ai-forever/MERA

Issues

• no targets in rummlu and others benchmarks

  #23 opened 6 months ago by thehir0
  12

• [Feature Request] Support for OpenAI ChatCompletion models

  #10 opened 10 months ago by kristaller486
  2

• Как добавить форматирование промпта?

  #4 opened a year ago by dmitrymailk
  2

• влияние промпта на результаты бенчмарков

  #20 opened 6 months ago by vlsav
  0

• 0.4.0 lm-evaluation-harness

  #15 opened 9 months ago by germanjke
  9

• Большие модели, не влезающие в одну карту, не параллелятся на несколько

  #19 opened 6 months ago by preduct0r
  2

• Не авторизоваться на сайте mera.a-ai.ru

  #17 opened 7 months ago by preduct0r
  0

• Скоринг GGUF моделей

  #16 opened 8 months ago by konductor000
  1

• Значения логов бенчмарка

  #14 opened 9 months ago by thehumit
  1

• Как бенчмарк закрытой модели, у которой нету метода loglikelihood?

  #11 opened 9 months ago by chorus12
  1

• empty value rummlu

  #3 opened 10 months ago by mizinovmv
  3

• tokenizer does not have a padding token

  #9 opened 10 months ago by razikov
  1

• Ошибка при сабмитах на mera.a-ai.ru

  #7 opened 10 months ago by GorbetskiyDmitriy
  2

• Как проскорить модель без метода loglikelihood?

  #5 opened 10 months ago by GorbetskiyDmitriy
  1