Cosmos-LLaVA: Chatting with the Visual Cosmos-LLaVA: G\"orselle Sohbet Etmek
Abstract
In this study, a Turkish visual instruction model was developed and various model architectures and dataset combinations were analysed to improve the performance of this model. The Cosmos-LLaVA model, which is built by combining different large language models and image coders, is designed to overcome the deficiencies in the Turkish language. In the experiments, the effects of fine-tuning with various datasets on the model performance are analysed in detail. The results show that model architecture and dataset selection have a significant impact on performance. Bu calsmada bir T\"urkce g\"orsel talimat modeli gelistirilerek bu modelin performansn artrmaya y\"onelik cesitli model mimarileri ve veri k\"umesi kombinasyonlar derinlemesine incelenmistir. Farkl b\"uy\"uk dil modelleri ve g\"or\"unt\"u kodlayclarnn bir araya getirilmesiyle olusturulan Cosmos-LLaVA modeli, T\"urkce dilindeki eksiklikleri gidermeye y\"onelik olarak tasarlanmstr. Yaplan deneylerde, cesitli veri k\"umeleri ile yaplan ince ayarlarn model performansn nasl etkiledigi detayl olarak ele alnmstr. Sonuclar, model mimarisi ve veri k\"umesi seciminin performans \"uzerinde \"onemli bir etkiye sahip oldugunu g\"ostermektedir.
Turn this paper into a full lesson
ArcXiv compiles a staged curriculum from this paper: 8-12 lessons across beginner → advanced, synthesised section guides, visuals, flashcards, a quiz, exercises, and on-demand deep dives per section. Grounded in the abstract, never invented.