Update app.py

app.py

@@ -40,9 +40,9 @@ def create_tts_fn(model, hps, speaker_ids):
         speaker_id = speaker_ids[speaker]
         stn_tst = get_text(text, hps, is_phoneme)
         with no_grad():
-            x_tst = stn_tst.unsqueeze(0)
-            x_tst_lengths = LongTensor([stn_tst.size(0)])
-            sid = LongTensor([speaker_id])
+            x_tst = stn_tst.cuda().unsqueeze(0)
+            x_tst_lengths = LongTensor([stn_tst.size(0)]).cuda()
+            sid = LongTensor([speaker_id]).cuda()
             audio = model.infer(x_tst, x_tst_lengths, sid=sid, noise_scale=.667, noise_scale_w=0.8,
                                 length_scale=1.0 / speed)[0][0, 0].data.cpu().float().numpy()
         del stn_tst, x_tst, x_tst_lengths, sid
@@ -96,7 +96,7 @@ if __name__ == '__main__':
         hps.data.filter_length // 2 + 1,
         hps.train.segment_size // hps.data.hop_length,
         n_speakers=hps.data.n_speakers,
-        **hps.model)
+        **hps.model).cuda()
     utils.load_checkpoint(model_path, model, None)
     model.eval()
     speaker_ids = [sid for sid, name in enumerate(hps.speakers) if name != "None"]