Differences from the paper author's official implementation

[binmakeswell]

The implementation of the PyTorch version you released is different from the official version of TensorFlow released by the author. According to the official implementation published in the paper, the author's code implementation skips some parameters according to their names when calculating. But in your implementation, it seems that all parameters are directly involved in the calculation.
Their implementation:
https://github.com/tensorflow/addons/blob/master/tensorflow_addons/optimizers/lamb.py

[8enmann]

Their implementation came out long after I stopped updating this repo. I recommend apex!
https://nvidia.github.io/apex/_modules/apex/optimizers/fused_lamb.html

[shahzad-ali]

Their implementation came out long after I stopped updating this repo. I recommend apex! https://nvidia.github.io/apex/_modules/apex/optimizers/fused_lamb.html

First of all, an important line from the official GitHub repo:

Windows support is experimental, and Linux is recommended.

Second, I tried compiling the above repo on Windows that ended with dozens of errors beginning with

C:\Users\USER_NAME\venv-torch181\lib\site-packages\torch\utils\cpp_extension.py:316: UserWarning: Error checking compiler version for cl: [WinError 2] The system cannot find the file specified
warnings.warn(f'Error checking compiler version for {compiler}: {error}')

Is there any workaround (or a recommended alternative)?