gpt3

KakaoBrain KoGPT (Korean Generative Pre-trained Transformer)

A PyTorch implementation of "Graph Wavelet Neural Network" (ICLR 2019)

🔍 Search Engine for a Procedural Simulation of the Web with GPT-3.

Pen.el stands for Prompt Engineering in emacs. It facilitates the creation, discovery and usage of prompts to language models. Pen supports OpenAI, EleutherAI, Aleph-Alpha, HuggingFace and others. It's the engine for the LookingGlass imaginary web browser.

FewCLUE 小样本学习测评基准，中文版

Guide: Finetune GPT2-XL (1.5 Billion Parameters) and finetune GPT-NEO (2.7 B) on a single GPU with Huggingface Transformers using DeepSpeed

A PyTorch implementation of "Signed Graph Convolutional Network" (ICDM 2018).

Curating the best GPT3 tools and resources

A C#/.NET SDK for accessing the OpenAI GPT-3 API

Generate SQL from Natural Language Sentences using OpenAI's GPT-3 Model

A GPT-J API to use with python3 to generate text, blogs, code, and more

A list dedicated to products, demos and articles related to 🤖 OpenAI's Codex.

Code for the ICLR 2021 paper "A Distributional Approach to Controlled Text Generation"

OpenAI GPT-3 Api Client in Java

A minimal TF2 re-implementation of the OpenAI GPT training

An OpenAI GPT-3 API client enabling Go/Golang programs to interact with the gpt3 APIs.

Lyrics generation with GPT2-based Transformer

OpenAI's GPT2 based Music AI Google Colab Notebooks for Music Generation/Composition and Capabilities Evaluation

A python code to interact with the GPT3 API to train the chatbot and use it.

This a very basic discord bot linking to GPT3 (openai and AI21). By defaults it is Kirby, but can be any real or virtual bot (just need more or less training prompt).

Using the GPT3 Language model, I create a web application that generates convincing looking emails and then sends them via Gmail. The user interacts with a frontend made with Streamlit

Translate radiology reports from clinical to lay text

GPT-3 Tailwind CSS Code Generator

🧠 Aquí mostramos como al escalar el tamaño del modelo mejora en gran medida el rendimiento el perfomance de pocos intentos o few shots, a veces incluso alcanzando niveles de competitividad con enfoques anteriores de vanguardia con fine tuning. Específicamente, entrenamos GPT-3, un modelo de lenguaje con 175 mil millones de parámetros, 10 veces más que cualquier modelo non-sparse anterior, y probamos su rendimiento en la configuración few shots.

🏃‍♂️ Natural language activity tracking with GPT-3

Generate headlines and descriptions for your LinkedIn profile with the help of GPT-3!

A python flask app that generates a spooky story using openai's gpt-3

Telegram bot integrated with ruGPT3

This is the code I used to create a small private SMS text chat system that employs GPT3 from OpenAI and "nonviolent communication", an algorithmically based method of conflict resolution. Hopefully the chat helps users process conflict.