Lightning-Universe/lightning-transformers

Issues

• HF compatibility issue

  #304 opened 2 years ago by BenoitDalFerro
  3

• Running inference on GPU

  #258 opened 2 years ago by griff4692
  3

• Ver0.2.4 compatibility PL v1.8

  #303 opened 2 years ago by BenoitDalFerro
  6

• `TransformerDataModule.setup()` run more than once unnecessarily

  #299 opened 2 years ago by RR-28023
  0

• language model load from checkpoint error

  #295 opened 2 years ago by omerarshad
  1

• Jointly train Question Answering and Multiple Choice

  #291 opened 2 years ago by tmquan
  1

• Question answering example throws an exception even if sanity check is skipped

  #233 opened 2 years ago by Pointy-Hat
  10

• Shuffle support

  #272 opened 2 years ago by mahnerak
  2

• How to fix ModuleNotFoundError: No module named 'habana_frameworks.torch'

  #282 opened 2 years ago by zolekode
  1

• pip installation issues

  #285 opened 2 years ago by omerarshad
  4

• Ignore padding tokens when using Translation Task + `padding='max_length'`

  #266 opened 2 years ago by SeanNaren
  6

• Can you demonstrate how to fine-tune a pretrained model on unlabeled data

  #287 opened 2 years ago by turian
  6

• Lightning 1.8 breaks lightning-transformers

  #300 opened 2 years ago by juliusfrost
  1

• Deepspeed sharding and load from checkpoint with custom lightning module - setup() not called during checkpoint loading

  #290 opened 2 years ago by maxzvyagin
  2

• Shuffling support

  #298 opened 2 years ago by juliusfrost
  0

• Continued ViT support

  #250 opened 2 years ago by tanmoyio
  6

• LightningCLI compatibility and Type["AutoModel"]

  #274 opened 2 years ago by mauvilsa
  4

• HFSaveCheckpoint does not work with deepspeed

  #273 opened 2 years ago by jessecambon
  5

• TextClassificationTransformer should log torchmetrics object instead of computed Tensors

  #276 opened 2 years ago by stefan-schroedl
  1

• Speech recognition tasks

  #247 opened 2 years ago by rafaelvp-db
  3

• DeepSpeed integration broken: Tensors must be CUDA and dense

  #259 opened 2 years ago by xinyangz
  3

• What do you guys think about having this in a method called set model?

  #267 opened 2 years ago by espoirMur
  2

• Same Val_precision, Val_recall, Val_accuracy in training.

  #253 opened 2 years ago by tanmoyio
  1

• Move away from configs, to just passing args to the class

  #246 opened 2 years ago by SeanNaren
  0

• omegaconf.errors.ConfigAttributeError: Missing key datafiles

  #236 opened 2 years ago by swcrazyfan
  3

• Add VIT documentation

  #254 opened 2 years ago by SeanNaren
  0

• Should we get rid of Hydra?

  #223 opened 2 years ago by SeanNaren
  8

• Support for `pytorch_lightning.Trainer.predict`

  #260 opened 2 years ago by RR-28023
  1

• Logger instantiation

  #232 opened 2 years ago by pietrolesci
  1

• Optimizers without `weight_decay` produce errors

  #231 opened 2 years ago by pietrolesci
  1

• Multiple Choice Dataset Files not Overwritten

  #224 opened 2 years ago by leonardtang
  2

• [Bug] Problems when training with hydra

  #248 opened 2 years ago by dongheehand
  3

• PL has changed the paramerters of Trainer

  #238 opened 2 years ago by ElderWanng
  4

• basic squad example is not working

  #218 opened 3 years ago by yuvalkirstain
  3

• [bug] trainer got unexpected keyword argument 'truncated_bptt_steps'

  #212 opened 3 years ago by mglowacki100
  4

• Problem with reading in large json files

  #191 opened 3 years ago by itamblyn
  6

• batch by maximum amount of tokens

  #208 opened 3 years ago by yuvalkirstain
  3

• Callback for writing predictions

  #205 opened 3 years ago by yuvalkirstain
  6

• Missing Compressed-Dictionary dependency

  #217 opened 3 years ago by pedrocolon93
  3

• Glue MNLI task fails due to missing 'validation' key in dataset

  #213 opened 3 years ago by mariomeissner
  10

• Cannot find file when using custom data in language modeling

  #202 opened 3 years ago by BramVanroy
  5

• Update imports to match with pl_bolts

  #219 opened 3 years ago by mathemusician
  0

• Add an option to use Huggingface metrics

  #206 opened 3 years ago by yuvalkirstain
  7

• How can we specify `num_classes` for the model?

  #215 opened 3 years ago by jmwoloso
  1

• What format does the model expect for the data(set)?

  #216 opened 3 years ago by jmwoloso
  1

• Mixed precision: scheduler called in the wrong order

  #201 opened 3 years ago by aleSuglia
  1

• Support saving the model so that it can be reloaded from Huggingface

  #188 opened 3 years ago by dfioravanti
  2

• Deprecate text metrics in favour of their counterparts from torchmetrics package.

  #194 opened 3 years ago by karthikrangasai
  1

• Sparseml Integration

  #196 opened 3 years ago by mathemusician
  5

• Can't use a logger

  #185 opened 3 years ago by yuvalkirstain
  4