diff options
author | 2018-03-26 14:33:10 -0700 | |
---|---|---|
committer | 2018-03-26 14:35:49 -0700 | |
commit | 2ff8e913ad000d379405c284857e7fc81eef9fed (patch) | |
tree | 49b43f92cc3e3a7977d72ef1f608b51e4b0e16d8 | |
parent | 72ed3c3b743e5feef99e37058dbd2f4344bcc5e3 (diff) |
Clarify eager gradient doc strings
PiperOrigin-RevId: 190526387
-rw-r--r-- | tensorflow/python/eager/backprop.py | 8 |
1 files changed, 4 insertions, 4 deletions
diff --git a/tensorflow/python/eager/backprop.py b/tensorflow/python/eager/backprop.py index a7837b8a7f..c54a5a1445 100644 --- a/tensorflow/python/eager/backprop.py +++ b/tensorflow/python/eager/backprop.py @@ -171,8 +171,8 @@ def implicit_val_and_grad(f): """Returns a function which differentiates f with respect to variables. The wrapped function returns the value and the gradient of f when called with - the same arguments. The gradient is with respect to all TFE variables which - are either trainable or have `variable.watch()` called on them by f. + the same arguments. The gradient is with respect to all trainable TFE + variables accessed by `f`. This function is useful when the exact set of variables to differentiate with is not known ahead of time. @@ -249,8 +249,8 @@ def implicit_grad(f): """Returns a function which differentiates f with respect to variables. The wrapped function returns the gradient of f when called with the same - arguments. The gradient is with respect to all TFE variables which are - either trainable or have `variable.watch()` called on them by f. + arguments. The gradient is with respect to all trainable TFE variables + accessed by `f`. This function is useful when the exact set of variables to differentiate with is not known ahead of time. |