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.

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

[NPU] Llama3, Qwen2 1.5b, MiniCPM 1/2B groupwise support #12327

Merged
merged 9 commits into from
Nov 5, 2024

Conversation

cyita
Copy link
Contributor

@cyita cyita commented Nov 4, 2024

Description

Llama3, Qwen2 1.5b, MiniCPM 1/2B GroupWise support.
TODOs:

  • OV path validation
  • Remove unused code

1. Why the change?

2. User API changes

3. Summary of the change

4. How to test?

  • N/A
  • Unit test: Please manually trigger the PR Validation here by inputting the PR number (e.g., 1234). And paste your action link here once it has been successfully finished.
  • Application test
  • Document test
  • ...

5. New dependencies

  • New Python dependencies
    - Dependency1
    - Dependency2
    - ...
  • New Java/Scala dependencies and their license
    - Dependency1 and license1
    - Dependency2 and license2
    - ...

@cyita cyita changed the title [WIP] [NPU] Model groupwise support [NPU] Llama3, Qwen2 1.5b, MiniCPM 1/2B groupwise support Nov 4, 2024
@cyita cyita marked this pull request as ready for review November 5, 2024 01:21
@cyita cyita requested a review from plusbang November 5, 2024 01:21
@cyita
Copy link
Contributor Author

cyita commented Nov 5, 2024

Test, driver 2956, Yoga05

model l0 cw l0 gw ov cw ov gw
Meta-Llama-3-8B-Instruct Y Y Y Y
Qwen2-1.5B-Instruct Y Y Y Y
Qwen2-7B-Instruct Y Y Y Y
MiniCPM-1B-sft-bf16 Y Y Y Y
MiniCPM-2B-sft-bf16 Y Y Y Y

if isinstance(module, (Qwen2Attention, LlamaAttention)):
if (
isinstance(module, (Qwen2Attention, LlamaAttention))
or module.__class__.__name__ in ['MiniCPMAttention', 'Attention']
Copy link
Contributor

Choose a reason for hiding this comment

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

shall we check more strictly? this 'Attention' seems a little wide range.

Copy link
Contributor Author

Choose a reason for hiding this comment

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

This is for Baichuan attention, I will add a model-type check in the next PR.

@cyita cyita merged commit d872639 into intel-analytics:main Nov 5, 2024
1 check passed
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

2 participants