LoreonLabsPlatform
DocsHome
  • Overview

Intelligence

  • Markets
  • Builders
  • Research
  • Ecosystems
  • Launchpads
  • Search
Ecosystems

Python

torchtitan

A PyTorch native platform for training generative AI models

PythonEmerging
GitHub
Stars
—
Forks
—
Contributors
8
Last push
2mo ago

Recent commits

Latest commits.

  • [rl] Add batched RL training with varlen sequence packing
    89d9a1ajoecummings2mo ago
  • [RL] Fix RL h100 workflow by installing FA3 and add explicit enable_gqa flag (#2891)
    8328876Jiani Wang2mo ago
  • fix(data-ckpting-resume): fix broken shuffling (#2865)
    a00ef7dSlim Frikha2mo ago
  • [GraphTrainer][AutoDev] Enable cudagraph in aot_fx_trace mode (#2868)
    d98fa54Sherlock Huang2mo ago
  • [compile] Share SimpleFSDP wrapper class across same-type module instances (#2754)
    e43f252Animesh Jain2mo ago
  • [graph_trainer] Add graph module dumping section to CLAUDE.md (#2902)
    2c6bb1cSherlock Huang2mo ago
  • [graph_trainer] Enable precompile with regional_inductor (#2883)
    4a57715Bob Ren2mo ago
  • Add A6000 and B300 peak FLOPS to get_peak_flops (#2884)
    ff95477Johnsonms2mo ago
  • Top contributors

    Builders behind this project.

    tianyu-l
    173 commits
    fegin
    138 commits
    wwwjn
    99 commits
    lessw2020
    63 commits
    wconstab
    58 commits
    wanchaol
    51 commits
    yiming0416
    42 commits
    H-Huang
    40 commits