Currently grappling with the WGCNA package for the first time and worked through the tutorials provided, and now trying to look at my own data with this technique.

However, im unsure what format my data needs to be in (log2/linear/etc)?,

Ive checked the website again and again, and cannot find the answer.

I usually recommend log-transformed data, although this may depend on pre-processing and normalization. The idea is to use data for which random errors are approximately symmetric and, if possible, not too heavy-tailed. Note that some microarray normalization methods, notably RMA for Affy arrays and some schemes for Agilent arrays, return log-transformed data. For RNA-seq, see the WGCNA FAQ at https://labs.genetics.ucla.edu/horvath/CoexpressionNetwork/Rpackages/WGCNA/faq.html, especially point 3.