Query-decompose-baselines / methods /searchain /ColBERT /process_hotpotqa_wiki-Copy1.py
Veblen34's picture
Upload folder using huggingface_hub
cd6775d verified
Raw
History Blame Contribute Delete
1.11 kB
import json
import bz2
import os
import csv
path = '/home/icml01/multi_rag/RAG/Search-in-the-Chain/data/enwiki-20171001-pages-meta-current-withlinks-abstracts'
def load_data(path=None):
assert path
corpus = []
count = 0
with open(
'/home/icml01/multi_rag/RAG/Search-in-the-Chain/data/enwiki-20171001-pages-meta-current-withlinks-abstracts.tsv',
'w', encoding='utf-8-sig') as w:
writer = csv.writer(w, delimiter="\t")
for root, dirs, files in os.walk(path):
for file in files:
file_path = root + '/' + file
with bz2.open(file_path, 'r') as fin:
for line in fin.readlines():
data = json.loads(line)
row_text = ''
for text in data['text']:
row_text = row_text + ' ' + text
writer.writerow([count, 'Title: '+data['title']+' Text: '+row_text]) #pid \t passage text
count += 1
print(count)
return corpus
load_data(path)