Unverified50% confidenceFactExact time
Turbo defaults to CFG=1, requiring only one forward inference pass per step.
1
Sources
50%
Confidence
Long-term
Relevance
10/11/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedCFG (Classifier-Free Guidance) requires running both conditional and unconditional forward passes simultaneously at each denoising step, doubling the computation per inference step when enabled.67% similarVerifiedTurbo version models are typically based on techniques such as consistency distillation or flow matching, compressing the tens of denoising inference steps of standard diffusion models to 4 to 8 steps or even fewer63% similarUnverifiedCFG performs conditional and unconditional predictions separately during inference and interpolates and amplifies them via the formula output = uncond + guidance_scale × (cond - uncond), at the cost of requiring two forward passes per step.62% similarUnverified从30步CFG 4到12步CFG 1的加速本质上是模型蒸馏技术的应用成果，与Consistency Model、SDXL-Turbo、LCM共享相似底层思路62% similarUnverifiedCommon inference acceleration techniques include consistency distillation, flow matching, and diffusion model step compression methods such as LCM and SDXL-Turbo.59% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/996801API
curl https://kongchang.com/api/v1/knowledge/claims/996801MCP
get_claim(id=996801)