Meta Rolls Out Multimodal Llama 3.2 — But Not in Europe
Meta unveils its Llama 3.2 models, but the multimodal versions are not available in Europe...
Maria Stasimioti
Array
(
[post_type] => post
[posts_per_page] => 12
[paged] => 19
[post_status] => publish
[author] => 2795
)
The First Large Language Model Supporting All EU Languages Is Here
Researchers from industry and academia present EuroLLM, the first large language model that supports all...
Sony Adds New Languages to Speech Translation Systems Without Retraining
Sony and Carnegie Mellon University researchers show how to expand language coverage in speech translation...
Google Aligns LLM Translation with Human Translation Processes
Google’s new approach mimics the human translation process, incorporating research, drafting, and iterative refinement in...
Translating Input in Prompts Improves LLM Performance for Low-Resource Languages
New research presents a translation-based prompting technique that can improve the performance of large language...
Fine-Tuning LLMs with Translation Memories Boosts Accuracy and Cuts Costs
New research demonstrates that fine-tuning large language models with translation memories improves translation quality, cuts...
Instruction-Tuned Large Language Models Outperform Baselines in Medical Translation
A new study highlights the potential of instruction tuning to enhance the accuracy and consistency...
Research Attempts to Bring Literary Machine Translation Closer to Human Quality
Researchers from Aalborg University and the University of Groningen show how to recover stylistic nuance...
New Research Reviews Key Challenges in Simultaneous Speech Translation
A new study provides a detailed analysis of the challenges and solutions in the field...
ByteDance Unveils a ‘Human-Like’ Speech Translation System
Researchers from ByteDance unveil a simultaneous speech translation system the Chinese tech company claims is...
Apple and USC Propose Solution for Gender Bias in Machine Translation
Researchers from Apple and USC propose tackling gender bias in MT by offering multiple gendered...