HelloClyde

> 在ik_max_word的分词结果任意组合不能拼接出原始词的情况下，是ik_smart的分词结果不包含于ik_max_word的分词结果的原因之一 那这个有参数可以关吗？

这个问题太严重了，我还是决定换分词插件

when i install Dependencies before build: ``` brew install \ git \ cmake \ boost \ eigen \ freeimage \ glog \ gflags \ metis \ suite-sparse \ ceres-solver \...

You might need to use the ImageCompositeMasked node to overlay the original image on the non-masked parts, and the mask could use some feathering.

 I guess it might have something to do with the model capacity and VAE? I tried it and it wasn't very exaggerated.

My understanding is that both Hires Fix and inpainting are generated through diffusion models, and even with the use of the LatentCompositionMasked node, the final result will definitely be different...

> Guys, sorry, where to get a proper CLIP G SDXL BF16? > and this: https://huggingface.co/stabilityai/stable-cascade/blob/main/text_encoder/model.safetensors goes in the ComfyUI/models/clip/ folder to be loaded with the Load CLIP node in...

The issue might be due to this commit; [Add InpaintModelConditioning node.](https://github.com/comfyanonymous/ComfyUI/commit/10f2609fdd2231faf40346df11372601385b6cbb) the inpaint model fails when generating images from text. Let's wait for the author to fix it. Of course,...

> @HelloClyde Thank you for provided the code snippet I needed to fix this! Have you notified the developer about this simple fix? @comfyanonymous Can you help with a quick...

The author has already fixed this issue. You just need to update your version of ComfyUI to the commit "[94a5a67](https://github.com/comfyanonymous/ComfyUI/commit/94a5a67c32ded77efbd1b47e8d0257ddab9ac3bc)" or later.