yisol comments

Results 17 comments of


                                            yisol

Gradio Demo

Hello, do you mean gradio codes for huggingface demo? You can check them on "Files" tab on https://huggingface.co/spaces/yisol/IDM-VTON. (app.py)

What is the minimum memory requirment?

Hello, >18GB VRAM is required for single image inference. You can use optimization scheme like offloading for further memory reducing.

Some questions about the number of channels of Unet and its' initialization

Hello, we initialized additional channels with zero weights and used sdxl inpainting model for initialization of TryOnNet.

Training questions

Hello, we used gradient checkpointing and 8 bit adam for training and fit batch size 6 to single A100 GPU. We didn't use precomputing latents and embeddings or gradient accumulation...

how could I use inference to generated masked images for inpaint in sd?

Hello, you can refer the 6. human agnostic part on this [page ](https://github.com/sangyun884/HR-VITON/issues/45)

the training error of pretrained model mismatch

Hello, you should use the pre-trained ip-adapter for sdxl **(IP-Adapter/sdxl_models/ip-adapter-plus_sdxl_vit-h.bin)** and corresponding image_encoder **(IP-Adapter/models/image_encoder)**. (Please refer [this](https://github.com/tencent-ailab/IP-Adapter/issues/66#issuecomment-1716983432) ) I will update the readme file.

yisol