Multitasking English-to-phoneme and TTS

v1.2.0

Bump version up (#55)

Training English ASR with more data

• Use LibriSpeech train-clean-360 instead of train-clean-100.
• Added audio augmentation with changing amplitude.

Multitasking English-to-phoneme and TTS

• This model generates audio and CMU phonemes from English text.

CMU phone model

• New CMU phone model added. Text data is encoded in CMU phonemes by g2p_en

New English ASR model

v1.0.1

Update English TTS audio pre-trained model (#46)

Initial release

v1.0

Bump version (#29)

Bigger ASR model

v0.8

Update pre-trained model. (#26)

Japanese TTS added

v0.7

Bump version (#24)

TTS with alignment model

v0.6

Bump version

LSTM alignment model

v0.5

LSTM alignment model (#8)