Dan Saattrup Nielsen's picture

Dan Saattrup Nielsen

saattrupdan

·

https://saattrupdan.com

AI & ML interests

NLP for low-resource languages.

Organizations

saattrupdan's activity

upvoted a paper 9 days ago

Magpie: Alignment Data Synthesis from Scratch by Prompting Aligned LLMs with Nothing

Paper • 2406.08464 • Published Jun 12 • 65

upvoted an article 9 days ago

Article

🇮🇹🇯🇵🇧🇷 Generating multilingual instruction datasets with Magpie 🐦‍⬛

By

•

20 days ago

• 18

upvoted a collection about 1 month ago

VPTQ Llama 3.1 70B Instruct without finetune

arxiv.org/abs/2409.17066, VPTQ Llama 3.1 70B without finetune • 9 items • Updated 23 days ago • 1

upvoted a collection 5 months ago

Nemotron 4 340B

Nemotron-4: open models for Synthetic Data Generation (SDG). Includes Base, Instruct, and Reward models. • 4 items • Updated 7 days ago • 157

upvoted an article 5 months ago

Article

Uncensor any LLM with abliteration

By

•

Jun 13

• 362

upvoted 2 collections 7 months ago

Meta Llama 3

This collection hosts the transformers and original repos of the Meta Llama 3 and Llama Guard 2 releases • 5 items • Updated Sep 25 • 681

Czech evaluation datasets

This collections should contain czech evaluation datasets • 8 items • Updated Jan 14 • 3

upvoted 5 collections 8 months ago

OpenCulture

A multilingual dataset of public domain books and newspapers. • 27 items • Updated 3 days ago • 113

Danish Benchmarks

Benchmarks for evaluating Danish Models. • 2 items • Updated Jun 9 • 4

State-of-the-art Danish Models

These models constitute state-of-the-art models for Danish within their respective domain (highlighted below the model). • 13 items • Updated Apr 11 • 10

occiglot-eu5-7b-v0.1

First release of 7B LLMs models for the 5 biggest European languages. All models initialised from mistral-7b-v0.1. • 10 items • Updated Mar 7 • 21

Transformers compatible Mamba

This release includes the `mamba` repositories compatible with the `transformers` library • 5 items • Updated Mar 6 • 36

upvoted a collection 9 months ago

Qwen1.5

Qwen1.5 is the improved version of Qwen, the large language model series developed by Alibaba Cloud. • 55 items • Updated Sep 18 • 206

upvoted a paper 10 months ago

Self-Rewarding Language Models

Paper • 2401.10020 • Published Jan 18 • 143

upvoted a collection about 1 year ago

Zephyr 7B

Models, datasets, and demos associated with Zephyr 7B. For code to train the models, see: https://github.com/huggingface/alignment-handbook • 9 items • Updated Apr 12 • 145