Views
No views yet
Dinamus/ml-danbooru-onnx-repo (image tagging / folder routing).manga-ocr when MANGA_TRANSLATOR_REAL_MODELS=1, else stub)opus-mt-ja-en when real models, else stub)1cd backend
2python -m venv .venv
3.\.venv\Scripts\Activate.ps1
4pip install -e ".[dev]"
5pytest
6uvicorn manga_translator.api.app:app --reload --port 8000manga-translator translate-batch .\path\to\pages1cd web
2npm install
3$env:NEXT_PUBLIC_API_BASE="http://127.0.0.1:8000"
4npm run dev.venv path can fail with WinError 206. Prefer a short venv:1python -m venv C:\mtvenv
2C:\mtvenv\Scripts\python -m pip install -e ".[ml]"
3C:\mtvenv\Scripts\python -m pip install -e .
4$env:MANGA_TRANSLATOR_REAL_MODELS="1"
5$env:MANGA_TRANSLATOR_OCR_BACKEND="manga_ocr"
6$env:MANGA_TRANSLATOR_TRANSLATE_BACKEND="opus_mt"
7C:\mtvenv\Scripts\python -m uvicorn manga_translator.api.app:app --port 8010docs/CHECKLIST-p1-ui-local-models.md1# Glossary: preserve names/terms (only entries present in the page are sent to the LLM)
2$env:MANGA_TRANSLATOR_GLOSSARY_PATH="data/dict/glossary.json" # {"太郎":"Taro"}
3# Pre-dict: fix OCR artefacts / strip SFX BEFORE translation (ordered regex rules)
4$env:MANGA_TRANSLATOR_PRE_DICT_PATH="data/dict/pre_dict.example.json"
5# Post-dict: fix recurring MT phrasing AFTER translation
6$env:MANGA_TRANSLATOR_POST_DICT_PATH="data/dict/post_dict.example.json"{"pattern": "replacement"} or ordered [["pattern","replacement"], ...]; keys are regular expressions. Lines already in the target language are passed through untranslated, and identical bubbles are translated once. Reading order follows manga convention (right-to-left, panel-banded).C:\mtvenv\Scripts\python backend/scripts/eval_accuracy.py ollama # CER (OCR) + chrF (translation)backend/ Python package, FastAPI, CLI, tests
web/ Next.js job UI
data/jobs/ Runtime job store (gitignored)
docs/superpowers/ Design spec + implementation plan