Instructions to use Tsagkas/vla_state_modulation_v3_variability with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- LeRobot
How to use Tsagkas/vla_state_modulation_v3_variability with LeRobot:
- Notebooks
- Google Colab
- Kaggle
| { | |
| "dataset": { | |
| "repo_id": "Tsagkas/dataset-test-v1", | |
| "root": null, | |
| "episodes": null, | |
| "image_transforms": { | |
| "enable": false, | |
| "max_num_transforms": 3, | |
| "random_order": false, | |
| "tfs": { | |
| "brightness": { | |
| "weight": 1.0, | |
| "type": "ColorJitter", | |
| "kwargs": { | |
| "brightness": [ | |
| 0.8, | |
| 1.2 | |
| ] | |
| } | |
| }, | |
| "contrast": { | |
| "weight": 1.0, | |
| "type": "ColorJitter", | |
| "kwargs": { | |
| "contrast": [ | |
| 0.8, | |
| 1.2 | |
| ] | |
| } | |
| }, | |
| "saturation": { | |
| "weight": 1.0, | |
| "type": "ColorJitter", | |
| "kwargs": { | |
| "saturation": [ | |
| 0.5, | |
| 1.5 | |
| ] | |
| } | |
| }, | |
| "hue": { | |
| "weight": 1.0, | |
| "type": "ColorJitter", | |
| "kwargs": { | |
| "hue": [ | |
| -0.05, | |
| 0.05 | |
| ] | |
| } | |
| }, | |
| "sharpness": { | |
| "weight": 1.0, | |
| "type": "SharpnessJitter", | |
| "kwargs": { | |
| "sharpness": [ | |
| 0.5, | |
| 1.5 | |
| ] | |
| } | |
| }, | |
| "affine": { | |
| "weight": 1.0, | |
| "type": "RandomAffine", | |
| "kwargs": { | |
| "degrees": [ | |
| -5.0, | |
| 5.0 | |
| ], | |
| "translate": [ | |
| 0.05, | |
| 0.05 | |
| ] | |
| } | |
| } | |
| } | |
| }, | |
| "revision": null, | |
| "use_imagenet_stats": true, | |
| "video_backend": "torchcodec", | |
| "streaming": false | |
| }, | |
| "env": null, | |
| "policy": { | |
| "type": "vla", | |
| "n_obs_steps": 1, | |
| "input_features": { | |
| "observation.state": { | |
| "type": "STATE", | |
| "shape": [ | |
| 7 | |
| ] | |
| }, | |
| "observation.images.wrist": { | |
| "type": "VISUAL", | |
| "shape": [ | |
| 3, | |
| 256, | |
| 256 | |
| ] | |
| }, | |
| "observation.images.front": { | |
| "type": "VISUAL", | |
| "shape": [ | |
| 3, | |
| 256, | |
| 256 | |
| ] | |
| }, | |
| "observation.vfm_features.wrist": { | |
| "type": "STATE", | |
| "shape": [ | |
| 16, | |
| 16, | |
| 768 | |
| ] | |
| }, | |
| "observation.vfm_features.front": { | |
| "type": "STATE", | |
| "shape": [ | |
| 16, | |
| 16, | |
| 768 | |
| ] | |
| }, | |
| "observation.skill_feature": { | |
| "type": "STATE", | |
| "shape": [ | |
| 1024 | |
| ] | |
| }, | |
| "observation.object_feature": { | |
| "type": "STATE", | |
| "shape": [ | |
| 1024 | |
| ] | |
| }, | |
| "observation.vlm_features.wrist": { | |
| "type": "STATE", | |
| "shape": [ | |
| 16, | |
| 16, | |
| 1024 | |
| ] | |
| }, | |
| "observation.vlm_features.front": { | |
| "type": "STATE", | |
| "shape": [ | |
| 16, | |
| 16, | |
| 1024 | |
| ] | |
| } | |
| }, | |
| "output_features": { | |
| "action": { | |
| "type": "ACTION", | |
| "shape": [ | |
| 7 | |
| ] | |
| } | |
| }, | |
| "device": "cuda:0", | |
| "use_amp": false, | |
| "push_to_hub": true, | |
| "repo_id": "Tsagkas/vla_state_modulation_v3_variability", | |
| "private": null, | |
| "tags": null, | |
| "license": null, | |
| "pretrained_path": null, | |
| "num_cameras": 2, | |
| "chunk_size": 100, | |
| "n_action_steps": 100, | |
| "state_dim": 7, | |
| "action_dim": 7, | |
| "state_hidden_dim": 256, | |
| "state_output_dim": 768, | |
| "num_transformer_heads": 12, | |
| "transformer_hidden_dim": 768, | |
| "transformer_expansion_factor": 4, | |
| "attention_probs_dropout_prob": 0.0, | |
| "hidden_dropout_prob": 0.1, | |
| "temporal_ensemble_coeff": 0.01, | |
| "optimizer_lr": 0.0001, | |
| "optimizer_weight_decay": 0.0001, | |
| "normalization_mapping": { | |
| "VISUAL": "IDENTITY", | |
| "STATE": "MEAN_STD", | |
| "ACTION": "MEAN_STD" | |
| } | |
| }, | |
| "output_dir": "outputs/train/vla_state_modulation_v3_variability", | |
| "job_name": "vla_state_modulation_v3_variability", | |
| "resume": false, | |
| "seed": 1000, | |
| "num_workers": 16, | |
| "batch_size": 8, | |
| "steps": 50000, | |
| "eval_freq": 20000, | |
| "log_freq": 50, | |
| "save_checkpoint": true, | |
| "save_freq": 25000, | |
| "use_policy_training_preset": true, | |
| "optimizer": { | |
| "type": "adamw", | |
| "lr": 0.0001, | |
| "weight_decay": 0.0001, | |
| "grad_clip_norm": 10.0, | |
| "betas": [ | |
| 0.9, | |
| 0.999 | |
| ], | |
| "eps": 1e-08 | |
| }, | |
| "scheduler": null, | |
| "eval": { | |
| "n_episodes": 50, | |
| "batch_size": 50, | |
| "use_async_envs": false | |
| }, | |
| "wandb": { | |
| "enable": true, | |
| "disable_artifact": false, | |
| "project": "lerobot", | |
| "entity": null, | |
| "notes": null, | |
| "run_id": "jv0vsm8a", | |
| "mode": null | |
| }, | |
| "checkpoint_path": null, | |
| "rename_map": {} | |
| } |