Docusearch
Welcome to Docusearch, the ultimate tool for document searching and processing using the power of vector embeddings. Follow the steps below to get started quickly!
Installation
First, you need to install the docusearch
package using pip
. Installation may take a minute or two. To avoid conflicts, it is recommend you use a virtual environment:
pip install docusearch
Usage
Step 1: Import the Module
After installing the package, import the process_query
function from the docusearch
module:
from docusearch import process_query
Step 2: Set the Parameters
Define the parameters for your query. You need to set your OpenAI API key, the path to the folder containing the documents, and the query for those documents
query = "What are some cool features of the Audi r8"
api_key = "your-openai-key"
folder_path = "path-to-your-folder"
Step 3: Call the Function and Print the Result
Now, call the process_query
function with the parameters you set and print the result:
result = process_query(query, api_key, folder_path)
print(result)
Example
Here is the complete example code:
from docusearch import process_query
query = "What are some cool features of the Audi r8"
api_key = "your-openai-key"
folder_path = "path-to-your-folder"
result = process_query(query, api_key, folder_path)
print(result)
And that's it! You have successfully used the docusearch package to process your query. Enjoy searching your documents with ease!
Additional Info
If you would like specific information about the result, you can use the extract_info
function. It will provide you the document source, answer, citations, and warnings for unsupported files.
from docusearch import extract_info
Use the extract_info function to extract the document source, answer, citations, and any warnings:
document_source, answer, citations, warning = extract_info(result)
print(f"Document Source: {document_source}")
print(f"Answer: {answer}")
print(f"Citations: {citations}")
if warning:
print(f"Warning: {warning}")
Complete Example
Here is the complete example code to query and extract specific information:
from docusearch import process_query, extract_info
query = "What are some cool features of the Audi r8"
api_key = "your-openai-key"
folder_path = "path-to-your-folder"
result = process_query(query, api_key, folder_path)
document_source, answer, citations, warning = extract_info(result)
print(f"Document Source: {document_source}")
print(f"Answer: {answer}")
print(f"Citations: {citations}")
if warning:
print(f"Warning: {warning}")
By following these steps, you can easily extract and use specific pieces of information from the result provided by the docusearch
package. Have fun!