Easy-to-use Speech Toolkit including SOTA/Streaming ASR with punctuation, influential TTS with text frontend, Speaker Verification System and End-to-End Speech Simultaneous Translation.

tts conformer speech-translation voice-recognition sound-classification transformer asr speech-synthesis voice-cloning punctuation-restoration streaming-tts speech-recognition vocoder kws streaming-asr speech-alignment

Go to file

Xinghai Sun a4c2dd7de2 Add function docs and comments to demo_server.py and demo_client.py.		7 years ago
data_utils	Add warming-up to demo_server.py for DS2 and clean codes.	7 years ago
datasets	change the wget method in run.sh of deep_speech2	7 years ago
lm	upload the language model	7 years ago
tests	update unittest with comments	7 years ago
README.md	add the requirement for cuDNN version in README	7 years ago
compute_mean_std.py	add mfcc feature for DS2	7 years ago
decoder.py	Add function docs for layer.py and model.py and update other details.	7 years ago
demo_client.py	Add function docs and comments to demo_server.py and demo_client.py.	7 years ago
demo_server.py	Add function docs and comments to demo_server.py and demo_client.py.	7 years ago
error_rate.py	Follow comments.	7 years ago
evaluate.py	Simplify train.py, evaluate.py, infer.py and tune.py by adding DeepSpeech2Model class.	7 years ago
infer.py	Add a realtime ASR demo for users to test their own voice with mic.	7 years ago
layer.py	Add function docs for layer.py and model.py and update other details.	7 years ago
model.py	Add a realtime ASR demo for users to test their own voice with mic.	7 years ago
requirements.txt	Add a realtime ASR demo for users to test their own voice with mic.	7 years ago
setup.sh	Add function docs for layer.py and model.py and update other details.	7 years ago
train.py	Add function docs for layer.py and model.py and update other details.	7 years ago
tune.py	Add function docs for layer.py and model.py and update other details.	7 years ago
utils.py	Add shuffle type of instance_shuffle and batch_shuffle_clipped.	7 years ago

README.md

Deep Speech 2 on PaddlePaddle

Installation

Prerequisites

Python = 2.7 only supported;
cuDNN >= 6.0 is required to utilize NVIDIA GPU platform in the installation of PaddlePaddle, and the CUDA toolkit with proper version suitable for cuDNN. The cuDNN library below 6.0 is found to yield a fatal error in batch normalization when handling utterances with long duration in inference.

Setup

sh setup.sh
export LD_LIBRARY_PATH=$PADDLE_INSTALL_DIR/Paddle/third_party/install/warpctc/lib:$LD_LIBRARY_PATH

Please replace $PADDLE_INSTALL_DIR with your own paddle installation directory.

Usage

Preparing Data

cd datasets
sh run_all.sh
cd ..

sh run_all.sh prepares all ASR datasets (currently, only LibriSpeech available). After running, we have several summarization manifest files in json-format.

A manifest file summarizes a speech data set, with each line containing the meta data (i.e. audio filepath, transcript text, audio duration) of each audio file within the data set, in json format. Manifest file serves as an interface informing our system of where and what to read the speech samples.

More help for arguments:

python datasets/librispeech/librispeech.py --help

Preparing for Training

python compute_mean_std.py

It will compute mean and stdandard deviation for audio features, and save them to a file with a default name ./mean_std.npz. This file will be used in both training and inferencing. The default feature of audio data is power spectrum, and the mfcc feature is also supported. To train and infer based on mfcc feature, please generate this file by

python compute_mean_std.py --specgram_type mfcc

and specify --specgram_type mfcc when running train.py, infer.py, evaluator.py or tune.py.

More help for arguments:

python compute_mean_std.py --help

Training

For GPU Training:

CUDA_VISIBLE_DEVICES=0,1,2,3,4,5,6,7 python train.py

For CPU Training:

python train.py --use_gpu False

More help for arguments:

python train.py --help

Preparing language model

The following steps, inference, parameters tuning and evaluating, will require a language model during decoding. A compressed language model is provided and can be accessed by

cd ./lm
sh run.sh
cd ..

Inference

For GPU inference

CUDA_VISIBLE_DEVICES=0 python infer.py

For CPU inference

python infer.py --use_gpu=False

More help for arguments:

python infer.py --help

Evaluating

CUDA_VISIBLE_DEVICES=0 python evaluate.py

More help for arguments:

python evaluate.py --help

Parameters tuning

Usually, the parameters \alpha and \beta for the CTC prefix beam search decoder need to be tuned after retraining the acoustic model.

For GPU tuning

CUDA_VISIBLE_DEVICES=0 python tune.py

For CPU tuning

python tune.py --use_gpu=False

More help for arguments:

python tune.py --help

Then reset parameters with the tuning result before inference or evaluating.