mixture-of-experts
16 projets partagent ce topic GitHub
Les traits relient les membres réellement apparentés entre eux. La taille des points suit les étoiles.
🧬 Membres
kimi-k3-in-c
A 2.78-trillion-parameter Kimi K3 running inference on a single CPU in 8.24 GB of RAM. Portable C99: no BLAS,…
★ 9.2k
ai-infra-book
《深入理解 AI Infra：量化分析与系统设计》（李博杰…
★ 7.1k · GitHub ↗
optillm
Optimizing inference proxy for LLMs
★ 4.3k
mixtral-offloading
Run Mixtral-8x7B models in Colab or consumer desktops
★ 2.3k
project-maya
Project Maya - run GLM-5.3-Flash (321B MoE), DeepSeek-V4.1-Flash-1017 (748B MoE) and MiniMax-M3.1-Flash (428B…
★ 2.1k · GitHub ↗
halogen-flash-server
Local inference that behaves like a frontier API: Qwen3.8-Flash-Next on AMD Strix Halo plus small NPU models,…
★ 1.1k · GitHub ↗
Aria
Codebase for Aria - an Open Multimodal Native MoE
★ 1.1k
makeMoE
From scratch implementation of a sparse mixture of experts language model inspired by Andrej Karpathy's…
★ 820
Chinese-Mixtral
中文Mixtral混合专家大模型（Chinese Mixtral MoE LLMs）
★ 613
krasis
Krasis is a Hybrid LLM runtime which focuses on efficient running of larger models on consumer grade VRAM…
★ 522
mergoo
A library for easily merging multiple LLM experts, and efficiently train the merged LLM.
★ 515
MoE-Infinity
PyTorch library for cost-effective, fast and easy serving of MoE models.
★ 363
soft-moe-pytorch
Implementation of Soft MoE, proposed by Brain's Vision team, in Pytorch
★ 348
Project_Chronos
⚡ Zero-Stall MoE Inference via Lookahead Prediction & Async DMA Prefetching. Optimized for SSD I/O with…
★ 207
llama.cpp-v100
llama.cpp optimized for NVIDIA Tesla V100 (SM70): Qwen3.8-Flash-Next with MTP (prefill 3402 tok/s at 200K…
★ 59 · GitHub ↗
neo-unify
Toy-scale unified multimodal model experiments — encoder-free understanding & generation with…
★ 51
🔗 Familles voisines
Mesuré à partir des topics GitHub communs aux deux projets, pondérés par leur rareté.