tf_agents.specs.bandit_spec_utils.get_context_dims_from_spec

Returns the global and per-arm context dimensions.

tf_agents.specs.bandit_spec_utils.get_context_dims_from_spec(
    context_spec: tf_agents.typing.types.NestedTensorSpec,
    accepts_per_arm_features: bool
) -> Tuple[int, int]

If the policy accepts per-arm features, this function returns the tuple of the global and per-arm context dimension. Otherwise, it returns the (global) context dim and zero.

Args
`context_spec`	A nest of tensor specs, containing the observation spec.
`accepts_per_arm_features`	(bool) Whether the context_spec is for a policy that accepts per-arm features.

Returns: A 2-tuple of ints, the global and per-arm context dimension. If the policy does not accept per-arm features, the per-arm context dim is 0.

Except as otherwise noted, the content of this page is licensed under the Creative Commons Attribution 4.0 License, and code samples are licensed under the Apache 2.0 License. For details, see the Google Developers Site Policies. Java is a registered trademark of Oracle and/or its affiliates.

Last updated 2024-04-26 UTC.

tf_agents.specs.bandit_spec_utils.get_context_dims_from_spec

Args