Tacotron2_ParallelWaveGAN_Korean

Nvidia , kan-bayashi 코드를 사용하여 제작했습니다.

개인 도커 컨테이너에서 환경을 구축 후 실험을 진행했습니다.

Requirements

This repository is tested on Ubuntu 16.04 with a GPU Titan V.

Name		Name	Last commit message	Last commit date
Latest commit History 5 Commits
audio		audio
common		common
exports		exports
filelists		filelists
img		img
mel		mel
notebooks		notebooks
parallel_output		parallel_output
parallel_wavegan		parallel_wavegan
phrases		phrases
platform		platform
scripts		scripts
tacotron2		tacotron2
trt		trt
.gitignore		.gitignore
.gitmodules		.gitmodules
LICENCE		LICENCE
README.md		README.md
Untitled.ipynb		Untitled.ipynb
data_functions.py		data_functions.py
decode.py		decode.py
inference.py		inference.py
inference_perf.py		inference_perf.py
loss_functions.py		loss_functions.py
main.py		main.py
models.py		models.py
multiproc.py		multiproc.py
output_mel		output_mel
plotting_utils.py		plotting_utils.py
preprocess_audio2mel.py		preprocess_audio2mel.py
run_latency_tests.sh		run_latency_tests.sh
taco_cleaner.ipynb		taco_cleaner.ipynb
taco_infer.ipynb		taco_infer.ipynb
taco_k_log.txt		taco_k_log.txt
taco_k_log2.txt		taco_k_log2.txt
taco_train.ipynb		taco_train.ipynb
taco_train_w.ipynb		taco_train_w.ipynb
test.py		test.py
test_infer.py		test_infer.py
test_infer.sh		test_infer.sh
text.txt		text.txt
train.py		train.py
train_w.py		train_w.py
utils.py		utils.py