refine xinference (#2521)

### What problem does this PR solve?

#1588

### Type of change

- [x] Refactoring

rag/llm/cv_model.py

@@ -449,6 +449,8 @@ class LocalAICV(GptV4):
 
 class XinferenceCV(Base):
     def __init__(self, key, model_name="", lang="Chinese", base_url=""):
+        if base_url.split("/")[-1] != "v1":
+            base_url = os.path.join(base_url, "v1")
         self.client = OpenAI(api_key="xxx", base_url=base_url)
         self.model_name = model_name
         self.lang = lang

rag/llm/embedding_model.py

@@ -268,6 +268,8 @@ class FastEmbed(Base):
 
 class XinferenceEmbed(Base):
     def __init__(self, key, model_name="", base_url=""):
+        if base_url.split("/")[-1] != "v1":
+            base_url = os.path.join(base_url, "v1")
         self.client = OpenAI(api_key="xxx", base_url=base_url)
         self.model_name = model_name
 

rag/llm/rerank_model.py

@@ -140,6 +140,8 @@ class YoudaoRerank(DefaultRerank):
 
 class XInferenceRerank(Base):
     def __init__(self, key="xxxxxxx", model_name="", base_url=""):
+        if base_url.split("/")[-1] != "v1":
+            base_url = os.path.join(base_url, "v1")
         self.model_name = model_name
         self.base_url = base_url
         self.headers = {

rag/llm/sequence2txt_model.py

@@ -93,6 +93,8 @@ class AzureSeq2txt(Base):
 
 class XinferenceSeq2txt(Base):
     def __init__(self, key, model_name="", base_url=""):
+        if base_url.split("/")[-1] != "v1":
+            base_url = os.path.join(base_url, "v1")
         self.client = OpenAI(api_key="xxx", base_url=base_url)
         self.model_name = model_name