PinnedRanko MosicWhy You Need a Custom Domain Chatbot: A DIY Guide Based on OpenAI’s ChatGPTWhat technologies / techniques are behind it ?5 min read·Feb 9, 2023----
Ranko MosicLlama3 — What We Know So FarLlama3 is #1 open source¹ and top 10 overall ranked model.1 min read·5 days ago----
Ranko MosicInfinite Context TransformersTransformer-based language models (LMs) are powerful and widely-applicable tools, but their usefulness is constrained by a fnite context…3 min read·Apr 21, 2024----
Ranko MosicQ* or What Comes Next in LLM LandIt is relatively obvious ( bar emerging surprises ) that brute force LLM scaling won’t get us to LLMs capable of planning and reasoning…4 min read·Apr 3, 2024----
Ranko MosicOpenAI Sora and Open Source Efforts to Catch UpMy mental model of Sora is that it is the “GPT-2 moment” for video generation.5 min read·Feb 17, 2024----
Ranko MosicOLMo — Interesting PointsOLMo was trained on both NVIDIA A100 and AMD MI250X hardware. This is of huge importance, given the current shortages of NVIDIA GPU and…2 min read·Feb 2, 2024----
Ranko MosicGoogle Gemini, OpenAI GPT-4 , Multimodality and Open Source Attempts to Catch UpJPEG is worth a thousand words ( old proverb ).8 min read·Dec 27, 2023----
Ranko MosicLLM Training —Google Hardware and Software StackWhy Does Specialized Hardware Make Sense for Deep Learning Models? Deep learning models have three properties that make them different than…3 min read·Oct 29, 2023----
Ranko MosicLLM Training — AMD Hardware and Software StackThe importance of the AMD stack is increasing due to the shortages of NVIDIA H100 GPUs and the closed-source, proprietary nature of CUDA…3 min read·Oct 2, 2023----