adipanda commited on
Commit
81076ab
1 Parent(s): 7de2983

Model card auto-generated by SimpleTuner

Browse files
Files changed (1) hide show
  1. README.md +175 -0
README.md ADDED
@@ -0,0 +1,175 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ license: other
3
+ base_model: "black-forest-labs/FLUX.1-dev"
4
+ tags:
5
+ - flux
6
+ - flux-diffusers
7
+ - text-to-image
8
+ - diffusers
9
+ - simpletuner
10
+ - safe-for-work
11
+ - lora
12
+ - template:sd-lora
13
+ - lycoris
14
+ inference: true
15
+ widget:
16
+ - text: 'unconditional (blank prompt)'
17
+ parameters:
18
+ negative_prompt: 'blurry, cropped, ugly'
19
+ output:
20
+ url: ./assets/image_0_0.png
21
+ - text: 'A scene from One Piece. Monkey D. Luffy holding a sign that says ''I LOVE PROMPTS!'', he is standing full body on a beach at sunset. He is wearing a red vest, yellow sash, and a straw hat. The setting sun casts a dynamic shadow on his face.'
22
+ parameters:
23
+ negative_prompt: 'blurry, cropped, ugly'
24
+ output:
25
+ url: ./assets/image_1_0.png
26
+ - text: 'A scene from One Piece. Monkey D. Luffy jumping out of a propeller airplane, sky diving. He looks excited and his hair is blowing in the wind. The sky is clear and blue, there are birds pictured in the distance.'
27
+ parameters:
28
+ negative_prompt: 'blurry, cropped, ugly'
29
+ output:
30
+ url: ./assets/image_2_0.png
31
+ - text: 'A scene from One Piece. Monkey D. Luffy spinning a basketball on his finger on a basketball court. He is wearing a lakers jersey with the #12 on it. The basketball hoop and crowd are in the background cheering him. He is smiling.'
32
+ parameters:
33
+ negative_prompt: 'blurry, cropped, ugly'
34
+ output:
35
+ url: ./assets/image_3_0.png
36
+ - text: 'A scene from One Piece. Monkey D. Luffy is wearing a suit in an office shaking the hand of a business woman. The woman has purple hair and is wearing professional attire. There is a Google logo in the background. It is during daytime, and the overall sentiment is one of accomplishment.'
37
+ parameters:
38
+ negative_prompt: 'blurry, cropped, ugly'
39
+ output:
40
+ url: ./assets/image_4_0.png
41
+ - text: 'A scene from One Piece. Monkey D. Luffy is fighting a large brown grizzly bear, deep in a forest. The bear is tall and standing on two legs, roaring. The bear is also wearing a crown because it is the king of all bears. Around them are tall trees and other animals watching.'
42
+ parameters:
43
+ negative_prompt: 'blurry, cropped, ugly'
44
+ output:
45
+ url: ./assets/image_5_0.png
46
+ ---
47
+
48
+ # luffy-simpletuner-lora-11
49
+
50
+ This is a LyCORIS adapter derived from [black-forest-labs/FLUX.1-dev](https://huggingface.co/black-forest-labs/FLUX.1-dev).
51
+
52
+
53
+ No validation prompt was used during training.
54
+
55
+
56
+
57
+
58
+ None
59
+
60
+
61
+ ## Validation settings
62
+ - CFG: `3.5`
63
+ - CFG Rescale: `0.0`
64
+ - Steps: `20`
65
+ - Sampler: `None`
66
+ - Seed: `42`
67
+ - Resolution: `1024x1024`
68
+
69
+ Note: The validation settings are not necessarily the same as the [training settings](#training-settings).
70
+
71
+ You can find some example images in the following gallery:
72
+
73
+
74
+ <Gallery />
75
+
76
+ The text encoder **was not** trained.
77
+ You may reuse the base model text encoder for inference.
78
+
79
+
80
+ ## Training settings
81
+
82
+ - Training epochs: 0
83
+ - Training steps: 300
84
+ - Learning rate: 5e-05
85
+ - Effective batch size: 8
86
+ - Micro-batch size: 8
87
+ - Gradient accumulation steps: 1
88
+ - Number of GPUs: 1
89
+ - Prediction type: flow-matching
90
+ - Rescaled betas zero SNR: False
91
+ - Optimizer: adamw_bf16
92
+ - Precision: Pure BF16
93
+ - Quantised: Yes: int8-quanto
94
+ - Xformers: Not used
95
+ - LyCORIS Config:
96
+ ```json
97
+ {
98
+ "algo": "lokr",
99
+ "multiplier": 1.0,
100
+ "linear_dim": 10000,
101
+ "linear_alpha": 1,
102
+ "factor": 12,
103
+ "apply_preset": {
104
+ "target_module": [
105
+ "Attention",
106
+ "FeedForward"
107
+ ],
108
+ "module_algo_map": {
109
+ "Attention": {
110
+ "factor": 12
111
+ },
112
+ "FeedForward": {
113
+ "factor": 6
114
+ }
115
+ }
116
+ }
117
+ }
118
+ ```
119
+
120
+ ## Datasets
121
+
122
+ ### luffy-1024
123
+ - Repeats: 1
124
+ - Total number of images: 306
125
+ - Total number of aspect buckets: 1
126
+ - Resolution: 1.048576 megapixels
127
+ - Cropped: False
128
+ - Crop style: None
129
+ - Crop aspect: None
130
+ ### luffy-768
131
+ - Repeats: 2
132
+ - Total number of images: 306
133
+ - Total number of aspect buckets: 1
134
+ - Resolution: 0.589824 megapixels
135
+ - Cropped: False
136
+ - Crop style: None
137
+ - Crop aspect: None
138
+ ### luffy-512
139
+ - Repeats: 2
140
+ - Total number of images: 306
141
+ - Total number of aspect buckets: 1
142
+ - Resolution: 0.262144 megapixels
143
+ - Cropped: False
144
+ - Crop style: None
145
+ - Crop aspect: None
146
+
147
+
148
+ ## Inference
149
+
150
+
151
+ ```python
152
+ import torch
153
+ from diffusers import DiffusionPipeline
154
+ from lycoris import create_lycoris_from_weights
155
+
156
+ model_id = 'black-forest-labs/FLUX.1-dev'
157
+ adapter_id = 'pytorch_lora_weights.safetensors' # you will have to download this manually
158
+ lora_scale = 1.0
159
+ wrapper, _ = create_lycoris_from_weights(lora_scale, adapter_id, pipeline.transformer)
160
+ wrapper.merge_to()
161
+
162
+ prompt = "An astronaut is riding a horse through the jungles of Thailand."
163
+
164
+ pipeline.to('cuda' if torch.cuda.is_available() else 'mps' if torch.backends.mps.is_available() else 'cpu')
165
+ image = pipeline(
166
+ prompt=prompt,
167
+ num_inference_steps=20,
168
+ generator=torch.Generator(device='cuda' if torch.cuda.is_available() else 'mps' if torch.backends.mps.is_available() else 'cpu').manual_seed(1641421826),
169
+ width=1024,
170
+ height=1024,
171
+ guidance_scale=3.5,
172
+ ).images[0]
173
+ image.save("output.png", format="PNG")
174
+ ```
175
+