theluyuan/DiffSynth-Studio
1
0
Fork 0
mirror of https://github.com/modelscope/DiffSynth-Studio.git synced 2026-03-24 18:28:10 +00:00
Code Issues Packages Projects Releases Wiki Activity

bug fix

Browse Source
This commit is contained in:
Artiprocher
2024-09-19 12:26:46 +08:00
parent a7197f846b
commit 5f12fd4346
2 changed files with 6 additions and 2 deletions
Download Patch File Download Diff File Expand all files Collapse all files

4
diffsynth/models/lora.py
View File

@@ -77,6 +77,8 @@ class LoRAFromCivitai:
state_dict_lora = model.__class__.state_dict_converter().from_diffusers(state_dict_lora) state_dict_lora = model.__class__.state_dict_converter().from_diffusers(state_dict_lora)
elif model_resource == "civitai": elif model_resource == "civitai":
state_dict_lora = model.__class__.state_dict_converter().from_civitai(state_dict_lora) state_dict_lora = model.__class__.state_dict_converter().from_civitai(state_dict_lora)
if isinstance(state_dict_lora, tuple):
state_dict_lora = state_dict_lora[0]
if len(state_dict_lora) > 0: if len(state_dict_lora) > 0:
print(f" {len(state_dict_lora)} tensors are updated.") print(f" {len(state_dict_lora)} tensors are updated.")
for name in state_dict_lora: for name in state_dict_lora:
@@ -96,6 +98,8 @@ class LoRAFromCivitai:
converter_fn = model.__class__.state_dict_converter().from_diffusers if model_resource == "diffusers" \ converter_fn = model.__class__.state_dict_converter().from_diffusers if model_resource == "diffusers" \
else model.__class__.state_dict_converter().from_civitai else model.__class__.state_dict_converter().from_civitai
state_dict_lora_ = converter_fn(state_dict_lora_) state_dict_lora_ = converter_fn(state_dict_lora_)
if isinstance(state_dict_lora_, tuple):
state_dict_lora_ = state_dict_lora_[0]
if len(state_dict_lora_) == 0: if len(state_dict_lora_) == 0:
continue continue
for name in state_dict_lora_: for name in state_dict_lora_:

4
examples/train/README.md
View File

@@ -156,7 +156,7 @@ After training, use `model_manager.load_lora` to load the LoRA for inference.
from diffsynth import ModelManager, FluxImagePipeline from diffsynth import ModelManager, FluxImagePipeline
import torch import torch
model_manager = ModelManager(torch_dtype=torch.float16, device="cuda", model_manager = ModelManager(torch_dtype=torch.bfloat16, device="cuda",
file_path_list=[ file_path_list=[
"models/FLUX/FLUX.1-dev/text_encoder/model.safetensors", "models/FLUX/FLUX.1-dev/text_encoder/model.safetensors",
"models/FLUX/FLUX.1-dev/text_encoder_2", "models/FLUX/FLUX.1-dev/text_encoder_2",
@@ -164,7 +164,7 @@ model_manager = ModelManager(torch_dtype=torch.float16, device="cuda",
"models/FLUX/FLUX.1-dev/flux1-dev.safetensors" "models/FLUX/FLUX.1-dev/flux1-dev.safetensors"
]) ])
model_manager.load_lora("models/lightning_logs/version_0/checkpoints/epoch=0-step=500.ckpt", lora_alpha=1.0) model_manager.load_lora("models/lightning_logs/version_0/checkpoints/epoch=0-step=500.ckpt", lora_alpha=1.0)
pipe = SDXLImagePipeline.from_model_manager(model_manager) pipe = FluxImagePipeline.from_model_manager(model_manager)
torch.manual_seed(0) torch.manual_seed(0)
image = pipe( image = pipe(