Makes a TensorFlow dataset.
tf_agents.replay_buffers.reverb_replay_buffer.make_reverb_dataset(
server_address: str,
table: str,
data_spec,
max_in_flight_samples_per_worker=10,
batch_size=None,
prefetch_size=None,
sequence_length=None,
cycle_length=None,
num_parallel_calls=None,
per_sequence_fn=None,
dataset_transformation=None,
num_workers_per_iterator=-1,
max_samples_per_stream=-1,
rate_limiter_timeout_ms=-1
) -> tf.data.Dataset
Args |
server_address
|
The server address of the replay server.
|
table
|
The name of the table to sample from replay.
|
data_spec
|
The data's spec.
|
max_in_flight_samples_per_worker
|
Optional, dataset buffer capacity.
|
batch_size
|
Optional. If specified the dataset returned will combine
consecutive elements into batches. This argument is also used to determine
the cycle_length for tf.data.Dataset.interleave -- if unspecified the
cycle length is set to tf.data.experimental.AUTOTUNE .
|
prefetch_size
|
How many batches to prefectch in the pipeline.
|
sequence_length
|
Optional. If specified consecutive elements of each
interleaved dataset will be combined into sequences.
|
cycle_length
|
Optional. When equal to batch_size it would make take a sample
from a different sequence. For reducing memory usage use a smaller number.
|
num_parallel_calls
|
Optional. If specified number of parallel calls in
iterleave. By default use tf.data.experimental.AUTOTUNE .
|
per_sequence_fn
|
Optional, per sequence function.
|
dataset_transformation
|
Optional, per dataset interleave transformation.
|
num_workers_per_iterator
|
(Defaults to -1, i.e auto selected) The number of
worker threads to create per dataset iterator. When the selected table
uses a FIFO sampler (i.e a queue) then exactly 1 worker must be used to
avoid races causing invalid ordering of items. For all other samplers,
this value should be roughly equal to the number of threads available on
the CPU.
|
max_samples_per_stream
|
(Defaults to -1, i.e auto selected) The maximum
number of samples to fetch from a stream before a new call is made.
Keeping this number low ensures that the data is fetched uniformly from
all server.
|
rate_limiter_timeout_ms
|
Timeout (in milliseconds) to wait on the rate
limiter when sampling from the table. If rate_limiter_timeout_ms >= 0 ,
this is the timeout passed to Table::Sample describing how long to wait
for the rate limiter to allow sampling.
|
Returns |
A tf.data.Dataset that streams data from the replay server.
|