Xiaomi MiMo open-sources MiMo-V2.6-RL-oss training dataset on Hugging Face under Apache-2.0

iScienceLuvr · x · 2026-09-26

Xiaomi's MiMo team has released its reinforcement learning dataset MiMo-V2.6-RL-oss on Hugging Face under the permissive Apache-2.0 license, a move the community is calling refreshingly open. The dataset spans text, image, and document modalities in parquet format with a few thousand samples, including real code test-case examples for RL training. It's a directly usable resource for anyone reproducing or studying post-training pipelines.

Related event: Xiaomi Open-Sources MiMo-V2.6 RL Dataset, Tops HuggingFace Trending(5 posts)→

Original post →

More from Research

Research channel →