3 projects
mooncake-transfer-engine-cuda13
A KVCache-centric Disaggregated Architecture for large-scale LLM inference and training. (CUDA 13 version)
mooncake-transfer-engine
A KVCache-centric Disaggregated Architecture for large-scale LLM inference and training.
mooncake-transfer-engine-non-cuda
A KVCache-centric Disaggregated Architecture for large-scale LLM inference and training. (Non-CUDA version)