TL;DR

Add "chonkie[semantic]" to the installs and add the new chunking params during init of the SemanticChunker. Also use potion-base-32M for improved chunk quality.

Copilot summary

This pull request includes updates to the tutorials/semantic_chunking.ipynb file to improve the tutorial's functionality and performance. The most important changes include updating the library installation command, modifying the model used for chunking, and adjusting the execution counts and outputs for various cells.

Library and model updates:

• Updated the library installation command to include the chonkie[semantic] package.
• Changed the embedding model from minishlab/potion-base-8M to minishlab/potion-base-32M and adjusted chunking parameters for better performance.

Execution counts and outputs:

• Updated execution counts for multiple cells to reflect the new order of execution. [1] [2] [3] [4]
• Modified the outputs of cells to reflect the new chunking results and performance metrics. [1] [2]

Metadata updates:

• Updated the kernel display name and Python version in the notebook metadata. [1] [2]