Category: LLM
-
Claude Code: AI-Powered Dev in Your Terminal
Claude Code runs AI in your terminal with full filesystem access. After 3 months in production: what breaks, what scales, what's worth it.
-
LoRA vs QLoRA vs DoRA ์๋ฒฝ ๋น๊ต: ํ๋ผ๋ฏธํฐ ํจ์จ์ ํ์ธํ๋(PEFT) ๋ฉ๋ชจ๋ฆฌ ์ต์ ํ ์ค์ ๊ฐ์ด๋
LoRA uses 60GB GPU, QLoRA needs 16GB, DoRA hits 12GB โ same model quality. Here's the memory breakdown and when to pick each method.
-
MoE ์ํคํ ์ฒ: Mixtral๋ถํฐ DeepSeek-MoE๊น์ง ์์ ๋ถ์
MoE architecture explained: how Mixtral and DeepSeek-MoE achieve 8x parameters with 2x compute. Implementation guide included.
-
RLHF vs DPO vs KTO: LLM ์ ๋ ฌ(Alignment) ๊ธฐ๋ฒ ์๋ฒฝ ๋น๊ต ๊ฐ์ด๋
RLHF, DPO, or KTO for LLM alignment? Compare training costs, data needs, and performance on real tasks to pick the right method.