dmlc--dgl
Quan (Andy) Gan
34 行
721 B
Markdown
34 行
721 B
Markdown
Junction Tree VAE - example for training
|
|
==========================================
|
|
|
|
This is a direct modification from https://github.com/wengong-jin/icml18-jtnn
|
|
|
|
Dependencies
|
|
--------------
|
|
* PyTorch 0.4.1+
|
|
* RDKit=2018.09.3.0
|
|
* requests
|
|
|
|
How to run
|
|
-----------
|
|
|
|
To run the model, use
|
|
```
|
|
python3 vaetrain_dgl.py
|
|
```
|
|
The script will automatically download the data, which is the same as the one in the
|
|
original repository.
|
|
|
|
To disable CUDA, run with `NOCUDA` variable set:
|
|
```
|
|
NOCUDA=1 python3 vaetrain_dgl.py
|
|
```
|
|
|
|
To decode for new molecules, run
|
|
```
|
|
python3 vaetrain_dgl.py -T
|
|
```
|
|
|
|
Currently, decoding involves encoding a training example, sampling from the posterior
|
|
distribution, and decoding a molecule from that.
|