vrr/kvcache-ai-ktransformers
2
0
Fork 0
mirror of https://github.com/kvcache-ai/ktransformers.git synced 2026-05-04 22:51:51 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions 1

fix load default max_new_tokens

Browse source
This commit is contained in:
qiyuxinlin 2025-04-25 04:20:12 +00:00
parent 67042d11e3
commit 7af83f9efb
4 changed files with 21 additions and 10 deletions
Download patch file Download diff file Expand all files Collapse all files

15
ktransformers/server/api/openai/endpoints/chat.py
View file

@ -138,12 +138,23 @@ async def chat_completion(request: Request, create: ChatCompletionCreate):
# Process messages with tool functionality if needed
enhanced_messages = list(create.messages)
if create.max_tokens<0 or create.max_completion_tokens<0:
if create.max_tokens is not None and create.max_tokens<0:
return JSONResponse(
status_code=400,
content={
"object": "error",
"message": f"max_new_tokens must be at least 0, got {create.max_tokens}.",
"message": f"max_tokens must be at least 0, got {create.max_tokens}.",
"type": "BadRequestError",
"param": None,
"code": 400
})
if create.max_completion_tokens is not None and create.max_completion_tokens<0:
return JSONResponse(
status_code=400,
content={
"object": "error",
"message": f"max_completion_tokens must be at least 0, got {create.max_completion_tokens}.",
"type": "BadRequestError",
"param": None,
"code": 400

8
ktransformers/server/api/openai/legacy/completions.py
View file

@ -14,22 +14,22 @@ router = APIRouter()
@router.post("/completions",tags=['openai'])
async def create_completion(request:Request, create:CompletionCreate):
id = str(uuid4())
if create.max_tokens<0:
if create.max_tokens is not None and create.max_tokens<0:
return JSONResponse(
status_code=400,
content={
"object": "error",
"message": f"max_new_tokens must be at least 0, got {create.max_tokens}.",
"message": f"max_tokens must be at least 0, got {create.max_tokens}.",
"type": "BadRequestError",
"param": None,
"code": 400
})
if create.max_completion_tokens<0:
if create.max_completion_tokens is not None and create.max_completion_tokens<0:
return JSONResponse(
status_code=400,
content={
"object": "error",
"message": f"max_new_tokens must be at least 0, got {create.max_completion_tokens}.",
"message": f"max_completion_tokens must be at least 0, got {create.max_completion_tokens}.",
"type": "BadRequestError",
"param": None,
"code": 400