ELI5: Nevergrad @FacebookOpenSource
ELI5: Nevergrad  @FacebookOpenSource
Uploaded June 2022 | Updated November 2026, 3 hours ago
In this video, Meta Open Source Developer Advocate Suraj explains Nevergrad, a Python library that offers an extensive collection of gradient-free algorithms, along with handy evaluation and benchmarking tools to see which one would work best for different kinds of machine learning and scientific engineering problems.

------- Project Links -------

Nevergrad Github: github.com/facebookresearch/nevergrad
Nevergrad Documentation: facebookresearch.github.io/nevergrad
Nevergrad Blog: engineering.fb.com/2018/12/20/ai-research/nevergrad
Nevergrad Facebook Group: facebook.com/groups/nevergradusers
Autograd (for gradient-based optimization) Starter guide: pytorch.org/tutorials/beginner/basics/autogradqs_tutorial.html

------- Social Accounts -------

Suraj’s Twitter: twitter.com/subramen

Meta Open Source Twitter: twitter.com/MetaOpenSource
Meta Open Source Facebook Page: facebook.com/metaOpenSource
Meta Open Source Website: opensource.facebook.com
Meta Open Source LinkedIn Page: linkedin.com/showcase/meta-open-source
ELI5: NevergradELI5: React DevToolsELI5: RecoilUnderstanding Physical Interaction from Video with Kristen GraumanThe Diff Episode 17: What you can use Velox for and what you shouldntProject Aria Case Study: University of BristolMigrate from MyPy to Pyrefly in 3 easy stepsProject Aria CVPR 2023 Tutorial: Aria Synthetic Environments Dataset (Section 9 of 10)The Diff Episode 15: Host-Migration in SharedSpacesELI5: DETR3 questions with the Lexical open source team at MetaAdopt Pyrefly gradually by suppressing errors
Meta Open Source |

ELI5: Nevergrad

SHARE TO X SHARE TO REDDIT SHARE TO FACEBOOK WALLPAPER