Albert-base-v2-emotion

Model description:

Albert is A Lite BERT architecture that has significantly fewer parameters than a traditional BERT architecture.

Albert-base-v2 finetuned on the emotion dataset using HuggingFace Trainer with below Hyperparameters

 learning rate 2e-5, 
 batch size 64,
 num_train_epochs=8,

Model Performance Comparision on Emotion Dataset from Twitter:

Model	Accuracy	F1 Score	Test Sample per Second
Distilbert-base-uncased-emotion	93.8	93.79	398.69
Bert-base-uncased-emotion	94.05	94.06	190.152
Roberta-base-emotion	93.95	93.97	195.639
Albert-base-v2-emotion	93.6	93.65	182.794

How to Use the model:

from transformers import pipeline
classifier = pipeline("text-classification",model='bhadresh-savani/albert-base-v2-emotion', return_all_scores=True)
prediction = classifier("I love using transformers. The best part is wide range of support and its easy to use", )
print(prediction)

"""
Output:
[[
{'label': 'sadness', 'score': 0.010403595864772797}, 
{'label': 'joy', 'score': 0.8902180790901184}, 
{'label': 'love', 'score': 0.042532723397016525}, 
{'label': 'anger', 'score': 0.041297927498817444}, 
{'label': 'fear', 'score': 0.011772023513913155}, 
{'label': 'surprise', 'score': 0.0037756056990474463}
]]
"""

Dataset:

Twitter-Sentiment-Analysis.

Training procedure

Colab Notebook

Eval results

{
'test_accuracy': 0.936,
 'test_f1': 0.9365658988006296,
 'test_loss': 0.15278364717960358,
 'test_runtime': 10.9413,
 'test_samples_per_second': 182.794,
 'test_steps_per_second': 2.925
 }

Reference:

• Natural Language Processing with Transformer By Lewis Tunstall, Leandro von Werra, Thomas Wolf