[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"news-slug-relora-base-model-lora-reuse-89pct-time":3,"topics-all":36,"news-related-4326dbe6-f7c1-4ce8-8ef1-8cd7aa1cbb97":55},{"id":4,"title":5,"summary":6,"content":6,"original_url":7,"source_id":8,"tags":9,"translations":23,"news_slug":29,"published_at":30,"created_at":31,"modified_at":32,"is_published":33,"publish_type":34,"image_url":13,"view_count":35},"4326dbe6-f7c1-4ce8-8ef1-8cd7aa1cbb97","ReLoRA：基础模型频繁更新下的LoRA适配器复用之道","大模型即服务（LLM-as-a-Service）模式已成为主流，但基础模型频繁更新让下游LoRA适配器面临两难：重新训练成本高，直接迁移到新基座又容易掉点。arXiv新论文ReLoRA提出一套知识复用框架来解决这个矛盾。\n\nReLoRA包含两步核心优化：第一步用贝叶斯优化融合旧适配器知识与新模型的演化信息，生成兼容性更高的初始化点；第二步采用先强后弱的正则化微调策略——先用高强度正则快速将适配器拉回高性能区域，再用低正则进行任务精细化。\n\n实验数据显示，ReLoRA相比从零训练减少89%的重新适配时间，同时精度提升4.6%。该工作的核心价值在于承认了一个现实：不是每个下游服务都有能力在基础模型更新后从零微调。\n\nReLoRA本质上是把旧适配器当作新任务的先验知识而非简单丢弃。对拥有大量下游模型的厂商（尤其是多租户SaaS场景），这意味着能以更低成本跟进最新基座。\n\n对从业者的启发是：微调资源有限时，与其每次从零训练，不如思考如何把历史适配器的积累知识迁移到新模型。ReLoRA框架已开源，详见arXiv:2606.02606。","https:\u002F\u002Farxiv.org\u002Fabs\u002F2606.02606","7437aeb9-930c-4866-a2e9-48003c1a792b",[10,14,17,20],{"id":11,"name":12,"slug":12,"description":13,"color":13},"e676a5cf-1f24-472f-a765-86fa21a1bc3c","ai-model",null,{"id":15,"name":16,"slug":16,"description":13,"color":13},"0ef8513a-0a26-42f0-b6f9-5b6dadded45c","efficiency",{"id":18,"name":19,"slug":19,"description":13,"color":13},"0a93ec8e-ea39-4693-81de-563ca8c173f7","inference",{"id":21,"name":22,"slug":22,"description":13,"color":13},"01598627-1ea6-4b27-a5d8-874971571a71","llm",[24],{"id":25,"lang":26,"title":27,"summary":28,"content":13},"4884563b-7b72-4fa8-9b78-dec1b0111dec","en","ReLoRA: the LoRA adapter reuse path under frequent base-model updates","arXiv 2606.02606 introduces ReLoRA, addressing a practical problem: when the base model updates frequently (weekly or even daily), the LoRA adapters fine-tuned for the old base become useless. ReLoRA enables \"incremental LoRA training\" — adapters trained on older bases can be efficiently ported to new bases with minimal retraining, saving the cost of full re-fine-tuning.","relora-base-model-lora-reuse-89pct-time","2026-06-03T08:10:00Z","2026-06-03T16:07:45.087404Z","2026-08-19T02:08:40.142862Z",true,"agent",160,[37,46],{"slug":38,"tag_slug":38,"title_zh":39,"title_en":40,"intro_zh":41,"intro_en":42,"id":43,"is_active":33,"created_at":44,"modified_at":45},"ai-for-science","AI for Science 2026：从 UniPert 到 GPT-Rosalind 的硬核进化","AI for Science 2026: from UniPert to GPT-Rosalind","生命科学、化学材料、物理世界模型——AI 正在从\"语言工具\"变成\"实验伙伴\"。本专题收录 AI 在三大科学方向的关键节点：UniPert 统一基因与化学扰动空间、GPT-Rosalind 端到端生命科学推理、达摩院 AI 智能体 28 小时找到 4 种超导新材料、Anthropic Claude Science 把工作台做成标准品。","From language tool to lab partner — AI is reshaping life sciences, chemistry\u002Fmaterials, and physical world models. This topic covers the key milestones: UniPert unifying genetic-chemical perturbation spaces, GPT-Rosalind's end-to-end life-sciences reasoning, DAMO's AI agent discovering 4 superconducting materials in 28 hours, and Anthropic's Claude Science workbench going mainstream.","988a4300-5fab-41c4-b5d8-63711a2dc757","2026-09-10T01:34:15.296649Z","2026-09-10T01:34:15.296663Z",{"slug":47,"tag_slug":47,"title_zh":48,"title_en":49,"intro_zh":50,"intro_en":51,"id":52,"is_active":33,"created_at":53,"modified_at":54},"h3-series","MiniMax H3 系列：从开源权重到 35 倍吞吐","MiniMax H3 Series: from open weights to 35x throughput","MiniMax H3 自 2026 年 8 月开源以来节奏密集：官方把生成、参考与编辑收回一个模型；ComfyUI 当天压进 RTX 3060；摩尔线程 3 小时完成国产 GPU 适配；fal 后训练版把吞吐拉到 35 倍；FastH3 蒸馏再砍推理成本。本专题持续追踪 H3 的发布—开源—蒸馏—部署全链路。","Since MiniMax open-sourced H3 in August 2026 the pace has been relentless: one unified omni-modal model, same-day ComfyUI support down to an RTX 3060, a 3-hour Day-0 port to Moore Threads GPUs, fal's post-trained H3 Max at 35x throughput, and FastH3 distillation cutting inference cost further. This topic tracks the full H3 chain — release, open weights, distillation, deployment.","83ef0daa-3c31-4cb3-86ed-e5ee58654d5f","2026-09-08T07:33:19.942193Z","2026-09-08T07:33:19.942209Z",{"items":56},[57,62,67,72,77,82],{"id":58,"title":59,"news_slug":60,"published_at":61},"a0db4732-7b67-45eb-b3df-92abc3cd61ad","SAP斥资逾10亿欧元收购Prior Labs：Tabular Foundation Models能否重塑企业AI","sap-prior-labs-tfm-1b-euro-acquisition","2026-05-05T17:30:00+00:00",{"id":63,"title":64,"news_slug":65,"published_at":66},"c3814f7d-2649-4660-a798-28fb03aa2b6d","SwitchSD 让投机解码学会「该抄才抄」:读内部信号,EAGLE3 之上再快 15%","switchsd-copy-intent-speculative-decoding","2026-09-20T23:09:25+00:00",{"id":68,"title":69,"news_slug":70,"published_at":71},"3cecce90-70b9-4bb3-b9b7-93e6b0c05105","D-Quant 用熵编码压 KV:2.26bit 近无损","d-quant-entropy-coding-kv-cache","2026-09-20T17:10:42+00:00",{"id":73,"title":74,"news_slug":75,"published_at":76},"813ad679-51dd-43d7-afcc-0baf48d2ef5f","When2Think:推理模型该想多久,先看题有多难","when2think-difficulty-aware-length-control","2026-09-19T19:08:00+00:00",{"id":78,"title":79,"news_slug":80,"published_at":81},"9a3cd449-e29a-4730-814b-f1be5c2685c6","复旦FFD让Flash Attention退役？11.6× kernel提速把长上下文推到256K","fudan-ffd-long-context-attention-sparsity","2026-09-15T07:15:46+00:00",{"id":83,"title":84,"news_slug":85,"published_at":86},"2638aeac-dc4d-4b73-b7fe-2b042015adee","OreoLook 开源:三层缓存把 AI 搜索搬进 8 核 CPU,重复问题 0.1 毫秒出答案","oreolook-three-layer-cpu-cache","2026-09-10T23:08:36+00:00"]