Ip-adapt error: AttributeError: 'dict' object has no attribute 'shape' 'dict' object has no attribute 'shape'

Nobody has claimed this yet.

the error seems solved, I use ip-adapter-faceid_sdxl [59ee31a3] replace ControlnetxlCNXL_2vxpswa7AnytestV4 [4bb64990].

sd-webui-forge version: dfdcbab - fix sd upscale ... (2025-06-27 01:53:55)
checkpoint:illustrious(sdxl)

This error occurs when using the ip-adapt in img2img mode. The preprocessor is InsightFace+CLIP-H (IPAdapter), and the model is ControlnetxlCNXL_2vxpswa7AnytestV4 [4bb64990] (from model integration https://civitai.com/models/136070/controlnetxl-cnxl ）

`Traceback (most recent call last):
File "E:\Program Files\sd webui forge\modules_forge\main_thread.py", line 30, in work
self.result = self.func(*self.args, **self.kwargs)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "E:\Program Files\sd webui forge\modules\img2img.py", line 248, in img2img_function
processed = process_images(p)
^^^^^^^^^^^^^^^^^
File "E:\Program Files\sd webui forge\modules\processing.py", line 842, in process_images
res = process_images_inner(p)
^^^^^^^^^^^^^^^^^^^^^^^
File "E:\Program Files\sd webui forge\modules\processing.py", line 990, in process_images_inner
samples_ddim = p.sample(conditioning=p.c, unconditional_conditioning=p.uc, seeds=p.seeds, subseeds=p.subseeds, subseed_strength=p.subseed_strength, prompts=p.prompts)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "E:\Program Files\sd webui forge\modules\processing.py", line 1865, in sample
samples = self.sampler.sample_img2img(self, self.init_latent, x, conditioning, unconditional_conditioning, image_conditioning=self.image_conditioning)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "E:\Program Files\sd webui forge\modules\sd_samplers_kdiffusion.py", line 188, in sample_img2img
samples = self.launch_sampling(t_enc + 1, lambda: self.func(self.model_wrap_cfg, xi, extra_args=self.sampler_extra_args, disable=False, callback=self.callback_state, **extra_params_kwargs))
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "E:\Program Files\sd webui forge\modules\sd_samplers_common.py", line 281, in launch_sampling
return func()
^^^^^^
File "E:\Program Files\sd webui forge\modules\sd_samplers_kdiffusion.py", line 188, in
samples = self.launch_sampling(t_enc + 1, lambda: self.func(self.model_wrap_cfg, xi, extra_args=self.sampler_extra_args, disable=False, callback=self.callback_state, **extra_params_kwargs))
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "E:\Program Files\sd webui forge\venv\Lib\site-packages\torch\utils\_contextlib.py", line 115, in decorate_context
return func(*args, **kwargs)
^^^^^^^^^^^^^^^^^^^^^
File "E:\Program Files\sd webui forge\k_diffusion\sampling.py", line 149, in sample_euler_ancestral
denoised = model(x, sigmas[i] * s_in, **extra_args)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "E:\Program Files\sd webui forge\venv\Lib\site-packages\torch\nn\modules\module.py", line 1532, in _wrapped_call_impl
return self._call_impl(*args, **kwargs)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "E:\Program Files\sd webui forge\venv\Lib\site-packages\torch\nn\modules\module.py", line 1541, in _call_impl
return forward_call(*args, **kwargs)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "E:\Program Files\sd webui forge\modules\sd_samplers_cfg_denoiser.py", line 199, in forward
denoised, cond_pred, uncond_pred = sampling_function(self, denoiser_params=denoiser_params, cond_scale=cond_scale, cond_composition=cond_composition)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "E:\Program Files\sd webui forge\backend\sampling\sampling_function.py", line 362, in sampling_function
denoised, cond_pred, uncond_pred = sampling_function_inner(model, x, timestep, uncond, cond, cond_scale, model_options, seed, return_full=True)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "E:\Program Files\sd webui forge\backend\sampling\sampling_function.py", line 303, in sampling_function_inner
cond_pred, uncond_pred = calc_cond_uncond_batch(model, cond, uncond_, x, timestep, model_options)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "E:\Program Files\sd webui forge\backend\sampling\sampling_function.py", line 267, in calc_cond_uncond_batch
c['control'] = control.get_control(input_x, timestep_, control_cond, len(cond_or_uncond))
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "E:\Program Files\sd webui forge\backend\patcher\controlnet.py", line 300, in get_control
control_prev = self.previous_controlnet.get_control(x_noisy, t, cond, batched_number)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "E:\Program Files\sd webui forge\backend\patcher\controlnet.py", line 318, in get_control
self.cond_hint = image_resize.adaptive_resize(self.cond_hint_original, x_noisy.shape[3] * 8, x_noisy.shape[2] * 8, 'nearest-exact', "center").to(dtype)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "E:\Program Files\sd webui forge\backend\misc\image_resize.py", line 94, in adaptive_resize
old_width = samples.shape[3]
^^^^^^^^^^^^^
AttributeError: 'dict' object has no attribute 'shape'
'dict' object has no attribute 'shape'`

The traceback reaches backend/patcher/controlnet.py and backend/misc/image_resize.py during img2img sampling. Reproduce the stated InsightFace+CLIP-H and ControlNet setup, then inspect the value passed from controlnet.py into image_resize.adaptive_resize; done means the reported AttributeError no longer occurs, although the issue says the error may already be solved by changing models.

Written by the indexing model from the issue text.