6ee7f5cf71
### What this PR does / why we need it? Backport: https://github.com/vllm-project/vllm-ascend/pull/252 This support speculative decoding in Ascend, including speculating with a draft model、by matching n-grams in the prompt、using MLP speculators and using EAGLE based draft models. Backport: https://github.com/vllm-project/vllm-ascend/pull/423 spec decode MultiStepWorker support TP1DraftModelRunner fully, support run the draft_model_runner with multi-step prepare on the NPU directly and support draft_model_runner use MLA. 1. before this pr, `MultiStepWorker` would not step into the branch using NPU prepare, but only into the branch using CPU prepare (`line 52` of `vllm_ascend/patch/patch_multi_step_worker.py`). Although this has `no effect` on the `correct operation` of speculative decoding and the performance of the two branches is basically the same as of the current version, I support entering this branch in this PR. In general, there are two main changes in `patch_multi_step_worker.py`: first, the `is_cuda_like()` check is removed and the `TP1DraftModelRunner` rewritten in vllm_ascend is used; second, the `supports_gpu_multi_step()` function is made to return true on NPU devices when outer Multi_step_worker could work correct. 3. before this pr, `TP1DraftModelRunner` only supports Attention on NPU, but not MLA. The relevant adaptation is in `vllm_ascend/worker/draft_model_runner.py`. Although I don’t know why the `input_positions` of `model_input.attn_metadata` in vllm-ascend needs to be added in `execute_model`, it is done in `model_runner.py`, so I also made corresponding changes. Otherwise, when atten_backend is MLA, it will prompt that input_positions cannot be found. 4. I commented out two lines in `draft_model_runner.py` in `line118` to support the scenario of K>1. ``` # lora_mapping=model_input.lora_mapping, # lora_requests=model_input.lora_requests, ``` I added comments. In the future, when vllm-ascend supports lora feature, the changes here can be restored. TODO： - [ ] revert the patch when the related issues are addressed in vllm ### How was this patch tested? CI passed with new added test. - e2e test for medusa proposer: tests/singlecard/spec_decode/e2e/test_medusa_correctness.py - e2e test for mlp proposer: tests/singlecard/spec_decode/e2e/test_mlp_correctness.py - e2e test for n-gram proposer: tests/singlecard/spec_decode/e2e/test_ngram_correctness.py Tests for patched files: - tests/singlecard/spec_decode/test_dynamic_spec_decode.py - tests/singlecard/spec_decode/test_multi_step_worker.py - tests/singlecard/spec_decode/test_ngram_worker.py - tests/singlecard/spec_decode/test_spec_decode_worker.py --------- Signed-off-by: MengqingCao <cmq0113@163.com> Co-authored-by: mengwei805 <mengwei25@huawei.com>
57 lines
1.9 KiB
Python
57 lines
1.9 KiB
Python
#
|
|
# Copyright (c) 2025 Huawei Technologies Co., Ltd. All Rights Reserved.
|
|
# 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.
|
|
# This file is a part of the vllm-ascend project.
|
|
# Adapted from vllm/tests/basic_correctness/test_basic_correctness.py
|
|
#
|
|
"""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 tests.conftest import VllmRunner
|
|
|
|
os.environ["PYTORCH_NPU_ALLOC_CONF"] = "max_split_size_mb:256"
|
|
|
|
|
|
@pytest.mark.parametrize("model, distributed_executor_backend", [
|
|
("Qwen/QwQ-32B", "mp"),
|
|
])
|
|
def test_models_distributed(model: str,
|
|
distributed_executor_backend: str) -> None:
|
|
example_prompts = [
|
|
"vLLM is a high-throughput and memory-efficient inference and serving engine for LLMs.",
|
|
"Briefly describe the major milestones in the development of artificial intelligence from 1950 to 2020.",
|
|
"Compare and contrast artificial intelligence with human intelligence in terms of processing information.",
|
|
]
|
|
dtype = "half"
|
|
max_tokens = 5
|
|
with VllmRunner(
|
|
model,
|
|
dtype=dtype,
|
|
tensor_parallel_size=4,
|
|
distributed_executor_backend=distributed_executor_backend,
|
|
) as vllm_model:
|
|
vllm_model.generate_greedy(example_prompts, max_tokens)
|
|
|
|
|
|
if __name__ == "__main__":
|
|
import pytest
|
|
pytest.main([__file__])
|