mirror of https://github.com/coqui-ai/TTS.git
bugfix in tacotron and tacotron 2 gst inference
This commit is contained in:
parent
d038a00602
commit
0fa2544fa5
|
@ -158,7 +158,7 @@ if __name__ == "__main__":
|
||||||
# check if gst_style string is a dict, if is dict convert else use string
|
# check if gst_style string is a dict, if is dict convert else use string
|
||||||
try:
|
try:
|
||||||
gst_style = json.loads(args.gst_style)
|
gst_style = json.loads(args.gst_style)
|
||||||
except:
|
except ValueError:
|
||||||
gst_style = args.gst_style
|
gst_style = args.gst_style
|
||||||
|
|
||||||
wav = tts(model, vocoder_model, args.text, C, args.use_cuda, ap, use_griffin_lim, args.speaker_fileid, speaker_embedding=speaker_embedding, gst_style=gst_style)
|
wav = tts(model, vocoder_model, args.text, C, args.use_cuda, ap, use_griffin_lim, args.speaker_fileid, speaker_embedding=speaker_embedding, gst_style=gst_style)
|
||||||
|
|
|
@ -40,7 +40,7 @@ class Tacotron(TacotronAbstract):
|
||||||
forward_attn, trans_agent, forward_attn_mask,
|
forward_attn, trans_agent, forward_attn_mask,
|
||||||
location_attn, attn_K, separate_stopnet,
|
location_attn, attn_K, separate_stopnet,
|
||||||
bidirectional_decoder, double_decoder_consistency,
|
bidirectional_decoder, double_decoder_consistency,
|
||||||
ddc_r, gst)
|
ddc_r, gst, gst_embedding_dim, gst_num_heads, gst_style_tokens)
|
||||||
|
|
||||||
# init layer dims
|
# init layer dims
|
||||||
decoder_in_features = 256
|
decoder_in_features = 256
|
||||||
|
|
|
@ -45,7 +45,7 @@ class Tacotron2(TacotronAbstract):
|
||||||
forward_attn, trans_agent, forward_attn_mask,
|
forward_attn, trans_agent, forward_attn_mask,
|
||||||
location_attn, attn_K, separate_stopnet,
|
location_attn, attn_K, separate_stopnet,
|
||||||
bidirectional_decoder, double_decoder_consistency,
|
bidirectional_decoder, double_decoder_consistency,
|
||||||
ddc_r, gst)
|
ddc_r, gst, gst_embedding_dim, gst_num_heads, gst_style_tokens)
|
||||||
|
|
||||||
# init layer dims
|
# init layer dims
|
||||||
decoder_in_features = 512
|
decoder_in_features = 512
|
||||||
|
|
Loading…
Reference in New Issue