> For the complete documentation index, see [llms.txt](https://linzhiqiu.gitbook.io/the-clear-benchmark/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://linzhiqiu.gitbook.io/the-clear-benchmark/documentation/download-clear-10-clear-100.md).

# Download CLEAR-10/CLEAR-100

Download links for CLEAR-10 and CLEAR-100

![CLEAR Statistics. Note that CLEAR10 only has 1st-10th bucket labeled.](https://2411580087-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FiPLWAhemH9JTpCCJxZ3p%2Fuploads%2Fj35qu3dXRvweH56E3Xld%2Fstats.png?alt=media\&token=38ecde83-ef73-4692-93fd-32b810b5356a)

### CLEAR-10 HuggingFace Download Links:

* [CLEAR-10 Trainset (1GB)](https://huggingface.co/datasets/elvishelvis6/CLEAR-Continual_Learning_Benchmark/resolve/main/clear10-train-image-only.zip)
* [CLEAR-10 Testset (200MB)](https://huggingface.co/datasets/elvishelvis6/CLEAR-Continual_Learning_Benchmark/resolve/main/clear10-test.zip)
* [CLEAR-10 Trainset & Unlabeled Metadata (3GB)](https://huggingface.co/datasets/elvishelvis6/CLEAR-Continual_Learning_Benchmark/resolve/main/clear10-train.zip)

### CLEAR-100 HuggingFace Download Links:

* [CLEAR-100 Trainset (3GB)](https://huggingface.co/datasets/elvishelvis6/CLEAR-Continual_Learning_Benchmark/resolve/main/clear100-train-image-only.zip)
* [CLEAR-100 Testset (1.6GB)](https://huggingface.co/datasets/elvishelvis6/CLEAR-Continual_Learning_Benchmark/resolve/main/clear100-test.zip)
* [CLEAR-100 Trainset & Unlabeled Metadata (13GB)](https://huggingface.co/datasets/elvishelvis6/CLEAR-Continual_Learning_Benchmark/resolve/main/clear100-train.zip)

### Instructions for downloading unlabeled images:

1. Download the CLEAR-10/-100 Trainset & Unlabeled Metadata.
2. Run the below Python script in the downloaded folder.

{% file src="/files/95wX3FU1VbMBLMEI9S6F" %}
Python script for downloading the unlabeled images from YFCC100M metadata.
{% endfile %}

### Folder Structure Preview:

```markup
root/
|   class_names.txt (each line is a class name)
|   labeled_metadata.json (with paths to labeled images' metadata)
|   all_metadata.json (with paths to all images' metadata)
|   download_all_images.py (for downloading unlabeled images)
└───labeled_images
|   └───1
|   |   └───computer
|   |   |   |   235821044.jpg
|   |   |   |   ...
|   |   └───camera
|   |   |   |   269400202.jpg
|   |   |   |   ...
|   |   └───...
|   └───2
|   └───...
└───labeled_metadata
|   └───1
|   |   |   computer.json (dict: key is flickr ID, value is metadata)
|   |   |   camera.json
|   |   |   ...
|   └───2
|   └───...
└───all_metadata
│   │   0.json
|   |   1.json
|   |   ...
└───features
|   └───moco_b0
|   |   |   features.json (with paths to labeled images' features files)
|   |   |   state_dict.pth.tar (a copy of the state_dict file)
|   |   └───1
|   |   |   |   computer.pth (dict: key is flickr ID, value is torch tensor)
|   |   |   |   camera.pth
|   |   |   |   ...
|   |   └───2
|   |   |   |   ...
|   |   └───...
```

### Loading the pre-trained MoCo model on 0th bucket

* Here is a short example script for loading the `state_dict.pth.tar` under `features/moco_b0` folder:

```python
'''
Load pretrain model for image features
'''
import torch
from torchvision.models import resnet50

def load_model(state_dict_path='./features/moco_b0/state_dict.pth.tar'):
    model = resnet50(pretrained=False)
    model.fc = torch.nn.Identity()
    state_dict = torch.load(state_dict_path)
    model.load_state_dict(state_dict)
    for p in model.parameters():
        p.requires_grad= False
    model.eval()
    return model
```

{% hint style="info" %}
**Good to know:** You can also download CLEAR-10 / CLEAR-100 on Avalanche with a few lines of code!
{% endhint %}