PinnedHarshit SharmainTowards AIThe Ever-evolving Pre-training Tasks for Language ModelsSelf-Supervised Learning (SSL) is the backbone of transformer-based pre-trained language models, and this paradigm involves solving…Dec 28, 20222Dec 28, 20222
Harshit SharmainLevel Up CodingHow does Langchain Expression Language (LCEL) work ?Classic use of Polymorphism and Operator overloadingJun 25Jun 25
Harshit SharmaA fun history about Git“the information manager from hell” — creator of gitMar 41Mar 41
Harshit SharmainLevel Up CodingMaximum Inner Product Search (MIPS) and it’s use in MLYou must have come across this popular problem statement:Jan 6Jan 6
Harshit SharmaHitchiker’s Guide to Galaxy and random.seed()Their uncanncy connectionSep 25, 20231Sep 25, 20231
Harshit SharmainTowards AIGradient CheckpointingTo “scale” new heights in model trainingJul 4, 2023Jul 4, 2023
Harshit SharmainTowards AIApproximate Nearest NeighborsAnd where to find them using Product QuantizationJun 20, 2023Jun 20, 2023
Harshit SharmaAccumulate Gradients and you are good to goA simple answer to “I can’t increase batch size because of limited GPU”Jun 4, 20232Jun 4, 20232
Harshit SharmainTowards AI4 Things to Know about Large Language ModelsAmidst the LLM hype, there are interesting “things” to know about the LLMs, as mentioned in a recent paper by Samuel from Anthropic:May 28, 20231May 28, 20231