David Bal

I just upgraded my 7 year-old Intel 4790k processor to a new Ryzen 5700g and I see training taking 18x longer while in general this CPU is 2.5x faster than...

Yes. 1.0.0-beta7

Yes, 1.0.0-beta7 on both machines. I'm now trying 1.0.0-M1.1, but I'm having trouble with my gradle build file: I'm getting this error with 1.0.0-M1.1 Caused by: java.lang.UnsatisfiedLinkError: /home/bal/.javacpp/cache/nd4j-native-1.0.0-M1.1-linux-x86_64.jar/org/nd4j/nativeblas/linux-x86_64/libjnind4jcpu.so: /usr/lib/x86_64-linux-gnu/libstdc++.so.6: version...

I set the environment variable on my AMD Ryzen compute: export MKL_DEBUG_CPU_TYPE=5 No effect. Still ~20x slower in training than my 7 year old Intel CPU.

I'll take a look at this issue soon. I'm swamped with other work right now, but I should get a chance in the next few days. Thanks for bringing this...

The best thing to help me fix a bug is ready-to-run example code that demonstrates the behavior. Could you prepare the example code? If the problem is only graphical, then...

Connor, I'm finally working on these issues.  I've confirmed the bugs existence - they are both bugs.  I'll fix them.  I'll also make some test cases to ensure no regressions...

Bug #1 was real easy.  I had a refresh() in the wrong place in ICalendarAgenda (around line 390) and it only affected deletes and not edits. Bug #2 is due...

Bug #2 is not a failure of equals.  It is a result of the copy of ByRules being an empty list instead of null, which is causing equals to fail. ...

I fixed bug #2 too.  The fixes are only in branch 8.0.  I'll migrate to other branches later. On Thursday, April 12, 2018, 1:42:08 PM MDT, ConorMurphy21 wrote: The example...