Safetensors
File size: 8,548 Bytes
ce51676
 
 
 
 
 
 
6443dcf
 
 
 
 
 
 
 
ce51676
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
f0c08bd
ce51676
 
 
f0c08bd
ce51676
 
 
 
 
 
 
 
 
f0c08bd
ce51676
 
 
f0c08bd
ce51676
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
---
license: apache-2.0
---
# Templates - Image Editing (FLUX.2-klein-base-4B)

This model is part of the open-source Diffusion Templates series by [DiffSynth-Studio](https://github.com/modelscope/DiffSynth-Studio). It is an image editing model capable of receiving an input image and precisely modifying specific objects, actions, or attributes within the image based on natural language instructions.

* Open-source code: [DiffSynth-Studio](https://github.com/modelscope/DiffSynth-Studio)
* Technical report: [arXiv](https://arxiv.org/abs/2604.24351)
* Project page: [GitHub](https://modelscope.github.io/diffusion-templates-web/)
* Documentation: [English Version](https://diffsynth-studio-doc.readthedocs.io/en/latest/Diffusion_Templates/Introducing_Diffusion_Templates.html)、[中文版](https://diffsynth-studio-doc.readthedocs.io/zh-cn/latest/Diffusion_Templates/Introducing_Diffusion_Templates.html)
* Online demo: [ModelScope](https://modelscope.cn/studios/DiffSynth-Studio/Diffusion-Templates)
* Models: [ModelScope](https://modelscope.cn/collections/DiffSynth-Studio/KleinBase4B-Templates)、[ModelScope International](https://modelscope.ai/collections/DiffSynth-Studio/KleinBase4B-Templates)、[HuggingFace](https://huggingface.co/collections/DiffSynth-Studio/kleinbase4b-templates)
* Datasets: [ModelScope](https://modelscope.cn/collections/DiffSynth-Studio/ImagePulseV2)、[ModelScope International](https://modelscope.cn/collections/DiffSynth-Studio/ImagePulseV2)、[HuggingFace](https://huggingface.co/collections/DiffSynth-Studio/imagepulsev2)

## Demo Results

|Reference|Prompt: Put a hat on this cat.|Prompt: Make the cat turn its head to look to the right.|
|-|-|-|
|![](./assets/cat.jpg)|![](./assets/cat_Edit_hat.jpg)|![](./assets/cat_Edit_head.jpg)|

|Reference|Prompt: Change the color of the car to matte black.|Prompt: Make it a rainy night.|
|-|-|-|
|![](./assets/car.jpg)|![](./assets/car_Edit_color.jpg)|![](./assets/car_Edit_rain.jpg)|

|Reference|Prompt: Make her hair long and blonde.|Prompt: Change the book in her hands into coffee.|
|-|-|-|
|![](./assets/girl.jpg)|![](./assets/girl_Edit_hair.jpg)|![](./assets/girl_Edit_coffee.jpg)|

## Inference Code

* Install [DiffSynth-Studio](https://github.com/modelscope/DiffSynth-Studio)

```
git clone https://github.com/modelscope/DiffSynth-Studio.git
cd DiffSynth-Studio
pip install -e .
```

* Direct inference (requires 40GB GPU memory)

```python
from diffsynth.diffusion.template import TemplatePipeline
from diffsynth.pipelines.flux2_image import Flux2ImagePipeline, ModelConfig
import torch
from modelscope import dataset_snapshot_download
from PIL import Image
```

```python
pipe = Flux2ImagePipeline.from_pretrained(
    torch_dtype=torch.bfloat16,
    device="cuda",
    model_configs=[
        ModelConfig(model_id="black-forest-labs/FLUX.2-klein-base-4B", origin_file_pattern="transformer/*.safetensors"),
        ModelConfig(model_id="black-forest-labs/FLUX.2-klein-4B", origin_file_pattern="text_encoder/*.safetensors"),
        ModelConfig(model_id="black-forest-labs/FLUX.2-klein-4B", origin_file_pattern="vae/diffusion_pytorch_model.safetensors"),
    ],
    tokenizer_config=ModelConfig(model_id="black-forest-labs/FLUX.2-klein-4B", origin_file_pattern="tokenizer/"),
)
template = TemplatePipeline.from_pretrained(
    torch_dtype=torch.bfloat16,
    device="cuda",
    model_configs=[ModelConfig(model_id="DiffSynth-Studio/Template-KleinBase4B-Edit")],
)
dataset_snapshot_download(
    "DiffSynth-Studio/examples_in_diffsynth",
    allow_file_pattern=["templates/*"],
    local_dir="data/examples",
)
image = template(
    pipe,
    prompt="Put a hat on this cat.",
    seed=0, cfg_scale=4, num_inference_steps=50,
    template_inputs = [{
        "image": Image.open("data/examples/templates/image_reference.jpg"),
        "prompt": "Put a hat on this cat.",
    }],
    negative_template_inputs = [{
        "image": Image.open("data/examples/templates/image_reference.jpg"),
        "prompt": "",
    }],
)
image.save("image_Edit_hat.jpg")
image = template(
    pipe,
    prompt="Make the cat turn its head to look to the right.",
    seed=0, cfg_scale=4, num_inference_steps=50,
    template_inputs = [{
        "image": Image.open("data/examples/templates/image_reference.jpg"),
        "prompt": "Make the cat turn its head to look to the right.",
    }],
    negative_template_inputs = [{
        "image": Image.open("data/examples/templates/image_reference.jpg"),
        "prompt": "",
    }],
)
image.save("image_Edit_head.jpg")
```

* Enable lazy loading and memory management, requires 24G GPU memory

```python
from diffsynth.diffusion.template import TemplatePipeline
from diffsynth.pipelines.flux2_image import Flux2ImagePipeline, ModelConfig
import torch
from modelscope import dataset_snapshot_download
from PIL import Image
```

```python
vram_config = {
    "offload_dtype": "disk",
    "offload_device": "disk",
    "onload_dtype": torch.float8_e4m3fn,
    "onload_device": "cpu",
    "preparing_dtype": torch.float8_e4m3fn,
    "preparing_device": "cuda",
    "computation_dtype": torch.bfloat16,
    "computation_device": "cuda",
}
pipe = Flux2ImagePipeline.from_pretrained(
    torch_dtype=torch.bfloat16,
    device="cuda",
    model_configs=[
        ModelConfig(model_id="black-forest-labs/FLUX.2-klein-base-4B", origin_file_pattern="transformer/*.safetensors", **vram_config),
        ModelConfig(model_id="black-forest-labs/FLUX.2-klein-4B", origin_file_pattern="text_encoder/*.safetensors", **vram_config),
        ModelConfig(model_id="black-forest-labs/FLUX.2-klein-4B", origin_file_pattern="vae/diffusion_pytorch_model.safetensors"),
    ],
    tokenizer_config=ModelConfig(model_id="black-forest-labs/FLUX.2-klein-4B", origin_file_pattern="tokenizer/"),
    vram_limit=torch.cuda.mem_get_info("cuda")[1] / (1024 ** 3) - 0.5,
)
template = TemplatePipeline.from_pretrained(
    torch_dtype=torch.bfloat16,
    device="cuda",
    model_configs=[ModelConfig(model_id="DiffSynth-Studio/Template-KleinBase4B-Edit")],
    lazy_loading=True,
)
dataset_snapshot_download(
    "DiffSynth-Studio/examples_in_diffsynth",
    allow_file_pattern=["templates/*"],
    local_dir="data/examples",
)
image = template(
    pipe,
    prompt="Put a hat on this cat.",
    seed=0, cfg_scale=4, num_inference_steps=50,
    template_inputs = [{
        "image": Image.open("data/examples/templates/image_reference.jpg"),
        "prompt": "Put a hat on this cat.",
    }],
    negative_template_inputs = [{
        "image": Image.open("data/examples/templates/image_reference.jpg"),
        "prompt": "",
    }],
)
image.save("image_Edit_hat.jpg")
image = template(
    pipe,
    prompt="Make the cat turn its head to look to the right.",
    seed=0, cfg_scale=4, num_inference_steps=50,
    template_inputs = [{
        "image": Image.open("data/examples/templates/image_reference.jpg"),
        "prompt": "Make the cat turn its head to look to the right.",
    }],
    negative_template_inputs = [{
        "image": Image.open("data/examples/templates/image_reference.jpg"),
        "prompt": "",
    }],
)
image.save("image_Edit_head.jpg")
```

## Training Code

After installing DiffSynth-Studio, use the following script to start training. For more information, please refer to the [DiffSynth-Studio Documentation](https://diffsynth-studio-doc.readthedocs.io/zh-cn/latest/).

```shell
modelscope download --dataset DiffSynth-Studio/diffsynth_example_dataset --include "flux2/Template-KleinBase4B-Edit/*" --local_dir ./data/diffsynth_example_dataset

accelerate launch examples/flux2/model_training/train.py \
  --dataset_base_path data/diffsynth_example_dataset/flux2/Template-KleinBase4B-Edit \
  --dataset_metadata_path data/diffsynth_example_dataset/flux2/Template-KleinBase4B-Edit/metadata.jsonl \
  --extra_inputs "template_inputs" \
  --max_pixels 1048576 \
  --dataset_repeat 50 \
  --model_id_with_origin_paths "black-forest-labs/FLUX.2-klein-4B:text_encoder/*.safetensors,black-forest-labs/FLUX.2-klein-base-4B:transformer/*.safetensors,black-forest-labs/FLUX.2-klein-4B:vae/diffusion_pytorch_model.safetensors" \
  --template_model_id_or_path "DiffSynth-Studio/Template-KleinBase4B-Edit:" \
  --tokenizer_path "black-forest-labs/FLUX.2-klein-4B:tokenizer/" \
  --learning_rate 1e-4 \
  --num_epochs 2 \
  --remove_prefix_in_ckpt "pipe.template_model." \
  --output_path "./models/train/Template-KleinBase4B-Edit_full" \
  --trainable_models "template_model" \
  --use_gradient_checkpointing \
  --find_unused_parameters
```