yatoubusha

baseline_predictor_demo.ipynb is empty

1

there is no code in baseline_predictor_demo.ipynb.

vevn中已经安装了xformers，启用后开始训练，报错说没有这个包

1

raceback (most recent call last): File "/root/lora-scripts-main/sd-scripts/library/train_util.py", line 2579, in replace_unet_modules Traceback (most recent call last): File "/root/lora-scripts-main/sd-scripts/library/train_util.py", line 2579, in replace_unet_modules import xformers.ops ModuleNotFoundError: No module named 'xformers'

为什么要先根据随机初始化的噪音latens，用随机去噪步数对它进行基于lora的去噪操作得到denoised_latents呢？后续的positive_latents, neutral_latents, unconditional_latents都是基于denoised_latens在原始sd上继续去噪的

How to deploy inference code to predict my own images instead of testing datasets like Coco

1

How to deploy inference code to predict your own images instead of testing datasets like Coco

Can dwpose output 25 keypoints aligned with openpose?

1


What should I do if I want to obtain the actual object map of the mask area？ 

Unable to successfully implement the function of 'convert widget to input

3

### Your question  As shown in the figure, clicking the button does not enable the conversion of this function ### Logs _No response_ ### Other _No response_