distilbert-base-uncased-finetuned-emotion

This model is a fine-tuned version of distilbert-base-uncased on the emotion dataset. It classifies an English tweet into one of six emotions: sadness, joy, love, anger, fear, surprise.

It was trained as a course assignment for an introductory AI course.

Results

On the validation set:

• Accuracy: 0.932
• Weighted F1: 0.9328

Training hyperparameters

• learning rate: 2e-5
• train batch size: 64
• optimizer: AdamW
• lr scheduler: linear with 10% warmup
• weight decay: 0.01
• epochs: 3 (best checkpoint kept by validation F1)

How to use

from transformers import pipeline

clf = pipeline("text-classification",
               model="ireallydk/distilbert-base-uncased-finetuned-emotion")
print(clf("i feel so happy today"))