Jonathan KernesinTowards Data ScienceDiffusion ModelsWhat are they, how do they work, and why now?Dec 13, 20225Dec 13, 20225
Jonathan KernesinNtropyThe False Promise of General Transaction Categorization and the inadequacy of in-house modelsWhy the Ntropy customized model approach is the only solution for high performanceMay 24, 2022May 24, 2022
Jonathan KernesinNtropyCustomized Transaction ClassificationSuccesses, benchmarking, and how to maximize your performance. Bringing meta-learning to production.Apr 19, 20221Apr 19, 20221
Jonathan KernesinNtropyQuickly finding recurring events in massive sequencesA scalable O(N*K) solution for sequences of size N and max intervals of size KFeb 20, 2022Feb 20, 2022
Jonathan KernesinTowards Data ScienceMaster Positional Encoding: Part IIWe upgrade to relative position, present a bi-directional relative encoding, and discuss the pros and cons of letting the model learn this…Feb 25, 20213Feb 25, 20213
Jonathan KernesinTowards Data ScienceMaster Positional Encoding: Part IWe present a “derivation” of the fixed positional encoding that powers Transformers, helping you get a full intuitive understanding.Feb 15, 202121Feb 15, 202121
Jonathan KernesinTowards Data ScienceBuild Your Own Automatic Differentiation ProgramFor anyone who’s completely lost at how graphs can be used to compute derivatives, or just wants to know how TensorFlow works at a…Feb 9, 20212Feb 9, 20212
Jonathan KernesinTowards Data ScienceHow to Overcome the Large Vocabulary Bottleneck Using an Adaptive Softmax LayerA detailed guide to how a simple GPU-optimized layer substitution can offer 2x-10x speedups, with little to no loss in performance.Feb 4, 20211Feb 4, 20211
Jonathan KernesinTowards Data ScienceLocality Sensitive Hashing: How to Find Similar Items in a Large Set, with PrecisionWe offer a guide to the art of locality sensitive hashing, with applications to document comparison and vector similarity.Feb 4, 20211Feb 4, 20211
Jonathan KernesinTowards Data ScienceSentencePiece Tokenizer DemystifiedAn in depth dive into the inner workings of the SentencePiece tokenizer, why it’s so powerful, and why it should be your go to tokenizer…Feb 4, 20216Feb 4, 20216