Convrot quants dillute loras, likeness is not preserved/ SOLVED
- Dominant language
- Python
- Stars
- 133k
- Forks
- 15.7k
- Avg merge
- 1d 6h
- Merged PRs (30d)
- 155
Description
### Custom Node Testing
- [x] I have tried disabling custom nodes and the issue persists (see [how to disable custom nodes](https://docs.comfy.org/troubleshooting/custom-node-issues#step-1%3A-test-with-all-custom-nodes-disabled) if you need help)
### Expected Behavior
it should preserve lora strength likeness like non convrot weights .
### Actual Behavior
i did non convrot vs convrot lora test - the non convrot one preserved likeness from 0.6 -1 strength no problem but with convrot even 1.0 likeness was at about 75 % !!! This is a big problem.
Test with zimage and convrots from hface.
### Steps to Reproduce
download non convrot and convrot version of z-image and test lora with likeness and frozen seed.
### Debug Logs
```powershell
...
```
### Other
_No response_
Contributor guide
Research direction
Reproduce the comparison described with non-convrot and convrot Z-Image weights, a LoRA, likeness strengths from 0.6 to 1.0, and a frozen seed. Start by tracing where ComfyUI loads or applies these weights; done means determining why convrot quantization reduces likeness preservation and verifying equivalent behavior against the non-convrot test.
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
- Quiet
- Clarity
- Needs clarification
- Newbie friendliness
- 35/100