-
I have been struggling with method RA.add_documents() with Azure OpenAI since it returned:
Error: 400 ‘$.input’ is invalid.
I found the problem was due to split_text() function in utils.py since i…
-
In conjunction with issue #6 (add marker/bookmark feature) it should be rather simple to also make it able to split files at marker positions, and delete portions between markers. A feature that could…
-
### Please describe the end goal of this project
In RFC https://github.com/opensearch-project/OpenSearch/issues/12918, we proposed to build shard level splitting without the need to stop write traf…
-
-
-
[Wiki 7.2b.3](https://github.com/keeleinstituut/tv-tolkevarav/wiki/A-11.-T%C3%B5lkekorraldaja-vaade-(Tellimuste-loomine,-korraldamine-ja-t%C3%B5lkimine-)#72-flow-generate-xliff-matecat-project-for-tra…
-
### Feature request
[Neuron 2.18](https://awsdocs-neuron.readthedocs-hosted.com/en/latest/release-notes/index.html#id2) introduced support for loading weights in SafeTensors format. Due to the nature…
-
### What is the bug or the crash?
Splitting polygons with lines doesn't work when the two different lines crosses in the middle of a polygon. It also depends on the chronological order. When you draw…
MTrim updated
3 months ago
-
-
Tesseract was trained on documents. Passing entire 1920x1080 screenshot will yield dubious performance.
Screenshots should be split into 3:4 aspect ratio images, ran in batches through tesseract, t…