Completed
Grasping with QT-Opt
Class Central Classrooms beta
YouTube videos curated by Class Central.
Classroom Contents
Deep Reinforcement Learning in the Real World - Sergey Levine
Automatically move to the next video in the Classroom when playback concludes
- 1 Intro
- 2 Deep learning helps us handle unstructured environments
- 3 Reinforcement learning provides a formalism for behavior
- 4 RL has a big problem
- 5 Off-policy RL with large datasets
- 6 Off-policy model-free learning
- 7 How to solve for the Q-function?
- 8 QT-Opt: off-policy Q-learning at scale
- 9 Grasping with QT-Opt
- 10 Emergent grasping strategies
- 11 So what's the problem?
- 12 How to stop training on garbage?
- 13 How well does it work?
- 14 Off-policy model-based reinforcement learning
- 15 High-level algorithm outline
- 16 Model-based RL for dexterous manipulation
- 17 Q-Functions (can) learn models
- 18 Temporal difference models
- 19 Optimizing over valid states
