File size: 3,578 Bytes
5bcc73a |
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 |
import streamlit as st
from pathlib import Path
from uuid import uuid4
import csv
from datetime import datetime, timezone
from huggingface_hub import CommitScheduler
CSV_DATASET_DIR = Path("flagged_rows")
CSV_DATASET_DIR.mkdir(parents=True, exist_ok=True)
CSV_DATASET_PATH = CSV_DATASET_DIR / f"train-{uuid4()}.csv"
wrote_header = False
def write_header(writer):
writer.writerow(
[
"date",
"grascii",
"longhand",
"incorrect_grascii",
"incorrect_longhand",
"incorrect_shorthand",
"improperly_cropped",
"extraneous_marks",
]
)
global wrote_header
wrote_header = True
scheduler = CommitScheduler(
repo_id=st.secrets.FEEDBACK_REPO,
repo_type="dataset",
folder_path=CSV_DATASET_DIR,
path_in_repo="data",
every=15,
token=st.secrets.HF_TOKEN,
)
@st.dialog("Flag Results for Review", width="large")
def report_dialog(data):
st.write("Please select one or more reasons for flagging each row:")
report_df = data
report_df["3"] = True
report_df["4"] = False
report_df["5"] = False
report_df["6"] = False
report_df["7"] = False
report_df["8"] = False
final_report = st.data_editor(
report_df,
hide_index=True,
column_config={
"0": "Grascii",
"1": "Longhand",
"2": st.column_config.ImageColumn("Shorthand", width="medium"),
"3": st.column_config.CheckboxColumn("Flag"),
"4": st.column_config.CheckboxColumn("Grascii is incorrect"),
"5": st.column_config.CheckboxColumn("Longhand is incorrect"),
"6": st.column_config.CheckboxColumn("Shorthand image is incorrect"),
"7": st.column_config.CheckboxColumn(
"Shorthand image is improperly cropped"
),
"8": st.column_config.CheckboxColumn(
"Shorthand image contains extraneous marks"
),
},
disabled=["0", "1", "2"],
use_container_width=True,
)
st.write(
"If you decide that a listed row does not need to be flagged, uncheck its 'Flag' box to prevent it from being included in the submission."
)
if st.button("Submit"):
with scheduler.lock:
with open(CSV_DATASET_PATH, "a", newline="") as f:
writer = csv.writer(f, dialect="unix")
def write_row(row):
if not wrote_header:
write_header(writer)
if row.iloc[3] and any(
[
row.iloc[4],
row.iloc[5],
row.iloc[6],
row.iloc[7],
row.iloc[8],
]
):
writer.writerow(
[
datetime.now(timezone.utc).date(),
row.iloc[0],
row.iloc[1],
1 if row.iloc[4] else 0,
1 if row.iloc[5] else 0,
1 if row.iloc[6] else 0,
1 if row.iloc[7] else 0,
1 if row.iloc[8] else 0,
]
)
final_report.apply(write_row, axis=1)
st.session_state["report_submitted"] = True
st.rerun()
|