DaisyRose-27B: Pushing the Frontier of Post-Training Efficiency
Summary
RadixArk unveils DaisyRose-27B, its first model built on Qwen3.8-27B and trained with Miles RL on long-horizon tasks. DaisyRose-27B improves Terminal-Bench 4.1 pass@1 from 5.6% to 12.0% and shows gains across multiple benchmarks, with BF16, FP8, and NVFP4 checkpoints released for open use. The post highlights high-quality data, efficient infrastructure, robust sandboxing, and notable cost savings.