load_data.py

import os
import random
import json
import shutil
from tqdm import tqdm
# from FlagEmbedding import FlagLLMModel

# model = FlagLLMModel('BAAI/bge-multilingual-gemma2',
#                      query_instruction_for_retrieval="Given a question, retrieve passages that answer the question.",
#                      query_instruction_format="<instruct>{}\n<query>{}",
#                      use_fp16=True,
#                      devices=['cuda:1']) # Setting use_fp16 to True speeds up computation with a slight performance degradation

avaliable_languages = ['ar', 'bn', 'en', 'es', 'fa', 'fi', 'fr', 'hi', 'id', 'ja', 'ko', 'ru', 'sw', 'te', 'th', 'zh', 'de', 'yo']

base_dir = '/share_2/chaofan/dataset/miracl/data'
new_dir = '/share/chaofan/code/bge_demo/data'
new_emb_dir = '/share/chaofan/code/bge_demo/emb'

for lang in tqdm(avaliable_languages, desc='language'):
    qrels_path = os.path.join(base_dir, lang, 'dev_qrels.jsonl')
    queries_path = os.path.join(base_dir, lang, 'dev_queries.jsonl')
    corpus_path = os.path.join(base_dir, lang, 'corpus.jsonl')

    os.makedirs(os.path.join(new_dir, lang), exist_ok=True)

    new_qrels_path = os.path.join(new_dir, lang, 'dev_qrels.jsonl')
    new_queries_path = os.path.join(new_dir, lang, 'dev_queries.jsonl')
    new_corpus_path = os.path.join(new_dir, lang, 'corpus.jsonl')

    useful_corpus = set()
    with open(qrels_path) as f:
        for line in f:
            data = json.loads(line)
            useful_corpus.add(data['docid'])

    corpus_ids = []
    corpus = {}
    with open(corpus_path) as f:
        for line in f:
            data = json.loads(line)
            corpus_ids.append(data['id'])
            corpus[data['id']] = data
    
    new_corpus = []
    random.shuffle(corpus_ids)
    for i in range(min(1000000, len(corpus_ids))):
        if corpus_ids[i] not in useful_corpus:
            useful_corpus.add(corpus_ids[i])
    
    print(f'language {lang}, all corpus {len(corpus_ids)}, use corpus {len(useful_corpus)}')


    with open(new_corpus_path, 'w') as f:
        for idx in useful_corpus:
            f.write(json.dumps(corpus[idx]) + '\n')
    
    shutil.copy(qrels_path, new_qrels_path)
    shutil.copy(queries_path, new_queries_path)