Carlos-Mesquita
|
4e05c4d913
|
ENCOA-255 gpt was grouping parts by sections and the reading passages were not updated with text.content instead of the old context field
|
2024-12-03 11:57:42 +00:00 |
|
Carlos-Mesquita
|
1603fa4ee6
|
ENCOA-254
|
2024-12-02 17:20:13 +00:00 |
|
Carlos-Mesquita
|
6681b2d0e9
|
Now grading is partitioned into smaller chunks so that whisper doesnt struggle
|
2024-11-27 08:07:54 +00:00 |
|
Carlos-Mesquita
|
a7da187ec6
|
Writing and speaking rework, some changes to module upload
|
2024-11-25 16:41:38 +00:00 |
|
Carlos-Mesquita
|
6daab0d9a7
|
Remove unnecessary section id's from reading and listening to retrieve questions since context is already on the post dto
|
2024-11-12 14:19:52 +00:00 |
|
Carlos-Mesquita
|
afeaf118c6
|
Fixed more or less reading import, attempted to do listening
|
2024-11-10 06:46:58 +00:00 |
|
Carlos-Mesquita
|
6909d75eb6
|
Fixed level issues
|
2024-11-10 04:21:36 +00:00 |
|
Carlos-Mesquita
|
cf1b676312
|
Avatar's can't be random on the video endpoint since these will be called in batch
|
2024-11-09 10:33:05 +00:00 |
|
Carlos-Mesquita
|
e955a16abf
|
Thought I had staged it
|
2024-11-09 09:30:54 +00:00 |
|
Carlos-Mesquita
|
136309120b
|
Mp3 uploading is now done on next, now doing concurrent reading and listening exercise calls with ayncio's gather to openai, should be faster
|
2024-11-07 11:09:56 +00:00 |
|
Carlos-Mesquita
|
a2e96f8e54
|
Batch import wasn't updated
|
2024-11-06 11:01:39 +00:00 |
|
Carlos-Mesquita
|
84ed2f2f6a
|
Changes to endpoints so they allow to only get context and then the exercises as well as tidying up a bit
|
2024-11-04 23:31:48 +00:00 |
|
Carlos-Mesquita
|
2a032c5aba
|
Fastapi refactor update
|
2024-10-01 19:31:01 +01:00 |
|
Carlos Mesquita
|
f92a803d96
|
Updated this to the latest version of develop, got rid of most of the duplication, might be missing some packages in toml, needs testing
|
2024-08-30 02:35:11 +01:00 |
|
Carlos Mesquita
|
3cf9fa5cba
|
Async release
|
2024-07-23 08:40:35 +01:00 |
|