DeepSeek-V3 is a 671B parameter Mixture-of-Experts language model with 37B activated parameters that...

Tokens:42,752
Snippets:252
Trust Score:6.8
License:MIT
Update:1 week ago
Tokens:
Raw