/samples-basic-text-generation-lstm

deeply insipired by https://www.kaggle.com/shivamb/beginners-guide-to-text-generation-using-lstms

Primary LanguageJupyter NotebookMIT LicenseMIT

samples-basic-text-generation-lstm

deeply insipired by Beginners Guide to Text Generation using LSTMs

Data Set

https://www.kaggle.com/aashita/nyt-comments/version/13

Unzip into text-generation-lstm/sources/data/nyt-comments folder.

RUN DOCKER

This repository contains a basic docker-compose configuration to run a Jupyter Notebook server with Keras support (Tensor backend), using the default conda environment. This configuration automatically mounts the text-generation-lstm/sources folder inside the container, and sets it as the working directory.

docker-compose run

The notebook server can be accessed by navigating to http://localhost:8900 on the host.

RUN CONDA ENVIRONMENT

This repository contains a conda environment file in text-generation-lstm/sources folder.

Create environment:

conda env create -f environment.yml

Activate environment:

conda activate text-generation-lstm

Register new Kernel for Jupyter Notebook:

python -m ipykernel install --user --name text-generation-lstm --display-name "Python (text-generation-lstm)"

Run Jupyter Notebook Server:

jupyter notebook

Select "Python (text-generation-lstm)" from Kernel menu.

TODO