Spaces:
Running
on
T4
Running
on
T4
Update app.py
Browse files
app.py
CHANGED
|
@@ -16,18 +16,12 @@ class TTSWebUI:
|
|
| 16 |
tts_model_path=None,
|
| 17 |
vocoder_model_path=None,
|
| 18 |
embedding_gan_path=None,
|
| 19 |
-
available_artificial_voices=
|
| 20 |
):
|
| 21 |
path_to_iso_list = hf_hub_download(repo_id="Flux9665/ToucanTTS", filename="iso_to_fullname.json")
|
| 22 |
iso_to_name = load_json_from_path(path_to_iso_list)
|
| 23 |
text_selection = [f"{iso_to_name[iso_code]} ({iso_code})" for iso_code in iso_to_name]
|
| 24 |
# accent_selection = [f"{iso_to_name[iso_code]} Accent ({iso_code})" for iso_code in iso_to_name]
|
| 25 |
-
if tts_model_path is None:
|
| 26 |
-
tts_model_path = hf_hub_download(repo_id="Flux9665/ToucanTTS", filename="ToucanTTS.pt")
|
| 27 |
-
if vocoder_model_path is None:
|
| 28 |
-
vocoder_model_path = hf_hub_download(repo_id="Flux9665/ToucanTTS", filename="Vocoder.pt")
|
| 29 |
-
if embedding_gan_path is None:
|
| 30 |
-
embedding_gan_path = hf_hub_download(repo_id="Flux9665/ToucanTTS", filename="embedding_gan.pt")
|
| 31 |
|
| 32 |
self.controllable_ui = ControllableInterface(gpu_id=gpu_id,
|
| 33 |
available_artificial_voices=available_artificial_voices,
|
|
@@ -45,7 +39,7 @@ class TTSWebUI:
|
|
| 45 |
label="Select the Language of the Text (type on your keyboard to find it quickly)"),
|
| 46 |
gr.Slider(minimum=0.0, maximum=0.8, step=0.1, value=0.5, label="Prosody Creativity"),
|
| 47 |
gr.Slider(minimum=0.7, maximum=1.3, step=0.1, value=1.0, label="Faster - Slower"),
|
| 48 |
-
gr.Slider(minimum=0, maximum=available_artificial_voices, step=1, value=
|
| 49 |
gr.Slider(minimum=-10.0, maximum=10.0, step=0.1, value=0.0, label="Gender of artificial Voice"),
|
| 50 |
gr.Audio(type="filepath", show_label=True, container=True, label="[OPTIONAL] Voice to Clone (if left empty, will use an artificial voice instead)"),
|
| 51 |
# gr.Slider(minimum=0.5, maximum=1.5, step=0.1, value=1.0, label="Pitch Variance Scale"),
|
|
|
|
| 16 |
tts_model_path=None,
|
| 17 |
vocoder_model_path=None,
|
| 18 |
embedding_gan_path=None,
|
| 19 |
+
available_artificial_voices=10 # be careful with this, if you want too many, it might lead to an endless loop
|
| 20 |
):
|
| 21 |
path_to_iso_list = hf_hub_download(repo_id="Flux9665/ToucanTTS", filename="iso_to_fullname.json")
|
| 22 |
iso_to_name = load_json_from_path(path_to_iso_list)
|
| 23 |
text_selection = [f"{iso_to_name[iso_code]} ({iso_code})" for iso_code in iso_to_name]
|
| 24 |
# accent_selection = [f"{iso_to_name[iso_code]} Accent ({iso_code})" for iso_code in iso_to_name]
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 25 |
|
| 26 |
self.controllable_ui = ControllableInterface(gpu_id=gpu_id,
|
| 27 |
available_artificial_voices=available_artificial_voices,
|
|
|
|
| 39 |
label="Select the Language of the Text (type on your keyboard to find it quickly)"),
|
| 40 |
gr.Slider(minimum=0.0, maximum=0.8, step=0.1, value=0.5, label="Prosody Creativity"),
|
| 41 |
gr.Slider(minimum=0.7, maximum=1.3, step=0.1, value=1.0, label="Faster - Slower"),
|
| 42 |
+
gr.Slider(minimum=0, maximum=available_artificial_voices, step=1, value=5, label="Random Seed for the artificial Voice"),
|
| 43 |
gr.Slider(minimum=-10.0, maximum=10.0, step=0.1, value=0.0, label="Gender of artificial Voice"),
|
| 44 |
gr.Audio(type="filepath", show_label=True, container=True, label="[OPTIONAL] Voice to Clone (if left empty, will use an artificial voice instead)"),
|
| 45 |
# gr.Slider(minimum=0.5, maximum=1.5, step=0.1, value=1.0, label="Pitch Variance Scale"),
|