def __init__(
self,
input_key: str = "targets",
output_key: str = "logits",
prefix: str = "accuracy",
topk_args: List[int] = None,
num_classes: int = None,
accuracy_args: List[int] = None,
**kwargs,
):
"""
Args:
input_key: input key to use for accuracy calculation;
specifies our `y_true`
output_key: output key to use for accuracy calculation;
specifies our `y_pred`
prefix: key for the metric's name
topk_args: specifies which accuracy@K to log:
[1] - accuracy
[1, 3] - accuracy at 1 and 3
[1, 3, 5] - accuracy at 1, 3 and 5
num_classes: number of classes to calculate ``topk_args``
if ``accuracy_args`` is None
"""
topk_args = (topk_args or accuracy_args
or get_default_topk_args(num_classes))
super().__init__(
prefix=prefix,
metric_fn=wrap_topk_metric2dict(accuracy, topk_args=topk_args),
input_key=input_key,
output_key=output_key,
**kwargs,
)
def __init__(
self,
embeddings_key: str = "logits",
labels_key: str = "targets",
is_query_key: str = "is_query",
prefix: str = "cmc",
topk_args: List[int] = None,
num_classes: int = None,
):
"""
This callback was designed to count
cumulative matching characteristics.
If current object is from query your dataset
should output `True` in `is_query_key`
and false if current object is from gallery.
You can see `QueryGalleryDataset` in
`catalyst.contrib.datasets.metric_learning` for more information.
On batch end callback accumulate all embeddings
Args:
embeddings_key: embeddings key in output dict
labels_key: labels key in output dict
is_query_key: bool key True if current
object is from query
prefix: key for the metric's name
topk_args: specifies which cmc@K to log.
[1] - cmc@1
[1, 3] - cmc@1 and cmc@3
[1, 3, 5] - cmc@1, cmc@3 and cmc@5
num_classes: number of classes to calculate ``accuracy_args``
if ``topk_args`` is None
"""
super().__init__(order=CallbackOrder.Metric)
self.list_args = topk_args or get_default_topk_args(num_classes)
self._metric_fn = cmc_score
self._prefix = prefix
self.embeddings_key = embeddings_key
self.labels_key = labels_key
self.is_query_key = is_query_key
self._gallery_embeddings: torch.Tensor = None
self._query_embeddings: torch.Tensor = None
self._gallery_labels: torch.Tensor = None
self._query_labels: torch.Tensor = None
self._gallery_idx = None
self._query_idx = None
self._query_size = None
self._gallery_size = None
def __init__(
self,
input_key: str = "targets",
output_key: str = "logits",
prefix: str = "accuracy",
topk_args: List[int] = None,
num_classes: int = None,
accuracy_args: List[int] = None,
**kwargs,
):
"""
Args:
input_key: input key to use for accuracy calculation;
specifies our `y_true`
output_key: output key to use for accuracy calculation;
specifies our `y_pred`
prefix: key for the metric's name
topk_args: specifies which accuracy@K to log:
[1] - accuracy
[1, 3] - accuracy at 1 and 3
[1, 3, 5] - accuracy at 1, 3 and 5
num_classes: number of classes to calculate ``topk_args``
if ``accuracy_args`` is None
**kwargs: key-value params to pass to the metric
.. note::
For ``**kwargs`` info, please follow
``catalyst.callbacks.metric.BatchMetricCallback`` and
``catalyst.metrics.accuracy.accuracy`` docs
"""
topk_args = (topk_args or accuracy_args
or get_default_topk_args(num_classes))
super().__init__(
prefix=prefix,
metric_fn=wrap_topk_metric2dict(accuracy, topk_args=topk_args),
input_key=input_key,
output_key=output_key,
**kwargs,
)
def __init__(
self,
input_key: str = "targets",
output_key: str = "logits",
prefix: str = "accuracy",
multiplier: float = 1.0,
topk_args: List[int] = None,
num_classes: int = None,
accuracy_args: List[int] = None,
**kwargs,
):
"""
Args:
input_key: input key to use for accuracy calculation;
specifies our `y_true`
output_key: output key to use for accuracy calculation;
specifies our `y_pred`
prefix: key for the metric's name
topk_args: specifies which accuracy@K to log:
[1] - accuracy
[1, 3] - accuracy at 1 and 3
[1, 3, 5] - accuracy at 1, 3 and 5
num_classes: number of classes to calculate ``topk_args``
if ``accuracy_args`` is None
activation: An torch.nn activation applied to the outputs.
Must be one of ``"none"``, ``"Sigmoid"``, or ``"Softmax"``
"""
topk_args = (topk_args or accuracy_args
or get_default_topk_args(num_classes))
super().__init__(
prefix=prefix,
metric_fn=wrap_topk_metric2dict(accuracy, topk_args=topk_args),
input_key=input_key,
output_key=output_key,
multiplier=multiplier,
**kwargs,
)