Update README.md

README.md

@@ -1,4 +1,3 @@
-
 ---
 pretty_name: MicroGen3D
 tags:
@@ -17,27 +16,118 @@ language:
   - en
 ---
 
-
 # microgen3D
 
 [![Code](https://img.shields.io/badge/GitHub-Code-black?logo=github)](https://github.com/baskargroup/MicroGen3D)
 
 ## Dataset Summary
 
-**microgen3D** is a dataset of 3D voxelized microstructures designed for training, evaluation, and benchmarking of generative models—especially Conditional Latent Diffusion Models (LDMs). It includes both synthetic (Cahn-Hilliard) and experimental microstructures with multiple phases (2 to 3). The voxel grids range from `64³` up to `128×128×64`.
+**microgen3D** is a dataset of 3D voxelized microstructures designed for training, evaluation, and benchmarking of generative models—especially Conditional Latent Diffusion Models (LDMs). It includes both synthetic (Cahn–Hilliard) and experimental microstructures with multiple phases (2 to 3). The voxel grids range from `64³` up to `128×128×64`.
 
 The dataset consists of three microstructure types:
 - **Experimental microstructures**
-- **2-phase Cahn-Hilliard microstructures**
-- **3-phase Cahn-Hilliard microstructures**
+- **2-phase Cahn–Hilliard microstructures**
+- **3-phase Cahn–Hilliard microstructures**
+
+The two Cahn–Hilliard datasets are thresholded versions of the same simulation source.  
+For each dataset type, we also provide pretrained generative model weights:
+- `vae.pt` – Variational Autoencoder
+- `fp.pt` – Feature Predictor
+- `ddpm.pt` – Denoising Diffusion Probabilistic Model
+
+---
+
+## 📂 Dataset Overview
+
+| File Name                         | Size     | Description |
+|------------------------------------|----------|-------------|
+| `CH_three_phase.tar.gz`            | ~5.57 GB | Full **three-phase Cahn–Hilliard** dataset with 3D microstructures and morphological descriptors. |
+| `CH_two_phase.tar.gz`              | ~4.37 GB | Full **two-phase Cahn–Hilliard** dataset with 3D microstructures and morphological descriptors. |
+| `experimental.tar.gz`              | ~843 MB  | **Experimental microstructure** dataset from real-world samples, voxelized for modeling. |
+| `sample_CH_three_phase.tar.gz`     | ~12.2 MB | Small subset of the three-phase dataset for testing/demo purposes. |
+| `sample_CH_two_phase.tar.gz`       | ~9.59 MB | Small subset of the two-phase dataset for testing/demo purposes. |
+
+---
+
+## 📊 Detailed Dataset Information
+
+### **CH Two-Phase Dataset**
+- **File:** `CH_two_phase.tar.gz`
+- **Total Microstructures:** 47,119
+- **Splits:** 10 (Train: 9, Validation: 1)
+- **Microstructure Shape:** `(128, 128, 64)`
+- **Attributes per Key:** 34
+- **Example Attributes:**
+  - `ABS_f_D`: 0.391171
+  - `CT_f_D_tort1`: 0.293271
+  - `phi`: 0.556
+  - `chi`: 2.33
+  - `source`: direct/data_chi_2.330_phi_0.556_step_235.txt
+
 
-The two Cahn-Hilliard datasets are thresholded versions of the same simulation source. For each dataset type, we also provide pretrained generative model weights, comprising:
-- `vae.ckpt` – Variational Autoencoder
-- `fp.ckpt` – Feature Predictor
-- `ddpm.ckpt` – Denoising Diffusion Probabilistic Model
+---
+
+### **CH Three-Phase Dataset**
+- **File:** `CH_three_phase.tar.gz`
+- **Total Microstructures:** 45,980
+- **Splits:** 10 (Train: 9, Validation: 1)
+- **Microstructure Shape:** `(128, 128, 64)`
+- **Attributes per Key:** 13
+- **Example Attributes:**
+  - `Interface_AM`: 113702.0
+  - `Interface_DM`: 96692.0
+  - `phi`: 0.514
+  - `chi`: 2.2
+  - `source`: data_chi_2.200_phi_0.514.h5____172.txt
+
+---
+
+### **Experimental Microstructure Dataset**
+- **File:** `experimental.tar.gz`
+- **Total Microstructures:** 21,421
+- **Train Samples:** 19,278  
+  **Validation Samples:** 2,143
+- **Microstructure Shape:** `(64, 64, 64)`
+- **Attributes per Key:** 23
+- **Example Attributes:**
+  - `ABS_f_D`: 0.591423
+  - `CT_f_D_tort1`: 0.159534
+  - `source`: /work/mech-ai-scratch/nirmal/generative_model_data/experimental/grid_cut/graspi/morphs/CB_120_260.txt
 
 ---
 
+## 📥 Download Examples
+
+### Using Python
+```python
+from huggingface_hub import hf_hub_download
+
+# Download sample dataset
+hf_hub_download(
+    repo_id="BGLab/microgen3D",
+    filename="data/experimental.tar.gz",
+    repo_type="dataset",
+    local_dir=""
+)
+
+# Download experimental pretrained weights
+import os
+weights_local_dir = ""
+os.makedirs(weights_local_dir, exist_ok=True)
+
+files = {
+    "weights/experimental/vae.pt": "vae.pt",
+    "weights/experimental/fp.pt": "fp.pt",
+    "weights/experimental/ddpm.pt": "ddpm.pt"
+}
+
+for remote_path, local_name in files.items():
+    downloaded_path = hf_hub_download(
+        repo_id="BGLab/microgen3D",
+        filename=remote_path,
+        repo_type="dataset"
+    )
+
 ## 📁 Repository Structure
 
 ```