I have created pytorch container within the Contianer, Create a conda virtual environment, namely “launcher” using Python 3.7 and installed tao 5.0.0 inside the virtual env, then runing TAO toolkit quick start notebook (ssd.ipynb) file as per the instructions .
Please confirm me
- Is it required to run bash setup/quickstart_launcher.sh --install ?
- Is it required to install docker-ce ?
When executing notebook cell one by one at the time of tao mode ssd datasets_covert, the followings error reported. Please guide .
The way i am trigger the pytorch container
apiVersion: apps/v1
kind: Deployment
metadata:
name: tao
spec:
selector:
matchLabels:
run: tao
replicas: 1
template:
metadata:
labels:
run: tao
spec:
nodeName: dgxa100
containers:
- name: tao
image: nvcr.io/nvidia/pytorch:23.03.01
command: [“/bin/sh”, “-c”]
args:
- jupyter lab --allow-root --ip=0.0.0.0 --NotebookApp.token=‘tao’;
resources:
limits:
nvidia.com/${profile}: 1
volumeMounts:
- name: tao-data
mountPath: /workspace/nemo/data
ports:
- containerPort: 8888
volumes:
- name: tao-data
hostPath:
path: /home/tao
type: Directory
When executing notebook cell one by one at the time of tao mode ssd datasets_covert, the followings error reported. Please guide .
Converting the training set to TFRecords.
Traceback (most recent call last):
File “/opt/conda/envs/tao/bin/tao”, line 8, in
sys.exit(main())
File “/opt/conda/envs/tao/lib/python3.7/site-packages/nvidia_tao_cli/entrypoint/tao_launcher.py”, line 137, in main
args[2:]
File “/opt/conda/envs/tao/lib/python3.7/site-packages/nvidia_tao_cli/components/instance_handler/local_instance.py”, line 356, in launch_command
docker_logged_in(required_registry=task_map[task].docker_registry)
File “/opt/conda/envs/tao/lib/python3.7/site-packages/nvidia_tao_cli/components/instance_handler/utils.py”, line 151, in docker_logged_in
data = load_config_file(docker_config)
File “/opt/conda/envs/tao/lib/python3.7/site-packages/nvidia_tao_cli/components/instance_handler/utils.py”, line 86, in load_config_file
“No file found at: {}. Did you run docker login?”.format(config_path)
AssertionError: Config path must be a valid unix path. No file found at: /root/.docker/config.json. Did you run docker login?
