Numpy-Neuron
A small, simple neural network framework built using only numpy and python (duh).
Install
pip install numpyneuron
Example
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 numpyneuron 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
Currently the learning rate in a NN object is static during training. I would like to work on developing at least the functionality for the Adam optimizer at some point. This would help prevent getting stuck in local minima of the loss function.
Metadata
Release files for numpyneuron 0.6
For a detailed explanation of source distributions (sdists) and built distributions (wheels), please see the package formats documentation.
Source distribution (sdist)
| File | Size | Uploaded | Provenance | Actions |
|---|---|---|---|---|
| numpyneuron-0.6.tar.gz | 5.5 kB | Without provenance | Details for numpyneuron-0.6.tar.gz |
Built distribution (wheel)
| File | Interpreter | ABI | Platform | Actions Reset |
|---|---|---|---|---|
| numpyneuron-0.6-py3-none-any.whl | Python 3 | none | any | Details for numpyneuron-0.6-py3-none-any.whl |
Total release size: 11.1 kB
Release files / numpyneuron-0.6.tar.gz
| Download URL | numpyneuron-0.6.tar.gz |
|---|---|
| Size | 5.5 kB |
| Tags | Source |
|
SHA-256 checksum How to use checksums |
6f2aecda0e8771cd2da936faf8eb422e38d749cbd0b35ab23e5fb5d5ca5a0f25
|
|
BLAKE2b-256 checksum How to use checksums |
7eb56a3b5bcf1757a691842169a9eea1411e3afb7c2043e0005d9a65643d5698
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
No |
| Uploaded via |
twine/5.0.0 CPython/3.12.3
|
Release files / numpyneuron-0.6-py3-none-any.whl
| Download URL | numpyneuron-0.6-py3-none-any.whl |
|---|---|
| Size | 5.6 kB |
| Tags | Python 3 |
|
SHA-256 checksum How to use checksums |
7e2d43dd50f973153ac82db488ef12862a87abdf8ab85e8b60044fe5c45d4893
|
|
BLAKE2b-256 checksum How to use checksums |
baf82851b3fcfbc3e0d1d309fd09ab8c0c8d2fd54bb6869be4d5b57509604d39
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
No |
| Uploaded via |
twine/5.0.0 CPython/3.12.3
|
Security for numpyneuron 0.6
Provenance
Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.
No provenance found
None of the files in this release were uploaded with provenance metadata.
A note on package security
PyPI is a repository for distributing packages, not a service for auditing them. While third-party security researchers and our staff do their best to detect and remove malicious packages, PyPI cannot guarantee the quality, security or suitability of any given package.
To learn more about protecting yourself from malicious actors, visit our package security documentation.