Commit
•
d1c3953
1
Parent(s):
7d4fd85
Update app.py
Browse files
app.py
CHANGED
@@ -69,12 +69,13 @@ def train(*inputs):
|
|
69 |
file_counter += 1
|
70 |
|
71 |
uses_custom = inputs[-1]
|
|
|
72 |
if(uses_custom):
|
73 |
Training_Steps = int(inputs[-3])
|
74 |
Train_text_encoder_for = int(inputs[-2])
|
75 |
else:
|
76 |
Training_Steps = file_counter*200
|
77 |
-
if(
|
78 |
class_data_dir = "mix"
|
79 |
Train_text_encoder_for=100
|
80 |
args_txt_encoder = argparse.Namespace(
|
@@ -124,41 +125,41 @@ def train(*inputs):
|
|
124 |
)
|
125 |
run_training(args_txt_encoder)
|
126 |
run_training(args_unet)
|
127 |
-
elif(
|
128 |
-
|
|
|
|
|
|
|
129 |
class_data_dir = None
|
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 |
-
run_training(args_general)
|
159 |
-
os.rmdir('instance_images')
|
160 |
-
shutil.make_archive("output_model.zip", 'zip', "output_model")
|
161 |
return gr.update(visible=True, value="output_model.zip")
|
|
|
162 |
with gr.Blocks(css=css) as demo:
|
163 |
with gr.Box():
|
164 |
# You can remove this part here for your local clone
|
|
|
69 |
file_counter += 1
|
70 |
|
71 |
uses_custom = inputs[-1]
|
72 |
+
type_of_thing = inputs[-4]
|
73 |
if(uses_custom):
|
74 |
Training_Steps = int(inputs[-3])
|
75 |
Train_text_encoder_for = int(inputs[-2])
|
76 |
else:
|
77 |
Training_Steps = file_counter*200
|
78 |
+
if(type_of_thing == "person"):
|
79 |
class_data_dir = "mix"
|
80 |
Train_text_encoder_for=100
|
81 |
args_txt_encoder = argparse.Namespace(
|
|
|
125 |
)
|
126 |
run_training(args_txt_encoder)
|
127 |
run_training(args_unet)
|
128 |
+
elif(type_of_thing == "object" or type_of_thing == "style"):
|
129 |
+
if(type_of_thing == "object"):
|
130 |
+
Train_text_encoder_for=30
|
131 |
+
elif(type_of_thing == "style"):
|
132 |
+
Train_text_encoder_for=15
|
133 |
class_data_dir = None
|
134 |
+
stptxt = int((Training_Steps*Train_text_encoder_for)/100)
|
135 |
+
args_general = argparse.Namespace(
|
136 |
+
image_captions_filename = True,
|
137 |
+
train_text_encoder = True,
|
138 |
+
stop_text_encoder_training = stptxt,
|
139 |
+
save_n_steps = 0,
|
140 |
+
pretrained_model_name_or_path = model_to_load,
|
141 |
+
instance_data_dir="instance_images",
|
142 |
+
class_data_dir=class_data_dir,
|
143 |
+
output_dir="output_model",
|
144 |
+
instance_prompt="",
|
145 |
+
seed=42,
|
146 |
+
resolution=512,
|
147 |
+
mixed_precision="fp16",
|
148 |
+
train_batch_size=1,
|
149 |
+
gradient_accumulation_steps=1,
|
150 |
+
use_8bit_adam=True,
|
151 |
+
learning_rate=2e-6,
|
152 |
+
lr_scheduler="polynomial",
|
153 |
+
lr_warmup_steps = 0,
|
154 |
+
max_train_steps=Training_Steps,
|
155 |
+
)
|
156 |
+
run_training(args_general)
|
157 |
+
|
158 |
+
shutil.rmtree('instance_images')
|
159 |
+
shutil.make_archive("output_model", 'zip', "output_model")
|
160 |
+
shutil.rmtree("output_model")
|
|
|
|
|
|
|
|
|
161 |
return gr.update(visible=True, value="output_model.zip")
|
162 |
+
|
163 |
with gr.Blocks(css=css) as demo:
|
164 |
with gr.Box():
|
165 |
# You can remove this part here for your local clone
|