Spaces:
Sleeping
Sleeping
Truong-Phuc Nguyen
commited on
Update app.py
Browse files
app.py
CHANGED
|
@@ -4,37 +4,37 @@ import re
|
|
| 4 |
|
| 5 |
st.set_page_config(page_icon='🍃', page_title='MRC for Legal Document Dataset checker', layout='wide', initial_sidebar_state="collapsed")
|
| 6 |
|
| 7 |
-
st.markdown("<h2 style='text-align: center;'>
|
| 8 |
|
| 9 |
df = pd.read_csv(filepath_or_buffer='./Legal_AbstractiveA.csv')
|
| 10 |
|
| 11 |
if 'idx' not in st.session_state:
|
| 12 |
st.session_state.idx = 0
|
| 13 |
|
| 14 |
-
st.markdown(f"<h4 style='text-align: center;'>
|
| 15 |
|
| 16 |
col_1, col_2, col_3, col_4, col_5, col_6, col_7, col_8, col_9, col_10 = st.columns([1, 1, 1, 1, 1, 1, 1, 1, 1, 1])
|
| 17 |
|
| 18 |
-
btn_previous = col_1.button(label=':arrow_backward:
|
| 19 |
-
btn_next = col_2.button(label='
|
| 20 |
-
btn_save = col_3.button(label=':heavy_check_mark:
|
| 21 |
txt_goto = col_5.selectbox(label='Sample', label_visibility='collapsed', options=list(range(1, len(df) + 1)))
|
| 22 |
-
btn_goto = col_6.button(label=':fast_forward:
|
| 23 |
|
| 24 |
if len(df) != 0:
|
| 25 |
col_x1, col_x2 = st.columns([8.5, 1.5])
|
| 26 |
-
txt_context = col_x1.text_area(height=300, label='
|
| 27 |
-
txt_question = st.text_area(height=100, label='
|
| 28 |
-
txt_answer = st.text_area(height=100, label='
|
| 29 |
|
| 30 |
-
options = ['Never been evaluated', '
|
| 31 |
criteria_1_value = df['criteria_1'][st.session_state.idx] if 'criteria_1' in df.columns else 'Never been evaluated'
|
| 32 |
criteria_2_value = df['criteria_2'][st.session_state.idx] if 'criteria_2' in df.columns else 'Never been evaluated'
|
| 33 |
criteria_3_value = df['criteria_3'][st.session_state.idx] if 'criteria_3' in df.columns else 'Never been evaluated'
|
| 34 |
|
| 35 |
-
criteria_1 = col_x2.selectbox(label='
|
| 36 |
-
criteria_2 = col_x2.selectbox(label='Is the answer correct, clear, and fluent?', options=['Never been evaluated', '
|
| 37 |
-
criteria_3 = col_x2.selectbox(label='Do the question and answer pairs match each other?', options=['Never been evaluated', '
|
| 38 |
|
| 39 |
if txt_answer.strip() and txt_context.strip():
|
| 40 |
highlighted_context = re.sub(re.escape(txt_answer), "<mark>" + txt_answer + "</mark>", txt_context, flags=re.IGNORECASE)
|
|
@@ -62,7 +62,7 @@ if len(df) != 0:
|
|
| 62 |
df['criteria_1'][st.session_state.idx] = criteria_1
|
| 63 |
df['criteria_2'][st.session_state.idx] = criteria_2
|
| 64 |
df['criteria_3'][st.session_state.idx] = criteria_3
|
| 65 |
-
btn_download = col_4.download_button(data=df.to_csv(), label=':arrow_down_small:
|
| 66 |
df.to_csv(path_or_buf='./Legal_AbstractiveA.csv', index=None)
|
| 67 |
|
| 68 |
if btn_goto:
|
|
|
|
| 4 |
|
| 5 |
st.set_page_config(page_icon='🍃', page_title='MRC for Legal Document Dataset checker', layout='wide', initial_sidebar_state="collapsed")
|
| 6 |
|
| 7 |
+
st.markdown("<h2 style='text-align: center;'>MRC Dataset Checker</h2>", unsafe_allow_html=True)
|
| 8 |
|
| 9 |
df = pd.read_csv(filepath_or_buffer='./Legal_AbstractiveA.csv')
|
| 10 |
|
| 11 |
if 'idx' not in st.session_state:
|
| 12 |
st.session_state.idx = 0
|
| 13 |
|
| 14 |
+
st.markdown(f"<h4 style='text-align: center;'>Bản ghi {st.session_state.idx + 1}/{len(df)}</h4>", unsafe_allow_html=True)
|
| 15 |
|
| 16 |
col_1, col_2, col_3, col_4, col_5, col_6, col_7, col_8, col_9, col_10 = st.columns([1, 1, 1, 1, 1, 1, 1, 1, 1, 1])
|
| 17 |
|
| 18 |
+
btn_previous = col_1.button(label=':arrow_backward: Bản ghi trước', use_container_width=True)
|
| 19 |
+
btn_next = col_2.button(label='Bản ghi sau :arrow_forward:', use_container_width=True)
|
| 20 |
+
btn_save = col_3.button(label=':heavy_check_mark: Lưu thay đổi', use_container_width=True)
|
| 21 |
txt_goto = col_5.selectbox(label='Sample', label_visibility='collapsed', options=list(range(1, len(df) + 1)))
|
| 22 |
+
btn_goto = col_6.button(label=':fast_forward: Chuyển đến', use_container_width=True)
|
| 23 |
|
| 24 |
if len(df) != 0:
|
| 25 |
col_x1, col_x2 = st.columns([8.5, 1.5])
|
| 26 |
+
txt_context = col_x1.text_area(height=300, label='Ngữ cảnh:', value=df['context'][st.session_state.idx])
|
| 27 |
+
txt_question = st.text_area(height=100, label='Câu hỏi:', value=df['question'][st.session_state.idx])
|
| 28 |
+
txt_answer = st.text_area(height=100, label='Câu trả lời:', value=df['answer'][st.session_state.idx])
|
| 29 |
|
| 30 |
+
options = ['Never been evaluated', 'Tệ', 'Chấp nhận được', 'Tốt']
|
| 31 |
criteria_1_value = df['criteria_1'][st.session_state.idx] if 'criteria_1' in df.columns else 'Never been evaluated'
|
| 32 |
criteria_2_value = df['criteria_2'][st.session_state.idx] if 'criteria_2' in df.columns else 'Never been evaluated'
|
| 33 |
criteria_3_value = df['criteria_3'][st.session_state.idx] if 'criteria_3' in df.columns else 'Never been evaluated'
|
| 34 |
|
| 35 |
+
criteria_1 = col_x2.selectbox(label='Câu hỏi có tự nhiên, bao quát, phù hợp với nội dung không?', options=['Never been evaluated', 'Tệ', 'Chấp nhận được', 'Tốt'], index=options.index(criteria_1_value))
|
| 36 |
+
criteria_2 = col_x2.selectbox(label='Is the answer correct, clear, and fluent?', options=['Never been evaluated', 'Tệ', 'Chấp nhận được', 'Tốt'], index=options.index(criteria_2_value))
|
| 37 |
+
criteria_3 = col_x2.selectbox(label='Do the question and answer pairs match each other?', options=['Never been evaluated', 'Tệ', 'Chấp nhận được', 'Tốt'], index=options.index(criteria_3_value))
|
| 38 |
|
| 39 |
if txt_answer.strip() and txt_context.strip():
|
| 40 |
highlighted_context = re.sub(re.escape(txt_answer), "<mark>" + txt_answer + "</mark>", txt_context, flags=re.IGNORECASE)
|
|
|
|
| 62 |
df['criteria_1'][st.session_state.idx] = criteria_1
|
| 63 |
df['criteria_2'][st.session_state.idx] = criteria_2
|
| 64 |
df['criteria_3'][st.session_state.idx] = criteria_3
|
| 65 |
+
btn_download = col_4.download_button(data=df.to_csv(), label=':arrow_down_small: Tải file', use_container_width=True, file_name="checked.csv", mime="text/csv")
|
| 66 |
df.to_csv(path_or_buf='./Legal_AbstractiveA.csv', index=None)
|
| 67 |
|
| 68 |
if btn_goto:
|