Subject archive for "data-science"

See how NVIDIA's RAPIDS v23.10 revolutionizes AI workloads with GPU-accelerated Pandas on Domino, offering massive speed boosts and data handling efficiency.

Domino's integrated AutoML offers scalable, reproducible, & flexible solutions with editable code, diverse team collaboration, & hybrid-cloud support.

Companies diving head-first into Generative AI must consider the risks it brings. Domino’s guide shows you how to harness GenAI responsibly. Download now!

Generative AI transforms industries, but LLM deployment is tough. See how Domino simplifies LLM hosting & inference.

This blog post explores the challenges of fine-tuning large language models (LLMs) and introduces resource-optimized and parameter-efficient techniques such as quantization, LoRA, and Zero Redundancy Optimization (ZeRO). By fine-tuning Falcon-7b, Falcon-40b, and GPTJ-6b, we demonstrate how these techniques offer improved performance, cost-effectiveness, and resource optimization in LLM fine-tuning. The blog post also discusses the future of fine-tuning and its potential for unlocking new possibilities in enterprise AI applications.

With the ongoing generative AI hype, one concept is becoming increasingly clear: giant, generic generative AI models, by themselves, are not the key to unlocking business value. While they are excellent for experimentation, entertainment, and some limited end-user work augmentation (ChatGPT might have helped with parts of this blog), they often fall short in terms of performance, accuracy, and risk when they aren't production grade.