Add files using upload-large-folder tool

.gitattributes

@@ -35,3 +35,5 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
 *tfevents* filter=lfs diff=lfs merge=lfs -text
 TDMPC2+HBench-h1-walk-v0+0/preview.mp4 filter=lfs diff=lfs merge=lfs -text
 TDMPC2+HBench-g1-walk-v0+0/preview.mp4 filter=lfs diff=lfs merge=lfs -text
+assets/tdmpc2-g1-walk.mp4 filter=lfs diff=lfs merge=lfs -text
+assets/tdmpc2-h1-walk.mp4 filter=lfs diff=lfs merge=lfs -text

README.md

@@ -39,12 +39,22 @@ _TD-MPC2 is a model-based RL algorithm combining a world model with sample-based
 
 `success_bar = 700` (HumanoidBench locomotion threshold). _Success = episode return ≥ success_bar._
 
-### 视频预览 / Video preview
+---
+
+## 🎬 演示 / Demos
+
+### H1-walk-v0 (Unitree H1, 19 DoF)
+
+<video controls width="720" src="https://huggingface.co/wsagi/HumanoidBench-TD-MPC2/resolve/main/assets/tdmpc2-h1-walk.mp4"></video>
+
+完整 walking cycle，1000 步不倒。_Full walking cycle, runs through 1000 steps without falling._
+
+### G1-walk-v0 (Unitree G1, 23 DoF with PD + BlockedHands wrappers)
 
-每个任务子目录下有 `preview.mp4` 演示 deterministic eval（最佳 seed, GUI viewer 录屏）：
+<video controls width="720" src="https://huggingface.co/wsagi/HumanoidBench-TD-MPC2/resolve/main/assets/tdmpc2-g1-walk.mp4"></video>
 
-- **`TDMPC2+HBench-h1-walk-v0+0/preview.mp4`** — H1 humanoid 完整 walking cycle，1000 步不倒
-- **`TDMPC2+HBench-g1-walk-v0+0/preview.mp4`** — G1 humanoid 走路，含偶发踉跄
+37D action 屏蔽 14D 手指剩 23D，配 PD 位置控制；含偶发踉跄但 50% 集 ≥ success_bar。
+_With 14D fingers masked (37→23 act dim) + PD position control; occasional stumbles but 50% of eps clear the success bar._
 
 ### 对比同任务 DR.Q 自训 ([wsagi/HumanoidBench-DR.Q](https://huggingface.co/wsagi/HumanoidBench-DR.Q))
 
@@ -112,16 +122,18 @@ DISPLAY=:0 conda run -n humanoidbench python scripts/tdmpc2_viewer.py \
 ## 📁 仓库结构 / Repo layout
 
 ```
-TDMPC2+HBench-h1-walk-v0+0/
-├── step_950000.pt          (32 MB — agent + world model + critic)
-├── train.log               (~370 KB — full training log)
-├── ckpt_eval.csv           (auto-eval per ckpt, N=3 quick)
-└── preview.mp4             (515 KB — GUI viewer recording)
-
-TDMPC2+HBench-g1-walk-v0+0/
-├── step_950000.pt          (32 MB)
-├── train.log               (~700 KB)
-└── preview.mp4             (257 KB)
+HumanoidBench-TD-MPC2/
+├── README.md                                  (this file)
+├── assets/
+│   ├── tdmpc2-h1-walk.mp4                     (515 KB — H1-walk GUI recording)
+│   └── tdmpc2-g1-walk.mp4                     (257 KB — G1-walk GUI recording)
+├── TDMPC2+HBench-h1-walk-v0+0/
+│   ├── step_950000.pt                         (32 MB — agent + world model + critic)
+│   ├── train.log                              (~370 KB — full training log)
+│   └── ckpt_eval.csv                          (auto-eval per ckpt, N=3 quick)
+└── TDMPC2+HBench-g1-walk-v0+0/
+    ├── step_950000.pt                         (32 MB)
+    └── train.log                              (~700 KB)
 ```
 
 `+0` 表示 seed=0。后续如果发其他 seed 会按 `+10` / `+20` 命名。

assets/tdmpc2-g1-walk.mp4

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:47b47c47d802d399552e6506d5faf781e8fab36395df8406b97704eef04b90bf
+size 256928

assets/tdmpc2-h1-walk.mp4

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:2bad368d1a3dad93006d1d0a56f9ead19fbbd8e9f2ec62f1b199ca96e1b5c70f
+size 515283