diff options
author | Peter Hawkins <phawkins@google.com> | 2018-06-28 20:24:15 -0700 |
---|---|---|
committer | TensorFlower Gardener <gardener@tensorflow.org> | 2018-06-28 20:26:49 -0700 |
commit | 50839154899377f89367d851f6d1e2c45fcd6431 (patch) | |
tree | 871ebff61994a33d6ae94b5a7fa9e89156942d45 /third_party/mkl | |
parent | f04400f18f8e73664ba08a1a2f4d4e4f5bd49c79 (diff) |
[TF:XLA] Add partial implementation of tf.FIFOQueue for XLA devices (e.g., TPU).
The idea is to have a host-side queue of device tensors.
Operators dequeue_many, enqueue_many, and dequeue_up_to are not yet implemented because they require splitting/concatenating tensors, which will require calling into a compiled XLA compilation.
Refactor queue operator implementations into libraries separate from the kernel registrations.
Add support for ResourceOpKernels that are placed on non-CPU devices. Add support for allocating host-memory tensors during OpKernel construction.
PiperOrigin-RevId: 202590292
Diffstat (limited to 'third_party/mkl')
0 files changed, 0 insertions, 0 deletions