Pytorch CUDA Image Recognition Trainer on local system

I'm interested in training an image recognition model on a large dataset using Pytorch and CUDA.

Key aspects of the project include:
- Utilizing Pytorch for the development of the image recognition model
- Exploiting the power of CUDA for faster training and processing

I need a skilled professional with:
- Strong background in deep learning and Pytorch
- Expertise in optimizing models for CUDA
- Prior experience with large-scale image recognition datasets

Please note that the dataset I'm working with is more than 100GB.