Spaces:
Running
Running
| title: README | |
| emoji: 🐠 | |
| colorFrom: yellow | |
| colorTo: yellow | |
| sdk: static | |
| pinned: false | |
| # Hugging Face Research | |
| The science team at Hugging Face is dedicated to advancing machine learning research in ways that maximize value for the whole community. Our work focuses on three core areas of tooling, datasets and open models. | |
| This is the release timeline so far (you can click on each element!): | |
| <style> | |
| @import url('https://fonts.googleapis.com/css2?family=DM+Sans:wght@400;600&display=swap'); | |
| .page-container { | |
| width: 100%; | |
| max-width: 100vw; | |
| overflow-x: hidden; | |
| } | |
| .timeline-container { | |
| max-width: 100%; | |
| margin: 120px auto 40px; | |
| padding: 30px; | |
| overflow-x: auto; | |
| font-family: 'DM Sans', sans-serif; | |
| scrollbar-width: auto; | |
| scrollbar-color: #888 #f1f1f1; | |
| } | |
| .timeline { | |
| position: relative; | |
| min-width: 1800px; | |
| } | |
| /* Rest of your existing CSS remains the same */ | |
| .line { | |
| position: absolute; | |
| width: 100%; | |
| height: 2px; | |
| background: #333; | |
| bottom: 30px; | |
| } | |
| .timeline-items { | |
| display: flex; | |
| justify-content: space-between; | |
| align-items: flex-end; | |
| position: relative; | |
| min-height: 320px; | |
| } | |
| .month-marker { | |
| display: flex; | |
| flex-direction: column; | |
| align-items: center; | |
| position: relative; | |
| flex: 1; | |
| } | |
| .month-dot { | |
| width: 12px; | |
| height: 12px; | |
| background: #000; | |
| border-radius: 50%; | |
| margin-bottom: -14px; | |
| position: relative; | |
| z-index: 1; | |
| } | |
| .month-label { | |
| font-weight: bold; | |
| margin-top: 10px; | |
| } | |
| .events-container { | |
| position: absolute; | |
| bottom: 40px; | |
| left: 50%; | |
| transform: translateX(-50%); | |
| width: 200px; | |
| text-align: left; | |
| } | |
| .event { | |
| position: relative; | |
| left: 92px; | |
| margin: 8px 0; | |
| font-size: 14px; | |
| display: flex; | |
| align-items: flex-start; | |
| gap: 5px; | |
| white-space: pre-wrap; | |
| max-width: 135px; | |
| } | |
| .event img { | |
| width: 20px; | |
| height: 20px; | |
| vertical-align: middle; | |
| } | |
| .event a { | |
| text-decoration: none; | |
| } | |
| .event a:hover { | |
| color: #686868; | |
| text-decoration: underline; | |
| } | |
| </style> | |
| <div class="page-container"> | |
| <div class="timeline-container"> | |
| <div class="timeline"> | |
| <div class="line"></div> | |
| <div class="timeline-items"> | |
| <div class="month-marker"> | |
| <div class="events-container"> | |
| <div class="event">🎥<a href="https://huggingface.co/blog/smolvlm2" target="_blank">SmolVLM2</a></div> | |
| <div class="event">🪐<a href="https://huggingface.co/spaces/nanotron/ultrascale-playbook" target="_blank">Ultra-Scale Playbook</a></div> | |
| <div class="event">🐋<a href="https://huggingface.co/datasets/open-r1/OpenR1-Math-220k" target="_blank">OpenR1-Math-220k</a></div> | |
| <div class="event">🕺<a href="https://huggingface.co/blog/dabstep" target="_blank">DABstep</a></div> | |
| <div class="event">🔎<a href="https://huggingface.co/blog/open-deep-research" target="_blank">Open Deep Research</a></div> | |
| <div class="event">📝<a href="https://huggingface.co/papers/2502.02737" target="_blank">SmolLM2 report</a></div> | |
| </div> | |
| <div class="month-dot"></div> | |
| <div class="month-label">Feb</div> | |
| </div> | |
| <div class="month-marker"> | |
| <div class="events-container"> | |
| <div class="event">🧮<a href="https://x.com/HKydlicek/status/1881734376696041659" target="_blank">Math Verify</a></div> | |
| <div class="event">🐳<a href="https://github.com/huggingface/open-r1" target="_blank">Open-R1</a></div> | |
| <div class="event">🛠️<a href="https://huggingface.co/docs/smolagents" target="_blank">SmolAgents</a></div> | |
| </div> | |
| <div class="month-dot"></div> | |
| <div class="month-label">Jan 2025</div> | |
| </div> | |
| <div class="month-marker"> | |
| <div class="events-container"> | |
| <div class="event">🔢<a href="https://huggingface.co/spaces/huggingface/number-tokenization-blog" target="_blank">Number Tokenization Blog</a></div> | |
| <div class="event">🥂<a href="https://huggingface.co/datasets/HuggingFaceFW/fineweb-2" target="_blank">FineWeb 2</a></div> | |
| <div class="event">📈<a href="https://huggingface.co/spaces/HuggingFaceH4/blogpost-scaling-test-time-compute" target="_blank">Scaling Test Time Compute</a></div> | |
| <div class="event">🤖<a href="https://github.com/huggingface/picotron" target="_blank">Picotron</a></div> | |
| <div class="event">📐<a href="https://huggingface.co/datasets/HuggingFaceTB/finemath" target="_blank">FineMath</a></div> | |
| </div> | |
| <div class="month-dot"></div> | |
| <div class="month-label">Dec 2024</div> | |
| </div> | |
| <div class="month-marker"> | |
| <div class="events-container"> | |
| <div class="event">🤏<a href="https://huggingface.co/HuggingFaceTB/SmolLM2-1.7B-Instruct" target="_blank">SmolLM2</a></div> | |
| <div class="event">🤓<a href="https://huggingface.co/HuggingFaceTB/SmolVLM-Instruct" target="_blank">SmolVLM</a></div> | |
| </div> | |
| <div class="month-dot"></div> | |
| <div class="month-label">Nov</div> | |
| </div> | |
| <div class="month-marker"> | |
| <div class="events-container"> | |
| <div class="event">🔎<a href="https://github.com/huggingface/evaluation-guidebook" target="_blank">LLM Evaluation Guidebook</a></div> | |
| <div class="event">🦾<a href="https://x.com/Thom_Wolf/status/1851557379294286176" target="_blank">LeRobot Hackathon</a></div> | |
| <div class="event">🗺️<a href="https://huggingface.co/spaces/HuggingFaceFW/blogpost-fine-tasks" target="_blank">FineTasks</a></div> | |
| <div class="event">⛅<a href="https://x.com/nathanhabib1011/status/1849160216438587466" target="_blank">Lighteval now Multilingual</a></div> | |
| </div> | |
| <div class="month-dot"></div> | |
| <div class="month-label">Oct</div> | |
| </div> | |
| <div class="month-marker"> | |
| <div class="events-container"> | |
| <div class="event">🦾<a href="https://x.com/alibert_s/status/1828760527730082024" target="_blank">LeRobot Video</a></div> | |
| <div class="event">🎥<a href="https://huggingface.co/spaces/HuggingFaceFV/FineVideo-Explorer" target="_blank">FineVideo</a></div> | |
| <div class="event">📣<a href="https://x.com/andi_marafioti/status/1830862304906268725" target="_blank">Speech-to-Speech Multilingual</a></div> | |
| </div> | |
| <div class="month-dot"></div> | |
| <div class="month-label">Sep</div> | |
| </div> | |
| <div class="month-marker"> | |
| <div class="events-container"> | |
| <div class="event">🦾<a href="https://github.com/huggingface/lerobot/blob/main/examples/7_get_started_with_real_robot.md" target="_blank">LeRobot Tutorial</a></div> | |
| <div class="event">📣<a href="https://github.com/huggingface/speech-to-speech" target="_blank">Speech-to-Speech</a></div> | |
| <div class="event">🐶<a href="https://huggingface.co/HuggingFaceM4/Idefics3-8B-Llama3" target="_blank">Idefics 3</a></div> | |
| <div class="event">🤏<a href="https://huggingface.co/spaces/HuggingFaceTB/instant-smollm" target="_blank">Instant SmolLM</a></div> | |
| </div> | |
| <div class="month-dot"></div> | |
| <div class="month-label">Aug</div> | |
| </div> | |
| <div class="month-marker"> | |
| <div class="events-container"> | |
| <div class="event">🛟<a href="https://x.com/Haojun_Zhao14/status/1815419356408336738" target="_blank">Ring attention</a></div> | |
| <div class="event">🦾<a href="https://x.com/RemiCadene/status/1805583409382932620" target="_blank">LeRobot TeleOps</a></div> | |
| <div class="event">🥇<a href="https://x.com/_lewtun/status/1808898804822720769" target="_blank">Win AIMO</a></div> | |
| <div class="event">🐶<a href="https://huggingface.co/blog/docmatix" target="_blank">Docmatix</a></div> | |
| <div class="event">🤏<a href="https://huggingface.co/blog/smollm" target="_blank">SmolLM</a></div> | |
| </div> | |
| <div class="month-dot"></div> | |
| <div class="month-label">Jul</div> | |
| </div> | |
| <div class="month-marker"> | |
| <div class="events-container"> | |
| <div class="event">🍷<a href="https://huggingface.co/spaces/HuggingFaceFW/blogpost-fineweb-v1f" target="_blank">FineWeb Report</a></div> | |
| <div class="event">🍷<a href="https://huggingface.co/datasets/HuggingFaceFW/fineweb-edu" target="_blank">FineWeb-Edu</a></div> | |
| <div class="event">🌺<a href="https://huggingface.co/blog/finetune-florence2" target="_blank">Florence 2 Blog</a></div> | |
| <div class="event">🏆<a href="https://huggingface.co/spaces/open-llm-leaderboard/blog" target="_blank">Open LLM Leaderboard v2</a></div> | |
| <div class="event">👩🏫<a href="https://www.youtube.com/watch?v=jm2hyJLFfN8" target="_blank">Stanford CS25</a></div> | |
| </div> | |
| <div class="month-dot"></div> | |
| <div class="month-label">Jun</div> | |
| </div> | |
| <div class="month-marker"> | |
| <div class="events-container"> | |
| <div class="event">🦾<a href="https://github.com/huggingface/lerobot" target="_blank">LeRobot Release</a></div> | |
| <div class="event">📈<a href="https://arxiv.org/abs/2405.18392" target="_blank">WSD Analysis</a></div> | |
| </div> | |
| <div class="month-dot"></div> | |
| <div class="month-label">May</div> | |
| </div> | |
| <div class="month-marker"> | |
| <div class="events-container"> | |
| <div class="event">🍷<a href="https://huggingface.co/datasets/HuggingFaceFW/fineweb" target="_blank">FineWeb</a></div> | |
| <div class="event">🕵️<a href="https://huggingface.co/blog/jat" target="_blank">JAT Agent</a></div> | |
| <div class="event">🪁<a href="HuggingFaceH4/zephyr-orpo-141b-A35b-v0.1" target="_blank">Zephyr Mixtral</a></div> | |
| <div class="event">🐶<a href="https://huggingface.co/HuggingFaceM4/idefics2-8bb" target="_blank">Idefics 2</a></div> | |
| <div class="event">🏆<a href="https://huggingface.co/blog/leaderboard-medicalllm" target="_blank">Community Leaderboards</a></div> | |
| </div> | |
| <div class="month-dot"></div> | |
| <div class="month-label">Apr</div> | |
| </div> | |
| <div class="month-marker"> | |
| <div class="events-container"> | |
| <div class="event">🪁<a href="https://huggingface.co/HuggingFaceH4/zephyr-7b-gemma-v0.1" target="_blank">Zephyr Gemma</a></div> | |
| <div class="event">🪐<a href="https://huggingface.co/datasets/HuggingFaceTB/cosmopedia" target="_blank">Cosmopedia</a></div> | |
| </div> | |
| <div class="month-dot"></div> | |
| <div class="month-label">Mar</div> | |
| </div> | |
| <div class="month-marker"> | |
| <div class="events-container"> | |
| <div class="event">⚙️<a href="https://github.com/huggingface/nanotron/" target="_blank">Nanotron Release</a></div> | |
| <div class="event">⛅<a href="https://github.com/huggingface/lighteval/" target="_blank">Lighteval Release</a></div> | |
| <div class="event">⚙️<a href="https://github.com/huggingface/datatrove/" target="_blank">Datatrove Release</a></div> | |
| <div class="event">⭐️<a href="https://huggingface.co/datasets/bigcode/the-stack-v2" target="_blank">The Stack v2</a></div> | |
| <div class="event">⭐️<a href="https://huggingface.co/bigcode/starcoder2-15b" target="_blank">StarCoder2</a></div> | |
| </div> | |
| <div class="month-dot"></div> | |
| <div class="month-label">Feb</div> | |
| </div> | |
| <div class="month-marker"> | |
| <div class="events-container"> | |
| <div class="event">🔥 Warming up</div> | |
| </div> | |
| <div class="month-dot"></div> | |
| <div class="month-label">Jan 2024</div> | |
| </div> | |
| </div> | |
| </div> | |
| </div> | |
| </div> | |
| ### 🛠️ Tooling & Infrastructure | |
| The foundation of ML research is tooling and infrastructure and we are working on a range of tools such as [datatrove](www.github.com/huggingface/datatrove), [nanotron](www.github.com/huggingface/nanotron), [TRL](www.github.com/huggingface/trl), [LeRobot](www.github.com/huggingface/lerobot), and [lighteval](www.github.com/huggingface/lighteval). | |
| ### 📑 Datasets | |
| High quality datasets are the powerhouse of LLMs and require special care and skills to build. We focus on building high-quality datasets such as [no-robots](https://huggingface.co/datasets/HuggingFaceH4/no_robots), [FineWeb](https://huggingface.co/datasets/HuggingFaceFW/fineweb), [The Stack](https://huggingface.co/datasets/bigcode/the-stack-v2), and [FineVideo](https://huggingface.co/datasets/HuggingFaceFV/finevideo). | |
| ### 🤖 Open Models | |
| The datatsets and training recipes of most state-of-the-art models are not released. We build cutting-edge models and release the full training pipeline as well fostering more innovation and reproducibility, such as [Zephyr](https://huggingface.co/HuggingFaceH4/zephyr-7b-beta), [StarCoder2](https://huggingface.co/bigcode/starcoder2-15b), or [SmolLM2](https://huggingface.co/HuggingFaceTB/SmolLM2-1.7B-Instruct). | |
| ### 🌸 Collaborations | |
| Research and collaboration go hand in hand. That's why we like to organize and participate in large open collaborations such as [BigScience](https://bigscience.huggingface.co) and [BigCode](https://www.bigcode-project.org), as well as lots of smaller partnerships such as [Leaderboards on the Hub](https://huggingface.co/blog?tag=leaderboard). | |
| ### ⚙️ Infrastructre | |
| The research team is organized in small teams with typically <4 people and the science cluster consists of 96 x 8xH100 nodes as well as an auto-scalable CPU cluster for dataset processing. In this setup, even a small research team can build and push out impactful artifacts. | |
| ### 📖 Educational material | |
| Besides writing tech reports of research projects we also like to write more educational content to help newcomers get started to the field or practitioners. We built for example the [alignment handbook](https://github.com/huggingface/alignment-handbook), the [evaluation guidebook](https://github.com/huggingface/evaluation-guidebook), the [pretraining tutorial](https://www.youtube.com/watch?v=2-SPH9hIKT8), or the [FineWeb blog](https://huggingface.co/spaces/HuggingFaceFW/blogpost-fineweb-v1). | |
| ### 🤗 Join us! | |
| We are actively hiring for both full-time and internships. Check out [hf.co/jobs](https://hf.co/jobs) | |