EpochOutputStore can optionally store the data to the state
See original GitHub issue🚀 Feature
The idea is to give an option to have the following:
eos = EpochOutputStore()
evaluator = create_supervised_evaluator(model, metrics)
eos.attach(evaluator, name="eval_data")
# Adding the name is equivalent to
# name = "eval_data"
# @evaluator.on(Events.EPOCH_COMPLETED, name)
# def store_data(engine, name):
# setattr(engine.state, name, eos.data)
state = evaluator.run(data)
print(len(state.eval_data))
In addition, it would make sense to move it from contrib to core as well, keeping BC.
Issue Analytics
- State:
- Created 2 years ago
- Reactions:1
- Comments:6 (4 by maintainers)
Top Results From Across the Web
EpochOutputStore — PyTorch-Ignite v0.4.10 Documentation
EpochOutputStore handler to save output prediction and target history after every ... If name is passed, will store self.data on engine.state under name...
Read more >High-level library to help with training neural networks in ...
The first argument can be optionally engine , but not necessary. Thus, we do not require to inherit from an interface and override...
Read more >
Top Related Medium Post
No results found
Top Related StackOverflow Question
No results found
Troubleshoot Live Code
Lightrun enables developers to add logs, metrics and snapshots to live code - no restarts or redeploys required.
Start Free
Top Related Reddit Thread
No results found
Top Related Hackernoon Post
No results found
Top Related Tweet
No results found
Top Related Dev.to Post
No results found
Top Related Hashnode Post
No results found
I opened #1974 but it is probably quite basic, happy to implement any changes that might be needed 🙂
This is only a draft - as a next step, will move it out of contrib keeping BC.
@vfdev-5 - would love to! Will read through the code today and either come here to ask questions if anything is unclear or will start working on a draft of the implementation 🙂