Pent/distributed-llama

Run LLMs on weak devices or make powerful devices even more powerful by distributing the workload and dividing the RAM usage.

C++MIT