Unverified50% confidenceFactExact time
Time Per Output Token (TPOT) measures the speed at which a model generates tokens one by one during the autoregressive decoding phase.
1
Sources
50%
Confidence
Long-term
Relevance
8/30/2026
First Seen
Sources
Related Entities
Related Claims
Unverified注意力机制计算当前 token 与所有历史 token 相关性的时间复杂度为 O(n²)，其中 n 为序列长度68% similarUnverifiedTimesFM adopts patch-based tokenization, packing several consecutive time steps into a patch and treating it as a single token, similar to the image patch strategy used by ViT.64% similarUnverified对于单用户交互式场景推理速度达到5-10 token/s通常就能提供流畅体验，而代码生成、长文档摘要或多轮Agent调用往往需要30 token/s以上63% similarUnverified20 tokens/s以上的输出速度可以实现流畅的实时阅读体验，低于5 tokens/s则会让用户明显感到等待62% similarUnverified自动标注的价值应以端到端时间衡量，公式为：节省时间 = 从零标注的时间 - (预标注生成时间 + 人工修正时间)61% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/823666API
curl https://kongchang.com/api/v1/knowledge/claims/823666MCP
get_claim(id=823666)