Missing tao models

Hello, this is gonna be probably the last question about TAO.

There are some missing TAO models.

I am trying to test gesturenet and it is not registered on tao model list.

!tao model gesturenet dataset_convert --dataset_spec $SPECS_DIR/dataset_config.json \
                                --k_folds 0 \
                                --experiment_spec $SPECS_DIR/dataset_experiment_config.json \
                                --output_filename $EXPERIMENT_DIR/data.json \
                                --experiment_name v1

then I get

usage: tao model [-h]
                 {list,stop,info,dataset,deploy,model} ...
tao model: error: invalid choice: 'gesturenet' (choose from 'list', 'stop', 'info', 'dataset', 'deploy', 'model')

should I just use other model or is there other way to access the missing models?

Refer to Release Notes - NVIDIA Docs, gesture Recognition is deprecated since TAO5.0.

If you are going to run gesturenet, you can use TAO4.0 docker from TAO Toolkit | NVIDIA NGC.
For example,

1 Like

Thank you for your quick answer :)

It’s been deprecated, but on the Git repo, it’s still updated for GestureNet and a few other models. Because commands like ‘tao model $(target_model)’ have been used since TAO 5.0, if I’m correct, and the examples of those deprecated models are using those updated command lines.

Anyway I thought since the repo is updated for those models, I thought there were still some ways to use those deprecated models xd.

Thank you again and have a nice one.

Could you please share the repo link? Thanks.

its the link you shared before.

Thanks for the info. It is unexpected to remain gesture notebook in 5.2 version. Will sync internally to update the 5.2 notebooks.

1 Like

Not just gesturenet. Heartratenet, Emotionnet are also on the same line with it.

Maybe check the deprecated models :p

Thank you for your work!

Yes, you are right. And also the repo is just for the latest 5.2 notebook.
To run gesturenet, Heartratenet or Emotionnet , you can still download old version of notebooks. For example, to download 4.0.1 notebook, refer to 4.0.1 user guide(TAO Toolkit Quick Start Guide - NVIDIA Docs).

wget --content-disposition https://api.ngc.nvidia.com/v2/resources/nvidia/tao/tao-getting-started/versions/4.0.1/zip -O getting_started_v4.0.1.zip
unzip -u getting_started_v4.0.1.zip  -d ./getting_started_v4.0.1 && rm -rf getting_started_v4.0.1.zip && cd ./getting_started_v4.0.1

Note: Run inside 4.0.1 docker: nvcr.io/nvidia/tao/tao-toolkit:4.0.1-tf1.15.5

1 Like

This topic was automatically closed 14 days after the last reply. New replies are no longer allowed.