PICa
by microsoft
An Empirical Study of GPT-3 for Few-Shot Knowledge-Based VQA, AAAI 2022 (Oral)
AI summary
Question Answering Experiment
An empirical study on using GPT-3 for multimodal question answering tasks with few-shot learning.
- stars
- 85
- forks
- 6
- watching
- 3
Similar projects
Found by comparing what the projects do, not just their names.
GPT-3 testing suite
A collection of test prompts and generated texts for evaluating OpenAI's GPT-3 API
Document Answerer
An application that utilizes language models and machine learning to provide instant citations and answers from uploaded documents.
Multimodal Chatbot
Trains a multimodal chatbot that combines visual and language instructions to generate responses
Movie QA system
This project explores question-answering in movies using various machine learning approaches.
Few-shot learning model
An implementation of the Optimization as a Model for Few-Shot Learning paper in PyTorch
Conversational AI model
Develops an empathy-based conversational AI model for emotional support and coaching
Paragraph QA tools
Tools and codebase for training neural question answering models on multiple paragraphs of text data
Information extractor
Extracts and stores information from various sources using AI models to generate answers.
Few-shot learner
A Python-based few-shot learning framework for medical applications utilizing a visual language model.
ScienceQA toolkit
A dataset and software framework for building multimodal reasoning systems to answer science questions.
VQA prompter
An implementation of a two-stage framework designed to prompt large language models with answer heuristics for knowledge-based visual question answering tasks.
Text classifier
A Python package implementing an interpretable machine learning model for text classification with visualization tools
Notion QA
A Python-based question answering system built on top of Notion's database and OpenAI's API for natural language processing.
Region-of-Interest Training
Training and deploying large language models on computer vision tasks using region-of-interest inputs
Q&A tool
An AI-powered question answering program that uses language models to generate and select the best response.
