tf.raw_ops.XlaSparseDenseMatmulGradWithAdagradMomentumAndCsrInput

tf.raw_ops.XlaSparseDenseMatmulGradWithAdagradMomentumAndCsrInput(
    row_pointers,
    sorted_sample_ids,
    sorted_token_ids,
    sorted_gains,
    activation_gradients,
    learning_rate,
    embedding_table,
    accumulator,
    momenta,
    num_minibatches_per_physical_sparse_core,
    use_nesterov,
    exponent,
    beta1,
    beta2,
    epsilon,
    table_name,
    clip_weight_min=float('-inf'),
    clip_weight_max=float('inf'),
    name=None
)

row_pointersA Tensor of type int32.
sorted_sample_idsA Tensor of type int32.
sorted_token_idsA Tensor of type int32.
sorted_gainsA Tensor of type float32.
activation_gradientsA Tensor of type float32.
learning_rateA Tensor of type float32.
embedding_tableA Tensor of type float32.
accumulatorA Tensor of type float32.
momentaA Tensor of type float32.
num_minibatches_per_physical_sparse_coreA Tensor of type int32.
use_nesterovA bool.
exponentA float.
beta1A float.
beta2A float.
epsilonA float.
table_nameA string.
clip_weight_minAn optional float. Defaults to float('-inf').
clip_weight_maxAn optional float. Defaults to float('inf').
nameA name for the operation (optional).

A tuple of Tensor objects (updated_embedding_table, updated_accumulator, updated_momenta).
updated_embedding_tableA Tensor of type float32.
updated_accumulatorA Tensor of type float32.
updated_momentaA Tensor of type float32.