Yakov Pechersky

Does https://github.com/tensorflow/tensorflow/issues/5343 help you solve the issue? Make sure your cuda is installed in an accessible way (possibly via conda).

I'd be interested to recapitulate GDB-n as well.

That's exactly right. Have you taken a look at the Gomez-Bombarelli paper linked in the README? On Wed, May 17, 2017 at 6:54 AM, hkmztrk wrote: > Hello, > >...

Those reasons are why we developed `sample_gen.py` and `train_gen.py`, which generate the one-hot representations on the fly and can train the model using the `fit_generator` functionality of Keras. To train...

The h5 files is expected to have two groups: "data_train" and "data_test". You might want to do something like this, from the repo's `preprocess.py`: ``` train_idx, test_idx = map(np.array, train_test_split(structures.index,...

What I mean is that you have to split your dataset of coulomb matrices into a train set and a test set. A helper function to do that is the...

That was the latent dimension that was reported in the Gomez-Bombarelli paper that is referenced in the README. On Tue, Jan 24, 2017 at 4:37 AM, jeffrey9909 wrote: > Just...

I'd be glad to work on this with you. I agree, it looks like they overfit their model to generate similar strings. This is especially evident in the fact that...

What are you using as the discriminator? SMILES validity or just absence of (possibly invalid) string from training set? On Tue, Jan 17, 2017 at 12:16 PM, Max Hodak wrote:...

One could do SMILES validity AND number of heavy atoms is I'm using presence/absence from the set. SMILES validity is an arguably > even easier metric, as CCCCCcccccccccccccccccccccccccccccccccc is valid...