Update README

README.md

@@ -1,46 +1,27 @@
----
-title: Stack 2.9 Code Assistant
-emoji: 💻
-colorFrom: blue
-colorTo: purple
-sdk: docker
-pinned: false
-tags:
-- code-generation
-- python
-- qwen
-- coding-assistant
-- fine-tuned
----
+# Stack 2.9 Code Assistant
 
-# 💻 Stack 2.9 - Fine-tuned Code Assistant
+A free HuggingFace Space demo running **Qwen2.5-Coder-1.5B-Instruct** — a capable open-source coding model.
 
-A **fine-tuned** coding assistant powered by Qwen2.5-Coder-1.5B, trained on Stack Overflow Q&A data.
+## About This Demo
 
-## Model
+This Space demonstrates a coding assistant built on Qwen2.5-Coder. The model is excellent at:
+- Python, JavaScript, TypeScript, and 50+ languages
+- Code completion and generation
+- Debugging and bug fixes
+- Explaining code concepts
 
-- **Base Model:** Qwen/Qwen2.5-Coder-1.5B
-- **Fine-tuned on:** Stack Overflow Q&A (Python-heavy)
-- **Context Length:** 32K tokens
-- **Parameters:** 1.5B
-- **License:** Apache 2.0
-- **Hub:** [my-ai-stack/Stack-2-9-finetuned](https://huggingface.co/my-ai-stack/Stack-2-9-finetuned)
+## Fine-Tuned Version
 
-## Features
+The full fine-tuned model (`my-ai-stack/Stack-2-9-finetuned`) was trained on Stack Exchange data but requires paid LFS storage on HuggingFace (6GB model file exceeds free tier). 
 
-- **Code Generation** - Write Python, SQL, JavaScript, TypeScript, and more
-- **Code Debugging** - Find and fix bugs in your code  
-- **Programming Help** - Get explanations and refactoring suggestions
-- **Chat Interface** - Easy-to-use Gradio UI
+For local deployment with the fine-tuned weights, use:
+```bash
+ollama pull walidsobhie/stack-2-9
+ollama run walidsobhie/stack-2-9
+```
 
-## Usage
+## Technical Details
 
-1. Enter your prompt in the text box
-2. Adjust settings (max tokens, temperature)
-3. Click "Generate" to get your response
-
-This demo runs the **actual fine-tuned model**, not the base Qwen2.5-Coder.
-
-## Hardware
-
-The 1.5B model fits on free T4 GPU on HuggingFace Spaces (~4GB VRAM FP16).
+- **Model:** Qwen/Qwen2.5-Coder-1.5B-Instruct
+- **Framework:** Gradio + Transformers
+- **Hardware:** Free T4 GPU (HuggingFace Spaces)