Fine-tune Translation Model
Budget: ₹1,500 – ₹12,500 INR
We want to fine-tune a NMT model with some translation data. We already have built the scraper to scrape files from `https://judgments.ecourts.gov.in/pdfsearch/index.php` and the pre-trained model is available at `https://github.com/ai4bharat/IndicTrans2`.
Data
The scraped data consists of English and regional language pdf texts extracted from pdfs with tesseract. You'll need to find some some way to structure the paragraphs before passing them to training.
Example files linked below.
Data
The scraped data consists of English and regional language pdf texts extracted from pdfs with tesseract. You'll need to find some some way to structure the paragraphs before passing them to training.
Example files linked below.