ROCm platform currently does not support the ability to do GPU stream level tracing / profiling.
This commit skip subtests (within python unit-tests) that this functionality. The "skip" is guarded by the call to "is_buil_with_rocm()", and hence these unit-tests will not be affected in any way when running with TF which was not built with ROCm support (i.e. `--config=rocm`)
In order to support tests running in eager mode we need to avoid
unnecessary use of Sessions in tests. This moves to remove some
of the uses of the `run` function in favor of `evaluate`.
PiperOrigin-RevId: 223009795
In order to support tests running in eager mode we need to avoid
unnecessary use of Sessions in tests. This moves to remove some
of the uses of the `run` function in favor of `evaluate`.
PiperOrigin-RevId: 222013881
1. Track the full allocation history of each tensor, visualized in timeline.
2. Better ProfileContext for tracing step selection.
3. Small bug fix.
PiperOrigin-RevId: 173988293
