fix: prevent false GPU detection when CUDA image runs without GPU

The STIRLING_GPU=true env var was baked into the :cuda Dockerfile,
which made gpu_available() return True without checking actual
hardware. On machines without a GPU, this would crash upscale.py
(torch.device("cuda") fails) and ocr.py (PaddleOCR use_gpu=True).

Fix: the env var can only disable GPU (set to false/0), never
force-enable it. Hardware detection always runs. Removed the
baked env var from the Dockerfile since it adds no value now.

docker/Dockerfile

@@ -190,8 +190,7 @@ ENV PORT=1349 \
     CONCURRENT_JOBS=3 \
     MAX_MEGAPIXELS=100 \
     RATE_LIMIT_PER_MIN=100 \
-    STIRLING_VARIANT=${VARIANT} \
-    STIRLING_GPU=${GPU}
+    STIRLING_VARIANT=${VARIANT}
 
 # Create non-root user for runtime
 RUN groupadd -r stirling && useradd -r -g stirling -d /app -s /sbin/nologin stirling

packages/ai/python/gpu.py

@@ -6,10 +6,15 @@ import os
 @functools.lru_cache(maxsize=1)
 def gpu_available():
     """Return True if a usable CUDA GPU is present at runtime."""
+    # Allow explicit disable via env var (set to "false" or "0")
     override = os.environ.get("STIRLING_GPU")
-    if override is not None:
-        return override.lower() in ("1", "true", "yes")
+    if override is not None and override.lower() in ("0", "false", "no"):
+        return False
 
+    # Always check actual hardware, even if STIRLING_GPU=true.
+    # The env var can disable GPU but never force-enable it,
+    # because the :cuda image bakes STIRLING_GPU=true and we
+    # still need to handle "no GPU attached" gracefully.
     try:
         import onnxruntime
         if "CUDAExecutionProvider" in onnxruntime.get_available_providers():