https://github.com/poppingtonic/semisupervised-rice-africa/blob/main/T-SNE%20Visualisation.ipynbI wanted to understand what the contribution of the RGN images w... See more

Congrats to the winners -- everyone is a winner in the end, and we all likely picked up one or two things from the experience. Well done @ZindiI definitely woul... See more

Guys, I've lost some of my result, because couldn't reproduce it . Is it any "magic" seeding formula, that guaranty repeating results, that i can use at the beg... See more

Just a single model (xcit_large) could have taken myself and my amazing teammate to 3rd place with 0.035xxxx on Private LB without ensembling this with another... See more

Hello everyone👋 Congrats to the winners first and special thanks to Zindi for hosting such competitions. I believe that this platform is for learning so I am s... See more

Respected Authors,Can you please release the test dataset which could be helpful for people who are later like me to evaluate the model?

Plato - my money was on you all the time. Congratulations.AZ - my goodness, what a finale! Congrats.What a comp, seems usage of TPU is growing.

I see no Cv lb thread in this comp , so i wanted to my share mine. You are free to comment your ownCV LB hereCV : 0.0669 (SKlearn SKF 5 Fold) LB : 0.03021 LB se... See more

I have created a custom log loss in Tensorflow. Though the below code is working but giving strange results like 80.06,75.04, etcdef log_loss_metric(y_true,y_pr... See more

Is there anybody who improves LB score with rgn images? In my experience, using rgn image solely makes LB worse so after that, I don't consider to use rgn image... See more

I need some help on tensorflow implementation of this, Anyone used tensorflow framework for this ?Mostly i'm seeing pytorch implementaion only.

Who has an idea of passing rgn images as one of the augmentations of the original image in fastai? Kindly help

Hi you all, i just found a dataset of a competition from kaggle ,and i was wondering if it's within the regulation of this competition to use this dataset.here... See more

This repository contains two training notebooks and notes showing how to pretrain from scratch, (or from imagenet weights) a self-supervised encoder that can be... See more

Do we have to remove rgn images before training? I can see rgn images are not included in testing csv and submittion csv files. Why to include rgn images and wh... See more

According to the competition's rule, we may use only the datasets provided for this competition. So.. can not we use the pre-trained model?

I don't know if i am missing something but i have a good local score of less than 0.10 but a bad lb?

I tried submitting but I keep getting this message error loading data

?

In this video I train a model to solve this competition from scratch, from end to end, using PyTorchhttps://www.youtube.com/watch?v=yq2fDQ4waY0Enjoy^_^Note: I h... See more

In this video I create folds for this competition :https://www.youtube.com/watch?v=rTK0JTli_XoEnjoy ^_^

The evaluation metric for this challenge is the Log Loss. Can you confirm the metric is : cross-entropy (loss) for the case of multi-class classification?I also... See more

Hii all, after reading this rule, I am not sure if we can use argmax to set the highest probability to 1 and the others to 0.

I found it interesting to inspect which parts of the image were most responsible for specific predictions. These are from a not-particularly-good model, and in... See more

Hey folks, I'm working on a post to help people get started on this challenge. Here's my take on a starter notebook (still a bit of a work in progress): https:/... See more