ML Research Wiki / Benchmarks / Molecule Captioning / ChEBI-20

ChEBI-20

Molecule Captioning Benchmark

Performance Over Time

📊 Showing 28 results | 📏 Metric: BLEU-2

Top Performing Models

Rank Model Paper BLEU-2 Date Code
1 BioT5+ BioT5+: Towards Generalized Biological Understanding with IUPAC Integration and Multi-task Tuning 59.10 2024-02-27 📦 QizhiPei/BioT5
2 BioT5 BioT5: Enriching Cross-modal Integration in Biology with Chemical Knowledge and Natural Language Associations 55.60 2023-10-11 📦 QizhiPei/BioT5
3 Text+Chem T5-augm-Base Unifying Molecular and Textual Representations via Multi-task Language Modelling 54.20 2023-01-29 📦 gt4sd/multitask_text_and_chemistry_t5
4 PEIT-GEN Property Enhanced Instruction Tuning for Multi-task Molecule Generation with Large Language Models 53.40 2024-12-24 📦 chenlong164/peit
5 MolCA, Galac1.3B MolCA: Molecular Graph-Language Modeling with Cross-Modal Projector and Uni-Modal Adapter 53.10 2023-10-19 📦 acharkq/molca
6 MolCA, Galac125M MolCA: Molecular Graph-Language Modeling with Cross-Modal Projector and Uni-Modal Adapter 52.90 2023-10-19 📦 acharkq/molca
7 MolReGPT (GPT-4-0314) Empowering Molecule Discovery for Molecule-Caption Translation with Large Language Models: A ChatGPT Perspective 52.50 2023-06-11 📦 phenixace/molregpt
8 LaMolT5-Large Automatic Annotation Augmentation Boosts Translation between Molecules and Natural Language 52.10 2025-02-10 📦 zhiqiangzhongddu/la3
9 MoMu+MolT5-Large A Molecular Multimodal Foundation Model Associating Molecule Graphs with Natural Language 51.50 2022-09-12 📦 ai-hpc-research-team/git-mol 📦 bingsu12/momu 📦 yangzhao1230/graphtextretrieval 📦 ai-hpc-research-team/slm4mol
10 Mol-LLM (SELFIES) Mol-LLM: Multimodal Generalist Molecular LLM with Improved Graph Utilization 51.50 2025-02-05 -

All Papers (28)