LLaVA-OneVision-1.5 is a series of fully open-source large multimodal models that achieve advanced performance at a lower cost by training on native resolution images. This model demonstrates excellent performance in multiple multimodal benchmark tests, surpassing competitors such as Qwen2.5-VL.

MultimodalTransformers

Dimple is the first discrete diffusion multimodal large language model (DMLLM) that combines autoregressive and diffusion training paradigms. After training on the same dataset as LLaVA-NEXT, it outperforms LLaVA-NEXT-7B by 3.9%.

MultimodalTransformersEnglish

This model is a GGUF format conversion of llava-hf/llava-1.5-7b-hf, supporting image-to-text generation tasks.

MultimodalGgufEnglish

Spatial-LLaVA-7B is a multimodal model fine-tuned based on the LLaVA model, focusing on improving the ability of spatial relationship reasoning and suitable for multimodal research and chatbot development.

MultimodalSafetensors

Table LLaVA 7B is an open-source multimodal chatbot specifically designed to understand table images and can perform various table-related tasks such as table question answering, table cell description, and structure understanding. This model is based on the LLaVA-v1.5 architecture, using CLIP-ViT-L-336px as the visual encoder and Vicuna-v1.5-7B as the base large language model.

MultimodalSafetensorsEnglish

This project provides weighted/matrix quantized versions of the llava-1.5-13b-hf model, including various quantization types to meet the usage requirements in different scenarios.

MultimodalTransformersEnglish

SESAME is an open-source multimodal model, fine-tuned on various instruction-based image localization (segmentation) datasets based on the LLaVA model.

MultimodalTransformers

This is a static quantized version of the llava-hf/llava-1.5-13b-hf model, offering multiple quantization type options to help users use this vision-language model more efficiently. The model supports image understanding and text generation tasks.

MultimodalTransformersEnglish

ViCA-7B is a vision-language model fine-tuned specifically for visual-spatial reasoning in indoor video environments. Built on the LLaVA-Video-7B-Qwen2 architecture and trained using the ViCA-322K dataset, it emphasizes structured spatial annotation and instruction-based complex reasoning tasks.

MultimodalTransformersEnglish

A fine-tuned version based on the lmms-lab/llava-onevision-qwen2-7b-ov model, supporting video-text-to-text conversion tasks.

MultimodalSafetensorsEnglish

Based on Microsoft LLaVA-Med v1.5 (Mistral 7B) architecture, customized for Turkish-language medical visual question answering (VQA)

MultimodalTransformersMultiple Languages

LLaVAction is a multimodal large language model evaluation and training framework for action recognition, based on the Qwen2 language model architecture, supporting first-person perspective video understanding.

MultimodalTransformersEnglish

LLaVAction is a multimodal large language model for action recognition, based on the Qwen2 language model, trained on the EPIC-KITCHENS-100-MQA dataset.

MultimodalTransformersEnglish

LLaVA-NeXT-Video-7B-hf is a video-based multimodal model capable of processing video and text inputs to generate text outputs.

MultimodalSafetensorsEnglish

This model is based on the Transformers library, and its specific purpose and functionality require further information to be supplemented.

Natural Language ProcessingTransformers

A multimodal foundation model fine-tuned based on LLaVA-Med v1.5 Mistral-7B, optimized for analyzing chest X-ray images and detecting pneumonia

MultimodalTransformers

A multimodal foundation model fine-tuned based on LLaVA-1.5-7B, optimized for plant leaf disease detection and interpretation

MultimodalTransformers

LLaVA-Med is an open-source large vision-language model optimized for biomedical applications, built on the LLaVA framework, enhanced through curriculum learning, and fine-tuned for open-ended biomedical question answering tasks.

MultimodalTransformers

LLaVE-7B is a 7-billion-parameter multimodal embedding model based on LLaVA-OneVision-7B, capable of embedding representations for text, images, multiple images, and videos.

MultimodalTransformersEnglish

LLaVE is a multimodal embedding model based on the LLaVA-OneVision-0.5B model, with a parameter scale of 0.5B, capable of embedding text, images, multiple images, and videos.

MultimodalTransformersEnglish