# Visual Language Understanding

Arabic image captioning model fine-tuned based on BLIP architecture, specifically optimized for the Flickr8k Arabic dataset

Skywork-R1V2-38B is currently the most advanced open-source multimodal reasoning model, demonstrating outstanding performance in multiple benchmark tests with robust visual reasoning and text comprehension capabilities.

EMOVA is an end-to-end omni-modal large language model that supports visual, auditory, and speech functions, capable of generating text and speech responses with emotional control.

This model is a multimodal vision-language model converted from Qwen2.5-VL-7B-Instruct, supporting visual question answering tasks.

A vision-language model based on the Qwen2.5-VL-7B-Instruct model with GPTQ-INT8 quantization

Qwen2.5-VL-72B-Instruct is a 72B-parameter multimodal large model that supports vision-language tasks, capable of understanding and generating text related to images.

This is an image-to-text conversion model capable of processing both image and text inputs to generate corresponding text outputs.

This model is a GGUF format conversion based on Qwen2-VL-7B-Captioner-Relaxed, optimized for image-to-text tasks and supports running via tools like llama.cpp and Koboldcpp.

DeepSeer is a vision-language model developed based on the DeepSeek-R1 model, supporting chain-of-thought reasoning and trained through dialogue templates for visual models.

Emu3 is a multimodal model developed by the Beijing Academy of Artificial Intelligence, trained solely by predicting the next token, supporting image, text, and video processing.

Llama-3-EvoVLM-JP-v2 is an experimental general-purpose Japanese vision-language model that supports interleaved input of text and images. This model was created using an evolutionary model fusion approach.

Cephalo is a series of vision-based large language models (V-LLMs) focused on multimodal materials science, designed to integrate visual and linguistic data to facilitate advanced understanding and interaction in human-computer or multi-agent AI frameworks.

An open-source chatbot model trained by fine-tuning the entire model on open-source data, which can be used for research on multimodal models and chatbots.

Cephalo is a series of vision-based large language models (V-LLMs) focused on multimodal materials science, designed to integrate visual and linguistic data to facilitate advanced understanding and interaction in human-computer or multi-agent AI frameworks.

LLaVA-JP is a multimodal vision-language model that supports Japanese, capable of understanding and generating descriptions and dialogues about input images.

This is a transformers-based image-to-text conversion model, specific functionalities require further details

LLaVA-v1.5-13B-DPO is a vision-language model based on the LLaVA framework, trained with Direct Preference Optimization (DPO) and converted to GGUF quantized format to improve inference efficiency.

LLaVA is an open-source multimodal chatbot, fine-tuned based on a large language model, supporting interactions with both images and text.

MoE-LLaVA is a large-scale vision-language model based on a mixture of experts architecture, achieving efficient multimodal learning through sparsely activated parameters.

TinyLLaVA is a compact large-scale multimodal model framework focused on vision-language tasks, featuring small parameter size yet excellent performance.

LLaVA-Lightning-7B is a multimodal model based on LLaMA-7B, achieving efficient vision-language task processing through delta parameter tuning.

Pix2Struct is a visual question answering model fine-tuned for OCR-VQA tasks, capable of parsing textual content in images and answering questions