[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"news-slug-ant-llada-image-open-generation-editing":3,"topics-all":38,"news-related-c480d2d0-9156-4aa7-826f-fba2f252b6b7":57},{"id":4,"title":5,"summary":6,"content":7,"original_url":8,"source_id":9,"tags":10,"translations":24,"news_slug":31,"published_at":32,"created_at":33,"modified_at":34,"is_published":35,"publish_type":36,"image_url":14,"view_count":37},"c480d2d0-9156-4aa7-826f-fba2f252b6b7","蚂蚁开源 LLaDA-Image:6B 参数生成编辑一体,Turbo 版 4 步出图","蚂蚁 inclusionAI 开源 6B 参数的统一图像生成与编辑模型 LLaDA-Image:单 checkpoint 同时支持文生图与指令编辑,Turbo 版蒸馏到 4 步出图,官方自报 Qwen-Image-Bench 英文 53.53、中文 53.38;训练代码暂未放出。","蚂蚁集团 inclusionAI 今日(9 月 4 日)开源 LLaDA-Image:一个 6B 参数的统一图像生成与编辑模型家族。Base 与 Turbo 两个 checkpoint 同步放出,BF16 和 FP8 双精度版本在 Hugging Face 与 ModelScope 同步上线,推理代码随 GitHub 仓库一并发布。\n\n## 一个 checkpoint 同时做生成和编辑\n\nLLaDA-Image 的核心卖点不在画质而在结构:单个 checkpoint 同时支持文生图与参考图指令编辑,不需要另挂一个编辑骨干网络。官方列出的能力覆盖文生图、VQ 条件生成、参考图编辑和中英双语文字渲染四种模式,展示样例也分写实图像、文字海报与指令编辑三类。\n\n技术路线上有两点值得注意。其一,文本骨干和 DiT 都是扩散模型,在统一框架内训练——这延续了蚂蚁 LLaDA 系列在扩散语言模型上的路线;VQ 模式直接用 LLaDA2 生成图像 VQ token,再经 SigVQ 嵌入进入扩散过程。其二,训练分阶段推进:先做纯图像预训练与中期训练建立视觉先验,之后才引入成对语言监督,最后进行生成与编辑的联合训练。\n\n## Turbo 版 4 步出图\n\nBase 模型推荐 50 步采样(guidance 5.0),Turbo 版经 Twin-DMD 蒸馏压到 4 步(guidance 1.0),官方称 2-4 步即可完成快速生成与编辑。跑分方面,官方报告在 Qwen-Image-Bench 上取得 53.53(英文)、53.38(中文)的总分,README 将其表述为 state-of-the-art——这是官方自报成绩,横向对比细节以技术报告为准。\n\n## 「Fully Open」还差最后一块\n\n论文标题写着 Fully Open Training Recipes,但仓库的开源计划显示:推理代码和模型权重已放出,训练代码仍标注 coming soon。也就是说,\"完全开放的训练配方\"目前兑现在权重与推理侧,想复现训练还要再等。技术报告已挂 arXiv(2609.03796),HF papers 页面上线一天拿到 84 个 upvote;工程实现基于 Python 3.11、PyTorch 2.8 与 Diffusers 0.39.0,官方致谢了字节 VeOmni 项目。\n\n对本地部署用户来说,6B 体量加上 FP8 版本,意味着消费级显卡就有机会跑起来;对行业观察者来说,LLaDA 这条扩散路线从语言模型延伸到图像生成,是比单次发布更值得留意的信号。等训练代码真正落地,「fully open」才算闭环。\n\n参考:github.com\u002FinclusionAI\u002FLLaDA-Image;huggingface.co\u002FinclusionAI\u002FLLaDA-Image;arXiv 2609.03796","https:\u002F\u002Fgithub.com\u002FinclusionAI\u002FLLaDA-Image","e7a2ac45-ed74-4858-8663-3bd2943959a2",[11,15,18,21],{"id":12,"name":13,"slug":13,"description":14,"color":14},"7b67033c-19e6-4052-a626-e681bba64c7a","diffusion",null,{"id":16,"name":17,"slug":17,"description":14,"color":14},"7e89b5cc-57db-4f37-bc6d-28919a73931c","model-release",{"id":19,"name":20,"slug":20,"description":14,"color":14},"b9bd9039-fcdb-41a8-b85b-fc1587def2b9","open-source",{"id":22,"name":23,"slug":23,"description":14,"color":14},"c883fd20-1d66-4fb7-9fc7-320fa7f87023","text-to-image",[25],{"id":26,"lang":27,"title":28,"summary":29,"content":30},"0f5034ed-ff79-4049-be59-343c236f7284","en","LLaDA-Image: Ant's 6B open model unifies generation and editing","Ant's inclusionAI open-sources LLaDA-Image, a 6B model that generates and edits images in one checkpoint; the distilled Turbo runs in 4 steps.","Ant Group's inclusionAI open-sourced LLaDA-Image today (September 4): a 6B-parameter unified image generation and editing model family. Base and Turbo checkpoints shipped together, with BF16 and FP8 builds on both Hugging Face and ModelScope, plus inference code in the GitHub repository.\n\n## One checkpoint for both generation and editing\n\nThe core selling point is structure rather than raw fidelity: a single checkpoint handles text-to-image generation and reference-image, instruction-guided editing, with no separate editing backbone. Officially listed capabilities span four modes: text-to-image, VQ-conditioned generation, reference-image editing, and Chinese-English text rendering. The showcase samples cover photorealistic images, text posters, and instruction editing.\n\nTwo technical choices stand out. First, the text backbone and the DiT are both diffusion models trained in a unified framework — extending Ant's LLaDA line of diffusion language models into vision; in VQ mode, LLaDA2 produces image VQ tokens that SigVQ embeds before diffusion. Second, training is staged: image-only pre-training and mid-training build the visual prior first, paired language supervision comes later, and joint generation-editing training finishes the recipe.\n\n## The Turbo variant generates in 4 steps\n\nThe Base model recommends 50 sampling steps (guidance 5.0); the Turbo variant, distilled with Twin-DMD, compresses this to 4 steps (guidance 1.0), with the team citing 2-4 steps for fast generation and editing. On scores, the official report lists 53.53 (English) and 53.38 (Chinese) overall on Qwen-Image-Bench, described in the README as state-of-the-art — a self-reported result; cross-model comparison details sit in the technical report.\n\n## \"Fully open\" is missing one piece\n\nThe paper is titled \"Building Strong Image Generators with Fully Open Training Recipes,\" yet the repository's opensource plan shows inference code and weights released while training code is still marked \"coming soon.\" In other words, the \"fully open training recipe\" currently covers weights and inference; reproducing training means waiting. The technical report is on arXiv (2609.03796), and its Hugging Face papers page picked up 84 upvotes within a day. The implementation uses Python 3.11, PyTorch 2.8, and Diffusers 0.39.0, and the team credits ByteDance's VeOmni project.\n\nFor local-deployment users, a 6B model with an FP8 build puts it within reach of consumer GPUs. For industry watchers, the LLaDA diffusion lineage stretching from language models into image generation is a stronger signal than any single release. The \"fully open\" loop closes only when the training code lands.\n\nReferences: github.com\u002FinclusionAI\u002FLLaDA-Image; huggingface.co\u002FinclusionAI\u002FLLaDA-Image; arXiv 2609.03796","ant-llada-image-open-generation-editing","2026-09-04T13:09:21Z","2026-09-04T13:09:33.623707Z","2026-09-04T13:09:33.623717Z",true,"agent",612,[39,48],{"slug":40,"tag_slug":40,"title_zh":41,"title_en":42,"intro_zh":43,"intro_en":44,"id":45,"is_active":35,"created_at":46,"modified_at":47},"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":49,"tag_slug":49,"title_zh":50,"title_en":51,"intro_zh":52,"intro_en":53,"id":54,"is_active":35,"created_at":55,"modified_at":56},"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":58},[59,64,69,74,79,84],{"id":60,"title":61,"news_slug":62,"published_at":63},"095917eb-02ae-4fd2-a1cb-17d0805442ee","微软 Mage-Flow 用 4B 跑赢 32B：原生分辨率 + 三件套协同设计把生成编辑都塞回单卡","microsoft-mage-flow-4b","2026-07-23T03:30:00+00:00",{"id":65,"title":66,"news_slug":67,"published_at":68},"f7287cac-6643-4f4a-8cbd-2b281d2d4d46","Krea 2 开源双发：12B DiT 把「2 秒出图」做进主流程，蒸馏后 8 步直出 2K","krea-2-12b-dit-2-second-turbo","2026-06-25T10:30:00+00:00",{"id":70,"title":71,"news_slug":72,"published_at":73},"72ee21ea-8a91-4dd3-88fa-f605551ff9ce","Qwen-Image-2.0 发布：7B 拿下原生 2K，把「图文一体 + 生成编辑统一」推到开源前沿","qwen-image-2-0-7b-native-2k-arena-no1","2026-06-18T08:00:00+00:00",{"id":75,"title":76,"news_slug":77,"published_at":78},"cd88ab8f-afff-4f8f-8edc-ab24715906c6","FLUX.2 [klein] 4B\u002F9B 发布：统一生图编辑，Apache 2.0","flux-2-klein-4b-9b-apache-2-sub-second","2026-06-12T06:30:00+00:00",{"id":80,"title":81,"news_slug":82,"published_at":83},"04d03b80-0a32-4ea1-87df-9248b36653c1","Ideogram 4.0 开源：9.3B 单流 DiT + Qwen3-VL 文本编码器，把排版与文字渲染做到开源第一","ideogram-4-0-9-3b-dit-qwen3-vl-text","2026-06-09T12:00:00+00:00",{"id":85,"title":86,"news_slug":87,"published_at":88},"6082cd23-0eca-40e0-9315-67318dc818ee","NovelAI Diffusion V5 发布:规模翻倍、32 通道 VAE,单次生成整页漫画","novelai-diffusion-v5-release","2026-08-22T13:10:00+00:00"]