datasets
Training and evaluation data, with the modality, task and licence stated up front. Listed live from the Hugging Face Hub.
deepscaler-preview-verl
DeepScaleR-Preview VERL
📊 Dataset Summary
This dataset contains 35,789 mathematical reasoning problems in VERL format, processed from agentica-org/DeepScaleR-Preview-Dataset.
Key Features:
35,789 high-quality math problems
Converted to VERL format for reward modeling
Verified ground truth answers
Ready for reinforcement learning training
🔗 Source Dataset
Original Repository
Repository:… See the full description on the dataset page: https://huggingface.co/datasets/sungyub/deepscaler-preview-verl.DeepScaleR-Preview-Dataset-verl-formataime-solution-hint-v6-deepscaler-respgenDeepScaleR-Easy-Medium-Hard-Gemma-26B-PT-10k
DeepScaleR Easy/Medium/Hard — Gemma 4 26B-A4B PT
This dataset contains 9,900 unique, deduplicated DeepScaleR math questions for
reinforcement-learning experiments. Difficulty is defined by how often the
pretrained google/gemma-4-26B-A4B teacher solved each question across eight
temperature-1 samples under the same rule-based grader used by the RL training
pipeline.
The Hub dataset has three configurations—easy, medium, and hard—and each
configuration has a train split with 3,000… See the full description on the dataset page: https://huggingface.co/datasets/JWei05/DeepScaleR-Easy-Medium-Hard-Gemma-26B-PT-10k.deepscaler-verl-aha-momentverl-deepscaler-clean
verl-deepscaler-clean
A cleaned, deduplicated release of the DeepScaleR math problem set, packaged in the parquet layout
that verl expects from its RLHFDataset, so it can be dropped
straight into a GRPO / rule-based-reward run (for example, when re-running an R1-style "aha moment"
experiment on a smaller model).
The whole dataset lives in a single parquet file at the repo root: verl_deepscaler.parquet.
Headline numbers
Total lines read
3038
Records… See the full description on the dataset page: https://huggingface.co/datasets/dusersad12/verl-deepscaler-clean.deepscaler-problem_only_qwen14bgenLaser-Deepscaler-DatasetDeepScaleR-Qwen3-1.7B-0-40kaime-solution-hint-v6-deepscaler-respgen__0_115DeepScaleR-EasyThe easy part (pass_rate > 4/8) of DeepScaleR
Check https://wutaiqiang.github.io/project/Art for more details
Citation:
@inproceedings{wu2026art,
title={The Art of Efficient Reasoning: Data, Reward, and Optimization},
author={Taiqiang Wu and Zenan Xu and Bo Zhou and Ngai Wong},
year={2026},
url={https://arxiv.org/pdf/2602.20945}
}
r1-qwen7b-deepscaler-n32
deepseek-r1-qwen-7b generations for deepscaler dataset
The original deepscaler dataset has been filtered:
we removed all synthetic data because their problem-answer may not match.
based on generations from Qwen/Qwen2.5-Math-7B-Instruct (pre-o1), we removed problems that has at least 5/32 correct generations.
We then use deepseek-ai/DeepSeek-R1-Distill-Qwen-7B to generate from this filtered dataset with num_generations=32 and max_tokens=8192
---
dataset_info:
features:
- name:… See the full description on the dataset page: https://huggingface.co/datasets/drproduck/r1-qwen7b-deepscaler-n32.open-deepscaler
Open-DeepScaleR Dataset
Summary
The open-deepscaler dataset comprises 21,044 challenging mathematical reasoning problems, sourced from the DeepScaleR dataset. It supports the Open RS project, enhancing reasoning in small LLMs via reinforcement learning.
Usage
Load the dataset using the Hugging Face datasets library:
from datasets import load_dataset
ds = load_dataset("knoveleng/open-deepscaler")["train"]
print(ds[0])
Dataset Structure… See the full description on the dataset page: https://huggingface.co/datasets/knoveleng/open-deepscaler.aime-solution-hint-v6-deepscaler-respgen__115_230DeepScaleR-Qwen3-1.7B-rl-wholeverl_deepscaler
DeepScaleR for verl (full cleaned build)
RL-ready dataset in verl parquet format, rebuilt from the
full three-shard DeepScaleR mirror dump.
Build process
Merged the three mirror shards (deepscaler_shard_00.json, deepscaler_shard_01.json,
deepscaler_shard_02.json) in ascending filename order, keeping each shard's record order.
Dropped records whose problem or answer was missing or whitespace-only (78 dropped).
An empty solution is normal in this export and is… See the full description on the dataset page: https://huggingface.co/datasets/dusersad12/verl_deepscaler.aime-solution-hint-v6-deepscaler-respgen__805_919aime-solution-hint-v6-deepscaler-respgen__230_345deepscaler_prepare_logp_inputswift-reasoning-rollouts-deepscaler-ministral8b
DeepScaleR Reasoning Rollouts (Ministral-8B)
This dataset contains reasoning rollouts used to train the SWIFT reward head.
Paper page: https://huggingface.co/papers/2505.12225
GitHub: https://github.com/aster2024/SWIFT/
Generator model: mistralai/Ministral-8B-Instruct-2410 (https://huggingface.co/mistralai/Ministral-8B-Instruct-2410)
Dataset Description
This dataset contains 10000 samples corresponding to the Generalization Test setup.
Source: DeepScaleR.
Generator:… See the full description on the dataset page: https://huggingface.co/datasets/Aster2024/swift-reasoning-rollouts-deepscaler-ministral8b.aime-solution-hint-v6-deepscaler-respgen__690_805aime-solution-hint-v6-deepscaler-respgen__345_460DeepScaleR-Qwen3-1.7B-23k-classifiedaime-solution-hint-v6-deepscaler-respgen__460_575aime-solution-hint-v6-deepscaler-respgen__575_690DeepScaleR-Qwen3-1.7B-0-40k-not-all-correctdeepscaler-easy10k-skyrl
deepscaler-easy10k-skyrl
Data for the short-horizon GRPO-vs-PPO study (Qwen2.5-3B, SkyRL). Parquets are in SkyRL's data.train_data / data.val_data format.
train/train.parquet — 9,800 DeepScaleR prompts, difficulty-balanced over buckets 1/5–4/5 (2,450 each); train/val.parquet (200 held-out) and train/val_x16.parquet are the original in-run val sets (legacy, do not use for reporting). train/manifest.json records the split.
eval/deepscaler_dedup500_x4.parquet — 500 deduplicated… See the full description on the dataset page: https://huggingface.co/datasets/JWei05/deepscaler-easy10k-skyrl.DeepScaleR-Qwen3-1.7B-22-40kDeepScaleR-40k-Prompt-Filtereddeepscaler-completion_solution_qwen14bgen