Skip to content

Latest commit

 

History

History
89 lines (48 loc) · 4.02 KB

dynamic_embedding.md

File metadata and controls

89 lines (48 loc) · 4.02 KB

Module: tfra.dynamic_embedding

View source on GitHub




Export dynamic_embedding APIs.

Modules

keras module

math module: math operations.

shadow_ops module: Dynamic Embedding is designed for Large-scale Sparse Weights Training.

Classes

class CuckooHashTable: A generic mutable hash table implementation.

class CuckooHashTableConfig

class CuckooHashTableCreator: A generic KV table creator.

class FrequencyRestrictPolicy: A derived policy to eliminate features in variable follow the

class GraphKeys: (Deprecated) extended standard names related to

class ModelMode: The global config of model modes.

class HkvHashTable: A generic mutable hash table implementation.

class HkvHashTableConfig: HkvHashTableConfig config init_capacity, max_capacity, max_hbm_for_values of HkvHashTable

class HkvHashTableCreator: A generic KV table creator.

class RedisTable: A generic mutable hash table implementation.

class RedisTableConfig: RedisTableConfig config json file for connecting Redis service and

class RedisTableCreator: RedisTableCreator will create a object to pass itself to the others classes

class RestrictPolicy: Base class of restrict policies. Never use this class directly, but

class TimestampRestrictPolicy: A derived policy to eliminate features in variable follow the

class TrainableWrapper: This class is a trainable wrapper of Dynamic Embedding,

class Variable: A Distributed version of HashTable(reference from lookup_ops.MutableHashTable)

Functions

DynamicEmbeddingOptimizer(...): An optimizer wrapper to make any TensorFlow optimizer capable of training

embedding_lookup(...): Provides a dynamic version of embedding_lookup

embedding_lookup_sparse(...): Provides a dynamic version of embedding_lookup_sparse

embedding_lookup_unique(...): Version of embedding_lookup that avoids duplicate lookups.

enable_inference_mode(...): set inference mode.

enable_train_mode(...): enable train mode.

get_model_mode(...): get model mode.

get_variable(...): Gets an Variable object with this name if it exists,

safe_embedding_lookup_sparse(...): Provides a dynamic version of tf.nn.safe_embedding_lookup_sparse.