41 feat batch inference for nitro

[tikikun]

** IMPORTANT

In order to enable continous batching (multi threading and multiple ccu) when load model need to enable cont_batching value

example

curl -X POST 'http://localhost:3928/inferences/llamacpp/loadModel' \
     -H 'Content-Type: application/json' \
     -d '{
          "llama_model_path": "/Users/alandao/Documents/codes/nitro.cpp_temp/models/llama2_7b_chat_uncensored.Q4_0.gguf",
          "ctx_len": 2048,
          "ngl": 100,
          "cont_batching": true
     }'