github.com/wbrown/gpt_bpe@v0.0.0-20250709161131-1571a6e8ad2d/resources/data/clip-tokenizer/special_config.json (about)

     1  {
     2    "punc_runes": [ ",", ".", "!", "?", "'" ],
     3    "normalizer": 	{ "‘": "'",
     4      "’": "'",
     5      "“": "\"",
     6      "”": "\""},
     7    "enclose_eos_bos": true,
     8    "prefix_space": false,
     9    "lower_case": true,
    10    "end_of_word": "</w>"
    11  }