| Author: | Mark Hall | |
| Category: | Python integration | |
| Changes: | Now determines the number of GPUs available and will prevent the use of mutli_gpu_model at prediction time (even if used at training time) if the number of GPUs in the prediction environment is not greater than 1. Also calls backend.clear_session() after training and each prediction batch to ensure a clean graph in case multiple clients are sharing the same python server for building DL models. | |
| Date: | 2019-10-21 | |
| Depends: | weka (>=3.8.0), wekaPython (>=1.0.9) | |
| Description: | Provides a wrapper classifier for invoking zoo models available in Keras. Python version 2.7.x or higher is required. Also requires the following packages to be installed in python: keras, numpy, pandas, matplotlib and scikit-learn. | |
| License: | GPL 3.0 | |
| Maintainer: | Weka team <wekalist{[at]}list.scms.waikato.ac.nz> | |
| MessageToDisplayOnInstallation: | This package requires python (2.7.x or 3.x) to be installed on your computer before it can be used. The python command must be in the PATH. Within python it requires the: following packages: keras, numpy, pandas, matplotlib and scikit-learn, scipy. | |
| PackageURL: | http://prdownloads.sourceforge.net/weka/kerasZoo1.0.1.zip?download | |
| URL: | ||
| Version: | 1.0.1 |