Herramientas

Tutoriales MNIST (cuadernos Jupyter)

• Keras: mnist.src.Keras.ipynb

• TensorFlow 2.x: mnist.src.TensorFlow2.ipynb

• TensorFlow 1.x: mnist.src.TensorFlow.ipynb

• MXNet: mnist.src.MXNet.ipynb

• PyTorch: mnist.src.PyTorch.ipynb

Bibliotecas software

• Keras (François Chollet), Python, marzo de 2015, https://keras.io/

  • Distributed Keras (CERN), 2016, https://github.com/cerndb/dist-keras

• TensorFlow (Google), Python/C++, noviembre de 2015, https://www.tensorflow.org/

• MXNet (Apache), diciembre de 2015, http://mxnet.io/

  • Gluon (Amazon & Microsoft), octubre de 2017, https://github.com/gluon-api/gluon-api/

• Torch (NYU), Lua, octubre de 2002, http://torch.ch/

  • PyTorch (Facebook), Python, octubre de 2016, http://pytorch.org/

    • DeepSpeed (Microsoft), Python (C++ & CUDA), mayo de 2020, https://www.deepspeed.ai/

• Caffe (University of California at Berkeley), diciembre de 2013, http://caffe.berkeleyvision.org/

  • Caffe2 (Facebook), abril de 2017, https://caffe2.ai/ (absorbido por PyTorch en marzo de 2018)

• Theano (Université de Montréal), Python, 2010-2017, https://github.com/Theano/Theano/

  • Lasagne, https://github.com/Lasagne/Lasagne

  • Blocks, https://github.com/mila-udem/blocks

• Chainer (Preferred Networks, Japan), Python, junio de 2015, https://chainer.org/

• DyNet (CMU), Python/C++, enero de 2017, http://dynet.io/

• Deeplearning4j (Skymind, San Francisco), Java, 2014, https://deeplearning4j.org/

• Microsoft Cognitive Toolkit, previamente conocida como CNTK, enero de 2016, https://www.microsoft.com/en-us/cognitive-toolkit/

• DSSTNE: Deep Scalable Sparse Tensor Network Engine (Amazon), mayo de 2016, https://github.com/amzn/amazon-dsstne

• PaddlePaddle (Baidu), septiembre de 2016, http://www.paddlepaddle.org/

• VELES (Samsung), junio de 2015, https://velesnet.ml/

• Minkowski Engine (Stanford), 2019, https://stanfordvl.github.io/MinkowskiEngine/

• Neon (Intel Nervana), mayo de 2015, https://www.intelnervana.com/neon/

• BigDL distributed deep learning library for Apache Spark (Intel), https://bigdl-project.github.io/master/

• PlaidML (Intel), octubre de 2017, https://www.intel.ai/plaidml/

• cuDNN (NVIDIA), septiembre de 2014, https://developer.nvidia.com/cudnn

• DNNL, a.k.a. MKL-DNN, Intel Math Kernel Library for Deep Neural Networks (Intel), C/C++, https://intel.github.io/mkl-dnn/index.html

• MATLAB Neural Network Toolbox (MathWorks), https://es.mathworks.com/help/nnet/index.html

• IBM PowerAI, https://www.ibm.com/us-en/marketplace/deep-learning-platform

• Deep Learning Studio (DeepCognition.ai), "no coding deep learning IDE", 2018, http://deepcognition.ai/

ONNX [Open Neural Network Exchange] format

https://github.com/onnx/onnx

Recomendaciones prácticas

• 37 Reasons why your Neural Network is not working, KDNuggets, August 2017. https://www.kdnuggets.com/2017/08/37-reasons-neural-network-not-working.html

• My Neural Network isn't working! What should I do?, http://theorangeduck.com/page/neural-network-not-working