DeepSeek-V2

核心与官方官方模型与基建中文文档MIT官方停更
Model Download | Evaluation Results | Model Architecture | API Platform | License | Citation

功能特性

  • 2024.05.16: We released the DeepSeek-V2-Lite.
  • 2024.05.06: We released the DeepSeek-V2.
  • For Feed-Forward Networks (FFNs), we adopt DeepSeekMoE architecture, a high-performance MoE architecture that enables training stronger models at lower costs.

项目预览

← 返回 核心与官方 列表