home.social

#language-modeling — Public Fediverse posts

Live and recent posts from across the Fediverse tagged #language-modeling, aggregated by home.social.

fetched live
  1. RT @NVIDIAAI: Wir haben ein 30B-Modell in zwei Hälften aufgeteilt, um Tokens parallel statt nacheinander zu verarbeiten. Wir stellen vor: Nemotron-Labs-TwoTower, ein Diffusions-Sprachmodell von NVIDIA Research, das auf Nemotron-3-Nano-30B-A3B basiert. So funktioniert es: Eine Hälfte hält den Kontext, die andere schreibt die Tokens, wobei beide die vortrainierte Modellarchitektur nutzen, anstatt ein neues Modell von Grund auf zu trainieren. Wir haben festgestellt, dass es 98,7 % der Qualität des Originalmodells bei 2,42× schnellerer Generierung beibehält. Video

    mehr auf Arint.info

    #AI #DiffusionModel #LanguageModeling #MachineLearning #Nemotron #NVIDIAResearch #arint_info

    https://x.com/NVIDIAAI/status/2072394812301480067#m

  2. RT @NVIDIAAI: Wir haben ein 30B-Modell in zwei Hälften aufgeteilt, um Tokens parallel statt nacheinander zu verarbeiten. Wir stellen vor: Nemotron-Labs-TwoTower, ein Diffusions-Sprachmodell von NVIDIA Research, das auf Nemotron-3-Nano-30B-A3B basiert. So funktioniert es: Eine Hälfte hält den Kontext, die andere schreibt die Tokens, wobei beide die vortrainierte Modellarchitektur nutzen, anstatt ein neues Modell von Grund auf zu trainieren. Wir haben festgestellt, dass es 98,7 % der Qualität des Originalmodells bei 2,42× schnellerer Generierung beibehält. Video

    mehr auf Arint.info

    #AI #DiffusionModel #LanguageModeling #MachineLearning #Nemotron #NVIDIAResearch #arint_info

    https://x.com/NVIDIAAI/status/2072394812301480067#m

  3. 🔮 Behold, the mystical prophecy of CS336! 🌟 Students are invited to glimpse the future of 2026, where they'll be serenaded by the dulcet tones of "Language Modeling from Scratch" every Monday and Wednesday. 🤖 No need to travel through time, just teleport to Skilling Auditorium and prepare to be mind-boggled by the repeat spectacle of springtime syllabi! 🎉
    cs336.stanford.edu/ #CS336 #LanguageModeling #Future2026 #SkillingAuditorium #SpringSyllabi #MindBoggled #HackerNews #ngated

  4. 🔮 Behold, the mystical prophecy of CS336! 🌟 Students are invited to glimpse the future of 2026, where they'll be serenaded by the dulcet tones of "Language Modeling from Scratch" every Monday and Wednesday. 🤖 No need to travel through time, just teleport to Skilling Auditorium and prepare to be mind-boggled by the repeat spectacle of springtime syllabi! 🎉
    cs336.stanford.edu/ #CS336 #LanguageModeling #Future2026 #SkillingAuditorium #SpringSyllabi #MindBoggled #HackerNews #ngated

  5. Andrey Markov & Claude Shannon Counted Letters to Build the First Language-Generation Models Shannon’s said: “OCRO HLI RGWR NMIELWIS” #Shannon #Markov #NLP #AIhistory #LanguageModeling

    Andrey Markov & Claude Shannon...

  6. 🎯 #OuteTTS introduces a novel approach to text-to-speech synthesis using pure #languagemodeling
    🔧 Built on #LLaMa architecture with just 350M parameters, featuring:

    Zero-shot #voicecloning capability
    Integration with #WavTokenizer (75 tokens/sec)
    Local deployment via #llamacpp
    #GGUF format compatibility

    🔍 Technical Implementation:

    Audio tokenization process
    CTC forced alignment
    Structured prompt system
    Temperature-adjustable outputs

    ⚠️ Current Limitations:

    Limited vocabulary range
    String-only input support
    Best performance with shorter sentences
    Variable temperature sensitivity

    github.com/edwko/OuteTTS
    huggingface.co/OuteAI/OuteTTS-

  7. New #languagemodeling #nlp #ai #paper, led by Angelica Chen! We break the steepest MLM training loss drop into *2* phase changes: first in internal grammatical structure, then external capabilities. Big implications for emergence, simplicity bias, and interpretability! arxiv.org/abs/2309.07311

  8. New #languagemodeling #nlp #ai #paper, led by Angelica Chen! We break the steepest MLM training loss drop into *2* phase changes: first in internal grammatical structure, then external capabilities. Big implications for emergence, simplicity bias, and interpretability! arxiv.org/abs/2309.07311

  9. #LanguageModeling is trending, to a large extent because of #ChatGPT. But did you know language modeling has been with us for more than a century? And that it was born of the collaboration of a poet and a mathematician?

    Our engineer Carsten Schnober tells us more:
    blog.esciencecenter.nl/languag

  10. #LanguageModeling is trending, to a large extent because of #ChatGPT. But did you know language modeling has been with us for more than a century? And that it was born of the collaboration of a poet and a mathematician?

    Our engineer Carsten Schnober tells us more:
    blog.esciencecenter.nl/languag