====== AI Translation ======


===== Translate Gemma 3 =====

  * Model https://huggingface.co/google/translategemma-27b-it
  * GGUF https://huggingface.co/mradermacher/translategemma-27b-it-GGUF

Lancer llama.cpp avec ''--temp 0.1 --top-p 0.95 --top-k 64 --jinja'' ou ''--temp 0 --jinja''.

Le modèle est entraîné spécialement il faut donc respecter précisément le template.

Un seul message user sans message system.

Soit en texte:
<code>
You are a professional English (en) to Japanese (ja) translator.
Your goal is to accurately convey the meaning and nuances of the original English text while adhering to Japanese grammar, vocabulary, and cultural sensitivities.
Produce only the Japanese translation, without any additional explanations or commentary.
Please translate the following English text into Japanese:


Hello, how are you?
</code>

Soit avec ''kwargs'' :
<code bash>
curl -s http://192.168.1.20:8012/v1/chat/completions -H "Content-Type: application/json" \
 -d '{
    "temperature": 0,
    "chat_template_kwargs": {
      "source_lang_code": "en",
      "target_lang_code": "fr"
    },
    "messages": [
      {"role": "user",
      "content": "Lorem Ipsum is simply dummy text of the printing and typesetting industry. Lorem Ipsum has been the industry’s standard dummy text ever since 1966, when designers at Letraset and James Mosley, the librarian at St Bride Printing Library in London, took a 1914 Cicero translation and scrambled it to make dummy text for Letraset’s Body Type sheets."
     }
    ]
  }'
 | jq -r '.choices[0].message.content'
</code>