epfml/landmark-attention

Issues

• Fine-tuned Weights Doesn't Work

  #16 opened 6 months ago by baotruyenthach
  0

• How is TransformerFAM different from Landmark Attention?

  #15 opened 7 months ago by Rock-Anderson
  0

• Question about training stability

  #14 opened a year ago by meme-virus
  3

• [Q]: Quantize Per-Trained model Using QLoRa or LoRa , PFET Technique

  #13 opened a year ago by deep-matter
  0

• Will this work with Qlora and 4bit inference?

  #1 opened a year ago by jordancole21
  36

• when using auto_insert_landmarks, it appears to be view problems

  #12 opened a year ago by EzTan1
  1

• Assertion `srcIndex < srcSelectDimSize` failed while running test.

  #11 opened a year ago by L16H7
  2

• Peft module

  #10 opened a year ago by NicolasMejiaPetit
  0

• How much VRAM do you need to run Inference?

  #9 opened a year ago by FFFiend
  0

• Is true attention (full-context) calculation needed during training?

  #2 opened a year ago by Andrey36652
  3

• Question about Training vs LLaMA Fine Tuning

  #6 opened a year ago by eugenepentland
  2

• Unlimited context length and no prompt token limit

  #8 opened a year ago by ekolawole
  0

• hello! i reached out on twitter before the release and received the link here when you guys dropped this

  #5 opened a year ago by Alignment-Lab-AI
  0

• Readme clarification

  #3 opened a year ago by StrangeTcy
  1

• model_name_or_path defaults

  #4 opened a year ago by StrangeTcy
  0