RL Training Infra Breakdown: Router Replay, Off-Policy Controls, Full-Vocab OPD on 40+ Teachers
nrehiew_ · x · 2026-09-11
nrehiew details the RL infra behind the new model: dispatch strategies eliminating long-tail stalls, router replay from previous checkpoints, dataset-level caps and discard schemes for short completions, off-policy ratio bounding with loss masking, persistent KV caches and routers on checkpoint updates, and full-vocab OPD distillation on over 40 teacher models at the final stage.
Related event: Leaked DeepSeek V4.1 Flash Tech Report Reveals KV Cache-Obsessed Flagship(44 posts)→
More from Infra
- AI chip startup Tenstorrent raises $1.5B at $6.2B valuation, more than doubling — Beth_Kindig · 2026-10-27
- Anthropic discloses up to $120.5B SpaceX compute deal, ~$3.1B monthly at full capacity — XFreeze · 2026-10-27
- Satire: swapping Raft for an LLM-style decision model tanks write throughput 99.98% — irombie · 2026-10-27
- Cadence joins the domain-specific AI model camp for chip design — karlfreund · 2026-10-27
- Qdrant builds Universal I/O to run vector search mmap code on object storage — qdrant_engine · 2026-10-27
- Nebius secures $3.5B Citi financing facility for GPU purchases and AI data centers — Beth_Kindig · 2026-10-26