view post Post 3794 there's a new multimodal retrieval model in town π€ LlamaIndex released vdr-2b-multi-v1> uses 70% less image tokens, yet outperforming other dse-qwen2 based models> 3x faster inference with less VRAM π¨> shrinkable with matryoshka πͺ> can do cross-lingual retrieval!Collection: llamaindex/visual-document-retrieval-678151d19d2758f78ce910e1 (with models and datasets)Demo: llamaindex/multimodal_vdr_demoLearn more from their blog post here https://huggingface.co/blog/vdr-2b-multilingual π See translation β€οΈ 9 9 π₯ 1 1 + Reply
view post Post 1627 microsoft just released Phi-4 , check it out here : Tonic/Phi-4 hope you like it :-) See translation π₯ 5 5 π 2 2 + Reply
REINFORCE++: A Simple and Efficient Approach for Aligning Large Language Models Paper β’ 2501.03262 β’ Published 16 days ago β’ 82
Deliberation in Latent Space via Differentiable Cache Augmentation Paper β’ 2412.17747 β’ Published 27 days ago β’ 29