Error occurred when executing SUPIR_Upscale

[wolfam0108]

No operator found for memory_efficient_attention_forward with inputs:
query : shape=(1, 4624, 1, 512) (torch.bfloat16)
key : shape=(1, 4624, 1, 512) (torch.bfloat16)
value : shape=(1, 4624, 1, 512) (torch.bfloat16)
attn_bias :
p : 0.0
decoderF is not supported because:
max(query.shape[-1] != value.shape[-1]) > 128
xFormers wasn't build with CUDA support
attn_bias type is
[email protected] is not supported because:
max(query.shape[-1] != value.shape[-1]) > 256
xFormers wasn't build with CUDA support
operator wasn't built - see python -m xformers.info for more info
tritonflashattF is not supported because:
max(query.shape[-1] != value.shape[-1]) > 128
xFormers wasn't build with CUDA support
operator wasn't built - see python -m xformers.info for more info
triton is not available
cutlassF is not supported because:
xFormers wasn't build with CUDA support
smallkF is not supported because:
max(query.shape[-1] != value.shape[-1]) > 32
xFormers wasn't build with CUDA support
dtype=torch.bfloat16 (supported: {torch.float32})
unsupported embed per head: 512

File "E:\test\UPRES\ComfyUI\execution.py", line 151, in recursive_execute
output_data, output_ui = get_output_data(obj, input_data_all)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "E:\test\UPRES\ComfyUI\execution.py", line 81, in get_output_data
return_values = map_node_over_list(obj, input_data_all, obj.FUNCTION, allow_interrupt=True)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "E:\test\UPRES\ComfyUI\execution.py", line 74, in map_node_over_list
results.append(getattr(obj, func)(**slice_dict(input_data_all, i)))
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "E:\test\UPRES\ComfyUI\custom_nodes\ComfyUI-SUPIR\nodes.py", line 336, in process
samples = self.model.batchify_sample(imgs, caps, num_steps=steps,
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "C:\Users\vovai\miniconda3\Lib\site-packages\torch\utils_contextlib.py", line 115, in decorate_context
return func(*args, **kwargs)
^^^^^^^^^^^^^^^^^^^^^
File "E:\test\UPRES\ComfyUI\custom_nodes\ComfyUI-SUPIR\SUPIR\models\SUPIR_model.py", line 121, in batchify_sample
_z = self.encode_first_stage_with_denoise(x, use_sample=False)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "C:\Users\vovai\miniconda3\Lib\site-packages\torch\utils_contextlib.py", line 115, in decorate_context
return func(*args, **kwargs)
^^^^^^^^^^^^^^^^^^^^^
File "E:\test\UPRES\ComfyUI\custom_nodes\ComfyUI-SUPIR\SUPIR\models\SUPIR_model.py", line 48, in encode_first_stage_with_denoise
h = self.first_stage_model.denoise_encoder(x)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "C:\Users\vovai\miniconda3\Lib\site-packages\torch\nn\modules\module.py", line 1511, in _wrapped_call_impl
return self._call_impl(*args, **kwargs)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "C:\Users\vovai\miniconda3\Lib\site-packages\torch\nn\modules\module.py", line 1520, in call_impl
return forward_call(*args, **kwargs)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "E:\test\UPRES\ComfyUI\custom_nodes\ComfyUI-SUPIR\SUPIR\utils\tilevae.py", line 704, in call
return self.vae_tile_forward(x)
^^^^^^^^^^^^^^^^^^^^^^^^
File "E:\test\UPRES\ComfyUI\custom_nodes\ComfyUI-SUPIR\SUPIR\utils\tilevae.py", line 587, in wrapper
ret = fn(*args, **kwargs)
^^^^^^^^^^^^^^^^^^^
File "C:\Users\vovai\miniconda3\Lib\site-packages\torch\utils_contextlib.py", line 115, in decorate_context
return func(*args, **kwargs)
^^^^^^^^^^^^^^^^^^^^^
File "E:\test\UPRES\ComfyUI\custom_nodes\ComfyUI-SUPIR\SUPIR\utils\tilevae.py", line 937, in vae_tile_forward
tile = task1
^^^^^^^^^^^^^
File "E:\test\UPRES\ComfyUI\custom_nodes\ComfyUI-SUPIR\SUPIR\utils\tilevae.py", line 372, in
('attn', lambda x, net=net: xformer_attn_forward(net, x)))
^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "E:\test\UPRES\ComfyUI\custom_nodes\ComfyUI-SUPIR\SUPIR\utils\tilevae.py", line 340, in xformer_attn_forward
out = xformers.ops.memory_efficient_attention(
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "C:\Users\vovai\miniconda3\Lib\site-packages\xformers\ops\fmha_init.py", line 223, in memory_efficient_attention
return memory_efficient_attention(
^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "C:\Users\vovai\miniconda3\Lib\site-packages\xformers\ops\fmha_init.py", line 321, in _memory_efficient_attention
return memory_efficient_attention_forward(
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "C:\Users\vovai\miniconda3\Lib\site-packages\xformers\ops\fmha_init.py", line 337, in _memory_efficient_attention_forward
op = _dispatch_fw(inp, False)
^^^^^^^^^^^^^^^^^^^^^^^^
File "C:\Users\vovai\miniconda3\Lib\site-packages\xformers\ops\fmha\dispatch.py", line 120, in _dispatch_fw
return _run_priority_list(
^^^^^^^^^^^^^^^^^^^
File "C:\Users\vovai\miniconda3\Lib\site-packages\xformers\ops\fmha\dispatch.py", line 63, in _run_priority_list
raise NotImplementedError(msg)

windows 11 23h2

[kijai]

Unsure why your xformers isn't working, but it's safe to uninstall.

[liuzhao6322]

重新安装一下 Torch2.2.0 CUDA12.1+ xFormers 0.0.24，就可以正常运行了

[liaosu0755]

重新安装一下 Torch2.2.0 CUDA12.1+ xFormers 0.0.24，就可以正常运行了

谢谢，用你的方法解决了这个问题。
不过我还多一个问题，SUPIR Model Loader (v2)加载模型中途会卡死，换回SUPIR Model Loader (Legacy)终于能用了