MyCaffe 0.11.6.86-beta1

MyCaffe AI Platform (CUDA 11.6, cuDNN 8.3.2) now supports Visual Studio 2022 and .NET 4.8!

CUDA 11.6.0.511, cuDNN 8.3.2.44, nvapi 510, Windows 10-21H2/Windows 11-21H2, Driver 511.65

MyCaffe[1] (a complete C# re-write of CAFFE[2]) now supports Visual Studio 2022 and CUDA 11.6/cuDNN 8.3.2 and Windows 11!

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 full version of Microsoft SQL Express 2016 (or later). NOTE: The full version of SQL Express must be installed as opposed to the light version included in Visual Studio. Microsoft SQL Express can be downloaded from https://www.microsoft.com/en-us/sql-server/sql-server-downloads

This release of the MyCaffe AI Platform and Test Applications has the following new additions: • CUDA 11.6.0.511/cuDNN 8.3.2.44/nvapi 510/driver 511.65 • Windows 11 21H2 • Windows 10 21H2, OS Build 19044.1469, SDK 10.0.19041.0 • Upgraded to Visual Studio 2022 Builds • Upgraded to .NET 4.8 • Updated online help with new Doxygen version 1.9 • Added Scale to SimpleDatum. • Added ScalePow to SimpleDatum. • Added POW Scale operator. • Added new InterpLayer for upsampling. • Added new CudaDnn interp2 function. • Added new ConvolutionOctaveLayer for dual frequency image processing. • Added new SERF activation function. • Added error catch-all dialog for MyCaffe Test Application. • Added driver version to CudaDnn.GetDriverInfo. • Optimized Convolution layer forward and backward for groups. • Optimized automated testing. • Improved autotest cleanup.

The following bug fixes are in this release: • Fixed bug compiling CudaDnnDll with proper nvcc versions. • Fixed auto-test bug where <double> version did not have correct CudaDnn dll version set. • Fixed bug in SimpleDatum.CreateMeanImage when data set contains varying image sizes. • Fixed bug in MyCaffeControl.Run related to resize property checks. • Fixed bug in DataTransformation occurring when using image masking. • Fixed bug in Run occurring when using image masking. • Fixed bugs in automated testing cleanup. • Fixed bugs in RESIZE dataset creator were restart always aborts. • Fixed bugs in RESIZE dataset where duplicate data sources could be created. • Fixed bugs in GetDeviceInfo. • Fixed bugs in low-level memory freeing, added synchronization. • CudaControl, fixed bug in RunFloatEx and RunDoubleEx related to path.

Easily run Single-Shot Multi-Box Nets[3][4], import/export ONNX AI Models, run Triplet Nets[5][6], run Siamese Nets[7][8], Neural Style, train Deep Q-Learning or Policy Gradient 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.

Schedule distributed AI work packages, or create and train Single-Shot Multi-Box[3][4], Triplet Net[5][6], Siamese Net[7][8], Deep Q-Learning with NoisyNet and Experienced Replay, Policy Gradient, 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 GPUs cool as you train, but also gives you detailed information on each of your GPUs (such as temperature, fan speed, overclock, and usage), and allows you to easily mine Ethereum. When not training AI, put those GPUs 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] SSD: Single Shot MultiBox Detector by Wei Liu, Dragomir Anguelov, Dumitru Erhan, Christian Szegedy, Scott Reed, Cheng-Yang Fu, Alexander C. Berg, 2016.

[4] GitHub: SSD: Single Shot MultiBox Detector, by weiliu89/caffe, 2016

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

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

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

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