Assignment 4 A: Natural Language Inference with BERT (Total: 50 Points out of 100)

Getting Started

• Connect to Google Colab: https://colab.research.google.com/
• Open NLI.ipynb in Collab by either uploading the file or follow:

File -> Open Notebook -> Github -> https://github.com/csci5832-f23/assignment_4 -> csci5832-f23/assignment_4 -> NLI.ipynb

• Using GPUs: Look at the top right side of Google Colab's pane

  

  • Click the button -> Change Runtime Type -> T4 GPU -> Save

  

You should now see:

Assignment 4 B: Natural Language Inference with BERT (Total: 50 Points out of 100)

Getting Started

• Install Anaconda

• Create a new conda environment (in the Terminal):

  conda create --name a4 python=3.9

• Activate the environment:

  conda activate a4

• Clone the repo, Enter the assignment directory

  git clone https://github.com/csci5832-f23/assignment_4.git

  cd assignment_4

• Install required libraries

  pip install -r requirements.txt

Task

• Irony Detection in Twitter tweets with Twitter GloVe Vectors and Binary Logistic Regression
  • Fill in the missing code in IronyDetection.ipynb

Submission

• Format: Single merged PDF
• Due Date: TBA