Article
FineBooks: are open OCR models good enough to unlock historical knowledge?
finebooks
• • 25
Ameya Joshi
ameya-extend
·
AI & ML interests
None yet
Recent Activity
upvoted an article 14 days ago
FineBooks: are open OCR models good enough to unlock historical knowledge? new activity 3 months ago
Extend-AI/RealDoc-Bench:[bot] Conversion to Parquet new activity 3 months ago
Extend-AI/RealDoc-Bench:Fix 4 ground-truth errors in qa_bank (mortgage + supply_chain)Organizations
upvoted an article 14 days ago
[bot] Conversion to Parquet
#1 opened 3 months ago
by 
parquet-converter
Fix 4 ground-truth errors in qa_bank (mortgage + supply_chain)
#6 opened 3 months ago
by 
joon-extend
fix: correct 7 gold answers and remove 3 bad questions Gold corrections: - finance_q224: fix date from 2026-03-10 to 26-03-10 (1925 tax receipt) - finance_q238: state=D is a postcode label not a state entry → state=null - finance_q206: labor_amount 9935 → 99.35 (missing decimal) - medical_healthcare_q57: urine checkbox is checked → true - medical_healthcare_q4: codes are -7/-8 (circled), not 77/98 → 7/8 - medical_healthcare_q5: neither value circled → self_pay_value=null - mortgage_q391: item l is 'No' not 'Yes' Removed questions: - finance_q355: unreadable text, misleading 'gotcha' question - mortgage_q352: misaligned column/row mapping, inconsistent gold - finance_q14: nonsensical question (Evidence is a sibling field, not umbrella)
#4 opened 3 months ago
by ameya-extend
updated a
dataset 3 months ago
published a
dataset 3 months ago
Fix mortgage_q208: drop unsourceable comp3_net_adjustment_total field
2
#3 opened 3 months ago
by 
brandonhawi
Fix mortgage_q222 gold: include second sentence of Cash to Close description
#2 opened 3 months ago
by brandonhawi
updated a
Space 3 months ago
published a
Space 3 months ago
updated a
collection 3 months ago
updated 2
datasets 3 months ago
published 2
datasets 3 months ago