Fix OMP_NUM_THREADS being overridden by HuggingFace CPU_CORES

v1.4.1 - CRITICAL FIX: HuggingFace overriding OMP_NUM_THREADS with CPU_CORES
- App was exiting because OMP_NUM_THREADS='3500m' instead of '1'
- HuggingFace environment was overriding our Dockerfile ENV
- Added explicit Python os.environ override BEFORE torch imports
- This should prevent app crashes and libgomp errors

🤖 Generated with [Claude Code](https://claude.ai/code)

Co-Authored-By: Claude <noreply@anthropic.com>

app.py

@@ -6,8 +6,13 @@ import os
 from typing import Optional
 from contextlib import asynccontextmanager
 
-# Environment variables now set in Dockerfile
-# OMP_NUM_THREADS=1, HF_HOME=/app/hf_cache, etc.
+# CRITICAL: Set OMP_NUM_THREADS before any torch/numpy imports
+# HuggingFace is overriding our Dockerfile ENV with CPU_CORES value
+os.environ['OMP_NUM_THREADS'] = '1'
+# Also ensure other environment variables are correct
+os.environ['HF_HOME'] = '/app/hf_cache'
+os.environ['HUGGINGFACE_HUB_CACHE'] = '/app/hf_cache'
+os.environ['TRANSFORMERS_CACHE'] = '/app/hf_cache'
 
 import torch
 import numpy as np
@@ -16,7 +21,7 @@ from fastapi.responses import JSONResponse, HTMLResponse
 import uvicorn
 
 # Version tracking
-VERSION = "1.4.0"
+VERSION = "1.4.1"
 COMMIT_SHA = "TBD"
 
 # Configure logging