[Bug]: Can't select VAE in settings: 'ONNXStableDiffusionModel' object has no attribute 'first_stage_model'

hlpientt commented 5 months ago

Checklist

[X] The issue exists after disabling all extensions
[X] The issue exists on a clean installation of webui
[ ] The issue is caused by an extension, but I believe it is caused by a bug in the webui
[X] The issue exists in the current version of the webui
[X] The issue has not been reported before recently
[ ] The issue has been reported before but has not been fixed yet

What happened?

When I select a VAE in the settings tab, it shows an error and the VAE isn't selected.

Steps to reproduce the problem

Start WebUI
Go to settings tab
Select the VAE menu
Select a VAE from the SD VAE drop-down list
Press Apply Settings

What should have happened?

It should select the VAE without an error.

What browsers do you use to access the UI ?

Google Chrome

Sysinfo

sysinfo-2024-01-14-15-53.json

Console logs

F:\stable-diffusion-webui-directml\modules\ui.py:1326: GradioDeprecationWarning: The `style` method is deprecated. Please set these arguments in the constructor instead.
  with gr.Row().style(equal_height=False):
F:\stable-diffusion-webui-directml\modules\ui.py:1448: GradioDeprecationWarning: The `style` method is deprecated. Please set these arguments in the constructor instead.
  with gr.Row().style(equal_height=False):
Running on local URL:  http://127.0.0.1:7860

To create a public link, set `share=True` in `launch()`.
Startup time: 0.4s (create ui: 0.3s).
Loading VAE weights specified in settings: F:\stable-diffusion-webui-directml\models\VAE\vae-ft-mse-840000-ema-pruned.safetensors
changing setting sd_vae to vae-ft-mse-840000-ema-pruned.safetensors: AttributeError
Traceback (most recent call last):
  File "F:\stable-diffusion-webui-directml\modules\options.py", line 146, in set
    option.onchange()
  File "F:\stable-diffusion-webui-directml\modules\call_queue.py", line 13, in f
    res = func(*args, **kwargs)
  File "F:\stable-diffusion-webui-directml\modules\initialize_util.py", line 175, in <lambda>
    shared.opts.onchange("sd_vae", wrap_queued_call(lambda: sd_vae.reload_vae_weights()), call=False)
  File "F:\stable-diffusion-webui-directml\modules\sd_vae.py", line 273, in reload_vae_weights
    load_vae(sd_model, vae_file, vae_source)
  File "F:\stable-diffusion-webui-directml\modules\sd_vae.py", line 209, in load_vae
    store_base_vae(model)
  File "F:\stable-diffusion-webui-directml\modules\sd_vae.py", line 49, in store_base_vae
    base_vae = deepcopy(model.first_stage_model.state_dict())
  File "F:\stable-diffusion-webui-directml\venv\lib\site-packages\torch\nn\modules\module.py", line 1269, in __getattr__
    raise AttributeError("'{}' object has no attribute '{}'".format(
AttributeError: 'ONNXStableDiffusionModel' object has no attribute 'first_stage_model'

Additional information

AMD Radeon RX 7900XT - Latest drivers Intel Core i7-12700F 32GB RAM

lshqqytiger commented 5 months ago

You can't load single file VAE with ONNX. If you want higher compatibility and stability, try olive branch of vladmandic/automatic which provides diffusers-integrated code base. It also supports .safetensors VAE loading.

hlpientt commented 5 months ago

You can't load single file VAE with ONNX. If you want higher compatibility and stability, try olive branch of vladmandic/automatic which provides diffusers-integrated code base. It also supports .safetensors VAE loading.

Thank you for the information, I will definitely check it out.

lshqqytiger / stable-diffusion-webui-amdgpu