Update README.md

README.md

@@ -5,17 +5,54 @@ tags:
 - transformers
 - unsloth
 - llama
+- education
 license: apache-2.0
 language:
 - en
 ---
 
-# Uploaded finetuned  model
+# 🦙 Uploaded Finetuned Model – Llama 3.1 (8B) by Matteo Angeloni
 
-- **Developed by:** matteoangeloni
-- **License:** apache-2.0
-- **Finetuned from model :** unsloth/meta-llama-3.1-8b-unsloth-bnb-4bit
+- **Developed by:** [matteoangeloni](https://huggingface.co/matteoangeloni)  
+- **License:** apache-2.0  
+- **Base model:** [unsloth/meta-llama-3.1-8b-unsloth-bnb-4bit](https://huggingface.co/unsloth/meta-llama-3.1-8b-unsloth-bnb-4bit)  
+- **Libraries used:** [Unsloth](https://github.com/unslothai/unsloth), Hugging Face TRL  
 
-This llama model was trained 2x faster with [Unsloth](https://github.com/unslothai/unsloth) and Huggingface's TRL library.
+This model is my **first finetuned Llama model**, built for **educational and legal-domain text generation**.  
+Training was accelerated with **Unsloth** (2x faster fine-tuning) and integrated with Hugging Face tools.
 
-[<img src="https://raw.githubusercontent.com/unslothai/unsloth/main/images/unsloth%20made%20with%20love.png" width="200"/>](https://github.com/unslothai/unsloth)
+---
+
+## 📚 Training Data
+
+The model was trained on:
+
+- **Dataset:** [louisbrulenaudet/code-education](https://huggingface.co/datasets/louisbrulenaudet/code-education)  
+  → educational dataset for code-related instructions.
+
+---
+
+## 🎯 Intended Use
+
+- Experimentation with **educational text generation**  
+- Testing **instruction-following capabilities** in code/education-related contexts  
+- Benchmarking performance of Unsloth-accelerated LLaMA models  
+
+⚠️ **Not suitable for production**. This is an **experimental finetune**.
+
+---
+
+## 🚀 Example Usage
+
+```python
+from transformers import AutoModelForCausalLM, AutoTokenizer
+
+model_name = "matteoangeloni/llama3-8b-edu"
+tokenizer = AutoTokenizer.from_pretrained(model_name)
+model = AutoModelForCausalLM.from_pretrained(model_name)
+
+prompt = "Summarize the main points of the Italian privacy law."
+inputs = tokenizer(prompt, return_tensors="pt")
+outputs = model.generate(**inputs, max_new_tokens=200)
+
+print(tokenizer.decode(outputs[0], skip_special_tokens=True))