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

Update convert.py

Browse source
This commit is contained in:
wonjun Jang 2023-11-12 03:23:41 +09:00 • committed by GitHub
parent f37a7d7028
commit 9f4dc236a9
No known key found for this signature in database
GPG key ID: 4AEE18F83AFDEB23
1 changed files with 36 additions and 12 deletions
Download patch file Download diff file Expand all files Collapse all files

46
convert.py
View file

@ -310,8 +310,16 @@ class VocabLoader:
self.tokenizer = AutoTokenizer.from_pretrained(str(fname_tokenizer)) self.tokenizer = AutoTokenizer.from_pretrained(str(fname_tokenizer))
vocab_set = {encoded_tok for encoded_tok, id in self.tokenizer.vocab.items()} vocab_set = {encoded_tok for encoded_tok, id in self.tokenizer.vocab.items()}
self.added_tokens_list = [] self.added_tokens_list = [tok for tok in self.tokenizer.get_added_vocab()]
self.added_tokens_dict = dict(self.tokenizer.get_added_vocab())
self.added_tokens_ids = set(self.tokenizer.get_added_vocab().values())
self.unk_token_id = self.tokenizer.unk_token_id self.unk_token_id = self.tokenizer.unk_token_id
self.specials = {
tok: self.tokenizer.vocab[tok]
for tok in self.tokenizer.all_special_tokens
}
print(self.specials)
self.special_ids = set(self.tokenizer.all_special_ids) self.special_ids = set(self.tokenizer.all_special_ids)
self.vocab_size_base: int = len(vocab_set) self.vocab_size_base: int = len(vocab_set)
self.vocab_size: int = len(vocab_set) self.vocab_size: int = len(vocab_set)
@ -321,6 +329,7 @@ class VocabLoader:
path_candidate = vocab_check_and_append_path(self.fname_tokenizer, vocab_file) path_candidate = vocab_check_and_append_path(self.fname_tokenizer, vocab_file)
if path_candidate is not None: if path_candidate is not None:
self.spm = SentencePieceProcessor(str(path_candidate)) self.spm = SentencePieceProcessor(str(path_candidate))
print(self.spm.vocab_size(), self.vocab_size_base)
else: else:
self.spm self.spm
@ -330,18 +339,16 @@ class VocabLoader:
special_ids = set(tokenizer.all_special_ids) special_ids = set(tokenizer.all_special_ids)
for i in range(self.vocab_size_base): for i in range(self.vocab_size_base):
if i in self.added_tokens_ids:
continue
text = reverse_vocab[i].encode("utf-8") text = reverse_vocab[i].encode("utf-8")
yield text, self.get_token_score(i), self.get_token_type(i) yield text, self.get_token_score(i), self.get_token_type(i)
def get_token_type(self, token_id): def get_token_type(self, token_id, default_type=gguf.TokenType.NORMAL):
toktype = gguf.TokenType.NORMAL toktype = default_type
if self.spm is None: if self.spm is not None and token_id < self.spm.vocab_size():
if i == self.unk_token_id:
toktype = gguf.TokenType.UNKNOWN
if i in self.special_ids:
toktype = gguf.TokenType.CONTROL
else:
if self.spm.is_unknown(token_id): if self.spm.is_unknown(token_id):
toktype = gguf.TokenType.UNKNOWN toktype = gguf.TokenType.UNKNOWN
if self.spm.is_control(token_id): if self.spm.is_control(token_id):
@ -350,18 +357,35 @@ class VocabLoader:
toktype = gguf.TokenType.UNUSED toktype = gguf.TokenType.UNUSED
if self.spm.is_byte(token_id): if self.spm.is_byte(token_id):
toktype = gguf.TokenType.BYTE toktype = gguf.TokenType.BYTE
else:
if token_id == self.unk_token_id:
toktype = gguf.TokenType.UNKNOWN
if token_id in self.special_ids:
toktype = gguf.TokenType.CONTROL
return toktype return toktype
def get_token_score(self, token_id): def get_token_score(self, token_id):
if self.spm is not None: if self.spm is not None and token_id < self.spm.vocab_size():
return self.spm.get_score(token_id) return self.spm.get_score(token_id)
else: else:
return 0.0 return 0.0
def added_tokens(self) -> Iterable[tuple[bytes, float, gguf.TokenType]]: def added_tokens(self) -> Iterable[tuple[bytes, float, gguf.TokenType]]:
default_toktype = gguf.TokenType.USER_DEFINED
for text in self.added_tokens_list: for text in self.added_tokens_list:
if text in self.specials:
toktype = self.get_token_type(self.specials[text], default_toktype)
score = self.get_token_score(self.specials[text])
else:
toktype = default_toktype
score = -1000.0 score = -1000.0
yield text.encode("utf-8"), score, gguf.TokenType.USER_DEFINED
yield text.encode("utf-8"), score, toktype
def has_newline_token(self): def has_newline_token(self):
return '<0x0A>' in self.tokenizer.vocab or '\n' in self.tokenizer.vocab return '<0x0A>' in self.tokenizer.vocab or '\n' in self.tokenizer.vocab