novastar112/pusht_96_norm4_visual_nomarker

这是一个96像素PushT PPO成功轨迹数据集。轨迹由1Hz PPO PushT求解器生成，使用norm4动作编码，在history[*].image、image_prev和image_next中渲染为96x96像素、JPEG质量90的图像。每个情节仅包含成功轨迹，且最多30个环境步骤。视觉标记模式为无，推动器在512空间中以半径11.0渲染，但物理碰撞仍使用环境原始半径。提示模式为first_full_then_step_only。历史图像约定：history[*].image_prev等于history[*].image；对于非终止步骤，history[i].image_next等于history[i+1].image；终止步骤保留其终止帧。数据分割包括50万条训练记录和200条测试记录，训练和测试的初始状态通过init_state_hash确保不重叠。move动作中的坐标使用归一化[0, 1]坐标，并四舍五入到四位小数，执行时解码回PushT 512x512物理坐标。96px PushT PPO successful trajectory dataset. The trajectories are generated by a 1Hz PPO PushT solver with action codec `norm4`, rendered images in `history[*].image`, `image_prev`, and `image_next` at `96x96` and JPEG quality `90`. Each episode is success-only and capped at `30` environment steps. Visual marker mode: `none`. The pusher is rendered with radius `11.0` in 512-space; physics still uses the environments collision radius. Prompt mode: `first_full_then_step_only`. History image convention: `history[*].image_prev == history[*].image`; `history[i].image_next == history[i+1].image` for non-terminal steps; the terminal step keeps its terminal frame. Splits include 500,000 train records and 200 test records, with disjoint initial states filtered by `init_state_hash`. Coordinates in `move` actions use normalized [0, 1] coordinates rounded to exactly four decimal places; execution decodes them back to PushT 512x512 physical coordinates.