{
  "fetched_at": "2026-10-19T21:57:04Z",
  "source_url": "https://www.kaggle.com/competitions/cs4670spring2020pa3/data",
  "pages": [
    {
      "name": "rules",
      "content": "<h2>Rules</h2>\n1. You may not hand-label the test data in **/test_unlabeled**. Doing so will be treated as an **Academic Integrity Violation**. We will check your model accuracies with the files you submit on CMS and flag any inconsistencies.\n2. You may not share models with members of other groups.\n3. You will have 10 max daily submissions, which should be more than enough.\n4. Your submissions may not extend past the slip days allowed by CMS. In other words, you **may not use more slip days than you currently have left**. We will use timestamps on your Kaggle submissions to verify.\n5. You may not have inappropriate or offensive team names.\n6. You may not use any **other dataset** than the one provided in the assignment.\n7. You may **not** use **pretrained models or model architectures from the internet**. We are expecting you to build one from scratch and explain the architecture choices you made.\n8. You may not ask why Bharath has a zoo. He just does.",
      "isPublished": false,
      "hasIsPublished": false,
      "postTitle": "",
      "hasPostTitle": false,
      "mimeType": "",
      "hasMimeType": false
    },
    {
      "name": "data-description",
      "content": "<h2>Welcome to The Data Page!</h2>\nHere you will find the test data for generating the submission csv file using Kaggle_Submission.ipynb. You find more details on how to integrate this dataset with your current code in the About this file section below.\n\n<h2>/test_labeled</h2>\nThis folder contains 16 subfolders corresponding to each of the 16 classes. Try exploring the folder by clicking on the down arrow next to the folder name below. You will see that each \"class folder\" has 25 images. We have included them because **the distribution that these images come from is not the same distribution as the training images**. You may find it helpful to perform validation on this data to get an idea of how well your model generalizes. You may also train on these images along with the provided training data. **Note**: this distribution is similar to the one being used by Kaggle for both public and private rankings.\n\n<h2>/test_unlabeled</h2>\nThis folder contains 1984 images on which you will perform class prediction. Make sure this folder is place inside /data so Kaggle_Submission.ipynb can locate it. Note: The distribution is similar to the labeled data in */test_labeled*. You **may not** hand-label the images as this is a **violation of Academic Integrity**. See **Rules** for more information.\n\nNow get cracking. Bharath needs your help sorting the animals.",
      "isPublished": false,
      "hasIsPublished": false,
      "postTitle": "",
      "hasPostTitle": false,
      "mimeType": "",
      "hasMimeType": false
    },
    {
      "name": "Description",
      "content": "<h1>Welcome to the CS 4670 Kaggle Competition!</h1>\n\nOn his adventures travelling the world, Bharath (and his crew of dedicated TAs) collected hundreds of photographs of different animals. Unfortunately, there's been a mix-up, and the pictures have all been shuffled together! Your mission - if you choose to accept it - is to classify all the animal photographs into the appropriate categories.\n\n<h2>Competition Rules</h2>\n- Open Source libraries such as cv2, numpy, scipy, skimage are allowed, although they must be cited in the writeup. Use of pre-trained models such as ResNet, Vgg, etc. is not allowed.\n- There is a computational budget of 5MB for the maximum size of the model weights - work within these limits.\n- Sharing personal code between teams or acquiring labels for the test dataset (by hand labeling/other means) is expressly forbidden, and will be considered a violation of academic integrity. To check consistency, we will run your model with your training script uploaded to CMS.\n- Have fun, and be creative!\n\n<h2>Acknowledgements</h2>\nTraining/Validation Datasets obtained from the tiny image",
      "isPublished": false,
      "hasIsPublished": false,
      "postTitle": "",
      "hasPostTitle": false,
      "mimeType": "",
      "hasMimeType": false
    }
  ]
}