feat: packaging, project structure + updated README.md

.gitignore

@@ -3,3 +3,5 @@ __pycache__
 *.npz
 *.npy
 .venv
+dist
+*.egg-info

README.md

@@ -2,24 +2,33 @@
 
 ## Usage 
 
-1. Install requirements :
+1. To install from source :
 ```sh
-$ pip install -r requirements.txt
+$ git clone git@github.com:lenoctambule/autoencoder.git
+$ pip install -e autoencoder/
 ```
 
-2. Optionally run mnist_test.py.
+2. Optionally, run mnist_test.py to see it in action on the MNIST dataset.
 ```sh
+$ cd examples
 $ py mnist_test.py 
 ```
 
 ## Training
 
-Instatiate an `Autoencoder` object :
+Instatiate an `ClassicalAutoencoder` or `VariationalAutoencoder` object :
 ```py
-from autoencoder import Autoencoder
-from activations import LeakyReLU
+from easyvae.autoencoder import ClassicalAutoencoder, VariationalAutoencoder
+from easyvae.activations import LeakyReLU
 
-autoencoder = Autoencoder(
+autoencoder = ClassicalAutoencoder(
+    [768, 64, 16],
+    [16, 64, 768],
+    0.01,
+    LeakyReLU()
+)
+# or
+autoencoder = VariationalAutoencoder(
     [768, 64, 16],
     [16, 64, 768],
     0.01,
@@ -30,11 +39,17 @@ And then via the `train_dataset` method to train over a dataset :
 ```py
 autoencoder.train_dataset(data)
 ```
-Or via the `train` to input each data points iteratively :
+Or via the `train` method to input each data points iteratively :
 ```py
 autoencoder.train(v)
 ```
 
+After training, you can save your model via the `save` method and load that model using `load` method :
+```
+autoencoder.save("mymodel.npy)
+autoencoder.load("mymodel.npy")
+```
+
 ## Inference
 
 Use your `Autoencoder` object with the `encode`, `decode`, `forward` methods like so :

examples/mnist_test.py

@@ -2,10 +2,12 @@ import matplotlib.pyplot as plt
 import numpy as np
 import os
 import signal
-from autoencoder import (VariationalAutoencoder, # noqa
-                         ClassicalAutoencoder,
-                         AAutoencoder)
-from activations import LeakyReLU
+from easyvae.autoencoder import ( # noqa
+        VariationalAutoencoder,
+        ClassicalAutoencoder,
+        AAutoencoder
+    )
+from easyvae.activations import LeakyReLU
 
 
 def load_mnist() -> list[np.ndarray]:

pyproject.toml

@@ -0,0 +1,23 @@
+[project]
+name = "easyvae"
+version = "1.0"
+authors = [
+  { name="Ravaka RALAMBOARIVONY", email="ravaka.rlb.pro@gmail.com" },
+]
+description = "Python implementation of a Classical and Variational Autoencoders "
+readme = "README.md"
+requires-python = ">=3.10"
+classifiers = [
+    "Programming Language :: Python :: 3",
+    "Operating System :: OS Independent",
+]
+license = "MIT"
+license-files = ["LICEN[CS]E*"]
+
+[project.urls]
+Homepage = "https://github.com/lenoctambule/autoencoder"
+Issues = "https://github.com/lenoctambule/autoencoder/issues"
+
+[build-system]
+requires = ["setuptools"]
+build-backend = "setuptools.build_meta"

src/easyvae/autoencoder.py

@@ -1,10 +1,12 @@
 import numpy as np
-from utils import (dynamic_loss_plot_init,
-                   dynamic_loss_plot_update,
-                   dynamic_loss_plot_finish)
+from .utils import (
+        dynamic_loss_plot_init,
+        dynamic_loss_plot_update,
+        dynamic_loss_plot_finish
+    )
 from tqdm import tqdm
-from layers import DeepNNLayer, SampleLayer
-from activations import ActivationFunc, Identity
+from .layers import DeepNNLayer, SampleLayer
+from .activations import ActivationFunc, Identity
 from abc import ABC, abstractmethod
 
 LOADER = ['⡿', '⣟', '⣯', '⣷', '⣾', '⣽', '⣻', '⢿']

src/easyvae/layers.py

@@ -1,5 +1,5 @@
 import numpy as np
-from activations import ActivationFunc, Identity
+from .activations import ActivationFunc, Identity
 
 
 class NNLayer: