110M BERT masked LM
Captured config identifies BertForMaskedLM and Safetensors metadata reports 109951602 parameters.
Open Source Model Profile · m3rg-iitd
matscibert
matscibert is a 110M-parameter BERT fill-mask model from m3rg-iitd. Its model card documents materials-science training for text mining and information extraction.
- Publisher
- m3rg-iitd
- Task
- fill-mask
- Model type
- bert
- License
- mit
- Library
- transformers
- Publication status
- Accepted · not indexed
Model overview
matscibert is published by m3rg-iitd as a BERT-based fill-mask model. Safetensors metadata reports 109951602 parameters, and the captured configuration identifies BertForMaskedLM. According to the model card, it is a materials-domain language model trained on materials science research papers.
Recorded capabilities
Materials-domain focus
According to the model card, the model is a BERT model trained on materials science research papers for text mining and information extraction.
Documented materials corpus
The model card describes a corpus spanning alloys, glasses, metallic glasses, cement, and concrete from ScienceDirect.
Research code and citation
The model card says pretraining and fine-tuning code is shared on GitHub and gives a 2022 npj Computational Materials citation.
Use cases in the source record
- Masked-language modeling over materials-science text for text mining and information extraction workflows.
- Domain fine-tuning studies that start from a BERT model already exposed to alloys, glasses, cement, and concrete literature.
Limitations and unknowns
- No evaluation results were extracted from this record.
- No context-window value was extracted from this record.
- Provider state is historical snapshot data, not independently refreshed current availability.
- Corpus and methodology details come from the publisher model card and linked paper and have not been independently verified by Ethen.
Source and provenance
Source: m3rg-iitd/matscibert
Captured: Unknown. Processed: 2026-09-07T19:34:50.660945+00:00.
MatSciBERT A Materials Domain Language Model for Text Mining and Information Extraction This is the pretrained model presented in MatSciBERT: A materials domain language model for text mining and information extraction , which is a BERT model trained on material science research papers. The training corpus comprises papers related to the broad category of materials: alloys, glasses, metallic glasses, cement and concrete. We have utilised the abstracts and full text of papers(when available). All the research papers have been downloaded from ScienceDirect using the Elsevier API . The detailed methodology is given in the paper. The co…
F001F002F003F004F005F006F007F009F010F011F012F013F014