Upload README.md with huggingface_hub

README.md

@@ -47,17 +47,26 @@ Xe is a hybrid AI system featuring:
 - **Input Size**: 64 features (bag-of-words representation)
 - **Hidden Layers**: 2 layers with 32 and 11 neurons respectively
 - **Activation**: ReLU for hidden layers, linear for output
-- **Learning Rate**: 0.03 (trained)
+- **Learning Rate**: 0.01 (optimized)
 - **Loss Function**: Mean Squared Error
 
 ## Training Data
 
 The model was trained on:
-- 197 conversation samples (classical training, up to 2026-07-07)
-- 81 universal truths samples (up to 2026-07-12)
-- Various domain-specific knowledge bases
-- Total interactions: 92 conversations
-- Last training: 2026-07-12
+- 80 conversation samples (expanded training)
+- Comprehensive knowledge base across 17 domains
+- Philosophical, scientific, mathematical, and cultural knowledge
+- Last training: 2026-07-22
+
+## Knowledge Domains
+
+The model is expert in:
+- **Humanities**: Philosophy, literature, history, art
+- **Sciences**: Physics, biology, chemistry, astronomy
+- **Mathematics**: Calculus, algebra, statistics, logic
+- **Social Sciences**: Psychology, law, economics, governance
+- **Technology**: Computer science, AI, current science
+- **Business**: Marketing, engineering, medicine
 
 ## Intended Use
 
@@ -81,8 +90,8 @@ import json
 import numpy as np
 
 # Download model files
-model_path = hf_hub_download(repo_id="username/xe-ai", filename="xe_model.json")
-brain_path = hf_hub_download(repo_id="username/xe-ai", filename="xe_brain.json")
+model_path = hf_hub_download(repo_id="Travellers/xe", filename="xe_model.json")
+brain_path = hf_hub_download(repo_id="Travellers/xe", filename="xe_brain.json")
 
 # Load model
 with open(model_path) as f: