README.md · daelba/vital2wikibase at 0f9d6c1f775f7aec148fa1c07c860468382e4a5a

metadata

license: cc-by-nc-sa-4.0
language:
  - cs
  - de
pipeline_tag: text2text-generation
tags:
  - code
  - wikibase
  - vital records
base_model: google/mt5-small
inference:
  parameters:
    max_new_tokens: 200
datasets:
  - daelba/hbma

A model for converting vital records to Wikibase. Trained on a transcription dataset extracted with HTR (eScriptorium) from 150 pages of the Jewish register book HBMa 2501 (Wikidata). The dataset contains 1071 vital records, which were uploaded to Wikibase Cloud after processing with this model.

Based on Google's mT5.

Processing ditto marks

mt5prefix = "dto"
upperCell = "18 Juny 1858 H. Barb. Kohn"
lowerCell = "dto dto H. Ros. Plowitz"
dataset_input = f"{mt5prefix}:{upperCell}|{lowerCell}"
dataset_target = "18 Juny 1858 H. Ros. Plowitz"

Processing a table cell

mt5prefix = "wbPre"
header = "Namen des Vaters mit Anführung des Standes, ob Familiant od. Gewerbsmann, u. seines Wohnortes, Schutzdom. Haus-Nr. und bei ehelichen Geburten seines Gubernial-Ehekonsenses Datum und Zahl"
cell = "Markus Gibianer, Lederhändler, S. d. Seligmann Gibianer u. der Magdalena geb. Motz"
dataset_input = f"{mt5prefix}:{header}|{cell}"
dataset_target = "[[P23:Markus Gibianer|P26:Lederhändler]] [[P27:Seligmann Gibianer]] [[P28:Magdalena Gibianer geb. Motz]]"

Processing a whole record

See the entry on Wikibase Cloud.

mt5prefix = "wbDef"
input_string = "[[P5:60]] [[P6:1865-05-05]] [[P7:H. Charlotte Löw]] [[P12|DD:23]] [[P12|MM:04]] [[P12|YY:1865]] [[P14|P8:1061/2]] [[P14:Q8]] [[P15|DD:06]] [[P15|MM:05]] [[P15|YY:1865]] [[P16:Q22]] [[P17:Friederika]] [[P20:4]] [[P21:Q13]] [[P22:Q15]] [[P23:Jakob Raubitschek|P26:Eisenhändler]] [[P27:Josef Raubitschek]] [[P28:Theresia Raubitschek geb. Latus]] [[P24:Pauline XXX]] [[P29:Markus Raudnitz]] [[P30:Rachel Raudnitz geb. Kuh]]"
dataset_input = f"{mt5prefix}:{input_string}"
dataset_target = "[[P12:1865-04-23]] [[P14:Q8|P8:1061/2]] [[P15:1865-05-06]] [[P16:Q22]] [[P17:Friederika]] [[P18:Raubitschek]] [[P19:Friederika Raubitschek]] [[P1:Q1]] [[P20:4]] [[P21:Q13]] [[P22:Q15]] [[P23:Jakob Raubitschek|P26:Eisenhändler]] [[P24:Pauline Raubitschek]] [[P27:Josef Raubitschek]] [[P28:Theresia Raubitschek geb. Latus]] [[P29:Markus Raudnitz]] [[P30:Rachel Raudnitz geb. Kuh]] [[P33:Charlotte Löw]] [[P38:Q170]] [[P5:60]] [[P6:1865-05-05]] [[P7:H. Charlotte Löw]]"