diva-dia/deepdiva

⛔️ DEPRECATED <Python Framework for Reproducible Deep Learning Experiments>

deepdiva's Issues

Downloading mnist

Setting up deep-diva on a new machine (thus possibly related to using [email protected]) we ran into an issue downloading the mnist dataset.

$ python3 util/data/get_a_dataset.py --dataset mnist --output-folder dataset/mnist
Downloading http://yann.lecun.com/exdb/mnist/train-images-idx3-ubyte.gz to dataset/mnist/MNIST/raw/train-images-idx3-ubyte.gz
100.1%Extracting dataset/mnist/MNIST/raw/train-images-idx3-ubyte.gz
Downloading http://yann.lecun.com/exdb/mnist/train-labels-idx1-ubyte.gz to dataset/mnist/MNIST/raw/train-labels-idx1-ubyte.gz
113.5%Extracting dataset/mnist/MNIST/raw/train-labels-idx1-ubyte.gz
Downloading http://yann.lecun.com/exdb/mnist/t10k-images-idx3-ubyte.gz to dataset/mnist/MNIST/raw/t10k-images-idx3-ubyte.gz
100.4%Extracting dataset/mnist/MNIST/raw/t10k-images-idx3-ubyte.gz
Downloading http://yann.lecun.com/exdb/mnist/t10k-labels-idx1-ubyte.gz to dataset/mnist/MNIST/raw/t10k-labels-idx1-ubyte.gz
180.4%Extracting dataset/mnist/MNIST/raw/t10k-labels-idx1-ubyte.gz
Processing...
Done!
Traceback (most recent call last):
  File "util/data/get_a_dataset.py", line 297, in <module>
    getattr(sys.modules[__name__], args.dataset)(args)
  File "util/data/get_a_dataset.py", line 46, in mnist
    'training.pt'))
  File "/home/ashlaban/.local/lib/python3.6/site-packages/torch/serialization.py", line 382, in load
    f = open(f, 'rb')
FileNotFoundError: [Errno 2] No such file or directory: 'dataset/mnist/processed/training.pt'

Semantic segmentation trainning error.

from: https://diva-dia.github.io/DeepDIVAweb/articles/use-image-segmentation/

I run code by using resnet18 model.
python RunMe.py --dataset-folder /path/to/dataset --runner-class semantic_segmentation --no-cuda --ignoregit --model-name resnet18.

For dataset, I use Diva-HisDB CB55 images

• data images from
  https://diuf.unifr.ch/main/hisdoc/sites/diuf.unifr.ch.main.hisdoc/files/uploads/diva-hisdb/hisdoc/img-CB55.zip

• gt images from
  https://diuf.unifr.ch/main/hisdoc/sites/diuf.unifr.ch.main.hisdoc/files/uploads/diva-hisdb/hisdoc/pixel-level-gt-CB55.zip

I arranged data to

dataset/train/data/
dataset/train/gt/

Trainning error is

[2019-10-04 10:04:58] [   ERROR] --- Unhandled error: RuntimeError('Given input size: (512x4x4). Calculated output size: (512x-2x-2). Output size is too small at /opt/conda/conda-bld/pytorch_1549636813070/work/aten/src/THNN/generic/SpatialAveragePooling.c:48') (RunMe.py:264)
[2019-10-04 10:04:58] [   ERROR] --- Traceback (most recent call last):
  File "template/RunMe.py", line 257, in _execute
    **args.__dict__)
  File "/home/nott/AI/ancient/DeepDIVA/template/runner/image_classification/image_classification.py", line 47, in single_run
    **kwargs)
  File "/home/nott/AI/ancient/DeepDIVA/template/runner/image_classification/image_classification.py", line 144, in train_routine
    val_value[-1] = cls._validate(epoch=-1, **kwargs)
  File "/home/nott/AI/ancient/DeepDIVA/template/runner/semantic_segmentation/semantic_segmentation.py", line 40, in _validate
    return evaluate.validate(class_encodings=cls.class_encoding, **kwargs)
  File "/home/nott/AI/ancient/DeepDIVA/template/runner/semantic_segmentation/evaluate.py", line 78, in validate
    output = model(input)
  File "/home/nott/anaconda3/envs/deepdiva/lib/python3.7/site-packages/torch/nn/modules/module.py", line 489, in __call__
    result = self.forward(*input, **kwargs)
  File "/home/nott/AI/ancient/DeepDIVA/models/image_classification/ResNet.py", line 156, in forward
    x = self.avgpool(x)
  File "/home/nott/anaconda3/envs/deepdiva/lib/python3.7/site-packages/torch/nn/modules/module.py", line 489, in __call__
    result = self.forward(*input, **kwargs)
  File "/home/nott/anaconda3/envs/deepdiva/lib/python3.7/site-packages/torch/nn/modules/pooling.py", line 565, in forward
    self.padding, self.ceil_mode, self.count_include_pad)
RuntimeError: Given input size: (512x4x4). Calculated output size: (512x-2x-2). Output size is too small at /opt/conda/conda-bld/pytorch_1549636813070/work/aten/src/THNN/generic/SpatialAveragePooling.c:48
 (RunMe.py:265)

I use latest git deepdiva and release version but still have this error

Visualizing embeddings - Format to specify output

Having trouble with creating a embedding to visualize. As per the DeepDIVA tutorial I use the following command,

python util/visualization/embedding.py --results-file /home/local/DFPOC/ghfiy/DeepDIVA/output/OSRAM3/MNIST/output_channels=10/no_cuda=True/04-06-19-10h-27m-58s/results.pkl --output-file /home/local/DFPOC/ghfiy/DeepDIVA/output/OSRAM3/MNIST/output_channels=10/no_cuda=True/04-06-19-10h-27m-58s --embedding tsne --tensorboard

whereby as per instructions I specify path to output folder. This is the error I get,

_usage: embedding.py [-h] [--results-file RESULTS_FILE]
[--embedding {_load_thumbnail,_main,_make_embedding,make_folder_if_not_exists,isomap,mds,pca,tsne}]
[--output OUTPUT] [--3d] [--tensorboard]
embedding.py: error: unrecognized arguments: --output-file /home/local/DFPOC/ghfiy/DeepDIVA/output/OSRAM3/MNIST/output_channels=10/no_cuda=True/04-06-19-10h-27m-58s

Since it complains of unrecognized argument I looked up the embedding.py script and saw the following lines,

parser.add_argument('--output', type=str, default='./output.png', help='path to generate output image')

Therefore I changed the command to,

python util/visualization/embedding.py --results-file /home/local/DFPOC/ghfiy/DeepDIVA/output/OSRAM3/MNIST/output_channels=10/no_cuda=True/04-06-19-10h-27m-58s/results.pkl --output /home/local/DFPOC/ghfiy/DeepDIVA/output/OSRAM3/MNIST/output_channels=10/no_cuda=True/04-06-19-10h-27m-58s --embedding tsne --tensorboard

Then the following error pops up,

_Traceback (most recent call last):
File "util/visualization/embedding.py", line 242, in
_main(args)
File "util/visualization/embedding.py", line 192, in _main
make_folder_if_not_exists(os.path.dirname(args.output_file))
AttributeError: 'Namespace' object has no attribute 'output_file'

Also tried the following,

(deepdiva) ghfiy@by-df547:~/DeepDIVA$ python util/visualization/embedding.py --results-file /home/local/DFPOC/ghfiy/DeepDIVA/output/OSRAM3/MNIST/output_channels=10/no_cuda=True/04-06-19-10h-27m-58s/results.pkl --output_file /home/local/DFPOC/ghfiy/DeepDIVA/output/OSRAM3/MNIST/output_channels=10/no_cuda=True/04-06-19-10h-27m-58s --embedding tsne --tensorboard

_usage: embedding.py [-h] [--results-file RESULTS_FILE]
[--embedding {_load_thumbnail,_main,_make_embedding,make_folder_if_not_exists,isomap,mds,pca,tsne}]
[--output OUTPUT] [--3d] [--tensorboard]
embedding.py: error: unrecognized arguments: --output_file /home/local/DFPOC/ghfiy/DeepDIVA/output/OSRAM3/MNIST/output_channels=10/no_cuda=True/04-06-19-10h-27m-58s

(deepdiva) ghfiy@by-df547:~/DeepDIVA$ python util/visualization/embedding.py --results-file /home/local/DFPOC/ghfiy/DeepDIVA/output/OSRAM3/MNIST/output_channels=10/no_cuda=True/04-06-19-10h-27m-58s/results.pkl --output-file ./output.png --embedding tsne --tensorboard

_usage: embedding.py [-h] [--results-file RESULTS_FILE]
[--embedding {_load_thumbnail,_main,_make_embedding,make_folder_if_not_exists,isomap,mds,pca,tsne}]
[--output OUTPUT] [--3d] [--tensorboard]
embedding.py: error: unrecognized arguments: --output-file ./output.png

and various other permutations and combinations. What would be the right way to specify this argument?

Failing setup and mnist installation when following 'Getting started'

Gets errors when running 'bash setup_environment.sh' which presumably causes further errors when attempting to Download mnist ('python util/data/get_a_dataset.py --dataset mnist --output-folder toy_dataset') when following the 'Getting started' instructions.

Printout from command line for Getting started process:

Collecting package metadata (repodata.json): done
Solving environment: done
Preparing transaction: done
Verifying transaction: done
Executing transaction: done
Ran pip subprocess with arguments:
['/opt/Anaconda/envs/deepdiva/bin/python', '-m', 'pip', 'install', '-U', '-r', '/nfs/staff/gusgru/Repositories/DeepDIVA/condaenv.hygei685.requirements.txt']
Pip subprocess output:
Requirement already up-to-date: dask==1.2.2 in /opt/Anaconda/envs/deepdiva/lib/python3.7/site-packages (from -r /nfs/staff/gusgru/Repositories/DeepDIVA/condaenv.hygei685.requirements.txt (line 1)) (1.2.2)
Collecting pydensecrf==1.0rc2 (from -r /nfs/staff/gusgru/Repositories/DeepDIVA/condaenv.hygei685.requirements.txt (line 2))
  Using cached https://files.pythonhosted.org/packages/dd/11/7a79e817078323893f8071bb3fda1f45e34e443ec4a0790f8542158bbced/pydensecrf-1.0rc2.tar.gz
Collecting rarfile==3.0 (from -r /nfs/staff/gusgru/Repositories/DeepDIVA/condaenv.hygei685.requirements.txt (line 3))
Requirement already up-to-date: torch==1.0.1.post2 in /opt/Anaconda/envs/deepdiva/lib/python3.7/site-packages (from -r /nfs/staff/gusgru/Repositories/DeepDIVA/condaenv.hygei685.requirements.txt (line 4)) (1.0.1.post2)
Requirement already up-to-date: wget==3.2 in /opt/Anaconda/envs/deepdiva/lib/python3.7/site-packages (from -r /nfs/staff/gusgru/Repositories/DeepDIVA/condaenv.hygei685.requirements.txt (line 5)) (3.2)
Building wheels for collected packages: pydensecrf
  Building wheel for pydensecrf (setup.py): started
  Building wheel for pydensecrf (setup.py): finished with status 'error'
  Complete output from command /opt/Anaconda/envs/deepdiva/bin/python -u -c "import setuptools, tokenize;__file__='/tmp/pip-install-xor3fcig/pydensecrf/setup.py';f=getattr(tokenize, 'open', open)(__file__);code=f.read().replace('\r\n', '\n');f.close();exec(compile(code, __file__, 'exec'))" bdist_wheel -d /tmp/pip-wheel-w0gp68c6 --python-tag cp37:
  running bdist_wheel
  running build
  running build_py
  creating build
  creating build/lib.linux-x86_64-3.7
  creating build/lib.linux-x86_64-3.7/pydensecrf
  copying pydensecrf/test.py -> build/lib.linux-x86_64-3.7/pydensecrf
  copying pydensecrf/utils.py -> build/lib.linux-x86_64-3.7/pydensecrf
  copying pydensecrf/__init__.py -> build/lib.linux-x86_64-3.7/pydensecrf
  copying pydensecrf/test_eigen.py -> build/lib.linux-x86_64-3.7/pydensecrf
  running build_ext
  building 'pydensecrf.eigen' extension
  creating build/temp.linux-x86_64-3.7
  creating build/temp.linux-x86_64-3.7/pydensecrf
  gcc -pthread -B /opt/Anaconda/envs/deepdiva/compiler_compat -Wl,--sysroot=/ -Wsign-compare -DNDEBUG -g -fwrapv -O3 -Wall -Wstrict-prototypes -fPIC -Ipydensecrf/densecrf/include -Ipydensecrf -I/opt/Anaconda/envs/deepdiva/include/python3.7m -c pydensecrf/eigen.cpp -o build/temp.linux-x86_64-3.7/pydensecrf/eigen.o
  cc1plus: warning: command line option ‘-Wstrict-prototypes’ is valid for C/ObjC but not for C++
  In file included from pydensecrf/densecrf/include/Eigen/Core:248:0,
                   from pydensecrf/densecrf/include/Eigen/Dense:1,
                   from pydensecrf/eigen.cpp:299:
  pydensecrf/densecrf/include/Eigen/src/Core/util/Memory.h: In function ‘Index Eigen::internal::first_aligned(const Scalar*, Index)’:
  pydensecrf/densecrf/include/Eigen/src/Core/util/Memory.h:454:48: warning: typedef ‘Packet’ locally defined but not used [-Wunused-local-typedefs]
     typedef typename packet_traits<Scalar>::type Packet;
                                                  ^
  In file included from pydensecrf/densecrf/include/Eigen/Core:324:0,
                   from pydensecrf/densecrf/include/Eigen/Dense:1,
                   from pydensecrf/eigen.cpp:299:
  pydensecrf/densecrf/include/Eigen/src/Core/products/SelfadjointMatrixVector.h: In static member function ‘static void Eigen::internal::selfadjoint_matrix_vector_product<Scalar, Index, StorageOrder, UpLo, ConjugateLhs, ConjugateRhs, Version>::run(Index, const Scalar*, Index, const Scalar*, Index, Scalar*, Scalar)’:
  pydensecrf/densecrf/include/Eigen/src/Core/products/SelfadjointMatrixVector.h:38:44: warning: typedef ‘RealScalar’ locally defined but not used [-Wunused-local-typedefs]
     typedef typename NumTraits<Scalar>::Real RealScalar;
                                              ^
  In file included from pydensecrf/densecrf/include/Eigen/QR:26:0,
                   from pydensecrf/densecrf/include/Eigen/Dense:4,
                   from pydensecrf/eigen.cpp:299:
  pydensecrf/densecrf/include/Eigen/src/QR/HouseholderQR.h: In function ‘void Eigen::internal::householder_qr_inplace_blocked(MatrixQR&, HCoeffs&, typename MatrixQR::Index, typename MatrixQR::Scalar*)’:
  pydensecrf/densecrf/include/Eigen/src/QR/HouseholderQR.h:235:41: warning: typedef ‘RealScalar’ locally defined but not used [-Wunused-local-typedefs]
     typedef typename MatrixQR::RealScalar RealScalar;
                                           ^
  In file included from pydensecrf/densecrf/include/Eigen/Eigenvalues:28:0,
                   from pydensecrf/densecrf/include/Eigen/Dense:7,
                   from pydensecrf/eigen.cpp:299:
  pydensecrf/densecrf/include/Eigen/src/Eigenvalues/Tridiagonalization.h: In function ‘void Eigen::internal::tridiagonalization_inplace(MatrixType&, DiagonalType&, SubDiagonalType&, bool)’:
  pydensecrf/densecrf/include/Eigen/src/Eigenvalues/Tridiagonalization.h:428:38: warning: typedef ‘Index’ locally defined but not used [-Wunused-local-typedefs]
     typedef typename MatrixType::Index Index;
                                        ^
  In file included from pydensecrf/densecrf/include/Eigen/Eigenvalues:34:0,
                   from pydensecrf/densecrf/include/Eigen/Dense:7,
                   from pydensecrf/eigen.cpp:299:
  pydensecrf/densecrf/include/Eigen/src/Eigenvalues/ComplexSchur.h: In static member function ‘static void Eigen::internal::complex_schur_reduce_to_hessenberg<MatrixType, false>::run(Eigen::ComplexSchur<MatrixType>&, const MatrixType&, bool)’:
  pydensecrf/densecrf/include/Eigen/src/Eigenvalues/ComplexSchur.h:312:66: warning: typedef ‘ComplexMatrixType’ locally defined but not used [-Wunused-local-typedefs]
       typedef typename ComplexSchur<MatrixType>::ComplexMatrixType ComplexMatrixType;
                                                                    ^
  pydensecrf/eigen.cpp: In function ‘void __Pyx__ExceptionSave(PyThreadState*, PyObject**, PyObject**, PyObject**)’:
  pydensecrf/eigen.cpp:17032:21: error: ‘PyThreadState {aka struct _ts}’ has no member named ‘exc_type’
       *type = tstate->exc_type;
                       ^
  pydensecrf/eigen.cpp:17033:22: error: ‘PyThreadState {aka struct _ts}’ has no member named ‘exc_value’
       *value = tstate->exc_value;
                        ^
  pydensecrf/eigen.cpp:17034:19: error: ‘PyThreadState {aka struct _ts}’ has no member named ‘exc_traceback’
       *tb = tstate->exc_traceback;
                     ^
  pydensecrf/eigen.cpp: In function ‘void __Pyx__ExceptionReset(PyThreadState*, PyObject*, PyObject*, PyObject*)’:
  pydensecrf/eigen.cpp:17041:24: error: ‘PyThreadState {aka struct _ts}’ has no member named ‘exc_type’
       tmp_type = tstate->exc_type;
                          ^
  pydensecrf/eigen.cpp:17042:25: error: ‘PyThreadState {aka struct _ts}’ has no member named ‘exc_value’
       tmp_value = tstate->exc_value;
                           ^
  pydensecrf/eigen.cpp:17043:22: error: ‘PyThreadState {aka struct _ts}’ has no member named ‘exc_traceback’
       tmp_tb = tstate->exc_traceback;
                        ^
  pydensecrf/eigen.cpp:17044:13: error: ‘PyThreadState {aka struct _ts}’ has no member named ‘exc_type’
       tstate->exc_type = type;
               ^
  pydensecrf/eigen.cpp:17045:13: error: ‘PyThreadState {aka struct _ts}’ has no member named ‘exc_value’
       tstate->exc_value = value;
               ^
  pydensecrf/eigen.cpp:17046:13: error: ‘PyThreadState {aka struct _ts}’ has no member named ‘exc_traceback’
       tstate->exc_traceback = tb;
               ^
  pydensecrf/eigen.cpp: In function ‘int __Pyx__GetException(PyThreadState*, PyObject**, PyObject**, PyObject**)’:
  pydensecrf/eigen.cpp:17101:24: error: ‘PyThreadState {aka struct _ts}’ has no member named ‘exc_type’
       tmp_type = tstate->exc_type;
                          ^
  pydensecrf/eigen.cpp:17102:25: error: ‘PyThreadState {aka struct _ts}’ has no member named ‘exc_value’
       tmp_value = tstate->exc_value;
                           ^
  pydensecrf/eigen.cpp:17103:22: error: ‘PyThreadState {aka struct _ts}’ has no member named ‘exc_traceback’
       tmp_tb = tstate->exc_traceback;
                        ^
  pydensecrf/eigen.cpp:17104:13: error: ‘PyThreadState {aka struct _ts}’ has no member named ‘exc_type’
       tstate->exc_type = local_type;
               ^
  pydensecrf/eigen.cpp:17105:13: error: ‘PyThreadState {aka struct _ts}’ has no member named ‘exc_value’
       tstate->exc_value = local_value;
               ^
  pydensecrf/eigen.cpp:17106:13: error: ‘PyThreadState {aka struct _ts}’ has no member named ‘exc_traceback’
       tstate->exc_traceback = local_tb;
               ^
  pydensecrf/eigen.cpp: In function ‘void __Pyx__ExceptionSwap(PyThreadState*, PyObject**, PyObject**, PyObject**)’:
  pydensecrf/eigen.cpp:17128:24: error: ‘PyThreadState {aka struct _ts}’ has no member named ‘exc_type’
       tmp_type = tstate->exc_type;
                          ^
  pydensecrf/eigen.cpp:17129:25: error: ‘PyThreadState {aka struct _ts}’ has no member named ‘exc_value’
       tmp_value = tstate->exc_value;
                           ^
  pydensecrf/eigen.cpp:17130:22: error: ‘PyThreadState {aka struct _ts}’ has no member named ‘exc_traceback’
       tmp_tb = tstate->exc_traceback;
                        ^
  pydensecrf/eigen.cpp:17131:13: error: ‘PyThreadState {aka struct _ts}’ has no member named ‘exc_type’
       tstate->exc_type = *type;
               ^
  pydensecrf/eigen.cpp:17132:13: error: ‘PyThreadState {aka struct _ts}’ has no member named ‘exc_value’
       tstate->exc_value = *value;
               ^
  pydensecrf/eigen.cpp:17133:13: error: ‘PyThreadState {aka struct _ts}’ has no member named ‘exc_traceback’
       tstate->exc_traceback = *tb;
               ^
  In file included from pydensecrf/eigen.cpp:300:0:
  pydensecrf/eigen_impl.cpp: At global scope:
  pydensecrf/eigen_impl.cpp:13:13: warning: ‘void vecf2buf(const VectorXf&, float*)’ defined but not used [-Wunused-function]
   static void vecf2buf(const Eigen::VectorXf& vec, float *mem)
               ^
  pydensecrf/eigen_impl.cpp:27:13: warning: ‘void matf2buf(const MatrixXf&, float*)’ defined but not used [-Wunused-function]
   static void matf2buf(const Eigen::MatrixXf& mat, float *mem)
               ^
  error: command 'gcc' failed with exit status 1

  ----------------------------------------
  Running setup.py clean for pydensecrf
Failed to build pydensecrf
Installing collected packages: pydensecrf, rarfile
  Running setup.py install for pydensecrf: started
    Running setup.py install for pydensecrf: finished with status 'error'
    Complete output from command /opt/Anaconda/envs/deepdiva/bin/python -u -c "import setuptools, tokenize;__file__='/tmp/pip-install-xor3fcig/pydensecrf/setup.py';f=getattr(tokenize, 'open', open)(__file__);code=f.read().replace('\r\n', '\n');f.close();exec(compile(code, __file__, 'exec'))" install --record /tmp/pip-record-6a_kt9ca/install-record.txt --single-version-externally-managed --compile:
    running install
    running build
    running build_py
    creating build
    creating build/lib.linux-x86_64-3.7
    creating build/lib.linux-x86_64-3.7/pydensecrf
    copying pydensecrf/test.py -> build/lib.linux-x86_64-3.7/pydensecrf
    copying pydensecrf/utils.py -> build/lib.linux-x86_64-3.7/pydensecrf
    copying pydensecrf/__init__.py -> build/lib.linux-x86_64-3.7/pydensecrf
    copying pydensecrf/test_eigen.py -> build/lib.linux-x86_64-3.7/pydensecrf
    running build_ext
    building 'pydensecrf.eigen' extension
    creating build/temp.linux-x86_64-3.7
    creating build/temp.linux-x86_64-3.7/pydensecrf
    gcc -pthread -B /opt/Anaconda/envs/deepdiva/compiler_compat -Wl,--sysroot=/ -Wsign-compare -DNDEBUG -g -fwrapv -O3 -Wall -Wstrict-prototypes -fPIC -Ipydensecrf/densecrf/include -Ipydensecrf -I/opt/Anaconda/envs/deepdiva/include/python3.7m -c pydensecrf/eigen.cpp -o build/temp.linux-x86_64-3.7/pydensecrf/eigen.o
    cc1plus: warning: command line option ‘-Wstrict-prototypes’ is valid for C/ObjC but not for C++
    In file included from pydensecrf/densecrf/include/Eigen/Core:248:0,
                     from pydensecrf/densecrf/include/Eigen/Dense:1,
                     from pydensecrf/eigen.cpp:299:
    pydensecrf/densecrf/include/Eigen/src/Core/util/Memory.h: In function ‘Index Eigen::internal::first_aligned(const Scalar*, Index)’:
    pydensecrf/densecrf/include/Eigen/src/Core/util/Memory.h:454:48: warning: typedef ‘Packet’ locally defined but not used [-Wunused-local-typedefs]
       typedef typename packet_traits<Scalar>::type Packet;
                                                    ^
    In file included from pydensecrf/densecrf/include/Eigen/Core:324:0,
                     from pydensecrf/densecrf/include/Eigen/Dense:1,
                     from pydensecrf/eigen.cpp:299:
    pydensecrf/densecrf/include/Eigen/src/Core/products/SelfadjointMatrixVector.h: In static member function ‘static void Eigen::internal::selfadjoint_matrix_vector_product<Scalar, Index, StorageOrder, UpLo, ConjugateLhs, ConjugateRhs, Version>::run(Index, const Scalar*, Index, const Scalar*, Index, Scalar*, Scalar)’:
    pydensecrf/densecrf/include/Eigen/src/Core/products/SelfadjointMatrixVector.h:38:44: warning: typedef ‘RealScalar’ locally defined but not used [-Wunused-local-typedefs]
       typedef typename NumTraits<Scalar>::Real RealScalar;
                                                ^
    In file included from pydensecrf/densecrf/include/Eigen/QR:26:0,
                     from pydensecrf/densecrf/include/Eigen/Dense:4,
                     from pydensecrf/eigen.cpp:299:
    pydensecrf/densecrf/include/Eigen/src/QR/HouseholderQR.h: In function ‘void Eigen::internal::householder_qr_inplace_blocked(MatrixQR&, HCoeffs&, typename MatrixQR::Index, typename MatrixQR::Scalar*)’:
    pydensecrf/densecrf/include/Eigen/src/QR/HouseholderQR.h:235:41: warning: typedef ‘RealScalar’ locally defined but not used [-Wunused-local-typedefs]
       typedef typename MatrixQR::RealScalar RealScalar;
                                             ^
    In file included from pydensecrf/densecrf/include/Eigen/Eigenvalues:28:0,
                     from pydensecrf/densecrf/include/Eigen/Dense:7,
                     from pydensecrf/eigen.cpp:299:
    pydensecrf/densecrf/include/Eigen/src/Eigenvalues/Tridiagonalization.h: In function ‘void Eigen::internal::tridiagonalization_inplace(MatrixType&, DiagonalType&, SubDiagonalType&, bool)’:
    pydensecrf/densecrf/include/Eigen/src/Eigenvalues/Tridiagonalization.h:428:38: warning: typedef ‘Index’ locally defined but not used [-Wunused-local-typedefs]
       typedef typename MatrixType::Index Index;
                                          ^
    In file included from pydensecrf/densecrf/include/Eigen/Eigenvalues:34:0,
                     from pydensecrf/densecrf/include/Eigen/Dense:7,
                     from pydensecrf/eigen.cpp:299:
    pydensecrf/densecrf/include/Eigen/src/Eigenvalues/ComplexSchur.h: In static member function ‘static void Eigen::internal::complex_schur_reduce_to_hessenberg<MatrixType, false>::run(Eigen::ComplexSchur<MatrixType>&, const MatrixType&, bool)’:
    pydensecrf/densecrf/include/Eigen/src/Eigenvalues/ComplexSchur.h:312:66: warning: typedef ‘ComplexMatrixType’ locally defined but not used [-Wunused-local-typedefs]
         typedef typename ComplexSchur<MatrixType>::ComplexMatrixType ComplexMatrixType;
                                                                      ^
    pydensecrf/eigen.cpp: In function ‘void __Pyx__ExceptionSave(PyThreadState*, PyObject**, PyObject**, PyObject**)’:
    pydensecrf/eigen.cpp:17032:21: error: ‘PyThreadState {aka struct _ts}’ has no member named ‘exc_type’
         *type = tstate->exc_type;
                         ^
    pydensecrf/eigen.cpp:17033:22: error: ‘PyThreadState {aka struct _ts}’ has no member named ‘exc_value’
         *value = tstate->exc_value;
                          ^
    pydensecrf/eigen.cpp:17034:19: error: ‘PyThreadState {aka struct _ts}’ has no member named ‘exc_traceback’
         *tb = tstate->exc_traceback;
                       ^
    pydensecrf/eigen.cpp: In function ‘void __Pyx__ExceptionReset(PyThreadState*, PyObject*, PyObject*, PyObject*)’:
    pydensecrf/eigen.cpp:17041:24: error: ‘PyThreadState {aka struct _ts}’ has no member named ‘exc_type’
         tmp_type = tstate->exc_type;
                            ^
    pydensecrf/eigen.cpp:17042:25: error: ‘PyThreadState {aka struct _ts}’ has no member named ‘exc_value’
         tmp_value = tstate->exc_value;
                             ^
    pydensecrf/eigen.cpp:17043:22: error: ‘PyThreadState {aka struct _ts}’ has no member named ‘exc_traceback’
         tmp_tb = tstate->exc_traceback;
                          ^
    pydensecrf/eigen.cpp:17044:13: error: ‘PyThreadState {aka struct _ts}’ has no member named ‘exc_type’
         tstate->exc_type = type;
                 ^
    pydensecrf/eigen.cpp:17045:13: error: ‘PyThreadState {aka struct _ts}’ has no member named ‘exc_value’
         tstate->exc_value = value;
                 ^
    pydensecrf/eigen.cpp:17046:13: error: ‘PyThreadState {aka struct _ts}’ has no member named ‘exc_traceback’
         tstate->exc_traceback = tb;
                 ^
    pydensecrf/eigen.cpp: In function ‘int __Pyx__GetException(PyThreadState*, PyObject**, PyObject**, PyObject**)’:
    pydensecrf/eigen.cpp:17101:24: error: ‘PyThreadState {aka struct _ts}’ has no member named ‘exc_type’
         tmp_type = tstate->exc_type;
                            ^
    pydensecrf/eigen.cpp:17102:25: error: ‘PyThreadState {aka struct _ts}’ has no member named ‘exc_value’
         tmp_value = tstate->exc_value;
                             ^
    pydensecrf/eigen.cpp:17103:22: error: ‘PyThreadState {aka struct _ts}’ has no member named ‘exc_traceback’
         tmp_tb = tstate->exc_traceback;
                          ^
    pydensecrf/eigen.cpp:17104:13: error: ‘PyThreadState {aka struct _ts}’ has no member named ‘exc_type’
         tstate->exc_type = local_type;
Setup completed!
Please run 'source ~/.bashrc' to refresh your environment
You can activate the deepdiva environment with 'source activate deepdiva'
(base) gusgru@B4375:~/Repositories/DeepDIVA$ source ~/.bashrc
(base) gusgru@B4375:~/Repositories/DeepDIVA$ source activate deepdiva
(deepdiva) gusgru@B4375:~/Repositories/DeepDIVA$ python util/data/get_a_dataset.py --dataset mnist --output-folder toy_dataset
Traceback (most recent call last):
  File "util/data/get_a_dataset.py", line 19, in <module>
    import rarfile
ModuleNotFoundError: No module named 'rarfile'

Improvements to the sigopt integration

Just a few notes from my interaction with the sigopt integration.

• We should probably report (best) validation scores, not test scores by defualt
• The runner should check if an experiment exists and is non-archived/deleted before creating a new one.
• In triplet training, the mAP evaluation is costly for large datasets. When doing hyper-parameter optimisation, we are generally only interested in the final value, so one could potentially skip all but the last evaluations. This also prefers a lower number of epochs if the results are comparable, saving on training time. (Maybe only enable this if n_epochs is part of the optimisation process?)

I will update this issue if I come across more points.

Triplet loss `invert_best=True` -> `False`

AFAIU, triplet training uses the MAP metric which ranges from 0 (bad) to 1 (good). I.e. the best value is the one with the maximum value. Hence I suspect this line of code:

DeepDIVA/template/runner/triplet/triplet.py

Line 130 in 6806103

invert_best=True,

should read

invert_best=False,

instead of True.

Segmentation fault running setup_environment.sh when reinstalling (ubuntu)

Hi.

I removed the DeepDIVA and miniconda3 folder to make a clean reinstall of DeepDiva, but everytime I run setup_environment.sh it get to line 42: echo 'source ~/.bash_functions' >> ~/.bashrc the terminal stops, prints 'segmentation fault' and crashes. If I then try to open a new terminal, it also crashes until I remove the source ~/.bash_functions from my .bashrc.

Initially, the segmentation fault showed up when I deteled everything within the >>>> initialized by conda block. I did this because I ran setup_environment.sh a second time (I did this really just to see what would happen). The result of running setup_enviroment.sh a second time was additional >>>> initialized by conda blocks in my .bashrc.

I am assuming that there is some .bash_whatever or some path variable that messes things up but I do not know where to start searching.

Model dict not found

When trying to extract features from a model I provide the path to the checkpoint as so,
python template/RunMe.py --runner-class apply_model --dataset-folder toy_dataset/MNIST --load-model /home/local/DFPOC/ghfiy/DeepDIVA/log/OSRAM/MNIST/epochs=1/lr=1e-06/no_cuda=True/03-06-19-13h-22m-02s/ckeckpoint.pth.tar --output-channels 10 --ignoregit --no-cuda

However when it attempts to search the location the following pops up,
[2019-06-03 20:13:10] [ INFO] --- Setup logging. Log file: ./output/OSRAM3/MNIST/output_channels=10/no_cuda=True/03-06-19-20h-13m-10s/logs.txt (setup.py:566)
[2019-06-03 20:13:10] [ INFO] --- Arguments saved to: ./output/OSRAM3/MNIST/output_channels=10/no_cuda=True/03-06-19-20h-13m-10s/args.txt (setup.py:569)
[2019-06-03 20:13:10] [ INFO] --- Initialize Tensorboard SummaryWriter (setup.py:574)
[2019-06-03 20:13:13] [ WARNING] --- Git status is ignored! (RunMe.py:205)
[2019-06-03 20:13:13] [ INFO] --- Randomly chosen seed is: 1936313780 (setup.py:651)
[2019-06-03 20:13:13] [ INFO] --- Model CNN_basic expects input size of (32, 32) (apply_model.py:70)
[2019-06-03 20:13:13] [ INFO] --- Loading MNIST from:toy_dataset/MNIST (setup.py:61)
[2019-06-03 20:13:14] [ INFO] --- Setting up model CNN_basic (setup.py:90)
[2019-06-03 20:13:14] [ INFO] --- Loading weights for data balancing (setup.py:229)
[2019-06-03 20:13:14] [ ERROR] --- No model dict found at '/home/local/DFPOC/ghfiy/DeepDIVA/log/OSRAM/MNIST/epochs=1/lr=1e-06/no_cuda=True/03-06-19-13h-22m-02s/ckeckpoint.pth.tar' (setup.py:118)

I have checked and the file checkpoint.pth.tar does exist in this location. Hopefully this is right file containing the checkpoint. What could be a possible workaround for this issue?

Issues with torch 1.1

Setting up diva on a new machine (without conda, but the problem should apply in the conda case as well since the torch version is not pinned and the latest conda version is [email protected]) we ran into some issues:

1. In util/data/get_a_dataset.py the dataset properties changed from

    train_data, train_labels = cifar_train.train_data, cifar_train.train_labels
    test_data, test_labels = cifar_test.test_data, cifar_test.test_labels

to

    train_data, train_labels = cifar_train.data, cifar_train.targets
    test_data, test_labels = cifar_test.data, cifar_test.targets

. This should apply to all datasets loaded through the same mechanism.

2. It seems torch no longer supports accessing 0-dim elements through array indexing. As such this change was required: loss.data[0] -> loss.data.item() in template/runner/image_classification/evaluate.py and template/runner/image_classification/train.py. Should be applicable also to apply_model etc.

3. Something is fishy with the arrays returned from _load_mean_std_from_file, the transforms.Normalize cannot convert the input mean and std to float tensors, instead the input is considered of dtype=np.object. Check setup_dataloader in template/setup.py.

If I find the time I will also provide a proper PR for these problems. For now, please see attached patch.

PATCH

diff --git a/template/runner/image_classification/evaluate.py b/template/runner/image_classification/evaluate.py
index 9205d73..e22f594 100644
--- a/template/runner/image_classification/evaluate.py
+++ b/template/runner/image_classification/evaluate.py
@@ -91,7 +91,7 @@ def _evaluate(data_loader, model, criterion, writer, epoch, logging_label, no_cu
 
         # Compute and record the loss
         loss = criterion(output, target_var)
-        losses.update(loss.data[0], input.size(0))
+        losses.update(loss.data.item(), input.size(0))
 
         # Compute and record the accuracy
         acc1 = accuracy(output.data, target, topk=(1,))[0]
@@ -103,10 +103,10 @@ def _evaluate(data_loader, model, criterion, writer, epoch, logging_label, no_cu
 
         # Add loss and accuracy to Tensorboard
         if multi_run is None:
-            writer.add_scalar(logging_label + '/mb_loss', loss.data[0], epoch * len(data_loader) + batch_idx)
+            writer.add_scalar(logging_label + '/mb_loss', loss.data.item(), epoch * len(data_loader) + batch_idx)
             writer.add_scalar(logging_label + '/mb_accuracy', acc1.cpu().numpy(), epoch * len(data_loader) + batch_idx)
         else:
-            writer.add_scalar(logging_label + '/mb_loss_{}'.format(multi_run), loss.data[0],
+            writer.add_scalar(logging_label + '/mb_loss_{}'.format(multi_run), loss.data.item(),
                               epoch * len(data_loader) + batch_idx)
             writer.add_scalar(logging_label + '/mb_accuracy_{}'.format(multi_run), acc1.cpu().numpy(),
                               epoch * len(data_loader) + batch_idx)
diff --git a/template/runner/image_classification/train.py b/template/runner/image_classification/train.py
index 9cf2324..497eee5 100644
--- a/template/runner/image_classification/train.py
+++ b/template/runner/image_classification/train.py
@@ -72,10 +72,10 @@ def train(train_loader, model, criterion, optimizer, writer, epoch, no_cuda=Fals
 
         # Add loss and accuracy to Tensorboard
         if multi_run is None:
-            writer.add_scalar('train/mb_loss', loss.data[0], epoch * len(train_loader) + batch_idx)
+            writer.add_scalar('train/mb_loss', loss.data.item(), epoch * len(train_loader) + batch_idx)
             writer.add_scalar('train/mb_accuracy', acc.cpu().numpy(), epoch * len(train_loader) + batch_idx)
         else:
-            writer.add_scalar('train/mb_loss_{}'.format(multi_run), loss.data[0],
+            writer.add_scalar('train/mb_loss_{}'.format(multi_run), loss.data.item(),
                               epoch * len(train_loader) + batch_idx)
             writer.add_scalar('train/mb_accuracy_{}'.format(multi_run), acc.cpu().numpy(),
                               epoch * len(train_loader) + batch_idx)
@@ -141,7 +141,7 @@ def train_one_mini_batch(model, criterion, optimizer, input_var, target_var, los
 
     # Compute and record the loss
     loss = criterion(output, target_var)
-    loss_meter.update(loss.data[0], len(input_var))
+    loss_meter.update(loss.data.item(), len(input_var))
 
     # Compute and record the accuracy
     acc = accuracy(output.data, target_var.data, topk=(1,))[0]
diff --git a/template/setup.py b/template/setup.py
index 2a87ccd..77c68a4 100644
--- a/template/setup.py
+++ b/template/setup.py
@@ -275,6 +275,8 @@ def set_up_dataloaders(model_expected_input_size, dataset_folder, batch_size, wo
 
         # Loads the analytics csv and extract mean and std
         mean, std = _load_mean_std_from_file(dataset_folder, inmem, workers)
+        mean = np.asarray([x for x in mean], dtype=np.float32)
+        std = np.asarray([x for x in std], dtype=np.float32)
 
         # Set up dataset transforms
         logging.debug('Setting up dataset transforms')
diff --git a/util/data/get_a_dataset.py b/util/data/get_a_dataset.py
index 5b78ef8..ea36794 100644
--- a/util/data/get_a_dataset.py
+++ b/util/data/get_a_dataset.py
@@ -150,9 +151,9 @@ def cifar10(args):
     cifar_test = torchvision.datasets.CIFAR10(root=args.output_folder, train=False, download=True)
 
     # Load the data into memory
-    train_data, train_labels = cifar_train.train_data, cifar_train.train_labels
+    train_data, train_labels = cifar_train.data, cifar_train.targets
 
-    test_data, test_labels = cifar_test.test_data, cifar_test.test_labels
+    test_data, test_labels = cifar_test.data, cifar_test.targets
 
     # Make output folders
     dataset_root = os.path.join(args.output_folder, 'CIFAR10')

Rename `self.train` -> `self.use_triplet_output` in `ImageFolderTriplet`

Renaming this member from train to use_triplet_output more clearly indicates that this is a switchable property (if one e.g. would want to eval both the triplet loss and the mAP "accuracy" metric on the same dataset).

DeepDIVA/datasets/image_folder_triplet.py

Line 103 in 4dccf3b

train : bool