Spaces:
Running
Running
Update app.py
Browse files
app.py
CHANGED
@@ -1,10 +1,23 @@
|
|
1 |
import streamlit as st
|
2 |
-
from transformers import AutoTokenizer, AutoModelForSequenceClassification
|
3 |
import torch
|
4 |
|
5 |
# Sidebar for user input
|
6 |
st.sidebar.header("Model Configuration")
|
7 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
8 |
|
9 |
# Load model and tokenizer on demand
|
10 |
@st.cache_resource
|
@@ -12,7 +25,10 @@ def load_model(model_name):
|
|
12 |
try:
|
13 |
# Load the model and tokenizer
|
14 |
tokenizer = AutoTokenizer.from_pretrained(model_name)
|
15 |
-
|
|
|
|
|
|
|
16 |
return tokenizer, model
|
17 |
except Exception as e:
|
18 |
st.error(f"Error loading model: {e}")
|
@@ -22,20 +38,28 @@ def load_model(model_name):
|
|
22 |
tokenizer, model = load_model(model_name)
|
23 |
|
24 |
# Input text box in the main panel
|
25 |
-
st.title("
|
26 |
-
user_input = st.text_area("Enter text
|
27 |
|
28 |
# Make prediction if user input is provided
|
29 |
if user_input and model and tokenizer:
|
30 |
-
|
31 |
-
|
32 |
-
|
|
|
|
|
|
|
|
|
33 |
|
34 |
-
|
35 |
-
|
36 |
-
|
37 |
-
|
38 |
-
|
39 |
-
|
40 |
-
|
|
|
|
|
41 |
|
|
|
|
|
|
1 |
import streamlit as st
|
2 |
+
from transformers import AutoTokenizer, AutoModelForSequenceClassification, AutoModelForSeq2SeqLM
|
3 |
import torch
|
4 |
|
5 |
# Sidebar for user input
|
6 |
st.sidebar.header("Model Configuration")
|
7 |
+
model_choice = st.sidebar.selectbox("Select a model", [
|
8 |
+
"CyberAttackDetection",
|
9 |
+
"text2shellcommands",
|
10 |
+
"pentest_ai"
|
11 |
+
])
|
12 |
+
|
13 |
+
# Define the model names
|
14 |
+
model_mapping = {
|
15 |
+
"CyberAttackDetection": "Canstralian/CyberAttackDetection",
|
16 |
+
"text2shellcommands": "Canstralian/text2shellcommands",
|
17 |
+
"pentest_ai": "Canstralian/pentest_ai"
|
18 |
+
}
|
19 |
+
|
20 |
+
model_name = model_mapping.get(model_choice, "Canstralian/CyberAttackDetection")
|
21 |
|
22 |
# Load model and tokenizer on demand
|
23 |
@st.cache_resource
|
|
|
25 |
try:
|
26 |
# Load the model and tokenizer
|
27 |
tokenizer = AutoTokenizer.from_pretrained(model_name)
|
28 |
+
if model_name == "Canstralian/text2shellcommands":
|
29 |
+
model = AutoModelForSeq2SeqLM.from_pretrained(model_name)
|
30 |
+
else:
|
31 |
+
model = AutoModelForSequenceClassification.from_pretrained(model_name)
|
32 |
return tokenizer, model
|
33 |
except Exception as e:
|
34 |
st.error(f"Error loading model: {e}")
|
|
|
38 |
tokenizer, model = load_model(model_name)
|
39 |
|
40 |
# Input text box in the main panel
|
41 |
+
st.title(f"{model_choice} Model")
|
42 |
+
user_input = st.text_area("Enter text:")
|
43 |
|
44 |
# Make prediction if user input is provided
|
45 |
if user_input and model and tokenizer:
|
46 |
+
if model_choice == "text2shellcommands":
|
47 |
+
# For text2shellcommands model, generate shell commands
|
48 |
+
inputs = tokenizer(user_input, return_tensors="pt", padding=True, truncation=True)
|
49 |
+
with torch.no_grad():
|
50 |
+
outputs = model.generate(**inputs)
|
51 |
+
generated_command = tokenizer.decode(outputs[0], skip_special_tokens=True)
|
52 |
+
st.write(f"Generated Shell Command: {generated_command}")
|
53 |
|
54 |
+
else:
|
55 |
+
# For CyberAttackDetection and pentest_ai models, perform classification
|
56 |
+
inputs = tokenizer(user_input, return_tensors="pt", padding=True, truncation=True)
|
57 |
+
with torch.no_grad():
|
58 |
+
outputs = model(**inputs)
|
59 |
+
logits = outputs.logits
|
60 |
+
predicted_class = torch.argmax(logits, dim=-1).item()
|
61 |
+
st.write(f"Predicted Class: {predicted_class}")
|
62 |
+
st.write(f"Logits: {logits}")
|
63 |
|
64 |
+
else:
|
65 |
+
st.info("Please enter some text for prediction.")
|