# Pruning Optimization

Sheared-LLaMA-2.7B is a lightweight language model derived from Llama-2-7b through pruning and continued pretraining, consuming only a 50B token budget.

A question-answering model fine-tuned and optimized on SQuAD v1 based on BERT-base uncased, retaining 43% of original weights through pruning techniques, achieving 1.96x faster inference speed

A Q&A system model fine-tuned on SQuAD v1 based on the BERT-base uncased model, optimized via the nn_pruning library, achieving 2.01x faster inference speed and a 0.69 improvement in F1 score.

This model is a text classification model fine-tuned from bert-base-uncased on the SST-2 dataset, optimized through pruning with the nn_pruning library, retaining 51% of the original model's weights while achieving 91.17% accuracy.