Datasets:

Kwai-Klear
/

KlearReasoner-CodeSub-15K

Modalities:

Formats:

Size:

ArXiv:

Libraries:

License:

Dataset card Data Studio Files Files and versions

Suu commited on Aug 13

Commit

0c6cafe

·

verified ·

1 Parent(s): 78617ac

Update README.md

Files changed (1) hide show

README.md +2 -3

README.md CHANGED Viewed

@@ -4,13 +4,12 @@ license: apache-2.0
 This dataset is a high-quality subset of the Klear-Reasoner Code RL dataset, derived from the RL data used in the [rllm project](https://github.com/agentica-project/rllm). Part of this data contributed to training Klear-Reasoner’s code reasoning models.
 The dataset is carefully cleaned and filtered to include only reliable samples suitable for reinforcement learning. Models trained with this dataset have shown substantial performance improvements across various code reasoning benchmarks.
 You can load the dataset via the Hugging Face datasets library:
-from datasets import load_dataset
 ```python
 from datasets import load_dataset
-dataset = load_dataset("Suu/KlearReasoner-MathSub-30K")
 ```
-See our paper and GitHub repository for more details.
 | Resource | Link |
 |---|---|

 This dataset is a high-quality subset of the Klear-Reasoner Code RL dataset, derived from the RL data used in the [rllm project](https://github.com/agentica-project/rllm). Part of this data contributed to training Klear-Reasoner’s code reasoning models.
 The dataset is carefully cleaned and filtered to include only reliable samples suitable for reinforcement learning. Models trained with this dataset have shown substantial performance improvements across various code reasoning benchmarks.
 You can load the dataset via the Hugging Face datasets library:
 ```python
 from datasets import load_dataset
+dataset = load_dataset("Suu/KlearReasoner-CodeSub-15K")
 ```
 | Resource | Link |
 |---|---|