tf.compat.v1.GPUOptions.Experimental

A ProtocolMessage

Attributes
`collective_ring_order`	`string collective_ring_order`
`disallow_retry_on_allocation_failure`	`bool disallow_retry_on_allocation_failure`
`gpu_host_mem_disallow_growth`	`bool gpu_host_mem_disallow_growth`
`gpu_host_mem_limit_in_mb`	`float gpu_host_mem_limit_in_mb`
`gpu_system_memory_size_in_mb`	`int32 gpu_system_memory_size_in_mb`
`internal_fragmentation_fraction`	`double internal_fragmentation_fraction`
`kernel_tracker_max_bytes`	`int32 kernel_tracker_max_bytes`
`kernel_tracker_max_interval`	`int32 kernel_tracker_max_interval`
`kernel_tracker_max_pending`	`int32 kernel_tracker_max_pending`
`num_dev_to_dev_copy_streams`	`int32 num_dev_to_dev_copy_streams`
`num_virtual_devices_per_gpu`	`int32 num_virtual_devices_per_gpu`
`timestamped_allocator`	`bool timestamped_allocator`
`use_cuda_malloc_async`	`bool use_cuda_malloc_async`
`use_unified_memory`	`bool use_unified_memory`
`virtual_devices`	`repeated VirtualDevices virtual_devices`

Child Classes