LLM Reference

LLaVA 1.6 Vicuna 7B

About

LLaVA 1.6 Vicuna 7B is a sophisticated multimodal large language model that integrates a Vicuna-7B language model with a high-performing vision encoder, such as CLIP's ViT-L/14. This innovative combination enables it to process and understand both text and image inputs, making it ideal for tasks like image captioning, visual question answering, and multimodal chatbot interactions. It features enhanced image resolution capabilities, improved OCR, and logical reasoning skills, all while maintaining data efficiency. Despite these advancements, users should be aware of its performance variability and computational demands when handling high-resolution images. Being open-source, it invites community collaboration for further enhancements.

Capabilities

MultimodalFunction CallingTool UseJSON Mode

Providers(1)

ProviderInput (per 1M)Output (per 1M)Type
Replicate API
Serverless

Specifications

FamilyLLaVA 1.6
Parameters7B
ArchitectureDecoder Only
Specializationgeneral