| from langchain_core.prompts import PromptTemplate | |
| from langchain_core.output_parsers import StrOutputParser | |
| insights_mind_map_prompt_template = """ | |
| Based on the provided excerpt, create a detailed outline in Markdown format that organizes the content into distinct sections for the experiment setup and separate branches for each analysis mentioned. | |
| This structure should provide a clear and comprehensive overview of the study, emphasizing the setup, methodologies, findings, and implications of each analysis independently. | |
| Structure the outline as follows: | |
| - **Experiment Setup**: | |
| - Description of the general setup for the experiments, including any datasets and baselines used. | |
| For each analysis mentioned in the excerpt, create individual branches as follows: | |
| - **Analysis [Name/Number]**: | |
| 1. **Objectives**: | |
| - The main goal or purpose of this specific analysis. | |
| 2. **Methods**: | |
| - Detailed overview of the approach, techniques, or methodologies employed in this analysis. | |
| 3. **Results and Conclusions**: | |
| - Key findings, data interpretation, and the implications drawn from this analysis. | |
| - Combine this section if the results and conclusions are closely aligned to streamline content. | |
| 4. **Figures and Tables**: | |
| - Mention and describe any figures or tables referenced, linking them to their corresponding insights. | |
| Ensure that the outline is structured in Markdown format for clarity, facilitating its integration into documents or presentations. This structured approach aims to provide a comprehensive view of the experimental framework and the individual analyses, highlighting their unique contributions to the overall study." | |
| excerpt: {paper} | |
| """ | |
| insights_mind_map_output_parser = StrOutputParser() | |
| insights_mind_map_prompt = PromptTemplate( | |
| template=insights_mind_map_prompt_template, | |
| input_variables=["paper"], | |
| ) | |
| insights_mind_map_chain = ( | |
| lambda model: insights_mind_map_prompt | model | insights_mind_map_output_parser | |
| ) | |