Lightnovel Translate Qwen2.5 32B GGUF

🚀 Qwen2.5 32B for Japanese to English Light Novel translation

This model is fine - tuned for translating Japanese light and web novels into English, capable of handling entire chapters with up to 32K tokens for both input and output.

🚀 Quick Start

This model was fine - tuned on light and web novel for Japanese to English translation. It can translate entire chapters (up to 32K tokens total for input and output).

✨ Features

• Large - scale translation: Capable of translating entire chapters with a combined input and output token limit of up to 32K.
• Glossary support: Allows users to provide custom translations for nouns and character names at runtime.

📦 Installation

Load in llama.cpp

💻 Usage Examples

Basic Usage

Prompt format

<|im_start|>system
Translate this text from Japanese to English.<|im_end|>
<|im_start|>user
{prompt}<|im_end|>
<|im_start|>assistant

Example:

<|im_start|>system
Translate this text from Japanese to English.<|im_end|>
<|im_start|>user
<GLOSSARY>
マイン : Myne
</GLOSSARY>
マイン、ルッツが迎えに来たよ<|im_end|>
<|im_start|>assistant
Myne, Lutz is here to take you home.

The glossary is optional. Remove it if not needed.

Advanced Usage

Text preprocessing

The Japanese text must be preprocessed with the following clean_string function that replaces some unicode characters with ASCII equivalents. Failure to do this may cause issues.

import ftfy

FTFY_ADDITIONAL_MAP = {
    "—": "--",
    "–": "-",
    "⸻": "----",
    "«": "\"",
    "»": "\"",
    "〝": "\"",
    "〟": "\"",
    "✧": "*",
    "✽": "*",
    "⬤": "*",
    "⭘": "*",
    "∴": "*",
    "∵": "*",
    "✩": "*",
    "【": "[",
    "】": "]",
    "「": "[",
    "」": "]",
    "〖": "[",
    "〗": "]",
    "〈": "<",
    "〉": ">",
    "《": "<<",
    "》": ">>",
}

def clean_string(text: str, strip: bool = True) -> str:
    config = ftfy.TextFixerConfig(normalization="NFC")
    s = ftfy.fix_text(text, config=config)
    s = "\n".join((x.strip() if strip else x.rstrip()) for x in s.splitlines())
    for b, g in FTFY_ADDITIONAL_MAP.items():
        s = s.replace(b, g)
    return s

Glossary

You can provide up to 30 custom translations for nouns and character names at runtime. Prefix your chapter with glossary terms (one per line) Japanese term : English term inside <GLOSSARY></GLOSSARY> tags.

glossary = [
    {"ja": "マイン", "en": "Myne"},
]
chapter_text = "マイン、ルッツが迎えに来たよ"

def make_glossary_str(glossary: list[dict[str, str]]) -> str:
    if glossart is None or len(glossary) == 0:
        return ""
    unique_glossary = {(term['ja'], term['en']) for term in glossary}
    terms = "\n".join([f"{ja} : {en}" for ja, en in unique_glossary])
    return f"<GLOSSARY>\n{terms}\n</GLOSSARY>\n"

user_prompt = f"{make_glossary_str(glossary)}{clean_string(chapter_text)}"

<GLOSSARY>
マイン : Myne
</GLOSSARY>
マイン、ルッツが迎えに来たよ

📄 License

This model is licensed under the apache - 2.0 license.

📚 Documentation