As others have said, a 78mm rise over 2035mm is 3.83%.
In general, this would be considered too high unless you are modelling something that requires that sort of grade, like a logging railway or possibly a streetcar layout or you are running very short or under-weighted trains.
While this is a bit of a 'how long is a piece of string', 2% grades are generally considered high but do-able, even for steam locos. There are a few things to keep in mind:
1. Diesel locos can certainly handle steeper grades than steam locos.
3. Consisting locos is always an option but requires that they be well speed matched otherwise you get other problems from having one engine surging against the other which acts a brake.
3. Things go wrong on even the best laid track. If a coupler breaks or disconnects, the steeper the grade, the faster everything hits the curve at the bottom if you don't catch it in time.
4. You will read about people having 4% helixes. I have never met anyone who was happy with grades anywhere near this (there are a lot of helixes in the US). It's normally more like 2-2.5% and there's normally a lot of time taken to improve the entry and exit vertical curves and the general alignment.
5. What are you building your subroadbed out of and how much does the humidity change over the year where you live? Basically, while your test train works now, what will happen in 6 months when the subroadbed has changed? Will you suddenly have a very sharp vertical curve somewhere?
6. Are there any curves in this elevation change? If yes, you are actually increasing the effective grade and increasing the liklihood of operational problems.
Assuming you don't have the space for a normal helix and depending on your layout and operational desires, you might want a vertical transfer table. This is basically shelf of track mounted on high quality drawer sliders that you can use to move from one level to another. It's certainly less than ideal but gets rid of the very high grades.
Alternatively, the prototype also built too-large grades around the place requiring the train to be broken into two (or more!) pieces and taken up the hill over multiple trips. This certainly increases the operating interest but is very annoying if you're more a run-the-trains sort of person.