anko-intel commented on a change in pull request #20606:
URL: https://github.com/apache/incubator-mxnet/pull/20606#discussion_r721266661
##########
File path: python/mxnet/contrib/quantization.py
##########
@@ -527,13 +527,13 @@ def quantize_model(sym, arg_params, aux_params,
data_names=('data',),
return qsym, qarg_params, aux_params
-def quantize_model_mkldnn(sym, arg_params, aux_params, data_names=('data',),
- ctx=cpu(), excluded_sym_names=None,
excluded_op_names=None,
- calib_mode='entropy', calib_data=None,
num_calib_batches=None,
- quantized_dtype='int8', quantize_mode='smart',
- quantize_granularity='tensor-wise', logger=None):
+def quantize_model_dnnl(sym, arg_params, aux_params, data_names=('data',),
+ ctx=cpu(), excluded_sym_names=None,
excluded_op_names=None,
+ calib_mode='entropy', calib_data=None,
num_calib_batches=None,
+ quantized_dtype='int8', quantize_mode='smart',
+ quantize_granularity='tensor-wise', logger=None):
"""User-level API for generating a fusion + quantized model from a FP32
model
- w/ or w/o calibration with Intel MKL-DNN.
+ w/ or w/o calibration with Intel DNNL.
Review comment:
@PawelGlomski-Intel - I think only names in documentation/description
should be oneDNN (but I am not stick to it) and other names should be DNNL as
OneDNN internally also use DNNL in runtime environment names
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
