Improve HuggingFace model error handling and startup robustness

- Add better error logging for HuggingFace model initialization failures
- Improve startup sequence to not crash if HuggingFace model fails to load
- Add warning message when V2 endpoints are disabled due to model loading failure
- Make application more resilient to HuggingFace model loading issues

This should fix the HuggingFace Spaces deployment startup issues.

app/main.py

@@ -107,6 +107,7 @@ async def startup_event():
             logger.info(f"✅ HuggingFace model warmup completed in {hf_time:.2f}s")
         except Exception as e:
             logger.warning(f"⚠️ HuggingFace model warmup failed: {e}")
+            logger.warning("V2 endpoints will be disabled until model loads successfully")
     else:
         logger.info("⏭️ Skipping V2 HuggingFace warmup (disabled)")
 

app/services/hf_streaming_summarizer.py

@@ -63,6 +63,9 @@ class HFStreamingSummarizer:
             
         except Exception as e:
             logger.error(f"❌ Failed to initialize HuggingFace model: {e}")
+            logger.error(f"Model ID: {settings.hf_model_id}")
+            logger.error(f"Cache dir: {settings.hf_cache_dir}")
+            logger.error(f"Device map: {settings.hf_device_map}")
             self.tokenizer = None
             self.model = None