[bug]: model manager doesn't download 16-bit weights #4127
Assignees
Labels
Comments
|
Thanks for noticing this. The referenced PR should fix it. |
StAlKeR7779
added a commit
that referenced
this issue
Aug 4, 2023
… config (#4129) ## What type of PR is this? (check all applicable) - [ ] Refactor - [ ] Feature - [X] Bug Fix - [ ] Optimization - [ ] Documentation Update - [ ] Community Node Submission ## Have you discussed this change with the InvokeAI team? - [X] Yes - [ ] No, because: ## Have you updated all relevant documentation? - [X] Yes - [ ] No ## Description At install time, when the user's config specified "auto" precision, the installer was downloading the fp32 models even when an fp16 model would be appropriate for the OS. ## Related Tickets & Documents <!-- For pull requests that relate or close an issue, please include them below. For example having the text: "closes #1234" would connect the current pull request to issue 1234. And when we merge the pull request, Github will automatically close the issue. --> - Closes #4127
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Is there an existing issue for this?
OS
Linux
GPU
cuda
VRAM
12GB
What version did you experience this issue on?
eb64265
What happened?
invokeai-configure downloaded only 32-bit safetensors, not 16-bit ones, even though the huggingface repo has
.fp16.safetensorsfile available and I have precision set toauto. (I don't think setting the precision explicitly tofloat16helped either.)Reproduction plan:
invokeai-configuremodels/sd-1/main/stable-diffusion-v1-5/unet/, see it has the 3.4 GB safetensors file, not the 1.7 GB fp16.safetensors file.The text was updated successfully, but these errors were encountered: