Published April 29, 2025 | Version v1
Dataset Open

HTR model: Scottish Custom Books V0.8

  • 1. ROR icon Trinity College Dublin

Description

This Transkribus HTR model has been trained on a text collection consisting of samples from the Scottish port books dated between 1660 and 1691 all from the collection of the National Records of Scotland. The model has been trained on photographs of 631 pages from Scottish port custom books from 1665-1691.

The model has been trained in order to provide a systematic analysis of the overseas’ marine commodity export of Scotland in the period following the Stuart restoration of 1660 and ending with the Act of Union of 1707. The model was trained and utilised to identify and extract all occurences of marine commodities being declared as exported from Scotland in this period. The model has been trained by Ph.D students Johannes Rom Dahl & Sophia Chapple of Trinity College Dublin and is a contribution to the 'Life in the Currents' project, a Prendergast Challenge Award funded multidisciplinary project investigating the role of naturally driven variability in the historical and contemporary exploitation of marine life in the Northeast Atlantic. 

The model has been trained on complete port customs books from nine different ports representing the different jurisdictions of seventeenth century Scotland. The books are all available as part of the collection of the National Records of Scotland (NRS) as part of the "The Exchequer records, custom books 2nd series (E72)".

The following port books has been included in the training material: Aberdeen (1690-91), Ayr (1667, 1681, 1681-82, 1682, 1682-83, 1684-85, 1685-86, 1689, 1689-90, 1690), Blackness and Bo’ness (1681-82), Edinburgh (1672-73, Inverness (1665-67, 1668-69, 1672-73, 1684-85, 1690), Kelso (1689-90), Kirkcaldy (Fife) (1672, 1673, 1680-81, 1681, 1681-1682, 1682-83, 1683-84, 1684-85, 1684-85, 1685-86, 1688-89, 1689-90), Leith (1671-72, 1681, 1682-83, 1683-84, 1684-85, 1684-85, 1685-86, 1688-89), Montrose (1672-73). The training material contains a decent variety of hands, styles and page layouts representative for the port books of the period. 

Samples from the training data are included as attachments. All training material used for the model are protected under Crown copyright. The included samled page are the following:

Crown copyright. National Records of Scotland, Aberdeen E72/1/5 p 8, Ayr E72/3/10 p 6, Blackness and Bo’ness E72/5/14 p 6, Edinburgh E72/8/5 p 18, Inverness E72/11/2 p 6, Kelso E72/14/18 pg 8, Kirkcaldy (Fife) E72/9/15 pg 15, Leith E72/15/41 pg 7, Montrose E72/16/1 pg 24, Port Glasgow E72/19/4 pg 5


Link to the model description on Transkribus.org:
https://www.transkribus.org/model/scottish-custom-books-v0.8 

 

Model details: 

Size (no. of words): 116.003
Training set size: 631 pages
CER Training data: 4.91 %
CER Validation set: 10.42 %
Model ID: 315953

Training settings:

5 % Validation set 
Base model: The English Eagle
Trained in 250 Epochs 
Learning Rate 0.0003

Files

Scottish customs books Training Set samples.zip

Files (107.2 MB)

Name Size Download all
md5:5229902c952386aa3b145c760ef6396e
107.2 MB Preview Download

Additional details

Funding

Trinity College Dublin
Prendergast Challenge Awards