Update README.md

README.md

@@ -1,14 +1,29 @@
----
-title: SHEFAA Arabic Medical QA
-emoji: 😻
-colorFrom: green
-colorTo: red
-sdk: gradio
-sdk_version: 5.13.1
-app_file: app.py
-pinned: false
-license: apache-2.0
-short_description: SHEFAA is an Arabic medical question-answering system
----
-
-Check out the configuration reference at https://huggingface.co/docs/hub/spaces-config-reference
+# 🚑 SHEFAA - Arabic Medical AI Assistant
+
+This is a **Hugging Face Space** that allows users to ask **medical questions in Arabic** and receive AI-generated responses. The model is based on **E-dd77/SHEFAA**, a fine-tuned version of **Jais-family-256m** using QLoRA.
+
+### 🔧 How It Works
+- Enter a medical question in **Arabic**.
+- The model processes the input and generates an answer.
+- The response is displayed on the interface.
+
+### 🚀 Deployment
+This app is hosted on **Hugging Face Spaces** using `gradio`.
+
+### 📌 Model Details
+- **Base Model:** `inceptionai/Jais-family-256m`
+- **Fine-tuned Model (LoRA):** `E-dd77/SHEFAA`
+- **Quantization:** 4-bit using `bitsandbytes`
+- **Inference:** Using `transformers` and `peft`
+
+### 📌 Requirements
+- `gradio`
+- `torch`
+- `transformers`
+- `accelerate`
+- `bitsandbytes`
+- `peft`
+- `tqdm`
+
+### 🔥 Try It Now!
+Click **"Open in Spaces"** to start using the model.