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/packaged_modules/json/json.py", line 291, in _generate_tables
pa_table = paj.read_json(
io.BytesIO(batch), read_options=paj.ReadOptions(block_size=block_size)
)
File "pyarrow/_json.pyx", line 364, in pyarrow._json.read_json
File "pyarrow/error.pxi", line 155, in pyarrow.lib.pyarrow_internal_check_status
return check_status(status)
File "pyarrow/error.pxi", line 92, in pyarrow.lib.check_status
raise convert_status(status)
pyarrow.lib.ArrowInvalid: JSON parse error: Column() changed from object to string in row 0
During handling of the above exception, another exception occurred:
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/json/json.py", line 101, in _split_generators
pa_table = next(iter(self._generate_tables(**splits[0].gen_kwargs, allow_full_read=False)))[1]
~~~~^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/usr/local/lib/python3.14/site-packages/datasets/packaged_modules/json/json.py", line 306, in _generate_tables
batch = json_encode_fields_in_json_lines(original_batch, json_field_paths)
File "/usr/local/lib/python3.14/site-packages/datasets/utils/json.py", line 111, in json_encode_fields_in_json_lines
examples = [ujson_loads(line) for line in original_batch.splitlines()]
~~~~~~~~~~~^^^^^^
File "/usr/local/lib/python3.14/site-packages/datasets/utils/json.py", line 20, in ujson_loads
return pd.io.json.ujson_loads(*args, **kwargs)
~~~~~~~~~~~~~~~~~~~~~~^^^^^^^^^^^^^^^^^
ValueError: Expected object or value
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 68, 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.
vla_framework — Blackwell 复现结果 (2026-07-07)
vla_framework(develop) 在 RTX PRO 4000 Blackwell ×2 (sm_120, driver 595) 上从零复现的评测结果与视频。
conda env vla_framework，torch 2.11.0+cu128。详见仓库 docs/REPRODUCTION_REPORT_2026-07-07.md。
1. SmolVLA 纯 VLA 基线 — place_bread_skillet
- 模型
arrow-hf/smolvla-robotwin-place-bread-skillet-50ep-multi，passthrough(无 MPC)，robotwin_gt，chunk_exec=50，max_steps=600，subprocess-per-episode。 - **env_success 4/10 (40%)**，对照文档 vtc benchmark multi 45% —— 采样噪声内吻合。
| seed | 结果 | ticks |
|---|---|---|
| 100000 | FAIL | 600 |
| 100001 | ✅ SUCCESS | 132 |
| 100002 | FAIL | 600 |
| 100003 | FAIL | 600 |
| 100004 | FAIL | 600 |
| 100005 | FAIL | 600 |
| 100006 | ✅ SUCCESS | 140 |
| 100007 | ✅ SUCCESS | 143 |
| 100008 | FAIL | 600 |
| 100009 | ✅ SUCCESS | 142 |
目录 smolvla_baseline_place_bread_skillet/：results.tsv + videos/ + constraint_results/。
2. SmolVLA + MPC + object_center — move_pillbottle_pad
- 模型
arrow-hf/smolvla-robotwin-move-pillbottle-pad-50ep-multi，corrector=mpc，grasp_source=object_center，robotwin_gt，seeds 100000–100004。 - **env_success 2/5 (40%)**。MPC 确实介入(
ENGAGE decision=replan,SEQ MPC scored=82/82)，但失败源于抓取不稳(holding margin≈-0.29)，与文档「MPC 相对纯 VLA 无稳定增益」一致。 - 抓取后端用 object_center 而非 EconomicGrasp：EG 依赖的 MinkowskiEngine 无法在 Blackwell(torch2.7/cu128)编译。
| seed | 结果 | ticks | replan |
|---|---|---|---|
| 100000 | ✅ SUCCESS | 132 | 0 |
| 100001 | FAIL | 600 | 480 |
| 100002 | FAIL | 600 | 423 |
| 100003 | FAIL | 600 | 470 |
| 100004 | ✅ SUCCESS | 141 | 0 |
目录 smolvla_mpc_oc_move_pillbottle_pad/：results.tsv + videos/ + constraint_results/。
视频命名
ep0_seed<SEED>_{success,fail}_<timestamp>.mp4 —— overlay 含约束/决策诊断。
- Downloads last month
- 107