Constantin Eichenberg
YOU?
Author Swipe
View article: Hierarchical Autoregressive Transformers: Combining Byte- and Word-Level Processing for Robust, Adaptable Language Models
Hierarchical Autoregressive Transformers: Combining Byte- and Word-Level Processing for Robust, Adaptable Language Models Open
Tokenization is a fundamental step in natural language processing, breaking text into units that computational models can process. While learned subword tokenizers have become the de-facto standard, they present challenges such as large vo…
View article: Divergent Token Metrics: Measuring degradation to prune away LLM components -- and optimize quantization
Divergent Token Metrics: Measuring degradation to prune away LLM components -- and optimize quantization Open
Large Language Models (LLMs) have reshaped natural language processing with their impressive capabilities. However, their ever-increasing size has raised concerns about their effective deployment and the need for LLM compression. This stud…
View article: MultiFusion: Fusing Pre-Trained Models for Multi-Lingual, Multi-Modal Image Generation
MultiFusion: Fusing Pre-Trained Models for Multi-Lingual, Multi-Modal Image Generation Open
The recent popularity of text-to-image diffusion models (DM) can largely be attributed to the intuitive interface they provide to users. The intended generation can be expressed in natural language, with the model producing faithful interp…
View article: M-VADER: A Model for Diffusion with Multimodal Context
M-VADER: A Model for Diffusion with Multimodal Context Open
We introduce M-VADER: a diffusion model (DM) for image generation where the output can be specified using arbitrary combinations of images and text. We show how M-VADER enables the generation of images specified using combinations of image…
View article: MAGMA – Multimodal Augmentation of Generative Models through Adapter-based Finetuning
MAGMA – Multimodal Augmentation of Generative Models through Adapter-based Finetuning Open
Large-scale pretraining is fast becoming the norm in Vision-Language (VL) modeling. However, prevailing VL approaches are limited by the requirement for labeled data and the use of complex multi-step pretraining objectives. We present MAGM…
View article: MAGMA -- Multimodal Augmentation of Generative Models through Adapter-based Finetuning
MAGMA -- Multimodal Augmentation of Generative Models through Adapter-based Finetuning Open
Large-scale pretraining is fast becoming the norm in Vision-Language (VL) modeling. However, prevailing VL approaches are limited by the requirement for labeled data and the use of complex multi-step pretraining objectives. We present MAGM…