Algorithm#
- class composer.Algorithm(*args, **kwargs)[source]#
Base class for algorithms.
Algorithms are pieces of code which run at specific events (see
Event
) in the training loop. Algorithms modify the trainerโsState
, generally with the effect of improving the modelโs quality or increasing the efficiency and throughput of the training loop.- Algorithms must implement the following two methods:
- abstract apply(event, state, logger)[source]#
Applies the algorithm to make an in-place change to the
State
.Can optionally return an exit code to be stored in a
Trace
. This exit code is made accessible for debugging.
- property backwards_create_graph#
Whether this algorithm requires the backwards pass to be differentiable. Defaults to
False
.If it returns
True
,create_graph=True
will be passed totorch.Tensor.backward()
which will result in the graph of the gradient also being constructed. This allows the computation of second order derivatives.
- property find_unused_parameters#
Indicates whether this algorithm may cause some model parameters to be unused. Defaults to False.
For example, it is used to tell
torch.nn.parallel.DistributedDataParallel
(DDP) that some parameters will be frozen during training, and hence it should not expect gradients from them. All algorithms which do any kind of parameter freezing should override this function to returnTrue
.Note
DeepSpeed integration with this function returning True is not tested. It may not work as expected.
- abstract match(event, state)[source]#
Determines whether this algorithm should run given the current
Event
andState
.Examples: To only run on a specific event (e.g., on
Event.BEFORE_LOSS
), override match as shown below:>>> class MyAlgorithm: ... def match(self, event, state): ... return event == Event.BEFORE_LOSS >>> MyAlgorithm().match(Event.BEFORE_LOSS, state) True
To run based on some value of a
State
attribute, override match as shown below:>>> class MyAlgorithm: ... def match(self, event, state): ... return state.timestamp.epoch > 30 >>> MyAlgorithm().match(Event.BEFORE_LOSS, state) False
See
State
for accessible attributes.