# Config file for event chains pipeline

PIPELINE_NAME=altavilla-fiction
HUMAN_READABLE_NAME="Altavilla (fiction)"

# Point to a script that will extract the input text for us
INPUT_EXTRACTOR=1-parse/preprocess/extract_altavilla.sh
# This is specific to the Altavilla corpus, used by the input extractor
ALTAVILLA_FILES=(/anfs/bigdisc/mtw29/corpora/altavilla/fiction/*.tar.gz)

# Number of processes to use in any stage
PROCESSES=12

# Not just protagonists
PROTAGONIST=

# Split the events files into training, dev and test sets
EVENTS_POSTPROCESS=3-events/altavilla/postprocess.sh
# After this splitting, we remove stopevents from the training set. This is not currently part of the pipeline,
#  so must be done manually
# Once it's been split, we use the training chapters only as training data for the model
POSTPROCESSED_CHAINS_FILE=/anfs/bigdisc/mtw29/chains/altavilla-fiction/chains_sets/stopevents/training_chapters.txt
