Skip to content

[fp8] add fp8 comm for low level zero #5961

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

Merged
merged 2 commits into from
Aug 2, 2024

Conversation

ver217
Copy link
Contributor

@ver217 ver217 commented Aug 2, 2024

📌 Checklist before creating the PR

  • I have created an issue for this PR for traceability
  • The title follows the standard format: [doc/gemini/tensor/...]: A concise description
  • I have added relevant tags if possible for us to better distinguish different PRs
  • I have installed pre-commit: pip install pre-commit && pre-commit install

🚨 Issue number

Link this PR to your issue with words like fixed to automatically close the linked issue upon merge

e.g. fixed #1234, closed #1234, resolved #1234

closed #5960

📝 What does this PR do?

Summarize your work here.
if you have any plots/diagrams/screenshots/tables, please attach them here.

add fp8 comm for low level zero.

Tested on BERT finetuning.

Plugin Accuracy F1-score GPU number
torch_ddp 84.4% 88.6% 2
low level zero-1 + fp8 84.3% 88.5% 2
low level zero-2 + fp8 84.8% 88.8% 2

💥 Checklist before requesting a review

  • I have linked my PR to an issue (instruction)
  • My issue clearly describes the problem/feature/proposal, with diagrams/charts/table/code if possible
  • I have performed a self-review of my code
  • I have added thorough tests.
  • I have added docstrings for all the functions/methods I implemented

⭐️ Do you enjoy contributing to Colossal-AI?

  • 🌝 Yes, I do.
  • 🌚 No, I don't.

Tell us more if you don't enjoy contributing to Colossal-AI.

@ver217 ver217 requested a review from a team as a code owner August 2, 2024 03:28
@BurkeHulk BurkeHulk self-requested a review August 2, 2024 12:32
@BurkeHulk BurkeHulk merged commit c297e21 into hpcaitech:feature/fp8_comm Aug 2, 2024
3 checks passed
@ver217 ver217 deleted the feature/zeor-fp8 branch August 5, 2024 02:06
if fp8_communication:
all_gather_into_tensor_flat_fp8(buffer, flat, output_shape=buffer.shape, group=group)
else:
dist.all_gather_into_tensor(buffer, flat, group=group)
Copy link
Contributor

Choose a reason for hiding this comment

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

Revert this change

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