Datasets:

sungyub
/

deepscaler-preview-verl

license: mit
language:
  - en
tags:
  - math
  - reasoning
  - verl
  - reinforcement-learning
  - math-reasoning
size_categories:
  - 10K<n<100K
task_categories:
  - text-generation
  - question-answering
dataset_info:
  features:
    - name: data_source
      dtype: string
    - name: prompt
      list:
        - name: role
          dtype: string
        - name: content
          dtype: string
    - name: ability
      dtype: string
    - name: reward_model
      struct:
        - name: style
          dtype: string
        - name: ground_truth
          dtype: string
    - name: extra_info
      struct:
        - name: index
          dtype: int64
        - name: original_dataset
          dtype: string
        - name: split
          dtype: string
  splits:
    - name: train
      num_bytes: 0
      num_examples: 35789
  download_size: 0
  dataset_size: 5000000

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: agentica-org/DeepScaleR-Preview-Dataset
License: MIT
Authors: Agentica Organization

Dataset Description

DeepScaleR-Preview-Dataset contains approximately 40,000 unique mathematics problem-answer pairs compiled from AIME (American Invitational Mathematics Examination, 1984-2023), AMC (American Mathematics Competition, pre-2023), Omni-MATH dataset, and Still dataset. The dataset was used to train the DeepScaleR-1.5B-Preview model using distributed reinforcement learning.

🔄 Preprocessing Pipeline

This dataset has been preprocessed and converted to the VERL (Verification and Reinforcement Learning) format for use in mathematical reasoning tasks with reward modeling.

Cleaning Methodology

Standard Processing:

URL filtering (samples containing URLs removed)
Format normalization to VERL schema
Basic text cleaning and validation

Deduplication

Intra-dataset Deduplication:

Method: SHA-256 hash-based with text normalization
Before deduplication: 40,000 samples
After deduplication: 35,789 samples
Reduction: 10.5%

Inter-dataset Deduplication (v3.0):

Priority level: 1
Cross-dataset duplicates removed: 2,000

💡 Preprocessing Examples

Example 1: Standard Processing (No Cleaning Preset)

Before Cleaning:

Calculate the derivative of $f(x) = x^2 + 3x + 2$ with respect to $x$.

After Cleaning:

Calculate the derivative of $f(x) = x^2 + 3x + 2$ with respect to $x$.

Changes Applied:

✓ Format normalization to VERL schema
✓ URL filtering (samples with URLs removed)
✓ Basic text validation
✓ No artifact removal applied

📐 VERL Schema

This dataset follows the standardized VERL (Verification and Reinforcement Learning) format:

{
  "data_source": "openai/gsm8k",
  "prompt": [
    {
      "content": "Calculate the sum of all odd numbers from 1 to 99.",
      "role": "user"
    }
  ],
  "ability": "math",
  "reward_model": {
    "style": "rule",
    "ground_truth": "\boxed{2500}",
    "hash": "sha256:abc123..."
  },
  "extra_info": {
    "split": "train"
  }
}

Field Descriptions

Field	Type	Description
`data_source`	`string`	Original dataset identifier (e.g., `openai/gsm8k`, `numina_aime`)
`prompt`	`list[dict]`	User query in chat format with role and content
`ability`	`string`	Task type (always `"math"` for this dataset)
`reward_model.style`	`string`	Reward computation method (`"rule"` for rule-based verification)
`reward_model.ground_truth`	`string`	Expected answer for verification (often in `\boxed{}` format)
`reward_model.hash`	`string`	SHA-256 hash of prompt content for deduplication
`extra_info.split`	`string`	Original split identifier (`"train"`, `"test"`, etc.)

📈 Dataset Statistics

Sample Distribution

Total Samples: 35,789
Dataset Size: 4.8 MB
Average Problem Length: N/A characters

Data Sources

Distribution of problems by original data source:

Source	Count	Percentage
Mixed Sources	35,789	100%

Note: Detailed source distribution statistics will be added in future updates.

🚀 Usage

Loading the Dataset

from datasets import load_dataset

# Load the full dataset
dataset = load_dataset("sungyub/deepscaler-preview-verl")

# Load with streaming (recommended for large datasets)
dataset = load_dataset("sungyub/deepscaler-preview-verl", streaming=True)

# Preview first few examples
for example in dataset['train'].take(5):
    print(example['prompt'][0]['content'])  # User question
    print(example['reward_model']['ground_truth'])  # Answer
    print("---")

Using with VERL

from datatrove.utils.reward_score import compute_score

# Compute reward score for a generated solution
score = compute_score(
    data_source=example['data_source'],
    solution_str=generated_solution,
    ground_truth=example['reward_model']['ground_truth'],
    format_type="auto"  # Auto-detect XML or GPT OSS format
)

print(f"Reward score: {score}")

Integration with DataTrove

from datatrove.pipeline.readers import ParquetReader
from datatrove.pipeline.filters import LambdaFilter
from datatrove.executor import LocalPipelineExecutor

pipeline = [
    ParquetReader("sungyub/deepscaler-preview-verl", text_key="prompt"),
    LambdaFilter(lambda doc: len(doc.text) > 100),  # Filter short problems
    # Add more processing steps...
]

executor = LocalPipelineExecutor(pipeline=pipeline, tasks=4)
executor.run()

📚 Citation

Original Dataset

@dataset{agentica_deepscaler_2025,
  author = {Agentica Organization},
  title = {DeepScaleR-Preview-Dataset: AIME and AMC Mathematics Collection},
  year = {2025},
  publisher = {Hugging Face},
  howpublished = {\url{https://huggingface.co/datasets/agentica-org/DeepScaleR-Preview-Dataset}}
}

This Processed Version

@dataset{sungyub_math_verl_deepscaler-preview-verl,
  author = {Sungyub Kim},
  title = DeepScaleR-Preview VERL,
  year = {2025},
  publisher = {Hugging Face},
  howpublished = {\url{https://huggingface.co/datasets/sungyub/deepscaler-preview-verl}}
}

⚖️ License

This processed dataset: mit
Original dataset: MIT

🙏 Acknowledgments

This dataset was processed using the DataTrove library.

Credits:

Original dataset authors: Agentica Organization
Processing and VERL conversion: Sungyub Kim
MathDatasetCleaner implementation: DataTrove contributors

Special thanks to: Agentica team for curating competition-level mathematics problems

📝 Version History

v1.0.0 (Initial Release)

Processed 35,789 samples from agentica-org/DeepScaleR-Preview-Dataset
Converted to VERL format
Standard processing applied
Ready for reinforcement learning training

🔗 Related Resources

Unified Collection: sungyub/math-verl-unified - All 9 math datasets with inter-dataset deduplication
DataTrove Documentation: https://github.com/huggingface/datatrove
VERL Format Specification: See VERL Schema section above

Questions or issues? Open an issue on the DataTrove GitHub repository