跳到正文
Hao AI Lab· @haoailab · X·· 2 小时前AI 评分38
AI 导读

JetSpec++ 让草稿时间几乎移出关键路径,在 GSM8K 上 Qwen3-8B 每轮草稿-验证延迟从 DSpark 的 9.92 ms 降至 6.84 ms(草稿时间 3.76 ms→0.61 ms),Qwen3.8-27B 从 31.88 ms 降至 28.47 ms(4.46 ms→0.71 ms)。

正文 · 原文

[2/5] With JetSpec++, drafting time is almost removed from the critical path. Walltime reported below shows how latency reduces for each draft-verification round, and its breakdown on GSM8K:
• Qwen3-8B: DSpark 9.92 ms (including 3.76 ms drafting time), JetSpec++ 6.84 ms (including 0.61 ms is drafting time)
• Qwen3.8-27B: DSpark 31.88 ms (4.46 ms drafting time), JetSpec++ 28.47 ms (0.71 ms drafting time)

The key is JetSpec++ has effective foresight drafting: it guesses the verification outcome correctly in 92.9% (8B) and 89.0% (27B) of rounds, so the next block is already drafted when verification ends.

来源:Hao AI Lab · x.com