Skip to content

Issues: ROCm/TransformerEngine

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

8 Open 4 Closed
8 Open 4 Closed
Author
Filter by author
Loading
Label
Filter by label
Loading
Use alt + click/return to exclude labels
or ⇧ + click/return for logical OR
Projects
Filter by project
Loading
Milestones
Filter by milestone
Loading
Assignee
Filter by who’s assigned
Sort
Sort by
Newest Oldest Most commented Least commented Recently updated Least recently updated Best match
Most reactions
👍 👎 😄 🎉 😕 ❤️ 🚀 👀

Issues list

[Build issue]: Building wheel for specific arch not supported without GPU but with all ROCm components installed
#92 opened Nov 13, 2024 by caaatch22
2
[Out of Box Experience]: ROCm Transformer Engine Should Be Included in AMD Pytorch Images
#82 opened Oct 18, 2024 by OrenLeung
2
[FSDP 8xMI300X] Llama3 8B FP8 is 21% slower than BF16 & OOMs on the same batch size
#79 opened Oct 15, 2024 by OrenLeung
@wenchenvincent
6
[FSDP 8xMI300X]: LLama3 70B 4 Layer Proxy Model GPU Core Dumps
#78 opened Oct 15, 2024 by OrenLeung
@wenchenvincent
24
[DDP 8xMI300X] GPT2-1.5B FP8 is 25% slower than BF16 & OOMs on the same batch size
#76 opened Oct 15, 2024 by OrenLeung
3
MI300X FP8 TE.Linear 2x Slower than AMP BF16 F.Linear
#73 opened Oct 13, 2024 by OrenLeung
@wenchenvincent
14
[1xMI300X] GPT-2 XL 1.5B FP8 Training ~30% slower than H100 FP8
#72 opened Oct 13, 2024 by OrenLeung
1
@wangye805
22
[TE] Investigate parallelism implementation in Transformer Engine
#34 opened Apr 23, 2024 by wangye805
4 of 5 tasks
@wangye805
1
ProTip! Type g p on any issue or pull request to go back to the pull request listing page.