app.py

import os
import torch
import torchaudio
import librosa
import streamlit as st
from huggingface_hub import login
from transformers import AutoProcessor, AutoModelForCTC
import numpy as np
# ================================
# 1️⃣ Authenticate with Hugging Face Hub (Securely)
# ================================
HF_TOKEN = os.getenv("hf_token")  

if HF_TOKEN is None:
    raise ValueError("❌ Hugging Face API token not found. Please set it in Secrets.")

login(token=HF_TOKEN)

# ================================
# 2️⃣ Load Conformer Model & Processor
# ================================
MODEL_NAME = "deepl-project/conformer-finetunning"
processor = AutoProcessor.from_pretrained(MODEL_NAME)
model = AutoModelForCTC.from_pretrained(MODEL_NAME)

device = "cuda" if torch.cuda.is_available() else "cpu"
model.to(device)
print(f"✅ Conformer Model loaded on {device}")

# ================================
# 3️⃣ Streamlit UI: Fine-Tuning Hyperparameter Selection
# ================================
st.sidebar.title("🔧 Fine-Tuning Hyperparameters")
num_epochs = st.sidebar.slider("Epochs", min_value=1, max_value=10, value=3)
learning_rate = st.sidebar.select_slider("Learning Rate", options=[5e-4, 1e-4, 5e-5, 1e-5], value=5e-5)
batch_size = st.sidebar.select_slider("Batch Size", options=[2, 4, 8, 16], value=8)
attack_strength = st.sidebar.slider("Attack Strength", 0.0, 0.9, 0.1)

# ================================
# 4️⃣ Streamlit ASR Web App (Fast Decoding & Security Features)
# ================================
st.title("🎙️ Speech-to-Text ASR Conformer Model Finetunned on Libri Speech with Security Features 🎶")

audio_file = st.file_uploader("Upload an audio file", type=["wav", "mp3", "flac"])

if audio_file:
    audio_path = "temp_audio.wav"
    with open(audio_path, "wb") as f:
        f.write(audio_file.read())

    speech, sr = librosa.load(audio_path, sr=16000)
    
    # Simulate an adversarial attack by injecting random noise
    adversarial_speech = speech + (attack_strength * np.random.randn(*speech.shape))
    adversarial_speech = np.clip(adversarial_speech, -1.0, 1.0)
    
    inputs = processor(adversarial_speech, sampling_rate=sr, return_tensors="pt", padding=True)
    input_values = inputs.input_values.to(device)
    
    with torch.no_grad():
        logits = model(input_values).logits
    
    predicted_ids = torch.argmax(logits, dim=-1)
    transcription = processor.batch_decode(predicted_ids, skip_special_tokens=True)
    
    if attack_strength > 0.2:
        st.warning("⚠️ Adversarial attack detected! Transcription may be affected.")
    
    st.success("📄 Secure Transcription:")
    st.write(transcription[0])