|
---
|
|
library_name: transformers
|
|
license: mit
|
|
base_model: base-uncased
|
|
tags:
|
|
- bert
|
|
- fine-tuning
|
|
- text-classification
|
|
model-index:
|
|
- name: NLP_with_Disaster_Tweets
|
|
results:
|
|
- task:
|
|
type: text-classification
|
|
name: Text Classification
|
|
metrics:
|
|
- name: Accuracy
|
|
type: accuracy
|
|
value: 0.835
|
|
language:
|
|
- en
|
|
---
|
|
|
|
# Disaster Tweets Classification
|
|
|
|
This model is fine-tuned BERT for classifying whether a tweet is about a real disaster or not.
|
|
|
|
## Model Description
|
|
|
|
- Based on `bert-base-uncased`
|
|
- Fine-tuned for binary classification task
|
|
- Achieves 83.5% accuracy on validation set
|
|
- Trained on Kaggle's "Natural Language Processing with Disaster Tweets" competition dataset
|
|
|
|
## How to Use
|
|
|
|
```python
|
|
from transformers import AutoConfig, AutoTokenizer, AutoModelForSequenceClassification
|
|
|
|
# Load model and tokenizer
|
|
tokenizer = AutoTokenizer.from_pretrained("real-jiakai/NLP_with_Disaster_Tweets")
|
|
model = AutoModelForSequenceClassification.from_pretrained("real-jiakai/NLP_with_Disaster_Tweets")
|
|
|
|
# Example usage
|
|
text = "There was a major earthquake in California"
|
|
inputs = tokenizer(text, return_tensors="pt", padding=True, truncation=True, max_length=512)
|
|
outputs = model(**inputs)
|
|
predicted_class = outputs.logits.argmax(-1).item()
|
|
```
|
|
|
|
## License
|
|
|
|
This model is licensed under the [MIT](https://opensource.org/license/mit) License.
|
|
|
|
## Citation
|
|
|
|
If you use this model in your work, please cite:
|
|
|
|
```
|
|
@misc{NLP_with_Disaster_Tweets,
|
|
author = {real-jiakai},
|
|
title = {NLP_with_Disaster_Tweets},
|
|
year = {2024},
|
|
url = {https://huggingface.co/real-jiakai/NLP_with_Disaster_Tweets},
|
|
publisher = {Hugging Face}
|
|
}
|
|
```
|
|
|
|
|
|
|
