Spaces:
Running
on
Zero
Running
on
Zero
subject fill
#23
by
Yuanshi
- opened
- README.md +1 -1
- app.py +23 -57
- src/transformer.py +7 -0
README.md
CHANGED
@@ -4,7 +4,7 @@ emoji: π
|
|
4 |
colorFrom: blue
|
5 |
colorTo: green
|
6 |
sdk: gradio
|
7 |
-
sdk_version: 5.
|
8 |
app_file: app.py
|
9 |
pinned: false
|
10 |
---
|
|
|
4 |
colorFrom: blue
|
5 |
colorTo: green
|
6 |
sdk: gradio
|
7 |
+
sdk_version: 5.6.0
|
8 |
app_file: app.py
|
9 |
pinned: false
|
10 |
---
|
app.py
CHANGED
@@ -97,65 +97,31 @@ header = """
|
|
97 |
|
98 |
|
99 |
def create_app():
|
100 |
-
# with gr.Blocks() as app:
|
101 |
-
# gr.Markdown(header, elem_id="header")
|
102 |
-
# # with gr.Tabs():
|
103 |
-
# # with gr.Tab("Subject-driven"):
|
104 |
-
# gr.Interface(
|
105 |
-
# fn=process_image_and_text,
|
106 |
-
# inputs=[
|
107 |
-
# gr.Image(type="pil", label="Condition Image", width=300, elem_id="input"),
|
108 |
-
# gr.Radio(
|
109 |
-
# [("512", 512), ("1024(beta)", 1024)],
|
110 |
-
# label="Resolution",
|
111 |
-
# value=512,
|
112 |
-
# elem_id="resolution",
|
113 |
-
# ),
|
114 |
-
# # gr.Slider(4, 16, 4, step=4, label="Inference Steps"),
|
115 |
-
# gr.Textbox(lines=2, label="Text Prompt", elem_id="text"),
|
116 |
-
# ],
|
117 |
-
# outputs=gr.Image(type="pil", elem_id="output"),
|
118 |
-
# examples=get_samples(),
|
119 |
-
# )
|
120 |
-
# # with gr.Tab("Fill"):
|
121 |
-
# # gr.Markdown("Coming soon")
|
122 |
-
# # with gr.Tab("Canny"):
|
123 |
-
# # gr.Markdown("Coming soon")
|
124 |
-
# # with gr.Tab("Depth"):
|
125 |
-
# # gr.Markdown("Coming soon")
|
126 |
-
|
127 |
with gr.Blocks() as app:
|
128 |
-
gr.Markdown(header
|
129 |
-
with gr.
|
130 |
-
with gr.
|
131 |
-
|
132 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
133 |
)
|
134 |
-
|
135 |
-
|
136 |
-
|
137 |
-
|
138 |
-
|
139 |
-
)
|
140 |
-
text = gr.Textbox(lines=2, label="Text Prompt", elem_id="text")
|
141 |
-
submit_btn = gr.Button("Run", elem_id="submit_btn")
|
142 |
-
|
143 |
-
with gr.Column(variant="panel", elem_classes="outputPanel"):
|
144 |
-
output_image = gr.Image(type="pil", elem_id="output")
|
145 |
-
|
146 |
-
with gr.Row():
|
147 |
-
examples = gr.Examples(
|
148 |
-
examples=get_samples(),
|
149 |
-
inputs=[original_image, resolution, text],
|
150 |
-
label="Examples",
|
151 |
-
)
|
152 |
-
|
153 |
-
submit_btn.click(
|
154 |
-
fn=process_image_and_text,
|
155 |
-
inputs=[original_image, resolution, text],
|
156 |
-
outputs=output_image,
|
157 |
-
)
|
158 |
-
|
159 |
return app
|
160 |
|
161 |
|
|
|
97 |
|
98 |
|
99 |
def create_app():
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
100 |
with gr.Blocks() as app:
|
101 |
+
gr.Markdown(header)
|
102 |
+
with gr.Tabs():
|
103 |
+
with gr.Tab("Subject-driven"):
|
104 |
+
gr.Interface(
|
105 |
+
fn=process_image_and_text,
|
106 |
+
inputs=[
|
107 |
+
gr.Image(type="pil", label="Condition Image", width=300),
|
108 |
+
gr.Radio(
|
109 |
+
[("512", 512), ("1024(beta)", 1024)],
|
110 |
+
label="Resolution",
|
111 |
+
value=512,
|
112 |
+
),
|
113 |
+
# gr.Slider(4, 16, 4, step=4, label="Inference Steps"),
|
114 |
+
gr.Textbox(lines=2, label="Text Prompt"),
|
115 |
+
],
|
116 |
+
outputs=gr.Image(type="pil"),
|
117 |
+
examples=get_samples(),
|
118 |
)
|
119 |
+
with gr.Tab("Fill"):
|
120 |
+
gr.Markdown("Coming soon")
|
121 |
+
with gr.Tab("Canny"):
|
122 |
+
gr.Markdown("Coming soon")
|
123 |
+
with gr.Tab("Depth"):
|
124 |
+
gr.Markdown("Coming soon")
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
125 |
return app
|
126 |
|
127 |
|
src/transformer.py
CHANGED
@@ -7,6 +7,7 @@ from diffusers.models.transformers.transformer_flux import (
|
|
7 |
FluxTransformer2DModel,
|
8 |
Transformer2DModelOutput,
|
9 |
USE_PEFT_BACKEND,
|
|
|
10 |
scale_lora_layers,
|
11 |
unscale_lora_layers,
|
12 |
logger,
|
@@ -154,6 +155,9 @@ def tranformer_forward(
|
|
154 |
|
155 |
return custom_forward
|
156 |
|
|
|
|
|
|
|
157 |
encoder_hidden_states, hidden_states = torch.utils.checkpoint.checkpoint(
|
158 |
create_custom_forward(block),
|
159 |
hidden_states,
|
@@ -200,6 +204,9 @@ def tranformer_forward(
|
|
200 |
|
201 |
return custom_forward
|
202 |
|
|
|
|
|
|
|
203 |
hidden_states = torch.utils.checkpoint.checkpoint(
|
204 |
create_custom_forward(block),
|
205 |
hidden_states,
|
|
|
7 |
FluxTransformer2DModel,
|
8 |
Transformer2DModelOutput,
|
9 |
USE_PEFT_BACKEND,
|
10 |
+
is_torch_version,
|
11 |
scale_lora_layers,
|
12 |
unscale_lora_layers,
|
13 |
logger,
|
|
|
155 |
|
156 |
return custom_forward
|
157 |
|
158 |
+
ckpt_kwargs: Dict[str, Any] = (
|
159 |
+
{"use_reentrant": False} if is_torch_version(">=", "1.11.0") else {}
|
160 |
+
)
|
161 |
encoder_hidden_states, hidden_states = torch.utils.checkpoint.checkpoint(
|
162 |
create_custom_forward(block),
|
163 |
hidden_states,
|
|
|
204 |
|
205 |
return custom_forward
|
206 |
|
207 |
+
ckpt_kwargs: Dict[str, Any] = (
|
208 |
+
{"use_reentrant": False} if is_torch_version(">=", "1.11.0") else {}
|
209 |
+
)
|
210 |
hidden_states = torch.utils.checkpoint.checkpoint(
|
211 |
create_custom_forward(block),
|
212 |
hidden_states,
|