diff --git a/.pre-commit-config.yaml b/.pre-commit-config.yaml
index 99472ac..f4b0fae 100644
--- a/.pre-commit-config.yaml
+++ b/.pre-commit-config.yaml
@@ -1,6 +1,6 @@
 repos:
 -   repo: https://github.com/psf/black
-    rev: '24.2.0'
+    rev: '24.4.2'
     hooks:
     -   id: black
     -   id: black-jupyter
@@ -24,13 +24,13 @@
     -   id: pydocstyle
         args: ['--convention=google', '--add-ignore=D100,D101,D102,D103,D104,D105,D106,D107']
 -   repo: https://github.com/pre-commit/pre-commit-hooks
-    rev: 'v4.5.0'
+    rev: 'v4.6.0'
     hooks:
     -   id: check-ast
     -   id: end-of-file-fixer
     -   id: trailing-whitespace
 -   repo: https://github.com/sqlfluff/sqlfluff
-    rev: '3.0.0a5'
+    rev: '3.0.5'
     hooks:
     -   id: sqlfluff-fix
     -   id: sqlfluff-lint
diff --git a/Makefile b/Makefile
index a874a05..784d87a 100644
--- a/Makefile
+++ b/Makefile
@@ -1,7 +1,7 @@
 .EXPORT_ALL_VARIABLES:
 .PHONY: venv install pre-commit clean
 
-GLOBAL_PYTHON = $(shell python3.9 -c 'import sys; print(sys.executable)')
+GLOBAL_PYTHON = $(shell python3.10 -c 'import sys; print(sys.executable)')
 LOCAL_PYTHON = ./.venv/bin/python
 LOCAL_PRE_COMMIT = ./.venv/lib/python3.10/site-packages/pre_commit
 
diff --git a/models/__init__.py b/models/__init__.py
new file mode 100644
index 0000000..e69de29
diff --git a/models/dataset.py b/models/dataset.py
new file mode 100644
index 0000000..dd7a22e
--- /dev/null
+++ b/models/dataset.py
@@ -0,0 +1,47 @@
+from typing import Tuple
+
+import numpy as np
+import tensorflow as tf
+from sklearn.model_selection import train_test_split
+
+
+def divide_data(data: np.array, test_size) -> Tuple[np.array]:
+    train, test = train_test_split(data, test_size=test_size, random_state=1900730)
+    train, val = train_test_split(train, test_size=test_size, random_state=1900730)
+    return train, val, test
+
+
+def normalize_standarize(data, mean, std):
+    return (data - mean) / (std + 1e-6)
+
+
+def flatten_images(images):
+    return images.reshape(images.shape[0], -1)
+
+
+def compute_mean_std(image_data):
+
+    # Compute mean and standard deviation along the examples axis
+    mean = np.mean(image_data, axis=0)
+    std = np.std(image_data, axis=0)
+
+    return mean.reshape(1, -1), std.reshape(1, -1)
+
+
+def get_data(test_size: float = 0.15) -> Tuple[np.array]:
+    # Load the MNIST dataset from TFDS
+    (x_train, _), (x_test, _) = tf.keras.datasets.mnist.load_data()
+    # Merge the data for division
+    data = np.vstack((x_train, x_test))  # (70000,28,28)
+
+    x_train, x_val, x_test = divide_data(data, test_size)
+    x_train, x_val, x_test = [flatten_images(data) for data in [x_train, x_val, x_test]]
+    mean, std = compute_mean_std(x_train)
+    # print(mean,std)
+    data = [normalize_standarize(data, mean, std) for data in [x_train, x_val, x_test]]
+    # print(x_train_norm.shape , x_val_norm.shape,x_test_norm.shape)
+    return data[0], data[1], data[-1]
+
+
+if __name__ == "__main__":
+    pass
diff --git a/models/model.py b/models/model.py
new file mode 100644
index 0000000..eb1b69b
--- /dev/null
+++ b/models/model.py
@@ -0,0 +1,73 @@
+from typing import Dict
+
+from tensorflow.python.keras.layers import Dense, Input
+from tensorflow.python.keras.models import Model
+
+
+class AutoEncoder(object):
+    def __init__(self, config: Dict[str, str]):
+        self.config = config
+        self.input_shape = self.config["train"]["input_shape"]
+        self.Encoder = None
+        self.Decoder = None
+        self.AE = None
+        self.Model_Encoder = None
+        self.input_layer = Input(shape=(self.input_shape,))
+        self.output_layer = Dense(units=self.input_shape)
+
+    def build_model(self):
+        encoder_layers = self._get_encoder_layers()
+
+        # initialize the Encoder Model
+        self.Encoder = encoder_layers[0](self.input_layer)
+        for layer in encoder_layers[1:]:
+            self.Encoder = layer(self.Encoder)
+
+        # initialize the Decoder Model
+        decoder_layers = self._get_decoder_layers()
+        self.Decoder = decoder_layers[0](self.Encoder)
+        for layer in decoder_layers[1:]:
+            self.Decoder = layer(self.Decoder)
+        self.Decoder = self.output_layer(self.Decoder)
+
+        self.Model_Encoder = Model(inputs=self.input_layer, outputs=self.Encoder)
+        self.AE = Model(inputs=self.input_layer, outputs=self.Decoder)
+
+    def fit(
+        self,
+        x_train,
+        x_val,
+    ):
+
+        if self.AE is None:
+            raise ValueError("Build The Model First")
+
+        self.AE.compile(
+            optimizer=self.config["train"]["optimizer"],
+            loss=self.config["train"]["loss"],
+            metrics=self.config["train"]["metrics"],
+        )
+        print(self.AE.summary())
+        return self.AE.fit(
+            x=x_train,
+            y=x_train,
+            batch_size=self.config["train"]["batches"],
+            epochs=self.config["train"]["epochs"],
+            validation_data=(x_val, x_val),
+            verbose=True,
+            use_multiprocessing=True,
+        )
+
+    def _get_encoder_layers(self):
+        layers = self.config["train"]["autoencoder"]["encoder"]["layers_units"]
+        activation = self.config["train"]["autoencoder"]["encoder"]["activations"]
+
+        return_layers = [Dense(units=unit, activation=activation) for unit in layers]
+        return return_layers
+
+    def _get_decoder_layers(self):
+        layers = self.config["train"]["autoencoder"]["decoder"]["layers_units"]
+        activation = self.config["train"]["autoencoder"]["decoder"]["activations"]
+
+        return_layers = [Dense(units=unit, activation=activation) for unit in layers]
+        return return_layers
diff --git a/models/params.yaml b/models/params.yaml
new file mode 100644
index 0000000..1876708
--- /dev/null
+++ b/models/params.yaml
@@ -0,0 +1,19 @@
+train:
+  input_shape: 784
+  batches: 32
+  optimizer: adam
+  metrics: ['accuracy']
+  loss: mean_squared_error
+  epochs: 50
+  autoencoder:
+    encoder:
+      layers_units: [512,256,128,64,32]
+      activations: relu
+    decoder:
+      layers_units: [64,128,256,512]
+      activations: relu
+
+data:
+  mean: 0
+  std: 255
+  random_status: 1900730
diff --git a/models/train.py b/models/train.py
new file mode 100644
index 0000000..c94005e
--- /dev/null
+++ b/models/train.py
@@ -0,0 +1,33 @@
+import argparse
+
+import pandas as pd
+import yaml
+from dataset import get_data
+from model import AutoEncoder
+from tensorflow.python.keras.models import save_model
+
+
+def read_params(config_path):
+    with open(config_path) as yaml_file:
+        config = yaml.safe_load(yaml_file)
+    return config
+
+
+if __name__ == "__main__":
+    args = argparse.ArgumentParser()
+    args.add_argument("--config", default="params.yaml")
+    parsed_args = args.parse_args()
+    print(parsed_args.config)
+    config = read_params(parsed_args.config)
+    (
+        train,
+        val,
+        test,
+    ) = get_data()
+    print(config)
+
+    model = AutoEncoder(config)
+    model.build_model()
+    history = model.fit(train, val)
+    pd.DataFrame(history.history).to_csv("./history.csv")
+    save_model(model.AE, "./model_files/")
diff --git a/pyproject.toml b/pyproject.toml
index 292a33b..0ec0498 100644
--- a/pyproject.toml
+++ b/pyproject.toml
@@ -5,7 +5,7 @@ description = "All You Need to Know about Computational Intelligence"
 authors = ["Ahmed Elsayed"]
 
 [tool.poetry.dependencies]
-python = ">=3.9,<3.10.dev0"
+python = ">=3.10,<3.11.dev0"
 pandas = "^1"
 tensorflow = "^2.15.0.post1"
 scipy = "^1.12.0"
@@ -16,6 +16,11 @@ pre-commit = "^3.7.0"
 visdom = "^0.2.4"
 notebook = "^7.1.3"
 jupyterlab = "^4.1.6"
+pipreqs = "0.4.0"
+tensorflow-datasets = "^4.9.4"
+update = "^0.0.1"
+scikit-learn = "^1.4.2"
+dvc = "^3.50.1"
 
 [tool.poetry.group.dev.dependencies]
 black = "*"