Stockholm, Sweden

Markus Sagen

Expert
@MarkusSagen

Curious & Problem Solver

Master-Thesis-Multilingual-Longformer. Master thesis with code investigating methods for incorporating long-context reasoning in low-resource languages, without the need to pre-train from scratch. We investigated if multilingual models could inherit these properties by making it an Efficient Transformer (s.a. the Longformer architecture).

35

Surmize. Web application for summarizing and asking questions on your own documents. Summarization and QA application in FastAPI, React and Transformers

8

rejx. Deal with `.rej` files with reasonable pain

6

xlm-roberta-longformer-base-4096.

3

Pytorch_Examples.

2

doom-emacs-for-writers. Doom Emacs config for non-programmers. Just make smart notes with Org and Org-Roam

2

SML_Project. Classifier of songs to predict good songs

2

cosFormer. Official implementation of cosformer-attention in cosFormer: Rethinking Softmax in Attention

1

datacards. Append missing model cards to datasets

1

docs. A collaborative note taking, wiki and documentation platform that scales. Built with Django and React. Opensource alternative to Notion or Outline.

1

sidekick. Sidekick code for the platform

1

lm-evaluation-harness. A framework for few-shot evaluation of autoregressive language models.

1

calibration. Code and datasets for the EMNLP 2020 paper "Calibration of Pre-trained Transformers"

1

comcrawl. A python utility for downloading Common Crawl data

1

mmf. A modular framework for vision & language multimodal research from Facebook AI Research (FAIR)

1

Awesome-Multimodal-Research. A curated list of Multimodal Related Research.

1

docarray. The data structure for unstructured data

1

NN_calibration. Calibration of Convolutional Neural Networks

1

TextGAIL. Python

1

TranslateAlignRetrieve. Python-based implementation of the Translate-Align-Retrieve method to automatically translate the SQuAD Dataset to Spanish.

1

CV. Information and collection about repos, reports and more used for CVs and Personal letters

1

example_cml. Python

1

budibase. Budibase is an open-source low-code platform for creating internal apps in minutes. Supports PostgreSQL, MySQL, MSSQL, MongoDB, Rest API, Docker, K8s 🚀

1

sucx3_ner. Python

1

pytorch-adapt. Domain adaptation made easy. Fully featured, modular, and customizable. Written in PyTorch.

1

papers-search. Neural search engine for AI papers

1

examples. Jina examples and demos to help you get started

1

MAT. The official implementation of the Molecule Attention Transformer.

1

Tatoeba-Challenge. Makefile

1

tensorflow-tabnet. Improved TabNet for TensorFlow

1

ScandEval. Evaluation of language models on mono- or multilingual Scandinavian language tasks.

1

community-code. Shared code for data transforms, calling your model or just having fun with deep learning

1

sms-explainable-antibiotics-VGR. Explainable Antibiotic Perscription for the Svensk Medicinskt Språkdatalabb Vinnova project (also known as MedBERT). A cooperation between Region Västra Götaland Folktandvården and Peltarion

1

NERDA. Framework for fine-tuning pretrained transformers for Named-Entity Recognition (NER) tasks

1

jiant. jiant is an nlp toolkit

1

automated-string-cleaning. Repository for my master thesis on automated string handling

1

protein_search. The neural search engine for proteins.

1

T5-QA. Python

1

pygaggle. a gaggle of deep neural architectures for text ranking and question answering, designed for Pyserini

1

quality.

1

MLOps-Basics. Jupyter Notebook

1

GAN-for-tabular-data. We well know GANs for success in the realistic image generation. However, they can be applied in tabular data generation. We will review and examine some recent papers about tabular GANs in action.

1

sms-fall-injury-prediction-RH. Fall injury prediction and identification from free text journal entries for the Svensk Medicinskt Språkdatalabb Vinnova project (also known as MedBERT). A cooperation between Region Halland, Peltarion, and RISE

1

HM-Fashion-image-neural-search. H&M Fashion Image similarity search with Weaviate and DocArray

1

chemicalx. A PyTorch and TorchDrug based deep learning library for drug pair scoring.

1

ai_nlp_transformer_models_benchmarking. Investigates what impacts the GPU allocation, inference and training time for different Transformer based models

1

keras-io-transformers. Jupyter Notebook

1

pyserini. Pyserini is a Python toolkit for reproducible information retrieval research with sparse and dense representations.

1

RATransformers. 🐭RATransformers - Make your transformer (like BERT, RoBERTa, GPT-2 and T5) Relation Aware!

1

stockpredictor-streamlit. Simple demo for predicting sample stock prices with streamlit and prothet

1

mltoolskit. Toolkit for ML or DS projects to reduce the number of repetitions made between projects

1

hoot-ot. Smart notifier for all your shopping or curiosity needs

1

sveltekit-tailwind-typescript-jit-template. CSS

1

huggingface-models-overview. Recommendation on the available deep learning models - and where to find them. Use what for where?

1

Transformers-LM-Benchmark. Benchmark training and inference time for Transformer models on Huggingface

1

keras_classifier. JavaScript

1

homepage. CSS

1

MarkusSagen.github.io. CSS

1
58
Apply