could you share the pretain ckpt？

[Rbrq03]

Hello @sunzhe09 thanks for interest. Which concept are u interested in ? And what are u want to use the pretrained ckpt for?

[sunzhe09]

I want to test my custom images，just combine the two image‘s subject。if it work ，I will make my ads picture，thx

[Rbrq03]

Hello @sunzhe09 , glad to hear that you are using our work! If you want to test your custom images, maybe you should train your own ckpt in custom images. I am not sure what do you mean by “pretain ckpt” (If it refers to the SD weights, you can run the train code and it will be downloaded automatically. ) , if you’re interested in the ckpt of concepts in data, just tell me and I am glad to share that. Reach out if you have further questions!

[sunzhe09]

I just want to test this code： import torch from diffusers import DiffusionPipeline

pipeline = DiffusionPipeline.from_pretrained("runwayml/stable-diffusion-v1-5", torch_dtype=torch.float16).to("cuda") pipeline.unet.load_attn_procs("path-to-save-model", weight_name="pytorch_custom_diffusion_weights.bin") pipeline.load_textual_inversion("path-to-save-model", weight_name=".bin") pipeline.load_textual_inversion("path-to-save-model", weight_name=".bin")

image = pipeline( "a teddy bear sitting in front of a barn", num_inference_steps=100, guidance_scale=6.0, eta=1.0, ).images[0] image.save("multi-subject.png")

but I can’t get path-to-save-model。it is appreciated if you could share the bin or gradio demo

[Rbrq03]

Hi @sunzhe09 , I see.

• if you have trained the model with our code, you will find all the weight in ckpt/${CLS_TOKEN1}_${CLS_TOKEN2}. You can use it as the path-to-save-model
• If you just want to test the code. You can use the following ckpt. Also I have updated it in README.

btw, I share a result generated by this ckpt.

If you have further questions, feel free to comment.