apache · reminisce · Mar 6, 2019 · Feb 25, 2019 · Feb 28, 2019 · Feb 28, 2019
diff --git a/src/operator/quantization/dequantize.cc b/src/operator/quantization/dequantize.cc
@@ -71,6 +71,7 @@ by keep zero centered for the quantized value:
 .set_attr<nnvm::FInferShape>("FInferShape", DequantizeShape)
 .set_attr<nnvm::FInferType>("FInferType", DequantizeType)
 .set_attr<FInferStorageType>("FInferStorageType", DequantizeStorageType)
+.set_attr<nnvm::FGradient>("FGradient", MakeZeroGradNodes)
 #if MXNET_USE_MKLDNN == 1
 .set_attr<bool>("TIsMKLDNN", true)
 .set_attr<FComputeEx>("FComputeEx<cpu>", MKLDNNDequantizeCompute)

diff --git a/src/operator/quantization/quantize.cc b/src/operator/quantization/quantize.cc
@@ -82,6 +82,7 @@ where
 .set_attr<nnvm::FInferShape>("FInferShape", QuantizeShape)
 .set_attr<nnvm::FInferType>("FInferType", QuantizeType)
 .set_attr<FInferStorageType>("FInferStorageType", QuantizeStorageType)
+.set_attr<nnvm::FGradient>("FGradient", MakeZeroGradNodes)
 #if MXNET_USE_MKLDNN == 1
 .set_attr<bool>("TIsMKLDNN", true)
 .set_attr<FComputeEx>("FComputeEx<cpu>", MKLDNNQuantizeCompute)

diff --git a/src/operator/quantization/quantized_concat.cc b/src/operator/quantization/quantized_concat.cc
@@ -127,6 +127,7 @@ If any input holds int8, then the output will be int8. Otherwise output will be
 .set_attr<nnvm::FListOutputNames>("FListOutputNames", [](const NodeAttrs& attrs) {
   return std::vector<std::string>{"output", "min_output", "max_output"};
 })
+.set_attr<nnvm::FGradient>("FGradient", MakeZeroGradNodes)
 .set_attr<nnvm::FInferType>("FInferType", ConcatType)
 .set_attr<nnvm::FInferShape>("FInferShape", ConcatShape)
 .set_attr<std::string>("key_var_num_args", "num_args")

diff --git a/src/operator/quantization/quantized_conv.cc b/src/operator/quantization/quantized_conv.cc
@@ -160,6 +160,7 @@ and max thresholds representing the threholds for quantizing the float32 output
 .set_attr<nnvm::FInferShape>("FInferShape", QuantizedConvShape)
 .set_attr<nnvm::FInferType>("FInferType", QuantizedConvType)
 .set_attr<FInferStorageType>("FInferStorageType", QuantizedConvStorageType)
+.set_attr<nnvm::FGradient>("FGradient", MakeZeroGradNodes)
 .set_attr<FResourceRequest>("FResourceRequest",
   [](const NodeAttrs& attrs) {
     return std::vector<ResourceRequest>(1, ResourceRequest::kTempSpace);

diff --git a/src/operator/quantization/quantized_flatten.cc b/src/operator/quantization/quantized_flatten.cc
@@ -34,6 +34,7 @@ NNVM_REGISTER_OP(_contrib_quantized_flatten)
 .set_attr<nnvm::FInferShape>("FInferShape", QuantizedFlattenShape)
 .set_attr<nnvm::FInferType>("FInferType", QuantizedFlattenType)
 .set_attr<FCompute>("FCompute<cpu>", QuantizedFlattenCompute<cpu>)
+.set_attr<nnvm::FGradient>("FGradient", MakeZeroGradNodes)
 .set_attr<nnvm::FListInputNames>("FListInputNames",
   [](const NodeAttrs& attrs) {
     return std::vector<std::string>{"data", "min_data", "max_data"};

diff --git a/src/operator/quantization/quantized_fully_connected.cc b/src/operator/quantization/quantized_fully_connected.cc
@@ -264,6 +264,7 @@ and max thresholds representing the threholds for quantizing the float32 output
 .set_attr<nnvm::FInferShape>("FInferShape", QuantizedFullyConnectedShape)
 .set_attr<nnvm::FInferType>("FInferType", QuantizedFullyConnectedType)
 .set_attr<FInferStorageType>("FInferStorageType", QuantizedFullyConnectedStorageType)
+.set_attr<nnvm::FGradient>("FGradient", MakeZeroGradNodes)
 .set_attr<FNeedRequantize>("FNeedRequantize", [](const NodeAttrs& attrs) { return true; })
 .set_attr<FComputeEx>("FComputeEx<cpu>",
     QuantizedFullyConnectedForward<int8_t>)

diff --git a/src/operator/quantization/quantized_pooling.cc b/src/operator/quantization/quantized_pooling.cc
@@ -157,6 +157,7 @@ the float32 data into int8.
 .set_attr<nnvm::FInferShape>("FInferShape", QuantizedPoolingShape)
 .set_attr<nnvm::FInferType>("FInferType", QuantizedPoolingType)
 .set_attr<FInferStorageType>("FInferStorageType", QuantizedPoolingStorageType)
+.set_attr<nnvm::FGradient>("FGradient", MakeZeroGradNodes)
 .set_attr<FNeedRequantize>("FNeedRequantize",
   [](const NodeAttrs& attrs) {
     const PoolingParam& param = nnvm::get<PoolingParam>(attrs.parsed);

diff --git a/src/operator/quantization/requantize.cc b/src/operator/quantization/requantize.cc
@@ -64,6 +64,7 @@ inference accuracy.
 .set_attr<nnvm::FInferShape>("FInferShape", QuantizeShape)
 .set_attr<nnvm::FInferType>("FInferType", RequantizeType)
 .set_attr<FInferStorageType>("FInferStorageType", RequantizeStorageType)
+.set_attr<nnvm::FGradient>("FGradient", MakeZeroGradNodes)
 #if MXNET_USE_MKLDNN == 1
 .set_attr<bool>("TIsMKLDNN", true)
 .set_attr<FComputeEx>("FComputeEx<cpu>", MKLDNNRequantizeForward)

diff --git a/src/operator/subgraph/mkldnn/mkldnn_conv.cc b/src/operator/subgraph/mkldnn/mkldnn_conv.cc
@@ -689,6 +689,7 @@ NNVM_REGISTER_OP(_sg_mkldnn_conv)
 .set_attr<FInferStorageType>("FInferStorageType", SgMKLDNNConvOpStorageType)
 .set_attr<FStatefulComputeEx>("FStatefulComputeEx<cpu>", SgMKLDNNConvOpForward)
 .set_attr<bool>("TIsMKLDNN", true)
+.set_attr<nnvm::FGradient>("FGradient", MakeZeroGradNodes)
 .set_attr<FResourceRequest>("FResourceRequest", [](const NodeAttrs& n) {
   return std::vector<ResourceRequest>{ResourceRequest::kTempSpace};
 })