mit-han-lab/QServe-benchmarks

QServe benchmarks This huggingface repository contains configurations and tokenizer files for all models benchmarked in our QServe project: Llama-3-8B Llama-2-7B Llama-2-13B Llama-2-70B Llama-30B Mistral-7B Yi-34B Qwen1.5-72B Please clone this repository if you wish to run our QServe benchmark code without cloning full models. Please consider citing our paper if it is helpful: @article{lin2024qserve, title={QServe: W4A8KV4 Quantization and System Co-design for Efficient LLM… See the full description on the dataset page: https://huggingface.co/datasets/mit-han-lab/QServe-benchmarks.

<iframe src="https://silytics.si/embed/item/e4f48026-eae9-499a-9108-e661bb00f303" width="420" height="190" style="border:0" loading="lazy" title="Silytics entry"></iframe>

Silytics shows public metadata from Hugging Face and links to the original. Information may be incomplete or out of date. Last refreshed 17 October 2026. Licenses and terms of the original apply to the work itself. Not financial, legal or professional advice.