I am not sure if this is the appropriate place to ask this question so forgive me if it is not.
I am a first year graduate student in mathematics. Often times when I face a relatively difficult problem or exercise, the first thing that I think of is "I can't do this!" which leads to "why am I so dumb?". This kind of unhealthy monologue is preventing me to actually attack the problem, and I have to spend a lot more time on problem sets than the others because of this.
I am wondering if anybody else has this kind of problem as well, if so how did you overcome it? Any advice would be appreciated.
I am not sure if this is the place, but let me try. I am a graduate student in physics who had second major in mathematics during undergraduate so I somewhat can understand your feeling (hopefully).
First thing first, if in graduate school you don't face difficult exercises or problems, generally you will not grow --- or rather, not quickly enough. During undergraduate, because exams are often the end of everything, our learning is very exam-oriented, which often does not help as much in graduate school. That said, graduate studies benefitted from undergraduate studies in terms of literacy. A lot of difficult concepts in graduate studies become somewhat easier to penetrate or even read when you have faced or seen the basics at least once (even if you do not fully understand them).
I suppose you will face many exercises involving proofs. I like the following sentences by John Howie in the preface of his Complex Analysis textbook:
While of course it is true that you cannot live off mathematics with just copying proofs, I think in mathematics critical thinking and "healthy imitation" are essential. A lot of technical strategies developed over the decades and centuries are extremely clever and ingenious, and mimicking them might give you a sense of how these older problems were dealt with, and oftentimes new problems can be bent similarly. Of course, some require completely new ideas, but usually very strong novel ideas come from experience, knowing what worked and what did not. That's why doing a lot of exercises, simple and hard, are essential even if you cannot deal with them at first. At least personally, I understand better the problem if I first know why I could not deal with it in certain ways, and then found out how to deal with it later (by myself or with some help).
In that sense math.stackexchange platform like this also helps; it gives new perspectives that some online users have, and surprisingly they can be really cool. For that same reason, do not do mathematics in isolation --- unless you have very strong reason not to (e.g. you are extremely introverted or find yourself really unable to talk to people).
Last but not least, at the beginning you should not overthink about "how fast" you should get the stuff. In fact, a lot of times you can learn or understand new things only way after you "finished". I first learn differential geometry in a course I took during study exchange, I learnt a lot, but I am pleasantly surprised that some concepts were actually misunderstood until I tried to use it elsewhere more than a year later (more specifically, when I was dealing with something in general relativity). And it occurred again 2.5 years later, when another part I thought I knew better was right but somewhat sloppy. I think not getting everything at first shot may be something one needs to be comfortable with. You only have finite time to learn anything.
A lot of things I said here are probably either common-sense or well-known, but I personally need to remind myself that graduate studies is hard. Of course some find it easier than others, but I guess we can work it out if we don't give up.