# Compact Model

An 800M-parameter vision-language model from H2O.ai, specializing in OCR and document understanding with excellent performance

SAM2-Hiera Small is a compact variant of Meta's SAM2 model for efficient mask generation tasks.

Image classification model fine-tuned on CIFAR-100 dataset based on Swin Transformer Tiny architecture

A Q&A model fine-tuned on the SQuAD2.0 dataset based on the BERT-Medium architecture, designed for resource-constrained environments

This model is a PyTorch pretrained model obtained by converting TensorFlow checkpoints from the official Google BERT repository, originating from the paper 'Well-Read Students Learn Better: On the Importance of Pre-training Compact Models,' and trained on the MNLI dataset.