• Home
  • Technology
  • Gaming
  • Entertainment
  • World & Business
  • Science
  • Sports
  • AI
HomeTechnologyGamingEntertainmentWorld & BusinessScienceSportsAI
  • HomeTechnologyGamingEntertainmentWorld & BusinessScienceSportsAI
    • Home
    • Technology
    • Gaming
    • Entertainment
    • World & Business
    • Science
    • Sports
    • AI
    AI

    Researcher Retweets Pangram Model Reveal Surprise

    AI researcher retweets post voicing surprise over Pangram model training and dataset details.

    LB
    AA
    KC
    3 Sources, 24d ago, first seen 24d ago

    TLDR

    Lucas Beyer, an AI researcher at Meta formerly with OpenAI and DeepMind, retweeted a post by @zainhas. The message states disbelief that all details behind the Pangram model were released. It points to information on how the model was trained and the makeup of its training dataset. The retweet circulates this reaction among visible replies. No further confirmation or additional context on the model appears in the packet. The original post focuses only on the scope of the released information about the model's construction.

    Combined views

    12.5K

    3 Sources, first seen 24d ago

    Combined views

    12.5K

    3 Sources, first seen 24d ago

    152 likes
    152 likes
    5 comments
    75 saves
    16 reposts

    Sentiment

    Positive——Negative

    Summary

    Not enough discussion yet.

    No sentiment analysis available yet.

    5 comments
    75 saves
    16 reposts

    Sentiment

    Positive——Negative

    Summary

    Not enough discussion yet.

    No sentiment analysis available yet.

    Today's Rank

    —

    Not ranked yet

    Today's Rank

    —

    Not ranked yet

    3 Sources

    @giffmanaRT @zainhas: i can't believe they revealed all the details behind how the Pangram model works > how it was trained > dataset composition >…
    @kroscooPangram uses old school structured prediction methods, I’m still holding out for a renaissance of the coolest ML tricks.
    @aryaman2020@kroscoo hmm, my personal AGI moment as a researcher was when I replaced BERT + LSTM + CRF with simple BERT + linear head for a parsing task and number went up. always thought CRFs are unnecessary after that

    3 Sources

    @giffmanaRT @zainhas: i can't believe they revealed all the details behind how the Pangram model works > how it was trained > dataset composition >…
    @kroscooPangram uses old school structured prediction methods, I’m still holding out for a renaissance of the coolest ML tricks.
    @aryaman2020@kroscoo hmm, my personal AGI moment as a researcher was when I replaced BERT + LSTM + CRF with simple BERT + linear head for a parsing task and number went up. always thought CRFs are unnecessary after that