Skip to content

TreeLSTM guide

Egor Spirin edited this page Feb 16, 2021 · 1 revision

TreeLSTM

PyTorch reimplementation of ChildSum version of TreeLSTM presented by Tai et al. in "Improved Semantic Representations From Tree-Structured Long Short-Term Memory Networks".

Training

A full list of train parameters is declared by the .yaml config. Examples of configs are store in the config folder.

To train the model with your config use

python train_treelstm.py -cn <config name>
Clone this wiki locally