ec2-gha/.github/workflows/demo-gpu-sweep.yml at v2 · Open-Athena/ec2-gha · 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
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
name: Demo – GPU instance sweep
on:
  workflow_dispatch:
  workflow_call:
permissions:
  id-token: write  # Required for AWS OIDC authentication
  contents: read   # Required for actions/checkout
jobs:
  # Launch GPU instances with PyTorch DLAMIs
  g4dn:
    name: 🚀 g4dn.xlarge
    uses: ./.github/workflows/runner.yml
    with:
      ec2_instance_type: g4dn.xlarge
      ec2_image_id: ami-00dddcf8fefea182f  # Deep Learning OSS PyTorch 2.5.1 Ubuntu 22.04
      instance_name: "gpu-sweep#$run g4dn"
    secrets: inherit
  g5:
    name: 🚀 g5.xlarge
    uses: ./.github/workflows/runner.yml
    with:
      ec2_instance_type: g5.xlarge
      ec2_image_id: ami-00dddcf8fefea182f  # Deep Learning OSS PyTorch 2.5.1 Ubuntu 22.04
      instance_name: "gpu-sweep#$run g5"
    secrets: inherit
  g6:
    name: 🚀 g6.xlarge
    uses: ./.github/workflows/runner.yml
    with:
      ec2_instance_type: g6.xlarge
      ec2_image_id: ami-00dddcf8fefea182f  # Deep Learning OSS PyTorch 2.5.1 Ubuntu 22.04
      instance_name: "gpu-sweep#$run g6"
    secrets: inherit
  g5g:
    name: 🚀 g5g.xlarge
    uses: ./.github/workflows/runner.yml
    with:
      ec2_instance_type: g5g.xlarge
      ec2_image_id: ami-00cbe74a3dff23b9f  # Deep Learning ARM64 OSS PyTorch 2.5.1 Ubuntu 22.04
      instance_name: "gpu-sweep#$run g5g"
    secrets: inherit

  # Test jobs for each GPU instance
  test-g4dn:
    name: 🔬 g4dn.xlarge
    needs: g4dn
    runs-on: ${{ needs.g4dn.outputs.id }}
    steps:
      - name: GPU Test
        run: |
          nvidia-smi
          # Activate PyTorch conda environment
          source /opt/conda/etc/profile.d/conda.sh
          conda activate pytorch
          python3 -c "import torch; print(f'PyTorch: {torch.__version__}, CUDA: {torch.cuda.is_available()}')"
          python3 -c "import torch; print(f'GPU: {torch.cuda.get_device_name(0)}')"

  test-g5:
    name: 🔬 g5.xlarge
    needs: g5
    runs-on: ${{ needs.g5.outputs.id }}
    steps:
      - name: GPU Test
        run: |
          nvidia-smi
          # Activate PyTorch conda environment
          source /opt/conda/etc/profile.d/conda.sh
          conda activate pytorch
          python3 -c "import torch; print(f'PyTorch: {torch.__version__}, GPU: {torch.cuda.get_device_name(0)}')"

  test-g6:
    name: 🔬 g6.xlarge
    needs: g6
    runs-on: ${{ needs.g6.outputs.id }}
    steps:
      - name: GPU Test
        run: |
          nvidia-smi
          # Activate PyTorch conda environment
          source /opt/conda/etc/profile.d/conda.sh
          conda activate pytorch
          python3 -c "import torch; print(f'PyTorch: {torch.__version__}, GPU: {torch.cuda.get_device_name(0)}')"

  test-g5g:
    name: 🔬 g5g.xlarge
    needs: g5g
    runs-on: ${{ needs.g5g.outputs.id }}
    steps:
      - name: GPU Info
        run: |
          echo "=== GPU Instance Information ==="
          echo "g5g.xlarge: AWS Graviton (ARM64) + NVIDIA T4g GPU"
          nvidia-smi
          echo ""
          echo "=== PyTorch Test ==="
          # Activate PyTorch conda environment
          source /opt/conda/etc/profile.d/conda.sh
          conda activate pytorch
          python3 -c "import torch; print(f'PyTorch: {torch.__version__}')"
          python3 -c "import torch; print(f'CUDA Available: {torch.cuda.is_available()}')"
          python3 -c "import torch; print(f'CUDA Version: {torch.version.cuda}')"
          python3 -c "import torch; print(f'GPU: {torch.cuda.get_device_name(0) if torch.cuda.is_available() else \"N/A\"}')"
      - name: Basic GPU Test
        run: |
          source /opt/conda/etc/profile.d/conda.sh
          conda activate pytorch
          python3 -c "
          import torch
          device = torch.device('cuda' if torch.cuda.is_available() else 'cpu')
          print(f'Using device: {device}')
          if device.type == 'cuda':
              x = torch.randn(1000, 1000).to(device)
              y = torch.randn(1000, 1000).to(device)
              z = torch.matmul(x, y)
              print(f'Matrix multiplication result shape: {z.shape}')
          "