Lightning-Universe / lightning-transformers

Flexible components pairing 🤗 Transformers with :zap: Pytorch Lightning

https://lightning-transformers.readthedocs.io

Lightning-Universe/lightning-transformers Issues

HF compatibility issue
Closed a year ago3
Running inference on GPU
Closed a year ago3
Ver0.2.4 compatibility PL v1.8
Closed a year ago6
`TransformerDataModule.setup()` run more than once unnecessarily
Updated a year ago
language model load from checkpoint error
Closed a year ago1
Jointly train Question Answering and Multiple Choice
Updated a year ago1
Question answering example throws an exception even if sanity check is skipped
Closed a year ago10
Shuffle support
Closed a year ago2
How to fix ModuleNotFoundError: No module named 'habana_frameworks.torch'
Closed a year ago1
pip installation issues
Closed a year ago4
Ignore padding tokens when using Translation Task + `padding='max_length'`
Updated a year ago6
Can you demonstrate how to fine-tune a pretrained model on unlabeled data
Updated a year ago6
Lightning 1.8 breaks lightning-transformers
Closed a year ago1
Deepspeed sharding and load from checkpoint with custom lightning module - setup() not called during checkpoint loading
Updated a year ago2
Shuffling support
Updated a year ago
Continued ViT support
Closed 2 years ago6
LightningCLI compatibility and Type["AutoModel"]
Closed 2 years ago4
HFSaveCheckpoint does not work with deepspeed
Updated 2 years ago5
TextClassificationTransformer should log torchmetrics object instead of computed Tensors
Updated 2 years ago1
Speech recognition tasks
Closed 2 years ago3
DeepSpeed integration broken: Tensors must be CUDA and dense
Closed 2 years ago3
What do you guys think about having this in a method called set model?
Closed 2 years ago2
Same Val_precision, Val_recall, Val_accuracy in training.
Closed 2 years ago1
Move away from configs, to just passing args to the class
Closed 2 years ago
omegaconf.errors.ConfigAttributeError: Missing key datafiles
Closed 2 years ago3
Add VIT documentation
Closed 2 years ago
Should we get rid of Hydra?
Closed 2 years ago8
Support for `pytorch_lightning.Trainer.predict`
Closed 2 years ago1
Logger instantiation
Closed 2 years ago1
Optimizers without `weight_decay` produce errors
Closed 2 years ago1
Multiple Choice Dataset Files not Overwritten
Closed 2 years ago2
[Bug] Problems when training with hydra
Closed 2 years ago3
PL has changed the paramerters of Trainer
Closed 2 years ago4
basic squad example is not working
Closed 2 years ago3
[bug] trainer got unexpected keyword argument 'truncated_bptt_steps'
Closed 2 years ago4
Problem with reading in large json files
Closed 2 years ago6
batch by maximum amount of tokens
Closed 2 years ago3
Callback for writing predictions
Closed 2 years ago6
Missing Compressed-Dictionary dependency
Closed 2 years ago3
Glue MNLI task fails due to missing 'validation' key in dataset
Closed 2 years ago10
Cannot find file when using custom data in language modeling
Closed 2 years ago5
Update imports to match with pl_bolts
Closed 2 years ago
Add an option to use Huggingface metrics
Closed 2 years ago7
How can we specify `num_classes` for the model?
Closed 2 years ago1
What format does the model expect for the data(set)?
Closed 2 years ago1
Mixed precision: scheduler called in the wrong order
Closed 2 years ago1
Support saving the model so that it can be reloaded from Huggingface
Closed 2 years ago2
Deprecate text metrics in favour of their counterparts from torchmetrics package.
Closed 3 years ago1
Sparseml Integration
Closed 3 years ago5
Can't use a logger
Closed 3 years ago4