vbatts/llama.cpp
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions

gguf : accept str for path in SpecialVocab.__init__

Browse source
This commit is contained in:
Cebtenzzre 2023-09-29 18:51:19 -04:00
parent ea90d2aa8c
commit 7fa5cbf8cc
1 changed files with 10 additions and 7 deletions
Download patch file Download diff file Expand all files Collapse all files

17
gguf-py/gguf/gguf.py
View file

@ -824,18 +824,21 @@ class SpecialVocab:
special_token_types: tuple[str, ...] = ('bos', 'eos', 'unk', 'sep', 'pad') special_token_types: tuple[str, ...] = ('bos', 'eos', 'unk', 'sep', 'pad')
special_token_ids: dict[str, int] = {} special_token_ids: dict[str, int] = {}
def __init__(self, path: Path, load_merges: bool = False, special_token_types: tuple[str, ...] | None = None): def __init__(
self, path: str | os.PathLike[str], load_merges: bool = False,
special_token_types: tuple[str, ...] | None = None,
):
self.special_token_ids = {} self.special_token_ids = {}
self.load_merges = load_merges self.load_merges = load_merges
if special_token_types is not None: if special_token_types is not None:
self.special_token_types = special_token_types self.special_token_types = special_token_types
self.load(path) self._load(Path(path))
def load(self, path: Path): def _load(self, path: Path) -> None:
if not self.try_load_from_tokenizer_json(path): if not self._try_load_from_tokenizer_json(path):
self.try_load_from_config_json(path) self._try_load_from_config_json(path)
def try_load_from_tokenizer_json(self, path: Path) -> bool: def _try_load_from_tokenizer_json(self, path: Path) -> bool:
tokenizer_file = path / 'tokenizer.json' tokenizer_file = path / 'tokenizer.json'
if not tokenizer_file.is_file(): if not tokenizer_file.is_file():
return False return False
@ -868,7 +871,7 @@ class SpecialVocab:
break break
return True return True
def try_load_from_config_json(self, path: Path) -> bool: def _try_load_from_config_json(self, path: Path) -> bool:
config_file = path / 'config.json' config_file = path / 'config.json'
if not config_file.is_file(): if not config_file.is_file():
return False return False