Add clarification to loghill and loggnls function documentation

[sedavid01]

The parameter descriptions for 'ga' and 'la' should be clarified for the loghill and loggnls to indicate these parameters are in log-scale and optimized/estimated as such. Then mention that the ga and la parameter are "back-transformed" from the log-scale to the linear/raw dose scale before being reported (will need to double check this in the code, but should be correct).

Needs:

• Update loghill return value description
• Update loggnls return value description

[sedavid01]

Though fithill and fitgnls already have similar documentation already, we do not want users to have to search in another location for the same answer of interest in the current function they are looking at.

• loggnls - append the following clarification to ps definition for this function: "ga and la are in log10-scale"
• loghill - append the following clarification to ps definition for this function: "ga is in log10-scale"

This is meant to mitigate confusion to users and future developers while ensuring clear understanding/interpretation of model parameters. For example, ga in gnls and ga in loggnls as they are, are not the same. That is, one is on the raw scale and one is on the log-scale.