Xiaomi Mimo 2.6 live post-training dashboard
- ID
- 25400
- Status
- summarized
- Published
- 17 Sep 2026, 4:09 AM
- Fetched
- 18 Sep 2026, 9:04 PM
- Provider
- Hacker News
- Category
- dev-community
- Original URL
- https://mimo.xiaomi.com/rl/
- Source URL
- https://hnrss.org/best
Summary
- Score
- 7.0
- Created
- 18 Sep 2026, 9:06 PM
- Tags
- Audience
- ai_ml_learnersdevelopersai_agent_users
What happened
Xiaomi is broadcasting a live dashboard of reinforcement-learning post-training for two Mimo models: v2.6-pro (at step 12-13, $829K spent so far, 25.1B total tokens) and v2.6-flash (at step 15-16, $365K spent, 35.2B total tokens). The dashboard exposes granular metrics including pass rates, entropy loss, gradient norms, KL divergence, per-step timing (~2h 30m per step for pro), and real-time infra errors like VRAM issues and undetected dataset errors requiring restarts.
Why it matters
This is one of the few public, real-time windows into what production-scale RL post-training actually costs and breaks at—$1.19M total and counting, with concrete failure modes (VRAM crashes, dataset infra errors) that forced restarts. If you are building or evaluating RL fine-tuning pipelines, study the metric definitions and failure logs here as a reference for what to instrument: pass rates per step, KL between inference and training distributions, staleness tracking, and infra error rates are all exposed.
Discussion angle
What this dashboard reveals about the real economics and failure modes of RL post-training—and which of these metrics (KL divergence, pass rate, staleness) you should be tracking if you ever do RL fine-tuning on your own agents.