README.md

---
datasets:
- fake_news_english
language:
- en
library_name: transformers
---

# Model Card for Model ID

<!-- Provide a quick summary of what the model is/does. -->

This modelcard aims to be a base template for new models. It has been generated using [this raw template](https://github.com/huggingface/huggingface_hub/blob/main/src/huggingface_hub/templates/modelcard_template.md?plain=1).

## Model Details

### Model Description

This model is created using the fake news dataset from kaggle. The custom model is a fine tuned distilbert model with additional layers. 
The code was written in pytorch. The dataset was processed with removing symbols and converting text to lower case. The train - validate - test datasets
are created in the ratio 60:20:20. The model was trained for two epochs and obtained an accuracy of 99%. 
However, the model has been shown to be overfitted on certain types of samples owing to lack of diversity in the samples. Please be cautious 
before using this model for a downstream use case

- **Developed by:** Aishwarya A. Nair
- **Shared by [optional]:** [More Information Needed]
- **Model type:** [More Information Needed]
- **Language(s) (NLP):** [More Information Needed]
- **License:** [More Information Needed]
- **Finetuned from model [optional]:** distilbert-base-uncased

### Model Sources [optional]

<!-- Provide the basic links for the model. -->

- **Notebook:** [https://colab.research.google.com/drive/1G65Ye1UC-QeQXAJN9WPkGHvM0Qgo4Mf5?usp=sharing]

## Uses
Fake news detection can be used in the cases when you need to verify the veracity of a news article or a tweet or other pieces of text.

## Bias, Risks, and Limitations

The model has been shown to be overfitted on certain types of samples owing to lack of diversity in the samples. Please be cautious 
before using this model for a downstream use case