Spaces:
Running
on
Zero
Running
on
Zero
Commit
·
156f2a1
1
Parent(s):
f522903
minor fixes
Browse files
app.py
CHANGED
@@ -6,6 +6,7 @@ import soundfile as sf
|
|
6 |
import torch
|
7 |
import traceback
|
8 |
import random
|
|
|
9 |
import spaces
|
10 |
|
11 |
import sys
|
@@ -98,8 +99,8 @@ def main(text_prompt, reference_paths, speed, denoise, avg_style, stabilize):
|
|
98 |
styles = model.get_styles(speaker, denoise, avg_style)
|
99 |
r = model.generate(phonemes, styles, stabilize, 18)
|
100 |
|
101 |
-
|
102 |
-
return
|
103 |
|
104 |
except Exception as e:
|
105 |
error_message = traceback.format_exc()
|
@@ -127,9 +128,9 @@ with gr.Blocks() as demo:
|
|
127 |
with gr.Row(equal_height=True):
|
128 |
with gr.Column():
|
129 |
speed = gr.Slider(0.0, 2.0, step=0.1, value=1.0, label="Speed")
|
130 |
-
denoise = gr.Slider(0.0, 1.0, step=0.1, value=0.
|
131 |
-
avg_style = gr.Checkbox(label="Use Average Styles", value=
|
132 |
-
stabilize = gr.Checkbox(label="Stabilize Speaking Speed", value=
|
133 |
|
134 |
text_prompt = gr.Textbox(label="Text Prompt", placeholder="Enter your text here...", lines=10, max_lines=10)
|
135 |
|
@@ -138,13 +139,15 @@ with gr.Blocks() as demo:
|
|
138 |
|
139 |
with gr.Column():
|
140 |
reference_audios = gr.Audio(label="Reference Audios", type='filepath')
|
141 |
-
synthesized_audio = gr.Audio(label="Generate Audio", type=
|
142 |
|
143 |
example_voices = gr.Dropdown(
|
144 |
label="Example voices",
|
145 |
choices=voice_choices,
|
146 |
-
value=
|
147 |
interactive=True,
|
|
|
|
|
148 |
)
|
149 |
|
150 |
with gr.Row(equal_height=False):
|
|
|
6 |
import torch
|
7 |
import traceback
|
8 |
import random
|
9 |
+
import numpy as np
|
10 |
import spaces
|
11 |
|
12 |
import sys
|
|
|
99 |
styles = model.get_styles(speaker, denoise, avg_style)
|
100 |
r = model.generate(phonemes, styles, stabilize, 18)
|
101 |
|
102 |
+
r = r / np.max(np.abs(r)) #Normalize
|
103 |
+
return (24000, r), "Audio generated successfully!"
|
104 |
|
105 |
except Exception as e:
|
106 |
error_message = traceback.format_exc()
|
|
|
128 |
with gr.Row(equal_height=True):
|
129 |
with gr.Column():
|
130 |
speed = gr.Slider(0.0, 2.0, step=0.1, value=1.0, label="Speed")
|
131 |
+
denoise = gr.Slider(0.0, 1.0, step=0.1, value=0.2, label="Denoise Strength")
|
132 |
+
avg_style = gr.Checkbox(label="Use Average Styles", value=True)
|
133 |
+
stabilize = gr.Checkbox(label="Stabilize Speaking Speed", value=True)
|
134 |
|
135 |
text_prompt = gr.Textbox(label="Text Prompt", placeholder="Enter your text here...", lines=10, max_lines=10)
|
136 |
|
|
|
139 |
|
140 |
with gr.Column():
|
141 |
reference_audios = gr.Audio(label="Reference Audios", type='filepath')
|
142 |
+
synthesized_audio = gr.Audio(label="Generate Audio", type='numpy')
|
143 |
|
144 |
example_voices = gr.Dropdown(
|
145 |
label="Example voices",
|
146 |
choices=voice_choices,
|
147 |
+
value=None,
|
148 |
interactive=True,
|
149 |
+
allow_custom_value=False,
|
150 |
+
filterable=False
|
151 |
)
|
152 |
|
153 |
with gr.Row(equal_height=False):
|