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

[Experiment] Basic Image Preprocessing x QCRPubMedCLIP #6

Open
7 tasks done
Nana2929 opened this issue May 12, 2023 · 3 comments
Open
7 tasks done

[Experiment] Basic Image Preprocessing x QCRPubMedCLIP #6

Nana2929 opened this issue May 12, 2023 · 3 comments
Assignees
@Nana2929

Comments

@Nana2929
Copy link
Owner

Nana2929 commented May 12, 2023 •
edited

前置作業：紀錄未經特殊處理的圖片的 VQA 正確率

  • Run with Visual Encoder RN50x4 and record acc
  • Run with Visual Encoder ViT and record acc

使用經過特殊處理的圖片跑三個實驗

  • Preprocess Images #9
  • Record the preprocessing steps for images (and the rationale) behind @liu @zhihao
  • Run with Visual Encoder RN50 (best) and record acc
  • Run with Visual Encoder RN50x4 and record acc
  • Run with Visual Encoder ViT and record acc
@Nana2929 Nana2929 self-assigned this May 12, 2023
@Nana2929
Copy link
Owner Author

2023-05-14 19:03
Preprocess Steps see #9.
Running with RN50 right now.

@Nana2929
Copy link
Owner Author

Nana2929 commented May 14, 2023 •
edited

[跑錯實驗] Copies Overwritten in collect_all_imgs.py (preprocess branch)

copilot 寫 code，多寫多錯 QQ
2023-05-15 00:29 Re-Running with RN50 right now.

@Nana2929
Copy link
Owner Author

Nana2929 commented May 14, 2023 •
edited

看他收斂的速度（大概在 89 epochs 可以收斂出最高）＋趕時間
Configs 內 TRAIN N_EPOCHS 改成 100 epochs

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Assignees

@Nana2929 Nana2929

Labels
None yet
Projects
No open projects
Medical VQA
Status: Done
Milestone
No milestone
Development

No branches or pull requests
1 participant
@Nana2929