Gahow Wang
445e491123
third_party/vllm/ now tracked in git for direct patch management.
Based on vLLM v0.18.1 release with one patch applied:
vllm/v1/core/sched/scheduler.py:
Replace fatal assert with graceful skip when KV transfer callback
arrives for an already-aborted request during PD disaggregated serving.
Future vLLM modifications should be made directly in third_party/vllm/
and committed normally. The patches/ directory is kept as documentation
of what changed from upstream.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
32 lines
1.1 KiB
Python
32 lines
1.1 KiB
Python
# SPDX-License-Identifier: Apache-2.0
|
|
# SPDX-FileCopyrightText: Copyright contributors to the vLLM project
|
|
|
|
import pytest
|
|
|
|
from vllm.config import ModelConfig, VllmConfig
|
|
from vllm.inputs.preprocess import InputPreprocessor
|
|
|
|
pytestmark = pytest.mark.cpu_test
|
|
|
|
|
|
@pytest.mark.parametrize("model_id", ["facebook/chameleon-7b"])
|
|
@pytest.mark.parametrize("prompt", ["", {"prompt_token_ids": []}])
|
|
@pytest.mark.skip(
|
|
reason=(
|
|
"Applying huggingface processor on text inputs results in "
|
|
"significant performance regression for multimodal models. "
|
|
"See https://github.com/vllm-project/vllm/issues/26320"
|
|
)
|
|
)
|
|
def test_preprocessor_always_mm_code_path(model_id, prompt):
|
|
model_config = ModelConfig(model=model_id)
|
|
vllm_config = VllmConfig(model_config=model_config)
|
|
input_preprocessor = InputPreprocessor(vllm_config)
|
|
|
|
# HF processor adds sep token
|
|
tokenizer = input_preprocessor.get_tokenizer()
|
|
sep_token_id = tokenizer.vocab[tokenizer.sep_token]
|
|
|
|
processed_inputs = input_preprocessor.preprocess(prompt)
|
|
assert sep_token_id in processed_inputs["prompt_token_ids"]
|