Back to News

vLLM Achieves Bitwise-Identical Logprobs for RL Rollout and Training Forward

#vllm#rl#logprobs#nvfp4

The vLLM blog reports that its rollout and forward training paths now produce bitwise identical logprobs on DeepSeek-V4-Flash and Nemotron-3.5-Lightning-30B-A3B with NVFP4 quantization. The implementation retains 91% of rollout throughput on DeepSeek-V4-Flash, and the agent skill has been open-sourced.

Coverage timeline

  1. vLLM BlogYan Bai (NVIDIA DevTech), Ao Shen (Inferact)
    Finish RL Training Faster: Bitwise-Consistent, High-Performance, Agent-Driven

    We make vLLM rollout and the training forward give bitwise identical logprobs on DeepSeek-V4-Flash and Nemotron-3.5-Lightning-30B-A3B (NVFP4), keep 91% of rollout throughput on DeepSeek-V4-Flash, and open-source the agent skill.

Related stories

  • vLLM Boosts DeepSeek-V4.1-Flash: 1.9x Low-Concurrency Speed, 5x Agentic Throughput

    According to a vLLM blog post dated October 7, 2026, within three weeks of DeepSeek-V4.1-Flash's release, vLLM optimizations made the model 1.9x faster at low concurrency and increased its throughput 5x on the SemiAnalysis AgentX benchmark. The improvements are attributed to SWA bounded replay, CUDA graphs, DeepSeek's new kernels, and vLLM kernel fusions.

    1 reports · 1 sources

  • Liquid AI releases LFM2.5-24B-A2B, an open-weight 24B reasoning model with 2B active parameters

    Liquid AI announced LFM2.5-24B-A2B, an open-weight 24B reasoning model with 2B active parameters, designed for fast, private on-device agents. The company claims it outperforms Qwen3.6-35B-A3B on the LiveClaw benchmark.

    2 reports · 2 sources284 points · 78 comments

  • DeepSeek details DSec sandbox system for agent training, 5000+ sandboxes/sec

    DeepSeek published a system paper, co-authored by Liang Wenfeng, detailing DSec (DeepSeek Elastic Compute), a production sandbox infrastructure for large-scale agentic RL training and evaluation. A single production unit comprises about 160 CPU nodes, 30,000 cores, and 250 TB DRAM, generating over 5,000 sandboxes per second, up to 3 million daily, with over 380,000 concurrent at peak. The paper covers four backend types (FnCall, Container, MicroVM, Full) and states that sandbox workloads from DeepSeek V3.2 to V4.1 run on DSec.

    2 reports · 2 sources

  • vLLM Adds Day-0 Support for NVIDIA Vera Rubin NVL72, Claims 7.8x Throughput over GB200

    vLLM announced support for NVIDIA Vera Rubin NVL72, claiming 7.8x throughput over GB200 NVL72. The platform offers daily container builds and day-0 model support for DeepSeek, Moonshot AI, Z.ai, and MiniMax, leveraging Blackwell kernel compatibility and Rubin-tuned kernels via FlashInfer 0.7.0.

    1 reports · 1 sources

News updates every hour. Subscribe in your RSS reader to get new stories automatically.

Subscribe