Layla-Server/.github/workflows/build.yml at main · l3utterfly/Layla-Server · GitHub

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
name: Build & Release

on:
  push:
    branches: [main]

permissions:
  contents: write

jobs:
  build-windows:
    runs-on: windows-latest

    steps:
      - name: Checkout repository
        uses: actions/checkout@v4

      - name: Setup Node.js
        uses: actions/setup-node@v4
        with:
          node-version: 22
          cache: yarn

      - name: Force git to use HTTPS
        run: git config --global url."https://github.com/".insteadOf "ssh://git@github.com/"

      - name: Install dependencies
        run: yarn install --frozen-lockfile

      - name: Download llama.cpp release binaries
        id: llama
        shell: bash
        run: |
          # Fetch all asset download URLs from the latest release
          RELEASE=$(curl -s https://api.github.com/repos/ggml-org/llama.cpp/releases/latest)
          TAG=$(echo "$RELEASE" | jq -r '.tag_name')
          echo "tag=$TAG" >> "$GITHUB_OUTPUT"
          echo "Latest llama.cpp release: $TAG"

          # List all assets so we can see exact names
          echo "$RELEASE" | jq -r '.assets[].name' | head -30

          # Find and download the CUDA binary (13.1 only)
          CUDA_URL=$(echo "$RELEASE" | jq -r '.assets[] | select(.name | test("bin-win-cuda-13\\.1-x64\\.zip$")) | select(.name | test("cudart") | not) | .browser_download_url')
          echo "CUDA binary URL: $CUDA_URL"
          curl -fL -o llama-cuda.zip "$CUDA_URL"

          # Find and download the CUDA runtime (13.1 only)
          CUDART_URL=$(echo "$RELEASE" | jq -r '.assets[] | select(.name | test("cudart.*cuda-13\\.1-x64\\.zip$")) | .browser_download_url')
          echo "CUDART URL: $CUDART_URL"
          curl -fL -o llama-cudart.zip "$CUDART_URL"

      - name: Extract and merge into assets/server
        shell: bash
        run: |
          mkdir -p assets/server

          # Extract CUDA binaries first
          unzip -o llama-cuda.zip -d llama-cuda-tmp
          # Extract CUDA runtime
          unzip -o llama-cudart.zip -d llama-cudart-tmp

          # Copy CUDA files (base layer)
          cp -r llama-cuda-tmp/*/* assets/server/ 2>/dev/null || cp -r llama-cuda-tmp/* assets/server/
          # Overlay CUDA runtime files (merges on top)
          cp -r llama-cudart-tmp/*/* assets/server/ 2>/dev/null || cp -r llama-cudart-tmp/* assets/server/

          echo "Contents of assets/server:"
          ls -la assets/server/

      - name: Build Electron app
        run: yarn dist -- --publish never

      - name: Upload artifact
        uses: actions/upload-artifact@v4
        with:
          name: layla-server-win
          path: |
            dist/*.exe
            dist/*.yml
          if-no-files-found: warn

      - name: Create GitHub Release
        if: github.ref == 'refs/heads/main'
        uses: softprops/action-gh-release@v2
        with:
          tag_name: v${{ github.run_number }}
          name: Layla Server Build ${{ github.run_number }}
          files: |
            dist/*.exe
          draft: false
          prerelease: false