Push model using huggingface_hub.
Browse files- README.md +12 -0
- config.json +83 -0
- model.safetensors +3 -0
README.md
ADDED
|
@@ -0,0 +1,12 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
library_name: lerobot
|
| 3 |
+
tags:
|
| 4 |
+
- act
|
| 5 |
+
- model_hub_mixin
|
| 6 |
+
- pytorch_model_hub_mixin
|
| 7 |
+
- robotics
|
| 8 |
+
---
|
| 9 |
+
|
| 10 |
+
This model has been pushed to the Hub using the [PytorchModelHubMixin](https://huggingface.co/docs/huggingface_hub/package_reference/mixins#huggingface_hub.PyTorchModelHubMixin) integration:
|
| 11 |
+
- Library: https://github.com/huggingface/lerobot
|
| 12 |
+
- Docs: [More Information Needed]
|
config.json
ADDED
|
@@ -0,0 +1,83 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"act_attn_layers": null,
|
| 3 |
+
"act_attn_loss_weights": null,
|
| 4 |
+
"act_attn_sigma": 25.0,
|
| 5 |
+
"chunk_size": 33,
|
| 6 |
+
"dim_feedforward": 3200,
|
| 7 |
+
"dim_gaze_decoder": 512,
|
| 8 |
+
"dim_gaze_decoder_feedforward": 3200,
|
| 9 |
+
"dim_model": 512,
|
| 10 |
+
"dinov2_attn_layers": null,
|
| 11 |
+
"dinov2_attn_loss_weights": null,
|
| 12 |
+
"dinov2_attn_sigma": 25.0,
|
| 13 |
+
"dropout": 0.1,
|
| 14 |
+
"eyes": {
|
| 15 |
+
"observation.left_eye": "observation.images.zed_cam_left"
|
| 16 |
+
},
|
| 17 |
+
"feedforward_activation": "relu",
|
| 18 |
+
"freeze_backbone": false,
|
| 19 |
+
"gaze_loss_weight": 1.0,
|
| 20 |
+
"gaze_sigma": 50.0,
|
| 21 |
+
"image_size": [
|
| 22 |
+
480,
|
| 23 |
+
640
|
| 24 |
+
],
|
| 25 |
+
"input_normalization_modes": {
|
| 26 |
+
"observation.images.left_eye_cam": "mean_std",
|
| 27 |
+
"observation.images.right_eye_cam": "mean_std",
|
| 28 |
+
"observation.images.wrist_cam_left": "mean_std",
|
| 29 |
+
"observation.images.wrist_cam_right": "mean_std",
|
| 30 |
+
"observation.state": "mean_std"
|
| 31 |
+
},
|
| 32 |
+
"input_shapes": {
|
| 33 |
+
"observation.images.left_eye_cam": [
|
| 34 |
+
3,
|
| 35 |
+
480,
|
| 36 |
+
640
|
| 37 |
+
],
|
| 38 |
+
"observation.images.right_eye_cam": [
|
| 39 |
+
3,
|
| 40 |
+
480,
|
| 41 |
+
640
|
| 42 |
+
],
|
| 43 |
+
"observation.images.wrist_cam_left": [
|
| 44 |
+
3,
|
| 45 |
+
480,
|
| 46 |
+
640
|
| 47 |
+
],
|
| 48 |
+
"observation.images.wrist_cam_right": [
|
| 49 |
+
3,
|
| 50 |
+
480,
|
| 51 |
+
640
|
| 52 |
+
],
|
| 53 |
+
"observation.state": [
|
| 54 |
+
21
|
| 55 |
+
]
|
| 56 |
+
},
|
| 57 |
+
"kl_weight": 10.0,
|
| 58 |
+
"latent_dim": 32,
|
| 59 |
+
"n_action_steps": 33,
|
| 60 |
+
"n_decoder_layers": 1,
|
| 61 |
+
"n_encoder_layers": 4,
|
| 62 |
+
"n_gaze_decoder_layers": 1,
|
| 63 |
+
"n_heads": 8,
|
| 64 |
+
"n_obs_steps": 1,
|
| 65 |
+
"n_vae_encoder_layers": 4,
|
| 66 |
+
"output_normalization_modes": {
|
| 67 |
+
"action": "mean_std"
|
| 68 |
+
},
|
| 69 |
+
"output_shapes": {
|
| 70 |
+
"action": [
|
| 71 |
+
21
|
| 72 |
+
]
|
| 73 |
+
},
|
| 74 |
+
"pre_norm": false,
|
| 75 |
+
"pretrained_backbone_weights": "ResNet18_Weights.IMAGENET1K_V1",
|
| 76 |
+
"replace_final_stride_with_dilation": false,
|
| 77 |
+
"temporal_ensemble_coeff": null,
|
| 78 |
+
"use_act_attn": false,
|
| 79 |
+
"use_dinov2_attn": false,
|
| 80 |
+
"use_gaze": false,
|
| 81 |
+
"use_vae": true,
|
| 82 |
+
"vision_backbone": "resnet18"
|
| 83 |
+
}
|
model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:4aaf4ed30c5a9232a725bbd4bf9fbd3eb29cbebfdaf0a2acd1104ff755bec35d
|
| 3 |
+
size 206550556
|