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

Merge branch 'main' of github.com:abetlen/llama_cpp_python into main

Browse source
This commit is contained in:
Andrei Betlen 2023-12-22 14:05:26 -05:00
commit 15ee2106f6
2 changed files with 5 additions and 3 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

4
llama_cpp/llama.py
View file

@ -1917,7 +1917,7 @@ class Llama:
completion_or_chunks = self._create_completion( completion_or_chunks = self._create_completion(
prompt=prompt, prompt=prompt,
suffix=suffix, suffix=suffix,
max_tokens=max_tokens, max_tokens=-1 if max_tokens is None else max_tokens,
temperature=temperature, temperature=temperature,
top_p=top_p, top_p=top_p,
min_p=min_p, min_p=min_p,
@ -1951,7 +1951,7 @@ class Llama:
self, self,
prompt: str, prompt: str,
suffix: Optional[str] = None, suffix: Optional[str] = None,
max_tokens: int = 128, max_tokens: Optional[int] = 16,
temperature: float = 0.8, temperature: float = 0.8,
top_p: float = 0.95, top_p: float = 0.95,
min_p: float = 0.05, min_p: float = 0.05,

4
llama_cpp/server/types.py
View file

@ -110,7 +110,9 @@ class CreateCompletionRequest(BaseModel):
default=None, default=None,
description="A suffix to append to the generated text. If None, no suffix is appended. Useful for chatbots.", description="A suffix to append to the generated text. If None, no suffix is appended. Useful for chatbots.",
) )
max_tokens: int = max_tokens_field max_tokens: Optional[int] = Field(
default=16, ge=0, description="The maximum number of tokens to generate."
)
temperature: float = temperature_field temperature: float = temperature_field
top_p: float = top_p_field top_p: float = top_p_field
min_p: float = min_p_field min_p: float = min_p_field