Error while training the model

Archives DeepSpeech
1

Hi,

I have converted mp3 training data to wav and created a train.csv file by using import_cv2.py. However, while training the model i am getting a strange error which reads -

File “/home/ubuntu/DeepSpeech/extracted1/language/archive/clips/train.csv”, line 2
common_voice_en_18885784.wav,208940,he was accorded a state funeral and was buried in drayton and toowoomba cemetery
^
SyntaxError: invalid syntax

Have used the following command for training - python3 /home/ubuntu/DeepSpeech/extracted1/language/archive/clips/train.csv --learning_rate 0.0001 --use_cudnn_rnn true --use_allow_g

Its seems to me that there is issue with the csv file that was created while running the import_cv2. py.

Any help will be appreciated.

Regards

2

Have you properly pasted this ? It lacks the positional arguments.

3

What is the full command you used for training? It looks like it was cutoff.

Also, can you post the line and surrounding context in train.csv that caused the error?