Remove GITHUB_BASE_URL env var and hard-code GitHub Models URL

Co-authored-by: pamelafox <[email protected]>

Author: Copilot

.env.sample

@@ -37,7 +37,6 @@ OLLAMA_EMBED_MODEL=nomic-embed-text
 OLLAMA_EMBEDDING_COLUMN=embedding_nomic
 # Needed for GitHub Models:
 GITHUB_TOKEN=YOUR-GITHUB-TOKEN
-GITHUB_BASE_URL=https://models.github.ai/inference
 GITHUB_MODEL=openai/gpt-4o
 GITHUB_EMBED_MODEL=openai/text-embedding-3-large
 GITHUB_EMBED_DIMENSIONS=1024

src/backend/fastapi_app/openai_clients.py

@@ -54,11 +54,10 @@ async def create_openai_chat_client(
         )
     elif OPENAI_CHAT_HOST == "github":
         logger.info("Setting up OpenAI client for chat completions using GitHub Models")
-        github_base_url = os.getenv("GITHUB_BASE_URL", "https://models.github.ai/inference")
         github_model = os.getenv("GITHUB_MODEL", "openai/gpt-4o")
-        logger.info(f"Using GitHub Models with base URL: {github_base_url}, model: {github_model}")
+        logger.info(f"Using GitHub Models with model: {github_model}")
         openai_chat_client = openai.AsyncOpenAI(
-            base_url=github_base_url,
+            base_url="https://models.github.ai/inference",
             api_key=os.getenv("GITHUB_TOKEN"),
         )
     else:
@@ -114,11 +113,10 @@ async def create_openai_embed_client(
         )
     elif OPENAI_EMBED_HOST == "github":
         logger.info("Setting up OpenAI client for embeddings using GitHub Models")
-        github_base_url = os.getenv("GITHUB_BASE_URL", "https://models.github.ai/inference")
         github_embed_model = os.getenv("GITHUB_EMBED_MODEL", "openai/text-embedding-3-small")
-        logger.info(f"Using GitHub Models with base URL: {github_base_url}, embedding model: {github_embed_model}")
+        logger.info(f"Using GitHub Models with embedding model: {github_embed_model}")
         openai_embed_client = openai.AsyncOpenAI(
-            base_url=github_base_url,
+            base_url="https://models.github.ai/inference",
             api_key=os.getenv("GITHUB_TOKEN"),
         )
     else:

tests/test_openai_clients.py

@@ -32,7 +32,7 @@ async def test_github_models_configuration(monkeypatch):
     monkeypatch.setenv("OPENAI_CHAT_HOST", "github")
     monkeypatch.setenv("OPENAI_EMBED_HOST", "github")
     monkeypatch.setenv("GITHUB_TOKEN", "fake-token")
-    # Don't set GITHUB_BASE_URL or GITHUB_MODEL to test defaults
+    # Don't set GITHUB_MODEL to test defaults
 
     # Test chat client configuration
     chat_client = await create_openai_chat_client(None)
@@ -57,7 +57,6 @@ async def test_github_models_with_custom_values(monkeypatch):
     monkeypatch.setenv("OPENAI_CHAT_HOST", "github")
     monkeypatch.setenv("OPENAI_EMBED_HOST", "github")
     monkeypatch.setenv("GITHUB_TOKEN", "fake-token")
-    monkeypatch.setenv("GITHUB_BASE_URL", "https://custom.github.ai/inference")
     monkeypatch.setenv("GITHUB_MODEL", "openai/gpt-4")
     monkeypatch.setenv("GITHUB_EMBED_MODEL", "openai/text-embedding-ada-002")