Update xtts_train.py
Browse files- xtts_train.py +6 -0
xtts_train.py
CHANGED
@@ -44,6 +44,12 @@ def run_tts(lang, tts_text, speaker_audio_file):
|
|
44 |
if XTTS_MODEL is None or not speaker_audio_file:
|
45 |
return "You need to run the previous step to load the model !!", None, None
|
46 |
|
|
|
|
|
|
|
|
|
|
|
|
|
47 |
gpt_cond_latent, speaker_embedding = XTTS_MODEL.get_conditioning_latents(audio_path=speaker_audio_file, gpt_cond_len=XTTS_MODEL.config.gpt_cond_len, max_ref_length=XTTS_MODEL.config.max_ref_len, sound_norm_refs=XTTS_MODEL.config.sound_norm_refs)
|
48 |
out = XTTS_MODEL.inference(
|
49 |
text=tts_text,
|
|
|
44 |
if XTTS_MODEL is None or not speaker_audio_file:
|
45 |
return "You need to run the previous step to load the model !!", None, None
|
46 |
|
47 |
+
# Access the list element
|
48 |
+
speaker_audio_file = speaker_audio_file[0]
|
49 |
+
|
50 |
+
# Remove brackets and quotes
|
51 |
+
speaker_audio_file = speaker_audio_file.replace("[", "").replace("]", "").replace("'", "")
|
52 |
+
|
53 |
gpt_cond_latent, speaker_embedding = XTTS_MODEL.get_conditioning_latents(audio_path=speaker_audio_file, gpt_cond_len=XTTS_MODEL.config.gpt_cond_len, max_ref_length=XTTS_MODEL.config.max_ref_len, sound_norm_refs=XTTS_MODEL.config.sound_norm_refs)
|
54 |
out = XTTS_MODEL.inference(
|
55 |
text=tts_text,
|