Model code for the SageMaker instance of the beular-api project.
These notebooks, with the exception of pipeline/Analysis.ipynb and pipeline/Eval.ipynb are meant to be run in an AWS Sagemaker Instance. If working with the former two noteboks, install the requirements with pip install -r requirements.txt.
If you're connecting this repo to the SageMaker instance in beular-api, ensure that this repo is public and add it's git url to config.py in beular-api. This will automatically git clone this project into the SageMaker instance once it builds. It will also let you source any code you develop while working within SageMaker.
The BlazingText model is around 1GB, so it's not suitable for GitHub. You can download that model from here. Place it in the pipeline/estimators/ directory with the other models. Once you're up and running with AWS, this (and other models) will be stored in S3.