Astra
Collection
Agent skill corpora and metadata, executable environments, and verified tool-use trajectories for agent research. • 6 items • Updated
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 84, in _split_generators
raise ValueError(
...<2 lines>...
)
ValueError: The TAR archives of the dataset should be in WebDataset format, but the files in the archive don't share the same prefix or the same types.
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.
🌳 Grove-500
The executable environment collection behind the end-to-end Grafting and Fathoming experiments
🤗 Hugging Face Dataset · 500 environments · Apache-2.0
Code · Grove-5K · Fathom-12K
Grove-500 contains the 500 self-contained executable environments used to generate the trajectory pool in Fathom-12K.
📦 At a Glance
| Field | Value |
|---|---|
| Environments | 500 |
| Archive | grove-500.tar.gz |
| World state | SQLite |
| Runtime | Deterministic Python |
| Related trajectories | 12,745 in Fathom-12K |
🚀 Quick Start
Download the dataset with the Hugging Face CLI:
uvx hf download zhangdw/grove-500 \
--type dataset \
--local-dir grove-500
Extract the environments:
tar -xzf grove-500/grove-500.tar.gz
Inspect one package:
ENV_DIR=$(find grove-500 -mindepth 1 -maxdepth 1 -type d | head -1)
cat "$ENV_DIR/spec/ENVIRONMENT.md"
head "$ENV_DIR/spec/tools.jsonl"
🧱 Environment Format
grove-500/
└── <environment_id>/
├── spec/
│ ├── ENVIRONMENT.md
│ ├── tool_graph.json
│ └── tools.jsonl
├── runtime/
│ └── backend.py
├── tests/
│ └── test_runtime.py
└── world/
├── schema.sql
└── world.db
| Component | Purpose |
|---|---|
tools.jsonl |
Tool names, descriptions, categories, and JSON schemas |
tool_graph.json |
Data dependencies and state preconditions between tools |
backend.py |
Deterministic tool execution against a supplied database |
world.db |
Populated seed world used to create session-local copies |
schema.sql |
Human-readable relational schema |
test_runtime.py |
Package-level runtime and workflow checks |
✅ Intended Use
Grove-500 is designed for:
- executable and stateful tool-use environment research;
- grounded task generation, trajectory synthesis, and replay;
- environment evolution and end-to-end Grafting experiments.
📚 Citation
If you use this dataset in your research, please cite our paper published in EMNLP 2026 (Main Conference):
From Agent Skills to Executable Environments: Frontier-Targeted Training for Tool-Use LLMs
@inproceedings{zhang-etal-2026-agent-skills,
title = "From Agent Skills to Executable Environments: Frontier-Targeted Training for Tool-Use {LLM}s",
author = "Zhang, Dawei and Huangfu, Yubin and Yao, Kyle Yusong and
Wei, Boyu and Han, Zongqi and Wu, Likang and
Zhao, Hongke and He, Ming and Fan, Jianping",
booktitle = "Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing",
year = "2026",
pages = "7026--7050",
url = "https://aclanthology.org/2026.emnlp-main.305/"
}
📄 License
Grove-500 is released under the Apache License 2.0.
- Downloads last month
- 44