[MM][Perf][CG] Support ViT full CUDA graph for Qwen3.5 (#42151)

Signed-off-by: shen-shanshan <467638484@qq.com>
This commit is contained in:
Shanshan Shen
2026-05-13 16:00:32 +08:00
committed by GitHub
parent 85b2fecab7
commit 92def124bc
4 changed files with 112 additions and 5 deletions
@@ -42,6 +42,18 @@ def qwen_vl_chat_template(content: str) -> str:
MODEL_CONFIGS: dict[str, VitCudagraphTestConfig] = {
"qwen2_5_vl": VitCudagraphTestConfig(
model="Qwen/Qwen2.5-VL-3B-Instruct",
image_prompt=qwen_vl_chat_template(
"<|vision_start|><|image_pad|><|vision_end|>What is in this image?"
),
video_prompt=qwen_vl_chat_template(
"<|vision_start|><|video_pad|><|vision_end|>"
"Describe this video in one sentence."
),
needs_video_metadata=False,
marks=[pytest.mark.core_model],
),
"qwen3_vl": VitCudagraphTestConfig(
model="Qwen/Qwen3-VL-2B-Instruct",
image_prompt=qwen_vl_chat_template(
@@ -54,8 +66,8 @@ MODEL_CONFIGS: dict[str, VitCudagraphTestConfig] = {
needs_video_metadata=True,
marks=[pytest.mark.core_model],
),
"qwen2_5_vl": VitCudagraphTestConfig(
model="Qwen/Qwen2.5-VL-3B-Instruct",
"qwen3_5": VitCudagraphTestConfig(
model="Qwen/Qwen3.5-0.8B",
image_prompt=qwen_vl_chat_template(
"<|vision_start|><|image_pad|><|vision_end|>What is in this image?"
),
@@ -63,7 +75,7 @@ MODEL_CONFIGS: dict[str, VitCudagraphTestConfig] = {
"<|vision_start|><|video_pad|><|vision_end|>"
"Describe this video in one sentence."
),
needs_video_metadata=False,
needs_video_metadata=True,
marks=[pytest.mark.core_model],
),
}