---
tags:
- fastai
---

# Model card

## Model description
This model was fine-tuned on `resnet18` for 10 epochs using training data described in [this blogpost](https://mlops.systems/fastai/redactionmodel/computervision/datalabelling/2021/09/06/redaction-classification-chapter-2.html). The model was trained using the [fastai](https://github.com/fastai/fastai) library.

## Model performance
The model achieved around 96% accuracy on a held-out validation dataset.

## Inference demonstration
You can test out the model's ability to detect whether a document is redacted [here](https://huggingface.co/spaces/strickvl/fastai_redaction_classifier).