Fastai

Bugs Squashed

New Features

- Enhancement: Let TextDataLoaders take in a custom `tok_text_col` ([3208](https://github.com/fastai/fastai/pull/3208)), thanks to [muellerzr](https://github.com/muellerzr)
- Changed dataloaders arguments to have consistent overrides ([3178](https://github.com/fastai/fastai/pull/3178)), thanks to [marii-moe](https://github.com/marii-moe)
- Better support for iterable datasets ([3173](https://github.com/fastai/fastai/pull/3173)), thanks to [jcaw](https://github.com/jcaw)

Bugs Squashed

- BrokenProcessPool in `download_images()` on Windows ([3196](https://github.com/fastai/fastai/issues/3196))
- error on predict() or using interp with resnet and MixUp ([3180](https://github.com/fastai/fastai/issues/3180))
- Fix 'cat' attribute with pandas dataframe: `AttributeError: Can only use .cat accessor with a 'category' dtype` ([3165](https://github.com/fastai/fastai/pull/3165)), thanks to [dreamflasher](https://github.com/dreamflasher)
- `cont_cat_split` does not support pandas types ([3156](https://github.com/fastai/fastai/issues/3156))
- `DataBlock.dataloaders` does not support the advertised "shuffle" argument ([3133](https://github.com/fastai/fastai/issues/3133))

New Features

- Calculate correct `nf` in `create_head` based on `concat_pool` ([3115](https://github.com/fastai/fastai/pull/3115)), thanks to [muellerzr](https://github.com/muellerzr)

Bugs Squashed

- wandb integration failing with latest wandb library ([3066](https://github.com/fastai/fastai/issues/3066))
- `Learner.load` and `LRFinder` not functioning properly for the optimizer states ([2892](https://github.com/fastai/fastai/issues/2892))

Bugs Squashed

- tensorboard and wandb can not access `smooth_loss` ([3131](https://github.com/fastai/fastai/issues/3131))

Breaking Changes

- Promote `NativeMixedPrecision` to default `MixedPrecision` (and similar for `Learner.to_fp16`); old `MixedPrecision` is now called `NonNativeMixedPrecision` ([3127](https://github.com/fastai/fastai/issues/3127))
- Use the new `GradientClip` callback instead of the `clip` parameter to use gradient clipping
- Adding a `Callback` which has the same name as an attribute no longer raises an exception ([3109](https://github.com/fastai/fastai/issues/3109))
- RNN training now requires `RNNCallback`, but does not require `RNNRegularizer`; `out` and `raw_out` have moved to `RNNRegularizer` ([3108](https://github.com/fastai/fastai/issues/3108))
- Call `rnn_cbs` to get all callbacks needed for RNN training, optionally with regularization
- replace callback `run_after` with `order`; do not run `after` cbs on exception ([3101](https://github.com/fastai/fastai/issues/3101))

New Features

- Add `GradientClip` callback ([3107](https://github.com/fastai/fastai/issues/3107))
- Make `Flatten` cast to `TensorBase` to simplify type compatibility ([3106](https://github.com/fastai/fastai/issues/3106))
- make flattened metrics compatible with all tensor subclasses ([3105](https://github.com/fastai/fastai/issues/3105))
- New class method `TensorBase.register_func` to register types for `__torch_function__` ([3097](https://github.com/fastai/fastai/issues/3097))
- new `dynamic` flag for controlling dynamic loss scaling in `NativeMixedPrecision` ([3096](https://github.com/fastai/fastai/issues/3096))
- remove need to call `to_native_fp32` before `predict`; set `skipped` in NativeMixedPrecision after NaN from dynamic loss scaling ([3095](https://github.com/fastai/fastai/issues/3095))
- make native fp16 extensible with callbacks ([3094](https://github.com/fastai/fastai/issues/3094))
- Calculate correct `nf` in `create_head` based on `concat_pool` ([3115](https://github.com/fastai/fastai/pull/3115)) thanks to [muellerzr](https://github.com/muellerzr)

New Features

- Small DICOM segmentation dataset ([3034](https://github.com/fastai/fastai/pull/3034)), thanks to [moritzschwyzer](https://github.com/moritzschwyzer)

Bugs Squashed

- `NoneType object has no attribute append` in fastbook chapter 6 BIWI example ([3091](https://github.com/fastai/fastai/issues/3091))