rasdani's picture
Exclude unresolved RTS image and VM failures
bd6c1e8 verified
|
Raw
History Blame Contribute Delete
4.24 kB
---
license: cc-by-4.0
task_categories:
- reinforcement-learning
language:
- en
tags:
- recursive-task-synthesis
- command-line
- synthetic-tasks
configs:
- config_name: default
data_files:
- split: train
path: metadata/tasks.parquet
---
# Recursive Task Synthesis
Tasks without completed platform artifacts or with unresolved VM validation
failures are temporarily excluded. `exclusions.json` records the exact IDs,
reasons, build IDs where available, and evidence dates/runs. Exclusions affect
both metadata rows and complete TAR task packages. Runtime failures are not
image-build failures or proof of incorrect gold solutions. This filter does
not establish that every retained task passes gold validation.
Restore a task by verifying its platform artifacts and the failed execution
stage, removing its exclusion, and regenerating from the pinned source.
The patch evidence below also describes excluded tasks; consult the sidecar
for current membership.
Source: [Zhongzhi1228/Recursive-Task-Synthesis](https://huggingface.co/datasets/Zhongzhi1228/Recursive-Task-Synthesis/tree/be44f96808d5a9b599d5cb024341ff00091adeb7),
revision `be44f96808d5a9b599d5cb024341ff00091adeb7`. **36,249 rows retained; 1235 temporarily excluded.**
Nine Dockerfile repairs preserve the original task problems while fixing tmux
installation. A further 47 tasks have verified repairs: 42 process checks, three
reward-output paths, and two PostgreSQL startup prerequisites. These 47 tasks
scored gold 1 and no-op 0 in live checks. Three reference scripts exit nonzero
but their verifier rewards are 1; validity follows verifier reward.
Repairs update task packages and corresponding metadata together. Instructions
and reference solutions remain unchanged. Dockerfile-only repairs retain
`validation_status=patched_unverified`; the verified repairs use
`patched_gold_verified`. This does not claim full-corpus correctness.
The Docker-daemon recovery is deferred until its platform image is rebuilt and
validated; the current release does not claim that repair. No images are built
by this preparation script.
`patches.json` records source hashes, patch evidence and the content-hash algorithm
used for modified tasks. Unmodified hashes retain their upstream meaning. TAR
sizes/checksums and task byte counts are updated. Preparation code lives in
[prime-data](https://github.com/PrimeIntellect-ai/prime-data/blob/fix/rts-verified-recoveries/datasets/recursive-task-synthesis/prepare.py).
Original attribution and CC-BY-4.0 license follow below.
---
This dataset contains 37,484 validated command-line task instances produced
through recursive task synthesis. Public identifiers are opaque and stable.
- `metadata/tasks.parquet`: one searchable row per task instance.
- `metadata/shard_manifest.jsonl`: TAR sizes and SHA256 checksums.
- `data/tasks-*.tar`: sanitized runnable task packages.
The searchable task rows include:
- `instruction`: contents of `instruction.md`.
- `task_toml`: contents of `task.toml`.
- `solution`: contents of `solution/solve.sh`, or null when absent.
- `dockerfile`: contents of `environment/Dockerfile`, or null when absent.
- `member_prefix` and `shard`: locations of the complete runnable task package.
```python
from datasets import load_dataset
tasks = load_dataset("Zhongzhi1228/Recursive-Task-Synthesis", split="train")
```
## License, attribution, and modifications
This dataset is released under the
[Creative Commons Attribution 4.0 International License](https://creativecommons.org/licenses/by/4.0/).
See `LICENSE` for the complete legal text.
The dataset was produced through recursive task synthesis and executable
validation. Depending on the task, instructions, solutions, verifiers,
environments, metadata, identifiers, and packaging were rewritten, extended,
sanitized, or regenerated. Each published task is paired with searchable
metadata and a complete runnable package.
This license covers the original contributions and
adaptations that the publisher has authority to license. It does not grant
rights over third-party materials beyond what is permitted by their original
terms.
Detailed methodology, provenance, and related-work attribution are documented
in the accompanying paper.