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.

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

Jump to bottom

Fix FSDP ci test issue with "TypeError: Object of type Tensor is not … #137

Open
wants to merge 1 commit into
base: habana-main
Choose a base branch
from
Open

Fix FSDP ci test issue with "TypeError: Object of type Tensor is not … #137

wants to merge 1 commit into from

Commits on Mar 28, 2024

  1. Fix FSDP ci test issue with "TypeError: Object of type Tensor is not … 

    …JSON serializable"

Transformer 4.38 logs the grad_norm in log_history. But FSDP doesn't have global grad norm function.
When logging non-scalar tensor, the .item fails. The solution now is not to log grad_norm in
logging_history for FSDP.
    @libinta @hsubramony
    libinta authored and hsubramony committed Mar 28, 2024
    Configuration menu
    Copy the full SHA
    c12c594 View commit details
    Browse the repository at this point in the history