shivangchopra11/maze_reasoning
收藏资源简介:
--- dataset_info: features: - name: Prompt dtype: string - name: Cot_Response dtype: string - name: Response dtype: string - name: raw_token dtype: string - name: maze_sizes dtype: int32 splits: - name: train num_bytes: 792552595 num_examples: 150000 - name: train_grpo num_bytes: 678345376 num_examples: 150000 - name: test_id num_bytes: 137994411 num_examples: 30000 - name: test_near_ood num_bytes: 819477059 num_examples: 30000 - name: test_far_ood num_bytes: 1494504366 num_examples: 20000 download_size: 387353094 dataset_size: 3922873807 configs: - config_name: default data_files: - split: train path: data/train-* - split: train_grpo path: data/train_grpo-* - split: test_id path: data/test_id-* - split: test_near_ood path: data/test_near_ood-* - split: test_far_ood path: data/test_far_ood-* ---



