pjreddie / darknet

Convolutional Neural Networks
http://pjreddie.com/darknet/
Other
25.78k stars 21.33k forks source link

I have got an accident and reinstall the whole system, however, none of the model can convergence after training. #521

Closed hahakid closed 6 years ago

hahakid commented 6 years ago

I use cuda8.0, cudnn6.0 and opencv3.2 in ubuntu16.04, they are same as what I used before. After reinstall my environment, only the pre-trained models can work, and if I train them, none of the models can have a classify accuracy than 10% ( I also try the example of cifar in https://pjreddie.com/darknet/train-cifar/) and also do not work. Before the system crush, my model can achieve 98% on my own set, however, the model can reach no more than 1%. I try to use cuda only by set the opencv=0, and other configurations for Makefile, nothing works. The following is the make results, does anyone can give me any suggestions?

gcc -Iinclude/ -Isrc/ -DOPENCV pkg-config --cflags opencv -DGPU -I/usr/local/cuda-8.0/include/ -DCUDNN -Wall -Wno-unknown-pragmas -Wfatal-errors -fPIC -Ofast -DOPENCV -DGPU -DCUDNN -c ./src/gemm.c -o obj/gemm.o gcc -Iinclude/ -Isrc/ -DOPENCV pkg-config --cflags opencv -DGPU -I/usr/local/cuda-8.0/include/ -DCUDNN -Wall -Wno-unknown-pragmas -Wfatal-errors -fPIC -Ofast -DOPENCV -DGPU -DCUDNN -c ./src/utils.c -o obj/utils.o gcc -Iinclude/ -Isrc/ -DOPENCV pkg-config --cflags opencv -DGPU -I/usr/local/cuda-8.0/include/ -DCUDNN -Wall -Wno-unknown-pragmas -Wfatal-errors -fPIC -Ofast -DOPENCV -DGPU -DCUDNN -c ./src/cuda.c -o obj/cuda.o gcc -Iinclude/ -Isrc/ -DOPENCV pkg-config --cflags opencv -DGPU -I/usr/local/cuda-8.0/include/ -DCUDNN -Wall -Wno-unknown-pragmas -Wfatal-errors -fPIC -Ofast -DOPENCV -DGPU -DCUDNN -c ./src/deconvolutional_layer.c -o obj/deconvolutional_layer.o gcc -Iinclude/ -Isrc/ -DOPENCV pkg-config --cflags opencv -DGPU -I/usr/local/cuda-8.0/include/ -DCUDNN -Wall -Wno-unknown-pragmas -Wfatal-errors -fPIC -Ofast -DOPENCV -DGPU -DCUDNN -c ./src/convolutional_layer.c -o obj/convolutional_layer.o gcc -Iinclude/ -Isrc/ -DOPENCV pkg-config --cflags opencv -DGPU -I/usr/local/cuda-8.0/include/ -DCUDNN -Wall -Wno-unknown-pragmas -Wfatal-errors -fPIC -Ofast -DOPENCV -DGPU -DCUDNN -c ./src/list.c -o obj/list.o gcc -Iinclude/ -Isrc/ -DOPENCV pkg-config --cflags opencv -DGPU -I/usr/local/cuda-8.0/include/ -DCUDNN -Wall -Wno-unknown-pragmas -Wfatal-errors -fPIC -Ofast -DOPENCV -DGPU -DCUDNN -c ./src/image.c -o obj/image.o gcc -Iinclude/ -Isrc/ -DOPENCV pkg-config --cflags opencv -DGPU -I/usr/local/cuda-8.0/include/ -DCUDNN -Wall -Wno-unknown-pragmas -Wfatal-errors -fPIC -Ofast -DOPENCV -DGPU -DCUDNN -c ./src/activations.c -o obj/activations.o gcc -Iinclude/ -Isrc/ -DOPENCV pkg-config --cflags opencv -DGPU -I/usr/local/cuda-8.0/include/ -DCUDNN -Wall -Wno-unknown-pragmas -Wfatal-errors -fPIC -Ofast -DOPENCV -DGPU -DCUDNN -c ./src/im2col.c -o obj/im2col.o ./src/utils.c: In function ‘read_file’: ./src/utils.c:267:5: warning: ignoring return value of ‘fread’, declared with attribute warn_unused_result [-Wunused-result] fread(text, 1, size, fp); ^ ./src/utils.c: In function ‘fgetl’: ./src/utils.c:355:9: warning: ignoring return value of ‘fgets’, declared with attribute warn_unused_result [-Wunused-result] fgets(&line[curr], readsize, fp); ^ gcc -Iinclude/ -Isrc/ -DOPENCV pkg-config --cflags opencv -DGPU -I/usr/local/cuda-8.0/include/ -DCUDNN -Wall -Wno-unknown-pragmas -Wfatal-errors -fPIC -Ofast -DOPENCV -DGPU -DCUDNN -c ./src/col2im.c -o obj/col2im.o ./src/image.c: In function ‘load_image_cv’: ./src/image.c:605:9: warning: ignoring return value of ‘system’, declared with attribute warn_unused_result [-Wunused-result] system(buff); ^ gcc -Iinclude/ -Isrc/ -DOPENCV pkg-config --cflags opencv -DGPU -I/usr/local/cuda-8.0/include/ -DCUDNN -Wall -Wno-unknown-pragmas -Wfatal-errors -fPIC -Ofast -DOPENCV -DGPU -DCUDNN -c ./src/blas.c -o obj/blas.o gcc -Iinclude/ -Isrc/ -DOPENCV pkg-config --cflags opencv -DGPU -I/usr/local/cuda-8.0/include/ -DCUDNN -Wall -Wno-unknown-pragmas -Wfatal-errors -fPIC -Ofast -DOPENCV -DGPU -DCUDNN -c ./src/crop_layer.c -o obj/crop_layer.o gcc -Iinclude/ -Isrc/ -DOPENCV pkg-config --cflags opencv -DGPU -I/usr/local/cuda-8.0/include/ -DCUDNN -Wall -Wno-unknown-pragmas -Wfatal-errors -fPIC -Ofast -DOPENCV -DGPU -DCUDNN -c ./src/dropout_layer.c -o obj/dropout_layer.o gcc -Iinclude/ -Isrc/ -DOPENCV pkg-config --cflags opencv -DGPU -I/usr/local/cuda-8.0/include/ -DCUDNN -Wall -Wno-unknown-pragmas -Wfatal-errors -fPIC -Ofast -DOPENCV -DGPU -DCUDNN -c ./src/maxpool_layer.c -o obj/maxpool_layer.o gcc -Iinclude/ -Isrc/ -DOPENCV pkg-config --cflags opencv -DGPU -I/usr/local/cuda-8.0/include/ -DCUDNN -Wall -Wno-unknown-pragmas -Wfatal-errors -fPIC -Ofast -DOPENCV -DGPU -DCUDNN -c ./src/softmax_layer.c -o obj/softmax_layer.o gcc -Iinclude/ -Isrc/ -DOPENCV pkg-config --cflags opencv -DGPU -I/usr/local/cuda-8.0/include/ -DCUDNN -Wall -Wno-unknown-pragmas -Wfatal-errors -fPIC -Ofast -DOPENCV -DGPU -DCUDNN -c ./src/data.c -o obj/data.o gcc -Iinclude/ -Isrc/ -DOPENCV pkg-config --cflags opencv -DGPU -I/usr/local/cuda-8.0/include/ -DCUDNN -Wall -Wno-unknown-pragmas -Wfatal-errors -fPIC -Ofast -DOPENCV -DGPU -DCUDNN -c ./src/matrix.c -o obj/matrix.o gcc -Iinclude/ -Isrc/ -DOPENCV pkg-config --cflags opencv -DGPU -I/usr/local/cuda-8.0/include/ -DCUDNN -Wall -Wno-unknown-pragmas -Wfatal-errors -fPIC -Ofast -DOPENCV -DGPU -DCUDNN -c ./src/network.c -o obj/network.o gcc -Iinclude/ -Isrc/ -DOPENCV pkg-config --cflags opencv -DGPU -I/usr/local/cuda-8.0/include/ -DCUDNN -Wall -Wno-unknown-pragmas -Wfatal-errors -fPIC -Ofast -DOPENCV -DGPU -DCUDNN -c ./src/connected_layer.c -o obj/connected_layer.o gcc -Iinclude/ -Isrc/ -DOPENCV pkg-config --cflags opencv -DGPU -I/usr/local/cuda-8.0/include/ -DCUDNN -Wall -Wno-unknown-pragmas -Wfatal-errors -fPIC -Ofast -DOPENCV -DGPU -DCUDNN -c ./src/cost_layer.c -o obj/cost_layer.o gcc -Iinclude/ -Isrc/ -DOPENCV pkg-config --cflags opencv -DGPU -I/usr/local/cuda-8.0/include/ -DCUDNN -Wall -Wno-unknown-pragmas -Wfatal-errors -fPIC -Ofast -DOPENCV -DGPU -DCUDNN -c ./src/parser.c -o obj/parser.o gcc -Iinclude/ -Isrc/ -DOPENCV pkg-config --cflags opencv -DGPU -I/usr/local/cuda-8.0/include/ -DCUDNN -Wall -Wno-unknown-pragmas -Wfatal-errors -fPIC -Ofast -DOPENCV -DGPU -DCUDNN -c ./src/option_list.c -o obj/option_list.o ./src/data.c: In function ‘load_regression_labels_paths’: ./src/data.c:558:9: warning: ignoring return value of ‘fscanf’, declared with attribute warn_unused_result [-Wunused-result] fscanf(file, "%f", &(y.vals[i][0])); ^ ./src/data.c: In function ‘load_cifar10_data’: ./src/data.c:1341:9: warning: ignoring return value of ‘fread’, declared with attribute warn_unused_result [-Wunused-result] fread(bytes, 1, 3073, fp); ^ ./src/data.c: In function ‘load_all_cifar10’: ./src/data.c:1404:13: warning: ignoring return value of ‘fread’, declared with attribute warn_unused_result [-Wunused-result] fread(bytes, 1, 3073, fp); ^ ./src/parser.c: In function ‘load_connected_weights’: ./src/parser.c:978:5: warning: ignoring return value of ‘fread’, declared with attribute warn_unused_result [-Wunused-result] fread(l.biases, sizeof(float), l.outputs, fp); ^ ./src/parser.c:979:5: warning: ignoring return value of ‘fread’, declared with attribute warn_unused_result [-Wunused-result] fread(l.weights, sizeof(float), l.outputs*l.inputs, fp); ^ ./src/parser.c:986:9: warning: ignoring return value of ‘fread’, declared with attribute warn_unused_result [-Wunused-result] fread(l.scales, sizeof(float), l.outputs, fp); ^ ./src/parser.c:987:9: warning: ignoring return value of ‘fread’, declared with attribute warn_unused_result [-Wunused-result] fread(l.rolling_mean, sizeof(float), l.outputs, fp); ^ ./src/parser.c:988:9: warning: ignoring return value of ‘fread’, declared with attribute warn_unused_result [-Wunused-result] fread(l.rolling_variance, sizeof(float), l.outputs, fp); ^ gcc -Iinclude/ -Isrc/ -DOPENCV pkg-config --cflags opencv -DGPU -I/usr/local/cuda-8.0/include/ -DCUDNN -Wall -Wno-unknown-pragmas -Wfatal-errors -fPIC -Ofast -DOPENCV -DGPU -DCUDNN -c ./src/detection_layer.c -o obj/detection_layer.o ./src/parser.c: In function ‘load_batchnorm_weights’: ./src/parser.c:1002:5: warning: ignoring return value of ‘fread’, declared with attribute warn_unused_result [-Wunused-result] fread(l.scales, sizeof(float), l.c, fp); ^ ./src/parser.c:1003:5: warning: ignoring return value of ‘fread’, declared with attribute warn_unused_result [-Wunused-result] fread(l.rolling_mean, sizeof(float), l.c, fp); ^ ./src/parser.c:1004:5: warning: ignoring return value of ‘fread’, declared with attribute warn_unused_result [-Wunused-result] fread(l.rolling_variance, sizeof(float), l.c, fp); ^ ./src/parser.c: In function ‘load_convolutional_weights_binary’: ./src/parser.c:1014:5: warning: ignoring return value of ‘fread’, declared with attribute warn_unused_result [-Wunused-result] fread(l.biases, sizeof(float), l.n, fp); ^ ./src/parser.c:1016:9: warning: ignoring return value of ‘fread’, declared with attribute warn_unused_result [-Wunused-result] fread(l.scales, sizeof(float), l.n, fp); ^ ./src/parser.c:1017:9: warning: ignoring return value of ‘fread’, declared with attribute warn_unused_result [-Wunused-result] fread(l.rolling_mean, sizeof(float), l.n, fp); ^ ./src/parser.c:1018:9: warning: ignoring return value of ‘fread’, declared with attribute warn_unused_result [-Wunused-result] fread(l.rolling_variance, sizeof(float), l.n, fp); ^ ./src/parser.c:1024:9: warning: ignoring return value of ‘fread’, declared with attribute warn_unused_result [-Wunused-result] fread(&mean, sizeof(float), 1, fp); ^ ./src/parser.c:1028:13: warning: ignoring return value of ‘fread’, declared with attribute warn_unused_result [-Wunused-result] fread(&c, sizeof(char), 1, fp); ^ ./src/parser.c: In function ‘load_convolutional_weights’: ./src/parser.c:1049:5: warning: ignoring return value of ‘fread’, declared with attribute warn_unused_result [-Wunused-result] fread(l.biases, sizeof(float), l.n, fp); ^ ./src/parser.c:1051:9: warning: ignoring return value of ‘fread’, declared with attribute warn_unused_result [-Wunused-result] fread(l.scales, sizeof(float), l.n, fp); ^ ./src/parser.c:1052:9: warning: ignoring return value of ‘fread’, declared with attribute warn_unused_result [-Wunused-result] fread(l.rolling_mean, sizeof(float), l.n, fp); ^ ./src/parser.c:1053:9: warning: ignoring return value of ‘fread’, declared with attribute warn_unused_result [-Wunused-result] fread(l.rolling_variance, sizeof(float), l.n, fp); ^ ./src/parser.c:1081:5: warning: ignoring return value of ‘fread’, declared with attribute warn_unused_result [-Wunused-result] fread(l.weights, sizeof(float), num, fp); ^ ./src/parser.c: In function ‘load_weights_upto’: ./src/parser.c:1110:5: warning: ignoring return value of ‘fread’, declared with attribute warn_unused_result [-Wunused-result] fread(&major, sizeof(int), 1, fp); ^ ./src/parser.c:1111:5: warning: ignoring return value of ‘fread’, declared with attribute warn_unused_result [-Wunused-result] fread(&minor, sizeof(int), 1, fp); ^ ./src/parser.c:1112:5: warning: ignoring return value of ‘fread’, declared with attribute warn_unused_result [-Wunused-result] fread(&revision, sizeof(int), 1, fp); ^ ./src/parser.c:1114:9: warning: ignoring return value of ‘fread’, declared with attribute warn_unused_result [-Wunused-result] fread(net->seen, sizeof(size_t), 1, fp); ^ ./src/parser.c:1117:9: warning: ignoring return value of ‘fread’, declared with attribute warn_unused_result [-Wunused-result] fread(&iseen, sizeof(int), 1, fp); ^ ./src/parser.c:1172:13: warning: ignoring return value of ‘fread’, declared with attribute warn_unused_result [-Wunused-result] fread(l.biases, sizeof(float), l.outputs, fp); ^ ./src/parser.c:1173:13: warning: ignoring return value of ‘fread’, declared with attribute warn_unused_result [-Wunused-result] fread(l.weights, sizeof(float), size, fp); ^ gcc -Iinclude/ -Isrc/ -DOPENCV pkg-config --cflags opencv -DGPU -I/usr/local/cuda-8.0/include/ -DCUDNN -Wall -Wno-unknown-pragmas -Wfatal-errors -fPIC -Ofast -DOPENCV -DGPU -DCUDNN -c ./src/route_layer.c -o obj/route_layer.o gcc -Iinclude/ -Isrc/ -DOPENCV pkg-config --cflags opencv -DGPU -I/usr/local/cuda-8.0/include/ -DCUDNN -Wall -Wno-unknown-pragmas -Wfatal-errors -fPIC -Ofast -DOPENCV -DGPU -DCUDNN -c ./src/box.c -o obj/box.o gcc -Iinclude/ -Isrc/ -DOPENCV pkg-config --cflags opencv -DGPU -I/usr/local/cuda-8.0/include/ -DCUDNN -Wall -Wno-unknown-pragmas -Wfatal-errors -fPIC -Ofast -DOPENCV -DGPU -DCUDNN -c ./src/normalization_layer.c -o obj/normalization_layer.o gcc -Iinclude/ -Isrc/ -DOPENCV pkg-config --cflags opencv -DGPU -I/usr/local/cuda-8.0/include/ -DCUDNN -Wall -Wno-unknown-pragmas -Wfatal-errors -fPIC -Ofast -DOPENCV -DGPU -DCUDNN -c ./src/avgpool_layer.c -o obj/avgpool_layer.o gcc -Iinclude/ -Isrc/ -DOPENCV pkg-config --cflags opencv -DGPU -I/usr/local/cuda-8.0/include/ -DCUDNN -Wall -Wno-unknown-pragmas -Wfatal-errors -fPIC -Ofast -DOPENCV -DGPU -DCUDNN -c ./src/layer.c -o obj/layer.o gcc -Iinclude/ -Isrc/ -DOPENCV pkg-config --cflags opencv -DGPU -I/usr/local/cuda-8.0/include/ -DCUDNN -Wall -Wno-unknown-pragmas -Wfatal-errors -fPIC -Ofast -DOPENCV -DGPU -DCUDNN -c ./src/local_layer.c -o obj/local_layer.o gcc -Iinclude/ -Isrc/ -DOPENCV pkg-config --cflags opencv -DGPU -I/usr/local/cuda-8.0/include/ -DCUDNN -Wall -Wno-unknown-pragmas -Wfatal-errors -fPIC -Ofast -DOPENCV -DGPU -DCUDNN -c ./src/shortcut_layer.c -o obj/shortcut_layer.o gcc -Iinclude/ -Isrc/ -DOPENCV pkg-config --cflags opencv -DGPU -I/usr/local/cuda-8.0/include/ -DCUDNN -Wall -Wno-unknown-pragmas -Wfatal-errors -fPIC -Ofast -DOPENCV -DGPU -DCUDNN -c ./src/activation_layer.c -o obj/activation_layer.o gcc -Iinclude/ -Isrc/ -DOPENCV pkg-config --cflags opencv -DGPU -I/usr/local/cuda-8.0/include/ -DCUDNN -Wall -Wno-unknown-pragmas -Wfatal-errors -fPIC -Ofast -DOPENCV -DGPU -DCUDNN -c ./src/rnn_layer.c -o obj/rnn_layer.o gcc -Iinclude/ -Isrc/ -DOPENCV pkg-config --cflags opencv -DGPU -I/usr/local/cuda-8.0/include/ -DCUDNN -Wall -Wno-unknown-pragmas -Wfatal-errors -fPIC -Ofast -DOPENCV -DGPU -DCUDNN -c ./src/gru_layer.c -o obj/gru_layer.o gcc -Iinclude/ -Isrc/ -DOPENCV pkg-config --cflags opencv -DGPU -I/usr/local/cuda-8.0/include/ -DCUDNN -Wall -Wno-unknown-pragmas -Wfatal-errors -fPIC -Ofast -DOPENCV -DGPU -DCUDNN -c ./src/crnn_layer.c -o obj/crnn_layer.o gcc -Iinclude/ -Isrc/ -DOPENCV pkg-config --cflags opencv -DGPU -I/usr/local/cuda-8.0/include/ -DCUDNN -Wall -Wno-unknown-pragmas -Wfatal-errors -fPIC -Ofast -DOPENCV -DGPU -DCUDNN -c ./src/demo.c -o obj/demo.o gcc -Iinclude/ -Isrc/ -DOPENCV pkg-config --cflags opencv -DGPU -I/usr/local/cuda-8.0/include/ -DCUDNN -Wall -Wno-unknown-pragmas -Wfatal-errors -fPIC -Ofast -DOPENCV -DGPU -DCUDNN -c ./src/batchnorm_layer.c -o obj/batchnorm_layer.o gcc -Iinclude/ -Isrc/ -DOPENCV pkg-config --cflags opencv -DGPU -I/usr/local/cuda-8.0/include/ -DCUDNN -Wall -Wno-unknown-pragmas -Wfatal-errors -fPIC -Ofast -DOPENCV -DGPU -DCUDNN -c ./src/region_layer.c -o obj/region_layer.o gcc -Iinclude/ -Isrc/ -DOPENCV pkg-config --cflags opencv -DGPU -I/usr/local/cuda-8.0/include/ -DCUDNN -Wall -Wno-unknown-pragmas -Wfatal-errors -fPIC -Ofast -DOPENCV -DGPU -DCUDNN -c ./src/reorg_layer.c -o obj/reorg_layer.o gcc -Iinclude/ -Isrc/ -DOPENCV pkg-config --cflags opencv -DGPU -I/usr/local/cuda-8.0/include/ -DCUDNN -Wall -Wno-unknown-pragmas -Wfatal-errors -fPIC -Ofast -DOPENCV -DGPU -DCUDNN -c ./src/tree.c -o obj/tree.o gcc -Iinclude/ -Isrc/ -DOPENCV pkg-config --cflags opencv -DGPU -I/usr/local/cuda-8.0/include/ -DCUDNN -Wall -Wno-unknown-pragmas -Wfatal-errors -fPIC -Ofast -DOPENCV -DGPU -DCUDNN -c ./src/lstm_layer.c -o obj/lstm_layer.o /usr/local/cuda-8.0/bin/nvcc -gencode arch=compute_30,code=sm_30 -gencode arch=compute_35,code=sm_35 -gencode arch=compute_50,code=[sm_50,compute_50] -gencode arch=compute_52,code=[sm_52,compute_52] -Iinclude/ -Isrc/ -DOPENCV pkg-config --cflags opencv -DGPU -I/usr/local/cuda-8.0/include/ -DCUDNN --compiler-options "-Wall -Wno-unknown-pragmas -Wfatal-errors -fPIC -Ofast -DOPENCV -DGPU -DCUDNN" -c ./src/convolutional_kernels.cu -o obj/convolutional_kernels.o /usr/local/cuda-8.0/bin/nvcc -gencode arch=compute_30,code=sm_30 -gencode arch=compute_35,code=sm_35 -gencode arch=compute_50,code=[sm_50,compute_50] -gencode arch=compute_52,code=[sm_52,compute_52] -Iinclude/ -Isrc/ -DOPENCV pkg-config --cflags opencv -DGPU -I/usr/local/cuda-8.0/include/ -DCUDNN --compiler-options "-Wall -Wno-unknown-pragmas -Wfatal-errors -fPIC -Ofast -DOPENCV -DGPU -DCUDNN" -c ./src/deconvolutional_kernels.cu -o obj/deconvolutional_kernels.o /usr/local/cuda-8.0/bin/nvcc -gencode arch=compute_30,code=sm_30 -gencode arch=compute_35,code=sm_35 -gencode arch=compute_50,code=[sm_50,compute_50] -gencode arch=compute_52,code=[sm_52,compute_52] -Iinclude/ -Isrc/ -DOPENCV pkg-config --cflags opencv -DGPU -I/usr/local/cuda-8.0/include/ -DCUDNN --compiler-options "-Wall -Wno-unknown-pragmas -Wfatal-errors -fPIC -Ofast -DOPENCV -DGPU -DCUDNN" -c ./src/activation_kernels.cu -o obj/activation_kernels.o /usr/local/cuda-8.0/bin/nvcc -gencode arch=compute_30,code=sm_30 -gencode arch=compute_35,code=sm_35 -gencode arch=compute_50,code=[sm_50,compute_50] -gencode arch=compute_52,code=[sm_52,compute_52] -Iinclude/ -Isrc/ -DOPENCV pkg-config --cflags opencv -DGPU -I/usr/local/cuda-8.0/include/ -DCUDNN --compiler-options "-Wall -Wno-unknown-pragmas -Wfatal-errors -fPIC -Ofast -DOPENCV -DGPU -DCUDNN" -c ./src/im2col_kernels.cu -o obj/im2col_kernels.o /usr/local/cuda-8.0/bin/nvcc -gencode arch=compute_30,code=sm_30 -gencode arch=compute_35,code=sm_35 -gencode arch=compute_50,code=[sm_50,compute_50] -gencode arch=compute_52,code=[sm_52,compute_52] -Iinclude/ -Isrc/ -DOPENCV pkg-config --cflags opencv -DGPU -I/usr/local/cuda-8.0/include/ -DCUDNN --compiler-options "-Wall -Wno-unknown-pragmas -Wfatal-errors -fPIC -Ofast -DOPENCV -DGPU -DCUDNN" -c ./src/col2im_kernels.cu -o obj/col2im_kernels.o /usr/local/cuda-8.0/bin/nvcc -gencode arch=compute_30,code=sm_30 -gencode arch=compute_35,code=sm_35 -gencode arch=compute_50,code=[sm_50,compute_50] -gencode arch=compute_52,code=[sm_52,compute_52] -Iinclude/ -Isrc/ -DOPENCV pkg-config --cflags opencv -DGPU -I/usr/local/cuda-8.0/include/ -DCUDNN --compiler-options "-Wall -Wno-unknown-pragmas -Wfatal-errors -fPIC -Ofast -DOPENCV -DGPU -DCUDNN" -c ./src/blas_kernels.cu -o obj/blas_kernels.o /usr/local/cuda-8.0/bin/nvcc -gencode arch=compute_30,code=sm_30 -gencode arch=compute_35,code=sm_35 -gencode arch=compute_50,code=[sm_50,compute_50] -gencode arch=compute_52,code=[sm_52,compute_52] -Iinclude/ -Isrc/ -DOPENCV pkg-config --cflags opencv -DGPU -I/usr/local/cuda-8.0/include/ -DCUDNN --compiler-options "-Wall -Wno-unknown-pragmas -Wfatal-errors -fPIC -Ofast -DOPENCV -DGPU -DCUDNN" -c ./src/crop_layer_kernels.cu -o obj/crop_layer_kernels.o /usr/local/cuda-8.0/bin/nvcc -gencode arch=compute_30,code=sm_30 -gencode arch=compute_35,code=sm_35 -gencode arch=compute_50,code=[sm_50,compute_50] -gencode arch=compute_52,code=[sm_52,compute_52] -Iinclude/ -Isrc/ -DOPENCV pkg-config --cflags opencv -DGPU -I/usr/local/cuda-8.0/include/ -DCUDNN --compiler-options "-Wall -Wno-unknown-pragmas -Wfatal-errors -fPIC -Ofast -DOPENCV -DGPU -DCUDNN" -c ./src/dropout_layer_kernels.cu -o obj/dropout_layer_kernels.o /usr/local/cuda-8.0/bin/nvcc -gencode arch=compute_30,code=sm_30 -gencode arch=compute_35,code=sm_35 -gencode arch=compute_50,code=[sm_50,compute_50] -gencode arch=compute_52,code=[sm_52,compute_52] -Iinclude/ -Isrc/ -DOPENCV pkg-config --cflags opencv -DGPU -I/usr/local/cuda-8.0/include/ -DCUDNN --compiler-options "-Wall -Wno-unknown-pragmas -Wfatal-errors -fPIC -Ofast -DOPENCV -DGPU -DCUDNN" -c ./src/maxpool_layer_kernels.cu -o obj/maxpool_layer_kernels.o /usr/local/cuda-8.0/bin/nvcc -gencode arch=compute_30,code=sm_30 -gencode arch=compute_35,code=sm_35 -gencode arch=compute_50,code=[sm_50,compute_50] -gencode arch=compute_52,code=[sm_52,compute_52] -Iinclude/ -Isrc/ -DOPENCV pkg-config --cflags opencv -DGPU -I/usr/local/cuda-8.0/include/ -DCUDNN --compiler-options "-Wall -Wno-unknown-pragmas -Wfatal-errors -fPIC -Ofast -DOPENCV -DGPU -DCUDNN" -c ./src/avgpool_layer_kernels.cu -o obj/avgpool_layer_kernels.o gcc -Iinclude/ -Isrc/ -DOPENCV pkg-config --cflags opencv -DGPU -I/usr/local/cuda-8.0/include/ -DCUDNN -Wall -Wno-unknown-pragmas -Wfatal-errors -fPIC -Ofast -DOPENCV -DGPU -DCUDNN -c ./examples/captcha.c -o obj/captcha.o gcc -Iinclude/ -Isrc/ -DOPENCV pkg-config --cflags opencv -DGPU -I/usr/local/cuda-8.0/include/ -DCUDNN -Wall -Wno-unknown-pragmas -Wfatal-errors -fPIC -Ofast -DOPENCV -DGPU -DCUDNN -c ./examples/lsd.c -o obj/lsd.o gcc -Iinclude/ -Isrc/ -DOPENCV pkg-config --cflags opencv -DGPU -I/usr/local/cuda-8.0/include/ -DCUDNN -Wall -Wno-unknown-pragmas -Wfatal-errors -fPIC -Ofast -DOPENCV -DGPU -DCUDNN -c ./examples/super.c -o obj/super.o gcc -Iinclude/ -Isrc/ -DOPENCV pkg-config --cflags opencv -DGPU -I/usr/local/cuda-8.0/include/ -DCUDNN -Wall -Wno-unknown-pragmas -Wfatal-errors -fPIC -Ofast -DOPENCV -DGPU -DCUDNN -c ./examples/art.c -o obj/art.o gcc -Iinclude/ -Isrc/ -DOPENCV pkg-config --cflags opencv -DGPU -I/usr/local/cuda-8.0/include/ -DCUDNN -Wall -Wno-unknown-pragmas -Wfatal-errors -fPIC -Ofast -DOPENCV -DGPU -DCUDNN -c ./examples/tag.c -o obj/tag.o gcc -Iinclude/ -Isrc/ -DOPENCV pkg-config --cflags opencv -DGPU -I/usr/local/cuda-8.0/include/ -DCUDNN -Wall -Wno-unknown-pragmas -Wfatal-errors -fPIC -Ofast -DOPENCV -DGPU -DCUDNN -c ./examples/cifar.c -o obj/cifar.o gcc -Iinclude/ -Isrc/ -DOPENCV pkg-config --cflags opencv -DGPU -I/usr/local/cuda-8.0/include/ -DCUDNN -Wall -Wno-unknown-pragmas -Wfatal-errors -fPIC -Ofast -DOPENCV -DGPU -DCUDNN -c ./examples/go.c -o obj/go.o gcc -Iinclude/ -Isrc/ -DOPENCV pkg-config --cflags opencv -DGPU -I/usr/local/cuda-8.0/include/ -DCUDNN -Wall -Wno-unknown-pragmas -Wfatal-errors -fPIC -Ofast -DOPENCV -DGPU -DCUDNN -c ./examples/rnn.c -o obj/rnn.o gcc -Iinclude/ -Isrc/ -DOPENCV pkg-config --cflags opencv -DGPU -I/usr/local/cuda-8.0/include/ -DCUDNN -Wall -Wno-unknown-pragmas -Wfatal-errors -fPIC -Ofast -DOPENCV -DGPU -DCUDNN -c ./examples/segmenter.c -o obj/segmenter.o gcc -Iinclude/ -Isrc/ -DOPENCV pkg-config --cflags opencv -DGPU -I/usr/local/cuda-8.0/include/ -DCUDNN -Wall -Wno-unknown-pragmas -Wfatal-errors -fPIC -Ofast -DOPENCV -DGPU -DCUDNN -c ./examples/regressor.c -o obj/regressor.o ./examples/rnn.c: In function ‘get_seq2seq_data’: ./examples/rnn.c:104:13: warning: unused variable ‘dlen’ [-Wunused-variable] int dlen = strlen(dest[index]); ^ ./examples/rnn.c:103:13: warning: unused variable ‘slen’ [-Wunused-variable] int slen = strlen(source[index]); ^ ./examples/go.c: In function ‘engine_go’: ./examples/go.c:844:9: warning: ignoring return value of ‘scanf’, declared with attribute warn_unused_result [-Wunused-result] scanf("%s", buff); ^ ./examples/go.c:861:13: warning: ignoring return value of ‘scanf’, declared with attribute warn_unused_result [-Wunused-result] scanf("%d %d %d", &main_time, &byo_yomi_time, &byo_yomi_stones); ^ ./examples/go.c:867:13: warning: ignoring return value of ‘scanf’, declared with attribute warn_unused_result [-Wunused-result] scanf("%s %d %d", color, &time, &stones); ^ ./examples/go.c:884:13: warning: ignoring return value of ‘scanf’, declared with attribute warn_unused_result [-Wunused-result] scanf("%s", comm); ^ ./examples/go.c:908:13: warning: ignoring return value of ‘scanf’, declared with attribute warn_unused_result [-Wunused-result] scanf("%d", &boardsize); ^ ./examples/go.c:921:13: warning: ignoring return value of ‘scanf’, declared with attribute warn_unused_result [-Wunused-result] scanf("%d", &handicap); ^ ./examples/go.c:938:13: warning: ignoring return value of ‘scanf’, declared with attribute warn_unused_result [-Wunused-result] scanf("%f", &komi); ^ ./examples/go.c:949:17: warning: ignoring return value of ‘scanf’, declared with attribute warn_unused_result [-Wunused-result] scanf("%s ", color); ^ ./examples/go.c:951:17: warning: ignoring return value of ‘scanf’, declared with attribute warn_unused_result [-Wunused-result] scanf(" "); ^ ./examples/go.c:995:17: warning: ignoring return value of ‘scanf’, declared with attribute warn_unused_result [-Wunused-result] scanf("%s", color); ^ ./examples/go.c:1043:13: warning: ignoring return value of ‘scanf’, declared with attribute warn_unused_result [-Wunused-result] scanf("%s", type); ^ ./examples/go.c:1068:13: warning: ignoring return value of ‘scanf’, declared with attribute warn_unused_result [-Wunused-result] scanf("%s", type); ^ gcc -Iinclude/ -Isrc/ -DOPENCV pkg-config --cflags opencv -DGPU -I/usr/local/cuda-8.0/include/ -DCUDNN -Wall -Wno-unknown-pragmas -Wfatal-errors -fPIC -Ofast -DOPENCV -DGPU -DCUDNN -c ./examples/classifier.c -o obj/classifier.o gcc -Iinclude/ -Isrc/ -DOPENCV pkg-config --cflags opencv -DGPU -I/usr/local/cuda-8.0/include/ -DCUDNN -Wall -Wno-unknown-pragmas -Wfatal-errors -fPIC -Ofast -DOPENCV -DGPU -DCUDNN -c ./examples/coco.c -o obj/coco.o gcc -Iinclude/ -Isrc/ -DOPENCV pkg-config --cflags opencv -DGPU -I/usr/local/cuda-8.0/include/ -DCUDNN -Wall -Wno-unknown-pragmas -Wfatal-errors -fPIC -Ofast -DOPENCV -DGPU -DCUDNN -c ./examples/yolo.c -o obj/yolo.o gcc -Iinclude/ -Isrc/ -DOPENCV pkg-config --cflags opencv -DGPU -I/usr/local/cuda-8.0/include/ -DCUDNN -Wall -Wno-unknown-pragmas -Wfatal-errors -fPIC -Ofast -DOPENCV -DGPU -DCUDNN -c ./examples/detector.c -o obj/detector.o gcc -Iinclude/ -Isrc/ -DOPENCV pkg-config --cflags opencv -DGPU -I/usr/local/cuda-8.0/include/ -DCUDNN -Wall -Wno-unknown-pragmas -Wfatal-errors -fPIC -Ofast -DOPENCV -DGPU -DCUDNN -c ./examples/nightmare.c -o obj/nightmare.o gcc -Iinclude/ -Isrc/ -DOPENCV pkg-config --cflags opencv -DGPU -I/usr/local/cuda-8.0/include/ -DCUDNN -Wall -Wno-unknown-pragmas -Wfatal-errors -fPIC -Ofast -DOPENCV -DGPU -DCUDNN -c ./examples/attention.c -o obj/attention.o ./examples/attention.c: In function ‘network_loss_data’: ./examples/attention.c:43:17: warning: unused variable ‘t’ [-Wunused-variable] int t = max_index(y + btest.y.cols, 1000); ^ ./examples/attention.c: In function ‘train_attention’: ./examples/attention.c:179:15: warning: unused variable ‘im’ [-Wunused-variable] image im = float_to_image(64,64,3,resized.X.vals[0]); ^ ./examples/attention.c: In function ‘run_attention’: ./examples/attention.c:452:11: warning: unused variable ‘layer_s’ [-Wunused-variable] char layer_s = (argc > 7) ? argv[7]: 0; ^ gcc -Iinclude/ -Isrc/ -DOPENCV pkg-config --cflags opencv -DGPU -I/usr/local/cuda-8.0/include/ -DCUDNN -Wall -Wno-unknown-pragmas -Wfatal-errors -fPIC -Ofast -DOPENCV -DGPU -DCUDNN -c ./examples/darknet.c -o obj/darknet.o gcc -Wall -Wno-unknown-pragmas -Wfatal-errors -fPIC -Ofast -DOPENCV -DGPU -DCUDNN -shared obj/gemm.o obj/utils.o obj/cuda.o obj/deconvolutional_layer.o obj/convolutional_layer.o obj/list.o obj/image.o obj/activations.o obj/im2col.o obj/col2im.o obj/blas.o obj/crop_layer.o obj/dropout_layer.o obj/maxpool_layer.o obj/softmax_layer.o obj/data.o obj/matrix.o obj/network.o obj/connected_layer.o obj/cost_layer.o obj/parser.o obj/option_list.o obj/detection_layer.o obj/route_layer.o obj/box.o obj/normalization_layer.o obj/avgpool_layer.o obj/layer.o obj/local_layer.o obj/shortcut_layer.o obj/activation_layer.o obj/rnn_layer.o obj/gru_layer.o obj/crnn_layer.o obj/demo.o obj/batchnorm_layer.o obj/region_layer.o obj/reorg_layer.o obj/tree.o obj/lstm_layer.o obj/convolutional_kernels.o obj/deconvolutional_kernels.o obj/activation_kernels.o obj/im2col_kernels.o obj/col2im_kernels.o obj/blas_kernels.o obj/crop_layer_kernels.o obj/dropout_layer_kernels.o obj/maxpool_layer_kernels.o obj/avgpool_layer_kernels.o -o libdarknet.so -lm -pthread pkg-config --libs opencv -L/usr/local/cuda-8.0/lib64 -lcuda -lcudart -lcublas -lcurand -lcudnn -lstdc++ ar rcs libdarknet.a obj/gemm.o obj/utils.o obj/cuda.o obj/deconvolutional_layer.o obj/convolutional_layer.o obj/list.o obj/image.o obj/activations.o obj/im2col.o obj/col2im.o obj/blas.o obj/crop_layer.o obj/dropout_layer.o obj/maxpool_layer.o obj/softmax_layer.o obj/data.o obj/matrix.o obj/network.o obj/connected_layer.o obj/cost_layer.o obj/parser.o obj/option_list.o obj/detection_layer.o obj/route_layer.o obj/box.o obj/normalization_layer.o obj/avgpool_layer.o obj/layer.o obj/local_layer.o obj/shortcut_layer.o obj/activation_layer.o obj/rnn_layer.o obj/gru_layer.o obj/crnn_layer.o obj/demo.o obj/batchnorm_layer.o obj/region_layer.o obj/reorg_layer.o obj/tree.o obj/lstm_layer.o obj/convolutional_kernels.o obj/deconvolutional_kernels.o obj/activation_kernels.o obj/im2col_kernels.o obj/col2im_kernels.o obj/blas_kernels.o obj/crop_layer_kernels.o obj/dropout_layer_kernels.o obj/maxpool_layer_kernels.o obj/avgpool_layer_kernels.o gcc -Iinclude/ -Isrc/ -DOPENCV pkg-config --cflags opencv -DGPU -I/usr/local/cuda-8.0/include/ -DCUDNN -Wall -Wno-unknown-pragmas -Wfatal-errors -fPIC -Ofast -DOPENCV -DGPU -DCUDNN obj/captcha.o obj/lsd.o obj/super.o obj/art.o obj/tag.o obj/cifar.o obj/go.o obj/rnn.o obj/segmenter.o obj/regressor.o obj/classifier.o obj/coco.o obj/yolo.o obj/detector.o obj/nightmare.o obj/attention.o obj/darknet.o libdarknet.a -o darknet -lm -pthread pkg-config --libs opencv -L/usr/local/cuda-8.0/lib64 -lcuda -lcudart -lcublas -lcurand -lcudnn -lstdc++ libdarknet.a

hahakid commented 6 years ago

I found the training process is all right, by copying the weights file to another computer, the valid is all right. For classifier predict on cifar, a very strange phenomenon happened: Loading weights from backup/cifar_11.weights...Done! data/cifar/test/2_ship.png: Predicted in 0.001040 seconds. 51.06%: airplane 41.01%: bird 3.68%: horse 1.29%: cat 0.66%: truck 0.50%: ship 0.49%: automobile 0.46%: deer 0.43%: frog 0.41%: dog

Loading weights from backup/cifar_11.weights...Done! data/cifar/test/23_truck.png: Predicted in 0.001037 seconds. 51.06%: airplane 41.01%: bird 3.68%: horse 1.29%: cat 0.66%: truck 0.50%: ship 0.49%: automobile 0.46%: deer 0.43%: frog 0.41%: dog

Two different images show the same result!!! That means, when using valid mode, all file may show same result and lead to very low accuracy. But why?

skazimax commented 6 years ago

hi, I tried to reproduce your issue following the guide https://pjreddie.com/darknet/train-cifar/ but seems that it's it's a little bit outdated. It's looking for names.list (we have labels.txt), also I wonder why we train here as classifier train while in main guide it's a detector train. However even after fixing config files(also changed valid to test), I don't have any results(even that you mention above). Can you share what you run to train and detect, to check if I running it correctly. Thanks!

hahakid commented 6 years ago

@skazimax @skazimax firstly, the classifier only give a result of a image, in contrast, the detector will told you the class and the bounding box of the target in the image. During the training of a detector, the bounding box info is also needed, and the Cifar data do not contain is kind of information, it is a classification dataset.

Here, I only considered the classifier in darknet. For several test, I am sure the training is all right, I copy the trained model to another machine and it works very well and can reach Top1 accuracy more than 80%. However, on my machine, it works as follows:

./darknet classifier valid cfg/cifar.data cfg/cifar.cfg backup/cifar_11.weights layer filters size input output 0 conv 128 3 x 3 / 1 28 x 28 x 3 -> 28 x 28 x 128 1 conv 128 3 x 3 / 1 28 x 28 x 128 -> 28 x 28 x 128 2 conv 128 3 x 3 / 1 28 x 28 x 128 -> 28 x 28 x 128 3 max 2 x 2 / 2 28 x 28 x 128 -> 14 x 14 x 128 4 dropout p = 0.50 25088 -> 25088 5 conv 256 3 x 3 / 1 14 x 14 x 128 -> 14 x 14 x 256 6 conv 256 3 x 3 / 1 14 x 14 x 256 -> 14 x 14 x 256 7 conv 256 3 x 3 / 1 14 x 14 x 256 -> 14 x 14 x 256 8 max 2 x 2 / 2 14 x 14 x 256 -> 7 x 7 x 256 9 dropout p = 0.50 12544 -> 12544 10 conv 512 3 x 3 / 1 7 x 7 x 256 -> 7 x 7 x 512 11 conv 512 3 x 3 / 1 7 x 7 x 512 -> 7 x 7 x 512 12 conv 512 3 x 3 / 1 7 x 7 x 512 -> 7 x 7 x 512 13 dropout p = 0.50 25088 -> 25088 14 conv 10 1 x 1 / 1 7 x 7 x 512 -> 7 x 7 x 10 15 avg 7 x 7 x 10 -> 10 16 softmax 10 17 type: Using default 'sse' cost 10 Loading weights from backup/cifar_11.weights...Done! 0: top 1: 0.000000, top 10: 1.000000 1: top 1: 0.000000, top 10: 1.000000 2: top 1: 0.333333, top 10: 1.000000 3: top 1: 0.250000, top 10: 1.000000 4: top 1: 0.200000, top 10: 1.000000 5: top 1: 0.166667, top 10: 1.000000 6: top 1: 0.142857, top 10: 1.000000 7: top 1: 0.125000, top 10: 1.000000 8: top 1: 0.111111, top 10: 1.000000 9: top 1: 0.100000, top 10: 1.000000 10: top 1: 0.090909, top 10: 1.000000 11: top 1: 0.083333, top 10: 1.000000 12: top 1: 0.153846, top 10: 1.000000 13: top 1: 0.142857, top 10: 1.000000 14: top 1: 0.133333, top 10: 1.000000 15: top 1: 0.125000, top 10: 1.000000 16: top 1: 0.117647, top 10: 1.000000 17: top 1: 0.111111, top 10: 1.000000 18: top 1: 0.105263, top 10: 1.000000 19: top 1: 0.100000, top 10: 1.000000 20: top 1: 0.142857, top 10: 1.000000 21: top 1: 0.136364, top 10: 1.000000 22: top 1: 0.130435, top 10: 1.000000 23: top 1: 0.125000, top 10: 1.000000 24: top 1: 0.120000, top 10: 1.000000 25: top 1: 0.153846, top 10: 1.000000 26: top 1: 0.185185, top 10: 1.000000 27: top 1: 0.178571, top 10: 1.000000 28: top 1: 0.172414, top 10: 1.000000 29: top 1: 0.166667, top 10: 1.000000 30: top 1: 0.161290, top 10: 1.000000 31: top 1: 0.187500, top 10: 1.000000 32: top 1: 0.181818, top 10: 1.000000 33: top 1: 0.176471, top 10: 1.000000 34: top 1: 0.171429, top 10: 1.000000 35: top 1: 0.166667, top 10: 1.000000 36: top 1: 0.162162, top 10: 1.000000 37: top 1: 0.157895, top 10: 1.000000 38: top 1: 0.153846, top 10: 1.000000 39: top 1: 0.150000, top 10: 1.000000 40: top 1: 0.146341, top 10: 1.000000 41: top 1: 0.142857, top 10: 1.000000 42: top 1: 0.139535, top 10: 1.000000 43: top 1: 0.136364, top 10: 1.000000 44: top 1: 0.133333, top 10: 1.000000 45: top 1: 0.130435, top 10: 1.000000 46: top 1: 0.127660, top 10: 1.000000 47: top 1: 0.125000, top 10: 1.000000 48: top 1: 0.122449, top 10: 1.000000 49: top 1: 0.120000, top 10: 1.000000 50: top 1: 0.117647, top 10: 1.000000 51: top 1: 0.115385, top 10: 1.000000 52: top 1: 0.113208, top 10: 1.000000 53: top 1: 0.111111, top 10: 1.000000 54: top 1: 0.109091, top 10: 1.000000 55: top 1: 0.107143, top 10: 1.000000 56: top 1: 0.105263, top 10: 1.000000 57: top 1: 0.103448, top 10: 1.000000 58: top 1: 0.101695, top 10: 1.000000 59: top 1: 0.100000, top 10: 1.000000 60: top 1: 0.098361, top 10: 1.000000 61: top 1: 0.096774, top 10: 1.000000 62: top 1: 0.095238, top 10: 1.000000 63: top 1: 0.093750, top 10: 1.000000 64: top 1: 0.092308, top 10: 1.000000 65: top 1: 0.090909, top 10: 1.000000 66: top 1: 0.089552, top 10: 1.000000 67: top 1: 0.102941, top 10: 1.000000 68: top 1: 0.101449, top 10: 1.000000 69: top 1: 0.100000, top 10: 1.000000 70: top 1: 0.098592, top 10: 1.000000 71: top 1: 0.097222, top 10: 1.000000 72: top 1: 0.095890, top 10: 1.000000 73: top 1: 0.094595, top 10: 1.000000 74: top 1: 0.093333, top 10: 1.000000 75: top 1: 0.092105, top 10: 1.000000 76: top 1: 0.090909, top 10: 1.000000 77: top 1: 0.102564, top 10: 1.000000 78: top 1: 0.113924, top 10: 1.000000 79: top 1: 0.112500, top 10: 1.000000 80: top 1: 0.111111, top 10: 1.000000 81: top 1: 0.109756, top 10: 1.000000 82: top 1: 0.108434, top 10: 1.000000 83: top 1: 0.107143, top 10: 1.000000 ..... 9992: top 1: 0.099970, top 10: 1.000000 9993: top 1: 0.099960, top 10: 1.000000 9994: top 1: 0.099950, top 10: 1.000000 9995: top 1: 0.099940, top 10: 1.000000 9996: top 1: 0.100030, top 10: 1.000000 9997: top 1: 0.100020, top 10: 1.000000 9998: top 1: 0.100010, top 10: 1.000000 9999: top 1: 0.100000, top 10: 1.000000

As we can see, the accuracy for ten-class classification is random guess (10%). here is the cifar.data. the model cfg also used the one from the website.

classes=10 train = data/cifar/train.list valid = data/cifar/test.list labels = data/cifar/labels.txt backup = backup/ top=10

hahakid commented 6 years ago

I find a very strange problem, the network.cfg file uses 6464 for training, and the valid and test may fail to deal with images. When I modified the input size as 5656, I can get a correct result!!!!!!!!!WTF???

6464===>>>> ./darknet classifier predict cfg/wpi/wpi.data cfg/wpi/light_tiny.cfg /media/kid/data/trafficlight/experiment/dataset/WPI/dataset/lights/models/light_tiny_74.weights /media/kid/data/trafficlight/experiment/dataset/WPI/dataset/lights/test/3122_goc.jpg layer filters size input output 0 conv 16 3 x 3 / 1 64 x 64 x 3 -> 64 x 64 x 16 1 max 2 x 2 / 2 64 x 64 x 16 -> 32 x 32 x 16 2 conv 32 3 x 3 / 1 32 x 32 x 16 -> 32 x 32 x 32 3 max 2 x 2 / 2 32 x 32 x 32 -> 16 x 16 x 32 4 conv 16 1 x 1 / 1 16 x 16 x 32 -> 16 x 16 x 16 5 conv 128 3 x 3 / 1 16 x 16 x 16 -> 16 x 16 x 128 6 conv 16 1 x 1 / 1 16 x 16 x 128 -> 16 x 16 x 16 7 conv 128 3 x 3 / 1 16 x 16 x 16 -> 16 x 16 x 128 8 max 2 x 2 / 2 16 x 16 x 128 -> 8 x 8 x 128 9 conv 32 1 x 1 / 1 8 x 8 x 128 -> 8 x 8 x 32 10 conv 256 3 x 3 / 1 8 x 8 x 32 -> 8 x 8 x 256 11 conv 32 1 x 1 / 1 8 x 8 x 256 -> 8 x 8 x 32 12 conv 256 3 x 3 / 1 8 x 8 x 32 -> 8 x 8 x 256 13 max 2 x 2 / 2 8 x 8 x 256 -> 4 x 4 x 256 14 conv 64 1 x 1 / 1 4 x 4 x 256 -> 4 x 4 x 64 15 conv 512 3 x 3 / 1 4 x 4 x 64 -> 4 x 4 x 512 16 conv 64 1 x 1 / 1 4 x 4 x 512 -> 4 x 4 x 64 17 conv 512 3 x 3 / 1 4 x 4 x 64 -> 4 x 4 x 512 18 conv 128 1 x 1 / 1 4 x 4 x 512 -> 4 x 4 x 128 19 conv 7 1 x 1 / 1 4 x 4 x 128 -> 4 x 4 x 7 20 avg 4 x 4 x 7 -> 7 21 softmax 7 22 cost 7 Loading weights from /media/kid/data/trafficlight/experiment/dataset/WPI/dataset/lights/models/light_tiny_74.weights...Done! /media/kid/data/trafficlight/experiment/dataset/WPI/dataset/lights/test/3122_goc.jpg: Predicted in 0.001452 seconds. 94.46%: others 5.46%: GC 0.06%: RC 0.01%: GAu 0.01%: RAl 0.00%: GAr 0.00%: GAl 5656===>>>> ./darknet classifier predict cfg/wpi/wpi.data cfg/wpi/light_tiny.cfg /media/kid/data/trafficlight/experiment/dataset/WPI/dataset/lights/models/light_tiny_74.weights /media/kid/data/trafficlight/experiment/dataset/WPI/dataset/lights/test/3122_goc.jpg layer filters size input output 0 conv 16 3 x 3 / 1 56 x 56 x 3 -> 56 x 56 x 16 1 max 2 x 2 / 2 56 x 56 x 16 -> 28 x 28 x 16 2 conv 32 3 x 3 / 1 28 x 28 x 16 -> 28 x 28 x 32 3 max 2 x 2 / 2 28 x 28 x 32 -> 14 x 14 x 32 4 conv 16 1 x 1 / 1 14 x 14 x 32 -> 14 x 14 x 16 5 conv 128 3 x 3 / 1 14 x 14 x 16 -> 14 x 14 x 128 6 conv 16 1 x 1 / 1 14 x 14 x 128 -> 14 x 14 x 16 7 conv 128 3 x 3 / 1 14 x 14 x 16 -> 14 x 14 x 128 8 max 2 x 2 / 2 14 x 14 x 128 -> 7 x 7 x 128 9 conv 32 1 x 1 / 1 7 x 7 x 128 -> 7 x 7 x 32 10 conv 256 3 x 3 / 1 7 x 7 x 32 -> 7 x 7 x 256 11 conv 32 1 x 1 / 1 7 x 7 x 256 -> 7 x 7 x 32 12 conv 256 3 x 3 / 1 7 x 7 x 32 -> 7 x 7 x 256 13 max 2 x 2 / 2 7 x 7 x 256 -> 3 x 3 x 256 14 conv 64 1 x 1 / 1 3 x 3 x 256 -> 3 x 3 x 64 15 conv 512 3 x 3 / 1 3 x 3 x 64 -> 3 x 3 x 512 16 conv 64 1 x 1 / 1 3 x 3 x 512 -> 3 x 3 x 64 17 conv 512 3 x 3 / 1 3 x 3 x 64 -> 3 x 3 x 512 18 conv 128 1 x 1 / 1 3 x 3 x 512 -> 3 x 3 x 128 19 conv 7 1 x 1 / 1 3 x 3 x 128 -> 3 x 3 x 7 20 avg 3 x 3 x 7 -> 7 21 softmax 7 22 cost 7 Loading weights from /media/kid/data/trafficlight/experiment/dataset/WPI/dataset/lights/models/light_tiny_74.weights...Done! /media/kid/data/trafficlight/experiment/dataset/WPI/dataset/lights/test/3122_goc.jpg: Predicted in 0.001535 seconds. 99.96%: GAu 0.02%: GAr 0.01%: others 0.01%: GC 0.00%: RAl 0.00%: RC 0.00%: GAl

And, If I use

Test

batch=1

subdivisions=1

, and the 64*64 input also works! This framework is so sick!!!!!!!