translation

....translating....

Trained on the public OPUS-100 (en–hi) dataset — quality is still rough, especially the knowledge-distilled LSTM — but we’re improving; treat this as fun and see what you get.