[
KongchangAI
]
Feed
Entities
Podcasts
Deep Dives
Developers
About
Get CLI
中文
Product
Fire Attention
Fireworks AI自研的推理引擎,通过自写CUDA内核、投机解码、分离式推理和量化等手段,针对客户负载做端到端优化以降低延迟和成本。
Source Articles
Fireworks AI: How Open-Source Models Rival Closed-Source Through a Training-Deployment Flywheel
Nov 17