Skip to content

Add option to train TE MLP layers only[再] #13

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
maruo555 merged 1 commit into from
May 23, 2025
Merged

Add option to train TE MLP layers only[再] #13

Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
9 changes: 8 additions & 1 deletion networks/lora.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -13,6 +13,7 @@
import re
from library.utils import setup_logging
from library.sdxl_original_unet import SdxlUNet2DConditionModel
import library.maruo_global_config as maruoCfg

setup_logging()
import logging
Expand Down Expand Up @@ -867,6 +868,7 @@ class LoRANetwork(torch.nn.Module):
UNET_TARGET_REPLACE_MODULE = ["Transformer2DModel"]
UNET_TARGET_REPLACE_MODULE_CONV2D_3X3 = ["ResnetBlock2D", "Downsample2D", "Upsample2D"]
TEXT_ENCODER_TARGET_REPLACE_MODULE = ["CLIPAttention", "CLIPSdpaAttention", "CLIPMLP"]
TEXT_ENCODER_TARGET_REPLACE_MODULE_MLPONLY = ["CLIPAttention", "CLIPMLP"] # 昔のバージョンの状態と同じにしたい場合用(実験用)
LORA_PREFIX_UNET = "lora_unet"
LORA_PREFIX_TEXT_ENCODER = "lora_te"

Expand Down Expand Up @@ -1029,7 +1031,12 @@ def create_modules(
index = None
logger.info(f"create LoRA for Text Encoder:")

text_encoder_loras, skipped = create_modules(False, index, text_encoder, LoRANetwork.TEXT_ENCODER_TARGET_REPLACE_MODULE)
if maruoCfg.te_mlp_fc_only:
# 改造ルート
text_encoder_loras, skipped = create_modules(False, index, text_encoder, LoRANetwork.TEXT_ENCODER_TARGET_REPLACE_MODULE_MLPONLY)
else:
# 通常ルート
text_encoder_loras, skipped = create_modules(False, index, text_encoder, LoRANetwork.TEXT_ENCODER_TARGET_REPLACE_MODULE)
self.text_encoder_loras.extend(text_encoder_loras)
skipped_te += skipped
logger.info(f"create LoRA for Text Encoder: {len(self.text_encoder_loras)} modules.")
Expand Down