About this Event
Join the global innovation movement at —a groundbreaking convergence of open-source pioneers, AI visionaries, and tech disruptors!
Gain exclusive insights from Hugging Face, Alibaba Cloud Qwen Team, NVIDIA, Huawei, Unitree Robotics, Google DeepMind, just to name a few a few.
AI Models Track
- Talks from AI Modelling team such as: Alibaba Qwen, Hugging Face Open-R1, ...
- Talks `about current AI Models limitation: Oxford, Minimax, ...
Embodied AI Track
- Talks about latest robotic open source hardware and software innovations such as: Unitree G1, Huggingface open source Humanoids, Pollen Robotics ...
AI Infra Track
- Talks include companies like: Nvidia, Khronos, Huawei, Docker...
AI Apps Track
- Talks include AI Agents such as: Google DeepMind AgentOS, Jina AI, Camel AI, OAKS
- As well as AI Apps development like: Dify, OpenManus, RAGFlow
Seeed Embodied AI Workshop
- Explore imitation learning with the open-source SO-ARM100 robotics arm kit, which should utilize the Hugging Face LeRobot framework for advanced AI architectures like ACT and Diffusion Policy.
- The ticket price includes a set of SO-ARM 101! to let you tinker at home after the event.
- Event Page: https://lu.ma/7djaix39
- The model built will be the latest version that should be faster to build than previous version.
Pytorch Day France
- Talks includes projects built with Pytorch such as: Nvidia scaling Mistral, HuggingFace UltraScale Playbook, IBM, ...
- Event Page: https://pytorchdayfrance2025.sched.com
Secure your spot now and be part of the community redefining technology’s boundaries. Tickets are limited—act fast to unlock early-bird discounts!
https://youtube.com/shorts/L-Z2bmesT3A?feature=share
Highlights 5th May
🕑: 10:30 AM - 11:30 AM
Open-R1: A fully open reproduction of DeepSeek-R1 by HuggingFace
Host: Guilherme Penedo / ML Research Engineer at Hugging Face
Info: A fully open reproduction of DeepSeek-R1. This repo is a work in progress, let's build it together!
🕑: 10:30 AM - 11:30 AM
AG2: The Open-Source AgentOS for Agentic AI
Host: Chi Wang / Research Scientist at Google DeepMind
🕑: 11:30 AM - 12:30 PM
Training and inferencing on Huawei Ascend CANN
Host: Xiyuan Wang / Senior Software Engineer at Huawei
🕑: 11:30 AM - 12:00 AM
The Zenoh Project
Host: Angelo Corsaro / CEO and CTO of ZettaScale Technology
🕑: 02:00 PM - 03:00 PM
Alibaba Apps: LLMs, Code Intelligence, Conversational AI and AI Agents
Host: Yongbin Li / Researcher at Alibaba
🕑: 02:30 PM - 03:30 PM
Linear Next: The evolution of LLM Architecture
Host: Yiran Zhong / Director of Research at MiniMax
🕑: 03:30 PM - 04:30 PM
Building robotic applications with open-source VLA models
Host: Ville Kuosmanen / Founder at Voyage Robotics
🕑: 04:30 PM - 05:00 PM
Datasets and Infrastructure for DeepSeek-R1 Style Reinforcement Learning (GRPO
Host: Greg Schoeninger / CEO at Oxen.ai
🕑: 04:30 PM - 05:30 PM
Dify: Open Source LLM app development platform
Host: Xinrui Liu / Developer Ecosystem Director at LangGenius
🕑: 04:30 PM - 05:30 PM
Going Beyond Tokens for Code Large Language Models
Host: Diego Rojas / Research Engineer at Zhipu.AI
🕑: 04:30 PM - 05:30 PM
Spatial Reasoning LLM: Enhancing 3D Understanding for Robotic Manipulation
Host: Huy Hoang Ha / LLM Researcher at Menlo Research
Highlights 6th May
🕑: 10:30 AM - 11:30 AM
Qwen2.5 Technical Report
Host: Baosong Yang / Qwen Team at Alibaba
Info: This report introduces the Qwen2 series, the latest addition to our large language models and large multimodal models. We release a comprehensive suite of foundational and instruction-tuned language models, encompassing a parameter range from 0.5 to 72 billion, featuring dense models and a Mixture-of-Experts model. Qwen2 surpasses most prior open-weight models, including its predecessor Qwen1.5, and exhibits competitive performance relative to proprietary models across diverse benchmarks on language understanding, generation, multilingual proficiency, coding, mathematics, and reasoning.
🕑: 10:30 AM - 11:30 AM
How to Build Your Humanoid
Host: Martino Russi / Embodied Engineer at Hugging Face
🕑: 11:30 AM - 12:00 PM
OpenManus: Open Source Framework for building general AI agents
Host: Sirui Hong / Co-Founder of OpenManus
🕑: 02:00 PM - 03:00 PM
G1 Open Source dataset and Humanoid Robot from Unitree Robotics
Host: Min Zhang / EU Director of Unitree Robotics
🕑: 02:00 PM - 03:00 PM
Pre-training of smol and large LLM
Host: Elie Bakouch / Research Engineer at Hugging Face
Info: While large language models have facilitated breakthroughs in many applications of artificial intelligence, their inherent largeness makes them computationally expensive and challenging to deploy in resource-constrained settings.
🕑: 03:30 PM - 04:30 PM
Khronos in the World of Open Source and Machine Learning
Host: Markus Tavenrath / Principal Engineer at NVIDIA
🕑: 04:30 PM - 05:30 PM
The Curse of Depth in Large Language Models
Host: Shiwei Liu / Fellow at University of Oxford
Agentic
🕑: 02:00 PM
Llama 4
Host: Christian Keller @ Meta
🕑: 02:30 PM
The Ultra-Scale Talk: Scaling Training to Thousands of GPUs
Host: Nouamane Tazi @ HuggingFace
🕑: 03:00 PM
Teaching Mistral to Reason: Post-Training with PyTorch and NVIDIA
Host: Meriem Bendris @ NVIDIA
🕑: 03:30 PM
DeepSpeed – Efficient Training Scalability for Deep Learning Models
Host: Olatunji Ruwase @ Deep Learnning Expert
🕑: 04:00 PM
Advancing Mamba in PyTorch
Host: Garrett Goon
🕑: 05:30 PM
PyTorch x Transformers journey: pythonicity, autodiff and modularity defining
Host: Pablo Montalvo @ HuggingFace
Event Venue & Nearby Stays
STATION F, 5 Parvis Alan Turing, Paris, France
EUR 53.90 to EUR 764.50
