invoke-ai / invoke-ai/InvokeAI
[enhancement]: convert model to 16-bit
- Dominant language
- Python
- Stars
- 28.2k
- Forks
- 3k
- Avg merge
- 6d 5h
- Merged PRs (30d)
- 19
Description
### Is there an existing issue for this?
- [X] I have searched the existing issues
### Contact Details
_No response_
### What should this feature add?
Sometimes models are only distributed in 32-bit format, but we're only interested in running them at 16-bit.
To save disk space and loading time, model manager should have a "convert to 16-bit" button that writes out float16 safetensors and deletes the old files.
### Alternatives
_No response_
### Aditional Content
In order to know whether models need converting to 16-bit or not, we first need to be able to see their data type:
- #4068
Contributor guide
No contributing guide indexed for this repository
Assessment
This issue has not been assessed yet.