GPUOptions

 The type of GPU allocation strategy to use.
 Allowed values:
 "": The empty string (default) uses a system-chosen default
     which may change over time.
 "BFC": A "Best-fit with coalescing" algorithm, simplified from a
        version of dlmalloc.
 

 The type of GPU allocation strategy to use.
 Allowed values:
 "": The empty string (default) uses a system-chosen default
     which may change over time.
 "BFC": A "Best-fit with coalescing" algorithm, simplified from a
        version of dlmalloc.
 

 If true, the allocator does not pre-allocate the entire specified
 GPU memory region, instead starting small and growing as needed.
 

 Delay deletion of up to this many bytes to reduce the number of
 interactions with gpu driver code.  If 0, the system chooses
 a reasonable default (several MBs).
 

 Everything inside experimental is subject to change and is not subject
 to API stability guarantees in
 https://www.tensorflow.org/guide/version_compat.
 

 Everything inside experimental is subject to change and is not subject
 to API stability guarantees in
 https://www.tensorflow.org/guide/version_compat.
 

 Force all tensors to be gpu_compatible. On a GPU-enabled TensorFlow,
 enabling this option forces all CPU tensors to be allocated with Cuda
 pinned memory. Normally, TensorFlow will infer which tensors should be
 allocated as the pinned memory. But in case where the inference is
 incomplete, this option can significantly speed up the cross-device memory
 copy performance as long as it fits the memory.
 Note that this option is not something that should be
 enabled by default for unknown or very large models, since all Cuda pinned
 memory is unpageable, having too much pinned memory might negatively impact
 the overall host system performance.
 

 Fraction of the available GPU memory to allocate for each process.
 1 means to allocate all of the GPU memory, 0.5 means the process
 allocates up to ~50% of the available GPU memory.
 GPU memory is pre-allocated unless the allow_growth option is enabled.
 If greater than 1.0, uses CUDA unified memory to potentially oversubscribe
 the amount of memory available on the GPU device by using host memory as a
 swap space. Accessing memory not available on the device will be
 significantly slower as that would require memory transfer between the host
 and the device. Options to reduce the memory requirement should be
 considered before enabling this option as this may come with a negative
 performance impact. Oversubscription using the unified memory requires
 Pascal class or newer GPUs and it is currently only supported on the Linux
 operating system. See
 https://docs.nvidia.com/cuda/cuda-c-programming-guide/index.html#um-requirements
 for the detailed requirements.
 

 In the event polling loop sleep this many microseconds between
 PollEvents calls, when the queue is not empty.  If value is not
 set or set to 0, gets set to a non-zero default.
 

 This field is deprecated and ignored.
 

 A comma-separated list of GPU ids that determines the 'visible'
 to 'virtual' mapping of GPU devices.  For example, if TensorFlow
 can see 8 GPU devices in the process, and one wanted to map
 visible GPU devices 5 and 3 as "/device:GPU:0", and "/device:GPU:1",
 then one would specify this field as "5,3".  This field is similar in
 spirit to the CUDA_VISIBLE_DEVICES environment variable, except
 it applies to the visible GPU devices in the process.
 NOTE:
 1. The GPU driver provides the process with the visible GPUs
    in an order which is not guaranteed to have any correlation to
    the *physical* GPU id in the machine.  This field is used for
    remapping "visible" to "virtual", which means this operates only
    after the process starts.  Users are required to use vendor
    specific mechanisms (e.g., CUDA_VISIBLE_DEVICES) to control the
    physical to visible device mapping prior to invoking TensorFlow.
 2. In the code, the ids in this list are also called "platform GPU id"s,
    and the 'virtual' ids of GPU devices (i.e. the ids in the device
    name "/device:GPU:<id>") are also called "TF GPU id"s. Please
    refer to third_party/tensorflow/core/common_runtime/gpu/gpu_id.h
    for more information.
 

 A comma-separated list of GPU ids that determines the 'visible'
 to 'virtual' mapping of GPU devices.  For example, if TensorFlow
 can see 8 GPU devices in the process, and one wanted to map
 visible GPU devices 5 and 3 as "/device:GPU:0", and "/device:GPU:1",
 then one would specify this field as "5,3".  This field is similar in
 spirit to the CUDA_VISIBLE_DEVICES environment variable, except
 it applies to the visible GPU devices in the process.
 NOTE:
 1. The GPU driver provides the process with the visible GPUs
    in an order which is not guaranteed to have any correlation to
    the *physical* GPU id in the machine.  This field is used for
    remapping "visible" to "virtual", which means this operates only
    after the process starts.  Users are required to use vendor
    specific mechanisms (e.g., CUDA_VISIBLE_DEVICES) to control the
    physical to visible device mapping prior to invoking TensorFlow.
 2. In the code, the ids in this list are also called "platform GPU id"s,
    and the 'virtual' ids of GPU devices (i.e. the ids in the device
    name "/device:GPU:<id>") are also called "TF GPU id"s. Please
    refer to third_party/tensorflow/core/common_runtime/gpu/gpu_id.h
    for more information.
 

 Everything inside experimental is subject to change and is not subject
 to API stability guarantees in
 https://www.tensorflow.org/guide/version_compat.