tldr

app.py

@@ -15,7 +15,7 @@ from chat_chains import (
     parse_context_and_question,
     ai_response_format,
 )
-from autoqa_chains import auto_qa_chain
+from autoqa_chain import auto_qa_chain
 from chain_of_density import chain_of_density_chain
 from insights_bullet_chain import insights_bullet_chain
 from insights_mind_map_chain import insights_mind_map_chain
@@ -23,6 +23,7 @@ from synopsis_chain import synopsis_chain
 from custom_exceptions import InvalidArgumentError, InvalidCommandError
 from openai_configuration import openai_parser
 from summary_chain import summary_chain
+from tldr_chain import tldr_chain
 
 st.set_page_config(layout="wide")
 
@@ -48,6 +49,7 @@ Here's a quick guide to getting started with me:
 | `/paper-synopsis <list of snippet ids>` | Generate a synopsis of the paper. |
 | `/deep-dive [<list of snippet ids>] <query>` | Query me with a specific context. |
 | `/summarise-section [<list of snippet ids>] <section name>` | Summarize a specific section of the paper. |
+| `/tldr <list of snippet ids>` | Generate a tldr summary of the paper. |
 
 
 <br>
@@ -249,6 +251,25 @@ def synopsis_wrapper(inputs):
     return (summary, "identity")
 
 
+def tldr_wrapper(inputs):
+    if inputs == []:
+        raise InvalidArgumentError("Please provide snippet ids")
+    document = "\n\n".join([st.session_state.documents[c].page_content for c in inputs])
+    llm = ChatOpenAI(model=st.session_state.model, temperature=0)
+    with get_openai_callback() as cb:
+        summary = tldr_chain(llm).invoke({"paper": document})
+        stats = cb
+    st.session_state.messages.append(("/tldr", summary, "identity"))
+    st.session_state.costing.append(
+        {
+            "prompt tokens": stats.prompt_tokens,
+            "completion tokens": stats.completion_tokens,
+            "cost": stats.total_cost,
+        }
+    )
+    return (summary, "identity")
+
+
 def insights_bullet_wrapper(inputs):
     if inputs == []:
         raise InvalidArgumentError("Please provide snippet ids")
@@ -381,6 +402,7 @@ if __name__ == "__main__":
         ("/insight-mind-map", list, insights_mind_map_wrapper),
         ("/paper-synopsis", list, synopsis_wrapper),
         ("/summarise-section", str, summarise_wrapper),
+        ("/tldr", list, tldr_wrapper),
     ]
     command_center = CommandCenter(
         default_input_type=str,

autoqa_chain.py

@@ -0,0 +1,15 @@
+from langchain_core.output_parsers import JsonOutputParser
+from langchain_core.prompts import PromptTemplate
+
+qa_prompt_template = """
+Create a mind map of questions (based on the given abstract) that will help understand a machine learning research paper.
+Ensure that the outline is structured in the following JSON array for clarity, such that each section should have two keys: "section_name" and "questions"
+
+Here is the research paper abstract: ####{paper}####
+"""
+
+qa_prompt = PromptTemplate(
+    template=qa_prompt_template,
+    input_variables=["paper"],
+)
+auto_qa_chain = lambda model: qa_prompt | model | JsonOutputParser()

tldr_chain.py

@@ -0,0 +1,26 @@
+from langchain_core.prompts import PromptTemplate
+from langchain_core.output_parsers import StrOutputParser
+
+tldr_prompt_template = """
+Create a mind map of the given research paper along the given lines:
+1. Background: A brief overview of what's being studied in what context
+2. Justification: Why the researchers conducted the study
+3. Method: How the researchers arrived at the result
+4. Major Findings: The main findings of the study
+5. Key Results: More details about the results of the study
+6. Conclusion: Significance of the findings and what they mean for future research
+
+The above sections may differ from paper to paper, hence you may need to adjust the structure accordingly by dropping / merging one or more sections.
+
+Here is the research paper abstract: ####{paper}####
+
+Ensure that the outline is structured in Markdown format for clarity, facilitating its integration into documents or presentations.
+"""
+
+
+tldr_output_parser = StrOutputParser()
+tldr_prompt = PromptTemplate(
+    template=tldr_prompt_template,
+    input_variables=["paper"],
+)
+tldr_chain = lambda model: tldr_prompt | model | tldr_output_parser