| { | |
| "best_metric": 26.5446, | |
| "best_model_checkpoint": "./ko-en_mbartLarge_exp5p/checkpoint-12000", | |
| "epoch": 7.424593967517401, | |
| "eval_steps": 1000, | |
| "global_step": 16000, | |
| "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": 1.8832, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.46, | |
| "learning_rate": 4.999580070906294e-05, | |
| "loss": 1.6447, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.46, | |
| "eval_bleu": 20.0927, | |
| "eval_gen_len": 18.3986, | |
| "eval_loss": 1.533764362335205, | |
| "eval_runtime": 290.0589, | |
| "eval_samples_per_second": 14.852, | |
| "eval_steps_per_second": 0.931, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "learning_rate": 4.998320424697532e-05, | |
| "loss": 1.5391, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.93, | |
| "learning_rate": 4.996221484543386e-05, | |
| "loss": 1.4737, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.93, | |
| "eval_bleu": 22.6168, | |
| "eval_gen_len": 18.5462, | |
| "eval_loss": 1.4056932926177979, | |
| "eval_runtime": 288.0701, | |
| "eval_samples_per_second": 14.955, | |
| "eval_steps_per_second": 0.937, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 1.16, | |
| "learning_rate": 4.993283955568685e-05, | |
| "loss": 1.4048, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 1.39, | |
| "learning_rate": 4.989508824616534e-05, | |
| "loss": 1.3708, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 1.39, | |
| "eval_bleu": 23.158, | |
| "eval_gen_len": 18.5132, | |
| "eval_loss": 1.364464521408081, | |
| "eval_runtime": 287.9921, | |
| "eval_samples_per_second": 14.959, | |
| "eval_steps_per_second": 0.938, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 1.62, | |
| "learning_rate": 4.984897359916788e-05, | |
| "loss": 1.3392, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 1.86, | |
| "learning_rate": 4.979451110660001e-05, | |
| "loss": 1.3357, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 1.86, | |
| "eval_bleu": 24.2178, | |
| "eval_gen_len": 18.4343, | |
| "eval_loss": 1.3166433572769165, | |
| "eval_runtime": 286.7402, | |
| "eval_samples_per_second": 15.024, | |
| "eval_steps_per_second": 0.942, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 2.09, | |
| "learning_rate": 4.9731719064769833e-05, | |
| "loss": 1.2855, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 2.32, | |
| "learning_rate": 4.966061856824153e-05, | |
| "loss": 1.2274, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 2.32, | |
| "eval_bleu": 24.8105, | |
| "eval_gen_len": 18.4761, | |
| "eval_loss": 1.2854443788528442, | |
| "eval_runtime": 287.0243, | |
| "eval_samples_per_second": 15.009, | |
| "eval_steps_per_second": 0.941, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 2.55, | |
| "learning_rate": 4.958123350274877e-05, | |
| "loss": 1.2255, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 2.78, | |
| "learning_rate": 4.949359053717043e-05, | |
| "loss": 1.2113, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 2.78, | |
| "eval_bleu": 25.4518, | |
| "eval_gen_len": 18.2672, | |
| "eval_loss": 1.2622113227844238, | |
| "eval_runtime": 282.5378, | |
| "eval_samples_per_second": 15.248, | |
| "eval_steps_per_second": 0.956, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 3.02, | |
| "learning_rate": 4.9397719114571386e-05, | |
| "loss": 1.2059, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 3.25, | |
| "learning_rate": 4.929365144231132e-05, | |
| "loss": 1.1392, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 3.25, | |
| "eval_bleu": 25.6184, | |
| "eval_gen_len": 18.4032, | |
| "eval_loss": 1.253997802734375, | |
| "eval_runtime": 282.541, | |
| "eval_samples_per_second": 15.247, | |
| "eval_steps_per_second": 0.956, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 3.48, | |
| "learning_rate": 4.918142248122488e-05, | |
| "loss": 1.1278, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 3.71, | |
| "learning_rate": 4.906106993387679e-05, | |
| "loss": 1.125, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 3.71, | |
| "eval_bleu": 25.3848, | |
| "eval_gen_len": 18.3781, | |
| "eval_loss": 1.2400809526443481, | |
| "eval_runtime": 282.7298, | |
| "eval_samples_per_second": 15.237, | |
| "eval_steps_per_second": 0.955, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 3.94, | |
| "learning_rate": 4.893263423189597e-05, | |
| "loss": 1.1295, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 4.18, | |
| "learning_rate": 4.879615852239274e-05, | |
| "loss": 1.0423, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 4.18, | |
| "eval_bleu": 25.9776, | |
| "eval_gen_len": 18.3387, | |
| "eval_loss": 1.2353969812393188, | |
| "eval_runtime": 282.3157, | |
| "eval_samples_per_second": 15.26, | |
| "eval_steps_per_second": 0.956, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 4.41, | |
| "learning_rate": 4.865168865346393e-05, | |
| "loss": 0.9937, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 4.64, | |
| "learning_rate": 4.849927315879044e-05, | |
| "loss": 1.011, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 4.64, | |
| "eval_bleu": 26.1619, | |
| "eval_gen_len": 18.4858, | |
| "eval_loss": 1.2417948246002197, | |
| "eval_runtime": 286.4621, | |
| "eval_samples_per_second": 15.039, | |
| "eval_steps_per_second": 0.943, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 4.87, | |
| "learning_rate": 4.833896324133269e-05, | |
| "loss": 1.0171, | |
| "step": 10500 | |
| }, | |
| { | |
| "epoch": 5.1, | |
| "learning_rate": 4.817081275612937e-05, | |
| "loss": 0.9493, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 5.1, | |
| "eval_bleu": 25.6398, | |
| "eval_gen_len": 18.2273, | |
| "eval_loss": 1.2616289854049683, | |
| "eval_runtime": 280.5938, | |
| "eval_samples_per_second": 15.353, | |
| "eval_steps_per_second": 0.962, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 5.34, | |
| "learning_rate": 4.799487819220517e-05, | |
| "loss": 0.8973, | |
| "step": 11500 | |
| }, | |
| { | |
| "epoch": 5.57, | |
| "learning_rate": 4.781121865359366e-05, | |
| "loss": 0.888, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 5.57, | |
| "eval_bleu": 26.5446, | |
| "eval_gen_len": 18.438, | |
| "eval_loss": 1.2328206300735474, | |
| "eval_runtime": 287.1698, | |
| "eval_samples_per_second": 15.002, | |
| "eval_steps_per_second": 0.94, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 5.8, | |
| "learning_rate": 4.761989583948173e-05, | |
| "loss": 0.8919, | |
| "step": 12500 | |
| }, | |
| { | |
| "epoch": 6.03, | |
| "learning_rate": 4.7420974023482126e-05, | |
| "loss": 0.8648, | |
| "step": 13000 | |
| }, | |
| { | |
| "epoch": 6.03, | |
| "eval_bleu": 26.0371, | |
| "eval_gen_len": 18.4074, | |
| "eval_loss": 1.2618447542190552, | |
| "eval_runtime": 284.5653, | |
| "eval_samples_per_second": 15.139, | |
| "eval_steps_per_second": 0.949, | |
| "step": 13000 | |
| }, | |
| { | |
| "epoch": 6.26, | |
| "learning_rate": 4.721452003204118e-05, | |
| "loss": 0.7633, | |
| "step": 13500 | |
| }, | |
| { | |
| "epoch": 6.5, | |
| "learning_rate": 4.700060322198889e-05, | |
| "loss": 0.776, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 6.5, | |
| "eval_bleu": 26.0043, | |
| "eval_gen_len": 18.4629, | |
| "eval_loss": 1.2669389247894287, | |
| "eval_runtime": 283.3474, | |
| "eval_samples_per_second": 15.204, | |
| "eval_steps_per_second": 0.953, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 6.73, | |
| "learning_rate": 4.6779295457239025e-05, | |
| "loss": 0.7702, | |
| "step": 14500 | |
| }, | |
| { | |
| "epoch": 6.96, | |
| "learning_rate": 4.6550671084646823e-05, | |
| "loss": 0.7856, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 6.96, | |
| "eval_bleu": 26.2716, | |
| "eval_gen_len": 18.403, | |
| "eval_loss": 1.2591649293899536, | |
| "eval_runtime": 283.0087, | |
| "eval_samples_per_second": 15.222, | |
| "eval_steps_per_second": 0.954, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 7.19, | |
| "learning_rate": 4.6314806909032766e-05, | |
| "loss": 0.6933, | |
| "step": 15500 | |
| }, | |
| { | |
| "epoch": 7.42, | |
| "learning_rate": 4.607178216738045e-05, | |
| "loss": 0.6997, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 7.42, | |
| "eval_bleu": 25.7842, | |
| "eval_gen_len": 18.3693, | |
| "eval_loss": 1.3154139518737793, | |
| "eval_runtime": 288.6701, | |
| "eval_samples_per_second": 14.924, | |
| "eval_steps_per_second": 0.935, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 7.42, | |
| "step": 16000, | |
| "total_flos": 5.548892425520415e+17, | |
| "train_loss": 1.1159878959655762, | |
| "train_runtime": 16243.1142, | |
| "train_samples_per_second": 84.895, | |
| "train_steps_per_second": 5.307 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 86200, | |
| "num_train_epochs": 40, | |
| "save_steps": 1000, | |
| "total_flos": 5.548892425520415e+17, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |