🤔Tired of noisy training data corrupting your LLM? Ultra-FineWeb introduces: • Efficient verification strategy with minimal compute cost • Automated seed data selection (less human bias!) • fastText-based lightweight classifiers • Applied to FineWeb & Chinese FineWeb

How do you make LLMs both long-context capable and super fast? Meet InfLLM-V2 from Tsinghua x OpenBMB — a breakthrough dense-sparse switchable attention system that: 1⃣ Seamlessly adapts from short to long sequences 2⃣Runs 4× faster than dense attention 3⃣Keeps 99%+ accuracy

🤔Loving the new DeepSeek-V3.2? 🔥Remember the first sparse-native models! ✨ InfLLM‑V2: Seamless Long‑Context Adaptation Paper: https://t.co/3krgcnMRRa 1️⃣ Ultra‑fast adaptation: only 5B long‑text tokens to train sparse attention (vs. ~1T in DSA from DeepSeek-V3.2) 2️⃣

absolutely 😛

Our reasoning models with trainable sparse attention!!!

Why does a realistic voice matter? 🤔🤔 The same robot that feels creepy can transform into a trusted companion just by having a human-like voice. Think about the movie, Her. 🔥VoxCPM is a new paradigm: continuous, context-aware, and incredibly lifelike. ✅ Small size: 0.5B

🔥Meet VoxCPM: Our game-changing, tokenizer-free TTS Model. Powered by MiniCPM-4, it delivers hyper-realistic speech, zero-shot voice cloning, and natural prosody. Trained on 1.8M+ hours, hitting SOTA. Try it now👉: https://t.co/9Am55qZMsG

VoxCPM Tokenizer-Free TTS for Context-Aware Speech Generation and True-to-Life Voice Cloning vibe coded a quick TTS app in anycoder

🧩New blog: From f(x) and g(x) to f(g(x)): LLMs Learn New Skills in RL by Composing Old Ones Do LLMs learn new skills through RL, or just activate existing patterns? Answer: RL teaches the powerful meta-skill of composition when properly incentivized. 🔗: https://t.co/4Ud8qsYrOT

We did it! We now have two models in the top two spots on the @huggingface trending charts. 🥇 Hunyuan-MT-7B 🥈 HunyuanWorld-Voyager Download and deploy the models for free on Hugging Face and GitHub. Your stars and feedback are welcome! 🌟👍❤️ This is just the beginning. Stay

MiniCPM4.1-8B Reasoning LLM with Trainable Sparse Attention with 8B parameters, support fusion thinking vibe coding a gradio app for it in anycoder link in the thread

@_akhaliq Anycoder is amazing! If you've enjoyed your experience with our model, feel free to check us out on Hugging Face: https://t.co/NNlWOXFjik

Introducing MiniCPM 4.1-8B: First Open-Source Reasoning LLM with Trainable Sparse Attention ✅ Strong Reasoning Capability: Surpasses similar-sized models on 15 tasks! ✅ Fast Generation: 3x decoding speedup for reasoning ✅ Efficient Architecture: Trainable sparse attention,

清华、OpenBMB等最新发布了一款首个基于MCP架构的RAG 框架：UltraRAG 2.0，只写YAML，就能以极低代码量快速实现多阶段推理系统 通过YAML文件即可声明串行、循环、条件分支等复杂逻辑，几十行代码即可构建DeepResearch类的复杂RAG流程 支持动态检索、条件判断、多轮交互等高级能力

🔥Finally found it! A new powerful RAG tool is now open source! 🚀 Introducing UltraRAG 2.0: The first MCP-based Retrieval-Augmented Generation (RAG) framework! 🧩 Modular MCP Servers: plug-and-play Retriever, Generator, Evaluator 📜 Low-code YAML pipelines: complex RAG in <100

MiniCPM-V 4.5 is very good! 🤗 it comes with hybrid thinking: it decides when to think on it's own 😍 it also can handle high res documents with odd aspect ratios, and super long videos efficiently 🙏🏻 see below hybrid results ⤵️ model is in comments!

vibe coding a MiniCPM-V 4.5 @OpenBMB chat app in anycoder MiniCPM-V 4.5 achieves an average score of 77.0 on OpenCompass, a comprehensive evaluation of 8 popular benchmarks. With only 8B parameters, it surpasses widely used proprietary models like GPT-4o-latest, Gemini-2.0 Pro,

🤯 MiniCPM-V4.5 solves math faster than I can even read. Developers, think of all the instant-recall scenarios! I sometimes think open source doesn't change the world in a single moment. Its power is built brick by brick, together, until we look back and see how different

🤔 Model missing the details like fleeting handwriting? Not anymore！ 🚀 Introducing MiniCPM-V 4.5 8B: accurately recognizing even the fastest scribbles. Get ready for the future of multimodal AI 👉 Huggingface｜ https://t.co/0NwYU3p81J Github｜ https://t.co/IivDNmxu4j

🚀 Introducing MiniCPM-V 4.5 8B: pushing the boundary of multimodal AI! ～ SOTA VL Capability: Surpasses GPT-4o, Gemini 2.0 Pro, Qwen2.5-VL 72B on OpenCompass! ～ "Eagle Eye" Video: 96x visual token compression for high refresh rate and long video understanding ～ Controllable