Nice poses and girl but way underexposed and too small a file to do much good in post processing. If you could post high res versions or better, the raw NEF files a lot could be done with them. Eyes, particularly the nearest eye is the key feature of any face image and we evolved to be very sensitive to eyes. If the eyes are sharp and detailed, the rest of the image can be greatly softened but our impression is of a very sharp image if the eyes are sharp because that is most of what our brains are looking for.
So pay attention to getting at least one eye sharp, both if they are looking at the viewer. You can make sure of good exposure even when against a background that is bright, by using spot metering on a middle level part of the face.
I played for a couple minutes with this and there was little in data to work with with such tiny files, but added a bit of smoothing to the skin where noise was adding roughness from the small fie size, and added a little color to the lips so there would be some color since none was visible in the eyes. I increased gain in post, and smoothed the busy background and cropped one of them. It is more flattering to have limbs or torso leaving the frame either narrowing or the same width because the viewer's brain fills in the missing information about size and proportion so if a body or limb is cut off by the framing, try to have the limb or body cropped at a point where it is tapering to less width. It really changes the impression of weight or bulk. The side view image is flattering because it appears slimmer because the shoulder is separated from the mass of her upper torso. The lower image gives the impression of the same girl being really bulky, more so than in person because with our binocular 3d vision can detect depth. With a camera being 2D, we as photographers have to give visual cues to the viewer about depth which if given those cues, we get the impression of slimmer proportions like we seen naturally with two eyes on a horizontal plane. That is one reason we pose subjects, to bend limbs and joints and use lights to create separation. Moving the limbs away from the torso and flexing, even if very slightly, gives additional cuing for the brain to understand in a 2d image, what they would normally see in a 3d image.
Please post the setting for the camera so we can make suggestions. Or before doing another set, let us know so we can suggest ways of getting what you and your model are seeking.
PS...the roughness of her skin in these photos is due to very few pixels defining the image in tiny JPGs, every time they are saved or edited so 1/2 the data is lost by compression. That is why you really should always shoot in RAW format and edit and render new JPGs as first generation JPGs from the original RAW file. Copy and resizing even a large detailed JPG a number of times and it loses so much data that it becomes useless. Only shoot JPG only when the file is only going to be edited and saved once. Making any change to a copy, means rendering and compressing again, which tosses out a great deal of data each time.