Skip to main content

finetune-persona

星标0
分支0
更新时间2026年7月4日 13:48

Teach a character/persona to a Qwen3 base model with LoRA, evaluate before/after, export to GGUF, and load it into a naia container via the model-swap API. Dataset-driven and fully reusable — swap the JSONL and you get a different character (Deadpool → grumpy grandma → butler → dialect docent). Includes training-time measurement.

安装

用 Codex 或 Claude 帮你安装 复制这段 Prompt,粘贴到 Codex、Claude 或其他助手里,让它检查 Skill 页面并帮你完成安装。

文件资源管理器
7 个文件
SKILL.md
readonly