Upload from Ropedia Academy

README.md

@@ -1,13 +1,54 @@
 ---
 license: mit
+library_name: pytorch
 tags:
 - ropedia-academy
 - educational
+- diffusion
 ---
 
-# A_motion_diffusion
+# Motion diffusion (DDPM)
 
-Trained from scratch in **Ropedia Academy** (https://chaoyue0307.github.io/ropedia-academy/) — an educational lab. The checkpoint and full loss/eval history (`metrics.json`) are included.
+A denoising-diffusion model that generates motion trajectories — the MDM recipe, trained from scratch, sampled from EMA weights.
 
+Trained from scratch in **[Ropedia Academy](https://chaoyue0307.github.io/ropedia-academy/)** — an interactive, bilingual course on embodied & spatial AI. **Educational model:** small and quick to train; the value is the *method* and a reproducible pipeline, not a leaderboard score.
 
-![results](figure.png)
+| | |
+|---|---|
+| **Task** | human-motion generation |
+| **Data** | synthetic 2D motion trajectories |
+| **Track** | A · Human modeling |
+| **Notebook** | [![Open In Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/github/ChaoYue0307/ropedia-academy/blob/main/notebooks/training/A_motion_diffusion.ipynb) |
+
+## Results
+
+| metric | value |
+|---|---|
+| loss (final) | 0.159 |
+
+
+![figure](figure.png)
+
+## How to use
+
+```python
+import torch
+state = torch.load("model.pt", map_location="cpu")   # some labs save pose.pt / gaussians.pt / transform.pt
+# Rebuild the model class from the Ropedia Academy notebook (linked above), then:
+# model.load_state_dict(state)
+```
+
+## Files
+
+- `denoiser.pt`
+- `figure.png`
+- `metrics.json`
+
+
+## Reproduce / train your own
+
+Open the [lab notebook in Colab](https://colab.research.google.com/github/ChaoYue0307/ropedia-academy/blob/main/notebooks/training/A_motion_diffusion.ipynb) → **Runtime → GPU → Run all**, then its *Publish to the Hugging Face Hub* cell. Browse every lab in the [Ropedia Academy Labs tab](https://chaoyue0307.github.io/ropedia-academy/labs).
+
+
+---
+*Part of the [Ropedia Academy](https://chaoyue0307.github.io/ropedia-academy/) trained-model collection.*

denoiser.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:c165e45a3daa3bbf0c6fe5ba195117bed603be44cd3c76669276d4d68e6ba44f
-size 927189
+oid sha256:6e36c1b51c7b80264b1f5e5c56c7e251c7f1b45e25ecf032f07d632911e73632
+size 927125

metrics.json

@@ -2,47 +2,47 @@
   "loss": [
     [
       0,
-      0.9922628998756409
+      1.0084
     ],
     [
-      800,
-      0.19711501896381378
+      200,
+      0.5319
     ],
     [
-      1600,
-      0.15966030955314636
+      400,
+      0.3503
     ],
     [
-      2400,
-      0.14236339926719666
+      600,
+      0.2463
     ],
     [
-      3200,
-      0.14732688665390015
+      800,
+      0.2112
     ],
     [
-      4000,
-      0.1111026480793953
+      1000,
+      0.178
     ],
     [
-      4800,
-      0.12580068409442902
+      1200,
+      0.1793
     ],
     [
-      5600,
-      0.13158488273620605
+      1400,
+      0.1548
     ],
     [
-      6400,
-      0.1079515814781189
+      1600,
+      0.1945
     ],
     [
-      7200,
-      0.09586234390735626
+      1800,
+      0.1445
     ],
     [
-      8000,
-      0.09420143067836761
+      2000,
+      0.159
     ]
   ]
 }