4 years, 5 months ago
Facebook's AI model can translate 100 languages without relying on English
Facebook on Monday introduced the first open-source multi-lingual machine translation model M2M-100 that can translate between any pair of 100 languages without relying on English data. Typical translation systems require building separate AI models for each language and task, but this approach doesn’t scale effectively on Facebook, where people post content in more than 160 languages across billions of posts. Facebook used novel mining strategies to create translation data, building a data set with 7.5 billion sentences for 100 languages. For years, AI researchers have been working toward building a single universal model that can understand all languages across different tasks.







Discover Related

1 day, 9 hours ago
Meta Says To Train AI Models On Public Data Of European Users

3 days, 9 hours ago
Small Language Models Are the New Rage, Researchers Say

1 week, 1 day ago
AI impact to change future workforce

1 week, 1 day ago
AI impact to change future workforce

1 week, 2 days ago
Evaluation of AI large language models in final stage: Ashwini Vaishnaw

1 week, 3 days ago
AI tracker: Three cases of AI ethics that gave us food for thought this week

55 years, 3 months ago
Databricks Has a Trick That Lets AI Models Improve Themselves

3 weeks, 6 days ago
Wipro brings sovereign AI services with NVIDIA to service global governments

4 weeks, 2 days ago
‘We’re still at the beginning of the AI journey’

1 month, 1 week ago
Toshiba America Business Solutions Launches Real-Time Translation Innovation

1 month, 2 weeks ago
AI models flunk language test that takes grammar out of the equation

1 month, 2 weeks ago
Google Translate app may get AI-powered ‘Follow up’ button to dive deeper: Report

1 month, 3 weeks ago
Google launches AI coding assistant, Gemini Code Assist, free for individual users

1 month, 3 weeks ago
India moves a step closer to desi Deepseek: 5 things you should know

1 month, 4 weeks ago
DeepSeek’s R1 may be the first of many AI super-apps to come

2 months ago
IIITH focuses on making AI to forget info

2 months ago
Proper data sharing essential for language models

2 months, 1 week ago
Can AI think on its own beyond the training parameters? Study finds evidence

2 months, 1 week ago
ChatGPT challenges DeepSeek R1, by launching Deep Research AI

2 months, 1 week ago
Seminar on potential of DeepSeek held in Thiruvananthapuram

2 months, 2 weeks ago
Why is Chinese AI startup DeepSeek stirring up the tech world?

2 months, 2 weeks ago
MCLI will create models to preserve classical Indian languages: Rohan Murty

2 months, 2 weeks ago
DeepSeek has rattled the AI industry. Here’s a quick look at other Chinese AI models

2 months, 2 weeks ago
36% of Indian enterprises started budgeting for Gen AI: E&Y report

55 years, 3 months ago
Chinese AI App DeepSeek Soars in Popularity, Startling Rivals

2 months, 2 weeks ago
‘Sorry, I didn’t get that’: AI misunderstands some people’s words more than others

2 months, 3 weeks ago
Indian IT services firms take a divergent AI approach

2 months, 3 weeks ago
How AI threatens linguistic diversity

2 months, 3 weeks ago
How AI threatens linguistic diversity

3 months, 2 weeks ago