Issue regarding dataset format

[Immabot]

Hey there,
I’m currently trying to implement the model for my native language.
Can anyone help me with the dataset format?
In particular what should be the format of train,test and dev files?
I’ve currently referred -

TUTORIAL : How I trained a specific french model to control my robot

Tutorial How to build your homemade deepspeech model from scratch Adapt links and params with your needs… For my robotic project, I needed to create a small monospeaker model, with nearly 1000 sentences orders (not just single word !) I recorded wav’s with a Respeaker Microphone Array : https://www.seeedstudio.com/ReSpeaker-Mic-Array-Far-field-w%2F-7-PDM-Microphones-p-2719.html Wav’s were recorder with the following params : mono / 16 bits / 16 k. The use of the google vad lib helped me t…

But it doesn’t seem to parse the files properly.

Here is the run command I used -
./DeepSpeech.py --train_files /content/drive/My\ Drive/stt_dataset/train/train.csv
–dev_files /content/drive/My\ Drive/stt_dataset/dev/dev.csv
–test_files /content/drive/My\ Drive/stt_dataset/test/test.csv
–export_dir /content/drive/My\ Drive/model_export/
–checkpoint_dir /content/drive/My\ Drive/checkout/
–alphabet_config_path /content/drive/My\ Drive/alphabet.txt \

Any kind of help is appreciated!!

Thank You!

[othiele (Olaf Thiele)]

Train, dev and test are usually split 80/10/10 or 75/20/5 and all have the same layout:

wav_filename,wav_filesize,transcript
/path/audio/common_voice_de_18530569.wav,185900,zahlen sie monatlich vierteljaehrlich halbjaehrlich oder jaehrlich ihren beitrag
/path/audio/common_voice_de_18530616.wav,184364,auf derartige belastungen ist mein gepaecktraeger nicht ausgelegt

But please read the docs first, you are missing a trie or scorer and use a higher batch size. Search the forum a bit, give more information, no images and read here:

https://deepspeech.readthedocs.io/en/v0.6.1/