Add python-multipart to requirements, GGUF support for CUDA backend

- Add python-multipart to requirements.txt, requirements-nvidia.txt, requirements-vulkan.txt
- Add llama-cpp-python to requirements-nvidia.txt for GGUF support
- When using CUDA/nvidia backend with GGUF file, automatically use llama-cpp-python

coderai

@@ -1808,11 +1808,14 @@ class ModelManager:
         
         Args:
             model_name: Model name or path
-            backend_type: 'nvidia', 'vulkan', or 'auto' to detect
+            backend_type: 'nvidia', 'vulkan', 'cuda', or 'auto' to detect
             **kwargs: Additional arguments for the specific backend
         """
         available = detect_available_backends()
         
+        # Check if model is a GGUF file
+        is_gguf = model_name.endswith('.gguf') or 'gguf' in model_name.lower()
+        
         # Determine backend
         if backend_type == "auto":
             if available.get('nvidia'):
@@ -1826,6 +1829,11 @@ class ModelManager:
                 print("For Vulkan, install llama-cpp-python with Vulkan support.")
                 raise RuntimeError("No suitable backend found")
         
+        # If GGUF file and backend is nvidia/cuda, use llama-cpp-python (vulkan backend)
+        if is_gguf and backend_type in ("nvidia", "cuda"):
+            print(f"GGUF model detected, using llama-cpp-python ({backend_type} backend)")
+            backend_type = "vulkan"  # Use llama-cpp-python for GGUF
+        
         self.backend_type = backend_type
         
         # Create appropriate backend

requirements-nvidia.txt

@@ -2,6 +2,8 @@
 fastapi>=0.104.0
 uvicorn[standard]>=0.24.0
 pydantic>=2.5.0
+python-multipart>=0.0.6  # for multipart form data parsing
+requests>=2.31.0  # for HTTP requests
 
 # ML dependencies (transformers-based for NVIDIA/CUDA)
 transformers>=4.35.0
@@ -22,6 +24,9 @@ tiktoken>=0.5.0
 tokenizers>=0.15.0
 protobuf>=3.20.0
 
+# llama-cpp-python with CUDA support (for GGUF files on CUDA backend)
+llama-cpp-python>=0.2.0
+
 # Optional: Flash Attention 2 for faster inference on supported NVIDIA GPUs
 # Requires specific CUDA versions and may need manual installation
 # Install with: pip install flash-attn --no-build-isolation

requirements-vulkan.txt

@@ -2,6 +2,8 @@
 fastapi>=0.104.0
 uvicorn[standard]>=0.24.0
 pydantic>=2.5.0
+python-multipart>=0.0.6  # for multipart form data parsing
+requests>=2.31.0  # for HTTP requests
 
 # llama-cpp-python is installed by build.sh with Vulkan support
 # CMAKE_ARGS="-DGGML_VULKAN=ON" pip install llama-cpp-python --no-cache-dir

requirements.txt

@@ -5,6 +5,7 @@ pydantic>=2.5.0
 
 # CLI dependencies
 requests>=2.31.0  # for the coder CLI tool
+python-multipart>=0.0.6  # for multipart form data parsing
 
 # PyTorch - Uncomment the appropriate version for your system.
 # IMPORTANT: Use quotes around version specifiers to prevent shell interpretation!