Unverified50% confidenceFactExact time
现代LLM通常经过专门的微调来学习何时、如何调用工具，训练数据中包含大量工具使用的示例对话
1
Sources
50%
Confidence
Long-term
Relevance
8/6/2026
First Seen
Sources
Related Claims
UnverifiedThe LLM Trainer Skill supports training of large language models and vision-language models, and can automatically calculate resource requirements, launch remote training, and automatically upload models to the Hub through conversational interaction.78% similarUnverified当前主流LLM普遍采用预训练加微调范式，先在互联网规模语料上无监督预训练，再通过RLHF等技术对齐人类偏好75% similarUnverifiedTALM and Toolformer are training-side tool-calling approaches that modify model weights; the currently more mainstream approach in industry is the prompt-side solution, which does not modify the model and instead relies on prompts to guide the model in deciding when to use tools.73% similarUnverified当前主流 LLM 采用指令遵循（Instruction Following）训练方式，会用概率最高的方式补全输出72% similarUnverifiedDuring training, LLMs learn grammatical rules and common patterns from large amounts of code and text data, but design judgment requires the combined capabilities of visual perception, user psychology, and brand consistency.72% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/699151API
curl https://kongchang.com/api/v1/knowledge/claims/699151MCP
get_claim(id=699151)