Semantic Chunking

Splits the text based on semantic similarity. Check the official LangChain documentation page

Settings

breakpoint_threshold_type must be one between:

the recommended values by langchain for breakpoint_threshold_amount are: "percentile": 95 "standard_deviation": 3 "interquartile": 1.5

NOTE: The SemanticChunker is an experimental component in LangChain, it couldn't work or produce some errors.

Name		Name	Last commit message	Last commit date
Latest commit History 8 Commits
.github		.github
.gitignore		.gitignore
IMG_6689.png		IMG_6689.png
README.md		README.md
plugin.json		plugin.json
requirements.txt		requirements.txt
semantic_chunking.py		semantic_chunking.py
settings.py		settings.py