The full dataset viewer is not available (click to read why). Only showing a preview of the rows.
Error code: DatasetGenerationCastError
Exception: DatasetGenerationCastError
Message: An error occurred while generating the dataset
All the data files must have the same columns, but at some point there are 36 new columns ({'0.000000.18', '1.099560', '1.100090', '0.000000.3', '1.499270', '0.249880.1', '0.000000.4', '0.149930', '0.200000', '0.000000.7', '0.000000.15', '1.499270.1', '1.000000', '-0.349830', '0.000000.5', '0.000000.9', '1.280000', '0.000000.12', '0.249880', '0.300200', '0.000000.6', '-1.449290', '-0.200000', '0.000000.17', '0.149930.1', '0.200000.1', '0.000000.13', '0.000000.2', '0.000000.10', '0.000000', '-1.449290.1', '0.000000.11', '0.000000.16', '0.000000.14', '0.000000.8', '0.000000.1'}) and 7 missing columns ({'0.1', '0', '0.4', '0.5', '0.2', '0.3', '1'}).
This happened while the csv dataset builder was generating data using
hf://datasets/KillerBoss/trainrobot-motionclips/ardy_g1/aufstehen.csv (at revision fa737893b3a33c4de01750e56394e6cd92e8c942), ['hf://datasets/KillerBoss/trainrobot-motionclips@fa737893b3a33c4de01750e56394e6cd92e8c942/ardy_g1/aufstehen.cmd.csv', 'hf://datasets/KillerBoss/trainrobot-motionclips@fa737893b3a33c4de01750e56394e6cd92e8c942/ardy_g1/aufstehen.csv', 'hf://datasets/KillerBoss/trainrobot-motionclips@fa737893b3a33c4de01750e56394e6cd92e8c942/ardy_g1/balance.cmd.csv', 'hf://datasets/KillerBoss/trainrobot-motionclips@fa737893b3a33c4de01750e56394e6cd92e8c942/ardy_g1/balance.csv', 'hf://datasets/KillerBoss/trainrobot-motionclips@fa737893b3a33c4de01750e56394e6cd92e8c942/ardy_g1/ducken.cmd.csv', 'hf://datasets/KillerBoss/trainrobot-motionclips@fa737893b3a33c4de01750e56394e6cd92e8c942/ardy_g1/ducken.csv', 'hf://datasets/KillerBoss/trainrobot-motionclips@fa737893b3a33c4de01750e56394e6cd92e8c942/ardy_g1/fusskick.cmd.csv', 'hf://datasets/KillerBoss/trainrobot-motionclips@fa737893b3a33c4de01750e56394e6cd92e8c942/ardy_g1/fusskick.csv', 'hf://datasets/KillerBoss/trainrobot-motionclips@fa737893b3a33c4de01750e56394e6cd92e8c942/ardy_g1/huepfen.cmd.csv', 'hf://datasets/KillerBoss/trainrobot-motionclips@fa737893b3a33c4de01750e56394e6cd92e8c942/ardy_g1/huepfen.csv', 'hf://datasets/KillerBoss/trainrobot-motionclips@fa737893b3a33c4de01750e56394e6cd92e8c942/ardy_g1/idle.cmd.csv', 'hf://datasets/KillerBoss/trainrobot-motionclips@fa737893b3a33c4de01750e56394e6cd92e8c942/ardy_g1/idle.csv', 'hf://datasets/KillerBoss/trainrobot-motionclips@fa737893b3a33c4de01750e56394e6cd92e8c942/ardy_g1/laufen.cmd.csv', 'hf://datasets/KillerBoss/trainrobot-motionclips@fa737893b3a33c4de01750e56394e6cd92e8c942/ardy_g1/laufen.csv', 'hf://datasets/KillerBoss/trainrobot-motionclips@fa737893b3a33c4de01750e56394e6cd92e8c942/ardy_g1/liegen.cmd.csv', 'hf://datasets/KillerBoss/trainrobot-motionclips@fa737893b3a33c4de01750e56394e6cd92e8c942/ardy_g1/liegen.csv', 'hf://datasets/KillerBoss/trainrobot-motionclips@fa737893b3a33c4de01750e56394e6cd92e8c942/ardy_g1/sprung.cmd.csv', 'hf://datasets/KillerBoss/trainrobot-motionclips@fa737893b3a33c4de01750e56394e6cd92e8c942/ardy_g1/sprung.csv', 'hf://datasets/KillerBoss/trainrobot-motionclips@fa737893b3a33c4de01750e56394e6cd92e8c942/ardy_g1/stopp.cmd.csv', 'hf://datasets/KillerBoss/trainrobot-motionclips@fa737893b3a33c4de01750e56394e6cd92e8c942/ardy_g1/stopp.csv', 'hf://datasets/KillerBoss/trainrobot-motionclips@fa737893b3a33c4de01750e56394e6cd92e8c942/ardy_g1/turn_l.cmd.csv', 'hf://datasets/KillerBoss/trainrobot-motionclips@fa737893b3a33c4de01750e56394e6cd92e8c942/ardy_g1/turn_l.csv', 'hf://datasets/KillerBoss/trainrobot-motionclips@fa737893b3a33c4de01750e56394e6cd92e8c942/ardy_g1/turn_r.cmd.csv', 'hf://datasets/KillerBoss/trainrobot-motionclips@fa737893b3a33c4de01750e56394e6cd92e8c942/ardy_g1/turn_r.csv', 'hf://datasets/KillerBoss/trainrobot-motionclips@fa737893b3a33c4de01750e56394e6cd92e8c942/ardy_g1/walk.cmd.csv', 'hf://datasets/KillerBoss/trainrobot-motionclips@fa737893b3a33c4de01750e56394e6cd92e8c942/ardy_g1/walk.csv', 'hf://datasets/KillerBoss/trainrobot-motionclips@fa737893b3a33c4de01750e56394e6cd92e8c942/ardy_g1/walk_back.cmd.csv', 'hf://datasets/KillerBoss/trainrobot-motionclips@fa737893b3a33c4de01750e56394e6cd92e8c942/ardy_g1/walk_back.csv', 'hf://datasets/KillerBoss/trainrobot-motionclips@fa737893b3a33c4de01750e56394e6cd92e8c942/ardy_g1/walk_side.cmd.csv', 'hf://datasets/KillerBoss/trainrobot-motionclips@fa737893b3a33c4de01750e56394e6cd92e8c942/ardy_g1/walk_side.csv', 'hf://datasets/KillerBoss/trainrobot-motionclips@fa737893b3a33c4de01750e56394e6cd92e8c942/ardy_g1/weitsprung.cmd.csv', 'hf://datasets/KillerBoss/trainrobot-motionclips@fa737893b3a33c4de01750e56394e6cd92e8c942/ardy_g1/weitsprung.csv', 'hf://datasets/KillerBoss/trainrobot-motionclips@fa737893b3a33c4de01750e56394e6cd92e8c942/ardy_g1/winken.cmd.csv', 'hf://datasets/KillerBoss/trainrobot-motionclips@fa737893b3a33c4de01750e56394e6cd92e8c942/ardy_g1/winken.csv']
Please either edit the data files to have matching columns, or separate them into different configurations (see docs at https://hf.co/docs/hub/datasets-manual-configuration#multiple-configurations)
Traceback: Traceback (most recent call last):
File "/usr/local/lib/python3.14/site-packages/datasets/builder.py", line 1848, in _prepare_split_single
writer.write_table(table)
~~~~~~~~~~~~~~~~~~^^^^^^^
File "/usr/local/lib/python3.14/site-packages/datasets/arrow_writer.py", line 765, in write_table
self._write_table(pa_table, writer_batch_size=writer_batch_size)
~~~~~~~~~~~~~~~~~^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/usr/local/lib/python3.14/site-packages/datasets/arrow_writer.py", line 773, in _write_table
pa_table = table_cast(pa_table, self._schema)
File "/usr/local/lib/python3.14/site-packages/datasets/table.py", line 2378, in table_cast
return cast_table_to_schema(table, schema)
File "/usr/local/lib/python3.14/site-packages/datasets/table.py", line 2306, in cast_table_to_schema
raise CastError(
...<3 lines>...
)
datasets.table.CastError: Couldn't cast
0.000000: double
0.000000.1: double
0.300200: double
1.000000: double
0.000000.2: double
0.000000.3: double
0.000000.4: double
-1.449290: double
0.149930: double
0.000000.5: double
1.499270: double
0.249880: double
0.000000.6: double
-1.449290.1: double
0.149930.1: double
0.000000.7: double
1.499270.1: double
0.249880.1: double
0.000000.8: double
0.000000.9: double
0.000000.10: double
-0.349830: double
0.200000: double
1.099560: double
0.000000.11: double
1.280000: double
0.000000.12: double
0.000000.13: double
0.000000.14: double
0.200000.1: double
-0.200000: double
0.000000.15: double
1.100090: double
0.000000.16: double
0.000000.17: double
0.000000.18: double
-- schema metadata --
pandas: '{"index_columns": [{"kind": "range", "name": null, "start": 0, "' + 4608
to
{'0': Value('int64'), '0.1': Value('int64'), '0.2': Value('int64'), '0.3': Value('int64'), '0.4': Value('int64'), '1': Value('int64'), '0.5': Value('int64')}
because column names don't match
During handling of the above exception, another exception occurred:
Traceback (most recent call last):
File "/src/services/worker/src/worker/job_runners/config/parquet_and_info.py", line 1369, in compute_config_parquet_and_info_response
parquet_operations, partial, estimated_dataset_info = stream_convert_to_parquet(
~~~~~~~~~~~~~~~~~~~~~~~~~^
builder, max_dataset_size_bytes=max_dataset_size_bytes
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
)
^
File "/src/services/worker/src/worker/job_runners/config/parquet_and_info.py", line 948, in stream_convert_to_parquet
builder._prepare_split(split_generator=splits_generators[split], file_format="parquet")
~~~~~~~~~~~~~~~~~~~~~~^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/usr/local/lib/python3.14/site-packages/datasets/builder.py", line 1694, in _prepare_split
for job_id, done, content in self._prepare_split_single(
~~~~~~~~~~~~~~~~~~~~~~~~~~^
gen_kwargs=gen_kwargs, job_id=job_id, **_prepare_split_args
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
):
^
File "/usr/local/lib/python3.14/site-packages/datasets/builder.py", line 1850, in _prepare_split_single
raise DatasetGenerationCastError.from_cast_error(
...<4 lines>...
)
datasets.exceptions.DatasetGenerationCastError: An error occurred while generating the dataset
All the data files must have the same columns, but at some point there are 36 new columns ({'0.000000.18', '1.099560', '1.100090', '0.000000.3', '1.499270', '0.249880.1', '0.000000.4', '0.149930', '0.200000', '0.000000.7', '0.000000.15', '1.499270.1', '1.000000', '-0.349830', '0.000000.5', '0.000000.9', '1.280000', '0.000000.12', '0.249880', '0.300200', '0.000000.6', '-1.449290', '-0.200000', '0.000000.17', '0.149930.1', '0.200000.1', '0.000000.13', '0.000000.2', '0.000000.10', '0.000000', '-1.449290.1', '0.000000.11', '0.000000.16', '0.000000.14', '0.000000.8', '0.000000.1'}) and 7 missing columns ({'0.1', '0', '0.4', '0.5', '0.2', '0.3', '1'}).
This happened while the csv dataset builder was generating data using
hf://datasets/KillerBoss/trainrobot-motionclips/ardy_g1/aufstehen.csv (at revision fa737893b3a33c4de01750e56394e6cd92e8c942), ['hf://datasets/KillerBoss/trainrobot-motionclips@fa737893b3a33c4de01750e56394e6cd92e8c942/ardy_g1/aufstehen.cmd.csv', 'hf://datasets/KillerBoss/trainrobot-motionclips@fa737893b3a33c4de01750e56394e6cd92e8c942/ardy_g1/aufstehen.csv', 'hf://datasets/KillerBoss/trainrobot-motionclips@fa737893b3a33c4de01750e56394e6cd92e8c942/ardy_g1/balance.cmd.csv', 'hf://datasets/KillerBoss/trainrobot-motionclips@fa737893b3a33c4de01750e56394e6cd92e8c942/ardy_g1/balance.csv', 'hf://datasets/KillerBoss/trainrobot-motionclips@fa737893b3a33c4de01750e56394e6cd92e8c942/ardy_g1/ducken.cmd.csv', 'hf://datasets/KillerBoss/trainrobot-motionclips@fa737893b3a33c4de01750e56394e6cd92e8c942/ardy_g1/ducken.csv', 'hf://datasets/KillerBoss/trainrobot-motionclips@fa737893b3a33c4de01750e56394e6cd92e8c942/ardy_g1/fusskick.cmd.csv', 'hf://datasets/KillerBoss/trainrobot-motionclips@fa737893b3a33c4de01750e56394e6cd92e8c942/ardy_g1/fusskick.csv', 'hf://datasets/KillerBoss/trainrobot-motionclips@fa737893b3a33c4de01750e56394e6cd92e8c942/ardy_g1/huepfen.cmd.csv', 'hf://datasets/KillerBoss/trainrobot-motionclips@fa737893b3a33c4de01750e56394e6cd92e8c942/ardy_g1/huepfen.csv', 'hf://datasets/KillerBoss/trainrobot-motionclips@fa737893b3a33c4de01750e56394e6cd92e8c942/ardy_g1/idle.cmd.csv', 'hf://datasets/KillerBoss/trainrobot-motionclips@fa737893b3a33c4de01750e56394e6cd92e8c942/ardy_g1/idle.csv', 'hf://datasets/KillerBoss/trainrobot-motionclips@fa737893b3a33c4de01750e56394e6cd92e8c942/ardy_g1/laufen.cmd.csv', 'hf://datasets/KillerBoss/trainrobot-motionclips@fa737893b3a33c4de01750e56394e6cd92e8c942/ardy_g1/laufen.csv', 'hf://datasets/KillerBoss/trainrobot-motionclips@fa737893b3a33c4de01750e56394e6cd92e8c942/ardy_g1/liegen.cmd.csv', 'hf://datasets/KillerBoss/trainrobot-motionclips@fa737893b3a33c4de01750e56394e6cd92e8c942/ardy_g1/liegen.csv', 'hf://datasets/KillerBoss/trainrobot-motionclips@fa737893b3a33c4de01750e56394e6cd92e8c942/ardy_g1/sprung.cmd.csv', 'hf://datasets/KillerBoss/trainrobot-motionclips@fa737893b3a33c4de01750e56394e6cd92e8c942/ardy_g1/sprung.csv', 'hf://datasets/KillerBoss/trainrobot-motionclips@fa737893b3a33c4de01750e56394e6cd92e8c942/ardy_g1/stopp.cmd.csv', 'hf://datasets/KillerBoss/trainrobot-motionclips@fa737893b3a33c4de01750e56394e6cd92e8c942/ardy_g1/stopp.csv', 'hf://datasets/KillerBoss/trainrobot-motionclips@fa737893b3a33c4de01750e56394e6cd92e8c942/ardy_g1/turn_l.cmd.csv', 'hf://datasets/KillerBoss/trainrobot-motionclips@fa737893b3a33c4de01750e56394e6cd92e8c942/ardy_g1/turn_l.csv', 'hf://datasets/KillerBoss/trainrobot-motionclips@fa737893b3a33c4de01750e56394e6cd92e8c942/ardy_g1/turn_r.cmd.csv', 'hf://datasets/KillerBoss/trainrobot-motionclips@fa737893b3a33c4de01750e56394e6cd92e8c942/ardy_g1/turn_r.csv', 'hf://datasets/KillerBoss/trainrobot-motionclips@fa737893b3a33c4de01750e56394e6cd92e8c942/ardy_g1/walk.cmd.csv', 'hf://datasets/KillerBoss/trainrobot-motionclips@fa737893b3a33c4de01750e56394e6cd92e8c942/ardy_g1/walk.csv', 'hf://datasets/KillerBoss/trainrobot-motionclips@fa737893b3a33c4de01750e56394e6cd92e8c942/ardy_g1/walk_back.cmd.csv', 'hf://datasets/KillerBoss/trainrobot-motionclips@fa737893b3a33c4de01750e56394e6cd92e8c942/ardy_g1/walk_back.csv', 'hf://datasets/KillerBoss/trainrobot-motionclips@fa737893b3a33c4de01750e56394e6cd92e8c942/ardy_g1/walk_side.cmd.csv', 'hf://datasets/KillerBoss/trainrobot-motionclips@fa737893b3a33c4de01750e56394e6cd92e8c942/ardy_g1/walk_side.csv', 'hf://datasets/KillerBoss/trainrobot-motionclips@fa737893b3a33c4de01750e56394e6cd92e8c942/ardy_g1/weitsprung.cmd.csv', 'hf://datasets/KillerBoss/trainrobot-motionclips@fa737893b3a33c4de01750e56394e6cd92e8c942/ardy_g1/weitsprung.csv', 'hf://datasets/KillerBoss/trainrobot-motionclips@fa737893b3a33c4de01750e56394e6cd92e8c942/ardy_g1/winken.cmd.csv', 'hf://datasets/KillerBoss/trainrobot-motionclips@fa737893b3a33c4de01750e56394e6cd92e8c942/ardy_g1/winken.csv']
Please either edit the data files to have matching columns, or separate them into different configurations (see docs at https://hf.co/docs/hub/datasets-manual-configuration#multiple-configurations)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.
0 int64 | 0.1 int64 | 0.2 int64 | 0.3 int64 | 0.4 int64 | 1 int64 | 0.5 int64 |
|---|---|---|---|---|---|---|
0 | 0 | 0 | 0 | 0 | 1 | 0 |
0 | 0 | 0 | 0 | 0 | 1 | 0 |
0 | 0 | 0 | 0 | 0 | 1 | 0 |
0 | 0 | 0 | 0 | 0 | 1 | 0 |
0 | 0 | 0 | 0 | 0 | 1 | 0 |
0 | 0 | 0 | 0 | 0 | 1 | 0 |
0 | 0 | 0 | 0 | 0 | 1 | 0 |
0 | 0 | 0 | 0 | 0 | 1 | 0 |
0 | 0 | 0 | 0 | 0 | 1 | 0 |
0 | 0 | 0 | 0 | 0 | 1 | 0 |
0 | 0 | 0 | 0 | 0 | 1 | 0 |
0 | 0 | 0 | 0 | 0 | 1 | 0 |
0 | 0 | 0 | 0 | 0 | 1 | 0 |
0 | 0 | 0 | 0 | 0 | 1 | 0 |
0 | 0 | 0 | 0 | 0 | 1 | 0 |
0 | 0 | 0 | 0 | 0 | 1 | 0 |
0 | 0 | 0 | 0 | 0 | 1 | 0 |
0 | 0 | 0 | 0 | 0 | 1 | 0 |
0 | 0 | 0 | 0 | 0 | 1 | 0 |
0 | 0 | 0 | 0 | 0 | 1 | 0 |
0 | 0 | 0 | 0 | 0 | 1 | 0 |
0 | 0 | 0 | 0 | 0 | 1 | 0 |
0 | 0 | 0 | 0 | 0 | 1 | 0 |
0 | 0 | 0 | 0 | 0 | 1 | 0 |
0 | 0 | 0 | 0 | 0 | 1 | 0 |
0 | 0 | 0 | 0 | 0 | 1 | 0 |
0 | 0 | 0 | 0 | 0 | 1 | 0 |
0 | 0 | 0 | 0 | 0 | 1 | 0 |
0 | 0 | 0 | 0 | 0 | 1 | 0 |
0 | 0 | 0 | 0 | 0 | 1 | 0 |
0 | 0 | 0 | 0 | 0 | 1 | 0 |
0 | 0 | 0 | 0 | 0 | 1 | 0 |
0 | 0 | 0 | 0 | 0 | 1 | 0 |
0 | 0 | 0 | 0 | 0 | 1 | 0 |
0 | 0 | 0 | 0 | 0 | 1 | 0 |
0 | 0 | 0 | 0 | 0 | 1 | 0 |
0 | 0 | 0 | 0 | 0 | 1 | 0 |
0 | 0 | 0 | 0 | 0 | 1 | 0 |
0 | 0 | 0 | 0 | 0 | 1 | 0 |
0 | 0 | 0 | 0 | 0 | 1 | 0 |
0 | 0 | 0 | 0 | 0 | 1 | 0 |
0 | 0 | 0 | 0 | 0 | 1 | 0 |
0 | 0 | 0 | 0 | 0 | 1 | 0 |
0 | 0 | 0 | 0 | 0 | 1 | 0 |
0 | 0 | 0 | 0 | 0 | 1 | 0 |
0 | 0 | 0 | 0 | 0 | 1 | 0 |
0 | 0 | 0 | 0 | 0 | 1 | 0 |
0 | 0 | 0 | 0 | 0 | 1 | 0 |
0 | 0 | 0 | 0 | 0 | 1 | 0 |
0 | 0 | 0 | 0 | 0 | 1 | 0 |
0 | 0 | 0 | 0 | 0 | 1 | 0 |
0 | 0 | 0 | 0 | 0 | 1 | 0 |
0 | 0 | 0 | 0 | 0 | 1 | 0 |
0 | 0 | 0 | 0 | 0 | 1 | 0 |
0 | 0 | 0 | 0 | 0 | 1 | 0 |
0 | 0 | 0 | 0 | 0 | 1 | 0 |
0 | 0 | 0 | 0 | 0 | 1 | 0 |
0 | 0 | 0 | 0 | 0 | 1 | 0 |
0 | 0 | 0 | 0 | 0 | 1 | 0 |
0 | 0 | 0 | 0 | 0 | 1 | 0 |
0 | 0 | 0 | 0 | 0 | 1 | 0 |
0 | 0 | 0 | 0 | 0 | 1 | 0 |
0 | 0 | 0 | 0 | 0 | 1 | 0 |
0 | 0 | 0 | 0 | 0 | 1 | 0 |
0 | 0 | 0 | 0 | 0 | 1 | 0 |
0 | 0 | 0 | 0 | 0 | 1 | 0 |
0 | 0 | 0 | 0 | 0 | 1 | 0 |
0 | 0 | 0 | 0 | 0 | 1 | 0 |
0 | 0 | 0 | 0 | 0 | 1 | 0 |
0 | 0 | 0 | 0 | 0 | 1 | 0 |
0 | 0 | 0 | 0 | 0 | 1 | 0 |
0 | 0 | 0 | 0 | 0 | 1 | 0 |
0 | 0 | 0 | 0 | 0 | 1 | 0 |
0 | 0 | 0 | 0 | 0 | 1 | 0 |
0 | 0 | 0 | 0 | 0 | 1 | 0 |
0 | 0 | 0 | 0 | 0 | 1 | 0 |
0 | 0 | 0 | 0 | 0 | 1 | 0 |
0 | 0 | 0 | 0 | 0 | 0 | 0 |
0 | 0 | 0 | 0 | 0 | 0 | 0 |
0 | 0 | 0 | 0 | 0 | 0 | 0 |
0 | 0 | 0 | 0 | 0 | 0 | 0 |
0 | 0 | 0 | 0 | 0 | 0 | 0 |
0 | 0 | 0 | 0 | 0 | 0 | 0 |
0 | 0 | 0 | 0 | 0 | 0 | 0 |
0 | 0 | 0 | 0 | 0 | 0 | 0 |
0 | 0 | 0 | 0 | 0 | 0 | 0 |
0 | 0 | 0 | 0 | 0 | 0 | 0 |
0 | 0 | 0 | 0 | 0 | 0 | 0 |
0 | 0 | 0 | 0 | 0 | 0 | 0 |
0 | 0 | 0 | 0 | 0 | 0 | 0 |
0 | 0 | 0 | 0 | 0 | 0 | 0 |
0 | 0 | 0 | 0 | 0 | 0 | 0 |
null | null | null | null | null | null | null |
null | null | null | null | null | null | null |
null | null | null | null | null | null | null |
null | null | null | null | null | null | null |
null | null | null | null | null | null | null |
null | null | null | null | null | null | null |
null | null | null | null | null | null | null |
null | null | null | null | null | null | null |
YAML Metadata Warning:empty or missing yaml metadata in repo card
Check out the documentation for more information.
Trainrobot Motion-Clips — Basis-Lehrer-Datensatz (v1.0.0)
Prozedural erzeugte Basis-Animationen für RL-Training als LEHRER (Belohnungs-Referenz). Wichtig: Die Animation dient NIE als Policy-Input — sie ist nur Belohnungssystem (Imitation-Reward). So kann der Lehrer später komplett entfernt werden, ohne dass die Policy ihr Verhalten ändert (Beobachtungsraum enthält die Animation nie; Lehrer-Gewicht wird im Training einfach auf 0 gefadet).
Kein CUDA nötig: alles deterministisch prozedural erzeugt (CPU, 30 fps, CC0-1.0).
Formate
| Pfad | Inhalt |
|---|---|
app/motionclips.json |
Alle 41 Clips für G1 + MicroDuck + Skydio X2 (App-nativ, inkl. Kommando-Spur pro Frame) |
ardy_g1/<skill>.csv |
G1-Clips im exakten NVIDIA-ARDY-QPOS-Format (36 Spalten: root xyz + quat wxyz + 29 DoF, z hoch, x vorwärts) — direkt in der Trainrobot-App per „.csv (ARDY)" importierbar |
ardy_g1/<skill>.cmd.csv |
Kommando-Spur je Frame (7 Spalten, gleiche Framezahl) |
kaggle_notebook_ardy.ipynb |
Echte ARDY-Clips (Text→Motion) auf Kaggle-GPU erzeugen und als CSV ergänzen |
Kommando-Spur (= Joystick-/Button-Antwort)
Jeder Frame trägt [vx, vy, wz, bA, bB, bC, bD]:
| Kanal | Bedeutung |
|---|---|
vx / vy |
Soll-Tempo vorwärts / seitwärts (m/s) — linke Stick-Achse |
wz |
Soll-Drehrate (rad/s) — rechte Stick-Achse |
bA |
Button A = Hüpfen/Sprung (0/1) |
bB |
Button B = Hinlegen (0/1) |
bC |
Button C = Aufstehen (0/1) |
bD |
Button D = Stopp (0/1) |
Antwort auf die Frage „Weiß der Datensatz, wo der Joystick zeigt / welcher Button gedrückt ist?": JA — beim Datensatz-Generieren ist die Kommandospur bekannt und wird mitgespeichert. Weil die Clips prozedural aus Kommandos erzeugt werden, ist jeder Frame mit der exakten Stick-/Button-Stellung verknüpft. Später mappen: Der Joystick in der App erzeugt denselben Kommando-Vektor → Match per nächstem Kommando-Vektor bzw. Skill-Auslöser. Für echte Mocap-Dateien ohne Spur: kleines Netz lernt offline die Kommandos aus der Bewegung (Kommando-Regression) — Notebook-Vorlage folgt.
Basis-Animations-Liste (fürs Training nötig)
Gemeinsam (Beiner): idle (Stehen), walk (Gehen vorwärts), walk_back (Rückwärts), walk_side (Seitwärts), turn_l/turn_r (Drehen), laufen (schnell), huepfen (Hüpfen), sprung (Vertikalsprung), weitsprung, liegen (Hinlegen), aufstehen, ducken (Hocke), stopp (Bremsen), balance (Gleichgewicht/Störung)
MicroDuck zusätzlich: sitzen (Hinsetzen), flattern (Kopf/Hals-Spiel)
G1 zusätzlich: winken (Arm winken), fusskick (Fußkick)
Skydio X2 (Drohne, Pfad-Lehrer): starten (Takeoff), schweben (Hover), vorwaerts, rueckwaerts, seitwaerts, kreisen (Kreisflug), steigen, sinken, rolle (Barrel Roll), salto (Flip), landen, notstopp
Nutzung in der App (Auto-Download von HuggingFace)
Die App lädt app/motionclips.json automatisch von HuggingFace (Cache im WebView),
bietet sie als Lehrer-Belohnung an (Slider „Lehrer-Gewicht", Fade-out auf 0 =
Lehrer weg, Verhalten bleibt) und ordnet sie per Skill den Experten/Router zu.
G1-CSVs zusätzlich manuell importierbar per „.csv (ARDY)"-Button.
Erweiterung mit echtem ARDY (optional, GPU)
kaggle_notebook_ardy.ipynbin einem Kaggle-Notebook öffnen (kostenlose GPU)- ARDY-Checkpoints laden, Prompts eingeben → QPOS-CSVs erzeugen
- CSVs in
ardy_g1/ergänzen (gleiche 36-Spalten-Konvention) und neu hochladen
Lizenz: CC0-1.0 (eigene Erzeugung; ARDY-Konvention-Referenz: NVIDIA ARDY, Apache-2.0).
- Downloads last month
- 106