Skip to content

Latest commit

 

History

History
19 lines (15 loc) · 735 Bytes

README.md

File metadata and controls

19 lines (15 loc) · 735 Bytes

asr-commands

Scripts run on Kaldi toolkit on Kaggle's Tensorflow Speech Recognition Challenge. The URL to the challenge is here.

To run the script, you have to copy this folder to $kaldi_path/egs/
Check the symbolic links of utils, steps, conf, local folders if it does work.
The kernel describe this repository is on kaggle.

Step 1:

Create training and validation data:

python data_prep.py

Step 2:

Create evaluation data:

python eval_prep.py

Step 3:

Training data (running from stage 0):

. ./path.sh
./run.sh --stage 0