Shadowgraph
About
Sign in

joshcarp/llm.go

83

stars

21

commits

Jupyter Notebook

primary language

Mar 21, 2026

updated

README

llm.go

Status GitHub Issues GitHub Pull Requests License

GPT-2 implementation written in go only using the standard library.

🪞 Quick start

Install python dependencies, output tokenized dataset

make setup

Run the training script:

make train

This will run go run ./cmd/traingpt2/main.go

Run the testing script:

make test

This will run go run ./cmd/testgpt2/main.go

TODO

  • Tokenize input text, the implementation of this is incorrect. Need to do pair matching not tries
  • Very slow, need to improve performance.
  • It runs in WASM but using WebGPU bindings might be fun.
  • More refactoring.
  • Running as CLI.

🖋️ License

See LICENSE for more details.

🎉 Acknowledgements

  • This is a fork of Andrej Karpathy's llm.c written in pure go.

Contributors

joshcarp

18 commits

albertpurnama

3 commits

Similar repos

Saimirbaci/llm.zig

54

nikolaydubina/llama2.go

LLaMA-2 in native Go

192

ezhoureal/myGPT

From-scratch LLM systems reimplementation for learning (tokenizer, GPT, manual autograd)

0

cookiengineer/gonano

:construction: Experimental port of Karpathy's nanochat to Go with SIMD

1

gevtushenko/llm.c

LLM training in simple, raw C/CUDA

113

karpathy/llm.c

LLM training in simple, raw C/CUDA

30,938

otabuzzman/llm.swift

A Swift port of Andrej Karpathy‘s llm.c.

28

anthonix/llm.c

LLM training in simple, raw C/HIP for AMD GPUs

61

Languages

Jupyter Notebook

61.0%

Go

22.5%

Python

12.2%

JavaScript

4.3%

joshcarp/llm.go

83

stars

21

commits

Jupyter Notebook

primary language

Mar 21, 2026

updated

README

llm.go

Status GitHub Issues GitHub Pull Requests License

GPT-2 implementation written in go only using the standard library.

🪞 Quick start

Install python dependencies, output tokenized dataset

make setup

Run the training script:

make train

This will run go run ./cmd/traingpt2/main.go

Run the testing script:

make test

This will run go run ./cmd/testgpt2/main.go

TODO

  • Tokenize input text, the implementation of this is incorrect. Need to do pair matching not tries
  • Very slow, need to improve performance.
  • It runs in WASM but using WebGPU bindings might be fun.
  • More refactoring.
  • Running as CLI.

🖋️ License

See LICENSE for more details.

🎉 Acknowledgements

  • This is a fork of Andrej Karpathy's llm.c written in pure go.

Similar repos

Saimirbaci/llm.zig

54

nikolaydubina/llama2.go

LLaMA-2 in native Go

192

ezhoureal/myGPT

From-scratch LLM systems reimplementation for learning (tokenizer, GPT, manual autograd)

0

cookiengineer/gonano

:construction: Experimental port of Karpathy's nanochat to Go with SIMD

1

gevtushenko/llm.c

LLM training in simple, raw C/CUDA

113

karpathy/llm.c

LLM training in simple, raw C/CUDA

30,938

otabuzzman/llm.swift

A Swift port of Andrej Karpathy‘s llm.c.

28

anthonix/llm.c

LLM training in simple, raw C/HIP for AMD GPUs

61

Contributors

joshcarp

18 commits

albertpurnama

3 commits

Languages

Jupyter Notebook

61.0%

Go

22.5%

Python

12.2%

JavaScript

4.3%