llama.cpp

convert : write tensors in parallel

#12837

Open