Great. I may give it a try.

@Ana Crosatto Thomsen I heard of the Fast.ai course before, but didn't have chance to try it. is it "Practical Deep Learning for Coders" one? How's it ? I finished Machine Learning Specialization from Andrew Ng, and I am doing the Deep Learning one. I will try couple practical projects in Kaggle on Deep Learning after the course. If I can't did well on Kaggle projects, I will see if there is any way to improve...

Andrew Ng's Machine Learning course is good. He is so good at explaining the basic and Math behind the ML and DL. He teach how to do all the ML and DL from scratch. But, for real practice in MLops world, I think I still need more hands-on. That's why, I wonder if Fast.ai will help in this sense. I just brought the "Deep Learning for Coders with Fastai and PyTorch: AI Applications Without a PhD" book (ebook). I don't if your course is following this book. I think I will spend some time tomorrow to read a bit of it. Tell me more how do you feel about the course? I think it is very good idea of you that keep working and fine tuning on the same Kaggle project. Instead of trying different project, trying different algorithms and technique, it helps improving self-taught stuff learn from different places, and save lot of time. That is what I am doing it too. I still trying the house price one. After improving with cross validation, I moved to rank of 19xx. I am planning somethings to try on it, but, just don't have time to start. It usually take me half to one day to try tweeting it. ML and DL is really consume me lots of time. But, I learn a lot it .... have a nice day, Joe

Interesting project. But, will it be great if it can solely use Llama 2 without using OpenAI, and the the whole thing running on-premises? Cause it will save lots of $$$ and it will work for company need to deal with some sensitive document. Just my comments, what do you guys think?

@Ana Crosatto Thomsen Great to hear that you tried Kaggle. I did the Titanic one too and I follow some of their tutorial. I especially like the Machine Learning tutorials, a short tutorial but included almost everything you need to know about Machine learning. Nice platform. I learn quite a lot from them. Let me know if you if found any cool on Kaggle.

@Ana Crosatto Thomsen I did the Titanic competition years ago by following the instruction of a tutorial online, and then I gave up my ML journey for busy with other stupid things for months and years. And now, I pick up my ML learning again. But I did another Kaggle project, Housing Price, I did not quite well the first try, with rank of 98xx with only Random Forest. Then, I tried different thing, like cross validation, pipeline, and GradientBoostRegressor. I got rank 2512. big improvement. But, I will keep fine-tuning it while I learn new Algorithm. I love learning ML this way on Kaggle. Pretty much like playing some sort of online game while studying... One thing I don't know understand is that why people with top ranks can get 0 error for their score. To me it is just impossible, according my understanding of ML- you just can't prefect prediction on your model, right? Let me know if you guys found something interesting in Kaggle.

@Ana Crosatto Thomsen I agree what you said about the Kaggle. As a beginner using Kaggle, I think it is very difficult and don't know where to start in those material in Kaggle, it is just like a forest to me. I wonder if there is some sort of suggested roadmap or guideline. @Marco Bottaro I think it is good idea to form a team to work on competition in Kaggle, but I think we need to find some good competition to start and have some sort of planing that we agree

@Marco Bottaro I am new to Kaggle and I really don't know what competition to start with. you have any idea?