convnets: A Python repository from carbonscott

Try a Colab demo

Add a shortcut to the shared demo data in your own Google Drive:

Open the link to access the shared data: https://drive.google.com/drive/folders/1YOKEYUWadK_DwKCRzelywr_ihQchb9FV?usp=sharing.
Add a shortcut to the shared data in your Google Drive.
Save it to "My Drive".
Open the Colab link: https://colab.research.google.com/drive/113uoe5Dwg_8gnNy5NWwWjKCa91qKd-25?usp=sharing
Follow the instructions in the Colab notebook to run the demo.

If you wanted to run peaknet on your own computers, please follow the installtation guide below.

Make sure your sytem has conda installed, otherwise please refer to the anaconda installation page for the installation.
Create a conda environment for running peaknet. Let'call this environment peaknet for now. Python version greater than 3.6 is required.
```
conda create --name peaknet python=3.9
```
Activate the conda environment.
```
conda activate peaknet
```
Install pip for your conda environment peaknet.
```
conda install pip
```
Then, you can use which pip to check if pip is a system-wide command or a command pointing to the conda environment.

Install torch and cupy that require GPUs. Our cuda version is 12.0.

# Install torch
conda install pytorch pytorch-cuda=11.8 -c pytorch -c nvidia

# Install cupy
conda install -c conda-forge cupy

Install tqdm, scikit-image and h5py.
```
conda install tqdm scikit-image h5py
```
(Demo specific) Install matplotlib for visualization using the demo.
```
conda install matplotlib
```
Finally, install peaknet directly with a github link (it should take less than one minute to install).
```
pip install git+https://github.com/carbonscott/peaknet --upgrade
```

Download the demo data from the shared folder: https://drive.google.com/drive/folders/1YOKEYUWadK_DwKCRzelywr_ihQchb9FV?usp=sharing
Download the jupyter notebook from the Colab link:
Copy and paste the downloaded jupyter notebook to the folder containing the demo data.
You can start running the notebook from the Demo section.