Upscaler not working after recent update

[Flonixcorn]

After the new update implemented Textual Inversion and new upscalers i get errors when upscaling
LDSR and SwinIR 4x work, the others dont do anything, except for upscaling with ESRGAN 4x it gives me this:
Arguments: (0, <PIL.Image.Image image mode=RGB size=512x512 at 0x150C33BBB80>, None, 0, 0, 0, 2, 4, 0, 1) {} Traceback (most recent call last): File "X:\AI\SUPER SD 2.0 Dependencies\stable-diffusion-webui\modules\ui.py", line 182, in f res = list(func(*args, **kwargs)) File "X:\AI\SUPER SD 2.0 Dependencies\stable-diffusion-webui\webui.py", line 59, in f res = func(*args, **kwargs) File "X:\AI\SUPER SD 2.0 Dependencies\stable-diffusion-webui\modules\extras.py", line 85, in run_extras res = upscale(image, extras_upscaler_1, upscaling_resize) File "X:\AI\SUPER SD 2.0 Dependencies\stable-diffusion-webui\modules\extras.py", line 79, in upscale c = upscaler.scaler.upscale(image, resize, upscaler.data_path) File "X:\AI\SUPER SD 2.0 Dependencies\stable-diffusion-webui\modules\upscaler.py", line 61, in upscale img = self.do_upscale(img, selected_model) File "X:\AI\SUPER SD 2.0 Dependencies\stable-diffusion-webui\modules\esrgan_model.py", line 97, in do_upscale model = self.load_model(selected_model) File "X:\AI\SUPER SD 2.0 Dependencies\stable-diffusion-webui\modules\esrgan_model.py", line 115, in load_model pretrained_net = torch.load(filename, map_location='cpu' if has_mps else None) File "X:\AI\SUPER SD 2.0 Dependencies\stable-diffusion-webui\venv\lib\site-packages\torch\serialization.py", line 713, in load return _legacy_load(opened_file, map_location, pickle_module, **pickle_load_args) File "X:\AI\SUPER SD 2.0 Dependencies\stable-diffusion-webui\venv\lib\site-packages\torch\serialization.py", line 920, in _legacy_load magic_number = pickle_module.load(f, **pickle_load_args) _pickle.UnpicklingError: invalid load key, '<'.

[d8ahazard]

Please delete the .pth file in \models\ESRGAN, then try downloading from
https://github.com/cszn/KAIR/releases/download/v1.0/ESRGAN.pth and replace.

I believe the issue is the URL currently being used has suddenly gone bad, so the pth file is invalid.

Submitted #1524 as a resolution.

[d8ahazard]

This should now be resolved with the latest commit.

[joshistoast]

@d8ahazard Not working for me (latest commit)

[bbecausereasonss]

Having issues /w SwinIR in latest update, havent tried others>

"Traceback (most recent call last):
File "C:\Users\chlyw\Desktop\Automatic1111\modules\ui.py", line 184, in f
res = list(func(*args, **kwargs))
File "C:\Users\chlyw\Desktop\Automatic1111\webui.py", line 64, in f
res = func(*args, **kwargs)
File "C:\Users\chlyw\Desktop\Automatic1111\modules\img2img.py", line 124, in img2img
processed = modules.scripts.scripts_img2img.run(p, *args)
File "C:\Users\chlyw\Desktop\Automatic1111\modules\scripts.py", line 159, in run
processed = script.run(p, *script_args)
File "C:\Users\chlyw\Desktop\Automatic1111\scripts\sd_upscale.py", line 39, in run
img = upscaler.scaler.upscale(init_img, 2, upscaler.data_path)
File "C:\Users\chlyw\Desktop\Automatic1111\modules\upscaler.py", line 62, in upscale
img = self.do_upscale(img, selected_model)
File "C:\Users\chlyw\Desktop\Automatic1111\modules\swinir_model.py", line 42, in do_upscale
model = self.load_model(model_file)
File "C:\Users\chlyw\Desktop\Automatic1111\modules\swinir_model.py", line 94, in load_model
model.load_state_dict(pretrained_model[params], strict=True)
KeyError: 'params_ema'"

[pomee4]

Having issues /w SwinIR in latest update, havent tried others>

"Traceback (most recent call last): File "C:\Users\chlyw\Desktop\Automatic1111\modules\ui.py", line 184, in f res = list(func(*args, **kwargs)) File "C:\Users\chlyw\Desktop\Automatic1111\webui.py", line 64, in f res = func(*args, **kwargs) File "C:\Users\chlyw\Desktop\Automatic1111\modules\img2img.py", line 124, in img2img processed = modules.scripts.scripts_img2img.run(p, *args) File "C:\Users\chlyw\Desktop\Automatic1111\modules\scripts.py", line 159, in run processed = script.run(p, *script_args) File "C:\Users\chlyw\Desktop\Automatic1111\scripts\sd_upscale.py", line 39, in run img = upscaler.scaler.upscale(init_img, 2, upscaler.data_path) File "C:\Users\chlyw\Desktop\Automatic1111\modules\upscaler.py", line 62, in upscale img = self.do_upscale(img, selected_model) File "C:\Users\chlyw\Desktop\Automatic1111\modules\swinir_model.py", line 42, in do_upscale model = self.load_model(model_file) File "C:\Users\chlyw\Desktop\Automatic1111\modules\swinir_model.py", line 94, in load_model model.load_state_dict(pretrained_model[params], strict=True) KeyError: 'params_ema'"

I'm having the same issue.

[gcp]

You're probably trying to use Swin2SR (not SwinIR) networks. You need the exact steps here: #2092 (comment)