Spaces:
Runtime error
Runtime error
init
Browse files- app.py +25 -0
- requirements.txt +4 -0
app.py
ADDED
@@ -0,0 +1,25 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
import gradio as gr
|
2 |
+
import numpy as np
|
3 |
+
import scipy.io.wavfile
|
4 |
+
import torch
|
5 |
+
import torch.nn.functional as F
|
6 |
+
from whisperspeech.pipeline import Pipeline
|
7 |
+
|
8 |
+
def process_audio(audio_elem):
|
9 |
+
|
10 |
+
scipy.io.wavfile.write('test.mp3', 48000, audio_elem[1])
|
11 |
+
|
12 |
+
# print out details about ut
|
13 |
+
pipe = Pipeline(s2a_ref='collabora/whisperspeech:s2a-q4-base-en+pl.model')
|
14 |
+
# save audio_elem as a file
|
15 |
+
speaker = pipe.extract_spk_emb("test.mp3")
|
16 |
+
speaker = speaker.cpu().numpy() # Move tensor from GPU to CPU and convert to numpy array
|
17 |
+
print(speaker)
|
18 |
+
#save it locally
|
19 |
+
np.savez_compressed("speaker", features=speaker)
|
20 |
+
return "speaker.npz"
|
21 |
+
|
22 |
+
|
23 |
+
# Define Gradio interface
|
24 |
+
with gr.Interface(fn=process_audio, inputs="audio", outputs="file") as iface:
|
25 |
+
iface.launch()
|
requirements.txt
ADDED
@@ -0,0 +1,4 @@
|
|
|
|
|
|
|
|
|
|
|
1 |
+
python ==3.10
|
2 |
+
WhisperSpeech==0.8
|
3 |
+
torch==2.0.1
|
4 |
+
gradio
|