A PyTorch implementation of the GPT-OSS-20B architecture. All components are coded from scratch: RoPE with YaRN, RMSNorm, SwiGLU with clamping and residual connection, Mixture-of-Experts (MoE), Self-Attention with learned sinks, banded attention, GQA, and KV-cache.
☆237Dec 2, 2025Updated 9 months ago
Alternatives and similar repositories for gpt-oss-20B
Users that are interested in gpt-oss-20B are comparing it to the libraries listed below. We may earn a commission when you buy through links labeled 'Ad' on this page.
Sorting:
- A series of high-performance GEMM (General Matrix Multiply) implementations Iteratively optimised for H100 GPUs in Pure CUDA.☆82Feb 18, 2026Updated 7 months ago
- A GPT-2 inference engine written from scratch in CUDA and C++. Implements custom CUDA kernels for tiled matrix multiplication, LayerNorm,…☆44May 17, 2026Updated 4 months ago
- Course on Flash-attention in Triton☆104Feb 9, 2026Updated 7 months ago
- ☆19Mar 3, 2025Updated last year
- Companion code for The Physics of LLM Inference book☆29Apr 21, 2026Updated 5 months ago
- Bare Metal GPUs on DigitalOcean Gradient AI • AdPurpose-built for serious AI teams training foundational models, running large-scale inference, and pushing the boundaries of what's possible.
- ☆15Mar 11, 2026Updated 6 months ago
- ☆19Mar 29, 2026Updated 5 months ago
- Qwen3-0.6B megakernel: 527 tok/s decode on RTX 3090 (3.8x faster than PyTorch)☆139Feb 10, 2026Updated 7 months ago
- we have ai at home☆119Jun 18, 2026Updated 3 months ago
- Row-wise block scaling for fp8 quantization matrix multiplication. Solution to GPU mode AMD challenge.☆19Feb 9, 2026Updated 7 months ago
- coding CUDA everyday!☆79Feb 5, 2026Updated 7 months ago
- Pure Triton kernels for Qwen3.5-27B inference on NVIDIA B200☆123Feb 28, 2026Updated 6 months ago
- 🌱 A little course on Reinforcement Learning Environments for evaluating and training Language Models☆226May 27, 2026Updated 3 months ago
- a personal collection of my notes for ml sys☆124Sep 1, 2026Updated 3 weeks ago
- Managed Database hosting by DigitalOcean • AdPostgreSQL, MySQL, MongoDB, Kafka, Valkey, and OpenSearch available. Automatically scale up storage and focus on building your apps.
- ☆133Dec 9, 2025Updated 9 months ago
- mHC kernels implemented in CUDA☆267Mar 9, 2026Updated 6 months ago
- a LLM inference engine to run on consumer hardware☆48Apr 15, 2026Updated 5 months ago
- Winner 🏆 (Agent-only) MLSys 2026 - FlashInfer AI Kernel Generation Contest for the DeepSeek Sparse Attention (DSA) track with an average…☆309Sep 12, 2026Updated last week
- ☆45May 4, 2025Updated last year
- SQL-like query language and CLI for Qdrant vector search engine☆47Jun 13, 2026Updated 3 months ago
- Hand-Rolled GPU communications library☆94Nov 25, 2025Updated 9 months ago
- An educational distributed training and inference library for neural nets using local computing☆84Jun 10, 2026Updated 3 months ago
- A compact implementation of SGLang, designed to demystify the complexities of modern LLM serving systems.☆5,147May 17, 2026Updated 4 months ago
- Deploy to Railway using AI coding agents - Free Credits Offer • AdUse Claude Code, Codex, OpenCode, and more. Autonomous software development now has the infrastructure to match with Railway.
- RVAA: Recursive Vision-Action Agent for Long Video Understanding. Implementation of the RLM paradigm (Zhang, Kraska, Khattab 2025)☆118Jan 15, 2026Updated 8 months ago
- a fun and educational take on vLLM☆228Jan 25, 2026Updated 7 months ago
- An extension of the nanoGPT repository for training small MOE models.☆286Mar 9, 2025Updated last year
- minimal pytorch 4D parallelism☆73Feb 16, 2026Updated 7 months ago
- ☆35Nov 11, 2025Updated 10 months ago
- Personal CUDA learning repo, built step by step from scratch.☆107Jun 14, 2026Updated 3 months ago
- ☆390May 4, 2026Updated 4 months ago
- MetaAttention: A Unified and Performant Attention Framework Across Hardware Backends(PPoPP'26)☆18Aug 6, 2026Updated last month
- A repository consisting of paper/architecture replications of classic/SOTA AI/ML papers in pytorch☆427Nov 11, 2025Updated 10 months ago
- Deploy open-source AI quickly and easily - Special Bonus Offer • AdRunpod Hub is built for open source. One-click deployment and autoscaling endpoints without provisioning your own infrastructure.
- ☆179May 29, 2026Updated 3 months ago
- My submission for the GPUMODE/AMD fp8 mm challenge☆29Jun 4, 2025Updated last year
- Tutorials of Extending and importing TVM with CMAKE Include dependency.☆16Oct 11, 2024Updated last year
- Implement a reasoning LLM in PyTorch from scratch, step by step☆5,279Updated this week
- dLLM: Simple Diffusion Language Modeling☆2,692Jul 17, 2026Updated 2 months ago
- ☆111Oct 16, 2025Updated 11 months ago
- 100 days of building GPU kernels!☆628Apr 27, 2025Updated last year