Rare find

LLM.C. Train AI models in plain C instead of heavy frameworks.

github.com/karpathy/llm.c

Problem

The full read on this project is a sign-in away — what it solves, who it serves, and where it sits in the market.

Audience

The full read on this project is a sign-in away — what it solves, who it serves, and where it sits in the market.

Market

The full read on this project is a sign-in away — what it solves, who it serves, and where it sits in the market.

Pricing

The full read on this project is a sign-in away — what it solves, who it serves, and where it sits in the market.

Vaya's read on this project

Problem, audience, market, and the verdict — sign in to see it.

Updates

May 2025

Merge pull request #801 from ngc92/ngc92/fix-test May 10

April 2025

adjusted error thresholds Apr 13
reorder checks to match actual tensor order Apr 13

October 2024

Merge pull request #766 from msaroufim/patch-1 Oct 2

September 2024

Update README.md Sep 24
cuda mode -> gpu mode Sep 24
change default params: use tinyshakespeare and decrease LR Sep 13

August 2024

Merge pull request #724 from GaoYusong/master Aug 26
Merge pull request #733 from zhangpiu/feature/llm.cpp Aug 26
Merge pull request #735 from gordicaleksa/minor_refactor3 Aug 26
Merge pull request #744 from dengl11/pr Aug 16
Merge pull request #745 from karpathy/feature/managed2 Aug 16
i misspelled reduced Aug 16
reduce across GPUs nicer Aug 16
fallback to memory allocation of m,v,master_weights on host automatic… Aug 16
fix makefile for multigpu setups Aug 16
hint to host Aug 15
just try to allocate on device; fallback if that fails Aug 15
fall back to cudaMallocManaged for optimizer states if we're out of m… Aug 15
fix a typo Aug 14