Commit History

Autor SHA1 Mensaxe Data
  Matthias Reso 5bceb44542 Fix fsdp_config.pure_bf16 flag in README hai 1 ano
  Matthias Reso 53fd82355f Add missing changes hai 1 ano
  Matthias Reso 07bcffbf50 clean up unit tests + add batching test hai 1 ano
  Matthias Reso 4c225c65eb Fix order of concat vs sampler hai 1 ano
  Matthias Reso f9756ca79d Added packing test for samsum hai 1 ano
  Matthias Reso 5a359b7bf2 Fix sampler vs batch_sampler hai 1 ano
  Matthias Reso fe8122daf1 Adapt alpaca dataset to ConcatDataset hai 1 ano
  Matthias Reso 5da84b2913 Fix usage of dataclass for train_config and fsdp_config hai 1 ano
  Matthias Reso aa5dee241a Fix unit test to reflect batch packing hai 1 ano
  Matthias Reso 8620ab8ac2 Fix invalid labels for context in custom dataset/oasst1 hai 1 ano
  Matthias Reso 52c417b7d5 Merge branch 'fix/invalidate_label_for_chat' into feature/length_based_batch_sampling hai 1 ano
  Matthias Reso 653a79e3dd Invalidate context in labels for samsum + grammar hai 1 ano
  Matthias Reso d3015b4c80 Remove max_word from alpaca; lets deal tokenizer deal with truncation hai 1 ano
  Matthias Reso a647955fc8 Make packing/padding a training setting hai 1 ano
  Matthias Reso eafea7b366 Invalidate labels in dialog dataset to disable loss hai 1 ano
  Matthias Reso cc8cc0d3c3 fix grammar dataset %!s(int64=2) %!d(string=hai) anos
  Matthias Reso 2e4bd2a665 Resize vocab size to fix idx error %!s(int64=2) %!d(string=hai) anos
  Matthias Reso 10f9367e56 fix missing labels in datasets %!s(int64=2) %!d(string=hai) anos
  Matthias Reso f2d02a9362 Add unit test for dis sampler %!s(int64=2) %!d(string=hai) anos
  Matthias Reso be63d9ec39 Remove padding in alpaca ds; remove concat in grammar %!s(int64=2) %!d(string=hai) anos
  Matthias Reso ddf58d205d Added dist length based batch sampler %!s(int64=2) %!d(string=hai) anos
  Matthias Reso ca41c1c697 Adjust tests to len based batch sampling %!s(int64=2) %!d(string=hai) anos
  Matthias Reso 97a7871f4b Fix seed in test %!s(int64=2) %!d(string=hai) anos
  Matthias Reso 17209cdabd Add license to test file %!s(int64=2) %!d(string=hai) anos
  Matthias Reso d5054ecae9 Move sampler test %!s(int64=2) %!d(string=hai) anos
  Matthias Reso 63ce4ce7f6 Moved sampler to data submodule %!s(int64=2) %!d(string=hai) anos
  Matthias Reso f620f3589d Adds length based batch sampler %!s(int64=2) %!d(string=hai) anos
  Matthias Reso 8ac44ef3be Fix vocab size mismatch in inference due to added pad token %!s(int64=2) %!d(string=hai) anos
  Geeta Chauhan 40b32ba559 Fix tqdm bar not change length after terminal is resized (#201) %!s(int64=2) %!d(string=hai) anos
  hongbo.mo 6217635e87 Fix tqdm bar not change length after terminal is resized %!s(int64=2) %!d(string=hai) anos