Code

Moltransf#85

Code

Draft

samadejacobs wants to merge 22 commits intomolecularsets:mastermolecularsets/moses:masterfrom samadejacobs:moltransfsamadejacobs/moses:moltransfCopy head branch name to clipboard

+31,270-111

Commits

Commits on Apr 2, 2019

• Extend JTNNVAE to support WAE,users have the choice of VAE or WAE in for the latent space

  samadejacobs

  committed
  388734d

  Copy full SHA for 388734d

• Extend JTNNVAE to support WAE,users have the choice of VAE or WAE in for the latent space

  samadejacobs

  committed
  464001a

  Copy full SHA for 464001a

Commits on Oct 16, 2019

• Extend CharRNN to load LBANN trained models

  samadejacobs

  committed
  4f4d31c

  Copy full SHA for 4f4d31c

• Fix typos

  samadejacobs

  committed
  6b881f7

  Copy full SHA for 6b881f7

• fix typos, import numpy etc

  samadejacobs

  committed
  9e7bdc6

  Copy full SHA for 9e7bdc6

Commits on Oct 17, 2019

• add vocab and config for char RNN

  samadejacobs

  committed
  b7c1a68

  Copy full SHA for b7c1a68

Commits on Oct 18, 2019

• More clean up and strict check

  samadejacobs

  committed
  79db018

  Copy full SHA for 79db018

Commits on Nov 22, 2019

• adds preprocessing script to featurize arbitrary smiles datasets. makes minor changes to run.py and sample.py that allow for the specification of arbitrary datasets and vocabs, also minor changes t…

  Show description for 161e892
  wderekjones

  committed
  161e892

  Copy full SHA for 161e892

Commits on Nov 25, 2019

• adding support for TF 2.0 to moses/metrics/utils_fcd.py

  wderekjones

  committed
  1527153

  Copy full SHA for 1527153

Commits on Feb 29, 2020

• adding utilities to preprocess arbitrary datasets, both python and slurm scripts

  wderekjones

  committed
  42bc7a7

  Copy full SHA for 42bc7a7

Commits on Mar 21, 2020

• adding scripts to prepare datasets

  wderekjones

  committed
  9a4f560

  Copy full SHA for 9a4f560

Commits on Mar 25, 2020

• addressing issues related to structure of example slurm data processing script (scripts/extract_enamine_vocab.sl), evalutation script (scripts/run.py), vocab extraction (scripts/char_vocab_utils.py…

  Show description for 28932ce
  wderekjones

  committed
  28932ce

  Copy full SHA for 28932ce

Commits on Mar 30, 2020

• Merge pull request #1 from samadejacobs/integrate_lbann_moses

  Show description for ac132b0
  samadejacobs

  authored
  ac132b0

  Copy full SHA for ac132b0

Commits on Jun 26, 2020

• Update model in response to changes in LBANN naming of dumping weights

  samadejacobs

  committed
  c59e4f4

  Copy full SHA for c59e4f4

Commits on Jul 14, 2020

• Make epoch count optional especially if loading 'final' model, one less arguments to worry about

  samadejacobs

  committed
  6e42081

  Copy full SHA for 6e42081

Commits on Aug 20, 2020

• Support for loading pretrained LBANN VAE model

  samadejacobs

  committed
  9fbfefb

  Copy full SHA for 9fbfefb

• Code clean up

  samadejacobs

  committed
  129d9e3

  Copy full SHA for 129d9e3

• Add small version of Zinc train and test data

  samadejacobs

  committed
  8a01982

  Copy full SHA for 8a01982

• Add ckpt for zinc10K

  samadejacobs

  committed
  0035ff7

  Copy full SHA for 0035ff7

Commits on Sep 24, 2020

• Support renaming of weights in LBANN, support evaluation of both reconstructed and sampled SMILES strings

  samadejacobs

  committed
  1fda9a0

  Copy full SHA for 1fda9a0

• Draft implementation of molecular transformer (GPT2) model

  samadejacobs

  committed
  1871098

  Copy full SHA for 1871098

Commits on Sep 27, 2020

• Add gpt to train and sample drivers

  samadejacobs

  committed
  648766d

  Copy full SHA for 648766d