README.md

---
license: apache-2.0
datasets:
- sem_eval_2018_task_1
language:
- en
metrics:
- accuracy
- f1
pipeline_tag: text-classification
widget:
  - text: We should lock the door and scream that curse word we know.
    example_title: Anger Tweet
  - text: You know what else barely touches the ground? Stray dogs, toenail clippings, road kill, hippies, dung beetles...
    example_title: Disgust Tweet
  - text: I sure am glad you told me earthquakes are a myth, Joy; otherwise, I’d be terrified right now.
    example_title: Fear Tweet
  - text: All right, everyone, fresh start. We are gonna have a good day, which will turn into a good week, which will turn into a good year, which turns into a good life!
    example_title: Joy Tweet
  - text: Crying helps me slow down and obsess over the weight of life's problems.
    example_title: Sadness Tweet
---

First posted on my [Kaggle](https://www.kaggle.com/code/wesleyacheng/twitter-emotion-multilabel-classification-w-bert/notebook#Create-Custom-Dataset).

Hello, I'm Wesley, nice to meet you! 👋

Since adding **Joy** and **Sadnesss** with **Anger** in my [Twitter Emotion MultiClass Classifier Notebook](https://www.kaggle.com/code/wesleyacheng/twitter-emotion-classification-with-bert), I wanted to complete the Inside Out group with **Fear** and **Disgust**!

Here I made a Twitter Emotion MultiLabel Classifier by doing transfer learning on [BERT](https://huggingface.co/distilbert-base-uncased) with the [SemEval Twitter Dataset](https://huggingface.co/datasets/sem_eval_2018_task_1) in PyTorch and HuggingFace.