The National Archives Netherlands have made this model for typewritten and print writings which were used in the Netherlands during the late 19th and early 20th century. 72100 words have been trained for this model and the CER on the validation set is 3.24%.
