Modular Multi-Task Learning for Chemical Reaction Prediction
Paper • 2602.10404 • Published
This is the task-specific LoRA fine-tuning adapter using the C-H dataset. To run inference of the C-H functionalisation, you will need to load the LoRA adapter + the base model byt5-small-finetuned-uspto-1k-tpl-chemical-reactions/
For further info, see https://github.com/jiayunpang/chem-peft
Reference Jiayun Pang, Ahmed M. Zaitoun, Xacobe Couso Cambeiro and Ivan Vulić. Modular Multi-Task Learning for Chemical Reaction Prediction. (2026) arXiv preprint https://arxiv.org/abs/2602.10404
Inference Providers NEW
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support