VictorMylle/Kohya-ss-sd-scripts
1
0
Fork 1
mirror of https://github.com/kohya-ss/sd-scripts.git synced 2026-04-09 06:45:09 +00:00
Code Issues Packages Projects Releases Wiki Activity

Fix some LoRA not trained if gradient checkpointing

Browse Source
This commit is contained in:
Kohya S
2023-01-19 20:39:33 +09:00
parent f2f2ce0d7d
commit e6a8c9d269
1 changed files with 3 additions and 0 deletions
Download Patch File Download Diff File Expand all files Collapse all files

3
train_network.py
View File

@@ -166,6 +166,9 @@ def train(args):
if args.gradient_checkpointing: # according to TI example in Diffusers, train is required if args.gradient_checkpointing: # according to TI example in Diffusers, train is required
unet.train() unet.train()
text_encoder.train() text_encoder.train()
# set top parameter requires_grad = True for gradient checkpointing works
text_encoder.text_model.embeddings.requires_grad_(True)
else: else:
unet.eval() unet.eval()
text_encoder.eval() text_encoder.eval()