Grabber live lab + Reacher archive lineage
A live browser-native continuous-control lab now sits beside the deeper Reacher archive, giving DRL a modern entrypoint without reviving Unity first.
How It Works
The new live lane is Grabber: a custom 2D arm task with continuous actions, machine playback, and bounded PPO training. Behind it, the archive still holds the course DDPG tutorials, multiple Reacher implementations, single-agent and twenty-agent project variants, vectorized environment experiments, PPO/REINFORCE side paths, sample transition CSVs, and the vendored ML-Agents Python package.
Why Reuse
This branch now connects a live, understandable continuous-control lab to the heavier historical Reacher work, so users can learn the control ideas first and then trace the lineage back into the archive.
Key Assets
Browser-rendered 2D continuous-control lab with PPO training, checkpoint playback, and learning snapshots.
drl_web
Main notebooks, data, docs, env zips, and multiple code paths.
source-material/classwork/project-reports/p2_continuous-control
Core Reacher DDPG agent, model, and training loop.
source-material/classwork/project-reports/p2_continuous-control/ddpg
Simpler Reacher variant for the one-agent environment.
source-material/classwork/project-reports/p2_continuous-control/ddpg_single_agent
Saved states, actions, rewards, dones, and next-state snapshots.
source-material/classwork/project-reports/p2_continuous-control/data
Bundled Windows executables for one-agent and twenty-agent Reacher.
source-material/classwork/project-reports/p2_continuous-control/zips