novastar112/pusht_96_int1

96px PushT PPO成功轨迹数据集。该数据集由1Hz PPO PushT求解器生成，使用int1动作编解码器，渲染图像存储在history[*].image、image_prev和image_next中，分辨率为96x96，JPEG质量为85。数据集包含训练集和测试集，分别有500,000和200条记录，格式为gzip压缩的JSONL。训练集和测试集的初始状态通过init_state_hash进行区分，确保不重叠。动作中的坐标使用PushT 512x512物理坐标，并四舍五入到一位小数。96px PushT PPO successful trajectory dataset. The trajectories are generated by a 1Hz PPO PushT solver with action codec `int1`, rendered images in `history[*].image`, `image_prev`, and `image_next` at `96x96` and JPEG quality `85`. The dataset includes train and test splits with 500,000 and 200 records respectively, in gzip-compressed JSONL format. Train/test initial states are filtered to be disjoint by `init_state_hash`. Coordinates in `move` actions use PushT 512x512 physical coordinates rounded to exactly one decimal place.