Comfy-Org

/

ace_step_1.5_ComfyUI_files

like 133

Follow

Comfy Org 8.26k

RuntimeError: Error(s) in loading state_dict for AceStepConditionGenerationModel:
size mismatch for decoder.scale_shift_table: copying a param with shape torch.Size([1, 2, 2560]) from checkpoint, the shape in current model is torch.Size([1, 2, 2048]).
size mismatch for decoder.proj_in.1.weight: copying a param with shape torch.Size([2560, 192, 2]) from checkpoint, the shape in current model is torch.Size([2048, 192, 2]).
size mismatch for decoder.proj_in.1.bias: copying a param with shape torch.Size([2560]) from checkpoint, the shape in current model is torch.Size([2048]).
size mismatch for decoder.layers.0.scale_shift_table: copying a param with shape torch.Size([1, 6, 2560]) from checkpoint, the shape in current model is torch.Size([1, 6, 2048]).
size mismatch for decoder.layers.0.self_attn_norm.weight: copying a param with shape torch.Size([2560]) from checkpoint, the shape in current model is torch.Size([2048]).
size mismatch for decoder.layers.0.cross_attn_norm.weight: copying a param with shape torch.Size([2560]) from checkpoint, the shape in current model is torch.Size([2048]).
size mismatch for decoder.layers.0.mlp_norm.weight: copying a param with shape torch.Size([2560]) from checkpoint, the shape in current model is torch.Size([2048]).
size mismatch for decoder.layers.1.scale_shift_table: copying a param with shape torch.Size([1, 6, 2560]) from checkpoint, the shape in current model is torch.Size([1, 6, 2048]).
size mismatch for decoder.layers.1.self_attn_norm.weight: copying a param with shape torch.Size([2560]) from checkpoint, the shape in current model is torch.Size([2048]).
size mismatch for decoder.layers.1.cross_attn_norm.weight: copying a param with shape torch.Size([2560]) from checkpoint, the shape in current model is torch.Size([2048]).
size mismatch for decoder.layers.1.mlp_norm.weight: copying a param with shape torch.Size([2560]) from checkpoint, the shape in current model is torch.Size([2048]).
size mismatch for decoder.layers.2.scale_shift_table: copying a param with shape torch.Size([1, 6, 2560]) from checkpoint, the shape in current model is torch.Size([1, 6, 2048]).
size mismatch for decoder.layers.2.self_attn_norm.weight: copying a param with shape torch.Size([2560]) from checkpoint, the shape in current model is torch.Size([2048]).
size mismatch for decoder.layers.2.cross_attn_norm.weight: copying a param with shape torch.Size([2560]) from checkpoint, the shape in current model is torch.Size([2048]).
size mismatch for decoder.layers.2.mlp_norm.weight: copying a param with shape torch.Size([2560]) from checkpoint, the shape in current model is torch.Size([2048]).
size mismatch for decoder.layers.3.scale_shift_table: copying a param with shape torch.Size([1, 6, 2560]) from checkpoint, the shape in current model is torch.Size([1, 6, 2048]).
size mismatch for decoder.layers.3.self_attn_norm.weight: copying a param with shape torch.Size([2560]) from checkpoint, the shape in current model is torch.Size([2048]).
size mismatch for decoder.layers.3.cross_attn_norm.weight: copying a param with shape torch.Size([2560]) from checkpoint, the shape in current model is torch.Size([2048]).
size mismatch for decoder.layers.3.mlp_norm.weight: copying a param with shape torch.Size([2560]) from checkpoint, the shape in current model is torch.Size([2048]).
size mismatch for decoder.layers.4.scale_shift_table: copying a param with shape torch.Size([1, 6, 2560]) from checkpoint, the shape in current model is torch.Size([1, 6, 2048]).
size mismatch for decoder.layers.4.self_attn_norm.weight: copying a param with shape torch.Size([2560]) from checkpoint, the shape in current model is torch.Size([2048]).
size mismatch for decoder.layers.4.cross_attn_norm.weight: copying a param with shape torch.Size([2560]) from checkpoint, the shape in current model is torch.Size([2048]).
size mismatch for decoder.layers.4.mlp_norm.weight: copying a param with shape torch.Size([2560]) from checkpoint, the shape in current model is torch.Size([2048]).
size mismatch for decoder.layers.5.scale_shift_table: copying a param with shape torch.Size([1, 6, 2560]) from checkpoint, the shape in current model is torch.Size([1, 6, 2048]).
size mismatch for decoder.layers.5.self_attn_norm.weight: copying a param with shape torch.Size([2560]) from checkpoint, the shape in current model is torch.Size([2048]).
size mismatch for decoder.layers.5.cross_attn_norm.weight: copying a param with shape torch.Size([2560]) from checkpoint, the shape in current model is torch.Size([2048]).
size mismatch for decoder.layers.5.mlp_norm.weight: copying a param with shape torch.Size([2560]) from checkpoint, the shape in current model is torch.Size([2048]).
size mismatch for decoder.layers.6.scale_shift_table: copying a param with shape torch.Size([1, 6, 2560]) from checkpoint, the shape in current model is torch.Size([1, 6, 2048]).
size mismatch for decoder.layers.6.self_attn_norm.weight: copying a param with shape torch.Size([2560]) from checkpoint, the shape in current model is torch.Size([2048]).
size mismatch for decoder.layers.6.cross_attn_norm.weight: copying a param with shape torch.Size([2560]) from checkpoint, the shape in current model is torch.Size([2048]).
size mismatch for decoder.layers.6.mlp_norm.weight: copying a param with shape torch.Size([2560]) from checkpoint, the shape in current model is torch.Size([2048]).
size mismatch for decoder.layers.7.scale_shift_table: copying a param with shape torch.Size([1, 6, 2560]) from checkpoint, the shape in current model is torch.Size([1, 6, 2048]).
size mismatch for decoder.layers.7.self_attn_norm.weight: copying a param with shape torch.Size([2560]) from checkpoint, the shape in current model is torch.Size([2048]).
size mismatch for decoder.layers.7.cross_attn_norm.weight: copying a param with shape torch.Size([2560]) from checkpoint, the shape in current model is torch.Size([2048]).
size mismatch for decoder.layers.7.mlp_norm.weight: copying a param with shape torch.Size([2560]) from checkpoint, the shape in current model is torch.Size([2048]).
size mismatch for decoder.layers.8.scale_shift_table: copying a param with shape torch.Size([1, 6, 2560]) from checkpoint, the shape in current model is torch.Size([1, 6, 2048]).
size mismatch for decoder.layers.8.self_attn_norm.weight: copying a param with shape torch.Size([2560]) from checkpoint, the shape in current model is torch.Size([2048]).
size mismatch for decoder.layers.8.cross_attn_norm.weight: copying a param with shape torch.Size([2560]) from checkpoint, the shape in current model is torch.Size([2048]).
size mismatch for decoder.layers.8.mlp_norm.weight: copying a param with shape torch.Size([2560]) from checkpoint, the shape in current model is torch.Size([2048]).
size mismatch for decoder.layers.9.scale_shift_table: copying a param with shape torch.Size([1, 6, 2560]) from checkpoint, the shape in current model is torch.Size([1, 6, 2048]).
size mismatch for decoder.layers.9.self_attn_norm.weight: copying a param with shape torch.Size([2560]) from checkpoint, the shape in current model is torch.Size([2048]).
size mismatch for decoder.layers.9.cross_attn_norm.weight: copying a param with shape torch.Size([2560]) from checkpoint, the shape in current model is torch.Size([2048]).
size mismatch for decoder.layers.9.mlp_norm.weight: copying a param with shape torch.Size([2560]) from checkpoint, the shape in current model is torch.Size([2048]).
size mismatch for decoder.layers.10.scale_shift_table: copying a param with shape torch.Size([1, 6, 2560]) from checkpoint, the shape in current model is torch.Size([1, 6, 2048]).
size mismatch for decoder.layers.10.self_attn_norm.weight: copying a param with shape torch.Size([2560]) from checkpoint, the shape in current model is torch.Size([2048]).
size mismatch for decoder.layers.10.cross_attn_norm.weight: copying a param with shape torch.Size([2560]) from checkpoint, the shape in current model is torch.Size([2048]).
size mismatch for decoder.layers.10.mlp_norm.weight: copying a param with shape torch.Size([2560]) from checkpoint, the shape in current model is torch.Size([2048]).
size mismatch for decoder.layers.11.scale_shift_table: copying a param with shape torch.Size([1, 6, 2560]) from checkpoint, the shape in current model is torch.Size([1, 6, 2048]).
size mismatch for decoder.layers.11.self_attn_norm.weight: copying a param with shape torch.Size([2560]) from checkpoint, the shape in current model is torch.Size([2048]).
size mismatch for decoder.layers.11.cross_attn_norm.weight: copying a param with shape torch.Size([2560]) from checkpoint, the shape in current model is torch.Size([2048]).
size mismatch for decoder.layers.11.mlp_norm.weight: copying a param with shape torch.Size([2560]) from checkpoint, the shape in current model is torch.Size([2048]).
size mismatch for decoder.layers.12.scale_shift_table: copying a param with shape torch.Size([1, 6, 2560]) from checkpoint, the shape in current model is torch.Size([1, 6, 2048]).
size mismatch for decoder.layers.12.self_attn_norm.weight: copying a param with shape torch.Size([2560]) from checkpoint, the shape in current model is torch.Size([2048]).
size mismatch for decoder.layers.12.cross_attn_norm.weight: copying a param with shape torch.Size([2560]) from checkpoint, the shape in current model is torch.Size([2048]).
size mismatch for decoder.layers.12.mlp_norm.weight: copying a param with shape torch.Size([2560]) from checkpoint, the shape in current model is torch.Size([2048]).
size mismatch for decoder.layers.13.scale_shift_table: copying a param with shape torch.Size([1, 6, 2560]) from checkpoint, the shape in current model is torch.Size([1, 6, 2048]).
size mismatch for decoder.layers.13.self_attn_norm.weight: copying a param with shape torch.Size([2560]) from checkpoint, the shape in current model is torch.Size([2048]).
size mismatch for decoder.layers.13.cross_attn_norm.weight: copying a param with shape torch.Size([2560]) from checkpoint, the shape in current model is torch.Size([2048]).
size mismatch for decoder.layers.13.mlp_norm.weight: copying a param with shape torch.Size([2560]) from checkpoint, the shape in current model is torch.Size([2048]).
size mismatch for decoder.layers.14.scale_shift_table: copying a param with shape torch.Size([1, 6, 2560]) from checkpoint, the shape in current model is torch.Size([1, 6, 2048]).
size mismatch for decoder.layers.14.self_attn_norm.weight: copying a param with shape torch.Size([2560]) from checkpoint, the shape in current model is torch.Size([2048]).
size mismatch for decoder.layers.14.cross_attn_norm.weight: copying a param with shape torch.Size([2560]) from checkpoint, the shape in current model is torch.Size([2048]).
size mismatch for decoder.layers.14.mlp_norm.weight: copying a param with shape torch.Size([2560]) from checkpoint, the shape in current model is torch.Size([2048]).
size mismatch for decoder.layers.15.scale_shift_table: copying a param with shape torch.Size([1, 6, 2560]) from checkpoint, the shape in current model is torch.Size([1, 6, 2048]).
size mismatch for decoder.layers.15.self_attn_norm.weight: copying a param with shape torch.Size([2560]) from checkpoint, the shape in current model is torch.Size([2048]).
size mismatch for decoder.layers.15.cross_attn_norm.weight: copying a param with shape torch.Size([2560]) from checkpoint, the shape in current model is torch.Size([2048]).
size mismatch for decoder.layers.15.mlp_norm.weight: copying a param with shape torch.Size([2560]) from checkpoint, the shape in current model is torch.Size([2048]).
size mismatch for decoder.layers.16.scale_shift_table: copying a param with shape torch.Size([1, 6, 2560]) from checkpoint, the shape in current model is torch.Size([1, 6, 2048]).
size mismatch for decoder.layers.16.self_attn_norm.weight: copying a param with shape torch.Size([2560]) from checkpoint, the shape in current model is torch.Size([2048]).
size mismatch for decoder.layers.16.cross_attn_norm.weight: copying a param with shape torch.Size([2560]) from checkpoint, the shape in current model is torch.Size([2048]).
size mismatch for decoder.layers.16.mlp_norm.weight: copying a param with shape torch.Size([2560]) from checkpoint, the shape in current model is torch.Size([2048]).
size mismatch for decoder.layers.17.scale_shift_table: copying a param with shape torch.Size([1, 6, 2560]) from checkpoint, the shape in current model is torch.Size([1, 6, 2048]).
size mismatch for decoder.layers.17.self_attn_norm.weight: copying a param with shape torch.Size([2560]) from checkpoint, the shape in current model is torch.Size([2048]).
size mismatch for decoder.layers.17.cross_attn_norm.weight: copying a param with shape torch.Size([2560]) from checkpoint, the shape in current model is torch.Size([2048]).
size mismatch for decoder.layers.17.mlp_norm.weight: copying a param with shape torch.Size([2560]) from checkpoint, the shape in current model is torch.Size([2048]).
size mismatch for decoder.layers.18.scale_shift_table: copying a param with shape torch.Size([1, 6, 2560]) from checkpoint, the shape in current model is torch.Size([1, 6, 2048]).
size mismatch for decoder.layers.18.self_attn_norm.weight: copying a param with shape torch.Size([2560]) from checkpoint, the shape in current model is torch.Size([2048]).
size mismatch for decoder.layers.18.cross_attn_norm.weight: copying a param with shape torch.Size([2560]) from checkpoint, the shape in current model is torch.Size([2048]).
size mismatch for decoder.layers.18.mlp_norm.weight: copying a param with shape torch.Size([2560]) from checkpoint, the shape in current model is torch.Size([2048]).
size mismatch for decoder.layers.19.scale_shift_table: copying a param with shape torch.Size([1, 6, 2560]) from checkpoint, the shape in current model is torch.Size([1, 6, 2048]).
size mismatch for decoder.layers.19.self_attn_norm.weight: copying a param with shape torch.Size([2560]) from checkpoint, the shape in current model is torch.Size([2048]).
size mismatch for decoder.layers.19.cross_attn_norm.weight: copying a param with shape torch.Size([2560]) from checkpoint, the shape in current model is torch.Size([2048]).
size mismatch for decoder.layers.19.mlp_norm.weight: copying a param with shape torch.Size([2560]) from checkpoint, the shape in current model is torch.Size([2048]).
size mismatch for decoder.layers.20.scale_shift_table: copying a param with shape torch.Size([1, 6, 2560]) from checkpoint, the shape in current model is torch.Size([1, 6, 2048]).
size mismatch for decoder.layers.20.self_attn_norm.weight: copying a param with shape torch.Size([2560]) from checkpoint, the shape in current model is torch.Size([2048]).
size mismatch for decoder.layers.20.cross_attn_norm.weight: copying a param with shape torch.Size([2560]) from checkpoint, the shape in current model is torch.Size([2048]).
size mismatch for decoder.layers.20.mlp_norm.weight: copying a param with shape torch.Size([2560]) from checkpoint, the shape in current model is torch.Size([2048]).
size mismatch for decoder.layers.21.scale_shift_table: copying a param with shape torch.Size([1, 6, 2560]) from checkpoint, the shape in current model is torch.Size([1, 6, 2048]).
size mismatch for decoder.layers.21.self_attn_norm.weight: copying a param with shape torch.Size([2560]) from checkpoint, the shape in current model is torch.Size([2048]).
size mismatch for decoder.layers.21.cross_attn_norm.weight: copying a param with shape torch.Size([2560]) from checkpoint, the shape in current model is torch.Size([2048]).
size mismatch for decoder.layers.21.mlp_norm.weight: copying a param with shape torch.Size([2560]) from checkpoint, the shape in current model is torch.Size([2048]).
size mismatch for decoder.layers.22.scale_shift_table: copying a param with shape torch.Size([1, 6, 2560]) from checkpoint, the shape in current model is torch.Size([1, 6, 2048]).
size mismatch for decoder.layers.22.self_attn_norm.weight: copying a param with shape torch.Size([2560]) from checkpoint, the shape in current model is torch.Size([2048]).
size mismatch for decoder.layers.22.cross_attn_norm.weight: copying a param with shape torch.Size([2560]) from checkpoint, the shape in current model is torch.Size([2048]).
size mismatch for decoder.layers.22.mlp_norm.weight: copying a param with shape torch.Size([2560]) from checkpoint, the shape in current model is torch.Size([2048]).
size mismatch for decoder.layers.23.scale_shift_table: copying a param with shape torch.Size([1, 6, 2560]) from checkpoint, the shape in current model is torch.Size([1, 6, 2048]).
size mismatch for decoder.layers.23.self_attn_norm.weight: copying a param with shape torch.Size([2560]) from checkpoint, the shape in current model is torch.Size([2048]).
size mismatch for decoder.layers.23.cross_attn_norm.weight: copying a param with shape torch.Size([2560]) from checkpoint, the shape in current model is torch.Size([2048]).
size mismatch for decoder.layers.23.mlp_norm.weight: copying a param with shape torch.Size([2560]) from checkpoint, the shape in current model is torch.Size([2048]).
size mismatch for decoder.norm_out.weight: copying a param with shape torch.Size([2560]) from checkpoint, the shape in current model is torch.Size([2048]).
size mismatch for decoder.proj_out.1.weight: copying a param with shape torch.Size([2560, 64, 2]) from checkpoint, the shape in current model is torch.Size([2048, 64, 2]).

File "C:\Users\sarah\AppData\Local\Programs\ComfyUI\resources\ComfyUI\execution.py", line 525, in execute
output_data, output_ui, has_subgraph, has_pending_tasks = await get_output_data(prompt_id, unique_id, obj, input_data_all, execution_block_cb=execution_block_cb, pre_execute_cb=pre_execute_cb, v3_data=v3_data)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^

File "C:\Users\sarah\AppData\Local\Programs\ComfyUI\resources\ComfyUI\execution.py", line 334, in get_output_data
return_values = await _async_map_node_over_list(prompt_id, unique_id, obj, input_data_all, obj.FUNCTION, allow_interrupt=True, execution_block_cb=execution_block_cb, pre_execute_cb=pre_execute_cb, v3_data=v3_data)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^

File "C:\Users\sarah\AppData\Local\Programs\ComfyUI\resources\ComfyUI\execution.py", line 308, in _async_map_node_over_list
await process_inputs(input_dict, i)

File "C:\Users\sarah\AppData\Local\Programs\ComfyUI\resources\ComfyUI\execution.py", line 296, in process_inputs
result = f(**inputs)
^^^^^^^^^^^

File "C:\Users\sarah\AppData\Local\Programs\ComfyUI\resources\ComfyUI\nodes.py", line 973, in load_unet
model = comfy.sd.load_diffusion_model(unet_path, model_options=model_options)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^

File "C:\Users\sarah\AppData\Local\Programs\ComfyUI\resources\ComfyUI\comfy\sd.py", line 1794, in load_diffusion_model
model = load_diffusion_model_state_dict(sd, model_options=model_options, metadata=metadata, disable_dynamic=disable_dynamic)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^

File "C:\Users\sarah\AppData\Local\Programs\ComfyUI\resources\ComfyUI\comfy\sd.py", line 1786, in load_diffusion_model_state_dict
model.load_model_weights(new_sd, "", assign=model_patcher.is_dynamic())

File "C:\Users\sarah\AppData\Local\Programs\ComfyUI\resources\ComfyUI\comfy\model_base.py", line 327, in load_model_weights
m, u = self.diffusion_model.load_state_dict(to_load, strict=False, assign=assign)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^

File "C:\2TC\New folder.venv\Lib\site-packages\torch\nn\modules\module.py", line 2624, in load_state_dict
raise RuntimeError(

Errors for me as well

You have to use the dual clip loader, that's what worked for me

I updated ComfyUI to Nightly version and solved, expect yours too.

You have to use the dual clip loader, that's what worked for me

I have this already.

I had the same error; the problem was an outdated workflow version. After downloading the updated script from the official ComfyUI website, it worked without any errors.
https://docs.comfy.org/tutorials/audio/ace-step/ace-step-v1-5 (Option 2: Split Model Files)
Direct link: https://raw.githubusercontent.com/Comfy-Org/workflow_templates/refs/heads/main/templates/audio_ace_step_1_5_split.json
SHA256: b99a35b6a74d9df4df53e1078749666439d174e294376c134bd8c0e623a4f6db

I had the same error; the problem was an outdated workflow version. After downloading the updated script from the official ComfyUI website, it worked without any errors.
https://docs.comfy.org/tutorials/audio/ace-step/ace-step-v1-5 (Option 2: Split Model Files)
Direct link: https://raw.githubusercontent.com/Comfy-Org/workflow_templates/refs/heads/main/templates/audio_ace_step_1_5_split.json
SHA256: b99a35b6a74d9df4df53e1078749666439d174e294376c134bd8c0e623a4f6db

Just tried this and it's the same version I have. It works on the ace-step gradio version but that isn't what I want lol

I had the same issue. Switch to the nightly version.