def _compute_metric_score(
metric: torchmetrics.Metric,
custom_metric_func: Callable,
label: torch.Tensor,
logits: Optional[torch.Tensor] = None,
embeddings1: Optional[torch.Tensor] = None,
embeddings2: Optional[torch.Tensor] = None,
reverse_prob: Optional[bool] = False,
):
if logits is not None:
if isinstance(metric, torchmetrics.AUROC):
prob = compute_probability(logits=logits)
metric.update(preds=prob, target=label) # only for binary classification
elif isinstance(metric, BaseAggregator):
metric.update(custom_metric_func(logits, label))
else:
metric.update(logits.squeeze(dim=1), label)
else:
if isinstance(metric, BaseAggregator):
metric.update(custom_metric_func(embeddings1, embeddings2, label))
else:
metric.update(
compute_probability(
embeddings1=embeddings1,
embeddings2=embeddings2,
reverse_prob=reverse_prob,
),
label,
)
def _assert_half_support(metric_module: Metric,
metric_functional: Callable,
preds: Tensor,
target: Tensor,
device: str = "cpu",
**kwargs_update):
"""
Test if an metric can be used with half precision tensors
Args:
metric_module: the metric module to test
metric_functional: the metric functional to test
preds: torch tensor with predictions
target: torch tensor with targets
device: determine device, either "cpu" or "cuda"
kwargs_update: Additional keyword arguments that will be passed with preds and
target when running update on the metric.
"""
y_hat = preds[0].half().to(
device) if preds[0].is_floating_point() else preds[0].to(device)
y = target[0].half().to(
device) if target[0].is_floating_point() else target[0].to(device)
kwargs_update = {
k: (v[0].half() if v.is_floating_point() else v[0]).to(device)
if isinstance(v, Tensor) else v
for k, v in kwargs_update.items()
}
metric_module = metric_module.to(device)
_assert_tensor(metric_module(y_hat, y, **kwargs_update))
_assert_tensor(metric_functional(y_hat, y, **kwargs_update))
def _compute_metric_score(
self,
metric: torchmetrics.Metric,
custom_metric_func: Callable,
logits: torch.Tensor,
label: torch.Tensor,
):
if isinstance(metric, (torchmetrics.AUROC, torchmetrics.AveragePrecision)):
prob = F.softmax(logits.float(), dim=1)
metric.update(preds=prob[:, 1], target=label) # only for binary classification
elif isinstance(metric, BaseAggregator):
metric.update(custom_metric_func(logits, label))
else:
metric.update(logits.squeeze(dim=1), label)
def _assert_half_support(
metric_module: Metric,
metric_functional: Callable,
preds: torch.Tensor,
target: torch.Tensor,
device: str = "cpu",
):
"""
Test if an metric can be used with half precision tensors
Args:
metric_module: the metric module to test
metric_functional: the metric functional to test
preds: torch tensor with predictions
target: torch tensor with targets
device: determine device, either "cpu" or "cuda"
"""
y_hat = preds[0].half().to(
device) if preds[0].is_floating_point() else preds[0].to(device)
y = target[0].half().to(
device) if target[0].is_floating_point() else target[0].to(device)
metric_module = metric_module.to(device)
_assert_tensor(metric_module(y_hat, y))
_assert_tensor(metric_functional(y_hat, y))