[Bugfix] Fix HunyuanVideo-1.5 CI failures (#2066)

Signed-off-by: lishunyang <lishunyang12@163.com>
This commit is contained in:
SYLAR authored and GitHub committed 2026-03-22 17:10:55 +13:00
1 parent 1bd1d9d618
commit a2a23a1499
2 files changed
+11 -6

No files matched your search

@@ -5,7 +5,7 @@ Tests of common diffusion feature combinations in online serving mode
for HunyuanVideo-1.5-T2V (480p).
Coverage (H100, since model cannot fit L4):
- TeaCache + Layerwise CPU offloading (1 GPU)
- CacheDiT + Layerwise CPU offloading (1 GPU)
- CacheDiT + TP=2 + VAE patch parallel=2 (2 GPUs)
"""
@@ -34,17 +34,17 @@ def _get_diffusion_feature_cases(model: str):
Designed for 2x H100 environment per issue #1832.
"""
return [
# (1 GPU) TeaCache + Layerwise CPU offloading
# (1 GPU) CacheDiT + Layerwise CPU offloading
pytest.param(
OmniServerParams(
model=model,
server_args=[
"--cache-backend",
"tea_cache",
"cache_dit",
"--enable-layerwise-offload",
],
),
id="single_card_teacache_layerwise",
id="single_card_cachedit_layerwise",
marks=SINGLE_CARD_MARKS,
),
# (2 GPUs) CacheDiT + TP=2 + VAE patch parallel=2