acshame

acshame

1. 14:32 · Apr 26, 2025 · Sat

   https://x.com/umesh_ai/status/1916117657373250041?t=bRUtWCbYYdMdZz6xlyalyw&s=35

   X (formerly Twitter)

   Umesh (@umesh_ai) on X

   A surreal illusionary prompt on ChatGPT 4o!
   
   Prompt : A surreal photograph showing a [BOTTOM OBJECT], with a [TOP OBJECT] floating clearly above it, leaving a visible empty space between them. The background is a warm, textured dark brown wall, and the lighting…

2. 13:31 · Apr 26, 2025 · Sat

   https://x.com/vista8/status/1915603984238465444?t=SgqbbbrpgOJ8wg3KEZSiWw&s=35

3. 13:13 · Apr 26, 2025 · Sat

   https://mp.weixin.qq.com/s/1MOevEyJMGeNRdQltn1P_g

4. 03:18 · Apr 26, 2025 · Sat

   让 AI 带你穿越古代
   提示词：超写实风格照片。你穿越到了【盛唐时代】，你的视角里正被一位美丽的古典东方女子牵着手向前拉着走。女孩带着你穿过繁华的街道，街上充满了时代气息。 ​​​https://mapp.api.weibo.cn/fx/766c4240d8dc030f238b94fffd63fa14.html

5. 02:44 · Apr 26, 2025 · Sat

   https://x.com/dotey/status/1915641057867399330?t=pxfPlzKiYTiEIrqhop3aNw&s=35

   X (formerly Twitter)

   宝玉 (@dotey) on X

   如果提示词用的对的话，Wan 做出来效果还不错的，但就是对提示词要求挺高的，视频用的提示词：
   Create a cinematic war scene featuring a realistic LEGO soldier (like the one in the image) in the middle of a gritty battlefield. The environment should have smoke, explosions in the

6. 02:31 · Apr 26, 2025 · Sat

   https://x.com/azed_ai/status/1915707665105469510?t=0P9rW4dOWEiNrS7CBPUMew&s=35

   X (formerly Twitter)

   Amira Zairi (@azed_ai) on X

   Prompt share: neon wireframe
   
   💬Prompt:
   A neon wireframe rendering of a [subject], outlined in glowing [color1] and [color2] lines. The scene is suspended in a digital void, with speed trails and geometric shadows creating a sense of motion.
   
   Check ATLs for…

7. 02:30 · Apr 26, 2025 · Sat

   https://x.com/GoogleCloudTech/status/1915847146244362602?t=sgRKSet377b172YBvpAT2w&s=35

8. 02:29 · Apr 26, 2025 · Sat

   https://x.com/jandotai/status/1915670642139226319?t=kUirBijj0CTMNkjxlmy2Bw&s=35

   X (formerly Twitter)

   👋 Jan (@jandotai) on X

   Tina: 1.5B model that beats other 1.5B reasoning models trained with full-parameter RL, at just $9 total cost.
   
   It hits 43.33% on AIME24, outperforming models trained with 20-40x more compute.
   
   Post-trained from R1's Qwen-1.5B distill using LoRA and RL. …

9. 02:28 · Apr 26, 2025 · Sat

   https://x.com/FinanceYF5/status/1915705586446197035?t=0neiRBcSp1tq3__mN-hQ7g&s=35

   X (formerly Twitter)

   AI Will (@FinanceYF5) on X

   这太疯狂了。
   
   一个1.5人规模的韩国团队刚刚发布了Dia 1.6 B
   
   这个AI可以直接从文本生成完整的对话（声音、笑声、咳嗽）🤯
   
   亲自试试看：👇

10. 02:27 · Apr 26, 2025 · Sat

    https://x.com/ZHO_ZHO_ZHO/status/1915700618129256796?t=RvHc-2f4Kp9vlkOfPJe2Rw&s=35

11. 02:27 · Apr 26, 2025 · Sat

    https://x.com/dotey/status/1915650504169865314?t=QqHxcmHSi2-FvS-mHp7r2w&s=35

    X (formerly Twitter)

    宝玉 (@dotey) on X

    一个 ChatGPT / Gemini 的实用技巧：检索并提取部分 PDF 内容
    
    ChatGPT 和 Gemini 应该都内置了 PDF 工具，可以读取 PDF 内容，也可以帮你检索。如果你选的是推理模型的话（o3/o4-mini/Gemini 2.5 pro等）由于它们都有推理能力，所以可以对 PDF 先进行检索，然后去提取。

12. 02:00 · Apr 26, 2025 · Sat

    #ai创造营#
    中文AI字体设计神器！附14组提示词 ​​​https://mapp.api.weibo.cn/fx/9cc819c16620bfe94c6e96b523d3129c.html

13. 01:59 · Apr 26, 2025 · Sat

    看到个好玩的调色板生成项目，RampenSau
    
    这个库利用色调循环和缓动函数来生成颜色渐变。它可以生成一系列色调，或者使用色调列表来生成颜色渐变。并且是开源的~
    
    github: github.com/meodai/rampensau
    网页直接用：meodai.github.io/rampensau ​​​https://mapp.api.weibo.cn/fx/83c4e2ecd0e8f101ffa29fa8052993e2.html

    GitHub

    GitHub - meodai/rampensau: Color palette generation function using hue cycling and simple easing functions.

    Color palette generation function using hue cycling and simple easing functions. - meodai/rampensau

14. 01:58 · Apr 26, 2025 · Sat

    我去，阶跃整个大活儿[哆啦A梦害怕]，开源了业界第一个gemini和gpt4o类似funtion能力的在线编辑生图模型，paper也写了，赶紧速刷一下去，大昕越来越牛B了，在国内这么卷的生意环境下走出了一条不一样的路，模型地址：http://t.cn/A6dML84X ​​​https://mapp.api.weibo.cn/fx/701b2cd97d1e90aa8be1befe436473a1.html

15. 01:49 · Apr 26, 2025 · Sat

    kimi现在开始玩开源了，这几个月也动作频出，刚刚发布了一个开源的voice模型，支持tts，stt，speech2speech等多种任务，encoder用的直接是whisper，然后多个模态融合，有机会可以测试一下，模型不大7b http://t.cn/A6dM292B ​​​https://mapp.api.weibo.cn/fx/c28ca207cf2b528076b6eea5e373a997.html

16. 15:30 · Apr 25, 2025 · Fri

    Docker 做这个就很直观，自然了。直接屏蔽环境差异，部署使用就很方便，虽然容器环境内会有一些限制，但大多数都是 API 调用，就不受影响。
    
    那么问题来了，会不会出现 mcp-compose， mcp-k8s，mcp-helm 的东西呢？
    
    https://youtu.be/98M_6njOnus?si=RI2OKgDC7cnuGMhn

    YouTube

    Announcing Docker MCP Catalog and Docker MCP Toolkit

    Coming soon! We’re introducing the Docker MCP Catalog and ToolKit to streamline how developers discover, install, authenticate, and connect MCP servers to their favorite clients.
    
    Whether you're building AI-powered apps or agents, the Docker MCP Catalog…

17. 15:11 · Apr 25, 2025 · Fri

    https://b23.tv/h3KeZcS

18. 04:35 · Apr 25, 2025 · Fri

    https://x.com/karminski3/status/1915579765160919281?t=M2zU492dibWvF5dkn-s5AQ&s=35

    X (formerly Twitter)

    karminski-牙医 (@karminski3) on X

    看到个开源的 AI 知识图谱生成框架——Graphiti
    
    这个知识图谱框架牛逼的地方是，可以持续将用户交互、结构化和非结构化数据以及外部信息整合到一个连贯的、可查询的图谱中。并且支持增量数据更新、高效检索、精确的历史查询。并且这些都无需重新计算整个图谱，非常适用于开发交互式、上下文感知的 AI

19. 01:26 · Apr 25, 2025 · Fri

    https://x.com/justone_he/status/1915197530729832613?s=35

20. 01:24 · Apr 25, 2025 · Fri

    https://x.com/karminski3/status/1915541513104199731?t=_9W3qVBXc17G1s743wMQTQ&s=35