pytorch/ignite

Warning using ConfusionMatrix and setup_tb_logging

Closed

#829 opened on Mar 4, 2020

View on GitHubContributor guide
 (2 comments) (0 reactions) (0 assignees)Python (602 forks)batch import
enhancementhelp wantedmodule: metrics

Repository metrics

Stars
 (4,313 stars)
PR merge metrics
 (Avg merge 18d 23h) (19 merged PRs in 30d)

Description

❓ Questions/Help/Support

logger configured by setup_tb_logging warns when a ConfusionMatrix metric is attached to the evaluator

tensorboard_logger.py:129: UserWarning: TensorboardLogger output_handler can not log metrics value type <class 'torch.Tensor'>
  "TensorboardLogger output_handler can not log " "metrics value type {}".format(type(value))

The reason is setup_tb_logging (via setup_any_logging) defines OutputHandler using metric_names="all" https://github.com/pytorch/ignite/blob/a22a0f5e909ac70d2a1f76a60b6e84b2134f196c/ignite/contrib/engines/common.py#L248-L254

What do you think to expose metric_names in the API to be able to fix the warning ?

Contributor guide

Research direction
Examine the setup tb logging function in ignite/contrib/engines/common.py (lines 248-254). The OutputHandler is instantiated with metric names="all", causing a warning when ConfusionMatrix returns tensors. Consider adding a parameter to setup tb logging to allow users to specify metric names, thereby avoiding the warning. Ensure backward compatibility and check if similar changes are needed in setup any logging.
Tech stack
python
Domain
machine learning
Issue type
Feature
DifficultyEstimated implementation difficulty for a new contributor, from 1 for very small changes to 5 for expert-level work.
3
Estimated timeA rough time range for an experienced contributor to investigate, implement, test, and prepare a pull request.
1-3 hours
Activity statusHow available the issue appears right now: fresh, active, stale, blocked, or waiting on maintainer input.
Stale
ClarityHow clearly the issue explains the expected change, acceptance criteria, and next step.
Clear
Prerequisites
PythonPyTorchIgnite basics
Newbie friendlinessA 1-100 score estimating how approachable this issue is for first-time contributors.
70