2 0 Regular - Pytorch Lightning

.. list-table:: reg. user 2.0 :widths: 40 40 20 :header-rows: 1

- If
- Then
- Ref
- used PyTorch 1.11
- upgrade to PyTorch 2.1 or higher
- PR18691_
- called self.trainer.model.parameters() in LightningModule.configure_optimizers() when using FSDP
- On PyTorch 2.0+, call self.parameters() from now on
- PR17309_
- used Trainer(accelerator="tpu", devices=[i])" to select the 1-based TPU core index
- the index is now 0-based
- PR17227_
- used torch_xla < 1.13
- upgrade to torch_xla >= 1.13
- PR17368_
- used trainer.num_val_batches to get the total size of all validation dataloaders
- use sum(trainer.num_val_batches)
- PR18441_
- used trainer.num_test_batches to get the total size of all test dataloaders
- use sum(trainer.num_test_batches)
- PR18441_
- used trainer.num_sanity_val_batches to get the total size of all validation dataloaders for sanity checking
- use sum(trainer.num_sanity_val_batches)
- PR18441_
- used Trainer(devices="auto") to auto-select all available GPUs in a Jupyter notebook
- use Trainer(devices=-1)
- PR18291_
- used Trainer(devices="auto") to auto-select all available GPUs in a Jupyter notebook
- use Trainer(devices=-1)
- PR18291_
- pip install lightning to install lightning.app dependencies
- use pip install lightning[app] if you need lightning.app
- PR18386_