We show for the first time that large-scale generative pretrained transformer (GPT) family models can be pruned to at least 50% sparsity in one-shot, without any retraining, at minimal loss of accuracy. [...] We can execute SparseGPT on the largest available open-source models, OPT-175B and BLOOM-176B, in under 4.5 hours, and can reach 60% unstructured sparsity with negligible increase in perplexity: remarkably, more than 100 billion weights from these models can be ignored at inference time.
Recent articles
- llm cmd undo last git commit - a new plugin for LLM - 26th March 2024
- Building and testing C extensions for SQLite with ChatGPT Code Interpreter - 23rd March 2024
- Claude and ChatGPT for ad-hoc sidequests - 22nd March 2024
- Weeknotes: the aftermath of NICAR - 16th March 2024
- The GPT-4 barrier has finally been broken - 8th March 2024
- Prompt injection and jailbreaking are not the same thing - 5th March 2024