Create QuestionRetrieverTool.py
Browse files- QuestionRetrieverTool.py +38 -0
QuestionRetrieverTool.py
ADDED
|
@@ -0,0 +1,38 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
from langchain.docstore.document import Document
|
| 2 |
+
from langchain.text_splitter import RecursiveCharacterTextSplitter
|
| 3 |
+
from smolagents import Tool
|
| 4 |
+
from langchain_community.retrievers import BM25Retriever
|
| 5 |
+
from smolagents import CodeAgent, InferenceClientModel
|
| 6 |
+
from datasets import load_dataset
|
| 7 |
+
import re
|
| 8 |
+
import pandas as pd
|
| 9 |
+
#%%
|
| 10 |
+
class QuestionRetrieverTool(Tool):
|
| 11 |
+
name = "Question_retriever"
|
| 12 |
+
description = "Uses semantic search to retrieve relevant question given the class, difficulty, and topic inputs by the user."
|
| 13 |
+
inputs = {
|
| 14 |
+
"query": {
|
| 15 |
+
"type": "string",
|
| 16 |
+
"description": "This tool returns relevant question and answer pairs based on the provided context.",
|
| 17 |
+
}
|
| 18 |
+
}
|
| 19 |
+
output_type = "string"
|
| 20 |
+
|
| 21 |
+
def __init__(self, docs, **kwargs):
|
| 22 |
+
super().__init__(**kwargs)
|
| 23 |
+
self.retriever = BM25Retriever.from_documents(
|
| 24 |
+
docs, k=5 # Retrieve the top 5 documents
|
| 25 |
+
)
|
| 26 |
+
|
| 27 |
+
def forward(self, query: str) -> str:
|
| 28 |
+
assert isinstance(query, str), "Your search query must be a string"
|
| 29 |
+
|
| 30 |
+
docs = self.retriever.invoke(
|
| 31 |
+
query,
|
| 32 |
+
)
|
| 33 |
+
return "\nRetrieved example question and answer pairs:\n" + "".join(
|
| 34 |
+
[
|
| 35 |
+
f"\n\n===== Q and A pairs {str(i)} =====\n" + doc.page_content
|
| 36 |
+
for i, doc in enumerate(docs)
|
| 37 |
+
]
|
| 38 |
+
)
|