CMPE346 - Assignment 02

Task

Model

• Base model: xlm-roberta-base
• Platform: Hugging Face Hub
• Note: Due to time and hardware limitations, the base pretrained model was used without fine-tuning.

Files

• main.py: Training pipeline (data load, preprocessing, training, saving)
• model.py: Model class with Huggingface Trainer support
• preprocessing.py: Tokenizer and dataset preparation
• run_test.py: Evaluation script using Hugging Face pipeline
• data/train.csv, data/valid.csv: Provided datasets

How to Run