Docker Instantiation Failed with Error when using Detectnet_v2 dataset_convert

Accelerated Computing Intelligent Video Analytics TAO Toolkit
1

Please provide the following information when requesting support.

• Hardware (T4/V100/Xavier/Nano/etc): Workstation with GPU - Quadro P5000
• Network Type (Detectnet_v2/Faster_rcnn/Yolo_v4/LPRnet/Mask_rcnn/Classification/etc): Detectnet_v2
• TLT Version (Please run “tlt info --verbose” and share “docker_tag” here): Tao 3.22.02 and docker tag: v3.21.11-tf1.15.4-py3
• Training spec file(If have, please share here): None
• How to reproduce the issue ? (This is for errors. Please share the command line and the detailed log here.)

Hi, I am currently trying to reproduce the technical blog on using Tao performing transfer learning using tao-toolkit (https://developer.nvidia.com/blog/fast-tracking-hand-gesture-recognition-ai-applications-with-pretrained-models-from-ngc/).

However, I am currently facing an issue while following the handdetect_training jupyter notebook that was provided in https://github.com/NVIDIA-AI-IOT/gesture_recognition_tlt_deepstream/blob/main/training_tlt/handdetect_training.ipynb.

When trying to convert the Egohands dataset from KITTI format to TfRecords using the command:
!tao detectnet_v2 dataset_convert ....

I received the following error:

docker_failed
docker_failed1087×451 126 KB

How do I overcome this error?

2

See TAO Toolkit Launcher — TAO Toolkit 3.22.05 documentation, usually when you run tao launcher for the first time, it will pull the required TAO container.

Can you run follow command in terminal instead of notebook to double check?
$ tao detectnet_v2

3

When running tao detectnet_v2 in terminal, the same error occurred.

docker_failed_cli
docker_failed_cli800×484 88.8 KB

I am not sure why the docker image nvcr.io/nvidia/tao/tao-toolkit-tf:v3.21.11-tf1.15.4-py3 cannot be found

4

Can you run below to check if docker is available? From the log, the docker is pulled completely.
$ docker images

5

docker_images
docker_images770×78 16.3 KB

This is weird, the docker image is not present.

6

Can you try another way?
$ docker pull nvcr.io/nvidia/tao/tao-toolkit-tf:v3.21.11-tf1.15.4-py3

7

It seems that the issue was caused due to having insufficient space on my drive. I will try to free up more space and work this out again. Thank you for the help and sorry for the inconvenience.

1 Like