ControlNet Reference Mode VAE Error!

[gohan2091]

Can someone please help? I have tried to fix this myself and I do not know how. I cannot use ControlNet Reference mode (1.5 models). No matter what model checkpoint I use, different VAEs, no VAEs I get the following errors: https://pastebin.com/44Di9DUG

The only way I can get it to work is to Use full precision for model (--no-half), Use full precision for VAE (--no-half-vae) and change Precision type from Autocast to full but then my SD.Next is much slower. Any ideas?

[gohan2091]

Maybe I need to post on Reddit or somewhere else

[vladmandic]

The log actually tells you - you're trying to use BF16 which is not supported by that controlnet model.
So don't.
And you should try just simple default before adding no-half or anything like that.

[gohan2091]

Almost all the settings are at default values. I do not have no-half enabled, it's disabled. I changed to FP16 and now my ControlNet does not work for any mode. Switched back to BF16 and still my ControlNet does not work. Both the preprocessor and model say "none" in the dropdown boxes. I can select them manually but it doesn't work. It gives me this error:

Traceback (most recent call last): File "C:\Users\USER\SD.Next\automatic\venv\lib\site-packages\gradio\queueing.py", line 388, in call_prediction output = await route_utils.call_process_api( File "C:\Users\USER\SD.Next\automatic\venv\lib\site-packages\gradio\route_utils.py", line 219, in call_process_api output = await app.get_blocks().process_api( File "C:\Users\USER\SD.Next\automatic\venv\lib\site-packages\gradio\blocks.py", line 1437, in process_api result = await self.call_function( File "C:\Users\USER\SD.Next\automatic\venv\lib\site-packages\gradio\blocks.py", line 1109, in call_function prediction = await anyio.to_thread.run_sync( File "C:\Users\USER\SD.Next\automatic\venv\lib\site-packages\anyio\to_thread.py", line 33, in run_sync return await get_asynclib().run_sync_in_worker_thread( File "C:\Users\USER\SD.Next\automatic\venv\lib\site-packages\anyio\_backends\_asyncio.py", line 877, in run_sync_in_worker_thread return await future File "C:\Users\USER\SD.Next\automatic\venv\lib\site-packages\anyio\_backends\_asyncio.py", line 807, in run result = context.run(func, *args) File "C:\Users\USER\SD.Next\automatic\venv\lib\site-packages\gradio\utils.py", line 641, in wrapper response = f(*args, **kwargs) File "C:\Users\USER\SD.Next\automatic\extensions-builtin\sd-webui-controlnet\scripts\controlnet_ui\controlnet_ui_group.py", line 609, in filter_selected ) = global_state.select_control_type(k, global_state.get_sd_version()) File "C:\Users\USER\SD.Next\automatic\extensions-builtin\sd-webui-controlnet\scripts\global_state.py", line 248, in get_sd_version elif shared.sd_model.is_sd2: File "C:\Users\USER\SD.Next\automatic\venv\lib\site-packages\torch\nn\modules\module.py", line 1695, in __getattr__ raise AttributeError(f"'{type(self).__name__}' object has no attribute '{name}'") AttributeError: 'LatentDiffusion' object has no attribute 'is_sd2' Traceback (most recent call last): File "C:\Users\USER\SD.Next\automatic\venv\lib\site-packages\gradio\queueing.py", line 388, in call_prediction output = await route_utils.call_process_api( File "C:\Users\USER\SD.Next\automatic\venv\lib\site-packages\gradio\route_utils.py", line 219, in call_process_api output = await app.get_blocks().process_api( File "C:\Users\USER\SD.Next\automatic\venv\lib\site-packages\gradio\blocks.py", line 1437, in process_api result = await self.call_function( File "C:\Users\USER\SD.Next\automatic\venv\lib\site-packages\gradio\blocks.py", line 1109, in call_function prediction = await anyio.to_thread.run_sync( File "C:\Users\USER\SD.Next\automatic\venv\lib\site-packages\anyio\to_thread.py", line 33, in run_sync return await get_asynclib().run_sync_in_worker_thread( File "C:\Users\USER\SD.Next\automatic\venv\lib\site-packages\anyio\_backends\_asyncio.py", line 877, in run_sync_in_worker_thread return await future File "C:\Users\USER\SD.Next\automatic\venv\lib\site-packages\anyio\_backends\_asyncio.py", line 807, in run result = context.run(func, *args) File "C:\Users\USER\SD.Next\automatic\venv\lib\site-packages\gradio\utils.py", line 641, in wrapper response = f(*args, **kwargs) File "C:\Users\USER\SD.Next\automatic\extensions-builtin\sd-webui-controlnet\scripts\controlnet_ui\controlnet_ui_group.py", line 609, in filter_selected ) = global_state.select_control_type(k, global_state.get_sd_version()) File "C:\Users\USER\SD.Next\automatic\extensions-builtin\sd-webui-controlnet\scripts\global_state.py", line 248, in get_sd_version elif shared.sd_model.is_sd2: File "C:\Users\USER\SD.Next\automatic\venv\lib\site-packages\torch\nn\modules\module.py", line 1695, in __getattr__ raise AttributeError(f"'{type(self).__name__}' object has no attribute '{name}'") AttributeError: 'LatentDiffusion' object has no attribute 'is_sd2'

[vladmandic]

thats a totally different error (and please look up how to quote and format logs, this is unreadable) which is introduced in latest controlnet update and just fixed.

[gohan2091]

I do not understand how to paste code on this forum. I use the code tags and it just formats it wrong. That's why in my original post, I quoted pastebin link because that keeps the formatting correct.

Just updated SD.Next and that ControlNet error has been fixed and now using FP16, I can use reference so problem is fixed now., Thanks for your help. So it looks like I have a choice between BF16 (without Reference CN) or FP16 (with Reference CN). I am hoping BF16 will be supported.