add readme better

README.md

@@ -9,5 +9,41 @@ app_file: app.py
 pinned: false
 short_description: 'A sarcastic GPT-Neo chatbot that roasts your questions with '
 ---
+# S.N.A.R.K – Sarcastic Neural Algorithm for Ridiculing Knowledge 🤖🖕
 
-An example chatbot using [Gradio](https://gradio.app), [`huggingface_hub`](https://huggingface.co/docs/huggingface_hub/v0.22.2/en/index), and the [Hugging Face Inference API](https://huggingface.co/docs/api-inference/index).
+**S.N.A.R.K** is a fine-tuned GPT-Neo language model trained to deliver witty, sarcastic responses to your prompts. Whether you're asking something profound or profoundly stupid, this model has something snarky to say about it.
+
+---
+
+## 🧠 Model
+This Space uses a locally fine-tuned **GPT-Neo** model, customized to produce sarcastic completions using carefully curated prompts and generation settings. It was trained with humor, bitterness, and just a pinch of disdain.
+
+---
+
+## 💬 How to Use
+Type your prompt into the input box (try something absurd), and S.N.A.R.K will generate a sarcastic response.
+
+Example prompts:
+- "What’s the best way to become a millionaire overnight?"
+- "Why is my code not working?"
+- "Is pineapple on pizza acceptable?"
+
+---
+
+## ⚙️ Technical Info
+- Model: GPT-Neo (fine-tuned)
+- Hosted on: Hugging Face Spaces with Gradio
+- Max new tokens: 60
+- Temperature: 0.9
+- Top-p: 0.95
+- Repetition penalty: 1.2
+- Stop token: `<|stop|>`
+
+---
+
+## 🚨 Warning
+This model is sarcastic by design. Don't take anything it says too seriously — or do, but don't blame us.
+
+---
+
+Created with 🤖 and a lot of eye rolls by [Alexandre Lagasse](https://huggingface.co/Alexandrelgss)