[KongchangAI]
Unverified85% confidenceFactTime unknown

继续预训练（CPT）后通常还需要再做一轮SFT才能恢复模型的对话能力，因为CPT可能会削弱模型在SFT阶段学到的指令遵循能力

1
Sources
85%
Confidence
Long-term
Relevance
6/3/2026
First Seen

Sources

Complete Guide to LLM Training: Pre-training, SFT Fine-tuning, and Preference Alignment Explained
bilibili小全栈

Related Entities

CPTconceptSFTconcept

Related Claims

UnverifiedThe learning rate during the SFT stage is typically 1–2 orders of magnitude lower than during pre-training, to avoid disrupting the general knowledge learned during the pre-training stage.70% similar

Cite This Claim

Stable URI
https://kongchang.com/claim/39835
API
curl https://kongchang.com/api/v1/knowledge/claims/39835
MCP
get_claim(id=39835)