Denoiser currently uses python's multiprocessing which is likely leading to inefficiency in processing. Look into and implement TensorFlow build in multiprocessing to speed up both training/inference. This will be working off of Scott's branch: https://github.com/danielsf/deepinterpolation/tree/staging/cleaner/ophys_etl
Denoiser currently uses python's multiprocessing which is likely leading to inefficiency in processing. Look into and implement TensorFlow build in multiprocessing to speed up both training/inference. This will be working off of Scott's branch: https://github.com/danielsf/deepinterpolation/tree/staging/cleaner/ophys_etl