mirror of
https://github.com/openai/CLIP.git
synced 2026-11-05 01:37:26 +00:00
[GH-ISSUE #128] L2 regularization strength for few-shot linear probe #116
Labels
No labels
pull-request
No milestone
No project
No assignees
1 participant
Notifications
Due date
No due date set.
Dependencies
No dependencies set.
Reference
openai/CLIP#116
Loading…
Add table
Add a link
Reference in a new issue
No description provided.
Delete branch "%!s()"
Deleting a branch is permanent. Although the deleted branch may continue to exist for a short time before it actually gets removed, it CANNOT be undone in most cases. Continue?
Originally created by @constanceferragu on GitHub (Jul 15, 2021).
Original GitHub issue: https://github.com/openai/CLIP/issues/128
Hello, could you please explain how you determined the L2 regularization strength for the few-shot linear probe? Did you use a hyperparameter sweep on the validation sets, and if so, how did you do that given that there isn't a lot of available training data in few-shot learning? Thank you in advance!
@jongwook commented on GitHub (Jul 18, 2021):
Across all compared models, we used the performance on same validation set that we used for the full-data linear probe. It's not truly few-shot learning in this sense, although we hoped to find a consistent benchmark to fairly compare against the models (Fig. 6), and also have a means to plot the interpolation line to get a sense of the effective robustness (Figure. 15).