Spaces:
Sleeping
Sleeping
File size: 3,647 Bytes
23ddfce 1c1f9d0 23ddfce 1c1f9d0 23ddfce 1c1f9d0 23ddfce 1c1f9d0 23ddfce 1c1f9d0 23ddfce 1c1f9d0 23ddfce 1c1f9d0 23ddfce 1c1f9d0 23ddfce 1c1f9d0 23ddfce 1c1f9d0 23ddfce 1c1f9d0 fae5744 1c1f9d0 fae5744 1c1f9d0 23ddfce |
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 |
import streamlit as st
import tensorflow as tf
import numpy as np
from tensorflow.keras.preprocessing.sequence import pad_sequences
import pickle
from lime import lime_text
from lime.lime_text import LimeTextExplainer
# Load the model
@st.cache_resource
def load_model(filepath):
return tf.keras.models.load_model(filepath)
# Load tokenizer
@st.cache_resource
def load_tokenizer(filepath):
with open(filepath, 'rb') as handle:
return pickle.load(handle)
# Preprocess prompt
def preprocess_prompt(prompt, tokenizer, max_length=100):
sequence = tokenizer.texts_to_sequences([prompt])
padded_sequence = pad_sequences(sequence, maxlen=max_length)
return padded_sequence
# Predict prompt class
def detect_prompt(prompt, tokenizer, model, max_length=100):
processed_prompt = preprocess_prompt(prompt, tokenizer, max_length)
prediction = model.predict(processed_prompt)[0][0]
class_label = "Malicious" if prediction >= 0.5 else "Valid"
confidence_score = prediction * 100 if prediction >= 0.5 else (1 - prediction) * 100
return class_label, confidence_score
# LIME explanation
def lime_explain(prompt, model, tokenizer, max_length=100):
def predict_fn(prompts):
sequences = tokenizer.texts_to_sequences(prompts)
padded_sequences = pad_sequences(sequences, maxlen=max_length)
predictions = model.predict(padded_sequences)
return np.hstack([1 - predictions, predictions])
class_names = ["Valid", "Malicious"]
explainer = LimeTextExplainer(class_names=class_names)
explanation = explainer.explain_instance(prompt, predict_fn, num_features=10)
return explanation
# Streamlit App
st.title("Prompt Injection Detection and Prevention")
st.write("Classify prompts as malicious or valid and understand predictions using LIME.")
# Model input
model_path = st.text_input("Enter the path to your trained model (.h5):")
if model_path:
try:
model = load_model(model_path)
st.success("Model Loaded Successfully!")
except Exception as e:
st.error(f"Error Loading Model: {e}")
model = None
else:
model = None
# Tokenizer input
tokenizer_path = st.text_input("Enter the path to your tokenizer file (.pickle):")
if tokenizer_path:
try:
tokenizer = load_tokenizer(tokenizer_path)
st.success("Tokenizer Loaded Successfully!")
except Exception as e:
st.error(f"Error Loading Tokenizer: {e}")
tokenizer = None
else:
tokenizer = None
# Prompt classification
if model and tokenizer:
user_prompt = st.text_input("Enter a prompt to classify:")
if user_prompt:
st.subheader("Model Prediction")
try:
# Classify the prompt
class_label, confidence_score = detect_prompt(user_prompt, tokenizer, model)
st.write(f"Predicted Class: **{class_label}**")
st.write(f"Confidence Score: **{confidence_score:.2f}%**")
# Debugging information
st.write("Debugging Information:")
st.write(f"Tokenized Sequence: {tokenizer.texts_to_sequences([user_prompt])}")
st.write(f"Padded Sequence: {preprocess_prompt(user_prompt, tokenizer)}")
st.write(f"Raw Model Output: {model.predict(preprocess_prompt(user_prompt, tokenizer))[0][0]}")
# Generate LIME explanation
explanation = lime_explain(user_prompt, model, tokenizer)
explanation_as_html = explanation.as_html()
st.components.v1.html(explanation_as_html, height=500)
except Exception as e:
st.error(f"Error during prediction: {e}")
|