Set more realistic goals. Having high ambitions is good, but be careful not to get held back by overly high ones. Practice hard, and try to progress naturally instead of forcing an unreasonable goal down your throat all the time. I did the same back when I started about an year ago. My goal was to become at least 1800+ or candidate master within a year from complete scratch (I didn't even know any programming languages or how to code basic stuff). I really don't like the rate of my progress, but that's mainly because of my lack of practice, I attended a lot of cf contests but didn't solve many problems outside them, also due to becoming overly comfortable with the cf interface I avoided other online judges as well. GM is on a whole new level of expertise even the most talented of people develop after years of hard work and dedication. The hard problems you solved will take you to candidate master level if you can solve them during contest time — but solving something during contest and upsolving them out of contest are two different things entirely. All I can say is keep practicing hard — solve problems of all kinds from different OJs and try to find your areas of expertise and your weaknesses, and practice accordingly, and don't let a numerical quantification of your progress (your CF rating) hold you back — just keep practicing and increase the difficulty level of your problems as your skill in CP grows, and remain hopeful to reach your goal someday to maximize your probability of actually doing it.

You can read this Link