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

Add three new test cases (#35502)

Browse Source 
Co-authored-by: HeYao <heyao@example.com>
This commit is contained in:
HeYao
2026-08-19 17:59:45 +08:00
committed by GitHub
co-authored by HeYao
parent adca19c497
commit f22442d3a4
6 changed files with 391 additions and 17 deletions
Download Patch File Download Diff File Expand all files Collapse all files
.github/workflows/nightly-test-npu.yml
+13
View File
@@ -368,6 +368,14 @@ jobs:
test_case: test/registered/npu/performance/mimo_v2_flash/test_npu_mimo_v2_flash_1p1d_12p_in16k_out1k_tpot_20ms.py
test_type: 'perf'
prefill_decode_deployment: 'separation'
# deepseek_v4_flash performance tests
- name: deepseek_v4_flash_w8a8_1p1d_16p_in8k_out1k_50ms
prefill_size: 1
decode_size: 1
router_size: 1
test_case: test/registered/npu/performance/deepseek_v4_flash/test_npu_deepseek_v4_flash_w8a8_1p1d_16p_in8k_out1k_50ms.py
test_type: 'perf'
prefill_decode_deployment: 'separation'
uses: ./.github/workflows/nightly-test-npu-e2e-multi-node.yml
with:
runner: linux-amd64-cpu-4
@@ -401,6 +409,11 @@ jobs:
node_size: 2
test_case: test/registered/npu/accuracy/kimi_k2_6/test_npu_kimi_k2_6_w4a8_16p_in64k_out1k_100ms_aime25.py
test_type: 'accuracy'
# glm_5_2 accuracy tests
- name: glm_5_2_w4a8_16p_gpqa
node_size: 2
test_case: test/registered/npu/accuracy/glm5_2/test_npu_glm_5_2_w4a8_16p_gpqa.py
test_type: 'accuracy'
uses: ./.github/workflows/nightly-test-npu-e2e-multi-node.yml
with:
runner: linux-amd64-cpu-4