Getting no command found error, while trying to train deep speech model with new dataset

Hi I am unable to train deep speech model with new data. While typing the below code,
I am getting facing the below problem :

python DeepSpeech.py --train_files ./customCode/clips/train.csv --dev_files ./customCode/clips/dev.csv --test_files ./customCode/clips/test.csv --beam_width 2000 --export_dir ./modelsDir --learning_rate 0.001 --lm_alpha 0.4 --lm_beta 1.5 --train_batch_size 3 --cudnn_checkpoint ./customCode/deepspeech-0.6.1-checkpoint/ -

This is the following output I am getting:

(deepspeech) nmishr27@myserv:/app/deepSpeech/DeepSpeech-0.6.1

$ Epoch 0 | Training | Elapsed Time: 0:00:00 | Steps: 0 | Loss: 0.000000

bash: Training: command not found

bash: Elapsed: command not found

bash: Steps:: command not found

bash: Loss:: command not found

bash: Epoch: command not found

O.S used is rhel7.

Please help.

What is this ?

It looks like you are passing the log output of the training to your bash and it’s trying to execute commands …

Sir, can you please clarify on what exactly u mean to say by stating this :

“passing the log output of the training to your bash and it’s trying to execute commands”

It looks like you haven’t done much work on Linux before. Maybe try running it in a file like the train_model.sh from this repo. It is for the 0.6 branch.

1 Like