ESM-2 is a protein language model optimized by NVIDIA using the TransformerEngine library based on the original ESM-2 model from Facebook Research. It has the same weights and outputs within the numerical precision range and is suitable for protein sequence analysis tasks.

Scientific ComputingTransformers

AMPLIFY is an efficient protein language model optimized using NVIDIA's TransformerEngine library. It can generate residue and protein embeddings, suggest mutations, and distinguish disordered proteins from non - protein sequences. This model maintains the same weights and outputs as the original model within the numerical precision range and can be used for both commercial and non - commercial purposes.

Scientific ComputingTransformersEnglish

An ultra-lightweight English text line recognition model developed by the PaddleOCR team, supporting the recognition of English and numeric characters

An ultra-lightweight Korean text line recognition model that supports the recognition of Korean and numeric characters, with an average accuracy of 60.21%.

An ultra-lightweight Japanese text line recognition model developed by the PaddleOCR team, supporting the recognition of Japanese and numeric characters.

An ultra-lightweight Latin text line recognition model developed by the PaddleOCR team, supporting the recognition of Latin and numeric characters.

An ultra-lightweight Arabic letter recognition model developed by the PaddleOCR team, supporting the recognition of Arabic letters and numeric characters.

ChatTS-14B is a large language model fine-tuned from Qwen2.5-14B-Instruct, specializing in understanding and reasoning about time series data with a focus on morphological and numerical feature analysis.

MultimodalTransformersMultiple Languages

OCR model optimized for printed digits and timestamp-like numerical values, achieving CER=0 recognition accuracy on single-line image test sets

Computer VisionTransformersEnglish

A Portuguese named entity recognition model based on Bertimbau-large, specifically optimized for Brazilian Portuguese, capable of recognizing 5 entity types: person, organization, location, time, and numerical value.

Natural Language ProcessingTransformersOther

An OCR model specifically designed for Persian language, based on CNN+LSTM architecture, optimized for printed/scanned documents, supporting numeric and special character recognition.

MatCha model fine-tuned on the PlotQA-v1 dataset, specializing in visual question answering tasks for charts, with excellent performance in chart deconstruction and numerical reasoning

MultimodalTransformersMultiple Languages

A vision-language model fine-tuned on the Chart2text-statista dataset, excelling in chart comprehension and numerical reasoning

MultimodalTransformersMultiple Languages

MatCha is a vision-language model based on the Pix2Struct architecture, specifically optimized for chart comprehension and numerical reasoning tasks, excelling in chart-based question answering.

MultimodalTransformersMultiple Languages

polyBERT is a chemical language model designed to achieve fully machine-driven ultrafast polymer informatics. It maps PSMILES strings into 600-dimensional dense fingerprints to numerically represent polymer chemical structures.

Natural Language ProcessingTransformers

A tabular data classification model based on Transformer architecture, supporting mixed input of numerical and categorical features

Scientific ComputingTensorflow

SEC-BERT-NUM is a BERT model tailored for the financial domain, which unifies numerical expressions by replacing numbers with [NUM] pseudo-tokens to avoid numerical fragmentation issues.

Natural Language ProcessingTransformersEnglish

A table question answering model based on BERT architecture, enhanced with intermediate pretraining for numerical reasoning, fine-tuned on the SQA dataset

Natural Language ProcessingTransformersEnglish

A BERT variant for the financial domain, preserving financial data integrity through numeric morphology pseudo-token processing

Natural Language ProcessingTransformersEnglish