https://github.com/vsooda/mxnet_kaldi

use mxnet and kaldi to train asr model
https://github.com/vsooda/mxnet_kaldi

Last synced: 2 months ago
JSON representation

use mxnet and kaldi to train asr model

• Host: GitHub
• URL: https://github.com/vsooda/mxnet_kaldi
• Owner: vsooda
• Created: 2018-01-23T02:31:50.000Z (almost 7 years ago)
• Default Branch: master
• Last Pushed: 2018-02-01T07:01:18.000Z (almost 7 years ago)
• Last Synced: 2024-08-01T22:41:52.571Z (5 months ago)
• Language: Python
• Homepage:
• Size: 80.1 KB
• Stars: 6
• Watchers: 3
• Forks: 2
• Open Issues: 0
• Metadata Files:
  • Readme: README.md

Awesome Lists containing this project

• Awesome-MXNet - mxnet_kaldi

README

        
### note

This is the speech demo from mxnet, but for some [reason](https://github.com/apache/incubator-mxnet/pull/9060), they removed this demo. I had contributed timit recipe for this demo, and make it work quite well.

I will update this demo for the latest kaldi and mxnet. currently, we use a old kaldi. we will upgraded to latest in a few day.

thanks for the original author @pluskid and @yzhang87.

### Build Kaldi

Build Kaldi as **shared libraties** if you have not already done so.

```bash

cd kaldi/src

./configure --shared # and other options that you need

make depend

make

```

### Build Python Wrapper

1. Copy or link the attached `python_wrap` folder to `kaldi/src`.

2. Compile python_wrap/

```

cd kaldi/src/python_wrap/

make

```

### usage (simple version:timit)

1. run timit recipe in Kaldi. 

2  modify egs/timit/s5/global.cfg

3. `./run.sh`

every thing should be ok.