Upload README.md with huggingface_hub
Browse files
README.md
ADDED
|
@@ -0,0 +1,27 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
library_name: lerobot
|
| 3 |
+
tags:
|
| 4 |
+
- lerobot
|
| 5 |
+
- robotics
|
| 6 |
+
- imitation-learning
|
| 7 |
+
- diffusion
|
| 8 |
+
---
|
| 9 |
+
|
| 10 |
+
# Diffusion Policy finetuned on SO-ARM100 dump-pocket-cleaning
|
| 11 |
+
|
| 12 |
+
This is a **Diffusion Policy** finetuned with LeRobot.
|
| 13 |
+
|
| 14 |
+
## Training summary
|
| 15 |
+
- policy: `diffusion`
|
| 16 |
+
- dataset: `Autobrik/SO-ARM100-dump-pocket-cleaning2`
|
| 17 |
+
- run_dir: `outputs/train/benchmark/20260216_022628_benchmark_diffusion`
|
| 18 |
+
- seed: `1000`
|
| 19 |
+
|
| 20 |
+
## Best offline validation
|
| 21 |
+
- best_val_step: `7252`
|
| 22 |
+
- best_val_loss: `0.014062351927789459`
|
| 23 |
+
|
| 24 |
+
## Included artifacts
|
| 25 |
+
- `checkpoints/best` (recommended for eval/deploy)
|
| 26 |
+
- `checkpoints/last` (resume/debug)
|
| 27 |
+
- `best_val.json`
|