Jan-Frederik Schulte
Jan-Frederik Schulte
Tests have only the "expected" failures now, so I think this is ok. I agree with Nicolo's comment on the pytest though, so if you could integrate that before merging...
Looks good now. @jmitrevs since you have the changes requested, you'll need to merge it.
Hi @sei-jgwohlbier, thanks for reporting this and the detailed studies you did. I was able to pinpoint the issue to a problem with the conversion to `channels_last` for the case...
Hi! Ah, no, this hasn't been fixed in the main branch yet. I'll get around to making a PR in the next few days. Thanks for reminding me!
I ended up including the fix in this PR that also deals with fixes to the pytorch parser: https://github.com/fastmachinelearning/hls4ml/pull/1086
I think the code in https://github.com/fastmachinelearning/hls4ml/pull/979 infers the precision from the QONNX model, so that should be covered there. The pytorch parser however will need to be updated to use...
Hi! Thanks for getting in touch. First of all, we just released version 1.0.0 of hls4ml and generally recommend switching to it. In your case, unfortunately this would not help...
Closed in favor of #1396
@sdubey11 @Beegorat Thanks for reporting the crash in tutorial. We found to be caused by the path to the files containing the model weights not being set correctly. Will be...
Thanks, this looks very nice to me! From the test logs, it seems to have triggered the synthesis tests successfully and the code looks good to me, so I think...