1774597379
This commit is contained in:
@@ -8,6 +8,8 @@
|
||||
|
||||
基于 vLLM 推理引擎部署 DeepSeek-R1-Distill-Qwen-32B
|
||||
|
||||
推荐:Qwen3.5-35B-A3B、Qwen3.5-35B-A3B-FP8
|
||||
|
||||
> 32K tokens(32768 约 2.4 万字中文)
|
||||
> 使用 vLLM 框架,注意需开启工具调用相关配置。
|
||||
|
||||
|
||||
Reference in New Issue
Block a user