0.07.804.147 I load_tensors: offloaded 47/47 layers to GPU
0.07.804.153 I load_tensors:        CUDA0 model buffer size = 74114.91 MiB
0.07.804.153 I load_tensors:        CUDA1 model buffer size = 71966.25 MiB
0.16.358.919 I llama_context:  CUDA_Host  output buffer size =     0.59 MiB
0.16.359.638 I llama_kv_cache:      CUDA0 KV buffer size =   240.00 MiB
0.16.359.903 I llama_kv_cache:      CUDA1 KV buffer size =   200.00 MiB
0.16.360.192 I llama_memory_recurrent:      CUDA0 RS buffer size =    77.06 MiB
0.16.360.320 I llama_memory_recurrent:      CUDA1 RS buffer size =    68.50 MiB
0.16.361.050 I llama_kv_cache:      CUDA0 KV buffer size =   180.00 MiB
0.16.361.281 I llama_kv_cache:      CUDA1 KV buffer size =   150.00 MiB
0.16.397.952 I sched_reserve:      CUDA0 compute buffer size =   109.33 MiB
0.16.397.956 I sched_reserve:      CUDA1 compute buffer size =   109.28 MiB
0.16.397.957 I sched_reserve:  CUDA_Host compute buffer size =    16.95 MiB
