Update README.md
Browse files
README.md
CHANGED
|
@@ -12,6 +12,7 @@ pipeline_tag: token-classification
|
|
| 12 |
|
| 13 |
bert-chunker-2 is a text chunker based on BERT with a classifier head to predict the start token of chunks (for use in RAG, etc), and using a sliding window it cuts documents of any size into chunks. It is a new experimental version of [bert-chunker](https://huggingface.co/tim1900/bert-chunker), enhanced for article structures, aiming to reach a balance between semantic chunking and structure chunking. It is a 0.1:0.9 linear weight merging of a trained semantic chunker and a trained structure chunker.
|
| 14 |
|
|
|
|
| 15 |
|
| 16 |
## Quickstart
|
| 17 |
Run the following:
|
|
|
|
| 12 |
|
| 13 |
bert-chunker-2 is a text chunker based on BERT with a classifier head to predict the start token of chunks (for use in RAG, etc), and using a sliding window it cuts documents of any size into chunks. It is a new experimental version of [bert-chunker](https://huggingface.co/tim1900/bert-chunker), enhanced for article structures, aiming to reach a balance between semantic chunking and structure chunking. It is a 0.1:0.9 linear weight merging of a trained semantic chunker and a trained structure chunker.
|
| 14 |
|
| 15 |
+
A new, better version is [bert-chunker-3](https://huggingface.co/tim1900/bert-chunker-3).
|
| 16 |
|
| 17 |
## Quickstart
|
| 18 |
Run the following:
|