Build a deepseek model (from scratch)
Auteur :
Dandekar, Raj
Éditeur :
Manning Publications
ISBN :
9781633434325
Date de publication :
7 oct. 2026
Langue :
Anglais
Pays d'origine :
USA
By creatively blending a variety of strategies and innovations like Mixture of Experts, Latent Attention, Multi-token Prediction, model distillation and efficient parallelisation, DeepSeek set a new standard for what’s possible in an open LLM.Â