File size: 6,470 Bytes
773685b |
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 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 |
import streamlit as st
import pandas as pd
import logging
import json
from dotenv import load_dotenv
import modeling
def show_launch(placeholder):
with placeholder.container():
st.divider()
st.markdown("""
## Before Using the App
### Disclaimer
This application is provided as-is, without any warranty or guarantee of any kind, expressed or implied. It is intended for educational, non-commercial use only.
The developers of this app shall not be held liable for any damages or losses incurred from its use. By using this application, you agree to the terms and conditions
outlined herein and acknowledge that any commercial use or reliance on its functionality is strictly prohibited.
""", unsafe_allow_html=True)
button_placeholder = st.empty()
if button_placeholder.button(label='Accept Disclaimer', type='primary', use_container_width=True):
st.session_state.show_launch = False
placeholder.empty()
button_placeholder.empty()
def show_demo(placeholder):
with placeholder:
with st.container():
st.divider()
st.markdown("""
## Try it yourself!
Use the input fields provided below to create items aimed at
assessing a particular psychological construct (e.g., personality
trait). If desired, employ the prefix option to generate items
that begin with a predetermined string. To manage the diversity
of the output, various sampling strategies may be applied.
For further information on these strategies, please refer to the
accompanying paper.
""")
modeling.load_model()
sampling_options = ['Greedy Search', 'Beam Search', 'Multinominal Sampling']
sampling_input = st.radio('Sampling', options=sampling_options, index=2, horizontal=True)
left_col, right_col = st.columns([1, 1])
with left_col:
prefix_input = st.text_input('Prefix', '')
construct_input = st.text_input('Construct', 'Pessimism')
with right_col:
if sampling_options.index(sampling_input) == 0:
num_beams = 1
num_return_sequences = 1
temperature = 1
top_k = 0
top_p = 1
if sampling_options.index(sampling_input) == 1:
num_beams = st.slider('Number of Search Beams', min_value=1, max_value=10, value=3, step=1)
num_return_sequences = st.slider('Number of Beams to Return', min_value=1, max_value=10, value=2, step=1)
temperature = 1
top_k = 0
top_p = 1
if sampling_options.index(sampling_input) == 2:
num_beams = 1
num_return_sequences = 1
temperature = st.slider('Temperature', min_value=0.1, max_value=1.5, value=1.0, step=0.1)
top_k = st.slider('Top k (0 = disabled)', min_value=0, max_value=1000, value=40, step=1)
top_p = st.slider('Top p (0 = disabled)', min_value=0.0, max_value=1.0, value=0.95, step=0.05)
message = st.empty()
if st.button(label='Generate Item', type='primary', use_container_width=True):
if num_return_sequences <= num_beams:
if len(construct_input) > 0:
kwargs = {
'num_return_sequences': num_return_sequences,
'num_beams': num_beams,
'do_sample': sampling_options.index(sampling_input) == 2,
'temperature': temperature,
'top_k': top_k,
'top_p': top_p
}
item_stems = modeling.generate_items(construct_input, prefix_input, **kwargs)
st.session_state.outputs.append({'construct': construct_input, 'item': item_stems})
else:
message.error('You have to enter a construct to proceed with item generation!')
else:
message.error('You cannot return more beams than to search for!')
if len(st.session_state.outputs) > 0:
tab1, tab2 = st.tabs(["Generated Items", "Details on last prompt"])
with tab1:
for output in st.session_state.outputs:
placeholder_outputs = st.empty()
with tab2:
pass
df = pd.DataFrame(st.session_state.outputs).explode(column='item').reset_index()
placeholder_outputs = st.dataframe(df.sort_values(by='index', ascending=False), use_container_width=True)
def initialize():
load_dotenv()
logging.basicConfig(format='%(asctime)s - %(name)s - %(levelname)s - %(message)s', level=logging.INFO)
if 'state_loaded' not in st.session_state:
st.session_state['state_loaded'] = True
with open('init.json') as json_data:
st.session_state.update(json.load(json_data))
def main():
st.set_page_config(page_title='Construct-Specific Automatic Item Generation')
col1, col2 = st.columns([2, 5])
with col1:
st.image('logo-130x130.svg')
with col2:
st.markdown("# Construct-Specific Automatic Item Generation")
st.markdown("""
This web application showcases item generation for psychological scale development
using natural language processing ("AI"), accompanying the paper
"Transformer-Based Deep Neural Language Modeling for Construct-Specific Automatic Item Generation".
📖 Paper (Open Access): https://link.springer.com/article/10.1007/s11336-021-09823-9
💾 Data: https://osf.io/rhe9w/
🖊️ Cite:<br> Hommel, B. E., Wollang, F.-J. M., Kotova, V., Zacher, H., & Schmukle, S. C. (2022). Transformer-Based Deep Neural Language Modeling for Construct-Specific Automatic Item Generation. Psychometrika, 87(2), 749–772. https://doi.org/10.1007/s11336-021-09823-9
#️⃣ Twitter/X: https://twitter.com/BjoernHommel
The web application is maintained by [magnolia psychometrics](https://www.magnolia-psychometrics.com/).
""", unsafe_allow_html=True)
placeholder_launch = st.empty()
placeholder_demo = st.empty()
if 'disclaimer' not in st.session_state:
show_launch(placeholder_launch)
st.session_state['disclaimer'] = True
else:
show_demo(placeholder_demo)
if __name__ == '__main__':
initialize()
main() |