FareedKhan-dev / 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, no framework, no GPU.
FareedKhan-dev / kimi-k3-in-c
sqliteai / warp
Run the full 2.78-trillion-parameter Kimi K3 model, DeepSeek V4.1 Flash or GLM-5.3-Flash beyond available RAM by streaming activated weights directly from NVMe. A dependency-free, embeddable C inference engine.
microsoft / Tutel
Tutel MoE: Optimized Mixture-of-Experts Library, Support GLM/Kimi/Gpt/DeepSeek/Qwen using FP8/NVFP4/MXFP4