[Feature] Conversation Summary Mode: Track Assistant Responses

.env.template

@@ -13,3 +13,8 @@ OPENAI_AZURE_DEPLOYMENT_ID=deployment-id-for-azure
 IMAGE_PROVIDER=dalle
 HUGGINGFACE_API_TOKEN=
 USE_MAC_OS_TTS=False
+
+# Step Summary Prompts
+STEP_SUMMARIZATION_PROMPT=outline the key steps and insights you plan to take and gain during the completion of your task in a concise and relevant manner
+FINAL_SUMMARIZATION_PROMPT=Consolidate and distill the essential information from the provided summaries of individual steps to create a concise and coherent summary of the overall process by creating an output structured using chapters, sub-chapters and bullet points. Don't include duplicate ideas. Leave only one instance if it's present in multiple steps:{}
+# {} is replaced with the summary chunk content, make sure to include it in your FINAL_SUMMARIZATION_PROMPT

README.md

@@ -43,6 +43,7 @@ Your support is greatly appreciated
   - [💀 Continuous Mode ⚠️](#-continuous-mode-️)
   - [GPT3.5 ONLY Mode](#gpt35-only-mode)
   - [🖼 Image Generation](#image-generation)
+  - [📜 Conversation Summary Mode](#-conversation-summary-mode)
   - [⚠️ Limitations](#️-limitations)
   - [🛡 Disclaimer](#-disclaimer)
   - [🐦 Connect with Us on Twitter](#-connect-with-us-on-twitter)
@@ -238,6 +239,18 @@ IMAGE_PROVIDER=sd
 HUGGINGFACE_API_TOKEN="YOUR_HUGGINGFACE_API_TOKEN"
 ```
 
+## 📜 Conversation Summary Mode
+Conversation Summary mode is designed to provide users with concise summaries of the AI's responses during a conversation. 
+This mode can be especially useful for long conversations where it's crucial to quickly understand the key points without reading through the entire response. 
+It enhances readability and helps users grasp important information more efficiently.
+
+How to Use Conversation Summary Mode:
+
+1. Define `STEP_SUMMARIZATION_PROMPT` and `FINAL_SUMMARIZATION_PROMPT` in the .env file (you can play around with these prompts to get the best results)
+2. Run: `python scripts/main.py --conversation-summary`
+3. View step summaries in the console and a summary file inside `scripts/logs` folder
+4. Upon exit, a final summary is printed to console and appended to the file
+
 ## ⚠️ Limitations
 This experiment aims to showcase the potential of GPT-4 but comes with some limitations:
 

scripts/config.py

@@ -36,6 +36,7 @@ def __init__(self):
         self.debug_mode = False
         self.continuous_mode = False
         self.speak_mode = False
+        self.conversation_summary_mode = False
 
         self.fast_llm_model = os.getenv("FAST_LLM_MODEL", "gpt-3.5-turbo")
         self.smart_llm_model = os.getenv("SMART_LLM_MODEL", "gpt-4")
@@ -67,6 +68,20 @@ def __init__(self):
         self.image_provider = os.getenv("IMAGE_PROVIDER")
         self.huggingface_api_token = os.getenv("HUGGINGFACE_API_TOKEN")
 
+        # Conversation Summary Mode - Prompts to use when summarizing
+        self.step_summarization_prompt = os.getenv(
+            "STEP_SUMMARIZATION_PROMPT",
+            "outline the key steps and insights you plan to take "
+            "and gain during the completion of your task in a concise and relevant manner"
+        )
+        self.final_summarization_prompt = os.getenv(
+            "FINAL_SUMMARIZATION_PROMPT",
+            "Consolidate and distill the essential information from the provided summaries of individual steps to "
+            "create a concise and coherent summary of the overall process by creating an output structured using "
+            "chapters, sub-chapters and bullet points. Don't include duplicate ideas. Leave only one instance if it's "
+            "present in multiple steps:{}"
+        )  # {} is replaced with the content
+
         # User agent headers to use when browsing web
         # Some websites might just completely deny request with an error code if no user agent was found.
         self.user_agent_header = {"User-Agent":"Mozilla/5.0 (Macintosh; Intel Mac OS X 10_15_4) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/83.0.4103.97 Safari/537.36"}
@@ -89,6 +104,9 @@ def set_speak_mode(self, value: bool):
         """Set the speak mode value."""
         self.speak_mode = value
 
+    def set_conversation_summary_mode(self, value: bool):
+        self.conversation_summary_mode = value
+
     def set_fast_llm_model(self, value: str):
         """Set the fast LLM model value."""
         self.fast_llm_model = value