---
license: apache-2.0
tags:
- generated_from_keras_callback
model-index:
- name: tmpjy56pamo
  results: []
---


## Model description

The rotten-tomatoes-model is a text-classification model. It used the `bert-base-cased` model, and was fine tuned on the `rotten_tomatoes` model. 

After inputting a movie review, the model will output its prediction of how positive/negative the review is. `LABEL_0` is Negative, while `LABEL_1` is Positive.   

## Intended uses & limitations

This model can be used to take in movie reviews and predict whether the overall sentiments of the review are positive or negative.

An example use case for this model is taking in reviews spanning from the start of the pandemic to the current time to see how sentiments surrounding movies might have been affected by when in the pandemic it was released (or other factors such as the method it was released).

## Training and evaluation data

As mentioned above, this model was fine-tuned on the `rotten_tomatoes` dataset, which contains 5,331 positive and 5,331 negative movie reviews from Rotten Tomatoes. 


### Training results

| Train Loss | Train Accuracy | Validation Loss | Validation Accuracy | Epoch |
|:----------:|:--------------:|:---------------:|:-------------------:|:-----:|
| 0.4028     | 0.8213         | 0.4626          | 0.8433              | 0     |
| 0.1628     | 0.9390         | 0.3498          | 0.8696              | 1     |
| 0.0386     | 0.9878         | 0.4790          | 0.8621              | 2     |


### Framework versions

- Transformers 4.18.0
- TensorFlow 2.8.0
- Datasets 2.1.0
- Tokenizers 0.12.1