For this project, a custom workspace on AI Studio using the Deep Learning GPU-based image.
For extra libraries and specific versions, are listed in the Project.
The memory configurations, was set to 4GB for GPU RAM and 2 for GPU VRAM.
AI Studio AI Studio
Notebook will be available in Google Drive and GitHub
Go to Practicum1
You can access the notebook directly from this PRACTICUM-1-MariliseCover. PRACTICUM2-MariliseCoverPRACTICUM-2-MariliseCover
The DIV2K dataset is a popular benchmark dataset for image super-resolution tasks. It consists of 800 high-resolution images (2560x2048 pixels) and their corresponding low-resolution versions (1280x1024 pixels). The images are diverse in content, covering a wide range of scenes and objects.
FSRCNN (Fast Super-Resolution Convolutional Neural Network) is a deep learning architecture specifically designed for image super-resolution tasks. It's known for its efficiency and speed compared to other super-resolution methods, making it a popular choice for real-time applications.
Jaiaid
https://github.com/pytorch/examples/blob/main/imagenet/main.py
Lornatag
https://github.com/Lornatang/FSRCNN-PyTorch/blob/master/model.py
PyTorch Tutorials 2.4.0 & documentation. You can access the notebook directly from this PyTorch Cheat Sheet. https://pytorch.org/tutorials/beginner/ptcheat.html
Rafael Borges
ML Engineer and Data Scientist – HP Brazil
https://www.linkedin.com/in/rafa-borges/
Morgana Dias Rodrigues
Data Scientist - Teais Labs
https://www.linkedin.com/in/morgana-dias-rodrigues/
You can access the notebook directly from this PyTorch Cheat Sheet.