Commit
•
9c2b1d0
1
Parent(s):
6964520
Push model using huggingface_hub.
Browse files- README.md +6 -36
- config.json +49 -0
- pytorch_model.bin +3 -0
README.md
CHANGED
@@ -1,39 +1,9 @@
|
|
1 |
---
|
2 |
-
|
|
|
|
|
3 |
---
|
4 |
-
# RDT-1B
|
5 |
-
|
6 |
-
RDT-1B is a 1B-parameter imitation learning Diffusion Transformer pre-trained on 1M+ multi-robot episodes. Given a language instruction and 3-view RGB image observations, RDT can predict the next
|
7 |
-
64 robot actions. RDT is inherently compatible with almost all kinds of modern mobile manipulators, from single-arm to dual-arm, joint to EEF, pos. to vel., and even with a mobile chassis.
|
8 |
-
|
9 |
-
All the code and model weights are licensed under MIT license.
|
10 |
-
|
11 |
-
Please refer to our [project page](), [github repository]() and [paper]() for more information.
|
12 |
-
|
13 |
-
## Model Details
|
14 |
-
|
15 |
-
- **Developed by** Thu-ml team
|
16 |
-
- **License:** MIT
|
17 |
-
- **Pretrain dataset:** [More Information Needed]
|
18 |
-
- **Finetune dataset:** [More Information Needed]
|
19 |
-
|
20 |
-
- **Repository:** [More Information Needed]
|
21 |
-
- **Paper :** [More Information Needed]
|
22 |
-
- **Project Page:** https://rdt-robotics.github.io/rdt-robotics/
|
23 |
-
|
24 |
-
## Uses
|
25 |
-
|
26 |
-
RDT-1B supports finetuning and pre-training on custom dataset, as well as deploying and inferencing on real-robots.
|
27 |
-
|
28 |
-
Please refer to [our repository](https://github.com/GeneralEmbodiedSystem/RoboticsDiffusionTransformer/blob/main/docs/pretrain.md) for all the above guides.
|
29 |
-
|
30 |
-
|
31 |
-
## Citation
|
32 |
-
|
33 |
-
<!-- If there is a paper or blog post introducing the model, the APA and Bibtex information for that should go in this section. -->
|
34 |
-
|
35 |
-
**BibTeX:**
|
36 |
-
|
37 |
-
[More Information Needed]
|
38 |
-
|
39 |
|
|
|
|
|
|
|
|
1 |
---
|
2 |
+
tags:
|
3 |
+
- pytorch_model_hub_mixin
|
4 |
+
- model_hub_mixin
|
5 |
---
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
6 |
|
7 |
+
This model has been pushed to the Hub using the [PytorchModelHubMixin](https://huggingface.co/docs/huggingface_hub/package_reference/mixins#huggingface_hub.PyTorchModelHubMixin) integration:
|
8 |
+
- Library: https://huggingface.co/robotics-diffusion-transformer/rdt-1b
|
9 |
+
- Docs: [More Information Needed]
|
config.json
ADDED
@@ -0,0 +1,49 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"action_dim": 128,
|
3 |
+
"ema": {
|
4 |
+
"inv_gamma": 1.0,
|
5 |
+
"max_value": 0.9999,
|
6 |
+
"min_value": 0.0,
|
7 |
+
"power": 0.75,
|
8 |
+
"update_after_step": 0
|
9 |
+
},
|
10 |
+
"img_adaptor": "mlp2x_gelu",
|
11 |
+
"img_cond_len": 4374,
|
12 |
+
"img_pos_embed_config": [
|
13 |
+
[
|
14 |
+
"image",
|
15 |
+
[
|
16 |
+
2,
|
17 |
+
3,
|
18 |
+
-729
|
19 |
+
]
|
20 |
+
]
|
21 |
+
],
|
22 |
+
"img_token_dim": 1152,
|
23 |
+
"lang_adaptor": "mlp2x_gelu",
|
24 |
+
"lang_pos_embed_config": [
|
25 |
+
[
|
26 |
+
"lang",
|
27 |
+
-1024
|
28 |
+
]
|
29 |
+
],
|
30 |
+
"lang_token_dim": 4096,
|
31 |
+
"max_lang_cond_len": 1024,
|
32 |
+
"noise_scheduler": {
|
33 |
+
"beta_schedule": "squaredcos_cap_v2",
|
34 |
+
"clip_sample": false,
|
35 |
+
"num_inference_timesteps": 5,
|
36 |
+
"num_train_timesteps": 1000,
|
37 |
+
"prediction_type": "sample",
|
38 |
+
"type": "ddpm"
|
39 |
+
},
|
40 |
+
"pred_horizon": 64,
|
41 |
+
"rdt": {
|
42 |
+
"cond_pos_embed_type": "multimodal",
|
43 |
+
"depth": 28,
|
44 |
+
"hidden_size": 2048,
|
45 |
+
"num_heads": 32
|
46 |
+
},
|
47 |
+
"state_adaptor": "mlp3x_gelu",
|
48 |
+
"state_token_dim": 128
|
49 |
+
}
|
pytorch_model.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:c5107b2ebbaa4edadb26cf40c4d1990a16e0f67d6137925523eb90cf4b8fdaca
|
3 |
+
size 2456755578
|