JonasGeiping/cramming

Issues

• How to load local data

  #46 opened 2 months ago by Doraemonzzz
  2

• Unable to replicate the results using the default command

  #45 opened 2 months ago by shiwenqin
  15

• From PR 43

  #44 opened 3 months ago by JonasGeiping
  5

• Configs for GPT?

  #41 opened 4 months ago by rexdu2003
  2

• Finetuning for token classification

  #40 opened 5 months ago by druskacik
  3

• I run the test command,got this error,how to fix it?looks like no dataset

  #30 opened 5 months ago by leescpeter
  12

• torch._dynamo error on step 2: calling compiler function 'inductor'

  #39 opened 5 months ago by ionutmodo
  7

• TypeError: _load_optimizer() missing 1 required positional argument: 'initial_time'

  #38 opened 6 months ago by vincent-163
  1

• try it on Mac M1 but failed

  #36 opened 6 months ago by yangboz
  2

• can't import cramming

  #37 opened 6 months ago by RobinRojowiec
  2

• Finetuning for SQuAD task

  #35 opened 7 months ago by kisacats
  2

• Uploading trained model to HF/saving in HF format locally

  #34 opened 8 months ago by NtaylorOX
  8

• Question about sparse token prediction

  #33 opened 9 months ago by leo-du
  1

• Issue with torch.compile / dynamo

  #32 opened 9 months ago by spencerfrei
  5

• GLUE evaluation numbers are very poor, if increase the sequence length to 512 and float 32

  #28 opened 10 months ago by tbaggu
  5

• Tutorial for pretrain RoBERTa with custom data

  #31 opened 10 months ago by iambestfeeddddd
  2

• Evaluation failed on MNLI and STSB Datasets for Last1.13release

  #29 opened a year ago by Labyrintbs
  3

• Errors with both the verify installation command as well as the final recipe

  #27 opened a year ago by tatami-galaxy
  2

• Pretraining on a single RTX 3060

  #26 opened a year ago by TahaBinhuraib
  2

• loading checkpoints for using as a huggingface model

  #9 opened a year ago by itay-nakash
  7

• Cola dataset evaluation

  #21 opened a year ago by TahaBinhuraib
  4

• RuntimeError: CUDA error: CUBLAS_STATUS_NOT_INITIALIZED when calling `cublasCreate(handle)` while running evaluation

  #24 opened a year ago by tbaggu
  10

• data preprocessing got failed during tokenization on single GPU

  #22 opened a year ago by tbaggu
  9

• Preprocessed files on S3/Google Drive

  #13 opened a year ago by tals
  2

• Verification command fails on macOS

  #12 opened a year ago by laclouis5
  4

• Can't evaluate

  #16 opened a year ago by TahaBinhuraib
  5

• Reproduce the result when freezing parameters

  #15 opened a year ago by sleeepeer
  8

• Training Step Count

  #11 opened a year ago by ekurtulus
  5

• Suggestion : support Maximal Update Parameterization

  #10 opened a year ago by tfisher98
  2

• TypeError: _new_shared() got an unexpected keyword argument 'device'

  #8 opened a year ago by wccccp
  2

• Preprocessing for final recipe

  #2 opened a year ago by florianmai
  10

• Storage space requirement

  #6 opened a year ago by okpatil4u
  4

• preprocessed c4 dataset?

  #7 opened a year ago by w32zhong
  3