chenweiye/llm-homework-sentiment-model

BERTweet sentiment classifier

This repository contains a vinai/bertweet-base model fine-tuned for binary sentiment classification on the small educational dataset included in the course submission.

Evaluation on the fixed 10-example test split after continued fine-tuning: accuracy 1.00, positive F1 1.00, negative F1 1.00, macro F1 1.00. The test split is very small, so this result should not be interpreted as broad generalization.