zoedepth/KeyError: 'version_name

[pondloso]

File "F:\rave\RAVE\annotator\zoe\zoedepth\utils\config.py", line 384, in get_config
version_name = overwrite_kwargs.get("version_name", config["version_name"])
KeyError: 'version_name

i try to change into line art CN it work but it out of memory
look like my cheap GPU cant do it
and you put xformers into requirements but i'm not see it run it on any command line
........................

a little feedback since other tread is closed

you guy seem to forgot to put gradio in requirements

if any one had bash error don't forget to put git into environment path

and for me when convert model i had to manual create "diffusers_models" folder and "model ID" folder inside "CIVIT_AI" folder
otherwise script will not work.

and in webUI i don't know how to change model it just show sd1.5 when i try to blind run it give error like
mkdir(name, mode)
OSError: [WinError 123] The filename, directory name, or volume label syntax is incorrect: 'F:\rave\RAVE/results/12-09-2023/F:'

maybe i need to change root folder.
........................
So bad look like i had to wait for midvram version
hope it give some help for you. keep it up

[ozgurkara99]

Thanks a lot for the feedback, we tested our code in Linux, and it seems the errors are related to using Windows. We are working on it and hoping to release the Windows version soon, thanks.

[bendeguzvaradi]

I got the same error when using a custom model - I was using RealisticVision - (Windows Ubuntu WSL) even from running it from cmd using python scripts/run_experiment.py:

File "/mnt/d/Dev/RAVE/pipelines/sd_controlnet_rave.py", line 140, in preprocess_control_grid
list_of_pils = [pu.pixel_perfect_process(np.array(frame_pil, dtype='uint8'), self.preprocess_name) for frame_pil in list_of_image_pils]
File "/mnt/d/Dev/RAVE/pipelines/sd_controlnet_rave.py", line 140, in
list_of_pils = [pu.pixel_perfect_process(np.array(frame_pil, dtype='uint8'), self.preprocess_name) for frame_pil in list_of_image_pils]
File "/mnt/d/Dev/RAVE/utils/preprocesser_utils.py", line 211, in pixel_perfect_process
detected_map, _ = preprocessors_dict[p_name](input_image, res=preprocessor_resolution)
File "/mnt/d/Dev/RAVE/utils/preprocesser_utils.py", line 187, in zoe_depth
result = model_zoe_depth(img)
File "/mnt/d/Dev/RAVE/annotator/zoe/init.py", line 38, in call
self.load_model()
File "/mnt/d/Dev/RAVE/annotator/zoe/init.py", line 26, in load_model
conf = get_config("zoedepth", "infer")
File "/mnt/d/Dev/RAVE/annotator/zoe/zoedepth/utils/config.py", line 384, in get_config
version_name = overwrite_kwargs.get("version_name", config["version_name"])
KeyError: 'version_name'

[pondloso]

I got the same error when using a custom model - I was using RealisticVision - (Windows Ubuntu WSL) even from running it from cmd using python scripts/run_experiment.py:

File "/mnt/d/Dev/RAVE/pipelines/sd_controlnet_rave.py", line 140, in preprocess_control_grid list_of_pils = [pu.pixel_perfect_process(np.array(frame_pil, dtype='uint8'), self.preprocess_name) for frame_pil in list_of_image_pils] File "/mnt/d/Dev/RAVE/pipelines/sd_controlnet_rave.py", line 140, in list_of_pils = [pu.pixel_perfect_process(np.array(frame_pil, dtype='uint8'), self.preprocess_name) for frame_pil in list_of_image_pils] File "/mnt/d/Dev/RAVE/utils/preprocesser_utils.py", line 211, in pixel_perfect_process detected_map, _ = preprocessors_dict[p_name](input_image, res=preprocessor_resolution) File "/mnt/d/Dev/RAVE/utils/preprocesser_utils.py", line 187, in zoe_depth result = model_zoe_depth(img) File "/mnt/d/Dev/RAVE/annotator/zoe/init.py", line 38, in call self.load_model() File "/mnt/d/Dev/RAVE/annotator/zoe/init.py", line 26, in load_model conf = get_config("zoedepth", "infer") File "/mnt/d/Dev/RAVE/annotator/zoe/zoedepth/utils/config.py", line 384, in get_config version_name = overwrite_kwargs.get("version_name", config["version_name"]) KeyError: 'version_name'

I change controlnet to line art and it work but out of memory for me you can try it.

[ozgurkara99]

Could you please try to reduce the batch size and try it with lineart please?

[ozgurkara99]

@pondloso hey, could you please try with Depth Zoe, now it shouls work. thanks