Fabrice-TIERCELIN
commited on
Commit
•
8fb74d6
1
Parent(s):
d565481
Validate parameters before
Browse files- gradio_demo.py +89 -28
gradio_demo.py
CHANGED
@@ -30,7 +30,7 @@ parser.add_argument("--no_llava", action='store_true', default=True)#False
|
|
30 |
parser.add_argument("--use_image_slider", action='store_true', default=False)
|
31 |
parser.add_argument("--log_history", action='store_true', default=False)
|
32 |
parser.add_argument("--loading_half_params", action='store_true', default=True)#False
|
33 |
-
parser.add_argument("--use_tile_vae", action='store_true', default=False
|
34 |
parser.add_argument("--encoder_tile_size", type=int, default=512)
|
35 |
parser.add_argument("--decoder_tile_size", type=int, default=64)
|
36 |
parser.add_argument("--load_8bit_llava", action='store_true', default=False)
|
@@ -67,15 +67,16 @@ if torch.cuda.device_count() > 0:
|
|
67 |
else:
|
68 |
llava_agent = None
|
69 |
|
70 |
-
|
|
|
|
|
|
|
|
|
71 |
def stage1_process(input_image, gamma_correction):
|
72 |
print('Start stage1_process')
|
73 |
if torch.cuda.device_count() == 0:
|
74 |
gr.Warning('Set this space to GPU config to make it work.')
|
75 |
return None
|
76 |
-
if input_image is None:
|
77 |
-
gr.Warning('Please provide an image to restore.')
|
78 |
-
return None
|
79 |
torch.cuda.set_device(SUPIR_device)
|
80 |
LQ = HWC3(input_image)
|
81 |
LQ = fix_resize(LQ, 512)
|
@@ -92,15 +93,12 @@ def stage1_process(input_image, gamma_correction):
|
|
92 |
print('End stage1_process')
|
93 |
return LQ
|
94 |
|
95 |
-
@spaces.GPU(duration=
|
96 |
def llave_process(input_image, temperature, top_p, qs=None):
|
97 |
print('Start llave_process')
|
98 |
if torch.cuda.device_count() == 0:
|
99 |
gr.Warning('Set this space to GPU config to make it work.')
|
100 |
return 'Set this space to GPU config to make it work.'
|
101 |
-
if input_image is None:
|
102 |
-
gr.Warning('Please provide an image to restore.')
|
103 |
-
return 'Please provide an image to restore.'
|
104 |
torch.cuda.set_device(LLaVA_device)
|
105 |
if use_llava:
|
106 |
LQ = HWC3(input_image)
|
@@ -111,7 +109,7 @@ def llave_process(input_image, temperature, top_p, qs=None):
|
|
111 |
print('End llave_process')
|
112 |
return captions[0]
|
113 |
|
114 |
-
@spaces.GPU(duration=
|
115 |
def stage2_process(input_image, prompt, a_prompt, n_prompt, num_samples, upscale, edm_steps, s_stage1, s_stage2,
|
116 |
s_cfg, seed, s_churn, s_noise, color_fix_type, diff_dtype, ae_dtype, gamma_correction,
|
117 |
linear_CFG, linear_s_stage2, spt_linear_CFG, spt_linear_s_stage2, model_select):
|
@@ -119,9 +117,6 @@ def stage2_process(input_image, prompt, a_prompt, n_prompt, num_samples, upscale
|
|
119 |
if torch.cuda.device_count() == 0:
|
120 |
gr.Warning('Set this space to GPU config to make it work.')
|
121 |
return None, None, None, None
|
122 |
-
if input_image is None:
|
123 |
-
gr.Warning('Please provide an image to restore.')
|
124 |
-
return None, None, None, None
|
125 |
torch.cuda.set_device(SUPIR_device)
|
126 |
event_id = str(time.time_ns())
|
127 |
event_dict = {'event_id': event_id, 'localtime': time.ctime(), 'prompt': prompt, 'a_prompt': a_prompt,
|
@@ -279,7 +274,7 @@ with gr.Blocks(title='SUPIR') as interface:
|
|
279 |
qs = gr.Textbox(label="Question", info="Describe the image and its style in a very detailed manner", placeholder="The image is a realistic photography, not an art painting.")
|
280 |
|
281 |
with gr.Accordion("Restoring options", open=False):
|
282 |
-
num_samples = gr.Slider(label="Num Samples", info="Number of generated results; I discourage to increase because the process is limited to
|
283 |
, value=1, step=1)
|
284 |
upscale = gr.Slider(label="Upscale", info="The resolution increase factor", minimum=1, maximum=8, value=1, step=1)
|
285 |
edm_steps = gr.Slider(label="Steps", info="lower=faster, higher=more details", minimum=1, maximum=200, value=default_setting.edm_steps if torch.cuda.device_count() > 0 else 1, step=1)
|
@@ -319,10 +314,10 @@ with gr.Blocks(title='SUPIR') as interface:
|
|
319 |
ae_dtype = gr.Radio(['fp32', 'bf16'], label="Auto-Encoder Data Type", value="bf16",
|
320 |
interactive=True)
|
321 |
with gr.Column():
|
322 |
-
color_fix_type = gr.Radio(["None", "AdaIn", "Wavelet"], label="Color-Fix Type", value="Wavelet",
|
323 |
interactive=True)
|
324 |
with gr.Column():
|
325 |
-
model_select = gr.Radio(["v0-Q", "v0-F"], label="Model Selection", value="v0-Q",
|
326 |
interactive=True)
|
327 |
|
328 |
with gr.Column():
|
@@ -352,17 +347,83 @@ with gr.Blocks(title='SUPIR') as interface:
|
|
352 |
with gr.Row():
|
353 |
gr.Markdown(claim_md)
|
354 |
event_id = gr.Textbox(label="Event ID", value="", visible=False)
|
355 |
-
|
356 |
-
|
357 |
-
|
358 |
-
|
359 |
-
|
360 |
-
|
361 |
-
|
362 |
-
|
363 |
-
|
364 |
-
|
365 |
-
|
366 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
367 |
|
368 |
interface.queue(10).launch()
|
|
|
30 |
parser.add_argument("--use_image_slider", action='store_true', default=False)
|
31 |
parser.add_argument("--log_history", action='store_true', default=False)
|
32 |
parser.add_argument("--loading_half_params", action='store_true', default=True)#False
|
33 |
+
parser.add_argument("--use_tile_vae", action='store_true', default=True)#False
|
34 |
parser.add_argument("--encoder_tile_size", type=int, default=512)
|
35 |
parser.add_argument("--decoder_tile_size", type=int, default=64)
|
36 |
parser.add_argument("--load_8bit_llava", action='store_true', default=False)
|
|
|
67 |
else:
|
68 |
llava_agent = None
|
69 |
|
70 |
+
def check(input_image):
|
71 |
+
if input_image is None:
|
72 |
+
raise gr.Error("Please provide an image to restore.")
|
73 |
+
|
74 |
+
@spaces.GPU(duration=180)
|
75 |
def stage1_process(input_image, gamma_correction):
|
76 |
print('Start stage1_process')
|
77 |
if torch.cuda.device_count() == 0:
|
78 |
gr.Warning('Set this space to GPU config to make it work.')
|
79 |
return None
|
|
|
|
|
|
|
80 |
torch.cuda.set_device(SUPIR_device)
|
81 |
LQ = HWC3(input_image)
|
82 |
LQ = fix_resize(LQ, 512)
|
|
|
93 |
print('End stage1_process')
|
94 |
return LQ
|
95 |
|
96 |
+
@spaces.GPU(duration=180)
|
97 |
def llave_process(input_image, temperature, top_p, qs=None):
|
98 |
print('Start llave_process')
|
99 |
if torch.cuda.device_count() == 0:
|
100 |
gr.Warning('Set this space to GPU config to make it work.')
|
101 |
return 'Set this space to GPU config to make it work.'
|
|
|
|
|
|
|
102 |
torch.cuda.set_device(LLaVA_device)
|
103 |
if use_llava:
|
104 |
LQ = HWC3(input_image)
|
|
|
109 |
print('End llave_process')
|
110 |
return captions[0]
|
111 |
|
112 |
+
@spaces.GPU(duration=180)
|
113 |
def stage2_process(input_image, prompt, a_prompt, n_prompt, num_samples, upscale, edm_steps, s_stage1, s_stage2,
|
114 |
s_cfg, seed, s_churn, s_noise, color_fix_type, diff_dtype, ae_dtype, gamma_correction,
|
115 |
linear_CFG, linear_s_stage2, spt_linear_CFG, spt_linear_s_stage2, model_select):
|
|
|
117 |
if torch.cuda.device_count() == 0:
|
118 |
gr.Warning('Set this space to GPU config to make it work.')
|
119 |
return None, None, None, None
|
|
|
|
|
|
|
120 |
torch.cuda.set_device(SUPIR_device)
|
121 |
event_id = str(time.time_ns())
|
122 |
event_dict = {'event_id': event_id, 'localtime': time.ctime(), 'prompt': prompt, 'a_prompt': a_prompt,
|
|
|
274 |
qs = gr.Textbox(label="Question", info="Describe the image and its style in a very detailed manner", placeholder="The image is a realistic photography, not an art painting.")
|
275 |
|
276 |
with gr.Accordion("Restoring options", open=False):
|
277 |
+
num_samples = gr.Slider(label="Num Samples", info="Number of generated results; I discourage to increase because the process is limited to 3 min", minimum=1, maximum=4 if not args.use_image_slider else 1
|
278 |
, value=1, step=1)
|
279 |
upscale = gr.Slider(label="Upscale", info="The resolution increase factor", minimum=1, maximum=8, value=1, step=1)
|
280 |
edm_steps = gr.Slider(label="Steps", info="lower=faster, higher=more details", minimum=1, maximum=200, value=default_setting.edm_steps if torch.cuda.device_count() > 0 else 1, step=1)
|
|
|
314 |
ae_dtype = gr.Radio(['fp32', 'bf16'], label="Auto-Encoder Data Type", value="bf16",
|
315 |
interactive=True)
|
316 |
with gr.Column():
|
317 |
+
color_fix_type = gr.Radio(["None", "AdaIn", "Wavelet"], label="Color-Fix Type", info="Wavelet=For JPEG artifacts", value="Wavelet",
|
318 |
interactive=True)
|
319 |
with gr.Column():
|
320 |
+
model_select = gr.Radio(["v0-Q", "v0-F"], label="Model Selection", info="Q=Quality, F=Fidelity", value="v0-Q",
|
321 |
interactive=True)
|
322 |
|
323 |
with gr.Column():
|
|
|
347 |
with gr.Row():
|
348 |
gr.Markdown(claim_md)
|
349 |
event_id = gr.Textbox(label="Event ID", value="", visible=False)
|
350 |
+
|
351 |
+
denoise_button.click(fn = check, inputs = [
|
352 |
+
input_image
|
353 |
+
], outputs = [], queue = False, show_progress = False).success(fn = stage1_process, inputs = [
|
354 |
+
input_image,
|
355 |
+
gamma_correction
|
356 |
+
], outputs=[
|
357 |
+
denoise_image
|
358 |
+
])
|
359 |
+
|
360 |
+
llave_button.click(fn = check, inputs = [
|
361 |
+
denoise_image
|
362 |
+
], outputs = [], queue = False, show_progress = False).success(fn = llave_process, inputs = [
|
363 |
+
denoise_image,
|
364 |
+
temperature,
|
365 |
+
top_p,
|
366 |
+
qs
|
367 |
+
], outputs = [
|
368 |
+
prompt
|
369 |
+
])
|
370 |
+
|
371 |
+
diffusion_button.click(fn = check, inputs = [
|
372 |
+
input_image
|
373 |
+
], outputs = [], queue = False, show_progress = False).success(fn=stage2_process, inputs = [
|
374 |
+
input_image,
|
375 |
+
prompt,
|
376 |
+
a_prompt,
|
377 |
+
n_prompt,
|
378 |
+
num_samples,
|
379 |
+
upscale,
|
380 |
+
edm_steps,
|
381 |
+
s_stage1,
|
382 |
+
s_stage2,
|
383 |
+
s_cfg,
|
384 |
+
seed,
|
385 |
+
s_churn,
|
386 |
+
s_noise,
|
387 |
+
color_fix_type,
|
388 |
+
diff_dtype,
|
389 |
+
ae_dtype,
|
390 |
+
gamma_correction,
|
391 |
+
linear_CFG,
|
392 |
+
linear_s_stage2,
|
393 |
+
spt_linear_CFG,
|
394 |
+
spt_linear_s_stage2,
|
395 |
+
model_select
|
396 |
+
], outputs = [
|
397 |
+
result_gallery,
|
398 |
+
event_id,
|
399 |
+
fb_score,
|
400 |
+
fb_text
|
401 |
+
])
|
402 |
+
|
403 |
+
restart_button.click(fn = load_and_reset, inputs = [
|
404 |
+
param_setting
|
405 |
+
], outputs = [
|
406 |
+
edm_steps,
|
407 |
+
s_cfg,
|
408 |
+
s_stage2,
|
409 |
+
s_stage1,
|
410 |
+
s_churn,
|
411 |
+
s_noise,
|
412 |
+
a_prompt,
|
413 |
+
n_prompt,
|
414 |
+
color_fix_type,
|
415 |
+
linear_CFG,
|
416 |
+
linear_s_stage2,
|
417 |
+
spt_linear_CFG,
|
418 |
+
spt_linear_s_stage2
|
419 |
+
])
|
420 |
+
|
421 |
+
submit_button.click(fn = submit_feedback, inputs = [
|
422 |
+
event_id,
|
423 |
+
fb_score,
|
424 |
+
fb_text
|
425 |
+
], outputs = [
|
426 |
+
fb_text
|
427 |
+
])
|
428 |
|
429 |
interface.queue(10).launch()
|