document_summarize

Launch

The created application can be used by starting it on a Docker container or by installing the library in the local environment.

Clone：

git clone https://github.com/YusukeOhnishi/document_summarize.git
cd ./document_summarize

Requires OpenAI API Key for use. Please obtain the KEY from the official website below.
https://platform.openai.com/api-keys

Docker

First, write your OpenAI API Key in the Dockerfile. Enter the obtained OpenAI API Key in the following part of the Dockerfile and save it.

After entering the OpenAI API Key, please build and run from the Dockerfile.

docker build -t document_summarize .
docker run -p 8501:8501 document_summarize

After executing the above command, access http:localhost:8501.

Local

Execute the following only when running in a virtual environment.

python venv venv
source ./source venv/bin/activate

Setting environment variables and install the necessary libraries.

export OPENAI_API_KEY='Your OpenAI API Key'
pip install -r requirement.txt
streamlit run main.py

After startup, a URL will be displayed on the console, so access this URL.

Usage

Overview

Outputs a summary of the selected file using OpenAI's gpt model. You can choose the language and gpt model for the final output result. Also, when summarizing a file, OpenAI will output how much it will cost. In addition, when outputting the results, it will output how much it will cost depending on the number of output tokens.

Set model

Select gpt model (Default is gpt-3.5-turbo-16k). Currently, the following models are available.

• gpt-3.5-turbo-16k
• gpt-3.5-turbo
• gpt-4
• gpt-4-32k
• gpt-4-1106-preview

Set output language

Specify in which language the summary results will be output (Default is English). Currently, the following languages can be specified.

• English
• Japanese

Select file

Select the document you want to summarize. You can select documents located locally or on the web. The supported file formats are as follows.

• pdf

If you select a local file, please upload the file with URL Load Mode turned off.

If you want to select a file on the web, please enter the URL in the text box with URL Load Mode turned on.

When the file is successfully loaded, the number of tokens for the input file and the cost of the input prompt in the gpt model will be displayed in the sidebar.

Click Summary button

Click the summary button.

Once the summarization process is complete, the summary results will be output. Also, the number of tokens used to output the results and the OpenAI API cost are output in the sidebar.