Scrapes, saves and tags fefe-blog-articles in a specific time frame. Also prints some tag stats.
Outputs the tagged articles as json and as jsonl ready for openai fine-tuning.
Disclaimer: this is just quickly hacked together.
# once after pull
npm install
# run
npm run dev# prepare data for training (validate, remove duplicates, etc.)
openai tools fine_tunes.prepare_data -f data/data.jsonl
# start training
openai api fine_tunes.create -t data/data_prepared.jsonl -m ada --suffix fefe-2
# view all fine tuning jobs, to check status of created job
openai api fine_tunes.list
# cancel a fine tuning job
openai api fine_tunes.cancel -i ${FINE_TUNE_JOB_ID}