Generic Callbacks

CHANGELOG.md

@@ -74,18 +74,35 @@ dataset at every epoch) and a `MultiTaskScheduler` (for ordering the instances w
   This makes it possible to use the `FileLock` class on a read-only file system.
 - Added a new learning rate scheduler: `CombinedLearningRateScheduler`. This can be used to combine different LR schedulers, using one after the other.
 - Moving `ModelCard` and `TaskCard` abstractions into the main repository.
+- The `TrainerCallback` constructor accepts `serialization_dir` provided by `Trainer`. This can be useeful for `Logger` callbacks those need to store files in the run directory.
+- The `TrainerCallback.on_start()` is fired at the start of the training.
+- The `TrainerCallback` event methods now accept `**kwargs`. This may be useful to maintain backwards-compability of callbacks easier in the future. E.g. we may decide to pass the exception/traceback object in case of failure to `on_end()` and this older callbacks may simply ignore the argument instead of raising a `TypeError`.
 
 ### Changed
 
 - 'master' branch renamed to 'main'
 - Torch version bumped to 1.7.1 in Docker images.
+- The `TrainerCallack.on_epoch()` does not fire with `epoch=-1` at the start of the training.
+  Instead, `TrainerCallback.on_start()` should be used for these cases.
+- `TensorBoardBatchMemoryUsage` is converted from `BatchCallback` into `TrainerCallback`.
+- `TrackEpochCallback` is converted from `EpochCallback` into `TrainerCallback`.
+- `Trainer` can accept callbacks simply with name `callbacks` instead of `trainer_callbacks`.
 
 ### Fixed
 
 - Fixed typo with `LabelField` string representation: removed trailing apostrophe.
 - `Vocabulary.from_files` and `cached_path` will issue a warning, instead of failing, when a lock on an existing resource
   can't be acquired because the file system is read-only.
 - `TrackEpochCallback` is now a `EpochCallback`.
+- Now Trainer always fires `TrainerCallback.on_end()` so all the resources can be cleaned up properly.
+- Fixed the misspelling, changed `TensoboardBatchMemoryUsage` to `TensorBoardBatchMemoryUsage`.
+- We set a value to `epoch` so in case of firing `TrainerCallback.on_end()` the variable is bound.
+  This could have lead to an error in case of trying to recover a run after it was finished training.
+
+### Removed
+
+- Removed `EpochCallback`, `BatchCallback` in favour of `TrainerCallback`.
+  The metaclass-wrapping implementation is removed as well.
 
 
 ## [v1.3.0](https://github.com/allenai/allennlp/releases/tag/v1.3.0) - 2020-12-15
@@ -113,7 +130,7 @@ dataset at every epoch) and a `MultiTaskScheduler` (for ordering the instances w
   were not passed to the constructor if the value of the parameter was equal to the default value.
   This caused bugs in some edge cases where a subclass that takes `**kwargs` needs to inspect
   `kwargs` before passing them to its superclass.
-- Improved the band-aid solution for segmentation faults and the "ImportError: dlopen: cannot load any more object with static TLS" 
+- Improved the band-aid solution for segmentation faults and the "ImportError: dlopen: cannot load any more object with static TLS"
   by adding a `transformers` import.
 - Added safety checks for extracting tar files
 - Turned superfluous warning to info when extending the vocab in the embedding matrix, if no pretrained file was provided

allennlp/training/__init__.py

@@ -4,8 +4,7 @@
 from allennlp.training.trainer import (
     Trainer,
     GradientDescentTrainer,
-    BatchCallback,
-    EpochCallback,
     TrainerCallback,
     TrackEpochCallback,
+    TensorBoardBatchMemoryUsage,
 )