Spaces:
Running
Running
burtenshaw
commited on
Commit
·
6f66039
1
Parent(s):
d28bf05
add rtc configuration
Browse files
README.md
CHANGED
@@ -60,16 +60,6 @@ This launches a Gradio UI at http://localhost:7860 with:
|
|
60 |
- A chat interface showing the conversation
|
61 |
- An image display panel for generated images
|
62 |
|
63 |
-
### With FastAPI (Advanced)
|
64 |
-
|
65 |
-
```bash
|
66 |
-
python app.py
|
67 |
-
```
|
68 |
-
|
69 |
-
This serves a FastAPI app at http://localhost:7860 that provides:
|
70 |
-
- WebRTC-based audio communication
|
71 |
-
- The same smolagents functionality in a more flexible API
|
72 |
-
|
73 |
## How to Use
|
74 |
|
75 |
1. Click the microphone button to start streaming your voice.
|
|
|
60 |
- A chat interface showing the conversation
|
61 |
- An image display panel for generated images
|
62 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
63 |
## How to Use
|
64 |
|
65 |
1. Click the microphone button to start streaming your voice.
|
app.py
CHANGED
@@ -2,7 +2,13 @@ from pathlib import Path
|
|
2 |
from typing import List, Dict
|
3 |
|
4 |
from dotenv import load_dotenv
|
5 |
-
from fastrtc import
|
|
|
|
|
|
|
|
|
|
|
|
|
6 |
from smolagents import CodeAgent, HfApiModel, DuckDuckGoSearchTool
|
7 |
|
8 |
# Load environment variables
|
@@ -73,7 +79,7 @@ def process_response(audio):
|
|
73 |
|
74 |
# Convert response to audio using TTS model
|
75 |
for audio_chunk in tts_model.stream_tts_sync(response_content or ""):
|
76 |
-
# Yield the audio chunk
|
77 |
yield audio_chunk
|
78 |
|
79 |
|
@@ -86,7 +92,8 @@ stream = Stream(
|
|
86 |
"icon_button_color": "rgb(255, 255, 255)",
|
87 |
"title": "🧑💻The Coworking Agent",
|
88 |
},
|
|
|
89 |
)
|
90 |
|
91 |
if __name__ == "__main__":
|
92 |
-
stream.ui.launch()
|
|
|
2 |
from typing import List, Dict
|
3 |
|
4 |
from dotenv import load_dotenv
|
5 |
+
from fastrtc import (
|
6 |
+
get_stt_model,
|
7 |
+
get_tts_model,
|
8 |
+
Stream,
|
9 |
+
ReplyOnPause,
|
10 |
+
get_twilio_turn_credentials,
|
11 |
+
)
|
12 |
from smolagents import CodeAgent, HfApiModel, DuckDuckGoSearchTool
|
13 |
|
14 |
# Load environment variables
|
|
|
79 |
|
80 |
# Convert response to audio using TTS model
|
81 |
for audio_chunk in tts_model.stream_tts_sync(response_content or ""):
|
82 |
+
# Yield the audio chunk
|
83 |
yield audio_chunk
|
84 |
|
85 |
|
|
|
92 |
"icon_button_color": "rgb(255, 255, 255)",
|
93 |
"title": "🧑💻The Coworking Agent",
|
94 |
},
|
95 |
+
rtc_configuration=get_twilio_turn_credentials(),
|
96 |
)
|
97 |
|
98 |
if __name__ == "__main__":
|
99 |
+
stream.ui.launch(server_port=7860)
|