xinkez / lpctron-tts-cpp

C++ implementation of End to End TTS which combines both Tacatron2 and LPCNET Vocoder.

This repo contains c/c++ version of Tacotron2 and LPCNet .

Prerequisite

Tensorflow bazel build

Merge Model( as github has limit for 100mb)

cd model bash split_merge - > generates inference_model.pb

Build and Run

1. git clone the repo
2. mkdir build && cd build
3. make
4. ./lpctron_cc sample.txt
5. ffplay -f s16le -ar 16k -ac 1 output.pcm

Requirment

• TensorFlow r1.8+
• Ubuntu 16.04
• C++ compiler + cmake

Compiling Tensorflow 1.13

COmpilation of this projects requires compiling tensorflow from source as it uses some headers files that are generated during compilation of tensorflow ( bazel compilation)

• Install Bazel 0.24

wget https://github.com/bazelbuild/bazel/releases/download/0.24.0/bazel-0.24.0-installer-linux-x86_64.sh
bash bazel-0.24.0-installer-linux-x86_64.sh

• Install Tensorflow

git clone https://github.com/tensorflow/tensorflow.git
cd tensorflow
git checkout r1.13
./configure ( i used CPU version)
bazel build --config=opt //tensorflow/tools/pip_package:build_pip_package
bazel build //tensorflow:libtensorflow_cc.so
bazel build //tensorflow:libtensorflow.so

More optimizaed build for the x86 platform without cuda
bazel build --config=opt --copt=-mavx --copt=-mavx2 --copt=-mfma --copt=-msse4.2 //tensorflow:libtensorflow_cc.so
bazel build --config=opt --copt=-mavx --copt=-mavx2 --copt=-mfma --copt=-msse4.2 //tensorflow:libtensorflow.so

• Native Compilation of TensorFlow on Aarch64( PX2)

sudo apt-get install openjdk-8-jdk
sudo apt-get install pkg-config zip g++ zlib1g-dev unzip
wget https://github.com/bazelbuild/bazel/releases/download/0.21.0/bazel-0.21.0-dist.zip
unzip 
./compile.sh
sudo cp output/bazel /usr/local/bin

git clone https://github.com/tensorflow/tensorflow.git
cd tensorflow
git checkout r1.13
./configure ( i used CPU version)

bazel build -c opt --jobs=4 --verbose_failures  --copt="-funsafe-math-optimizations" --copt="-O3" --copt="-ftree-vectorize" --copt="-fomit-frame-pointer" //tensorflow:libtensorflow_cc.so

• Note : if you abort or build failure error , increase swap space . ( OOM killer message in dmesg)

• ABSL library

cd /usr/local/inclue
git clone https://github.com/abseil/abseil-cpp.git
mv abseil-cpp/absl .

Tensorflow compilation Links

https://gist.github.com/EKami/9869ae6347f68c592c5b5cd181a3b205

http://wangkejie.me/2018/03/01/tensorflow-cplusplus-installation/

https://tuanphuc.github.io/standalone-tensorflow-cpp/

https://gist.github.com/EKami/9869ae6347f68c592c5b5cd181a3b205 ( For ARM Linux)

https://www.pytorials.com/how-to-install-tensorflow-gpu-with-cuda-10-0-for-python-on-ubuntu/2/

https://medium.com/@sometimescasey/building-tensorflow-from-source-for-sse-avx-fma-instructions-worth-the-effort-fbda4e30eec3

The major revision of the Android NDK referenced by android_ndk_repository rule 'androidndk' is 19. The major revisions supported by Bazel are [10, 11, 12, 13, 14, 15, 16, 17, 18]. Bazel will attempt to treat the NDK as if it was r18. This may cause compilation and linkage problems. Please download a supported NDK version.