Matthew GuntoninTowards Data ScienceExploring Medusa and Multi-Token PredictionThis blog post will go into detail on the “MEDUSA: Simple LLM Inference Acceleration Framework with Multiple Decoding Heads” paper2d ago2d ago
Matthew GuntoninTowards Data ScienceDiving Deep into AutoGen and Agentic FrameworksThis blog post will go into the details of the “AutoGen: Enabling Next-Gen LLM Applications via Multi-Agent Conversation” paperJun 28Jun 28
Matthew GuntoninTowards AIUnderstanding Mamba and Selective State Space Models (SSMs)This blog post will go in detail on the “Mamba: Linear-Time Sequence Modeling with Selective State Spaces” paperJun 241Jun 241
Matthew GuntoninTowards Data ScienceUnderstanding You Only Cache OnceThis blog post will go in detail on the “You Only Cache Once: Decoder-Decoder Architectures for Language Models” Paper and its findingsJun 4Jun 4
Matthew GuntoninTowards Data ScienceUnderstanding Low Rank Adaptation (LoRA) in Fine Tuning LLMsHow LoRA works to fine-tune LLMs, following the methodology set out in the “LoRA: Low-Rank Adaptation of Large Language Models” paperMay 241May 241
Matthew GuntoninTowards Data ScienceUnderstanding Long RoPE in LLMsThis blog post will go in detail about the new Long RoPE Methodology used to expand the context lengths LLMs can support without…May 155May 155
Matthew GuntoninTowards Data SciencePhi-3 and the Beginning of Highly Performant iPhone ModelsThis blog post will go into the findings of the Phi-3 paper, as well as some of the implications of models like Phi-3 being releasedMay 9May 9
Matthew GuntoninTowards Data ScienceTool Use, Agents, and the Voyager PaperA detailed exploration of the Voyager Paper and its findings on tool usageMay 12May 12
Matthew GuntoninTowards Data ScienceMultimodal Large Language Models & Apple’s MM1This blog post will go into the architecture and findings behind Apple’s “MM1: Methods, Analysis & Insights from Multimodal LLM…Apr 13Apr 13
Matthew GuntoninTowards Data ScienceFrugalGPT and Reducing LLM Operating CostsThis blog post will go into detail about a cost-saving architecture for LLM-driven apps as seen in the “FrugalGPT” paperMar 271Mar 271