| { | |
| "cls_token": "[CLS]", | |
| "do_basic_tokenize": true, | |
| "do_lower_case": false, | |
| "mask_token": "[MASK]", | |
| "max_len": 1024, | |
| "model_max_length": 1024, | |
| "name_or_path": "/data/ouga/home/ag_gagneur/hingerl/BERTADN/preprocessing/tokenizer_upstream_species_k1", | |
| "never_split": null, | |
| "pad_token": "[PAD]", | |
| "sep_token": "[SEP]", | |
| "special_tokens_map_file": "/data/ouga/home/ag_gagneur/hingerl/.cache/huggingface/hub/models--zhihan1996--DNA_bert_6/snapshots/a79a8fd96ad172f964a4dbef3f4d7545a5034baa/special_tokens_map.json", | |
| "strip_accents": null, | |
| "tokenize_chinese_chars": true, | |
| "tokenizer_class": "BertTokenizer", | |
| "unk_token": "[UNK]" | |
| } | |