(self, *args, **kwargs)
| 363 | assert_str = "make sure to set PATH for openwebtext data_utils/corpora.py" |
| 364 | |
| 365 | def __init__(self, *args, **kwargs): |
| 366 | import fasttext |
| 367 | super().__init__(*args, **kwargs) |
| 368 | self.model = fasttext.load_model('/dataset/fd5061f6/english_data/lid.176.bin') |
| 369 | print_rank_0("Load language detection model") |
| 370 | |
| 371 | def process_line(self, data, tokenizer, tokenize): |
| 372 | text = data['text'] |
nothing calls this directly
no test coverage detected