Text encoder not saved in pipeline folder #1309
Answered
by
bghira
MohamedAziz-Khezami
asked this question in
Q&A
-
|
Hello, I am trying to train a full sdxl model, and once it finishes it gives me checkpoints folder that have unet safetensors and a pipeline folder that dosen't have the text encoder in it, so i cant convert it with the python script included. How to solve this problem please?? |
Beta Was this translation helpful? Give feedback.
Answered by
bghira
Feb 3, 2025
Replies: 1 comment
-
|
see the documentation where it mentions |
Beta Was this translation helpful? Give feedback.
0 replies
Answer selected by
MohamedAziz-Khezami
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
see the documentation where it mentions
--save_text_encoder, this is done to save disk space as you typically already started with a copy..