Week 5 Journal
Learning
yt-watch-history: After 3 iterations I have understand how to deal with ML models.yt-watch-history: New PR forchannel_recoinfrontend. This PR just take subscription data and shows the recommendation of selected channel.
Thoughts
- PROJECT IDEA: I am thinking about to create a Python Package from which you can scrape website using
cURL. - In
yt-watch-history, I don’t want to give so much time onfrontendinstead more focus onbackend.
CRITICISM: If you don’t focus on
frontendthen it is bad to present and maintain. That’s why you should also take care offrontendpart.
Week 6 Journal
Learning
campusx: I have built new project where you can get all the notes and links provided in the description of sessions of DSMP course.- I have created this in almost 3 days.
- Done many things like follow conventional commits, written tests and more. See here
- I have released
v0.1.0ofcampusxbut I use 🤖 ChatGPT to generate release note (using all commits) and I don’t think that it is good release note like I see for other projects. - 🤓 I want to learn how to write and manage Release Notes and CHANGELOG.
Thoughts
- Conventional Commits: For a long time I have been thinking about to follow a convention for GIT commits but
where-ever I check I don’t understand then I came across
vivaxy.vscode-conventional-commitsa VSCode extension and now I know how to write better commits. 😉 - Mkdocs:
mkdocs-materialis a very powerful tool to create static web pages using markdowns. - 🎉 The Open Source Contributions of Indians ft. @ApnaCollege.
- I am thinking about to switch to
pixiorryebecause I also want to manage my projects like never before; like professional 🤯.
Week 7 Journal
Learning
rye: My new way to manage projects like pro. I have used it and this is very convenient to use and manage.rye: I have raised issue #639 while configuring thecampusxproject. For more info see issue itself.astral-sh/uv: Another blazing fast tool for python devs to manage their packages (written in 🦀 Rust).- 🤯 This is actually powerful (I have seen it 👀). It is almost 8x faster than
pipwhen I build myyt-watch-historyproject usingdocker-composeit takes 🐢 ~200 seconds withpipbut now it takes 🐇 ~25 seconds.
- 🤯 This is actually powerful (I have seen it 👀). It is almost 8x faster than
starship: Another blazing fast tool for terminal prompts (written in 🦀 Rust).- 🎉 Yay, I have learned and configured my terminal prompt using
starship.
- 🎉 Yay, I have learned and configured my terminal prompt using
- Social Cards: I have added social card preview on the website. Here, I have hard coded the
<meta>tags (which is not recommended). You can usesocialbuiltin plugin ofmkdocs-material.
Thoughts
- Praise by Nitish Sir for
campusxproject. See my LinkedIn Post on this. yt-watch-history: I have been configuring this project withryeand there are many challenges I have faced while doing this like:- This is a virtual project and there are some limitations comes with virtual packages check the issue #639.
- This project contains two separate repos
backendandfrontendwhich make this a monorepo and for nowryedoes not well support a monorepo because I have to create one workspace level environment for both repos (and I don’t kinda like this).
Week 8 Journal
Learning
Starship: Now I have learnedstarshipin a way that I can customize my prompt easily however I want to be. I written a blog a on it to. You can check mystarship.tomlon Github.notebooks/spotify-analysis: I have started a new project which analyses User’s Spotify’s Streaming History data and give insights from them.- I have done some analysis on
StreamingHistoryandPlaylistdataset and gain some interesting insights from it.
- I have done some analysis on
Thoughts
- backdropbuild.com: I have received a mail from
@JoeyDeBruin to participate in their program (which is a kind of Hack-a-thon) and
mentioned my project
yt-watch-historybecause he thinks that the project good fit their program (and I am doing this project to learn MLOps concepts like Docker, MLFlow).

Note
Hey Joey, Thanks to appreciate my project yt-watch-history and also for considering me to participate in your
program.
Honestly, I am just learning “MLOps” concepts while building this project and for now, I am not sure that I am a good fit for your program. And by keeping this in mind I am not able to join but I am sure that I’ll join it in future for sure.
Thanks Joey
Anshul Raj Verma
Week 9 Journal
Learning
- 😞 Not Much!
- But I have watched some sessions of the DSMP 2.0.
Thoughts
- Before Wedding: I am going to wedding for approx. 3-4 days.
- After Wedding: I have absent for 11 days from 25 Feb — 6 Mar.
Week 10 Journal
Learning
ryetransferred toastral-shGithub Organization.ruffreleased itv0.3.0with new formatting rules2024-02. Read the Blog- I have simplified the complex url of blogs in my website.
Thoughts
- I have to bring my laptop to where I go I go because without it I lose many things and make it hard to catch up.