---
library_name: lerobot
tags:
- lerobot
- robotics
- imitation-learning
- diffusion
---

# Diffusion Policy finetuned on SO-ARM100 dump-pocket-cleaning

This is a **Diffusion Policy** finetuned with LeRobot.

## Training summary
- policy: `diffusion`
- dataset: `Autobrik/SO-ARM100-dump-pocket-cleaning2`
- run_dir: `outputs/train/benchmark/20260216_022628_benchmark_diffusion`
- seed: `1000`

## Best offline validation
- best_val_step: `7252`
- best_val_loss: `0.014062351927789459`

## Included artifacts
- `checkpoints/best` (recommended for eval/deploy)
- `checkpoints/last` (resume/debug)
- `best_val.json`