Unify tools across various examples - Githubissues

vincekurtz / rddp

Reward-Driven Diffusion Policy

1 stars 0 forks source link

Unify tools across various examples #7

Open vincekurtz opened 3 months ago

vincekurtz commented 3 months ago

Currently there is a lot of duplicate code in the examples folder. Ideally each example would be quite minimal, with common utilities for:

[x] generating a dataset
[ ] visualizing the generated dataset before training
[x] training the diffusion policy
[ ] rolling out the trained model open loop and visualizing the result
[ ] deploying the diffusion policy in a closed-loop simulation