English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Tools annual

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Tools monthly

Global: Zendy Plus monthly

Presents the access of premium content as premium feature

Premium Content

Global: Zendy Plus annual

Global: Zendy Free Plan

How to efficiently transform large language models (LLMs) into instructionfollowers is recently a popular research direction, while training LLM formulti-modal reasoning remains less explored. Although the recent LLaMA-Adapterdemonstrates the potential to handle visual inputs with LLMs, it still cannotgeneralize well to open-ended visual instructions and lags behind GPT-4. Inthis paper, we present LLaMA-Adapter V2, a parameter-efficient visualinstruction model. Specifically, we first augment LLaMA-Adapter by unlockingmore learnable parameters (e.g., norm, bias and scale), which distribute theinstruction-following ability across the entire LLaMA model besides adapters.Secondly, we propose an early fusion strategy to feed visual tokens only intothe early LLM layers, contributing to better visual knowledge incorporation.Thirdly, a joint training paradigm of image-text pairs andinstruction-following data is introduced by optimizing disjoint groups oflearnable parameters. This strategy effectively alleviates the interferencebetween the two tasks of image-text alignment and instruction following andachieves strong multi-modal reasoning with only a small-scale image-text andinstruction dataset. During inference, we incorporate additional expert models(e.g. captioning/OCR systems) into LLaMA-Adapter to further enhance its imageunderstanding capability without incurring training costs. Compared to theoriginal LLaMA-Adapter, our LLaMA-Adapter V2 can perform open-ended multi-modalinstructions by merely introducing 14M parameters over LLaMA. The newlydesigned framework also exhibits stronger language-only instruction-followingcapabilities and even excels in chat interactions. Our code and models areavailable at https://github.com/ZrrSkywalker/LLaMA-Adapter.

LLaMA-Adapter V2: Parameter-Efficient Visual Instruction Model