SparkNLP introducing Phi-3

[prabod]

This PR introduces Phi-3

Description
This pull request integrates the Phi-3 model into the existing suite of NLP tools. Phi-3 is a Transformer-based model with 3 billion parameters, optimized for advanced natural language understanding tasks. This model builds upon the success of its predecessor, Phi-2, by incorporating additional training on a wider array of texts and improved architecture tweaks that enhance performance on a variety of NLP benchmarks.

Motivation and Context

The introduction of Phi-3 addresses the growing demand for more efficient and powerful NLP models that require less computational resources without compromising on performance. Phi-3 offers a significant improvement in handling complex NLP tasks such as text summarization, sentiment analysis, and language inference, making it a valuable addition for developers and researchers looking to push the boundaries of what's possible in NLP.

Screenshots (if appropriate):

Types of changes

• New feature (non-breaking change which adds functionality)

Checklist:

• My code follows the code style of this project.
• My change requires a change to the documentation.
• I have updated the documentation accordingly.
• I have read the CONTRIBUTING page.
• I have added tests to cover my changes.
• All new and existing tests passed.