[Feature] Support OCR-VQA dataset

[InvincibleWyq]

Modification

Support OCR-VQA dataset.

A example config for BLIP zero-shot inference is provided. It perform an ACC of 28.3.

Example command:

sh tools/slurm_test.sh <patrition> <jobname> configs/blip/blip-base_8xb32_ocrvqa.py /mnt/petrelfs/share_data/wangyiqin/blip_ckpt/blip-base_3rdparty-capflit_vqa_20230505-81488941.pth

[codecov]

Codecov Report

Patch coverage: 44.33% and project coverage change: -1.17 ⚠️

Comparison is base (f9dcae2) 68.16% compared to head (0e0c42f) 67.00%.

❗ Current head 0e0c42f differs from pull request most recent head 859d8d3. Consider uploading reports for the commit 859d8d3 to get more accurate results

Additional details and impacted files

@@            Coverage Diff             @@
##              dev    #1621      +/-   ##
==========================================
- Coverage   68.16%   67.00%   -1.17%     
==========================================
  Files         295      309      +14     
  Lines       23372    24608    +1236     
  Branches     3713     3903     +190     
==========================================
+ Hits        15932    16488     +556     
- Misses       6880     7518     +638     
- Partials      560      602      +42     

Flag	Coverage Δ
unittests	67.00% <44.33%> (-1.17%)	⬇️

Flags with carried forward coverage won't be shown. Click here to find out more.

Impacted Files	Coverage Δ
mmpretrain/apis/image_retrieval.py	21.42% <ø> (ø)
mmpretrain/datasets/__init__.py	66.66% <0.00%> (-7.62%)	⬇️
mmpretrain/datasets/gqa_dataset.py	0.00% <0.00%> (ø)
mmpretrain/datasets/nocaps.py	0.00% <0.00%> (ø)
mmpretrain/datasets/ocr_vqa.py	0.00% <0.00%> (ø)
mmpretrain/datasets/scienceqa.py	0.00% <0.00%> (ø)
mmpretrain/datasets/textvqa.py	0.00% <0.00%> (ø)
mmpretrain/models/multimodal/__init__.py	40.00% <0.00%> (-4.45%)	⬇️
...retrain/models/multimodal/chinese_clip/__init__.py	0.00% <0.00%> (ø)
mmpretrain/models/multimodal/chinese_clip/bert.py	0.00% <0.00%> (ø)
... and 29 more

... and 1 file with indirect coverage changes

☔ View full report in Codecov by Sentry.
📢 Do you have feedback about the report comment? Let us know in this issue.