Datasets:
Update README.md
Browse files
README.md
CHANGED
@@ -6,4 +6,31 @@ task_categories:
|
|
6 |
tags:
|
7 |
- chemistry
|
8 |
- biology
|
9 |
-
---
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
6 |
tags:
|
7 |
- chemistry
|
8 |
- biology
|
9 |
+
---
|
10 |
+
|
11 |
+
# ChEBI-20-MM Dataset
|
12 |
+
|
13 |
+
## Overview
|
14 |
+
|
15 |
+
The ChEBI-20-MM is an extensive and multi-modal benchmark developed from the ChEBI-20 dataset. It is designed to provide a comprehensive benchmark for evaluating various models' capabilities in the field of molecular science. This benchmark integrates multi-modal data, including InChI, IUPAC, SELFIES, and images, making it a versatile tool for a wide range of molecular tasks.
|
16 |
+
|
17 |
+
## Dataset Description
|
18 |
+
|
19 |
+
ChEBI-20-MM is an expansion of the original ChEBI-20 dataset, with a focus on incorporating diverse modalities of molecular data. This benchmark is tailored to assess models in several key areas:
|
20 |
+
|
21 |
+
- **Molecule Generation**: Evaluating the ability of models to generate accurate molecular structures.
|
22 |
+
- **Image and IUPAC Recognition**: Testing models on their proficiency in interpreting and converting molecular images and IUPAC names into other representational formats.
|
23 |
+
- **Molecular Captioning**: Assessing the capability of models to generate descriptive captions for molecular structures.
|
24 |
+
- **Retrieval Tasks**: Measuring the effectiveness of models in retrieving molecular information accurately and efficiently.
|
25 |
+
|
26 |
+
## Utility and Significance
|
27 |
+
|
28 |
+
By expanding the data modality variety, this benchmark enables a more comprehensive evaluation of models' performance in multi-modal data handling. It provides an opportunity to explore the intersection of molecular science and advanced computational models.
|
29 |
+
|
30 |
+
## How to Use
|
31 |
+
|
32 |
+
Model reviews and evaluations related to this dataset can be directly accessed and used via the LLM4Mol link: [LLM4Mol](https://github.com/AI-HPC-Research-Team/LLM4Mol).
|
33 |
+
|
34 |
+
## Acknowledgments
|
35 |
+
|
36 |
+
The development of the ChEBI-20-MM dataset was inspired by the ChEBI-20 in molecule generation and captioning initiated by MolT5. Additional data information supplements are derived from PubChem.
|