examples/text_to_image/train_text_to_image.py · 8de78001df95a641bf6ef942bee9553921d44490 · chenpangpang / diffusers

"git@developer.sourcefind.cn:zhaoyu6/sglang.git" did not exist on "32ec68faf9f11723753e2fb8b89e215f49413faa"

Correct SNR weighted loss in v-prediction case by only adding 1 to SNR on the denominator (#6307) · c1010662

Yingtian Liu authored Jan 27, 2024



* fix minsnr implementation for v-prediction case

* format code

* always compute snr when snr_gamma is specified

---------
Co-authored-by: Sayak Paul <spsayakpaul@gmail.com>

c1010662

train_text_to_image.py 43.6 KB

Replace train_text_to_image.py