|
Fix Version: |
0.7 |
Documentation in Release Notes: |
N/A |
Documentation: |
N/A |
Description: |
In particular when, e.g., the actual embedding instantly fails or chunks don't need to be re-embedded , it is noticeable that indexing is quite slow. I found that this is due to two reasons: # The regular expression for finding the last heading at the end of a chunk is surprisingly slow. # The Solr index is committed very frequently (after every document).
Both can easily be improved. |
Assignee: |
Michael Hamann |
Resolution: |
Fixed |
Status: |
Open Closed |
|