ComVint
Instruction generator
Creating synthetic visual reasoning instructions to improve the performance of large language models on image-related tasks
The official GitHub page for ''What Makes for Good Visual Instructions? Synthesizing Complex Visual Reasoning Instructions for Visual Instruction Tuning''
18 stars
2 watching
0 forks
Language: Python
last commit: almost 2 years ago Related projects:
| Repository | Description | Stars |
|---|---|---|
 circleradon/osprey | This project presents a new approach to fine-grained visual understanding using pixel-wise mask regions in language instructions | 781 |
 opendatalab/vigc | Autonomously generates high-quality image-text instruction fine-tuning datasets | 91 |
 rowanz/r2c | An open-source project providing PyTorch code and data for a deep learning model that enables visual commonsense reasoning. | 466 |
 deepcs233/visual-cot | A framework for training multi-modal language models with a focus on visual inputs and providing interpretable thoughts. | 162 |
 salt-nlp/llavar | An open-source project that enhances visual instruction tuning for text-rich image understanding by integrating GPT-4 models with multimodal datasets. | 259 |
 rbbrdckybk/ai-art-generator | Automates large batches of AI-generated artwork locally using GPU acceleration. | 633 |
 baai-dcai/visual-instruction-tuning | A dataset and model designed to scale visual instruction tuning using language-only GPT-4 models. | 164 |
| rucaibox/pope | An evaluation framework for detecting object hallucinations in vision-language models | 187 |
 kunpengli1994/vsrn | An open-source PyTorch implementation of a visual semantic reasoning model for image-text matching | 294 |
 aidc-ai/parrot | A method and toolkit for fine-tuning large language models to perform visual instruction tasks in multiple languages. | 34 |
 bigredt/vico | Multi-sense word embeddings learned from visual cooccurrences | 25 |
 rubocop/rubocop-rspec | Analyzes Ruby code for style and syntax errors in RSpec files | 810 |
 lxtgh/omg-seg | Develops an end-to-end model for multiple visual perception and reasoning tasks using a single encoder, decoder, and large language model. | 1,336 |
 kacky24/stylenet | A PyTorch implementation of a framework for generating captions with styles for images and videos. | 63 |
 jtoy/sketchnet | Generates code in a visual programming language using images as input | 40 |