Files
lerobot/src/lerobot/policies/pi0fast/processor_pi0fast.py
2025-09-24 11:11:53 +02:00

93 lines
3.4 KiB
Python

#!/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 typing import Any
import torch
from lerobot.policies.pi0fast.configuration_pi0fast import PI0FASTConfig
from lerobot.processor import (
AddBatchDimensionProcessorStep,
DeviceProcessorStep,
NormalizerProcessorStep,
PolicyAction,
PolicyProcessorPipeline,
RenameObservationsProcessorStep,
UnnormalizerProcessorStep,
)
from lerobot.processor.converters import policy_action_to_transition, transition_to_policy_action
from lerobot.utils.constants import POLICY_POSTPROCESSOR_DEFAULT_NAME, POLICY_PREPROCESSOR_DEFAULT_NAME
def make_pi0fast_pre_post_processors(
config: PI0FASTConfig,
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 PI0Fast 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. 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 PI0Fast 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.
"""
input_steps = [
RenameObservationsProcessorStep(rename_map={}), # To mimic the same processor as pretrained one
AddBatchDimensionProcessorStep(),
DeviceProcessorStep(device=config.device),
NormalizerProcessorStep(
features={**config.input_features, **config.output_features},
norm_map=config.normalization_mapping,
stats=dataset_stats,
),
]
output_steps = [
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,
),
)