OOM Error

[jakethesnake1126]

It seems my 12 GB VRAM RTX 4070 Super is not enough to run this. Is there any possibility for further optimization such as quantization?

[zhuang2002]

Thank you so much for your attention and support. To help reduce memory overhead, since ColorFlow does not rely on text conditions, I would kindly suggest avoiding the CFG strategy and instead using empty text for inference. Additionally, you could pre-save the T5 model’s output for empty text inputs, which would allow you to skip loading the T5 model entirely during inference. I’ve had quite a lot on my plate recently, but I’ll make sure to update the code with these optimizations as soon as I can. Thank you for your understanding and patience!