deepseek-v3:671b-fp16

2.5M Downloads Updated 9 months ago

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

671b

Updated 9 months ago

9 months ago

7770bf5a5ed8 · 1.3TB ·

model

archdeepseek2

parameters671B

quantizationF16

1.3TB

template

{{- range $i, $_ := .Messages }} {{- if eq .Role "user" }}<｜User｜> {{- else if eq .Role "assista

359B

license

14kB

params

{ "stop": [ "<｜begin▁of▁sentence｜>", "<｜end▁of▁sentence｜>",

148B

Readme

Note: this model requires Ollama 0.5.5 or later.

DeepSeek-V3 achieves a significant breakthrough in inference speed over previous models. It tops the leaderboard among open-source models and rivals the most advanced closed-source models globally.

References

GitHub

Paper