 |
ML Newsnew | past | best | rss | submit | about |
| Nemotron-4 15B large multilingual language model trained on 8T tokens (arxiv.org) | |
| 3 points by hack_ml on Feb 27, 2024 | hide | past | pdf | 1 comment on HN | |
|
In plain words: A 15-billion-parameter language model trained on 8 trillion text tokens, built to handle English, many languages, and coding. It beat every similar-sized open model in most test areas, and its multilingual skills topped even models over four times larger. Abstract · Nemotron-4 15B Technical ReportWe introduce Nemotron-4 15B, a 15-billion-parameter large multilingual language model trained on 8 trillion text tokens. Nemotron-4 15B demonstrates strong performance when assessed on English, multilingual, and coding tasks: it outperforms all existing similarly-sized open models on 4 out of 7 downstream evaluation areas and achieves competitive performance to the leading open models in the remaining ones. Specifically, Nemotron-4 15B exhibits the best multilingual capabilities of all similarly-sized models, even outperforming models over four times larger and those explicitly specialized for multilingual tasks. | |
| add comment on HN |
introduce Nemotron-4 15B, a 15-billion-parameter large multilingual language model trained on 8 trillion text tokens. Nemotron-4 15B demonstrates strong performance when assessed on English, multilingual, and coding tasks: it outperforms all existing similarly-sized open models on 4 out of 7 downstream evaluation areas and achieves competitive performance to the leading open models in the remaining ones. Specifically, Nemotron-4 15B exhibits the best multilingual capabilities of all similarly-sized models, even outperforming models over four times larger and those explicitly specialized for multilingual tasks.