README.md

---
license: cc-by-sa-4.0
language:
- en
tags:
- biology
- medical
size_categories:
- 100K<n<1M

dataset_info:
  - config_name: default
    features:
      - name: query-id
        dtype: string
      - name: corpus-id
        dtype: string
      - name: score
        dtype: float64
  - config_name: corpus
    features:
      - name: _id
        dtype: string
      - name: text
        dtype: string
  - config_name: queries
    features:
      - name: _id
        dtype: string
      - name: caption
        dtype: string
      - name: image_path
        dtype: string
# configs:
#   - config_name: default
#     data_files:
#       - split: test
#         path: "Pubmed/qrels/test.tsv"
#   - config_name: corpus
#     data_files:
#       - split: corpus
#         path: "Pubmed/corpus.jsonl"
#   - config_name: queries
#     data_files:
#       - split: queries
#         path: "Pubmed/queries.jsonl"
---
# M3Retrieve: Benchmarking Multimodal Retrieval for Medicine
Arkadeep Acharya, Akash Ghosh, Pradeepika Verma, Sriparna Saha, Kitsuchart Pashupa

Each of the folder contains 3 files ```queries.jsonl```, ```corpus.jsonl```, ```qrels/test.tsv``` containing the queries, corpus and relevancy mapping files respectively.
The images can be found in the ```images.zip``` file, where the images have been sorted based on their medical disciplines.