Tags: WilliamBy/LMCache
Tags
[Doc]: Consolidate contributing guide into 1 guide (LMCache#731) Consolidate contributing guide into 1 guide - Make online guide be the main guide - Improve the usability of the guide - Make the contributor guide in the repo be a placeholder/reference to the main guide Signed-off-by: Martin Hickey <[email protected]>
[release] Bumping version to 0.2.1 (LMCache#526) bumping version to 0.2.1 Signed-off-by: ApostaC <[email protected]>
[CI] Make LMCache pip-installable via bdists (and fallback sdists) bu… …ilt by CIBuildWheel (LMCache#516) * make sure to check contains() and not just exists_in_put_tasks() * Skeleton of LocalCPUBackend * Adding LocalCPUBackend to CreateStorageBackends() Hierarchy * Updated StorageManager Constructor * Finalized design * refactor StorageManager allocate() * refactor StorageManager put() with LocalCPUBackend touch() * accidentally deleted LocalDiskBackend import with the LocalCPUBackend import * accidentally deleted LocalDiskBackend import with the LocalCPUBackend import * accidentally deleted LocalDiskBackend import with the LocalCPUBackend import * adding hot cache lock * rename make_space_for() as allocate() * refactor get() in StorageManager * added some comment questions * first draft, still need to run tests * format.sh lint * first draft of publish.yml * pytorch 2.6.0 needs CUDA 12.4 * lmcache_test experiment * remove changes corrupting from refactor/hot-cache * sidestep installing torch in publish.yml * avoid cuda when building sdists * try again * sdist please * v1.0.1 * make sure not to link CUDA and torch * typo in NO_CUDA_EXT for sdist * try again * locally works, now trying to fix github workflows * added dockerfile to docker/ folder * removing the strategy matrix * change lmcache_test back to lmcache * add github release * trying again with 1.0.2 * remove matrix python * license udpate * fix release * forgot to add test to name for local workflow check * change name back to lmcache * Fix: Use softprops/action-gh-release with token * Fix: Use softprops/action-gh-release with token * lmcache_test -> lmcache * lmcache_test final try * 1.0.3 hopefully I stop letting this drag on * GH Release finally worked! changing back to lmcache * 1.0.3 -> 0.2.0 * Update PULL_REQUEST_TEMPLATE.md --------- Co-authored-by: Samuel Shen <[email protected]>
vllm 0.6.0 integration and support for pip install (LMCache#101) * Fix unused torchac dependency * Separate torchac to a standalone pip package * Fix buildkite configuration error * [refactor] merge the lmcache-server * Prepare to release lmcache 0.1.2 with pip * Update README.md with new quick start example
[Add] support for redis.Sentinel (LMCache#93) * [Add] support for redis.Sentinel * remove useless codes --------- Co-authored-by: ApostaC <[email protected]>
Add coverage report for CI pipeline (LMCache#89) * Add coverage report for CI pipeline * Fix the branch of lmcache-server --------- Co-authored-by: ApostaC <[email protected]>