diff --git a/dl/RNN/.ipynb_checkpoints/RNN-checkpoint.ipynb b/dl/RNN/.ipynb_checkpoints/RNN-checkpoint.ipynb
new file mode 100644
index 000000000..5000bf835
--- /dev/null
+++ b/dl/RNN/.ipynb_checkpoints/RNN-checkpoint.ipynb
@@ -0,0 +1,247 @@
+{
+ "cells": [
+  {
+   "cell_type": "code",
+   "execution_count": 1,
+   "metadata": {
+    "id": "zwzjFi8-CR4a"
+   },
+   "outputs": [],
+   "source": [
+    "import tensorflow as tf\r\n",
+    "from tensorflow.keras.models import Sequential\r\n",
+    "from tensorflow.keras.layers import Dense, Dropout, LSTM"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 2,
+   "metadata": {
+    "colab": {
+     "base_uri": "https://localhost:8080/"
+    },
+    "id": "yGzOxAAjCVld",
+    "outputId": "6a0e52cc-7071-4a64-ea60-a2801097915d"
+   },
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Downloading data from https://storage.googleapis.com/tensorflow/tf-keras-datasets/mnist.npz\n",
+      "11493376/11490434 [==============================] - 0s 0us/step\n"
+     ]
+    }
+   ],
+   "source": [
+    "mnist = tf.keras.datasets.mnist\r\n",
+    "(x_train, y_train), (x_test, y_test) = mnist.load_data()"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 3,
+   "metadata": {
+    "colab": {
+     "base_uri": "https://localhost:8080/"
+    },
+    "id": "GjR_FRu3CYFa",
+    "outputId": "2ecd447c-79e0-44c4-93d3-0f52fad26eec"
+   },
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "(60000, 28, 28)\n"
+     ]
+    }
+   ],
+   "source": [
+    "print(x_train.shape)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 4,
+   "metadata": {
+    "id": "dW_UOayCCYy3"
+   },
+   "outputs": [],
+   "source": [
+    "# normalise data\r\n",
+    "x_train = x_train/255.0\r\n",
+    "x_test = x_test/255.0"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 5,
+   "metadata": {
+    "id": "Ev3eg2PcCYps"
+   },
+   "outputs": [],
+   "source": [
+    "model = Sequential()"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 6,
+   "metadata": {
+    "id": "W7OnTS_UCYem"
+   },
+   "outputs": [],
+   "source": [
+    "model.add(LSTM(128, input_shape = (28, 28), activation='relu', return_sequences=True))\r\n",
+    "model.add(Dropout(0.2))\r\n",
+    "\r\n",
+    "model.add(LSTM(128, activation='relu'))\r\n",
+    "model.add(Dropout(0.2))\r\n",
+    "\r\n",
+    "model.add(Dense(32, activation='relu'))\r\n",
+    "model.add(Dropout(0.2))\r\n",
+    "\r\n",
+    "model.add(Dense(10, activation='softmax'))"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 11,
+   "metadata": {
+    "colab": {
+     "base_uri": "https://localhost:8080/"
+    },
+    "id": "-y_DlifKEVUu",
+    "outputId": "3bfb3e71-d4d8-4267-8d20-ce10c3b173d0"
+   },
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Model: \"sequential\"\n",
+      "_________________________________________________________________\n",
+      "Layer (type)                 Output Shape              Param #   \n",
+      "=================================================================\n",
+      "lstm (LSTM)                  (None, 28, 128)           80384     \n",
+      "_________________________________________________________________\n",
+      "dropout (Dropout)            (None, 28, 128)           0         \n",
+      "_________________________________________________________________\n",
+      "lstm_1 (LSTM)                (None, 128)               131584    \n",
+      "_________________________________________________________________\n",
+      "dropout_1 (Dropout)          (None, 128)               0         \n",
+      "_________________________________________________________________\n",
+      "dense (Dense)                (None, 32)                4128      \n",
+      "_________________________________________________________________\n",
+      "dropout_2 (Dropout)          (None, 32)                0         \n",
+      "_________________________________________________________________\n",
+      "dense_1 (Dense)              (None, 10)                330       \n",
+      "=================================================================\n",
+      "Total params: 216,426\n",
+      "Trainable params: 216,426\n",
+      "Non-trainable params: 0\n",
+      "_________________________________________________________________\n"
+     ]
+    }
+   ],
+   "source": [
+    "model.summary()"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 7,
+   "metadata": {
+    "id": "6O9qZ5WRCl2d"
+   },
+   "outputs": [],
+   "source": [
+    "optimizer = tf.keras.optimizers.Adam(lr=0.001, decay=0.00005)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 8,
+   "metadata": {
+    "id": "BMz_PlqqCn3Q"
+   },
+   "outputs": [],
+   "source": [
+    "model.compile(loss='sparse_categorical_crossentropy', \r\n",
+    "    optimizer=optimizer,\r\n",
+    "    metrics=['accuracy'])"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 10,
+   "metadata": {
+    "colab": {
+     "base_uri": "https://localhost:8080/"
+    },
+    "id": "ZGyzfMlYCq5e",
+    "outputId": "e755ae08-d56b-4fac-8abd-ae991363409d"
+   },
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 1/5\n",
+      "1875/1875 [==============================] - 101s 54ms/step - loss: 0.2994 - accuracy: 0.9152 - val_loss: 0.1330 - val_accuracy: 0.9612\n",
+      "Epoch 2/5\n",
+      "1875/1875 [==============================] - 99s 53ms/step - loss: 0.1400 - accuracy: 0.9634 - val_loss: 0.0844 - val_accuracy: 0.9764\n",
+      "Epoch 3/5\n",
+      "1875/1875 [==============================] - 102s 54ms/step - loss: 0.0969 - accuracy: 0.9743 - val_loss: 0.0621 - val_accuracy: 0.9812\n",
+      "Epoch 4/5\n",
+      "1875/1875 [==============================] - 100s 53ms/step - loss: 0.0791 - accuracy: 0.9793 - val_loss: 0.0698 - val_accuracy: 0.9811\n",
+      "Epoch 5/5\n",
+      "1875/1875 [==============================] - 99s 53ms/step - loss: 0.0643 - accuracy: 0.9830 - val_loss: 0.0502 - val_accuracy: 0.9850\n"
+     ]
+    },
+    {
+     "data": {
+      "text/plain": [
+       "<tensorflow.python.keras.callbacks.History at 0x7f40fc0f4e80>"
+      ]
+     },
+     "execution_count": 10,
+     "metadata": {
+      "tags": []
+     },
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "model.fit(x_train, y_train, epochs=5, \r\n",
+    "    validation_data=(x_test, y_test))"
+   ]
+  }
+ ],
+ "metadata": {
+  "colab": {
+   "name": "RNN.ipynb",
+   "provenance": []
+  },
+  "kernelspec": {
+   "display_name": "Python 3",
+   "language": "python",
+   "name": "python3"
+  },
+  "language_info": {
+   "codemirror_mode": {
+    "name": "ipython",
+    "version": 3
+   },
+   "file_extension": ".py",
+   "mimetype": "text/x-python",
+   "name": "python",
+   "nbconvert_exporter": "python",
+   "pygments_lexer": "ipython3",
+   "version": "3.7.6"
+  }
+ },
+ "nbformat": 4,
+ "nbformat_minor": 4
+}
diff --git a/dl/RNN/RNN.ipynb b/dl/RNN/RNN.ipynb
new file mode 100644
index 000000000..5000bf835
--- /dev/null
+++ b/dl/RNN/RNN.ipynb
@@ -0,0 +1,247 @@
+{
+ "cells": [
+  {
+   "cell_type": "code",
+   "execution_count": 1,
+   "metadata": {
+    "id": "zwzjFi8-CR4a"
+   },
+   "outputs": [],
+   "source": [
+    "import tensorflow as tf\r\n",
+    "from tensorflow.keras.models import Sequential\r\n",
+    "from tensorflow.keras.layers import Dense, Dropout, LSTM"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 2,
+   "metadata": {
+    "colab": {
+     "base_uri": "https://localhost:8080/"
+    },
+    "id": "yGzOxAAjCVld",
+    "outputId": "6a0e52cc-7071-4a64-ea60-a2801097915d"
+   },
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Downloading data from https://storage.googleapis.com/tensorflow/tf-keras-datasets/mnist.npz\n",
+      "11493376/11490434 [==============================] - 0s 0us/step\n"
+     ]
+    }
+   ],
+   "source": [
+    "mnist = tf.keras.datasets.mnist\r\n",
+    "(x_train, y_train), (x_test, y_test) = mnist.load_data()"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 3,
+   "metadata": {
+    "colab": {
+     "base_uri": "https://localhost:8080/"
+    },
+    "id": "GjR_FRu3CYFa",
+    "outputId": "2ecd447c-79e0-44c4-93d3-0f52fad26eec"
+   },
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "(60000, 28, 28)\n"
+     ]
+    }
+   ],
+   "source": [
+    "print(x_train.shape)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 4,
+   "metadata": {
+    "id": "dW_UOayCCYy3"
+   },
+   "outputs": [],
+   "source": [
+    "# normalise data\r\n",
+    "x_train = x_train/255.0\r\n",
+    "x_test = x_test/255.0"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 5,
+   "metadata": {
+    "id": "Ev3eg2PcCYps"
+   },
+   "outputs": [],
+   "source": [
+    "model = Sequential()"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 6,
+   "metadata": {
+    "id": "W7OnTS_UCYem"
+   },
+   "outputs": [],
+   "source": [
+    "model.add(LSTM(128, input_shape = (28, 28), activation='relu', return_sequences=True))\r\n",
+    "model.add(Dropout(0.2))\r\n",
+    "\r\n",
+    "model.add(LSTM(128, activation='relu'))\r\n",
+    "model.add(Dropout(0.2))\r\n",
+    "\r\n",
+    "model.add(Dense(32, activation='relu'))\r\n",
+    "model.add(Dropout(0.2))\r\n",
+    "\r\n",
+    "model.add(Dense(10, activation='softmax'))"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 11,
+   "metadata": {
+    "colab": {
+     "base_uri": "https://localhost:8080/"
+    },
+    "id": "-y_DlifKEVUu",
+    "outputId": "3bfb3e71-d4d8-4267-8d20-ce10c3b173d0"
+   },
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Model: \"sequential\"\n",
+      "_________________________________________________________________\n",
+      "Layer (type)                 Output Shape              Param #   \n",
+      "=================================================================\n",
+      "lstm (LSTM)                  (None, 28, 128)           80384     \n",
+      "_________________________________________________________________\n",
+      "dropout (Dropout)            (None, 28, 128)           0         \n",
+      "_________________________________________________________________\n",
+      "lstm_1 (LSTM)                (None, 128)               131584    \n",
+      "_________________________________________________________________\n",
+      "dropout_1 (Dropout)          (None, 128)               0         \n",
+      "_________________________________________________________________\n",
+      "dense (Dense)                (None, 32)                4128      \n",
+      "_________________________________________________________________\n",
+      "dropout_2 (Dropout)          (None, 32)                0         \n",
+      "_________________________________________________________________\n",
+      "dense_1 (Dense)              (None, 10)                330       \n",
+      "=================================================================\n",
+      "Total params: 216,426\n",
+      "Trainable params: 216,426\n",
+      "Non-trainable params: 0\n",
+      "_________________________________________________________________\n"
+     ]
+    }
+   ],
+   "source": [
+    "model.summary()"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 7,
+   "metadata": {
+    "id": "6O9qZ5WRCl2d"
+   },
+   "outputs": [],
+   "source": [
+    "optimizer = tf.keras.optimizers.Adam(lr=0.001, decay=0.00005)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 8,
+   "metadata": {
+    "id": "BMz_PlqqCn3Q"
+   },
+   "outputs": [],
+   "source": [
+    "model.compile(loss='sparse_categorical_crossentropy', \r\n",
+    "    optimizer=optimizer,\r\n",
+    "    metrics=['accuracy'])"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 10,
+   "metadata": {
+    "colab": {
+     "base_uri": "https://localhost:8080/"
+    },
+    "id": "ZGyzfMlYCq5e",
+    "outputId": "e755ae08-d56b-4fac-8abd-ae991363409d"
+   },
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Epoch 1/5\n",
+      "1875/1875 [==============================] - 101s 54ms/step - loss: 0.2994 - accuracy: 0.9152 - val_loss: 0.1330 - val_accuracy: 0.9612\n",
+      "Epoch 2/5\n",
+      "1875/1875 [==============================] - 99s 53ms/step - loss: 0.1400 - accuracy: 0.9634 - val_loss: 0.0844 - val_accuracy: 0.9764\n",
+      "Epoch 3/5\n",
+      "1875/1875 [==============================] - 102s 54ms/step - loss: 0.0969 - accuracy: 0.9743 - val_loss: 0.0621 - val_accuracy: 0.9812\n",
+      "Epoch 4/5\n",
+      "1875/1875 [==============================] - 100s 53ms/step - loss: 0.0791 - accuracy: 0.9793 - val_loss: 0.0698 - val_accuracy: 0.9811\n",
+      "Epoch 5/5\n",
+      "1875/1875 [==============================] - 99s 53ms/step - loss: 0.0643 - accuracy: 0.9830 - val_loss: 0.0502 - val_accuracy: 0.9850\n"
+     ]
+    },
+    {
+     "data": {
+      "text/plain": [
+       "<tensorflow.python.keras.callbacks.History at 0x7f40fc0f4e80>"
+      ]
+     },
+     "execution_count": 10,
+     "metadata": {
+      "tags": []
+     },
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "model.fit(x_train, y_train, epochs=5, \r\n",
+    "    validation_data=(x_test, y_test))"
+   ]
+  }
+ ],
+ "metadata": {
+  "colab": {
+   "name": "RNN.ipynb",
+   "provenance": []
+  },
+  "kernelspec": {
+   "display_name": "Python 3",
+   "language": "python",
+   "name": "python3"
+  },
+  "language_info": {
+   "codemirror_mode": {
+    "name": "ipython",
+    "version": 3
+   },
+   "file_extension": ".py",
+   "mimetype": "text/x-python",
+   "name": "python",
+   "nbconvert_exporter": "python",
+   "pygments_lexer": "ipython3",
+   "version": "3.7.6"
+  }
+ },
+ "nbformat": 4,
+ "nbformat_minor": 4
+}
diff --git a/dl/RNN/requirements.txt b/dl/RNN/requirements.txt
new file mode 100644
index 000000000..4cff948f3
--- /dev/null
+++ b/dl/RNN/requirements.txt
@@ -0,0 +1,5 @@
+1 keras == 2.3.1
+2 scikit-learn == 0.22.1
+3 matplotlib == 3.1.3
+4 numpy == 1.18.1
+5 pandas == 1.0.3
\ No newline at end of file
diff --git a/dl/RNN/rnn.py b/dl/RNN/rnn.py
new file mode 100644
index 000000000..7d532388e
--- /dev/null
+++ b/dl/RNN/rnn.py
@@ -0,0 +1,32 @@
+import tensorflow as tf
+from tensorflow.keras.models import Sequential
+from tensorflow.keras.layers import Dense, Dropout, LSTM
+
+def rnn_model(input_shape, x_train, y_train, x_test, y_test, epochs=0, loss='sparse_categorical_crossentropy', metric='accuracy', lr=0):
+    model = Sequential([
+
+            LSTM(128, input_shape=input_shape, activation='relu', return_sequences=True),
+            Dropout(0.2),
+
+            LSTM(128, activation='relu'), 
+            Dropout(0.2),
+
+            Dense(32, activation='relu'),
+            Dropout(0.2),
+
+            Dense(10, activation='softmax')
+
+        ])
+    print(model.summary())
+
+
+    optimizer = tf.keras.optimizers.Adam(lr=lr, decay=0.00005)
+    model.compile(loss=loss, 
+            optimizer=optimizer, 
+            metrics=[metric])
+
+    model.fit(x_train, y_train, epochs=epochs, validation_data=(x_test, y_test))
+
+
+# example
+# rnn_model((28, 28),x_train, y_train, x_test, y_test, epochs=5, loss='sparse_categorical_crossentropy', metric='accuracy', lr=0.01)
\ No newline at end of file