novastar112/pusht_96_int1_visual_nomarker

这是一个96像素的PushT PPO成功轨迹数据集。轨迹由1Hz PPO PushT求解器生成，使用动作编码int1，在history[*].image、image_prev和image_next中渲染为96x96像素、JPEG质量90的图像。每个轨迹仅包含成功案例，最多30个环境步骤。视觉标记模式为无，推动器在512空间中渲染半径为11.0，物理碰撞半径仍使用环境默认值。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。数据分割包括训练集（500,000条记录）和测试集（200条记录），均为gzip压缩的JSONL格式，初始状态通过init_state_hash过滤以确保不重叠。动作中的坐标使用PushT 512x512物理坐标，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 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 train (500,000 records) and test (200 records) in gzip-compressed JSONL format, with initial states filtered to be disjoint by init_state_hash. Coordinates in move actions use PushT 512x512 physical coordinates rounded to exactly one decimal place. Rendered images are 96px.