Richard Wong
|
ac340f6fd2
|
triplet loss with classification as a regularizer
- best record of 82.57 for esAppMod
|
2025-01-18 23:53:08 +09:00 |
Richard Wong
|
94ee7beba7
|
added cosines code
|
2025-01-18 12:14:06 +09:00 |
Richard Wong
|
b6cf2d4416
|
Implemented dynamic data re-sampling at each epoch
|
2025-01-16 19:41:03 +09:00 |
Richard Wong
|
5312cfa06f
|
added more augmentations to finally beat sota
- class_bert_augmentation is now the reference training code
|
2025-01-15 20:09:15 +09:00 |
Richard Wong
|
e90bc69ea9
|
Added datasets
- included code for augmentation
|
2025-01-14 17:34:17 +09:00 |
Richard Wong
|
a1d000d9c8
|
First commit
- added classification-based mapping for esAppMod data
|
2025-01-13 19:05:13 +09:00 |