princeton-nlp/LESS

Issues

• step 2 when run "/get_train_lora_grads.sh", load the optimizer.pt error is happend

  #4 opened a year ago by victorjiax
  19

• Discrepancy in Model Performance When Reproducing Experiment

  #25 opened 5 months ago by tangzhy
  2

• Is this normal? There are 15011 examples in the dataset，but it only completed up to 199/15011?

  #31 opened 3 months ago by kuang1216
  0

• load the optimizer.bin

  #28 opened 3 months ago by kuang1216
  0

• `open_instruct` used in `evaluation/eval/utils.py` is out-of-date

  #30 opened 3 months ago by cafeii
  0

• RuntimeError: Expected all tensors to be on the same device, but found at least two devices, cuda:6 and cuda:0!

  #19 opened 3 months ago by Haruka1307
  3

• RDS baseline?

  #29 opened 3 months ago by hamishivi
  2

• Why does the CSV extracted according to the steps show all the scores as nan?

  #27 opened 4 months ago by SsmallSong
  1

• At step1, single GPU works while multiple GPUs get stuck.

  #22 opened 6 months ago by timturing
  1

• How to get tydiqa-one-shot.json

  #26 opened 4 months ago by ppalantir
  0

• Question about influence scoring (regarding cosine-similarity)

  #24 opened 5 months ago by DSKSD
  2

• OOM Errors in Step 3.1

  #20 opened 7 months ago by ChangyuChen347
  3

• Error with fast_jl

  #16 opened 7 months ago by simplelifetime
  1

• a small mistake in the code

  #15 opened 7 months ago by ZZZZZccccc123
  3

• Question about accuracy

  #12 opened 7 months ago by fyf3
  1

• some question about calulate socre？

  #14 opened 7 months ago by smashfan
  1

• Question about running stage 3 script

  #9 opened 7 months ago by 2003pro
  3

• Using Multiple GPUs

  #7 opened 7 months ago by SD325
  3

• Are samples used for warmup training and gradient calculation the same?

  #5 opened 7 months ago by ZigeW
  1

• Questions about the influence value

  #21 opened 7 months ago by simplelifetime
  1

• No optimizer.bin in Step 2

  #18 opened 8 months ago by mihara-bot
  3

• Step 2: Building the gradient datastore: FileNotFoundError: [Errno 2] No such file or directory: '../out/llama2-7b-p0.05-lora-seed3/checkpoint-1688/optimizer.bin'

  #13 opened 9 months ago by minchaoyue
  3

• Error while deserializing header: InvalidHeaderDeserialization

  #17 opened 8 months ago by shangqing-liu
  0

• FileNotFoundError: optimizer.bin and KeyError: 'base_model.model.model.layers.0.self_attn.q_proj.lora_A.default.weight'

  #10 opened 9 months ago by yinochaos
  2

• data selection results

  #6 opened 10 months ago by wang-debug
  2

• What if we have the optim states of specific tasks at the beginning?

  #2 opened a year ago by suhmily
  2

• Does it support baichuan or chatglm?

  #1 opened a year ago by amulil
  2

• A concerned question about adding a new baseline.

  #3 opened a year ago by FanqingM
  2