Dataeval

🌟 **Feature Release**
- `7ad22a7c` - Switch all stats classes to use dataset inputs and change `datasetstats` to `imagestats`
- `e1da1768` - Integrate meta_distribution_compare as metadata_distance

🛠️ **Improvements and Enhancements**
- `fc8b68c2` - Spike: Define which metrics comprise completeness metrics

📝 **Miscellaneous**
- `70a8e1e2` - Create lightweight dataset wrapper factory functions
- `fef938d5` - Simplify typing and add more docstrings to dataeval.typing module

🌟 **Feature Release**
- `101fea34` - Add selection feature for datasets

📝 **Miscellaneous**
- `efc90630` - Add index based selection helper
- `666e2866` - Add collate helper functions for Dataloaders
- `4e92ae90` - Restructure typing for Datasets to allow better extensibility for other data classes

🌟 **Feature Release**
- `94a10b03` - Refactor Images, Embeddings and Metadata as a stateful classes using Dataset inputs
- `353ac6bd` - Add DataProcessor class to handle extraction of images, embeddings, targets and metadata from datasets

🛠️ **Improvements and Enhancements**
- `529e5595` - Merging current WIP state of OODdetector, based on universal embeddings from sigma-optimal-VAE, along with OOD_VAE_minimal notebook.

👾 **Fixes**
- `a2a7057e` - Adjust merge to handle numpy arrays

Adds in a numpy array check and if true returns the array as a list prior to normal processing

📝 **Miscellaneous**
- `e55cb3f2` - Use cpu as default torch device

🌟 **Feature Release**
- `841425ff` - Release Metadata OOD function most_deviated_factors

Adds a new explanatory function using Metadata and an OODOutput

🌟 **Feature Release**
- `bff82522` - Add collate function and convert packaged datasets to MAITE protocols

* Changes all dataset utility classes to use `MAITE` protocol formats (`MNIST`, `CIFAR10`, and `VOCDetection`)
* Addes `collate` to aggregate (and encode) `MAITE`datasets into images/embeddings, targets, and metadata

🛠️ **Improvements and Enhancements**
- `d9e0f8b0` - Enforce embeddings on functions/methods that take embedding inputs

🛠️ **Improvements and Enhancements**
- `9a420f7d` - Update Assess the data space tutorial to fit JATIC DR-2.3
- `3ab63f3e` - Integrate clusterer speed improvements with numba