youngkingdom/dify
1
0
Fork 0
mirror of https://github.com/langgenius/dify.git synced 2026-05-06 02:18:08 +08:00
Code Issues Packages Projects Releases Wiki Activity

feat: optimize xinference request max token key and stop reason (#998)

Browse Source
This commit is contained in:
takatost
2023-08-24 18:11:15 +08:00
committed by GitHub
parent 276d3d10a0
commit 9ae91a2ec3
2 changed files with 24 additions and 21 deletions
Download Patch File Download Diff File Expand all files Collapse all files

3
api/core/model_providers/providers/xinference_provider.py
View File

@ -2,7 +2,6 @@ import json
from typing import Type
import requests
from xinference.client import RESTfulGenerateModelHandle, RESTfulChatModelHandle, RESTfulChatglmCppChatModelHandle
from core.helper import encrypter
from core.model_providers.models.embedding.xinference_embedding import XinferenceEmbedding
@ -73,7 +72,7 @@ class XinferenceProvider(BaseModelProvider):
top_p=KwargRule[float](min=0, max=1, default=0.7),
presence_penalty=KwargRule[float](enabled=False),
frequency_penalty=KwargRule[float](enabled=False),
max_tokens=KwargRule[int](alias='max_new_tokens', min=10, max=4000, default=256),
max_tokens=KwargRule[int](min=10, max=4000, default=256),
)