Global Terrorism Database Assessment

• This project was created using data derived from the Global Terrorism Databse. The specific Kaggle Dataset used consists of a large portion of the data used, spanning over 180000 entries between the years 1970 and 2017, with the omission of data from 1993.
• The project is intended to predict the primary attack type of terror attacks based on a variety of the over 130 variables counted within this dataset.

Import and Usage Instructions

• In order to properly utilize this model, take note of the below directions and be sure to follow the steps to avoid issue(s):
  • pip install:
    • gradio
    • HuggingFace-hub
    • torch
  • These are the three main packages that will be needed to ensure proper function of the import and execution of the HuggingFace model.
  • To import the model, you will need to ensure that the following import statements are present in your module:
    • import torch
    • from HuggingFace_hub import hf_hub_download
    • from model.model import DeepGRU
    • import torch
    • import gradio as gr
    • import numpy as np
• Once these import statements are setup, the following is the line that would allow you to import the specific HuggingFace model. We will include the generic and specific code statements for context and clarity.
  • GENERIC (DO NOT USE THIS, IT IS AN USE-CASE EXAMPLE): weights_path = hf_hub_download(repo_id=, filename=)
  • SPECIFIC: weights_path = hf_hub_download(repo_id="OGrieco/CMPSC-405-GRU", filename="GRU_h64_l3_d0.6_s50.pth")
• Now at this point, the .pth file containing the data of the pre-trained model can be loaded using the torch.load command. You are now free to use the model for your own means.