[Deep learning] Translate Tensorflow into Pytorch source code of a research paper

Translation paper: MoDL: Model-Based Deep Learning Architecture for Inverse Problems Tensorflow version source code: [login to view URL] (available 2020.10.06) Budget: 80 – 120$ Timeline: 1 week Detail requirements: Development environment and techniques: - Pytorch, Python 3, OOP style - Translation source code: [login to view URL] Evaluation criteria: - Conjugate gradient algorithm with backward in Pytorch. - Use datasets provided in the source code to produce the results. - Reproduced results: PSNR > 38.5 dB for an 8-fold acceleration with random noise of sigma = 0.01 ( tensorflow achieves 39.5 dB) with visualization of results.