Your curated collection of saved posts and media
Interesting. @slow is looking to address this as well. https://t.co/0AmG69Gxda
Interesting. @slow is looking to address this as well. https://t.co/0AmG69Gxda
Looking for your next AI Engineering Role? Aa bunch of VC-backed startups are hiring founding engineers, in ai and backend systems. Don't miss out on some great opportunities and your next big challenge Apply now! https://t.co/1wqJdsRLpS
Ring-mini-2.0 With only 16B total parameters and 1.4B activated parameters, it achieves comprehensive reasoning capabilities comparable to dense models below the 10B scale On multiple benchmarks (LiveCodeBench, AIME 2025, GPQA, ARC-AGI-v1, etc.), outperforms dense models below 10B and even rivals larger MoE models (e.g., gpt-oss-20B-medium) with comparable output lengths, particularly excelling in logical reasoning vibe coded a quick chat app in anycoder
vibe coded app: https://t.co/ladQBBCmWQ
vibe coding app: https://t.co/esPDyHE1YC
VoxCPM Tokenizer-Free TTS for Context-Aware Speech Generation and True-to-Life Voice Cloning vibe coded a quick TTS app in anycoder https://t.co/eOP5BshtNR
vibe coded app: https://t.co/NCY2IXGZmR
vibe coding app: https://t.co/esPDyHE1YC
๐ฅ amazing anycoder China Nation day 8 day trip to Hugging face AnyCoder, I will use anycoder to create a Hugging Face Daily Paper Quiz๏ผgenerate a quiz for every HF Papers, to help researchers understand the paper better @_akhaliq https://t.co/azapzNhK7O https://t.co/PzbXBNdf8p
1/2 Unveilingย theย scienceย behindย ourย Tongyiย DeepResearchย Agent!ย Theย 6ย coreย researchย papersย behindย itย haveย sweptย theย topย 6ย onย @huggingface Dailyย Papers.ย Weย detailedย everything:ย data,ย agenticย trainingย (CPT,ย SFT,ย RL),ย andย inference.ย Check them in the following thread!!! https://t.co/5j9MbGMQ6D
@huggingface Cheak all the trending paper here ๐https://t.co/CbedMHC97M
@huggingface Cheak all the trending paper here ๐https://t.co/CbedMHC97M
Scaling Agents via Continual Pre-training https://t.co/8nnmldwwpd
discuss with author: https://t.co/zAm6AROMJp
WebSailor-V2 Bridging the Chasm to Proprietary Agents via Synthetic Data and Scalable Reinforcement Learning https://t.co/D5ppNRMp4f
discuss with author: https://t.co/xczllYYpRj
Alibaba's WebSailor-V2: SOTA open-source web agents arrive A groundbreaking framework, powered by synthetic data and a dual-environment RL pipeline, achieves state-of-the-art results on BrowseComp & HLE. It outperforms existing open-source models and closes the gap to proprietary systems.
discuss with author: https://t.co/G5M0VYe61t
Stable Part Diffusion 4D Multi-View RGB and Kinematic Parts Video Generation https://t.co/E0dAzMRKuC
discuss with author: https://t.co/Z5VkJQte1t
Announcing FIRST ever AI World Tour - KION 2026. With the FIRST FULL AI Music Video: Kion - Stay Mad. Made ENTIRELY in Higgsfield x KLING. AND Introducing Kion's SOUL ID character in Higgsfield! Create Unlimited FREE SOUL images with Kion now. Creators & Kion lovers! EARN 1 YEAR Creator Plan - create your own content featuring Kion, post with #KION and let her take over your stage.
Alibaba's Tongyi Lab unveils AgentFounder Introducing Agentic Continual Pre-training (CPT) to build robust foundation models for LLM agents. AgentFounder-30B achieves state-of-the-art performance on 10 benchmarks, solving deep information-seeking tasks more effectively. https://t.co/8vNxnpAPaN
BOOM! Starting today you can use open source frontier LLMs in @code with HF Inference Providers! ๐ฅ Use your inference credits on SoTA llms like GLM 4.5, Qwen3 Coder, DeepSeek 3.1 and more All of it packaged in one simple extension - try it out today ๐ค https://t.co/1DaHUYXHl1
Starting today, you can use Hugging Face Inference Providers directly in GitHub Copilot Chat on @code! ๐ฅ which means you can access frontier open-source LLMs like Qwen3-Coder, gpt-oss and GLM-4.5 directly in VS Code, powered by our world-class inference partners - @CerebrasSystems, @Cohere_Labs, @FireworksAI_HQ, @GroqInc, @novita_labs, @togethercompute & more! give it a try today! ๐งต๐
Qwen3-Next-80B-A3B is out 80B params, but only 3B activated per token โ 10x cheaper training, 10x faster inference than Qwen3-32B.(esp. @ 32K+ context!) Qwen3-Next-80B-A3B-Instruct approaches our 235B flagship. Qwen3-Next-80B-A3B-Thinking outperforms Gemini-2.5-Flash-Thinking both now available in anycoder for vibe coding
Super excited to bring hundreds of state-of-the-art open models (Kimi K2, Qwen3 Next, gpt-oss, Aya, GLM 4.5, Deepseek 3.1, Hermes 4, and dozens new ones every day) directly into @code & @Copilot, thanks to @huggingface inference providers! This is powered by our amazing partners @CerebrasSystems, @FireworksAI_HQ, @Cohere_Labs, @GroqInc, @novita_labs, @togethercompute, and others who make this possible. ๐ช Hereโs why this is different than other APIs: ๐ง Open weights - models you can truly own, so theyโll never get nerfed or taken away from you โก Multiple providers - automatically routing to get you the best speed, latency, and reliability ๐ธ Fair pricing - competitive rates with generous free tiers to experiment and build ๐ Seamless switching - swap models on the fly without touching your code ๐งฉ Full transparency - know exactly whatโs running and customize it however you want The future of AI copilots is open and this is a big first step! ๐
You can now access Groq models directly in VS @code with @huggingface. Just BYOK. ๐ https://t.co/mGCtYl1yMY
You can now access Groq models directly in VS @code with @huggingface. Just BYOK. ๐ https://t.co/mGCtYl1yMY
we've been pushing commits to transformers discretely, time to talk about we've been cooking the last few months: โก๏ธ Continuous Batching is in transformers โก๏ธ this will simplify, most notably, evaluation and your training loop: no need for extra dependencies or infra to get fast inference, and no need for convoluted code to update your weights note that speed is currently not on par with the best inference frameworks and servers out there and probably never will be the goal is *not* to become as fast: we want to complement the existing landscape with features like these, aiming for transformers to be the toolbox for tinkering with and building models
Weโve been cooking this summer: Holo1.5 is here! SOTA UI localization + QA, 3ร gains vs Qwen-2.5 VL ๐ณ Now up to 72B ๐ฅ โ a strong base for computer-use agents like Surfer. โข Open weights on HuggingFace ๐ค https://t.co/JKj9dKbHZ8 โข Blog post ๐ https://t.co/PIQBV9wIA8 (1/n ๐งต)
Talking about the state of Open Source LLMs at @aiDotEngineer next week! ๐ฅ Quite excited for the talk and meeting everyone - let's goo! ๐ค https://t.co/IO21y0E1vp
