Do I need to use cache when testing a model in vowpal wabbit?

Question

I have only a vague understanding of the caching option -c in vowpal wabbit. My best guess from reading the online documentation is that it's caching the training data during the model training stage, and is especially helpful if one needs to run multiple passes.

Does this mean that there is no reason to be concerned with the caching option when one is testing on new data?

Answer 1

  

这是否意味着在测试新数据时，没有理由与缓存选项有关？

使用缓存选项非常有理由-如果您打算进行多次通过或再次尝试使用相同的训练数据。很少在同一个数据集上尝试不同的选项是很罕见的，因此创建缓存文件通常是一个好主意。