open-r1/DAPO-Math-17k-Processed
Dataset Card for DAPO-Math-17k-Processed This is a processed version of BytedTsinghua-SIA/DAPO-Math-17k where we have: Deduplicated the prompts Reformatted the prompts and ground truth answers to be compatible with TRL's GRPO trainer We have also derived pure English and Chinese subsets. The full dataset processing logic can be found in create_dataset.py. If you find this dataset useful in your work, please cite the original source with:… See the full description on the dataset page: https://huggingface.co/datasets/open-r1/DAPO-Math-17k-Processed.
| Author / organization | open-r1 |
|---|---|
| Downloads | 5.3K (as reported by Hugging Face) |
| Likes | 89 |
| Created | 10 April 2025 |
| Last updated at source | 10 November 2025 |
Popularity, last 2 days
5,336 now (-71 since 2026-10-16). Counts as reported by the source, snapshotted daily by Silytics.
Connected in the knowledge graph
- Palind/Qwen3-4B-TIS-repro-bestModel · used to train
- Palind/Qwen3-4B-SC-TIS-repro-bestModel · used to train
- mradermacher/Llama3.1-8B-Thinking-R1-GGUFModel · used to train
- Palind/Qwen3-4B-PG-repro-bestModel · used to train
- Palind/Qwen3-4B-TIS-repro-finalModel · used to train
- Palind/Qwen3-4B-SC-repro-bestModel · used to train
- hdong0/Qwen3-8B-base-Open-R1-GRPO_dapo_acc_4096_noklModel · used to train
- Palind/Qwen3-4B-PG-repro-finalModel · used to train
- hdong0/Qwen3-8B-base-Open-R1-GRPO_dapo_acc_2048_noklModel · used to train
- Palind/Qwen3-4B-SC-repro-finalModel · used to train
- mradermacher/Llama3.1-8B-Thinking-R1-i1-GGUFModel · used to train
- Palind/Qwen3-4B-SC-TIS-repro-finalModel · used to train
Recent changes
- 17 Oct 2026 Description: Dataset Card for DAPO-Math-17k-Processed This is a processed version of BytedTsinghua-SIA/DAPO-Math-17k where we have: Deduplicated the prompts Reformatted the prompts and ground truth answers to be compatible with TRL's GRPO trainer We hav → Dataset Card for DAPO-Math-17k-Processed This is a processed version of BytedTsinghua-SIA/DAPO-Math-17k where we have: Deduplicated the prompts Reformatted the prompts and ground truth answers to be compatible with TRL's GRPO trainer We hav
Changes detected by Silytics between refreshes of the source.
More from open-r1
- open-r1/OpenR1-Math-220kDataset
- open-r1/codeforcesDataset
- open-r1/Mixture-of-ThoughtsDataset
- open-r1/OpenThoughts-114k-mathDataset
- open-r1/codeforces-submissionsDataset
- open-r1/codeforces-cotsDataset
Silytics shows public metadata from Hugging Face and links to the original. Information may be incomplete or out of date. Last refreshed 17 October 2026. Licenses and terms of the original apply to the work itself. Not financial, legal or professional advice.