Update SDXL v2 Lora 12GB.json
Browse filesUpdate to remove personal model paths
- SDXL v2 Lora 12GB.json +292 -295
SDXL v2 Lora 12GB.json
CHANGED
@@ -1,296 +1,293 @@
|
|
1 |
-
{
|
2 |
-
"__version": 3,
|
3 |
-
"training_method": "LORA",
|
4 |
-
"model_type": "STABLE_DIFFUSION_XL_10_BASE",
|
5 |
-
"debug_mode": false,
|
6 |
-
"debug_dir": "debug",
|
7 |
-
"workspace_dir": "workspace/run",
|
8 |
-
"cache_dir": "workspace-cache/run",
|
9 |
-
"tensorboard": true,
|
10 |
-
"tensorboard_expose": false,
|
11 |
-
"continue_last_backup": false,
|
12 |
-
"include_train_config": "NONE",
|
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 |
-
"
|
198 |
-
"
|
199 |
-
"
|
200 |
-
"
|
201 |
-
"
|
202 |
-
"
|
203 |
-
"
|
204 |
-
"
|
205 |
-
"
|
206 |
-
"
|
207 |
-
"
|
208 |
-
"
|
209 |
-
"
|
210 |
-
"
|
211 |
-
"
|
212 |
-
"
|
213 |
-
"
|
214 |
-
"
|
215 |
-
"
|
216 |
-
"
|
217 |
-
"
|
218 |
-
"
|
219 |
-
"
|
220 |
-
"
|
221 |
-
"
|
222 |
-
"
|
223 |
-
"
|
224 |
-
"
|
225 |
-
|
226 |
-
|
227 |
-
"
|
228 |
-
|
229 |
-
|
230 |
-
|
231 |
-
"
|
232 |
-
"
|
233 |
-
"
|
234 |
-
"
|
235 |
-
"
|
236 |
-
"
|
237 |
-
"
|
238 |
-
"
|
239 |
-
"
|
240 |
-
"
|
241 |
-
"
|
242 |
-
"
|
243 |
-
"
|
244 |
-
"
|
245 |
-
"
|
246 |
-
"
|
247 |
-
"
|
248 |
-
"
|
249 |
-
"
|
250 |
-
"
|
251 |
-
"
|
252 |
-
"
|
253 |
-
"
|
254 |
-
"
|
255 |
-
"
|
256 |
-
"
|
257 |
-
"
|
258 |
-
"
|
259 |
-
"
|
260 |
-
"
|
261 |
-
"
|
262 |
-
"
|
263 |
-
"
|
264 |
-
"
|
265 |
-
"
|
266 |
-
"
|
267 |
-
"
|
268 |
-
"
|
269 |
-
"
|
270 |
-
"
|
271 |
-
"
|
272 |
-
"
|
273 |
-
"
|
274 |
-
"
|
275 |
-
"
|
276 |
-
|
277 |
-
|
278 |
-
|
279 |
-
|
280 |
-
|
281 |
-
"
|
282 |
-
"
|
283 |
-
"
|
284 |
-
"
|
285 |
-
"
|
286 |
-
"
|
287 |
-
"
|
288 |
-
"
|
289 |
-
"
|
290 |
-
"
|
291 |
-
"
|
292 |
-
"
|
293 |
-
"save_after": 10,
|
294 |
-
"save_after_unit": "EPOCH",
|
295 |
-
"save_filename_prefix": ""
|
296 |
}
|
|
|
1 |
+
{
|
2 |
+
"__version": 3,
|
3 |
+
"training_method": "LORA",
|
4 |
+
"model_type": "STABLE_DIFFUSION_XL_10_BASE",
|
5 |
+
"debug_mode": false,
|
6 |
+
"debug_dir": "debug",
|
7 |
+
"workspace_dir": "workspace/run",
|
8 |
+
"cache_dir": "workspace-cache/run",
|
9 |
+
"tensorboard": true,
|
10 |
+
"tensorboard_expose": false,
|
11 |
+
"continue_last_backup": false,
|
12 |
+
"include_train_config": "NONE",
|
13 |
+
"weight_dtype": "FLOAT_16",
|
14 |
+
"output_dtype": "FLOAT_32",
|
15 |
+
"output_model_format": "SAFETENSORS",
|
16 |
+
"gradient_checkpointing": true,
|
17 |
+
"force_circular_padding": false,
|
18 |
+
"concepts": null,
|
19 |
+
"circular_mask_generation": false,
|
20 |
+
"random_rotate_and_crop": false,
|
21 |
+
"aspect_ratio_bucketing": true,
|
22 |
+
"latent_caching": true,
|
23 |
+
"clear_cache_before_training": true,
|
24 |
+
"learning_rate_scheduler": "CONSTANT",
|
25 |
+
"custom_learning_rate_scheduler": null,
|
26 |
+
"scheduler_params": [],
|
27 |
+
"learning_rate": 0.0003,
|
28 |
+
"learning_rate_warmup_steps": 200,
|
29 |
+
"learning_rate_cycles": 1,
|
30 |
+
"epochs": 20,
|
31 |
+
"batch_size": 4,
|
32 |
+
"gradient_accumulation_steps": 1,
|
33 |
+
"ema": "OFF",
|
34 |
+
"ema_decay": 0.999,
|
35 |
+
"ema_update_step_interval": 5,
|
36 |
+
"dataloader_threads": 2,
|
37 |
+
"train_device": "cuda",
|
38 |
+
"temp_device": "cpu",
|
39 |
+
"train_dtype": "FLOAT_16",
|
40 |
+
"fallback_train_dtype": "BFLOAT_16",
|
41 |
+
"enable_autocast_cache": true,
|
42 |
+
"only_cache": false,
|
43 |
+
"resolution": "896",
|
44 |
+
"attention_mechanism": "XFORMERS",
|
45 |
+
"align_prop": false,
|
46 |
+
"align_prop_probability": 0.1,
|
47 |
+
"align_prop_loss": "AESTHETIC",
|
48 |
+
"align_prop_weight": 0.01,
|
49 |
+
"align_prop_steps": 20,
|
50 |
+
"align_prop_truncate_steps": 0.5,
|
51 |
+
"align_prop_cfg_scale": 7.0,
|
52 |
+
"mse_strength": 1.0,
|
53 |
+
"mae_strength": 0.0,
|
54 |
+
"vb_loss_strength": 1.0,
|
55 |
+
"loss_weight_fn": "CONSTANT",
|
56 |
+
"loss_weight_strength": 5.0,
|
57 |
+
"dropout_probability": 0.0,
|
58 |
+
"loss_scaler": "NONE",
|
59 |
+
"learning_rate_scaler": "NONE",
|
60 |
+
"offset_noise_weight": 0.0,
|
61 |
+
"perturbation_noise_weight": 0.0,
|
62 |
+
"rescale_noise_scheduler_to_zero_terminal_snr": false,
|
63 |
+
"force_v_prediction": false,
|
64 |
+
"force_epsilon_prediction": false,
|
65 |
+
"min_noising_strength": 0.0,
|
66 |
+
"max_noising_strength": 1.0,
|
67 |
+
"noising_weight": 0.0,
|
68 |
+
"noising_bias": 0.5,
|
69 |
+
"unet": {
|
70 |
+
"__version": 0,
|
71 |
+
"model_name": "",
|
72 |
+
"train": true,
|
73 |
+
"stop_training_after": 100000,
|
74 |
+
"stop_training_after_unit": "EPOCH",
|
75 |
+
"learning_rate": null,
|
76 |
+
"weight_dtype": "NONE"
|
77 |
+
},
|
78 |
+
"prior": {
|
79 |
+
"__version": 0,
|
80 |
+
"model_name": "",
|
81 |
+
"train": true,
|
82 |
+
"stop_training_after": 10000,
|
83 |
+
"stop_training_after_unit": "EPOCH",
|
84 |
+
"learning_rate": null,
|
85 |
+
"weight_dtype": "NONE"
|
86 |
+
},
|
87 |
+
"text_encoder": {
|
88 |
+
"__version": 0,
|
89 |
+
"model_name": "",
|
90 |
+
"train": false,
|
91 |
+
"stop_training_after": 30,
|
92 |
+
"stop_training_after_unit": "EPOCH",
|
93 |
+
"learning_rate": null,
|
94 |
+
"weight_dtype": "NONE"
|
95 |
+
},
|
96 |
+
"text_encoder_layer_skip": 0,
|
97 |
+
"text_encoder_2": {
|
98 |
+
"__version": 0,
|
99 |
+
"model_name": "",
|
100 |
+
"train": false,
|
101 |
+
"stop_training_after": 30,
|
102 |
+
"stop_training_after_unit": "EPOCH",
|
103 |
+
"learning_rate": null,
|
104 |
+
"weight_dtype": "NONE"
|
105 |
+
},
|
106 |
+
"text_encoder_2_layer_skip": 0,
|
107 |
+
"vae": {
|
108 |
+
"__version": 0,
|
109 |
+
"model_name": "",
|
110 |
+
"train": true,
|
111 |
+
"stop_training_after": null,
|
112 |
+
"stop_training_after_unit": "NEVER",
|
113 |
+
"learning_rate": null,
|
114 |
+
"weight_dtype": "FLOAT_32"
|
115 |
+
},
|
116 |
+
"effnet_encoder": {
|
117 |
+
"__version": 0,
|
118 |
+
"model_name": "",
|
119 |
+
"train": true,
|
120 |
+
"stop_training_after": null,
|
121 |
+
"stop_training_after_unit": "NEVER",
|
122 |
+
"learning_rate": null,
|
123 |
+
"weight_dtype": "NONE"
|
124 |
+
},
|
125 |
+
"decoder": {
|
126 |
+
"__version": 0,
|
127 |
+
"model_name": "",
|
128 |
+
"train": true,
|
129 |
+
"stop_training_after": null,
|
130 |
+
"stop_training_after_unit": "NEVER",
|
131 |
+
"learning_rate": null,
|
132 |
+
"weight_dtype": "NONE"
|
133 |
+
},
|
134 |
+
"decoder_text_encoder": {
|
135 |
+
"__version": 0,
|
136 |
+
"model_name": "",
|
137 |
+
"train": true,
|
138 |
+
"stop_training_after": null,
|
139 |
+
"stop_training_after_unit": "NEVER",
|
140 |
+
"learning_rate": null,
|
141 |
+
"weight_dtype": "NONE"
|
142 |
+
},
|
143 |
+
"decoder_vqgan": {
|
144 |
+
"__version": 0,
|
145 |
+
"model_name": "",
|
146 |
+
"train": true,
|
147 |
+
"stop_training_after": null,
|
148 |
+
"stop_training_after_unit": "NEVER",
|
149 |
+
"learning_rate": null,
|
150 |
+
"weight_dtype": "NONE"
|
151 |
+
},
|
152 |
+
"masked_training": false,
|
153 |
+
"unmasked_probability": 0.1,
|
154 |
+
"unmasked_weight": 0.1,
|
155 |
+
"normalize_masked_area_loss": false,
|
156 |
+
"embedding_learning_rate": null,
|
157 |
+
"preserve_embedding_norm": false,
|
158 |
+
"embedding": {
|
159 |
+
"__version": 0,
|
160 |
+
"uuid": "fed98d8d-5c8f-470e-8861-73a603175824",
|
161 |
+
"model_name": "",
|
162 |
+
"placeholder": "<embedding>",
|
163 |
+
"train": true,
|
164 |
+
"stop_training_after": null,
|
165 |
+
"stop_training_after_unit": "NEVER",
|
166 |
+
"token_count": 1,
|
167 |
+
"initial_embedding_text": "*"
|
168 |
+
},
|
169 |
+
"additional_embeddings": [],
|
170 |
+
"embedding_weight_dtype": "FLOAT_32",
|
171 |
+
"lora_model_name": "",
|
172 |
+
"lora_rank": 16,
|
173 |
+
"lora_alpha": 1.0,
|
174 |
+
"lora_weight_dtype": "FLOAT_32",
|
175 |
+
"bundle_additional_embeddings": true,
|
176 |
+
"optimizer": {
|
177 |
+
"__version": 0,
|
178 |
+
"optimizer": "ADAMW",
|
179 |
+
"adam_w_mode": false,
|
180 |
+
"alpha": null,
|
181 |
+
"amsgrad": false,
|
182 |
+
"beta1": 0.9,
|
183 |
+
"beta2": 0.999,
|
184 |
+
"beta3": null,
|
185 |
+
"bias_correction": false,
|
186 |
+
"block_wise": false,
|
187 |
+
"capturable": false,
|
188 |
+
"centered": false,
|
189 |
+
"clip_threshold": null,
|
190 |
+
"d0": null,
|
191 |
+
"d_coef": null,
|
192 |
+
"dampening": null,
|
193 |
+
"decay_rate": null,
|
194 |
+
"decouple": false,
|
195 |
+
"differentiable": false,
|
196 |
+
"eps": 1e-08,
|
197 |
+
"eps2": null,
|
198 |
+
"foreach": false,
|
199 |
+
"fsdp_in_use": false,
|
200 |
+
"fused": true,
|
201 |
+
"fused_back_pass": false,
|
202 |
+
"growth_rate": null,
|
203 |
+
"initial_accumulator_value": null,
|
204 |
+
"is_paged": false,
|
205 |
+
"log_every": null,
|
206 |
+
"lr_decay": null,
|
207 |
+
"max_unorm": null,
|
208 |
+
"maximize": false,
|
209 |
+
"min_8bit_size": null,
|
210 |
+
"momentum": null,
|
211 |
+
"nesterov": false,
|
212 |
+
"no_prox": false,
|
213 |
+
"optim_bits": null,
|
214 |
+
"percentile_clipping": null,
|
215 |
+
"r": null,
|
216 |
+
"relative_step": false,
|
217 |
+
"safeguard_warmup": false,
|
218 |
+
"scale_parameter": false,
|
219 |
+
"stochastic_rounding": true,
|
220 |
+
"use_bias_correction": false,
|
221 |
+
"use_triton": false,
|
222 |
+
"warmup_init": false,
|
223 |
+
"weight_decay": 0.01,
|
224 |
+
"weight_lr_power": null
|
225 |
+
},
|
226 |
+
"optimizer_defaults": {
|
227 |
+
"ADAMW": {
|
228 |
+
"__version": 0,
|
229 |
+
"optimizer": "ADAMW",
|
230 |
+
"adam_w_mode": false,
|
231 |
+
"alpha": null,
|
232 |
+
"amsgrad": false,
|
233 |
+
"beta1": 0.9,
|
234 |
+
"beta2": 0.999,
|
235 |
+
"beta3": null,
|
236 |
+
"bias_correction": false,
|
237 |
+
"block_wise": false,
|
238 |
+
"capturable": false,
|
239 |
+
"centered": false,
|
240 |
+
"clip_threshold": null,
|
241 |
+
"d0": null,
|
242 |
+
"d_coef": null,
|
243 |
+
"dampening": null,
|
244 |
+
"decay_rate": null,
|
245 |
+
"decouple": false,
|
246 |
+
"differentiable": false,
|
247 |
+
"eps": 1e-08,
|
248 |
+
"eps2": null,
|
249 |
+
"foreach": false,
|
250 |
+
"fsdp_in_use": false,
|
251 |
+
"fused": true,
|
252 |
+
"fused_back_pass": false,
|
253 |
+
"growth_rate": null,
|
254 |
+
"initial_accumulator_value": null,
|
255 |
+
"is_paged": false,
|
256 |
+
"log_every": null,
|
257 |
+
"lr_decay": null,
|
258 |
+
"max_unorm": null,
|
259 |
+
"maximize": false,
|
260 |
+
"min_8bit_size": null,
|
261 |
+
"momentum": null,
|
262 |
+
"nesterov": false,
|
263 |
+
"no_prox": false,
|
264 |
+
"optim_bits": null,
|
265 |
+
"percentile_clipping": null,
|
266 |
+
"r": null,
|
267 |
+
"relative_step": false,
|
268 |
+
"safeguard_warmup": false,
|
269 |
+
"scale_parameter": false,
|
270 |
+
"stochastic_rounding": false,
|
271 |
+
"use_bias_correction": false,
|
272 |
+
"use_triton": false,
|
273 |
+
"warmup_init": false,
|
274 |
+
"weight_decay": 0.01,
|
275 |
+
"weight_lr_power": null
|
276 |
+
}
|
277 |
+
},
|
278 |
+
"sample_definition_file_name": "training_samples/samples.json",
|
279 |
+
"samples": null,
|
280 |
+
"sample_after": 10,
|
281 |
+
"sample_after_unit": "NEVER",
|
282 |
+
"sample_image_format": "JPG",
|
283 |
+
"samples_to_tensorboard": true,
|
284 |
+
"non_ema_sampling": true,
|
285 |
+
"backup_after": 1,
|
286 |
+
"backup_after_unit": "HOUR",
|
287 |
+
"rolling_backup": false,
|
288 |
+
"rolling_backup_count": 3,
|
289 |
+
"backup_before_save": true,
|
290 |
+
"save_after": 10,
|
291 |
+
"save_after_unit": "EPOCH",
|
292 |
+
"save_filename_prefix": ""
|
|
|
|
|
|
|
293 |
}
|