Spaces:
Sleeping
Sleeping
Delete crewai_flashcard.py
#3
by
Pavaas
- opened
- crewai_flashcard.py +0 -104
crewai_flashcard.py
DELETED
|
@@ -1,104 +0,0 @@
|
|
| 1 |
-
import os
|
| 2 |
-
from crewai import Agent, Task, Crew
|
| 3 |
-
from crewai_tools import PDFSearchTool
|
| 4 |
-
import openai
|
| 5 |
-
|
| 6 |
-
# Ensure your OpenAI API key and model settings are set
|
| 7 |
-
openai.api_key = os.getenv("OPENAI_API_KEY")
|
| 8 |
-
|
| 9 |
-
# Initialize the PDFSearchTool without a preset PDF
|
| 10 |
-
pdf_tool = PDFSearchTool()
|
| 11 |
-
|
| 12 |
-
# Agent 1: PDF Extractor Agent β extracts text from the specified page range.
|
| 13 |
-
pdf_extractor = Agent(
|
| 14 |
-
role="PDF Extractor",
|
| 15 |
-
goal="Extract the text content from the specified page range of the provided PDF.",
|
| 16 |
-
backstory="You specialize in reading and processing PDFs to return useful study content.",
|
| 17 |
-
verbose=True,
|
| 18 |
-
tools=[pdf_tool]
|
| 19 |
-
)
|
| 20 |
-
|
| 21 |
-
# Agent 2: Flashcard Generator Agent β generates flashcards based on extracted text.
|
| 22 |
-
flashcard_generator = Agent(
|
| 23 |
-
role="Flashcard Generator",
|
| 24 |
-
goal="Generate flashcards with questions and answers based on the extracted text.",
|
| 25 |
-
backstory="You create study flashcards to test understanding, ensuring each flashcard has a clear question and answer.",
|
| 26 |
-
verbose=True
|
| 27 |
-
)
|
| 28 |
-
|
| 29 |
-
# Task 1: Extraction Task β Extract text from pages {page_range} of the PDF.
|
| 30 |
-
extraction_task = Task(
|
| 31 |
-
description=(
|
| 32 |
-
"Extract and return the text content from the PDF located at {pdf_file_path}."
|
| 33 |
-
),
|
| 34 |
-
expected_output="A string containing the extracted text from the specified pages.",
|
| 35 |
-
agent=pdf_extractor
|
| 36 |
-
)
|
| 37 |
-
|
| 38 |
-
# Task 2: Flashcard Generation Task β Generate {flashcard_count} flashcards from the extracted text.
|
| 39 |
-
flashcard_task = Task(
|
| 40 |
-
description=(
|
| 41 |
-
"Based on the following extracted text"
|
| 42 |
-
"Generate {flashcard_count} flashcards. Each flashcard should have a question and a corresponding answer. "
|
| 43 |
-
"Return the flashcards as a JSON list of objects with keys 'question' and 'answer'."
|
| 44 |
-
),
|
| 45 |
-
expected_output="A JSON list of flashcards.",
|
| 46 |
-
output_file="flashcards.json",
|
| 47 |
-
agent=flashcard_generator,
|
| 48 |
-
context=[extraction_task]
|
| 49 |
-
)
|
| 50 |
-
|
| 51 |
-
|
| 52 |
-
# In crewai_flashcard.py
|
| 53 |
-
|
| 54 |
-
# # Task 1: Extraction Task β Extract text from pages {page_range} of the PDF.
|
| 55 |
-
# extraction_task = Task(
|
| 56 |
-
# description=(
|
| 57 |
-
# "Extract and return the text content from pages page_range of the PDF located at pdf_file_path."
|
| 58 |
-
# ),
|
| 59 |
-
# expected_output="A string containing the extracted text from the specified pages.",
|
| 60 |
-
# agent=pdf_extractor
|
| 61 |
-
# )
|
| 62 |
-
|
| 63 |
-
# # Task 2: Flashcard Generation Task β Generate {flashcard_count} flashcards from the extracted text.
|
| 64 |
-
# flashcard_task = Task(
|
| 65 |
-
# description=(
|
| 66 |
-
# "Based on the following extracted text:"
|
| 67 |
-
# "Generate {flashcard_count} flashcards. Each flashcard should have a question and a corresponding answer. "
|
| 68 |
-
# "Return the flashcards as a JSON list of objects with keys 'question' and 'answer'."
|
| 69 |
-
# ),
|
| 70 |
-
# expected_output="A JSON list of flashcards.",
|
| 71 |
-
# agent=flashcard_generator,
|
| 72 |
-
# context=[extraction_task]
|
| 73 |
-
# )
|
| 74 |
-
|
| 75 |
-
|
| 76 |
-
|
| 77 |
-
|
| 78 |
-
# Assemble the Crew
|
| 79 |
-
flashcard_crew = Crew(
|
| 80 |
-
agents=[pdf_extractor, flashcard_generator],
|
| 81 |
-
tasks=[extraction_task, flashcard_task],
|
| 82 |
-
verbose=True
|
| 83 |
-
)
|
| 84 |
-
|
| 85 |
-
def generate_flashcards( pdf_file_path: str,flashcard_count: int):
|
| 86 |
-
#generate_flashcards(pdf_file_path: str, page_range: str, flashcard_count: int):
|
| 87 |
-
"""
|
| 88 |
-
Run the CrewAI system to extract text from specified pages and generate flashcards.
|
| 89 |
-
|
| 90 |
-
Args:
|
| 91 |
-
pdf_file_path (str): Path to the PDF file.
|
| 92 |
-
page_range (str): Page range to extract text from (e.g., "1-5").
|
| 93 |
-
flashcard_count (int): Number of flashcards to generate.
|
| 94 |
-
|
| 95 |
-
Returns:
|
| 96 |
-
str: JSON string containing the flashcards.
|
| 97 |
-
"""
|
| 98 |
-
inputs = {
|
| 99 |
-
"pdf_file_path": pdf_file_path,
|
| 100 |
-
#"page_range": page_range,
|
| 101 |
-
"flashcard_count": flashcard_count
|
| 102 |
-
}
|
| 103 |
-
results = flashcard_crew.kickoff(inputs=inputs)
|
| 104 |
-
return results.raw #[flashcard_task.id]
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|