Hugging Face's logo

bghira
/
terminus-xl-velocity-v2

Text-to-Image
Diffusers
Safetensors
stable-diffusion
stable-diffusion-diffusers
full
Model card Files Community
PseudoTerminal X commited on
Commit
a8efc6b
·
verified ·
1 Parent(s): 5bbd765

Trained for 0 epochs and 550 steps.

Browse files

Trained with datasets ['text-embeds', 'photocb-clip-embeds', 'mj-60', 'laion-hd-exif-1024', 'laion-aesthetics-1024', 'experimental', 'ethnic', 'sports', 'gay', 'architecture', 'shutterstock', 'cinemamix-1mp', 'nsfw-1024', 'anatomy', 'bg20k-1024', 'yoga', 'photo-aesthetics', 'text-1mp', 'photo-concept-bucket']
Learning rate 1e-06, batch size 4, and 1 gradient accumulation steps.
Used DDPM noise scheduler for training with v_prediction prediction type and rescaled_betas_zero_snr=True
Using 'trailing' timestep spacing.
Base model: ptx0/terminus-xl-velocity-v1
VAE: madebyollin/sdxl-vae-fp16-fix

Files changed (24) hide show
  1. README.md +192 -4
  2. optimizer.bin +3 -0
  3. random_states_0.pkl +3 -0
  4. scheduler.bin +3 -0
  5. training_state-anatomy.json +3 -0
  6. training_state-architecture.json +3 -0
  7. training_state-bg20k-1024.json +3 -0
  8. training_state-cinemamix-1mp.json +3 -0
  9. training_state-ethnic.json +3 -0
  10. training_state-experimental.json +3 -0
  11. training_state-gay.json +3 -0
  12. training_state-laion-aesthetics-1024.json +3 -0
  13. training_state-laion-hd-exif-1024.json +3 -0
  14. training_state-mj-60.json +3 -0
  15. training_state-nsfw-1024.json +3 -0
  16. training_state-photo-aesthetics.json +3 -0
  17. training_state-photo-concept-bucket.json +3 -0
  18. training_state-shutterstock.json +3 -0
  19. training_state-sports.json +3 -0
  20. training_state-text-1mp.json +3 -0
  21. training_state-yoga.json +3 -0
  22. training_state.json +3 -0
  23. unet/config.json +2 -2
  24. unet/diffusion_pytorch_model.safetensors +1 -1
README.md CHANGED
@@ -1,9 +1,197 @@
1
  ---
2
- license: openrail++
 
 
 
 
 
 
 
 
 
 
3
  ---
4
 
5
- # Terminus XL - Gamma (v2 preview)
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
6
 
7
- This is an in-progress checkpoint of [the "Gamma" model](/ptx0/terminus-xl-gamma-v1) from the Terminus XL series.
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
8
 
9
- It's updated randomly for evaluation as progress rolls on.
 
1
  ---
2
+ license: creativeml-openrail-m
3
+ base_model: "ptx0/terminus-xl-velocity-v1"
4
+ tags:
5
+ - stable-diffusion
6
+ - stable-diffusion-diffusers
7
+ - text-to-image
8
+ - diffusers
9
+ - full
10
+
11
+ inference: true
12
+
13
  ---
14
 
15
+ # terminus-xl-velocity-v2
16
+
17
+ This is a full rank finetuned model derived from [ptx0/terminus-xl-velocity-v1](https://huggingface.co/ptx0/terminus-xl-velocity-v1).
18
+
19
+ The main validation prompt used during training was:
20
+
21
+ ```
22
+ a cute anime character named toast
23
+ ```
24
+
25
+ ## Validation settings
26
+ - CFG: `7.5`
27
+ - CFG Rescale: `0.7`
28
+ - Steps: `30`
29
+ - Sampler: `euler`
30
+ - Seed: `420420420`
31
+ - Resolution: `1024`
32
+
33
+ Note: The validation settings are not necessarily the same as the [training settings](#training-settings).
34
+
35
+
36
+
37
+
38
+ <Gallery />
39
+
40
+ The text encoder **was not** trained.
41
+ You may reuse the base model text encoder for inference.
42
+
43
+
44
+ ## Training settings
45
+
46
+ - Training epochs: 0
47
+ - Training steps: 550
48
+ - Learning rate: 1e-06
49
+ - Effective batch size: 4
50
+ - Micro-batch size: 4
51
+ - Gradient accumulation steps: 1
52
+ - Prediction type: v_prediction
53
+ - Rescaled betas zero SNR: True
54
+ - Optimizer: AdamW, stochastic bf16
55
+ - Precision: Pure BF16
56
+ - Xformers: Enabled
57
+
58
+
59
+ ## Datasets
60
 
61
+ ### mj-60
62
+ - Repeats: 0
63
+ - Total number of images: 3
64
+ - Total number of aspect buckets: 1
65
+ - Resolution: 1.0 megapixels
66
+ - Cropped: False
67
+ - Crop style: None
68
+ - Crop aspect: None
69
+ ### laion-hd-exif-1024
70
+ - Repeats: 0
71
+ - Total number of images: 102
72
+ - Total number of aspect buckets: 27
73
+ - Resolution: 1.0 megapixels
74
+ - Cropped: False
75
+ - Crop style: None
76
+ - Crop aspect: None
77
+ ### laion-aesthetics-1024
78
+ - Repeats: 0
79
+ - Total number of images: 180
80
+ - Total number of aspect buckets: 47
81
+ - Resolution: 1.0 megapixels
82
+ - Cropped: False
83
+ - Crop style: None
84
+ - Crop aspect: None
85
+ ### experimental
86
+ - Repeats: 0
87
+ - Total number of images: 74
88
+ - Total number of aspect buckets: 19
89
+ - Resolution: 1.0 megapixels
90
+ - Cropped: False
91
+ - Crop style: None
92
+ - Crop aspect: None
93
+ ### ethnic
94
+ - Repeats: 0
95
+ - Total number of images: 90
96
+ - Total number of aspect buckets: 23
97
+ - Resolution: 1.0 megapixels
98
+ - Cropped: False
99
+ - Crop style: None
100
+ - Crop aspect: None
101
+ ### sports
102
+ - Repeats: 0
103
+ - Total number of images: 47
104
+ - Total number of aspect buckets: 12
105
+ - Resolution: 1.0 megapixels
106
+ - Cropped: False
107
+ - Crop style: None
108
+ - Crop aspect: None
109
+ ### gay
110
+ - Repeats: 0
111
+ - Total number of images: 62
112
+ - Total number of aspect buckets: 16
113
+ - Resolution: 1.0 megapixels
114
+ - Cropped: False
115
+ - Crop style: None
116
+ - Crop aspect: None
117
+ ### architecture
118
+ - Repeats: 0
119
+ - Total number of images: 85
120
+ - Total number of aspect buckets: 22
121
+ - Resolution: 1.0 megapixels
122
+ - Cropped: False
123
+ - Crop style: None
124
+ - Crop aspect: None
125
+ ### shutterstock
126
+ - Repeats: 0
127
+ - Total number of images: 130
128
+ - Total number of aspect buckets: 34
129
+ - Resolution: 1.0 megapixels
130
+ - Cropped: False
131
+ - Crop style: None
132
+ - Crop aspect: None
133
+ ### cinemamix-1mp
134
+ - Repeats: 0
135
+ - Total number of images: 23
136
+ - Total number of aspect buckets: 6
137
+ - Resolution: 1.0 megapixels
138
+ - Cropped: False
139
+ - Crop style: None
140
+ - Crop aspect: None
141
+ ### nsfw-1024
142
+ - Repeats: 0
143
+ - Total number of images: 16
144
+ - Total number of aspect buckets: 4
145
+ - Resolution: 1.0 megapixels
146
+ - Cropped: False
147
+ - Crop style: None
148
+ - Crop aspect: None
149
+ ### anatomy
150
+ - Repeats: 5
151
+ - Total number of images: 98
152
+ - Total number of aspect buckets: 25
153
+ - Resolution: 1.0 megapixels
154
+ - Cropped: False
155
+ - Crop style: None
156
+ - Crop aspect: None
157
+ ### bg20k-1024
158
+ - Repeats: 0
159
+ - Total number of images: 160
160
+ - Total number of aspect buckets: 42
161
+ - Resolution: 1.0 megapixels
162
+ - Cropped: False
163
+ - Crop style: None
164
+ - Crop aspect: None
165
+ ### yoga
166
+ - Repeats: 0
167
+ - Total number of images: 82
168
+ - Total number of aspect buckets: 21
169
+ - Resolution: 1.0 megapixels
170
+ - Cropped: False
171
+ - Crop style: None
172
+ - Crop aspect: None
173
+ ### photo-aesthetics
174
+ - Repeats: 0
175
+ - Total number of images: 116
176
+ - Total number of aspect buckets: 30
177
+ - Resolution: 1.0 megapixels
178
+ - Cropped: False
179
+ - Crop style: None
180
+ - Crop aspect: None
181
+ ### text-1mp
182
+ - Repeats: 5
183
+ - Total number of images: 90
184
+ - Total number of aspect buckets: 23
185
+ - Resolution: 1.0 megapixels
186
+ - Cropped: False
187
+ - Crop style: None
188
+ - Crop aspect: None
189
+ ### photo-concept-bucket
190
+ - Repeats: 1
191
+ - Total number of images: 149
192
+ - Total number of aspect buckets: 39
193
+ - Resolution: 1.0 megapixels
194
+ - Cropped: False
195
+ - Crop style: None
196
+ - Crop aspect: None
197
 
 
optimizer.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ba222172c8d318016ee666af1d042fc955a8c64fa086feffb691f2ddae364220
3
+ size 15406336826
random_states_0.pkl ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d3b1f08a580bf8af491faba51628f06e7a8230d1d858d094470fe1016250801c
3
+ size 14344
scheduler.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:6ba98ccf0ba4182e6dd389305e01f09ba519041c8bf9f46714dd649cb6b39bcc
3
+ size 1000
training_state-anatomy.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3ffe889f4a2492529ab38428aab76117936ad5d2ea85336d45e93b0217e3f8d4
3
+ size 2206091
training_state-architecture.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:80c986f0a438e07dbc1a39dd0f687efdc875f922050060a1e5a9946ddf5dff01
3
+ size 219538
training_state-bg20k-1024.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:0cc30b7063affd3b418586375d38b0805319499b212b1911e23408790038211e
3
+ size 14669495
training_state-cinemamix-1mp.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:81632abbda6d7ef0d894431c409f062b2876c65910abfa6272cdbca2ef3149b5
3
+ size 1101298
training_state-ethnic.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:55e00ca6882b41834489ab0dc9902cb81e0f54db151520583d10cde43966b16a
3
+ size 740979
training_state-experimental.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:0167c24a74df5b2513fa152c71d967884d14e8e9f05c9c58cbb9d495b248ef92
3
+ size 399437
training_state-gay.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:21158f24c46629aae110b3694abf1582ecfc853e3454e099d18dcd745d640bbd
3
+ size 229911
training_state-laion-aesthetics-1024.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:4be6944a65230c64b392602fdcd9babfed4004439308c667cfc4f874fbca3df6
3
+ size 14084398
training_state-laion-hd-exif-1024.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:332c34b2916b925bfa4736ac54aab50552348ec08cb22ee8f93a9c9fb734f84b
3
+ size 12948320
training_state-mj-60.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:73c573d69ee41c0834f15d410271c33dac7f302739fc8facbfe8b8e4fafbefab
3
+ size 2286911
training_state-nsfw-1024.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:289e1cc284bde3a17d0cb8579e6f382f90b0f62ffdc1f76b087d82aba714ac86
3
+ size 276780
training_state-photo-aesthetics.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:37482adcdf01da50ab2e5f8b24a41931414c1d3b8cc0c9fad535ff2932546ba9
3
+ size 4583448
training_state-photo-concept-bucket.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:6ce6cc91f1448966b29a699d3645c42f085d5724f1984de7617b4fd9fcc4f9e4
3
+ size 15134511
training_state-shutterstock.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:2e85d72b9eda2cb436e812f10d87d1589cdc2a76e79ebcd6afe333bcc9c870b5
3
+ size 4447262
training_state-sports.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ca1b0b7815e9510254777ec1854cdc3c773968d9079f3c5ec92d68353b45a055
3
+ size 153450
training_state-text-1mp.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:0dd8c8bb04ee5113d3d3df95b9fad26f6b4c32c61824e1ed2a691fd0ce81638f
3
+ size 1664195
training_state-yoga.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3f1423db5dc5ed17b4ce840613032705f7491c2e6f1afa9754f1d2ad5ab487ad
3
+ size 664338
training_state.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:1ad9d17f34fddcaad4393eeaead99aaa74c6553f03000df98be75a721fd18165
3
+ size 92
unet/config.json CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:dccec8ddabc55c6b694afdbd130d241fb8e18c42157bbc8d457c8cf8e5f116f1
3
- size 1837
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:bf92b2b6742c6981c1070e26a51e4731c6820fa9c90fcdc5eca87f09e15c8ed3
3
+ size 1836
unet/diffusion_pytorch_model.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:b4428f1efcbe3a0ad4c80b97cea849dc05a0560e8802970f486db6fa84bf8626
3
  size 5135151440
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:70a5e43f090f63de4ec7e60b4fded46a972caab91877e233eca49b59892363aa
3
  size 5135151440