aidotse/LeakPro

Loss in training loops

Open

#239 opened on Feb 24, 2025

View on GitHub
 (1 comment) (0 reactions) (0 assignees)Python (26 forks)auto 404
good first issue

Repository metrics

Stars
 (23 stars)
PR merge metrics
 (PR metrics pending)

Description

Issue

Problem Description

The last batch of each epoch is usually smaller than the others. One approach is to sum the losses and divide by the total number of samples.

Currently, loss calculations in handlers sum the average loss of each batch and divide by the length of the dataloader.

Expected Behavior

What Needs to be Done

How Can It Be Tested or Reproduced

Contributor guide

Research direction
Investigate the loss calculation in the training handlers and modify it to sum individual batch losses and divide by total number of samples instead of averaging batch losses and dividing by number of batches.
Tech stack
python
Domain
machine learning
Issue type
Bug
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.
Active
ClarityHow clearly the issue explains the expected change, acceptance criteria, and next step.
Clear
Prerequisites
Python
Newbie friendlinessA 1-100 score estimating how approachable this issue is for first-time contributors.
70