Input type (struct c10::Half) and bias type (float) should be the same

[alexbofa]

Hello, it gives an error, I can't figure out what the problem is

Error occurred when executing CCSR_Upscale:

Input type (struct c10::Half) and bias type (float) should be the same

File "Z:\ComfyUI\ComfyUI\execution.py", line 152, in recursive_execute
output_data, output_ui = get_output_data(obj, input_data_all)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "Z:\ComfyUI\ComfyUI\execution.py", line 82, in get_output_data
return_values = map_node_over_list(obj, input_data_all, obj.FUNCTION, allow_interrupt=True)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "Z:\ComfyUI\ComfyUI\execution.py", line 75, in map_node_over_list
results.append(getattr(obj, func)(**slice_dict(input_data_all, i)))
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "Z:\ComfyUI\python_embeded\Lib\site-packages\torch\utils\_contextlib.py", line 115, in decorate_context
return func(*args, **kwargs)
^^^^^^^^^^^^^^^^^^^^^
File "Z:\ComfyUI\ComfyUI\custom_nodes\ComfyUI-CCSR\nodes.py", line 120, in process
samples = sampler.sample_with_mixdiff_ccsr(
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "Z:\ComfyUI\python_embeded\Lib\site-packages\torch\utils\_contextlib.py", line 115, in decorate_context
return func(*args, **kwargs)
^^^^^^^^^^^^^^^^^^^^^
File "Z:\ComfyUI\ComfyUI\custom_nodes\ComfyUI-CCSR\model\q_sampler.py", line 718, in sample_with_mixdiff_ccsr
"c_latent": [self.model.apply_condition_encoder(tile_cond_img)],
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "Z:\ComfyUI\ComfyUI\custom_nodes\ComfyUI-CCSR\model\ccsr_stage2.py", line 320, in apply_condition_encoder
c_latent_meanvar = self.cond_encoder(control * 2 - 1)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "Z:\ComfyUI\python_embeded\Lib\site-packages\torch\nn\modules\module.py", line 1511, in _wrapped_call_impl
return self._call_impl(*args, **kwargs)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "Z:\ComfyUI\python_embeded\Lib\site-packages\torch\nn\modules\module.py", line 1520, in _call_impl
return forward_call(*args, **kwargs)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "Z:\ComfyUI\python_embeded\Lib\site-packages\torch\nn\modules\container.py", line 217, in forward
input = module(input)
^^^^^^^^^^^^^
File "Z:\ComfyUI\python_embeded\Lib\site-packages\torch\nn\modules\module.py", line 1511, in _wrapped_call_impl
return self._call_impl(*args, **kwargs)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "Z:\ComfyUI\python_embeded\Lib\site-packages\torch\nn\modules\module.py", line 1520, in _call_impl
return forward_call(*args, **kwargs)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "Z:\ComfyUI\ComfyUI\comfy\ldm\modules\diffusionmodules\model.py", line 520, in forward
h = self.conv_in(x)
^^^^^^^^^^^^^^^
File "Z:\ComfyUI\python_embeded\Lib\site-packages\torch\nn\modules\module.py", line 1511, in _wrapped_call_impl
return self._call_impl(*args, **kwargs)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "Z:\ComfyUI\python_embeded\Lib\site-packages\torch\nn\modules\module.py", line 1520, in _call_impl
return forward_call(*args, **kwargs)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "Z:\ComfyUI\ComfyUI\comfy\ops.py", line 42, in forward
return super().forward(*args, **kwargs)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "Z:\ComfyUI\python_embeded\Lib\site-packages\torch\nn\modules\conv.py", line 460, in forward
return self._conv_forward(input, self.weight, self.bias)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "Z:\ComfyUI\python_embeded\Lib\site-packages\torch\nn\modules\conv.py", line 456, in _conv_forward
return F.conv2d(input, weight, bias, self.stride,
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^

[Chryseus]

I'm having the exact same problem, tried everything I could think of to fix it.
Edit: Never mind it fixed itself, try restart it if you're having problems.

[unexpectedly]

Having the same issue. Tried restart & reinstall - still the same error:

Error occurred when executing CCSR_Upscale:
Input type (c10::Half) and bias type (float) should be the same

[RigzTD]

Yep same here, tried to also reinstall and restart, but its not fixing it.

Error occurred when executing CCSR_Upscale:
Input type (struct c10::Half) and bias type (float) should be the same

[pmajor74]

I'm using the CCSR node in the AP Workflow 8.0

I am also having the same issue. To reproduce, I've created an image and have enabled are:

• Enable Image Generator (I'm making the 3 clowns image)
• Enable Refiner (for SDXL)
• Enable Upscaler

When the flow makes its way over to the "Upscaler" group, it dies on the CCSR_Update with the error I pasted below this message.

• I am using the CCSR_Model_Select to select the "real-world_ccsr.ckpt" and have the following settings:
• resize method: lanczos
• scale by: 2.0
• steps: 3
• t_max: 0.67
• t_min: 0.33
• sampling method: ccsr_tiled_vae_gaussian_weights
• tile_size: 512
• tile_stride: 256
• vae_tile_size_encode: 1024
• vae_tile_size_decode: 1024
• color_fix_type: wavelet
• keep model loaded: false
• seed: 2
• contol_after_generate: fixed

I'm a programmer as well, so am digging a bit deeper, but am half hoping someone else sees some of this info and it helps create a solution quicker.

=================== error output ==================
Using gaussian weights
timesteps used in spaced sampler:
[0, 53, 105, 158, 210, 263, 315, 368, 421, 473, 526, 578, 631, 684, 736, 789, 841, 894, 946, 999]
Process tile with location (0 64) (0 64): 0%| | 0/527 [00:00<?, ?it/s]
ERROR:root:!!! Exception during processing !!!| | 0/527 [00:00<?, ?it/s]
ERROR:root:Traceback (most recent call last):
File "G:_APWorkFlow_ComfyUI\ComfyUI\execution.py", line 152, in recursive_execute
output_data, output_ui = get_output_data(obj, input_data_all)
File "G:_APWorkFlow_ComfyUI\ComfyUI\execution.py", line 82, in get_output_data
return_values = map_node_over_list(obj, input_data_all, obj.FUNCTION, allow_interrupt=True)
File "G:_APWorkFlow_ComfyUI\ComfyUI\execution.py", line 75, in map_node_over_list
results.append(getattr(obj, func)(**slice_dict(input_data_all, i)))
File "G:_APWorkFlow_ComfyUI\ComfyUI\venv\lib\site-packages\torch\utils_contextlib.py", line 115, in decorate_context
return func(*args, **kwargs)
File "G:_APWorkFlow_ComfyUI\ComfyUI\custom_nodes\ComfyUI-CCSR\nodes.py", line 130, in process
samples = sampler.sample_with_tile_ccsr(
File "G:_APWorkFlow_ComfyUI\ComfyUI\venv\lib\site-packages\torch\utils_contextlib.py", line 115, in decorate_context
return func(*args, **kwargs)
File "G:_APWorkFlow_ComfyUI\ComfyUI\custom_nodes\ComfyUI-CCSR\model\q_sampler.py", line 562, in sample_with_tile_ccsr
"c_latent": [self.model.apply_condition_encoder(tile_cond_img)],
File "G:_APWorkFlow_ComfyUI\ComfyUI\custom_nodes\ComfyUI-CCSR\model\ccsr_stage2.py", line 320, in apply_condition_encoder
c_latent_meanvar = self.cond_encoder(control * 2 - 1)
File "G:_APWorkFlow_ComfyUI\ComfyUI\venv\lib\site-packages\torch\nn\modules\module.py", line 1511, in _wrapped_call_impl
return self._call_impl(*args, **kwargs)
File "G:_APWorkFlow_ComfyUI\ComfyUI\venv\lib\site-packages\torch\nn\modules\module.py", line 1520, in _call_impl
return forward_call(*args, **kwargs)
File "G:_APWorkFlow_ComfyUI\ComfyUI\venv\lib\site-packages\torch\nn\modules\container.py", line 217, in forward
input = module(input)
File "G:_APWorkFlow_ComfyUI\ComfyUI\venv\lib\site-packages\torch\nn\modules\module.py", line 1511, in _wrapped_call_impl
return self._call_impl(*args, **kwargs)
File "G:_APWorkFlow_ComfyUI\ComfyUI\venv\lib\site-packages\torch\nn\modules\module.py", line 1520, in _call_impl
return forward_call(*args, **kwargs)
File "G:_APWorkFlow_ComfyUI\ComfyUI\comfy\ldm\modules\diffusionmodules\model.py", line 520, in forward
h = self.conv_in(x)
File "G:_APWorkFlow_ComfyUI\ComfyUI\venv\lib\site-packages\torch\nn\modules\module.py", line 1511, in _wrapped_call_impl
return self._call_impl(*args, **kwargs)
File "G:_APWorkFlow_ComfyUI\ComfyUI\venv\lib\site-packages\torch\nn\modules\module.py", line 1520, in _call_impl
return forward_call(*args, **kwargs)
File "G:_APWorkFlow_ComfyUI\ComfyUI\comfy\ops.py", line 60, in forward
return super().forward(*args, **kwargs)
File "G:_APWorkFlow_ComfyUI\ComfyUI\venv\lib\site-packages\torch\nn\modules\conv.py", line 460, in forward
return self._conv_forward(input, self.weight, self.bias)
File "G:_APWorkFlow_ComfyUI\ComfyUI\venv\lib\site-packages\torch\nn\modules\conv.py", line 456, in _conv_forward
return F.conv2d(input, weight, bias, self.stride,
RuntimeError: Input type (struct c10::Half) and bias type (float) should be the same

Prompt executed in 52.22 seconds
Spaced Sampler: 0%| | 0/20 [00:00<?, ?it/s]

[Limbicnation]

Please try restarting your ComfyUI to resolve the issue.

[qianjun2659]

An error occurred when using real-world_ccsr-fp16.safetensors with the message "Input type (struct c10::Half) and bias type (float) should be the same." However, there is no issue when using real-world_ccsr-fp32.safetensors. The environment is an M1 Mac with macOS