the most basic text workflow produces nothing but black on m2 apple with sd 1.5 workflow
- Dominant language
- Python
- Stars
- 133k
- Forks
- 15.7k
- Avg merge
- 1d 10h
- Merged PRs (30d)
- 153
Description
### Expected Behavior
I expect that when I follow this tutorial at timecode 5:46 https://youtu.be/_C7kR2TFIX0?si=JUsLSZep24l4A1vU&t=346 that I'll get the most basic image when using an SD 1.5 checkpoint
### Actual Behavior
ComfyUI on mac m2 produces an all black image. I hit update all and update comfy in manager, restart the browser, re-run the workflow and get the same all black image. But if I use the "crystal clear" sdxl checkpoint i dont get an all black image.
### Steps to Reproduce
start brand new workflow from scratch. add all the nodes discussed in Matteo's tutorial: load check point, empty latent, clip text encoder, ksampler, vae decode and preview image window. try 2 separate checkpoints: an sdxl checkpoint and an sd 1.5 checkpoint
### Debug Logs
```powershell
Starting server
To see the GUI go to: http://127.0.0.1:8188
FETCH DATA from: /Volumes/RachelComfyUI/ComfyUI/custom_nodes/ComfyUI-Manager/extension-node-map.json [DONE]
got prompt
model weight dtype torch.float16, manual cast: None
model_type EPS
Using split attention in VAE
Using split attention in VAE
Requested to load SD1ClipModel
Loading 1 new model
loaded completely 0.0 235.84423828125 True
Requested to load BaseModel
Loading 1 new model
loaded completely 0.0 1639.406135559082 True
```
### Other
SDXL workflow
Contributor guide
Assessment
This issue has not been assessed yet.