pcyin / pytorch_neural_symbolic_machines

A PyTorch Implementation of Neural Symbolic Machines by Liang et al. (2018)

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

Interpret evaluation results

todpole3 opened this issue · comments

Should we find dev evaluation results in runs/demo_run/dev.log?

My log says

{"accuracy": 0.6031746031746031, "oracle_accuracy": 0.7380952380952381}

which seems to be too high for WikiTableQuestions..

The config file is updated, with which I got a reasonable results. You can try it.

Thanks, looks like the original evaluation script points to a slice of training data.