Dataset Viewer
Duplicate
The dataset viewer is not available for this split.
Cannot load the dataset split (in streaming mode) to extract the first rows.
Error code:   StreamingRowsError
Exception:    CastError
Message:      Couldn't cast
chunks: list<item: struct<chunk_bytes: int64, chunk_size: int64, dim: null, filename: string>>
  child 0, item: struct<chunk_bytes: int64, chunk_size: int64, dim: null, filename: string>
      child 0, chunk_bytes: int64
      child 1, chunk_size: int64
      child 2, dim: null
      child 3, filename: string
config: struct<chunk_bytes: int64, chunk_size: null, compression: null, data_format: list<item: string>, dat (... 54 chars omitted)
  child 0, chunk_bytes: int64
  child 1, chunk_size: null
  child 2, compression: null
  child 3, data_format: list<item: string>
      child 0, item: string
  child 4, data_spec: string
  child 5, encryption: null
  child 6, item_loader: string
updated_at: string
platform: string
config_snapshot: struct<model: struct<backend: string, depthfm_checkpoint: string, vae_id: string, use_lora: bool, fr (... 2433 chars omitted)
  child 0, model: struct<backend: string, depthfm_checkpoint: string, vae_id: string, use_lora: bool, freeze_encoder:  (... 144 chars omitted)
      child 0, backend: string
      child 1, depthfm_checkpoint: string
      child 2, vae_id: string
      child 3, use_lora: bool
      child 4, freeze_encoder: bool
      child 5, gradient_checkpointing: bool
      child 6, torch_compile: bool
      child 7, torch_compile_mode: string
      child 8, full_graph: bool
      child 9, use_checkpoint: bool
      child 10, model_type: string
  child 1, data: struct<source: string, parallel_load: bool, use_wds: bool, hirise: struct<root:
...
_every_steps: int64
      child 16, log_every_steps: int64
      child 17, num_val_samples: int64
      child 18, flow_vis_every_steps: int64
      child 19, train_vis_every_steps: int64
      child 20, test_euler_steps: int64
      child 21, test_choice: string
      child 22, early_stopping_patience: null
      child 23, logger: string
      child 24, project_name: string
      child 25, run_name: string
      child 26, output_dir: string
      child 27, resume_from: null
      child 28, use_ema: bool
      child 29, ema_decay: double
      child 30, ema_device: null
      child 31, num_workers: int64
      child 32, pin_memory: bool
      child 33, prefetch_factor: int64
file_checksums: struct<_SUCCESS: string, chunk-0-0.bin: string, chunk-12-0.bin: string, chunk-16-0.bin: string, chun (... 113 chars omitted)
  child 0, _SUCCESS: string
  child 1, chunk-0-0.bin: string
  child 2, chunk-12-0.bin: string
  child 3, chunk-16-0.bin: string
  child 4, chunk-2-0.bin: string
  child 5, chunk-23-0.bin: string
  child 6, chunk-27-0.bin: string
  child 7, chunk-30-0.bin: string
  child 8, chunk-6-0.bin: string
git_commit: string
created_at: string
cache_hash: string
key_packages: struct<torch: string, lightning: string, litdata: string, numpy: string, rasterio: string, omegaconf (... 9 chars omitted)
  child 0, torch: string
  child 1, lightning: string
  child 2, litdata: string
  child 3, numpy: string
  child 4, rasterio: string
  child 5, omegaconf: string
python_version: string
to
{'cache_hash': Value('string'), 'config_snapshot': {'model': {'backend': Value('string'), 'depthfm_checkpoint': Value('string'), 'vae_id': Value('string'), 'use_lora': Value('bool'), 'freeze_encoder': Value('bool'), 'gradient_checkpointing': Value('bool'), 'torch_compile': Value('bool'), 'torch_compile_mode': Value('string'), 'full_graph': Value('bool'), 'use_checkpoint': Value('bool'), 'model_type': Value('string')}, 'data': {'source': Value('string'), 'parallel_load': Value('bool'), 'use_wds': Value('bool'), 'hirise': {'root': Value('string'), 'include_ortho': Value('bool'), 'ortho_type': Value('string'), 'ortho_scale': Value('null'), 'download': Value('bool'), 'reuse_cache': Value('bool'), 'target': Value('null'), 'bbox': List(Value('int64'))}, 'sampler': {'size': Value('float64'), 'length': Value('null'), 'center_mode': Value('string')}, 'split_fractions': List(Value('float64')), 'split_method': Value('string'), 'split_axis': Value('string'), 'n_folds': Value('null'), 'fold_idx': Value('int64'), 'latent_dir': Value('null'), 'resolution': Value('int64'), 'dtm_normalization': Value('string'), 'stats_path': Value('string'), 'random_flip': Value('bool'), 'brightness_jitter': Value('float64'), 'clip': Value('bool'), 'elev_ref_scale': Value('float64'), 'signal_boost': Value('float64')}, 'training': {'enable': Value('bool'), 'per_gpu_batch_size': Value('int64'), 'gradient_accumulation_steps': Value('int64'), 'num_gpus': Value('int64'), 'max_steps': Value('int64'), 'show_train_vi
...
acian_weight': Value('float64'), 'laplacian_start_step': Value('float64'), 'ordinal_weight': Value('float64'), 'ordinal_start_step': Value('float64'), 'use_confidence_weighting': Value('bool')}, 'val_every_steps': Value('float64'), 'save_every_steps': Value('int64'), 'log_every_steps': Value('int64'), 'num_val_samples': Value('int64'), 'flow_vis_every_steps': Value('int64'), 'train_vis_every_steps': Value('int64'), 'test_euler_steps': Value('int64'), 'test_choice': Value('string'), 'early_stopping_patience': Value('null'), 'logger': Value('string'), 'project_name': Value('string'), 'run_name': Value('string'), 'output_dir': Value('string'), 'resume_from': Value('null'), 'use_ema': Value('bool'), 'ema_decay': Value('float64'), 'ema_device': Value('null'), 'num_workers': Value('int64'), 'pin_memory': Value('bool'), 'prefetch_factor': Value('int64')}}, 'created_at': Value('string'), 'git_commit': Value('string'), 'python_version': Value('string'), 'platform': Value('string'), 'key_packages': {'torch': Value('string'), 'lightning': Value('string'), 'litdata': Value('string'), 'numpy': Value('string'), 'rasterio': Value('string'), 'omegaconf': Value('string')}, 'file_checksums': {'_SUCCESS': Value('string'), 'chunk-0-0.bin': Value('string'), 'chunk-12-0.bin': Value('string'), 'chunk-16-0.bin': Value('string'), 'chunk-2-0.bin': Value('string'), 'chunk-23-0.bin': Value('string'), 'chunk-27-0.bin': Value('string'), 'chunk-30-0.bin': Value('string'), 'chunk-6-0.bin': Value('string')}}
because column names don't match
Traceback:    Traceback (most recent call last):
                File "/src/services/worker/src/worker/utils.py", line 99, in get_rows_or_raise
                  return get_rows(
                         ^^^^^^^^^
                File "/src/libs/libcommon/src/libcommon/utils.py", line 272, in decorator
                  return func(*args, **kwargs)
                         ^^^^^^^^^^^^^^^^^^^^^
                File "/src/services/worker/src/worker/utils.py", line 77, in get_rows
                  rows_plus_one = list(itertools.islice(ds, rows_max_number + 1))
                                  ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
                File "/usr/local/lib/python3.12/site-packages/datasets/iterable_dataset.py", line 2690, in __iter__
                  for key, example in ex_iterable:
                                      ^^^^^^^^^^^
                File "/usr/local/lib/python3.12/site-packages/datasets/iterable_dataset.py", line 2227, in __iter__
                  for key, pa_table in self._iter_arrow():
                                       ^^^^^^^^^^^^^^^^^^
                File "/usr/local/lib/python3.12/site-packages/datasets/iterable_dataset.py", line 2251, in _iter_arrow
                  for key, pa_table in self.ex_iterable._iter_arrow():
                                       ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
                File "/usr/local/lib/python3.12/site-packages/datasets/iterable_dataset.py", line 494, in _iter_arrow
                  for key, pa_table in iterator:
                                       ^^^^^^^^
                File "/usr/local/lib/python3.12/site-packages/datasets/iterable_dataset.py", line 384, in _iter_arrow
                  for key, pa_table in self.generate_tables_fn(**gen_kwags):
                                       ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
                File "/usr/local/lib/python3.12/site-packages/datasets/packaged_modules/json/json.py", line 299, in _generate_tables
                  self._cast_table(pa_table, json_field_paths=json_field_paths),
                  ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
                File "/usr/local/lib/python3.12/site-packages/datasets/packaged_modules/json/json.py", line 128, in _cast_table
                  pa_table = table_cast(pa_table, self.info.features.arrow_schema)
                             ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
                File "/usr/local/lib/python3.12/site-packages/datasets/table.py", line 2321, in table_cast
                  return cast_table_to_schema(table, schema)
                         ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
                File "/usr/local/lib/python3.12/site-packages/datasets/table.py", line 2249, in cast_table_to_schema
                  raise CastError(
              datasets.table.CastError: Couldn't cast
              chunks: list<item: struct<chunk_bytes: int64, chunk_size: int64, dim: null, filename: string>>
                child 0, item: struct<chunk_bytes: int64, chunk_size: int64, dim: null, filename: string>
                    child 0, chunk_bytes: int64
                    child 1, chunk_size: int64
                    child 2, dim: null
                    child 3, filename: string
              config: struct<chunk_bytes: int64, chunk_size: null, compression: null, data_format: list<item: string>, dat (... 54 chars omitted)
                child 0, chunk_bytes: int64
                child 1, chunk_size: null
                child 2, compression: null
                child 3, data_format: list<item: string>
                    child 0, item: string
                child 4, data_spec: string
                child 5, encryption: null
                child 6, item_loader: string
              updated_at: string
              platform: string
              config_snapshot: struct<model: struct<backend: string, depthfm_checkpoint: string, vae_id: string, use_lora: bool, fr (... 2433 chars omitted)
                child 0, model: struct<backend: string, depthfm_checkpoint: string, vae_id: string, use_lora: bool, freeze_encoder:  (... 144 chars omitted)
                    child 0, backend: string
                    child 1, depthfm_checkpoint: string
                    child 2, vae_id: string
                    child 3, use_lora: bool
                    child 4, freeze_encoder: bool
                    child 5, gradient_checkpointing: bool
                    child 6, torch_compile: bool
                    child 7, torch_compile_mode: string
                    child 8, full_graph: bool
                    child 9, use_checkpoint: bool
                    child 10, model_type: string
                child 1, data: struct<source: string, parallel_load: bool, use_wds: bool, hirise: struct<root:
              ...
              _every_steps: int64
                    child 16, log_every_steps: int64
                    child 17, num_val_samples: int64
                    child 18, flow_vis_every_steps: int64
                    child 19, train_vis_every_steps: int64
                    child 20, test_euler_steps: int64
                    child 21, test_choice: string
                    child 22, early_stopping_patience: null
                    child 23, logger: string
                    child 24, project_name: string
                    child 25, run_name: string
                    child 26, output_dir: string
                    child 27, resume_from: null
                    child 28, use_ema: bool
                    child 29, ema_decay: double
                    child 30, ema_device: null
                    child 31, num_workers: int64
                    child 32, pin_memory: bool
                    child 33, prefetch_factor: int64
              file_checksums: struct<_SUCCESS: string, chunk-0-0.bin: string, chunk-12-0.bin: string, chunk-16-0.bin: string, chun (... 113 chars omitted)
                child 0, _SUCCESS: string
                child 1, chunk-0-0.bin: string
                child 2, chunk-12-0.bin: string
                child 3, chunk-16-0.bin: string
                child 4, chunk-2-0.bin: string
                child 5, chunk-23-0.bin: string
                child 6, chunk-27-0.bin: string
                child 7, chunk-30-0.bin: string
                child 8, chunk-6-0.bin: string
              git_commit: string
              created_at: string
              cache_hash: string
              key_packages: struct<torch: string, lightning: string, litdata: string, numpy: string, rasterio: string, omegaconf (... 9 chars omitted)
                child 0, torch: string
                child 1, lightning: string
                child 2, litdata: string
                child 3, numpy: string
                child 4, rasterio: string
                child 5, omegaconf: string
              python_version: string
              to
              {'cache_hash': Value('string'), 'config_snapshot': {'model': {'backend': Value('string'), 'depthfm_checkpoint': Value('string'), 'vae_id': Value('string'), 'use_lora': Value('bool'), 'freeze_encoder': Value('bool'), 'gradient_checkpointing': Value('bool'), 'torch_compile': Value('bool'), 'torch_compile_mode': Value('string'), 'full_graph': Value('bool'), 'use_checkpoint': Value('bool'), 'model_type': Value('string')}, 'data': {'source': Value('string'), 'parallel_load': Value('bool'), 'use_wds': Value('bool'), 'hirise': {'root': Value('string'), 'include_ortho': Value('bool'), 'ortho_type': Value('string'), 'ortho_scale': Value('null'), 'download': Value('bool'), 'reuse_cache': Value('bool'), 'target': Value('null'), 'bbox': List(Value('int64'))}, 'sampler': {'size': Value('float64'), 'length': Value('null'), 'center_mode': Value('string')}, 'split_fractions': List(Value('float64')), 'split_method': Value('string'), 'split_axis': Value('string'), 'n_folds': Value('null'), 'fold_idx': Value('int64'), 'latent_dir': Value('null'), 'resolution': Value('int64'), 'dtm_normalization': Value('string'), 'stats_path': Value('string'), 'random_flip': Value('bool'), 'brightness_jitter': Value('float64'), 'clip': Value('bool'), 'elev_ref_scale': Value('float64'), 'signal_boost': Value('float64')}, 'training': {'enable': Value('bool'), 'per_gpu_batch_size': Value('int64'), 'gradient_accumulation_steps': Value('int64'), 'num_gpus': Value('int64'), 'max_steps': Value('int64'), 'show_train_vi
              ...
              acian_weight': Value('float64'), 'laplacian_start_step': Value('float64'), 'ordinal_weight': Value('float64'), 'ordinal_start_step': Value('float64'), 'use_confidence_weighting': Value('bool')}, 'val_every_steps': Value('float64'), 'save_every_steps': Value('int64'), 'log_every_steps': Value('int64'), 'num_val_samples': Value('int64'), 'flow_vis_every_steps': Value('int64'), 'train_vis_every_steps': Value('int64'), 'test_euler_steps': Value('int64'), 'test_choice': Value('string'), 'early_stopping_patience': Value('null'), 'logger': Value('string'), 'project_name': Value('string'), 'run_name': Value('string'), 'output_dir': Value('string'), 'resume_from': Value('null'), 'use_ema': Value('bool'), 'ema_decay': Value('float64'), 'ema_device': Value('null'), 'num_workers': Value('int64'), 'pin_memory': Value('bool'), 'prefetch_factor': Value('int64')}}, 'created_at': Value('string'), 'git_commit': Value('string'), 'python_version': Value('string'), 'platform': Value('string'), 'key_packages': {'torch': Value('string'), 'lightning': Value('string'), 'litdata': Value('string'), 'numpy': Value('string'), 'rasterio': Value('string'), 'omegaconf': Value('string')}, 'file_checksums': {'_SUCCESS': Value('string'), 'chunk-0-0.bin': Value('string'), 'chunk-12-0.bin': Value('string'), 'chunk-16-0.bin': Value('string'), 'chunk-2-0.bin': Value('string'), 'chunk-23-0.bin': Value('string'), 'chunk-27-0.bin': Value('string'), 'chunk-30-0.bin': Value('string'), 'chunk-6-0.bin': Value('string')}}
              because column names don't match

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.

Mars HiRISE DTM — LitData Streaming Dataset

Pre-processed Mars HiRISE orthoimage + DTM patches in LitData optimized streaming format.

Quick Start

from litdata import StreamingDataset

# Stream directly from HuggingFace — no full download needed
train_ds = StreamingDataset(input_dir="hf://datasets/SuperComputer/mars_hirise_dtm_processed-8e53c64ff6f6a79f/train")
val_ds   = StreamingDataset(input_dir="hf://datasets/SuperComputer/mars_hirise_dtm_processed-8e53c64ff6f6a79f/val")
test_ds  = StreamingDataset(input_dir="hf://datasets/SuperComputer/mars_hirise_dtm_processed-8e53c64ff6f6a79f/test")

sample = train_ds[0]

# Core Tensors
print(sample["image"].shape)        # (3, H, W) float16
print(sample["dtm"].shape)          # (3, H, W) float16
print(sample["confidence"].shape)   # (1, H, W) float16

# Physical / Lighting Parameters
print(sample["sun_vector"].shape)   # (3,)      float32
print(sample["intensity"])          # float32
print(sample["ambient"])            # float32

# Original / Unnormalized Data
print(sample["original_image"].shape) # (C, H, W) float16
print(sample["original_dtm"].shape)   # (1, H, W) float16
print(sample["trend_params"].shape)   # (3,)      float16

# Normalization / Processing Metadata
print(sample["residual_scale"])     # float32
print(sample["raw_residual_p98"])   # float32
print(sample["key"])                # str (e.g., "left_red")

Fields

During extraction, model inputs are quantized to float16 to optimize streaming bandwidth. Physical lighting parameters remain float32.

Key Dtype Shape Description
image float16 (3, H, W) Normalized HiRISE orthoimage [-1, 1]
dtm float16 (3, H, W) Normalized DTM elevation
confidence float16 (1, H, W) Binary valid data mask (eroded/cleaned)
sun_vector float32 (3,) Estimated sun direction (OLS, unit-normalized)
intensity float32 scalar Estimated sun intensity
ambient float32 scalar Estimated ambient light
original_image float16 (C, H, W) Unnormalized resized orthoimage
original_dtm float16 (1, H, W) Unnormalized resized DTM elevation
trend_params float16 (3,) LSQR detrend parameters for the DTM plane
residual_scale float32 scalar Normalization scale applied to the DTM residual
raw_residual_p98 float32 scalar 98th percentile of the raw topographic residual
key string scalar Orthoimage source used (e.g., left_red)

Preprocessing Configuration

Config Hash: 8e53c64ff6f6a79f

This dataset was generated with the following pipeline parameters:

data:
  source: hirise
  parallel_load: false
  use_wds: true
  hirise:
    root: /scratch/mars_hirise_dtm
    include_ortho: true
    ortho_type: RED
    ortho_scale: null
    download: false
    reuse_cache: true
    target: null
    bbox:
    - -120
    - -30
    - 150
    - 30
  sampler:
    size: 0.018
    length: null
    center_mode: optimal
  split_fractions:
  - 0.8
  - 0.1
  - 0.1
  split_method: geographic
  split_axis: longitude
  n_folds: null
  fold_idx: 0
  latent_dir: null
  resolution: 512
  dtm_normalization: relative
  stats_path: dataset_stats/dtm/dataset_stats.json
  random_flip: true
  brightness_jitter: 0.0
  clip: false
  elev_ref_scale: 45.90752235993998
  signal_boost: 1.0
Downloads last month
40