做大规模LoRA训练和推理的团队终于有了正经的工程方案——MinT解决了策略数量爆炸时的资源浪费问题，用适配器分离和调度大幅降低成本，搞大模型服务的建议点开看看。
MinT（MindLab Toolkit）是一个专为低秩适配（LoRA）后训练和在线推理设计的托管基础设施系统。它针对在少量昂贵基座模型上产生大量训练策略的场景，通过保持基座模型常驻内存，仅移动导出的LoRA适配器，避免了合并完整检查点的开销。MinT沿三个维度扩展：向上支持超过1T总参数的前沿密集和MoE架构；向下实现适配器仅占基座模型1%以下大小，在4B密集模型上步骤时间减少18.3倍；向外支持百万级可寻址策略目录和千级适配器并发波次。该系统使得在共享的1T级基座模型上训练和推理数百万LoRA策略成为可能。
MinT: Managed Infrastructure for Training and Serving Millions of LLMs
We present MindLab Toolkit (MinT), a managed infrastructure system for Low-Rank Adaptation (LoRA) post-training and online serving. MinT targets a setting where many trained policies are produced over a small number of expensive base-model deployments. Instead of materializing each policy as a merged full checkpoint, MinT keeps the base model resident and moves exported LoRA adapter revisions through rollout, update, export, evaluation, serving, and rollback, hiding distributed training, serving, scheduling, and data movement behind a service interface. MinT scales this path along three axes. Scale Up extends LoRA RL to frontier-scale dense and MoE architectures, including MLA and DSA attention paths, with training and serving validated beyond 1T total parameters. Scale Down moves only the exported LoRA adapter, which can be under 1% of base-model size in rank-1 settings; adapter-only handoff reduces the measured step by 18.3x on a 4B dense model and 2.85x on a 30B MoE, while concurrent multi-policy GRPO shortens wall time by 1.77x and 1.45x without raising peak memory. Scale Out separates durable policy addressability from CPU/GPU working sets: a tensor-parallel deployment supports 10^6-scale addressable catalogs (measured single-engine sweeps through 100K) and thousand-adapter active waves at cluster scale, with cold loading treated as scheduled service work and packed MoE LoRA tensors improving live engine loading by 8.5-8.7x. MinT thus manages million-scale LoRA policy catalogs while training and serving selected adapter revisions over shared 1T-class base models.