create app.py

app.py

@@ -0,0 +1,76 @@
+import requests
+from PIL import Image
+import gradio as gr
+
+def upload_button_config():
+    return gr.update(visible=False)
+
+def upload_textbox_config(text_in):
+    return gr.update(visible=True)
+
+#takes input and generates the Response
+def predict(btn_upload, counter,image_hid, input, history):
+    
+    if counter == 0:
+      image_in = Image.open(btn_upload)
+      #Resizing the image
+      basewidth = 512
+      wpercent = (basewidth/float(image_in.size[0]))
+      hsize = int((float(image_in.size[1])*float(wpercent)))
+      image_in = image_in.resize((basewidth,hsize)) #, Image.Resampling.LANCZOS)
+      # Save the image to the file-like object
+      #seed = random.randint(0, 1000000)
+      img_name = "uploaded_image.png" #f"./edited_image_{seed}.png"
+      image_in.save(img_name)
+      #add state
+      history = history or []
+      response = '<img src="/file=' + img_name + '">'
+      history.append((input, response))
+      counter += 1
+      return history, history, img_name, counter, image_in
+
+    #process the input prompt and image
+    #image = Image.open(btn_upload)
+    print(f"prompt is :{input}") #Question: Is this photo unusual? Answer:
+    input = f"Question: {input} Answer:"
+    inputs = processor(image_hid, text=input, return_tensors="pt").to(device, torch.float16)
+    
+    #generte the response
+    generated_ids = model.generate(**inputs, max_new_tokens=10)
+    generated_text = processor.batch_decode(generated_ids, skip_special_tokens=True)[0].strip()
+    print(f"generated_text is : {generated_text}")
+
+    #add state
+    history = history or []
+    response = generated_text #'<img src="/file=' + img_name + '">'
+    history.append((input, response))
+    counter += 1
+    return history, history, "uploaded_image.png", counter, image_hid
+
+#Blocks Layout
+with gr.Blocks() as demo:
+  with gr.Column():
+    #text_in = gr.Textbox(value='', placeholder="Type your questions here and press enter", elem_id = "input_prompt", visible=False, label='Great! Now you can ask questions to get more information about the image')
+    btn_upload = gr.UploadButton("Upload image!", file_types=["image"], file_count="single", elem_id="upload_button")
+    chatbot = gr.Chatbot(elem_id = 'chatbot-component', label='Conversational with Images')
+    text_in = gr.Textbox(value='', placeholder="Type your questions here and press enter", elem_id = "input_prompt", visible=False, label='Great! Now you can ask questions to get more information about the image')
+    state_in = gr.State()
+    counter_out = gr.Number(visible=False, value=0, precision=0)
+    text_out = gr.Textbox(visible=False)  #getting imag name out
+    image_hid = gr.Image() #visible=False) #, type='pil')
+
+  #Using Event Listeners
+  btn_upload.upload(predict, [btn_upload, counter_out, image_hid, text_in, state_in], [chatbot, state_in, text_out, counter_out, image_hid])
+  btn_upload.upload(fn = upload_textbox_config, inputs=text_in, outputs = text_in)
+
+  text_in.submit(predict, [btn_upload, counter_out, image_hid, text_in, state_in], [chatbot, state_in, text_out, counter_out, image_hid])
+  #text_in.submit(previous, [image_hid], [image_oneup])
+  
+  chatbot.change(fn = upload_button_config, outputs=btn_upload) #, scroll_to_output = True)
+  #text_in.submit(None, [], [], _js = "() => document.getElementById('#chatbot-component').scrollTop = document.getElementById('#chatbot-component').scrollHeight")
+
+  #with gr.Accordion("Release Notes", open=False):
+  #gr.Markdown(help_text)
+    
+demo.queue(concurrency_count=10)
+demo.launch(debug=True) #, width="80%", height=2000)