Compress large language models (LLMs) with SparseGPT to make your machine learning inference fast and efficient. Prune in one-shot with minimal accuracy loss.