UnboundLocalError: local variable 'image_latents' referenced before assignment

[shubhdotai]

Describe the bug

Getting this error with strength = 1.0 (is_strength_max) other wise it's working fine in the StableDiffusionXLControlNetInpaintPipeline

Reproduction

sdxl_inpaint_pipe(
        prompt=prompt,
        negative_prompt= negative_prompt,
        width=1024,
        height=1024,
        num_inference_steps=30,
        image= image.convert("RGB"),
        mask_image= mask.convert("L"),
        control_image=control_image,
        controlnet_conditioning_scale=0.5,
        guidance_scale=5,
        num_images_per_prompt=1,
        strength=1.0,
).images

Logs

783     image = image.to(device=device, dtype=dtype)
    784     image_latents = self._encode_vae_image(image=image, generator=generator)
--> 786 image_latents = image_latents.repeat(batch_size // image_latents.shape[0], 1, 1, 1)
    788 if latents is None and add_noise:
    789     noise = randn_tensor(shape, generator=generator, device=device, dtype=dtype)

System Info

• diffusers version: 0.22.0.dev0
• Platform: Linux-5.15.0-1041-aws-x86_64-with-glibc2.31
• Python version: 3.10.12
• PyTorch version (GPU?): 2.0.1 (True)
• Huggingface_hub version: 0.16.4
• Transformers version: 4.30.2
• Accelerate version: 0.21.0
• xFormers version: 0.0.20
• Using GPU in script?: Yes
• Using distributed or parallel set-up in script?: No

Who can help?

@sayakpaul @yiyixuxu @DN6 @patrickvonplaten

[DN6]

cc: @yiyixuxu Could you please take a look here?

[patrickvonplaten]

@ShubhamWorks we would need a fully reproducible code snippet here I think

[shubhdotai]

@patrickvonplaten The issue got fixed in the last commit
Thanks team. Closing it!