Conversation
CLA Signature Passlicy666, thanks for your pull request. All authors of the commits have signed the CLA. 👍 |
Contributor
|
📖 文档预览：https://ascend.github.io/docs/pr-preview/pr-166/sources/lmcache/ 也可打开整站首页。 构建自 |
CLA Signature Passlicy666, thanks for your pull request. All authors of the commits have signed the CLA. 👍 |
CLA Signature Passlicy666, thanks for your pull request. All authors of the commits have signed the CLA. 👍 |
licy666
force-pushed
the
feat/lmcache-quick-start
branch
from
September 21, 2026 02:25
0a9f412 to
d21f28c
Compare
CLA Signature Passlicy666, thanks for your pull request. All authors of the commits have signed the CLA. 👍 |
把 LMCache-Ascend 的昇腾入门文档接到 Ascend/docs 的 doc-test 看护：用户文档、测试子类、薄触发器和首页卡片同一提交。Release v0.4.4 在 CANN 9.1 / vLLM 0.23 上仍需本地 CMake include 与 KV connector 补丁，CI 通过隐藏 #test-setup 落盘后再跑可见 python。
v0.4.4 的 setup.py 优先读 npu-smi，香港 A2 是 Ascend910B3，命令行 SOC_VERSION=Ascend910B4 不会进 _build_info，预期输出对不上。
#test-result 开头单独一行 ... 会要求下一句前面有换行，但补丁脚本的 stdout 第一行就是 patched_ok True，比对会失败。
licy666
force-pushed
the
feat/lmcache-quick-start
branch
from
September 21, 2026 03:18
d21f28c to
96f80d6
Compare
CLA Signature Passlicy666, thanks for your pull request. All authors of the commits have signed the CLA. 👍 |
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Summary
LMCache/LMCache-Ascend的昇腾 Quick Start 迁到Ascend/docs：用户文档、测试子类、薄触发器、首页卡片和 toctree 同一 PR。主仓LMCache/LMCache没有昇腾实现，monitor 的upstream_repo是LMCache/LMCache-Ascend。Qwen/Qwen2.5-0.5B-Instruct。薄触发器不传container_options/--volume，镜像钉cann:9.1.0-910b-ubuntu22.04-py3.12，runner 单卡linux-aarch64-a2-1。vllm==0.23.0→vllm-ascend==0.23.0→--force-reinstall --no-deps triton-ascend==3.2.2。补丁脚本和离线卸载脚本拆成无标签.py+ 隐藏#test-setupcat+ 可见python path.py。工作负载#test-result含设备锚点Platform plugin ascend is activated/Using NPU for LMCache engine/current_device npu:0。选型（不进用户正文）
v0.4.4，不跟main。这个 tag 在 CANN 9.1 上编 HIXL 缺pkg_inc这一层 include（PR 288 已合进 main，尚未进 Release）；子类LMCacheAscendConnectorV1Dynamic到 main 仍是两参数构造，vLLM 0.23 工厂会因缺少kv_cache_config直接ValueError。因此克隆后必须先跑本地补丁再编译；源码已有对应内容时脚本跳过。#test-setup，不能把#test藏进去。可见无标签 Python 围栏给人复制保存，CI 靠注释里的cat落盘，再由可见#test真正执行python patch_lmcache_ascend.py。quick-start-template.yml/tests/doc_test/。Test plan
actionlint薄触发器py_compile测试子类#test（check-tools/install-system-prereqs/install-vllm/install-lmcache/clone-lmcache-ascend/install-lmcache-ascend/offload），5 个隐藏#test-setup；注释里没有#test/#test-resultcosdt-ci-test//root/.cache/modelscope/gh-proxy/python - <<patch_lmcache_ascend.py与offload_qs.py的可见围栏与隐藏cat字节一致workflow_dispatch：clone + 编译 HIXL + Hub 下载 + 设备锚点/root/.cache/huggingface/hub已有模型后重跑卸载段coder 出口网络和香港 A2 runner 不同，GitHub clone / Hugging Face 下载没有在 coder 上当作 e2e 证据。合入默认分支前对新 workflow 做
workflow_dispatch会 404；合入后在香港 A2 上补冷/热两次。