Pyepo

<p align="center"><img width="30%" src="images/logo1.png" /></p>

We're happy to announce the 0.3.3 release.

We fix the sign bug of `pyepo.func.NCE`, and add modules `pyepo.func.contrastiveMAP` and `pyepo.func.negativeIdentity`. See details in our [docs](https://khalil-research.github.io/PyEPO/build/html/content/examples/function.html#noise-contrastive-estimation-nce).

There are the corresponding papers [Contrastive losses and solution caching for predict-and-optimize](https://www.ijcai.org/proceedings/2021/390) and [Backpropagation through combinatorial algorithms: Identity with projection works](https://arxiv.org/abs/2205.15213).

<p align="center"><img width="30%" src="images/logo1.png" /></p>

We're happy to announce the 0.3.0 release.

Thank ijskar to add new end-to-end predict-then-optimize methods: noise contrastive estimation and learning to rank. We add modules `pyepo.func.NCE`, `pyepo.func.pointwiseLTR`, `pyepo.func.pairwiseLTR` and `pyepo.func.listwiseLTR`. See details in our [docs](https://khalil-research.github.io/PyEPO/build/html/content/examples/function.html#noise-contrastive-estimation-nce).

There are the corresponding papers [Contrastive losses and solution caching for predict-and-optimize](https://www.ijcai.org/proceedings/2021/390) and [Decision-focused learning: through the lens of learning to rank](https://proceedings.mlr.press/v162/mandi22a.html).

<p align="center"><img width="30%" src="images/logo1.png" /></p>

We're happy to announce the 0.2.4 release.

This new version includes several bug fixes and improvements that should enhance the stability and usability of the package.

<p align="center"><img width="30%" src="images/logo1.png" /></p>

We're happy to announce the 0.2.0 release.

We add two end-to-end predict-then-optimize methods with stochastical perturbation: Differentiable Perturbed Optimizers and Fenchel-Young loss with Perturbation into our package "PyEPO."

People now are allowed to use PyTorch module `pyepo.func.perturbedOpt` and `pyepo.func.perturbedFenchelYoung`. See details in our [docs](https://khalil-research.github.io/PyEPO/build/html/content/examples/function.html#differentiable-perturbed-optimizer-dpo).

Both approaches come from Google Research's awesome project [Differentiable Optimizers with Perturbations in Tensorflow](https://github.com/google-research/google-research/tree/master/perturbations), and there is the corresponding paper [Learning with differentiable perturbed optimizers](https://papers.nips.cc/paper/2020/hash/6bb56208f672af0dd65451f869fedfd9-Abstract.html).

We're happy to announce the 0.1.0 release.
It's the first release version of "PyEPO".