1. X
  2. Moritz Borrett-Laurer
Log inSign up
Moritz Borrett-Laurer
651 posts
Image
user avatar
Moritz Borrett-Laurer
@MoritzLaurer
Machine Learning Engineer, MTS @Cohere, post-training. Ex-@HuggingFace
Paris, France
moritzlaurer.com
Joined June 2017
1,080
Following
2,131
Followers
RepliesRepliesArticlesArticlesMediaMedia

New to X?

Sign up now to get your own personalized timeline!

Create account

By signing up, you agree to the Terms of Service and Privacy Policy, including Cookie Use.

TermsĀ·PrivacyĀ·CookiesĀ·AccessibilityĀ·Ads InfoĀ·Ā© 2026 X Corp.
Don't miss what's happening
People on X are the first to know.
Log inSign up
  • Pinned
    user avatar
    Moritz Borrett-Laurer
    @MoritzLaurer
    Feb 2, 2023
    The amazing thing about "AI" today is that people w limited resources can have real impact. My multilingual 0-shot model was downloaded 432k+ times last month. It cost 0€ to train, built purely on #opensource from @huggingface & others. Happy its useful!
    Image
    MoritzLaurer/mDeBERTa-v3-base-mnli-xnli Ā· Hugging Face
    From huggingface.co
    50K050K
  • user avatar
    Moritz Borrett-Laurer
    @MoritzLaurer
    Jul 22, 2023
    Microsoft and Tsinghua U. claim to have found the "Successor to Transformer for Large Language Models": RetNet. They claim better language modelling performance, with 3.4x lower memory consumption, 8.4x higher throughput, 15.6x lower latency. 1/2
    Image
    Image
    Image
    Image
    222K0222K
  • user avatar
    Moritz Borrett-Laurer
    @MoritzLaurer
    Dec 11, 2023
    Professional update: I've started as a ML Engineer at @HuggingFace šŸ¤—! The photo is from 3 years ago when the HF team sent me stickers for a small project on COVID-19 news I built with Transformers. The sticker has remained on my computer ever since. 3 years, some model
    Image
    54K054K
  • user avatar
    Moritz Borrett-Laurer
    @MoritzLaurer
    Mar 9, 2023
    Never finetune BERT-base! Take a model from the list below. @IBMResearch @LChoshen have ranked 2500+ #opensource models from the @huggingface hub. The best models are +8% better than BERT-base & similar size. Happy that 2 of my models are in the top 10: ibm.github.io/model-recyclin… 1/
    Image
    microsoft_deberta-v3-base
    From ibm.github.io
    24K024K
  • user avatar
    Moritz Borrett-Laurer
    @MoritzLaurer
    Sep 8, 2022
    šŸ†•Dataset for multilingual zero-shot classification & NLIšŸ†•: 2.7 million NLI texts in 26 languages spoken by more than 4 billion people including šŸ‡ØšŸ‡³šŸ‡®šŸ‡³šŸ‡·šŸ‡ŗšŸ‡§šŸ‡·šŸ‡«šŸ‡·šŸ‡©šŸ‡ŖšŸ‡ŖšŸ‡øšŸ‡®šŸ‡·šŸ‡ÆšŸ‡µšŸ‡®šŸ‡©šŸ‡»šŸ‡³šŸ‡§šŸ‡©šŸ‡®šŸ‡¹šŸ‡°šŸ‡µšŸ‡µšŸ‡±šŸ‡ŗšŸ‡¦šŸ‡³šŸ‡±šŸ‡øšŸ‡ŖšŸ‡¹šŸ‡·šŸ‡µšŸ‡°šŸ‡ŖšŸ‡¬šŸ‡®šŸ‡±šŸ‡°šŸ‡ŖšŸ‡¹šŸ‡æšŸ‡±šŸ‡°. Freely available on @huggingface: huggingface.co/datasets/Morit… #opensource #NLProc
    Image
    MoritzLaurer/multilingual-NLI-26lang-2mil7 Ā· Datasets at Hugging Face
    From huggingface.co
  • user avatar
    Moritz Borrett-Laurer
    @MoritzLaurer
    Feb 16, 2024
    Should you fine-tune your own model or use an LLM API? We show how you can combine the best of both worlds in a new @huggingface blog post: ā€œSynthetic data: save money, time and carbon with open sourceā€ By training a specialized model with synthetic data, you can: šŸ’ø reduce
    28K028K
  • user avatar
    Moritz Borrett-Laurer
    @MoritzLaurer
    Jul 12, 2023
    I have no idea who has been downloading my 0-shot model 4+ million times this month via @huggingface but that's v motivating. I'm working on an update, should be done in a few weeks. Encoder 0-shot models are very efficient: they can run on a Raspberry Pi, no A100 GPU needed 1/2
    Image
    40K040K
  • user avatar
    Moritz Borrett-Laurer
    @MoritzLaurer
    Mar 28, 2020
    Thanks @huggingface for democratising machine learning! Their new BART model enabled me to summarise the Communist Manifesto, Orwell's 1984 and Darwin's Origin of Species in a few hours. Results are impressive! See here & try it yourself: colab.research.google.com/drive/1iAIFX1Q… #nlp #python
    user avatar
    Hugging Face
    @huggingface
    Mar 24, 2020
    Bored at home? Need a new friend? Hang out with BART, the newest model available in transformers (thx @sam_shleifer) , with the hefty 2.6 release (notes: github.com/huggingface/tr…). Now you can get state-of-the-art summarization with a few lines of code: šŸ‘‡šŸ‘‡šŸ‘‡
    Image
  • user avatar
    Moritz Borrett-Laurer
    @MoritzLaurer
    Mar 2, 2023
    šŸ†• Chinese @BaiduResearch and @PaddlePaddle recently open-sourced their multilingual ERNIE-m model, outperforming @metaai's XLM-RoBERTa-large. You can now download the 0-shot version for classifying text in 100+ languages on @huggingface here huggingface.co/MoritzLaurer/e… (1/2)
    Image
    MoritzLaurer/ernie-m-large-mnli-xnli Ā· Hugging Face
    From huggingface.co
    26K026K
  • user avatar
    Moritz Borrett-Laurer
    @MoritzLaurer
    Jan 8, 2024
    šŸ†• Sharing a new paper and efficient 0.1B zeroshot classifiers, directly compatible with the @huggingface zeroshot pipeline. The universal classifiers are trained on 33 datasets with 389 diverse classes. The paper provides a step-by-step guide with reusable Jupyter notebooks for
    Image
    Image
    12K012K
  • user avatar
    Moritz Borrett-Laurer
    @MoritzLaurer
    Feb 10, 2023
    šŸ†• multilingual 0-shot model for 116 languages now available on @huggingface: It's based on @MetaAI's newest XLM-V model, which has a larger and better vocabulary of 1 million tokens to better represent more languages. You can test it here: huggingface.co/MoritzLaurer/x… (1/2)
    Image
    MoritzLaurer/xlm-v-base-mnli-xnli Ā· Hugging Face
    From huggingface.co
    18K018K
  • user avatar
    Moritz Borrett-Laurer
    @MoritzLaurer
    Sep 29, 2023
    šŸ†• Releasing a new 0-shot model on the @huggingface hub, trained on 27 tasks, 310 classes, ~1.3 million texts! šŸ¤– My new deberta-v3-zeroshot-v1 is specifically designed for 0-shot classification. Free download: huggingface.co/MoritzLaurer/d… āš™ļø Key properties:
    Image
    Image
    23K023K
  • user avatar
    Moritz Borrett-Laurer
    @MoritzLaurer
    Jul 22, 2023
    Replying to @MoritzLaurer
    Their new RetNet combines lessons from Transformers with RNNs. This new architecture would actually be a big deal, if other teams can reproduce this. Paper: arxiv.org/pdf/2307.08621…
    7.9K07.9K
  • user avatar
    Moritz Borrett-Laurer
    @MoritzLaurer
    Jan 11, 2024
    šŸ¤ New 0.02B, 25 MB tiny zeroshot classifiers for edge device use-cases on @huggingface! The xtremedistil ONNX quantized version is only 13 MB and very fast on CPUs. Without quantization, it has a throughput of ~4000 full sentences (! not just tokens) per second on an A10G with
    29K029K
Advertisement
Advertisement