[Errno 2] No such file or directory:when 'tao classification train' is run

Please provide the following information when requesting support.

• Hardware (RGX 3080)
• Network Type (Classification)
• TLT Version format_version: 2.0
toolkit_version: 3.22.05
published_date: 05/25/2022()
• Training spec file
• How to reproduce the issue ?
Login Succeeded
When I run the following command I get a [Errno 2] error, despite using the absolute path to the directory where the ‘training _spec.json’ is located.


Is the path wrong?
Is the ‘training_spec.json’ not the file it is looking for?

The documentation at:
https://docs.nvidia.com/tao/tao-toolkit/text/image_classification.html
does not specify a name for the Spec File
Please advise

It depends on your ~/.tao_mount.json.

Make sure the ““destination”” path is available.

Thank you Morganh.

I have a further question but I will open a new thread.

Hi there,

I still have this same problem.

It depends on your ~/.tao_mount.json. Make sure the ““destination”” path is available.

TAO Toolkit Launcher — TAO Toolkit 3.22.05 documentation refers to ~/.tao_mounts.json and that is how I have named the file (‘mounts’, not ‘mount’)
Should I rename the file?

The path referred to in the Errno 2 is in the ~/.tao_mounts.json. and, as in post 1, the path is available.

The experiment spec is there, called 'training_spec.json.
Is this the filename that is being looked for?
TAO Toolkit Launcher — TAO Toolkit 3.22.05 documentation refers to ’ -e /workspace/tlt_drive/spec/spec.cfg

If not, what should it be named?

Hi I am still awaiting a response.

I have exactly the same issue when working with the byom Jupyter notebook


Same Errno 2 can’t find the training_spec.json

Here is the bash ‘$ ls’ showing ‘training_spec.json’ located in /home/peter/TAO_toolkit/config/files
And below is the same thing in ‘.tao_mounts.json’

Please advise

Hi,
Can you share the result of
$ cat ~/.tao_mounts.json

From your .tao_mounts.json, you already mount “/home/perter/TAO_toolkit” to “/workspace/tao-experiments”.
In all command such as " ! tao classification train -e spec.txt " , please note that the spec.txt should be in the path of docker. In other words, the spec.txt should be available in “destination” of the tao_mounts.josn.

If I understand this correctly, I need to:

  1. add another “source”: and “destination”; pair to my tao_mounts.json

  2. spec.txt should be in “destination” i.e “destination”: “/home/peter/TAO_toolkit/config/files/training_spec.json”

  3. What comes after “source”:

/workspace/tao-experiments/classification?

Please advise

Refer to TAO Toolkit Launcher — TAO Toolkit 3.22.05 documentation.

For ease, I suggest you to set “source” and “destination” to the same.

Thank you Morganh, that did it.

Can you give me any idea why that was necessary (please)?

Not necessary to be same. Just a simplest way for end users.

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