@@ -135,7 +136,7 @@ Optional name for the instance.
from_config
-View
+View
source
@@ -176,7 +177,7 @@ A `Loss` instance.
get_config
-View
+View
source
@@ -187,7 +188,7 @@ Returns the config dictionary for a `Loss` instance.
__call__
-View
+View
source
diff --git a/docs/api_docs/python/tfr/keras/losses/ApproxNDCGLoss.md b/docs/api_docs/python/tfr/keras/losses/ApproxNDCGLoss.md
index e431c60..9a0c16d 100644
--- a/docs/api_docs/python/tfr/keras/losses/ApproxNDCGLoss.md
+++ b/docs/api_docs/python/tfr/keras/losses/ApproxNDCGLoss.md
@@ -15,7 +15,7 @@ description: Computes approximate NDCG loss between y_true and y_pred.
-
+
View source on GitHub
@@ -117,13 +117,14 @@ $$
|
Type of `tf.keras.losses.Reduction` to apply to
-loss. Default value is `AUTO`. `AUTO` indicates that the reduction
-option will be determined by the usage context. For almost all cases
-this defaults to `SUM_OVER_BATCH_SIZE`. When used under a
-`tf.distribute.Strategy`, except via `Model.compile()` and
-`Model.fit()`, using `AUTO` or `SUM_OVER_BATCH_SIZE`
-will raise an error. Please see this custom training [tutorial](https://www.tensorflow.org/tutorials/distribute/custom_training)
- for more details.
+loss. Default value is `AUTO`. `AUTO` indicates that the
+reduction option will be determined by the usage context. For
+almost all cases this defaults to `SUM_OVER_BATCH_SIZE`. When
+used under a `tf.distribute.Strategy`, except via
+`Model.compile()` and `Model.fit()`, using `AUTO` or
+`SUM_OVER_BATCH_SIZE` will raise an error. Please see this
+custom training [tutorial](https://www.tensorflow.org/tutorials/distribute/custom_training)
+for more details.
|
@@ -139,7 +140,7 @@ Optional name for the instance.
from_config
-View
+View
source
@@ -180,7 +181,7 @@ A `Loss` instance.
get_config
-View
+View
source
@@ -191,7 +192,7 @@ Returns the config dictionary for a `Loss` instance.
__call__
-View
+View
source
diff --git a/docs/api_docs/python/tfr/keras/losses/ClickEMLoss.md b/docs/api_docs/python/tfr/keras/losses/ClickEMLoss.md
index 3659d92..4094e38 100644
--- a/docs/api_docs/python/tfr/keras/losses/ClickEMLoss.md
+++ b/docs/api_docs/python/tfr/keras/losses/ClickEMLoss.md
@@ -15,7 +15,7 @@ description: Computes click EM loss between y_true and y_pred.
-
+
View source on GitHub
@@ -173,7 +173,7 @@ A `Loss` instance.
get_config
-View
+View
source
@@ -184,7 +184,7 @@ Returns the config dictionary for a `Loss` instance.
__call__
-View
+View
source
diff --git a/docs/api_docs/python/tfr/keras/losses/CoupledRankDistilLoss.md b/docs/api_docs/python/tfr/keras/losses/CoupledRankDistilLoss.md
index 38072f9..c1a6fb3 100644
--- a/docs/api_docs/python/tfr/keras/losses/CoupledRankDistilLoss.md
+++ b/docs/api_docs/python/tfr/keras/losses/CoupledRankDistilLoss.md
@@ -15,7 +15,7 @@ description: Computes the Rank Distil loss between y_true and y_pred.
-
+
View source on GitHub
@@ -86,7 +86,6 @@ The Coupled-RankDistil loss is defined as: $$ \mathcal{L}(y, s) = -\sum_{\pi}
[reddi2021]: https://research.google/pubs/pub50695/
-
Args |
|---|
@@ -156,7 +155,6 @@ all entries in the list.
Instantiates a `Loss` from its config (output of `get_config()`).
-
| Args |
|---|
@@ -172,7 +170,6 @@ Output of `get_config()`.
-
| Returns |
|---|
@@ -186,7 +183,7 @@ A `Loss` instance.
get_config
-View
+View
source
@@ -197,7 +194,7 @@ Returns the config dictionary for a `Loss` instance.
__call__
-View
+View
source
diff --git a/docs/api_docs/python/tfr/keras/losses/DCGLambdaWeight.md b/docs/api_docs/python/tfr/keras/losses/DCGLambdaWeight.md
index de99c5a..f66eadf 100644
--- a/docs/api_docs/python/tfr/keras/losses/DCGLambdaWeight.md
+++ b/docs/api_docs/python/tfr/keras/losses/DCGLambdaWeight.md
@@ -15,7 +15,7 @@ description: Keras serializable class for DCG.
-
+
View source on GitHub
@@ -85,7 +85,7 @@ LambdaMART.
get_config
-View
+View
source
diff --git a/docs/api_docs/python/tfr/keras/losses/GumbelApproxNDCGLoss.md b/docs/api_docs/python/tfr/keras/losses/GumbelApproxNDCGLoss.md
index a42c7bd..c9310f2 100644
--- a/docs/api_docs/python/tfr/keras/losses/GumbelApproxNDCGLoss.md
+++ b/docs/api_docs/python/tfr/keras/losses/GumbelApproxNDCGLoss.md
@@ -16,7 +16,7 @@ y_pred.
|
-
+
View source on GitHub
@@ -125,13 +125,14 @@ $$
|
Type of `tf.keras.losses.Reduction` to apply to
-loss. Default value is `AUTO`. `AUTO` indicates that the reduction
-option will be determined by the usage context. For almost all cases
-this defaults to `SUM_OVER_BATCH_SIZE`. When used under a
-`tf.distribute.Strategy`, except via `Model.compile()` and
-`Model.fit()`, using `AUTO` or `SUM_OVER_BATCH_SIZE`
-will raise an error. Please see this custom training [tutorial](https://www.tensorflow.org/tutorials/distribute/custom_training)
- for more details.
+loss. Default value is `AUTO`. `AUTO` indicates that the
+reduction option will be determined by the usage context. For
+almost all cases this defaults to `SUM_OVER_BATCH_SIZE`. When
+used under a `tf.distribute.Strategy`, except via
+`Model.compile()` and `Model.fit()`, using `AUTO` or
+`SUM_OVER_BATCH_SIZE` will raise an error. Please see this
+custom training [tutorial](https://www.tensorflow.org/tutorials/distribute/custom_training)
+for more details.
|
@@ -147,7 +148,7 @@ Optional name for the instance.
from_config
-View
+View
source
@@ -188,7 +189,7 @@ A `Loss` instance.
get_config
-View
+View
source
@@ -199,7 +200,7 @@ Returns the config dictionary for a `Loss` instance.
__call__
-View
+View
source
diff --git a/docs/api_docs/python/tfr/keras/losses/LabelDiffLambdaWeight.md b/docs/api_docs/python/tfr/keras/losses/LabelDiffLambdaWeight.md
index 62ac7d5..23716d7 100644
--- a/docs/api_docs/python/tfr/keras/losses/LabelDiffLambdaWeight.md
+++ b/docs/api_docs/python/tfr/keras/losses/LabelDiffLambdaWeight.md
@@ -15,7 +15,7 @@ description: Keras serializable class for LabelDiffLambdaWeight.
-
+
View source on GitHub
@@ -36,7 +36,7 @@ Keras serializable class for LabelDiffLambdaWeight.
get_config
-View
+View
source
diff --git a/docs/api_docs/python/tfr/keras/losses/ListMLELambdaWeight.md b/docs/api_docs/python/tfr/keras/losses/ListMLELambdaWeight.md
index f3a7f05..aea61de 100644
--- a/docs/api_docs/python/tfr/keras/losses/ListMLELambdaWeight.md
+++ b/docs/api_docs/python/tfr/keras/losses/ListMLELambdaWeight.md
@@ -15,7 +15,7 @@ description: LambdaWeight for ListMLE cost function.
-
+
View source on GitHub
@@ -52,7 +52,7 @@ LambdaWeight for ListMLE cost function.
get_config
-View
+View
source
@@ -61,7 +61,7 @@ source
individual_weights
-View
+View
source
@@ -74,7 +74,7 @@ See `_LambdaWeight`.
pair_weights
-View
+View
source
diff --git a/docs/api_docs/python/tfr/keras/losses/ListMLELoss.md b/docs/api_docs/python/tfr/keras/losses/ListMLELoss.md
index bba6532..eda113e 100644
--- a/docs/api_docs/python/tfr/keras/losses/ListMLELoss.md
+++ b/docs/api_docs/python/tfr/keras/losses/ListMLELoss.md
@@ -15,7 +15,7 @@ description: Computes ListMLE loss between y_true and y_pred.
-
+
View source on GitHub
@@ -153,7 +153,7 @@ False, this loss will accept dense tensors.
from_config
-View
+View
source
@@ -194,7 +194,7 @@ A `Loss` instance.
get_config
-View
+View
source
@@ -205,7 +205,7 @@ Returns the config dictionary for a `Loss` instance.
__call__
-View
+View
source
diff --git a/docs/api_docs/python/tfr/keras/losses/MeanSquaredLoss.md b/docs/api_docs/python/tfr/keras/losses/MeanSquaredLoss.md
index d38cb25..a3a9a99 100644
--- a/docs/api_docs/python/tfr/keras/losses/MeanSquaredLoss.md
+++ b/docs/api_docs/python/tfr/keras/losses/MeanSquaredLoss.md
@@ -15,7 +15,7 @@ description: Computes mean squared loss between y_true and y_pred.
-
+
View source on GitHub
@@ -142,7 +142,7 @@ A `Loss` instance.
get_config
-View
+View
source
@@ -153,7 +153,7 @@ Returns the config dictionary for a `Loss` instance.
__call__
-View
+View
source
diff --git a/docs/api_docs/python/tfr/keras/losses/MixtureEMLoss.md b/docs/api_docs/python/tfr/keras/losses/MixtureEMLoss.md
index 6a77a62..55891f9 100644
--- a/docs/api_docs/python/tfr/keras/losses/MixtureEMLoss.md
+++ b/docs/api_docs/python/tfr/keras/losses/MixtureEMLoss.md
@@ -15,7 +15,7 @@ description: Computes mixture EM loss between y_true and y_pred.
-
+
View source on GitHub
@@ -145,7 +145,7 @@ False, this loss will accept dense tensors.
from_config
-View
+View
source
@@ -186,7 +186,7 @@ A `Loss` instance.
get_config
-View
+View
source
@@ -197,7 +197,7 @@ Returns the config dictionary for a `Loss` instance.
__call__
-View
+View
source
diff --git a/docs/api_docs/python/tfr/keras/losses/NDCGLambdaWeight.md b/docs/api_docs/python/tfr/keras/losses/NDCGLambdaWeight.md
index 7f9b7f3..83b096b 100644
--- a/docs/api_docs/python/tfr/keras/losses/NDCGLambdaWeight.md
+++ b/docs/api_docs/python/tfr/keras/losses/NDCGLambdaWeight.md
@@ -15,7 +15,7 @@ description: Keras serializable class for NDCG.
-
+
View source on GitHub
@@ -86,7 +86,7 @@ LambdaMART.
get_config
-View
+View
source
diff --git a/docs/api_docs/python/tfr/keras/losses/NDCGLambdaWeightV2.md b/docs/api_docs/python/tfr/keras/losses/NDCGLambdaWeightV2.md
index 6a499ba..c75e7ac 100644
--- a/docs/api_docs/python/tfr/keras/losses/NDCGLambdaWeightV2.md
+++ b/docs/api_docs/python/tfr/keras/losses/NDCGLambdaWeightV2.md
@@ -15,7 +15,7 @@ description: Keras serializable class for NDCG LambdaWeight V2 for topn.
-
+
View source on GitHub
@@ -75,7 +75,7 @@ Keras serializable class for NDCG LambdaWeight V2 for topn.
get_config
-View
+View
source
diff --git a/docs/api_docs/python/tfr/keras/losses/OrdinalLoss.md b/docs/api_docs/python/tfr/keras/losses/OrdinalLoss.md
index d225493..8e443c9 100644
--- a/docs/api_docs/python/tfr/keras/losses/OrdinalLoss.md
+++ b/docs/api_docs/python/tfr/keras/losses/OrdinalLoss.md
@@ -15,7 +15,7 @@ description: Computes the Ordinal loss between y_true and y_pred.
|
-
+
View source on GitHub
@@ -85,13 +85,14 @@ $$
|
Type of `tf.keras.losses.Reduction` to apply to
-loss. Default value is `AUTO`. `AUTO` indicates that the reduction
-option will be determined by the usage context. For almost all cases
-this defaults to `SUM_OVER_BATCH_SIZE`. When used under a
-`tf.distribute.Strategy`, except via `Model.compile()` and
-`Model.fit()`, using `AUTO` or `SUM_OVER_BATCH_SIZE`
-will raise an error. Please see this custom training [tutorial](https://www.tensorflow.org/tutorials/distribute/custom_training)
- for more details.
+loss. Default value is `AUTO`. `AUTO` indicates that the
+reduction option will be determined by the usage context. For
+almost all cases this defaults to `SUM_OVER_BATCH_SIZE`. When
+used under a `tf.distribute.Strategy`, except via
+`Model.compile()` and `Model.fit()`, using `AUTO` or
+`SUM_OVER_BATCH_SIZE` will raise an error. Please see this
+custom training [tutorial](https://www.tensorflow.org/tutorials/distribute/custom_training)
+for more details.
|
@@ -145,7 +146,7 @@ A `Loss` instance.
get_config
-View
+View
source
@@ -156,7 +157,7 @@ Returns the config dictionary for a `Loss` instance.
__call__
-View
+View
source
diff --git a/docs/api_docs/python/tfr/keras/losses/PairwiseHingeLoss.md b/docs/api_docs/python/tfr/keras/losses/PairwiseHingeLoss.md
index 2e168c1..9b384ff 100644
--- a/docs/api_docs/python/tfr/keras/losses/PairwiseHingeLoss.md
+++ b/docs/api_docs/python/tfr/keras/losses/PairwiseHingeLoss.md
@@ -15,7 +15,7 @@ description: Computes pairwise hinge loss between y_true and y_pred.
-
+
View source on GitHub
@@ -126,7 +126,7 @@ False, this loss will accept dense tensors.
from_config
-View
+View
source
@@ -167,7 +167,7 @@ A `Loss` instance.
get_config
-View
+View
source
@@ -178,7 +178,7 @@ Returns the config dictionary for a `Loss` instance.
__call__
-View
+View
source
diff --git a/docs/api_docs/python/tfr/keras/losses/PairwiseLogisticLoss.md b/docs/api_docs/python/tfr/keras/losses/PairwiseLogisticLoss.md
index a640d3f..0416c0a 100644
--- a/docs/api_docs/python/tfr/keras/losses/PairwiseLogisticLoss.md
+++ b/docs/api_docs/python/tfr/keras/losses/PairwiseLogisticLoss.md
@@ -15,7 +15,7 @@ description: Computes pairwise logistic loss between y_true and y_pred.
-
+
View source on GitHub
@@ -126,7 +126,7 @@ False, this loss will accept dense tensors.
from_config
-View
+View
source
@@ -167,7 +167,7 @@ A `Loss` instance.
get_config
-View
+View
source
@@ -178,7 +178,7 @@ Returns the config dictionary for a `Loss` instance.
__call__
-View
+View
source
diff --git a/docs/api_docs/python/tfr/keras/losses/PairwiseMSELoss.md b/docs/api_docs/python/tfr/keras/losses/PairwiseMSELoss.md
index 9eb9c33..9744ee4 100644
--- a/docs/api_docs/python/tfr/keras/losses/PairwiseMSELoss.md
+++ b/docs/api_docs/python/tfr/keras/losses/PairwiseMSELoss.md
@@ -16,7 +16,7 @@ y_pred.
-
+
View source on GitHub
@@ -128,7 +128,7 @@ False, this loss will accept dense tensors.
from_config
-View
+View
source
@@ -169,7 +169,7 @@ A `Loss` instance.
get_config
-View
+View
source
@@ -180,7 +180,7 @@ Returns the config dictionary for a `Loss` instance.
__call__
-View
+View
source
diff --git a/docs/api_docs/python/tfr/keras/losses/PairwiseSoftZeroOneLoss.md b/docs/api_docs/python/tfr/keras/losses/PairwiseSoftZeroOneLoss.md
index f0af422..d6822a2 100644
--- a/docs/api_docs/python/tfr/keras/losses/PairwiseSoftZeroOneLoss.md
+++ b/docs/api_docs/python/tfr/keras/losses/PairwiseSoftZeroOneLoss.md
@@ -15,7 +15,7 @@ description: Computes pairwise soft zero-one loss between y_true and y_pred.
-
+
View source on GitHub
@@ -127,7 +127,7 @@ False, this loss will accept dense tensors.
from_config
-View
+View
source
@@ -168,7 +168,7 @@ A `Loss` instance.
get_config
-View
+View
source
@@ -179,7 +179,7 @@ Returns the config dictionary for a `Loss` instance.
__call__
-View
+View
source
diff --git a/docs/api_docs/python/tfr/keras/losses/PrecisionLambdaWeight.md b/docs/api_docs/python/tfr/keras/losses/PrecisionLambdaWeight.md
index 167bed0..16361ca 100644
--- a/docs/api_docs/python/tfr/keras/losses/PrecisionLambdaWeight.md
+++ b/docs/api_docs/python/tfr/keras/losses/PrecisionLambdaWeight.md
@@ -15,7 +15,7 @@ description: Keras serializable class for Precision.
-
+
View source on GitHub
@@ -61,7 +61,7 @@ for positive examples. The rest are negative examples.
get_config
-View
+View
source
@@ -118,7 +118,7 @@ A `Tensor` that can weight individual examples.
pair_weights
-View
+View
source
diff --git a/docs/api_docs/python/tfr/keras/losses/RankingLossKey.md b/docs/api_docs/python/tfr/keras/losses/RankingLossKey.md
index 9ffc302..8da9d24 100644
--- a/docs/api_docs/python/tfr/keras/losses/RankingLossKey.md
+++ b/docs/api_docs/python/tfr/keras/losses/RankingLossKey.md
@@ -18,6 +18,7 @@ description: Ranking loss key strings.
+
# tfr.keras.losses.RankingLossKey
@@ -26,7 +27,7 @@ description: Ranking loss key strings.
-
+
View source on GitHub
@@ -41,7 +42,7 @@ Ranking loss key strings.
all_keys
-View
+View
source
@@ -152,5 +153,12 @@ UNIQUE_SOFTMAX_LOSS
|
`'unique_softmax_loss'`
|
+
+|
+YETI_LOGISTIC_LOSS
+ |
+
+`'yeti_logistic_loss'`
+ |
|
diff --git a/docs/api_docs/python/tfr/keras/losses/SigmoidCrossEntropyLoss.md b/docs/api_docs/python/tfr/keras/losses/SigmoidCrossEntropyLoss.md
index 768ad2d..ee1cb37 100644
--- a/docs/api_docs/python/tfr/keras/losses/SigmoidCrossEntropyLoss.md
+++ b/docs/api_docs/python/tfr/keras/losses/SigmoidCrossEntropyLoss.md
@@ -15,7 +15,7 @@ description: Computes the Sigmoid cross-entropy loss between y_true and y_pred.
|
-
+
View source on GitHub
@@ -85,13 +85,14 @@ $$
|
Type of `tf.keras.losses.Reduction` to apply to
-loss. Default value is `AUTO`. `AUTO` indicates that the reduction
-option will be determined by the usage context. For almost all cases
-this defaults to `SUM_OVER_BATCH_SIZE`. When used under a
-`tf.distribute.Strategy`, except via `Model.compile()` and
-`Model.fit()`, using `AUTO` or `SUM_OVER_BATCH_SIZE`
-will raise an error. Please see this custom training [tutorial](https://www.tensorflow.org/tutorials/distribute/custom_training)
- for more details.
+loss. Default value is `AUTO`. `AUTO` indicates that the
+reduction option will be determined by the usage context. For
+almost all cases this defaults to `SUM_OVER_BATCH_SIZE`. When
+used under a `tf.distribute.Strategy`, except via
+`Model.compile()` and `Model.fit()`, using `AUTO` or
+`SUM_OVER_BATCH_SIZE` will raise an error. Please see this
+custom training [tutorial](https://www.tensorflow.org/tutorials/distribute/custom_training)
+for more details.
|
@@ -145,7 +146,7 @@ A `Loss` instance.
get_config
-View
+View
source
@@ -156,7 +157,7 @@ Returns the config dictionary for a `Loss` instance.
__call__
-View
+View
source
diff --git a/docs/api_docs/python/tfr/keras/losses/SoftmaxLoss.md b/docs/api_docs/python/tfr/keras/losses/SoftmaxLoss.md
index 8f25a30..e05b23f 100644
--- a/docs/api_docs/python/tfr/keras/losses/SoftmaxLoss.md
+++ b/docs/api_docs/python/tfr/keras/losses/SoftmaxLoss.md
@@ -15,7 +15,7 @@ description: Computes Softmax cross-entropy loss between y_true and y_pred.
-
+
View source on GitHub
@@ -126,7 +126,7 @@ False, this loss will accept dense tensors.
from_config
-View
+View
source
@@ -167,7 +167,7 @@ A `Loss` instance.
get_config
-View
+View
source
@@ -178,7 +178,7 @@ Returns the config dictionary for a `Loss` instance.
__call__
-View
+View
source
diff --git a/docs/api_docs/python/tfr/keras/losses/UniqueSoftmaxLoss.md b/docs/api_docs/python/tfr/keras/losses/UniqueSoftmaxLoss.md
index 9b8bfc8..7e1d7ec 100644
--- a/docs/api_docs/python/tfr/keras/losses/UniqueSoftmaxLoss.md
+++ b/docs/api_docs/python/tfr/keras/losses/UniqueSoftmaxLoss.md
@@ -16,7 +16,7 @@ y_pred.
|
-
+
View source on GitHub
@@ -104,13 +104,14 @@ $$
|
Type of `tf.keras.losses.Reduction` to apply to
-loss. Default value is `AUTO`. `AUTO` indicates that the reduction
-option will be determined by the usage context. For almost all cases
-this defaults to `SUM_OVER_BATCH_SIZE`. When used under a
-`tf.distribute.Strategy`, except via `Model.compile()` and
-`Model.fit()`, using `AUTO` or `SUM_OVER_BATCH_SIZE`
-will raise an error. Please see this custom training [tutorial](https://www.tensorflow.org/tutorials/distribute/custom_training)
- for more details.
+loss. Default value is `AUTO`. `AUTO` indicates that the
+reduction option will be determined by the usage context. For
+almost all cases this defaults to `SUM_OVER_BATCH_SIZE`. When
+used under a `tf.distribute.Strategy`, except via
+`Model.compile()` and `Model.fit()`, using `AUTO` or
+`SUM_OVER_BATCH_SIZE` will raise an error. Please see this
+custom training [tutorial](https://www.tensorflow.org/tutorials/distribute/custom_training)
+for more details.
|
@@ -126,7 +127,7 @@ Optional name for the instance.
from_config
-View
+View
source
@@ -167,7 +168,7 @@ A `Loss` instance.
get_config
-View
+View
source
@@ -178,7 +179,7 @@ Returns the config dictionary for a `Loss` instance.
__call__
-View
+View
source
diff --git a/docs/api_docs/python/tfr/keras/losses/YetiDCGLambdaWeight.md b/docs/api_docs/python/tfr/keras/losses/YetiDCGLambdaWeight.md
new file mode 100644
index 0000000..eb0cba8
--- /dev/null
+++ b/docs/api_docs/python/tfr/keras/losses/YetiDCGLambdaWeight.md
@@ -0,0 +1,110 @@
+description: Keras serializable class for YetiDCGLambdaWeight.
+
+
+
+
+
+
+
+
+
+
+# tfr.keras.losses.YetiDCGLambdaWeight
+
+
+
+
+
+Keras serializable class for YetiDCGLambdaWeight.
+
+
+tfr.keras.losses.YetiDCGLambdaWeight(
+ topn: Optional[int] = None,
+ gain_fn: Optional[tfr.keras.utils.GainFunction] = None,
+ rank_discount_fn: Optional[tfr.keras.utils.GainFunction] = None,
+ normalized: bool = False,
+ **kwargs
+)
+
+
+
+
+
+
+
+Args |
|---|
+
+
+|
+`topn`
+ |
+
+(int) The topn for the DCG metric.
+ |
+
+|
+`gain_fn`
+ |
+
+(function) Transforms labels.
+ |
+
+|
+`rank_discount_fn`
+ |
+
+(function) The rank discount function.
+ |
+
+|
+`normalized`
+ |
+
+(bool) If True, normalize weight by the max DCG.
+ |
+
+
+
+## Methods
+
+get_config
+
+View
+source
+
+
+get_config() -> Dict[str, Any]
+
+
+individual_weights
+
+View
+source
+
+
+individual_weights(
+ labels, ranks
+)
+
+
+See `_LambdaWeight`.
+
+pair_weights
+
+View
+source
+
+
+pair_weights(
+ labels: tf.Tensor, ranks: tf.Tensor
+) -> tf.Tensor
+
+
+See `_LambdaWeight`.
diff --git a/docs/api_docs/python/tfr/keras/losses/YetiLogisticLoss.md b/docs/api_docs/python/tfr/keras/losses/YetiLogisticLoss.md
new file mode 100644
index 0000000..ab094cd
--- /dev/null
+++ b/docs/api_docs/python/tfr/keras/losses/YetiLogisticLoss.md
@@ -0,0 +1,204 @@
+description: Computes Yeti logistic loss between y_true and y_pred.
+
+
+
+
+
+
+
+
+
+
+# tfr.keras.losses.YetiLogisticLoss
+
+
+
+
+
+Computes Yeti logistic loss between `y_true` and `y_pred`.
+
+
+tfr.keras.losses.YetiLogisticLoss(
+ reduction: tf.losses.Reduction = tf.losses.Reduction.AUTO,
+ name: Optional[str] = None,
+ lambda_weight: Optional[tfr.keras.losses.YetiDCGLambdaWeight] = None,
+ temperature: float = 0.1,
+ sample_size: int = 8,
+ gumbel_temperature: float = 1.0,
+ seed: Optional[int] = None,
+ ragged: bool = False
+)
+
+
+
+
+Adapted to neural network models from the Yeti loss implemenation for GBDT in
+([Lyzhin et al, 2022][lyzhin2022]).
+
+In this code base, we support Yeti loss with the DCG lambda weight option. The
+default uses the YetiDCGLambdaWeight with default settings. To customize, please
+set the lambda_weight to YetiDCGLambdaWeight.
+
+For each list of scores `s` in `y_pred` and list of labels `y` in `y_true`:
+
+```
+loss = sum_a sum_i I[y_i > y_{i\pm 1}] * log(1 + exp(-(s^a_i - s^a_{i\pm 1})))
+```
+
+where `s^a_i = s_i + gumbel(0, 1)^a`
+
+#### Standalone usage:
+
+```
+>>> y_true = [[1., 0.]]
+>>> y_pred = [[0.6, 0.8]]
+>>> loss = tfr.keras.losses.YetiLogisticLoss(sample_size=2, seed=1)
+>>> loss(y_true, y_pred).numpy()
+0.90761846
+```
+
+```
+>>> # Using ragged tensors
+>>> y_true = tf.ragged.constant([[1., 0.], [0., 1., 0.]])
+>>> y_pred = tf.ragged.constant([[0.6, 0.8], [0.5, 0.8, 0.4]])
+>>> loss = tfr.keras.losses.YetiLogisticLoss(seed=1, ragged=True)
+>>> loss(y_true, y_pred).numpy()
+0.43420443
+```
+
+Usage with the `compile()` API:
+
+```python
+model.compile(optimizer='sgd', loss=tfr.keras.losses.YetiLogisticLoss())
+```
+
+#### Definition:
+
+$$
+\mathcal{L}(\{y\}, \{s\}) =
+\sum_a \sum_i \sum_{j=i\pm 1}I[y_i > y_j] \log(1 + \exp(-(s^a_i - s^a_j)))
+$$
+
+
+
+
+
+References |
|---|
+
+|
+- [Which Tricks are Important for Learning to Rank?, Lyzhin et al, 2022][lyzhin2022]
+ |
+
+
+
+
+[lyzhin2022]: https://arxiv.org/abs/2204.01500
+
+
+
+
+
+Args |
|---|
+
+
+|
+`reduction`
+ |
+
+Type of `tf.keras.losses.Reduction` to apply to
+loss. Default value is `AUTO`. `AUTO` indicates that the
+reduction option will be determined by the usage context. For
+almost all cases this defaults to `SUM_OVER_BATCH_SIZE`. When
+used under a `tf.distribute.Strategy`, except via
+`Model.compile()` and `Model.fit()`, using `AUTO` or
+`SUM_OVER_BATCH_SIZE` will raise an error. Please see this
+custom training [tutorial](https://www.tensorflow.org/tutorials/distribute/custom_training)
+for more details.
+ |
+
+|
+`name`
+ |
+
+Optional name for the instance.
+ |
+
+
+
+## Methods
+
+from_config
+
+View
+source
+
+
+@classmethod
+from_config(
+ config, custom_objects=None
+)
+
+
+Instantiates a `Loss` from its config (output of `get_config()`).
+
+
+
+
+
+| Args |
|---|
+
+
+|
+`config`
+ |
+
+Output of `get_config()`.
+ |
+
+
+
+
+
+
+
+| Returns |
|---|
+
+|
+A `Loss` instance.
+ |
+
+
+
+
+get_config
+
+View
+source
+
+
+get_config() -> Dict[str, Any]
+
+
+Returns the config dictionary for a `Loss` instance.
+
+__call__
+
+View
+source
+
+
+__call__(
+ y_true: tfr.keras.model.TensorLike,
+ y_pred: tfr.keras.model.TensorLike,
+ sample_weight: Optional[utils.TensorLike] = None
+) -> tf.Tensor
+
+
+See _RankingLoss.
diff --git a/docs/api_docs/python/tfr/keras/losses/get.md b/docs/api_docs/python/tfr/keras/losses/get.md
index f1ac68a..29c5d0e 100644
--- a/docs/api_docs/python/tfr/keras/losses/get.md
+++ b/docs/api_docs/python/tfr/keras/losses/get.md
@@ -11,7 +11,7 @@ description: Factory method to get a ranking loss class.
|
-
+
View source on GitHub
diff --git a/docs/api_docs/python/tfr/keras/metrics/ARPMetric.md b/docs/api_docs/python/tfr/keras/metrics/ARPMetric.md
index 70738bb..c9d9334 100644
--- a/docs/api_docs/python/tfr/keras/metrics/ARPMetric.md
+++ b/docs/api_docs/python/tfr/keras/metrics/ARPMetric.md
@@ -7,7 +7,6 @@ description: Average relevance position (ARP).
-
@@ -17,9 +16,11 @@ description: Average relevance position (ARP).
+
+
@@ -192,20 +193,8 @@ gradients back to the corresponding variables.
```
| | `metrics` | List of metrics
-added using the `add_metric()` API.
-
-```
->>> input = tf.keras.layers.Input(shape=(3,))
->>> d = tf.keras.layers.Dense(2)
->>> output = d(input)
->>> d.add_metric(tf.reduce_max(output), name='max')
->>> d.add_metric(tf.reduce_min(output), name='min')
->>> [m.name for m in d.metrics]
-['max', 'min']
-```
-
- | | `name` | Name of the layer
-(string), set in the constructor. | |
+attached to the layer. | | `name` |
+Name of the layer (string), set in the constructor. | |
`name_scope` | Returns a `tf.name_scope`
instance for this class. | |
`non_trainable_weights` | List of all
@@ -352,89 +341,6 @@ Used for backwards compatibility only.
|
-add_metric
-
-
-add_metric(
- value, name=None, **kwargs
-)
-
-
-Adds metric tensor to the layer.
-
-This method can be used inside the `call()` method of a subclassed layer or
-model.
-
-```python
-class MyMetricLayer(tf.keras.layers.Layer):
- def __init__(self):
- super(MyMetricLayer, self).__init__(name='my_metric_layer')
- self.mean = tf.keras.metrics.Mean(name='metric_1')
-
- def call(self, inputs):
- self.add_metric(self.mean(inputs))
- self.add_metric(tf.reduce_sum(inputs), name='metric_2')
- return inputs
-```
-
-This method can also be called directly on a Functional Model during
-construction. In this case, any tensor passed to this Model must be symbolic and
-be able to be traced back to the model's `Input`s. These metrics become part of
-the model's topology and are tracked when you save the model via `save()`.
-
-```python
-inputs = tf.keras.Input(shape=(10,))
-x = tf.keras.layers.Dense(10)(inputs)
-outputs = tf.keras.layers.Dense(1)(x)
-model = tf.keras.Model(inputs, outputs)
-model.add_metric(math_ops.reduce_sum(x), name='metric_1')
-```
-
-Note: Calling `add_metric()` with the result of a metric object on a Functional
-Model, as shown in the example below, is not supported. This is because we
-cannot trace the metric result tensor back to the model's inputs.
-
-```python
-inputs = tf.keras.Input(shape=(10,))
-x = tf.keras.layers.Dense(10)(inputs)
-outputs = tf.keras.layers.Dense(1)(x)
-model = tf.keras.Model(inputs, outputs)
-model.add_metric(tf.keras.metrics.Mean()(x), name='metric_1')
-```
-
-
-
-
-| Args |
|---|
-
-
-|
-`value`
- |
-
-Metric tensor.
- |
-
-|
-`name`
- |
-
-String metric name.
- |
-
-|
-`**kwargs`
- |
-
-Additional keyword arguments for backward compatibility.
-Accepted values:
-`aggregation` - When the `value` tensor provided is not the result
-of calling a `keras.Metric` instance, it will be aggregated by
-default using a `keras.Metric.Mean`.
- |
-
-
-
build
@@ -477,6 +383,29 @@ Instance of `TensorShape`, or list of instances of
)
+Builds the layer's states with the supplied config dict.
+
+By default, this method calls the `build(config["input_shape"])` method, which
+creates weights based on the layer's input shape in the supplied config. If your
+config contains other information needed to load the layer's state, you should
+override this method.
+
+
+
+
+
+| Args |
|---|
+
+
+|
+`config`
+ |
+
+Dict containing the input shape associated with this layer.
+ |
+
+
+
compute_mask
@@ -653,6 +582,30 @@ A layer instance.
get_build_config()
+Returns a dictionary with the layer's input shape.
+
+This method returns a config dict that can be used by
+`build_from_config(config)` to create all states (e.g. Variables and Lookup
+tables) needed by the layer.
+
+By default, the config only contains the input shape that the layer was built
+with. If you're writing a custom layer that creates state in an unusual way, you
+should override this method to make sure this state is already created when
+Keras attempts to load its value upon model loading.
+
+
+
+
+
+| Returns |
|---|
+
+|
+A dict containing the input shape associated with the layer.
+ |
+
+
+
+
get_config
View
@@ -714,6 +667,35 @@ Weights values as a list of NumPy arrays.
|
+load_own_variables
+
+
+load_own_variables(
+ store
+)
+
+
+Loads the state of the layer.
+
+You can override this method to take full control of how the state of the layer
+is loaded upon calling `keras.models.load_model()`.
+
+
+
+
+
+| Args |
|---|
+
+
+|
+`store`
+ |
+
+Dict from which the state of the model will be loaded.
+ |
+
+
+
merge_state
@@ -814,6 +796,35 @@ A scalar tensor, or a dictionary of scalar tensors.
|
+save_own_variables
+
+
+save_own_variables(
+ store
+)
+
+
+Saves the state of the layer.
+
+You can override this method to take full control of how the state of the layer
+is saved upon calling `model.save()`.
+
+
+
+
+
+| Args |
|---|
+
+
+|
+`store`
+ |
+
+Dict where the state of the model will be saved.
+ |
+
+
+
set_weights
diff --git a/docs/api_docs/python/tfr/keras/metrics/AlphaDCGMetric.md b/docs/api_docs/python/tfr/keras/metrics/AlphaDCGMetric.md
index 6cc6598..7ed1ff5 100644
--- a/docs/api_docs/python/tfr/keras/metrics/AlphaDCGMetric.md
+++ b/docs/api_docs/python/tfr/keras/metrics/AlphaDCGMetric.md
@@ -7,7 +7,6 @@ description: Alpha discounted cumulative gain (alphaDCG).
-
@@ -17,9 +16,11 @@ description: Alpha discounted cumulative gain (alphaDCG).
+
+
@@ -309,20 +310,8 @@ gradients back to the corresponding variables.
```
| | `metrics` | List of metrics
-added using the `add_metric()` API.
-
-```
->>> input = tf.keras.layers.Input(shape=(3,))
->>> d = tf.keras.layers.Dense(2)
->>> output = d(input)
->>> d.add_metric(tf.reduce_max(output), name='max')
->>> d.add_metric(tf.reduce_min(output), name='min')
->>> [m.name for m in d.metrics]
-['max', 'min']
-```
-
- | | `name` | Name of the layer
-(string), set in the constructor. | |
+attached to the layer. | | `name` |
+Name of the layer (string), set in the constructor. | |
`name_scope` | Returns a `tf.name_scope`
instance for this class. | |
`non_trainable_weights` | List of all
@@ -469,89 +458,6 @@ Used for backwards compatibility only.
|
-add_metric
-
-
-add_metric(
- value, name=None, **kwargs
-)
-
-
-Adds metric tensor to the layer.
-
-This method can be used inside the `call()` method of a subclassed layer or
-model.
-
-```python
-class MyMetricLayer(tf.keras.layers.Layer):
- def __init__(self):
- super(MyMetricLayer, self).__init__(name='my_metric_layer')
- self.mean = tf.keras.metrics.Mean(name='metric_1')
-
- def call(self, inputs):
- self.add_metric(self.mean(inputs))
- self.add_metric(tf.reduce_sum(inputs), name='metric_2')
- return inputs
-```
-
-This method can also be called directly on a Functional Model during
-construction. In this case, any tensor passed to this Model must be symbolic and
-be able to be traced back to the model's `Input`s. These metrics become part of
-the model's topology and are tracked when you save the model via `save()`.
-
-```python
-inputs = tf.keras.Input(shape=(10,))
-x = tf.keras.layers.Dense(10)(inputs)
-outputs = tf.keras.layers.Dense(1)(x)
-model = tf.keras.Model(inputs, outputs)
-model.add_metric(math_ops.reduce_sum(x), name='metric_1')
-```
-
-Note: Calling `add_metric()` with the result of a metric object on a Functional
-Model, as shown in the example below, is not supported. This is because we
-cannot trace the metric result tensor back to the model's inputs.
-
-```python
-inputs = tf.keras.Input(shape=(10,))
-x = tf.keras.layers.Dense(10)(inputs)
-outputs = tf.keras.layers.Dense(1)(x)
-model = tf.keras.Model(inputs, outputs)
-model.add_metric(tf.keras.metrics.Mean()(x), name='metric_1')
-```
-
-
-
-
-| Args |
|---|
-
-
-|
-`value`
- |
-
-Metric tensor.
- |
-
-|
-`name`
- |
-
-String metric name.
- |
-
-|
-`**kwargs`
- |
-
-Additional keyword arguments for backward compatibility.
-Accepted values:
-`aggregation` - When the `value` tensor provided is not the result
-of calling a `keras.Metric` instance, it will be aggregated by
-default using a `keras.Metric.Mean`.
- |
-
-
-
build
@@ -594,6 +500,29 @@ Instance of `TensorShape`, or list of instances of
)
+Builds the layer's states with the supplied config dict.
+
+By default, this method calls the `build(config["input_shape"])` method, which
+creates weights based on the layer's input shape in the supplied config. If your
+config contains other information needed to load the layer's state, you should
+override this method.
+
+
+
+
+
+| Args |
|---|
+
+
+|
+`config`
+ |
+
+Dict containing the input shape associated with this layer.
+ |
+
+
+
compute_mask
@@ -770,6 +699,30 @@ A layer instance.
get_build_config()
+Returns a dictionary with the layer's input shape.
+
+This method returns a config dict that can be used by
+`build_from_config(config)` to create all states (e.g. Variables and Lookup
+tables) needed by the layer.
+
+By default, the config only contains the input shape that the layer was built
+with. If you're writing a custom layer that creates state in an unusual way, you
+should override this method to make sure this state is already created when
+Keras attempts to load its value upon model loading.
+
+
+
+
+
+| Returns |
|---|
+
+|
+A dict containing the input shape associated with the layer.
+ |
+
+
+
+
get_config
View
@@ -831,6 +784,35 @@ Weights values as a list of NumPy arrays.
|
+load_own_variables
+
+
+load_own_variables(
+ store
+)
+
+
+Loads the state of the layer.
+
+You can override this method to take full control of how the state of the layer
+is loaded upon calling `keras.models.load_model()`.
+
+
+
+
+
+| Args |
|---|
+
+
+|
+`store`
+ |
+
+Dict from which the state of the model will be loaded.
+ |
+
+
+
merge_state
@@ -931,6 +913,35 @@ A scalar tensor, or a dictionary of scalar tensors.
|
+save_own_variables
+
+
+save_own_variables(
+ store
+)
+
+
+Saves the state of the layer.
+
+You can override this method to take full control of how the state of the layer
+is saved upon calling `model.save()`.
+
+
+
+
+
+| Args |
|---|
+
+
+|
+`store`
+ |
+
+Dict where the state of the model will be saved.
+ |
+
+
+
set_weights
diff --git a/docs/api_docs/python/tfr/keras/metrics/DCGMetric.md b/docs/api_docs/python/tfr/keras/metrics/DCGMetric.md
index 84ae3da..5b18b83 100644
--- a/docs/api_docs/python/tfr/keras/metrics/DCGMetric.md
+++ b/docs/api_docs/python/tfr/keras/metrics/DCGMetric.md
@@ -7,7 +7,6 @@ description: Discounted cumulative gain (DCG).
-
@@ -17,9 +16,11 @@ description: Discounted cumulative gain (DCG).
+
+
@@ -222,20 +223,8 @@ gradients back to the corresponding variables.
```
| | `metrics` | List of metrics
-added using the `add_metric()` API.
-
-```
->>> input = tf.keras.layers.Input(shape=(3,))
->>> d = tf.keras.layers.Dense(2)
->>> output = d(input)
->>> d.add_metric(tf.reduce_max(output), name='max')
->>> d.add_metric(tf.reduce_min(output), name='min')
->>> [m.name for m in d.metrics]
-['max', 'min']
-```
-
- | | `name` | Name of the layer
-(string), set in the constructor. | |
+attached to the layer. | | `name` |
+Name of the layer (string), set in the constructor. | |
`name_scope` | Returns a `tf.name_scope`
instance for this class. | |
`non_trainable_weights` | List of all
@@ -382,89 +371,6 @@ Used for backwards compatibility only.
|
-add_metric
-
-
-add_metric(
- value, name=None, **kwargs
-)
-
-
-Adds metric tensor to the layer.
-
-This method can be used inside the `call()` method of a subclassed layer or
-model.
-
-```python
-class MyMetricLayer(tf.keras.layers.Layer):
- def __init__(self):
- super(MyMetricLayer, self).__init__(name='my_metric_layer')
- self.mean = tf.keras.metrics.Mean(name='metric_1')
-
- def call(self, inputs):
- self.add_metric(self.mean(inputs))
- self.add_metric(tf.reduce_sum(inputs), name='metric_2')
- return inputs
-```
-
-This method can also be called directly on a Functional Model during
-construction. In this case, any tensor passed to this Model must be symbolic and
-be able to be traced back to the model's `Input`s. These metrics become part of
-the model's topology and are tracked when you save the model via `save()`.
-
-```python
-inputs = tf.keras.Input(shape=(10,))
-x = tf.keras.layers.Dense(10)(inputs)
-outputs = tf.keras.layers.Dense(1)(x)
-model = tf.keras.Model(inputs, outputs)
-model.add_metric(math_ops.reduce_sum(x), name='metric_1')
-```
-
-Note: Calling `add_metric()` with the result of a metric object on a Functional
-Model, as shown in the example below, is not supported. This is because we
-cannot trace the metric result tensor back to the model's inputs.
-
-```python
-inputs = tf.keras.Input(shape=(10,))
-x = tf.keras.layers.Dense(10)(inputs)
-outputs = tf.keras.layers.Dense(1)(x)
-model = tf.keras.Model(inputs, outputs)
-model.add_metric(tf.keras.metrics.Mean()(x), name='metric_1')
-```
-
-
-
-
-| Args |
|---|
-
-
-|
-`value`
- |
-
-Metric tensor.
- |
-
-|
-`name`
- |
-
-String metric name.
- |
-
-|
-`**kwargs`
- |
-
-Additional keyword arguments for backward compatibility.
-Accepted values:
-`aggregation` - When the `value` tensor provided is not the result
-of calling a `keras.Metric` instance, it will be aggregated by
-default using a `keras.Metric.Mean`.
- |
-
-
-
build
@@ -507,6 +413,29 @@ Instance of `TensorShape`, or list of instances of
)
+Builds the layer's states with the supplied config dict.
+
+By default, this method calls the `build(config["input_shape"])` method, which
+creates weights based on the layer's input shape in the supplied config. If your
+config contains other information needed to load the layer's state, you should
+override this method.
+
+
+
+
+
+| Args |
|---|
+
+
+|
+`config`
+ |
+
+Dict containing the input shape associated with this layer.
+ |
+
+
+
compute_mask
@@ -683,6 +612,30 @@ A layer instance.
get_build_config()
+Returns a dictionary with the layer's input shape.
+
+This method returns a config dict that can be used by
+`build_from_config(config)` to create all states (e.g. Variables and Lookup
+tables) needed by the layer.
+
+By default, the config only contains the input shape that the layer was built
+with. If you're writing a custom layer that creates state in an unusual way, you
+should override this method to make sure this state is already created when
+Keras attempts to load its value upon model loading.
+
+
+
+
+
+| Returns |
|---|
+
+|
+A dict containing the input shape associated with the layer.
+ |
+
+
+
+
get_config
View
@@ -744,6 +697,35 @@ Weights values as a list of NumPy arrays.
|
+load_own_variables
+
+
+load_own_variables(
+ store
+)
+
+
+Loads the state of the layer.
+
+You can override this method to take full control of how the state of the layer
+is loaded upon calling `keras.models.load_model()`.
+
+
+
+
+
+| Args |
|---|
+
+
+|
+`store`
+ |
+
+Dict from which the state of the model will be loaded.
+ |
+
+
+
merge_state
@@ -844,6 +826,35 @@ A scalar tensor, or a dictionary of scalar tensors.
|
+save_own_variables
+
+
+save_own_variables(
+ store
+)
+
+
+Saves the state of the layer.
+
+You can override this method to take full control of how the state of the layer
+is saved upon calling `model.save()`.
+
+
+
+
+
+| Args |
|---|
+
+
+|
+`store`
+ |
+
+Dict where the state of the model will be saved.
+ |
+
+
+
set_weights
diff --git a/docs/api_docs/python/tfr/keras/metrics/HitsMetric.md b/docs/api_docs/python/tfr/keras/metrics/HitsMetric.md
index 23d59c1..643c92f 100644
--- a/docs/api_docs/python/tfr/keras/metrics/HitsMetric.md
+++ b/docs/api_docs/python/tfr/keras/metrics/HitsMetric.md
@@ -7,7 +7,6 @@ description: Hits@k metric.
-
@@ -17,9 +16,11 @@ description: Hits@k metric.
+
+
@@ -202,20 +203,8 @@ gradients back to the corresponding variables.
```
| | `metrics` | List of metrics
-added using the `add_metric()` API.
-
-```
->>> input = tf.keras.layers.Input(shape=(3,))
->>> d = tf.keras.layers.Dense(2)
->>> output = d(input)
->>> d.add_metric(tf.reduce_max(output), name='max')
->>> d.add_metric(tf.reduce_min(output), name='min')
->>> [m.name for m in d.metrics]
-['max', 'min']
-```
-
- | | `name` | Name of the layer
-(string), set in the constructor. | |
+attached to the layer. | | `name` |
+Name of the layer (string), set in the constructor. | |
`name_scope` | Returns a `tf.name_scope`
instance for this class. | |
`non_trainable_weights` | List of all
@@ -362,89 +351,6 @@ Used for backwards compatibility only.
|
-add_metric
-
-
-add_metric(
- value, name=None, **kwargs
-)
-
-
-Adds metric tensor to the layer.
-
-This method can be used inside the `call()` method of a subclassed layer or
-model.
-
-```python
-class MyMetricLayer(tf.keras.layers.Layer):
- def __init__(self):
- super(MyMetricLayer, self).__init__(name='my_metric_layer')
- self.mean = tf.keras.metrics.Mean(name='metric_1')
-
- def call(self, inputs):
- self.add_metric(self.mean(inputs))
- self.add_metric(tf.reduce_sum(inputs), name='metric_2')
- return inputs
-```
-
-This method can also be called directly on a Functional Model during
-construction. In this case, any tensor passed to this Model must be symbolic and
-be able to be traced back to the model's `Input`s. These metrics become part of
-the model's topology and are tracked when you save the model via `save()`.
-
-```python
-inputs = tf.keras.Input(shape=(10,))
-x = tf.keras.layers.Dense(10)(inputs)
-outputs = tf.keras.layers.Dense(1)(x)
-model = tf.keras.Model(inputs, outputs)
-model.add_metric(math_ops.reduce_sum(x), name='metric_1')
-```
-
-Note: Calling `add_metric()` with the result of a metric object on a Functional
-Model, as shown in the example below, is not supported. This is because we
-cannot trace the metric result tensor back to the model's inputs.
-
-```python
-inputs = tf.keras.Input(shape=(10,))
-x = tf.keras.layers.Dense(10)(inputs)
-outputs = tf.keras.layers.Dense(1)(x)
-model = tf.keras.Model(inputs, outputs)
-model.add_metric(tf.keras.metrics.Mean()(x), name='metric_1')
-```
-
-
-
-
-| Args |
|---|
-
-
-|
-`value`
- |
-
-Metric tensor.
- |
-
-|
-`name`
- |
-
-String metric name.
- |
-
-|
-`**kwargs`
- |
-
-Additional keyword arguments for backward compatibility.
-Accepted values:
-`aggregation` - When the `value` tensor provided is not the result
-of calling a `keras.Metric` instance, it will be aggregated by
-default using a `keras.Metric.Mean`.
- |
-
-
-
build
@@ -487,6 +393,29 @@ Instance of `TensorShape`, or list of instances of
)
+Builds the layer's states with the supplied config dict.
+
+By default, this method calls the `build(config["input_shape"])` method, which
+creates weights based on the layer's input shape in the supplied config. If your
+config contains other information needed to load the layer's state, you should
+override this method.
+
+
+
+
+
+| Args |
|---|
+
+
+|
+`config`
+ |
+
+Dict containing the input shape associated with this layer.
+ |
+
+
+
compute_mask
@@ -663,6 +592,30 @@ A layer instance.
get_build_config()
+Returns a dictionary with the layer's input shape.
+
+This method returns a config dict that can be used by
+`build_from_config(config)` to create all states (e.g. Variables and Lookup
+tables) needed by the layer.
+
+By default, the config only contains the input shape that the layer was built
+with. If you're writing a custom layer that creates state in an unusual way, you
+should override this method to make sure this state is already created when
+Keras attempts to load its value upon model loading.
+
+
+
+
+
+| Returns |
|---|
+
+|
+A dict containing the input shape associated with the layer.
+ |
+
+
+
+
get_config
View
@@ -724,6 +677,35 @@ Weights values as a list of NumPy arrays.
|
+load_own_variables
+
+
+load_own_variables(
+ store
+)
+
+
+Loads the state of the layer.
+
+You can override this method to take full control of how the state of the layer
+is loaded upon calling `keras.models.load_model()`.
+
+
+
+
+
+| Args |
|---|
+
+
+|
+`store`
+ |
+
+Dict from which the state of the model will be loaded.
+ |
+
+
+
merge_state
@@ -824,6 +806,35 @@ A scalar tensor, or a dictionary of scalar tensors.
|
+save_own_variables
+
+
+save_own_variables(
+ store
+)
+
+
+Saves the state of the layer.
+
+You can override this method to take full control of how the state of the layer
+is saved upon calling `model.save()`.
+
+
+
+
+
+| Args |
|---|
+
+
+|
+`store`
+ |
+
+Dict where the state of the model will be saved.
+ |
+
+
+
set_weights
diff --git a/docs/api_docs/python/tfr/keras/metrics/MRRMetric.md b/docs/api_docs/python/tfr/keras/metrics/MRRMetric.md
index 4bd26c1..41b4926 100644
--- a/docs/api_docs/python/tfr/keras/metrics/MRRMetric.md
+++ b/docs/api_docs/python/tfr/keras/metrics/MRRMetric.md
@@ -7,7 +7,6 @@ description: Mean reciprocal rank (MRR).
-
@@ -17,9 +16,11 @@ description: Mean reciprocal rank (MRR).
+
+
@@ -201,20 +202,8 @@ gradients back to the corresponding variables.
```
| | `metrics` | List of metrics
-added using the `add_metric()` API.
-
-```
->>> input = tf.keras.layers.Input(shape=(3,))
->>> d = tf.keras.layers.Dense(2)
->>> output = d(input)
->>> d.add_metric(tf.reduce_max(output), name='max')
->>> d.add_metric(tf.reduce_min(output), name='min')
->>> [m.name for m in d.metrics]
-['max', 'min']
-```
-
- | | `name` | Name of the layer
-(string), set in the constructor. | |
+attached to the layer. | | `name` |
+Name of the layer (string), set in the constructor. | |
`name_scope` | Returns a `tf.name_scope`
instance for this class. | |
`non_trainable_weights` | List of all
@@ -361,89 +350,6 @@ Used for backwards compatibility only.
|
-add_metric
-
-
-add_metric(
- value, name=None, **kwargs
-)
-
-
-Adds metric tensor to the layer.
-
-This method can be used inside the `call()` method of a subclassed layer or
-model.
-
-```python
-class MyMetricLayer(tf.keras.layers.Layer):
- def __init__(self):
- super(MyMetricLayer, self).__init__(name='my_metric_layer')
- self.mean = tf.keras.metrics.Mean(name='metric_1')
-
- def call(self, inputs):
- self.add_metric(self.mean(inputs))
- self.add_metric(tf.reduce_sum(inputs), name='metric_2')
- return inputs
-```
-
-This method can also be called directly on a Functional Model during
-construction. In this case, any tensor passed to this Model must be symbolic and
-be able to be traced back to the model's `Input`s. These metrics become part of
-the model's topology and are tracked when you save the model via `save()`.
-
-```python
-inputs = tf.keras.Input(shape=(10,))
-x = tf.keras.layers.Dense(10)(inputs)
-outputs = tf.keras.layers.Dense(1)(x)
-model = tf.keras.Model(inputs, outputs)
-model.add_metric(math_ops.reduce_sum(x), name='metric_1')
-```
-
-Note: Calling `add_metric()` with the result of a metric object on a Functional
-Model, as shown in the example below, is not supported. This is because we
-cannot trace the metric result tensor back to the model's inputs.
-
-```python
-inputs = tf.keras.Input(shape=(10,))
-x = tf.keras.layers.Dense(10)(inputs)
-outputs = tf.keras.layers.Dense(1)(x)
-model = tf.keras.Model(inputs, outputs)
-model.add_metric(tf.keras.metrics.Mean()(x), name='metric_1')
-```
-
-
-
-
-| Args |
|---|
-
-
-|
-`value`
- |
-
-Metric tensor.
- |
-
-|
-`name`
- |
-
-String metric name.
- |
-
-|
-`**kwargs`
- |
-
-Additional keyword arguments for backward compatibility.
-Accepted values:
-`aggregation` - When the `value` tensor provided is not the result
-of calling a `keras.Metric` instance, it will be aggregated by
-default using a `keras.Metric.Mean`.
- |
-
-
-
build
@@ -486,6 +392,29 @@ Instance of `TensorShape`, or list of instances of
)
+Builds the layer's states with the supplied config dict.
+
+By default, this method calls the `build(config["input_shape"])` method, which
+creates weights based on the layer's input shape in the supplied config. If your
+config contains other information needed to load the layer's state, you should
+override this method.
+
+
+
+
+
+| Args |
|---|
+
+
+|
+`config`
+ |
+
+Dict containing the input shape associated with this layer.
+ |
+
+
+
compute_mask
@@ -662,6 +591,30 @@ A layer instance.
get_build_config()
+Returns a dictionary with the layer's input shape.
+
+This method returns a config dict that can be used by
+`build_from_config(config)` to create all states (e.g. Variables and Lookup
+tables) needed by the layer.
+
+By default, the config only contains the input shape that the layer was built
+with. If you're writing a custom layer that creates state in an unusual way, you
+should override this method to make sure this state is already created when
+Keras attempts to load its value upon model loading.
+
+
+
+
+
+| Returns |
|---|
+
+|
+A dict containing the input shape associated with the layer.
+ |
+
+
+
+
get_config
View
@@ -723,6 +676,35 @@ Weights values as a list of NumPy arrays.
|
+load_own_variables
+
+
+load_own_variables(
+ store
+)
+
+
+Loads the state of the layer.
+
+You can override this method to take full control of how the state of the layer
+is loaded upon calling `keras.models.load_model()`.
+
+
+
+
+
+| Args |
|---|
+
+
+|
+`store`
+ |
+
+Dict from which the state of the model will be loaded.
+ |
+
+
+
merge_state
@@ -823,6 +805,35 @@ A scalar tensor, or a dictionary of scalar tensors.
|
+save_own_variables
+
+
+save_own_variables(
+ store
+)
+
+
+Saves the state of the layer.
+
+You can override this method to take full control of how the state of the layer
+is saved upon calling `model.save()`.
+
+
+
+
+
+| Args |
|---|
+
+
+|
+`store`
+ |
+
+Dict where the state of the model will be saved.
+ |
+
+
+
set_weights
diff --git a/docs/api_docs/python/tfr/keras/metrics/MeanAveragePrecisionMetric.md b/docs/api_docs/python/tfr/keras/metrics/MeanAveragePrecisionMetric.md
index 7b3bd63..b1e4671 100644
--- a/docs/api_docs/python/tfr/keras/metrics/MeanAveragePrecisionMetric.md
+++ b/docs/api_docs/python/tfr/keras/metrics/MeanAveragePrecisionMetric.md
@@ -7,7 +7,6 @@ description: Mean average precision (MAP).
-
@@ -17,9 +16,11 @@ description: Mean average precision (MAP).
+
+
@@ -209,20 +210,8 @@ gradients back to the corresponding variables.
```
| | `metrics` | List of metrics
-added using the `add_metric()` API.
-
-```
->>> input = tf.keras.layers.Input(shape=(3,))
->>> d = tf.keras.layers.Dense(2)
->>> output = d(input)
->>> d.add_metric(tf.reduce_max(output), name='max')
->>> d.add_metric(tf.reduce_min(output), name='min')
->>> [m.name for m in d.metrics]
-['max', 'min']
-```
-
- | | `name` | Name of the layer
-(string), set in the constructor. | |
+attached to the layer. | | `name` |
+Name of the layer (string), set in the constructor. | |
`name_scope` | Returns a `tf.name_scope`
instance for this class. | |
`non_trainable_weights` | List of all
@@ -369,89 +358,6 @@ Used for backwards compatibility only.
|
-add_metric
-
-
-add_metric(
- value, name=None, **kwargs
-)
-
-
-Adds metric tensor to the layer.
-
-This method can be used inside the `call()` method of a subclassed layer or
-model.
-
-```python
-class MyMetricLayer(tf.keras.layers.Layer):
- def __init__(self):
- super(MyMetricLayer, self).__init__(name='my_metric_layer')
- self.mean = tf.keras.metrics.Mean(name='metric_1')
-
- def call(self, inputs):
- self.add_metric(self.mean(inputs))
- self.add_metric(tf.reduce_sum(inputs), name='metric_2')
- return inputs
-```
-
-This method can also be called directly on a Functional Model during
-construction. In this case, any tensor passed to this Model must be symbolic and
-be able to be traced back to the model's `Input`s. These metrics become part of
-the model's topology and are tracked when you save the model via `save()`.
-
-```python
-inputs = tf.keras.Input(shape=(10,))
-x = tf.keras.layers.Dense(10)(inputs)
-outputs = tf.keras.layers.Dense(1)(x)
-model = tf.keras.Model(inputs, outputs)
-model.add_metric(math_ops.reduce_sum(x), name='metric_1')
-```
-
-Note: Calling `add_metric()` with the result of a metric object on a Functional
-Model, as shown in the example below, is not supported. This is because we
-cannot trace the metric result tensor back to the model's inputs.
-
-```python
-inputs = tf.keras.Input(shape=(10,))
-x = tf.keras.layers.Dense(10)(inputs)
-outputs = tf.keras.layers.Dense(1)(x)
-model = tf.keras.Model(inputs, outputs)
-model.add_metric(tf.keras.metrics.Mean()(x), name='metric_1')
-```
-
-
-
-
-| Args |
|---|
-
-
-|
-`value`
- |
-
-Metric tensor.
- |
-
-|
-`name`
- |
-
-String metric name.
- |
-
-|
-`**kwargs`
- |
-
-Additional keyword arguments for backward compatibility.
-Accepted values:
-`aggregation` - When the `value` tensor provided is not the result
-of calling a `keras.Metric` instance, it will be aggregated by
-default using a `keras.Metric.Mean`.
- |
-
-
-
build
@@ -494,6 +400,29 @@ Instance of `TensorShape`, or list of instances of
)
+Builds the layer's states with the supplied config dict.
+
+By default, this method calls the `build(config["input_shape"])` method, which
+creates weights based on the layer's input shape in the supplied config. If your
+config contains other information needed to load the layer's state, you should
+override this method.
+
+
+
+
+
+| Args |
|---|
+
+
+|
+`config`
+ |
+
+Dict containing the input shape associated with this layer.
+ |
+
+
+
compute_mask
@@ -670,6 +599,30 @@ A layer instance.
get_build_config()
+Returns a dictionary with the layer's input shape.
+
+This method returns a config dict that can be used by
+`build_from_config(config)` to create all states (e.g. Variables and Lookup
+tables) needed by the layer.
+
+By default, the config only contains the input shape that the layer was built
+with. If you're writing a custom layer that creates state in an unusual way, you
+should override this method to make sure this state is already created when
+Keras attempts to load its value upon model loading.
+
+
+
+
+
+| Returns |
|---|
+
+|
+A dict containing the input shape associated with the layer.
+ |
+
+
+
+
get_config
View
@@ -731,6 +684,35 @@ Weights values as a list of NumPy arrays.
|
+load_own_variables
+
+
+load_own_variables(
+ store
+)
+
+
+Loads the state of the layer.
+
+You can override this method to take full control of how the state of the layer
+is loaded upon calling `keras.models.load_model()`.
+
+
+
+
+
+| Args |
|---|
+
+
+|
+`store`
+ |
+
+Dict from which the state of the model will be loaded.
+ |
+
+
+
merge_state
@@ -831,6 +813,35 @@ A scalar tensor, or a dictionary of scalar tensors.
|
+save_own_variables
+
+
+save_own_variables(
+ store
+)
+
+
+Saves the state of the layer.
+
+You can override this method to take full control of how the state of the layer
+is saved upon calling `model.save()`.
+
+
+
+
+
+| Args |
|---|
+
+
+|
+`store`
+ |
+
+Dict where the state of the model will be saved.
+ |
+
+
+
set_weights
diff --git a/docs/api_docs/python/tfr/keras/metrics/NDCGMetric.md b/docs/api_docs/python/tfr/keras/metrics/NDCGMetric.md
index 92f088a..c9c52ad 100644
--- a/docs/api_docs/python/tfr/keras/metrics/NDCGMetric.md
+++ b/docs/api_docs/python/tfr/keras/metrics/NDCGMetric.md
@@ -7,7 +7,6 @@ description: Normalized discounted cumulative gain (NDCG).
-
@@ -17,9 +16,11 @@ description: Normalized discounted cumulative gain (NDCG).
+
+
@@ -227,20 +228,8 @@ gradients back to the corresponding variables.
```
| | `metrics` | List of metrics
-added using the `add_metric()` API.
-
-```
->>> input = tf.keras.layers.Input(shape=(3,))
->>> d = tf.keras.layers.Dense(2)
->>> output = d(input)
->>> d.add_metric(tf.reduce_max(output), name='max')
->>> d.add_metric(tf.reduce_min(output), name='min')
->>> [m.name for m in d.metrics]
-['max', 'min']
-```
-
- | | `name` | Name of the layer
-(string), set in the constructor. | |
+attached to the layer. | | `name` |
+Name of the layer (string), set in the constructor. | |
`name_scope` | Returns a `tf.name_scope`
instance for this class. | |
`non_trainable_weights` | List of all
@@ -387,89 +376,6 @@ Used for backwards compatibility only.
|
-add_metric
-
-
-add_metric(
- value, name=None, **kwargs
-)
-
-
-Adds metric tensor to the layer.
-
-This method can be used inside the `call()` method of a subclassed layer or
-model.
-
-```python
-class MyMetricLayer(tf.keras.layers.Layer):
- def __init__(self):
- super(MyMetricLayer, self).__init__(name='my_metric_layer')
- self.mean = tf.keras.metrics.Mean(name='metric_1')
-
- def call(self, inputs):
- self.add_metric(self.mean(inputs))
- self.add_metric(tf.reduce_sum(inputs), name='metric_2')
- return inputs
-```
-
-This method can also be called directly on a Functional Model during
-construction. In this case, any tensor passed to this Model must be symbolic and
-be able to be traced back to the model's `Input`s. These metrics become part of
-the model's topology and are tracked when you save the model via `save()`.
-
-```python
-inputs = tf.keras.Input(shape=(10,))
-x = tf.keras.layers.Dense(10)(inputs)
-outputs = tf.keras.layers.Dense(1)(x)
-model = tf.keras.Model(inputs, outputs)
-model.add_metric(math_ops.reduce_sum(x), name='metric_1')
-```
-
-Note: Calling `add_metric()` with the result of a metric object on a Functional
-Model, as shown in the example below, is not supported. This is because we
-cannot trace the metric result tensor back to the model's inputs.
-
-```python
-inputs = tf.keras.Input(shape=(10,))
-x = tf.keras.layers.Dense(10)(inputs)
-outputs = tf.keras.layers.Dense(1)(x)
-model = tf.keras.Model(inputs, outputs)
-model.add_metric(tf.keras.metrics.Mean()(x), name='metric_1')
-```
-
-
-
-
-| Args |
|---|
-
-
-|
-`value`
- |
-
-Metric tensor.
- |
-
-|
-`name`
- |
-
-String metric name.
- |
-
-|
-`**kwargs`
- |
-
-Additional keyword arguments for backward compatibility.
-Accepted values:
-`aggregation` - When the `value` tensor provided is not the result
-of calling a `keras.Metric` instance, it will be aggregated by
-default using a `keras.Metric.Mean`.
- |
-
-
-
build
@@ -512,6 +418,29 @@ Instance of `TensorShape`, or list of instances of
)
+Builds the layer's states with the supplied config dict.
+
+By default, this method calls the `build(config["input_shape"])` method, which
+creates weights based on the layer's input shape in the supplied config. If your
+config contains other information needed to load the layer's state, you should
+override this method.
+
+
+
+
+
+| Args |
|---|
+
+
+|
+`config`
+ |
+
+Dict containing the input shape associated with this layer.
+ |
+
+
+
compute_mask
@@ -688,6 +617,30 @@ A layer instance.
get_build_config()
+Returns a dictionary with the layer's input shape.
+
+This method returns a config dict that can be used by
+`build_from_config(config)` to create all states (e.g. Variables and Lookup
+tables) needed by the layer.
+
+By default, the config only contains the input shape that the layer was built
+with. If you're writing a custom layer that creates state in an unusual way, you
+should override this method to make sure this state is already created when
+Keras attempts to load its value upon model loading.
+
+
+
+
+
+| Returns |
|---|
+
+|
+A dict containing the input shape associated with the layer.
+ |
+
+
+
+
get_config
View
@@ -749,6 +702,35 @@ Weights values as a list of NumPy arrays.
+load_own_variables
+
+
+load_own_variables(
+ store
+)
+
+
+Loads the state of the layer.
+
+You can override this method to take full control of how the state of the layer
+is loaded upon calling `keras.models.load_model()`.
+
+
+
+
+
+| Args |
|---|
+
+
+|
+`store`
+ |
+
+Dict from which the state of the model will be loaded.
+ |
+
+
+
merge_state
@@ -849,6 +831,35 @@ A scalar tensor, or a dictionary of scalar tensors.
+save_own_variables
+
+
+save_own_variables(
+ store
+)
+
+
+Saves the state of the layer.
+
+You can override this method to take full control of how the state of the layer
+is saved upon calling `model.save()`.
+
+
+
+
+
+| Args |
|---|
+
+
+|
+`store`
+ |
+
+Dict where the state of the model will be saved.
+ |
+
+
+
set_weights
diff --git a/docs/api_docs/python/tfr/keras/metrics/OPAMetric.md b/docs/api_docs/python/tfr/keras/metrics/OPAMetric.md
index 9977e47..7719964 100644
--- a/docs/api_docs/python/tfr/keras/metrics/OPAMetric.md
+++ b/docs/api_docs/python/tfr/keras/metrics/OPAMetric.md
@@ -7,7 +7,6 @@ description: Ordered pair accuracy (OPA).
-
@@ -17,9 +16,11 @@ description: Ordered pair accuracy (OPA).
+
+
@@ -200,20 +201,8 @@ gradients back to the corresponding variables.
```
| | `metrics` | List of metrics
-added using the `add_metric()` API.
-
-```
->>> input = tf.keras.layers.Input(shape=(3,))
->>> d = tf.keras.layers.Dense(2)
->>> output = d(input)
->>> d.add_metric(tf.reduce_max(output), name='max')
->>> d.add_metric(tf.reduce_min(output), name='min')
->>> [m.name for m in d.metrics]
-['max', 'min']
-```
-
- | | `name` | Name of the layer
-(string), set in the constructor. | |
+attached to the layer. | | `name` |
+Name of the layer (string), set in the constructor. | |
`name_scope` | Returns a `tf.name_scope`
instance for this class. | |
`non_trainable_weights` | List of all
@@ -360,89 +349,6 @@ Used for backwards compatibility only.
|
-add_metric
-
-
-add_metric(
- value, name=None, **kwargs
-)
-
-
-Adds metric tensor to the layer.
-
-This method can be used inside the `call()` method of a subclassed layer or
-model.
-
-```python
-class MyMetricLayer(tf.keras.layers.Layer):
- def __init__(self):
- super(MyMetricLayer, self).__init__(name='my_metric_layer')
- self.mean = tf.keras.metrics.Mean(name='metric_1')
-
- def call(self, inputs):
- self.add_metric(self.mean(inputs))
- self.add_metric(tf.reduce_sum(inputs), name='metric_2')
- return inputs
-```
-
-This method can also be called directly on a Functional Model during
-construction. In this case, any tensor passed to this Model must be symbolic and
-be able to be traced back to the model's `Input`s. These metrics become part of
-the model's topology and are tracked when you save the model via `save()`.
-
-```python
-inputs = tf.keras.Input(shape=(10,))
-x = tf.keras.layers.Dense(10)(inputs)
-outputs = tf.keras.layers.Dense(1)(x)
-model = tf.keras.Model(inputs, outputs)
-model.add_metric(math_ops.reduce_sum(x), name='metric_1')
-```
-
-Note: Calling `add_metric()` with the result of a metric object on a Functional
-Model, as shown in the example below, is not supported. This is because we
-cannot trace the metric result tensor back to the model's inputs.
-
-```python
-inputs = tf.keras.Input(shape=(10,))
-x = tf.keras.layers.Dense(10)(inputs)
-outputs = tf.keras.layers.Dense(1)(x)
-model = tf.keras.Model(inputs, outputs)
-model.add_metric(tf.keras.metrics.Mean()(x), name='metric_1')
-```
-
-
-
-
-| Args |
|---|
-
-
-|
-`value`
- |
-
-Metric tensor.
- |
-
-|
-`name`
- |
-
-String metric name.
- |
-
-|
-`**kwargs`
- |
-
-Additional keyword arguments for backward compatibility.
-Accepted values:
-`aggregation` - When the `value` tensor provided is not the result
-of calling a `keras.Metric` instance, it will be aggregated by
-default using a `keras.Metric.Mean`.
- |
-
-
-
build
@@ -485,6 +391,29 @@ Instance of `TensorShape`, or list of instances of
)
+Builds the layer's states with the supplied config dict.
+
+By default, this method calls the `build(config["input_shape"])` method, which
+creates weights based on the layer's input shape in the supplied config. If your
+config contains other information needed to load the layer's state, you should
+override this method.
+
+
+
+
+
+| Args |
|---|
+
+
+|
+`config`
+ |
+
+Dict containing the input shape associated with this layer.
+ |
+
+
+
compute_mask
@@ -661,6 +590,30 @@ A layer instance.
get_build_config()
+Returns a dictionary with the layer's input shape.
+
+This method returns a config dict that can be used by
+`build_from_config(config)` to create all states (e.g. Variables and Lookup
+tables) needed by the layer.
+
+By default, the config only contains the input shape that the layer was built
+with. If you're writing a custom layer that creates state in an unusual way, you
+should override this method to make sure this state is already created when
+Keras attempts to load its value upon model loading.
+
+
+
+
+
+| Returns |
|---|
+
+|
+A dict containing the input shape associated with the layer.
+ |
+
+
+
+
get_config
View
@@ -722,6 +675,35 @@ Weights values as a list of NumPy arrays.
|
+load_own_variables
+
+
+load_own_variables(
+ store
+)
+
+
+Loads the state of the layer.
+
+You can override this method to take full control of how the state of the layer
+is loaded upon calling `keras.models.load_model()`.
+
+
+
+
+
+| Args |
|---|
+
+
+|
+`store`
+ |
+
+Dict from which the state of the model will be loaded.
+ |
+
+
+
merge_state
@@ -822,6 +804,35 @@ A scalar tensor, or a dictionary of scalar tensors.
|
+save_own_variables
+
+
+save_own_variables(
+ store
+)
+
+
+Saves the state of the layer.
+
+You can override this method to take full control of how the state of the layer
+is saved upon calling `model.save()`.
+
+
+
+
+
+| Args |
|---|
+
+
+|
+`store`
+ |
+
+Dict where the state of the model will be saved.
+ |
+
+
+
set_weights
diff --git a/docs/api_docs/python/tfr/keras/metrics/PrecisionIAMetric.md b/docs/api_docs/python/tfr/keras/metrics/PrecisionIAMetric.md
index 8ab15f2..00f6204 100644
--- a/docs/api_docs/python/tfr/keras/metrics/PrecisionIAMetric.md
+++ b/docs/api_docs/python/tfr/keras/metrics/PrecisionIAMetric.md
@@ -7,7 +7,6 @@ description: Precision-IA@k (Pre-IA@k).
-
@@ -17,9 +16,11 @@ description: Precision-IA@k (Pre-IA@k).
+
+
@@ -266,20 +267,8 @@ gradients back to the corresponding variables.
```
|