Reset_comeback
Learned for: ???
Hello it has been a long time since I came back here. I was so consistent, and then I fell off. A string of various vacations and too many easy excuses to fall out of my habits of pushing to github and getting work done. I hate to be corny about this, but it changes now. I am putting a real acountability system in place to push myself to work as hard as possible for the next few weeks, until it is time to go back to school. Don’t worry, I won’t give up when I get to school, things will just be different.
I can’t count the hours that I’ve thrown away watching youtube, reading manhwa, and playing games. It sucks me away so easily that I forget my motivations and purpose for trying so hard here. I want to be great. And, right now is the perfect time to put all my focus towards that greatness, a time when I can leave all those meaningless distractions behind. I’m not saying I’m going to try to be a god or anything, just that I can sure as hell work a bit harder to limit the unnecessary screentime.
I think another thing that has been holding me back is Spaced. The big thing, the thing that I’ve invested so much time and even some money into, has turned into something that I am not in love with. And today, I have finally gotten it to a point where I am satisfied (might still add a couple of features tbh). Here’s the real breaking point: I have vibe coded up 99% of this, and don’t feel like I’ve learned a ton that I care about and think is cool. Further, anything complex brings major headaches, and a lot of the optimizations that I would like done are exactly this. I am just not motivated enough to do fullstack in this way I guess. So, as of now, I will end the journey of Spaced (unless I get super famous overnight when I do some posting) and move onto bigger and better things. It’s been a great experience.
Now, what are those bigger and better things, you may ask? Reinforcement Learning! It’s super cool and I reallllyyy want to be able to do RL engineering or something similar in the future, so the grind begins now. I know it’s gonna suck a lot at the start, but it’ll pay off. The big things I’ll start with will consist of reading papers, looking at implementations, doing my own implementations, and eventually some sort of project to tie things together. Additionally, I’m going to be on the lookout for a lab position next semester for RL, so hopefully that can boost things up for me. Always looking for more advice on what I should do here, so feel free to reach out. I’m also going to try to use Spaced for learning to see how much my own software can help me with the learning process.
So, what will this really look like then? I’m thinking that I’ll dedicate time every single day to this cause, and a solid chunk of it at that. I want this to be a mix of things depending on where I’m at, and since it’s so early now I want a lot of time spect learning and listening and taking notes. So, tomorrow I want to watch a couple of the David Silver RL lectures on YT (2-3 hours), watch a video about GRPO, and dig around on how to do paper implementations, and look into more conventions on writing RL code / what implementations will even look like.
Anyways that’s my update. I’ll come back here for daily updates again for the forseeable future, so I’m sure this will be refined. See you tomorrow!