MyCaffe 0.11.0.65-beta1

MyCaffe AI Platform and Test Application (CUDA 11.0.2, cuDNN 8.0.2) with TripletNet and ONNX AI Model Support (onnx.ai).

CUDA 11.0.2, cuDNN 8.0.2, nvapi 440, Native Caffe up to 10/24/2018, Windows 10-1909, Driver 451.77

MyCaffe[1] (a complete C# re-write of CAFFE[2]) now supports TripletNet AI Models!

IMPORTANT NOTE: When using TCC mode, we recommend that ALL headless GPU’s are placed in TCC mode for we have experienced stability issues when using a mix of TCC and WDM modes with headless GPU’s.

REQUIRED SOFTWARE to use MyCaffe: 1.) Download and install Microsoft SQL Express 2016 (or later).

REQUIRED SOFTWARE to build MyCaffe: 1.) Install NVIDIA CUDA 11.0.2 which you can download from https://developer.nvidia.com/cuda-downloads 2.) Install NVIDIA cuDNN 8.0.2 which you can download from https://developer.nvidia.com/cudnn

This release of the MyCaffe AI Platform and Test Applications has the following new additions: • CUDA 11.0.2/cuDNN 8.0.2 supported (with driver 451.77 or above). • Windows 1909, OS Build 18363.959 now supported. • Added new samples user interface with new TripletNet sample. • Added new MyCaffeConversionControl to easily load and save *.onnx InceptionV1 model files. • Added support to MyCaffeImageDatabase for LoadLimit now allowing the image database to support very large datasets. • Added download button to easily download and extract the CIFAR-10 dataset. • Added optimized ImageTools.AdjustContrast • Added optimized ImageTools distortion on GPU. • Added distortion to Data Transformer. • Added LOAD_ON_DEMAND_NOCACHE support for low memory usage. • Added new DataSequenceLayer for k-n triplet based models. • Added updated and improved TripletLossLayer for triplet based models. • Added new low-level functions: copy_expand, minmaxvec and set_bounds. • Added TripletNet to MyCaffe Test Application for MNIST. • Added ability to save hand drawn images from MyCaffe Test Application. • Removed TripletSelectLayer, replaced by DataSequenceLayer. • Removed TripletDataLayer, replaced by DataSequenceLayer. • Removed TripletLossSimpleLayer, replaced by upgraded TripletLossLayer. • Removed BinaryHashLayer.

The following bug fixes are in this release: • Fixed bugs in samples now allowing use with SQLEXPRESS databases. • Fixed bugs in RnnForward and RnnForwardEx. • Fixed bugs in RnnBackwardData and RnnBackwardDataEx. • Fixed bugs in RnnBackwardWeights and RnnBackwardWeightsEx. • Fixed bugs in Image Database related to loading small datasets.

Easily import/export ONNX AI Models, run Triplet Nets[3][4], run Siamese Nets[5][6], Neural Style, train Deep Q-Learning or Policy Gradient[1] models to beat Pong or Cart-Pole, or create the CIFAR-10 and MNIST datasets using the MyCaffe Test Application which you can download from the MyCaffe GitHub site.

Create and train the Triplet Net[3][4], Siamese Net[5][6], Deep Q-Learning with NoisyNet and Experienced Replay, Policy Gradient[1], Neural Style Transfer, Recurrent Learning, Policy Gradient Reinforcement Learning, Auto-Encoder, DANN and ResNet models by following step-by-step instructions in the SignalPop Tutorials. And, to see other cool examples that show what MyCaffe can do, see the SignalPop Examples.

If you would like to visually design, develop, test and debug your models, see the SignalPop AI Designer specifically designed to enhance your MyCaffe deep learning.

Also, check out the SignalPop Universal Miner that not only keeps your GPU's cool as you train, but also gives you detailed information on each of your GPU's (such as temperature, fan speed, overclock, and usage), and allows you to easily mine Ethereum. When not training AI, put those GPU's to use making some Ether - never let a good GPU go to waste!

Happy ‘deep’ learning!

[1] MyCaffe: A Complete C# Re-Write of Caffe with Reinforcement Learning by D. Brown, 2018.

[2] Caffe: Convolutional Architecture for Fast Feature Embedding by Yangqing Jia, Evan Shelhamer, Jeff Donahue, Sergey Karayev, Jonathan Long, Ross Girshick, Sergio Guadarrama, Trevor Darrell, 2014, arXiv:1408.5093

[3] Deep metric learning using Triplet network by Elad Hoffer and Nir Ailon, 2018, arXiv:1412.6622

[4] In Defense of the Triplet Loss for Person Re-Identification by Alexander Hermans, Lucas Beyer and Bastian Liebe, 2017, arXiv:1703.07737v2

[5] Siamese Network Training with Caffe by Berkeley Artificial Intelligence (BAIR)

[6] Siamese Neural Network for One-shot Image Recognition by G. Koch, R. Zemel and R. Salakhutdinov, ICML 2015 Deep Learning Workshop, 2015.