LLAMA 3.3 70B Fully Tested ( Coding / Logic and reasoning / Math )

Поделиться
HTML-код
  • Опубликовано: 15 дек 2024

Комментарии • 4

  • @rickhoro
    @rickhoro 5 дней назад

    I like the format. Rather than wasting a bunch of time you tested a bunch of real world related stuff in short order. Thanks. Subscribed.

  • @NLPprompter
    @NLPprompter 5 дней назад

    next year prompt:
    here is my dna, generate me my new kidney
    openai_tissue_sequencer

  • @gaiustacitus4242
    @gaiustacitus4242 День назад

    The tests always ask AI to produce an application that multiple versions of source code already exist for. They never ask it to code a truly unique application. Why not? Because AI fails miserably at creating applications for which it has no sample code to use as a baseline model.

  • @nguyenanhnguyen7658
    @nguyenanhnguyen7658 6 дней назад

    Never trust a statistical model :D haha. Funny 3 years of hundreds of billions of US that go nowhere. Let the Chinese do the job.