Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

Optimize speculative decoding PVC memory usage #10329

Merged
merged 7 commits into from
Mar 6, 2024
Merged

Optimize speculative decoding PVC memory usage #10329

merged 7 commits into from
Mar 6, 2024

Conversation

cyita
Copy link
Contributor

@cyita cyita commented Mar 5, 2024 •
edited
Loading

Description

Optimize speculative decoding PVC memory usage.

3. Summary of the change

  1. Call empty_cache when necessary.
  2. Fix the draft&verify model repetitive extend kv cache issue to save memory.

4. How to test?

  • local test (llama, mistral, gptj, baichuan, qwen, chatglm)

@cyita cyita requested a review from rnwang04 March 5, 2024 09:17
rnwang04
rnwang04 approved these changes Mar 5, 2024
View reviewed changes
Copy link
Contributor

@rnwang04 rnwang04 left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

LGTM

@cyita cyita merged commit 786254a into intel:main Mar 6, 2024
19 checks passed
liu-shaojun pushed a commit that referenced this pull request Mar 25, 2024
@cyita
Optimize speculative decoding PVC memory usage (#10329)
9ea499c 
* optimize memory

* update

* update

* update

* support other models

* update

* fix style
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Reviewers

@rnwang04 rnwang04 rnwang04 approved these changes

Assignees
No one assigned
Labels
None yet
Projects
None yet
Milestone
No milestone
Development

Successfully merging this pull request may close these issues.
2 participants
@cyita @rnwang04