661, | |
| "eval_runtime": 12.2209, | |
| "eval_samples_per_second": 514.201, | |
| "eval_steps_per_second": 64.316, | |
| "step": 6800 | |
| }, | |
| { | |
| "epoch": 3.23, | |
| "learning_rate": 3.38337182448037e-05, | |
| "loss": 0.4606, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 3.33, | |
| "eval_accuracy": 0.7237428426742554, | |
| "eval_loss": 0.6991223096847534, | |
| "eval_runtime": 12.2324, | |
| "eval_samples_per_second": 513.719, | |
| "eval_steps_per_second": 64.256, | |
| "step": 7200 | |
| }, | |
| { | |
| "epoch": 3.46, | |
| "learning_rate": 3.2678983833718243e-05, | |
| "loss": 0.4629, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 3.51, | |
| "eval_accuracy": 0.7248567938804626, | |
| "eval_loss": 0.7126758694648743, | |
| "eval_runtime": 12.2245, | |
| "eval_samples_per_second": 514.051, | |
| "eval_steps_per_second": 64.297, | |
| "step": 7600 | |
| }, | |
| { | |
| "epoch": 3.7, | |
| "learning_rate": 3.1524249422632794e-05, | |
| "loss": 0.4729, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 3.7, | |
| "eval_accuracy": 0.7277212142944336, | |
| "eval_loss": 0.7072490453720093, | |
| "eval_runtime": 12.2205, | |
| "eval_samples_per_second": 514.217, | |
| "eval_steps_per_second": 64.318, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 3.7, | |
| "step": 8000, | |
| "total_flos": 8.010663974208e+16, | |
| "train_loss": 0.649251142501831, | |
| "train_runtime": 4415.1363, | |
| "train_samples_per_second": 588.385, | |
| "train_steps_per_second": 4.904 | |
| } | |
| ], | |
| "max_steps": 21650, | |
| "num_train_epochs": 10, | |
| "total_flos": 8.010663974208e+16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |