File size: 215 Bytes
31dc8dc | 1 2 3 4 5 6 7 8 9 | train: reflow-dataset
valid: openwebtext-valid
tokenizer_name_or_path: gpt2
cache_dir: /PATH/TO/TRAIN/DATASET
wrap: True
streaming: False
insert_train_eos: True
insert_valid_eos: True
save_dir: /PATH/TO/SAVE/DATASET |