fix error "TypeError: 'zip' object is not subscriptable" in deepspeec…

[moneypi]

…h.cfg training

Description

When traing speech_recognition by command python main.py --configfile deepspeech.cfg, I will get an error:
Traceback (most recent call last): File "main.py", line 299, in <module> data_train, data_val, args = load_data(args) File "main.py", line 173, in load_data save_feature_as_csvfile=save_feature_as_csvfile) File "/home/luke/tmp/mxnet/example/speech_recognition/stt_io_bucketingiter.py", line 105, in __init__ self.data[buck].append(self.trainDataList[i]) TypeError: 'zip' object is not subscriptable

Checklist

Essentials

Please feel free to remove inapplicable items for your PR.

• All changes have test coverage:
• To the best of my knowledge, examples are either not affected by this change, or have been fixed to be compatible with this change

Comments

Just like file https://github.com/apache/incubator-mxnet/blob/master/example/speech_recognition/stt_io_iter.py, use the same logic:
self.trainDataList = list(zip(durations, audio_paths, texts))

[ChaiBapchya]

One of those py2 vs py3 changes right?
If run in py2 this example would have worked right? Else I'm curious as to why this example was not checked/tested at the time of introduction to the repo

[leezu]

@ChaiBapchya yes, it's a "Py3 only bug".