LoreonLabsPlatform

Overview

Intelligence

Markets
Builders
Research
Narratives
Ecosystems
Launchpads

Discover

Search
Sources

Other

torchtitan

A native PyTorch Library for large model training

OtherEmerging

Stars

—

Forks

—

Contributors

8

Last push

24mo ago

Recent commits

Latest commits.

Added type annotations and more stylistic changes
040ea1dAndrew Gu24mo ago
Renamed parallel styles for transformer block weights
0929280Andrew Gu24mo ago
compiled RMSNorm
7afe902Tianyu Liu24mo ago
Refactored activation checkpointing
bc3ec02Andrew Gu24mo ago
Made some stylistic changes to `apply_dp`
c7a6a3eAndrew Gu24mo ago

Reordered TP parallel plan to follow execution order

f0ca3e8Andrew Gu24mo ago

Removed `_experimental_support_context_fn_in_torch_utils_checkpoint`

2f5285bAndrew Gu24mo ago

switch float8 logic from Float8DynamicLinear to Float8Linear (#436)

3fca883Vasiliy Kuznetsov24mo ago

Top contributors

Builders behind this project.