| --- |
| title: Sample-Efficient Imitation |
| emoji: 🎯 |
| colorFrom: green |
| colorTo: blue |
| sdk: static |
| pinned: true |
| license: mit |
| short_description: Does regularization buy sample-efficiency? (PushT) |
| --- |
| |
| # Sample-Efficient Imitation — live results |
|
|
| Does aggressive regularization buy sample-efficiency for imitation-learning policies? |
| Diffusion Policy on PushT, trained & evaluated on AMD ROCm (Radeon AI PRO R9700). |
|
|
| The page (`index.html`) shows the policy solving the task (real eval rollouts) and the |
| data-efficiency comparison. Code: https://github.com/enyolanev-bit/sample-efficient-imitation |
|
|
