Prototype calculating floating point operations

[brendan-ai2]

Is your feature request related to a problem? Please describe.
Please see the Green AI paper for full context.

Describe the solution you'd like
A tool which computes FPO for a subset of the PyTorch operations which are used in the AllenNLP models. This will likely take the form of a small library that determines how many FPOs are used by a module's forward method. Initially perhaps called by allennlp benchmark-fpo, later integrated into allennlp train.

[github-actions]

This issue is being closed due to lack of activity. If you think it still needs to be addressed, please comment on this thread 👇