Spaces:
Sleeping
Sleeping
File size: 3,261 Bytes
4cd8562 4d88240 4cd8562 8caa19b 4d88240 4cd8562 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 |
import os
import gradio as gr
import whisper
from gtts import gTTS
import io
from groq import Groq
os.environ["GROQ_API_KEY"] = "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"
# Initialize the Groq client
client = Groq(api_key=os.environ.get("GROQ_API_KEY"))
# Load the Whisper model
model = whisper.load_model("base") # You can choose other models like "small", "medium", "large"
def process_audio(file_path):
try:
# Load the audio file
audio = whisper.load_audio(file_path)
# Transcribe the audio using Whisper
result = model.transcribe(audio)
text = result["text"]
# Generate a response using Groq
chat_completion = client.chat.completions.create(
messages=[{"role": "user", "content": text}],
model="llama3-8b-8192", # Replace with the correct model if necessary
)
# Access the response using dot notation
response_message = chat_completion.choices[0].message.content.strip()
# Convert the response text to speech
tts = gTTS(response_message)
response_audio_io = io.BytesIO()
tts.write_to_fp(response_audio_io) # Save the audio to the BytesIO object
response_audio_io.seek(0)
# Save audio to a file to ensure it's generated correctly
with open("response.mp3", "wb") as audio_file:
audio_file.write(response_audio_io.getvalue())
# Return the response text and the path to the saved audio file
return response_message, "response.mp3"
except Exception as e:
return f"An error occurred: {e}", None
# Define custom CSS for improved appearance
custom_css = """
.gradio-container {
background-color: #87ceeb; /* Sky blue background */
color: #333; /* Dark text color for contrast */
font-family: Arial, sans-serif; /* Better font */
text-align: center; /* Center align text */
}
.gradio-title {
color: #333; /* Title color */
font-size: 36px; /* Large font size */
font-weight: bold; /* Bold text */
margin-bottom: 20px; /* Space below the title */
}
.gradio-input, .gradio-output {
background-color: #e0f7fa; /* Slightly lighter background for inputs/outputs */
color: #333; /* Text color for inputs/outputs */
border: 1px solid #80deea; /* Border color */
border-radius: 5px; /* Rounded corners */
padding: 10px; /* Padding */
}
.gradio-button {
background-color: #4fc3f7; /* Button background color */
color: white; /* Button text color */
border: none; /* Remove default border */
border-radius: 5px; /* Rounded corners */
padding: 10px 20px; /* Padding */
cursor: pointer; /* Pointer cursor on hover */
}
.gradio-button:hover {
background-color: #29b6f6; /* Button hover color */
}
"""
# Define the Gradio interface with custom CSS
iface = gr.Interface(
fn=process_audio,
inputs=gr.Audio(type="filepath"), # Use type="filepath"
outputs=[gr.Textbox(label="Response Text"), gr.Audio(label="Response Audio")],
live=True,
title="Audio to Audio ChatBot", # Add title here
description="Convert audio input to text, generate a response, and convert it back to audio.",
css=custom_css # Apply custom CSS
)
iface.launch() |