This commit is contained in:
Ying Sheng
2024-07-05 10:06:17 -07:00
committed by GitHub
parent 5a57b8addd
commit dc1b8bcfaa
21 changed files with 487 additions and 354 deletions

View File

@@ -264,7 +264,9 @@ class TiktokenTokenizer:
return self.tokenizer.decode_batch(batch)
def apply_chat_template(self, messages, tokenize, add_generation_prompt):
ret = self.chat_template.render(messages=messages, add_generation_prompt=add_generation_prompt)
ret = self.chat_template.render(
messages=messages, add_generation_prompt=add_generation_prompt
)
return self.encode(ret) if tokenize else ret
@@ -297,5 +299,7 @@ class SentencePieceTokenizer:
return self.tokenizer.decode(batch)
def apply_chat_template(self, messages, tokenize, add_generation_prompt):
ret = self.chat_template.render(messages=messages, add_generation_prompt=add_generation_prompt)
return self.encode(ret) if tokenize else ret
ret = self.chat_template.render(
messages=messages, add_generation_prompt=add_generation_prompt
)
return self.encode(ret) if tokenize else ret