update README and add metadata

README.md

@@ -1,40 +1,18 @@
-# Heterogeneous World Modeling with Actions
+---
+title: HMA
+emoji: 📈
+colorFrom: purple
+colorTo: indigo
+sdk: gradio
+python_version: 3.9.13
+sdk_version: 3.41.2
+app_file: app.py
+pinned: false
+license: apache-2.0
+---
 
-Progress in video generation may soon make it possible to evaluate robot policies in a completely learned world model.
-Modified from [here](https://github.com/1x-technologies/1xgpt)
+# Generative Simulation Interactive Demo
 
-## Getting Started
-We require `Python 3.10` or later. This code was tested with `Python 3.10.12`.
+This demo is from the project:
 
-```
-# Install dependencies and download data
-./build.sh
-
-# Source the Python environment
-source venv/bin/activate
-```
-
-## File Structures
-```angular2html
-├── ...
-├── HPT-Video
-|   |── data 			# cached token datasets and model checkpoints
-|   |── genie 			# main modeling code
-|   |   |── diffusion   # diffusion loss related
-|   |   |── evaluate.py   # evaluate a trained model
-|   |   |── st_mar.py   # spatial time MAR
-|   |   |── generate.py   # generate tokens from trained model
-|   |   |── st_maskgit.py   # spatial-time maskgit
-|   |── magvit 			# magvit code
-|   |── sim 			# simulation related codebase
-|   |── experiments
-|   |   |── cmd     # handy commands
-|   |   |── datasplit # dataset split
-|   |   |── scripts # ablation and training scripts.
-|   |── common 		# common utility and plot scripts
-|   |── train.py # train using magvit
-|   |── train_diffusion.py # train using mar
-|   |── train_multi.py # train on multiple datasets jointly
-|   |── visualize.py # visualize generated tokens
-└── ...
-```
+**HMA**