|
--- |
|
license: apache-2.0 |
|
datasets: |
|
- tweet_eval |
|
language: |
|
- en |
|
metrics: |
|
- accuracy |
|
- f1 |
|
pipeline_tag: text-classification |
|
widget: |
|
- text: Yay! |
|
example_title: Joy Example |
|
- text: There is no meaning in life. |
|
example_title: Sadness Example |
|
- text: I hate you! |
|
example_title: Anger Example |
|
--- |
|
|
|
First posted in my [Kaggle](https://www.kaggle.com/code/wesleyacheng/twitter-emotion-classification-with-bert). |
|
|
|
Hello, I'm **Wesley**, nice to meet you! 👋 |
|
|
|
While I was making my **[Angry Birds Classifier](https://www.kaggle.com/code/wesleyacheng/angry-birds-classifier)** to classify if tweets are angry or not, I thought why don't we add **2** more emotions! **Joy and Sadness** into the mix! |
|
|
|
Here I created a **Multiclass Text Classifier** that classifies tweets as either having **JOY, SADNESS, or ANGER**. |
|
|
|
I used the [Twitter Emotion Dataset](https://huggingface.co/datasets/tweet_eval/viewer/emotion/train) and [BERT](https://huggingface.co/distilbert-base-uncased) to do [Transfer Learning](https://en.wikipedia.org/wiki/Transfer_learning) with [PyTorch](https://pytorch.org) and [HuggingFace](https://huggingface.co). |
