Load a checkpoint and start the viewer.

usage: ns-viewer [-h] [OPTIONS]


-h, --help

show this help message and exit


Path to config YAML file. (required)


Possible choices: viewer, viewer_legacy

Type of viewer (default: viewer)

viewer options#

Viewer configuration


Filename to use for the log file. (default: viewer_log_filename.txt)


The websocket port to connect to. If None, find an available port. (default: None)


The default websocket port to connect to if websocket_port is not specified (default: 7007)


The host address to bind the websocket server to. (default:


Maximum number of training images to display in the viewer, to avoid lag. This does not change which images are actually used in training/evaluation. If -1, display all. (default: 512)


Possible choices: True, False

Whether to kill the training job when it has completed. Note this will stop rendering in the viewer. (default: False)


Possible choices: jpeg, png

Image format viewer should use; jpeg is lossy compression, while png is lossless. (default: jpeg)


Quality tradeoff to use for jpeg compression. (default: 75)


Possible choices: True, False

Viewer beta feature: print a shareable URL. This flag is ignored in the legacy version of the viewer. (default: False)


Scale for the camera frustums in the viewer. (default: 0.1)


Possible choices: True, False

The default value for compositing depth. Turn off if you want to see the camera frustums without occlusions. (default: True)