mirror of
https://github.com/huggingface/lerobot.git
synced 2026-07-29 12:39:41 +00:00
Add OpenPi, Pi0 and Pi0.5 (#1910)
* initial commit * change device in test * do detailed import * adhere to python 3.11 syntax * fix autodocstring * additionally * do same in other files * add model. prefix to all keys in state dict * use dummy stats * add pi05 * also shorten action_steps * fix test * all test pass! and fix tokenizer max length between 05 and 0 * remove test * fix transformer dependency * fix test * split pi0 and pi05 policy in seperate files * fix test * fix push to hub test * add some comments, license and readme * remove warning in config * add pi05 to factory * remove check * rename action_horizon to chunk_size * clean up padding of state and action (more in line with lerobot pi0) * add openpi image transforms for training and add more flexibility to _preprocess_images similar to lerobot pi0 * fix key match from pytorch state dict (similar keys to openpi implementation now) * also for pi05 * update to python 3.11 * revert to openpi transformer replace python 3.11 * fix(modeling pi0): nit warning message * use safeauto_docstring * fix: remove unused param * fix from pretrained * add preprocess tests * also compile forward method * Do not add model prefix to normalization * use same name for action and state dim as lerobot pi0 and remove fixed image keys * load from pretrained_path * temp: hardcode base model * fix override self.pretrained_path = None overwrite * rename to loss * remove additional image augmentations, lerobot dataset already does this * Add docs * put tests in test folder * Add test to instatiate all base models * go back to python 3.10 * update docs * adapt docs pi05 * change docs: finetune base model options * minor docs fixes and dependencies * remove todo * cast float64 to float32 for mps * skip if no transformers * fix tests * add new models to modelcard * add back init * fix circular input * feat: only run pi test on GPU * remove require_nightly_gpu * replace decorator test_pi0_openpi * rename action_dim, state_dim to max_action_dim, max_state_dim * fix doc and constants * cleanup tests * fix from pretrained * fix tests * add comment pi0 pi05 tests, add image features to pi0 pi05 hub tests * fix, state is included in language not in flow head * Move test to specific folder * and paligemma task with newline * remove add_special_tokens, not needed * feedback pr * Remove previous pi0 and rename pi0_openpi and pi05_openpi * Add Quantile stats to LeRobotDataset (#1985) * - Add RunningQuantileStats class for efficient histogram-based quantile computation - Integrate quantile parameters (compute_quantiles, quantiles) into LeRobotDataset - Support quantile computation during episode collection and aggregation - Add comprehensive function-based test suite (24 tests) for quantile functionality - Maintain full backward compatibility with existing stats computation - Enable configurable quantiles (default: [0.01, 0.99]) for robust normalization * style fixes, make quantiles computation by default to new datasets * fix tests * - Added DEFAULT_QUANTILES=[0.01, 0.10, 0.50, 0.90, 0.99] to be computed for each features instead of being chosen by the user - Fortified tests. * - add helper functions to reshape stats - add missing test for quantiles * - Add QUANTILE normalization mode to normalize the data with the 1st and 99th percentiles. - Add QUANTILE10 normalization mode to normalize the data with the 10th and 90th percentiles. * style fixes * Added missing lisence * Simplify compute_stats * - added script `augment_dataset_quantile_stats.py` so that we can add quantile stats to existing v3 datasets that dont have quatniles - modified quantile computation instead of using the edge for the value, interpolate the values in the bin * rename pi0/pi05 files * Remove open pi patch and use custom transformer branch for now * renaming * fix * Revert "fix" This reverts commit1ea65730ac. * fix naming * feet(pi0/pi0.5): add pipeline (#2009) * feat(processor): convert openpi model with processor * TODO: Make test works * fix(modeling_pi0openpi): update attention mask value and time scaling; improve task handling in tests - Changed the attention mask value from `self.config.attention_mask_value` to a fixed value of `-2.3819763e38`. - Updated time scaling in the `sample_noise` method to use a constant factor of `0.999` and an offset of `0.001`. - Enhanced task handling in tests to ensure proper formatting and batch size consistency. - Cleaned up commented-out test code for clarity. * refactor(pi0): rename PI0OpenPIConfig and PI0OpenPIPolicy to PI0Config and PI0Policy - Updated imports and references throughout the codebase to reflect the new naming convention. - Introduced a new processor file for PI0 to handle pre-processing and post-processing steps. - Adjusted tests to utilize the renamed classes, ensuring consistency and functionality. - Enhanced clarity and maintainability by removing outdated naming conventions. * refactor(pi05): rename PI0OpenPIPolicy to PI0Policy and update configuration - Renamed `PI0OpenPIPolicy` to `PI0Policy` for consistency with naming conventions. - Updated the `PI05OpenPIConfig` to include a new `tokenizer_max_length` attribute and changed the normalization mode for state from `MEAN_STD` to `QUANTILES`. - Simplified model initialization in `PI05OpenPIPolicy` by removing unused `dataset_stats` parameter. - Added a new processor class for `Pi05PrepareStateTokenizerProcessorStep` with `@dataclass` for improved readability. - Introduced a test script to compare the integration of the PI0OpenPI policy with the original implementation, ensuring local testing compatibility. * feat(processor): convert openpi model with processor * TODO: Make test works * fix(modeling_pi0openpi): update attention mask value and time scaling; improve task handling in tests - Changed the attention mask value from `self.config.attention_mask_value` to a fixed value of `-2.3819763e38`. - Updated time scaling in the `sample_noise` method to use a constant factor of `0.999` and an offset of `0.001`. - Enhanced task handling in tests to ensure proper formatting and batch size consistency. - Cleaned up commented-out test code for clarity. * refactor(pi0): rename PI0OpenPIConfig and PI0OpenPIPolicy to PI0Config and PI0Policy - Updated imports and references throughout the codebase to reflect the new naming convention. - Introduced a new processor file for PI0 to handle pre-processing and post-processing steps. - Adjusted tests to utilize the renamed classes, ensuring consistency and functionality. - Enhanced clarity and maintainability by removing outdated naming conventions. * refactor(pi05): rename PI0OpenPIPolicy to PI0Policy and update configuration - Renamed `PI0OpenPIPolicy` to `PI0Policy` for consistency with naming conventions. - Updated the `PI05OpenPIConfig` to include a new `tokenizer_max_length` attribute and changed the normalization mode for state from `MEAN_STD` to `QUANTILES`. - Simplified model initialization in `PI05OpenPIPolicy` by removing unused `dataset_stats` parameter. - Added a new processor class for `Pi05PrepareStateTokenizerProcessorStep` with `@dataclass` for improved readability. - Introduced a test script to compare the integration of the PI0OpenPI policy with the original implementation, ensuring local testing compatibility. * refactor(pi05): update imports and rename configuration classes - Changed imports to reflect the new naming convention for PI05 configuration and policy classes. - Renamed `PI05OpenPIConfig` to `PI05Config` and `PI05OpenPIPolicy` to `PI05Policy` for consistency. - Introduced a new processor file for PI05, implementing pre-processing and post-processing steps. - Updated tests to utilize the renamed classes, ensuring functionality and consistency across the codebase. * update(pi05): increase tokenizer_max_length for improved processing - Changed the `tokenizer_max_length` from 48 to 200 to enhance the model's capability in handling longer sequences. - This adjustment aims to improve the overall performance and flexibility of the PI05 configuration. * add default for state (max_state_dim) * correct naming * fix import * cleanup code * remove unused test * us quantiles for action * move to device * remove discrete state assert * fix pi05 test * move pi05 to device * use base models in comparison tests * small renames for tests * change number of tokens pi05 test * fix openpi tokenization in test * fix hub test * fix test * assert lerobot vs openpi tests --------- Co-authored-by: Pepijn <pepijn@huggingface.co> * add headers * add back previously removed imports * update if statement load processor with dataset stats * remove to avoid circular import * inject dataset stats for pretrained models * check normalization before applying * add link to quantile augument script * fix(policies): transformers import for ci in PI0 & PI05 (#2039) * fix(policies): transformers import for ci in PI0 * fix(policies): transformers import for ci in PI05 * test(processor): fix expected raise when normalization types are missing (#2040) * switch normalization order pipeline for pi05 * Fix/quantiles script (#2064) * refactor augment stats with quantiles script add parallelization for faster processing shift the quantile normalization between -1 1 * fix replay buffer tests * fix comment * overwrite the pipeline normalization features with the policy features * remove double normalization overwrite * cleanup from pretrained * remove typo * also set norm_map * fix(augment_quantiles) images incorrectly divided by 255 * clamp quantiles * link to lerobot base models * rename tests * encorperate PR feedback * update docstring for RunningQuantileStats * update doc links * Revert "clamp quantiles" This reverts commit172207471c. * fix self.paligemma * fix tests related to quantiles that were scaled to [0,1], the new range is [-1, 1] * fix libero doc and use different transformer branch * use fix branch instead of feat * update results libero * add new line * fix formatting * precommit * update results libero * update libero doc * update title * final changes * add quantiles to test * run pre commit --------- Signed-off-by: Steven Palma <imstevenpmwork@ieee.org> Co-authored-by: Michel Aractingi <michel.aractingi@huggingface.co> Co-authored-by: Adil Zouitine <adilzouitinegm@gmail.com> Co-authored-by: Steven Palma <imstevenpmwork@ieee.org> Co-authored-by: Steven Palma <steven.palma@huggingface.co>
This commit is contained in:
Pepijn
@@ -0,0 +1,49 @@
|
||||
# π₀.₅ (pi05)
|
||||
|
||||
This repository contains the Hugging Face port of **π₀.₅**, adapted from [OpenPI](https://github.com/Physical-Intelligence/openpi) by the Physical Intelligence.
|
||||
It is designed as a **Vision-Language-Action model with open-world generalization**.
|
||||
|
||||
---
|
||||
|
||||
## Model Overview
|
||||
|
||||
| Feature | π₀ | π₀.₅ |
|
||||
| -------------------- | ------------------------------------------------------ | ----------------------------------------- |
|
||||
| Time Conditioning | Concatenates time with actions via `action_time_mlp_*` | Uses `time_mlp_*` for AdaRMS conditioning |
|
||||
| AdaRMS | Not used | Used in action expert |
|
||||
| Tokenizer Length | 48 tokens | 200 tokens |
|
||||
| Discrete State Input | False (Uses `state_proj` layer) | True |
|
||||
| Parameter Count | Higher (includes state embedding) | Lower (no state embedding) |
|
||||
|
||||
---
|
||||
|
||||
## Citation
|
||||
|
||||
If you use this work, please cite both **OpenPI** and the π₀.₅ paper:
|
||||
|
||||
```bibtex
|
||||
@misc{openpi2024,
|
||||
author = {Physical Intelligence Lab},
|
||||
title = {OpenPI: PyTorch Implementation of π0 and π0.5 Policies},
|
||||
year = {2024},
|
||||
publisher = {GitHub},
|
||||
howpublished = {\url{https://github.com/Physical-Intelligence/openpi}},
|
||||
license = {Apache-2.0}
|
||||
}
|
||||
|
||||
@misc{intelligence2025pi05visionlanguageactionmodelopenworld,
|
||||
title = {π₀.₅: a Vision-Language-Action Model with Open-World Generalization},
|
||||
author = {Physical Intelligence and Kevin Black and Noah Brown and James Darpinian and Karan Dhabalia and Danny Driess and Adnan Esmail and Michael Equi and Chelsea Finn and Niccolo Fusai and Manuel Y. Galliker and Dibya Ghosh and Lachy Groom and Karol Hausman and Brian Ichter and Szymon Jakubczak and Tim Jones and Liyiming Ke and Devin LeBlanc and Sergey Levine and Adrian Li-Bell and Mohith Mothukuri and Suraj Nair and Karl Pertsch and Allen Z. Ren and Lucy Xiaoyang Shi and Laura Smith and Jost Tobias Springenberg and Kyle Stachowicz and James Tanner and Quan Vuong and Homer Walke and Anna Walling and Haohuan Wang and Lili Yu and Ury Zhilinsky},
|
||||
year = {2025},
|
||||
eprint = {2504.16054},
|
||||
archivePrefix= {arXiv},
|
||||
primaryClass = {cs.LG},
|
||||
url = {https://arxiv.org/abs/2504.16054},
|
||||
}
|
||||
```
|
||||
|
||||
---
|
||||
|
||||
## License
|
||||
|
||||
This port follows the **Apache 2.0 License**, consistent with the original [OpenPI repository](https://github.com/Physical-Intelligence/openpi).
|
||||
@@ -0,0 +1,21 @@
|
||||
#!/usr/bin/env python
|
||||
|
||||
# Copyright 2025 Physical Intelligence and The HuggingFace Inc. team. All rights reserved.
|
||||
#
|
||||
# Licensed under the Apache License, Version 2.0 (the "License");
|
||||
# you may not use this file except in compliance with the License.
|
||||
# You may obtain a copy of the License at
|
||||
#
|
||||
# http://www.apache.org/licenses/LICENSE-2.0
|
||||
#
|
||||
# Unless required by applicable law or agreed to in writing, software
|
||||
# distributed under the License is distributed on an "AS IS" BASIS,
|
||||
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
|
||||
# See the License for the specific language governing permissions and
|
||||
# limitations under the License.
|
||||
|
||||
from .configuration_pi05 import PI05Config
|
||||
from .modeling_pi05 import PI05Policy
|
||||
from .processor_pi05 import make_pi05_pre_post_processors
|
||||
|
||||
__all__ = ["PI05Config", "PI05Policy", "make_pi05_pre_post_processors"]
|
||||
@@ -0,0 +1,153 @@
|
||||
#!/usr/bin/env python
|
||||
|
||||
# Copyright 2025 Physical Intelligence and The HuggingFace Inc. team. All rights reserved.
|
||||
#
|
||||
# Licensed under the Apache License, Version 2.0 (the "License");
|
||||
# you may not use this file except in compliance with the License.
|
||||
# You may obtain a copy of the License at
|
||||
#
|
||||
# http://www.apache.org/licenses/LICENSE-2.0
|
||||
#
|
||||
# Unless required by applicable law or agreed to in writing, software
|
||||
# distributed under the License is distributed on an "AS IS" BASIS,
|
||||
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
|
||||
# See the License for the specific language governing permissions and
|
||||
# limitations under the License.
|
||||
|
||||
from dataclasses import dataclass, field
|
||||
|
||||
from lerobot.configs.policies import PreTrainedConfig
|
||||
from lerobot.configs.types import FeatureType, NormalizationMode, PolicyFeature
|
||||
from lerobot.optim.optimizers import AdamWConfig
|
||||
from lerobot.optim.schedulers import CosineDecayWithWarmupSchedulerConfig
|
||||
|
||||
|
||||
@PreTrainedConfig.register_subclass("pi05")
|
||||
@dataclass
|
||||
class PI05Config(PreTrainedConfig):
|
||||
paligemma_variant: str = "gemma_2b"
|
||||
action_expert_variant: str = "gemma_300m"
|
||||
dtype: str = "float32" # Options: "bfloat16", "float32"
|
||||
|
||||
n_obs_steps: int = 1
|
||||
chunk_size: int = 50 # Number of action steps to predict, in openpi called "action_horizon"
|
||||
n_action_steps: int = 50 # Number of action steps to execute
|
||||
|
||||
# Shorter state and action vectors will be padded to these dimensions
|
||||
max_state_dim: int = 32
|
||||
max_action_dim: int = 32
|
||||
|
||||
# Flow matching parameters: see openpi `PI0Pytorch`
|
||||
num_inference_steps: int = 10
|
||||
time_sampling_beta_alpha: float = 1.5
|
||||
time_sampling_beta_beta: float = 1.0
|
||||
time_sampling_scale: float = 0.999
|
||||
time_sampling_offset: float = 0.001
|
||||
min_period: float = 4e-3
|
||||
max_period: float = 4.0
|
||||
|
||||
image_resolution: tuple[int, int] = (224, 224) # see openpi `preprocessing_pytorch.py`
|
||||
|
||||
# Add empty images. Used to add empty cameras when no image features are present.
|
||||
empty_cameras: int = 0
|
||||
|
||||
tokenizer_max_length: int = 200 # see openpi `__post_init__`
|
||||
|
||||
normalization_mapping: dict[str, NormalizationMode] = field(
|
||||
default_factory=lambda: {
|
||||
"VISUAL": NormalizationMode.IDENTITY,
|
||||
"STATE": NormalizationMode.QUANTILES, # Pi0.5 uses quantiles for state
|
||||
"ACTION": NormalizationMode.QUANTILES, # Pi0.5 uses quantiles for action
|
||||
}
|
||||
)
|
||||
|
||||
# Training settings
|
||||
gradient_checkpointing: bool = False # Enable gradient checkpointing for memory optimization
|
||||
compile_model: bool = False # Whether to use torch.compile for model optimization
|
||||
compile_mode: str = "max-autotune" # Torch compile mode
|
||||
device: str | None = None # Device to use for the model (None = auto-detect)
|
||||
|
||||
# Optimizer settings: see openpi `AdamW`
|
||||
optimizer_lr: float = 2.5e-5 # see openpi `CosineDecaySchedule: peak_lr`
|
||||
optimizer_betas: tuple[float, float] = (0.9, 0.95)
|
||||
optimizer_eps: float = 1e-8
|
||||
optimizer_weight_decay: float = 0.01
|
||||
optimizer_grad_clip_norm: float = 1.0
|
||||
|
||||
# Scheduler settings: see openpi `CosineDecaySchedule`
|
||||
scheduler_warmup_steps: int = 1_000
|
||||
scheduler_decay_steps: int = 30_000
|
||||
scheduler_decay_lr: float = 2.5e-6
|
||||
|
||||
tokenizer_max_length: int = 200 # see openpi `__post_init__`
|
||||
|
||||
def __post_init__(self):
|
||||
super().__post_init__()
|
||||
|
||||
# Validate configuration
|
||||
if self.n_action_steps > self.chunk_size:
|
||||
raise ValueError(
|
||||
f"n_action_steps ({self.n_action_steps}) cannot be greater than chunk_size ({self.chunk_size})"
|
||||
)
|
||||
|
||||
if self.paligemma_variant not in ["gemma_300m", "gemma_2b"]:
|
||||
raise ValueError(f"Invalid paligemma_variant: {self.paligemma_variant}")
|
||||
|
||||
if self.action_expert_variant not in ["gemma_300m", "gemma_2b"]:
|
||||
raise ValueError(f"Invalid action_expert_variant: {self.action_expert_variant}")
|
||||
|
||||
if self.dtype not in ["bfloat16", "float32"]:
|
||||
raise ValueError(f"Invalid dtype: {self.dtype}")
|
||||
|
||||
def validate_features(self) -> None:
|
||||
"""Validate and set up input/output features."""
|
||||
for i in range(self.empty_cameras):
|
||||
key = f"observation.images.empty_camera_{i}"
|
||||
empty_camera = PolicyFeature(
|
||||
type=FeatureType.VISUAL,
|
||||
shape=(3, *self.image_resolution), # Use configured image resolution
|
||||
)
|
||||
self.input_features[key] = empty_camera
|
||||
|
||||
if "observation.state" not in self.input_features:
|
||||
state_feature = PolicyFeature(
|
||||
type=FeatureType.STATE,
|
||||
shape=(self.max_state_dim,), # Padded to max_state_dim
|
||||
)
|
||||
self.input_features["observation.state"] = state_feature
|
||||
|
||||
if "action" not in self.output_features:
|
||||
action_feature = PolicyFeature(
|
||||
type=FeatureType.ACTION,
|
||||
shape=(self.max_action_dim,), # Padded to max_action_dim
|
||||
)
|
||||
self.output_features["action"] = action_feature
|
||||
|
||||
def get_optimizer_preset(self) -> AdamWConfig:
|
||||
return AdamWConfig(
|
||||
lr=self.optimizer_lr,
|
||||
betas=self.optimizer_betas,
|
||||
eps=self.optimizer_eps,
|
||||
weight_decay=self.optimizer_weight_decay,
|
||||
grad_clip_norm=self.optimizer_grad_clip_norm,
|
||||
)
|
||||
|
||||
def get_scheduler_preset(self):
|
||||
return CosineDecayWithWarmupSchedulerConfig(
|
||||
peak_lr=self.optimizer_lr,
|
||||
decay_lr=self.scheduler_decay_lr,
|
||||
num_warmup_steps=self.scheduler_warmup_steps,
|
||||
num_decay_steps=self.scheduler_decay_steps,
|
||||
)
|
||||
|
||||
@property
|
||||
def observation_delta_indices(self) -> None:
|
||||
return None
|
||||
|
||||
@property
|
||||
def action_delta_indices(self) -> list:
|
||||
return list(range(self.chunk_size))
|
||||
|
||||
@property
|
||||
def reward_delta_indices(self) -> None:
|
||||
return None
|
||||
File diff suppressed because it is too large
Load Diff
@@ -0,0 +1,171 @@
|
||||
#!/usr/bin/env python
|
||||
|
||||
# Copyright 2025 Physical Intelligence and The HuggingFace Inc. team. All rights reserved.
|
||||
#
|
||||
# Licensed under the Apache License, Version 2.0 (the "License");
|
||||
# you may not use this file except in compliance with the License.
|
||||
# You may obtain a copy of the License at
|
||||
#
|
||||
# http://www.apache.org/licenses/LICENSE-2.0
|
||||
#
|
||||
# Unless required by applicable law or agreed to in writing, software
|
||||
# distributed under the License is distributed on an "AS IS" BASIS,
|
||||
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
|
||||
# See the License for the specific language governing permissions and
|
||||
# limitations under the License.
|
||||
|
||||
from copy import deepcopy
|
||||
from dataclasses import dataclass
|
||||
from typing import Any
|
||||
|
||||
import numpy as np
|
||||
import torch
|
||||
|
||||
from lerobot.configs.types import PipelineFeatureType, PolicyFeature
|
||||
from lerobot.policies.pi05.configuration_pi05 import PI05Config
|
||||
from lerobot.policies.pi05.modeling_pi05 import pad_vector
|
||||
from lerobot.processor import (
|
||||
AddBatchDimensionProcessorStep,
|
||||
DeviceProcessorStep,
|
||||
NormalizerProcessorStep,
|
||||
PolicyAction,
|
||||
PolicyProcessorPipeline,
|
||||
ProcessorStep,
|
||||
ProcessorStepRegistry,
|
||||
RenameObservationsProcessorStep,
|
||||
TokenizerProcessorStep,
|
||||
UnnormalizerProcessorStep,
|
||||
)
|
||||
from lerobot.processor.converters import policy_action_to_transition, transition_to_policy_action
|
||||
from lerobot.processor.core import EnvTransition, TransitionKey
|
||||
from lerobot.utils.constants import (
|
||||
OBS_STATE,
|
||||
POLICY_POSTPROCESSOR_DEFAULT_NAME,
|
||||
POLICY_PREPROCESSOR_DEFAULT_NAME,
|
||||
)
|
||||
|
||||
|
||||
@ProcessorStepRegistry.register(name="pi05_prepare_state_tokenizer_processor_step")
|
||||
@dataclass
|
||||
class Pi05PrepareStateTokenizerProcessorStep(ProcessorStep):
|
||||
"""
|
||||
Processor step to prepare the state and tokenize the language input.
|
||||
"""
|
||||
|
||||
max_state_dim: int = 32
|
||||
task_key: str = "task"
|
||||
|
||||
def __call__(self, transition: EnvTransition) -> EnvTransition:
|
||||
transition = transition.copy()
|
||||
|
||||
state = transition.get(TransitionKey.OBSERVATION, {}).get(OBS_STATE)
|
||||
if state is None:
|
||||
raise ValueError("State is required for PI05")
|
||||
tasks = transition.get(TransitionKey.COMPLEMENTARY_DATA, {}).get(self.task_key)
|
||||
if tasks is None:
|
||||
raise ValueError("No task found in complementary data")
|
||||
|
||||
# TODO: check if this necessary
|
||||
state = deepcopy(state)
|
||||
|
||||
# Prepare state (pad to max_state_dim)
|
||||
state = pad_vector(state, self.max_state_dim)
|
||||
|
||||
# State should already be normalized to [-1, 1] by the NormalizerProcessorStep that runs before this step
|
||||
# Discretize into 256 bins (see openpi `PaligemmaTokenizer.tokenize()`)
|
||||
state_np = state.cpu().numpy()
|
||||
discretized_states = np.digitize(state_np, bins=np.linspace(-1, 1, 256 + 1)[:-1]) - 1
|
||||
|
||||
full_prompts = []
|
||||
for i, task in enumerate(tasks):
|
||||
cleaned_text = task.strip().replace("_", " ").replace("\n", " ")
|
||||
state_str = " ".join(map(str, discretized_states[i]))
|
||||
full_prompt = f"Task: {cleaned_text}, State: {state_str};\nAction: "
|
||||
full_prompts.append(full_prompt)
|
||||
|
||||
transition[TransitionKey.COMPLEMENTARY_DATA][self.task_key] = full_prompts
|
||||
# Normalize state to [-1, 1] range if needed (assuming it's already normalized by normalizer processor step!!)
|
||||
# Discretize into 256 bins (see openpi `PaligemmaTokenizer.tokenize()`)
|
||||
return transition
|
||||
|
||||
def transform_features(
|
||||
self, features: dict[PipelineFeatureType, dict[str, PolicyFeature]]
|
||||
) -> dict[PipelineFeatureType, dict[str, PolicyFeature]]:
|
||||
"""
|
||||
This step does not alter the feature definitions.
|
||||
"""
|
||||
return features
|
||||
|
||||
|
||||
def make_pi05_pre_post_processors(
|
||||
config: PI05Config,
|
||||
dataset_stats: dict[str, dict[str, torch.Tensor]] | None = None,
|
||||
) -> tuple[
|
||||
PolicyProcessorPipeline[dict[str, Any], dict[str, Any]],
|
||||
PolicyProcessorPipeline[PolicyAction, PolicyAction],
|
||||
]:
|
||||
"""
|
||||
Constructs pre-processor and post-processor pipelines for the PI0 policy.
|
||||
|
||||
The pre-processing pipeline prepares input data for the model by:
|
||||
1. Renaming features to match pretrained configurations.
|
||||
2. Normalizing input and output features based on dataset statistics.
|
||||
3. Adding a batch dimension.
|
||||
4. Appending a newline character to the task description for tokenizer compatibility.
|
||||
5. Tokenizing the text prompt using the PaliGemma tokenizer.
|
||||
6. Moving all data to the specified device.
|
||||
|
||||
The post-processing pipeline handles the model's output by:
|
||||
1. Moving data to the CPU.
|
||||
2. Unnormalizing the output features to their original scale.
|
||||
|
||||
Args:
|
||||
config: The configuration object for the PI0 policy.
|
||||
dataset_stats: A dictionary of statistics for normalization.
|
||||
preprocessor_kwargs: Additional arguments for the pre-processor pipeline.
|
||||
postprocessor_kwargs: Additional arguments for the post-processor pipeline.
|
||||
|
||||
Returns:
|
||||
A tuple containing the configured pre-processor and post-processor pipelines.
|
||||
"""
|
||||
|
||||
# Add remaining processors
|
||||
input_steps: list[ProcessorStep] = [
|
||||
RenameObservationsProcessorStep(rename_map={}), # To mimic the same processor as pretrained one
|
||||
AddBatchDimensionProcessorStep(),
|
||||
# NOTE: NormalizerProcessorStep MUST come before Pi05PrepareStateTokenizerProcessorStep
|
||||
# because the tokenizer step expects normalized state in [-1, 1] range for discretization
|
||||
NormalizerProcessorStep(
|
||||
features={**config.input_features, **config.output_features},
|
||||
norm_map=config.normalization_mapping,
|
||||
stats=dataset_stats,
|
||||
),
|
||||
Pi05PrepareStateTokenizerProcessorStep(max_state_dim=config.max_state_dim),
|
||||
TokenizerProcessorStep(
|
||||
tokenizer_name="google/paligemma-3b-pt-224",
|
||||
max_length=config.tokenizer_max_length,
|
||||
padding_side="right",
|
||||
padding="max_length",
|
||||
),
|
||||
DeviceProcessorStep(device=config.device),
|
||||
]
|
||||
|
||||
output_steps: list[ProcessorStep] = [
|
||||
UnnormalizerProcessorStep(
|
||||
features=config.output_features, norm_map=config.normalization_mapping, stats=dataset_stats
|
||||
),
|
||||
DeviceProcessorStep(device="cpu"),
|
||||
]
|
||||
|
||||
return (
|
||||
PolicyProcessorPipeline[dict[str, Any], dict[str, Any]](
|
||||
steps=input_steps,
|
||||
name=POLICY_PREPROCESSOR_DEFAULT_NAME,
|
||||
),
|
||||
PolicyProcessorPipeline[PolicyAction, PolicyAction](
|
||||
steps=output_steps,
|
||||
name=POLICY_POSTPROCESSOR_DEFAULT_NAME,
|
||||
to_transition=policy_action_to_transition,
|
||||
to_output=transition_to_policy_action,
|
||||
),
|
||||
)
|
||||
Reference in New Issue
Block a user