mw66

Need to fix for all the np float types: np.float32, numpy.float16 / np.single, np.half: ``` $ python Python 3.6.7 |Anaconda, Inc.| (default, Oct 23 2018, 19:16:44) [GCC 7.3.0] on linux...

> My bot ... adjusting every 20 seconds if needed for a week Maybe that is too much? Think about the server load if there are many people doing this.

Related issue: https://github.com/cszn/KAIR/issues/200

if options/swinir/train_swinir_sr_lightweight.json, change: ``` "H_size": 256 ``` then, got the following error: ``` Traceback (most recent call last): File "F:\Projects\KAIR\main_train_psnr.py", line 252, in main() File "F:\Projects\KAIR\main_train_psnr.py", line 172, in main...

if options/swinir/train_swinir_sr_lightweight.json, change: ``` "H_size": 256 ``` then, got the following error: ``` Traceback (most recent call last): File "F:\Projects\KAIR\main_train_psnr.py", line 252, in main() File "F:\Projects\KAIR\main_train_psnr.py", line 172, in main...

@maitry63 Thanks for the info. Can you share your tried working code with Torch or JAX backend? Along with the Python version and freezed requirements.txt of all the library versions?...

> Hi [@mw66](https://github.com/mw66) , I am adding a simple example with Mnist dataset with multigpu parallel training with jax & Torch using multiprocessing. Find the [gist ](https://colab.sandbox.google.com/gist/maitry63/44bf4a978a84beb89ae6021dda157224/-21862-train-two-different-models-on-two-different-gpus-in-different-threads-of-the-same-process.ipynb)here. Hi @maitry63 Thanks...

Actually, these data are more or less static, can we add a data file in this package? Maybe only updated weekly/ monthly ?

Related issue: https://github.com/JingyunLiang/SwinIR/issues/181

Looks like that temp cache directory is not created for `gdc` only.