Team 3 project repo for CS 492 - Introduction to Deep Learning course
We implemented two-fold predictor for coming trading day's stock price fluctuation
Look at below diagram
Download two pretrained BERT models from my google drive and place them on _pretrainedmodels directory
Link here, those two would take 1.3GB disk in total
Install required python modules using below command
Please set up your own venv before downloading these modules if you have your own setups!
$ pip install -r requirements.txt
Run below command on main directory of this project
Downloading pretrained multilingual BERT may take 7GB in disk, thus, make enough space before running this command
$ streamlit run simple_web.py
Our default setting is '삼성전자' on 2021-12-17. You may type in other companies or date to get a prediction. |
Samsung Electronics | KAKAO |
---|---|---|
Caution If the date to predict is too far, you may fail to get related articles!