cmn  common_param:   - MTL0    : Synthetic GPU (12288 MiB, 12287 MiB free)
common_memory_breakdown_print: |   - MTL0 (Synthetic GPU)    | 12288 = 12287 + (4385 =  4167 +     133 +      83) +       -4384 |
common_params_fit_impl: projected to use 4385 MiB of device memory vs. 12287 MiB of free device memory
common_params_fit_impl: will leave 7902 >= 1024 MiB of free device memory, no changes needed
common_fit_params: successfully fit params to free device memory
load_tensors:  MTL0_Mapped model buffer size =  4167.21 MiB
llama_kv_cache:       MTL0 KV buffer size =    76.50 MiB
llama_kv_cache:       MTL0 KV buffer size =    57.38 MiB
sched_reserve:       MTL0 compute buffer size =    83.99 MiB
common_memory_breakdown_print: |   - MTL0 (Synthetic GPU)    | 12288 = 7902 + (4385 =  4167 +     133 +      83) +           0 |
srv  llama_server: listening
