Skip to content
Projects
Groups
Snippets
Help
Loading...
Help
Submit feedback
Contribute to GitLab
Sign in / Register
Toggle navigation
S
speech-to-text
Project
Project
Details
Activity
Releases
Cycle Analytics
Repository
Repository
Files
Commits
Branches
Tags
Contributors
Graph
Compare
Charts
Issues
0
Issues
0
List
Board
Labels
Milestones
Merge Requests
0
Merge Requests
0
CI / CD
CI / CD
Pipelines
Jobs
Schedules
Charts
Wiki
Wiki
Snippets
Snippets
Members
Members
Collapse sidebar
Close sidebar
Activity
Graph
Charts
Create a new issue
Jobs
Commits
Issue Boards
Open sidebar
servicios-cognitivos
speech-to-text
Commits
71b551d1
Commit
71b551d1
authored
May 08, 2019
by
cristian Quezada
Browse files
Options
Browse Files
Download
Email Patches
Plain Diff
Agregado log
parent
4f9592e7
Changes
2
Hide whitespace changes
Inline
Side-by-side
Showing
2 changed files
with
50 additions
and
23 deletions
+50
-23
README.md
README.md
+2
-23
log_ubuntu16_04.txt
log_ubuntu16_04.txt
+48
-0
No files found.
README.md
View file @
71b551d1
...
...
@@ -5,11 +5,7 @@ git : [DeepSpeech](https://github.com/mozilla/DeepSpeech)
*
Python 3.5 o 3.6
*
Linux o Mac
Algunos comandos:
```
sudo apt-get install build-essential libboost-all-dev cmake zlib1g-dev libbz2-dev liblzma-dev
sudo apt-get install sox libsox-fmt-mp3
```
Seguir los pasos en log_ubuntu16_04.txt primero, para tener listo las dependencias.
## Uso con modelos entrenados:
...
...
@@ -26,26 +22,9 @@ deepspeech --model models/output_graph.pbmm --alphabet models/alphabet.txt --lm
```
# Entrenar un modelo propio:
*
Clonar la repo
`git clone https://github.com/mozilla/DeepSpeech.git`
## Requisitos :
numpy == 1.15.4
tensorflow == 1.13.1
progressbar2
pandas
six
pyxdg
attrdict
sox
bs4
requests
librosa
soundfile
# Miscellaneous scripts
paramiko >= 2.1
scipy
matplotlib
seguir los pasos en log_ubuntu16_04.txt
## Generar archivos de entrenamiento
...
...
log_ubuntu16_04.txt
0 → 100644
View file @
71b551d1
log
sudo apt install python3-pip
pip3 install deepspeech
sudo apt install git
git clone http://192.168.21.51/servicios-cognitivos/speech-to-text.git
# Situarse dentro de la carpeta speech-to-text
# Probando modelo en inglés
deepspeech --model models/output_graph.pbmm --alphabet models/alphabet.txt --lm models/lm.binary --trie models/trie --audio LDC93S1.wav
Resultado : she had a duck suit and greasy wash or all year
deepspeech --model models/output_graph.pbmm --alphabet models/alphabet.txt --lm models/lm.binary --trie models/trie --audio nnn.wav
Resultado : a divine flame it is the burning fire of la
## TRaining
* Dependencias
pip install tensorflow==1.13.1
sudo apt-get install sox libsox-fmt-mp3
pip3 install progressbar2
pip3 install pandas
pip3 install six
pip3 install pyxdg
pip3 install attrdict
pip3 install sox
pip3 install bs4
pip3 install requests
pip3 install librosa
pip3 install soundfile
sudo apt-get install build-essential libboost-all-dev cmake zlib1g-dev libbz2-dev liblzma-dev
sudo apt-get install libffi-dev libssl-dev libxml2-dev libxslt1-dev libjpeg8-dev
pip3 install paramiko
pip3 install scipy
pip3 install matplotlib
* Modelo de Lenguaje
//situarse en una nueva carpeta
wget -O - https://kheafield.com/code/kenlm.tar.gz |tar xz
mkdir kenlm/build
cd kenlm/build
sudo apt install libeigen3-dev
cmake ..
make -j2
// los archivos binarios deben estar en kenlm/build/bin ,usar lmplz y build_binary
* Generar trie
//situarse en una nueva carpeta
git clone https://github.com/mozilla/DeepSpeech.git
cd DeepSpeech/
python3 util/taskcluster.py --target ./native_client
// esto generará el archivo ejecutable "generate_trie" en la carpeta DeepSpeech/native_client
./generate_trie
Write
Preview
Markdown
is supported
0%
Try again
or
attach a new file
Attach a file
Cancel
You are about to add
0
people
to the discussion. Proceed with caution.
Finish editing this message first!
Cancel
Please
register
or
sign in
to comment