|
Formatron v0.4.2
Formatron empowers everyone to control the output format of language models with minimal overhead.
|
Functions | |
| _multiple_replace (replacements, regex, text) | |
| _autodetect_processors (typing.Dict[str, int] vocab) | |
| typing.Dict[bytes, int] | get_original_characters (typing.Dict[str, int] vocab) |
| huggingface_bytelevel_decoder () | |
| I hate legacy code. | |
Variables | |
| Processors = set[typing.Literal["sentencepiece", "<0xHH>", "dot_G"]] | |
|
protected |
|
protected |
| typing.Dict[bytes, int] formatron.integrations._utils.get_original_characters | ( | typing.Dict[str, int] | vocab | ) |
| formatron.integrations._utils.huggingface_bytelevel_decoder | ( | ) |