File size: 6,959 Bytes
456c9d7 | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 | #!/usr/bin/env python
"""Exercise the RLinf CFG dataset/transform path without launching Ray/FSDP.
The script intentionally limits the local LeRobot view to the rows for which a
smoke advantage sidecar exists. It validates the exact transform stack used by
``FSDPCfgWorker.build_dataloader`` and prints one collated CFG observation.
"""
from __future__ import annotations
import argparse
import json
import logging
from pathlib import Path
import numpy as np
import pandas as pd
import torch
from torch.utils.data import DataLoader
import lerobot.common.datasets.lerobot_dataset as lerobot_dataset
import openpi.training.data_loader as openpi_data_loader
import openpi.transforms as transforms
from rlinf.data.datasets.recap.cfg_model import (
AdvantagePreservingDataset,
TokenizePromptWithGuidance,
)
from rlinf.data.datasets.recap.utils import cast_image_features
from rlinf.data.lerobot_paths import resolve_lerobot_dataset_root
from rlinf.models.embodiment.openpi.dataconfig import get_openpi_config
class _LookupLimitedDataset:
"""Expose only sidecar-covered rows while retaining LeRobot lookahead.
LeRobot's delta-timestamp query for row 0 reads the following nine rows.
Keeping the full decoder as the delegate avoids truncating those rows, but
the small ``hf_dataset`` view makes the advantage join check exactly the
rows present in the smoke sidecar.
"""
def __init__(self, delegate, count: int):
self._delegate = delegate
self.hf_dataset = delegate.hf_dataset.select(range(count))
def __len__(self):
return len(self.hf_dataset)
def __getitem__(self, index):
return self._delegate[index]
def parse_args() -> argparse.Namespace:
parser = argparse.ArgumentParser()
parser.add_argument(
"--dataset",
type=Path,
default=Path(
"/data/myx/RL_post_train/datasets/recap_value_smoke/"
"libero10_task0_train"
),
)
parser.add_argument(
"--model",
type=Path,
default=Path(
"/data/myx/RL_post_train/checkpoints/base/RLinf-Pi05-LIBERO-SFT"
),
)
parser.add_argument("--advantage-tag", default="smoke")
parser.add_argument("--max-samples", type=int, default=8)
return parser.parse_args()
def main() -> int:
args = parse_args()
logging.basicConfig(level=logging.INFO)
for name, path in (("dataset", args.dataset), ("model", args.model)):
if not path.exists():
raise FileNotFoundError(f"{name} path does not exist: {path}")
if args.max_samples < 1:
raise ValueError("--max-samples must be positive")
data_path = str(args.dataset.resolve())
model_path = str(args.model.resolve())
train_config = get_openpi_config(
"pi05_libero", model_path=model_path, batch_size=1, repo_id=data_path
)
data_config = train_config.data.create(train_config.assets_dirs, train_config.model)
dataset_root = resolve_lerobot_dataset_root(data_path)
dataset_meta = lerobot_dataset.LeRobotDatasetMetadata(
data_path, root=dataset_root
)
base_dataset = lerobot_dataset.LeRobotDataset(
data_path,
root=dataset_root,
delta_timestamps={
key: [t / dataset_meta.fps for t in range(train_config.model.action_horizon)]
for key in data_config.action_sequence_keys
},
)
base_dataset.hf_dataset = cast_image_features(base_dataset.hf_dataset)
# The smoke sidecar contains only the first N rows. Restrict the lookup
# view, while retaining the real decoder's full dataset for the 10-step
# delta-timestamp query used by pi05_libero.
n = min(args.max_samples, len(base_dataset.hf_dataset))
base_dataset = _LookupLimitedDataset(base_dataset, n)
if data_config.prompt_from_task:
base_dataset = openpi_data_loader.TransformedDataset(
base_dataset,
[transforms.PromptFromLeRobotTask(dataset_meta.tasks)],
)
tokenizer = None
for transform in data_config.model_transforms.inputs:
if hasattr(transform, "tokenizer"):
tokenizer = transform.tokenizer
break
if tokenizer is None:
raise RuntimeError("Could not find tokenizer in pi05_libero model transforms")
model_transforms = []
for transform in data_config.model_transforms.inputs:
if type(transform).__name__ == "TokenizePrompt":
model_transforms.append(
TokenizePromptWithGuidance(
tokenizer=tokenizer,
discrete_state_input=getattr(
transform, "discrete_state_input", False
),
)
)
else:
model_transforms.append(transform)
transforms_list = [
*data_config.repack_transforms.inputs,
*data_config.data_transforms.inputs,
transforms.Normalize(
data_config.norm_stats, use_quantiles=data_config.use_quantile_norm
),
*model_transforms,
]
transformed_dataset = openpi_data_loader.TransformedDataset(
base_dataset, transforms_list
)
adv_path = args.dataset / "meta" / f"advantages_{args.advantage_tag}.parquet"
adv_df = pd.read_parquet(adv_path)
advantages_lookup = dict(
zip(
zip(
adv_df["episode_index"].astype(int),
adv_df["frame_index"].astype(int),
),
adv_df["advantage"].astype(bool),
)
)
final_dataset = AdvantagePreservingDataset(
base_dataset=base_dataset,
transformed_dataset=transformed_dataset,
advantages_lookup=advantages_lookup,
)
loader = DataLoader(
final_dataset,
batch_size=1,
shuffle=False,
num_workers=0,
drop_last=True,
)
batch = next(iter(loader))
image_shapes = {key: list(value.shape) for key, value in batch["image"].items()}
token_shapes = {
key: list(value.shape)
for key, value in batch.items()
if key.startswith("tokenized_")
}
advantages = torch.as_tensor(batch["advantage"], dtype=torch.bool)
print(
"CFG_DATALOADER_PASS",
json.dumps(
{
"dataset_rows": len(final_dataset),
"image_shapes": image_shapes,
"state_shape": list(batch["state"].shape),
"actions_shape": list(batch["actions"].shape),
"token_shapes": token_shapes,
"advantage_shape": list(advantages.shape),
"advantage_values": advantages.tolist(),
"finite_images": all(
bool(np.isfinite(value.numpy()).all())
for value in batch["image"].values()
),
},
sort_keys=True,
),
)
return 0
if __name__ == "__main__":
raise SystemExit(main())
|