psychedelicious
commented
4 months ago Hi, you are several versions behind. Please update to the latest version and see if that fixes the issue.
Open KISSEsWHISPERsFEEtBACKHUGs opened 4 months ago
psychedelicious
commented
4 months ago Hi, you are several versions behind. Please update to the latest version and see if that fixes the issue.
Is there an existing issue for this problem?
Operating system
Linux
GPU vendor
AMD (ROCm)
GPU model
AMD R4750G APU
GPU VRAM
14.9GiB UMA
Version number
4.2.0
Browser
Microsoft Edge 128.0.2708.0-1
Python dependencies
{ "accelerate": "0.30.1", "compel": "2.0.2", "cuda": null, "diffusers": "0.27.2", "numpy": "1.26.4", "opencv": "4.9.0.80", "onnx": "1.15.0", "pillow": "10.4.0", "python": "3.11.9", "torch": "2.2.2+rocm5.6", "torchvision": "0.17.2+rocm5.6", "transformers": "4.41.1", "xformers": null }
What happened
During normal image generations started by the workflows as usual, the local server crashes somewhy, and the most results are empty and a few results are just noises
What you expected to happen
All the 30 generations by the workflows have been done w/o local server command exiting often, and the image results are better reflecting the prompts from the workflows
How to reproduce the problem
(1) Install InvokeAI on your custom path (2) Write and run a workflow with 9527DetailRealistic checkpoint and KoreanInstagramXL downloaded LoRA downloaded from CivitAI (3) Run 30 batches and obseve what’s going on and oversee if freezes and/or local server command exits have been often (4) Also check if there are ‘‘illegal’’ shadings which maybe also cause local server command exits
Additional context
No response
Discord username
No response