Classification --init_epoch not

[Mangesh_mane]

Please provide the following information when requesting support.

• HardwareDGpu
• Network Type Classification
• TLT Version docker tag name v3.22.05-tf1.15.5-py3
in tao classification train --init_epoch not working on insider of docker contianer when i call the tao
classification train -e $SPECS_DIR/classification_spec.cfg
"-r $SPECS_DIR/output -k $KEY --init_epoch 10
its working
but inside of docker
classification train -e $SPECS_DIR/classification_spec.cfg
"-r $SPECS_DIR/output -k $KEY --init_epoch 10
showing blow error:
classification train: error: invalid choice: ‘–init_epoch 80’ (choose from ‘calibration_tensorfile’, ‘evaluate’, ‘export’, ‘inference’, ‘prune’, ‘train’)

[Morganh]

It does not make sense. Please share the full command and full log.
And also please double check.

[Mangesh_mane]

I specifically run the following command inside of the Docker Container
classification train -e $SPECS_DIR/classification_spec.cfg
"-r $SPECS_DIR/output -k $KEY --init_epoch 10
and the error classification train given below error: invalid selection using “-init epoch 80” (Select one of the following: “calibration tensorfile,” “evaluate,” “export,” “inference,” “prune,” or “train”)

[Morganh]

There is no update from you for a period, assuming this is not an issue anymore.
Hence we are closing this topic. If need further support, please open a new one.
Thanks

I cannot reproduce your issue inside the docker.
Suggest you to double check.

Why there is an additional character in "-r ? Should be -r, right?

Share my example as below.
$ tao classification run /bin/bash
# classification train -e notebook.cfg -r result_forum -k mykey --init_epoch 2