update provider api, and max token in batch, exclude jupyter

app/modules/github/github_service.py

@@ -60,7 +60,7 @@ def get_github_repo_details(self, repo_name: str) -> Tuple[Github, Dict, str]:
         }
         response = requests.get(url, headers=headers)
         if response.status_code != 200:
-            logger.error(
+            logger.info(
                 f"Failed to get installation ID for {repo_name}. Status code: {response.status_code}, Response: {response.text}"
             )
             raise HTTPException(

app/modules/intelligence/provider/provider_service.py

@@ -38,16 +38,32 @@ async def list_available_llms(self) -> List[ProviderInfo]:
         ]
 
     async def set_global_ai_provider(self, user_id: str, provider: str):
+        provider = provider.lower()
+        # First check if preferences exist
         preferences = self.db.query(UserPreferences).filter_by(user_id=user_id).first()
+
         if not preferences:
-            preferences = UserPreferences(user_id=user_id, preferences={})
+            # Create new preferences if they don't exist
+            preferences = UserPreferences(user_id=user_id, preferences={"llm_provider": provider})
             self.db.add(preferences)
+        else:
+            # Initialize preferences dict if None
+            if preferences.preferences is None:
+                preferences.preferences = {}
+
+            # Update the provider in preferences
+            preferences.preferences["llm_provider"] = provider
+
+            # Explicit update query
+            self.db.query(UserPreferences).filter_by(user_id=user_id).update(
+                {"preferences": preferences.preferences}
+            )
+
         PostHogClient().send_event(
             user_id, "provider_change_event", {"provider": provider}
         )
-        preferences.preferences["llm_provider"] = provider
+
         self.db.commit()
-
         return {"message": f"AI provider set to {provider}"}
 
     def get_large_llm(self):

app/modules/parsing/graph_construction/parsing_helper.py

@@ -97,6 +97,7 @@ def open_text_file(file_path):
             "mov",
             "wmv",
             "flv",
+            "ipynb",
         ]
         include_extensions = [
             "py",

app/modules/parsing/graph_construction/parsing_repomap.py

@@ -540,7 +540,6 @@ def create_graph(self, repo_dir):
 
             for file in files:
                 file_count += 1
-                logging.info(f"Processing file number: {file_count}")
 
                 file_path = os.path.join(root, file)
                 rel_path = os.path.relpath(file_path, repo_dir)

app/modules/parsing/knowledge_graph/inference_service.py

@@ -189,7 +189,7 @@ def get_entry_points_for_nodes(
             }
 
     def batch_nodes(
-        self, nodes: List[Dict], max_tokens: int = 32000, model: str = "gpt-4"
+        self, nodes: List[Dict], max_tokens: int = 16000, model: str = "gpt-4"
     ) -> List[List[DocstringRequest]]:
         batches = []
         current_batch = []
@@ -294,7 +294,7 @@ def batch_entry_points(
         self,
         entry_points_neighbors: Dict[str, List[str]],
         docstring_lookup: Dict[str, str],
-        max_tokens: int = 32000,
+        max_tokens: int = 16000,
         model: str = "gpt-4",
     ) -> List[List[Dict[str, str]]]:
         batches = []