SD3 Lora support

[AUTOMATIC1111]

• tested to work on all currently existing SD3 loras on civitai
• should probably add more diffusers layer name mappings, but I'd rather do id when there's a lora for me to work with

[tritant]

Lora working with sd3, but if you load a sdxl model lora does not work anymore.
You can find more lora here: https://www.shakker.ai/search?keyword=sd3

    return forward_call(*args, **kwargs)
  File "C:\sd.webui\webui\extensions-builtin\Lora\networks.py", line 639, in network_MultiheadAttention_forward
    network_apply_weights(self)
  File "C:\sd.webui\webui\extensions-builtin\Lora\networks.py", line 440, in network_apply_weights
    bias_backup = store_weights_backup(self.out_proj)
  File "C:\sd.webui\webui\extensions-builtin\Lora\networks.py", line 380, in store_weights_backup
    return weight.to(devices.cpu, copy=True)
  File "C:\sd.webui\webui\venv\lib\site-packages\torch\nn\modules\module.py", line 1141, in to
    device, dtype, non_blocking, convert_to_format = torch._C._nn._parse_to(*args, **kwargs)
RuntimeError: .to() does not accept copy argument

[SunGreen777]

https://www.shakker.ai/search?keyword=sd3

Thanks for this resource. It’s unclear what’s happening on CivitAi