Panml

What's changed

wanoz
- adding prompt templates
- added future annotation import for supprting Python 3.8 or lower
- added GPU setting for training and inference for HuggingFace models
- updated setup.py

What's changed

Hanshan1988:
- Fixed error in flan t5 embeddings
- Added HuggingFace Hub model support for bert, distilbert, roberta and distilroberta base models

What's changed

wanoz:
- added support for more llms from HuggingFace Hub
- added separate function to save models loaded from HuggingFace Hub
- general refactoring (converted base_models into submodules of core, and then llms and clustering)
- allow x and y inputs to ModelPack.fit to be pd.Series type
- fix: removed model argument from datacollator
- fix: huggingface transformer and accelerate library versioning

What's changed

- updated FAISS to include IVFFlat and IVFPQ implementations, with unittests (wanoz)
- added setup.py and setup.cfg files for pypi publishing (wanoz)
- refactored models and search class into individual submodules (wanoz)

What's changed
- added and updated files for pypi distribution and packaging (wanoz)
- updated gitignore for distribution and packaging (wanoz)
- add exception handling for OpenAI predict function (wanoz)
- added storing and loading vector store, updated doc string (wanoz)
- added more performant selection of HF embedding models for vector search (wanoz)

What's changed

- added ability to predict code using pandas dataframe (wanoz)
- add text-davinci-003 to supported OpenAI models for code generation (wanoz)
- updated requirements.txt (wanoz)