MycroftAI
/
mimic2
mirror of https://github.com/MycroftAI/mimic2.git
1
0
Fork 0
Code Issues Projects Releases Wiki Activity

Update README.md

pull/15/head
Michael Nguyen 2018-07-06 19:02:51 -05:00 committed by GitHub
parent ccd3207a09
commit 4eac585dca
No known key found for this signature in database
GPG Key ID: 4AEE18F83AFDEB23
1 changed files with 8 additions and 7 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

15
README.md
View File

@ -4,10 +4,9 @@ This is a fork of [keithito/tacotron](https://github.com/keithito/tacotron)
with changes specific to Mimic 2 applied.
## Background
Earlier this year, Google published a paper, [Tacotron: A Fully End-to-End Text-To-Speech Synthesis Model](https://arxiv.org/pdf/1703.10135.pdf),
Google published a paper, [Tacotron: A Fully End-to-End Text-To-Speech Synthesis Model](https://arxiv.org/pdf/1703.10135.pdf),
where they present a neural text-to-speech model that learns to synthesize speech directly from
(text, audio) pairs. However, they didn't release their source code or training data. This is an
attempt to provide an open-source implementation of the model described in their paper.
@ -16,7 +15,6 @@ The quality isn't as good as Google's demo yet, but hopefully it will get there
Pull requests are welcome!
## Quick Start
### Installing dependencies
@ -51,7 +49,7 @@ Pull requests are welcome!
### Using a pre-trained model
**NOTE** this model will only work if you switch out the LocationSensitiveAttention layer for the BahdanauAttention layer in tacotron.py
**NOTE this model will only work if you switch out the LocationSensitiveAttention layer for the BahdanauAttention layer in tacotron.py
1. **Download and unpack a model**:
```
@ -63,7 +61,7 @@ Pull requests are welcome!
python3 demo_server.py --checkpoint /tmp/tacotron-20170720/model.ckpt
```
3. **Point your browser at localhost:9000**
3. **Point your browser at localhost:3000**
* Type what you want to synthesize
@ -77,7 +75,7 @@ Pull requests are welcome!
The following are supported out of the box:
* [LJ Speech](https://keithito.com/LJ-Speech-Dataset/) (Public Domain)
* [Blizzard 2012](http://www.cstr.ed.ac.uk/projects/blizzard/2012/phase_one) (Creative Commons Attribution Share-Alike)
* [M-ailabs](http://www.m-ailabs.bayern/en/the-mailabs-speech-dataset/)
You can use other datasets if you convert them to the right format. See [TRAINING_DATA.md](TRAINING_DATA.md) for more info.
@ -104,12 +102,15 @@ Pull requests are welcome!
|- lab
|- wav
```
For M-AILABS follow the directory structure from [here](http://www.m-ailabs.bayern/en/the-mailabs-speech-dataset/)
3. **Preprocess the data**
```
python3 preprocess.py --dataset ljspeech
```
* Use `--dataset blizzard` for Blizzard data
* other datasets can be used i.e. `--dataset blizzard` for Blizzard data
* for the mailabs dataset, do `preprocess.py --help` for options. Also note that mailabs uses sample_size of 16000
4. **Train a model**
```