Spaces:
Runtime error
Runtime error
Fabrice-TIERCELIN
commited on
Comment more code
Browse files
app.py
CHANGED
@@ -94,81 +94,79 @@ class Tango:
|
|
94 |
return outputs
|
95 |
return list(self.chunks(outputs, samples))
|
96 |
|
97 |
-
|
98 |
-
|
99 |
-
tango = Tango(device = "cpu")
|
100 |
-
tango.vae.to(device_type)
|
101 |
-
tango.stft.to(device_type)
|
102 |
-
tango.model.to(device_type)
|
103 |
-
|
104 |
-
def update_seed(is_randomize_seed, seed):
|
105 |
-
if is_randomize_seed:
|
106 |
-
return random.randint(0, max_64_bit_int)
|
107 |
-
return seed
|
108 |
-
|
109 |
-
def check(
|
110 |
-
prompt,
|
111 |
-
output_number,
|
112 |
-
steps,
|
113 |
-
guidance,
|
114 |
-
is_randomize_seed,
|
115 |
-
seed
|
116 |
-
):
|
117 |
-
if prompt is None or prompt == "":
|
118 |
-
raise gr.Error("Please provide a prompt input.")
|
119 |
-
if not output_number in [1, 2, 3]:
|
120 |
-
raise gr.Error("Please ask for 1, 2 or 3 output files.")
|
121 |
-
|
122 |
-
def update_output(output_format, output_number):
|
123 |
-
return [
|
124 |
-
gr.update(format = output_format),
|
125 |
-
gr.update(format = output_format, visible = (2 <= output_number)),
|
126 |
-
gr.update(format = output_format, visible = (output_number == 3)),
|
127 |
-
gr.update(visible = False)
|
128 |
-
]
|
129 |
-
|
130 |
-
def text2audio(
|
131 |
-
prompt,
|
132 |
-
output_number,
|
133 |
-
steps,
|
134 |
-
guidance,
|
135 |
-
is_randomize_seed,
|
136 |
-
seed
|
137 |
-
):
|
138 |
-
start = time.time()
|
139 |
-
|
140 |
-
if seed is None:
|
141 |
-
seed = random.randint(0, max_64_bit_int)
|
142 |
-
|
143 |
-
random.seed(seed)
|
144 |
-
torch.manual_seed(seed)
|
145 |
-
|
146 |
-
output_wave = tango.generate(prompt, steps, guidance, output_number)
|
147 |
-
|
148 |
-
output_wave_1 = gr.make_waveform((16000, output_wave[0]))
|
149 |
-
output_wave_2 = gr.make_waveform((16000, output_wave[1])) if (2 <= output_number) else None
|
150 |
-
output_wave_3 = gr.make_waveform((16000, output_wave[2])) if (output_number == 3) else None
|
151 |
-
|
152 |
-
end = time.time()
|
153 |
-
secondes = int(end - start)
|
154 |
-
minutes = secondes // 60
|
155 |
-
secondes = secondes - (minutes * 60)
|
156 |
-
hours = minutes // 60
|
157 |
-
minutes = minutes - (hours * 60)
|
158 |
-
return [
|
159 |
-
output_wave_1,
|
160 |
-
output_wave_2,
|
161 |
-
output_wave_3,
|
162 |
-
gr.update(visible = True, value = "Start again to get a different result. The output have been generated in " + ((str(hours) + " h, ") if hours != 0 else "") + ((str(minutes) + " min, ") if hours != 0 or minutes != 0 else "") + str(secondes) + " sec.")
|
163 |
-
]
|
164 |
-
|
165 |
-
if is_space_imported:
|
166 |
-
text2audio = spaces.GPU(text2audio, duration = 420)
|
167 |
|
168 |
# Old code
|
169 |
net=BriaRMBG()
|
170 |
-
# model_path = "./model1.pth"
|
171 |
-
#model_path = hf_hub_download("briaai/RMBG-1.4", 'model.pth')
|
172 |
model_path = hf_hub_download("cocktailpeanut/gbmr", 'model.pth')
|
173 |
if torch.cuda.is_available():
|
174 |
net.load_state_dict(torch.load(model_path))
|
@@ -220,36 +218,8 @@ def process(image):
|
|
220 |
# paste the mask on the original image
|
221 |
new_im = Image.new("RGBA", pil_im.size, (0,0,0,0))
|
222 |
new_im.paste(orig_image, mask=pil_im)
|
223 |
-
# new_orig_image = orig_image.convert('RGBA')
|
224 |
|
225 |
return new_im
|
226 |
-
# return [new_orig_image, new_im]
|
227 |
-
|
228 |
-
|
229 |
-
# block = gr.Blocks().queue()
|
230 |
-
|
231 |
-
# with block:
|
232 |
-
# gr.Markdown("## BRIA RMBG 1.4")
|
233 |
-
# gr.HTML('''
|
234 |
-
# <p style="margin-bottom: 10px; font-size: 94%">
|
235 |
-
# This is a demo for BRIA RMBG 1.4 that using
|
236 |
-
# <a href="https://huggingface.co/briaai/RMBG-1.4" target="_blank">BRIA RMBG-1.4 image matting model</a> as backbone.
|
237 |
-
# </p>
|
238 |
-
# ''')
|
239 |
-
# with gr.Row():
|
240 |
-
# with gr.Column():
|
241 |
-
# input_image = gr.Image(sources=None, type="pil") # None for upload, ctrl+v and webcam
|
242 |
-
# # input_image = gr.Image(sources=None, type="numpy") # None for upload, ctrl+v and webcam
|
243 |
-
# run_button = gr.Button(value="Run")
|
244 |
-
|
245 |
-
# with gr.Column():
|
246 |
-
# result_gallery = gr.Gallery(label='Output', show_label=False, elem_id="gallery", columns=[1], height='auto')
|
247 |
-
# ips = [input_image]
|
248 |
-
# run_button.click(fn=process, inputs=ips, outputs=[result_gallery])
|
249 |
-
|
250 |
-
# block.launch(debug = True)
|
251 |
-
|
252 |
-
# block = gr.Blocks().queue()
|
253 |
|
254 |
gr.Markdown("## BRIA RMBG 1.4")
|
255 |
gr.HTML('''
|
@@ -263,8 +233,6 @@ description = r"""Background removal model developed by <a href='https://BRIA.AI
|
|
263 |
For test upload your image and wait. Read more at model card <a href='https://huggingface.co/briaai/RMBG-1.4' target='_blank'><b>briaai/RMBG-1.4</b></a>.<br>
|
264 |
"""
|
265 |
examples = [['./input.jpg'],]
|
266 |
-
# output = ImageSlider(position=0.5,label='Image without background', type="pil", show_download_button=True)
|
267 |
-
# demo = gr.Interface(fn=process,inputs="image", outputs=output, examples=examples, title=title, description=description)
|
268 |
demo = gr.Interface(fn=process,inputs="image", outputs="image", examples=examples, title=title, description=description)
|
269 |
|
270 |
if __name__ == "__main__":
|
|
|
94 |
return outputs
|
95 |
return list(self.chunks(outputs, samples))
|
96 |
|
97 |
+
## Initialize TANGO
|
98 |
+
#
|
99 |
+
#tango = Tango(device = "cpu")
|
100 |
+
#tango.vae.to(device_type)
|
101 |
+
#tango.stft.to(device_type)
|
102 |
+
#tango.model.to(device_type)
|
103 |
+
#
|
104 |
+
#def update_seed(is_randomize_seed, seed):
|
105 |
+
# if is_randomize_seed:
|
106 |
+
# return random.randint(0, max_64_bit_int)
|
107 |
+
# return seed
|
108 |
+
#
|
109 |
+
#def check(
|
110 |
+
# prompt,
|
111 |
+
# output_number,
|
112 |
+
# steps,
|
113 |
+
# guidance,
|
114 |
+
# is_randomize_seed,
|
115 |
+
# seed
|
116 |
+
#):
|
117 |
+
# if prompt is None or prompt == "":
|
118 |
+
# raise gr.Error("Please provide a prompt input.")
|
119 |
+
# if not output_number in [1, 2, 3]:
|
120 |
+
# raise gr.Error("Please ask for 1, 2 or 3 output files.")
|
121 |
+
#
|
122 |
+
#def update_output(output_format, output_number):
|
123 |
+
# return [
|
124 |
+
# gr.update(format = output_format),
|
125 |
+
# gr.update(format = output_format, visible = (2 <= output_number)),
|
126 |
+
# gr.update(format = output_format, visible = (output_number == 3)),
|
127 |
+
# gr.update(visible = False)
|
128 |
+
# ]
|
129 |
+
#
|
130 |
+
#def text2audio(
|
131 |
+
# prompt,
|
132 |
+
# output_number,
|
133 |
+
# steps,
|
134 |
+
# guidance,
|
135 |
+
# is_randomize_seed,
|
136 |
+
# seed
|
137 |
+
#):
|
138 |
+
# start = time.time()
|
139 |
+
#
|
140 |
+
# if seed is None:
|
141 |
+
# seed = random.randint(0, max_64_bit_int)
|
142 |
+
#
|
143 |
+
# random.seed(seed)
|
144 |
+
# torch.manual_seed(seed)
|
145 |
+
#
|
146 |
+
# output_wave = tango.generate(prompt, steps, guidance, output_number)
|
147 |
+
#
|
148 |
+
# output_wave_1 = gr.make_waveform((16000, output_wave[0]))
|
149 |
+
# output_wave_2 = gr.make_waveform((16000, output_wave[1])) if (2 <= output_number) else None
|
150 |
+
# output_wave_3 = gr.make_waveform((16000, output_wave[2])) if (output_number == 3) else None
|
151 |
+
#
|
152 |
+
# end = time.time()
|
153 |
+
# secondes = int(end - start)
|
154 |
+
# minutes = secondes // 60
|
155 |
+
# secondes = secondes - (minutes * 60)
|
156 |
+
# hours = minutes // 60
|
157 |
+
# minutes = minutes - (hours * 60)
|
158 |
+
# return [
|
159 |
+
# output_wave_1,
|
160 |
+
# output_wave_2,
|
161 |
+
# output_wave_3,
|
162 |
+
# gr.update(visible = True, value = "Start again to get a different result. The output have been generated in " + ((str(hours) + " h, ") if hours != 0 else "") + ((str(minutes) + " min, ") if hours != 0 or minutes != 0 else "") + str(secondes) + " sec.")
|
163 |
+
# ]
|
164 |
+
#
|
165 |
+
#if is_space_imported:
|
166 |
+
# text2audio = spaces.GPU(text2audio, duration = 420)
|
167 |
|
168 |
# Old code
|
169 |
net=BriaRMBG()
|
|
|
|
|
170 |
model_path = hf_hub_download("cocktailpeanut/gbmr", 'model.pth')
|
171 |
if torch.cuda.is_available():
|
172 |
net.load_state_dict(torch.load(model_path))
|
|
|
218 |
# paste the mask on the original image
|
219 |
new_im = Image.new("RGBA", pil_im.size, (0,0,0,0))
|
220 |
new_im.paste(orig_image, mask=pil_im)
|
|
|
221 |
|
222 |
return new_im
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
223 |
|
224 |
gr.Markdown("## BRIA RMBG 1.4")
|
225 |
gr.HTML('''
|
|
|
233 |
For test upload your image and wait. Read more at model card <a href='https://huggingface.co/briaai/RMBG-1.4' target='_blank'><b>briaai/RMBG-1.4</b></a>.<br>
|
234 |
"""
|
235 |
examples = [['./input.jpg'],]
|
|
|
|
|
236 |
demo = gr.Interface(fn=process,inputs="image", outputs="image", examples=examples, title=title, description=description)
|
237 |
|
238 |
if __name__ == "__main__":
|