Shivanandroy/simpleT5

Issues

• Settings for Multiple GPU Ports usage

  #63 opened 4 months ago by Abubakkar13
  0

• AttributeError: 'LightningDataModule' object has no attribute '_has_setup_TrainerFn.FITTING'

  #57 opened a year ago by tr4wzified
  2

• import error

  #62 opened 5 months ago by PranshuKala
  0

• AttributeError: 'LightningDataModule' object has no attribute '_has_setup_TrainerFn.FITTING'

  #61 opened 7 months ago by lucbuijs
  1

• How to resume training?

  #43 opened 2 years ago by RK-BAKU
  3

• how to train in multi-gpus

  #31 opened 3 years ago by jiangliqin
  2

• How can we evaluate model using some metric?

  #60 opened 9 months ago by MuFazil
  0

• about labels and decoder_attention_mask

  #59 opened a year ago by gitfor20
  0

• 'SimpleT5' object has no attribute 'device'

  #20 opened 3 years ago by athack
  4

• Optimizer

  #58 opened a year ago by peregilk
  0

• Training model on low RAM GPU

  #48 opened 2 years ago by mbledkowski
  1

• Data parallelism technique for Training simplet5 model – CUDA out of memory proplem

  #56 opened a year ago by NashaatRJ
  0

• Pytorch two devices error

  #55 opened 2 years ago by mystsec
  1

• how to generate attention mask seperately?

  #52 opened 2 years ago by Liujingxiu23
  2

• How to use flan-t5?

  #51 opened 2 years ago by Bachstelze
  3

• TPU support

  #27 opened 3 years ago by peregilk
  4

• test and eval sets the same?

  #50 opened 2 years ago by mgh1
  2

• ValueError: text input must of type `str` (single example), `List[str]` (batch or single pretokenized example) or `List[List[str]]` (batch of pretokenized examples).

  #10 opened 3 years ago by j0st
  5

• `do_sample` does nothing

  #49 opened 2 years ago by mgh1
  0

• Would this work to train flan-t5-xxl on multiple GPUs?

  #46 opened 2 years ago by experimarketing
  5

• Possible to feed target as array

  #47 opened 2 years ago by cclegend90
  3

• Push finished model

  #44 opened 2 years ago by peregilk
  2

• Kernel dies every time when I start training the model

  #45 opened 2 years ago by kkrishnan90
  1

• Model.predict on vector of strings

  #29 opened 3 years ago by Nagakiran1
  1

• Does this rep support DeepSpeed?

  #41 opened 2 years ago by benjpau
  0

• Unicode Charecter training issue

  #40 opened 2 years ago by rahat10120141
  5

• ValueError: text input must of type `str` (single example), `List[str]` (batch or single pretokenized example) or `List[List[str]]` (batch of pretokenized examples).

  #39 opened 2 years ago by Ushanjay
  1

• Saved model name not customizable

  #37 opened 2 years ago by ke-lara
  0

• How can i train my t5 (not t5.1 or not mt5) model from scratch?

  #34 opened 2 years ago by 520jefferson
  0

• No Model.Save method?

  #33 opened 3 years ago by simonhughes22
  0

• Train on specific cuda device

  #32 opened 3 years ago by iknoorjobs
  0

• Suppress the Output Models

  #21 opened 3 years ago by bayismet
  6

• how to use simple t5 to build a new language model

  #30 opened 3 years ago by AswiN-7
  0

• Metrics and logging.

  #11 opened 3 years ago by Sripaad
  4

• Is Byt5 Supported?

  #28 opened 3 years ago by tbetth01
  0

• ValueError: transformers.__spec__ is None

  #12 opened 3 years ago by seregadgl20-oss
  9

• Azure Deployment

  #26 opened 3 years ago by mandulasandeep
  0

• model.predict

  #25 opened 3 years ago by saloyiana
  1

• FineTuned Model ( Compute Metrics & Push to Hub )

  #24 opened 3 years ago by saloyiana
  0

• simpleT5 for Grammatical Error Correction

  #15 opened 3 years ago by pradeepdev-1995
  5

• shows object has no attribute 'convert_and_load_onnx_model'

  #16 opened 3 years ago by pradeepdev-1995
  4

• TypeError: forward() got an unexpected keyword argument 'cross_attn_head_mask In onnx_predict function

  #8 opened 3 years ago by farshadfiruzi
  6

• CUDA error or IndexError (on CPU)

  #14 opened 3 years ago by djstrong
  0

• byT5 with version 0.1.2

  #9 opened 3 years ago by kimgerdes
  2

• colab error

  #7 opened 3 years ago by yzhang-github-pub
  6

• Is there any option for fine-tuning mt5 models instead of training from scratch?

  #6 opened 3 years ago by farshadfiruzi
  2

• Add License

  #3 opened 3 years ago by derdanielb
  1

• Support ByT5

  #1 opened 3 years ago by seregadgl
  2

• Is the task string necessary?

  #2 opened 3 years ago by nikogamulin
  1