LoreonLabsPlatform

Overview

Intelligence

Markets
Builders
Research
Narratives
Ecosystems
Launchpads

Discover

Search
Sources

Other

benchmark

TorchBench is a collection of open source benchmarks used to evaluate PyTorch performance.

OtherEmerging

Stars

—

Forks

—

Contributors

8

Last push

21mo ago

Recent commits

Latest commits.

Fix a metrics reporting issue for FP8 row-wise
10e4bc4Hongtao Yu24mo ago
Always run baseline benchmark first, even with --only
95b4202Jez Ng24mo ago
Use cuda graphs for benchmarking
b91caadJez Ng24mo ago
Extend support for x_only = True extra metrics through canonical name retrieval (#2300)
86904caJanani Sriram24mo ago
Add jagged_sum operator for unpadded nested tensors to TritonBench (#2299)
576b2b2Janani Sriram24mo ago

Change the llama max_batch_size larger than default eval batch size (#2283)

a7ae1fbleslie-fang-intel24mo ago

Change Dynamo's custom ops warning message to be less spammy (#128456)

ab687a0rzou (Meta Employee)24mo ago

enable fx graph cache on torchbench (#128239)

04a1339Sam Larsen (Meta Employee)25mo ago

Top contributors

Builders behind this project.