ByteDance-Seed / ByteDance-Seed/Bagel

About text2img generation finetuning

Open
#294 1 comment 0 reactions 0 assignees View on GitHub
Dominant language
Python
Stars
6.2k
Forks
545
PR merge metrics
No merged PRs in 30d

Description

Hi @wangwei1237 @haooooooqi @helio9cn @TsuTikgiau @Luodian

Thank you for releasing such an impactful work.

I try to fine-tune my text2img data on Bagel.
But I found that the generated image suffering from strange color and its shape seems correct.
I have checked the input data which is valid. I also checked the inference code that is correct.
Image

What could be causing this result?

Looking forward to your reply.

Contributor guide

No contributing guide indexed for this repository

Research direction

The report concerns text-to-image fine-tuning in Bagel and mentions the input data and inference code, but names no files or tests. Start by reproducing the strange-color output and comparing the fine-tuning inputs with the inference path. Done means identifying the cause and confirming that valid fine-tuning data produces correctly colored images.

Written by the indexing model from the issue text.

Assessment

Tech stack
python
Domain
machine-learning
Issue type
Bug
Difficulty
4/5
Estimated time
3-5 days
Activity status
Stale
Clarity
Needs clarification
Newbie friendliness
25/100

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.