IBM open sources Granite 4.2 reasoning models with full training recipe
krvarshney · x · 2026-08-28
IBM released the Granite 4.2 reasoning model family (3B, 8B, 30B) under Apache 2.0, including the full training recipe—data mixture proportions, hyperparameters, RL environments, and chat templates. The models were pre-trained on 15T tokens with 512K context and post-trained via multi-stage RL in real software engineering, terminal, and web search environments. They feature three modes: thinking (default), non-thinking, and a low-effort mode.
More from Models
- Users face 'Forbidden' errors when downloading MiniMax H3 — musashiasano · 2026-08-28
- User Opinion: GPT Sol Would Be the Best Model With Better Front-End Skills — BLUECOW009 · 2026-08-28
- Anima-3.8B Trends on Hugging Face — lylogummy · 2026-08-28
- Evolution strategies beat GRPO on reasoning diversity and Pass@K, paper finds — Yunpeng Ba · 2026-08-28
- Opinion: Qwen's n-gram innovation pops the AI bubble — Acrobatic_Stress1388 · 2026-08-28
- Agnes 2.5 Pro Beta Joins Frontier Tier with Major Agentic Gains, Higher Cost — ArtificialAnlys · 2026-08-28