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

Bit 614 gptneox compatibility check #1038

Merged
merged 9 commits into from
Jan 9, 2023

Conversation

isabella618033
Copy link
Contributor

@isabella618033 isabella618033 commented Jan 4, 2023

Some models (eg. gptneox / pythia / finetuned opt) are not compatible with half precision, and outputing hidden state / logits with nan causing the error message mentioned here
Which is not an internal bug.

Updated warning message to suggest users to turn off --neuron.autocast in case of using half precision non compatible models
image

@isabella618033 isabella618033 requested review from Eugene-hu and removed request for Eugene-hu January 4, 2023 17:29
@opentaco opentaco self-requested a review January 5, 2023 07:04
@isabella618033 isabella618033 merged commit 9ae55a3 into nobunaga Jan 9, 2023
@ifrit98 ifrit98 deleted the BIT-614-gptneox_compatibility_check branch May 24, 2023 14:46
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