Zhen Li
Zhen Li
@akk-123 @chille9 An important note: For those GPUs that do not support bfloat16, please change [this line](https://github.com/TencentARC/PhotoMaker/blob/6ec44fc13909d64a65c635b9e3b6f238eb1de9fe/gradio_demo/app.py#L39) to torch_dtype = torch.float16, the speed will be greatly improved (1min/img (before) vs....
Hello, can you please tell me the main functions of this PR? Please write it clearly in the commit.
> Hello, I didn't use the notebook but followed the pure python instructions. It said that it was not able to find photo.maker from photomakerPipeline Hello @ilisparrow, first of all,...
> The correct syntax is the following unless you download the model directly from the hugging face repo here: https://huggingface.co/TencentARC/PhotoMaker/tree/main, in which case the notebook suggests placing the .bin file...
Hi all, We have released the code, demo, and model of our PhotoMaker.
Hello, could you show me some cases? Here are our **user tips**: 1. Upload more photos of the person to be customized to improve ID fidelty. If the input is...
Hi all, We have released the code, demo, and model of our PhotoMaker. You could refer to our code for more details.
Hi, We have released the code, demo, and model of our PhotoMaker.
Hi all, We have released the code, demo, and model of our PhotoMaker. You could try our model and compare it with IP-Adapter-FaceID 😉
Try this photo without cropping. 