| {"lr": 0.0003, "betas": [0.9, 0.98], "weight_decay": 0.01, "eps": 1e-06, "num_warmup_steps": 2340, "num_training_steps": 23400, "overwrite_output_dir": true, "num_train_epochs": 2, "per_device_train_batch_size": 40, "per_device_eval_batch_size": 40, "do_eval": true, "do_train": true, "evaluation_strategy": "steps", "eval_steps": 500, "save_steps": 5000, "save_total_limit": 5, "logging_steps": 100, "logging_dir": "globertise-v02/logs", "gradient_accumulation_steps": 10, "eval_accumulation_steps": 10, "fp16": true, "path_to_traindata": "data/train/linebyline_transcr_v2.txt", "path_to_eval": "data/val/val.txt", "train_tokenizer": false, "outfile_tokenizer": "tokenizer_from_scratch", "start_from_checkpoint": false, "path_to_checkpoint": "globertise-v02/models/checkpoint_10000", "output_dir": "globertise-v02/models"} |