SkyworkAI / SkyworkAI/UniPic

have a question regarding the training resources and time cost of different stages:

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

Nobody has claimed this yet.

Dominant language
Python
Stars
875
Forks
43
PR merge metrics
No merged PRs in 30d

Description

Hi, thanks for your great work on UniPic2.0! I have a question regarding the training resources and time cost of different stages:

For UniPic2-MetaQuery (Stage 1), when training the connector (with frozen MLLM and DiT), how many computational resources (e.g., GPUs, type, total GPU hours) were used, and approximately how many days did it take?

For UniPic2-MetaQuery (Stage 2), when jointly fine-tuning the connector and the SD3.5M-Kontext (DiT), how many resources were used and how long did this stage take?

It would be very helpful to know the approximate training setup (e.g., number/type of GPUs, training days) for both stages.

Thanks!

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 checking the repository's UniPic2-MetaQuery training materials and configuration references for Stage 1 and Stage 2. Done means documenting the GPU type and count, total GPU hours or training days, and approximate setup for both stages; no specific file or test is mentioned in the issue.

Written by the indexing model from the issue text.

Assessment

Tech stack
python
Domain
machine-learning
Issue type
Documentation
Difficulty
3/5
Estimated time
1-2 days
Activity status
Stale
Clarity
Clearly specified
Newbie friendliness
25/100

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.