diff options
| author |  Suharsh Sivakumar <suharshs@google.com> | 2017-04-27 20:50:36 -0800 |
|---|---|---|
| committer |  TensorFlower Gardener <gardener@tensorflow.org> | 2017-04-27 22:07:22 -0700 |
| commit | d84aacfb7c8b5e561af7414b220a37bb1209ad77 (patch) | |
| tree | 1678e1217db4a0765ee334fba37919c2a0714e97 /tensorflow/python/training/quantize_training.i | |
| parent | 141e8d652eb113319301fff6d30452b01903b771 (diff) | |
Allow FakeQuantWithMinMaxVars to be used with quantize_training rewriter.
Change: 154507631
Diffstat (limited to 'tensorflow/python/training/quantize_training.i')
| -rw-r--r-- | tensorflow/python/training/quantize_training.i | 4 |
1 files changed, 3 insertions, 1 deletions
diff --git a/tensorflow/python/training/quantize_training.i b/tensorflow/python/training/quantize_training.i index 59cc895084..40c6076973 100644 --- a/tensorflow/python/training/quantize_training.i +++ b/tensorflow/python/training/quantize_training.i @@ -24,8 +24,10 @@ static PyObject* DoQuantizeTrainingOnGraphDefHelper( int num_bits, TF_Status* out_status) { string result; + // TODO(suharshs): Make the QuantizeAndDequantizeV2 configurable. tensorflow::Status status = - tensorflow::DoQuantizeTrainingOnSerializedGraphDef(input_graph, num_bits, &result); + tensorflow::DoQuantizeTrainingOnSerializedGraphDef(input_graph, num_bits, + "QuantizeAndDequantizeV2", &result); if (!status.ok()) { Set_TF_Status_from_Status(out_status, status); Py_RETURN_NONE; |