Dataset Viewer
The dataset viewer is not available for this subset.
Cannot get the split names for the config 'default' of the dataset.
Exception:    SplitsNotFoundError
Message:      The split names could not be parsed from the dataset config.
Traceback:    Traceback (most recent call last):
                File "/usr/local/lib/python3.14/site-packages/datasets/inspect.py", line 286, in get_dataset_config_info
                  for split_generator in builder._split_generators(
                                         ~~~~~~~~~~~~~~~~~~~~~~~~~^
                      StreamingDownloadManager(base_path=builder.base_path, download_config=download_config)
                      ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
                  )
                  ^
                File "/usr/local/lib/python3.14/site-packages/datasets/packaged_modules/webdataset/webdataset.py", line 81, in _split_generators
                  first_examples = list(islice(pipeline, self.NUM_EXAMPLES_FOR_FEATURES_INFERENCE))
                File "/usr/local/lib/python3.14/site-packages/datasets/packaged_modules/webdataset/webdataset.py", line 32, in _get_pipeline_from_tar
                  fs: fsspec.AbstractFileSystem = fsspec.filesystem("memory")
                                                  ~~~~~~~~~~~~~~~~~^^^^^^^^^^
                File "/usr/local/lib/python3.14/site-packages/fsspec/registry.py", line 302, in filesystem
                  cls = get_filesystem_class(protocol)
                File "/usr/local/lib/python3.14/site-packages/fsspec/registry.py", line 239, in get_filesystem_class
                  raise ValueError(f"Protocol not known: {protocol}")
              ValueError: Protocol not known: memory
              
              The above exception was the direct cause of the following exception:
              
              Traceback (most recent call last):
                File "/src/services/worker/src/worker/job_runners/config/split_names.py", line 66, in compute_split_names_from_streaming_response
                  for split in get_dataset_split_names(
                               ~~~~~~~~~~~~~~~~~~~~~~~^
                      path=dataset,
                      ^^^^^^^^^^^^^
                      config_name=config,
                      ^^^^^^^^^^^^^^^^^^^
                      token=hf_token,
                      ^^^^^^^^^^^^^^^
                  )
                  ^
                File "/usr/local/lib/python3.14/site-packages/datasets/inspect.py", line 340, in get_dataset_split_names
                  info = get_dataset_config_info(
                      path,
                  ...<6 lines>...
                      **config_kwargs,
                  )
                File "/usr/local/lib/python3.14/site-packages/datasets/inspect.py", line 291, in get_dataset_config_info
                  raise SplitsNotFoundError("The split names could not be parsed from the dataset config.") from err
              datasets.inspect.SplitsNotFoundError: The split names could not be parsed from the dataset config.

Need help to make the dataset viewer work? Make sure to review how to configure the dataset viewer, and open a discussion for direct support.

YAML Metadata Warning:empty or missing yaml metadata in repo card

Check out the documentation for more information.

VLN-CE Annotation Archive Tools

Utilities for packaging VLN-CE semantic annotation sidecars into plain tar archives that are easier to upload to Hugging Face and restore on a cluster.

Suggested Hugging Face dataset repo:

shaoyoubo/internnav-vlnce-semantic-bbox-annotations

Archive Layout

The packer writes two artifact families:

semantic_maps/
  vln_ce_semantic_metadata.tar
  r2r/<scene-id>.semantic.tar
  rxr/<scene-id>.semantic.tar
bbox_parquet/
  r2r/<scene-id>.bbox.tar
  rxr/<scene-id>.bbox.tar
manifest.jsonl

Semantic scene archives contain:

traj_data/<dataset>/<scene-id>/videos/
traj_data/<dataset>/<scene-id>/meta/

BBox archives contain:

traj_data/<dataset>/<scene-id>/data/

vln_ce_semantic_metadata.tar contains semantic_metadata/, which is needed when regenerating bbox parquet from semantic maps.

Pack

Pack all R2R and RxR annotations:

cd /root/InternNav

python scripts/vln_ce_semantic/archive_tools/pack_vlnce_annotations.py \
  --annotation-root data/annotations/vln_ce \
  --output-root /workspace/vlnce_annotation_archives \
  --dataset r2r \
  --dataset rxr \
  --artifact all \
  --jobs 8 \
  --checksum sha256

Dry-run one scene:

python scripts/vln_ce_semantic/archive_tools/pack_vlnce_annotations.py \
  --annotation-root data/annotations/vln_ce \
  --output-root /workspace/vlnce_annotation_archives \
  --dataset r2r \
  --scene-id 1pXnuDYAj8r \
  --artifact all \
  --dry-run

Extract

Extract on a cluster into the current annotation layout:

cd /root/InternNav

python scripts/vln_ce_semantic/archive_tools/extract_vlnce_annotations.py \
  /workspace/vlnce_annotation_archives \
  --annotation-root data/annotations/vln_ce \
  --jobs 8

Use --overwrite to replace archive-owned roots such as one scene's videos/, meta/, or data/. Without --overwrite, existing roots cause extraction to fail before tar is run.

Download

Download the uploaded archive folder from Hugging Face:

cd /root/InternNav

/root/miniconda3/envs/internnav-habitat/bin/python \
  scripts/vln_ce_semantic/archive_tools/download_vlnce_annotation_archives.py \
  --archive-root /workspace/vlnce_annotation_archives

Resume is handled by huggingface_hub through the local download cache. Add --verify to check downloaded tar files against the sha256 values in manifest.jsonl.

You can download a smaller subset before extracting:

/root/miniconda3/envs/internnav-habitat/bin/python \
  scripts/vln_ce_semantic/archive_tools/download_vlnce_annotation_archives.py \
  --archive-root /workspace/vlnce_annotation_archives \
  --dataset r2r \
  --artifact bbox

Upload

After packing finishes and manifest.jsonl exists, upload with:

/root/miniconda3/envs/internnav-habitat/bin/python \
  scripts/vln_ce_semantic/archive_tools/upload_vlnce_annotation_archives.py \
  --archive-root /workspace/vlnce_annotation_archives

The script creates the dataset repo by default if it does not exist. Add --private if you want the newly created repo to be private. It uses huggingface_hub's resumable upload_large_folder path by default; rerunning the same command resumes from .cache/.huggingface/ under the archive root.

Downloads last month
5