Error while using the Skyreel img2vid workflow #371
Comments
|
the same problem |
|
make sure you are using his model and his vae |
I am. Error still occurs. |
|
It looks like the nodes are not up to date for you as it doesn't detect the model type. |
Updated the nodes but I'm now getting a new error: |
That would happen if the latent wasn't fed in the
|
|
I work on WSL2.0 ComfyUI with RTX 4090. after update I got the same error: Given groups=1, weight of size [3072, 32, 1, 2, 2], expected input[2, 16, 25, 80, 80] to have 32 channels, ComfyUI Error:Report Error Details
Stack Trace
|
The input image for this model should be plugged in the |
Thanks. Set TORCH_LOGS="+dynamo" and TORCHDYNAMO_VERBOSE=1 for more information You can suppress this exception and fall back to eager by setting: |
|
Torch compile errors are harder to troubleshoot, can't really help there |
So.... I must wait that ComfyUI will bring a lighter version as they done to Hunyuan t2v |
Compile isn't a necessity, it's a bonus, and the error isn't specific to these nodes and you'd face same issue using compile with the native nodes. |
This configuration works for me:
|
|
In my case i'm getting this: The size of tensor a (32) must match the size of tensor b (16) at non-singleton dimension 1 Any sight? @kijai
](url) |
Don't put anything in the samples input, it's for video2video processes only (and leapfusion as it uses similar approach). |
|
@kijai Thanks for your reply. Now the output is completely black. Any sight why is this happening? No matter what I do, it doesnt work. I lower the number of frames just to try different settings to see if works but it just doesnt work.
|
Which version of pytorch are you using? This is known issue with versions prior to 2.5.0. |
|
pytorch version: 2.3.1+cu121 |
|
I just updated the pytorch version: 2.5.1+cu121, also tried with 2.5.1+cu124
|
Hmm... that I don't have experience with, try with minimum of 17 frames and without the enhance a video. |
|
@kijai I can't thank you enough!! I was getting crazy as I spend two days trying to make it work and nothing worked, but it is working now! Thanks! |
|
Failed to import transformers.models.timm_wrapper.configuration_timm_wrapper because of the following error (look up to see its traceback): Hello there i am getting this issue while trying your workflow any ideas @kijai ? |
That error would come from having too old |
Hi, when I use the skyreel img2vid workflow, I get this error when trying to run it:
!!! Exception during processing !!! Trying to set a tensor of shape torch.Size([3072, 32, 1, 2, 2]) in "weight" (which has shape torch.Size([3072, 16, 1, 2, 2])), this looks incorrect.
Traceback (most recent call last):
File "C:\pinokio\api\comfy.git\app\execution.py", line 327, in execute
output_data, output_ui, has_subgraph = get_output_data(obj, input_data_all, execution_block_cb=execution_block_cb, pre_execute_cb=pre_execute_cb)
File "C:\pinokio\api\comfy.git\app\execution.py", line 202, in get_output_data
return_values = _map_node_over_list(obj, input_data_all, obj.FUNCTION, allow_interrupt=True, execution_block_cb=execution_block_cb, pre_execute_cb=pre_execute_cb)
File "C:\pinokio\api\comfy.git\app\execution.py", line 174, in _map_node_over_list
process_inputs(input_dict, i)
File "C:\pinokio\api\comfy.git\app\execution.py", line 163, in process_inputs
results.append(getattr(obj, func)(**inputs))
File "C:\pinokio\api\comfy.git\app\custom_nodes\ComfyUI-HunyuanVideoWrapper\nodes.py", line 363, in loadmodel
set_module_tensor_to_device(transformer, name, device=transformer_load_device, dtype=dtype_to_use, value=sd[name])
File "C:\pinokio\api\comfy.git\app\env\lib\site-packages\accelerate\utils\modeling.py", line 286, in set_module_tensor_to_device
raise ValueError(
ValueError: Trying to set a tensor of shape torch.Size([3072, 32, 1, 2, 2]) in "weight" (which has shape torch.Size([3072, 16, 1, 2, 2])), this looks incorrect.
What could be causing this error?
The text was updated successfully, but these errors were encountered: