| { |
| "type": "pi05", |
| "n_obs_steps": 1, |
| "input_features": { |
| "observation.images.image": { |
| "type": "VISUAL", |
| "shape": [ |
| 3, |
| 256, |
| 256 |
| ] |
| }, |
| "observation.images.image2": { |
| "type": "VISUAL", |
| "shape": [ |
| 3, |
| 256, |
| 256 |
| ] |
| }, |
| "observation.state": { |
| "type": "STATE", |
| "shape": [ |
| 8 |
| ] |
| }, |
| "observation.images.empty_camera_0": { |
| "type": "VISUAL", |
| "shape": [ |
| 3, |
| 224, |
| 224 |
| ] |
| } |
| }, |
| "output_features": { |
| "action": { |
| "type": "ACTION", |
| "shape": [ |
| 7 |
| ] |
| } |
| }, |
| "device": "cuda", |
| "use_amp": false, |
| "use_peft": false, |
| "push_to_hub": false, |
| "repo_id": "GSR/PI05-LIBERO", |
| "private": null, |
| "tags": null, |
| "license": null, |
| "pretrained_path": null, |
| "paligemma_variant": "gemma_2b", |
| "action_expert_variant": "gemma_300m", |
| "dtype": "bfloat16", |
| "chunk_size": 50, |
| "n_action_steps": 50, |
| "max_state_dim": 32, |
| "max_action_dim": 32, |
| "num_inference_steps": 10, |
| "time_sampling_beta_alpha": 1.5, |
| "time_sampling_beta_beta": 1.0, |
| "time_sampling_scale": 0.999, |
| "time_sampling_offset": 0.001, |
| "min_period": 0.004, |
| "max_period": 4.0, |
| "use_relative_actions": false, |
| "relative_exclude_joints": [ |
| "gripper" |
| ], |
| "action_feature_names": [ |
| "actions" |
| ], |
| "rtc_config": null, |
| "image_resolution": [ |
| 224, |
| 224 |
| ], |
| "empty_cameras": 1, |
| "tokenizer_max_length": 64, |
| "normalization_mapping": { |
| "ACTION": "MEAN_STD", |
| "STATE": "MEAN_STD", |
| "VISUAL": "IDENTITY" |
| }, |
| "gradient_checkpointing": true, |
| "compile_model": false, |
| "compile_mode": "max-autotune", |
| "freeze_vision_encoder": true, |
| "train_expert_only": true, |
| "pi05_t5_bottleneck": true, |
| "pi05_neutral_prompt": "perform the task", |
| "pi05_t5_model": "google-t5/t5-large", |
| "pi05_t5_max_length": 16, |
| "pi05_paligemma_tokenizer": "google/paligemma-3b-pt-224", |
| "pi05_reinit_binding_after_load": true, |
| "pi05_t5_cross_attention": false, |
| "pi05_t5_cross_attention_gate_init": 0.05, |
| "pi05_t5_cross_attention_strict_vla_adapter": false, |
| "pi05_ground_t5_in_native_slots": true, |
| "pi05_ground_t5_preserve_native_prompt": true, |
| "pi05_native_grounder_num_heads": 8, |
| "pi05_native_grounder_gate_init": 1.0, |
| "pi05_t5_eval_mode": "normal", |
| "pi05_wrong_t5_text": "turn on the stove", |
| "pi05_t5_preserve_native_prompt": false, |
| "pi05_motor_preserving_distillation": false, |
| "pi05_motor_distillation_weight": 1.0, |
| "pi05_motor_preserving_curriculum": false, |
| "pi05_motor_curriculum_start_step": 0, |
| "pi05_motor_curriculum_end_step": 3500, |
| "pi05_motor_curriculum_flow_weight": 1.0, |
| "pi05_motor_rollout_distillation_weight": 1.0, |
| "pi05_motor_rollout_distillation_every": 1, |
| "pi05_motor_rollout_distillation_batch_size": 1, |
| "pi05_interface_matching": false, |
| "pi05_interface_loss_weight": 1.0, |
| "pi05_interface_velocity_weight": 0.0, |
| "pi05_native_soft_prefix": false, |
| "pi05_native_soft_prefix_tokens": 32, |
| "pi05_native_soft_prefix_layers": 2, |
| "pi05_native_soft_prefix_heads": 16, |
| "pi05_native_soft_prefix_mode": "append", |
| "pi05_native_soft_prefix_start_index": 3, |
| "pi05_native_soft_prefix_matching": false, |
| "pi05_native_soft_prefix_interface_weight": 1.0, |
| "pi05_native_soft_prefix_velocity_weight": 0.1, |
| "pi05_native_soft_prefix_flow_weight": 0.0, |
| "pi05_native_soft_prefix_embedding_weight": 0.0, |
| "pi05_train_paraphrase_distillation": false, |
| "pi05_train_paraphrases": "", |
| "pi05_eval_paraphrases": "", |
| "pi05_train_canonical_probability": 0.25, |
| "pi05_translator_consistency_weight": 0.0, |
| "pi05_native_soft_prefix_inference_blend": 1.0, |
| "pi05_native_soft_prefix_embedding_blend": 1.0, |
| "pi05_native_soft_prefix_training_blend_min": 1.0, |
| "pi05_native_soft_prefix_training_blend_max": 1.0, |
| "pi05_native_soft_prefix_geometry_preserving": false, |
| "pi05_native_soft_prefix_sentence_bottleneck": false, |
| "pi05_native_soft_prefix_zero_init": false, |
| "pi05_native_kv_transplant": false, |
| "pi05_native_kv_matching": false, |
| "pi05_native_kv_neutral_conditioning": false, |
| "pi05_native_kv_visual_conditioning": false, |
| "pi05_native_kv_visual_summary_tokens": 8, |
| "pi05_native_kv_resampler_dim": 512, |
| "pi05_native_kv_resampler_layers": 2, |
| "pi05_native_kv_resampler_heads": 8, |
| "pi05_native_kv_weight": 1.0, |
| "pi05_native_kv_interface_weight": 0.1, |
| "pi05_native_kv_velocity_weight": 0.1, |
| "optimizer_lr": 0.0001, |
| "optimizer_betas": [ |
| 0.9, |
| 0.95 |
| ], |
| "optimizer_eps": 1e-08, |
| "optimizer_weight_decay": 0.01, |
| "optimizer_grad_clip_norm": 1.0, |
| "scheduler_warmup_steps": 125, |
| "scheduler_decay_steps": 6250, |
| "scheduler_decay_lr": 5e-06, |
| "gsr_enable": true, |
| "gsr_preserve_native_language": true, |
| "gsr_t5_model": "google-t5/t5-large", |
| "gsr_t5_max_length": 16, |
| "gsr_paligemma_tokenizer": "google/paligemma-3b-pt-224", |
| "gsr_reinitialize_action_binding": true, |
| "gsr_grounder_num_heads": 8, |
| "gsr_grounder_gate_init": 1.0 |
| } |
|
|
