Hi folks, trust you had a great time in this competition. I've just made available a single-model-notebook for this competition that contains all my FastAI tricks. It gets you as low as 0.19+ without ensembles. I'm hoping you can learn one or two and get better at computer vision.......and maybe beat me in the next challenge........😁.
Find here the link to the notebook: https://github.com/osinkolu/Task-Mate-Kenyan-Sign-Language-Classification-Challenge
I've tried to explain every block of code as best as I can.......cheers. please feel free to reach out on how It could have gone better.🤝
Thanks for sharing!
Anytime Daniel.
Thank you
Sure, anytime.
Prof, I can't help but say thank you. Your bag of tricks were helpful, especially dealing with wrong labels.
Sure bro, your model is only as good as your data.....😉
Very helpful, thank you !
You're welcome.
Thanks Victor for sharing your knowledge. Very insightful and helpful.
you are welcome....
Thank you for sharing, very much appreciated
Welcome....
Prof how can I chat with you.
Reasons being that you use Colab Pro and you are Nigerian. Can you give me access to talk with you?
🤝
Prof.... wow!! thanks so much!!!
Much appreciation