yuruny/agentic-sudoku-NonMarkov_qwen2.5-3B-it-5e-6_9x9_6-6_gt-SFT_ans1-non_markovian-eval_results_dev

--- dataset_info: features: - name: steps list: - name: action dtype: string - name: chat_completions list: - name: content dtype: string - name: role dtype: string - name: done dtype: bool - name: mc_return dtype: float64 - name: model_response dtype: string - name: observation dtype: string - name: reward dtype: float64 - name: reward dtype: float64 splits: - name: train num_bytes: 575261 num_examples: 100 download_size: 63895 dataset_size: 575261 configs: - config_name: default data_files: - split: train path: data/train-* ---