|
7 | 7 | schedule: |
8 | 8 | - cron: "0 0 * * *" # Nightly |
9 | 9 | workflow_dispatch: # On-demand |
10 | | - inputs: |
11 | | - torch_examples: |
12 | | - description: "PyTorch examples to test (space-separated, empty to skip)" |
13 | | - type: string |
14 | | - default: "" |
15 | | - trtllm_examples: |
16 | | - description: "TensorRT-LLM examples to test (space-separated, empty to skip)" |
17 | | - type: string |
18 | | - default: "llm_ptq vlm_ptq" |
19 | | - onnx_examples: |
20 | | - description: "ONNX/TensorRT examples to test (space-separated, empty to skip)" |
21 | | - type: string |
22 | | - default: "diffusers" |
23 | 10 |
|
24 | 11 | # Cancel previous runs if new commit is pushed to the same PR |
25 | 12 | concurrency: |
@@ -73,71 +60,83 @@ jobs: |
73 | 60 | example-torch-pr: |
74 | 61 | needs: [check-file-changes, wait-checks] |
75 | 62 | if: startsWith(github.ref, 'refs/heads/pull-request/') && needs.check-file-changes.outputs.any_changed == 'true' |
| 63 | + strategy: |
| 64 | + matrix: |
| 65 | + example: [llm_distill, llm_qat, llm_sparsity, speculative_decoding] |
76 | 66 | uses: ./.github/workflows/_example_tests_runner.yml |
77 | 67 | secrets: inherit |
78 | 68 | with: |
79 | 69 | docker_image: "nvcr.io/nvidia/pytorch:25.06-py3" |
80 | | - examples: "llm_distill llm_qat llm_sparsity speculative_decoding" |
81 | | - timeout_minutes: 60 |
| 70 | + example: ${{ matrix.example }} |
82 | 71 | pip_install_extras: "[hf,dev-test]" |
83 | 72 | runner: linux-amd64-gpu-l4-latest-1 |
84 | 73 |
|
85 | 74 | example-torch-non-pr: |
86 | | - if: ${{ !startsWith(github.ref, 'refs/heads/pull-request/') && (github.event_name != 'workflow_dispatch' || inputs.torch_examples != '') }} |
| 75 | + if: ${{ !startsWith(github.ref, 'refs/heads/pull-request/') }} |
| 76 | + strategy: |
| 77 | + matrix: |
| 78 | + example: [llm_distill, llm_qat, llm_sparsity, speculative_decoding] |
87 | 79 | uses: ./.github/workflows/_example_tests_runner.yml |
88 | 80 | secrets: inherit |
89 | 81 | with: |
90 | 82 | docker_image: "nvcr.io/nvidia/pytorch:25.06-py3" |
91 | | - examples: ${{ inputs.torch_examples || 'llm_distill llm_qat llm_sparsity speculative_decoding' }} |
92 | | - timeout_minutes: 60 |
| 83 | + example: ${{ matrix.example }} |
93 | 84 | pip_install_extras: "[hf,dev-test]" |
94 | 85 | runner: linux-amd64-gpu-h100-latest-2 |
95 | 86 |
|
96 | 87 | ##### TensorRT-LLM Example Tests ##### |
97 | 88 | example-trtllm-pr: |
98 | 89 | needs: [check-file-changes, wait-checks] |
99 | 90 | if: startsWith(github.ref, 'refs/heads/pull-request/') && needs.check-file-changes.outputs.any_changed == 'true' |
| 91 | + strategy: |
| 92 | + matrix: |
| 93 | + example: [llm_ptq, vlm_ptq] |
100 | 94 | uses: ./.github/workflows/_example_tests_runner.yml |
101 | 95 | secrets: inherit |
102 | 96 | with: |
103 | 97 | docker_image: "nvcr.io/nvidia/tensorrt-llm/release:1.1.0rc2.post2" |
104 | | - examples: "llm_autodeploy llm_eval llm_ptq vlm_ptq" |
105 | | - timeout_minutes: 90 |
| 98 | + example: ${{ matrix.example }} |
106 | 99 | pip_install_extras: "[hf,dev-test]" |
107 | 100 | runner: linux-amd64-gpu-h100-latest-1 |
108 | 101 |
|
109 | 102 | example-trtllm-non-pr: |
110 | | - if: ${{ !startsWith(github.ref, 'refs/heads/pull-request/') && (github.event_name != 'workflow_dispatch' || inputs.trtllm_examples != '') }} |
| 103 | + if: ${{ !startsWith(github.ref, 'refs/heads/pull-request/') }} |
| 104 | + strategy: |
| 105 | + matrix: |
| 106 | + example: [llm_autodeploy, llm_eval, llm_ptq, vlm_ptq] |
111 | 107 | uses: ./.github/workflows/_example_tests_runner.yml |
112 | 108 | secrets: inherit |
113 | 109 | with: |
114 | 110 | docker_image: "nvcr.io/nvidia/tensorrt-llm/release:1.1.0rc2.post2" |
115 | | - examples: ${{ inputs.trtllm_examples || 'llm_autodeploy llm_eval llm_ptq vlm_ptq' }} |
116 | | - timeout_minutes: 90 |
| 111 | + example: ${{ matrix.example }} |
117 | 112 | pip_install_extras: "[hf,dev-test]" |
118 | 113 | runner: linux-amd64-gpu-h100-latest-2 |
119 | 114 |
|
120 | 115 | ##### ONNX/TensorRT Example Tests ##### |
121 | 116 | example-onnx-pr: |
122 | 117 | needs: [check-file-changes, wait-checks] |
123 | 118 | if: startsWith(github.ref, 'refs/heads/pull-request/') && needs.check-file-changes.outputs.any_changed == 'true' |
| 119 | + strategy: |
| 120 | + matrix: |
| 121 | + example: [diffusers, onnx_ptq] |
124 | 122 | uses: ./.github/workflows/_example_tests_runner.yml |
125 | 123 | secrets: inherit |
126 | 124 | with: |
127 | 125 | docker_image: "nvcr.io/nvidia/tensorrt:25.08-py3" |
128 | | - examples: "diffusers onnx_ptq" |
129 | | - timeout_minutes: 60 |
| 126 | + example: ${{ matrix.example }} |
130 | 127 | pip_install_extras: "[all,dev-test]" |
131 | 128 | runner: linux-amd64-gpu-l4-latest-1 |
132 | 129 |
|
133 | 130 | example-onnx-non-pr: |
134 | | - if: ${{ !startsWith(github.ref, 'refs/heads/pull-request/') && (github.event_name != 'workflow_dispatch' || inputs.onnx_examples != '') }} |
| 131 | + if: ${{ !startsWith(github.ref, 'refs/heads/pull-request/') }} |
| 132 | + strategy: |
| 133 | + matrix: |
| 134 | + example: [diffusers, onnx_ptq] |
135 | 135 | uses: ./.github/workflows/_example_tests_runner.yml |
136 | 136 | secrets: inherit |
137 | 137 | with: |
138 | 138 | docker_image: "nvcr.io/nvidia/tensorrt:25.08-py3" |
139 | | - examples: ${{ inputs.onnx_examples || 'diffusers onnx_ptq' }} |
140 | | - timeout_minutes: 60 |
| 139 | + example: ${{ matrix.example }} |
141 | 140 | pip_install_extras: "[all,dev-test]" |
142 | 141 | runner: linux-amd64-gpu-l4-latest-1 |
143 | 142 |
|
|
0 commit comments