Lightning-AI / Lightning-AI/lightning-thunder

Add trace dataflow correctness tests

Open
#2,145 0 comments 0 reactions 0 assignees View on GitHub
enhancement testing
Dominant language
Python
Stars
1.5k
Forks
121
PR merge metrics
No merged PRs in 30d

Description

## 🚀 Feature

As one of the core principles of Thunder is having a dataflow from input to output, there should be some tests to verify that the transformations respect that.

Simple example of test might be to run dce on the trace we wish to check. If the trace is not runnable anymore after dce, then the dataflow was not correct.

cc @mruberry

Contributor guide

No contributing guide indexed for this repository

Research direction

The issue names no files or existing tests; start by locating the trace transformation tests and the DCE entry point. Define the checks around whether transformed traces remain runnable after DCE, and consider the work done when tests cover dataflow correctness across the relevant transformations.

Written by the indexing model from the issue text.

Assessment

Tech stack
python
Domain
compilers, testing
Issue type
Feature
Difficulty
4/5
Estimated time
3-5 days
Activity status
Stale
Clarity
Needs clarification
Newbie friendliness
35/100

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.