Project-MONAI / Project-MONAI/tutorials
MAISI: Quantizer compression for the embedding feature generation by autoencoder
Open
Nobody has claimed this yet.
- Dominant language
- Jupyter Notebook
- Stars
- 2.5k
- Forks
- 803
- Avg merge
- 6d 22h
- Merged PRs (30d)
- 3
Description
Hi,
When I generate the embedding features by the trained autoencoder for the lung CT case with size of 836x512x512, I often encounter the OOM issue with 48GB GPU. Even though I have adjust the slices_num_max=256, it still suffers from the OOM issue, which may result in the stitch artifact issue. Thanks.
Contributor guide
First steps
- Read the whole issue, then the project's contributing guide.
- Comment on the issue to say you are picking it up — it saves two people doing the same work.
- Fork the repository and make your change on a branch.
- Open a pull request that references the issue number.
Research direction
No file or test is named. Start by locating the MAISI autoencoder embedding-feature generation workflow and reproduce it with the 836x512x512 lung CT case on a 48GB GPU. Investigate quantizer compression, with completion measured by avoiding OOM while preventing stitch artifacts.
Written by the indexing model from the issue text.
Assessment
- Tech stack
- pytorch
- Domain
- machine-learning, performance
- Issue type
- Feature
- Difficulty
- 5/5
- Estimated time
- Over a week
- Activity status
- Stale
- Clarity
- Needs clarification
- Newbie friendliness
- 25/100