Spaces:
Sleeping
Sleeping
Add Streamlit app and requirements
Browse files- app.py +34 -0
- requirements.txt +3 -0
app.py
ADDED
@@ -0,0 +1,34 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
import streamlit as st
|
2 |
+
from transformers import AutoModelForTokenClassification, AutoTokenizer
|
3 |
+
import torch
|
4 |
+
|
5 |
+
# Load the model and tokenizer from Hugging Face
|
6 |
+
model_name = "fajjos/Keyword_v1" # Replace with the actual model name
|
7 |
+
tokenizer = AutoTokenizer.from_pretrained(model_name)
|
8 |
+
model = AutoModelForTokenClassification.from_pretrained(model_name)
|
9 |
+
|
10 |
+
# Streamlit interface
|
11 |
+
st.title("Keyword Extractor")
|
12 |
+
user_input = st.text_area("Enter text for keyword extraction")
|
13 |
+
|
14 |
+
if user_input:
|
15 |
+
# Tokenize the input
|
16 |
+
inputs = tokenizer(user_input, return_tensors="pt")
|
17 |
+
|
18 |
+
# Get model predictions
|
19 |
+
with torch.no_grad():
|
20 |
+
outputs = model(**inputs)
|
21 |
+
|
22 |
+
# Process the predictions (this will depend on your specific model output)
|
23 |
+
tokens = tokenizer.convert_ids_to_tokens(inputs['input_ids'][0])
|
24 |
+
predictions = torch.argmax(outputs.logits, dim=2)
|
25 |
+
|
26 |
+
# Display extracted keywords
|
27 |
+
st.write("Extracted Keywords:")
|
28 |
+
for token, pred in zip(tokens, predictions[0]):
|
29 |
+
if pred == 1: # Assuming label '1' corresponds to a keyword
|
30 |
+
st.write(token)
|
31 |
+
|
32 |
+
# # Add a slider for interaction (example)
|
33 |
+
# x = st.slider('Select a value')
|
34 |
+
# st.write(f"{x} squared is {x * x}")
|
requirements.txt
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
transformers
|
2 |
+
torch
|
3 |
+
streamlit
|