Instructions to use maximellerbach/omx_multicubes_lingbot_va with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- LeRobot
How to use maximellerbach/omx_multicubes_lingbot_va with LeRobot:
- Notebooks
- Google Colab
- Kaggle
| { | |
| "type": "lingbot_va", | |
| "n_obs_steps": 1, | |
| "input_features": { | |
| "observation.images.cam_high": { | |
| "type": "VISUAL", | |
| "shape": [ | |
| 3, | |
| 256, | |
| 256 | |
| ] | |
| }, | |
| "observation.images.cam_left_wrist": { | |
| "type": "VISUAL", | |
| "shape": [ | |
| 3, | |
| 256, | |
| 256 | |
| ] | |
| }, | |
| "observation.images.cam_right_wrist": { | |
| "type": "VISUAL", | |
| "shape": [ | |
| 3, | |
| 256, | |
| 256 | |
| ] | |
| } | |
| }, | |
| "output_features": { | |
| "action": { | |
| "type": "ACTION", | |
| "shape": [ | |
| 6 | |
| ] | |
| } | |
| }, | |
| "device": "cuda", | |
| "use_amp": false, | |
| "use_peft": false, | |
| "push_to_hub": true, | |
| "repo_id": "maximellerbach/omx_multicubes_lingbot_va", | |
| "private": null, | |
| "tags": null, | |
| "license": null, | |
| "pretrained_path": "lerobot/lingbot_va_base", | |
| "pretrained_revision": null, | |
| "patch_size": [ | |
| 1, | |
| 2, | |
| 2 | |
| ], | |
| "num_attention_heads": 24, | |
| "attention_head_dim": 128, | |
| "in_channels": 48, | |
| "out_channels": 48, | |
| "action_dim": 30, | |
| "text_dim": 4096, | |
| "freq_dim": 256, | |
| "ffn_dim": 14336, | |
| "num_layers": 30, | |
| "cross_attn_norm": true, | |
| "eps": 1e-06, | |
| "rope_max_seq_len": 1024, | |
| "attn_mode": "flex", | |
| "wan_pretrained_path": "robbyant/lingbot-va-base", | |
| "dtype": "bfloat16", | |
| "text_encoder_device": "cuda", | |
| "obs_cam_keys": [ | |
| "observation.images.cam_high", | |
| "observation.images.cam_left_wrist" | |
| ], | |
| "image_hflip": false, | |
| "camera_layout": "width_concat", | |
| "height": 256, | |
| "width": 320, | |
| "action_per_frame": 16, | |
| "frame_chunk_size": 2, | |
| "attn_window": 72, | |
| "num_inference_steps": 25, | |
| "video_exec_step": -1, | |
| "action_num_inference_steps": 50, | |
| "guidance_scale": 5.0, | |
| "action_guidance_scale": 1.0, | |
| "snr_shift": 5.0, | |
| "action_snr_shift": 1.0, | |
| "max_sequence_length": 512, | |
| "used_action_channel_ids": [ | |
| 0, | |
| 1, | |
| 2, | |
| 3, | |
| 4, | |
| 5 | |
| ], | |
| "save_predicted_video": false, | |
| "normalization_mapping": { | |
| "VISUAL": "IDENTITY", | |
| "STATE": "IDENTITY", | |
| "ACTION": "QUANTILES" | |
| }, | |
| "optimizer_lr": 1e-05, | |
| "optimizer_betas": [ | |
| 0.9, | |
| 0.95 | |
| ], | |
| "optimizer_eps": 1e-08, | |
| "optimizer_weight_decay": 0.0001, | |
| "optimizer_grad_clip_norm": 1.0, | |
| "scheduler_warmup_steps": 1000 | |
| } |