Hugging Face

Hugging Face (Twitter) RT @mervenoyann: I need to follow more people, please send your @huggingface usernames 🤝 (also flexing the ✨)

Hugging Face (Twitter) RT @Thom_Wolf: We’re releasing a new, extremely high-quality large dataset: *FinePDF-edu* This continues our work on open-sourcing everything possible about the science of pre-training datasets. In this new data, we applied the same filtering techniques we used for the widely adopted *FineWeb-edu*, but this time to our recently published PDF dataset called FinePDF — 3 trillion tokens of Common Crawl PDFs released a few months ago. The result is *FinePDF-edu* a high-quality dataset of about 350B tokens (130B English), representing the most educational slice of all Common Crawl PDFs and outperforming our previously released pretraining dataset on our benchmark. Some details 👇 High-level: - 350B+ highly educational tokens in 69 languages with strong performance - 69 education classifiers powered by ModernBERT and mmBERT - 300k+ EDU annotations per language, generated with Qwen3-235B Method: We filtered this dataset by asking an LLM to score the of a... Перейти на оригинальный пост

Hugging Face (Twitter) RT @HKydlicek: We releasing a large update to 📄FinePDFs! - 350B+ highly education tokens in 69 languages, with incredible perf 🚀 - 69 edu classifiers, powered by ModernBert and mmBERT - 300k+ EDU annotations for each of 69 languages from Qwen3-235B

Hugging Face (Twitter) RT @Thom_Wolf: Open-source/science is the most effective way to reduce civilization boot time

‌Hugging Face (Twitter) RT @RedHat_AI: Paris, we’re coming your way 🇫🇷 Join us for the vLLM meetup on 19 November, hosted by @RedHat, @AMD & @huggingface. Expect awesome community vibes plus deep dives into @vllm_project, @_llm_d_, and how teams are scaling GenAI in production. Sign up → luma.com/ryaao30v

Hugging Face (Twitter) RT @Xianbao_QIAN: Welcome Weibo to the open source AI community! They just released a 1.5B reasoning model that surpasses DeepSeek R1-0120 on math while being 100x smaller. Model: https://huggingface.co/WeiboAI/VibeThinker-1.5Bhttps://twitter.com/WeiboLLM/status/1988109435902832896#m

Hugging Face (Twitter) 👀https://twitter.com/abidlabs/status/1988312855767294330#m

‌Hugging Face (Twitter) RT @mattcline__: @huggingface has a library of free courses to learn about LLMs, MCP, etc.: huggingface.co/learn

Hugging Face (Twitter) RT @samhogan: @samhogan@laion_ai@Wyndlabs_ai@AmarSVS@0xdrej@bonham_sol@francescodvirga@ChristophSchuh6@inference_net: In order to complete this project, we trained 2 custom models: Aella-Nemotron-12b, and Aella-Qwen-14b. Available today on @huggingface These models achieve frontier performance on our extraction task at 98% lower cost.

Hugging Face (Twitter) RT @teortaxesTex: I admit I was skeptical about Alexander's small model obsession, but this is a powerful flex. https://twitter.com/Dorialexander/status/1987930819021635964#m

‌Hugging Face (Twitter) RT @ErnieforDevs: 🔥We just open-sourced ERNIE-4.5-VL-28B-A3B-Thinking! Only 3B activated parameters, yet delivers TOP-TIER visual performance: 💙Visual Reasoning: Multi-step reasoning, chart analysis, and causal reasoning in complex visual tasks. 💙STEM Reasoning: Snap a photo to solve complex question. 💙Visual Grounding: Pinpoint objects accurately in any scene. 💙Thinking with Images: Zoom and catch every detail instantly. 💙Tool Calling: Use functions like Image Search and more. 💙Video Comprehension: Track timelines and events with ease. 👉Model: 👉Full compatibility with vLLM, Transformers, and FastDeploy! Give us a like💗 if you find it useful!

Hugging Face (Twitter) RT @alexandrabotez: Met Eddy when he was 17, working on AI glasses for chess. Six months later he’s a CEO, open-sourcing the largest first-person robotics dataset ever. The new generation moves fast. Props to the team! https://twitter.com/eddybuild/status/1987951619804414416#m