diff --git a/docs/index.md b/docs/index.md index 3698e6e..d2ccb02 100644 --- a/docs/index.md +++ b/docs/index.md @@ -119,7 +119,7 @@ step behind a slow link. -*70 seconds, no sound: one value changed on its way to the trainer, a learning +*67 seconds, no sound: one value changed on its way to the trainer, a learning curve that rises anyway, and the step ledger that finds the change.* The same bridge let us ask how a broken boundary shows itself. While SB3's diff --git a/docs/index.zh.md b/docs/index.zh.md index 4259564..9dd4058 100644 --- a/docs/index.zh.md +++ b/docs/index.zh.md @@ -94,7 +94,7 @@ RLinf 还训练了跑在一台笔记本上的 16 个 HalfCheetah 环境端,那 -*70 秒,无声:一个值在送往训练端的路上被改掉,学习曲线照样上升,步账本找到了这处改动。* +*67 秒,无声:一个值在送往训练端的路上被改掉,学习曲线照样上升,步账本找到了这处改动。* 借同一个 bridge,我们还能看边界出错时会有什么表现。Stable-Baselines3 的 PPO 训练 Pendulum、HalfCheetah、Hopper 和 7 关节机械臂 Pusher 时,环境端把发出去的某一个字段按 diff --git a/docs/media/boundary-faults.mp4 b/docs/media/boundary-faults.mp4 index 00e9c78..b73de00 100644 Binary files a/docs/media/boundary-faults.mp4 and b/docs/media/boundary-faults.mp4 differ