d5e7756028
### What this PR does / why we need it?
vLLM Ascend plugin (vllm-ascend) is a backend plugin for running vLLM on
the Ascend NPU.
This plugin is the recommended approach for supporting the Ascend
backend within the vLLM community. It adheres to the principles outlined
in the [RFC]: Hardware pluggable, providing a hardware-pluggable
interface that decouples the integration of the Ascend NPU with vLLM.
This patch also include changes to make CI work and use cache speed up
e2e test, including:
1. Change push (post merge ci) and pull_request (pr ci) trigger branch
to main
2. Make mypy work by ignore base_communicator and clear unused deps
3. Several improvements for vllm_ascend_test:
- use cache (pip, ms, hf) speed up e2e test (25mins --> 5mins)
- switch `git clone` command to `action/checkout` to speedup checkout
and
- Enable sv for pytest for better info dump
- Remove network host to resole `docker: conflicting ontions: cannot
attach both user-defined and non-user-definednetwork-modes`, which is a
problem on docker 1.45 but not on 1.39.
4. Adapt MLA decode optimizations:
cabaf4eff3
### Does this PR introduce _any_ user-facing change?
Yes, init the PR.
### How was this patch tested?
- This is the first PR to make ascend NPU work on vLLM. All code is
tested on ascend with vLLM V0 Engine.
- CI passed
---------
Signed-off-by: wangxiyuan <wangxiyuan1007@gmail.com>
Signed-off-by: Yikun Jiang <yikunkero@gmail.com>
Co-authored-by: wangxiyuan <wangxiyuan1007@gmail.com>
Co-authored-by: MengqingCao <cmq0113@163.com>
Co-authored-by: wangshuai09 <391746016@qq.com>
Co-authored-by: Shanshan Shen <467638484@qq.com>
Co-authored-by: wangli <wangli858794774@gmail.com>
62 lines
2.0 KiB
Python
62 lines
2.0 KiB
Python
#
|
|
# Copyright (c) 2025 Huawei Technologies Co., Ltd. All Rights Reserved.
|
|
# This file is a part of the vllm-ascend project.
|
|
# Adapted from vllm/tests/basic_correctness/test_basic_correctness.py
|
|
# Copyright 2023 The vLLM team.
|
|
#
|
|
# Licensed under the Apache License, Version 2.0 (the "License");
|
|
# you may not use this file except in compliance with the License.
|
|
# You may obtain a copy of the License at
|
|
#
|
|
# http://www.apache.org/licenses/LICENSE-2.0
|
|
#
|
|
# Unless required by applicable law or agreed to in writing, software
|
|
# distributed under the License is distributed on an "AS IS" BASIS,
|
|
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
|
|
# See the License for the specific language governing permissions and
|
|
# limitations under the License.
|
|
#
|
|
"""Compare the short outputs of HF and vLLM when using greedy sampling.
|
|
|
|
Run `pytest tests/test_offline_inference.py`.
|
|
"""
|
|
import os
|
|
|
|
import pytest
|
|
import vllm # noqa: F401
|
|
from conftest import VllmRunner
|
|
|
|
import vllm_ascend # noqa: F401
|
|
|
|
MODELS = [
|
|
"Qwen/Qwen2.5-0.5B-Instruct",
|
|
]
|
|
os.environ["VLLM_USE_MODELSCOPE"] = "True"
|
|
|
|
TARGET_TEST_SUITE = os.environ.get("TARGET_TEST_SUITE", "L4")
|
|
|
|
|
|
@pytest.mark.parametrize("model", MODELS)
|
|
@pytest.mark.parametrize("dtype", ["half", "float16"])
|
|
@pytest.mark.parametrize("max_tokens", [5])
|
|
def test_models(
|
|
model: str,
|
|
dtype: str,
|
|
max_tokens: int,
|
|
) -> None:
|
|
os.environ["VLLM_ATTENTION_BACKEND"] = "ASCEND"
|
|
|
|
# 5042 tokens for gemma2
|
|
# gemma2 has alternating sliding window size of 4096
|
|
# we need a prompt with more than 4096 tokens to test the sliding window
|
|
prompt = "The following numbers of the sequence " + ", ".join(
|
|
str(i) for i in range(1024)) + " are:"
|
|
example_prompts = [prompt]
|
|
|
|
with VllmRunner(model,
|
|
max_model_len=8192,
|
|
dtype=dtype,
|
|
enforce_eager=False,
|
|
gpu_memory_utilization=0.7) as vllm_model:
|
|
vllm_model.generate_greedy(example_prompts, max_tokens)
|