igreck/

deepseek_grpo_correct_6144

Filtered GRPO training subset generated from deepseek-reasoner math generations.

train.jsonl: filtered training examples with prompt, solution, dataset_index, and DeepSeek metadata. metadata.json: filtering metadata.

Rows are kept when the raw generation is successful, stopped, correct, deduplicated by dataset_index, and has usage_total_tokens <= 6144.

Rows: 7576 Max total tokens: 6144 Source raw… See the full description on the dataset page: https://huggingface.co/datasets/igreck/deepseek_grpo_correct_6144.