metadata

license: mit
library_name: transformers
pipeline_tag: fill-mask
tags:
  - cheminformatics
  - ChemBERTa
  - masked-lm
  - roberta

ChemBERTa-100M-MLM

ChemBERTa model pretrained on a subset of 100M molecules from ZINC20 dataset using masked language modeling (MLM).

Usage

from transformers import AutoTokenizer, AutoModelForMaskedLM

tokenizer = AutoTokenizer.from_pretrained("DeepChem/ChemBERTa-100M-MLM")
model = AutoModelForMaskedLM.from_pretrained("DeepChem/ChemBERTa-100M-MLM")