Chunkit allows you to scrape and convert webpages into markdown chunks, for RAG applications.
- Install
pip install chunkit- Start chunking
from chunkit import Chunker
# Initialize the Chunker
chunker = Chunker()
# Define URLs to process
urls = ["https://en.wikipedia.org/wiki/Chunking_(psychology)"]
# Process the URLs into markdown chunks
chunkified_urls = chunker.process(urls)
# Output the resulting chunks
for url in chunkified_urls:
if url['success']:
for chunk in url['chunks']:
print(chunk)Example results for above Wikipedia page
### Chunking (psychology)
In cognitive psychology, **chunking** is a process by which small individual pieces of a set of information are bound together to create a meaningful whole later on in memory. The chunks, by which the information is grouped, are meant to improve short-term retention of the material, thus bypassing the limited capacity of working memory...### Modality effect
A modality effect is present in chunking. That is, the mechanism used to convey the list of items to the individual affects how much "chunking" occurs. Experimentally, it has been found that auditory presentation results in a larger amount of grouping in the responses of individuals than visual presentation does...### Memory training systems, mnemonic
Various kinds of memory training systems and mnemonics include training and drills in specially-designed recoding or chunking schemes. Such systems existed before Miller's paper, but there was no convenient term to describe the general strategy and no substantive and reliable research...Etc.
Most chunkers:
- Perform a naive chunking based on the number of words in the content.
- For example, they may split content every 200 words, and have a 30 word overlap between each.
- This leads to messy chunks that are noisy and have unnecessary extra data.
- Additionally, the chunked sentences are usually split in the middle, with lost meaning.
- This leads to poor LLM performance, with incorrect answers and hallucinations.
Chunkit however, converts HTML to Markdown, and then determines split points based on the most common header levels.
This gives you better results because:
- Online content tends to be logically split in paragraphs delimited by headers.
- By chunking based on headers, this method preserves semantic meaning better.
- You get a much cleaner, semantically cohesive paragraph of data.
This free open source package primarily chunks webpages and html.
This project is licensed under GPL v3 - see the LICENSE file for details.
For questions or support, please open an issue.