mirror of
https://github.com/huggingface/lerobot.git
synced 2026-05-22 03:59:42 +00:00
Revert accidental changes to dataset files
This commit is contained in:
@@ -234,8 +234,6 @@ def merge_datasets(
|
|||||||
datasets: list[LeRobotDataset],
|
datasets: list[LeRobotDataset],
|
||||||
output_repo_id: str,
|
output_repo_id: str,
|
||||||
output_dir: str | Path | None = None,
|
output_dir: str | Path | None = None,
|
||||||
data_files_size_in_mb: float | None = None,
|
|
||||||
video_files_size_in_mb: float | None = None,
|
|
||||||
) -> LeRobotDataset:
|
) -> LeRobotDataset:
|
||||||
"""Merge multiple LeRobotDatasets into a single dataset.
|
"""Merge multiple LeRobotDatasets into a single dataset.
|
||||||
|
|
||||||
@@ -259,8 +257,6 @@ def merge_datasets(
|
|||||||
aggr_repo_id=output_repo_id,
|
aggr_repo_id=output_repo_id,
|
||||||
roots=roots,
|
roots=roots,
|
||||||
aggr_root=output_dir,
|
aggr_root=output_dir,
|
||||||
data_files_size_in_mb=data_files_size_in_mb,
|
|
||||||
video_files_size_in_mb=video_files_size_in_mb,
|
|
||||||
)
|
)
|
||||||
|
|
||||||
merged_dataset = LeRobotDataset(
|
merged_dataset = LeRobotDataset(
|
||||||
@@ -751,11 +747,11 @@ def _copy_and_reindex_videos(
|
|||||||
f"videos/{video_key}/to_timestamp"
|
f"videos/{video_key}/to_timestamp"
|
||||||
]
|
]
|
||||||
else:
|
else:
|
||||||
# Episodes are already in order by old episode index (from episode_mapping iteration),
|
# Build list of time ranges to keep, in sorted order.
|
||||||
# which equals from_timestamp order since episodes are created sequentially.
|
sorted_keep_episodes = sorted(episodes_in_file, key=lambda x: episode_mapping[x])
|
||||||
episodes_to_keep_ranges: list[tuple[float, float]] = []
|
episodes_to_keep_ranges: list[tuple[float, float]] = []
|
||||||
|
|
||||||
for old_idx in episodes_in_file:
|
for old_idx in sorted_keep_episodes:
|
||||||
src_ep = src_dataset.meta.episodes[old_idx]
|
src_ep = src_dataset.meta.episodes[old_idx]
|
||||||
from_ts = src_ep[f"videos/{video_key}/from_timestamp"]
|
from_ts = src_ep[f"videos/{video_key}/from_timestamp"]
|
||||||
to_ts = src_ep[f"videos/{video_key}/to_timestamp"]
|
to_ts = src_ep[f"videos/{video_key}/to_timestamp"]
|
||||||
@@ -785,7 +781,7 @@ def _copy_and_reindex_videos(
|
|||||||
)
|
)
|
||||||
|
|
||||||
cumulative_ts = 0.0
|
cumulative_ts = 0.0
|
||||||
for old_idx in episodes_in_file:
|
for old_idx in sorted_keep_episodes:
|
||||||
new_idx = episode_mapping[old_idx]
|
new_idx = episode_mapping[old_idx]
|
||||||
src_ep = src_dataset.meta.episodes[old_idx]
|
src_ep = src_dataset.meta.episodes[old_idx]
|
||||||
ep_length = src_ep["length"]
|
ep_length = src_ep["length"]
|
||||||
|
|||||||
@@ -557,7 +557,7 @@ class LeRobotDataset(torch.utils.data.Dataset):
|
|||||||
episodes: list[int] | None = None,
|
episodes: list[int] | None = None,
|
||||||
image_transforms: Callable | None = None,
|
image_transforms: Callable | None = None,
|
||||||
delta_timestamps: dict[str, list[float]] | None = None,
|
delta_timestamps: dict[str, list[float]] | None = None,
|
||||||
tolerance_s: float = 1e-2,
|
tolerance_s: float = 1e-4,
|
||||||
revision: str | None = None,
|
revision: str | None = None,
|
||||||
force_cache_sync: bool = False,
|
force_cache_sync: bool = False,
|
||||||
download_videos: bool = True,
|
download_videos: bool = True,
|
||||||
|
|||||||
Reference in New Issue
Block a user