pytorch / pytorch/vision

Include implementation of MViTV2-B and MViTV2-L models

Open
#8,675 2 comments 0 reactions 0 assignees View on GitHub

Nobody has claimed this yet.

Dominant language
Python
Stars
17.9k
Forks
7.3k
Avg merge
1d 15h
Merged PRs (30d)
13

Description

🚀 The feature, motivation and pitch

Currently Pytorch includes an implementation of MViTV2-S, showing that the model is useful. MViTV2-B and MViTV2-L are even more powerful. Weights are publicly available at https://github.com/facebookresearch/SlowFast/blob/main/projects/mvitv2/README.md.

Alternatives

The SlowFast library has an implementation but it's hard to use outside of the slowfast library. Would be much much easier to have something built in to pytorch.

Additional context

No response

Contributor guide

Open the contributing guide

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 by reading the existing MViTV2-S implementation in PyTorch and compare it with the MViTV2-B and MViTV2-L implementation and publicly available weights in the linked SlowFast README. Confirm the model scope and integration expectations before proceeding; done means both requested models are implemented in PyTorch with the referenced weights supported.

Written by the indexing model from the issue text.

Assessment

Tech stack
python, pytorch
Domain
computer-vision, machine-learning
Issue type
Feature
Difficulty
5/5
Estimated time
Over a week
Activity status
Stale
Clarity
Mostly clear
Newbie friendliness
35/100

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.