RadixArk open-sources DaisyRose-27B: 18 days on 144 H200s, Terminal-Bench 5.6% to 12.0%
BanghuaZ · x · 2026-11-19
At SGLang Summit, RadixArk launched DaisyRose-27B, its first open-weight model: RL training on Qwen3-27B with 1,200 curated long-horizon agent tasks over 18 days on 144 H200 GPUs. Using Miles and its in-house training backend, the team stretched agent sessions up to 1M tokens and lifted Terminal-Bench 4.1 from 5.6% to 12.0%. The Day 0 pin shown was a summit Easter egg.
More from coding & agent
- GPT-6.2 Rigs and Animates a 3D Game Character End-to-End, Dev Puts It Straight Into Production — OpenAIDevs · 2026-11-19
- Developer Builds a Planet-Themed Dashboard to Monitor All His Agents at a Glance — kevinkern · 2026-11-19
- SpaceX engineer launches Grok Bots directory with 100 community-built agents — VibeMarketer_ · 2026-11-19
- Open-source MCP server controls Logitech Harmony hubs locally with LLM-safe guardrails — SDNick484 · 2026-11-19
- AWS case study: Social Department's QuickSearch finds memeable moments across video libraries — AWS ML Blog · 2026-11-19
- Elicit on agent swarms: pharma winners of 2036 will robustly deploy them without going rogue — jungofthewon · 2026-11-19