Multi-speaker sample wav (with wavernn vocoder)

begeekmyfriend / tacotron2

Forked from NVIDIA/tacotron2 and merged with Rayhane-mamah/Tacotron-2

BSD 3-Clause "New" or "Revised" License

81 stars 38 forks source link

Multi-speaker sample wav (with wavernn vocoder) #1

Open begeekmyfriend opened 5 years ago

begeekmyfriend commented 5 years ago

eval_multi_speaker_tacotron2_wavernn.zip

tugstugi commented 4 years ago

@begeekmyfriend sounds really nice. Each voice has its own WaveRNN model or is it a single WaveRNN model for all voices?

begeekmyfriend commented 4 years ago

Trained with multi speaker GTA mels together. By the way, in fact those samples produced by the WaveRNN are not the perfectly trained (only 400K steps) we can train more steps to improve the evaluation.