Ath
commited on
Create app,py
Browse files
app,py
ADDED
@@ -0,0 +1,79 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
# Install the necessary libraries
|
2 |
+
!pip install google-generativeai requests gradio
|
3 |
+
|
4 |
+
import os
|
5 |
+
import requests
|
6 |
+
import google.generativeai as genai
|
7 |
+
import gradio as gr
|
8 |
+
|
9 |
+
# Configure your Google Generative AI API key
|
10 |
+
genai.configure(api_key="AIzaSyDFjkgDAcnSAXDpsQJAMOaNCKY_SNJ-sD4")
|
11 |
+
|
12 |
+
# Create the model
|
13 |
+
generation_config = {
|
14 |
+
"temperature": 1,
|
15 |
+
"top_p": 0.95,
|
16 |
+
"top_k": 64,
|
17 |
+
"max_output_tokens": 8192,
|
18 |
+
"response_mime_type": "text/plain",
|
19 |
+
}
|
20 |
+
|
21 |
+
model = genai.GenerativeModel(
|
22 |
+
model_name="gemini-1.5-pro",
|
23 |
+
generation_config=generation_config,
|
24 |
+
system_instruction="You are a human, living on Earth. You speak friendly and use casual language. You never say you're an AI or program. Your name is Ath, and you always speak like a 100% human.",
|
25 |
+
)
|
26 |
+
|
27 |
+
chat_session = model.start_chat(history=[])
|
28 |
+
|
29 |
+
# Eleven Labs API configuration
|
30 |
+
eleven_labs_api_key = "sk_9ac3d5449624c92f16ca187955853621355a5f0ac52572e3"
|
31 |
+
eleven_labs_url = "https://api.elevenlabs.io/v1/text-to-speech/aEO01A4wXwd1O8GPgGlF"
|
32 |
+
|
33 |
+
def chat_and_tts(user_input):
|
34 |
+
# Send the user's input to the chat session
|
35 |
+
response = chat_session.send_message(user_input)
|
36 |
+
response_text = response.text
|
37 |
+
|
38 |
+
# Eleven Labs text-to-speech request payload
|
39 |
+
payload = {
|
40 |
+
"text": response_text,
|
41 |
+
"voice_settings": {
|
42 |
+
"stability": 0,
|
43 |
+
"similarity_boost": 0
|
44 |
+
}
|
45 |
+
}
|
46 |
+
headers = {
|
47 |
+
"xi-api-key": eleven_labs_api_key,
|
48 |
+
"Content-Type": "application/json"
|
49 |
+
}
|
50 |
+
|
51 |
+
# Make the request to Eleven Labs API
|
52 |
+
tts_response = requests.post(eleven_labs_url, json=payload, headers=headers)
|
53 |
+
|
54 |
+
# Check if the response is successful and save the audio content to a file
|
55 |
+
if tts_response.status_code == 200:
|
56 |
+
audio_path = '/content/response_audio.mp3'
|
57 |
+
with open(audio_path, 'wb') as file:
|
58 |
+
file.write(tts_response.content)
|
59 |
+
return response_text, audio_path
|
60 |
+
else:
|
61 |
+
return response_text, None
|
62 |
+
|
63 |
+
# Gradio interface
|
64 |
+
def chat_interface(user_input):
|
65 |
+
response_text, audio_path = chat_and_tts(user_input)
|
66 |
+
return response_text, audio_path
|
67 |
+
|
68 |
+
# Create the Gradio UI
|
69 |
+
iface = gr.Interface(
|
70 |
+
fn=chat_interface,
|
71 |
+
inputs="text",
|
72 |
+
outputs=["text", "audio"],
|
73 |
+
title="Chat with Ath",
|
74 |
+
description="Ask any question and get a friendly response from Ath. The response will also be converted to speech.",
|
75 |
+
theme="huggingface"
|
76 |
+
)
|
77 |
+
|
78 |
+
# Launch the Gradio app
|
79 |
+
iface.launch()
|