Mayowa Osibodu@mayowa_osibodu

Language is the medium via which we communicate with other people. Given the large number of languages in the world today, language learning is a task we all encounter at various points in our lives. Llama Lingo uses Meta's Llama 3.2 Vision Model to process images you take of your surroundings, identify objects in those images, and then quiz you by asking you to name those objects in the language you're trying to learn. Language resources (language dictionaries etc) can be uploaded to Llama Lingo (The Llama LLM is given access to these resources using Llama Index). This modular approach means Llama Lingo can be applied to any language - both languages Llama can speak by default, as well as low-resource languages where Llama would need access to specialized language resources. Llama Lingo provides a more practical way to learn a language, because you're focused on learning words for objects you encounter in your day-to-day life (for example words for objects in your bedroom, or words describing people on the subway). It also makes for a more memorable learning experience because it requires using multiple senses - You visually and spatially identify subjects in your surrounding, while also cognitively thinking about their corresponding words in the new language.