
Latent Space 对谈 Baseten 的 Philip Kiely 与 Ali Taha,解释 KV cache、量化、speculative decoding 和硬件协同如何把新模型从能运行推进到可稳定服务。
| 字段 | 内容 |
|---|---|
| 播客 | Latent Space: The AI Engineer Podcast 1 |
| 主持人 | swyx、Vibhu;逐字稿中的主要提问者标注为 swyx 2 |
| 嘉宾 | Philip Kiely、Ali Taha,Baseten |
| 集标题 | The Inference Engineering Masterclass — Philip Kiely & Ali Taha, Baseten |
| 发布日期 | 2026 年 8 月 4 日(北京时间;Apple Podcasts 记录为 2026-08-03T21:44:03Z) 1 |
| 原集链接 | The Inference Engineering Masterclass |
Three years ago, inference engineering barely existed as a category. Today, it is one of the most critical disciplines in AI.Philip Kiely。2
The LLM is not capable of doing anything. It’s only capable of making suggestions of what to do.Philip Kiely。2
Quantization errors have canceled out.Ali Taha。2
latent.space
api.substack.com
This story was produced automatically by a channel. One sentence is all it takes for Neodrop to keep producing for you.