Do I use full power of my server?

[IAkumaI]

I'm using localAI on VPS with GPU RTX A4000.

I have tried different images of localAI docker image, different model, but whatever I do i get messages like this:

8:18PM INF Trying to load the model 'Meta-Llama-3-8B-Instruct.Q6_K.gguf' with the backend '[llama-cpp llama-ggml gpt4all llama-cpp-fallback piper rwkv stablediffusion whisper huggingface bert-embeddings /build/backend/python/petals/run.sh /build/backend/python/parler-tts/run.sh /build/backend/python/exllama/run.sh /build/backend/python/rerankers/run.sh /build/backend/python/bark/run.sh /build/backend/python/openvoice/run.sh /build/backend/python/sentencetransformers/run.sh /build/backend/python/transformers/run.sh /build/backend/python/sentencetransformers/run.sh /build/backend/python/diffusers/run.sh /build/backend/python/exllama2/run.sh /build/backend/python/transformers-musicgen/run.sh /build/backend/python/vllm/run.sh /build/backend/python/mamba/run.sh /build/backend/python/vall-e-x/run.sh /build/backend/python/coqui/run.sh /build/backend/python/autogptq/run.sh]'
8:18PM INF [llama-cpp] Attempting to load
8:18PM INF Loading model 'Meta-Llama-3-8B-Instruct.Q6_K.gguf' with backend llama-cpp
8:18PM INF GPU device found but no CUDA backend present
8:18PM INF [llama-cpp] attempting to load with AVX2 variant
8:18PM INF [llama-cpp] Loads OK

The most interesting line is "GPU device found but no CUDA backend present". Does it mean something bad?
In fact everything works with a fair speed.

nvidia-smi shows GPU is using:

+-----------------------------------------------------------------------------------------+
| NVIDIA-SMI 550.90.07              Driver Version: 550.90.07      CUDA Version: 12.4     |
|-----------------------------------------+------------------------+----------------------+
| GPU  Name                 Persistence-M | Bus-Id          Disp.A | Volatile Uncorr. ECC |
| Fan  Temp   Perf          Pwr:Usage/Cap |           Memory-Usage | GPU-Util  Compute M. |
|                                         |                        |               MIG M. |
|=========================================+========================+======================|
|   0  NVIDIA RTX A4000               Off |   00000000:07:00.0 Off |                  Off |
| 49%   57C    P8             10W /  140W |    7792MiB /  16376MiB |      0%      Default |
|                                         |                        |                  N/A |
+-----------------------------------------+------------------------+----------------------+

+-----------------------------------------------------------------------------------------+
| Processes:                                                                              |
|  GPU   GI   CI        PID   Type   Process name                              GPU Memory |
|        ID   ID                                                               Usage      |
|=========================================================================================|
|    0   N/A  N/A     80113      C   .../backend-assets/grpc/llama-cpp-avx2       7786MiB |
+-----------------------------------------------------------------------------------------

The questions is Does it mean something goes wrong and I have to fix it or it's just fine? Or I should left it as is and I can not speed it up more?