use_accel: false exploratory_grad_updates: true num_edits: 2 score_function: MaxMC level_buffer_capacity: 4000 replay_prob: 0.5 staleness_coeff: 0.3 temperature: 1.0 topk_k: 8 minimum_fill_ratio: 0.5 prioritization: rank buffer_duplicate_check: false buffer_train: false mode: train checkpoint_directory: checkpoints/physicsenv/ued max_number_of_checkpoints: 5 accel_start_from_empty: True