AraElectra-Arabic-SQuADv2-QA / tokenizer_config.json
ZeyadAhmed's picture
add tokenizer
3e3e903
raw
history blame
481 Bytes
{
"cls_token": "[CLS]",
"do_basic_tokenize": true,
"do_lower_case": false,
"mask_token": "[MASK]",
"max_len": 512,
"name_or_path": "/content/drive/MyDrive/AraElectra-ASQuADv2-QA",
"never_split": [
"[بريد]",
"[مستخدم]",
"[رابط]"
],
"pad_token": "[PAD]",
"sep_token": "[SEP]",
"special_tokens_map_file": null,
"strip_accents": null,
"tokenize_chinese_chars": true,
"tokenizer_class": "ElectraTokenizer",
"unk_token": "[UNK]"
}