Spaces:
Runtime error
Runtime error
Update app.py
Browse files
app.py
CHANGED
@@ -33,47 +33,59 @@ from PyPDF2 import PdfReader
|
|
33 |
from templates import bot_template, css, user_template
|
34 |
from xml.etree import ElementTree as ET
|
35 |
|
36 |
-
|
|
|
37 |
from datasets import load_dataset
|
38 |
|
39 |
# π Load USMLE Step 1 dataset
|
40 |
-
dataset = load_dataset("augtoma/usmle_step_1")
|
41 |
|
42 |
# π Streamlit App π
|
43 |
st.title("USMLE Step 1 Dataset Viewer")
|
44 |
-
|
45 |
-
π
|
46 |
-
|
47 |
-
|
48 |
-
# π©βπ¬ Search Box
|
49 |
-
search_term = st.text_input("Search for a specific question:", "")
|
50 |
-
|
51 |
-
# π Pagination
|
52 |
-
records_per_page = 100
|
53 |
-
num_records = len(dataset['test'])
|
54 |
-
num_pages = int(num_records / records_per_page) + 1
|
55 |
-
page_number = st.select_slider("Select page:", options=list(range(1, num_pages + 1)))
|
56 |
-
|
57 |
-
# π Display Data
|
58 |
-
start_idx = (page_number - 1) * records_per_page
|
59 |
-
end_idx = start_idx + records_per_page
|
60 |
-
|
61 |
-
# π§ͺ Apply the Search Filter
|
62 |
-
if search_term:
|
63 |
-
filtered_data = [record for record in dataset['test'][start_idx:end_idx] if search_term.lower() in record['text'].lower()]
|
64 |
else:
|
65 |
-
|
66 |
-
|
67 |
-
|
68 |
-
|
69 |
-
|
70 |
-
st.
|
71 |
-
|
72 |
-
|
73 |
-
|
74 |
-
|
75 |
-
|
76 |
-
st.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
77 |
|
78 |
# 1. Constants and Top Level UI Variables
|
79 |
|
|
|
33 |
from templates import bot_template, css, user_template
|
34 |
from xml.etree import ElementTree as ET
|
35 |
|
36 |
+
|
37 |
+
import streamlit as st
|
38 |
from datasets import load_dataset
|
39 |
|
40 |
# π Load USMLE Step 1 dataset
|
41 |
+
dataset = load_dataset("augtoma/usmle_step_1")['test'] # Using 'test' split
|
42 |
|
43 |
# π Streamlit App π
|
44 |
st.title("USMLE Step 1 Dataset Viewer")
|
45 |
+
|
46 |
+
# π Check if dataset is empty
|
47 |
+
if len(dataset) == 0:
|
48 |
+
st.write("π’ The dataset is empty.")
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
49 |
else:
|
50 |
+
st.write("""
|
51 |
+
π Use the search box to filter questions or use the grid to scroll through the dataset.
|
52 |
+
""")
|
53 |
+
|
54 |
+
# π©βπ¬ Search Box
|
55 |
+
search_term = st.text_input("Search for a specific question:", "")
|
56 |
+
|
57 |
+
# π Pagination
|
58 |
+
records_per_page = 100
|
59 |
+
num_records = len(dataset)
|
60 |
+
num_pages = max(int(num_records / records_per_page), 1)
|
61 |
+
page_number = st.select_slider("Select page:", options=list(range(1, num_pages + 1)))
|
62 |
+
|
63 |
+
# π Display Data
|
64 |
+
start_idx = (page_number - 1) * records_per_page
|
65 |
+
end_idx = start_idx + records_per_page
|
66 |
+
|
67 |
+
# π§ͺ Apply the Search Filter
|
68 |
+
filtered_data = []
|
69 |
+
for record in dataset[start_idx:end_idx]:
|
70 |
+
if isinstance(record, dict) and 'text' in record and 'id' in record:
|
71 |
+
if search_term:
|
72 |
+
if search_term.lower() in record['text'].lower():
|
73 |
+
filtered_data.append(record)
|
74 |
+
else:
|
75 |
+
filtered_data.append(record)
|
76 |
+
|
77 |
+
# π Render the Grid
|
78 |
+
for record in filtered_data:
|
79 |
+
st.write(f"## Question ID: {record['id']}")
|
80 |
+
st.write(f"### Question:")
|
81 |
+
st.write(f"{record['text']}")
|
82 |
+
st.write(f"### Answer:")
|
83 |
+
st.write(f"{record['answer']}")
|
84 |
+
st.write("---")
|
85 |
+
|
86 |
+
st.write(f"π Total Records: {num_records} | π Displaying {start_idx+1} to {min(end_idx, num_records)}")
|
87 |
+
|
88 |
+
|
89 |
|
90 |
# 1. Constants and Top Level UI Variables
|
91 |
|