openai/CLIP
1
0
Fork 0
mirror of https://github.com/openai/CLIP.git synced 2026-11-05 01:37:26 +00:00
Code Issues 274 Projects Packages Wiki Activity

[GH-ISSUE #75] Evaluation with Pascal VOC2007 classification #67

New issue
Closed
opened 2026-10-31 21:10:27 +00:00 by bsakaguchi · 1 comment
bsakaguchi commented 2026-10-31 21:10:27 +00:00
Owner
Copy link

Originally created by @meigaoms on GitHub (Apr 1, 2021).
Original GitHub issue: https://github.com/openai/CLIP/issues/75

Hi there,
For this multi-label datasets, I used sklearn MultiOutputRegressor with estimator logisticRegression and tested it with several models. All scores I got are significantly higher than corresponding score in Table 10: ResNet50 89.1 (83.8 CLIP), ResNet101 88.7 (84.4 CLIP), Efficientnet_b0 88.6 (84.4 CLIP), CLIP_VIT_B32 89.2 (87.7 CLIP).
I'm wondering if my evaluation logic for this multi-label dataset is different from yours, since MultiOutputRegressor trains 20 estimators under the hood. However, with a single logisticRegression model, it handles single label problem only. Could you share more information of how you do linear probe with this multi-label dataset?

Originally created by @meigaoms on GitHub (Apr 1, 2021). Original GitHub issue: https://github.com/openai/CLIP/issues/75 Hi there, For this multi-label datasets, I used sklearn MultiOutputRegressor with estimator logisticRegression and tested it with several models. All scores I got are significantly higher than corresponding score in Table 10: ResNet50 89.1 (83.8 CLIP), ResNet101 88.7 (84.4 CLIP), Efficientnet_b0 88.6 (84.4 CLIP), CLIP_VIT_B32 89.2 (87.7 CLIP). I'm wondering if my evaluation logic for this multi-label dataset is different from yours, since MultiOutputRegressor trains 20 estimators under the hood. However, with a single logisticRegression model, it handles single label problem only. Could you share more information of how you do linear probe with this multi-label dataset?
bsakaguchi commented 2026-10-31 21:10:28 +00:00
Author
Owner
Copy link

@Sierkinhane commented on GitHub (Jul 1, 2021):

Hi, can you provide your evaluation code in computing accuracy score? @meigaoms Maybe I have some mistakes in calculating mAP score.

<!-- gh-comment-id:871940197 --> @Sierkinhane commented on GitHub (Jul 1, 2021): Hi, can you provide your evaluation code in computing accuracy score? @meigaoms Maybe I have some mistakes in calculating mAP score.
Sign in to join this conversation.
No Branch/Tag specified
Branches Tags
main
codex/pin-github-workflow-refs-20260325-132838
jongwook/issue-396
jongwook-patch-1
No results found.
Labels
Clear labels   
pull-request
Mirrored from GitHub Pull Request

No labels
pull-request
Milestone
Clear milestone
No items
No milestone
Projects
Clear projects
No items
No project
Assignees
Clear assignees
No assignees
1 participant
Notifications
Due date
The due date is invalid or out of range. Please use the format "yyyy-mm-dd".

No due date set.

Dependencies

No dependencies set.

Reference
openai/CLIP#67
No description provided.