# Visual Reasoning

Cosmos-Reason1-7B is a 7B-parameter foundational model released by NVIDIA, specializing in image-to-text tasks.

A vision-language model based on ViLT architecture, fine-tuned specifically for GQA visual reasoning tasks

This is a multimodal model based on Qwen2.5-VL-7B-Instruct, supporting visual question answering tasks, converted to MLX format for efficient operation on Apple chips.

This model is a multimodal vision-language model converted from Qwen2.5-VL-7B-Instruct, supporting 8-bit quantization and suitable for visual question-answering tasks.

Idefics3 is an open-source multimodal model capable of processing arbitrary sequences of image and text inputs to generate text outputs. It shows significant improvements in OCR, document understanding, and visual reasoning.

ChartQA is a visual question answering model focused on extracting information from charts and answering related questions.