Error occurred when executing T5TextEncode #ELLA: (RX580 i39100f Windows11 32gb Ram)

[KillyTheNetTerminal]

Error occurred when executing T5TextEncode #ELLA:

"addmm_impl_cpu_" not implemented for 'Half'

File "C:\Users\WarMa\OneDrive\Escritorio\ComfyUI\ComfyUI\execution.py", line 151, in recursive_execute
output_data, output_ui = get_output_data(obj, input_data_all)
File "C:\Users\WarMa\OneDrive\Escritorio\ComfyUI\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 "C:\Users\WarMa\OneDrive\Escritorio\ComfyUI\ComfyUI\execution.py", line 74, in map_node_over_list
results.append(getattr(obj, func)(**slice_dict(input_data_all, i)))
File "C:\Users\WarMa\OneDrive\Escritorio\ComfyUI\ComfyUI\custom_nodes\ComfyUI-ELLA\ella.py", line 228, in encode
cond = text_encoder_model(text, max_length=max_length)
File "C:\Users\WarMa\OneDrive\Escritorio\ComfyUI\ComfyUI\custom_nodes\ComfyUI-ELLA\model.py", line 158, in call
outputs = self.model(text_input_ids, attention_mask=attention_mask) # type: ignore
File "C:\Users\WarMa\AppData\Local\Programs\Python\Python310\lib\site-packages\torch\nn\modules\module.py", line 1501, in _call_impl
return forward_call(*args, **kwargs)
File "C:\Users\WarMa\AppData\Local\Programs\Python\Python310\lib\site-packages\transformers\models\t5\modeling_t5.py", line 1980, in forward
encoder_outputs = self.encoder(
File "C:\Users\WarMa\AppData\Local\Programs\Python\Python310\lib\site-packages\torch\nn\modules\module.py", line 1501, in _call_impl
return forward_call(*args, **kwargs)
File "C:\Users\WarMa\AppData\Local\Programs\Python\Python310\lib\site-packages\transformers\models\t5\modeling_t5.py", line 1115, in forward
layer_outputs = layer_module(
File "C:\Users\WarMa\AppData\Local\Programs\Python\Python310\lib\site-packages\torch\nn\modules\module.py", line 1501, in _call_impl
return forward_call(*args, **kwargs)
File "C:\Users\WarMa\AppData\Local\Programs\Python\Python310\lib\site-packages\transformers\models\t5\modeling_t5.py", line 695, in forward
self_attention_outputs = self.layer[0](
File "C:\Users\WarMa\AppData\Local\Programs\Python\Python310\lib\site-packages\torch\nn\modules\module.py", line 1501, in _call_impl
return forward_call(*args, **kwargs)
File "C:\Users\WarMa\AppData\Local\Programs\Python\Python310\lib\site-packages\transformers\models\t5\modeling_t5.py", line 602, in forward
attention_output = self.SelfAttention(
File "C:\Users\WarMa\AppData\Local\Programs\Python\Python310\lib\site-packages\torch\nn\modules\module.py", line 1501, in _call_impl
return forward_call(*args, **kwargs)
File "C:\Users\WarMa\AppData\Local\Programs\Python\Python310\lib\site-packages\transformers\models\t5\modeling_t5.py", line 521, in forward
query_states = shape(self.q(hidden_states)) # (batch_size, n_heads, seq_length, dim_per_head)
File "C:\Users\WarMa\AppData\Local\Programs\Python\Python310\lib\site-packages\torch\nn\modules\module.py", line 1501, in _call_impl
return forward_call(*args, **kwargs)
File "C:\Users\WarMa\AppData\Local\Programs\Python\Python310\lib\site-packages\torch\nn\modules\linear.py", line 114, in forward
return F.linear(input, self.weight, self.bias)

[JettHu]

use --fp32-text-enc refer to

[KillyTheNetTerminal]

oh my god thanks it worked!

[KillyTheNetTerminal]

Exactly the same workflow with the same model but this is the output, something I missing?

[JettHu]

It looks like using --fp32-text-enc affects the results. refer to

The results on my machine are similar to yours.

[JettHu]

It looks like using --fp32-text-enc affects the results. refer to

The results on my machine are similar to yours.

The effect is somewhat different on some GPU models that cannot run fp16. This may be something we need to pay attention to in the future.
cc @budui

[KillyTheNetTerminal]

there is a way to solve this? RX580 can't use fp16?