Views
No views yet
make dllama
./dllama chat --model model.m --tokenizer tokenizer.t --buffer-float-type q80 --nthreads 4 --max-seq-len 8192