Update config.json
Browse files- config.json +48 -48
config.json
CHANGED
|
@@ -1,49 +1,49 @@
|
|
| 1 |
-
{
|
| 2 |
-
"chunk_size": 100,
|
| 3 |
-
"dim_feedforward": 3200,
|
| 4 |
-
"dim_model": 512,
|
| 5 |
-
"dropout": 0.1,
|
| 6 |
-
"feedforward_activation": "relu",
|
| 7 |
-
"input_normalization_modes": {
|
| 8 |
-
"observation.images.laptop": "mean_std",
|
| 9 |
-
"observation.images.phone": "mean_std",
|
| 10 |
-
"observation.state": "mean_std"
|
| 11 |
-
},
|
| 12 |
-
"input_shapes": {
|
| 13 |
-
"observation.images.laptop": [
|
| 14 |
-
3,
|
| 15 |
-
480,
|
| 16 |
-
640
|
| 17 |
-
],
|
| 18 |
-
"observation.images.phone": [
|
| 19 |
-
3,
|
| 20 |
-
480,
|
| 21 |
-
640
|
| 22 |
-
],
|
| 23 |
-
"observation.state": [
|
| 24 |
-
6
|
| 25 |
-
]
|
| 26 |
-
},
|
| 27 |
-
"kl_weight": 10.0,
|
| 28 |
-
"latent_dim": 32,
|
| 29 |
-
"n_action_steps": 100,
|
| 30 |
-
"n_decoder_layers": 1,
|
| 31 |
-
"n_encoder_layers": 4,
|
| 32 |
-
"n_heads": 8,
|
| 33 |
-
"n_obs_steps": 1,
|
| 34 |
-
"n_vae_encoder_layers": 4,
|
| 35 |
-
"output_normalization_modes": {
|
| 36 |
-
"action": "mean_std"
|
| 37 |
-
},
|
| 38 |
-
"output_shapes": {
|
| 39 |
-
"action": [
|
| 40 |
-
6
|
| 41 |
-
]
|
| 42 |
-
},
|
| 43 |
-
"pre_norm": false,
|
| 44 |
-
"pretrained_backbone_weights": "ResNet18_Weights.IMAGENET1K_V1",
|
| 45 |
-
"replace_final_stride_with_dilation": false,
|
| 46 |
-
"temporal_ensemble_coeff": null,
|
| 47 |
-
"use_vae": true,
|
| 48 |
-
"vision_backbone": "resnet18"
|
| 49 |
}
|
|
|
|
| 1 |
+
{ "type": "act",
|
| 2 |
+
"chunk_size": 100,
|
| 3 |
+
"dim_feedforward": 3200,
|
| 4 |
+
"dim_model": 512,
|
| 5 |
+
"dropout": 0.1,
|
| 6 |
+
"feedforward_activation": "relu",
|
| 7 |
+
"input_normalization_modes": {
|
| 8 |
+
"observation.images.laptop": "mean_std",
|
| 9 |
+
"observation.images.phone": "mean_std",
|
| 10 |
+
"observation.state": "mean_std"
|
| 11 |
+
},
|
| 12 |
+
"input_shapes": {
|
| 13 |
+
"observation.images.laptop": [
|
| 14 |
+
3,
|
| 15 |
+
480,
|
| 16 |
+
640
|
| 17 |
+
],
|
| 18 |
+
"observation.images.phone": [
|
| 19 |
+
3,
|
| 20 |
+
480,
|
| 21 |
+
640
|
| 22 |
+
],
|
| 23 |
+
"observation.state": [
|
| 24 |
+
6
|
| 25 |
+
]
|
| 26 |
+
},
|
| 27 |
+
"kl_weight": 10.0,
|
| 28 |
+
"latent_dim": 32,
|
| 29 |
+
"n_action_steps": 100,
|
| 30 |
+
"n_decoder_layers": 1,
|
| 31 |
+
"n_encoder_layers": 4,
|
| 32 |
+
"n_heads": 8,
|
| 33 |
+
"n_obs_steps": 1,
|
| 34 |
+
"n_vae_encoder_layers": 4,
|
| 35 |
+
"output_normalization_modes": {
|
| 36 |
+
"action": "mean_std"
|
| 37 |
+
},
|
| 38 |
+
"output_shapes": {
|
| 39 |
+
"action": [
|
| 40 |
+
6
|
| 41 |
+
]
|
| 42 |
+
},
|
| 43 |
+
"pre_norm": false,
|
| 44 |
+
"pretrained_backbone_weights": "ResNet18_Weights.IMAGENET1K_V1",
|
| 45 |
+
"replace_final_stride_with_dilation": false,
|
| 46 |
+
"temporal_ensemble_coeff": null,
|
| 47 |
+
"use_vae": true,
|
| 48 |
+
"vision_backbone": "resnet18"
|
| 49 |
}
|