4 ms·
The ipex-llm implementation extends llama.cpp and includes additonal CPU-GPU hybrid optimizations for sparse MoE
by colorant 2y ago
The ipex-llm implementation extends llama.cpp and includes additonal CPU-GPU hybrid optimizations for sparse MoE