awesome-prompts

awesome for your to collections

LLMs

based on: 总结当下可用的大模型LLMs

Model	作者	Size	类型	开源？
LLaMa	Meta AI	7B-65B	Decoder	open
OPT	Meta AI	125M-175B	Decoder	open
T5	Google	220M-11B	Encoder-Decoder	open
mT5	Google	235M-13B	Encoder-Decoder	open
UL2	Google	20B	Encoder-Decoder	open
PaLM	Google	540B	Decoder	no
LaMDA	Google	2B-137B	Decoder	no
FLAN-T5	Google	同T5	Encoder-Decoder	open
FLAN-UL2	Google	同U2	Encoder-Decoder	open
FLAN-PaLM	Google	同PaLM	Decoder	no
FLAN	Google	同LaMDA	Decoder	no
BLOOM	BigScience	176B	Decoder	open
T0	BigScience	3B	Decoder	open
BLOOMZ	BigScience	同BLOOM	Decoder	open
mT0	BigScience	同T0	Decoder	open
GPT-Neo	EleutherAI	125M-2.7B	Decoder	open
GPT-NeoX	EleutherAI	20B	Decoder	open
GPT3	OpenAI	175B (davinci)	Decoder	no
GPT4	OpenAI	unknown	OpenAI	no
InstructGPT	OpenAI	1.3B	Decoder	no
Alpaca	Stanford	同LLaMa	Decoder	open
ChatGLM-6B	Tsinghua University	6B	Decoder	open
GLM-130B	Tsinghua University	130B	Decoder	open

Instruct/Prompt Tuning Data

based on 总结开源可用的Instruct/Prompt Tuning数据

数据集/项目名称	组织/作者	简介
Natural Instruction / Super-Natural Instruction	Allen AI	包含61个NLP任务（Natural Instruction）和1600个NLP任务（Super-Natural Instruction）的指令数据
PromptSource / P3	BigScience	包含270个NLP任务的2000多个prompt模版（PromptSource）和规模在100M-1B之间的P3数据集
xMTF	BigScience	包含13个NLP任务、46种语言的多语言prompt数据
HH-RLHF	Anthropic	旨在训练Helpful and Harmless（HH）的LLMs的RLHF数据集
Unnatural Instruction	orhonovich	使用GPT3生成64k的instruction prompt数据，经改写后得到240k条instruction数据
Self-Instruct	yizhongw	使用LLMs生成prompt进行instruct-tuning的方法，引入Task pool和Quality filtering等概念
UnifiedSKG	HKU	在Text-to-Text框架中加入knowledge grounding，将结构化数据序列化并嵌入到prompt中
Flan Collection	Google	将Flan 2021数据与一些开源的instruction数据（P3，super-natural instruction等）进行合并
InstructDial	prakharguptaz	在特定的一种任务类型（对话指令）上进行指令微调的尝试
Alpaca-LoRA	tloen	Low-Rank LLaMA Instruct-Tuning

Projects:

• Prompt Engineering Guide - Guides, papers, lecture, and resources for prompt engineering
  • notebooks
• LangChain - ⚡ Building applications with LLMs through composability ⚡
  • ChatLangChain is an implementation of a locally hosted chatbot specifically focused on question answering over the LangChain documentation. Built with LangChain and FastAPI.
  • Langflow - LangFlow is a GUI for LangChain, designed with react-flow to provide an effortless way to experiment and prototype flows with drag-and-drop components and a chat box.
• [Alpaca]
  • Alpaca-LoRA - Low-Rank LLaMA Instruct-Tuning, Instruct-tune LLaMA on consumer hardware .
  • Alpaca-LoRA as a Chatbot Service
  • AlpacaDataCleaned
  • BELLE - Bloom-Enhanced Large Language model Engine，针对 Stanford Alpaca 中文做了优化，模型调优仅使用由ChatGPT生产的数据（不包含任何其他数据）。
  • https://github.com/LC1332/Chinese-alpaca-lora - A Chinese finetuned instruction LLaMA.
  • Dolly - fine-tunes the GPT-J 6B model on the Alpaca dataset using a Databricks notebook.
• ChatRWKV - ChatRWKV is like ChatGPT but powered by RWKV (100% RNN) language model, and open source.
• [GPT-j]

Infra

• PEFT - Parameter-Efficient Fine-Tuning (PEFT) methods enable efficient adaptation of pre-trained language models (PLMs) to various downstream applications without fine-tuning all the model's parameters.

Patterns

Semantic Kernel

Semantic Kernel is designed to support and encapsulate several design patterns from the latest in AI research, such that developers can infuse their applications with complex skills like prompt chaining, recursive reasoning, summarization, zero/few-shot learning, contextual memory, long-term memory, embeddings, semantic indexing, planning, and accessing external knowledge stores as well as your own data.

Prompter Papers

• MathPrompter: Mathematical Reasoning using Large Language Models