I do it for work, but that's because they pay me.
Find out if you really don't care about the end goal if the end goal is a project itself. If that's the case, maybe your actual end goal is to keep learning new things for personal pleasure. If you do decide that this is it, no reason to be "ashamed" or worried that you don't keep going after the technical challenges.
Just be honest with yourself to see if you really don't care about the end goal or you are not very familiar/comfortable with what comes after the technical challenges and are rationalizing ways to keep in this comfort zone.
If you do realize that you are actually _afraid_ to care about the end goal because of what is needed to get there, then work on learning something you are completely unfamiliar with. Maybe this will 'break the spell' and reduce your fear (that leads to a possible procastination).
If you - after a honest look at yourself - decide that you are _really_ not afraid, and that you _really_ don't care, then the next step is to ask yourself if you are happy with it. If you are, problem solved =).
To have a project as a proof of understanding. Its one thing to say, "Yes I understand how to do dec64 math." Its another to say, "Yes I understand dec64 math, here's my implementation."
Its kinda stupid I know, but that's literally my frustration with not completing projects. Its petty, so I don't indulge it.
Why don't you just release the library as open-source? This would give you a sense of accomplishment (because you did complete something), and possibly pride, if the library becomes popular.
A key point is it being open-source, though: so you can focus only on releasing the "proof of concept", but letting other people focus on the details, the edge cases, and the fixes to mundane but show-stopping bugs that make the project battle-hardened, production-ready.
One day, you'll want to start a company, or get a particular job, and whatever side projects you're doing, however unfinished, will help you do that.