Hugging Face opens courses
Hugging Face released a set of 12 internal courses on LLMs, robotics, and model-centred practices this week, and the collection is already being shared widely as a free resource for production-ready skills. The release is getting traction among engineers looking to move from experiments to deployed ML systems. (x.com)
Hugging Face’s Learn hub lists a dozen course tracks by name — LLM, Robotics (LeRobot), MCP, Agents, Deep RL, Computer Vision, Audio, Diffusion, ML for Games, ML for 3D, an Open-Source AI Cookbook, and a compact “smol” course. (huggingface.co)) Multiple course pages advertise hands‑on assignments and formal completion certificates, and several units run exercises on Hugging Face–hosted infrastructure such as MCP servers for practical labs. (aihola.com)) The Robotics track is accompanied by a public GitHub repository that contains unit materials and exercises (the repo snapshot shows about 48 commits and organized unit content). (github.com)) Hugging Face’s MCP course explicitly teaches the Model Context Protocol stack — covering MCP theory, SDK usage, and integrations for connecting LLMs to external data and tools — linking its curriculum to the MCP standard popularized by Anthropic. (marktechpost.com)) The LLM Course is run as a cohorted series with scheduled chapter releases and certification milestones (Chapter 1 reboot April 23, 2025; Chapter 1 certification and quiz May 10, 2025; Chapter 2 reboot June 18, 2025; Chapter 3 and final certification slated for July 2025). (huggingface.co)) Hugging Face also announced a formal partnership with DataCamp to package a “Hugging Face Fundamentals” skill track combining interactive course content with a real‑world capstone project. (huggingface.co)) Several course codebases and guides are published openly (for example the “101‑course” and robotics course repos), enabling contribution, forked projects, and reuse of course notebooks and demos on GitHub. (github.com))
