Brian Coutinho

@agkphysics would this catch this issue https://github.com/pytorch/kineto/commit/51fd6e6a1caccbbbf0e23e5cf2b68cc6afd79ad4

Monotonic clock does make sense. Also @aaronenyeshi your comment here further agrees with this > The implementation in PyTorch Profiler is monotonic, fast, and would help align Kineto timestamps to...

@pytorchbot rebase

> > I guess USE_DISTRIBUTED and USE_C10D are checking different things, would like to get some eyes from distributed team on whether this is okay > > Honestly no idea...

Is this a feature request for tensorboard plugin? looks like the trace from libkineto has the information right?

@apcameron We don't have any RISC-V systems at handto develop on, but we are open to contributions in that space By the way, when you say support are you talking...

Hi @apcameron, thanks for sharing the fix to the build, we had to add a similar one for Arm architecture too at the same spot. Would you be comfortable contributing...

@jon-chuang sorry we missed this, if you can have the dynolog process run within the container and then have dynolog's port exported I think it should work. There is an...

Let me know if this issue can be closed 👍

Sorry about the delay, yes this could be addressed. The server logic is fairly simpe, i think this part needs updating on IPv4 https://github.com/facebookincubator/dynolog/blob/main/dynolog/src/rpc/SimpleJsonServer.cpp#L34-L52