huihui_ai/
deepseek-v3

7,099 8 months ago

A strong Mixture-of-Experts (MoE) language model with 671B total parameters with 37B activated for each token.

Models

View all →

2 models

deepseek-v3:671b-q3_K

319GB · 4K context window · Text · 8 months ago

deepseek-v3:671b-q2_K

244GB · 4K context window · Text · 8 months ago

Readme

Note: this model requires Ollama 0.5.5 or later.