mzio/

aprm-tw-cooking-game-easy-gpt5m_med-gs4-s0-train

	Act-PRM Rollout Dataset




	
	Run Metadata

env_config: textworld/cooking_game_easy model_config: oai_gpt5m_med model: gpt-5-mini split: train group_size: 4 seed: 0 num_samples: 80 num_trajectories: 320 accuracy: 99.7% mean_reward: 0.997 run_cmd: main_api.py --env_config textworld/cooking_game_easy --model_config oai_gpt5m_med --split train --group_size 4 --max_concurrent 5 --checkpoint_every 100 --dataset_name mzio/aprm-tw-cooking-game-easy-gpt5m_med-gs4-s0-train --seed 0… See the full description on the dataset page: https://huggingface.co/datasets/mzio/aprm-tw-cooking-game-easy-gpt5m_med-gs4-s0-train.