Fixed macos issue

main/main.py

@@ -1,6 +1,8 @@
 """
 LLM Inference Server main application using LitServe framework.
 """
+from sys import platform
+
 import litserve as ls
 import logging
 import os
@@ -52,13 +54,21 @@ def create_app():
     # Initialize router with API instance
     init_router(api, config)
 
-    # Create LitServer instance
-    server = ls.LitServer(
-        api,
-        timeout=server_config.get('timeout', 60),
-        max_batch_size=server_config.get('max_batch_size', 1),
-        track_requests=True
-    )
+    if platform == "darwin":  # Darwin is macOS
+        server = ls.LitServer(
+            api,
+            timeout=server_config.get('timeout', 60),
+            max_batch_size=server_config.get('max_batch_size', 1),
+            track_requests=True,
+            accelerator="cpu"  # Force CPU on Mac
+        )
+    else:
+        server = ls.LitServer(
+            api,
+            timeout=server_config.get('timeout', 60),
+            max_batch_size=server_config.get('max_batch_size', 1),
+            track_requests=True
+        )
 
     # Launch inference workers (assuming single uvicorn worker for now)
     _MANAGER, _WORKER_PROCESSES = server.launch_inference_worker(num_uvicorn_servers=1)