Is XLA enabled in 19.03-py2 ?

AI & Data Science Deep Learning (Training & Inference) Frameworks
1

I see references to XLA in the output as follows :

2019-04-22 16:36:40.982977: I tensorflow/core/platform/profile_utils/cpu_utils.cc:94] CPU Frequency: 2200000000 Hz
2019-04-22 16:36:40.983551: I tensorflow/compiler/xla/service/service.cc:161] XLA service 0xb389cb0 executing computations on platform Host. Devices:
2019-04-22 16:36:40.983606: I tensorflow/compiler/xla/service/service.cc:168] StreamExecutor device (0): ,
2019-04-22 16:36:41.769539: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:998] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA nod$
zero
2019-04-22 16:36:41.770292: I tensorflow/compiler/xla/service/service.cc:161] XLA service 0xb449c60 executing computations on platform CUDA. Devices:
2019-04-22 16:36:41.770329: I tensorflow/compiler/xla/service/service.cc:168] StreamExecutor device (0): xxxxxx
2019-04-22 16:36:41.770972: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1433] Found device 0 with properties: xxxxxx

2

TensorFlow in the NGC containers is build with support for XLA. Whether XLA is actually used to execute a given model depends on the specific model scripts. For more information see XLA：优化机器学习编译器  |  TensorFlow