Create onnxruntime.py

tests/onnxruntime.py

@@ -0,0 +1,14 @@
+import onnxruntime as ort
+
+# Check if CUDA execution provider is available
+if 'CUDAExecutionProvider' in ort.get_available_providers():
+    print("CUDA is available")
+
+    # Get CUDA device information
+    cuda_providers = [provider for provider in ort.get_available_providers() if provider.startswith('CUDA')]
+    for provider in cuda_providers:
+        print(f"Provider: {provider}")
+        print(f"Device ID: {ort.get_device()}")
+
+else:
+    print("CUDA is not available")