Some confusion in your visualization TensorBoard about fastspeech2

[ming024]

I am not sure of the meaning of the sentence "Are your models really output a speech or not?"
The output of my FastSpeech2 implementation is a mel-spectrogram, which can be conversed to wavfiles by vocoders such as WaveGlow and MelGAN.

If you are asking that why there is a large gap between the training and validation mel_loss and mel_postnet_loss curves, it is because that in evaluate.py the model synthesizes mel-spectrograms without ground-truth F0 and energy labels.

FastSpeech2/evaluate.py

Lines 63 to 64 in 172d2ea

	mel_output, mel_postnet_output, duration_output, f0_output, energy_output = model(
	text, src_pos, mel_pos, max_len, D)

[ming024]

closed #4