microsoft / microsoft/TRELLIS

Finetuning a pretrained checkpoint

Open
#255 1 comment 10 reactions 0 assignees View on GitHub

Nobody has claimed this yet.

Dominant language
Python
Stars
13.7k
Forks
1.3k
PR merge metrics
No merged PRs in 30d

Description

Right now, the training scripts only allow for resuming a training run from a previous step.
I see from the code that trainers have a from_finetuned method, which is not used anywhere. Is there a straightforward way to use the provided safetensors files for finetuning ?

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 by reading the training scripts and locating the trainers' from_finetuned method. Then trace how the provided safetensors files are loaded and how resuming from a previous step is currently handled. Done means establishing a documented, working path for finetuning from a pretrained checkpoint, with the relevant training behavior verified.

Written by the indexing model from the issue text.

Assessment

Tech stack
python
Domain
machine-learning
Issue type
Feature
Difficulty
4/5
Estimated time
3-5 days
Activity status
Stale
Clarity
Mostly clear
Newbie friendliness
35/100

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.