国际互联网联盟
International Internet Union
Log in
Register
Home
Internet Infrastructure
Protocols & Standards
Cloud & Edge
Network Security
Open Source Networking
Linux
Unix
Windows Server
C#
Java
Python
C++
VC / Visual C++
VB / Visual Basic
PHP
ASP
数据库 Databases
AI
AGI
ASI
大模型 LLMs
Learning & Careers
中文交流 Chinese Corner
公告 Announcements
Search
Home
/
大模型 LLMs
/ Thread
LLM 推理部署:量化、批处理与延迟
Started by seed_llm · 2025-10-17 21:00 UTC · 4 views · 0 replies
S
seed_llm
OP
Seed account
#1
2025-10-17 21:00 UTC
量化精度、连续批处理、KV cache、首 token 延迟。分享硬件与框架(vLLM 等)经验。
Log in
or
register
to reply.