From: Yangqing Jia Date: Fri, 27 Sep 2013 18:53:12 +0000 (-0700) Subject: more cpplint X-Git-Tag: submit/tizen/20180823.020014~987 X-Git-Url: http://review.tizen.org/git/?a=commitdiff_plain;h=8de8006b46140c282c26700be22c3ea19f13d22c;p=platform%2Fupstream%2Fcaffeonacl.git more cpplint --- diff --git a/src/caffe/test/lenet.hpp b/src/caffe/test/lenet.hpp index 266f0b2..017463a 100644 --- a/src/caffe/test/lenet.hpp +++ b/src/caffe/test/lenet.hpp @@ -1,3 +1,9 @@ +// Copyright Yangqing Jia 2013 + +// This file is merely here so we can easily get a string of the lenet. +// It is actually not the very original LeNet, but with the sigmoid layers +// replaced by ReLU layers. + #ifndef CAFFE_TEST_LENET_HPP_ #define CAFFE_TEST_LENET_HPP_ @@ -5,119 +11,119 @@ namespace caffe { -const char* kLENET = "name: \"LeNet\"\n\ -bottom: \"data\"\n\ -bottom: \"label\"\n\ -layers {\n\ - layer {\n\ - name: \"conv1\"\n\ - type: \"conv\"\n\ - num_output: 20\n\ - kernelsize: 5\n\ - stride: 1\n\ - weight_filler {\n\ - type: \"xavier\"\n\ - }\n\ - bias_filler {\n\ - type: \"constant\"\n\ - }\n\ - }\n\ - bottom: \"data\"\n\ - top: \"conv1\"\n\ -}\n\ -layers {\n\ - layer {\n\ - name: \"pool1\"\n\ - type: \"pool\"\n\ - kernelsize: 2\n\ - stride: 2\n\ - pool: MAX\n\ - }\n\ - bottom: \"conv1\"\n\ - top: \"pool1\"\n\ -}\n\ -layers {\n\ - layer {\n\ - name: \"conv2\"\n\ - type: \"conv\"\n\ - num_output: 50\n\ - kernelsize: 5\n\ - stride: 1\n\ - weight_filler {\n\ - type: \"xavier\"\n\ - }\n\ - bias_filler {\n\ - type: \"constant\"\n\ - }\n\ - }\n\ - bottom: \"pool1\"\n\ - top: \"conv2\"\n\ -}\n\ -layers {\n\ - layer {\n\ - name: \"pool2\"\n\ - type: \"pool\"\n\ - kernelsize: 2\n\ - stride: 2\n\ - pool: MAX\n\ - }\n\ - bottom: \"conv2\"\n\ - top: \"pool2\"\n\ -}\n\ -layers {\n\ - layer {\n\ - name: \"ip1\"\n\ - type: \"innerproduct\"\n\ - num_output: 500\n\ - weight_filler {\n\ - type: \"xavier\"\n\ - }\n\ - bias_filler {\n\ - type: \"constant\"\n\ - }\n\ - }\n\ - bottom: \"pool2\"\n\ - top: \"ip1\"\n\ -}\n\ -layers {\n\ - layer {\n\ - name: \"relu1\"\n\ - type: \"relu\"\n\ - }\n\ - bottom: \"ip1\"\n\ - top: \"relu1\"\n\ -}\n\ -layers {\n\ - layer {\n\ - name: \"ip2\"\n\ - type: \"innerproduct\"\n\ - num_output: 10\n\ - weight_filler {\n\ - type: \"xavier\"\n\ - }\n\ - bias_filler {\n\ - type: \"constant\"\n\ - }\n\ - }\n\ - bottom: \"relu1\"\n\ - top: \"ip2\"\n\ -}\n\ -layers {\n\ - layer {\n\ - name: \"prob\"\n\ - type: \"softmax\"\n\ - }\n\ - bottom: \"ip2\"\n\ - top: \"prob\"\n\ -}\n\ -layers {\n\ - layer {\n\ - name: \"loss\"\n\ - type: \"multinomial_logistic_loss\"\n\ - }\n\ - bottom: \"prob\"\n\ - bottom: \"label\"\n\ -}"; +const char* kLENET = "name: \"LeNet\"\n" +"bottom: \"data\"\n" +"bottom: \"label\"\n" +"layers {\n" +" layer {\n" +" name: \"conv1\"\n" +" type: \"conv\"\n" +" num_output: 20\n" +" kernelsize: 5\n" +" stride: 1\n" +" weight_filler {\n" +" type: \"xavier\"\n" +" }\n" +" bias_filler {\n" +" type: \"constant\"\n" +" }\n" +" }\n" +" bottom: \"data\"\n" +" top: \"conv1\"\n" +"}\n" +"layers {\n" +" layer {\n" +" name: \"pool1\"\n" +" type: \"pool\"\n" +" kernelsize: 2\n" +" stride: 2\n" +" pool: MAX\n" +" }\n" +" bottom: \"conv1\"\n" +" top: \"pool1\"\n" +"}\n" +"layers {\n" +" layer {\n" +" name: \"conv2\"\n" +" type: \"conv\"\n" +" num_output: 50\n" +" kernelsize: 5\n" +" stride: 1\n" +" weight_filler {\n" +" type: \"xavier\"\n" +" }\n" +" bias_filler {\n" +" type: \"constant\"\n" +" }\n" +" }\n" +" bottom: \"pool1\"\n" +" top: \"conv2\"\n" +"}\n" +"layers {\n" +" layer {\n" +" name: \"pool2\"\n" +" type: \"pool\"\n" +" kernelsize: 2\n" +" stride: 2\n" +" pool: MAX\n" +" }\n" +" bottom: \"conv2\"\n" +" top: \"pool2\"\n" +"}\n" +"layers {\n" +" layer {\n" +" name: \"ip1\"\n" +" type: \"innerproduct\"\n" +" num_output: 500\n" +" weight_filler {\n" +" type: \"xavier\"\n" +" }\n" +" bias_filler {\n" +" type: \"constant\"\n" +" }\n" +" }\n" +" bottom: \"pool2\"\n" +" top: \"ip1\"\n" +"}\n" +"layers {\n" +" layer {\n" +" name: \"relu1\"\n" +" type: \"relu\"\n" +" }\n" +" bottom: \"ip1\"\n" +" top: \"relu1\"\n" +"}\n" +"layers {\n" +" layer {\n" +" name: \"ip2\"\n" +" type: \"innerproduct\"\n" +" num_output: 10\n" +" weight_filler {\n" +" type: \"xavier\"\n" +" }\n" +" bias_filler {\n" +" type: \"constant\"\n" +" }\n" +" }\n" +" bottom: \"relu1\"\n" +" top: \"ip2\"\n" +"}\n" +"layers {\n" +" layer {\n" +" name: \"prob\"\n" +" type: \"softmax\"\n" +" }\n" +" bottom: \"ip2\"\n" +" top: \"prob\"\n" +"}\n" +"layers {\n" +" layer {\n" +" name: \"loss\"\n" +" type: \"multinomial_logistic_loss\"\n" +" }\n" +" bottom: \"prob\"\n" +" bottom: \"label\"\n" +"}"; } // namespace caffe diff --git a/src/caffe/test/test_caffe_main.hpp b/src/caffe/test/test_caffe_main.hpp index 9ee11a3..a8c1657 100644 --- a/src/caffe/test/test_caffe_main.hpp +++ b/src/caffe/test/test_caffe_main.hpp @@ -5,13 +5,14 @@ #ifndef CAFFE_TEST_TEST_CAFFE_MAIN_HPP_ #define CAFFE_TEST_TEST_CAFFE_MAIN_HPP_ +#include +#include +#include + #include #include #include -#include -#include -#include namespace caffe { diff --git a/src/caffe/test/test_gradient_check_util.hpp b/src/caffe/test/test_gradient_check_util.hpp index 0c34861..c540549 100644 --- a/src/caffe/test/test_gradient_check_util.hpp +++ b/src/caffe/test/test_gradient_check_util.hpp @@ -3,10 +3,13 @@ #ifndef CAFFE_TEST_GRADIENT_CHECK_UTIL_H_ #define CAFFE_TEST_GRADIENT_CHECK_UTIL_H_ -#include -#include #include #include + +#include +#include +#include + #include "caffe/layer.hpp" using std::max; @@ -22,7 +25,7 @@ class GradientChecker { const unsigned int seed = 1701, const Dtype kink = 0., const Dtype kink_range = -1) : stepsize_(stepsize), threshold_(threshold), seed_(seed), - kink_(kink), kink_range_(kink_range) {}; + kink_(kink), kink_range_(kink_range) {} // Checks the gradient of a layer, with provided bottom layers and top // layers. The gradient checker will check the gradient with respect to // the parameters of the layer, as well as the input blobs if check_through @@ -41,6 +44,7 @@ class GradientChecker { void CheckGradientSingle(Layer& layer, vector*>& bottom, vector*>& top, int check_bottom, int top_id, int top_data_id); + protected: Dtype GetObjAndGradient(vector*>& top, int top_id = -1, int top_data_id = -1); @@ -73,11 +77,11 @@ void GradientChecker::CheckGradientSingle(Layer& layer, blobs_to_check.push_back(bottom[check_bottom]); } // go through the bottom and parameter blobs - //LOG(ERROR) << "Checking " << blobs_to_check.size() << " blobs."; + // LOG(ERROR) << "Checking " << blobs_to_check.size() << " blobs."; for (int blobid = 0; blobid < blobs_to_check.size(); ++blobid) { Blob* current_blob = blobs_to_check[blobid]; - //LOG(ERROR) << "Blob " << blobid << ": checking " << current_blob->count() - // << " parameters."; + // LOG(ERROR) << "Blob " << blobid << ": checking " << current_blob->count() + // << " parameters."; // go through the values for (int feat_id = 0; feat_id < current_blob->count(); ++feat_id) { // First, obtain the original data @@ -104,13 +108,13 @@ void GradientChecker::CheckGradientSingle(Layer& layer, Dtype estimated_gradient = (positive_objective - negative_objective) / stepsize_ / 2.; Dtype feature = current_blob->cpu_data()[feat_id]; - //LOG(ERROR) << "debug: " << current_blob->cpu_data()[feat_id] << " " - // << current_blob->cpu_diff()[feat_id]; + // LOG(ERROR) << "debug: " << current_blob->cpu_data()[feat_id] << " " + // << current_blob->cpu_diff()[feat_id]; if (kink_ - kink_range_ > feature || feature > kink_ + kink_range_) { // We check relative accuracy, but for too small values, we threshold // the scale factor by 1. - Dtype scale = max(max(fabs(computed_gradient), fabs(estimated_gradient)), - 1.); + Dtype scale = max( + max(fabs(computed_gradient), fabs(estimated_gradient)), 1.); EXPECT_GT(computed_gradient, estimated_gradient - threshold_ * scale) << "debug: (top_id, top_data_id, blob_id, feat_id)=" << top_id << "," << top_data_id << "," << blobid << "," << feat_id; @@ -118,22 +122,23 @@ void GradientChecker::CheckGradientSingle(Layer& layer, << "debug: (top_id, top_data_id, blob_id, feat_id)=" << top_id << "," << top_data_id << "," << blobid << "," << feat_id; } - //LOG(ERROR) << "Feature: " << current_blob->cpu_data()[feat_id]; - //LOG(ERROR) << "computed gradient: " << computed_gradient - // << " estimated_gradient: " << estimated_gradient; + // LOG(ERROR) << "Feature: " << current_blob->cpu_data()[feat_id]; + // LOG(ERROR) << "computed gradient: " << computed_gradient + // << " estimated_gradient: " << estimated_gradient; } } } template void GradientChecker::CheckGradientExhaustive(Layer& layer, - vector*>& bottom, vector*>& top, int check_bottom) { + vector*>& bottom, vector*>& top, + int check_bottom) { layer.SetUp(bottom, &top); - //LOG(ERROR) << "Exhaustive Mode."; + // LOG(ERROR) << "Exhaustive Mode."; for (int i = 0; i < top.size(); ++i) { - //LOG(ERROR) << "Exhaustive: blob " << i << " size " << top[i]->count(); + // LOG(ERROR) << "Exhaustive: blob " << i << " size " << top[i]->count(); for (int j = 0; j < top[i]->count(); ++j) { - //LOG(ERROR) << "Exhaustive: blob " << i << " data " << j; + // LOG(ERROR) << "Exhaustive: blob " << i << " data " << j; CheckGradientSingle(layer, bottom, top, check_bottom, i, j); } }