Seppo Enarvi
Seppo Enarvi
If data samples contain a Python list with a variable number of elements, type inference will fail
## 🐛 Bug When optimizing a dataset, [`BinaryWriter.serialize()`](https://github.com/Lightning-AI/litdata/blob/main/src/litdata/streaming/writer.py#L167) will first flatten the sample dictionary and infer the type of each element. Then it will assume that all data samples have...
A callback that updates an [AveragedModel](https://pytorch.org/docs/stable/generated/torch.optim.swa_utils.AveragedModel.html) after every training step ## What does this PR do? This is similar to the existing [StochasticWeightAveraging](https://lightning.ai/docs/pytorch/stable/api/lightning.pytorch.callbacks.StochasticWeightAveraging.html) callback, but wraps the [AveragedModel](https://pytorch.org/docs/stable/generated/torch.optim.swa_utils.AveragedModel.html) class from...
I installed Byobu from the Git repository, added the `bin` directory to `$PATH` and launched `byobu`. Everything works correctly. However, it appends one line to the end of `~/.bashrc` that...