diff options
| author |  Jacques Pienaar <jpienaar@google.com> | 2018-03-15 12:58:08 -0700 |
|---|---|---|
| committer |  TensorFlower Gardener <gardener@tensorflow.org> | 2018-03-15 13:03:19 -0700 |
| commit | ccd8079e579604547f4b4d8a6b061cfdc6ce49bf (patch) | |
| tree | 0d498e84ca32a101afcada0993a30a5e3b0452a2 /tensorflow/contrib/lite/interpreter.h | |
| parent | 61032e9ca7bf9849cb65db9b646381d124080856 (diff) | |
Merge changes from github.
PiperOrigin-RevId: 189231636
Diffstat (limited to 'tensorflow/contrib/lite/interpreter.h')
| -rw-r--r-- | tensorflow/contrib/lite/interpreter.h | 2 |
1 files changed, 1 insertions, 1 deletions
diff --git a/tensorflow/contrib/lite/interpreter.h b/tensorflow/contrib/lite/interpreter.h index af143370ee..3749869f58 100644 --- a/tensorflow/contrib/lite/interpreter.h +++ b/tensorflow/contrib/lite/interpreter.h @@ -481,7 +481,7 @@ class Interpreter { // During Invoke(), Interpreter will allocate input tensors first, which are // known to be fixed size. Then it will allocate outputs from nodes as many // as possible. When there is a node that produces dynamic sized tensor. - // Intepreter will stop allocating tensors, set the value of next allocate + // Interpreter will stop allocating tensors, set the value of next allocate // node id, and execute the node to generate the output tensor before continue // to allocate successors. This process repeats until all nodes are executed. // NOTE: this relies on the order of nodes that is in topological order. |