Update app.py

app.py

@@ -3,19 +3,23 @@ import base64
 import requests
 import tempfile
 import os
+import openai
+
+prompt = "Type and press Enter"
 
 def encode_image(image_path):
     with open(image_path, "rb") as image_file:
         return base64.b64encode(image_file.read()).decode('utf-8')
 
-def upload_img(image,api_key,img_ques):
-
+def api_calling(image, prompt, api_key):
     base64_image = encode_image(image)
+    if len(prompt) == 0:
+        prompt = "Provide a brief description of the image."
     
     
     headers = {
         "Content-Type": "application/json",
-        "Authorization": f"Bearer {api_key}" 
+        "Authorization": f"Bearer {api_key}"
     }
     payload = {
         "model": "gpt-4-vision-preview",
@@ -25,7 +29,7 @@ def upload_img(image,api_key,img_ques):
                 "content": [
                     {
                         "type": "text",
-                        "text": img_ques
+                        "text": prompt
                     },
                     {
                         "type": "image_url",
@@ -36,31 +40,40 @@ def upload_img(image,api_key,img_ques):
                 ]
             }
         ],
-        "max_tokens": 300
+        "max_tokens": 500
     }
     response = requests.post("https://api.openai.com/v1/chat/completions", headers=headers, json=payload)
     img_preview = response.json()
     return img_preview["choices"][0]["message"]["content"]
 
-title = """<h1 align="center">GPT-4 Image Preview</h1>"""
-
-with gr.Blocks() as demo:
-    gr.Markdown(title)
+def message_and_history(img, input, history, api_key):
+    history = history or []
+    s = list(sum(history, ()))
+    s.append(input)
+    inp = '  '.join(s)
+    output = api_calling(img,inp, api_key)
+    if len(input) == 0:
+        input = "Brief description of the image."
+        history.append((input, output))
+    else:
+        history.append((input, output))
+    return history, history
 
+block = gr.Blocks(theme=gr.themes.Monochrome())
+with block:
+    gr.Markdown("""<h1><center>Visual Explorer - Ask and Learn about Images</center></h1> """)
     with gr.Row():
         with gr.Column(scale=0.5):
-            image_input = gr.Image(type="filepath", label="Upload Image")
-            key_input = gr.Textbox(label="Enter API-key")
+            img_input = gr.Image(type="filepath", label="Upload Image")
+            api_input = gr.Textbox(label= "Enter Api-key")
             upload_button = gr.Button(value="Upload & Start Chat", interactive=True, variant="primary")
-            btn_clear = gr.ClearButton([image_input,key_input])
-        
         with gr.Column():
-            text = gr.Textbox(label="Output")
-            ques_input = gr.Textbox(label="Enter the question")
-            ques_clear = gr.ClearButton([ques_input])
-
-    upload_button.click(upload_img, inputs=[image_input,key_input,ques_input], outputs=text)
-    btn_clear.add(text)
-    ques_clear.add(text)
-
-demo.launch(share=True)
+            chatbot = gr.Chatbot(label="Chat with Image")
+            message = gr.Textbox(label="User", placeholder=prompt)
+            btn_clear = gr.ClearButton([message])
+            state = gr.State()
+            
+    upload_button.click(message_and_history, inputs=[img_input, message, state, api_input], outputs=[chatbot, state])
+    message.submit(message_and_history, inputs=[img_input, message, state, api_input], outputs=[chatbot, state])
+    btn_clear.add(message)
+block.launch(share = True)