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

[2/4][Refactor] Refactor torchair utils (#1892)

Browse Source 
There is a lot torchair specified logic in common code. It results hard
code maintenance. We will create a new torchair module to launch
torchair related logic there. I plan to add 4 PR.

1. Refactor worker
2. Refactor utils (this PR)
- simple change that move all torchair related util function to torchair
module
3. Refactor model_runner
4. Refactor attention

- vLLM version: v0.9.2
- vLLM main:
8188196a1c

Signed-off-by: wangxiyuan <wangxiyuan1007@gmail.com>
This commit is contained in:
wangxiyuan
2025-07-21 19:43:30 +08:00
committed by GitHub
parent 957b0b611f
commit 7265dc090d
11 changed files with 142 additions and 136 deletions
Download Patch File Download Diff File Expand all files Collapse all files

0
tests/ut/__init__.py Normal file
View File

21
tests/ut/test_utils.py
View File

@@ -290,27 +290,6 @@ class TestUtils(TestBase):
3,
len(test_vllm_config.compilation_config.cudagraph_capture_sizes))
def test_get_torchair_current_work_dir(self):
cache_dir = utils.TORCHAIR_CACHE_DIR
work_dir = utils.get_torchair_current_work_dir()
self.assertEqual(cache_dir, work_dir)
work_dir = utils.get_torchair_current_work_dir("test")
self.assertEqual(os.path.join(cache_dir, "test"), work_dir)
def test_torchair_cache_dir(self):
utils.write_kv_cache_bytes_to_file(0, 100)
self.assertTrue(utils.check_torchair_cache_exist(),
"Create torchair cache dir failed")
self.assertTrue(utils.check_kv_cache_bytes_cache_exist(),
"Create kv cache bytes cache dir failed")
kv_cache_bytes = utils.read_kv_cache_bytes_from_file(0)
self.assertEqual(100, kv_cache_bytes)
utils.delete_torchair_cache_file()
self.assertFalse(utils.check_torchair_cache_exist(),
"Delete torchair cache dir failed")
self.assertFalse(utils.check_kv_cache_bytes_cache_exist(),
"Delete kv cache bytes cache dir failed")
@mock.patch("vllm.model_executor.custom_op.CustomOp")
@mock.patch("vllm_ascend.ops.activation.AscendQuickGELU")
@mock.patch("vllm_ascend.ops.activation.AscendSiluAndMul")

0
tests/ut/torchair/__init__.py Normal file
View File

28
tests/ut/torchair/test_utils.py Normal file
View File

@@ -0,0 +1,28 @@
import os
from tests.ut.base import TestBase
from vllm_ascend.torchair import utils
class TestTorchairUtils(TestBase):
def test_get_torchair_current_work_dir(self):
cache_dir = utils.TORCHAIR_CACHE_DIR
work_dir = utils._get_torchair_current_work_dir()
self.assertEqual(cache_dir, work_dir)
work_dir = utils._get_torchair_current_work_dir("test")
self.assertEqual(os.path.join(cache_dir, "test"), work_dir)
def test_torchair_cache_dir(self):
utils.write_kv_cache_bytes_to_file(0, 100)
self.assertTrue(utils.check_torchair_cache_exist(),
"Create torchair cache dir failed")
self.assertTrue(utils.check_kv_cache_bytes_cache_exist(),
"Create kv cache bytes cache dir failed")
kv_cache_bytes = utils.read_kv_cache_bytes_from_file(0)
self.assertEqual(100, kv_cache_bytes)
utils.delete_torchair_cache_file()
self.assertFalse(utils.check_torchair_cache_exist(),
"Delete torchair cache dir failed")
self.assertFalse(utils.check_kv_cache_bytes_cache_exist(),
"Delete kv cache bytes cache dir failed")