theluyuan/DiffSynth-Studio
1
0
Fork 0
mirror of https://github.com/modelscope/DiffSynth-Studio.git synced 2026-03-24 10:18:12 +00:00
Code Issues Packages Projects Releases Wiki Activity

support training for eligen and nexusgen

Browse Source
This commit is contained in:
mi804
2025-07-29 13:28:42 +08:00
parent 2861ec4d9f
commit 8ef91b3672
14 changed files with 218 additions and 17 deletions
Download Patch File Download Diff File Expand all files Collapse all files

33
examples/flux/model_training/validate_lora/FLUX.1-dev-EliGen.py Normal file
View File

@@ -0,0 +1,33 @@
import torch
from PIL import Image
from diffsynth.pipelines.flux_image_new import FluxImagePipeline, ModelConfig
pipe = FluxImagePipeline.from_pretrained(
torch_dtype=torch.bfloat16,
device="cuda",
model_configs=[
ModelConfig(model_id="black-forest-labs/FLUX.1-dev", origin_file_pattern="flux1-dev.safetensors"),
ModelConfig(model_id="black-forest-labs/FLUX.1-dev", origin_file_pattern="text_encoder/model.safetensors"),
ModelConfig(model_id="black-forest-labs/FLUX.1-dev", origin_file_pattern="text_encoder_2/"),
ModelConfig(model_id="black-forest-labs/FLUX.1-dev", origin_file_pattern="ae.safetensors"),
],
)
pipe.load_lora(pipe.dit, "models/train/FLUX.1-dev-EliGen_lora/epoch-4.safetensors", alpha=1)
entity_prompts = ["A beautiful girl", "sign 'Entity Control'", "shorts", "shirt"]
global_prompt = "A beautiful girl wearing shirt and shorts in the street, holding a sign 'Entity Control'"
masks = [Image.open(f"data/example_image_dataset/eligen/{i}.png").convert('RGB') for i in range(len(entity_prompts))]
# generate image
image = pipe(
prompt=global_prompt,
cfg_scale=1.0,
num_inference_steps=50,
embedded_guidance=3.5,
seed=42,
height=1024,
width=1024,
eligen_entity_prompts=entity_prompts,
eligen_entity_masks=masks,
)
image.save(f"EliGen_lora.png")

26
examples/flux/model_training/validate_lora/Nexus-Gen-Editing.py Normal file
View File

@@ -0,0 +1,26 @@
import torch
from PIL import Image
from diffsynth.pipelines.flux_image_new import FluxImagePipeline, ModelConfig
pipe = FluxImagePipeline.from_pretrained(
torch_dtype=torch.bfloat16,
device="cuda",
model_configs=[
ModelConfig(model_id="DiffSynth-Studio/Nexus-GenV2", origin_file_pattern="model*.safetensors"),
ModelConfig(model_id="DiffSynth-Studio/Nexus-GenV2", origin_file_pattern="edit_decoder.bin"),
ModelConfig(model_id="black-forest-labs/FLUX.1-dev", origin_file_pattern="text_encoder/model.safetensors"),
ModelConfig(model_id="black-forest-labs/FLUX.1-dev", origin_file_pattern="text_encoder_2/"),
ModelConfig(model_id="black-forest-labs/FLUX.1-dev", origin_file_pattern="ae.safetensors"),
],
)
pipe.load_lora(pipe.dit, "models/train/FLUX.1-NexusGen-Edit_lora/epoch-4.safetensors", alpha=1)
ref_image = Image.open("data/example_image_dataset/nexus_gen/image_1.png").convert("RGB")
prompt = "Add a pair of sunglasses."
image = pipe(
prompt=prompt, negative_prompt="",
seed=42, cfg_scale=1.0, num_inference_steps=50,
nexus_gen_reference_image=ref_image,
height=512, width=512,
)
image.save("NexusGen-Edit_lora.jpg")