Spaces:
Sleeping
Sleeping
File size: 3,833 Bytes
614c1d4 cf5d3de 614c1d4 83c6221 3fcc7dc 614c1d4 3fcc7dc 614c1d4 83c6221 614c1d4 83c6221 614c1d4 cf5d3de 614c1d4 cf5d3de 614c1d4 29a5603 614c1d4 83c6221 614c1d4 83c6221 |
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 99 100 101 102 103 104 105 106 107 108 |
import streamlit as st
from transformers import pipeline
import whisper
from gtts import gTTS
import tempfile
import os
import logging
from pydub import AudioSegment
import openai
# Set up logging
logging.basicConfig(level=logging.INFO)
logger = logging.getLogger(__name__)
# Load Hugging Face model for text generation (instead of Google Cloud)
def load_hf_model():
# Load a model for heart health-related questions
return pipeline("text-generation", model="gpt2")
# Load Whisper model for transcription
def load_whisper_model():
return whisper.load_model("base")
# Function to generate response using Hugging Face model
def generate_hf_response(model, prompt):
result = model(prompt, max_length=100, num_return_sequences=1)
return result[0]["generated_text"]
# Function to process audio input using Whisper and Hugging Face
def process_audio(audio_file, hf_model, whisper_model):
try:
# Transcribe audio using Whisper
result = whisper_model.transcribe(audio_file)
user_text = result['text']
logger.info(f"Transcription successful: {user_text}")
except Exception as e:
logger.error(f"Error in transcribing audio: {e}")
return "Error in transcribing audio.", None
try:
# Generate response using Hugging Face model
response_text = generate_hf_response(hf_model, user_text)
logger.info(f"Generated response: {response_text}")
except Exception as e:
logger.error(f"Error in generating response: {e}")
return "Error in generating response.", None
try:
# Convert the response text to speech
tts = gTTS(text=response_text, lang='en')
audio_file = tempfile.NamedTemporaryFile(delete=False, suffix='.mp3')
tts.save(audio_file.name)
logger.info("Text-to-speech conversion successful.")
except Exception as e:
logger.error(f"Error in text-to-speech conversion: {e}")
return "Error in text-to-speech conversion.", None
return response_text, audio_file.name
# Main application layout
def main():
st.title("Heart Health & Audio Processing App 🫀🎙️ (Hugging Face Edition)")
# Load models
hf_model = load_hf_model()
whisper_model = load_whisper_model()
# Two tabs: one for the chatbot and one for audio processing
tab1, tab2 = st.tabs(["Heart Health Chatbot", "Audio Processing"])
# Tab 1: Heart Health Chatbot
with tab1:
st.header("Chat with Heart Health Specialist")
if "history" not in st.session_state:
st.session_state.history = []
user_input = st.text_input("Ask about heart health:", placeholder="Type here...")
if st.button("Send") and user_input:
bot_response = generate_hf_response(hf_model, user_input)
st.session_state.history.append({"role": "user", "content": user_input})
st.session_state.history.append({"role": "bot", "content": bot_response})
for chat in st.session_state.history:
if chat["role"] == "user":
st.write(f"**You:** {chat['content']}")
else:
st.write(f"**Bot:** {chat['content']}")
# Tab 2: Audio Processing
with tab2:
st.header("Audio Processing with Whisper and Hugging Face")
uploaded_audio = st.file_uploader("Upload an audio file for transcription and response", type=["mp3", "wav", "ogg"])
if uploaded_audio:
with st.spinner("Processing audio..."):
response_text, audio_file_path = process_audio(uploaded_audio, hf_model, whisper_model)
if response_text:
st.write(f"**Response:** {response_text}")
st.audio(audio_file_path)
# Run the app
if __name__ == "__main__":
main()
|