File size: 1,214 Bytes
7c0ded9
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
---
library_name: lerobot
tags:
- model_hub_mixin
- pytorch_model_hub_mixin
- robotics
- dot
license: apache-2.0
datasets:
- lerobot/aloha_sim_insertion_human
pipeline_tag: robotics
---

# Model Card for "Decoder Only Transformer (DOT) Policy" for ALOHA bimanual insert problem

Read more about the model and implementation details in the [DOT Policy repository](https://github.com/IliaLarchenko/dot_policy).

This model is trained using the [LeRobot library](https://huggingface.co/lerobot) and achieves state-of-the-art results on behavior cloning on ALOHA bimanual insert dataset. It achieves 29.6% success rate vs. 21% for the previous state-of-the-art model (ACT).

This result is achieved without the checkpoint selection and is easy to reproduce.

You can use this model by installing LeRobot from [this branch](https://github.com/IliaLarchenko/lerobot/tree/dot)

To train the model:

```bash
python lerobot/scripts/train.py policy=dot_insert env=aloha env.episode_length=500
```

To evaluate the model:

```bash
python lerobot/scripts/eval.py -p IliaLarchenko/dot_bimanual_insert eval.n_episodes=1000 eval.batch_size=100 seed=1000000
```

Model size:
- Total parameters: 14.1m
- Trainable parameters: 2.9m