Datasetpublic

LIF1014/ptdbench-task-hparam-overlong-023-dataset

PTDBench task_hparam_overlong_023 dataset snapshot This repository stores the exact case-specific DAPO-Math-17k_verl snapshot used by PTDBench task qwen_dapo_hparam/task_hparam_overlong_023. The files are published separately from other PTDBench tasks because processed dataset snapshots may differ between tasks. Consumers should pin a concrete repository commit and verify each file's SHA-256 digest. Storage format The original Parquet bytes are Base64-encoded as… See the full description on the dataset page: https://huggingface.co/datasets/LIF1014/ptdbench-task-hparam-overlong-023-dataset.

sourceHugging Faceapache-2.0updated 1mo agoView on Hugging Face

0likes52downloads

Dataset Card

PTDBench taskhparamoverlong_023 dataset snapshot

This repository stores the exact case-specific DAPO-Math-17k_verl snapshot used by PTDBench task qwen_dapo_hparam/task_hparam_overlong_023.

The files are published separately from other PTDBench tasks because processed dataset snapshots may differ between tasks. Consumers should pin a concrete repository commit and verify each file's SHA-256 digest.

Storage format

The original Parquet bytes are Base64-encoded as regular Hub text artifacts:

—artifacts/train.parquet.b64.part-00000.txt through artifacts/train.parquet.b64.part-00009.txt, concatenated in lexical order;
—artifacts/test.parquet.b64.txt.

PTDBench reconstructs the two Parquet files during the task image build and rejects them unless the resulting byte streams match the hashes below.

Files

File	Rows	Size (bytes)	SHA-256
`train.parquet`	16,125	5,885,886	`2897263b99e9b4af7523edeb5ec1ab9cac6e1bba63f0bfa78d4cd9356426ca5b`
`test.parquet`	1,792	675,271	`f860d1c3bfb0ce4a937b5e91bb43549a0a7ded51b1725c58c9c10f9f075d2aeb`

Provenance and license

This is a processed snapshot of DAPO-Math-17k prepared for the verl training data contract. The upstream and processed dataset are distributed under the Apache License 2.0.