removed the message['text'] from the input_prompt as there is no multimodal

app.py

@@ -96,7 +96,7 @@ def multimodal_and_generation(message, history):
                 image_path = hist[0][0]
 
     if image_path is None:
-        input_prompt = message["text"]
+        input_prompt = message
         client = openai.OpenAI(api_key=API_KEY)
         stream = client.chat.completions.create(
             model="gpt-3.5-turbo",
@@ -198,12 +198,12 @@ def bot_comms(message, history):
     #     # logger.debug(f"\nimagery mode: {mode}\n")
     #     return "Imagery On! Type your prompt to make the image 🖼️"
     
-    if message["text"] == "chatting":
-        logger.debug("Switching to chatting mode.")
-        # mode_manager.set_mode("chatting")
-        mode += "chatting"
-        # logger.debug(f"\nchatting mode: {mode}\n")
-        return "Imagery Off. Ask me any questions. ☄️"
+    # if message["text"] == "chatting":
+    #     logger.debug("Switching to chatting mode.")
+    #     # mode_manager.set_mode("chatting")
+    #     mode += "chatting"
+    #     # logger.debug(f"\nchatting mode: {mode}\n")
+    #     return "Imagery Off. Ask me any questions. ☄️"
     
     # if mode == "imagery":
     #     logger.debug("Processing imagery prompt.")
@@ -216,21 +216,21 @@ def bot_comms(message, history):
         
     buffer = ""
     gpt_outputs = []
-    if mode == "chatting" or mode == "":
-        # logger.debug("On chatting or no mode.\n\n")
-        stream = multimodal_and_generation(message, history)
-        mode += "chatting"
-        for chunk in stream:
-            if chunk is not None and hasattr(chunk.choices[0].delta, "content"):
-                logger.debug("\n\nFound the chunk in stream for gpt-3.5\n\n")
-                text = chunk.choices[0].delta.content
-                if text:
-                    gpt_outputs.append(text)
-                    buffer += text
-                yield "".join(gpt_outputs)
+    # if mode == "chatting" or mode == "":
+    # logger.debug("On chatting or no mode.\n\n")
+    stream = multimodal_and_generation(message, history)
+    mode += "chatting"
+    for chunk in stream:
+        if chunk is not None and hasattr(chunk.choices[0].delta, "content"):
+            logger.debug("\n\nFound the chunk in stream for gpt-3.5\n\n")
+            text = chunk.choices[0].delta.content
+            if text:
+                gpt_outputs.append(text)
+                buffer += text
+            yield "".join(gpt_outputs)
 
 chatbot = gr.Chatbot(height=600, label="Chimera AI")
-chat_input = gr.MultimodalTextbox(interactive=True, file_types=["images"], placeholder="Enter your question or upload an image.", show_label=False)
+# chat_input = gr.MultimodalTextbox(interactive=True, file_types=["images"], placeholder="Enter your question or upload an image.", show_label=False)
 
 # with gr.Blocks(fill_height=True) as demo:
 #     gr.Markdown(DESCRIPTION)