code-oracle-laya-421m / base-164m /tokenizer /tokenizer_config.json
wxsys's picture
Add lightweight ModernBERT-base 164M model variant in BF16 (~312MB)
603a3e0 verified
Raw History Blame Contribute Delete
337 Bytes
{
"clean_up_tokenization_spaces": true,
"cls_token": "[CLS]",
"local_files_only": false,
"mask_token": "[MASK]",
"model_input_names": [
"input_ids",
"attention_mask"
],
"model_max_length": 8192,
"pad_token": "[PAD]",
"sep_token": "[SEP]",
"tokenizer_class": "PreTrainedTokenizerFast",
"unk_token": "[UNK]"
}