oxkitsune/rerun-ml-depth-pro

even smaller batch size?

don't clear cuda cache

limit batch size even more

limit batch size

batch model inference

batch model inference

use gradio error + fix support for webcam

webm webcam output?

use gradio-rerun-viewer 0.0.8

fix visualization

rescale depth

clean up logging

improve

add cat

add cat example

set plane distance

fix examples

remove torch

clean up code

simple requirements.txt

convert to images + clean up code

os.path.exists

set limit back

increase limit

clean up video files

change layout

continue, don't break

log every 10 frames

move to device on stateless gpu

put image on correct device

fix code

clean up logging

vertical view

fix double numpy array

move image to correct device

unpin torch

remove pinned version?

spaces

load model once

download pretrained models

drop hashes in requirements.txt

initial commit

initial commit