I am trying to run YOLOv4 on TLT 3.0. I am recieving an error when trying to use costume channel means in the augmentation config:
google.protobuf.text_format.ParseError: 74:7 : Message type “AugmentationConfig” has no field named “image_mean”.
Attached below is my experiment config file.
experiment_spec.txt (4.1 KB)
I also had the same issue with using the augmentation field “output_channel”
Which version of tlt did you use? The latest one should be 3.0-py3 . That should be compatible with “image_mean”.
Please find the docker_tag when you run “tlt info --verbose”.
The docker tag that I have for TLT 3.0 is v3.0-dp-py3
Please update to TLT 3.0-py3 instead of 3.0-dp-py3.
See TLT Quick Start Guide — Transfer Learning Toolkit 3.0 documentation
The nvidia-tlt package is hosted in the nvidia-pyindex , which has to be installed as a pre-requisite to install nvidia-tlt .
If you had installed an older version of nvidia-tlt launcher, you may upgrade to the latest version by running the following command.
pip3 install --upgrade nvidia-tlt
This topic was automatically closed 60 days after the last reply. New replies are no longer allowed.