EleutherAI/lm-evaluation-harness

Allow Task objects to defer dataset download

Open

#1,558 opened on Mar 11, 2024

View on GitHub
 (3 comments) (1 reaction) (0 assignees)Python (3,306 forks)auto 404
feature requestgood first issuehelp wanted

Repository metrics

Stars
 (12,755 stars)
PR merge metrics
 (Avg merge 15d 7h) (11 merged PRs in 30d)

Description

currently Task object initialization causes the tasks to download their datasets upon initialization.

This is not always desirable, so we should allow users to defer download and perform it manually.

Should add a defer_download: bool = False flag to Task and ConfigurableTask init() methods which, when set to true, has the dataset not downloaded, and allow users to easily call a task.download() method with no args that performs the download.

If users attempt to run the task without first downloading the dataset, we should raise an error.

Contributor guide

Research direction
Examine the current Task and ConfigurableTask init methods to understand dataset download logic. Implement a defer download flag that skips download and add a download() method. Ensure error raised if task run without download.
Tech stack
python
Domain
backend
Issue type
Feature
DifficultyEstimated implementation difficulty for a new contributor, from 1 for very small changes to 5 for expert-level work.
2
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
GitPython
Newbie friendlinessA 1-100 score estimating how approachable this issue is for first-time contributors.
70