Model-Optimizer/.github/workflows/example_tests.yml at main · Distillative-AI/Model-Optimizer · 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
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
name: Example tests

on:
  push:
    branches: ["pull-request/[0-9]+"]
    # NOTE: paths cannot be used since push happens to copied PR and only latest commit to PR is used
  schedule:
    - cron: "0 0 * * *" # Nightly
  workflow_dispatch: # On-demand

# Cancel previous runs if new commit is pushed to the same PR
concurrency:
  group: ${{ github.workflow }}-${{ startsWith(github.ref, 'refs/heads/pull-request/') && github.ref || github.sha }}
  cancel-in-progress: true

jobs:
  check-file-changes:
    if: startsWith(github.ref, 'refs/heads/pull-request/')
    runs-on: ubuntu-latest
    outputs:
      any_changed: ${{ steps.changed-tests.outputs.any_changed }}
    steps:
      - uses: actions/checkout@v6
        with:
          fetch-depth: 0
      - id: get-pr-info
        uses: nv-gha-runners/get-pr-info@main
      # Get commit from main branch that is present in the PR to use as base for changed files
      - id: calculate-merge-base
        env:
          PR_SHA: ${{ fromJSON(steps.get-pr-info.outputs.pr-info).head.sha }}
          BASE_SHA: ${{ fromJSON(steps.get-pr-info.outputs.pr-info).base.sha }}
        run: |
          (echo -n "merge-base="; git merge-base "$BASE_SHA" "$PR_SHA") | tee --append "${GITHUB_OUTPUT}"
      - name: Check for changes in test-relevant directories
        id: changed-tests
        uses: step-security/changed-files@v46.0.5
        with:
          base_sha: ${{ steps.calculate-merge-base.outputs.merge-base }}
          sha: ${{ fromJSON(steps.get-pr-info.outputs.pr-info).head.sha }}
          files: |
            .github/workflows/example_tests.yml
            examples/**
            modelopt/**
            pyproject.toml
            tests/examples/**
          fail_on_initial_diff_error: true
  wait-checks:
    needs: [check-file-changes]
    if: needs.check-file-changes.outputs.any_changed == 'true'
    uses: ./.github/workflows/_wait_for_checks.yml
    permissions:
      checks: read
    secrets: inherit
    with:
      match_pattern: "^DCO$|^linux$" # Wait for DCO and Unit tests / linux to pass
      delay: 300s

  ##### PyTorch Example Tests (speculative_decoding requires 26.01 image) #####
  torch-pr:
    needs: [check-file-changes, wait-checks]
    if: startsWith(github.ref, 'refs/heads/pull-request/') && needs.check-file-changes.outputs.any_changed == 'true'
    strategy: &torch_strategy
      fail-fast: false
      matrix:
        example: [llm_distill, llm_qat, llm_sparsity]
        include:
          - example: speculative_decoding
            docker_image: "26.01"
    uses: ./.github/workflows/_example_tests_runner.yml
    secrets: inherit
    with:
      docker_image: "nvcr.io/nvidia/pytorch:${{ matrix.docker_image || '26.01' }}-py3"
      example: ${{ matrix.example }}
      timeout_minutes: 30
      pip_install_extras: "[hf,dev-test]"
      runner: linux-amd64-gpu-h100-latest-1

  torch-non-pr:
    if: ${{ !startsWith(github.ref, 'refs/heads/pull-request/') }}
    strategy: *torch_strategy
    uses: ./.github/workflows/_example_tests_runner.yml
    secrets: inherit
    with:
      docker_image: "nvcr.io/nvidia/pytorch:${{ matrix.docker_image || '26.01' }}-py3"
      example: ${{ matrix.example }}
      timeout_minutes: 30
      pip_install_extras: "[hf,dev-test]"
      runner: linux-amd64-gpu-rtxpro6000-latest-2

  ##### TensorRT-LLM Example Tests #####
  trtllm-pr:
    needs: [check-file-changes, wait-checks]
    if: startsWith(github.ref, 'refs/heads/pull-request/') && needs.check-file-changes.outputs.any_changed == 'true'
    strategy:
      fail-fast: false
      matrix:
        example: [llm_ptq, vlm_ptq]
    uses: ./.github/workflows/_example_tests_runner.yml
    secrets: inherit
    with:
      docker_image: "nvcr.io/nvidia/tensorrt-llm/release:1.3.0rc5"
      example: ${{ matrix.example }}
      pip_install_extras: "[hf,dev-test]"
      runner: linux-amd64-gpu-rtxpro6000-latest-1

  trtllm-non-pr:
    if: ${{ !startsWith(github.ref, 'refs/heads/pull-request/') }}
    strategy:
      fail-fast: false
      matrix:
        example: [llm_autodeploy, llm_eval, llm_ptq, vlm_ptq]
    uses: ./.github/workflows/_example_tests_runner.yml
    secrets: inherit
    with:
      docker_image: "nvcr.io/nvidia/tensorrt-llm/release:1.3.0rc5"
      example: ${{ matrix.example }}
      pip_install_extras: "[hf,dev-test]"
      runner: linux-amd64-gpu-rtxpro6000-latest-2

  ##### ONNX/TensorRT Example Tests #####
  onnx-pr:
    needs: [check-file-changes, wait-checks]
    if: startsWith(github.ref, 'refs/heads/pull-request/') && needs.check-file-changes.outputs.any_changed == 'true'
    strategy: &onnx_strategy
      fail-fast: false
      matrix:
        example: [diffusers, torch_onnx]
    uses: ./.github/workflows/_example_tests_runner.yml
    secrets: inherit
    with:
      docker_image: "nvcr.io/nvidia/tensorrt:26.01-py3"
      example: ${{ matrix.example }}
      pip_install_extras: "[all,dev-test]"
      runner: linux-amd64-gpu-l4-latest-1

  onnx-non-pr:
    if: ${{ !startsWith(github.ref, 'refs/heads/pull-request/') }}
    strategy: *onnx_strategy
    uses: ./.github/workflows/_example_tests_runner.yml
    secrets: inherit
    with:
      docker_image: "nvcr.io/nvidia/tensorrt:26.01-py3"
      example: ${{ matrix.example }}
      pip_install_extras: "[all,dev-test]"
      runner: linux-amd64-gpu-rtxpro6000-latest-2

  ##### Required Check for PR #####
  example-pr-required-check:
    # Run even if example tests are skipped
    if: ${{ startsWith(github.ref, 'refs/heads/pull-request/') && always() }}
    needs: [check-file-changes, torch-pr, trtllm-pr, onnx-pr]
    runs-on: ubuntu-latest
    steps:
      - name: Required GPU tests did not succeed
        if: |
          needs.check-file-changes.result != 'success' ||
          (needs.check-file-changes.outputs.any_changed == 'true' && (
            needs.torch-pr.result != 'success' ||
            needs.trtllm-pr.result != 'success' ||
            needs.onnx-pr.result != 'success'
          ))
        run: exit 1