efeslab/Atom

Issues

• e2e demonstration for bigger models

  #24 opened a month ago by jiwonsong-dev
  1

• Question about KV Cache quantization

  #23 opened 2 months ago by SherrySwift
  3

• Quention about end-to-end efficiency evaluation of Atom

  #22 opened 3 months ago by cokeshao
  2

• Is it possible to add support for other models?

  #21 opened 3 months ago by wlll123456
  1

• Question about the synchronazation in low-precision kernel

  #20 opened 3 months ago by cat538
  2

• TypeError: QLlamaDecoderLayer.forward() got an unexpected keyword argument 'cache_position'

  #19 opened 3 months ago by galenyu
  2

• LLM model load hanging problem

  #18 opened 3 months ago by jimmy-adams
  2

• Question regarding the efficiency evaluation

  #17 opened 6 months ago by FlyFoxPlayer
  3

• The question about calib data

  #15 opened 7 months ago
  3

• How to load quantized weight?

  #14 opened 7 months ago
  1

• RuntimeError when quant llama model

  #12 opened 7 months ago
  10

• AssertionError

  #9 opened 8 months ago by muzi0111
  1

• error：same device

  #8 opened 8 months ago by muzi0111
  1

• the ppl for llama-7b is very large

  #7 opened 8 months ago by priscilla-pan
  3

• not including dynamic quantizaiton when reproducing results, why?

  #6 opened 8 months ago by priscilla-pan
  3

• how to compare the performance with vllm/tgit/lightllm or other llm serving framework?

  #4 opened 9 months ago by irasin
  3

• ppl on ptb

  #3 opened 10 months ago by MrDoghead
  2

• issue with `c4` dataset for eval

  #2 opened 10 months ago by HamidShojanazeri
  1