| { | |
| "best_metric": 26.2764, | |
| "best_model_checkpoint": "./ko-en_mbartLarge_exp5p_linear/checkpoint-15000", | |
| "epoch": 8.816705336426914, | |
| "eval_steps": 1000, | |
| "global_step": 19000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.23, | |
| "learning_rate": 5e-05, | |
| "loss": 1.9159, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.46, | |
| "learning_rate": 4.970828471411903e-05, | |
| "loss": 1.7665, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.46, | |
| "eval_bleu": 17.6773, | |
| "eval_gen_len": 18.7196, | |
| "eval_loss": 1.6563962697982788, | |
| "eval_runtime": 295.5881, | |
| "eval_samples_per_second": 14.574, | |
| "eval_steps_per_second": 0.913, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "learning_rate": 4.941656942823804e-05, | |
| "loss": 1.6536, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.93, | |
| "learning_rate": 4.912485414235706e-05, | |
| "loss": 1.5688, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.93, | |
| "eval_bleu": 20.8837, | |
| "eval_gen_len": 18.3983, | |
| "eval_loss": 1.493910789489746, | |
| "eval_runtime": 286.1303, | |
| "eval_samples_per_second": 15.056, | |
| "eval_steps_per_second": 0.944, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 1.16, | |
| "learning_rate": 4.883313885647608e-05, | |
| "loss": 1.4925, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 1.39, | |
| "learning_rate": 4.8541423570595104e-05, | |
| "loss": 1.457, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 1.39, | |
| "eval_bleu": 21.9168, | |
| "eval_gen_len": 18.458, | |
| "eval_loss": 1.4349554777145386, | |
| "eval_runtime": 286.2559, | |
| "eval_samples_per_second": 15.049, | |
| "eval_steps_per_second": 0.943, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 1.62, | |
| "learning_rate": 4.824970828471412e-05, | |
| "loss": 1.4164, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 1.86, | |
| "learning_rate": 4.795799299883314e-05, | |
| "loss": 1.4107, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 1.86, | |
| "eval_bleu": 22.8881, | |
| "eval_gen_len": 18.4826, | |
| "eval_loss": 1.3751940727233887, | |
| "eval_runtime": 290.8825, | |
| "eval_samples_per_second": 14.81, | |
| "eval_steps_per_second": 0.928, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 2.09, | |
| "learning_rate": 4.7666277712952163e-05, | |
| "loss": 1.3584, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 2.32, | |
| "learning_rate": 4.737456242707118e-05, | |
| "loss": 1.3039, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 2.32, | |
| "eval_bleu": 23.8115, | |
| "eval_gen_len": 18.4348, | |
| "eval_loss": 1.332729458808899, | |
| "eval_runtime": 285.9035, | |
| "eval_samples_per_second": 15.068, | |
| "eval_steps_per_second": 0.944, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 2.55, | |
| "learning_rate": 4.70828471411902e-05, | |
| "loss": 1.2963, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 2.78, | |
| "learning_rate": 4.679113185530922e-05, | |
| "loss": 1.282, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 2.78, | |
| "eval_bleu": 24.235, | |
| "eval_gen_len": 18.3561, | |
| "eval_loss": 1.3079338073730469, | |
| "eval_runtime": 284.0397, | |
| "eval_samples_per_second": 15.167, | |
| "eval_steps_per_second": 0.951, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 3.02, | |
| "learning_rate": 4.649941656942824e-05, | |
| "loss": 1.2761, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 3.25, | |
| "learning_rate": 4.620770128354726e-05, | |
| "loss": 1.2133, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 3.25, | |
| "eval_bleu": 24.8877, | |
| "eval_gen_len": 18.5204, | |
| "eval_loss": 1.2819560766220093, | |
| "eval_runtime": 284.2962, | |
| "eval_samples_per_second": 15.153, | |
| "eval_steps_per_second": 0.95, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 3.48, | |
| "learning_rate": 4.5915985997666276e-05, | |
| "loss": 1.1875, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 3.71, | |
| "learning_rate": 4.56242707117853e-05, | |
| "loss": 1.1787, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 3.71, | |
| "eval_bleu": 25.2719, | |
| "eval_gen_len": 18.415, | |
| "eval_loss": 1.2579996585845947, | |
| "eval_runtime": 283.3691, | |
| "eval_samples_per_second": 15.203, | |
| "eval_steps_per_second": 0.953, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 3.94, | |
| "learning_rate": 4.5332555425904324e-05, | |
| "loss": 1.175, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 4.18, | |
| "learning_rate": 4.5040840140023335e-05, | |
| "loss": 1.1154, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 4.18, | |
| "eval_bleu": 25.5507, | |
| "eval_gen_len": 18.3528, | |
| "eval_loss": 1.2543020248413086, | |
| "eval_runtime": 282.3342, | |
| "eval_samples_per_second": 15.259, | |
| "eval_steps_per_second": 0.956, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 4.41, | |
| "learning_rate": 4.474912485414236e-05, | |
| "loss": 1.0893, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 4.64, | |
| "learning_rate": 4.445740956826138e-05, | |
| "loss": 1.0956, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 4.64, | |
| "eval_bleu": 25.7284, | |
| "eval_gen_len": 18.5348, | |
| "eval_loss": 1.2415348291397095, | |
| "eval_runtime": 283.6466, | |
| "eval_samples_per_second": 15.188, | |
| "eval_steps_per_second": 0.952, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 4.87, | |
| "learning_rate": 4.41656942823804e-05, | |
| "loss": 1.0869, | |
| "step": 10500 | |
| }, | |
| { | |
| "epoch": 5.1, | |
| "learning_rate": 4.387397899649942e-05, | |
| "loss": 1.023, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 5.1, | |
| "eval_bleu": 25.7912, | |
| "eval_gen_len": 18.3347, | |
| "eval_loss": 1.2409833669662476, | |
| "eval_runtime": 281.8959, | |
| "eval_samples_per_second": 15.282, | |
| "eval_steps_per_second": 0.958, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 5.34, | |
| "learning_rate": 4.358226371061844e-05, | |
| "loss": 0.9568, | |
| "step": 11500 | |
| }, | |
| { | |
| "epoch": 5.57, | |
| "learning_rate": 4.329054842473746e-05, | |
| "loss": 0.95, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 5.57, | |
| "eval_bleu": 25.9921, | |
| "eval_gen_len": 18.2593, | |
| "eval_loss": 1.2326716184616089, | |
| "eval_runtime": 280.949, | |
| "eval_samples_per_second": 15.334, | |
| "eval_steps_per_second": 0.961, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 5.8, | |
| "learning_rate": 4.299883313885648e-05, | |
| "loss": 0.9611, | |
| "step": 12500 | |
| }, | |
| { | |
| "epoch": 6.03, | |
| "learning_rate": 4.2707117852975496e-05, | |
| "loss": 0.9476, | |
| "step": 13000 | |
| }, | |
| { | |
| "epoch": 6.03, | |
| "eval_bleu": 25.829, | |
| "eval_gen_len": 18.3686, | |
| "eval_loss": 1.2631200551986694, | |
| "eval_runtime": 282.7475, | |
| "eval_samples_per_second": 15.236, | |
| "eval_steps_per_second": 0.955, | |
| "step": 13000 | |
| }, | |
| { | |
| "epoch": 6.26, | |
| "learning_rate": 4.241540256709452e-05, | |
| "loss": 0.887, | |
| "step": 13500 | |
| }, | |
| { | |
| "epoch": 6.5, | |
| "learning_rate": 4.212368728121354e-05, | |
| "loss": 0.9061, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 6.5, | |
| "eval_bleu": 25.8316, | |
| "eval_gen_len": 18.7481, | |
| "eval_loss": 1.254765272140503, | |
| "eval_runtime": 286.2873, | |
| "eval_samples_per_second": 15.048, | |
| "eval_steps_per_second": 0.943, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 6.73, | |
| "learning_rate": 4.1831971995332556e-05, | |
| "loss": 0.8963, | |
| "step": 14500 | |
| }, | |
| { | |
| "epoch": 6.96, | |
| "learning_rate": 4.154025670945157e-05, | |
| "loss": 0.9037, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 6.96, | |
| "eval_bleu": 26.2764, | |
| "eval_gen_len": 18.3888, | |
| "eval_loss": 1.2307679653167725, | |
| "eval_runtime": 281.3307, | |
| "eval_samples_per_second": 15.313, | |
| "eval_steps_per_second": 0.96, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 7.19, | |
| "learning_rate": 4.12485414235706e-05, | |
| "loss": 0.7613, | |
| "step": 15500 | |
| }, | |
| { | |
| "epoch": 7.42, | |
| "learning_rate": 4.095682613768962e-05, | |
| "loss": 0.7431, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 7.42, | |
| "eval_bleu": 25.9268, | |
| "eval_gen_len": 18.256, | |
| "eval_loss": 1.2716145515441895, | |
| "eval_runtime": 279.8965, | |
| "eval_samples_per_second": 15.391, | |
| "eval_steps_per_second": 0.965, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 7.66, | |
| "learning_rate": 4.066511085180863e-05, | |
| "loss": 0.7538, | |
| "step": 16500 | |
| }, | |
| { | |
| "epoch": 7.89, | |
| "learning_rate": 4.037339556592766e-05, | |
| "loss": 0.7526, | |
| "step": 17000 | |
| }, | |
| { | |
| "epoch": 7.89, | |
| "eval_bleu": 25.9883, | |
| "eval_gen_len": 18.2052, | |
| "eval_loss": 1.265455722808838, | |
| "eval_runtime": 280.4684, | |
| "eval_samples_per_second": 15.36, | |
| "eval_steps_per_second": 0.963, | |
| "step": 17000 | |
| }, | |
| { | |
| "epoch": 8.12, | |
| "learning_rate": 4.0081680280046675e-05, | |
| "loss": 0.7074, | |
| "step": 17500 | |
| }, | |
| { | |
| "epoch": 8.35, | |
| "learning_rate": 3.97899649941657e-05, | |
| "loss": 0.6654, | |
| "step": 18000 | |
| }, | |
| { | |
| "epoch": 8.35, | |
| "eval_bleu": 25.6866, | |
| "eval_gen_len": 18.2217, | |
| "eval_loss": 1.311843991279602, | |
| "eval_runtime": 281.3073, | |
| "eval_samples_per_second": 15.314, | |
| "eval_steps_per_second": 0.96, | |
| "step": 18000 | |
| }, | |
| { | |
| "epoch": 8.58, | |
| "learning_rate": 3.949824970828472e-05, | |
| "loss": 0.666, | |
| "step": 18500 | |
| }, | |
| { | |
| "epoch": 8.82, | |
| "learning_rate": 3.9206534422403734e-05, | |
| "loss": 0.6953, | |
| "step": 19000 | |
| }, | |
| { | |
| "epoch": 8.82, | |
| "eval_bleu": 25.8958, | |
| "eval_gen_len": 18.3387, | |
| "eval_loss": 1.3050304651260376, | |
| "eval_runtime": 280.1508, | |
| "eval_samples_per_second": 15.377, | |
| "eval_steps_per_second": 0.964, | |
| "step": 19000 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 86200, | |
| "num_train_epochs": 40, | |
| "save_steps": 1000, | |
| "total_flos": 6.589808201024143e+17, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |