experiments/STT-tensorflow
1
0
Fork 0
Code Issues Pull Requests Projects Releases Wiki Activity

Consider FakeQuantWithMinMaxArgs to be a train-time quant op in the TFLite converter

Browse Source
This commit is contained in:
Adam Hillier 2021-01-26 18:23:46 +00:00 committed by GitHub
parent 2833b3d945
commit 9db9d08855
No known key found for this signature in database
GPG Key ID: 4AEE18F83AFDEB23
1 changed files with 1 additions and 0 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

1
tensorflow/lite/python/lite.py
View File

@ -411,6 +411,7 @@ class QuantizationMode(object):
"""Checks if the graph contains any training-time quantization ops."""
training_quant_ops = frozenset({
"FakeQuantWithMinMaxVars", "FakeQuantWithMinMaxVarsPerChannel",
"FakeQuantWithMinMaxArgs", "FakeQuantWithMinMaxArgsPerChannel",
"QuantizeAndDequantizeV2", "QuantizeAndDequantizeV3"
})