Keep on to blur preview images; turn off to show them clearly
Spending 40 years in a job you dislike so you can retire and finally live is the worst deal in human history. I'd rather fail 100x starting my own thing than spend half my life feeling resentful.
The $10M Solopreneur | Helping 100,000+ burned-out corporate professionals build six-figure, one-person online businesses at https://t.co/27OAdtwjh5
RT @polymutex: Just wanted to give a public shout-out to @wslyvh. He has helped me navigate the whole conference/side-event meta, and answe…
Privacy @PrivacyEthereum | Events @ethereum | OSS 🛠️ Indie Maker @eth_gastracker https://t.co/xNlTAwRRp1
Sharing the playbook that turned years of failure into an $8m acquisition and multiple $100k+ MRR SaaS creations actionable growth tactic behind-the-scene stories no sponsors • no ads • always free Join 60k+ founders & makers here → https://t.co/U4V7DEKzYG See you in your inbox 👇
Built Tweet Hunter, Taplio (sold $8m) Growing https://t.co/OyNJ8ZUyOh - https://t.co/jS9GQJ5Ps8 - https://t.co/EFUcKeBbpU - https://t.co/JkVOl1O0S1 - https://t.co/KG9PgxJabg Sharing weekly tips about growth: https://t.co/ereQodN3Ov
百度开源了 ERNIE-4.5-VL-28B-A3B 的Thinking 版本模型 专注于提升多模态理解和跨模态推理能力。 优势是通过高效的模型结构设计和强化学习策略，在较小参数规模下实现了高精度的视觉理解与复杂推理。 他们自己的测试大部分指标多模态理解指标与达Gemini-2.5-Pro与GPT-5-High 两个模型接近。文档/图表理解任务还超了点。 这次更新引入了 Thinking with Images 能力，类似之前的 O3，模型可以自动对图片进行放大、缩小、裁切等能力辅助发现图片细节进行视觉推理。 从前几天的 OCR 模型到现在的 ERNIE-4.5-VL-28B-A3B 的Thinking 百度好像找到了自己的发力方向和擅长的点。 现在在多模态这个层面国内的竞争确实不多，好用的多模态模型也少，大家都盯着代码的时候他们搞多模态也不错。 模型已经上线 HuggingFace/GitHub / 飞桨社区，权重、推理代码都公开了。
模型下载：https://t.co/eFD6tcWBRr
GitHub 上一份精心整理的 Claude 技能合集：Awesome Claude Skills，可以直接导入使用。 涵盖了文档处理、开发工具、数据分析、内容创作、生产力工具等 9 大类别的实用技能。 每个技能都是标准化的工作流程，可以让 Claude 安装我们的特定需求执行任务。 GitHub：https://t.co/h2o2vCPiOK 而且这些技能可以通过 Claude 官网、Claude Code 或者 API 几种方式使用。 项目还提供了详细的技能创建与使用指南，教我们如何编写自定义 Claude 技能，值得收藏备用。
💡 挖掘开源的价值 🧑🏻💻 坚持分享 GitHub 上高质量、有趣、实用的教程、AI工具、前沿 AI 技术 🧐 A list cool, interesting projects of GitHub. ✏️ 公众号：GitHubDaily
No doubt that they can add strong vision to K2, we've seen very strong precedents (Qwen3, but also Rednote turned actual DeepSeek V3 into a decent VLM). But this extends the K2 life cycle. I was wrong about K3 coming soon. Hopefully they fix overthinking in "K2.1"
We're in a race. It's not USA vs China but humans and AGIs vs ape power centralization. @deepseek_ai stan #1, 2023–Deep Time «C’est la guerre.» ®1