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

Attempt for cleverer auto batch_prefill values (some simplifications). #2808

Merged
merged 7 commits into from
Dec 9, 2024

Conversation

Narsil
Copy link
Collaborator

@Narsil Narsil commented Dec 8, 2024

What does this PR do?

Fixes # (issue)

Before submitting

  • This PR fixes a typo or improves the docs (you can dismiss the other checks if that's the case).
  • Did you read the contributor guideline,
    Pull Request section?
  • Was this discussed/approved via a Github issue or the forum? Please add a link
    to it if that's the case.
  • Did you make sure to update the documentation with your changes? Here are the
    documentation guidelines, and
    here are tips on formatting docstrings.
  • Did you write any new necessary tests?

Who can review?

Anyone in the community is free to review the PR once the tests have passed. Feel free to tag
members/contributors who may be interested in your PR.

@Narsil Narsil requested review from danieldk and drbh December 9, 2024 04:03
launcher/src/main.rs Outdated Show resolved Hide resolved
// TODO This calculation depends on the actual implementation
let dtype_size = 2;
let mlp_size = self.intermediate_size?;
Some((mlp_size + mlp_size / 2) * self.num_experts * dtype_size * 3)
Copy link
Member

Choose a reason for hiding this comment

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

Maybe a comment where mlp_size / 2 comes from?

Copy link
Collaborator Author

Choose a reason for hiding this comment

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

Done.

enum Gpu {
RTX4090,
T4,
L4,
Copy link
Member

Choose a reason for hiding this comment

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

We probably want L40 as well, since we are using them now?

Copy link
Collaborator Author

Choose a reason for hiding this comment

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

Done.

# switch on equivalent logits based on the position in the batch.
# 1 output being different is not uncommon
if sum(equals) < len(equals) - 1:
assert outputs == expected
Copy link
Member

Choose a reason for hiding this comment

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

I'm missing something. Can't this only be true when sum(equals) == len(equals)?

Copy link
Collaborator Author

Choose a reason for hiding this comment

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

I want the error message to be about the content and containing the diff.

Copy link
Collaborator

@drbh drbh left a comment

Choose a reason for hiding this comment

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

lgtm!

also the qwen2-vl tests should be resolved shortly in this PR #2802

@Narsil Narsil merged commit a04356f into main Dec 9, 2024
10 of 12 checks passed
@Narsil Narsil deleted the update_max_prefill_auto_with_vram_reqs branch December 9, 2024 18:44
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