Upload 3 files

README.md

@@ -1,188 +1,67 @@
----
-title: ATLES-ECHO System
-emoji: 🧠
-colorFrom: blue
-colorTo: purple
-sdk: static
-pinned: true
-license: mit
-tags:
-- semantic-memory
-- personal-ai
-- embeddings
-- privacy-first
-- digital-twin
----
-
-# ATLES-ECHO 🧠
-
-**Your Semantic Digital Twin** - A privacy-first AI system that remembers everything you do.
-
-## 🌟 Overview
-
-ATLES-ECHO is an intelligent semantic monitoring and memory system powered by the **[ATLES Champion Embedding Model](https://huggingface.co/spartan8806/atles-champion-embedding)** (Top-10 worldwide on MTEB).
-
-### What It Does
-
-- 📝 **Captures** - Files, screens, code, notes, conversations
-- 🔍 **Understands** - Uses advanced embeddings for semantic comprehension
-- 🧠 **Learns** - Builds behavioral patterns and interest profiles
-- 🔐 **Protects** - 100% local storage, zero cloud uploads
-- ⚡ **Delivers** - Real-time semantic search across your entire digital life
-
-## 🚀 Quick Facts
-
-| Feature | Details |
-|---------|---------|
-| **Embedding Model** | [spartan8806/atles-champion-embedding](https://huggingface.co/spartan8806/atles-champion-embedding) |
-| **Performance** | STS-B Pearson: 0.8445, Spearman: 0.8374 (Top-10 MTEB) |
-| **Dimensions** | 768-dim MPNet-base architecture |
-| **Speed** | ~200 embeddings/sec (GPU) |
-| **Vector DB** | FAISS (Facebook AI Similarity Search) |
-| **Backend** | FastAPI (Python 3.11+) |
-| **Frontend** | React 18 + TypeScript |
-| **Privacy** | 100% local, encrypted, open source |
-
-## 🎯 Core Capabilities
-
-### Semantic Monitoring
-- 📁 **File Changes** - Track all code and document edits
-- 🖥️ **Screen Content** - OCR-based extraction (optional)
-- 📋 **Clipboard** - Save copied text and snippets
-- 🪟 **App Usage** - Monitor focus time and patterns
-- ⌨️ **Typing Patterns** - Context-aware analysis (opt-in)
-
-### AI-Powered Search
-- **Natural Language** - "How did I implement authentication?"
-- **Semantic Similarity** - Find related content without exact matches
-- **Context Retrieval** - Get relevant background for any topic
-- **Pattern Detection** - Discover productivity trends
-
-### Auto-Generated Insights
-- Daily activity summaries
-- Interest profiling
-- Usage analytics
-- Behavioral pattern analysis
-
-## 🏗️ Architecture
-
-```
-Web UI (React) 
-    ↓
-FastAPI Backend 
-    ↓
-┌─────────────────┬──────────────────┐
-│ Embedding Engine│   Vector DB      │
-│   (Champion)    │    (FAISS)       │
-│   768-dim       │  Similarity      │
-└─────────────────┴──────────────────┘
-         ↓
-   Knowledge Base (SQLite)
-         ↓
-   Watchers (File, Screen, Clipboard, etc.)
-```
-
-## 📖 Usage Example
-
-```python
-import requests
-
-# Semantic search
-response = requests.get(
-    "http://localhost:5001/api/search",
-    params={
-        "query": "authentication implementation",
-        "limit": 5
-    }
-)
-
-results = response.json()
-for item in results["results"]:
-    print(f"{item['score']:.3f} - {item['content'][:100]}...")
-```
-
-## 🔒 Privacy
-
-ATLES-ECHO is **privacy-first by design**:
-
-✅ **100% Local** - All data stays on your machine  
-✅ **No Cloud** - Zero uploads, ever  
-✅ **Encrypted** - AES-256 encryption at rest  
-✅ **Open Source** - Audit the code yourself  
-✅ **Full Control** - Disable any feature anytime  
-
-## 🚀 Installation
-
-```bash
-# Clone repository
-git clone https://github.com/spartan8806/atles-echo.git
-cd atles-echo
-
-# Install backend
-cd backend && pip install -r requirements.txt
-
-# Install frontend
-cd ../frontend && npm install
-
-# Run (Windows)
-.\start_echo.bat
-```
-
-Access dashboard at: **http://localhost:3000**
-
-## 📊 Performance
-
-| Dataset Size | Search Latency | Storage |
-|--------------|----------------|---------|
-| 1K entries | 5ms | 4.5 MB |
-| 10K entries | 8ms | 45 MB |
-| 100K entries | 15ms | 450 MB |
-| 1M entries | 50ms | 4.5 GB |
-
-## 🎨 Use Cases
-
-- **Developers**: Code search, debug history, research notes
-- **Writers**: Document tracking, research management, idea capture
-- **Researchers**: Paper organization, experiment notes, literature review
-- **Knowledge Workers**: Second brain, meeting notes, project memory
-
-## 🗺️ Roadmap
-
-- [x] Core semantic monitoring
-- [x] Real-time search
-- [x] Interest profiling
-- [ ] Browser history integration
-- [ ] Email integration (local)
-- [ ] Voice memo capture
-- [ ] Mobile companion app
-
-## 🤝 Part of ATLES Ecosystem
-
-ATLES-ECHO is one component of the **ATLES (Advanced Thinking & Learning Execution System)**:
-
-- **ATLES Brain** - Central AI coordinator
-- **ATLES-ECHO** - Semantic memory (this project)
-- **Phoenix** - AI introspection research system *(private, not public)*
-- **SENTINEL** - Documentation-focused semantic monitoring *(like ECHO for docs)*
-- **ATLES-MENTOR** - MoE code assistance system *(private, not public)*
-
-## ��� License
-
-MIT License - Copyright (c) 2025 Conner (spartan8806)
-
-## 🙏 Credits
-
-Powered by **[ATLES Champion Embedding](https://huggingface.co/spartan8806/atles-champion-embedding)**
-
-Built with: FAISS, FastAPI, React, Sentence Transformers
-
----
-
-<div align="center">
-
-**"Your digital life, remembered."**
-
-[📚 Full Documentation](https://github.com/spartan8806/atles-echo) | [🐛 Report Issues](https://github.com/spartan8806/atles-echo/issues) | [⭐ Star on GitHub](https://github.com/spartan8806/atles-echo)
-
-</div>
-
+---
+title: ATLES-ECHO Embedding Service
+emoji: 🧠
+colorFrom: blue
+colorTo: indigo
+sdk: gradio
+sdk_version: 4.44.0
+app_file: app.py
+pinned: true
+license: apache-2.0
+short_description: Semantic embeddings with ATLES Champion model
+---
+
+# 🧠 ATLES-ECHO Embedding Service
+
+Generate high-quality semantic embeddings using the **ATLES Champion** embedding model.
+
+## Features
+
+- **🔤 Single Embedding**: Generate embedding for any text
+- **⚖️ Compare Similarity**: Compare semantic similarity between two texts  
+- **📦 Batch Embed**: Process multiple texts at once (up to 10)
+
+## Model Details
+
+| Property | Value |
+|----------|-------|
+| **Model** | [spartan8806/atles-champion-embedding](https://huggingface.co/spartan8806/atles-champion-embedding) |
+| **Dimension** | 768 |
+| **Base Model** | all-mpnet-base-v2 |
+| **Parameters** | 110M |
+| **Training** | H200 GPU (30 minutes) |
+
+## Performance (MTEB STS-B)
+
+- **Pearson**: 0.8445 (Top-10 worldwide)
+- **Spearman**: 0.8374
+
+## About ATLES
+
+ATLES-ECHO is the semantic memory core of the ATLES ecosystem - your AI digital twin that learns from your digital life while keeping everything private and local.
+
+**Ecosystem Components:**
+- 🧠 **ECHO** - Semantic memory and embeddings
+- 🦅 **Phoenix** - AI council for decision making
+- 🔬 **SENTINEL** - Research and knowledge gathering
+- 📚 **MENTOR** - Code understanding and assistance
+
+## API Usage
+
+This space provides a visual interface. For programmatic access, use the model directly:
+
+```python
+from sentence_transformers import SentenceTransformer
+
+model = SentenceTransformer("spartan8806/atles-champion-embedding")
+embedding = model.encode("Your text here", normalize_embeddings=True)
+```
+
+## License
+
+Apache 2.0 - Free for commercial and personal use.
+
+---
+
+Built with ❤️ by [spartan8806](https://huggingface.co/spartan8806)
+

app.py

@@ -0,0 +1,185 @@
+"""
+ATLES-ECHO - Semantic Embedding Service
+A Hugging Face Space for generating embeddings using the ATLES Champion model.
+"""
+
+import gradio as gr
+from sentence_transformers import SentenceTransformer
+import numpy as np
+
+# Load the ATLES Champion embedding model
+print("Loading ATLES Champion Embedding model...")
+model = SentenceTransformer("spartan8806/atles-champion-embedding")
+print(f"Model loaded! Dimension: {model.get_sentence_embedding_dimension()}")
+
+def generate_embedding(text: str) -> dict:
+    """Generate embedding for input text"""
+    if not text or not text.strip():
+        return {"error": "Please enter some text", "embedding": None, "dimension": None}
+    
+    # Generate embedding
+    embedding = model.encode(text, normalize_embeddings=True)
+    
+    return {
+        "text_preview": text[:100] + "..." if len(text) > 100 else text,
+        "dimension": len(embedding),
+        "embedding_preview": embedding[:10].tolist(),  # First 10 values
+        "embedding_full": embedding.tolist()
+    }
+
+def compare_texts(text1: str, text2: str) -> dict:
+    """Compare similarity between two texts"""
+    if not text1.strip() or not text2.strip():
+        return {"error": "Please enter both texts", "similarity": None}
+    
+    # Generate embeddings
+    embeddings = model.encode([text1, text2], normalize_embeddings=True)
+    
+    # Calculate cosine similarity
+    similarity = float(np.dot(embeddings[0], embeddings[1]))
+    
+    return {
+        "text1_preview": text1[:50] + "..." if len(text1) > 50 else text1,
+        "text2_preview": text2[:50] + "..." if len(text2) > 50 else text2,
+        "similarity": round(similarity, 4),
+        "similarity_percent": f"{similarity * 100:.1f}%",
+        "interpretation": get_similarity_interpretation(similarity)
+    }
+
+def get_similarity_interpretation(score: float) -> str:
+    """Interpret similarity score"""
+    if score >= 0.9:
+        return "🟢 Nearly identical meaning"
+    elif score >= 0.7:
+        return "🟡 Very similar"
+    elif score >= 0.5:
+        return "🟠 Somewhat related"
+    elif score >= 0.3:
+        return "🔴 Loosely related"
+    else:
+        return "⚫ Different topics"
+
+def batch_embed(texts: str) -> dict:
+    """Generate embeddings for multiple texts (one per line)"""
+    lines = [l.strip() for l in texts.split('\n') if l.strip()]
+    
+    if not lines:
+        return {"error": "Please enter at least one text (one per line)", "embeddings": None}
+    
+    if len(lines) > 10:
+        return {"error": "Maximum 10 texts at a time", "embeddings": None}
+    
+    # Generate embeddings
+    embeddings = model.encode(lines, normalize_embeddings=True)
+    
+    results = []
+    for i, (text, emb) in enumerate(zip(lines, embeddings)):
+        results.append({
+            "index": i + 1,
+            "text": text[:50] + "..." if len(text) > 50 else text,
+            "embedding_preview": emb[:5].tolist()
+        })
+    
+    return {
+        "count": len(lines),
+        "dimension": len(embeddings[0]),
+        "results": results
+    }
+
+# Create Gradio interface
+with gr.Blocks(
+    title="ATLES-ECHO Embedding Service",
+    theme=gr.themes.Soft(primary_hue="blue", secondary_hue="cyan")
+) as demo:
+    
+    gr.Markdown("""
+    # 🧠 ATLES-ECHO Embedding Service
+    
+    Generate high-quality semantic embeddings using the **ATLES Champion** model.
+    
+    - **Model**: [spartan8806/atles-champion-embedding](https://huggingface.co/spartan8806/atles-champion-embedding)
+    - **Dimension**: 768
+    - **Top-10 MTEB Performance**: Pearson 0.8445, Spearman 0.8374
+    """)
+    
+    with gr.Tabs():
+        # Tab 1: Single Embedding
+        with gr.TabItem("🔤 Single Embedding"):
+            gr.Markdown("Generate an embedding for a single piece of text.")
+            
+            with gr.Row():
+                with gr.Column():
+                    single_input = gr.Textbox(
+                        label="Input Text",
+                        placeholder="Enter text to embed...",
+                        lines=3
+                    )
+                    single_btn = gr.Button("Generate Embedding", variant="primary")
+                
+                with gr.Column():
+                    single_output = gr.JSON(label="Embedding Result")
+            
+            single_btn.click(
+                fn=generate_embedding,
+                inputs=single_input,
+                outputs=single_output
+            )
+        
+        # Tab 2: Compare Texts
+        with gr.TabItem("⚖️ Compare Similarity"):
+            gr.Markdown("Compare the semantic similarity between two texts.")
+            
+            with gr.Row():
+                text1_input = gr.Textbox(label="Text 1", placeholder="First text...", lines=2)
+                text2_input = gr.Textbox(label="Text 2", placeholder="Second text...", lines=2)
+            
+            compare_btn = gr.Button("Compare Similarity", variant="primary")
+            compare_output = gr.JSON(label="Similarity Result")
+            
+            compare_btn.click(
+                fn=compare_texts,
+                inputs=[text1_input, text2_input],
+                outputs=compare_output
+            )
+        
+        # Tab 3: Batch Embedding
+        with gr.TabItem("📦 Batch Embed"):
+            gr.Markdown("Generate embeddings for multiple texts (one per line, max 10).")
+            
+            with gr.Row():
+                with gr.Column():
+                    batch_input = gr.Textbox(
+                        label="Texts (one per line)",
+                        placeholder="Text 1\nText 2\nText 3...",
+                        lines=6
+                    )
+                    batch_btn = gr.Button("Generate Batch Embeddings", variant="primary")
+                
+                with gr.Column():
+                    batch_output = gr.JSON(label="Batch Results")
+            
+            batch_btn.click(
+                fn=batch_embed,
+                inputs=batch_input,
+                outputs=batch_output
+            )
+    
+    gr.Markdown("""
+    ---
+    ### About ATLES-ECHO
+    
+    ATLES-ECHO is the semantic memory core of the ATLES ecosystem - your AI digital twin that learns from your digital life.
+    
+    **Features:**
+    - 🧠 High-quality semantic embeddings (768 dimensions)
+    - ⚡ Fast inference with normalized vectors
+    - 🎯 Top-10 MTEB benchmark performance
+    - 🔒 Built for the ATLES privacy-first ecosystem
+    
+    [View Model Card](https://huggingface.co/spartan8806/atles-champion-embedding) | [ATLES GitHub](https://github.com/spartan8806)
+    """)
+
+# Launch the app
+if __name__ == "__main__":
+    demo.launch()
+

requirements.txt

@@ -0,0 +1,5 @@
+gradio>=4.0.0
+sentence-transformers>=2.2.2
+torch>=2.0.0
+numpy>=1.24.0
+