Scheduled Commit
Browse files- ai-toolkit.log +63 -1
ai-toolkit.log
CHANGED
@@ -100,4 +100,66 @@ Loading transformer
|
|
100 |
Quantizing transformer
|
101 |
Loading vae
|
102 |
Loading t5
|
103 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
104 |
0%| | 0/9 [00:00<?, ?it/s]
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
105 |
0%| | 0/9 [00:00<?, ?it/s]
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
106 |
0%| | 0/9 [00:00<?, ?it/s]
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
100 |
Quantizing transformer
|
101 |
Loading vae
|
102 |
Loading t5
|
103 |
+
|
104 |
+
|
105 |
+
Quantizing T5
|
106 |
+
Loading clip
|
107 |
+
making pipe
|
108 |
+
preparing
|
109 |
+
create LoRA network. base dim (rank): 16, alpha: 16
|
110 |
+
neuron dropout: p=None, rank dropout: p=None, module dropout: p=None
|
111 |
+
create LoRA for Text Encoder: 0 modules.
|
112 |
+
create LoRA for U-Net: 494 modules.
|
113 |
+
enable LoRA for U-Net
|
114 |
+
Dataset: /workspace/ai-toolkit/images
|
115 |
+
- Preprocessing image dimensions
|
116 |
+
|
117 |
0%| | 0/9 [00:00<?, ?it/s]
|
118 |
+
- Found 9 images
|
119 |
+
Bucket sizes for /workspace/ai-toolkit/images:
|
120 |
+
128x256: 1 files
|
121 |
+
384x640: 2 files
|
122 |
+
256x832: 1 files
|
123 |
+
256x1024: 1 files
|
124 |
+
256x960: 1 files
|
125 |
+
448x576: 1 files
|
126 |
+
192x448: 1 files
|
127 |
+
256x896: 1 files
|
128 |
+
8 buckets made
|
129 |
+
Caching latents for /workspace/ai-toolkit/images
|
130 |
+
- Saving latents to disk
|
131 |
+
|
132 |
+
Dataset: /workspace/ai-toolkit/images
|
133 |
+
- Preprocessing image dimensions
|
134 |
+
|
135 |
0%| | 0/9 [00:00<?, ?it/s]
|
136 |
+
- Found 9 images
|
137 |
+
Bucket sizes for /workspace/ai-toolkit/images:
|
138 |
+
128x256: 1 files
|
139 |
+
512x1024: 2 files
|
140 |
+
256x896: 2 files
|
141 |
+
384x1536: 1 files
|
142 |
+
384x1408: 1 files
|
143 |
+
640x832: 1 files
|
144 |
+
192x448: 1 files
|
145 |
+
7 buckets made
|
146 |
+
Caching latents for /workspace/ai-toolkit/images
|
147 |
+
- Saving latents to disk
|
148 |
+
|
149 |
+
Dataset: /workspace/ai-toolkit/images
|
150 |
+
- Preprocessing image dimensions
|
151 |
+
|
152 |
0%| | 0/9 [00:00<?, ?it/s]
|
153 |
+
- Found 9 images
|
154 |
+
Bucket sizes for /workspace/ai-toolkit/images:
|
155 |
+
128x256: 1 files
|
156 |
+
576x1088: 1 files
|
157 |
+
256x896: 2 files
|
158 |
+
512x2048: 1 files
|
159 |
+
512x1856: 1 files
|
160 |
+
768x1344: 1 files
|
161 |
+
896x1152: 1 files
|
162 |
+
192x448: 1 files
|
163 |
+
8 buckets made
|
164 |
+
Caching latents for /workspace/ai-toolkit/images
|
165 |
+
- Saving latents to disk
|
166 |
+
|
167 |
+
Generating baseline samples before training
|
168 |
+
|