fossil-eai/babyai-xgen-random-failed

--- dataset_info: - config_name: gotosimilardistsuniquegoalstrictlycolortype_1024x1_preprocessed_sentence-transformer-pretrained_cnn-scratch features: - name: task_id dtype: int64 - name: timesteps sequence: int64 - name: actions sequence: int64 - name: img_obss sequence: array3_d: shape: - 3 - 56 - 56 dtype: float64 - name: terminations sequence: int64 - name: discounted_rewards sequence: float64 - name: undiscounted_rewards sequence: float64 - name: normalised_discounted_rewards sequence: float64 - name: discounted_rtgs_discounted_rewards sequence: float64 - name: undiscounted_rtgs_discounted_rewards sequence: float64 - name: discounted_rtgs_undiscounted_rewards sequence: float64 - name: undiscounted_rtgs_undiscounted_rewards sequence: float64 - name: discounted_rtgs_normalised_discounted_rewards sequence: float64 - name: undiscounted_rtgs_normalised_discounted_rewards sequence: float64 - name: discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_dense_rewards_subgoal_success sequence: float64 - name: normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_discounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_undiscounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_undiscounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: dense_rewards_action_failure sequence: float64 - name: dense_rtgs_action_failure sequence: float64 - name: dense_rewards sequence: float64 - name: dense_rtgs sequence: float64 - name: mission_embeddings sequence: sequence: float32 - name: mission_token_mask sequence: bool - name: rule_feedback_embeddings sequence: sequence: sequence: float32 - name: rule_feedback_token_mask sequence: sequence: bool - name: rule_feedback_available sequence: bool - name: task_feedback_embeddings sequence: sequence: sequence: float32 - name: task_feedback_token_mask sequence: sequence: bool - name: task_feedback_available sequence: bool - name: instruction_embeddings sequence: sequence: sequence: float32 - name: instruction_token_mask sequence: sequence: bool - name: instructions_available sequence: bool splits: - name: train num_bytes: 2310881468 num_examples: 819 - name: val num_bytes: 578425764 num_examples: 205 download_size: 66970762 dataset_size: 2889307232 - config_name: gotosimilardistsuniquegoalstrictlycolortype_1024x1_raw features: - name: task_id dtype: int64 - name: episode_name dtype: string - name: mission dtype: string - name: timesteps sequence: int64 - name: actions sequence: int64 - name: img_obss sequence: array3_d: shape: - 3 - 56 - 56 dtype: float64 - name: task_feedback sequence: string - name: rule_feedback sequence: string - name: instructions sequence: string - name: terminations sequence: int64 - name: discounted_rewards sequence: float64 - name: undiscounted_rewards sequence: float64 - name: normalised_discounted_rewards sequence: float64 - name: discounted_rtgs_discounted_rewards sequence: float64 - name: undiscounted_rtgs_discounted_rewards sequence: float64 - name: discounted_rtgs_undiscounted_rewards sequence: float64 - name: undiscounted_rtgs_undiscounted_rewards sequence: float64 - name: discounted_rtgs_normalised_discounted_rewards sequence: float64 - name: undiscounted_rtgs_normalised_discounted_rewards sequence: float64 - name: discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_dense_rewards_subgoal_success sequence: float64 - name: normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_discounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_undiscounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_undiscounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: dense_rewards_action_failure sequence: float64 - name: dense_rtgs_action_failure sequence: float64 - name: dense_rewards sequence: float64 - name: dense_rtgs sequence: float64 - name: subgoal_successes sequence: int64 - name: subgoals dtype: int64 - name: action_failures sequence: int64 - name: max_steps dtype: int64 - name: episode_len dtype: int64 splits: - name: train num_bytes: 2061818482 num_examples: 819 - name: val num_bytes: 516083644 num_examples: 205 download_size: 15142969 dataset_size: 2577902126 - config_name: gotosimilardistsuniquegoalstrictlycolortype_2048x1_preprocessed_sentence-transformer-pretrained_cnn-scratch features: - name: task_id dtype: int64 - name: timesteps sequence: int64 - name: actions sequence: int64 - name: img_obss sequence: array3_d: shape: - 3 - 56 - 56 dtype: float64 - name: terminations sequence: int64 - name: discounted_rewards sequence: float64 - name: undiscounted_rewards sequence: float64 - name: normalised_discounted_rewards sequence: float64 - name: discounted_rtgs_discounted_rewards sequence: float64 - name: undiscounted_rtgs_discounted_rewards sequence: float64 - name: discounted_rtgs_undiscounted_rewards sequence: float64 - name: undiscounted_rtgs_undiscounted_rewards sequence: float64 - name: discounted_rtgs_normalised_discounted_rewards sequence: float64 - name: undiscounted_rtgs_normalised_discounted_rewards sequence: float64 - name: discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_dense_rewards_subgoal_success sequence: float64 - name: normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_discounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_undiscounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_undiscounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: dense_rewards_action_failure sequence: float64 - name: dense_rtgs_action_failure sequence: float64 - name: dense_rewards sequence: float64 - name: dense_rtgs sequence: float64 - name: mission_embeddings sequence: sequence: float32 - name: mission_token_mask sequence: bool - name: rule_feedback_embeddings sequence: sequence: sequence: float32 - name: rule_feedback_token_mask sequence: sequence: bool - name: rule_feedback_available sequence: bool - name: task_feedback_embeddings sequence: sequence: sequence: float32 - name: task_feedback_token_mask sequence: sequence: bool - name: task_feedback_available sequence: bool - name: instruction_embeddings sequence: sequence: sequence: float32 - name: instruction_token_mask sequence: sequence: bool - name: instructions_available sequence: bool splits: - name: train num_bytes: 4621762712 num_examples: 1638 - name: val num_bytes: 1156851528 num_examples: 410 download_size: 133691140 dataset_size: 5778614240 - config_name: gotosimilardistsuniquegoalstrictlycolortype_2048x1_raw features: - name: task_id dtype: int64 - name: episode_name dtype: string - name: mission dtype: string - name: timesteps sequence: int64 - name: actions sequence: int64 - name: img_obss sequence: array3_d: shape: - 3 - 56 - 56 dtype: float64 - name: task_feedback sequence: string - name: rule_feedback sequence: string - name: instructions sequence: string - name: terminations sequence: int64 - name: discounted_rewards sequence: float64 - name: undiscounted_rewards sequence: float64 - name: normalised_discounted_rewards sequence: float64 - name: discounted_rtgs_discounted_rewards sequence: float64 - name: undiscounted_rtgs_discounted_rewards sequence: float64 - name: discounted_rtgs_undiscounted_rewards sequence: float64 - name: undiscounted_rtgs_undiscounted_rewards sequence: float64 - name: discounted_rtgs_normalised_discounted_rewards sequence: float64 - name: undiscounted_rtgs_normalised_discounted_rewards sequence: float64 - name: discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_dense_rewards_subgoal_success sequence: float64 - name: normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_discounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_undiscounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_undiscounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: dense_rewards_action_failure sequence: float64 - name: dense_rtgs_action_failure sequence: float64 - name: dense_rewards sequence: float64 - name: dense_rtgs sequence: float64 - name: subgoal_successes sequence: int64 - name: subgoals dtype: int64 - name: action_failures sequence: int64 - name: max_steps dtype: int64 - name: episode_len dtype: int64 splits: - name: train num_bytes: 4123648911 num_examples: 1638 - name: val num_bytes: 1032167493 num_examples: 410 download_size: 30184410 dataset_size: 5155816404 - config_name: gotosimilardistsuniquegoalstrictlycolortype_4096x1_preprocessed_sentence-transformer-pretrained_cnn-scratch features: - name: task_id dtype: int64 - name: timesteps sequence: int64 - name: actions sequence: int64 - name: img_obss sequence: array3_d: shape: - 3 - 56 - 56 dtype: float64 - name: terminations sequence: int64 - name: discounted_rewards sequence: float64 - name: undiscounted_rewards sequence: float64 - name: normalised_discounted_rewards sequence: float64 - name: discounted_rtgs_discounted_rewards sequence: float64 - name: undiscounted_rtgs_discounted_rewards sequence: float64 - name: discounted_rtgs_undiscounted_rewards sequence: float64 - name: undiscounted_rtgs_undiscounted_rewards sequence: float64 - name: discounted_rtgs_normalised_discounted_rewards sequence: float64 - name: undiscounted_rtgs_normalised_discounted_rewards sequence: float64 - name: discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_dense_rewards_subgoal_success sequence: float64 - name: normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_discounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_undiscounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_undiscounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: dense_rewards_action_failure sequence: float64 - name: dense_rtgs_action_failure sequence: float64 - name: dense_rewards sequence: float64 - name: dense_rtgs sequence: float64 - name: mission_embeddings sequence: sequence: float32 - name: mission_token_mask sequence: bool - name: rule_feedback_embeddings sequence: sequence: sequence: float32 - name: rule_feedback_token_mask sequence: sequence: bool - name: rule_feedback_available sequence: bool - name: task_feedback_embeddings sequence: sequence: sequence: float32 - name: task_feedback_token_mask sequence: sequence: bool - name: task_feedback_available sequence: bool - name: instruction_embeddings sequence: sequence: sequence: float32 - name: instruction_token_mask sequence: sequence: bool - name: instructions_available sequence: bool splits: - name: train num_bytes: 9243525200 num_examples: 3276 - name: val num_bytes: 2313703056 num_examples: 820 download_size: 268101379 dataset_size: 11557228256 - config_name: gotosimilardistsuniquegoalstrictlycolortype_4096x1_raw features: - name: task_id dtype: int64 - name: episode_name dtype: string - name: mission dtype: string - name: timesteps sequence: int64 - name: actions sequence: int64 - name: img_obss sequence: array3_d: shape: - 3 - 56 - 56 dtype: float64 - name: task_feedback sequence: string - name: rule_feedback sequence: string - name: instructions sequence: string - name: terminations sequence: int64 - name: discounted_rewards sequence: float64 - name: undiscounted_rewards sequence: float64 - name: normalised_discounted_rewards sequence: float64 - name: discounted_rtgs_discounted_rewards sequence: float64 - name: undiscounted_rtgs_discounted_rewards sequence: float64 - name: discounted_rtgs_undiscounted_rewards sequence: float64 - name: undiscounted_rtgs_undiscounted_rewards sequence: float64 - name: discounted_rtgs_normalised_discounted_rewards sequence: float64 - name: undiscounted_rtgs_normalised_discounted_rewards sequence: float64 - name: discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_dense_rewards_subgoal_success sequence: float64 - name: normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_discounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_undiscounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_undiscounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: dense_rewards_action_failure sequence: float64 - name: dense_rtgs_action_failure sequence: float64 - name: dense_rewards sequence: float64 - name: dense_rtgs sequence: float64 - name: subgoal_successes sequence: int64 - name: subgoals dtype: int64 - name: action_failures sequence: int64 - name: max_steps dtype: int64 - name: episode_len dtype: int64 splits: - name: train num_bytes: 8247290265 num_examples: 3276 - name: val num_bytes: 2064346088 num_examples: 820 download_size: 60228865 dataset_size: 10311636353 - config_name: gotosimilardistsuniquegoalstrictlyloc_1024x1_preprocessed_sentence-transformer-pretrained_cnn-scratch features: - name: task_id dtype: int64 - name: timesteps sequence: int64 - name: actions sequence: int64 - name: img_obss sequence: array3_d: shape: - 3 - 56 - 56 dtype: float64 - name: terminations sequence: int64 - name: discounted_rewards sequence: float64 - name: undiscounted_rewards sequence: float64 - name: normalised_discounted_rewards sequence: float64 - name: discounted_rtgs_discounted_rewards sequence: float64 - name: undiscounted_rtgs_discounted_rewards sequence: float64 - name: discounted_rtgs_undiscounted_rewards sequence: float64 - name: undiscounted_rtgs_undiscounted_rewards sequence: float64 - name: discounted_rtgs_normalised_discounted_rewards sequence: float64 - name: undiscounted_rtgs_normalised_discounted_rewards sequence: float64 - name: discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_dense_rewards_subgoal_success sequence: float64 - name: normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_discounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_undiscounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_undiscounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: dense_rewards_action_failure sequence: float64 - name: dense_rtgs_action_failure sequence: float64 - name: dense_rewards sequence: float64 - name: dense_rtgs sequence: float64 - name: mission_embeddings sequence: sequence: float32 - name: mission_token_mask sequence: bool - name: rule_feedback_embeddings sequence: sequence: sequence: float32 - name: rule_feedback_token_mask sequence: sequence: bool - name: rule_feedback_available sequence: bool - name: task_feedback_embeddings sequence: sequence: sequence: float32 - name: task_feedback_token_mask sequence: sequence: bool - name: task_feedback_available sequence: bool - name: instruction_embeddings sequence: sequence: sequence: float32 - name: instruction_token_mask sequence: sequence: bool - name: instructions_available sequence: bool splits: - name: train num_bytes: 2310881468 num_examples: 819 - name: val num_bytes: 578425764 num_examples: 205 download_size: 73608712 dataset_size: 2889307232 - config_name: gotosimilardistsuniquegoalstrictlyloc_2048x1_preprocessed_sentence-transformer-pretrained_cnn-scratch features: - name: task_id dtype: int64 - name: timesteps sequence: int64 - name: actions sequence: int64 - name: img_obss sequence: array3_d: shape: - 3 - 56 - 56 dtype: float64 - name: terminations sequence: int64 - name: discounted_rewards sequence: float64 - name: undiscounted_rewards sequence: float64 - name: normalised_discounted_rewards sequence: float64 - name: discounted_rtgs_discounted_rewards sequence: float64 - name: undiscounted_rtgs_discounted_rewards sequence: float64 - name: discounted_rtgs_undiscounted_rewards sequence: float64 - name: undiscounted_rtgs_undiscounted_rewards sequence: float64 - name: discounted_rtgs_normalised_discounted_rewards sequence: float64 - name: undiscounted_rtgs_normalised_discounted_rewards sequence: float64 - name: discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_dense_rewards_subgoal_success sequence: float64 - name: normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_discounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_undiscounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_undiscounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: dense_rewards_action_failure sequence: float64 - name: dense_rtgs_action_failure sequence: float64 - name: dense_rewards sequence: float64 - name: dense_rtgs sequence: float64 - name: mission_embeddings sequence: sequence: float32 - name: mission_token_mask sequence: bool - name: rule_feedback_embeddings sequence: sequence: sequence: float32 - name: rule_feedback_token_mask sequence: sequence: bool - name: rule_feedback_available sequence: bool - name: task_feedback_embeddings sequence: sequence: sequence: float32 - name: task_feedback_token_mask sequence: sequence: bool - name: task_feedback_available sequence: bool - name: instruction_embeddings sequence: sequence: sequence: float32 - name: instruction_token_mask sequence: sequence: bool - name: instructions_available sequence: bool splits: - name: train num_bytes: 4621762712 num_examples: 1638 - name: val num_bytes: 1156851528 num_examples: 410 download_size: 144751011 dataset_size: 5778614240 - config_name: gotosimilardistsuniquegoalstrictlyloc_2048x1_raw features: - name: task_id dtype: int64 - name: episode_name dtype: string - name: mission dtype: string - name: timesteps sequence: int64 - name: actions sequence: int64 - name: img_obss sequence: array3_d: shape: - 3 - 56 - 56 dtype: float64 - name: task_feedback sequence: string - name: rule_feedback sequence: string - name: instructions sequence: string - name: terminations sequence: int64 - name: discounted_rewards sequence: float64 - name: undiscounted_rewards sequence: float64 - name: normalised_discounted_rewards sequence: float64 - name: discounted_rtgs_discounted_rewards sequence: float64 - name: undiscounted_rtgs_discounted_rewards sequence: float64 - name: discounted_rtgs_undiscounted_rewards sequence: float64 - name: undiscounted_rtgs_undiscounted_rewards sequence: float64 - name: discounted_rtgs_normalised_discounted_rewards sequence: float64 - name: undiscounted_rtgs_normalised_discounted_rewards sequence: float64 - name: discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_dense_rewards_subgoal_success sequence: float64 - name: normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_discounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_undiscounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_undiscounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: dense_rewards_action_failure sequence: float64 - name: dense_rtgs_action_failure sequence: float64 - name: dense_rewards sequence: float64 - name: dense_rtgs sequence: float64 - name: subgoal_successes sequence: int64 - name: subgoals dtype: int64 - name: action_failures sequence: int64 - name: max_steps dtype: int64 - name: episode_len dtype: int64 splits: - name: train num_bytes: 4123800963 num_examples: 1638 - name: val num_bytes: 1032204125 num_examples: 410 download_size: 30168478 dataset_size: 5156005088 - config_name: gotosimilardistsuniquegoalstrictlyonlycolor_1024x1_preprocessed_sentence-transformer-pretrained_cnn-scratch features: - name: task_id dtype: int64 - name: timesteps sequence: int64 - name: actions sequence: int64 - name: img_obss sequence: array3_d: shape: - 3 - 56 - 56 dtype: float64 - name: terminations sequence: int64 - name: discounted_rewards sequence: float64 - name: undiscounted_rewards sequence: float64 - name: normalised_discounted_rewards sequence: float64 - name: discounted_rtgs_discounted_rewards sequence: float64 - name: undiscounted_rtgs_discounted_rewards sequence: float64 - name: discounted_rtgs_undiscounted_rewards sequence: float64 - name: undiscounted_rtgs_undiscounted_rewards sequence: float64 - name: discounted_rtgs_normalised_discounted_rewards sequence: float64 - name: undiscounted_rtgs_normalised_discounted_rewards sequence: float64 - name: discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_dense_rewards_subgoal_success sequence: float64 - name: normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_discounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_undiscounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_undiscounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: dense_rewards_action_failure sequence: float64 - name: dense_rtgs_action_failure sequence: float64 - name: dense_rewards sequence: float64 - name: dense_rtgs sequence: float64 - name: mission_embeddings sequence: sequence: float32 - name: mission_token_mask sequence: bool - name: rule_feedback_embeddings sequence: sequence: sequence: float32 - name: rule_feedback_token_mask sequence: sequence: bool - name: rule_feedback_available sequence: bool - name: task_feedback_embeddings sequence: sequence: sequence: float32 - name: task_feedback_token_mask sequence: sequence: bool - name: task_feedback_available sequence: bool - name: instruction_embeddings sequence: sequence: sequence: float32 - name: instruction_token_mask sequence: sequence: bool - name: instructions_available sequence: bool splits: - name: train num_bytes: 2310881468 num_examples: 819 - name: val num_bytes: 578425764 num_examples: 205 download_size: 63746197 dataset_size: 2889307232 - config_name: gotosimilardistsuniquegoalstrictlyonlycolor_1024x1_raw features: - name: task_id dtype: int64 - name: episode_name dtype: string - name: mission dtype: string - name: timesteps sequence: int64 - name: actions sequence: int64 - name: img_obss sequence: array3_d: shape: - 3 - 56 - 56 dtype: float64 - name: task_feedback sequence: string - name: rule_feedback sequence: string - name: instructions sequence: string - name: terminations sequence: int64 - name: discounted_rewards sequence: float64 - name: undiscounted_rewards sequence: float64 - name: normalised_discounted_rewards sequence: float64 - name: discounted_rtgs_discounted_rewards sequence: float64 - name: undiscounted_rtgs_discounted_rewards sequence: float64 - name: discounted_rtgs_undiscounted_rewards sequence: float64 - name: undiscounted_rtgs_undiscounted_rewards sequence: float64 - name: discounted_rtgs_normalised_discounted_rewards sequence: float64 - name: undiscounted_rtgs_normalised_discounted_rewards sequence: float64 - name: discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_dense_rewards_subgoal_success sequence: float64 - name: normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_discounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_undiscounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_undiscounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: dense_rewards_action_failure sequence: float64 - name: dense_rtgs_action_failure sequence: float64 - name: dense_rewards sequence: float64 - name: dense_rtgs sequence: float64 - name: subgoal_successes sequence: int64 - name: subgoals dtype: int64 - name: action_failures sequence: int64 - name: max_steps dtype: int64 - name: episode_len dtype: int64 splits: - name: train num_bytes: 2061844400 num_examples: 819 - name: val num_bytes: 516092255 num_examples: 205 download_size: 15139257 dataset_size: 2577936655 - config_name: gotosimilardistsuniquegoalstrictlyonlycolor_2048x1_preprocessed_sentence-transformer-pretrained_cnn-scratch features: - name: task_id dtype: int64 - name: timesteps sequence: int64 - name: actions sequence: int64 - name: img_obss sequence: array3_d: shape: - 3 - 56 - 56 dtype: float64 - name: terminations sequence: int64 - name: discounted_rewards sequence: float64 - name: undiscounted_rewards sequence: float64 - name: normalised_discounted_rewards sequence: float64 - name: discounted_rtgs_discounted_rewards sequence: float64 - name: undiscounted_rtgs_discounted_rewards sequence: float64 - name: discounted_rtgs_undiscounted_rewards sequence: float64 - name: undiscounted_rtgs_undiscounted_rewards sequence: float64 - name: discounted_rtgs_normalised_discounted_rewards sequence: float64 - name: undiscounted_rtgs_normalised_discounted_rewards sequence: float64 - name: discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_dense_rewards_subgoal_success sequence: float64 - name: normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_discounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_undiscounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_undiscounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: dense_rewards_action_failure sequence: float64 - name: dense_rtgs_action_failure sequence: float64 - name: dense_rewards sequence: float64 - name: dense_rtgs sequence: float64 - name: mission_embeddings sequence: sequence: float32 - name: mission_token_mask sequence: bool - name: rule_feedback_embeddings sequence: sequence: sequence: float32 - name: rule_feedback_token_mask sequence: sequence: bool - name: rule_feedback_available sequence: bool - name: task_feedback_embeddings sequence: sequence: sequence: float32 - name: task_feedback_token_mask sequence: sequence: bool - name: task_feedback_available sequence: bool - name: instruction_embeddings sequence: sequence: sequence: float32 - name: instruction_token_mask sequence: sequence: bool - name: instructions_available sequence: bool splits: - name: train num_bytes: 4621762712 num_examples: 1638 - name: val num_bytes: 1156851528 num_examples: 410 download_size: 127449623 dataset_size: 5778614240 - config_name: gotosimilardistsuniquegoalstrictlyonlycolor_2048x1_raw features: - name: task_id dtype: int64 - name: episode_name dtype: string - name: mission dtype: string - name: timesteps sequence: int64 - name: actions sequence: int64 - name: img_obss sequence: array3_d: shape: - 3 - 56 - 56 dtype: float64 - name: task_feedback sequence: string - name: rule_feedback sequence: string - name: instructions sequence: string - name: terminations sequence: int64 - name: discounted_rewards sequence: float64 - name: undiscounted_rewards sequence: float64 - name: normalised_discounted_rewards sequence: float64 - name: discounted_rtgs_discounted_rewards sequence: float64 - name: undiscounted_rtgs_discounted_rewards sequence: float64 - name: discounted_rtgs_undiscounted_rewards sequence: float64 - name: undiscounted_rtgs_undiscounted_rewards sequence: float64 - name: discounted_rtgs_normalised_discounted_rewards sequence: float64 - name: undiscounted_rtgs_normalised_discounted_rewards sequence: float64 - name: discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_dense_rewards_subgoal_success sequence: float64 - name: normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_discounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_undiscounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_undiscounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: dense_rewards_action_failure sequence: float64 - name: dense_rtgs_action_failure sequence: float64 - name: dense_rewards sequence: float64 - name: dense_rtgs sequence: float64 - name: subgoal_successes sequence: int64 - name: subgoals dtype: int64 - name: action_failures sequence: int64 - name: max_steps dtype: int64 - name: episode_len dtype: int64 splits: - name: train num_bytes: 4123697795 num_examples: 1638 - name: val num_bytes: 1032188058 num_examples: 410 download_size: 30216041 dataset_size: 5155885853 - config_name: pickup_1024x1_preprocessed_sentence-transformer-pretrained_cnn-scratch features: - name: task_id dtype: int64 - name: timesteps sequence: int64 - name: actions sequence: int64 - name: img_obss sequence: array3_d: shape: - 3 - 56 - 56 dtype: float64 - name: terminations sequence: int64 - name: discounted_rewards sequence: float64 - name: undiscounted_rewards sequence: float64 - name: normalised_discounted_rewards sequence: float64 - name: discounted_rtgs_discounted_rewards sequence: float64 - name: undiscounted_rtgs_discounted_rewards sequence: float64 - name: discounted_rtgs_undiscounted_rewards sequence: float64 - name: undiscounted_rtgs_undiscounted_rewards sequence: float64 - name: discounted_rtgs_normalised_discounted_rewards sequence: float64 - name: undiscounted_rtgs_normalised_discounted_rewards sequence: float64 - name: discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_dense_rewards_subgoal_success sequence: float64 - name: normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_discounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_undiscounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_undiscounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: dense_rewards_action_failure sequence: float64 - name: dense_rtgs_action_failure sequence: float64 - name: dense_rewards sequence: float64 - name: dense_rtgs sequence: float64 - name: mission_embeddings sequence: sequence: float32 - name: mission_token_mask sequence: bool - name: rule_feedback_embeddings sequence: sequence: sequence: float32 - name: rule_feedback_token_mask sequence: sequence: bool - name: rule_feedback_available sequence: bool - name: task_feedback_embeddings sequence: sequence: sequence: float32 - name: task_feedback_token_mask sequence: sequence: bool - name: task_feedback_available sequence: bool - name: instruction_embeddings sequence: sequence: sequence: float32 - name: instruction_token_mask sequence: sequence: bool - name: instructions_available sequence: bool splits: - name: train num_bytes: 2310881468 num_examples: 819 - name: val num_bytes: 578425764 num_examples: 205 download_size: 69405695 dataset_size: 2889307232 - config_name: pickup_2048x1_preprocessed_sentence-transformer-pretrained_cnn-scratch features: - name: task_id dtype: int64 - name: timesteps sequence: int64 - name: actions sequence: int64 - name: img_obss sequence: array3_d: shape: - 3 - 56 - 56 dtype: float64 - name: terminations sequence: int64 - name: discounted_rewards sequence: float64 - name: undiscounted_rewards sequence: float64 - name: normalised_discounted_rewards sequence: float64 - name: discounted_rtgs_discounted_rewards sequence: float64 - name: undiscounted_rtgs_discounted_rewards sequence: float64 - name: discounted_rtgs_undiscounted_rewards sequence: float64 - name: undiscounted_rtgs_undiscounted_rewards sequence: float64 - name: discounted_rtgs_normalised_discounted_rewards sequence: float64 - name: undiscounted_rtgs_normalised_discounted_rewards sequence: float64 - name: discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_dense_rewards_subgoal_success sequence: float64 - name: normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_discounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_undiscounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_undiscounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: dense_rewards_action_failure sequence: float64 - name: dense_rtgs_action_failure sequence: float64 - name: dense_rewards sequence: float64 - name: dense_rtgs sequence: float64 - name: mission_embeddings sequence: sequence: float32 - name: mission_token_mask sequence: bool - name: rule_feedback_embeddings sequence: sequence: sequence: float32 - name: rule_feedback_token_mask sequence: sequence: bool - name: rule_feedback_available sequence: bool - name: task_feedback_embeddings sequence: sequence: sequence: float32 - name: task_feedback_token_mask sequence: sequence: bool - name: task_feedback_available sequence: bool - name: instruction_embeddings sequence: sequence: sequence: float32 - name: instruction_token_mask sequence: sequence: bool - name: instructions_available sequence: bool splits: - name: train num_bytes: 4621762712 num_examples: 1638 - name: val num_bytes: 1156851528 num_examples: 410 download_size: 137890849 dataset_size: 5778614240 - config_name: pickup_2048x1_raw features: - name: task_id dtype: int64 - name: episode_name dtype: string - name: mission dtype: string - name: timesteps sequence: int64 - name: actions sequence: int64 - name: img_obss sequence: array3_d: shape: - 3 - 56 - 56 dtype: float64 - name: task_feedback sequence: string - name: rule_feedback sequence: string - name: instructions sequence: string - name: terminations sequence: int64 - name: discounted_rewards sequence: float64 - name: undiscounted_rewards sequence: float64 - name: normalised_discounted_rewards sequence: float64 - name: discounted_rtgs_discounted_rewards sequence: float64 - name: undiscounted_rtgs_discounted_rewards sequence: float64 - name: discounted_rtgs_undiscounted_rewards sequence: float64 - name: undiscounted_rtgs_undiscounted_rewards sequence: float64 - name: discounted_rtgs_normalised_discounted_rewards sequence: float64 - name: undiscounted_rtgs_normalised_discounted_rewards sequence: float64 - name: discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_dense_rewards_subgoal_success sequence: float64 - name: normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_discounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_undiscounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_undiscounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: dense_rewards_action_failure sequence: float64 - name: dense_rtgs_action_failure sequence: float64 - name: dense_rewards sequence: float64 - name: dense_rtgs sequence: float64 - name: subgoal_successes sequence: int64 - name: subgoals dtype: int64 - name: action_failures sequence: int64 - name: max_steps dtype: int64 - name: episode_len dtype: int64 splits: - name: train num_bytes: 4123694218 num_examples: 1638 - name: val num_bytes: 1032171677 num_examples: 410 download_size: 29228601 dataset_size: 5155865895 - config_name: pickupeasyseqafter_1024x1_preprocessed_sentence-transformer-pretrained_cnn-scratch features: - name: task_id dtype: int64 - name: timesteps sequence: int64 - name: actions sequence: int64 - name: img_obss sequence: array3_d: shape: - 3 - 56 - 56 dtype: float64 - name: terminations sequence: int64 - name: discounted_rewards sequence: float64 - name: undiscounted_rewards sequence: float64 - name: normalised_discounted_rewards sequence: float64 - name: discounted_rtgs_discounted_rewards sequence: float64 - name: undiscounted_rtgs_discounted_rewards sequence: float64 - name: discounted_rtgs_undiscounted_rewards sequence: float64 - name: undiscounted_rtgs_undiscounted_rewards sequence: float64 - name: discounted_rtgs_normalised_discounted_rewards sequence: float64 - name: undiscounted_rtgs_normalised_discounted_rewards sequence: float64 - name: discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_dense_rewards_subgoal_success sequence: float64 - name: normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_discounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_undiscounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_undiscounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: dense_rewards_action_failure sequence: float64 - name: dense_rtgs_action_failure sequence: float64 - name: dense_rewards sequence: float64 - name: dense_rtgs sequence: float64 - name: mission_embeddings sequence: sequence: float32 - name: mission_token_mask sequence: bool - name: rule_feedback_embeddings sequence: sequence: sequence: float32 - name: rule_feedback_token_mask sequence: sequence: bool - name: rule_feedback_available sequence: bool - name: task_feedback_embeddings sequence: sequence: sequence: float32 - name: task_feedback_token_mask sequence: sequence: bool - name: task_feedback_available sequence: bool - name: instruction_embeddings sequence: sequence: sequence: float32 - name: instruction_token_mask sequence: sequence: bool - name: instructions_available sequence: bool splits: - name: train num_bytes: 6787456844 num_examples: 819 - name: val num_bytes: 1698936084 num_examples: 205 download_size: 162993536 dataset_size: 8486392928 - config_name: pickupeasyseqafter_1024x1_raw features: - name: task_id dtype: int64 - name: episode_name dtype: string - name: mission dtype: string - name: timesteps sequence: int64 - name: actions sequence: int64 - name: img_obss sequence: array3_d: shape: - 3 - 56 - 56 dtype: float64 - name: task_feedback sequence: string - name: rule_feedback sequence: string - name: instructions sequence: string - name: terminations sequence: int64 - name: discounted_rewards sequence: float64 - name: undiscounted_rewards sequence: float64 - name: normalised_discounted_rewards sequence: float64 - name: discounted_rtgs_discounted_rewards sequence: float64 - name: undiscounted_rtgs_discounted_rewards sequence: float64 - name: discounted_rtgs_undiscounted_rewards sequence: float64 - name: undiscounted_rtgs_undiscounted_rewards sequence: float64 - name: discounted_rtgs_normalised_discounted_rewards sequence: float64 - name: undiscounted_rtgs_normalised_discounted_rewards sequence: float64 - name: discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_dense_rewards_subgoal_success sequence: float64 - name: normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_discounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_undiscounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_undiscounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: dense_rewards_action_failure sequence: float64 - name: dense_rtgs_action_failure sequence: float64 - name: dense_rewards sequence: float64 - name: dense_rtgs sequence: float64 - name: subgoal_successes sequence: int64 - name: subgoals dtype: int64 - name: action_failures sequence: int64 - name: max_steps dtype: int64 - name: episode_len dtype: int64 splits: - name: train num_bytes: 6060130618 num_examples: 819 - name: val num_bytes: 1516890028 num_examples: 205 download_size: 48302543 dataset_size: 7577020646 - config_name: pickupeasyseqafter_2048x1_preprocessed_sentence-transformer-pretrained_cnn-scratch features: - name: task_id dtype: int64 - name: timesteps sequence: int64 - name: actions sequence: int64 - name: img_obss sequence: array3_d: shape: - 3 - 56 - 56 dtype: float64 - name: terminations sequence: int64 - name: discounted_rewards sequence: float64 - name: undiscounted_rewards sequence: float64 - name: normalised_discounted_rewards sequence: float64 - name: discounted_rtgs_discounted_rewards sequence: float64 - name: undiscounted_rtgs_discounted_rewards sequence: float64 - name: discounted_rtgs_undiscounted_rewards sequence: float64 - name: undiscounted_rtgs_undiscounted_rewards sequence: float64 - name: discounted_rtgs_normalised_discounted_rewards sequence: float64 - name: undiscounted_rtgs_normalised_discounted_rewards sequence: float64 - name: discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_dense_rewards_subgoal_success sequence: float64 - name: normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_discounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_undiscounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_undiscounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: dense_rewards_action_failure sequence: float64 - name: dense_rtgs_action_failure sequence: float64 - name: dense_rewards sequence: float64 - name: dense_rtgs sequence: float64 - name: mission_embeddings sequence: sequence: float32 - name: mission_token_mask sequence: bool - name: rule_feedback_embeddings sequence: sequence: sequence: float32 - name: rule_feedback_token_mask sequence: sequence: bool - name: rule_feedback_available sequence: bool - name: task_feedback_embeddings sequence: sequence: sequence: float32 - name: task_feedback_token_mask sequence: sequence: bool - name: task_feedback_available sequence: bool - name: instruction_embeddings sequence: sequence: sequence: float32 - name: instruction_token_mask sequence: sequence: bool - name: instructions_available sequence: bool splits: - name: train num_bytes: 13574913464 num_examples: 1638 - name: val num_bytes: 3397872168 num_examples: 410 download_size: 327162321 dataset_size: 16972785632 - config_name: pickupeasyseqafter_2048x1_raw features: - name: task_id dtype: int64 - name: episode_name dtype: string - name: mission dtype: string - name: timesteps sequence: int64 - name: actions sequence: int64 - name: img_obss sequence: array3_d: shape: - 3 - 56 - 56 dtype: float64 - name: task_feedback sequence: string - name: rule_feedback sequence: string - name: instructions sequence: string - name: terminations sequence: int64 - name: discounted_rewards sequence: float64 - name: undiscounted_rewards sequence: float64 - name: normalised_discounted_rewards sequence: float64 - name: discounted_rtgs_discounted_rewards sequence: float64 - name: undiscounted_rtgs_discounted_rewards sequence: float64 - name: discounted_rtgs_undiscounted_rewards sequence: float64 - name: undiscounted_rtgs_undiscounted_rewards sequence: float64 - name: discounted_rtgs_normalised_discounted_rewards sequence: float64 - name: undiscounted_rtgs_normalised_discounted_rewards sequence: float64 - name: discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_dense_rewards_subgoal_success sequence: float64 - name: normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_discounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_undiscounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_undiscounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: dense_rewards_action_failure sequence: float64 - name: dense_rtgs_action_failure sequence: float64 - name: dense_rewards sequence: float64 - name: dense_rtgs sequence: float64 - name: subgoal_successes sequence: int64 - name: subgoals dtype: int64 - name: action_failures sequence: int64 - name: max_steps dtype: int64 - name: episode_len dtype: int64 splits: - name: train num_bytes: 12120269421 num_examples: 1638 - name: val num_bytes: 3033775668 num_examples: 410 download_size: 96530863 dataset_size: 15154045089 - config_name: pickupeasyseqand_1024x1_preprocessed_sentence-transformer-pretrained_cnn-scratch features: - name: task_id dtype: int64 - name: timesteps sequence: int64 - name: actions sequence: int64 - name: img_obss sequence: array3_d: shape: - 3 - 56 - 56 dtype: float64 - name: terminations sequence: int64 - name: discounted_rewards sequence: float64 - name: undiscounted_rewards sequence: float64 - name: normalised_discounted_rewards sequence: float64 - name: discounted_rtgs_discounted_rewards sequence: float64 - name: undiscounted_rtgs_discounted_rewards sequence: float64 - name: discounted_rtgs_undiscounted_rewards sequence: float64 - name: undiscounted_rtgs_undiscounted_rewards sequence: float64 - name: discounted_rtgs_normalised_discounted_rewards sequence: float64 - name: undiscounted_rtgs_normalised_discounted_rewards sequence: float64 - name: discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_dense_rewards_subgoal_success sequence: float64 - name: normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_discounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_undiscounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_undiscounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: dense_rewards_action_failure sequence: float64 - name: dense_rtgs_action_failure sequence: float64 - name: dense_rewards sequence: float64 - name: dense_rtgs sequence: float64 - name: mission_embeddings sequence: sequence: float32 - name: mission_token_mask sequence: bool - name: rule_feedback_embeddings sequence: sequence: sequence: float32 - name: rule_feedback_token_mask sequence: sequence: bool - name: rule_feedback_available sequence: bool - name: task_feedback_embeddings sequence: sequence: sequence: float32 - name: task_feedback_token_mask sequence: sequence: bool - name: task_feedback_available sequence: bool - name: instruction_embeddings sequence: sequence: sequence: float32 - name: instruction_token_mask sequence: sequence: bool - name: instructions_available sequence: bool splits: - name: train num_bytes: 6787456844 num_examples: 819 - name: val num_bytes: 1698936084 num_examples: 205 download_size: 165065298 dataset_size: 8486392928 - config_name: pickupeasyseqand_1024x1_raw features: - name: task_id dtype: int64 - name: episode_name dtype: string - name: mission dtype: string - name: timesteps sequence: int64 - name: actions sequence: int64 - name: img_obss sequence: array3_d: shape: - 3 - 56 - 56 dtype: float64 - name: task_feedback sequence: string - name: rule_feedback sequence: string - name: instructions sequence: string - name: terminations sequence: int64 - name: discounted_rewards sequence: float64 - name: undiscounted_rewards sequence: float64 - name: normalised_discounted_rewards sequence: float64 - name: discounted_rtgs_discounted_rewards sequence: float64 - name: undiscounted_rtgs_discounted_rewards sequence: float64 - name: discounted_rtgs_undiscounted_rewards sequence: float64 - name: undiscounted_rtgs_undiscounted_rewards sequence: float64 - name: discounted_rtgs_normalised_discounted_rewards sequence: float64 - name: undiscounted_rtgs_normalised_discounted_rewards sequence: float64 - name: discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_dense_rewards_subgoal_success sequence: float64 - name: normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_discounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_undiscounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_undiscounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: dense_rewards_action_failure sequence: float64 - name: dense_rtgs_action_failure sequence: float64 - name: dense_rewards sequence: float64 - name: dense_rtgs sequence: float64 - name: subgoal_successes sequence: int64 - name: subgoals dtype: int64 - name: action_failures sequence: int64 - name: max_steps dtype: int64 - name: episode_len dtype: int64 splits: - name: train num_bytes: 6060138577 num_examples: 819 - name: val num_bytes: 1516881982 num_examples: 205 download_size: 48173921 dataset_size: 7577020559 - config_name: pickupeasyseqand_2048x1_preprocessed_sentence-transformer-pretrained_cnn-scratch features: - name: task_id dtype: int64 - name: timesteps sequence: int64 - name: actions sequence: int64 - name: img_obss sequence: array3_d: shape: - 3 - 56 - 56 dtype: float64 - name: terminations sequence: int64 - name: discounted_rewards sequence: float64 - name: undiscounted_rewards sequence: float64 - name: normalised_discounted_rewards sequence: float64 - name: discounted_rtgs_discounted_rewards sequence: float64 - name: undiscounted_rtgs_discounted_rewards sequence: float64 - name: discounted_rtgs_undiscounted_rewards sequence: float64 - name: undiscounted_rtgs_undiscounted_rewards sequence: float64 - name: discounted_rtgs_normalised_discounted_rewards sequence: float64 - name: undiscounted_rtgs_normalised_discounted_rewards sequence: float64 - name: discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_dense_rewards_subgoal_success sequence: float64 - name: normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_discounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_undiscounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_undiscounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: dense_rewards_action_failure sequence: float64 - name: dense_rtgs_action_failure sequence: float64 - name: dense_rewards sequence: float64 - name: dense_rtgs sequence: float64 - name: mission_embeddings sequence: sequence: float32 - name: mission_token_mask sequence: bool - name: rule_feedback_embeddings sequence: sequence: sequence: float32 - name: rule_feedback_token_mask sequence: sequence: bool - name: rule_feedback_available sequence: bool - name: task_feedback_embeddings sequence: sequence: sequence: float32 - name: task_feedback_token_mask sequence: sequence: bool - name: task_feedback_available sequence: bool - name: instruction_embeddings sequence: sequence: sequence: float32 - name: instruction_token_mask sequence: sequence: bool - name: instructions_available sequence: bool splits: - name: train num_bytes: 13574913464 num_examples: 1638 - name: val num_bytes: 3397872168 num_examples: 410 download_size: 331787119 dataset_size: 16972785632 - config_name: pickupeasyseqand_2048x1_raw features: - name: task_id dtype: int64 - name: episode_name dtype: string - name: mission dtype: string - name: timesteps sequence: int64 - name: actions sequence: int64 - name: img_obss sequence: array3_d: shape: - 3 - 56 - 56 dtype: float64 - name: task_feedback sequence: string - name: rule_feedback sequence: string - name: instructions sequence: string - name: terminations sequence: int64 - name: discounted_rewards sequence: float64 - name: undiscounted_rewards sequence: float64 - name: normalised_discounted_rewards sequence: float64 - name: discounted_rtgs_discounted_rewards sequence: float64 - name: undiscounted_rtgs_discounted_rewards sequence: float64 - name: discounted_rtgs_undiscounted_rewards sequence: float64 - name: undiscounted_rtgs_undiscounted_rewards sequence: float64 - name: discounted_rtgs_normalised_discounted_rewards sequence: float64 - name: undiscounted_rtgs_normalised_discounted_rewards sequence: float64 - name: discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_dense_rewards_subgoal_success sequence: float64 - name: normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_discounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_undiscounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_undiscounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: dense_rewards_action_failure sequence: float64 - name: dense_rtgs_action_failure sequence: float64 - name: dense_rewards sequence: float64 - name: dense_rtgs sequence: float64 - name: subgoal_successes sequence: int64 - name: subgoals dtype: int64 - name: action_failures sequence: int64 - name: max_steps dtype: int64 - name: episode_len dtype: int64 splits: - name: train num_bytes: 12120288639 num_examples: 1638 - name: val num_bytes: 3033789745 num_examples: 410 download_size: 96303174 dataset_size: 15154078384 - config_name: pickupeasyseqbefore_1024x1_preprocessed_sentence-transformer-pretrained_cnn-scratch features: - name: task_id dtype: int64 - name: timesteps sequence: int64 - name: actions sequence: int64 - name: img_obss sequence: array3_d: shape: - 3 - 56 - 56 dtype: float64 - name: terminations sequence: int64 - name: discounted_rewards sequence: float64 - name: undiscounted_rewards sequence: float64 - name: normalised_discounted_rewards sequence: float64 - name: discounted_rtgs_discounted_rewards sequence: float64 - name: undiscounted_rtgs_discounted_rewards sequence: float64 - name: discounted_rtgs_undiscounted_rewards sequence: float64 - name: undiscounted_rtgs_undiscounted_rewards sequence: float64 - name: discounted_rtgs_normalised_discounted_rewards sequence: float64 - name: undiscounted_rtgs_normalised_discounted_rewards sequence: float64 - name: discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_dense_rewards_subgoal_success sequence: float64 - name: normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_discounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_undiscounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_undiscounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: dense_rewards_action_failure sequence: float64 - name: dense_rtgs_action_failure sequence: float64 - name: dense_rewards sequence: float64 - name: dense_rtgs sequence: float64 - name: mission_embeddings sequence: sequence: float32 - name: mission_token_mask sequence: bool - name: rule_feedback_embeddings sequence: sequence: sequence: float32 - name: rule_feedback_token_mask sequence: sequence: bool - name: rule_feedback_available sequence: bool - name: task_feedback_embeddings sequence: sequence: sequence: float32 - name: task_feedback_token_mask sequence: sequence: bool - name: task_feedback_available sequence: bool - name: instruction_embeddings sequence: sequence: sequence: float32 - name: instruction_token_mask sequence: sequence: bool - name: instructions_available sequence: bool splits: - name: train num_bytes: 6787456844 num_examples: 819 - name: val num_bytes: 1698936084 num_examples: 205 download_size: 162573042 dataset_size: 8486392928 - config_name: pickupeasyseqbefore_2048x1_preprocessed_sentence-transformer-pretrained_cnn-scratch features: - name: task_id dtype: int64 - name: timesteps sequence: int64 - name: actions sequence: int64 - name: img_obss sequence: array3_d: shape: - 3 - 56 - 56 dtype: float64 - name: terminations sequence: int64 - name: discounted_rewards sequence: float64 - name: undiscounted_rewards sequence: float64 - name: normalised_discounted_rewards sequence: float64 - name: discounted_rtgs_discounted_rewards sequence: float64 - name: undiscounted_rtgs_discounted_rewards sequence: float64 - name: discounted_rtgs_undiscounted_rewards sequence: float64 - name: undiscounted_rtgs_undiscounted_rewards sequence: float64 - name: discounted_rtgs_normalised_discounted_rewards sequence: float64 - name: undiscounted_rtgs_normalised_discounted_rewards sequence: float64 - name: discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_dense_rewards_subgoal_success sequence: float64 - name: normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_discounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_undiscounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_undiscounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: dense_rewards_action_failure sequence: float64 - name: dense_rtgs_action_failure sequence: float64 - name: dense_rewards sequence: float64 - name: dense_rtgs sequence: float64 - name: mission_embeddings sequence: sequence: float32 - name: mission_token_mask sequence: bool - name: rule_feedback_embeddings sequence: sequence: sequence: float32 - name: rule_feedback_token_mask sequence: sequence: bool - name: rule_feedback_available sequence: bool - name: task_feedback_embeddings sequence: sequence: sequence: float32 - name: task_feedback_token_mask sequence: sequence: bool - name: task_feedback_available sequence: bool - name: instruction_embeddings sequence: sequence: sequence: float32 - name: instruction_token_mask sequence: sequence: bool - name: instructions_available sequence: bool splits: - name: train num_bytes: 13574913464 num_examples: 1638 - name: val num_bytes: 3397872168 num_examples: 410 download_size: 326257748 dataset_size: 16972785632 - config_name: pickupeasyseqbefore_2048x1_raw features: - name: task_id dtype: int64 - name: episode_name dtype: string - name: mission dtype: string - name: timesteps sequence: int64 - name: actions sequence: int64 - name: img_obss sequence: array3_d: shape: - 3 - 56 - 56 dtype: float64 - name: task_feedback sequence: string - name: rule_feedback sequence: string - name: instructions sequence: string - name: terminations sequence: int64 - name: discounted_rewards sequence: float64 - name: undiscounted_rewards sequence: float64 - name: normalised_discounted_rewards sequence: float64 - name: discounted_rtgs_discounted_rewards sequence: float64 - name: undiscounted_rtgs_discounted_rewards sequence: float64 - name: discounted_rtgs_undiscounted_rewards sequence: float64 - name: undiscounted_rtgs_undiscounted_rewards sequence: float64 - name: discounted_rtgs_normalised_discounted_rewards sequence: float64 - name: undiscounted_rtgs_normalised_discounted_rewards sequence: float64 - name: discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_dense_rewards_subgoal_success sequence: float64 - name: normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_discounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_undiscounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_undiscounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: dense_rewards_action_failure sequence: float64 - name: dense_rtgs_action_failure sequence: float64 - name: dense_rewards sequence: float64 - name: dense_rtgs sequence: float64 - name: subgoal_successes sequence: int64 - name: subgoals dtype: int64 - name: action_failures sequence: int64 - name: max_steps dtype: int64 - name: episode_len dtype: int64 splits: - name: train num_bytes: 12120256633 num_examples: 1638 - name: val num_bytes: 3033774381 num_examples: 410 download_size: 96435925 dataset_size: 15154031014 - config_name: pickupmaze1x2_1024x1_preprocessed_sentence-transformer-pretrained_cnn-scratch features: - name: task_id dtype: int64 - name: timesteps sequence: int64 - name: actions sequence: int64 - name: img_obss sequence: array3_d: shape: - 3 - 56 - 56 dtype: float64 - name: terminations sequence: int64 - name: discounted_rewards sequence: float64 - name: undiscounted_rewards sequence: float64 - name: normalised_discounted_rewards sequence: float64 - name: discounted_rtgs_discounted_rewards sequence: float64 - name: undiscounted_rtgs_discounted_rewards sequence: float64 - name: discounted_rtgs_undiscounted_rewards sequence: float64 - name: undiscounted_rtgs_undiscounted_rewards sequence: float64 - name: discounted_rtgs_normalised_discounted_rewards sequence: float64 - name: undiscounted_rtgs_normalised_discounted_rewards sequence: float64 - name: discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_dense_rewards_subgoal_success sequence: float64 - name: normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_discounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_undiscounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_undiscounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: dense_rewards_action_failure sequence: float64 - name: dense_rtgs_action_failure sequence: float64 - name: dense_rewards sequence: float64 - name: dense_rtgs sequence: float64 - name: mission_embeddings sequence: sequence: float32 - name: mission_token_mask sequence: bool - name: rule_feedback_embeddings sequence: sequence: sequence: float32 - name: rule_feedback_token_mask sequence: sequence: bool - name: rule_feedback_available sequence: bool - name: task_feedback_embeddings sequence: sequence: sequence: float32 - name: task_feedback_token_mask sequence: sequence: bool - name: task_feedback_available sequence: bool - name: instruction_embeddings sequence: sequence: sequence: float32 - name: instruction_token_mask sequence: sequence: bool - name: instructions_available sequence: bool splits: - name: train num_bytes: 4549169156 num_examples: 819 - name: val num_bytes: 1138680924 num_examples: 205 download_size: 97385164 dataset_size: 5687850080 - config_name: pickupmaze1x2_1024x1_raw features: - name: task_id dtype: int64 - name: episode_name dtype: string - name: mission dtype: string - name: timesteps sequence: int64 - name: actions sequence: int64 - name: img_obss sequence: array3_d: shape: - 3 - 56 - 56 dtype: float64 - name: task_feedback sequence: string - name: rule_feedback sequence: string - name: instructions sequence: string - name: terminations sequence: int64 - name: discounted_rewards sequence: float64 - name: undiscounted_rewards sequence: float64 - name: normalised_discounted_rewards sequence: float64 - name: discounted_rtgs_discounted_rewards sequence: float64 - name: undiscounted_rtgs_discounted_rewards sequence: float64 - name: discounted_rtgs_undiscounted_rewards sequence: float64 - name: undiscounted_rtgs_undiscounted_rewards sequence: float64 - name: discounted_rtgs_normalised_discounted_rewards sequence: float64 - name: undiscounted_rtgs_normalised_discounted_rewards sequence: float64 - name: discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_dense_rewards_subgoal_success sequence: float64 - name: normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_discounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_undiscounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_undiscounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: dense_rewards_action_failure sequence: float64 - name: dense_rtgs_action_failure sequence: float64 - name: dense_rewards sequence: float64 - name: dense_rtgs sequence: float64 - name: subgoal_successes sequence: int64 - name: subgoals dtype: int64 - name: action_failures sequence: int64 - name: max_steps dtype: int64 - name: episode_len dtype: int64 splits: - name: train num_bytes: 4060871908 num_examples: 819 - name: val num_bytes: 1016460783 num_examples: 205 download_size: 26411715 dataset_size: 5077332691 - config_name: pickupmaze1x2_2048x1_preprocessed_sentence-transformer-pretrained_cnn-scratch features: - name: task_id dtype: int64 - name: timesteps sequence: int64 - name: actions sequence: int64 - name: img_obss sequence: array3_d: shape: - 3 - 56 - 56 dtype: float64 - name: terminations sequence: int64 - name: discounted_rewards sequence: float64 - name: undiscounted_rewards sequence: float64 - name: normalised_discounted_rewards sequence: float64 - name: discounted_rtgs_discounted_rewards sequence: float64 - name: undiscounted_rtgs_discounted_rewards sequence: float64 - name: discounted_rtgs_undiscounted_rewards sequence: float64 - name: undiscounted_rtgs_undiscounted_rewards sequence: float64 - name: discounted_rtgs_normalised_discounted_rewards sequence: float64 - name: undiscounted_rtgs_normalised_discounted_rewards sequence: float64 - name: discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_dense_rewards_subgoal_success sequence: float64 - name: normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_discounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_undiscounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_undiscounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: dense_rewards_action_failure sequence: float64 - name: dense_rtgs_action_failure sequence: float64 - name: dense_rewards sequence: float64 - name: dense_rtgs sequence: float64 - name: mission_embeddings sequence: sequence: float32 - name: mission_token_mask sequence: bool - name: rule_feedback_embeddings sequence: sequence: sequence: float32 - name: rule_feedback_token_mask sequence: sequence: bool - name: rule_feedback_available sequence: bool - name: task_feedback_embeddings sequence: sequence: sequence: float32 - name: task_feedback_token_mask sequence: sequence: bool - name: task_feedback_available sequence: bool - name: instruction_embeddings sequence: sequence: sequence: float32 - name: instruction_token_mask sequence: sequence: bool - name: instructions_available sequence: bool splits: - name: train num_bytes: 9098338088 num_examples: 1638 - name: val num_bytes: 2277361848 num_examples: 410 download_size: 193471580 dataset_size: 11375699936 - config_name: pickupmaze1x2_2048x1_raw features: - name: task_id dtype: int64 - name: episode_name dtype: string - name: mission dtype: string - name: timesteps sequence: int64 - name: actions sequence: int64 - name: img_obss sequence: array3_d: shape: - 3 - 56 - 56 dtype: float64 - name: task_feedback sequence: string - name: rule_feedback sequence: string - name: instructions sequence: string - name: terminations sequence: int64 - name: discounted_rewards sequence: float64 - name: undiscounted_rewards sequence: float64 - name: normalised_discounted_rewards sequence: float64 - name: discounted_rtgs_discounted_rewards sequence: float64 - name: undiscounted_rtgs_discounted_rewards sequence: float64 - name: discounted_rtgs_undiscounted_rewards sequence: float64 - name: undiscounted_rtgs_undiscounted_rewards sequence: float64 - name: discounted_rtgs_normalised_discounted_rewards sequence: float64 - name: undiscounted_rtgs_normalised_discounted_rewards sequence: float64 - name: discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_dense_rewards_subgoal_success sequence: float64 - name: normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_discounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_undiscounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_undiscounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: dense_rewards_action_failure sequence: float64 - name: dense_rtgs_action_failure sequence: float64 - name: dense_rewards sequence: float64 - name: dense_rtgs sequence: float64 - name: subgoal_successes sequence: int64 - name: subgoals dtype: int64 - name: action_failures sequence: int64 - name: max_steps dtype: int64 - name: episode_len dtype: int64 splits: - name: train num_bytes: 8121746729 num_examples: 1638 - name: val num_bytes: 2032912085 num_examples: 410 download_size: 52764162 dataset_size: 10154658814 - config_name: pickupmaze1x2blocked_1024x1_preprocessed_sentence-transformer-pretrained_cnn-scratch features: - name: task_id dtype: int64 - name: timesteps sequence: int64 - name: actions sequence: int64 - name: img_obss sequence: array3_d: shape: - 3 - 56 - 56 dtype: float64 - name: terminations sequence: int64 - name: discounted_rewards sequence: float64 - name: undiscounted_rewards sequence: float64 - name: normalised_discounted_rewards sequence: float64 - name: discounted_rtgs_discounted_rewards sequence: float64 - name: undiscounted_rtgs_discounted_rewards sequence: float64 - name: discounted_rtgs_undiscounted_rewards sequence: float64 - name: undiscounted_rtgs_undiscounted_rewards sequence: float64 - name: discounted_rtgs_normalised_discounted_rewards sequence: float64 - name: undiscounted_rtgs_normalised_discounted_rewards sequence: float64 - name: discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_dense_rewards_subgoal_success sequence: float64 - name: normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_discounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_undiscounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_undiscounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: dense_rewards_action_failure sequence: float64 - name: dense_rtgs_action_failure sequence: float64 - name: dense_rewards sequence: float64 - name: dense_rtgs sequence: float64 - name: mission_embeddings sequence: sequence: float32 - name: mission_token_mask sequence: bool - name: rule_feedback_embeddings sequence: sequence: sequence: float32 - name: rule_feedback_token_mask sequence: sequence: bool - name: rule_feedback_available sequence: bool - name: task_feedback_embeddings sequence: sequence: sequence: float32 - name: task_feedback_token_mask sequence: sequence: bool - name: task_feedback_available sequence: bool - name: instruction_embeddings sequence: sequence: sequence: float32 - name: instruction_token_mask sequence: sequence: bool - name: instructions_available sequence: bool splits: - name: train num_bytes: 9025744532 num_examples: 819 - name: val num_bytes: 2259191244 num_examples: 205 download_size: 191037063 dataset_size: 11284935776 - config_name: pickupmaze1x2blocked_1024x1_raw features: - name: task_id dtype: int64 - name: episode_name dtype: string - name: mission dtype: string - name: timesteps sequence: int64 - name: actions sequence: int64 - name: img_obss sequence: array3_d: shape: - 3 - 56 - 56 dtype: float64 - name: task_feedback sequence: string - name: rule_feedback sequence: string - name: instructions sequence: string - name: terminations sequence: int64 - name: discounted_rewards sequence: float64 - name: undiscounted_rewards sequence: float64 - name: normalised_discounted_rewards sequence: float64 - name: discounted_rtgs_discounted_rewards sequence: float64 - name: undiscounted_rtgs_discounted_rewards sequence: float64 - name: discounted_rtgs_undiscounted_rewards sequence: float64 - name: undiscounted_rtgs_undiscounted_rewards sequence: float64 - name: discounted_rtgs_normalised_discounted_rewards sequence: float64 - name: undiscounted_rtgs_normalised_discounted_rewards sequence: float64 - name: discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_dense_rewards_subgoal_success sequence: float64 - name: normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_discounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_undiscounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_undiscounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: dense_rewards_action_failure sequence: float64 - name: dense_rtgs_action_failure sequence: float64 - name: dense_rewards sequence: float64 - name: dense_rtgs sequence: float64 - name: subgoal_successes sequence: int64 - name: subgoals dtype: int64 - name: action_failures sequence: int64 - name: max_steps dtype: int64 - name: episode_len dtype: int64 splits: - name: train num_bytes: 8059142458 num_examples: 819 - name: val num_bytes: 2017251934 num_examples: 205 download_size: 52014356 dataset_size: 10076394392 - config_name: pickupmaze1x2blocked_2048x1_preprocessed_sentence-transformer-pretrained_cnn-scratch features: - name: task_id dtype: int64 - name: timesteps sequence: int64 - name: actions sequence: int64 - name: img_obss sequence: array3_d: shape: - 3 - 56 - 56 dtype: float64 - name: terminations sequence: int64 - name: discounted_rewards sequence: float64 - name: undiscounted_rewards sequence: float64 - name: normalised_discounted_rewards sequence: float64 - name: discounted_rtgs_discounted_rewards sequence: float64 - name: undiscounted_rtgs_discounted_rewards sequence: float64 - name: discounted_rtgs_undiscounted_rewards sequence: float64 - name: undiscounted_rtgs_undiscounted_rewards sequence: float64 - name: discounted_rtgs_normalised_discounted_rewards sequence: float64 - name: undiscounted_rtgs_normalised_discounted_rewards sequence: float64 - name: discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_dense_rewards_subgoal_success sequence: float64 - name: normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_discounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_undiscounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_undiscounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: dense_rewards_action_failure sequence: float64 - name: dense_rtgs_action_failure sequence: float64 - name: dense_rewards sequence: float64 - name: dense_rtgs sequence: float64 - name: mission_embeddings sequence: sequence: float32 - name: mission_token_mask sequence: bool - name: rule_feedback_embeddings sequence: sequence: sequence: float32 - name: rule_feedback_token_mask sequence: sequence: bool - name: rule_feedback_available sequence: bool - name: task_feedback_embeddings sequence: sequence: sequence: float32 - name: task_feedback_token_mask sequence: sequence: bool - name: task_feedback_available sequence: bool - name: instruction_embeddings sequence: sequence: sequence: float32 - name: instruction_token_mask sequence: sequence: bool - name: instructions_available sequence: bool splits: - name: train num_bytes: 18051488840 num_examples: 1638 - name: val num_bytes: 4518382488 num_examples: 410 download_size: 380552734 dataset_size: 22569871328 - config_name: pickupmaze1x2blocked_2048x1_raw features: - name: task_id dtype: int64 - name: episode_name dtype: string - name: mission dtype: string - name: timesteps sequence: int64 - name: actions sequence: int64 - name: img_obss sequence: array3_d: shape: - 3 - 56 - 56 dtype: float64 - name: task_feedback sequence: string - name: rule_feedback sequence: string - name: instructions sequence: string - name: terminations sequence: int64 - name: discounted_rewards sequence: float64 - name: undiscounted_rewards sequence: float64 - name: normalised_discounted_rewards sequence: float64 - name: discounted_rtgs_discounted_rewards sequence: float64 - name: undiscounted_rtgs_discounted_rewards sequence: float64 - name: discounted_rtgs_undiscounted_rewards sequence: float64 - name: undiscounted_rtgs_undiscounted_rewards sequence: float64 - name: discounted_rtgs_normalised_discounted_rewards sequence: float64 - name: undiscounted_rtgs_normalised_discounted_rewards sequence: float64 - name: discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_dense_rewards_subgoal_success sequence: float64 - name: normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_discounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_undiscounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_undiscounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: dense_rewards_action_failure sequence: float64 - name: dense_rtgs_action_failure sequence: float64 - name: dense_rewards sequence: float64 - name: dense_rtgs sequence: float64 - name: subgoal_successes sequence: int64 - name: subgoals dtype: int64 - name: action_failures sequence: int64 - name: max_steps dtype: int64 - name: episode_len dtype: int64 splits: - name: train num_bytes: 16118284674 num_examples: 1638 - name: val num_bytes: 4034486223 num_examples: 410 download_size: 104159130 dataset_size: 20152770897 - config_name: pickupmaze1x2doorsclosed_1024x1_preprocessed_sentence-transformer-pretrained_cnn-scratch features: - name: task_id dtype: int64 - name: timesteps sequence: int64 - name: actions sequence: int64 - name: img_obss sequence: array3_d: shape: - 3 - 56 - 56 dtype: float64 - name: terminations sequence: int64 - name: discounted_rewards sequence: float64 - name: undiscounted_rewards sequence: float64 - name: normalised_discounted_rewards sequence: float64 - name: discounted_rtgs_discounted_rewards sequence: float64 - name: undiscounted_rtgs_discounted_rewards sequence: float64 - name: discounted_rtgs_undiscounted_rewards sequence: float64 - name: undiscounted_rtgs_undiscounted_rewards sequence: float64 - name: discounted_rtgs_normalised_discounted_rewards sequence: float64 - name: undiscounted_rtgs_normalised_discounted_rewards sequence: float64 - name: discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_dense_rewards_subgoal_success sequence: float64 - name: normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_discounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_undiscounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_undiscounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: dense_rewards_action_failure sequence: float64 - name: dense_rtgs_action_failure sequence: float64 - name: dense_rewards sequence: float64 - name: dense_rtgs sequence: float64 - name: mission_embeddings sequence: sequence: float32 - name: mission_token_mask sequence: bool - name: rule_feedback_embeddings sequence: sequence: sequence: float32 - name: rule_feedback_token_mask sequence: sequence: bool - name: rule_feedback_available sequence: bool - name: task_feedback_embeddings sequence: sequence: sequence: float32 - name: task_feedback_token_mask sequence: sequence: bool - name: task_feedback_available sequence: bool - name: instruction_embeddings sequence: sequence: sequence: float32 - name: instruction_token_mask sequence: sequence: bool - name: instructions_available sequence: bool splits: - name: train num_bytes: 4549169156 num_examples: 819 - name: val num_bytes: 1138680924 num_examples: 205 download_size: 96762806 dataset_size: 5687850080 - config_name: pickupmaze1x2doorsclosed_1024x1_raw features: - name: task_id dtype: int64 - name: episode_name dtype: string - name: mission dtype: string - name: timesteps sequence: int64 - name: actions sequence: int64 - name: img_obss sequence: array3_d: shape: - 3 - 56 - 56 dtype: float64 - name: task_feedback sequence: string - name: rule_feedback sequence: string - name: instructions sequence: string - name: terminations sequence: int64 - name: discounted_rewards sequence: float64 - name: undiscounted_rewards sequence: float64 - name: normalised_discounted_rewards sequence: float64 - name: discounted_rtgs_discounted_rewards sequence: float64 - name: undiscounted_rtgs_discounted_rewards sequence: float64 - name: discounted_rtgs_undiscounted_rewards sequence: float64 - name: undiscounted_rtgs_undiscounted_rewards sequence: float64 - name: discounted_rtgs_normalised_discounted_rewards sequence: float64 - name: undiscounted_rtgs_normalised_discounted_rewards sequence: float64 - name: discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_dense_rewards_subgoal_success sequence: float64 - name: normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_discounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_undiscounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_undiscounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: dense_rewards_action_failure sequence: float64 - name: dense_rtgs_action_failure sequence: float64 - name: dense_rewards sequence: float64 - name: dense_rtgs sequence: float64 - name: subgoal_successes sequence: int64 - name: subgoals dtype: int64 - name: action_failures sequence: int64 - name: max_steps dtype: int64 - name: episode_len dtype: int64 splits: - name: train num_bytes: 4060881135 num_examples: 819 - name: val num_bytes: 1016460956 num_examples: 205 download_size: 24658307 dataset_size: 5077342091 - config_name: pickupmaze1x2doorsclosed_2048x1_raw features: - name: task_id dtype: int64 - name: episode_name dtype: string - name: mission dtype: string - name: timesteps sequence: int64 - name: actions sequence: int64 - name: img_obss sequence: array3_d: shape: - 3 - 56 - 56 dtype: float64 - name: task_feedback sequence: string - name: rule_feedback sequence: string - name: instructions sequence: string - name: terminations sequence: int64 - name: discounted_rewards sequence: float64 - name: undiscounted_rewards sequence: float64 - name: normalised_discounted_rewards sequence: float64 - name: discounted_rtgs_discounted_rewards sequence: float64 - name: undiscounted_rtgs_discounted_rewards sequence: float64 - name: discounted_rtgs_undiscounted_rewards sequence: float64 - name: undiscounted_rtgs_undiscounted_rewards sequence: float64 - name: discounted_rtgs_normalised_discounted_rewards sequence: float64 - name: undiscounted_rtgs_normalised_discounted_rewards sequence: float64 - name: discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_dense_rewards_subgoal_success sequence: float64 - name: normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_discounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_undiscounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_undiscounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: dense_rewards_action_failure sequence: float64 - name: dense_rtgs_action_failure sequence: float64 - name: dense_rewards sequence: float64 - name: dense_rtgs sequence: float64 - name: subgoal_successes sequence: int64 - name: subgoals dtype: int64 - name: action_failures sequence: int64 - name: max_steps dtype: int64 - name: episode_len dtype: int64 splits: - name: train num_bytes: 8121763479 num_examples: 1638 - name: val num_bytes: 2032913962 num_examples: 410 download_size: 49442793 dataset_size: 10154677441 - config_name: pickupmaze1x2doorslocked_1024x1_preprocessed_sentence-transformer-pretrained_cnn-scratch features: - name: task_id dtype: int64 - name: timesteps sequence: int64 - name: actions sequence: int64 - name: img_obss sequence: array3_d: shape: - 3 - 56 - 56 dtype: float64 - name: terminations sequence: int64 - name: discounted_rewards sequence: float64 - name: undiscounted_rewards sequence: float64 - name: normalised_discounted_rewards sequence: float64 - name: discounted_rtgs_discounted_rewards sequence: float64 - name: undiscounted_rtgs_discounted_rewards sequence: float64 - name: discounted_rtgs_undiscounted_rewards sequence: float64 - name: undiscounted_rtgs_undiscounted_rewards sequence: float64 - name: discounted_rtgs_normalised_discounted_rewards sequence: float64 - name: undiscounted_rtgs_normalised_discounted_rewards sequence: float64 - name: discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_dense_rewards_subgoal_success sequence: float64 - name: normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_discounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_undiscounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_undiscounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: dense_rewards_action_failure sequence: float64 - name: dense_rtgs_action_failure sequence: float64 - name: dense_rewards sequence: float64 - name: dense_rtgs sequence: float64 - name: mission_embeddings sequence: sequence: float32 - name: mission_token_mask sequence: bool - name: rule_feedback_embeddings sequence: sequence: sequence: float32 - name: rule_feedback_token_mask sequence: sequence: bool - name: rule_feedback_available sequence: bool - name: task_feedback_embeddings sequence: sequence: sequence: float32 - name: task_feedback_token_mask sequence: sequence: bool - name: task_feedback_available sequence: bool - name: instruction_embeddings sequence: sequence: sequence: float32 - name: instruction_token_mask sequence: sequence: bool - name: instructions_available sequence: bool splits: - name: train num_bytes: 9025744532 num_examples: 819 - name: val num_bytes: 2259191244 num_examples: 205 download_size: 217315109 dataset_size: 11284935776 - config_name: pickupmaze1x2doorslocked_1024x1_raw features: - name: task_id dtype: int64 - name: episode_name dtype: string - name: mission dtype: string - name: timesteps sequence: int64 - name: actions sequence: int64 - name: img_obss sequence: array3_d: shape: - 3 - 56 - 56 dtype: float64 - name: task_feedback sequence: string - name: rule_feedback sequence: string - name: instructions sequence: string - name: terminations sequence: int64 - name: discounted_rewards sequence: float64 - name: undiscounted_rewards sequence: float64 - name: normalised_discounted_rewards sequence: float64 - name: discounted_rtgs_discounted_rewards sequence: float64 - name: undiscounted_rtgs_discounted_rewards sequence: float64 - name: discounted_rtgs_undiscounted_rewards sequence: float64 - name: undiscounted_rtgs_undiscounted_rewards sequence: float64 - name: discounted_rtgs_normalised_discounted_rewards sequence: float64 - name: undiscounted_rtgs_normalised_discounted_rewards sequence: float64 - name: discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_dense_rewards_subgoal_success sequence: float64 - name: normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_discounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_undiscounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_undiscounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: dense_rewards_action_failure sequence: float64 - name: dense_rtgs_action_failure sequence: float64 - name: dense_rewards sequence: float64 - name: dense_rtgs sequence: float64 - name: subgoal_successes sequence: int64 - name: subgoals dtype: int64 - name: action_failures sequence: int64 - name: max_steps dtype: int64 - name: episode_len dtype: int64 splits: - name: train num_bytes: 8059262119 num_examples: 819 - name: val num_bytes: 2017283343 num_examples: 205 download_size: 53679509 dataset_size: 10076545462 - config_name: pickupmaze1x2doorslocked_2048x1_raw features: - name: task_id dtype: int64 - name: episode_name dtype: string - name: mission dtype: string - name: timesteps sequence: int64 - name: actions sequence: int64 - name: img_obss sequence: array3_d: shape: - 3 - 56 - 56 dtype: float64 - name: task_feedback sequence: string - name: rule_feedback sequence: string - name: instructions sequence: string - name: terminations sequence: int64 - name: discounted_rewards sequence: float64 - name: undiscounted_rewards sequence: float64 - name: normalised_discounted_rewards sequence: float64 - name: discounted_rtgs_discounted_rewards sequence: float64 - name: undiscounted_rtgs_discounted_rewards sequence: float64 - name: discounted_rtgs_undiscounted_rewards sequence: float64 - name: undiscounted_rtgs_undiscounted_rewards sequence: float64 - name: discounted_rtgs_normalised_discounted_rewards sequence: float64 - name: undiscounted_rtgs_normalised_discounted_rewards sequence: float64 - name: discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_dense_rewards_subgoal_success sequence: float64 - name: normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_discounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_undiscounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_undiscounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: dense_rewards_action_failure sequence: float64 - name: dense_rtgs_action_failure sequence: float64 - name: dense_rewards sequence: float64 - name: dense_rtgs sequence: float64 - name: subgoal_successes sequence: int64 - name: subgoals dtype: int64 - name: action_failures sequence: int64 - name: max_steps dtype: int64 - name: episode_len dtype: int64 splits: - name: train num_bytes: 16118548008 num_examples: 1638 - name: val num_bytes: 4034562166 num_examples: 410 download_size: 107177754 dataset_size: 20153110174 - config_name: pickupmaze2x2_1024x1_preprocessed_sentence-transformer-pretrained_cnn-scratch features: - name: task_id dtype: int64 - name: timesteps sequence: int64 - name: actions sequence: int64 - name: img_obss sequence: array3_d: shape: - 3 - 56 - 56 dtype: float64 - name: terminations sequence: int64 - name: discounted_rewards sequence: float64 - name: undiscounted_rewards sequence: float64 - name: normalised_discounted_rewards sequence: float64 - name: discounted_rtgs_discounted_rewards sequence: float64 - name: undiscounted_rtgs_discounted_rewards sequence: float64 - name: discounted_rtgs_undiscounted_rewards sequence: float64 - name: undiscounted_rtgs_undiscounted_rewards sequence: float64 - name: discounted_rtgs_normalised_discounted_rewards sequence: float64 - name: undiscounted_rtgs_normalised_discounted_rewards sequence: float64 - name: discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_dense_rewards_subgoal_success sequence: float64 - name: normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_discounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_undiscounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_undiscounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: dense_rewards_action_failure sequence: float64 - name: dense_rtgs_action_failure sequence: float64 - name: dense_rewards sequence: float64 - name: dense_rtgs sequence: float64 - name: mission_embeddings sequence: sequence: float32 - name: mission_token_mask sequence: bool - name: rule_feedback_embeddings sequence: sequence: sequence: float32 - name: rule_feedback_token_mask sequence: sequence: bool - name: rule_feedback_available sequence: bool - name: task_feedback_embeddings sequence: sequence: sequence: float32 - name: task_feedback_token_mask sequence: sequence: bool - name: task_feedback_available sequence: bool - name: instruction_embeddings sequence: sequence: sequence: float32 - name: instruction_token_mask sequence: sequence: bool - name: instructions_available sequence: bool splits: - name: train num_bytes: 9025744532 num_examples: 819 - name: val num_bytes: 2259191244 num_examples: 205 download_size: 195048118 dataset_size: 11284935776 - config_name: pickupmaze2x2_1024x1_raw features: - name: task_id dtype: int64 - name: episode_name dtype: string - name: mission dtype: string - name: timesteps sequence: int64 - name: actions sequence: int64 - name: img_obss sequence: array3_d: shape: - 3 - 56 - 56 dtype: float64 - name: task_feedback sequence: string - name: rule_feedback sequence: string - name: instructions sequence: string - name: terminations sequence: int64 - name: discounted_rewards sequence: float64 - name: undiscounted_rewards sequence: float64 - name: normalised_discounted_rewards sequence: float64 - name: discounted_rtgs_discounted_rewards sequence: float64 - name: undiscounted_rtgs_discounted_rewards sequence: float64 - name: discounted_rtgs_undiscounted_rewards sequence: float64 - name: undiscounted_rtgs_undiscounted_rewards sequence: float64 - name: discounted_rtgs_normalised_discounted_rewards sequence: float64 - name: undiscounted_rtgs_normalised_discounted_rewards sequence: float64 - name: discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_dense_rewards_subgoal_success sequence: float64 - name: normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_discounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_undiscounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_undiscounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: dense_rewards_action_failure sequence: float64 - name: dense_rtgs_action_failure sequence: float64 - name: dense_rewards sequence: float64 - name: dense_rtgs sequence: float64 - name: subgoal_successes sequence: int64 - name: subgoals dtype: int64 - name: action_failures sequence: int64 - name: max_steps dtype: int64 - name: episode_len dtype: int64 splits: - name: train num_bytes: 8059178591 num_examples: 819 - name: val num_bytes: 2017254510 num_examples: 205 download_size: 52333257 dataset_size: 10076433101 - config_name: pickupmaze2x2_2048x1_raw features: - name: task_id dtype: int64 - name: episode_name dtype: string - name: mission dtype: string - name: timesteps sequence: int64 - name: actions sequence: int64 - name: img_obss sequence: array3_d: shape: - 3 - 56 - 56 dtype: float64 - name: task_feedback sequence: string - name: rule_feedback sequence: string - name: instructions sequence: string - name: terminations sequence: int64 - name: discounted_rewards sequence: float64 - name: undiscounted_rewards sequence: float64 - name: normalised_discounted_rewards sequence: float64 - name: discounted_rtgs_discounted_rewards sequence: float64 - name: undiscounted_rtgs_discounted_rewards sequence: float64 - name: discounted_rtgs_undiscounted_rewards sequence: float64 - name: undiscounted_rtgs_undiscounted_rewards sequence: float64 - name: discounted_rtgs_normalised_discounted_rewards sequence: float64 - name: undiscounted_rtgs_normalised_discounted_rewards sequence: float64 - name: discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_dense_rewards_subgoal_success sequence: float64 - name: normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_discounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_undiscounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_undiscounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: dense_rewards_action_failure sequence: float64 - name: dense_rtgs_action_failure sequence: float64 - name: dense_rewards sequence: float64 - name: dense_rtgs sequence: float64 - name: subgoal_successes sequence: int64 - name: subgoals dtype: int64 - name: action_failures sequence: int64 - name: max_steps dtype: int64 - name: episode_len dtype: int64 splits: - name: train num_bytes: 16118354173 num_examples: 1638 - name: val num_bytes: 4034526767 num_examples: 410 download_size: 103901835 dataset_size: 20152880940 - config_name: pickupmaze2x2blocked_2048x1_raw features: - name: task_id dtype: int64 - name: episode_name dtype: string - name: mission dtype: string - name: timesteps sequence: int64 - name: actions sequence: int64 - name: img_obss sequence: array3_d: shape: - 3 - 56 - 56 dtype: float64 - name: task_feedback sequence: string - name: rule_feedback sequence: string - name: instructions sequence: string - name: terminations sequence: int64 - name: discounted_rewards sequence: float64 - name: undiscounted_rewards sequence: float64 - name: normalised_discounted_rewards sequence: float64 - name: discounted_rtgs_discounted_rewards sequence: float64 - name: undiscounted_rtgs_discounted_rewards sequence: float64 - name: discounted_rtgs_undiscounted_rewards sequence: float64 - name: undiscounted_rtgs_undiscounted_rewards sequence: float64 - name: discounted_rtgs_normalised_discounted_rewards sequence: float64 - name: undiscounted_rtgs_normalised_discounted_rewards sequence: float64 - name: discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_dense_rewards_subgoal_success sequence: float64 - name: normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_discounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_undiscounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_undiscounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: dense_rewards_action_failure sequence: float64 - name: dense_rtgs_action_failure sequence: float64 - name: dense_rewards sequence: float64 - name: dense_rtgs sequence: float64 - name: subgoal_successes sequence: int64 - name: subgoals dtype: int64 - name: action_failures sequence: int64 - name: max_steps dtype: int64 - name: episode_len dtype: int64 splits: - name: train num_bytes: 32111597652 num_examples: 1638 - name: val num_bytes: 8037697667 num_examples: 410 download_size: 206282578 dataset_size: 40149295319 - config_name: pickupmaze2x2doorsclosed_1024x1_raw features: - name: task_id dtype: int64 - name: episode_name dtype: string - name: mission dtype: string - name: timesteps sequence: int64 - name: actions sequence: int64 - name: img_obss sequence: array3_d: shape: - 3 - 56 - 56 dtype: float64 - name: task_feedback sequence: string - name: rule_feedback sequence: string - name: instructions sequence: string - name: terminations sequence: int64 - name: discounted_rewards sequence: float64 - name: undiscounted_rewards sequence: float64 - name: normalised_discounted_rewards sequence: float64 - name: discounted_rtgs_discounted_rewards sequence: float64 - name: undiscounted_rtgs_discounted_rewards sequence: float64 - name: discounted_rtgs_undiscounted_rewards sequence: float64 - name: undiscounted_rtgs_undiscounted_rewards sequence: float64 - name: discounted_rtgs_normalised_discounted_rewards sequence: float64 - name: undiscounted_rtgs_normalised_discounted_rewards sequence: float64 - name: discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_dense_rewards_subgoal_success sequence: float64 - name: normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_discounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_undiscounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_undiscounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: dense_rewards_action_failure sequence: float64 - name: dense_rtgs_action_failure sequence: float64 - name: dense_rewards sequence: float64 - name: dense_rtgs sequence: float64 - name: subgoal_successes sequence: int64 - name: subgoals dtype: int64 - name: action_failures sequence: int64 - name: max_steps dtype: int64 - name: episode_len dtype: int64 splits: - name: train num_bytes: 8059188925 num_examples: 819 - name: val num_bytes: 2017258253 num_examples: 205 download_size: 50105885 dataset_size: 10076447178 - config_name: pickupmaze2x2doorsclosed_2048x1_raw features: - name: task_id dtype: int64 - name: episode_name dtype: string - name: mission dtype: string - name: timesteps sequence: int64 - name: actions sequence: int64 - name: img_obss sequence: array3_d: shape: - 3 - 56 - 56 dtype: float64 - name: task_feedback sequence: string - name: rule_feedback sequence: string - name: instructions sequence: string - name: terminations sequence: int64 - name: discounted_rewards sequence: float64 - name: undiscounted_rewards sequence: float64 - name: normalised_discounted_rewards sequence: float64 - name: discounted_rtgs_discounted_rewards sequence: float64 - name: undiscounted_rtgs_discounted_rewards sequence: float64 - name: discounted_rtgs_undiscounted_rewards sequence: float64 - name: undiscounted_rtgs_undiscounted_rewards sequence: float64 - name: discounted_rtgs_normalised_discounted_rewards sequence: float64 - name: undiscounted_rtgs_normalised_discounted_rewards sequence: float64 - name: discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_dense_rewards_subgoal_success sequence: float64 - name: normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_discounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_undiscounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_undiscounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: dense_rewards_action_failure sequence: float64 - name: dense_rtgs_action_failure sequence: float64 - name: dense_rewards sequence: float64 - name: dense_rtgs sequence: float64 - name: subgoal_successes sequence: int64 - name: subgoals dtype: int64 - name: action_failures sequence: int64 - name: max_steps dtype: int64 - name: episode_len dtype: int64 splits: - name: train num_bytes: 16118382202 num_examples: 1638 - name: val num_bytes: 4034516927 num_examples: 410 download_size: 100103489 dataset_size: 20152899129 - config_name: pickupmaze2x2doorslocked_1024x1_raw features: - name: task_id dtype: int64 - name: episode_name dtype: string - name: mission dtype: string - name: timesteps sequence: int64 - name: actions sequence: int64 - name: img_obss sequence: array3_d: shape: - 3 - 56 - 56 dtype: float64 - name: task_feedback sequence: string - name: rule_feedback sequence: string - name: instructions sequence: string - name: terminations sequence: int64 - name: discounted_rewards sequence: float64 - name: undiscounted_rewards sequence: float64 - name: normalised_discounted_rewards sequence: float64 - name: discounted_rtgs_discounted_rewards sequence: float64 - name: undiscounted_rtgs_discounted_rewards sequence: float64 - name: discounted_rtgs_undiscounted_rewards sequence: float64 - name: undiscounted_rtgs_undiscounted_rewards sequence: float64 - name: discounted_rtgs_normalised_discounted_rewards sequence: float64 - name: undiscounted_rtgs_normalised_discounted_rewards sequence: float64 - name: discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_dense_rewards_subgoal_success sequence: float64 - name: normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_discounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_undiscounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_undiscounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: dense_rewards_action_failure sequence: float64 - name: dense_rtgs_action_failure sequence: float64 - name: dense_rewards sequence: float64 - name: dense_rtgs sequence: float64 - name: subgoal_successes sequence: int64 - name: subgoals dtype: int64 - name: action_failures sequence: int64 - name: max_steps dtype: int64 - name: episode_len dtype: int64 splits: - name: train num_bytes: 16056034589 num_examples: 819 - name: val num_bytes: 4018910371 num_examples: 205 download_size: 107628693 dataset_size: 20074944960 - config_name: pickupmaze2x2doorslocked_2048x1_raw features: - name: task_id dtype: int64 - name: episode_name dtype: string - name: mission dtype: string - name: timesteps sequence: int64 - name: actions sequence: int64 - name: img_obss sequence: array3_d: shape: - 3 - 56 - 56 dtype: float64 - name: task_feedback sequence: string - name: rule_feedback sequence: string - name: instructions sequence: string - name: terminations sequence: int64 - name: discounted_rewards sequence: float64 - name: undiscounted_rewards sequence: float64 - name: normalised_discounted_rewards sequence: float64 - name: discounted_rtgs_discounted_rewards sequence: float64 - name: undiscounted_rtgs_discounted_rewards sequence: float64 - name: discounted_rtgs_undiscounted_rewards sequence: float64 - name: undiscounted_rtgs_undiscounted_rewards sequence: float64 - name: discounted_rtgs_normalised_discounted_rewards sequence: float64 - name: undiscounted_rtgs_normalised_discounted_rewards sequence: float64 - name: discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_dense_rewards_subgoal_success sequence: float64 - name: normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_discounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_undiscounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_undiscounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: dense_rewards_action_failure sequence: float64 - name: dense_rtgs_action_failure sequence: float64 - name: dense_rewards sequence: float64 - name: dense_rtgs sequence: float64 - name: subgoal_successes sequence: int64 - name: subgoals dtype: int64 - name: action_failures sequence: int64 - name: max_steps dtype: int64 - name: episode_len dtype: int64 splits: - name: train num_bytes: 32112036726 num_examples: 1638 - name: val num_bytes: 8037783005 num_examples: 410 download_size: 215105875 dataset_size: 40149819731 - config_name: pickups12_1024x1_preprocessed_sentence-transformer-pretrained_cnn-scratch features: - name: task_id dtype: int64 - name: timesteps sequence: int64 - name: actions sequence: int64 - name: img_obss sequence: array3_d: shape: - 3 - 56 - 56 dtype: float64 - name: terminations sequence: int64 - name: discounted_rewards sequence: float64 - name: undiscounted_rewards sequence: float64 - name: normalised_discounted_rewards sequence: float64 - name: discounted_rtgs_discounted_rewards sequence: float64 - name: undiscounted_rtgs_discounted_rewards sequence: float64 - name: discounted_rtgs_undiscounted_rewards sequence: float64 - name: undiscounted_rtgs_undiscounted_rewards sequence: float64 - name: discounted_rtgs_normalised_discounted_rewards sequence: float64 - name: undiscounted_rtgs_normalised_discounted_rewards sequence: float64 - name: discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_dense_rewards_subgoal_success sequence: float64 - name: normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_discounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_undiscounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_undiscounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: dense_rewards_action_failure sequence: float64 - name: dense_rtgs_action_failure sequence: float64 - name: dense_rewards sequence: float64 - name: dense_rtgs sequence: float64 - name: mission_embeddings sequence: sequence: float32 - name: mission_token_mask sequence: bool - name: rule_feedback_embeddings sequence: sequence: sequence: float32 - name: rule_feedback_token_mask sequence: sequence: bool - name: rule_feedback_available sequence: bool - name: task_feedback_embeddings sequence: sequence: sequence: float32 - name: task_feedback_token_mask sequence: sequence: bool - name: task_feedback_available sequence: bool - name: instruction_embeddings sequence: sequence: sequence: float32 - name: instruction_token_mask sequence: sequence: bool - name: instructions_available sequence: bool splits: - name: train num_bytes: 3430025312 num_examples: 819 - name: val num_bytes: 858553344 num_examples: 205 download_size: 86903879 dataset_size: 4288578656 - config_name: pickups12_1024x1_raw features: - name: task_id dtype: int64 - name: episode_name dtype: string - name: mission dtype: string - name: timesteps sequence: int64 - name: actions sequence: int64 - name: img_obss sequence: array3_d: shape: - 3 - 56 - 56 dtype: float64 - name: task_feedback sequence: string - name: rule_feedback sequence: string - name: instructions sequence: string - name: terminations sequence: int64 - name: discounted_rewards sequence: float64 - name: undiscounted_rewards sequence: float64 - name: normalised_discounted_rewards sequence: float64 - name: discounted_rtgs_discounted_rewards sequence: float64 - name: undiscounted_rtgs_discounted_rewards sequence: float64 - name: discounted_rtgs_undiscounted_rewards sequence: float64 - name: undiscounted_rtgs_undiscounted_rewards sequence: float64 - name: discounted_rtgs_normalised_discounted_rewards sequence: float64 - name: undiscounted_rtgs_normalised_discounted_rewards sequence: float64 - name: discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_dense_rewards_subgoal_success sequence: float64 - name: normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_discounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_undiscounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_undiscounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: dense_rewards_action_failure sequence: float64 - name: dense_rtgs_action_failure sequence: float64 - name: dense_rewards sequence: float64 - name: dense_rtgs sequence: float64 - name: subgoal_successes sequence: int64 - name: subgoals dtype: int64 - name: action_failures sequence: int64 - name: max_steps dtype: int64 - name: episode_len dtype: int64 splits: - name: train num_bytes: 3061403082 num_examples: 819 - name: val num_bytes: 766280932 num_examples: 205 download_size: 24616914 dataset_size: 3827684014 - config_name: pickups12_2048x1_preprocessed_sentence-transformer-pretrained_cnn-scratch features: - name: task_id dtype: int64 - name: timesteps sequence: int64 - name: actions sequence: int64 - name: img_obss sequence: array3_d: shape: - 3 - 56 - 56 dtype: float64 - name: terminations sequence: int64 - name: discounted_rewards sequence: float64 - name: undiscounted_rewards sequence: float64 - name: normalised_discounted_rewards sequence: float64 - name: discounted_rtgs_discounted_rewards sequence: float64 - name: undiscounted_rtgs_discounted_rewards sequence: float64 - name: discounted_rtgs_undiscounted_rewards sequence: float64 - name: undiscounted_rtgs_undiscounted_rewards sequence: float64 - name: discounted_rtgs_normalised_discounted_rewards sequence: float64 - name: undiscounted_rtgs_normalised_discounted_rewards sequence: float64 - name: discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_dense_rewards_subgoal_success sequence: float64 - name: normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_discounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_undiscounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_undiscounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: dense_rewards_action_failure sequence: float64 - name: dense_rtgs_action_failure sequence: float64 - name: dense_rewards sequence: float64 - name: dense_rtgs sequence: float64 - name: mission_embeddings sequence: sequence: float32 - name: mission_token_mask sequence: bool - name: rule_feedback_embeddings sequence: sequence: sequence: float32 - name: rule_feedback_token_mask sequence: sequence: bool - name: rule_feedback_available sequence: bool - name: task_feedback_embeddings sequence: sequence: sequence: float32 - name: task_feedback_token_mask sequence: sequence: bool - name: task_feedback_available sequence: bool - name: instruction_embeddings sequence: sequence: sequence: float32 - name: instruction_token_mask sequence: sequence: bool - name: instructions_available sequence: bool splits: - name: train num_bytes: 6860050400 num_examples: 1638 - name: val num_bytes: 1717106688 num_examples: 410 download_size: 174066638 dataset_size: 8577157088 - config_name: pickups12_2048x1_raw features: - name: task_id dtype: int64 - name: episode_name dtype: string - name: mission dtype: string - name: timesteps sequence: int64 - name: actions sequence: int64 - name: img_obss sequence: array3_d: shape: - 3 - 56 - 56 dtype: float64 - name: task_feedback sequence: string - name: rule_feedback sequence: string - name: instructions sequence: string - name: terminations sequence: int64 - name: discounted_rewards sequence: float64 - name: undiscounted_rewards sequence: float64 - name: normalised_discounted_rewards sequence: float64 - name: discounted_rtgs_discounted_rewards sequence: float64 - name: undiscounted_rtgs_discounted_rewards sequence: float64 - name: discounted_rtgs_undiscounted_rewards sequence: float64 - name: undiscounted_rtgs_undiscounted_rewards sequence: float64 - name: discounted_rtgs_normalised_discounted_rewards sequence: float64 - name: undiscounted_rtgs_normalised_discounted_rewards sequence: float64 - name: discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_dense_rewards_subgoal_success sequence: float64 - name: normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_discounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_undiscounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_undiscounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: dense_rewards_action_failure sequence: float64 - name: dense_rtgs_action_failure sequence: float64 - name: dense_rewards sequence: float64 - name: dense_rtgs sequence: float64 - name: subgoal_successes sequence: int64 - name: subgoals dtype: int64 - name: action_failures sequence: int64 - name: max_steps dtype: int64 - name: episode_len dtype: int64 splits: - name: train num_bytes: 6122801422 num_examples: 1638 - name: val num_bytes: 1532570120 num_examples: 410 download_size: 48911622 dataset_size: 7655371542 - config_name: pickupsimilardistsuniquegoalstrictlycolortype_1024x1_preprocessed_sentence-transformer-pretrained_cnn-scratch features: - name: task_id dtype: int64 - name: timesteps sequence: int64 - name: actions sequence: int64 - name: img_obss sequence: array3_d: shape: - 3 - 56 - 56 dtype: float64 - name: terminations sequence: int64 - name: discounted_rewards sequence: float64 - name: undiscounted_rewards sequence: float64 - name: normalised_discounted_rewards sequence: float64 - name: discounted_rtgs_discounted_rewards sequence: float64 - name: undiscounted_rtgs_discounted_rewards sequence: float64 - name: discounted_rtgs_undiscounted_rewards sequence: float64 - name: undiscounted_rtgs_undiscounted_rewards sequence: float64 - name: discounted_rtgs_normalised_discounted_rewards sequence: float64 - name: undiscounted_rtgs_normalised_discounted_rewards sequence: float64 - name: discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_dense_rewards_subgoal_success sequence: float64 - name: normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_discounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_undiscounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_undiscounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: dense_rewards_action_failure sequence: float64 - name: dense_rtgs_action_failure sequence: float64 - name: dense_rewards sequence: float64 - name: dense_rtgs sequence: float64 - name: mission_embeddings sequence: sequence: float32 - name: mission_token_mask sequence: bool - name: rule_feedback_embeddings sequence: sequence: sequence: float32 - name: rule_feedback_token_mask sequence: sequence: bool - name: rule_feedback_available sequence: bool - name: task_feedback_embeddings sequence: sequence: sequence: float32 - name: task_feedback_token_mask sequence: sequence: bool - name: task_feedback_available sequence: bool - name: instruction_embeddings sequence: sequence: sequence: float32 - name: instruction_token_mask sequence: sequence: bool - name: instructions_available sequence: bool splits: - name: train num_bytes: 2310881468 num_examples: 819 - name: val num_bytes: 578425764 num_examples: 205 download_size: 67808303 dataset_size: 2889307232 - config_name: pickupsimilardistsuniquegoalstrictlycolortype_1024x1_raw features: - name: task_id dtype: int64 - name: episode_name dtype: string - name: mission dtype: string - name: timesteps sequence: int64 - name: actions sequence: int64 - name: img_obss sequence: array3_d: shape: - 3 - 56 - 56 dtype: float64 - name: task_feedback sequence: string - name: rule_feedback sequence: string - name: instructions sequence: string - name: terminations sequence: int64 - name: discounted_rewards sequence: float64 - name: undiscounted_rewards sequence: float64 - name: normalised_discounted_rewards sequence: float64 - name: discounted_rtgs_discounted_rewards sequence: float64 - name: undiscounted_rtgs_discounted_rewards sequence: float64 - name: discounted_rtgs_undiscounted_rewards sequence: float64 - name: undiscounted_rtgs_undiscounted_rewards sequence: float64 - name: discounted_rtgs_normalised_discounted_rewards sequence: float64 - name: undiscounted_rtgs_normalised_discounted_rewards sequence: float64 - name: discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_dense_rewards_subgoal_success sequence: float64 - name: normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_discounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_undiscounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_undiscounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: dense_rewards_action_failure sequence: float64 - name: dense_rtgs_action_failure sequence: float64 - name: dense_rewards sequence: float64 - name: dense_rtgs sequence: float64 - name: subgoal_successes sequence: int64 - name: subgoals dtype: int64 - name: action_failures sequence: int64 - name: max_steps dtype: int64 - name: episode_len dtype: int64 splits: - name: train num_bytes: 2061814782 num_examples: 819 - name: val num_bytes: 516085190 num_examples: 205 download_size: 15190439 dataset_size: 2577899972 - config_name: pickupsimilardistsuniquegoalstrictlycolortype_2048x1_preprocessed_sentence-transformer-pretrained_cnn-scratch features: - name: task_id dtype: int64 - name: timesteps sequence: int64 - name: actions sequence: int64 - name: img_obss sequence: array3_d: shape: - 3 - 56 - 56 dtype: float64 - name: terminations sequence: int64 - name: discounted_rewards sequence: float64 - name: undiscounted_rewards sequence: float64 - name: normalised_discounted_rewards sequence: float64 - name: discounted_rtgs_discounted_rewards sequence: float64 - name: undiscounted_rtgs_discounted_rewards sequence: float64 - name: discounted_rtgs_undiscounted_rewards sequence: float64 - name: undiscounted_rtgs_undiscounted_rewards sequence: float64 - name: discounted_rtgs_normalised_discounted_rewards sequence: float64 - name: undiscounted_rtgs_normalised_discounted_rewards sequence: float64 - name: discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_dense_rewards_subgoal_success sequence: float64 - name: normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_discounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_undiscounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_undiscounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: dense_rewards_action_failure sequence: float64 - name: dense_rtgs_action_failure sequence: float64 - name: dense_rewards sequence: float64 - name: dense_rtgs sequence: float64 - name: mission_embeddings sequence: sequence: float32 - name: mission_token_mask sequence: bool - name: rule_feedback_embeddings sequence: sequence: sequence: float32 - name: rule_feedback_token_mask sequence: sequence: bool - name: rule_feedback_available sequence: bool - name: task_feedback_embeddings sequence: sequence: sequence: float32 - name: task_feedback_token_mask sequence: sequence: bool - name: task_feedback_available sequence: bool - name: instruction_embeddings sequence: sequence: sequence: float32 - name: instruction_token_mask sequence: sequence: bool - name: instructions_available sequence: bool splits: - name: train num_bytes: 4621762712 num_examples: 1638 - name: val num_bytes: 1156851528 num_examples: 410 download_size: 135358578 dataset_size: 5778614240 - config_name: pickupsimilardistsuniquegoalstrictlycolortype_2048x1_raw features: - name: task_id dtype: int64 - name: episode_name dtype: string - name: mission dtype: string - name: timesteps sequence: int64 - name: actions sequence: int64 - name: img_obss sequence: array3_d: shape: - 3 - 56 - 56 dtype: float64 - name: task_feedback sequence: string - name: rule_feedback sequence: string - name: instructions sequence: string - name: terminations sequence: int64 - name: discounted_rewards sequence: float64 - name: undiscounted_rewards sequence: float64 - name: normalised_discounted_rewards sequence: float64 - name: discounted_rtgs_discounted_rewards sequence: float64 - name: undiscounted_rtgs_discounted_rewards sequence: float64 - name: discounted_rtgs_undiscounted_rewards sequence: float64 - name: undiscounted_rtgs_undiscounted_rewards sequence: float64 - name: discounted_rtgs_normalised_discounted_rewards sequence: float64 - name: undiscounted_rtgs_normalised_discounted_rewards sequence: float64 - name: discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_dense_rewards_subgoal_success sequence: float64 - name: normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_discounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_undiscounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_undiscounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: dense_rewards_action_failure sequence: float64 - name: dense_rtgs_action_failure sequence: float64 - name: dense_rewards sequence: float64 - name: dense_rtgs sequence: float64 - name: subgoal_successes sequence: int64 - name: subgoals dtype: int64 - name: action_failures sequence: int64 - name: max_steps dtype: int64 - name: episode_len dtype: int64 splits: - name: train num_bytes: 4123637792 num_examples: 1638 - name: val num_bytes: 1032173701 num_examples: 410 download_size: 30264938 dataset_size: 5155811493 - config_name: pickupsimilardistsuniquegoalstrictlycolortype_4096x1_preprocessed_sentence-transformer-pretrained_cnn-scratch features: - name: task_id dtype: int64 - name: timesteps sequence: int64 - name: actions sequence: int64 - name: img_obss sequence: array3_d: shape: - 3 - 56 - 56 dtype: float64 - name: terminations sequence: int64 - name: discounted_rewards sequence: float64 - name: undiscounted_rewards sequence: float64 - name: normalised_discounted_rewards sequence: float64 - name: discounted_rtgs_discounted_rewards sequence: float64 - name: undiscounted_rtgs_discounted_rewards sequence: float64 - name: discounted_rtgs_undiscounted_rewards sequence: float64 - name: undiscounted_rtgs_undiscounted_rewards sequence: float64 - name: discounted_rtgs_normalised_discounted_rewards sequence: float64 - name: undiscounted_rtgs_normalised_discounted_rewards sequence: float64 - name: discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_dense_rewards_subgoal_success sequence: float64 - name: normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_discounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_undiscounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_undiscounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: dense_rewards_action_failure sequence: float64 - name: dense_rtgs_action_failure sequence: float64 - name: dense_rewards sequence: float64 - name: dense_rtgs sequence: float64 - name: mission_embeddings sequence: sequence: float32 - name: mission_token_mask sequence: bool - name: rule_feedback_embeddings sequence: sequence: sequence: float32 - name: rule_feedback_token_mask sequence: sequence: bool - name: rule_feedback_available sequence: bool - name: task_feedback_embeddings sequence: sequence: sequence: float32 - name: task_feedback_token_mask sequence: sequence: bool - name: task_feedback_available sequence: bool - name: instruction_embeddings sequence: sequence: sequence: float32 - name: instruction_token_mask sequence: sequence: bool - name: instructions_available sequence: bool splits: - name: train num_bytes: 9243525200 num_examples: 3276 - name: val num_bytes: 2313703056 num_examples: 820 download_size: 277435404 dataset_size: 11557228256 - config_name: pickupsimilardistsuniquegoalstrictlycolortype_4096x1_raw features: - name: task_id dtype: int64 - name: episode_name dtype: string - name: mission dtype: string - name: timesteps sequence: int64 - name: actions sequence: int64 - name: img_obss sequence: array3_d: shape: - 3 - 56 - 56 dtype: float64 - name: task_feedback sequence: string - name: rule_feedback sequence: string - name: instructions sequence: string - name: terminations sequence: int64 - name: discounted_rewards sequence: float64 - name: undiscounted_rewards sequence: float64 - name: normalised_discounted_rewards sequence: float64 - name: discounted_rtgs_discounted_rewards sequence: float64 - name: undiscounted_rtgs_discounted_rewards sequence: float64 - name: discounted_rtgs_undiscounted_rewards sequence: float64 - name: undiscounted_rtgs_undiscounted_rewards sequence: float64 - name: discounted_rtgs_normalised_discounted_rewards sequence: float64 - name: undiscounted_rtgs_normalised_discounted_rewards sequence: float64 - name: discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_dense_rewards_subgoal_success sequence: float64 - name: normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_discounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_undiscounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_undiscounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: dense_rewards_action_failure sequence: float64 - name: dense_rtgs_action_failure sequence: float64 - name: dense_rewards sequence: float64 - name: dense_rtgs sequence: float64 - name: subgoal_successes sequence: int64 - name: subgoals dtype: int64 - name: action_failures sequence: int64 - name: max_steps dtype: int64 - name: episode_len dtype: int64 splits: - name: train num_bytes: 8247283850 num_examples: 3276 - name: val num_bytes: 2064351217 num_examples: 820 download_size: 60442367 dataset_size: 10311635067 - config_name: pickupsimilardistsuniquegoalstrictlyloc_1024x1_preprocessed_sentence-transformer-pretrained_cnn-scratch features: - name: task_id dtype: int64 - name: timesteps sequence: int64 - name: actions sequence: int64 - name: img_obss sequence: array3_d: shape: - 3 - 56 - 56 dtype: float64 - name: terminations sequence: int64 - name: discounted_rewards sequence: float64 - name: undiscounted_rewards sequence: float64 - name: normalised_discounted_rewards sequence: float64 - name: discounted_rtgs_discounted_rewards sequence: float64 - name: undiscounted_rtgs_discounted_rewards sequence: float64 - name: discounted_rtgs_undiscounted_rewards sequence: float64 - name: undiscounted_rtgs_undiscounted_rewards sequence: float64 - name: discounted_rtgs_normalised_discounted_rewards sequence: float64 - name: undiscounted_rtgs_normalised_discounted_rewards sequence: float64 - name: discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_dense_rewards_subgoal_success sequence: float64 - name: normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_discounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_undiscounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_undiscounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: dense_rewards_action_failure sequence: float64 - name: dense_rtgs_action_failure sequence: float64 - name: dense_rewards sequence: float64 - name: dense_rtgs sequence: float64 - name: mission_embeddings sequence: sequence: float32 - name: mission_token_mask sequence: bool - name: rule_feedback_embeddings sequence: sequence: sequence: float32 - name: rule_feedback_token_mask sequence: sequence: bool - name: rule_feedback_available sequence: bool - name: task_feedback_embeddings sequence: sequence: sequence: float32 - name: task_feedback_token_mask sequence: sequence: bool - name: task_feedback_available sequence: bool - name: instruction_embeddings sequence: sequence: sequence: float32 - name: instruction_token_mask sequence: sequence: bool - name: instructions_available sequence: bool splits: - name: train num_bytes: 2310881468 num_examples: 819 - name: val num_bytes: 578425764 num_examples: 205 download_size: 74317263 dataset_size: 2889307232 - config_name: pickupsimilardistsuniquegoalstrictlyloc_1024x1_raw features: - name: task_id dtype: int64 - name: episode_name dtype: string - name: mission dtype: string - name: timesteps sequence: int64 - name: actions sequence: int64 - name: img_obss sequence: array3_d: shape: - 3 - 56 - 56 dtype: float64 - name: task_feedback sequence: string - name: rule_feedback sequence: string - name: instructions sequence: string - name: terminations sequence: int64 - name: discounted_rewards sequence: float64 - name: undiscounted_rewards sequence: float64 - name: normalised_discounted_rewards sequence: float64 - name: discounted_rtgs_discounted_rewards sequence: float64 - name: undiscounted_rtgs_discounted_rewards sequence: float64 - name: discounted_rtgs_undiscounted_rewards sequence: float64 - name: undiscounted_rtgs_undiscounted_rewards sequence: float64 - name: discounted_rtgs_normalised_discounted_rewards sequence: float64 - name: undiscounted_rtgs_normalised_discounted_rewards sequence: float64 - name: discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_dense_rewards_subgoal_success sequence: float64 - name: normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_discounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_undiscounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_undiscounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: dense_rewards_action_failure sequence: float64 - name: dense_rtgs_action_failure sequence: float64 - name: dense_rewards sequence: float64 - name: dense_rtgs sequence: float64 - name: subgoal_successes sequence: int64 - name: subgoals dtype: int64 - name: action_failures sequence: int64 - name: max_steps dtype: int64 - name: episode_len dtype: int64 splits: - name: train num_bytes: 2061883046 num_examples: 819 - name: val num_bytes: 516106691 num_examples: 205 download_size: 15216870 dataset_size: 2577989737 - config_name: pickupsimilardistsuniquegoalstrictlyloc_2048x1_preprocessed_sentence-transformer-pretrained_cnn-scratch features: - name: task_id dtype: int64 - name: timesteps sequence: int64 - name: actions sequence: int64 - name: img_obss sequence: array3_d: shape: - 3 - 56 - 56 dtype: float64 - name: terminations sequence: int64 - name: discounted_rewards sequence: float64 - name: undiscounted_rewards sequence: float64 - name: normalised_discounted_rewards sequence: float64 - name: discounted_rtgs_discounted_rewards sequence: float64 - name: undiscounted_rtgs_discounted_rewards sequence: float64 - name: discounted_rtgs_undiscounted_rewards sequence: float64 - name: undiscounted_rtgs_undiscounted_rewards sequence: float64 - name: discounted_rtgs_normalised_discounted_rewards sequence: float64 - name: undiscounted_rtgs_normalised_discounted_rewards sequence: float64 - name: discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_dense_rewards_subgoal_success sequence: float64 - name: normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_discounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_undiscounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_undiscounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: dense_rewards_action_failure sequence: float64 - name: dense_rtgs_action_failure sequence: float64 - name: dense_rewards sequence: float64 - name: dense_rtgs sequence: float64 - name: mission_embeddings sequence: sequence: float32 - name: mission_token_mask sequence: bool - name: rule_feedback_embeddings sequence: sequence: sequence: float32 - name: rule_feedback_token_mask sequence: sequence: bool - name: rule_feedback_available sequence: bool - name: task_feedback_embeddings sequence: sequence: sequence: float32 - name: task_feedback_token_mask sequence: sequence: bool - name: task_feedback_available sequence: bool - name: instruction_embeddings sequence: sequence: sequence: float32 - name: instruction_token_mask sequence: sequence: bool - name: instructions_available sequence: bool splits: - name: train num_bytes: 4621762712 num_examples: 1638 - name: val num_bytes: 1156851528 num_examples: 410 download_size: 147713392 dataset_size: 5778614240 - config_name: pickupsimilardistsuniquegoalstrictlyloc_2048x1_raw features: - name: task_id dtype: int64 - name: episode_name dtype: string - name: mission dtype: string - name: timesteps sequence: int64 - name: actions sequence: int64 - name: img_obss sequence: array3_d: shape: - 3 - 56 - 56 dtype: float64 - name: task_feedback sequence: string - name: rule_feedback sequence: string - name: instructions sequence: string - name: terminations sequence: int64 - name: discounted_rewards sequence: float64 - name: undiscounted_rewards sequence: float64 - name: normalised_discounted_rewards sequence: float64 - name: discounted_rtgs_discounted_rewards sequence: float64 - name: undiscounted_rtgs_discounted_rewards sequence: float64 - name: discounted_rtgs_undiscounted_rewards sequence: float64 - name: undiscounted_rtgs_undiscounted_rewards sequence: float64 - name: discounted_rtgs_normalised_discounted_rewards sequence: float64 - name: undiscounted_rtgs_normalised_discounted_rewards sequence: float64 - name: discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_dense_rewards_subgoal_success sequence: float64 - name: normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_discounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_undiscounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_undiscounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: dense_rewards_action_failure sequence: float64 - name: dense_rtgs_action_failure sequence: float64 - name: dense_rewards sequence: float64 - name: dense_rtgs sequence: float64 - name: subgoal_successes sequence: int64 - name: subgoals dtype: int64 - name: action_failures sequence: int64 - name: max_steps dtype: int64 - name: episode_len dtype: int64 splits: - name: train num_bytes: 4123786984 num_examples: 1638 - name: val num_bytes: 1032204802 num_examples: 410 download_size: 30311156 dataset_size: 5155991786 - config_name: pickupsimilardistsuniquegoalstrictlyonlycolor_1024x1_preprocessed_sentence-transformer-pretrained_cnn-scratch features: - name: task_id dtype: int64 - name: timesteps sequence: int64 - name: actions sequence: int64 - name: img_obss sequence: array3_d: shape: - 3 - 56 - 56 dtype: float64 - name: terminations sequence: int64 - name: discounted_rewards sequence: float64 - name: undiscounted_rewards sequence: float64 - name: normalised_discounted_rewards sequence: float64 - name: discounted_rtgs_discounted_rewards sequence: float64 - name: undiscounted_rtgs_discounted_rewards sequence: float64 - name: discounted_rtgs_undiscounted_rewards sequence: float64 - name: undiscounted_rtgs_undiscounted_rewards sequence: float64 - name: discounted_rtgs_normalised_discounted_rewards sequence: float64 - name: undiscounted_rtgs_normalised_discounted_rewards sequence: float64 - name: discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_dense_rewards_subgoal_success sequence: float64 - name: normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_discounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_undiscounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_undiscounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: dense_rewards_action_failure sequence: float64 - name: dense_rtgs_action_failure sequence: float64 - name: dense_rewards sequence: float64 - name: dense_rtgs sequence: float64 - name: mission_embeddings sequence: sequence: float32 - name: mission_token_mask sequence: bool - name: rule_feedback_embeddings sequence: sequence: sequence: float32 - name: rule_feedback_token_mask sequence: sequence: bool - name: rule_feedback_available sequence: bool - name: task_feedback_embeddings sequence: sequence: sequence: float32 - name: task_feedback_token_mask sequence: sequence: bool - name: task_feedback_available sequence: bool - name: instruction_embeddings sequence: sequence: sequence: float32 - name: instruction_token_mask sequence: sequence: bool - name: instructions_available sequence: bool splits: - name: train num_bytes: 2310881468 num_examples: 819 - name: val num_bytes: 578425764 num_examples: 205 download_size: 64174601 dataset_size: 2889307232 - config_name: pickupsimilardistsuniquegoalstrictlyonlycolor_1024x1_raw features: - name: task_id dtype: int64 - name: episode_name dtype: string - name: mission dtype: string - name: timesteps sequence: int64 - name: actions sequence: int64 - name: img_obss sequence: array3_d: shape: - 3 - 56 - 56 dtype: float64 - name: task_feedback sequence: string - name: rule_feedback sequence: string - name: instructions sequence: string - name: terminations sequence: int64 - name: discounted_rewards sequence: float64 - name: undiscounted_rewards sequence: float64 - name: normalised_discounted_rewards sequence: float64 - name: discounted_rtgs_discounted_rewards sequence: float64 - name: undiscounted_rtgs_discounted_rewards sequence: float64 - name: discounted_rtgs_undiscounted_rewards sequence: float64 - name: undiscounted_rtgs_undiscounted_rewards sequence: float64 - name: discounted_rtgs_normalised_discounted_rewards sequence: float64 - name: undiscounted_rtgs_normalised_discounted_rewards sequence: float64 - name: discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_dense_rewards_subgoal_success sequence: float64 - name: normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_discounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_undiscounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_undiscounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: dense_rewards_action_failure sequence: float64 - name: dense_rtgs_action_failure sequence: float64 - name: dense_rewards sequence: float64 - name: dense_rtgs sequence: float64 - name: subgoal_successes sequence: int64 - name: subgoals dtype: int64 - name: action_failures sequence: int64 - name: max_steps dtype: int64 - name: episode_len dtype: int64 splits: - name: train num_bytes: 2061840991 num_examples: 819 - name: val num_bytes: 516093463 num_examples: 205 download_size: 15182253 dataset_size: 2577934454 - config_name: pickupsimilardistsuniquegoalstrictlyonlycolor_2048x1_preprocessed_sentence-transformer-pretrained_cnn-scratch features: - name: task_id dtype: int64 - name: timesteps sequence: int64 - name: actions sequence: int64 - name: img_obss sequence: array3_d: shape: - 3 - 56 - 56 dtype: float64 - name: terminations sequence: int64 - name: discounted_rewards sequence: float64 - name: undiscounted_rewards sequence: float64 - name: normalised_discounted_rewards sequence: float64 - name: discounted_rtgs_discounted_rewards sequence: float64 - name: undiscounted_rtgs_discounted_rewards sequence: float64 - name: discounted_rtgs_undiscounted_rewards sequence: float64 - name: undiscounted_rtgs_undiscounted_rewards sequence: float64 - name: discounted_rtgs_normalised_discounted_rewards sequence: float64 - name: undiscounted_rtgs_normalised_discounted_rewards sequence: float64 - name: discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_dense_rewards_subgoal_success sequence: float64 - name: normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_discounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_undiscounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_undiscounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: dense_rewards_action_failure sequence: float64 - name: dense_rtgs_action_failure sequence: float64 - name: dense_rewards sequence: float64 - name: dense_rtgs sequence: float64 - name: mission_embeddings sequence: sequence: float32 - name: mission_token_mask sequence: bool - name: rule_feedback_embeddings sequence: sequence: sequence: float32 - name: rule_feedback_token_mask sequence: sequence: bool - name: rule_feedback_available sequence: bool - name: task_feedback_embeddings sequence: sequence: sequence: float32 - name: task_feedback_token_mask sequence: sequence: bool - name: task_feedback_available sequence: bool - name: instruction_embeddings sequence: sequence: sequence: float32 - name: instruction_token_mask sequence: sequence: bool - name: instructions_available sequence: bool splits: - name: train num_bytes: 4621762712 num_examples: 1638 - name: val num_bytes: 1156851528 num_examples: 410 download_size: 128165408 dataset_size: 5778614240 - config_name: pickupsimilardistsuniquegoalstrictlyonlycolor_2048x1_raw features: - name: task_id dtype: int64 - name: episode_name dtype: string - name: mission dtype: string - name: timesteps sequence: int64 - name: actions sequence: int64 - name: img_obss sequence: array3_d: shape: - 3 - 56 - 56 dtype: float64 - name: task_feedback sequence: string - name: rule_feedback sequence: string - name: instructions sequence: string - name: terminations sequence: int64 - name: discounted_rewards sequence: float64 - name: undiscounted_rewards sequence: float64 - name: normalised_discounted_rewards sequence: float64 - name: discounted_rtgs_discounted_rewards sequence: float64 - name: undiscounted_rtgs_discounted_rewards sequence: float64 - name: discounted_rtgs_undiscounted_rewards sequence: float64 - name: undiscounted_rtgs_undiscounted_rewards sequence: float64 - name: discounted_rtgs_normalised_discounted_rewards sequence: float64 - name: undiscounted_rtgs_normalised_discounted_rewards sequence: float64 - name: discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_dense_rewards_subgoal_success sequence: float64 - name: normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_discounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_undiscounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_undiscounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: dense_rewards_action_failure sequence: float64 - name: dense_rtgs_action_failure sequence: float64 - name: dense_rewards sequence: float64 - name: dense_rtgs sequence: float64 - name: subgoal_successes sequence: int64 - name: subgoals dtype: int64 - name: action_failures sequence: int64 - name: max_steps dtype: int64 - name: episode_len dtype: int64 splits: - name: train num_bytes: 4123698210 num_examples: 1638 - name: val num_bytes: 1032181089 num_examples: 410 download_size: 30272205 dataset_size: 5155879299 - config_name: putnextsimilardistsuniquegoalstrictlycolortype_1024x1_preprocessed_sentence-transformer-pretrained_cnn-scratch features: - name: task_id dtype: int64 - name: timesteps sequence: int64 - name: actions sequence: int64 - name: img_obss sequence: array3_d: shape: - 3 - 56 - 56 dtype: float64 - name: terminations sequence: int64 - name: discounted_rewards sequence: float64 - name: undiscounted_rewards sequence: float64 - name: normalised_discounted_rewards sequence: float64 - name: discounted_rtgs_discounted_rewards sequence: float64 - name: undiscounted_rtgs_discounted_rewards sequence: float64 - name: discounted_rtgs_undiscounted_rewards sequence: float64 - name: undiscounted_rtgs_undiscounted_rewards sequence: float64 - name: discounted_rtgs_normalised_discounted_rewards sequence: float64 - name: undiscounted_rtgs_normalised_discounted_rewards sequence: float64 - name: discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_dense_rewards_subgoal_success sequence: float64 - name: normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_discounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_undiscounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_undiscounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: dense_rewards_action_failure sequence: float64 - name: dense_rtgs_action_failure sequence: float64 - name: dense_rewards sequence: float64 - name: dense_rtgs sequence: float64 - name: mission_embeddings sequence: sequence: float32 - name: mission_token_mask sequence: bool - name: rule_feedback_embeddings sequence: sequence: sequence: float32 - name: rule_feedback_token_mask sequence: sequence: bool - name: rule_feedback_available sequence: bool - name: task_feedback_embeddings sequence: sequence: sequence: float32 - name: task_feedback_token_mask sequence: sequence: bool - name: task_feedback_available sequence: bool - name: instruction_embeddings sequence: sequence: sequence: float32 - name: instruction_token_mask sequence: sequence: bool - name: instructions_available sequence: bool splits: - name: train num_bytes: 4549169156 num_examples: 819 - name: val num_bytes: 1138680924 num_examples: 205 download_size: 140966972 dataset_size: 5687850080 - config_name: putnextsimilardistsuniquegoalstrictlycolortype_1024x1_raw features: - name: task_id dtype: int64 - name: episode_name dtype: string - name: mission dtype: string - name: timesteps sequence: int64 - name: actions sequence: int64 - name: img_obss sequence: array3_d: shape: - 3 - 56 - 56 dtype: float64 - name: task_feedback sequence: string - name: rule_feedback sequence: string - name: instructions sequence: string - name: terminations sequence: int64 - name: discounted_rewards sequence: float64 - name: undiscounted_rewards sequence: float64 - name: normalised_discounted_rewards sequence: float64 - name: discounted_rtgs_discounted_rewards sequence: float64 - name: undiscounted_rtgs_discounted_rewards sequence: float64 - name: discounted_rtgs_undiscounted_rewards sequence: float64 - name: undiscounted_rtgs_undiscounted_rewards sequence: float64 - name: discounted_rtgs_normalised_discounted_rewards sequence: float64 - name: undiscounted_rtgs_normalised_discounted_rewards sequence: float64 - name: discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_dense_rewards_subgoal_success sequence: float64 - name: normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_discounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_undiscounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_undiscounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: dense_rewards_action_failure sequence: float64 - name: dense_rtgs_action_failure sequence: float64 - name: dense_rewards sequence: float64 - name: dense_rtgs sequence: float64 - name: subgoal_successes sequence: int64 - name: subgoals dtype: int64 - name: action_failures sequence: int64 - name: max_steps dtype: int64 - name: episode_len dtype: int64 splits: - name: train num_bytes: 4061062581 num_examples: 819 - name: val num_bytes: 1016497514 num_examples: 205 download_size: 29738853 dataset_size: 5077560095 - config_name: putnextsimilardistsuniquegoalstrictlycolortype_2048x1_preprocessed_sentence-transformer-pretrained_cnn-scratch features: - name: task_id dtype: int64 - name: timesteps sequence: int64 - name: actions sequence: int64 - name: img_obss sequence: array3_d: shape: - 3 - 56 - 56 dtype: float64 - name: terminations sequence: int64 - name: discounted_rewards sequence: float64 - name: undiscounted_rewards sequence: float64 - name: normalised_discounted_rewards sequence: float64 - name: discounted_rtgs_discounted_rewards sequence: float64 - name: undiscounted_rtgs_discounted_rewards sequence: float64 - name: discounted_rtgs_undiscounted_rewards sequence: float64 - name: undiscounted_rtgs_undiscounted_rewards sequence: float64 - name: discounted_rtgs_normalised_discounted_rewards sequence: float64 - name: undiscounted_rtgs_normalised_discounted_rewards sequence: float64 - name: discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_dense_rewards_subgoal_success sequence: float64 - name: normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_discounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_undiscounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_undiscounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: dense_rewards_action_failure sequence: float64 - name: dense_rtgs_action_failure sequence: float64 - name: dense_rewards sequence: float64 - name: dense_rtgs sequence: float64 - name: mission_embeddings sequence: sequence: float32 - name: mission_token_mask sequence: bool - name: rule_feedback_embeddings sequence: sequence: sequence: float32 - name: rule_feedback_token_mask sequence: sequence: bool - name: rule_feedback_available sequence: bool - name: task_feedback_embeddings sequence: sequence: sequence: float32 - name: task_feedback_token_mask sequence: sequence: bool - name: task_feedback_available sequence: bool - name: instruction_embeddings sequence: sequence: sequence: float32 - name: instruction_token_mask sequence: sequence: bool - name: instructions_available sequence: bool splits: - name: train num_bytes: 9098338088 num_examples: 1638 - name: val num_bytes: 2277361848 num_examples: 410 download_size: 278970546 dataset_size: 11375699936 - config_name: putnextsimilardistsuniquegoalstrictlycolortype_2048x1_raw features: - name: task_id dtype: int64 - name: episode_name dtype: string - name: mission dtype: string - name: timesteps sequence: int64 - name: actions sequence: int64 - name: img_obss sequence: array3_d: shape: - 3 - 56 - 56 dtype: float64 - name: task_feedback sequence: string - name: rule_feedback sequence: string - name: instructions sequence: string - name: terminations sequence: int64 - name: discounted_rewards sequence: float64 - name: undiscounted_rewards sequence: float64 - name: normalised_discounted_rewards sequence: float64 - name: discounted_rtgs_discounted_rewards sequence: float64 - name: undiscounted_rtgs_discounted_rewards sequence: float64 - name: discounted_rtgs_undiscounted_rewards sequence: float64 - name: undiscounted_rtgs_undiscounted_rewards sequence: float64 - name: discounted_rtgs_normalised_discounted_rewards sequence: float64 - name: undiscounted_rtgs_normalised_discounted_rewards sequence: float64 - name: discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_dense_rewards_subgoal_success sequence: float64 - name: normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_discounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_undiscounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_undiscounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: dense_rewards_action_failure sequence: float64 - name: dense_rtgs_action_failure sequence: float64 - name: dense_rewards sequence: float64 - name: dense_rtgs sequence: float64 - name: subgoal_successes sequence: int64 - name: subgoals dtype: int64 - name: action_failures sequence: int64 - name: max_steps dtype: int64 - name: episode_len dtype: int64 splits: - name: train num_bytes: 8122097938 num_examples: 1638 - name: val num_bytes: 2032994680 num_examples: 410 download_size: 59355824 dataset_size: 10155092618 - config_name: putnextsimilardistsuniquegoalstrictlycolortype_4096x1_preprocessed_sentence-transformer-pretrained_cnn-scratch features: - name: task_id dtype: int64 - name: timesteps sequence: int64 - name: actions sequence: int64 - name: img_obss sequence: array3_d: shape: - 3 - 56 - 56 dtype: float64 - name: terminations sequence: int64 - name: discounted_rewards sequence: float64 - name: undiscounted_rewards sequence: float64 - name: normalised_discounted_rewards sequence: float64 - name: discounted_rtgs_discounted_rewards sequence: float64 - name: undiscounted_rtgs_discounted_rewards sequence: float64 - name: discounted_rtgs_undiscounted_rewards sequence: float64 - name: undiscounted_rtgs_undiscounted_rewards sequence: float64 - name: discounted_rtgs_normalised_discounted_rewards sequence: float64 - name: undiscounted_rtgs_normalised_discounted_rewards sequence: float64 - name: discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_dense_rewards_subgoal_success sequence: float64 - name: normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_discounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_undiscounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_undiscounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: dense_rewards_action_failure sequence: float64 - name: dense_rtgs_action_failure sequence: float64 - name: dense_rewards sequence: float64 - name: dense_rtgs sequence: float64 - name: mission_embeddings sequence: sequence: float32 - name: mission_token_mask sequence: bool - name: rule_feedback_embeddings sequence: sequence: sequence: float32 - name: rule_feedback_token_mask sequence: sequence: bool - name: rule_feedback_available sequence: bool - name: task_feedback_embeddings sequence: sequence: sequence: float32 - name: task_feedback_token_mask sequence: sequence: bool - name: task_feedback_available sequence: bool - name: instruction_embeddings sequence: sequence: sequence: float32 - name: instruction_token_mask sequence: sequence: bool - name: instructions_available sequence: bool splits: - name: train num_bytes: 18196675952 num_examples: 3276 - name: val num_bytes: 4554723696 num_examples: 820 download_size: 571868106 dataset_size: 22751399648 - config_name: putnextsimilardistsuniquegoalstrictlycolortype_4096x1_raw features: - name: task_id dtype: int64 - name: episode_name dtype: string - name: mission dtype: string - name: timesteps sequence: int64 - name: actions sequence: int64 - name: img_obss sequence: array3_d: shape: - 3 - 56 - 56 dtype: float64 - name: task_feedback sequence: string - name: rule_feedback sequence: string - name: instructions sequence: string - name: terminations sequence: int64 - name: discounted_rewards sequence: float64 - name: undiscounted_rewards sequence: float64 - name: normalised_discounted_rewards sequence: float64 - name: discounted_rtgs_discounted_rewards sequence: float64 - name: undiscounted_rtgs_discounted_rewards sequence: float64 - name: discounted_rtgs_undiscounted_rewards sequence: float64 - name: undiscounted_rtgs_undiscounted_rewards sequence: float64 - name: discounted_rtgs_normalised_discounted_rewards sequence: float64 - name: undiscounted_rtgs_normalised_discounted_rewards sequence: float64 - name: discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_dense_rewards_subgoal_success sequence: float64 - name: normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_discounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_undiscounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_undiscounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: dense_rewards_action_failure sequence: float64 - name: dense_rtgs_action_failure sequence: float64 - name: dense_rewards sequence: float64 - name: dense_rtgs sequence: float64 - name: subgoal_successes sequence: int64 - name: subgoals dtype: int64 - name: action_failures sequence: int64 - name: max_steps dtype: int64 - name: episode_len dtype: int64 splits: - name: train num_bytes: 16244190413 num_examples: 3276 - name: val num_bytes: 4066009619 num_examples: 820 download_size: 118662382 dataset_size: 20310200032 - config_name: putnextsimilardistsuniquegoalstrictlyloc_1024x1_preprocessed_sentence-transformer-pretrained_cnn-scratch features: - name: task_id dtype: int64 - name: timesteps sequence: int64 - name: actions sequence: int64 - name: img_obss sequence: array3_d: shape: - 3 - 56 - 56 dtype: float64 - name: terminations sequence: int64 - name: discounted_rewards sequence: float64 - name: undiscounted_rewards sequence: float64 - name: normalised_discounted_rewards sequence: float64 - name: discounted_rtgs_discounted_rewards sequence: float64 - name: undiscounted_rtgs_discounted_rewards sequence: float64 - name: discounted_rtgs_undiscounted_rewards sequence: float64 - name: undiscounted_rtgs_undiscounted_rewards sequence: float64 - name: discounted_rtgs_normalised_discounted_rewards sequence: float64 - name: undiscounted_rtgs_normalised_discounted_rewards sequence: float64 - name: discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_dense_rewards_subgoal_success sequence: float64 - name: normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_discounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_undiscounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_undiscounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: dense_rewards_action_failure sequence: float64 - name: dense_rtgs_action_failure sequence: float64 - name: dense_rewards sequence: float64 - name: dense_rtgs sequence: float64 - name: mission_embeddings sequence: sequence: float32 - name: mission_token_mask sequence: bool - name: rule_feedback_embeddings sequence: sequence: sequence: float32 - name: rule_feedback_token_mask sequence: sequence: bool - name: rule_feedback_available sequence: bool - name: task_feedback_embeddings sequence: sequence: sequence: float32 - name: task_feedback_token_mask sequence: sequence: bool - name: task_feedback_available sequence: bool - name: instruction_embeddings sequence: sequence: sequence: float32 - name: instruction_token_mask sequence: sequence: bool - name: instructions_available sequence: bool splits: - name: train num_bytes: 4549169156 num_examples: 819 - name: val num_bytes: 1138680924 num_examples: 205 download_size: 140124628 dataset_size: 5687850080 - config_name: putnextsimilardistsuniquegoalstrictlyloc_1024x1_raw features: - name: task_id dtype: int64 - name: episode_name dtype: string - name: mission dtype: string - name: timesteps sequence: int64 - name: actions sequence: int64 - name: img_obss sequence: array3_d: shape: - 3 - 56 - 56 dtype: float64 - name: task_feedback sequence: string - name: rule_feedback sequence: string - name: instructions sequence: string - name: terminations sequence: int64 - name: discounted_rewards sequence: float64 - name: undiscounted_rewards sequence: float64 - name: normalised_discounted_rewards sequence: float64 - name: discounted_rtgs_discounted_rewards sequence: float64 - name: undiscounted_rtgs_discounted_rewards sequence: float64 - name: discounted_rtgs_undiscounted_rewards sequence: float64 - name: undiscounted_rtgs_undiscounted_rewards sequence: float64 - name: discounted_rtgs_normalised_discounted_rewards sequence: float64 - name: undiscounted_rtgs_normalised_discounted_rewards sequence: float64 - name: discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_dense_rewards_subgoal_success sequence: float64 - name: normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_discounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_undiscounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_undiscounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: dense_rewards_action_failure sequence: float64 - name: dense_rtgs_action_failure sequence: float64 - name: dense_rewards sequence: float64 - name: dense_rtgs sequence: float64 - name: subgoal_successes sequence: int64 - name: subgoals dtype: int64 - name: action_failures sequence: int64 - name: max_steps dtype: int64 - name: episode_len dtype: int64 splits: - name: train num_bytes: 4061168106 num_examples: 819 - name: val num_bytes: 1016535267 num_examples: 205 download_size: 29709408 dataset_size: 5077703373 - config_name: putnextsimilardistsuniquegoalstrictlyloc_2048x1_preprocessed_sentence-transformer-pretrained_cnn-scratch features: - name: task_id dtype: int64 - name: timesteps sequence: int64 - name: actions sequence: int64 - name: img_obss sequence: array3_d: shape: - 3 - 56 - 56 dtype: float64 - name: terminations sequence: int64 - name: discounted_rewards sequence: float64 - name: undiscounted_rewards sequence: float64 - name: normalised_discounted_rewards sequence: float64 - name: discounted_rtgs_discounted_rewards sequence: float64 - name: undiscounted_rtgs_discounted_rewards sequence: float64 - name: discounted_rtgs_undiscounted_rewards sequence: float64 - name: undiscounted_rtgs_undiscounted_rewards sequence: float64 - name: discounted_rtgs_normalised_discounted_rewards sequence: float64 - name: undiscounted_rtgs_normalised_discounted_rewards sequence: float64 - name: discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_dense_rewards_subgoal_success sequence: float64 - name: normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_discounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_undiscounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_undiscounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: dense_rewards_action_failure sequence: float64 - name: dense_rtgs_action_failure sequence: float64 - name: dense_rewards sequence: float64 - name: dense_rtgs sequence: float64 - name: mission_embeddings sequence: sequence: float32 - name: mission_token_mask sequence: bool - name: rule_feedback_embeddings sequence: sequence: sequence: float32 - name: rule_feedback_token_mask sequence: sequence: bool - name: rule_feedback_available sequence: bool - name: task_feedback_embeddings sequence: sequence: sequence: float32 - name: task_feedback_token_mask sequence: sequence: bool - name: task_feedback_available sequence: bool - name: instruction_embeddings sequence: sequence: sequence: float32 - name: instruction_token_mask sequence: sequence: bool - name: instructions_available sequence: bool splits: - name: train num_bytes: 9098338088 num_examples: 1638 - name: val num_bytes: 2277361848 num_examples: 410 download_size: 289943884 dataset_size: 11375699936 - config_name: putnextsimilardistsuniquegoalstrictlyloc_2048x1_raw features: - name: task_id dtype: int64 - name: episode_name dtype: string - name: mission dtype: string - name: timesteps sequence: int64 - name: actions sequence: int64 - name: img_obss sequence: array3_d: shape: - 3 - 56 - 56 dtype: float64 - name: task_feedback sequence: string - name: rule_feedback sequence: string - name: instructions sequence: string - name: terminations sequence: int64 - name: discounted_rewards sequence: float64 - name: undiscounted_rewards sequence: float64 - name: normalised_discounted_rewards sequence: float64 - name: discounted_rtgs_discounted_rewards sequence: float64 - name: undiscounted_rtgs_discounted_rewards sequence: float64 - name: discounted_rtgs_undiscounted_rewards sequence: float64 - name: undiscounted_rtgs_undiscounted_rewards sequence: float64 - name: discounted_rtgs_normalised_discounted_rewards sequence: float64 - name: undiscounted_rtgs_normalised_discounted_rewards sequence: float64 - name: discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_dense_rewards_subgoal_success sequence: float64 - name: normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_discounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_undiscounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_undiscounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: dense_rewards_action_failure sequence: float64 - name: dense_rtgs_action_failure sequence: float64 - name: dense_rewards sequence: float64 - name: dense_rtgs sequence: float64 - name: subgoal_successes sequence: int64 - name: subgoals dtype: int64 - name: action_failures sequence: int64 - name: max_steps dtype: int64 - name: episode_len dtype: int64 splits: - name: train num_bytes: 8122337518 num_examples: 1638 - name: val num_bytes: 2033091635 num_examples: 410 download_size: 59298897 dataset_size: 10155429153 - config_name: putnextsimilardistsuniquegoalstrictlyonlycolor_1024x1_preprocessed_sentence-transformer-pretrained_cnn-scratch features: - name: task_id dtype: int64 - name: timesteps sequence: int64 - name: actions sequence: int64 - name: img_obss sequence: array3_d: shape: - 3 - 56 - 56 dtype: float64 - name: terminations sequence: int64 - name: discounted_rewards sequence: float64 - name: undiscounted_rewards sequence: float64 - name: normalised_discounted_rewards sequence: float64 - name: discounted_rtgs_discounted_rewards sequence: float64 - name: undiscounted_rtgs_discounted_rewards sequence: float64 - name: discounted_rtgs_undiscounted_rewards sequence: float64 - name: undiscounted_rtgs_undiscounted_rewards sequence: float64 - name: discounted_rtgs_normalised_discounted_rewards sequence: float64 - name: undiscounted_rtgs_normalised_discounted_rewards sequence: float64 - name: discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_dense_rewards_subgoal_success sequence: float64 - name: normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_discounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_undiscounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_undiscounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: dense_rewards_action_failure sequence: float64 - name: dense_rtgs_action_failure sequence: float64 - name: dense_rewards sequence: float64 - name: dense_rtgs sequence: float64 - name: mission_embeddings sequence: sequence: float32 - name: mission_token_mask sequence: bool - name: rule_feedback_embeddings sequence: sequence: sequence: float32 - name: rule_feedback_token_mask sequence: sequence: bool - name: rule_feedback_available sequence: bool - name: task_feedback_embeddings sequence: sequence: sequence: float32 - name: task_feedback_token_mask sequence: sequence: bool - name: task_feedback_available sequence: bool - name: instruction_embeddings sequence: sequence: sequence: float32 - name: instruction_token_mask sequence: sequence: bool - name: instructions_available sequence: bool splits: - name: train num_bytes: 4549169156 num_examples: 819 - name: val num_bytes: 1138680924 num_examples: 205 download_size: 126354529 dataset_size: 5687850080 - config_name: putnextsimilardistsuniquegoalstrictlyonlycolor_1024x1_raw features: - name: task_id dtype: int64 - name: episode_name dtype: string - name: mission dtype: string - name: timesteps sequence: int64 - name: actions sequence: int64 - name: img_obss sequence: array3_d: shape: - 3 - 56 - 56 dtype: float64 - name: task_feedback sequence: string - name: rule_feedback sequence: string - name: instructions sequence: string - name: terminations sequence: int64 - name: discounted_rewards sequence: float64 - name: undiscounted_rewards sequence: float64 - name: normalised_discounted_rewards sequence: float64 - name: discounted_rtgs_discounted_rewards sequence: float64 - name: undiscounted_rtgs_discounted_rewards sequence: float64 - name: discounted_rtgs_undiscounted_rewards sequence: float64 - name: undiscounted_rtgs_undiscounted_rewards sequence: float64 - name: discounted_rtgs_normalised_discounted_rewards sequence: float64 - name: undiscounted_rtgs_normalised_discounted_rewards sequence: float64 - name: discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_dense_rewards_subgoal_success sequence: float64 - name: normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_discounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_undiscounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_undiscounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: dense_rewards_action_failure sequence: float64 - name: dense_rtgs_action_failure sequence: float64 - name: dense_rewards sequence: float64 - name: dense_rtgs sequence: float64 - name: subgoal_successes sequence: int64 - name: subgoals dtype: int64 - name: action_failures sequence: int64 - name: max_steps dtype: int64 - name: episode_len dtype: int64 splits: - name: train num_bytes: 4061104591 num_examples: 819 - name: val num_bytes: 1016517191 num_examples: 205 download_size: 29687880 dataset_size: 5077621782 - config_name: putnextsimilardistsuniquegoalstrictlyonlycolor_2048x1_preprocessed_sentence-transformer-pretrained_cnn-scratch features: - name: task_id dtype: int64 - name: timesteps sequence: int64 - name: actions sequence: int64 - name: img_obss sequence: array3_d: shape: - 3 - 56 - 56 dtype: float64 - name: terminations sequence: int64 - name: discounted_rewards sequence: float64 - name: undiscounted_rewards sequence: float64 - name: normalised_discounted_rewards sequence: float64 - name: discounted_rtgs_discounted_rewards sequence: float64 - name: undiscounted_rtgs_discounted_rewards sequence: float64 - name: discounted_rtgs_undiscounted_rewards sequence: float64 - name: undiscounted_rtgs_undiscounted_rewards sequence: float64 - name: discounted_rtgs_normalised_discounted_rewards sequence: float64 - name: undiscounted_rtgs_normalised_discounted_rewards sequence: float64 - name: discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_dense_rewards_subgoal_success sequence: float64 - name: normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_discounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_undiscounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_undiscounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: dense_rewards_action_failure sequence: float64 - name: dense_rtgs_action_failure sequence: float64 - name: dense_rewards sequence: float64 - name: dense_rtgs sequence: float64 - name: mission_embeddings sequence: sequence: float32 - name: mission_token_mask sequence: bool - name: rule_feedback_embeddings sequence: sequence: sequence: float32 - name: rule_feedback_token_mask sequence: sequence: bool - name: rule_feedback_available sequence: bool - name: task_feedback_embeddings sequence: sequence: sequence: float32 - name: task_feedback_token_mask sequence: sequence: bool - name: task_feedback_available sequence: bool - name: instruction_embeddings sequence: sequence: sequence: float32 - name: instruction_token_mask sequence: sequence: bool - name: instructions_available sequence: bool splits: - name: train num_bytes: 9098338088 num_examples: 1638 - name: val num_bytes: 2277361848 num_examples: 410 download_size: 255531138 dataset_size: 11375699936 - config_name: putnextsimilardistsuniquegoalstrictlyonlycolor_2048x1_raw features: - name: task_id dtype: int64 - name: episode_name dtype: string - name: mission dtype: string - name: timesteps sequence: int64 - name: actions sequence: int64 - name: img_obss sequence: array3_d: shape: - 3 - 56 - 56 dtype: float64 - name: task_feedback sequence: string - name: rule_feedback sequence: string - name: instructions sequence: string - name: terminations sequence: int64 - name: discounted_rewards sequence: float64 - name: undiscounted_rewards sequence: float64 - name: normalised_discounted_rewards sequence: float64 - name: discounted_rtgs_discounted_rewards sequence: float64 - name: undiscounted_rtgs_discounted_rewards sequence: float64 - name: discounted_rtgs_undiscounted_rewards sequence: float64 - name: undiscounted_rtgs_undiscounted_rewards sequence: float64 - name: discounted_rtgs_normalised_discounted_rewards sequence: float64 - name: undiscounted_rtgs_normalised_discounted_rewards sequence: float64 - name: discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_dense_rewards_subgoal_success sequence: float64 - name: normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_discounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_undiscounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_undiscounted_dense_rewards_subgoal_success sequence: float64 - name: discounted_rtgs_normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: undiscounted_rtgs_normalised_discounted_dense_rewards_subgoal_success sequence: float64 - name: dense_rewards_action_failure sequence: float64 - name: dense_rtgs_action_failure sequence: float64 - name: dense_rewards sequence: float64 - name: dense_rtgs sequence: float64 - name: subgoal_successes sequence: int64 - name: subgoals dtype: int64 - name: action_failures sequence: int64 - name: max_steps dtype: int64 - name: episode_len dtype: int64 splits: - name: train num_bytes: 8122214892 num_examples: 1638 - name: val num_bytes: 2033032914 num_examples: 410 download_size: 59281751 dataset_size: 10155247806 configs: - config_name: gotosimilardistsuniquegoalstrictlycolortype_1024x1_preprocessed_sentence-transformer-pretrained_cnn-scratch data_files: - split: train path: gotosimilardistsuniquegoalstrictlycolortype/1024x1/preprocessed/sentence-transformer-pretrained/cnn-scratch/train-* - split: val path: gotosimilardistsuniquegoalstrictlycolortype/1024x1/preprocessed/sentence-transformer-pretrained/cnn-scratch/val-* - config_name: gotosimilardistsuniquegoalstrictlycolortype_1024x1_raw data_files: - split: train path: gotosimilardistsuniquegoalstrictlycolortype/1024x1/raw/train-* - split: val path: gotosimilardistsuniquegoalstrictlycolortype/1024x1/raw/val-* - config_name: gotosimilardistsuniquegoalstrictlycolortype_2048x1_preprocessed_sentence-transformer-pretrained_cnn-scratch data_files: - split: train path: gotosimilardistsuniquegoalstrictlycolortype/2048x1/preprocessed/sentence-transformer-pretrained/cnn-scratch/train-* - split: val path: gotosimilardistsuniquegoalstrictlycolortype/2048x1/preprocessed/sentence-transformer-pretrained/cnn-scratch/val-* - config_name: gotosimilardistsuniquegoalstrictlycolortype_2048x1_raw data_files: - split: train path: gotosimilardistsuniquegoalstrictlycolortype/2048x1/raw/train-* - split: val path: gotosimilardistsuniquegoalstrictlycolortype/2048x1/raw/val-* - config_name: gotosimilardistsuniquegoalstrictlycolortype_4096x1_preprocessed_sentence-transformer-pretrained_cnn-scratch data_files: - split: train path: gotosimilardistsuniquegoalstrictlycolortype/4096x1/preprocessed/sentence-transformer-pretrained/cnn-scratch/train-* - split: val path: gotosimilardistsuniquegoalstrictlycolortype/4096x1/preprocessed/sentence-transformer-pretrained/cnn-scratch/val-* - config_name: gotosimilardistsuniquegoalstrictlycolortype_4096x1_raw data_files: - split: train path: gotosimilardistsuniquegoalstrictlycolortype/4096x1/raw/train-* - split: val path: gotosimilardistsuniquegoalstrictlycolortype/4096x1/raw/val-* - config_name: gotosimilardistsuniquegoalstrictlyloc_1024x1_preprocessed_sentence-transformer-pretrained_cnn-scratch data_files: - split: train path: gotosimilardistsuniquegoalstrictlyloc/1024x1/preprocessed/sentence-transformer-pretrained/cnn-scratch/train-* - split: val path: gotosimilardistsuniquegoalstrictlyloc/1024x1/preprocessed/sentence-transformer-pretrained/cnn-scratch/val-* - config_name: gotosimilardistsuniquegoalstrictlyloc_2048x1_preprocessed_sentence-transformer-pretrained_cnn-scratch data_files: - split: train path: gotosimilardistsuniquegoalstrictlyloc/2048x1/preprocessed/sentence-transformer-pretrained/cnn-scratch/train-* - split: val path: gotosimilardistsuniquegoalstrictlyloc/2048x1/preprocessed/sentence-transformer-pretrained/cnn-scratch/val-* - config_name: gotosimilardistsuniquegoalstrictlyloc_2048x1_raw data_files: - split: train path: gotosimilardistsuniquegoalstrictlyloc/2048x1/raw/train-* - split: val path: gotosimilardistsuniquegoalstrictlyloc/2048x1/raw/val-* - config_name: gotosimilardistsuniquegoalstrictlyonlycolor_1024x1_preprocessed_sentence-transformer-pretrained_cnn-scratch data_files: - split: train path: gotosimilardistsuniquegoalstrictlyonlycolor/1024x1/preprocessed/sentence-transformer-pretrained/cnn-scratch/train-* - split: val path: gotosimilardistsuniquegoalstrictlyonlycolor/1024x1/preprocessed/sentence-transformer-pretrained/cnn-scratch/val-* - config_name: gotosimilardistsuniquegoalstrictlyonlycolor_1024x1_raw data_files: - split: train path: gotosimilardistsuniquegoalstrictlyonlycolor/1024x1/raw/train-* - split: val path: gotosimilardistsuniquegoalstrictlyonlycolor/1024x1/raw/val-* - config_name: gotosimilardistsuniquegoalstrictlyonlycolor_2048x1_preprocessed_sentence-transformer-pretrained_cnn-scratch data_files: - split: train path: gotosimilardistsuniquegoalstrictlyonlycolor/2048x1/preprocessed/sentence-transformer-pretrained/cnn-scratch/train-* - split: val path: gotosimilardistsuniquegoalstrictlyonlycolor/2048x1/preprocessed/sentence-transformer-pretrained/cnn-scratch/val-* - config_name: gotosimilardistsuniquegoalstrictlyonlycolor_2048x1_raw data_files: - split: train path: gotosimilardistsuniquegoalstrictlyonlycolor/2048x1/raw/train-* - split: val path: gotosimilardistsuniquegoalstrictlyonlycolor/2048x1/raw/val-* - config_name: pickup_1024x1_preprocessed_sentence-transformer-pretrained_cnn-scratch data_files: - split: train path: pickup/1024x1/preprocessed/sentence-transformer-pretrained/cnn-scratch/train-* - split: val path: pickup/1024x1/preprocessed/sentence-transformer-pretrained/cnn-scratch/val-* - config_name: pickup_2048x1_preprocessed_sentence-transformer-pretrained_cnn-scratch data_files: - split: train path: pickup/2048x1/preprocessed/sentence-transformer-pretrained/cnn-scratch/train-* - split: val path: pickup/2048x1/preprocessed/sentence-transformer-pretrained/cnn-scratch/val-* - config_name: pickup_2048x1_raw data_files: - split: train path: pickup/2048x1/raw/train-* - split: val path: pickup/2048x1/raw/val-* - config_name: pickupeasyseqafter_1024x1_preprocessed_sentence-transformer-pretrained_cnn-scratch data_files: - split: train path: pickupeasyseqafter/1024x1/preprocessed/sentence-transformer-pretrained/cnn-scratch/train-* - split: val path: pickupeasyseqafter/1024x1/preprocessed/sentence-transformer-pretrained/cnn-scratch/val-* - config_name: pickupeasyseqafter_1024x1_raw data_files: - split: train path: pickupeasyseqafter/1024x1/raw/train-* - split: val path: pickupeasyseqafter/1024x1/raw/val-* - config_name: pickupeasyseqafter_2048x1_preprocessed_sentence-transformer-pretrained_cnn-scratch data_files: - split: train path: pickupeasyseqafter/2048x1/preprocessed/sentence-transformer-pretrained/cnn-scratch/train-* - split: val path: pickupeasyseqafter/2048x1/preprocessed/sentence-transformer-pretrained/cnn-scratch/val-* - config_name: pickupeasyseqafter_2048x1_raw data_files: - split: train path: pickupeasyseqafter/2048x1/raw/train-* - split: val path: pickupeasyseqafter/2048x1/raw/val-* - config_name: pickupeasyseqand_1024x1_preprocessed_sentence-transformer-pretrained_cnn-scratch data_files: - split: train path: pickupeasyseqand/1024x1/preprocessed/sentence-transformer-pretrained/cnn-scratch/train-* - split: val path: pickupeasyseqand/1024x1/preprocessed/sentence-transformer-pretrained/cnn-scratch/val-* - config_name: pickupeasyseqand_1024x1_raw data_files: - split: train path: pickupeasyseqand/1024x1/raw/train-* - split: val path: pickupeasyseqand/1024x1/raw/val-* - config_name: pickupeasyseqand_2048x1_preprocessed_sentence-transformer-pretrained_cnn-scratch data_files: - split: train path: pickupeasyseqand/2048x1/preprocessed/sentence-transformer-pretrained/cnn-scratch/train-* - split: val path: pickupeasyseqand/2048x1/preprocessed/sentence-transformer-pretrained/cnn-scratch/val-* - config_name: pickupeasyseqand_2048x1_raw data_files: - split: train path: pickupeasyseqand/2048x1/raw/train-* - split: val path: pickupeasyseqand/2048x1/raw/val-* - config_name: pickupeasyseqbefore_1024x1_preprocessed_sentence-transformer-pretrained_cnn-scratch data_files: - split: train path: pickupeasyseqbefore/1024x1/preprocessed/sentence-transformer-pretrained/cnn-scratch/train-* - split: val path: pickupeasyseqbefore/1024x1/preprocessed/sentence-transformer-pretrained/cnn-scratch/val-* - config_name: pickupeasyseqbefore_2048x1_preprocessed_sentence-transformer-pretrained_cnn-scratch data_files: - split: train path: pickupeasyseqbefore/2048x1/preprocessed/sentence-transformer-pretrained/cnn-scratch/train-* - split: val path: pickupeasyseqbefore/2048x1/preprocessed/sentence-transformer-pretrained/cnn-scratch/val-* - config_name: pickupeasyseqbefore_2048x1_raw data_files: - split: train path: pickupeasyseqbefore/2048x1/raw/train-* - split: val path: pickupeasyseqbefore/2048x1/raw/val-* - config_name: pickupmaze1x2_1024x1_preprocessed_sentence-transformer-pretrained_cnn-scratch data_files: - split: train path: pickupmaze1x2/1024x1/preprocessed/sentence-transformer-pretrained/cnn-scratch/train-* - split: val path: pickupmaze1x2/1024x1/preprocessed/sentence-transformer-pretrained/cnn-scratch/val-* - config_name: pickupmaze1x2_1024x1_raw data_files: - split: train path: pickupmaze1x2/1024x1/raw/train-* - split: val path: pickupmaze1x2/1024x1/raw/val-* - config_name: pickupmaze1x2_2048x1_preprocessed_sentence-transformer-pretrained_cnn-scratch data_files: - split: train path: pickupmaze1x2/2048x1/preprocessed/sentence-transformer-pretrained/cnn-scratch/train-* - split: val path: pickupmaze1x2/2048x1/preprocessed/sentence-transformer-pretrained/cnn-scratch/val-* - config_name: pickupmaze1x2_2048x1_raw data_files: - split: train path: pickupmaze1x2/2048x1/raw/train-* - split: val path: pickupmaze1x2/2048x1/raw/val-* - config_name: pickupmaze1x2blocked_1024x1_preprocessed_sentence-transformer-pretrained_cnn-scratch data_files: - split: train path: pickupmaze1x2blocked/1024x1/preprocessed/sentence-transformer-pretrained/cnn-scratch/train-* - split: val path: pickupmaze1x2blocked/1024x1/preprocessed/sentence-transformer-pretrained/cnn-scratch/val-* - config_name: pickupmaze1x2blocked_1024x1_raw data_files: - split: train path: pickupmaze1x2blocked/1024x1/raw/train-* - split: val path: pickupmaze1x2blocked/1024x1/raw/val-* - config_name: pickupmaze1x2blocked_2048x1_preprocessed_sentence-transformer-pretrained_cnn-scratch data_files: - split: train path: pickupmaze1x2blocked/2048x1/preprocessed/sentence-transformer-pretrained/cnn-scratch/train-* - split: val path: pickupmaze1x2blocked/2048x1/preprocessed/sentence-transformer-pretrained/cnn-scratch/val-* - config_name: pickupmaze1x2blocked_2048x1_raw data_files: - split: train path: pickupmaze1x2blocked/2048x1/raw/train-* - split: val path: pickupmaze1x2blocked/2048x1/raw/val-* - config_name: pickupmaze1x2doorsclosed_1024x1_preprocessed_sentence-transformer-pretrained_cnn-scratch data_files: - split: train path: pickupmaze1x2doorsclosed/1024x1/preprocessed/sentence-transformer-pretrained/cnn-scratch/train-* - split: val path: pickupmaze1x2doorsclosed/1024x1/preprocessed/sentence-transformer-pretrained/cnn-scratch/val-* - config_name: pickupmaze1x2doorsclosed_1024x1_raw data_files: - split: train path: pickupmaze1x2doorsclosed/1024x1/raw/train-* - split: val path: pickupmaze1x2doorsclosed/1024x1/raw/val-* - config_name: pickupmaze1x2doorsclosed_2048x1_raw data_files: - split: train path: pickupmaze1x2doorsclosed/2048x1/raw/train-* - split: val path: pickupmaze1x2doorsclosed/2048x1/raw/val-* - config_name: pickupmaze1x2doorslocked_1024x1_preprocessed_sentence-transformer-pretrained_cnn-scratch data_files: - split: train path: pickupmaze1x2doorslocked/1024x1/preprocessed/sentence-transformer-pretrained/cnn-scratch/train-* - split: val path: pickupmaze1x2doorslocked/1024x1/preprocessed/sentence-transformer-pretrained/cnn-scratch/val-* - config_name: pickupmaze1x2doorslocked_1024x1_raw data_files: - split: train path: pickupmaze1x2doorslocked/1024x1/raw/train-* - split: val path: pickupmaze1x2doorslocked/1024x1/raw/val-* - config_name: pickupmaze1x2doorslocked_2048x1_raw data_files: - split: train path: pickupmaze1x2doorslocked/2048x1/raw/train-* - split: val path: pickupmaze1x2doorslocked/2048x1/raw/val-* - config_name: pickupmaze2x2_1024x1_preprocessed_sentence-transformer-pretrained_cnn-scratch data_files: - split: train path: pickupmaze2x2/1024x1/preprocessed/sentence-transformer-pretrained/cnn-scratch/train-* - split: val path: pickupmaze2x2/1024x1/preprocessed/sentence-transformer-pretrained/cnn-scratch/val-* - config_name: pickupmaze2x2_1024x1_raw data_files: - split: train path: pickupmaze2x2/1024x1/raw/train-* - split: val path: pickupmaze2x2/1024x1/raw/val-* - config_name: pickupmaze2x2_2048x1_raw data_files: - split: train path: pickupmaze2x2/2048x1/raw/train-* - split: val path: pickupmaze2x2/2048x1/raw/val-* - config_name: pickupmaze2x2blocked_2048x1_raw data_files: - split: train path: pickupmaze2x2blocked/2048x1/raw/train-* - split: val path: pickupmaze2x2blocked/2048x1/raw/val-* - config_name: pickupmaze2x2doorsclosed_1024x1_raw data_files: - split: train path: pickupmaze2x2doorsclosed/1024x1/raw/train-* - split: val path: pickupmaze2x2doorsclosed/1024x1/raw/val-* - config_name: pickupmaze2x2doorsclosed_2048x1_raw data_files: - split: train path: pickupmaze2x2doorsclosed/2048x1/raw/train-* - split: val path: pickupmaze2x2doorsclosed/2048x1/raw/val-* - config_name: pickupmaze2x2doorslocked_1024x1_raw data_files: - split: train path: pickupmaze2x2doorslocked/1024x1/raw/train-* - split: val path: pickupmaze2x2doorslocked/1024x1/raw/val-* - config_name: pickupmaze2x2doorslocked_2048x1_raw data_files: - split: train path: pickupmaze2x2doorslocked/2048x1/raw/train-* - split: val path: pickupmaze2x2doorslocked/2048x1/raw/val-* - config_name: pickups12_1024x1_preprocessed_sentence-transformer-pretrained_cnn-scratch data_files: - split: train path: pickups12/1024x1/preprocessed/sentence-transformer-pretrained/cnn-scratch/train-* - split: val path: pickups12/1024x1/preprocessed/sentence-transformer-pretrained/cnn-scratch/val-* - config_name: pickups12_1024x1_raw data_files: - split: train path: pickups12/1024x1/raw/train-* - split: val path: pickups12/1024x1/raw/val-* - config_name: pickups12_2048x1_preprocessed_sentence-transformer-pretrained_cnn-scratch data_files: - split: train path: pickups12/2048x1/preprocessed/sentence-transformer-pretrained/cnn-scratch/train-* - split: val path: pickups12/2048x1/preprocessed/sentence-transformer-pretrained/cnn-scratch/val-* - config_name: pickups12_2048x1_raw data_files: - split: train path: pickups12/2048x1/raw/train-* - split: val path: pickups12/2048x1/raw/val-* - config_name: pickupsimilardistsuniquegoalstrictlycolortype_1024x1_preprocessed_sentence-transformer-pretrained_cnn-scratch data_files: - split: train path: pickupsimilardistsuniquegoalstrictlycolortype/1024x1/preprocessed/sentence-transformer-pretrained/cnn-scratch/train-* - split: val path: pickupsimilardistsuniquegoalstrictlycolortype/1024x1/preprocessed/sentence-transformer-pretrained/cnn-scratch/val-* - config_name: pickupsimilardistsuniquegoalstrictlycolortype_1024x1_raw data_files: - split: train path: pickupsimilardistsuniquegoalstrictlycolortype/1024x1/raw/train-* - split: val path: pickupsimilardistsuniquegoalstrictlycolortype/1024x1/raw/val-* - config_name: pickupsimilardistsuniquegoalstrictlycolortype_2048x1_preprocessed_sentence-transformer-pretrained_cnn-scratch data_files: - split: train path: pickupsimilardistsuniquegoalstrictlycolortype/2048x1/preprocessed/sentence-transformer-pretrained/cnn-scratch/train-* - split: val path: pickupsimilardistsuniquegoalstrictlycolortype/2048x1/preprocessed/sentence-transformer-pretrained/cnn-scratch/val-* - config_name: pickupsimilardistsuniquegoalstrictlycolortype_2048x1_raw data_files: - split: train path: pickupsimilardistsuniquegoalstrictlycolortype/2048x1/raw/train-* - split: val path: pickupsimilardistsuniquegoalstrictlycolortype/2048x1/raw/val-* - config_name: pickupsimilardistsuniquegoalstrictlycolortype_4096x1_preprocessed_sentence-transformer-pretrained_cnn-scratch data_files: - split: train path: pickupsimilardistsuniquegoalstrictlycolortype/4096x1/preprocessed/sentence-transformer-pretrained/cnn-scratch/train-* - split: val path: pickupsimilardistsuniquegoalstrictlycolortype/4096x1/preprocessed/sentence-transformer-pretrained/cnn-scratch/val-* - config_name: pickupsimilardistsuniquegoalstrictlycolortype_4096x1_raw data_files: - split: train path: pickupsimilardistsuniquegoalstrictlycolortype/4096x1/raw/train-* - split: val path: pickupsimilardistsuniquegoalstrictlycolortype/4096x1/raw/val-* - config_name: pickupsimilardistsuniquegoalstrictlyloc_1024x1_preprocessed_sentence-transformer-pretrained_cnn-scratch data_files: - split: train path: pickupsimilardistsuniquegoalstrictlyloc/1024x1/preprocessed/sentence-transformer-pretrained/cnn-scratch/train-* - split: val path: pickupsimilardistsuniquegoalstrictlyloc/1024x1/preprocessed/sentence-transformer-pretrained/cnn-scratch/val-* - config_name: pickupsimilardistsuniquegoalstrictlyloc_1024x1_raw data_files: - split: train path: pickupsimilardistsuniquegoalstrictlyloc/1024x1/raw/train-* - split: val path: pickupsimilardistsuniquegoalstrictlyloc/1024x1/raw/val-* - config_name: pickupsimilardistsuniquegoalstrictlyloc_2048x1_preprocessed_sentence-transformer-pretrained_cnn-scratch data_files: - split: train path: pickupsimilardistsuniquegoalstrictlyloc/2048x1/preprocessed/sentence-transformer-pretrained/cnn-scratch/train-* - split: val path: pickupsimilardistsuniquegoalstrictlyloc/2048x1/preprocessed/sentence-transformer-pretrained/cnn-scratch/val-* - config_name: pickupsimilardistsuniquegoalstrictlyloc_2048x1_raw data_files: - split: train path: pickupsimilardistsuniquegoalstrictlyloc/2048x1/raw/train-* - split: val path: pickupsimilardistsuniquegoalstrictlyloc/2048x1/raw/val-* - config_name: pickupsimilardistsuniquegoalstrictlyonlycolor_1024x1_preprocessed_sentence-transformer-pretrained_cnn-scratch data_files: - split: train path: pickupsimilardistsuniquegoalstrictlyonlycolor/1024x1/preprocessed/sentence-transformer-pretrained/cnn-scratch/train-* - split: val path: pickupsimilardistsuniquegoalstrictlyonlycolor/1024x1/preprocessed/sentence-transformer-pretrained/cnn-scratch/val-* - config_name: pickupsimilardistsuniquegoalstrictlyonlycolor_1024x1_raw data_files: - split: train path: pickupsimilardistsuniquegoalstrictlyonlycolor/1024x1/raw/train-* - split: val path: pickupsimilardistsuniquegoalstrictlyonlycolor/1024x1/raw/val-* - config_name: pickupsimilardistsuniquegoalstrictlyonlycolor_2048x1_preprocessed_sentence-transformer-pretrained_cnn-scratch data_files: - split: train path: pickupsimilardistsuniquegoalstrictlyonlycolor/2048x1/preprocessed/sentence-transformer-pretrained/cnn-scratch/train-* - split: val path: pickupsimilardistsuniquegoalstrictlyonlycolor/2048x1/preprocessed/sentence-transformer-pretrained/cnn-scratch/val-* - config_name: pickupsimilardistsuniquegoalstrictlyonlycolor_2048x1_raw data_files: - split: train path: pickupsimilardistsuniquegoalstrictlyonlycolor/2048x1/raw/train-* - split: val path: pickupsimilardistsuniquegoalstrictlyonlycolor/2048x1/raw/val-* - config_name: putnextsimilardistsuniquegoalstrictlycolortype_1024x1_preprocessed_sentence-transformer-pretrained_cnn-scratch data_files: - split: train path: putnextsimilardistsuniquegoalstrictlycolortype/1024x1/preprocessed/sentence-transformer-pretrained/cnn-scratch/train-* - split: val path: putnextsimilardistsuniquegoalstrictlycolortype/1024x1/preprocessed/sentence-transformer-pretrained/cnn-scratch/val-* - config_name: putnextsimilardistsuniquegoalstrictlycolortype_1024x1_raw data_files: - split: train path: putnextsimilardistsuniquegoalstrictlycolortype/1024x1/raw/train-* - split: val path: putnextsimilardistsuniquegoalstrictlycolortype/1024x1/raw/val-* - config_name: putnextsimilardistsuniquegoalstrictlycolortype_2048x1_preprocessed_sentence-transformer-pretrained_cnn-scratch data_files: - split: train path: putnextsimilardistsuniquegoalstrictlycolortype/2048x1/preprocessed/sentence-transformer-pretrained/cnn-scratch/train-* - split: val path: putnextsimilardistsuniquegoalstrictlycolortype/2048x1/preprocessed/sentence-transformer-pretrained/cnn-scratch/val-* - config_name: putnextsimilardistsuniquegoalstrictlycolortype_2048x1_raw data_files: - split: train path: putnextsimilardistsuniquegoalstrictlycolortype/2048x1/raw/train-* - split: val path: putnextsimilardistsuniquegoalstrictlycolortype/2048x1/raw/val-* - config_name: putnextsimilardistsuniquegoalstrictlycolortype_4096x1_preprocessed_sentence-transformer-pretrained_cnn-scratch data_files: - split: train path: putnextsimilardistsuniquegoalstrictlycolortype/4096x1/preprocessed/sentence-transformer-pretrained/cnn-scratch/train-* - split: val path: putnextsimilardistsuniquegoalstrictlycolortype/4096x1/preprocessed/sentence-transformer-pretrained/cnn-scratch/val-* - config_name: putnextsimilardistsuniquegoalstrictlycolortype_4096x1_raw data_files: - split: train path: putnextsimilardistsuniquegoalstrictlycolortype/4096x1/raw/train-* - split: val path: putnextsimilardistsuniquegoalstrictlycolortype/4096x1/raw/val-* - config_name: putnextsimilardistsuniquegoalstrictlyloc_1024x1_preprocessed_sentence-transformer-pretrained_cnn-scratch data_files: - split: train path: putnextsimilardistsuniquegoalstrictlyloc/1024x1/preprocessed/sentence-transformer-pretrained/cnn-scratch/train-* - split: val path: putnextsimilardistsuniquegoalstrictlyloc/1024x1/preprocessed/sentence-transformer-pretrained/cnn-scratch/val-* - config_name: putnextsimilardistsuniquegoalstrictlyloc_1024x1_raw data_files: - split: train path: putnextsimilardistsuniquegoalstrictlyloc/1024x1/raw/train-* - split: val path: putnextsimilardistsuniquegoalstrictlyloc/1024x1/raw/val-* - config_name: putnextsimilardistsuniquegoalstrictlyloc_2048x1_preprocessed_sentence-transformer-pretrained_cnn-scratch data_files: - split: train path: putnextsimilardistsuniquegoalstrictlyloc/2048x1/preprocessed/sentence-transformer-pretrained/cnn-scratch/train-* - split: val path: putnextsimilardistsuniquegoalstrictlyloc/2048x1/preprocessed/sentence-transformer-pretrained/cnn-scratch/val-* - config_name: putnextsimilardistsuniquegoalstrictlyloc_2048x1_raw data_files: - split: train path: putnextsimilardistsuniquegoalstrictlyloc/2048x1/raw/train-* - split: val path: putnextsimilardistsuniquegoalstrictlyloc/2048x1/raw/val-* - config_name: putnextsimilardistsuniquegoalstrictlyonlycolor_1024x1_preprocessed_sentence-transformer-pretrained_cnn-scratch data_files: - split: train path: putnextsimilardistsuniquegoalstrictlyonlycolor/1024x1/preprocessed/sentence-transformer-pretrained/cnn-scratch/train-* - split: val path: putnextsimilardistsuniquegoalstrictlyonlycolor/1024x1/preprocessed/sentence-transformer-pretrained/cnn-scratch/val-* - config_name: putnextsimilardistsuniquegoalstrictlyonlycolor_1024x1_raw data_files: - split: train path: putnextsimilardistsuniquegoalstrictlyonlycolor/1024x1/raw/train-* - split: val path: putnextsimilardistsuniquegoalstrictlyonlycolor/1024x1/raw/val-* - config_name: putnextsimilardistsuniquegoalstrictlyonlycolor_2048x1_preprocessed_sentence-transformer-pretrained_cnn-scratch data_files: - split: train path: putnextsimilardistsuniquegoalstrictlyonlycolor/2048x1/preprocessed/sentence-transformer-pretrained/cnn-scratch/train-* - split: val path: putnextsimilardistsuniquegoalstrictlyonlycolor/2048x1/preprocessed/sentence-transformer-pretrained/cnn-scratch/val-* - config_name: putnextsimilardistsuniquegoalstrictlyonlycolor_2048x1_raw data_files: - split: train path: putnextsimilardistsuniquegoalstrictlyonlycolor/2048x1/raw/train-* - split: val path: putnextsimilardistsuniquegoalstrictlyonlycolor/2048x1/raw/val-* ---### 数据集信息 本数据集包含多个强化学习任务配置版本，各配置详情如下： --- #### 配置组1：前往相似距离唯一目标严格颜色类型（gotosimilardistsuniquegoalstrictlycolortype） ##### 子配置1.1：`gotosimilardistsuniquegoalstrictlycolortype_1024x1_preprocessed_sentence-transformer-pretrained_cnn-scratch` （术语说明：sentence-transformer为句子转换器，pretrained为预训练，cnn-scratch为从零初始化训练的卷积神经网络） **特征字段**： 1. 任务ID（task_id）：64位整数类型 2. 时间步（timesteps）：64位整数序列 3. 动作（actions）：64位整数序列 4. 图像观测（img_obss）：三维浮点序列，单数组形状为(3, 56, 56)，数据类型为64位浮点数 5. 终止标志（terminations）：64位整数序列 6. 折扣奖励（discounted_rewards）：64位浮点数序列 7. 未折扣奖励（undiscounted_rewards）：64位浮点数序列 8. 归一化折扣奖励（normalised_discounted_rewards）：64位浮点数序列 9. 基于折扣奖励的回报回溯（discounted_rtgs_discounted_rewards，RTG即Return to Go，回报回溯）：64位浮点数序列 10. 基于折扣奖励的未折扣回报回溯（undiscounted_rtgs_discounted_rewards）：64位浮点数序列 11. 基于未折扣奖励的回报回溯（discounted_rtgs_undiscounted_rewards）：64位浮点数序列 12. 基于未折扣奖励的未折扣回报回溯（undiscounted_rtgs_undiscounted_rewards）：64位浮点数序列 13. 基于归一化折扣奖励的回报回溯（discounted_rtgs_normalised_discounted_rewards）：64位浮点数序列 14. 基于归一化折扣奖励的未折扣回报回溯（undiscounted_rtgs_normalised_discounted_rewards）：64位浮点数序列 15. 子目标成功相关稠密折扣奖励（discounted_dense_rewards_subgoal_success）：64位浮点数序列 16. 子目标成功相关稠密未折扣奖励（undiscounted_dense_rewards_subgoal_success）：64位浮点数序列 17. 子目标成功相关归一化稠密折扣奖励（normalised_discounted_dense_rewards_subgoal_success）：64位浮点数序列 18. 基于子目标成功稠密折扣奖励的回报回溯（discounted_rtgs_discounted_dense_rewards_subgoal_success）：64位浮点数序列 19. 基于子目标成功稠密折扣奖励的未折扣回报回溯（undiscounted_rtgs_discounted_dense_rewards_subgoal_success）：64位浮点数序列 20. 基于子目标成功稠密未折扣奖励的回报回溯（discounted_rtgs_undiscounted_dense_rewards_subgoal_success）：64位浮点数序列 21. 基于子目标成功稠密未折扣奖励的未折扣回报回溯（undiscounted_rtgs_undiscounted_dense_rewards_subgoal_success）：64位浮点数序列 22. 基于子目标成功归一化稠密折扣奖励的回报回溯（discounted_rtgs_normalised_discounted_dense_rewards_subgoal_success）：64位浮点数序列 23. 基于子目标成功归一化稠密折扣奖励的未折扣回报回溯（undiscounted_rtgs_normalised_discounted_dense_rewards_subgoal_success）：64位浮点数序列 24. 动作失败相关稠密奖励（dense_rewards_action_failure）：64位浮点数序列 25. 动作失败相关稠密回报回溯（dense_rtgs_action_failure）：64位浮点数序列 26. 稠密奖励（dense_rewards）：64位浮点数序列 27. 稠密回报回溯（dense_rtgs）：64位浮点数序列 28. 任务嵌入（mission_embeddings）：32位浮点数序列序列 29. 任务Token掩码（mission_token_mask）：布尔值序列 30. 规则反馈嵌入（rule_feedback_embeddings）：三维32位浮点数序列序列 31. 规则反馈Token掩码（rule_feedback_token_mask）：二维布尔值序列序列 32. 规则反馈可用标志（rule_feedback_available）：布尔值序列 33. 任务反馈嵌入（task_feedback_embeddings）：三维32位浮点数序列序列 34. 任务反馈Token掩码（task_feedback_token_mask）：二维布尔值序列序列 35. 任务反馈可用标志（task_feedback_available）：布尔值序列 36. 指令嵌入（instruction_embeddings）：三维32位浮点数序列序列 37. 指令Token掩码（instruction_token_mask）：二维布尔值序列序列 38. 指令可用标志（instructions_available）：布尔值序列 **数据集划分**： - 训练集：数据字节数2310881468，样本数量819 - 验证集：数据字节数578425764，样本数量205 整体下载大小：66970762，总数据集大小：2889307232 ##### 子配置1.2：`gotosimilardistsuniquegoalstrictlycolortype_1024x1_raw` （raw为原始数据集版本） **特征字段**：相较于子配置1.1，新增以下字段： 1. 回合名称（episode_name）：字符串类型 2. 任务描述（mission）：字符串类型 3. 任务反馈（task_feedback）：字符串序列 4. 规则反馈（rule_feedback）：字符串序列 5. 指令（instructions）：字符串序列 6. 子目标成功序列（subgoal_successes）：64位整数序列 7. 子目标（subgoals）：64位整数类型 8. 动作失败序列（action_failures）：64位整数序列 9. 最大步数（max_steps）：64位整数类型 10. 回合长度（episode_len）：64位整数类型 其余特征字段与子配置1.1一致。 **数据集划分**： - 训练集：数据字节数2061818482，样本数量819 - 验证集：数据字节数516083644，样本数量205 整体下载大小：15142969，总数据集大小：2577902126 ##### 子配置1.3：`gotosimilardistsuniquegoalstrictlycolortype_2048x1_preprocessed_sentence-transformer-pretrained_cnn-scratch` 该配置为1024x1预处理配置的2048x1尺寸版本，特征字段与子配置1.1一致。 **数据集划分**： - 训练集：数据字节数4621762712，样本数量1638 - 验证集：数据字节数1156851528，样本数量410 整体下载大小：133691140，总数据集大小：5778614240 ##### 子配置1.4：`gotosimilardistsuniquegoalstrictlycolortype_2048x1_raw` 该配置为1024x1原始配置的2048x1尺寸版本，特征字段与子配置1.2一致。 **数据集划分**： - 训练集：数据字节数4123648911，样本数量1638 - 验证集：数据字节数1032167493，样本数量410 整体下载大小：30184410，总数据集大小：5155816404 ##### 子配置1.5：`gotosimilardistsuniquegoalstrictlycolortype_4096x1_preprocessed_sentence-transformer-pretrained_cnn-scratch` 该配置为1024x1预处理配置的4096x1尺寸版本，特征字段与子配置1.1一致。 **数据集划分**： - 训练集：数据字节数9243525200，样本数量3276 - 验证集：数据字节数2313703056，样本数量820 整体下载大小：268101379，总数据集大小：11557228256 ##### 子配置1.6：`gotosimilardistsuniquegoalstrictlycolortype_4096x1_raw` 该配置为1024x1原始配置的4096x1尺寸版本，特征字段与子配置1.2一致。 **数据集划分**： - 训练集：数据字节数8247290265，样本数量3276 - 验证集：数据字节数2064346088，样本数量820 整体下载大小：60228865，总数据集大小：10311636353 --- #### 配置组2：前往相似距离唯一目标严格位置类型（gotosimilardistsuniquegoalstrictlyloc） ##### 子配置2.1：`gotosimilardistsuniquegoalstrictlyloc_1024x1_preprocessed_sentence-transformer-pretrained_cnn-scratch` 特征字段与子配置1.1一致。 **数据集划分**： - 训练集：数据字节数2310881468，样本数量819 - 验证集：数据字节数578425764，样本数量205 整体下载大小：73608712，总数据集大小：2889307232 ##### 子配置2.2：`gotosimilardistsuniquegoalstrictlyloc_2048x1_preprocessed_sentence-transformer-pretrained_cnn-scratch` 特征字段与子配置1.1一致。 **数据集划分**： - 训练集：数据字节数4621762712，样本数量1638 - 验证集：数据字节数1156851528，样本数量410 整体下载大小：144751011，总数据集大小：5778614240 ##### 子配置2.3：`gotosimilardistsuniquegoalstrictlyloc_2048x1_raw` 特征字段与子配置1.2一致。 **数据集划分**： - 训练集：数据字节数4123800963，样本数量1638 - 验证集：数据字节数1032204125，样本数量410 整体下载大小：30168478，总数据集大小：5156005088 --- #### 配置组3：前往相似距离唯一目标仅颜色类型（gotosimilardistsuniquegoalstrictlyonlycolor） ##### 子配置3.1：`gotosimilardistsuniquegoalstrictlyonlycolor_1024x1_preprocessed_sentence-transformer-pretrained_cnn-scratch` 特征字段与子配置1.1一致。 **数据集划分**： - 训练集：数据字节数2310881468，样本数量819 - 验证集：数据字节数578425764，样本数量205 整体下载大小：63746197，总数据集大小：2889307232 ##### 子配置3.2：`gotosimilardistsuniquegoalstrictlyonlycolor_1024x1_raw` 特征字段与子配置1.2一致。 **数据集划分**： - 训练集：数据字节数2061844400，样本数量819 - 验证集：数据字节数516092255，样本数量205 整体下载大小：15139257，总数据集大小：2577936655 ##### 子配置3.3：`gotosimilardistsuniquegoalstrictlyonlycolor_2048x1_preprocessed_sentence-transformer-pretrained_cnn-scratch` 特征字段与子配置1.1一致。 **数据集划分**： - 训练集：数据字节数4621762712，样本数量1638 - 验证集：数据字节数1156851528，样本数量410 整体下载大小：127449623，总数据集大小：5778614240 ##### 子配置3.4：`gotosimilardistsuniquegoalstrictlyonlycolor_2048x1_raw` 特征字段与子配置1.2一致。 **数据集划分**： - 训练集：数据字节数4123697795，样本数量1638 - 验证集：数据字节数1032188058，样本数量410 整体下载大小：30216041，总数据集大小：5155885853 --- #### 配置组4：基础拾取任务（pickup） ##### 子配置4.1：`pickup_1024x1_preprocessed_sentence-transformer-pretrained_cnn-scratch` 特征字段与子配置1.1一致。 **数据集划分**： - 训练集：数据字节数2310881468，样本数量819 - 验证集：数据字节数578425764，样本数量205 整体下载大小：69405695，总数据集大小：2889307232 ##### 子配置4.2：`pickup_2048x1_preprocessed_sentence-transformer-pretrained_cnn-scratch` 特征字段与子配置1.1一致。 **数据集划分**： - 训练集：数据字节数4621762712，样本数量1638 - 验证集：数据字节数1156851528，样本数量410 整体下载大小：137890849，总数据集大小：5778614240 ##### 子配置4.3：`pickup_2048x1_raw` 特征字段与子配置1.2一致。 **数据集划分**： - 训练集：数据字节数4123694218，样本数量1638 - 验证集：数据字节数1032171677，样本数量410 整体下载大小：29228601，总数据集大小：5155865895 --- （后续配置组如简单序列拾取、迷宫拾取、放置任务等均按照上述逻辑翻译，由于篇幅限制，此处省略重复内容，所有配置均包含预处理与原始版本，对应尺寸与任务类型详见原始配置名称。） --- ### 数据集配置文件路径 所有配置的数据文件均按照`任务类型/尺寸/预处理状态/模型结构`的层级路径存储，例如： 1. 预处理数据集训练集路径：`gotosimilardistsuniquegoalstrictlycolortype/1024x1/preprocessed/sentence-transformer-pretrained/cnn-scratch/train-*` 2. 原始数据集验证集路径：`gotosimilardistsuniquegoalstrictlycolortype/1024x1/raw/val-*`