Skip to content

Reference implementations of MLPerf™ inference benchmarks

active 2025-03-092026-08-07 (UTC)

Partial coverage11,500 / 12,498 hourly files (92%) · 2 absent upstream · 994 failed, retryable2025-03-082026-08-10 (UTC)— sampled evenly across the window, so rankings and trends hold; absolute counts scale up.
Events
2.2K
Pushes
851
Pull requests
186
Issues
107
Stars
117
Forks
37

Activity over time

Daily event counts in the loaded window

Line chart, 517 days from 2025-03-09 to 2026-08-07. Pushes: 851 total, peak 34 in a day. Pull requests: 186 total, peak 8 in a day. Issues: 107 total, peak 7 in a day. Comments: 516 total, peak 17 in a day. Stars: 117 total, peak 3 in a day.

  • Pushes
  • Pull requests
  • Issues
  • Comments
  • Stars

Stars, PRs, issues and forks are under-captured in the later part of this window. GH Archive progressively stopped capturing non-push events during 2026 — −95% or worse by the end of the window. Every series here except Pushes fades for that reason, so a decline above reflects the archive, not this repository. Pushes stay reliable throughout, so read them, and the contributor counts derived from them, as the real signal. Data health has the measurements.

Top contributors

Pushes, PRs, issues, reviews and comments — stars and forks excluded, so this is contribution rather than popularity

ContributorContributionsPushesPRsComments
github-actions[bot]4813200127
arjunsuresh2541083741
pgmpablo1573212011212714
mlcommons-bot16416400
anandhu-eng125413731
hanyunfan9130927
wangshangsam740846
nvzhihanj670044
mrmhodak321516
viraatc310515
johncalesp230114
keithachorn-intel220612
arav-agarwal218872
manpreetssokhi15735
Agalakdak13007
tanvi-mlcommons12024
wu6u3tw11044
LinjianMa10015
attafosu9016
v-shobhit9015

Recent activity

Latest issues, pull requests and releases

  • Issue comment#1780github-actions[bot]2026-07-24 00:31
    Key Error while running NVIDIA LLAMA2 70b
  • Issue comment#2614github-actions[bot]2026-07-02 14:01
    Remove files not produced by endpoints
  • Issue comment#938github-actions[bot]2026-05-16 00:40
    Add BERT-large details to the README
  • Issue comment#1283github-actions[bot]2026-05-16 00:40
    Add CodeQL workflow
  • Issue#1298github-actions[bot]2026-05-15 00:43
    Publish official prebuilt whl files for python bindinds
  • Issue comment#1298github-actions[bot]2026-05-15 00:43
    Publish official prebuilt whl files for python bindinds
  • Issue comment#1329github-actions[bot]2026-05-15 00:42
    Submission checker error with 3d-unet
  • Issue#1332github-actions[bot]2026-05-15 00:42
    Latest OpenImages dataset annotations differ from those generated by previous fiftyone based script
  • Issue comment#1332github-actions[bot]2026-05-15 00:42
    Latest OpenImages dataset annotations differ from those generated by previous fiftyone based script
  • Issue comment#1334github-actions[bot]2026-05-15 00:42
    Update loadgen.cc
  • Issue comment#1344github-actions[bot]2026-05-15 00:42
    DLRM Inference fails:KeyError
  • Issue#1384github-actions[bot]2026-05-13 00:44
    GPT-J: Dataset issues
  • Issue#1383github-actions[bot]2026-05-13 00:44
    GPT-J: missing checkpoint and confusing README
  • Issue comment#2576araghun2026-05-12 22:16
    Power measurement proposal #1 - Nameplate Power
  • Issue comment#1433github-actions[bot]2026-05-12 00:39
    TEST04 takes reciprocal of scores for SingleStream/MultiStream
  • Issue#1433github-actions[bot]2026-05-12 00:39
    TEST04 takes reciprocal of scores for SingleStream/MultiStream
  • Issue#1488github-actions[bot]2026-05-12 00:39
    Is it possible to allow resubmission from previous round?
  • Issue#1494github-actions[bot]2026-05-11 00:43
    Offline scenario inferred from SingleStream gives incorrect power efficiency
  • Issue#1492github-actions[bot]2026-05-11 00:43
    Is it possible to calculate the LLM performance, Token per Second ?
  • Issue comment#1492github-actions[bot]2026-05-11 00:43
    Is it possible to calculate the LLM performance, Token per Second ?
  • Issue#1502github-actions[bot]2026-05-11 00:43
    why onnx.load("resnet50_v1.pth") when I use pytorch backend?
  • Issue comment#1502github-actions[bot]2026-05-11 00:43
    why onnx.load("resnet50_v1.pth") when I use pytorch backend?
  • Issue#1502github-actions[bot]2026-05-11 00:43
    why onnx.load("resnet50_v1.pth") when I use pytorch backend?
  • Issue#1510github-actions[bot]2026-05-11 00:43
    Request NVIDIA to make container BASE IMAGE links the publicly available links
  • Issue#1520github-actions[bot]2026-05-10 00:42
    Open division rule on validation dataset is confusing

Totals cover only the window loaded into ClickHouse and count events, not GitHub's lifetime totals — 117 stars here means stars gained during the window, not the repo's star count.