Spaces:
Runtime error
Runtime error
New file creation
Browse files
app.py
ADDED
@@ -0,0 +1,47 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
import gradio as gr
|
2 |
+
import transformers
|
3 |
+
import librosa
|
4 |
+
import numpy as np
|
5 |
+
|
6 |
+
# Load the model pipeline
|
7 |
+
pipe = transformers.pipeline(
|
8 |
+
model='fixie-ai/ultravox-v0_5-llama-3_1-8b',
|
9 |
+
trust_remote_code=True
|
10 |
+
)
|
11 |
+
|
12 |
+
def transcribe(audio):
|
13 |
+
if audio is None:
|
14 |
+
return "No audio provided."
|
15 |
+
|
16 |
+
# Load audio using librosa
|
17 |
+
audio_array, sr = librosa.load(audio, sr=16000)
|
18 |
+
|
19 |
+
# Define initial system prompt
|
20 |
+
turns = [
|
21 |
+
{
|
22 |
+
"role": "system",
|
23 |
+
"content": "You are a friendly and helpful character. You love to answer questions for people."
|
24 |
+
},
|
25 |
+
]
|
26 |
+
|
27 |
+
# Run inference
|
28 |
+
result = pipe(
|
29 |
+
{'audio': audio_array, 'turns': turns, 'sampling_rate': sr},
|
30 |
+
max_new_tokens=30
|
31 |
+
)
|
32 |
+
|
33 |
+
# Return result content
|
34 |
+
return result[0]['content'] if isinstance(result, list) else str(result)
|
35 |
+
|
36 |
+
# Build Gradio Interface
|
37 |
+
demo = gr.Interface(
|
38 |
+
fn=transcribe,
|
39 |
+
inputs=gr.Audio(source="upload", type="filepath", label="Upload an Audio File"),
|
40 |
+
outputs=gr.Textbox(label="Ultravox Response"),
|
41 |
+
title="🎙️ Ultravox AI Voicebot",
|
42 |
+
description="Upload an audio file and Ultravox will respond intelligently!"
|
43 |
+
)
|
44 |
+
|
45 |
+
# Launch app
|
46 |
+
if __name__ == "__main__":
|
47 |
+
demo.launch()
|