A Python Keras model API
Project description
aitk.keras
An implementation of the main Keras API with the layers in numpy.
UNDER DEVELOPMENT
Why?
- useful to explain deep learning
- can be used where tensorflow is not available (eg, JupterLite)
Features
- supports Keras's Sequential and functional APIs
- alternative dataset downloader for JupyterLite
Examples:
# Classic XOR
from aitk.keras.layers import Input, Dense
from aitk.keras.models import Sequential
inputs = [[0, 0], [0, 1], [1, 0], [1, 1]]
targets = [[0], [1], [1], [0]]
model = Sequential()
model.add(Input(2, name="input"))
model.add(Dense(8, activation="tanh", name="hidden"))
model.add(Dense(1, activation="sigmoid", name="output"))
model.compile(optimizer="adam", loss="mse")
outputs = model.predict(inputs)
model.fit(inputs, targets, epochs=epochs, verbose=0, shuffle=False)
See the notebook directory for additional examples.
See also the examples in the tests folder.
Development
- implement shuffle
- report metrics to logs/history
- probably lots of edge cases ar broken
- see "FIXME" items in code
To run the tests:
$ pytest -vvv tests
Please feel free to report issues and make Pull Requests!
References
Lowlevel numpy code based on numpy_ml.
Release history Release notifications | RSS feed
Download files
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.
Source Distributions
No source distribution files available for this release.See tutorial on generating distribution archives.
Built Distribution
aitk.keras-0.3.5-py3-none-any.whl
(172.3 kB
view details)
File details
Details for the file aitk.keras-0.3.5-py3-none-any.whl.
File metadata
- Download URL: aitk.keras-0.3.5-py3-none-any.whl
- Upload date:
- Size: 172.3 kB
- Tags: Python 3
- Uploaded using Trusted Publishing? No
- Uploaded via: twine/3.4.1 importlib_metadata/4.5.0 pkginfo/1.7.0 requests/2.24.0 requests-toolbelt/0.9.1 tqdm/4.61.0 CPython/3.8.5
File hashes
| Algorithm | Hash digest | |
|---|---|---|
| SHA256 | d6b0a326a6fa8dfc8cc9b363bbf70ffa66620ae4dca9076db7fb92f137eb3e5c |
|
| MD5 | d5e0176718b9c41b4c95fb2e5a4152ef |
|
| BLAKE2b-256 | 7ecf31bd422f6a534a082d6b617b6236c2565d7e7128a05a04d5dcccd5057630 |
