关于langchain的一些总结和思考, 各类资源汇总

Overview 文章出处在这里 LLM powered autonomous agents: planning + memeoy + tool use 1. planning Chain of thought: 想法 prior work A: 可以通过产生用自然语言表达的中间步骤使得LLM获得更好的推理能力, techniques for arithmetic reasoning can benefit from generating natural language rationales that lead to the final answer。 缺点: 要想训练具有良好推理能力的大模型(或者fine-tune a pretrained model), 需要大量的高质量推理数据集, 成本太高, 远远超过简单的input-output pairs。 prior work B: 大模型已经可以通过prompt工程获得出色的少样本学习能力, large language models offer the exciting prospect of in-context few-shot learning via prompting. That is, instead of finetuning a separate language model checkpoint for each new task, one can simply “prompt” the model with a few input–output exemplars demonstrating the task...

This post provides an overview of Hugo (PaperMod theme) and details the steps I took in setting up this website.