[tf.data] Increase the roll out percentage of optimization enable_gradient_descent
to 100%.
PiperOrigin-RevId: 340252424 Change-Id: Ic5a5d347c38fd94b68e10373c0b66ea3043f2e9b
This commit is contained in:
parent
a77e4aec3f
commit
aeafc6a567
@ -84,7 +84,7 @@ void OptimizeDatasetOp::MakeDataset(OpKernelContext* ctx, DatasetBase* input,
|
||||
// of the Borg jobs, the experiments will be randomly turned on.
|
||||
// clang-format off
|
||||
absl::flat_hash_map<string, uint64> live_experiments = {
|
||||
{"enable_gradient_descent", 50},
|
||||
{"enable_gradient_descent", 100},
|
||||
{"map_parallelization", 1}
|
||||
};
|
||||
// clang-format on
|
||||
|
Loading…
Reference in New Issue
Block a user