pytorch/ignite

Checkpoint stateful handlers and metrics

Aperta

#966 aperta il 22 apr 2020

 (3 commenti) (4 reazioni) (0 assegnatari)Python (602 fork)batch import
enhancementhelp wantedmodule: metrics

Metriche repository

Star
 (4313 stelle)
Metriche merge PR
 (Metriche PR in attesa)

Descrizione

🚀 Feature

Things that are attached to the Engine might have state that would ideally be checkpointed and restored using as part of the Engine's state_dict.

An example is a Checkpoint handler when it has a score_function. Currently, the priorities the Checkpoint class stores is not saved anywhere. It is not able to recover gracefully from failure without manual intervention to parse the checkpoint path names, and directly setting the internals of the class.

Handlers and Metrics should have state_dict and load_state_dict methods (empty by default), and I think it should be possible for these to automatically make it into/restored from the Engine's state_dict when it's attached to an Engine.

Guida contributor