Add Windows CUDA llama.cpp release asset

.github/workflows/build-llamacpp-binaries.yml

@@ -82,13 +82,43 @@ jobs:
               -DGGML_FMA=ON
               -DGGML_F16C=ON
               -DGGML_BMI2=ON
+          - os: windows-latest
+            name: windows-x64-cuda
+            cuda: true
+            cuda_version: '13.2.0'
+            cmake_flags: >-
+              -DGGML_NATIVE=OFF
+              -DGGML_CUDA=ON
+              -DGGML_CUDA_FA=OFF
+              -DGGML_CUDA_NCCL=OFF
+              -DGGML_CUDA_GRAPHS=OFF
+              -DGGML_AVX=OFF
+              -DGGML_AVX2=OFF
+              -DGGML_AVX_VNNI=OFF
+              -DGGML_AVX512=OFF
+              -DGGML_AVX512_VBMI=OFF
+              -DGGML_AVX512_VNNI=OFF
+              -DGGML_AVX512_BF16=OFF
+              -DGGML_FMA=OFF
+              -DGGML_F16C=OFF
+              -DGGML_BMI2=OFF
     runs-on: ${{ matrix.os }}
     defaults:
       run:
         shell: bash
         working-directory: runtime/llama.cpp
     steps:
       - uses: actions/checkout@v4
+      - name: Install CUDA Toolkit
+        if: matrix.cuda
+        uses: Jimver/cuda-toolkit@v0.2.35
+        with:
+          cuda: ${{ matrix.cuda_version }}
+          method: network
+          log-file-suffix: ${{ matrix.name }}.txt
+      - name: Show CUDA compiler
+        if: matrix.cuda
+        run: nvcc --version
       - name: Configure and build
         run: |
           # Release binaries must be portable across user machines, not tuned to
@@ -135,4 +165,4 @@ jobs:
           gh release create "$tag" dist/* \
             --repo "${{ github.repository }}" \
             --title "FunASR llama.cpp runtime $version" \
-            --notes "Prebuilt self-contained binaries for the FunASR llama.cpp / GGUF runtime: SenseVoice, Paraformer and Fun-ASR-Nano with built-in FSMN-VAD. Download the default quantized model with \`bash download-funasr-model.sh <sensevoice|paraformer|nano>\` (the helper requires the Hugging Face CLI: \`pip install -U huggingface_hub\`), then run \`llama-funasr-cli\` / \`llama-funasr-sensevoice\` / \`llama-funasr-paraformer\`. Use the default x64 asset for maximum CPU compatibility; use the x64-avx2 asset on CPUs with AVX2/FMA/F16C/BMI2 for higher throughput. No Python ASR runtime or local build is required. Docs: $docs"
+            --notes "Prebuilt self-contained binaries for the FunASR llama.cpp / GGUF runtime: SenseVoice, Paraformer and Fun-ASR-Nano with built-in FSMN-VAD. Download the default quantized model with \`bash download-funasr-model.sh <sensevoice|paraformer|nano>\` (the helper requires the Hugging Face CLI: \`pip install -U huggingface_hub\`), then run \`llama-funasr-cli\` / \`llama-funasr-sensevoice\` / \`llama-funasr-paraformer\`. Use the default x64 asset for maximum CPU compatibility; use the x64-avx2 asset on CPUs with AVX2/FMA/F16C/BMI2 for higher throughput. The Windows CUDA asset is \`windows-x64-cuda\`; it requires an NVIDIA driver compatible with CUDA 13.2 and enables SenseVoiceSmall graph execution with \`llama-funasr-sensevoice ... --backend cuda\`. No Python ASR runtime or local build is required. Docs: $docs"

runtime/llama.cpp/README.md

@@ -72,9 +72,12 @@ cmake --build build-cuda -j --target llama-funasr-sensevoice
   -m sensevoice-small-f16.gguf -a sample.wav --backend cuda
 ```
 
-`--backend cpu` remains the default and is what the current cross-platform
-prebuilt binaries use. A binary built without `-DGGML_CUDA=ON` exits with a clear
-message if `--backend cuda` is requested.
+`--backend cpu` remains the default and is what the portable cross-platform
+prebuilt binaries use. Tagged runtime releases also publish a
+`funasr-llamacpp-windows-x64-cuda.zip` package when the Windows CUDA release job
+passes; it requires an NVIDIA driver compatible with CUDA 13.2. A binary built
+without `-DGGML_CUDA=ON` exits with a clear message if `--backend cuda` is
+requested.
 
 ## Build (shared)
 ```bash

runtime/llama.cpp/tests/test_release_workflow.py

@@ -0,0 +1,31 @@
+from pathlib import Path
+
+
+ROOT = Path(__file__).resolve().parents[3]
+WORKFLOW = ROOT / ".github" / "workflows" / "build-llamacpp-binaries.yml"
+
+
+def test_windows_cuda_release_asset_is_in_matrix():
+    workflow = WORKFLOW.read_text(encoding="utf-8")
+
+    assert "name: windows-x64-cuda" in workflow
+    assert "cuda: true" in workflow
+    assert "windows-x64-cuda" in workflow
+
+
+def test_windows_cuda_build_uses_cuda_toolkit_and_flags():
+    workflow = WORKFLOW.read_text(encoding="utf-8")
+
+    assert "Jimver/cuda-toolkit@v0.2.35" in workflow
+    assert "if: matrix.cuda" in workflow
+    assert "-DGGML_CUDA=ON" in workflow
+    assert "-DGGML_CUDA_FA=OFF" in workflow
+    assert "-DGGML_CUDA_NCCL=OFF" in workflow
+
+
+def test_release_notes_explain_cpu_and_cuda_windows_assets():
+    workflow = WORKFLOW.read_text(encoding="utf-8")
+
+    assert "windows-x64-cuda" in workflow
+    assert "--backend cuda" in workflow
+    assert "Windows CUDA" in workflow