Skip to content

narrative_manager.md

Latest commit

 

History

History
188 lines (135 loc) · 6.37 KB

narrative_manager.md

File metadata and controls

188 lines (135 loc) · 6.37 KB
Raw

🚀 Extract Reviewer

Welcome to the Extract Reviewer module documentation! The ExtractReviewer class is essential for reviewing and refining extracted information, ensuring its accuracy and completeness. This guide provides detailed information about its attributes, methods, and usage examples.

📚 Class: ExtractReviewer

Description:

The ExtractReviewer class is responsible for reviewing and refining the extracted information from EMS transcripts. It leverages the ModelLoader and PromptManager classes to interact with the AI model and manage prompts effectively.

🏗️ Attributes:

  • model_loader (ModelLoader): An instance of ModelLoader to interact with the AI model.
  • prompt_manager (PromptManager): An instance of PromptManager to manage prompts.

🚀 Methods:

__init__(self, model_loader: ModelLoader, prompt_manager: PromptManager)

Description: Initializes the ExtractReviewer with a ModelLoader and PromptManager instance.

Parameters:

  • model_loader (ModelLoader): An instance of ModelLoader to interact with the AI model.
  • prompt_manager (PromptManager): An instance of PromptManager to manage prompts.

Example:

from modules.model_loader import ModelLoader
from modules.prompt_manager import PromptManager
from modules.extract_reviewer import ExtractReviewer

# Initialize ModelLoader and PromptManager
model_loader = ModelLoader(model_name="llama3.1")
prompt_manager = PromptManager(prompts={"review": "Review and refine the following information: {info}"})

# Initialize ExtractReviewer
reviewer = ExtractReviewer(model_loader=model_loader, prompt_manager=prompt_manager)

review_section(self, section: str, user_input: Optional[str] = None) -> str

Description: Reviews a section of extracted data using the AI model.

Parameters:

  • section (str): The section of extracted data to review.
  • user_input (Optional[str]): The user's input for modifications, if any.

Returns:

  • str: The AI model's response.

Example:

section = "Incident Information\n- Unit: [No Info]\n- Response Mode: emergent\n..."
reviewed_section = reviewer.review_section(section)
print("Reviewed Section:")
print(reviewed_section)

final_review(self, updated_section: str) -> str

Description: Performs a final review of a section after changes have been made.

Parameters:

  • updated_section (str): The section of data after user modifications.

Returns:

  • str: The AI model's response after final review.

Example:

updated_section = "Incident Information\n- Unit: 123\n- Response Mode: emergent\n..."
final_reviewed_section = reviewer.final_review(updated_section)
print("Final Reviewed Section:")
print(final_reviewed_section)

_get_review_prompt(self, prompt_key: str, section: str, user_input: Optional[str] = None) -> str

Description: Retrieves and formats the review prompt.

Parameters:

  • prompt_key (str): The key for the review prompt.
  • section (str): The section of extracted data.
  • user_input (Optional[str]): The user's input for modifications, if any.

Returns:

  • str: The formatted review prompt.

Example:

prompt_key = "review"
section = "Incident Information\n- Unit: [No Info]\n- Response Mode: emergent\n..."
user_input = "Unit: 123"
formatted_prompt = reviewer._get_review_prompt(prompt_key, section, user_input)
print("Formatted Prompt:")
print(formatted_prompt)

_generate_response(self, prompt: str) -> str

Description: Generates a response from the AI model based on the provided prompt.

Parameters:

  • prompt (str): The input prompt for the model.

Returns:

  • str: The AI model's response.

Example:

prompt = "Review and refine the following information: Incident Information\n- Unit: 123\n- Response Mode: emergent\n..."
response = reviewer._generate_response(prompt)
print("Generated Response:")
print(response)

🌟 Usage Examples

Example 1: Initializing the ExtractReviewer

from modules.model_loader import ModelLoader
from modules.prompt_manager import PromptManager
from modules.extract_reviewer import ExtractReviewer

# Initialize ModelLoader with a specific model name
model_loader = ModelLoader(model_name="llama3.1")

# Initialize PromptManager with predefined prompts
prompts = {
    "review": "Review and refine the following information: {info}"
}
prompt_manager = PromptManager(prompts=prompts, context_window_size=32000)

# Initialize ExtractReviewer
reviewer = ExtractReviewer(model_loader=model_loader, prompt_manager=prompt_manager)

Example 2: Reviewing a Section of Extracted Data

from modules.model_loader import ModelLoader
from modules.prompt_manager import PromptManager
from modules.extract_reviewer import ExtractReviewer

# Initialize components
model_loader = ModelLoader(model_name="llama3.1")
prompt_manager = PromptManager(prompts={"review": "Review and refine the following information: {info}"})
reviewer = ExtractReviewer(model_loader=model_loader, prompt_manager=prompt_manager)

# Review a section of extracted data
section = "Incident Information\n- Unit: [No Info]\n- Response Mode: emergent\n..."
reviewed_section = reviewer.review_section(section)
print("Reviewed Section:")
print(reviewed_section)

Example 3: Performing a Final Review of a Section

from modules.model_loader import ModelLoader
from modules.prompt_manager import PromptManager
from modules.extract_reviewer import ExtractReviewer

# Initialize components
model_loader = ModelLoader(model_name="llama3.1")
prompt_manager = PromptManager(prompts={"review": "Review and refine the following information: {info}"})
reviewer = ExtractReviewer(model_loader=model_loader, prompt_manager=prompt_manager)

# Perform a final review of a section after user modifications
updated_section = "Incident Information\n- Unit: 123\n- Response Mode: emergent\n..."
final_reviewed_section = reviewer.final_review(updated_section)
print("Final Reviewed Section:")
print(final_reviewed_section)

🎉 Conclusion

The ExtractReviewer class is an essential tool for reviewing and refining extracted information, ensuring its accuracy and completeness. By understanding its attributes and methods, you can effectively improve the quality of extracted data in EMS documentation. For any questions or support, please contact ScottSucks.