Support paged kv cache for single batch chat module #1651
Merged
Conversation
This file contains 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
|
cc @MasterJH5574 . would be great to have two things
|
This PR adds the support for paged kv cache for single batch chat module.
cyx-6
force-pushed
the
single-kv-cache
branch
from
a74c545 to
6a2cb33
Compare
MasterJH5574
approved these changes
Feb 1, 2024
|
cc @CharlieFRuan on webllm runtime update |
|
This PR breaks llama on ROCm (single GPU): Compiling and running Edit: above fixed by #1710, not just ROCm; but ROCm still observes accuracy issue Compiling and running Confirmed that no issues is observed if rebased to #1627. Update: all issues fixed now. |
CharlieFRuan
added a commit
to mlc-ai/web-llm
that referenced
this pull request
Feb 13, 2024
PagedKVCache is introduced in MLC-LLM a while back to unite the interface for KVCache. This PR makes WebLLM compatible with the new PagedKVCache interface, encapsulating it with the goal that WebLLM users will not notice any difference. This PR is equivalent to the changes to `llm_chat.cc` in mlc-ai/mlc-llm#1651, and should address issues like mlc-ai/mlc-llm#1628. There are still existing model compilation issues regarding `workgroup_size` (since WebGPU, unlike most other backends, can only support 256 number of threads). We will address this issue more elegantly soon; for now, compiling llama-based models require manually changing kernel sizes as shown in [this branch](https://github.com/CharlieFRuan/mlc-llm/tree/local-workgroupSize-webLLM-kvCache). This PR is also largely dependent on apache/tvm#16554.
This was referenced Feb 13, 2024
4 tasks
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
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Outdated suggestions cannot be applied.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.
This PR adds the support for paged kv cache for single batch chat module.