测试集结果文件md5

链接：https://pan.baidu.com/s/1uFOFjj5QdCvaCw-8MYTFhA 提取码：jpmn --来自百度网盘超级会员V5的分享

(base) ➜  visualbert git:(kk) ✗ md5sum logs/vqa_finetune_0831/result.json 
22e0a4ce40f0ec92d8fa730b697101cd  logs/vqa_finetune_0831/result.json

下载提交到竞赛网站上的 Submitted file, 同样检查md5

(base) ➜  visualbert git:(kk) ✗ md5sum logs/vqa_finetune_0831/result.json 
22e0a4ce40f0ec92d8fa730b697101cd  logs/vqa_finetune_0831/result.json
(base) ➜  visualbert git:(kk) ✗ wget https://evalai.s3.amazonaws.com/media/submission_files/submission_156749/2ae1c2c6-7f72-4dd6-994a-b0adf5867983.json--2021-09-13 16:01:56--  https://evalai.s3.amazonaws.com/media/submission_files/submission_156749/2ae1c2c6-7f72-4dd6-994a-b0adf5867983.json
Resolving evalai.s3.amazonaws.com... 52.217.14.60
Connecting to evalai.s3.amazonaws.com|52.217.14.60|:443... connected.
HTTP request sent, awaiting response... 200 OK
Length: 20517922 (20M) [application/json]
Saving to: ‘2ae1c2c6-7f72-4dd6-994a-b0adf5867983.json’

2ae1c2c6-7f72-4dd6-994a-b0adf5867983.json       100%[======================================================================================================>]  19.57M  1.70MB/s    in 17s     

2021-09-13 16:02:30 (1.15 MB/s) - ‘2ae1c2c6-7f72-4dd6-994a-b0adf5867983.json’ saved [20517922/20517922]

(base) ➜  visualbert git:(kk) ✗ md5sum 2ae1c2c6-7f72-4dd6-994a-b0adf5867983.json 
22e0a4ce40f0ec92d8fa730b697101cd  2ae1c2c6-7f72-4dd6-994a-b0adf5867983.json

This repository contains code for the following two papers:

• VisualBERT: A Simple and Performant Baseline for Vision and Language (arxiv) with a short version titiled What Does BERT with Vision Look At? published on ACL 2020.

  Under the folder visualbert is code (the original VisualBERT), where we pre-train a Transformer for vision-and-language (V&L) tasks on image-caption data.

• Unsupervised Vision-and-Language Pre-training Without Parallel Images and Captions published on NAACL 2021.

  Under the folder unsupervised_visualbert is code (Unsupervised VisualBERT), where we pre-train a V&L transformer without aligned image-captions pairs. Rather, we pre-training only using unaligned images and text, and achieve competitive performance with many models supervised with aligned data.

The model VisualBERT has been also integrated into several libararies such as Huggingface Transformer (many thanks to Gunjan Chhablani who made it work) and Facebook MMF.

Thanks~