Aleksei Petrenko

Also found this: https://github.com/marklakata/mqueue-w32

@Sniper199999 I know very little about IPC in Windows. I am not at all confident that pthread-win32 has full support for all POSIX features, such as inter-process mutexes (in fact,...

Wow, that's pretty cool! :D Awesome idea @beasteers I didn't expect that a simple overloading of loads would enable functionality like that! I'm pretty sure you can save some more...

@no2chem @DariusSG I'm not able to work on this repository atm for personal reasons. Comments made by @no2chem above are fair and this requires a fix.

Hi @tendermonster There's nothing dumb about this question! :) There's indeed no documentation website. The API is almost identical to the standard Queue except for get_many() and put_many() which are...

Hi @folkien Thank you for submitting this. I believe this is a functionality of [JoinableQueue](https://docs.python.org/3/library/multiprocessing.html#multiprocessing.JoinableQueue). Should be possible to do it using a wrapper around the regular faster-fifo Queue. Contributions...

I had to bump from 11 to 14 during the development at some point, but it can be the latest version of the code does not use anything C++-14 specific...

Not sure actually, never had a chance to work with TPUs. I imagine it would be trivial for serial training usecase `--serial_mode=True`, just replacing "cuda" with TPU everywhere. Although SF2...

Apologies for radio silence! Yes, I believe that would be a fine contribution! I've never done this myself. Wonder if it would be enough to just type the faster_fifo.pyx file?

@Abc11c thank you for using the code! Generally, I didn't have many issues debugging the version 1 (nor the non-serial V2 for that matter) You don't want to rely on...