making it a python package on PyPI

README.md

@@ -5,15 +5,97 @@ colorFrom: yellow
 colorTo: blue
 sdk: gradio
 sdk_version: 4.26.0
-app_file: app.py
+app_file: gradio_app.py
 pinned: false
 license: mit
 ---
 
-## Dev Notes
+
+# Numpy-Neuron
+
+A small, simple neural network framework built using only [numpy](https://numpy.org) and python (duh).
+Here is an example of how to use the package for training a classifier.
+
+```py
+from sklearn import datasets
+from sklearn.preprocessing import OneHotEncoder
+from sklearn.model_selection import train_test_split
+from sklearn.metrics import accuracy_score, precision_score, recall_score
+import numpy as np
+from nn import (
+    NN,
+    Relu,
+    Sigmoid,
+    CrossEntropyWithLogits,
+)
+
+
+RANDOM_SEED = 2
+
+
+def _preprocess_digits(
+    seed: int,
+) -> tuple[np.ndarray, ...]:
+    digits = datasets.load_digits(as_frame=False)
+    n_samples = len(digits.images)
+    data = digits.images.reshape((n_samples, -1))
+    y = OneHotEncoder().fit_transform(digits.target.reshape(-1, 1)).toarray()
+    X_train, X_test, y_train, y_test = train_test_split(
+        data,
+        y,
+        test_size=0.2,
+        random_state=seed,
+    )
+    return X_train, X_test, y_train, y_test
+
+
+def train_nn_classifier() -> None:
+    X_train, X_test, y_train, y_test = _preprocess_digits(seed=RANDOM_SEED)
+
+    nn_classifier = NN(
+        epochs=2_000,
+        hidden_size=16,
+        batch_size=1,
+        learning_rate=0.01,
+        loss_fn=CrossEntropyWithLogits(),
+        hidden_activation_fn=Relu(),
+        output_activation_fn=Sigmoid(),
+        input_size=64,  # 8x8 pixel grid images
+        output_size=10,  # digits 0-9
+        seed=2,
+    )
+
+    nn_classifier.train(
+        X_train=X_train,
+        y_train=y_train,
+    )
+
+    pred = nn_classifier.predict(X_test=X_test)
+
+    pred = np.argmax(pred, axis=1)
+    y_test = np.argmax(y_test, axis=1)
+
+    accuracy = accuracy_score(y_true=y_test, y_pred=pred)
+
+    print(f"accuracy on validation set: {accuracy:.4f}")
+
+
+if __name__ == "__main__":
+    train_nn_classifier()
+```
+
+
+## Roadmap
+
+**Optimizers**
+I would love to add the ability to modify the learning rate over each epoch to ensure
+that the gradient descent algorithm does not get stuck in local minima as easily.
+
+
+## Gradio app demo development notes
 
 The remote added to this repo so that it runs on hugging face spaces
 `git remote add space git@hf.co:spaces/Jensen-holm/Numpy-Neuron` 
 
 The command to force push to that space
-`git push --force space main`
+`git push --force space main`

about_package.md

@@ -0,0 +1,81 @@
+# Numpy-Neuron
+
+A small, simple neural network framework built using only [numpy](https://numpy.org) and python (duh).
+Here is an example of how to use the package for training a classifier.
+
+```py
+from sklearn import datasets
+from sklearn.preprocessing import OneHotEncoder
+from sklearn.model_selection import train_test_split
+from sklearn.metrics import accuracy_score, precision_score, recall_score
+import numpy as np
+from nn import (
+    NN,
+    Relu,
+    Sigmoid,
+    CrossEntropyWithLogits,
+)
+
+
+RANDOM_SEED = 2
+
+
+def _preprocess_digits(
+    seed: int,
+) -> tuple[np.ndarray, ...]:
+    digits = datasets.load_digits(as_frame=False)
+    n_samples = len(digits.images)
+    data = digits.images.reshape((n_samples, -1))
+    y = OneHotEncoder().fit_transform(digits.target.reshape(-1, 1)).toarray()
+    X_train, X_test, y_train, y_test = train_test_split(
+        data,
+        y,
+        test_size=0.2,
+        random_state=seed,
+    )
+    return X_train, X_test, y_train, y_test
+
+
+def train_nn_classifier() -> None:
+    X_train, X_test, y_train, y_test = _preprocess_digits(seed=RANDOM_SEED)
+
+    nn_classifier = NN(
+        epochs=2_000,
+        hidden_size=16,
+        batch_size=1,
+        learning_rate=0.01,
+        loss_fn=CrossEntropyWithLogits(),
+        hidden_activation_fn=Relu(),
+        output_activation_fn=Sigmoid(),
+        input_size=64,  # 8x8 pixel grid images
+        output_size=10,  # digits 0-9
+        seed=2,
+    )
+
+    nn_classifier.train(
+        X_train=X_train,
+        y_train=y_train,
+    )
+
+    pred = nn_classifier.predict(X_test=X_test)
+
+    pred = np.argmax(pred, axis=1)
+    y_test = np.argmax(y_test, axis=1)
+
+    accuracy = accuracy_score(y_true=y_test, y_pred=pred)
+
+    print(f"accuracy on validation set: {accuracy:.4f}")
+
+
+if __name__ == "__main__":
+    train_nn_classifier()
+```
+
+
+## Roadmap
+
+**Optimizers**
+I would love to add the ability to modify the learning rate over each epoch to ensure
+that the gradient descent algorithm does not get stuck in local minima as easily.
+
+

app.py → gradio_app.py

@@ -4,7 +4,7 @@ from sklearn.model_selection import train_test_split
 import numpy as np
 import gradio as gr
 
-import nn  # custom neural network module
+import nn
 from vis import (  # classification visualization funcitons
     show_digits,
     hits_and_misses,
@@ -84,7 +84,7 @@ def classification(
 if __name__ == "__main__":
 
     def _open_warning() -> str:
-        with open("warning.md", "r") as f:
+        with open("gradio_warning.md", "r") as f:
             return f.read()
 
     with gr.Blocks() as interface:

requirements.txt

@@ -3,4 +3,5 @@ matplotlib==3.8.4
 numpy==1.26.4
 plotly==5.21.0
 scikit_learn==1.4.2
+setuptools==69.5.1
 tqdm==4.66.2

setup.py

@@ -0,0 +1,21 @@
+from setuptools import setup, find_packages
+
+setup(
+    name="numpy_neuron",
+    version="0.3",
+    author="Jensen Holm",
+    author_email="jensen.dev.01@gmail.com",
+    description="Simple, lightweight neural network framework built in numpy",
+    long_description=open("about_package.md").read(),
+    long_description_content_type="text/markdown",
+    url="https://github.com/Jensen-holm/Numpy-Neuron",
+    project_urls={"Bug Tracker": "https://github.com/Jensen-holm/Numpy-Neuron/issues"},
+    package_dir={"": "nn"},
+    packages=find_packages(where="nn"),
+    classifiers=[
+        "Programming Language :: Python :: 3",
+        "License :: OSI Approved :: MIT License",
+        "Operating System :: OS Independent",
+    ],
+    python_requires=">=3.6",
+)