added keras tuner and updated readme

Model Tuning/Keras Tuner/assignment.ipynb

@@ -0,0 +1,70 @@
+{
+ "cells": [
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "## 📝 Assignment: Hyperparameter Tuning with Keras Tuner\n",
+    "\n",
+    "### Objective:\n",
+    "Learn how to use *Keras Tuner* to optimize the hyperparameters of a simple Convolutional Neural Network (CNN) on the MNIST dataset.\n",
+    "\n",
+    "---\n",
+    "\n",
+    "### 🚀 Tasks:\n",
+    "\n",
+    "1. *Install Keras Tuner*  \n",
+    "   - Ensure you have keras-tuner installed using:  \n",
+    "     \n",
+    "     !pip install keras-tuner\n",
+    "     \n",
+    "   \n",
+    "2. *Load and Normalize MNIST Dataset*  \n",
+    "   - Load the MNIST dataset using keras.datasets.mnist.load_data().  \n",
+    "   - Normalize pixel values between 0 and 1.  \n",
+    "   - Expand dimensions to match CNN input format.\n",
+    "\n",
+    "3. *Define a CNN Model for Hyperparameter Tuning*  \n",
+    "   - Use keras.Sequential() to define a CNN.  \n",
+    "   - Add two Conv2D layers with tunable filter sizes (hp.Choice()).\n",
+    "   - Add a Dropout layer with a tunable dropout rate.\n",
+    "   - Compile the model with a tunable learning rate.\n",
+    "\n",
+    "4. *Use Keras Tuner to Find the Best Hyperparameters*  \n",
+    "   - Use kt.RandomSearch to optimize hyperparameters.  \n",
+    "   - Run the search for 3 trials to save time.  \n",
+    "   - Set epochs=2 during tuning.\n",
+    "\n",
+    "5. *Train the Best Model*  \n",
+    "   - Retrieve the best hyperparameters using tuner.get_best_hyperparameters(num_trials=1).  \n",
+    "   - Train the best model on the dataset for 3 epochs.  \n",
+    "\n",
+    "6. *Compare Results Before and After Tuning*  \n",
+    "   - Print the *accuracy* before and after tuning.  \n",
+    "   - Plot the loss curves using matplotlib.\n",
+    "\n",
+    "---\n",
+    "\n",
+    "### 📌 Bonus Challenge:\n",
+    "- Try using *Hyperband Tuner* instead of RandomSearch and compare results.\n",
+    "- Adjust max_trials and executions_per_trial to see how it impacts training time.\n",
+    "\n",
+    "---\n",
+    "\n",
+    "💡 *Hint:* Refer to the main notebook for code structure. Modify parameters and observe the impact on model accuracy! 🎯"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": []
+  }
+ ],
+ "metadata": {
+  "language_info": {
+   "name": "python"
+  }
+ },
+ "nbformat": 4,
+ "nbformat_minor": 2
+}