| { |
| "type": "lavla", |
| "n_obs_steps": 1, |
| "input_features": { |
| "observation.state": { |
| "type": "STATE", |
| "shape": [ |
| 6 |
| ] |
| }, |
| "observation.images.camera1": { |
| "type": "VISUAL", |
| "shape": [ |
| 3, |
| 480, |
| 640 |
| ] |
| }, |
| "observation.images.camera2": { |
| "type": "VISUAL", |
| "shape": [ |
| 3, |
| 480, |
| 640 |
| ] |
| }, |
| "observation.images.camera3": { |
| "type": "VISUAL", |
| "shape": [ |
| 3, |
| 480, |
| 640 |
| ] |
| } |
| }, |
| "output_features": { |
| "action": { |
| "type": "ACTION", |
| "shape": [ |
| 6 |
| ] |
| } |
| }, |
| "device": "cuda", |
| "use_amp": false, |
| "use_peft": false, |
| "push_to_hub": true, |
| "repo_id": "Alkatt/LAVLA_PickStackSort_RedBlueCube_test", |
| "private": null, |
| "tags": null, |
| "license": null, |
| "pretrained_path": null, |
| "s1_only": false, |
| "s2_backend_type": "local", |
| "s2_remote_server_ip": "0.0.0.0", |
| "s2_remote_server_port": "0000", |
| "s2_blocking_mode": false, |
| "normalization_mapping": { |
| "VISUAL": "MEAN_STD", |
| "STATE": "MEAN_STD", |
| "ACTION": "MEAN_STD" |
| }, |
| "n_action_steps": 50, |
| "chunk_size": 50, |
| "max_action_dim": 6, |
| "s1_freeze_vision_encoder": true, |
| "s1_image_resolution": [ |
| 256, |
| 256 |
| ], |
| "s1_mlp_dim": 2048, |
| "s1_embedding_dim": 1024, |
| "s1_num_attention_heads": 8, |
| "s1_num_layers": 8, |
| "s1_flow_integration_steps": 10, |
| "s1_efficient_net": "google/efficientnet-b0", |
| "s1_add_action_pos_embeddings": true, |
| "s2_vlm_model": "HuggingFaceTB/SmolVLM2-500M-Video-Instruct", |
| "s2_freeze_vlm": true, |
| "optimizer_type": "linearwarmup_cosinedecay", |
| "optimizer_lr": 0.0001, |
| "optimizer_weight_decay": 1e-10, |
| "optimzer_grad_clip_norm": 10, |
| "optimizer_decay_lr": 2.5e-06, |
| "optimizer_betas": [ |
| 0.9, |
| 0.95 |
| ], |
| "optimizer_eps": 1e-10 |
| } |
