-
Notifications
You must be signed in to change notification settings - Fork 224
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
float8 training: make the "config from recipe" API polished #1731
Conversation
Stack from ghstack (oldest at bottom): |
🔗 Helpful Links🧪 See artifacts and rendered test results at hud.pytorch.org/pr/pytorch/ao/1731
Note: Links to docs will display an error until the docs builds have been completed. This comment was automatically generated by Dr. CI and updates every 15 minutes. |
Summary: This PR makes the API that takes a recipe name (enum or string) and returns a `Float8LinearConfig` instance more polished and ready for usage in README.md docs and by partner callsites such as torchtitan and torchtune. Test Plan: ``` ./test/float8/test_everything.sh ``` Reviewers: Subscribers: Tasks: Tags: ghstack-source-id: 4f72eeb19603d6e1203fa9bf6ce8235bf431ecad ghstack-comment-id: 2667010633 Pull Request resolved: #1731
class Float8LinearRecipeName(enum.Enum): | ||
TENSORWISE = "tensorwise" | ||
ROWWISE = "rowwise" | ||
ROWWISE_WITH_GW_HP = "rowwise_with_gw_hp" |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
nit: wondering if gw_hp should be clarified somewhere as far as what it means?
it might also make sense to allow a fully written out version
Summary:
This PR makes the API that takes a recipe name (enum or string) and
returns a
Float8LinearConfig
instance more polished and ready forusage in README.md docs and by partner callsites such as torchtitan and
torchtune.
Test Plan:
Reviewers:
Subscribers:
Tasks:
Tags: