minke.yu/sglang
1
0
Fork 0
Code Issues Pull Requests Actions 12 Packages Projects Releases Wiki Activity

[Diffusion] [NPU] Fix CI run (#18921)

Browse Source
This commit is contained in:
Makcum888e
2026-02-17 16:54:19 +03:00
committed by GitHub
parent 8bb1037796
commit 2aa0db7d9c
4 changed files with 9 additions and 9 deletions
Download Patch File Download Diff File Expand all files Collapse all files
.github/workflows/pr-test-npu.yml
+3 -3
View File
@@ -28,9 +28,9 @@ jobs:
check-changes: check-changes:
runs-on: ubuntu-latest runs-on: ubuntu-latest
outputs: outputs:
changes_exist: ${{ steps.filter.outputs.main_package || steps.filter.outputs.multimodal_gen || steps.run-mode.outputs.run_all_tests}} changes_exist: ${{ steps.filter.outputs.main_package == 'true' || steps.filter.outputs.multimodal_gen == 'true' || steps.run-mode.outputs.run_all_tests == 'true'}}
main_package: ${{ steps.filter.outputs.main_package || steps.run-mode.outputs.run_all_tests }} main_package: ${{ steps.filter.outputs.main_package == 'true' || steps.run-mode.outputs.run_all_tests == 'true' }}
multimodal_gen: ${{ steps.filter.outputs.multimodal_gen || steps.run-mode.outputs.run_all_tests }} multimodal_gen: ${{ steps.filter.outputs.multimodal_gen == 'true' || steps.run-mode.outputs.run_all_tests == 'true' }}
steps: steps:
- name: Checkout code - name: Checkout code
uses: actions/checkout@v4 uses: actions/checkout@v4
python/sglang/multimodal_gen/runtime/managers/gpu_worker.py
+2 -2
View File
@@ -154,8 +154,8 @@ class GPUWorker:
output_batch.timings.record_memory_snapshot("mem_analysis", final_snapshot) output_batch.timings.record_memory_snapshot("mem_analysis", final_snapshot)
# for details on max_memory_reserved: https://docs.pytorch.org/docs/stable/generated/torch.cuda.memory.max_memory_reserved.html # for details on max_memory_reserved: https://docs.pytorch.org/docs/stable/generated/torch.cuda.memory.max_memory_reserved.html
peak_reserved_bytes = torch.cuda.max_memory_reserved() peak_reserved_bytes = torch.get_device_module().max_memory_reserved()
peak_allocated_bytes = torch.cuda.max_memory_allocated() peak_allocated_bytes = torch.get_device_module().max_memory_allocated()
output_batch.peak_memory_mb = peak_reserved_bytes / (1024**2) output_batch.peak_memory_mb = peak_reserved_bytes / (1024**2)
peak_reserved_gb = peak_reserved_bytes / (1024**3) peak_reserved_gb = peak_reserved_bytes / (1024**3)
python/sglang/multimodal_gen/test/server/ascend/perf_baselines_npu.json
+3 -3
View File
@@ -17,9 +17,9 @@
"per_frame_generation": null "per_frame_generation": null
}, },
"denoise_step_ms": { "denoise_step_ms": {
"0": 195.27, "0": 364.97,
"1": 329.05, "1": 542.66,
"2": 545.43, "2": 542.1,
"3": 541.3, "3": 541.3,
"4": 537.07, "4": 537.07,
"5": 537.21, "5": 537.21,
python/sglang/multimodal_gen/test/server/ascend/testcase_configs_npu.py
+1 -1
View File
@@ -12,7 +12,7 @@ ONE_NPU_CASES: list[DiffusionTestCase] = [
DiffusionServerArgs( DiffusionServerArgs(
model_path="/root/.cache/modelscope/hub/models/Wan-AI/Wan2.1-T2V-1.3B-Diffusers", model_path="/root/.cache/modelscope/hub/models/Wan-AI/Wan2.1-T2V-1.3B-Diffusers",
modality="video", modality="video",
warmup=0, warmup=True,
custom_validator="video", custom_validator="video",
), ),
DiffusionSamplingParams( DiffusionSamplingParams(