merlin.models.tf.InBatchSampler

class merlin.models.tf.InBatchSampler(*args, **kwargs)

Bases: merlin.models.tf.blocks.sampling.base.ItemSampler

Provides in-batch sampling 1 for two-tower item retrieval models. The implementation is very simple, as it just returns the current item embeddings and metadata, but it is necessary to have InBatchSampler under the same interface of other more advanced samplers (e.g. CachedCrossBatchSampler). In a nutshell, for a given (user,item) embeddings pair, the other in-batch item embeddings are used as negative items, rather than computing different embeddings exclusively for negative items. This is a popularity-biased sampling as popular items are observed more often in training batches. P.s. Ignoring the false negatives (negative items equal to the positive ones) is managed by ItemRetrievalScorer(…, sampling_downscore_false_negatives=True)

References

1

Yi, Xinyang, et al. “Sampling-bias-corrected neural modeling for large corpus item recommendations.” Proceedings of the 13th ACM Conference on Recommender Systems. 2019.

Parameters

batch_size (int, optional) – The batch size. If not set it is inferred when the layer is built (first call())

__init__(batch_size: Optional[int] = None, **kwargs)

Methods

__init__([batch_size])
add(inputs[, training])
add_loss(losses, **kwargs)	Add loss tensor(s), potentially dependent on layer inputs.
add_metric(value[, name])	Adds metric tensor to the layer.
add_update(updates)	Add update op(s), potentially dependent on layer inputs.
add_variable(*args, **kwargs)	Deprecated, do NOT use! Alias for add_weight.
add_weight([name, shape, dtype, ...])	Adds a new variable to the layer.
build(input_shapes)
build_from_config(config)
call(inputs[, training])	Returns the item embeddings and item metadata from the current batch.
compute_mask(inputs[, mask])	Computes an output mask tensor.
compute_output_shape(input_shape)	Computes the output shape of the layer.
compute_output_signature(input_signature)	Compute the output tensor signature of the layer based on the inputs.
count_params()	Count the total number of scalars composing the weights.
finalize_state()	Finalizes the layers state after updating layer weights.
from_config(config)	Creates a layer from its config.
get_build_config()
get_config()
get_input_at(node_index)	Retrieves the input tensor(s) of a layer at a given node.
get_input_mask_at(node_index)	Retrieves the input mask tensor(s) of a layer at a given node.
get_input_shape_at(node_index)	Retrieves the input shape(s) of a layer at a given node.
get_output_at(node_index)	Retrieves the output tensor(s) of a layer at a given node.
get_output_mask_at(node_index)	Retrieves the output mask tensor(s) of a layer at a given node.
get_output_shape_at(node_index)	Retrieves the output shape(s) of a layer at a given node.
get_weights()	Returns the current weights of the layer, as NumPy arrays.
sample()
set_batch_size(value)
set_max_num_samples(value)
set_weights(weights)	Sets the weights of the layer, from NumPy arrays.
with_name_scope(method)	Decorator to automatically enter the module name scope.

Attributes

activity_regularizer	Optional regularizer function for the output of this layer.
batch_size
compute_dtype	The dtype of the layer's computations.
dtype	The dtype of the layer weights.
dtype_policy	The dtype policy associated with this layer.
dynamic	Whether the layer is dynamic (eager-only); set in the constructor.
inbound_nodes	Return Functional API nodes upstream of this layer.
input	Retrieves the input tensor(s) of a layer.
input_mask	Retrieves the input mask tensor(s) of a layer.
input_shape	Retrieves the input shape(s) of a layer.
input_spec	InputSpec instance(s) describing the input format for this layer.
losses	List of losses added using the add_loss() API.
max_num_samples
metrics	List of metrics added using the add_metric() API.
name	Name of the layer (string), set in the constructor.
name_scope	Returns a tf.name_scope instance for this class.
non_trainable_variables
non_trainable_weights	List of all non-trainable weights tracked by this layer.
outbound_nodes	Return Functional API nodes downstream of this layer.
output	Retrieves the output tensor(s) of a layer.
output_mask	Retrieves the output mask tensor(s) of a layer.
output_shape	Retrieves the output shape(s) of a layer.
required_features
stateful
submodules	Sequence of all sub-modules.
supports_masking	Whether this layer supports computing a mask using compute_mask.
trainable
trainable_variables
trainable_weights	List of all trainable weights tracked by this layer.
updates
variable_dtype	Alias of Layer.dtype, the dtype of the weights.
variables	Returns the list of all layer variables/weights.
weights	Returns the list of all layer variables/weights.

property batch_size: int

set_batch_size(value)

build(input_shapes: Dict[str, tensorflow.python.framework.ops.Tensor]) → None

call(inputs: Dict[str, tensorflow.python.framework.ops.Tensor], training=True) → merlin.models.tf.core.base.EmbeddingWithMetadata

Returns the item embeddings and item metadata from the current batch. The implementation is very simple, as it just returns the current item embeddings and metadata, but it is necessary to have InBatchSampler under the same interface of other more advanced samplers (e.g. CachedCrossBatchSampler).

Parameters

• inputs (TabularData) –

  Dict with two keys:

  ”items_embeddings”: Items embeddings tensor “items_metadata”: Dict like {“<feature name>”: “<feature tensor>”} which contains features that might be relevant for the sampler. The InBatchSampler does not use metadata features specifically, but “item_id” is required when using in combination with ItemRetrievalScorer(…, sampling_downscore_false_negatives=True), so that false negatives are identified and downscored.

• training (bool, optional) – Flag indicating if on training mode, by default True

Returns

Value object with the sampled item embeddings and item metadata

Return type

EmbeddingWithMetadata

add(inputs: Dict[str, tensorflow.python.framework.ops.Tensor], training=True) → None

sample() → merlin.models.tf.core.base.EmbeddingWithMetadata

get_config()