Show HN: kulyk-uk-en and kulyk-en-uk

Heykuki News

2 points

a year ago

A new lightweight model to do machine translation Ukrainian<->English using recently published LFM2 model by LiquidAI.

With llama-server it gives 60 tokens/sec. on Apple M1, with NVIDIA RTX 3090 Ti it's faster around 230 tokens/sec with batch size = 1.

1 comment

Show HN: kulyk-uk-en and kulyk-en-uk | Heykuki News