WizardLM: Evolving Instruction Datasets to Create a Better Model

Colab: https://colab.research.google.com/drive/1H308Mj11PTMCUm_TxTj8DF189ujDG_1w?usp=sharing

Demo: https://261f01fdd31bfe1ca0.gradio.live/
Github: https://github.com/nlpxucan/WizardLM/tree/main
Paper: https://arxiv.org/abs/2304.12244

In this paper I look at the WizardLM project which goes beyond just a new model and introduces a new way to distill a dataset for fine tuning.

For more tutorials on using LLMs and building Agents, check out my Patreon:
Patreon: https://www.patreon.com/SamWitteveen
Twitter: https://twitter.com/Sam_Witteveen

My Links:
Linkedin: https://www.linkedin.com/in/samwitteveen/

Github:
https://github.com/samwit/langchain-tutorials
https://github.com/samwit/llm-tutorials

00:00 Intro
02:57 Paper
08:32 Colab Walkthrough

Видео WizardLM: Evolving Instruction Datasets to Create a Better Model канала Sam Witteveen