pytorch/ignite

Checkpoint stateful handlers and metrics

Open

#966 创建于 2020年4月22日

在 GitHub 查看
 (3 评论) (4 反应) (0 负责人)Python (602 fork)batch import
enhancementhelp wantedmodule: metrics

仓库指标

Star
 (4,313 star)
PR 合并指标
 (平均合并 18天 23小时) (30 天内合并 19 个 PR)

描述

🚀 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.

贡献者指南