Pytorch - question about an open source utility

Job ID: 31415302

Budget: $30 – $250 USD

I2L-MeshNet is a network for 3D human pose and mesh estimation from a single RGB image. I've got myself familiar with the process of predicting the mesh, though I need to understand a couple more features.
I've found the variables and functions in the implementation responsible for the estimated pose, but I'm struggling to extract them in a proper format.
I would like to receive help to tune the network for my needs.

I2L-MeshNet (https://github.com/mks0601/I2L-MeshNet_RELEASE)
Related categories: Pytorch Computer Vision