Skip to content

[SW-218526] Updated Readme files for explicite lazy mode part2 (#177)#1889

Closed
jasi306 wants to merge 2 commits into
huggingface:mainfrom
HabanaAI:auto-pr-b56bafa
Closed

[SW-218526] Updated Readme files for explicite lazy mode part2 (#177)#1889
jasi306 wants to merge 2 commits into
huggingface:mainfrom
HabanaAI:auto-pr-b56bafa

Conversation

@jasi306
Copy link
Copy Markdown
Contributor

@jasi306 jasi306 commented Mar 26, 2025

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 make sure to update the documentation with your changes?
  • Did you write any new necessary tests?

@jasi306 jasi306 requested a review from regisss as a code owner March 26, 2025 09:26
@regisss
Copy link
Copy Markdown
Collaborator

regisss commented Mar 28, 2025

@jasi306 Is it possible to merge this PR and #1890 ? Unless there is a good reason to keep them separated of course

@emascarenhas
Copy link
Copy Markdown
Contributor

@jasi306 , Since --use_hpu_graphs works only with Lazy mode, shouldn't the PT_HPU_LAZY_MODE=1 be used anywhere we see use_hpu_graphs? Or are there cases where it makes sense to not use Lazy mode while setting use_hpu_graphs for inference/finetuning/training? See some cases in speech-recognition.

@emascarenhas
Copy link
Copy Markdown
Contributor

@jasi306 , Since --use_hpu_graphs works only with Lazy mode, shouldn't the PT_HPU_LAZY_MODE=1 be used anywhere we see use_hpu_graphs? Or are there cases where it makes sense to not use Lazy mode while setting use_hpu_graphs for inference/finetuning/training? See some cases in speech-recognition.

Hi @jasi306 , Any update?

@jasi306
Copy link
Copy Markdown
Contributor Author

jasi306 commented Apr 10, 2025

Just created new PR aggregating part 1 and 2:
#1921

@jasi306 jasi306 closed this Apr 10, 2025
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