microsoft / microsoft/MoGe

About loss in every refinement step

Open
#161 0 comments 0 reactions 0 assignees View on GitHub

Nobody has claimed this yet.

Dominant language
JavaScript
Stars
3k
Forks
234
PR merge metrics
No merged PRs in 30d

Description

Hello authors,

Thank you so much for your amazing work!

I have a question regarding the training of MoGe-3: do the losses from different refinement steps vary significantly? I try to train a SSR my self, but the losses of different refinement step are similar. Is it normal? Or is there a way to fix it?

Thanks you very much!

Chengqi

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

The issue mentions MoGe-3, SSR, and losses across refinement steps but names no files, tests, or entry points. Start by reviewing the MoGe-3 training setup and how refinement-step losses are recorded, then compare the reported behavior with the expected training process. Done means documenting whether similar losses are normal and explaining any corrective steps.

Written by the indexing model from the issue text.

Assessment

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

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.