PuLID icon indicating copy to clipboard operation
PuLID copied to clipboard

[NeurIPS 2024] Official code for PuLID: Pure and Lightning ID Customization via Contrastive Alignment

Results 165 PuLID issues
Sort by recently updated
recently updated
newest added

In my practice, I found that when training flux pulid, the results of fake guidance always have worse performance (less similar but good quality) , how do you guys train...

What is uncond_id_embedding used for? It doesn’t seem to be mentioned in your paper. Could you explain it? In pulid/pipeline.py file, `uncond_id_embedding = self.id_adapter(id_uncond, id_vit_hidden_uncond)`

Can you provide a requirement file for running app_v1_1.py. I tried the two requirements in your repo but there are always errors due to the dependencies....

How to implement img2img? ` sigmas = self.get_sigmas_karras(steps).to(self.device) print(type(init_image)) # latents # noise = torch.randn((size[0], 4, size[1] // 8, size[2] // 8), device="cpu", generator=torch.manual_seed(seed)) # noise = noise.to(dtype=self.pipe.unet.dtype, device=self.device) #...

I was not able to make it work in comfy, so I am using your gradio program (https://github.com/ToTheBeginning/PuLID/blob/main/README.md#wrench-dependencies-and-installation) Could you add possibility to read LORAS please? This would LIFE CHANGING....

运行后,点击示例报错:Could not parse server response: SyntaxError: Unexpected token 'I', "Internal S"... is not valid JSON

![cmd_YqzoIaD7q2](https://github.com/user-attachments/assets/4f91f65f-1f3c-46e4-a616-e025ffb7824c)

I used with streamize/leosam-lightning XL model but I got black image

Hello @ToTheBeginning , My problem is that, I was only able to run pulid flux thought these instructions (https://github.com/ToTheBeginning/PuLID/blob/main/docs/pulid_for_flux.md) But for some reason it CANNOT run on comfy. The reason...