Skip to content

Conversation

rnwang04
Copy link
Contributor

@rnwang04 rnwang04 commented Oct 29, 2024

Description

1. Why the change?

  • Support basic save & load for NPU pipelines (only save torch quantized weight now)

2. User API changes

See examples.

3. Summary of the change

  • Support basic save & load for NPU pipelines (only save torch quantized weight now)
  • Fix wrong setting of max_context_len (max_output_len -> max_context_len - 1)
  • Fix abnormal loading performance caused by split_linears (92s -> 0.037s)
  • Update all existing examples

4. How to test?

  • 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

@rnwang04
Copy link
Contributor Author

@rnwang04 rnwang04 requested review from cyita and jason-dai October 29, 2024 14:14
Copy link
Contributor

@jason-dai jason-dai left a comment

Choose a reason for hiding this comment

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

LGTM

Copy link
Contributor

@cyita cyita left a comment

Choose a reason for hiding this comment

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

LGTM

@rnwang04 rnwang04 merged commit 2b2cb9c into intel:main Oct 30, 2024
1 check passed
@rnwang04 rnwang04 deleted the L0_save_load branch October 30, 2024 02:02
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.

3 participants