Update app.py

app.py

@@ -16,44 +16,119 @@ openai.api_key = st.secrets["openai_api_key"]
 # File uploader for images
 uploaded_file = st.file_uploader("Upload an image", type=["jpg", "png", "jpeg"])
 
-if uploaded_file:
-    with st.expander("Image", expanded=True):
-        st.image(uploaded_file, caption=uploaded_file.name, use_column_width=True)
+# if uploaded_file:
+#     with st.expander("Image", expanded=True):
+#         st.image(uploaded_file, caption=uploaded_file.name, use_column_width=True)
 
-    # Toggle for additional details input
-    show_details = st.checkbox("Add details about the image")
+#     # Toggle for additional details input
+#     show_details = st.checkbox("Add details about the image")
 
-    if show_details:
-        additional_details = st.text_area("Add any additional details or context about the image here:")
+#     if show_details:
+#         additional_details = st.text_area("Add any additional details or context about the image here:")
 
-    # Button to trigger the analysis
-    analyze_button = st.button("Analyse the MTSS Image")
+#     # Button to trigger the analysis
+#     analyze_button = st.button("Analyse the MTSS Image")
 
-    if analyze_button:
-        with st.spinner("Analyzing the image..."):
-            base64_image = encode_image(uploaded_file)
-            prompt_text = (
-                "You are a highly knowledgeable accessibility expert. "
+#     if analyze_button:
+#         with st.spinner("Analyzing the image..."):
+#             base64_image = encode_image(uploaded_file)
+#             prompt_text = (
+#                 "You are a highly knowledgeable accessibility expert. "
+#                 "Your task is to examine the following image in detail. "
+#                 "Provide a comprehensive, factual, and accurate explanation of what the image depicts. "
+#                 "Highlight key elements and their significance, and present your analysis in clear, well-structured format. "
+#                 "Create a detailed image caption in explaining in 150 words or less."
+#             )
+
+#             if show_details and additional_details:
+#                 prompt_text += f"\n\nAdditional Context Provided by the User:\n{additional_details}"
+
+#             # Define the messages payload
+#             messages = [{
+#                 "role": "user",
+#                 "content": [{"type": "text", "text": prompt_text}, {"type": "image_url", "image_url": f"data:image/jpeg;base64,{base64_image}"}]
+#             }]
+
+#             try:
+#                 # Make the request to OpenAI and handle streaming if required
+#                 response = openai.chat.completions.create(model="gpt-4-vision-preview", messages=messages, max_tokens=150)
+#                 st.write(response.choices[0].message.content)
+#             except Exception as e:
+#                 st.error(f"An error occurred: {e}")
+# else:
+#     st.warning("Please upload an image.")
+
+
+
+
+
+
+
+
+# Check if an image has been uploaded, if the API key is available, and if the button has been pressed
+if uploaded_file is not None and api_key and analyze_button:
+
+    with st.spinner("Analysing the image ..."):
+        # Encode the image
+        base64_image = encode_image(uploaded_file)
+    
+        # Optimized prompt for additional clarity and detail
+        prompt_text = (
+            "You are a highly knowledgeable accessibility expert. "
                 "Your task is to examine the following image in detail. "
                 "Provide a comprehensive, factual, and accurate explanation of what the image depicts. "
                 "Highlight key elements and their significance, and present your analysis in clear, well-structured format. "
                 "Create a detailed image caption in explaining in 150 words or less."
+        )
+    
+        if show_details and additional_details:
+            prompt_text += (
+                f"\n\nAdditional Context Provided by the User:\n{additional_details}"
             )
-
-            if show_details and additional_details:
-                prompt_text += f"\n\nAdditional Context Provided by the User:\n{additional_details}"
-
-            # Define the messages payload
-            messages = [{
+    
+        # Create the payload for the completion request
+        messages = [
+            {
                 "role": "user",
-                "content": [{"type": "text", "text": prompt_text}, {"type": "image_url", "image_url": f"data:image/jpeg;base64,{base64_image}"}]
-            }]
-
-            try:
-                # Make the request to OpenAI and handle streaming if required
-                response = openai.chat.completions.create(model="gpt-4-vision-preview", messages=messages, max_tokens=150)
-                st.write(response.choices[0].message.content)
-            except Exception as e:
-                st.error(f"An error occurred: {e}")
+                "content": [
+                    {"type": "text", "text": prompt_text},
+                    {
+                        "type": "image_url",
+                        "image_url": f"data:image/jpeg;base64,{base64_image}",
+                    },
+                ],
+            }
+        ]
+    
+        # Make the request to the OpenAI API
+        try:
+            # Without Stream
+            
+            # response = client.chat.completions.create(
+            #     model="gpt-4-vision-preview", messages=messages, max_tokens=500, stream=False
+            # )
+    
+            # Stream the response
+            full_response = ""
+            message_placeholder = st.empty()
+            for completion in openai.chat.completions.create(
+                model="gpt-4-vision-preview", messages=messages, 
+                max_tokens=150, stream=True
+            ):
+                # Check if there is content to display
+                if completion.choices[0].delta.content is not None:
+                    full_response += completion.choices[0].delta.content
+                    message_placeholder.markdown(full_response + "▌")
+            # Final update to placeholder after the stream ends
+            message_placeholder.markdown(full_response)
+    
+            # Display the response in the app
+            # st.write(response.choices[0].message.content)
+        except Exception as e:
+            st.error(f"An error occurred: {e}")
 else:
-    st.warning("Please upload an image.")
+    # Warnings for user action required
+    if not uploaded_file and analyze_button:
+        st.warning("Please upload an image.")
+    if not api_key:
+        st.warning("Please enter your OpenAI API key.")