MrGiovanni / MrGiovanni/ModelsGenesis

Not able to fine tune nnUNet with the "competition" style fine-tuning method

Open
#55 0 comments 0 reactions 0 assignees View on GitHub

Nobody has claimed this yet.

Dominant language
Jupyter Notebook
Stars
791
Forks
141
PR merge metrics
No merged PRs in 30d

Description

Hi @MrGiovanni @sbajpai2,

Thank you so much for all your work, it is fantastic!

I'm following the method "Usage of the pre-trained nnU-Net" from the competition folder, but I'm having my own dataset. Essentially, I'm trying to fine-tune nnUNet + Model genesis to my own data. I've followed all the steps as per your usage here.

When I start my nnUNet fine-tuning at step 3, I don't see any message like ""############################################### Loading pre-trained Models Genesis from " or even '############################################### Training from scratch'.

I'm only able to see the nnUNet's usual log messages. I have ensured that I'm correctly following all the steps shown.

Could you please suggest where I'm going wrong?

Contributor guide

No contributing guide indexed for this repository

First steps

  1. Read the whole issue, then the project's contributing guide.
  2. Comment on the issue to say you are picking it up — it saves two people doing the same work.
  3. Fork the repository and make your change on a branch.
  4. Open a pull request that references the issue number.

Research direction

Start with competition/README.md, especially the documented Usage of the pre-trained nnU-Net example and step 3, then reproduce the fine-tuning command with the custom dataset. Compare the expected pre-trained or from-scratch messages with the nnU-Net logs; done means identifying a concrete configuration or documentation correction.

Written by the indexing model from the issue text.

Assessment

Domain
machine-learning
Issue type
Bug
Difficulty
4/5
Estimated time
3-5 days
Activity status
Stale
Clarity
Needs clarification
Newbie friendliness
20/100

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.