I'm currently incorporating SentEval in my training procedure, so it evaluates intermediate model parameters every x training steps. However I'm now finding myself hacking the code to make it work with Python 3.6, which I need for other components in my model. It would be wonderful if SentEval had out of the box support for Python 3.x.
I'm currently incorporating SentEval in my training procedure, so it evaluates intermediate model parameters every
x
training steps. However I'm now finding myself hacking the code to make it work with Python 3.6, which I need for other components in my model. It would be wonderful if SentEval had out of the box support for Python 3.x.