It's nearly 3 weeks past the competition end. Can the top teams share their approaches even if its not code. Atleast a writeup would be considerate. Most of the... See more

when will the top solutions be available?

I have implemented my model with Tensorflow but I'm getting terrible scores.I have tried:* Class weights to handle the imbalanced dataset* Training with differe... See more

Hi @zindiMy code is based on downloading the images from S3, but I'm not able to download it anymore. Could you fix it? or just send the link to the images for... See more

In the spirit of sharing knowledge can top 5 share their solutions and insights to help us get better and improve our knowledge@YangTea@Mohamed_abdelrazik@sinch... See more

How could the top of LB get 0.1x :)I wonder how they were able to achieve such a score without using data leakage.

Hi Zindi (@Zindi, @amyflorida626)I think best Public LB was auto selected for Private scoring, regardless of your chosen 2 tags. Please, can you fix this.

Is possible to extend the deadline for 1 week?

looking forward to learning about what the top 3 teams found and the rest of us missed.

Compute RestrictionsAdditionally, the winning solutions will be implemented on a 4-core Virtual Machine with up to 16GB of RAM.1. Kindly, Can anyone explain it... See more

for those who done data cleaning is the percentage of mislabeled data worth cleaning it?

HI,I posted the model with 72.94% accuracy on hugging face here: https://huggingface.co/gianlab/swin-tiny-patch4-window7-224-finetuned-crop-classification . If... See more

I can't even import the images. Every time I do this, a FileNotFound error appears.I try to load images using ImageDataLoaders.from_df() and It always raise an... See more

Upon learning that the evaluation system for this competition is based on log loss, my initial inclination was to explore the common practice of utilizing thres... See more

I would like to confirm whether TensorFlow is an acceptable tool for the competition. Additionally, is the use of transfer learning allowed in the competition?

The third Image is definitely mislabelled right?

I did oversampling with skf and checked my validation set twice there's no data leaks in it, I used data that weren't oversample to not introduce data leak, how... See more

I wish to participate in this comp, but some times you must just face reality that it is not possible due to hardware and other resource constraints. Goodluck t... See more

It's strange that there are 3 teams whose scores are so different from the other teams :) I can observe that the previous competition had a team with a similar... See more

cv 0.62lb 0.59

Please this is my kaggle notebook https://www.kaggle.com/code/ridwanadejumo/agric-competition, I am facing the following error when trying to run the starter co... See more

I need advice on where to train, since I dont have access to a GPU.

Hello everyone. Recently I've got 0.25 score on leaderboard, and some people asked me about any advices. So, there is the main advice I can give you:Look deep i... See more

Does anyone knows why under the same circumstances ,FastAI is better than pure pytorch ?!!

Seeking clarification on competition time constraints. With inference taking just 5 minutes, can additional time be allocated to training, up to 8 hours and mor... See more