Upload README.md with huggingface_hub

README.md

@@ -1,46 +1,42 @@
 ---
-title: Autoencoder General Purpose 2D
+title: Autoencoder (Transcriptome-centric, 2D)
 emoji: 🧬
-colorFrom: blue
-colorTo: green
-sdk: pytorch
+colorFrom: purple
+colorTo: blue
+sdk: python
 tags:
 - transcriptomics
 - dimensionality-reduction
 - ae
-- general
 license: mit
 ---
 
-# Autoencoder (General Purpose, 2D)
+# Autoencoder (Transcriptome-centric, 2D)
 
-This model is part of the TRACERx Datathon 2025 transcriptomics analysis pipeline.
+Pre-trained Autoencoder model for transcriptomics data compression, part of the TRACERx Datathon 2025 project.
 
 ## Model Details
 
-- **Model Type**: Autoencoder
-- **Dataset**: General Purpose
-- **Latent Dimensions**: 2
-- **Compression Mode**: samples
-- **Framework**: PyTorch
+- **Method**: Autoencoder
+- **Compression Mode**: Transcriptome-centric
+- **Output Dimensions**: 2
+- **Training Data**: TRACERx open dataset (VST-normalized counts)
 
 ## Usage
 
 This model is designed to be used with the TRACERx Datathon 2025 analysis pipeline.
 It will be automatically downloaded and cached when needed.
 
-## Model Architecture
+```python
+import joblib
 
-- Input: Gene expression data
-- Hidden layers: [input_size, 512, 256, 128, 2]
-- Output: 2-dimensional latent representation
-- Activation: ELU with batch normalization
+# Load the model bundle
+model_data = joblib.load("model.joblib")
 
-## Training Data
-
-Trained on broader open transcriptomics datasets
+# Access components based on model type
+# See documentation for specific usage
+```
 
 ## Files
 
-- `autoencoder_2_latent_dims_oos_mode.pt`: Main model weights
-- `latent_df.csv`: Example latent representations (if available)
+- `model.joblib`: Model bundle containing fitted model and preprocessing parameters