## Amazon SageMaker fine-tuning seq2seq
### Objective
This repository intends to provide a sample script for fine tuning a sequence to sequence model for summarization tasks with domain specific data. See the sample_data folder for an example of a fine-tuning dataset and code comments for descriptions of different components used.

### Process
The `summ-fine-tune.ipynb` notebook walks through the process of importing the dataset to be used for fine-tuning, tokenizing the data, setting up necessary training components, running the training loop, then saving the training artifacts to a given S3 bucket. The notebook exposes a number of parameters and hyperparameters to simplify the configuration of the environment and the tuning of model training. To use the model for inference, use the `deploy-fine-tuned-model.ipynb` notebook. See the accompanying blog for a more detailed walk through of this project.

## Security

See [CONTRIBUTING](CONTRIBUTING.md#security-issue-notifications) for more information.

## License

This library is licensed under the MIT-0 License. See the LICENSE file.