nanoeuler

GPT-2-style LLM built from scratch in C/CUDA with hand-written backprop, BPE tokenizer, FlashAttention, pretraining, and SFT.

JustVugg/nanoeuler

View on GitHub ↗

GPT-2-style LLM built from scratch in C/CUDA with hand-written backprop, BPE tokenizer, FlashAttention, pretraining, and SFT.

nanoeuler grew +0.8% in the last 30 days, ranks 95th of 113 in Pretraining & Training, last commit 157 days ago.

Star trajectory

82 days tracked
2026-08-31127 stars today2026-11-22
Momentum

+0.8% in 30d, above the Pretraining & Training median of +0.3%.

Maintenance

Last commit 157 days ago, staler than the median of 103d.

How it compares in Pretraining & Training

nanoeuler
127
stars
+0.8% 30d
commit 157 days ago
Pretraining & Training median
342
stars
+0.3% 30d
commit ~103d ago
Leader · maxtext
2.5k
stars
+1.0% 30d
commit today

Ranked 95th of 113 by stars.

Ranked by shared tags, then keywords, then stars
  1. GPU 性能与 AI Infra 学习项目:CUDA/Triton 算子、NCU/NSYS、vLLM/SGLang/TRT-LLM/ms-swift、PyTorch/DeepSpeed/ms-swift 训练、并行架构

    +2 (+1.6%) in 30d · Committed 75d ago · Python · Apache-2.0

    Shares Training Toolkits, GPU Kernels, Deep Learning Frameworks

  2. Pure Rust implementation of a minimal Generative Pretrained Transformer

    +0 (+0.0%) in 30d · Committed 396d ago · Rust · MIT

    Shares Pretraining & Training, Deep Learning Frameworks, GPU Kernels

  3. Omega-AI is a Java-based deep learning framework that helps you quickly build neural networks for inference…

    +1 (+0.1%) in 30d · Committed 22d ago · Java · Apache-2.0

    Shares Deep Learning Frameworks, Pretraining & Training, Training Toolkits

  4. Quantized LLM training in pure CUDA/C++.

    +3 (+1.1%) in 30d · Committed 8d ago · C++ · Apache-2.0

    Shares Pretraining & Training, Training Toolkits, GPU Kernels

  5. GleamLM大语言模型全栈训练平台——核心代码用 PyTorch 从零实现,最低仅需 12GB 消费级显卡即可完整训练,另有一套工业代码与手写实现消费同一份数据逐层对照。模型训练支持CLI和可视化训练。

    +1 (+1.5%) in 30d · Committed 65d ago · Python · Apache-2.0

    Shares Pretraining & Training, Training Toolkits, Deep Learning Frameworks

  6. The most atomic way to train and inference a GPT in pure, dependency-free C

    +5 (+0.6%) in 30d · Committed 96d ago · C · MIT

    Shares Pretraining & Training, GPU Kernels, Training Toolkits

  7. seqax = sequence modeling + JAX

    -1 (-0.5%) in 30d · Committed 487d ago · Python · BSD-3-Clause

    Shares Pretraining & Training, Training Toolkits, Deep Learning Frameworks

  8. Following Karpathy with GPT-2 implementation and training, writing lots of comments cause I have memory of a…

    +0 (+0.0%) in 30d · Committed 843d ago · Jupyter Notebook

    Shares Pretraining & Training, Training Toolkits, Deep Learning Frameworks

  9. 🦀 Decoder-only LLM built from scratch in pure Rust using Candle — no Python, no PyTorch. Gated DeltaNet +…

    +2 (+1.8%) in 30d · Committed 92d ago · Rust · Apache-2.0

    Shares Pretraining & Training, Training Toolkits, Deep Learning Frameworks

  10. A high-performance framework for training LLMs, VLMs, diffusion, and embodied models on NVIDIA GPUs and Kunlun…

    Committed yesterday · Python · Apache-2.0

    Shares Training Toolkits, Pretraining & Training

Language
Cuda
License
MIT License
Contributors
1
Forks
26
Watchers
3
Open issues
0
Releases
0
Created
2026-06-15
Stars this week
+0

You've seen where nanoeuler stands today.

Get the moves before everyone else: the 5 Pretraining & Training repos climbing fastest, every Tuesday.

Free weekly digest. Double opt-in, unsubscribe anytime.

Maintain nanoeuler?

Claim this page to add your own call-to-action.

Claim this page