Skip to content

barncastle/CASC-Predictor

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

2 Commits
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

CASC Predictor

Blizzard Filenames + Machine Learning...
An example model, training data and arguments can be found under Example.

Prerequisites

  • .Net Core 2.2
  • Microsoft CTNK v2.6 GPU (Setup Guide)
    • The binary path needs to be added to Windows' Path environment variable
    • CTNK is optimised for Intel CPUs and CUDA GPUs expect AMD CPUs to be slow!

Arguments

Argument Description Mode
--mode 0 = Sample, 1 = Train (Required) -
--modelprefix Model name prefix (Required) -
--datapath Path to the Training Data (Required) -
--samplesize Sample character length. Defaults to 100 -
--epochs Amount of times to train. Defaults to 50 0
--samplefrequency Amount of minibatches between each sample. Defaults to 1000 0
--samplecount Amount of samples to produce. Defaults to 100 1
--sampleprime Sample prefix. Randomly selected if not provided. Defaults to random 1

About

A one-hot RNN for Blizzard filenames

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published