[Frontend][Core] Standardize request error handling with VLLMError hierarchy (#49665)

Signed-off-by: Zach Zhu <zzqshu@126.com>
This commit is contained in:
Zach Zhu
2026-07-29 04:20:40 +00:00
committed by GitHub
parent f37f03db4a
commit 58f9659397
46 changed files with 370 additions and 248 deletions
+5 -1
View File
@@ -3,6 +3,8 @@
import pytest
from vllm.exceptions import VLLMValidationError
from ..conftest import IMAGE_ASSETS
HF_IMAGE_PROMPTS = IMAGE_ASSETS.prompts(
@@ -19,7 +21,9 @@ models = ["llava-hf/llava-1.5-7b-hf"]
def test_context_length_too_short(vllm_runner, image_assets, model):
images = [asset.pil_image for asset in image_assets]
with pytest.raises(ValueError, match="longer than the maximum model length"):
with pytest.raises(
VLLMValidationError, match="longer than the maximum model length"
):
vllm_model = vllm_runner(
model,
# LLaVA has a feature size of 576