Pip3 install nvidia-tao-byom dependencies

Please provide the following information when requesting support.

• Hardware (T4/V100/Xavier/Nano/etc)
T4
• Network Type (Detectnet_v2/Faster_rcnn/Yolo_v4/LPRnet/Mask_rcnn/Classification/etc)
• TLT Version (Please run “tlt info --verbose” and share “docker_tag” here)
• Training spec file(If have, please share here)
• How to reproduce the issue ? (This is for errors. Please share the command line and the detailed log here.)

Hi,

I’m trying to install the BYOM package for TLT, but it seems that there are several dependencies missing when trying to install:

pip3 install nvidia-tao-byom

ERROR: Could not find a version that satisfies the requirement netron==5.7.7 (from nvidia-tao-byom) (from versions: 5.9.3, 5.9.4, 5.9.5, 5.9.6, 5.9.7, 5.9.8, 5.9.9, 6.0.0)

Can you update the package to include a netron 5.9.3?

Thanks,

/M

Can you try below?
pip3 install --upgrade nvidia-tao-byom

Same problem with --upgrade flag.
Seems that 5.7.7 is no longer available? netron · PyPI

/M

Can you double check the requirement and steps in
BYOM Converter — TAO Toolkit 3.22.05 documentation and GitHub - NVIDIA-AI-IOT/tao_byom_examples: Examples of converting different open-source deep learning models to TAO compatible format through TAO BYOM package. ?

Seems a problem with conda env using the wrong pip… replacing pip3 ... with python-3 m pip ....seems to work, however netron 5.7.7 is still not available;

Running python3 -m pip install nvidia-tao-byom --ignore-installed netron --no-deps will force the installation with netron==6.0.2:

However another exception is then hit…

Driver Version: 516.94
CUDA Version: 11.7
Python 3.6.13 :: Anaconda, Inc.

Suggestions…?

I can reproduce the error.

Please check if below commands work.
$ pip3 install https://urm.nvidia.com/artifactory/api/pypi/sw-colossus-v3-pypi/packages/packages/6a/ee/89d97dbc806637b7f0ac2e3d195f63f0f63feb7d96c230463d4814a08e53/netron-5.7.7-py2.py3-none-any.whl
$ pip3 install nvidia-tao-byom

Seems to be a DNS issue?

/M

So, please ignore above way.
I will sync with internal team for other solution.

1 Like

Is there an update on this?

I will update to you if the new wheel is available. Sorry for the inconvenience.

Hi, the new version of nvidia-tao-byom is available now. https://pypi.org/project/nvidia-tao-byom/0.0.9/.
Please update. Thanks.
conda activate byom_dev
pip3 install nvidia-pyindex
pip3 install nvidia-tensorflow
pip3 install nvidia-tao-byom==0.0.9

Thanks @Morganh - everything is now working as expected!

Cheers,

/M

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