Metadata-Version: 2.4
Name: keras-proba
Version: 0.1.0
Summary: Uncertainty awareness in deep learning with multi-backend Keras 3
Author-email: Muhammad Anas Raza <memanasraza@gmail.com>
License: MIT
Requires-Python: >=3.10
Description-Content-Type: text/markdown
License-File: LICENSE
Requires-Dist: keras>=3.0.0
Requires-Dist: numpy
Provides-Extra: torch
Requires-Dist: torch; extra == "torch"
Provides-Extra: jax
Requires-Dist: jax; extra == "jax"
Requires-Dist: jaxlib; extra == "jax"
Provides-Extra: tensorflow
Requires-Dist: tensorflow; extra == "tensorflow"
Provides-Extra: dev
Requires-Dist: pytest; extra == "dev"
Requires-Dist: tensorflow; extra == "dev"
Requires-Dist: tf_keras; extra == "dev"
Requires-Dist: jax; extra == "dev"
Requires-Dist: torch; extra == "dev"
Dynamic: license-file

# Keras Probability: Uncertainty Awareness in Deep Learning

Documentation: https://anas-rz.github.io/keras-proba/

Install directly from GitHub:

```bash
pip install "git+https://github.com/anas-rz/keras-proba.git"
```

## Layers

- [ ] `RandomFourierFeatures`
- [ ] `GaussianProcess` output layer
- [ ] `SpectralNormalization`
- [ ] `GaussianHead` for probabilistic regression

## Post-processing

- [ ] Temperature scaling for classification calibration
- [ ] Split conformal prediction for classification and regression

## Models and Examples

- [ ] SNGP classifier example
- [ ] SNGP regressor example
- [ ] Calibrated classifier example
- [ ] Conformal prediction example

## Tests

- [ ] Parity tests with original implementation
- [ ] Training tests across backends
- [ ] Smoke tests across backends
- [ ] Serialization and reload tests
