Update app.py

app.py

@@ -4,11 +4,10 @@ warnings.filterwarnings('ignore')
 
 import subprocess, io, os, sys, time
 
-# os.system("pip install -q gradio==3.48.0")
 os.system("pip install -q gradio")
 os.system("pip install -q diffusers")
 os.system("pip install -q segment_anything")
-os.environ["CUDA_VISIBLE_DEVICES"] = "0"
+# os.environ["CUDA_VISIBLE_DEVICES"] = "0"
 
 result = subprocess.run(['pip', 'install', '-e', 'GroundingDINO'], check=True)
 print(f'pip install GroundingDINO = {result}')
@@ -84,7 +83,7 @@ class CFG:
         "vit_l": "https://dl.fbaipublicfiles.com/segment_anything/sam_vit_l_0b3195.pth",
         "vit_b": "https://dl.fbaipublicfiles.com/segment_anything/sam_vit_b_01ec64.pth"
     }
-    device = 'cuda'
+    device = 'cuda' if cuda.is_available() else 'cpu'
     ckpt_repo_id = "ShilongLiu/GroundingDINO"
     ckpt_filename = "groundingdino_swinb_cogcoor.pth"
     ckpt_config_filename = "GroundingDINO_SwinB.cfg.py"
@@ -251,147 +250,6 @@ def main_fun(image_pil, x_units, y_units, text_prompt):
 
 import gradio as gr
 
-# with gr.Blocks() as demo:
-#     gr.Markdown("Segmentation and shift")
-#     with gr.Tab("Txt2Img"):
-#         with gr.Row():
-#             with gr.Column(scale=1):
-# #                 image_pil, x_units, y_units, text_prompt
-
-#                 text_prompt = gr.Textbox(lines=1, label="Prompt")
-#                 image_pil = gr.Image(type="filepath", label='Image')
-# #                 negative_prompt = gr.Textbox(lines=1, label="Negative Prompt")
-# #                 width = gr.Dropdown(choices=number_choices, value=704, label="Width")
-# #                 height = gr.Dropdown(choices=number_choicess, value=408, label="Height")
-#                 x_units = gr.Slider(minimum=50, maximum=300, step=1, value=10, label="x_units")
-#                 y_units = gr.Slider(minimum=30, maximum=300, step=0.1, value=5, label="y_units")
-# #                 seed = gr.Textbox(label="Seed (Leave empty for random seed)")
-#             with gr.Column(scale=2):
-#                 output = gr.Image(type="pil", label="Output Image")
-# #                 generate = gr.Button("Generate")
-# #                 output_seed = gr.Textbox(label="Current Seed")
-
-#     # Create the txt2img function
-    
-# #     generate.click(fn=main_fun, inputs=[image_pil, x_units, y_units, text_prompt], outputs=[output_image])
-
-# x_units = gr.Slider(minimum=50, maximum=300, step=1, value=10, label="x_units")
-# y_units = gr.Slider(minimum=30, maximum=300, step=0.1, value=5, label="y_units")
-# text_prompt = gr.Textbox(lines=1, label="Prompt")
-# image_pil = gr.components.Image(type="filepath", label="Input Image")
-# output = gr.components.Image(type="numpy", label="Output Image")
-
-# # Launch the Gradio UI
-# gr.Interface(fn=main_fun, inputs=[image_pil, x_units, y_units, text_prompt], outputs=[output]).launch(share=True, debug=True)
-
-
-# block = gr.Blocks().queue()
-#     with block:
-#         with gr.Row():
-#             with gr.Column():
-#                 # task_types = ["detection"]
-#                 # if sam_enable:
-#                 #     task_types.append("segment")
-#                 # if inpainting_enable:
-#                 #     task_types.append("inpainting")
-#                 #     task_types.append("outpainting")
-#                 # if lama_cleaner_enable:
-#                 #     task_types.append("remove")
-#                 # if ram_enable:
-#                 #     task_types.append("relate anything")
-#                 # if kosmos_enable:
-#                 #     task_types.append("Kosmos-2")           
-         
-#                 input_image = gr.Image(source='upload', elem_id="image_upload", tool='sketch', type='pil', label="Upload",
-#                                     height=512, brush_color='#00FFFF', mask_opacity=0.6)    
-#                 # task_type = gr.Radio(task_types,  value="detection", 
-#                 #                                 label='Task type', visible=True) 
-#                 # mask_source_radio = gr.Radio([mask_source_draw, mask_source_segment], 
-#                 #                     value=mask_source_segment, label="Mask from",
-#                 #                     visible=False) 
-#                 x_units = gr.Slider(minimum=50, maximum=300, step=1, value=10, label="x_units")
-#                 y_units = gr.Slider(minimum=30, maximum=300, step=0.1, value=5, label="y_units")
-#                 text_prompt = gr.Textbox(lines=1, label="Prompt")
-#                 # image_pil = gr.components.Image(type="filepath", label="Input Image")
-#                 # output = gr.components.Image(type="numpy", label="Output Image")
-                
-#                 run_button = gr.Button(label="Run", visible=True)
-#                 # with gr.Accordion("Advanced options", open=False) as advanced_options:
-#                 #     box_threshold = gr.Slider(
-#                 #         label="Box Threshold", minimum=0.0, maximum=1.0, value=0.3, step=0.001
-#                 #     )
-#                 #     text_threshold = gr.Slider(
-#                 #         label="Text Threshold", minimum=0.0, maximum=1.0, value=0.25, step=0.001
-#                 #     )
-#                 #     iou_threshold = gr.Slider(
-#                 #         label="IOU Threshold", minimum=0.0, maximum=1.0, value=0.8, step=0.001
-#                 #     )                    
-#                 #     inpaint_mode = gr.Radio(["merge", "first"], value="merge", label="inpaint_mode")
-#                 #     with gr.Row():
-#                 #         with gr.Column(scale=1):
-#                 #             remove_mode = gr.Radio(["segment", "rectangle"],  value="segment", label='remove mode') 
-#                 #         with gr.Column(scale=1):
-#                 #             remove_mask_extend = gr.Textbox(label="remove_mask_extend", value='10')
-
-#             with gr.Column():
-#                 image_gallery = gr.Gallery(label="result images", show_label=True, elem_id="gallery", height=512, visible=True
-#                     ).style(preview=True, columns=[5], object_fit="scale-down", height="auto")   
-#                 # time_cost = gr.Textbox(label="Time cost by step (ms):", visible=False, interactive=False)
-
-#                 # kosmos_output = gr.Image(type="pil", label="result images", visible=False)
-#                 # kosmos_text_output = gr.HighlightedText(
-#                 #                     label="Generated Description",
-#                 #                     combine_adjacent=False,
-#                 #                     show_legend=True,
-#                 #                     visible=False,
-#                 #                 ).style(color_map=color_map)
-#                 # # record which text span (label) is selected
-#                 # selected = gr.Number(-1, show_label=False, placeholder="Selected", visible=False)
-
-#                 # # record the current `entities`
-#                 # entity_output = gr.Textbox(visible=False)
-
-#                 # # get the current selected span label
-#                 # def get_text_span_label(evt: gr.SelectData):
-#                 #     if evt.value[-1] is None:
-#                 #         return -1
-#                 #     return int(evt.value[-1])
-#                 # # and set this information to `selected`
-#                 # kosmos_text_output.select(get_text_span_label, None, selected)
-                
-#                 # update output image when we change the span (enity) selection
-#                 # def update_output_image(img_input, image_output, entities, idx):
-#                 #     entities = ast.literal_eval(entities)
-#                 #     updated_image = draw_entity_boxes_on_image(img_input, entities, entity_index=idx)
-#                 #     return updated_image
-#                 # selected.change(update_output_image, [kosmos_output, kosmos_output, entity_output, selected], [kosmos_output])
-
-#             run_button.click(fn=main_fun, inputs=[image_pil, x_units, y_units, text_prompt], outputs=[image_gallery], show_progress=True, queue=True)
-            
-#             # mask_source_radio.change(fn=change_radio_display, inputs=[task_type, mask_source_radio], 
-#             #                 outputs=[text_prompt, inpaint_prompt, mask_source_radio, num_relation])
-#             # task_type.change(fn=change_radio_display, inputs=[task_type, mask_source_radio], 
-#             #                 outputs=[text_prompt, inpaint_prompt, mask_source_radio, num_relation,
-#             #                 image_gallery, kosmos_input, kosmos_output, kosmos_text_output
-#             #                 ])
-
-#         # DESCRIPTION = f'### This demo from [Grounded-Segment-Anything](https://github.com/IDEA-Research/Grounded-Segment-Anything). <br>'
-#         # if lama_cleaner_enable:
-#         #     DESCRIPTION += f'Remove(cleaner) from [lama-cleaner](https://github.com/Sanster/lama-cleaner). <br>'
-#         # if kosmos_enable:
-#         #     DESCRIPTION += f'Kosmos-2 from [Kosmos-2](https://github.com/microsoft/unilm/tree/master/kosmos-2). <br>'
-#         # if ram_enable:
-#         #     DESCRIPTION += f'RAM from [RelateAnything](https://github.com/Luodian/RelateAnything). <br>'
-#         # DESCRIPTION += f'Thanks for their excellent work.'
-#         # DESCRIPTION += f'<p>For faster inference without waiting in queue, you may duplicate the space and upgrade to GPU in settings. \
-#         #                 <a href="https://huggingface.co/spaces/yizhangliu/Grounded-Segment-Anything?duplicate=true"><img style="display: inline; margin-top: 0em; margin-bottom: 0em" src="https://bit.ly/3gLdBN6" alt="Duplicate Space" /></a></p>'
-#         # gr.Markdown(DESCRIPTION)
-
-#     # print(f'device = {device}')
-#     # print(f'torch.cuda.is_available = {torch.cuda.is_available()}')
-#     # computer_info()
-#     block.launch(server_name='0.0.0.0')
-
 
 image_blocks = gr.Blocks()
 with image_blocks as demo: