llama_model_load: offloaded 33/33 layers to GPU
common_params_fit_impl: projected to use 6912 MiB of device memory vs. 5324 MiB of free device memory
common_params_fit: failed to fit params to free device memory: n_gpu_layers already set by user to 99, abort
load_tensors: Vulkan0 model buffer size = 3072.00 MiB
llama_kv_cache: Vulkan0 KV buffer size = 3584.00 MiB
llama_context: Vulkan0 compute buffer size = 256.00 MiB
