machine-reasoning-ufrgs / gnn-gcp Goto Github PK

Hi

I wanted to run training using the cmd line you share but I keep getting this error. Can you please help?

`python3 dataset.py -path adversarial-training --train

Creating 32768 instances
0.0% Complete
Exception('CSP is unsure about the problem')
`

Hi

Unfortunately, the model does not reproduce the results in the paper.
I wanted to train it on a larger dataset, but the data-generation script takes an extremely long time (would need more than 1 week).

Is there a way for you to upload the pre-trained weights? Do you have any two cents as to why the data generation is taking so long?

I can't download pysat=0.1.3. After downloading pysat=0.1.7, the pysolvers function in solvers.py in the pysat package (as in other versions) cannot be successfully imported. How should I deal with this situation? Can you share the package pysat=0.1.3 with us

Hello,

Can you upload the pre-trained weights?

Thanks