Multi-speaker speech synthesis

We obtain synthesized speech from the following multi-speaker TTS model trained on the VCTK data set:
ClariNet model with 40 layer WaveNet vocoder (Multi-speaker ClariNet)
Deep Voice 2 model with 80 layer WaveNet vocoder (Deep Voice 2)
Deep Voice 3 model with 30 layer WaveNet vocoder (Deep Voice 3).

	Multi-speaker ClariNet	Deep Voice 2	Deep Voice 3
1: Prosecutors have opened a massive investigation into allegations of fixing games and illegal betting.
Speaker 1
Speaker 2
Speaker 3
Speaker 4
Speaker 5
Speaker 6
Speaker 7
Speaker 8
2: We can continue to strengthen the education of good lawyers.
Speaker 1
Speaker 2
Speaker 3
Speaker 4
Speaker 5
Speaker 6
Speaker 7
Speaker 8
3: Humans also judge distance by using the relative sizes of objects.
Speaker 1
Speaker 2
Speaker 3
Speaker 4
Speaker 5
Speaker 6
Speaker 7
Speaker 8