NVlabs / NVlabs/FastGen

Question about DMDv2

Open
#23 3 comments 0 reactions 0 assignees View on GitHub

Nobody has claimed this yet.

Dominant language
Python
Stars
1k
Forks
83
PR merge metrics
No merged PRs in 30d

Description

I find in the original DMDv2 implementation, gan loss uses fake model to extract hidden states feature, but in your code, use Teacher model to extract. Why?

Image Image

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

Begin by comparing the original DMDv2 implementation with FastGen's current teacher-model usage, using the two screenshots and the relevant training code as references. Identify the rationale for extracting hidden-state features from the teacher rather than the fake model, then document a confirmed explanation in this issue or the project's documentation.

Written by the indexing model from the issue text.

Assessment

Domain
machine-learning
Issue type
Documentation
Difficulty
5/5
Estimated time
Over a week
Activity status
Quiet
Clarity
Needs clarification
Newbie friendliness
25/100

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.