| import json | |
| from glob import glob | |
| import random | |
| from tqdm import tqdm | |
| # cnts = set() | |
| # for fname in tqdm(glob('/etc/ssd1/jiangzhongtao/baai_embedding_tune/data/all_collect/mteb-Retrieval_aug/*.jsonl')): | |
| # examples = [] | |
| # with open(fname, 'r') as f: | |
| # for line in f: | |
| # line = json.loads(line) | |
| # if type(line['neg']) == list: | |
| # cnts.add(len(line['neg'])) | |
| # try: | |
| # neg = random.sample(line['neg'][30:210], k=10) | |
| # line['neg'] = neg | |
| # del line['sims'] | |
| # except Exception: | |
| # pass | |
| # examples.append(line) | |
| # with open(fname, 'w') as f: | |
| # for example in examples: | |
| # f.write(json.dumps(example, ensure_ascii=False) + '\n') | |
| # import os | |
| # import json | |
| # from glob import glob | |
| # from tqdm import tqdm | |
| # src_path = '/etc/ssd1/jiangzhongtao/baai_embedding_tune/data/all_collect/mteb-Retrieval' | |
| # for dst_file in tqdm(glob('/etc/ssd1/jiangzhongtao/baai_embedding_tune/data/all_collect/mteb-Retrieval_aug/*.jsonl')): | |
| # _, fname = os.path.split(dst_file) | |
| # src_file = os.path.join(src_path, fname) | |
| # if os.path.exists(src_file): | |
| # f_src = open(src_file, 'r') | |
| # f_dst = open(dst_file, 'a') | |
| # for line in f_src: | |
| # f_dst.write(line) | |
| # f_src.close() | |
| # f_dst.close() | |
| # src_files = [ | |
| # "/etc/ssd1/dengjingcheng/general_embedding/mteb_all_data/retrieval/arguana/mteb_ArguAna_en_Aug.jsonl", | |
| # "/etc/ssd1/dengjingcheng/general_embedding/mteb_all_data/retrieval/climate-fever/mteb_ClimateFEVER_en_Aug.jsonl", | |
| # "/etc/ssd1/dengjingcheng/general_embedding/mteb_all_data/retrieval/dbpedia/mteb_DBPedia_en_Aug.jsonl", | |
| # "/etc/ssd1/dengjingcheng/general_embedding/mteb_all_data/retrieval/fever/mteb_FEVER_en_Aug.jsonl", | |
| # "/etc/ssd1/dengjingcheng/general_embedding/mteb_all_data/retrieval/fiqa/mteb_FiQA2018_en_Aug.jsonl", | |
| # "/etc/ssd1/dengjingcheng/general_embedding/mteb_all_data/retrieval/hotpotqa/mteb_HotpotQA_en_Aug.jsonl", | |
| # "/etc/ssd1/dengjingcheng/general_embedding/mteb_all_data/retrieval/msmarco/mteb_MSMARCO_en_Aug.jsonl", | |
| # "/etc/ssd1/dengjingcheng/general_embedding/mteb_all_data/retrieval/nfcorpus/mteb_NFCorpus_en_Aug.jsonl", | |
| # "/etc/ssd1/dengjingcheng/general_embedding/mteb_all_data/retrieval/nq/mteb_NQ_en_Aug.jsonl", | |
| # "/etc/ssd1/dengjingcheng/general_embedding/mteb_all_data/retrieval/quora/mteb_QuoraRetrieval_en_Aug.jsonl", | |
| # "/etc/ssd1/dengjingcheng/general_embedding/mteb_all_data/retrieval/scidocs/mteb_SCIDOCS_en_Aug.jsonl", | |
| # "/etc/ssd1/dengjingcheng/general_embedding/mteb_all_data/retrieval/scifact/mteb_SciFact_en_Aug.jsonl", | |
| # "/etc/ssd1/dengjingcheng/general_embedding/mteb_all_data/retrieval/touche2020/mteb_Touche2020_en_Aug.jsonl", | |
| # "/etc/ssd1/dengjingcheng/general_embedding/mteb_all_data/retrieval/treccovid/mteb_TRECCOVID_en_Aug.jsonl" | |
| # ] | |
| # from tqdm import tqdm | |
| # import json | |
| # import os | |
| # import random | |
| # dst_path = '/etc/ssd1/jiangzhongtao/baai_embedding_tune/data/all_collect/mteb-Retrieval_aug' | |
| # for src_file in tqdm(src_files): | |
| # _, fname = os.path.split(src_file) | |
| # fname = fname.replace('_Aug', '') | |
| # dst_file = os.path.join(dst_path, fname) | |
| # src_f = open(src_file, 'r') | |
| # dst_f = open(dst_file, 'w') | |
| # for line in src_f: | |
| # line = json.loads(line) | |
| # if len(line['neg']) != 1: | |
| # line['neg'] = random.sample(line['neg'][30:210], k=10) | |
| # del line['sims'] | |
| # dst_f.write(json.dumps(line, ensure_ascii=False) + '\n') | |
| # src_f.close() | |
| # dst_f.close() | |
| # import os | |
| # import json | |
| # from glob import glob | |
| # for file in glob('/etc/ssd1/jiangzhongtao/baai_embedding_tune/data/all_collect/mteb-Retrieval_aug/*.jsonl'): | |
| # nnegs2cnt = dict() | |
| # _, fname = os.path.split(file) | |
| # with open(file, 'r') as f: | |
| # for line in f: | |
| # line = json.loads(line) | |
| # assert 'sims' not in set(line.keys()) | |
| # n_negs = len(line['neg']) | |
| # nnegs2cnt[n_negs] = nnegs2cnt.get(n_negs, 0) + 1 | |
| # print(fname, nnegs2cnt) | |