fix bugs (#200)

Author: BytedanceFu

README.md

@@ -121,7 +121,8 @@ Volcengine supports both model names and endpoint IDs. Using model names is reco
   "vlm": {
     "provider": "volcengine",
     "model": "doubao-seed-1-6-240615",
-    "api_key": "your-api-key"
+    "api_key": "your-api-key",
+    "api_base" : "https://ark.cn-beijing.volces.com/api/v3",
   }
 }
 ```
@@ -133,7 +134,8 @@ You can also use endpoint IDs (found in [Volcengine ARK Console](https://console
   "vlm": {
     "provider": "volcengine",
     "model": "ep-20241220174930-xxxxx",
-    "api_key": "your-api-key"
+    "api_key": "your-api-key",
+    "api_base" : "https://ark.cn-beijing.volces.com/api/v3",
   }
 }
 ```

openviking/models/vlm/backends/litellm_vlm.py

@@ -72,6 +72,13 @@ def _resolve_model(self, model: str) -> str:
                 model = f"{prefix}/{model}"
             return model
 
+        if self._provider_name == "openai" and self.api_base:
+            from openviking.models.vlm.registry import find_by_name
+            openai_spec = find_by_name("openai")
+            is_openai_official = "api.openai.com" in self.api_base
+            if openai_spec and not is_openai_official and not model.startswith("openai/"):
+                return f"openai/{model}"
+
         spec = find_by_model(model)
         if spec and spec.litellm_prefix:
             if not any(model.startswith(s) for s in spec.skip_prefixes):