|
--- |
|
tags: |
|
- distilbert |
|
- health |
|
- tweet |
|
datasets: |
|
- custom-phm-tweets |
|
metrics: |
|
- accuracy |
|
model-index: |
|
- name: distilbert-phmtweets-sutd |
|
results: |
|
- task: |
|
name: Text Classification |
|
type: text-classification |
|
dataset: |
|
name: custom-phm-tweets |
|
type: labelled |
|
metrics: |
|
- name: Accuracy |
|
type: accuracy |
|
value: 0.877 |
|
--- |
|
|
|
# distilbert-phmtweets-sutd |
|
|
|
This model is a fine-tuned version of [distilbert-base-uncased](https://huggingface.co/distilbert-base-uncased) for text classification to identify public health events through tweets. The project was based on an [Emory University Study on Detection of Personal Health Mentions in Social Media paper](https://arxiv.org/pdf/1802.09130v2.pdf), that worked with this [custom dataset](https://github.com/emory-irlab/PHM2017). |
|
|
|
It achieves the following results on the evaluation set: |
|
- Accuracy: 0.877 |
|
|
|
## Usage |
|
|
|
```Python |
|
from transformers import AutoTokenizer, AutoModelForSequenceClassification |
|
tokenizer = AutoTokenizer.from_pretrained("dibsondivya/distilbert-phmtweets-sutd") |
|
model = AutoModelForSequenceClassification.from_pretrained("dibsondivya/distilbert-phmtweets-sutd") |
|
``` |
|
|
|
|
|
### Model Evaluation Results |
|
With Validation Set |
|
- Accuracy: 0.8708661417322835 |
|
|
|
With Test Set |
|
- Accuracy: 0.8772961058045555 |
|
|
|
# Reference for distilbert-base-uncased Model |
|
@article{Sanh2019DistilBERTAD, |
|
title={DistilBERT, a distilled version of BERT: smaller, faster, cheaper and lighter}, |
|
author={Victor Sanh and Lysandre Debut and Julien Chaumond and Thomas Wolf}, |
|
journal={ArXiv}, |
|
year={2019}, |
|
volume={abs/1910.01108} |
|
} |
|
|
|
|
