Document the 4th (vision-based) extractor family

Add gpt-4o vision-based capture (6,510 fields/399 docs)

Add cord_qwen2.5-14b.json

Add cord_claude-haiku-4-5.json

Add xfund_zh_claude-sonnet-5.json

Add xfund_fr_claude-sonnet-5.json

Add xfund_es_claude-sonnet-5.json

Add xfund_de_claude-sonnet-5.json

Add funsd_claude-sonnet-5.json

Add cord_claude-sonnet-5.json

Add dataset card

initial commit