Upload policy with preprocessor, postprocessor, and model card

README.md

@@ -0,0 +1,70 @@
+---
+datasets: sengi/flatten_towel_pi05_adv
+library_name: lerobot
+license: apache-2.0
+model_name: pi05
+pipeline_tag: robotics
+tags:
+- lerobot
+- pi05
+- robotics
+---
+
+# Model Card for pi05
+
+<!-- Provide a quick summary of what the model is/does. -->
+
+**π₀.₅ (Pi05) Policy**
+
+π₀.₅ (Pi05) Policy
+
+π₀.₅ is a Vision-Language-Action model with open-world generalization, from Physical Intelligence. The LeRobot implementation is adapted from their open source OpenPI repository.
+
+**Model Overview**
+
+π₀.₅ represents a significant evolution from π₀, developed by Physical Intelligence to address a big challenge in robotics: open-world generalization. While robots can perform impressive tasks in controlled environments, π₀.₅ is designed to generalize to entirely new environments and situations that were never seen during training.
+
+For more details, see the [Physical Intelligence π₀.₅ blog post](https://www.physicalintelligence.company/blog/pi05).
+
+This policy has been trained and pushed to the Hub using [LeRobot](https://github.com/huggingface/lerobot).
+See the full documentation at [LeRobot Docs](https://huggingface.co/docs/lerobot/index).
+
+## Model Details
+
+- **License:** apache-2.0
+
+---
+
+## How to Get Started with the Model
+
+For a complete walkthrough, see the [training guide](https://huggingface.co/docs/lerobot/il_robots#train-a-policy).
+Below is the short version on how to train and run inference/eval:
+
+### Train from scratch
+
+```bash
+lerobot-train \
+  --dataset.repo_id=${HF_USER}/<dataset> \
+  --policy.type=pi05 \
+  --output_dir=outputs/train/<desired_policy_repo_id> \
+  --job_name=lerobot_training \
+  --policy.device=cuda \
+  --policy.repo_id=${HF_USER}/<desired_policy_repo_id>
+  --wandb.enable=true
+```
+
+_Writes checkpoints to `outputs/train/<desired_policy_repo_id>/checkpoints/`._
+
+### Evaluate the policy/run inference
+
+```bash
+lerobot-record \
+  --robot.type=so100_follower \
+  --dataset.repo_id=<hf_user>/eval_<dataset> \
+  --policy.path=<hf_user>/<desired_policy_repo_id> \
+  --episodes=10
+```
+
+Prefix the dataset repo with **eval\_** and supply `--policy.path` pointing to a local or hub checkpoint.
+
+---

config.json

@@ -0,0 +1,78 @@
+{
+    "type": "pi05",
+    "n_obs_steps": 1,
+    "normalization_mapping": {
+        "VISUAL": "IDENTITY",
+        "STATE": "QUANTILES",
+        "ACTION": "QUANTILES"
+    },
+    "input_features": {
+        "observation.state": {
+            "type": "STATE",
+            "shape": [
+                14
+            ]
+        },
+        "observation.images.right": {
+            "type": "VISUAL",
+            "shape": [
+                3,
+                224,
+                224
+            ]
+        },
+        "observation.images.left": {
+            "type": "VISUAL",
+            "shape": [
+                3,
+                224,
+                224
+            ]
+        },
+        "observation.images.top": {
+            "type": "VISUAL",
+            "shape": [
+                3,
+                224,
+                224
+            ]
+        }
+    },
+    "output_features": {
+        "action": {
+            "type": "ACTION",
+            "shape": [
+                14
+            ]
+        }
+    },
+    "device": "cuda:0",
+    "use_amp": false,
+    "push_to_hub": true,
+    "repo_id": "sengi/rabc_pi05",
+    "private": false,
+    "tags": null,
+    "license": null,
+    "pretrained_path": "thomas0829/folding_towel_pi05",
+    "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,
+    "rtc_config": null,
+    "image_resolution": [
+        224,
+        224
+    ],
+    "empty_cameras": 0,
+    "tokenizer_max_length": 200
+}

model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:62cf17961c1f1fe2045dfdd3317a5b33e8d8d6547b466ee70bd2cd99c5f75e5a
+size 7473096344

policy_postprocessor.json

@@ -0,0 +1,32 @@
+{
+  "name": "policy_postprocessor",
+  "steps": [
+    {
+      "registry_name": "unnormalizer_processor",
+      "config": {
+        "eps": 1e-08,
+        "features": {
+          "action": {
+            "type": "ACTION",
+            "shape": [
+              14
+            ]
+          }
+        },
+        "norm_map": {
+          "VISUAL": "IDENTITY",
+          "STATE": "QUANTILES",
+          "ACTION": "QUANTILES"
+        }
+      },
+      "state_file": "policy_postprocessor_step_0_unnormalizer_processor.safetensors"
+    },
+    {
+      "registry_name": "device_processor",
+      "config": {
+        "device": "cpu",
+        "float_dtype": null
+      }
+    }
+  ]
+}

policy_postprocessor_step_0_unnormalizer_processor.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a94373fec64a95619f0af85e29f6c74a69e2674b06c055c8434da48d028df153
+size 9168

policy_preprocessor.json

@@ -0,0 +1,87 @@
+{
+  "name": "policy_preprocessor",
+  "steps": [
+    {
+      "registry_name": "rename_observations_processor",
+      "config": {
+        "rename_map": {}
+      }
+    },
+    {
+      "registry_name": "to_batch_processor",
+      "config": {}
+    },
+    {
+      "registry_name": "normalizer_processor",
+      "config": {
+        "eps": 1e-08,
+        "features": {
+          "observation.state": {
+            "type": "STATE",
+            "shape": [
+              14
+            ]
+          },
+          "observation.images.right": {
+            "type": "VISUAL",
+            "shape": [
+              3,
+              224,
+              224
+            ]
+          },
+          "observation.images.left": {
+            "type": "VISUAL",
+            "shape": [
+              3,
+              224,
+              224
+            ]
+          },
+          "observation.images.top": {
+            "type": "VISUAL",
+            "shape": [
+              3,
+              224,
+              224
+            ]
+          },
+          "action": {
+            "type": "ACTION",
+            "shape": [
+              14
+            ]
+          }
+        },
+        "norm_map": {
+          "VISUAL": "IDENTITY",
+          "STATE": "QUANTILES",
+          "ACTION": "QUANTILES"
+        }
+      },
+      "state_file": "policy_preprocessor_step_2_normalizer_processor.safetensors"
+    },
+    {
+      "registry_name": "pi05_prepare_state_tokenizer_processor_step",
+      "config": {}
+    },
+    {
+      "registry_name": "tokenizer_processor",
+      "config": {
+        "max_length": 200,
+        "task_key": "task",
+        "padding_side": "right",
+        "padding": "max_length",
+        "truncation": true,
+        "tokenizer_name": "google/paligemma-3b-pt-224"
+      }
+    },
+    {
+      "registry_name": "device_processor",
+      "config": {
+        "device": "cuda",
+        "float_dtype": null
+      }
+    }
+  ]
+}

policy_preprocessor_step_2_normalizer_processor.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a94373fec64a95619f0af85e29f6c74a69e2674b06c055c8434da48d028df153
+size 9168

train_config.json

@@ -0,0 +1,244 @@
+{
+    "dataset": {
+        "repo_id": "sengi/flatten_towel_pi05_adv",
+        "repo_ids": null,
+        "root": null,
+        "episodes": null,
+        "image_transforms": {
+            "enable": true,
+            "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",
+        "force_cache_sync": true,
+        "use_annotated_tasks": false
+    },
+    "num_datasets": 100,
+    "env": null,
+    "policy": {
+        "type": "pi05",
+        "n_obs_steps": 1,
+        "normalization_mapping": {
+            "VISUAL": "IDENTITY",
+            "STATE": "QUANTILES",
+            "ACTION": "QUANTILES"
+        },
+        "input_features": {
+            "observation.state": {
+                "type": "STATE",
+                "shape": [
+                    14
+                ]
+            },
+            "observation.images.right": {
+                "type": "VISUAL",
+                "shape": [
+                    3,
+                    224,
+                    224
+                ]
+            },
+            "observation.images.left": {
+                "type": "VISUAL",
+                "shape": [
+                    3,
+                    224,
+                    224
+                ]
+            },
+            "observation.images.top": {
+                "type": "VISUAL",
+                "shape": [
+                    3,
+                    224,
+                    224
+                ]
+            }
+        },
+        "output_features": {
+            "action": {
+                "type": "ACTION",
+                "shape": [
+                    14
+                ]
+            }
+        },
+        "device": "cuda:2",
+        "use_amp": false,
+        "compiled": false,
+        "push_to_hub": true,
+        "repo_id": "sengi/rabc_pi05",
+        "private": false,
+        "tags": null,
+        "license": null,
+        "pretrained_path": "thomas0829/folding_towel_pi05",
+        "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,
+        "rtc_config": null,
+        "image_resolution": [
+            224,
+            224
+        ],
+        "empty_cameras": 0,
+        "tokenizer_max_length": 200,
+        "gradient_checkpointing": true,
+        "compile_model": false,
+        "compile_mode": "max-autotune",
+        "attention_implementation": "eager",
+        "use_lora": false,
+        "lora_rank": 16,
+        "lora_alpha": 32.0,
+        "lora_dropout": 0.1,
+        "lora_target_modules": null,
+        "optimizer_lr": 1e-05,
+        "optimizer_betas": [
+            0.9,
+            0.95
+        ],
+        "optimizer_eps": 1e-08,
+        "optimizer_weight_decay": 0.01,
+        "optimizer_grad_clip_norm": 1.0,
+        "scheduler_warmup_steps": 1000,
+        "scheduler_decay_steps": 10000,
+        "scheduler_decay_lr": 1e-07
+    },
+    "compile": true,
+    "strict": true,
+    "loss_threshold": 3.0,
+    "output_dir": "outputs/train/2026-02-20/20-11-34_rabc_pi05_finetuning",
+    "job_name": "rabc_pi05_finetuning",
+    "resume": false,
+    "resume_scheduler": true,
+    "seed": 3407,
+    "num_workers": 4,
+    "batch_size": 16,
+    "gradient_accumulation_steps": 1,
+    "steps": 10000,
+    "eval_freq": 20000,
+    "log_freq": 5,
+    "save_checkpoint": true,
+    "push_to_hub": false,
+    "repo_id": null,
+    "save_freq": 1000,
+    "use_policy_training_preset": true,
+    "optimizer": {
+        "type": "adamw",
+        "lr": 1e-05,
+        "weight_decay": 0.01,
+        "grad_clip_norm": 1.0,
+        "betas": [
+            0.9,
+            0.95
+        ],
+        "eps": 1e-08
+    },
+    "scheduler": {
+        "type": "cosine_decay_with_warmup",
+        "num_warmup_steps": 1000,
+        "num_decay_steps": 10000,
+        "peak_lr": 1e-05,
+        "decay_lr": 1e-07
+    },
+    "eval": {
+        "n_episodes": 50,
+        "batch_size": 50,
+        "use_async_envs": false
+    },
+    "wandb": {
+        "enable": true,
+        "disable_artifact": true,
+        "project": "lerobot",
+        "entity": "carl_uw",
+        "notes": "Fine-tuning pi05 on put_the_doll_into_the_box_adv0_ft dataset",
+        "run_id": null,
+        "mode": null
+    },
+    "test_dataloader": false,
+    "num_epochs": 20,
+    "ddp_timeout_s": 6000,
+    "rename_map": {
+        "observation.images.front_camera": "observation.images.top",
+        "observation.images.left_camera": "observation.images.left",
+        "observation.images.right_camera": "observation.images.right"
+    }
+}