EngineX/xc-llm-ascend
3
0
Fork 0
Code Issues Pull Requests Projects Releases Wiki Activity

[CI] Fix mypy CI (#443)

Browse Source 
### What this PR does / why we need it?
Fix CI by updating mypy and pining numpy version

_the modification of model_runner_v1 is just to make CI happy_

### Does this PR introduce _any_ user-facing change?
N/A

### How was this patch tested?
CI passed

Signed-off-by: MengqingCao <cmq0113@163.com>
This commit is contained in:
Mengqing Cao
2025-04-01 09:25:33 +08:00
committed by GitHub
parent c42e21a5aa
commit 2dbd763584
4 changed files with 8 additions and 9 deletions
Download Patch File Download Diff File Expand all files Collapse all files

6
.github/workflows/vllm_ascend_test_main.yaml vendored
View File

@@ -63,10 +63,6 @@ jobs:
run: | run: |
apt-get -y install `cat packages.txt` apt-get -y install `cat packages.txt`
- name: Install dependencies
run: |
pip install -r requirements-dev.txt
- name: Checkout vllm-project/vllm repo - name: Checkout vllm-project/vllm repo
uses: actions/checkout@v4 uses: actions/checkout@v4
with: with:
@@ -80,6 +76,8 @@ jobs:
- name: Install vllm-project/vllm-ascend - name: Install vllm-project/vllm-ascend
run: | run: |
pip uninstall -y numpy
pip install -r requirements-dev.txt
pip install -e . pip install -e .
- name: Install pta - name: Install pta

2
requirements-lint.txt
View File

@@ -9,7 +9,7 @@ clang-format==18.1.5
sphinx-lint==1.0.0 sphinx-lint==1.0.0
# type checking # type checking
mypy==1.11.1 mypy==1.15.0
types-PyYAML types-PyYAML
types-requests types-requests
types-setuptools types-setuptools

1
requirements.txt
View File

@@ -3,3 +3,4 @@ pyyaml
scipy scipy
setuptools setuptools
setuptools-scm setuptools-scm
numpy==1.26.4

8
vllm_ascend/worker/model_runner_v1.py
View File

@@ -21,6 +21,7 @@ import gc
from typing import TYPE_CHECKING, Dict, List, Optional, Union from typing import TYPE_CHECKING, Dict, List, Optional, Union
import numpy as np import numpy as np
import numpy.typing as npt
import torch import torch
import torch.distributed import torch.distributed
import torch.nn as nn import torch.nn as nn
@@ -171,10 +172,9 @@ class NPUModelRunner:
device=self.device) device=self.device)
# OPTIMIZATION: Cache the tensors rather than creating them every step. # OPTIMIZATION: Cache the tensors rather than creating them every step.
self.arange_np = np.arange(max(self.max_num_reqs + 1, self.arange_np: npt.NDArray[np.int32] = np.arange(max(
self.max_model_len, self.max_num_reqs + 1, self.max_model_len, self.max_num_tokens),
self.max_num_tokens), dtype=np.int32)
dtype=np.int32)
# NOTE(woosuk): These tensors are "stateless", i.e., they are literally # NOTE(woosuk): These tensors are "stateless", i.e., they are literally
# a faster version of creating a new tensor every time. Thus, we should # a faster version of creating a new tensor every time. Thus, we should
# not make any assumptions about the values in these tensors. # not make any assumptions about the values in these tensors.