feat: frequency, min_new_tokens, presence, and repetition penalties (#973)

This commit is contained in:
Juwan Yoo
2024-08-08 04:21:08 -07:00
committed by GitHub
parent 228cf47547
commit ab7875941b
20 changed files with 1898 additions and 18 deletions

View File

@@ -162,6 +162,9 @@ class CompletionRequest(BaseModel):
# Extra parameters for SRT backend only and will be ignored by OpenAI models.
regex: Optional[str] = None
ignore_eos: Optional[bool] = False
min_tokens: Optional[int] = 0
repetition_penalty: Optional[float] = 1.0
stop_token_ids: Optional[List[int]] = Field(default_factory=list)
class CompletionResponseChoice(BaseModel):
@@ -259,6 +262,9 @@ class ChatCompletionRequest(BaseModel):
# Extra parameters for SRT backend only and will be ignored by OpenAI models.
regex: Optional[str] = None
min_tokens: Optional[int] = 0
repetition_penalty: Optional[float] = 1.0
stop_token_ids: Optional[List[int]] = Field(default_factory=list)
class ChatMessage(BaseModel):