gia2-small / tokenizer_config.json
ClementRomac's picture
ClementRomac HF staff
Upload processor
2e8ba46
raw
history blame contribute delete
486 Bytes
{
"auto_map": {
"AutoProcessor": "processor.GIAProcessor"
},
"clean_up_tokenization_spaces": true,
"cls_token": "[CLS]",
"do_lower_case": true,
"mask_token": "[MASK]",
"model_input_names": [
"input_ids",
"attention_mask"
],
"model_max_length": 512,
"pad_token": "[PAD]",
"processor_class": "GIAProcessor",
"sep_token": "[SEP]",
"strip_accents": null,
"tokenize_chinese_chars": true,
"tokenizer_class": "BertTokenizer",
"unk_token": "[UNK]"
}