Add -DLLAMA_CURL=OFF

Seems to be causing build errors: https://github.com/ggml-org/llama.cpp/issues/12925

```
-- Detecting CUDA compile features
-- Detecting CUDA compile features - done
-- CUDA host compiler is GNU 13.3.0

-- Including CUDA backend
-- Could NOT find CURL (missing: CURL_LIBRARY CURL_INCLUDE_DIR)
CMake Error at common/CMakeLists.txt:90 (message):
Could NOT find CURL. Hint: to disable this feature, set -DLLAMA_CURL=OFF


-- Configuring incomplete, errors occurred!
gmake: Makefile: No such file or directory
gmake: *** No rule to make target 'Makefile'. Stop.
cp: cannot stat './build/bin/llama-*': No such file or directory
* Running on local URL: http://0.0.0.0:7860
```

start.sh

@@ -12,7 +12,7 @@ if [[ -z "${RUN_LOCALLY}" ]]; then
 fi
 
 cd llama.cpp
-cmake -B build -DBUILD_SHARED_LIBS=OFF -DGGML_CUDA=${GGML_CUDA}
+cmake -B build -DBUILD_SHARED_LIBS=OFF -DGGML_CUDA=${GGML_CUDA} -DLLAMA_CURL=OFF
 cmake --build build --config Release -j --target llama-quantize llama-gguf-split llama-imatrix
 cp ./build/bin/llama-* .
 rm -rf build