lmsys/mt_bench_human_judgments download history

lmsys/mt_bench_human_judgments is a question answering dataset on the Hugging Face Hub. In the last 30 days it was downloaded 13,922 times (10,906 in the last 7 days), and 119,371 times in total. It ranks #2,999 among datasets by monthly downloads.

Content This dataset contains 3.3K expert-level pairwise human preferences for model responses generated by 6 models in response to 80 MT-bench questions. The 6 models are GPT-4, GPT-3.5, Claud-v1, Vicuna-13B, Alpaca-13B, and LLaMA-13B. The annotators are mostly graduate students with

Models trained on mt_bench_human_judgments

3 models list it as training data.

Downloads by month

Downloads in the 30 days up to the end of each month, and the change from the month before.

MonthDownloadsChange
Up to 2026-11-1813,922+229%
October 20264,229+89%
September 20262,236+1%
August 20262,212-16%
July 20262,649+10%
June 20262,407-35%
May 20263,694+35%
April 20262,730+108%
March 20261,311+138%
February 2026552+9%
January 2026506-41%
December 2025852-42%

Its busiest 30 days so far ended on 2026-11-18, with 13,922 downloads.

Open lmsys/mt_bench_human_judgments on Hugging Face

Sister project: Paper Pulse, the upvote history of every Hugging Face Daily Paper.