- Download PubMedCentral Author Manuscript Collection
into
~/pmc_dataset
folder.
Expected folder content:author_manuscript_xml.PMC001xxxxxx.baseline.2022-06-16.filelist.csv author_manuscript_xml.PMC001xxxxxx.baseline.2022-06-16.filelist.txt author_manuscript_xml.PMC001xxxxxx.baseline.2022-06-16.tar.gz ...
- Extract all downloaded
tar.gz
files.
Expected extracted folders:PMC001xxxxxx PMC002xxxxxx PMC003xxxxxx PMC004xxxxxx PMC005xxxxxx PMC006xxxxxx PMC007xxxxxx PMC008xxxxxx PMC009xxxxxx
- Launch
dataset/analyze_archive.ipynb
to analyze downloaded archive. - Launch
dataset/prepare_dataset.ipynb
to create tables required for model training. - Refer to
train/README.md
for instruction on model training.