package syft_proto.frameworks.torch.tensors.interpreters.v1

Get desktop application:
View/edit binary Protocol Buffers messages

Used in: FixedPrecisionTensor

• optional types.syft.v1.Id id = 1

• oneof field_size

  • int64 field_int = 2

  • string field_str = 3

• string dtype = 4

• optional types.syft.v1.Id crypto_provider_id = 5

• repeated types.syft.v1.Id location_ids = 6

  It would be natural to capture a {location_id->share} dictionary as a map, but the key types of a map can't be messages. The ids use the Id type because they can be either integers or strings. Capturing them as two parallel lists/arrays isn't ideal but ought to work okay (given that constraint.)

• repeated generic.pointers.v1.PointerTensor shares = 7

• optional types.syft.v1.Id id = 1

• string field = 3

• string dtype = 4

• int32 base = 5

• int32 kappa = 6

• int32 precision_fractional = 7

• repeated string tags = 8

• string description = 9

• optional AdditiveSharingTensor child = 10

• optional types.syft.v1.Id id = 1

• bytes serialized_model = 2

• repeated string tags = 3

• string description = 4

• optional types.syft.v1.Id id = 1

• oneof ring_size

  • int64 ring_int = 2

  • string ring_str = 3

• string dtype = 4

• repeated types.syft.v1.Id location_ids = 5

  It would be natural to capture a {location_id->share} dictionary as a map, but the key types of a map can't be messages. The ids use the Id type because they can be either integers or strings. Capturing them as two parallel lists/arrays isn't ideal but ought to work okay (given that constraint.)

• repeated generic.pointers.v1.PointerTensor shares = 6