Welcome to fastai…I have been a stay at a home dad for 3 years(2017 to 2020) and it it was the best decision I ever made. I really think you will enjoy it. It’s a great time to develop a bond with your child.
I also have limited amount of time per week to learn fastai but with your background in computer engineering and the new fastai API it should come easy for you. A lot of people don’t have your background. It’s an advantage for you I think. Especially if you want to move from experiment to production environment.
So far I experimented with tabular data and vision. I Successfully install a local environment but lost a lot of time “fighting” with the hardware…I suggest you open an account on Paperspace and use their free GPU.
Thanks for the support David. I loaded paperspace yesterday. I was following along the 1st video. I had some issue of not being able to find v4 folder there. But I will look again today. If it doesn’t work. I will search the forum.
I feel like I am already something because I am also learning reinforcement learning on the side. And I realize using virtual GPU will be helpful for that. So the fast.AI course is already helping me. If there any beginners with limited or none focus time ( I am talking like 15mins a day of focus time). Feel free to message me.
I plan to use deeplearning with tabular of finance data.
I already tried playing with tabular data and it’s tedious to my taste. You have to learn pandas(useful but there is a learning curve) and then clean/organize the date (especially for time series)… if you can put you data into a graph I find that it’s easier to go with vision because you don’t have to parse/clean the data to much.
Let me check it out. I will come back to you on this for finance, I have a lot of indicators. So may be I can do a classification task but I think in finance most signals testing should be done as a rolling window because relationships change. I am not sure how feasible that is via deep learning.
My name is Sanath. I am a high school Junior. I have been trying to learn ML and DL using fast.ai and various youtube tutorials (lol we love youtube ). I really want to gain real life experience in applying ML and DL to solve actual problems and hope to find other people with the same interest to work together with. If anyone has tips or opportunities they could recommend it would be greatly appreciated.
Thanks to Jermey for making it free and easy to access courses about ML.
Look for problems to solve around you. It’s a good start. I am going through the same process. I was trying the find ‘‘THE’’ impact project …but really start with a small project to build your confidence and then move to the next one
Hello! My name is Deepshika Sharma. I am in the beginning stages of my deep learning journey. I’m interested in applying ML/DL to the fields of Economics & Finance. Thank you to everyone at fast.ai who made this such a wonderful learning resource. Excited to learn through creating projects and deploying models!
My name is Darren. I am brand new to this stuff - although I have done a small bit of Python3 but for the most part, its all new.
I was doing a Coursera course with Andrew Ng but I simply could not grasp all of the math. Some of it made sense. Some of it is actually helpful, like seeing sigmoid functions and going “oh I know what that is!”. Most of it was Greek.
I may return to that course at some point. However, what I really like is fast.ai. These functions are fantastic! And the whole idea of ‘transfer learning’, now that is amazing - at least to me.
It shows me how this field is evolving. I hope to be a part of it…some day.
For now I am going through Jeremy’s material at super-turtle speed. There is alot to cover but I am hopeful I can get through the book by the summer, perhaps.
I think that would be really cool. Let me ask if anyone knows how one would go about doing this in fast.ai?
For example, do we have any good tweaks for audio?
I can certainly use the same sources as the previous author. But I am keen to use fast.ai mainly because I like the idea of doing things efficiently.
Let me know if you do know of any good references or developments.
Thanks Guys and Gals,
Hope to talk to you all!
-Darren
Hello, I have some experience of training deep learning models.
In order to further my experience, I’m looking to add location information to my images in order to improve my models accuracy. I don’t seem to be able to find anything online though.
Hi all! I’m a director of customer success working with large-scale customer experience design projects and platforms to support the same.
I don’t do Twitter or most other social media but I’m looking forward to pushing the cutting edge of neural networks a bit further in some niche areas of text analytics.
I am a professional programmer with minimal experience in MLAI. So, far I found most courses and tools inadequate. I like fast.ai because of the practical approach taken by the course makers. Let us see how it goes.
My ultimate goal is to come up with solutions for better predictive modelling for agriculture. Hopefully enabling more efficient crop cycles.
Looking forward to practical applications and broader horizons.
My name is Tom with a background in statistics. I currently work as a data scientist in the Finance industry and am going through the Fast AI course to continue my learning. I have done several other courses on Deep Learning and Machine Learning before but they tend to be too theoretical and less applied for me. I thought the Fast AI was awesome and going to keep going through course work.
I am Edwin with some background in mathematics. Currently, I work for a fintech startup doing data work. I am fascinated by how powerful and efficient fast.ai is, and I hope to study the fast.ai course as a starting point to break into data science. Very happy to join the community.
I am Venky (Vyankatesh Rege) and new to this forum. I have just started going through fastai-v2 this week.
I am based out of India and my background is ETL and Data Engineering. I have taken some courses over past few years in AI/ML but as Jeremy mentioned in his videos, the theoretical knowledge gets lost without practice. So, planning to do it the other way and see the benefits myself. Since my role is managerial and I don’t code regularly, it will definitely take some effort to learn and implement but feedback from some senior members on old fastai forums etc was very encouraging. Options like gradient are super helpful as well, already opened an account there and ran basic cat vs dog images code from first jupyter notebook.
Happy to start this journey !!!
Anyone who has done the course recently, if you have some tips on time to be spent per week and how long it takes, what practices helped you stay on course to complete the course will be extremely helpful. If there is any existing thread on the “best practices” of sorts, please point me there as well.
). I really want to gain real life experience in applying ML and DL to solve actual problems and hope to find other people with the same interest to work together with. If anyone has tips or opportunities they could recommend it would be greatly appreciated.
