解决xinference内嵌模型类使用问题。由于目前xinference组件的版本和llamaindex最新版有冲突，所以未更新支持xinference的内嵌模型的版本

backend/app/settings.py

@@ -3,11 +3,10 @@ from typing import Dict
 
 from llama_index.core.constants import DEFAULT_TEMPERATURE
 from llama_index.core.settings import Settings
+from app.xinference.base import XinferenceEmbedding, XinferenceRerank
 from llama_index.llms.xinference import Xinference
 from llama_index.llms.xinference.base import DEFAULT_XINFERENCE_TEMP
 
-from app.xinference.base import XinferenceEmbedding, XinferenceRerank
-
 
 def get_node_postprocessors():
     rerank_enabled = os.getenv("RERANK_ENABLED").title()
@@ -232,4 +231,4 @@ def init_mistral():
     #
     # Settings.llm = MistralAI(model=os.getenv("MODEL"))
     # Settings.embed_model = MistralAIEmbedding(model_name=os.getenv("EMBEDDING_MODEL"))
-    pass
+    pass