tflsxyy
/

DeepSeek-V2-Lite-BiMoE-w1g16

Text Generation

text-generation-inference

Model card Files Files and versions

Attention and shared experts: 4-bit

All routed experts: 1-bit

Quantized by https://github.com/tflsxyy/BiMoE

Downloads last month: 14

Safetensors

Model size

1.98B params

Tensor type

F16

·

I32

·

Model tree for tflsxyy/DeepSeek-V2-Lite-BiMoE-w1g16

Base model

deepseek-ai/DeepSeek-V2-Lite

Finetuned

(19)

this model