Dataset · Text generation
qwen25-math-sft-long-think7k-v1
by Hoang Hai Dang danghoang2005/qwen25-math-sft-long-think7k-v1
2,500 train and 250 validation rows. Long examples have 4,096–7,168 thinking tokens; complete ChatML samples including final answer and EOS fit in 8,192 tokens.
Dataset Card
2,500 train and 250 validation rows. Long examples have 4,096–7,168 thinking tokens; complete ChatML samples including final answer and EOS fit in 8,192 tokens. Long source: open-r1/OpenR1-Math-220k at e4e141ec9dea9f8326f4d347be56105859b2bd68 (only math-verified generations). Short replay source: danghoang2005/qwen25-math-sft-long-15k-v1 at 32f0afc54a09c4963ae2478cf93871025abda901. Tokenizer: Qwen/Qwen2.5-Math-1.5B at 4a83ca6e4526a4f2da3aa259ec36c259f66b2ab2. Deduplicated across splits and screened against GSM8K, SVAMP, MATH-500, and AIME 2026. Review the source datasets' licenses and terms before reuse or redistribution. See manifest.json for counts and reproducibility settings.
Excerpt from the card by Hoang Hai Dang.
Structure
default 2,750 rows
| Split | Rows | Size |
|---|---|---|
| train | 2,500 | 80.8 MB |
| validation | 250 | 8.3 MB |
Details
- Repository
- danghoang2005/qwen25-math-sft-long-think7k-v1
- Publisher
- Hoang Hai Dang
- Task category
- Text generation
- Tags
- math, long-cot, qwen2.5-math
- Size category
- Not stated by the source
- Languages
- en
- Revision
- 57dbe28b85535a7394b4502abae036cdd4860ac6
- Last updated
- 2026-09-24
Files
5 files, 39.8 MB in total.
Every file
| File | Type | Size | SHA-256 |
|---|---|---|---|
| data/train-00000-of-00001.parquet | Data | 36.1 MB | 18c0c5e2475d |
| data/validation-00000-of-00001.parquet | Data | 3.7 MB | d3356b84c85a |
| manifest.json | Data | 762 B | — |
| README.md | Documentation | 823 B | — |
| .gitattributes | Repository | 2.5 KB | — |
License and Download
- License
- Not stated by the source
- Access
- No access gate
Released by Hoang Hai Dang through its official repository on Hugging Face.