clean up llama2-70B draft

clean up llama2-70B draft

clean llama-70B and remove encodings from data files

add new GPT4 IE synthie data, without dot

add fixed CP predictions

fix palm filename typo

add plam, claude etc

remove semicolon from draft text data

add processed wikinre data

restructure data files

add wikinre-1k and synthie-1k

add wikinre constrained world

add wikinre with underscore fixed dataset

add wikinre dataset

add cp data with draft merged

add test_small_ordered_only_gpt3.5_draft.jsonl

fix ambiguous [e] marker in drafts generated by GPT4

fix ambiguous [e] marker in drafts generated by GPT4

fix json utf-8 encoding error in jsonl with draft

add synthie_only_draft data

add data from GCD

set up gitignore

add synthie_test with draft output

initial commit